diff --git "a/checkpoint-2611/trainer_state.json" "b/checkpoint-2611/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2611/trainer_state.json" @@ -0,0 +1,18310 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.999808539153743, + "eval_steps": 500, + "global_step": 2611, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0003829216925138809, + "grad_norm": 1.5127280950546265, + "learning_rate": 5.0000000000000004e-08, + "loss": 0.9601, + "step": 1 + }, + { + "epoch": 0.0007658433850277618, + "grad_norm": 1.4570214748382568, + "learning_rate": 1.0000000000000001e-07, + "loss": 0.9752, + "step": 2 + }, + { + "epoch": 0.0011487650775416428, + "grad_norm": 1.4853640794754028, + "learning_rate": 1.5000000000000002e-07, + "loss": 0.9485, + "step": 3 + }, + { + "epoch": 0.0015316867700555237, + "grad_norm": 1.6019998788833618, + "learning_rate": 2.0000000000000002e-07, + "loss": 0.9711, + "step": 4 + }, + { + "epoch": 0.0019146084625694046, + "grad_norm": 1.4343098402023315, + "learning_rate": 2.5000000000000004e-07, + "loss": 0.947, + "step": 5 + }, + { + "epoch": 0.0022975301550832855, + "grad_norm": 1.5140701532363892, + "learning_rate": 3.0000000000000004e-07, + "loss": 0.9413, + "step": 6 + }, + { + "epoch": 0.0026804518475971664, + "grad_norm": 1.5071738958358765, + "learning_rate": 3.5000000000000004e-07, + "loss": 0.9399, + "step": 7 + }, + { + "epoch": 0.0030633735401110473, + "grad_norm": 1.6219124794006348, + "learning_rate": 4.0000000000000003e-07, + "loss": 0.9393, + "step": 8 + }, + { + "epoch": 0.0034462952326249283, + "grad_norm": 1.548447847366333, + "learning_rate": 4.5000000000000003e-07, + "loss": 0.978, + "step": 9 + }, + { + "epoch": 0.003829216925138809, + "grad_norm": 1.5101799964904785, + "learning_rate": 5.000000000000001e-07, + "loss": 0.968, + "step": 10 + }, + { + "epoch": 0.00421213861765269, + "grad_norm": 1.3384488821029663, + "learning_rate": 5.5e-07, + "loss": 0.9415, + "step": 11 + }, + { + "epoch": 0.004595060310166571, + "grad_norm": 1.3894020318984985, + "learning_rate": 6.000000000000001e-07, + "loss": 0.985, + "step": 12 + }, + { + "epoch": 0.004977982002680452, + "grad_norm": 1.3622794151306152, + "learning_rate": 6.5e-07, + "loss": 0.9326, + "step": 13 + }, + { + "epoch": 0.005360903695194333, + "grad_norm": 1.3284450769424438, + "learning_rate": 7.000000000000001e-07, + "loss": 0.9442, + "step": 14 + }, + { + "epoch": 0.005743825387708214, + "grad_norm": 1.2537797689437866, + "learning_rate": 7.5e-07, + "loss": 0.9426, + "step": 15 + }, + { + "epoch": 0.006126747080222095, + "grad_norm": 1.330793857574463, + "learning_rate": 8.000000000000001e-07, + "loss": 0.965, + "step": 16 + }, + { + "epoch": 0.006509668772735976, + "grad_norm": 1.1742899417877197, + "learning_rate": 8.500000000000001e-07, + "loss": 0.9486, + "step": 17 + }, + { + "epoch": 0.0068925904652498565, + "grad_norm": 1.1290744543075562, + "learning_rate": 9.000000000000001e-07, + "loss": 0.9652, + "step": 18 + }, + { + "epoch": 0.0072755121577637374, + "grad_norm": 1.0592819452285767, + "learning_rate": 9.500000000000001e-07, + "loss": 0.9177, + "step": 19 + }, + { + "epoch": 0.007658433850277618, + "grad_norm": 0.9967736005783081, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.9321, + "step": 20 + }, + { + "epoch": 0.0080413555427915, + "grad_norm": 0.9203951358795166, + "learning_rate": 1.0500000000000001e-06, + "loss": 0.869, + "step": 21 + }, + { + "epoch": 0.00842427723530538, + "grad_norm": 0.8977105021476746, + "learning_rate": 1.1e-06, + "loss": 0.9206, + "step": 22 + }, + { + "epoch": 0.008807198927819261, + "grad_norm": 0.8208876252174377, + "learning_rate": 1.1500000000000002e-06, + "loss": 0.8671, + "step": 23 + }, + { + "epoch": 0.009190120620333142, + "grad_norm": 0.8222814202308655, + "learning_rate": 1.2000000000000002e-06, + "loss": 0.9111, + "step": 24 + }, + { + "epoch": 0.009573042312847023, + "grad_norm": 0.7808475494384766, + "learning_rate": 1.25e-06, + "loss": 0.9222, + "step": 25 + }, + { + "epoch": 0.009955964005360904, + "grad_norm": 0.7570652961730957, + "learning_rate": 1.3e-06, + "loss": 0.9151, + "step": 26 + }, + { + "epoch": 0.010338885697874785, + "grad_norm": 0.7235645651817322, + "learning_rate": 1.3500000000000002e-06, + "loss": 0.9068, + "step": 27 + }, + { + "epoch": 0.010721807390388666, + "grad_norm": 0.6783406138420105, + "learning_rate": 1.4000000000000001e-06, + "loss": 0.863, + "step": 28 + }, + { + "epoch": 0.011104729082902547, + "grad_norm": 0.6298149228096008, + "learning_rate": 1.45e-06, + "loss": 0.8956, + "step": 29 + }, + { + "epoch": 0.011487650775416428, + "grad_norm": 0.6101565957069397, + "learning_rate": 1.5e-06, + "loss": 0.8933, + "step": 30 + }, + { + "epoch": 0.011870572467930308, + "grad_norm": 0.5784317255020142, + "learning_rate": 1.5500000000000002e-06, + "loss": 0.8739, + "step": 31 + }, + { + "epoch": 0.01225349416044419, + "grad_norm": 0.5533347725868225, + "learning_rate": 1.6000000000000001e-06, + "loss": 0.8753, + "step": 32 + }, + { + "epoch": 0.01263641585295807, + "grad_norm": 0.5175893306732178, + "learning_rate": 1.6500000000000003e-06, + "loss": 0.8757, + "step": 33 + }, + { + "epoch": 0.013019337545471951, + "grad_norm": 0.4827329218387604, + "learning_rate": 1.7000000000000002e-06, + "loss": 0.8784, + "step": 34 + }, + { + "epoch": 0.013402259237985832, + "grad_norm": 0.4603060781955719, + "learning_rate": 1.75e-06, + "loss": 0.8488, + "step": 35 + }, + { + "epoch": 0.013785180930499713, + "grad_norm": 0.429062157869339, + "learning_rate": 1.8000000000000001e-06, + "loss": 0.8601, + "step": 36 + }, + { + "epoch": 0.014168102623013594, + "grad_norm": 0.4230949878692627, + "learning_rate": 1.85e-06, + "loss": 0.8592, + "step": 37 + }, + { + "epoch": 0.014551024315527475, + "grad_norm": 0.3980255126953125, + "learning_rate": 1.9000000000000002e-06, + "loss": 0.8387, + "step": 38 + }, + { + "epoch": 0.014933946008041356, + "grad_norm": 0.39369726181030273, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.8565, + "step": 39 + }, + { + "epoch": 0.015316867700555237, + "grad_norm": 0.3655986189842224, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.8545, + "step": 40 + }, + { + "epoch": 0.015699789393069118, + "grad_norm": 0.3628844916820526, + "learning_rate": 2.05e-06, + "loss": 0.8191, + "step": 41 + }, + { + "epoch": 0.016082711085583, + "grad_norm": 0.3656681180000305, + "learning_rate": 2.1000000000000002e-06, + "loss": 0.8506, + "step": 42 + }, + { + "epoch": 0.01646563277809688, + "grad_norm": 0.35703447461128235, + "learning_rate": 2.15e-06, + "loss": 0.8395, + "step": 43 + }, + { + "epoch": 0.01684855447061076, + "grad_norm": 0.34360966086387634, + "learning_rate": 2.2e-06, + "loss": 0.8379, + "step": 44 + }, + { + "epoch": 0.01723147616312464, + "grad_norm": 0.3302212357521057, + "learning_rate": 2.25e-06, + "loss": 0.8214, + "step": 45 + }, + { + "epoch": 0.017614397855638522, + "grad_norm": 0.32497096061706543, + "learning_rate": 2.3000000000000004e-06, + "loss": 0.8152, + "step": 46 + }, + { + "epoch": 0.017997319548152403, + "grad_norm": 0.3161090910434723, + "learning_rate": 2.35e-06, + "loss": 0.8113, + "step": 47 + }, + { + "epoch": 0.018380241240666284, + "grad_norm": 0.30970415472984314, + "learning_rate": 2.4000000000000003e-06, + "loss": 0.8126, + "step": 48 + }, + { + "epoch": 0.018763162933180165, + "grad_norm": 0.3097487986087799, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.8378, + "step": 49 + }, + { + "epoch": 0.019146084625694046, + "grad_norm": 0.28779536485671997, + "learning_rate": 2.5e-06, + "loss": 0.8102, + "step": 50 + }, + { + "epoch": 0.019529006318207927, + "grad_norm": 0.2880657911300659, + "learning_rate": 2.55e-06, + "loss": 0.8065, + "step": 51 + }, + { + "epoch": 0.019911928010721808, + "grad_norm": 0.283342182636261, + "learning_rate": 2.6e-06, + "loss": 0.8131, + "step": 52 + }, + { + "epoch": 0.02029484970323569, + "grad_norm": 0.26725032925605774, + "learning_rate": 2.6500000000000005e-06, + "loss": 0.8193, + "step": 53 + }, + { + "epoch": 0.02067777139574957, + "grad_norm": 0.2662472128868103, + "learning_rate": 2.7000000000000004e-06, + "loss": 0.8171, + "step": 54 + }, + { + "epoch": 0.02106069308826345, + "grad_norm": 0.25971609354019165, + "learning_rate": 2.7500000000000004e-06, + "loss": 0.8067, + "step": 55 + }, + { + "epoch": 0.02144361478077733, + "grad_norm": 0.2470993995666504, + "learning_rate": 2.8000000000000003e-06, + "loss": 0.7906, + "step": 56 + }, + { + "epoch": 0.021826536473291212, + "grad_norm": 0.24413476884365082, + "learning_rate": 2.85e-06, + "loss": 0.813, + "step": 57 + }, + { + "epoch": 0.022209458165805093, + "grad_norm": 0.2281661480665207, + "learning_rate": 2.9e-06, + "loss": 0.8004, + "step": 58 + }, + { + "epoch": 0.022592379858318974, + "grad_norm": 0.2510649263858795, + "learning_rate": 2.95e-06, + "loss": 0.8144, + "step": 59 + }, + { + "epoch": 0.022975301550832855, + "grad_norm": 0.22505076229572296, + "learning_rate": 3e-06, + "loss": 0.7959, + "step": 60 + }, + { + "epoch": 0.023358223243346736, + "grad_norm": 0.22958475351333618, + "learning_rate": 3.05e-06, + "loss": 0.7918, + "step": 61 + }, + { + "epoch": 0.023741144935860617, + "grad_norm": 0.2242196947336197, + "learning_rate": 3.1000000000000004e-06, + "loss": 0.8004, + "step": 62 + }, + { + "epoch": 0.024124066628374498, + "grad_norm": 0.23278626799583435, + "learning_rate": 3.1500000000000003e-06, + "loss": 0.8133, + "step": 63 + }, + { + "epoch": 0.02450698832088838, + "grad_norm": 0.230966255068779, + "learning_rate": 3.2000000000000003e-06, + "loss": 0.8119, + "step": 64 + }, + { + "epoch": 0.02488991001340226, + "grad_norm": 0.21737371385097504, + "learning_rate": 3.2500000000000002e-06, + "loss": 0.8, + "step": 65 + }, + { + "epoch": 0.02527283170591614, + "grad_norm": 0.21784932911396027, + "learning_rate": 3.3000000000000006e-06, + "loss": 0.813, + "step": 66 + }, + { + "epoch": 0.02565575339843002, + "grad_norm": 0.21986863017082214, + "learning_rate": 3.3500000000000005e-06, + "loss": 0.7739, + "step": 67 + }, + { + "epoch": 0.026038675090943902, + "grad_norm": 0.2227390855550766, + "learning_rate": 3.4000000000000005e-06, + "loss": 0.7865, + "step": 68 + }, + { + "epoch": 0.026421596783457783, + "grad_norm": 0.2248571366071701, + "learning_rate": 3.45e-06, + "loss": 0.7959, + "step": 69 + }, + { + "epoch": 0.026804518475971664, + "grad_norm": 0.22709882259368896, + "learning_rate": 3.5e-06, + "loss": 0.8137, + "step": 70 + }, + { + "epoch": 0.027187440168485545, + "grad_norm": 0.2120588719844818, + "learning_rate": 3.5500000000000003e-06, + "loss": 0.7869, + "step": 71 + }, + { + "epoch": 0.027570361860999426, + "grad_norm": 0.20743295550346375, + "learning_rate": 3.6000000000000003e-06, + "loss": 0.7929, + "step": 72 + }, + { + "epoch": 0.027953283553513307, + "grad_norm": 0.20250606536865234, + "learning_rate": 3.65e-06, + "loss": 0.7967, + "step": 73 + }, + { + "epoch": 0.028336205246027188, + "grad_norm": 0.200389102101326, + "learning_rate": 3.7e-06, + "loss": 0.7764, + "step": 74 + }, + { + "epoch": 0.02871912693854107, + "grad_norm": 0.19319339096546173, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.7723, + "step": 75 + }, + { + "epoch": 0.02910204863105495, + "grad_norm": 0.19402000308036804, + "learning_rate": 3.8000000000000005e-06, + "loss": 0.8049, + "step": 76 + }, + { + "epoch": 0.02948497032356883, + "grad_norm": 0.1952792853116989, + "learning_rate": 3.85e-06, + "loss": 0.7721, + "step": 77 + }, + { + "epoch": 0.02986789201608271, + "grad_norm": 0.18888434767723083, + "learning_rate": 3.900000000000001e-06, + "loss": 0.7846, + "step": 78 + }, + { + "epoch": 0.030250813708596593, + "grad_norm": 0.1821848452091217, + "learning_rate": 3.95e-06, + "loss": 0.7753, + "step": 79 + }, + { + "epoch": 0.030633735401110473, + "grad_norm": 0.1820901334285736, + "learning_rate": 4.000000000000001e-06, + "loss": 0.7826, + "step": 80 + }, + { + "epoch": 0.031016657093624354, + "grad_norm": 0.19272948801517487, + "learning_rate": 4.05e-06, + "loss": 0.7989, + "step": 81 + }, + { + "epoch": 0.031399578786138235, + "grad_norm": 0.184035524725914, + "learning_rate": 4.1e-06, + "loss": 0.791, + "step": 82 + }, + { + "epoch": 0.031782500478652116, + "grad_norm": 0.18398649990558624, + "learning_rate": 4.15e-06, + "loss": 0.7768, + "step": 83 + }, + { + "epoch": 0.032165422171166, + "grad_norm": 0.18076884746551514, + "learning_rate": 4.2000000000000004e-06, + "loss": 0.7729, + "step": 84 + }, + { + "epoch": 0.03254834386367988, + "grad_norm": 0.18133755028247833, + "learning_rate": 4.25e-06, + "loss": 0.7854, + "step": 85 + }, + { + "epoch": 0.03293126555619376, + "grad_norm": 0.18535292148590088, + "learning_rate": 4.3e-06, + "loss": 0.774, + "step": 86 + }, + { + "epoch": 0.03331418724870764, + "grad_norm": 0.1858111023902893, + "learning_rate": 4.350000000000001e-06, + "loss": 0.7942, + "step": 87 + }, + { + "epoch": 0.03369710894122152, + "grad_norm": 0.1836749166250229, + "learning_rate": 4.4e-06, + "loss": 0.7923, + "step": 88 + }, + { + "epoch": 0.0340800306337354, + "grad_norm": 0.1837168186903, + "learning_rate": 4.450000000000001e-06, + "loss": 0.7763, + "step": 89 + }, + { + "epoch": 0.03446295232624928, + "grad_norm": 0.17813347280025482, + "learning_rate": 4.5e-06, + "loss": 0.7667, + "step": 90 + }, + { + "epoch": 0.034845874018763164, + "grad_norm": 0.1816832423210144, + "learning_rate": 4.5500000000000005e-06, + "loss": 0.7816, + "step": 91 + }, + { + "epoch": 0.035228795711277044, + "grad_norm": 0.1764446496963501, + "learning_rate": 4.600000000000001e-06, + "loss": 0.755, + "step": 92 + }, + { + "epoch": 0.035611717403790925, + "grad_norm": 0.18107102811336517, + "learning_rate": 4.65e-06, + "loss": 0.7808, + "step": 93 + }, + { + "epoch": 0.035994639096304806, + "grad_norm": 0.17818239331245422, + "learning_rate": 4.7e-06, + "loss": 0.7682, + "step": 94 + }, + { + "epoch": 0.03637756078881869, + "grad_norm": 0.1828058362007141, + "learning_rate": 4.75e-06, + "loss": 0.7313, + "step": 95 + }, + { + "epoch": 0.03676048248133257, + "grad_norm": 0.18249830603599548, + "learning_rate": 4.800000000000001e-06, + "loss": 0.7704, + "step": 96 + }, + { + "epoch": 0.03714340417384645, + "grad_norm": 0.177218496799469, + "learning_rate": 4.85e-06, + "loss": 0.7691, + "step": 97 + }, + { + "epoch": 0.03752632586636033, + "grad_norm": 0.17561639845371246, + "learning_rate": 4.9000000000000005e-06, + "loss": 0.7679, + "step": 98 + }, + { + "epoch": 0.03790924755887421, + "grad_norm": 0.18068744242191315, + "learning_rate": 4.95e-06, + "loss": 0.7807, + "step": 99 + }, + { + "epoch": 0.03829216925138809, + "grad_norm": 0.1780369132757187, + "learning_rate": 5e-06, + "loss": 0.7685, + "step": 100 + }, + { + "epoch": 0.03867509094390197, + "grad_norm": 0.1759374588727951, + "learning_rate": 4.99999979369311e-06, + "loss": 0.7534, + "step": 101 + }, + { + "epoch": 0.039058012636415854, + "grad_norm": 0.1728181093931198, + "learning_rate": 4.999999174772476e-06, + "loss": 0.7606, + "step": 102 + }, + { + "epoch": 0.039440934328929735, + "grad_norm": 0.17361775040626526, + "learning_rate": 4.999998143238196e-06, + "loss": 0.7729, + "step": 103 + }, + { + "epoch": 0.039823856021443615, + "grad_norm": 0.17070014774799347, + "learning_rate": 4.999996699090443e-06, + "loss": 0.7979, + "step": 104 + }, + { + "epoch": 0.040206777713957496, + "grad_norm": 0.19051475822925568, + "learning_rate": 4.999994842329455e-06, + "loss": 0.7766, + "step": 105 + }, + { + "epoch": 0.04058969940647138, + "grad_norm": 0.17698678374290466, + "learning_rate": 4.99999257295554e-06, + "loss": 0.7636, + "step": 106 + }, + { + "epoch": 0.04097262109898526, + "grad_norm": 0.17492914199829102, + "learning_rate": 4.999989890969069e-06, + "loss": 0.7558, + "step": 107 + }, + { + "epoch": 0.04135554279149914, + "grad_norm": 0.17513518035411835, + "learning_rate": 4.9999867963704875e-06, + "loss": 0.7501, + "step": 108 + }, + { + "epoch": 0.04173846448401302, + "grad_norm": 0.1762644648551941, + "learning_rate": 4.9999832891603055e-06, + "loss": 0.7677, + "step": 109 + }, + { + "epoch": 0.0421213861765269, + "grad_norm": 0.1700354665517807, + "learning_rate": 4.999979369339102e-06, + "loss": 0.7462, + "step": 110 + }, + { + "epoch": 0.04250430786904078, + "grad_norm": 0.17454922199249268, + "learning_rate": 4.999975036907523e-06, + "loss": 0.7501, + "step": 111 + }, + { + "epoch": 0.04288722956155466, + "grad_norm": 0.1729852706193924, + "learning_rate": 4.9999702918662855e-06, + "loss": 0.7917, + "step": 112 + }, + { + "epoch": 0.043270151254068544, + "grad_norm": 0.171586275100708, + "learning_rate": 4.99996513421617e-06, + "loss": 0.7846, + "step": 113 + }, + { + "epoch": 0.043653072946582425, + "grad_norm": 0.17341765761375427, + "learning_rate": 4.99995956395803e-06, + "loss": 0.7652, + "step": 114 + }, + { + "epoch": 0.044035994639096306, + "grad_norm": 0.1708601862192154, + "learning_rate": 4.999953581092784e-06, + "loss": 0.7605, + "step": 115 + }, + { + "epoch": 0.044418916331610186, + "grad_norm": 0.17076106369495392, + "learning_rate": 4.999947185621419e-06, + "loss": 0.7434, + "step": 116 + }, + { + "epoch": 0.04480183802412407, + "grad_norm": 0.1779450923204422, + "learning_rate": 4.999940377544991e-06, + "loss": 0.7754, + "step": 117 + }, + { + "epoch": 0.04518475971663795, + "grad_norm": 0.17263121902942657, + "learning_rate": 4.999933156864624e-06, + "loss": 0.7504, + "step": 118 + }, + { + "epoch": 0.04556768140915183, + "grad_norm": 0.17754054069519043, + "learning_rate": 4.999925523581508e-06, + "loss": 0.7669, + "step": 119 + }, + { + "epoch": 0.04595060310166571, + "grad_norm": 0.1822337955236435, + "learning_rate": 4.999917477696906e-06, + "loss": 0.769, + "step": 120 + }, + { + "epoch": 0.04633352479417959, + "grad_norm": 0.17491310834884644, + "learning_rate": 4.999909019212144e-06, + "loss": 0.7679, + "step": 121 + }, + { + "epoch": 0.04671644648669347, + "grad_norm": 0.17979766428470612, + "learning_rate": 4.999900148128617e-06, + "loss": 0.7584, + "step": 122 + }, + { + "epoch": 0.04709936817920735, + "grad_norm": 0.17815561592578888, + "learning_rate": 4.999890864447791e-06, + "loss": 0.7806, + "step": 123 + }, + { + "epoch": 0.047482289871721234, + "grad_norm": 0.18036983907222748, + "learning_rate": 4.999881168171197e-06, + "loss": 0.7639, + "step": 124 + }, + { + "epoch": 0.047865211564235115, + "grad_norm": 0.1740260273218155, + "learning_rate": 4.999871059300436e-06, + "loss": 0.7672, + "step": 125 + }, + { + "epoch": 0.048248133256748996, + "grad_norm": 0.17238426208496094, + "learning_rate": 4.9998605378371765e-06, + "loss": 0.762, + "step": 126 + }, + { + "epoch": 0.04863105494926288, + "grad_norm": 0.17532694339752197, + "learning_rate": 4.999849603783155e-06, + "loss": 0.7559, + "step": 127 + }, + { + "epoch": 0.04901397664177676, + "grad_norm": 0.18054932355880737, + "learning_rate": 4.999838257140176e-06, + "loss": 0.7727, + "step": 128 + }, + { + "epoch": 0.04939689833429064, + "grad_norm": 0.18592335283756256, + "learning_rate": 4.999826497910112e-06, + "loss": 0.741, + "step": 129 + }, + { + "epoch": 0.04977982002680452, + "grad_norm": 0.1814471334218979, + "learning_rate": 4.999814326094904e-06, + "loss": 0.7467, + "step": 130 + }, + { + "epoch": 0.0501627417193184, + "grad_norm": 0.17422868311405182, + "learning_rate": 4.999801741696561e-06, + "loss": 0.7415, + "step": 131 + }, + { + "epoch": 0.05054566341183228, + "grad_norm": 0.17902833223342896, + "learning_rate": 4.99978874471716e-06, + "loss": 0.7771, + "step": 132 + }, + { + "epoch": 0.05092858510434616, + "grad_norm": 0.17519019544124603, + "learning_rate": 4.999775335158845e-06, + "loss": 0.769, + "step": 133 + }, + { + "epoch": 0.05131150679686004, + "grad_norm": 0.17666171491146088, + "learning_rate": 4.999761513023831e-06, + "loss": 0.7368, + "step": 134 + }, + { + "epoch": 0.051694428489373924, + "grad_norm": 0.17427121102809906, + "learning_rate": 4.999747278314398e-06, + "loss": 0.743, + "step": 135 + }, + { + "epoch": 0.052077350181887805, + "grad_norm": 0.17496870458126068, + "learning_rate": 4.9997326310328954e-06, + "loss": 0.7399, + "step": 136 + }, + { + "epoch": 0.052460271874401686, + "grad_norm": 0.17835672199726105, + "learning_rate": 4.999717571181742e-06, + "loss": 0.7491, + "step": 137 + }, + { + "epoch": 0.05284319356691557, + "grad_norm": 0.17227713763713837, + "learning_rate": 4.999702098763422e-06, + "loss": 0.7351, + "step": 138 + }, + { + "epoch": 0.05322611525942945, + "grad_norm": 0.17822884023189545, + "learning_rate": 4.999686213780489e-06, + "loss": 0.765, + "step": 139 + }, + { + "epoch": 0.05360903695194333, + "grad_norm": 0.17335212230682373, + "learning_rate": 4.999669916235566e-06, + "loss": 0.7558, + "step": 140 + }, + { + "epoch": 0.05399195864445721, + "grad_norm": 0.18416562676429749, + "learning_rate": 4.999653206131342e-06, + "loss": 0.7746, + "step": 141 + }, + { + "epoch": 0.05437488033697109, + "grad_norm": 0.1746789515018463, + "learning_rate": 4.999636083470574e-06, + "loss": 0.7386, + "step": 142 + }, + { + "epoch": 0.05475780202948497, + "grad_norm": 0.17718607187271118, + "learning_rate": 4.99961854825609e-06, + "loss": 0.7565, + "step": 143 + }, + { + "epoch": 0.05514072372199885, + "grad_norm": 0.17604880034923553, + "learning_rate": 4.999600600490783e-06, + "loss": 0.7601, + "step": 144 + }, + { + "epoch": 0.05552364541451273, + "grad_norm": 0.18366846442222595, + "learning_rate": 4.999582240177615e-06, + "loss": 0.7615, + "step": 145 + }, + { + "epoch": 0.055906567107026614, + "grad_norm": 0.18398283421993256, + "learning_rate": 4.999563467319616e-06, + "loss": 0.7554, + "step": 146 + }, + { + "epoch": 0.056289488799540495, + "grad_norm": 0.1746983826160431, + "learning_rate": 4.9995442819198856e-06, + "loss": 0.7382, + "step": 147 + }, + { + "epoch": 0.056672410492054376, + "grad_norm": 0.17418019473552704, + "learning_rate": 4.99952468398159e-06, + "loss": 0.7298, + "step": 148 + }, + { + "epoch": 0.05705533218456826, + "grad_norm": 0.17950575053691864, + "learning_rate": 4.999504673507964e-06, + "loss": 0.7603, + "step": 149 + }, + { + "epoch": 0.05743825387708214, + "grad_norm": 0.18124008178710938, + "learning_rate": 4.999484250502308e-06, + "loss": 0.7541, + "step": 150 + }, + { + "epoch": 0.05782117556959602, + "grad_norm": 0.1779160499572754, + "learning_rate": 4.999463414967994e-06, + "loss": 0.7268, + "step": 151 + }, + { + "epoch": 0.0582040972621099, + "grad_norm": 0.17948390543460846, + "learning_rate": 4.999442166908462e-06, + "loss": 0.7427, + "step": 152 + }, + { + "epoch": 0.05858701895462378, + "grad_norm": 0.17860113084316254, + "learning_rate": 4.999420506327219e-06, + "loss": 0.7661, + "step": 153 + }, + { + "epoch": 0.05896994064713766, + "grad_norm": 0.18167608976364136, + "learning_rate": 4.999398433227838e-06, + "loss": 0.7534, + "step": 154 + }, + { + "epoch": 0.05935286233965154, + "grad_norm": 0.1777266561985016, + "learning_rate": 4.999375947613963e-06, + "loss": 0.7631, + "step": 155 + }, + { + "epoch": 0.05973578403216542, + "grad_norm": 0.18054281175136566, + "learning_rate": 4.999353049489306e-06, + "loss": 0.75, + "step": 156 + }, + { + "epoch": 0.060118705724679304, + "grad_norm": 0.1803414523601532, + "learning_rate": 4.999329738857645e-06, + "loss": 0.7527, + "step": 157 + }, + { + "epoch": 0.060501627417193185, + "grad_norm": 0.18072110414505005, + "learning_rate": 4.999306015722828e-06, + "loss": 0.7661, + "step": 158 + }, + { + "epoch": 0.060884549109707066, + "grad_norm": 0.18042054772377014, + "learning_rate": 4.99928188008877e-06, + "loss": 0.7686, + "step": 159 + }, + { + "epoch": 0.06126747080222095, + "grad_norm": 0.17903763055801392, + "learning_rate": 4.9992573319594554e-06, + "loss": 0.7239, + "step": 160 + }, + { + "epoch": 0.06165039249473483, + "grad_norm": 0.17662514746189117, + "learning_rate": 4.999232371338934e-06, + "loss": 0.7362, + "step": 161 + }, + { + "epoch": 0.06203331418724871, + "grad_norm": 0.1817161589860916, + "learning_rate": 4.999206998231326e-06, + "loss": 0.7553, + "step": 162 + }, + { + "epoch": 0.06241623587976259, + "grad_norm": 0.1797408014535904, + "learning_rate": 4.9991812126408215e-06, + "loss": 0.7302, + "step": 163 + }, + { + "epoch": 0.06279915757227647, + "grad_norm": 0.19102029502391815, + "learning_rate": 4.999155014571673e-06, + "loss": 0.7447, + "step": 164 + }, + { + "epoch": 0.06318207926479034, + "grad_norm": 0.18745839595794678, + "learning_rate": 4.9991284040282064e-06, + "loss": 0.76, + "step": 165 + }, + { + "epoch": 0.06356500095730423, + "grad_norm": 0.18368470668792725, + "learning_rate": 4.999101381014813e-06, + "loss": 0.7423, + "step": 166 + }, + { + "epoch": 0.0639479226498181, + "grad_norm": 0.18375608325004578, + "learning_rate": 4.999073945535952e-06, + "loss": 0.7235, + "step": 167 + }, + { + "epoch": 0.064330844342332, + "grad_norm": 0.17745068669319153, + "learning_rate": 4.9990460975961525e-06, + "loss": 0.7538, + "step": 168 + }, + { + "epoch": 0.06471376603484587, + "grad_norm": 0.18004310131072998, + "learning_rate": 4.999017837200011e-06, + "loss": 0.7487, + "step": 169 + }, + { + "epoch": 0.06509668772735976, + "grad_norm": 0.1802518218755722, + "learning_rate": 4.998989164352191e-06, + "loss": 0.7311, + "step": 170 + }, + { + "epoch": 0.06547960941987363, + "grad_norm": 0.18313384056091309, + "learning_rate": 4.998960079057426e-06, + "loss": 0.7379, + "step": 171 + }, + { + "epoch": 0.06586253111238752, + "grad_norm": 0.1821993589401245, + "learning_rate": 4.9989305813205134e-06, + "loss": 0.7465, + "step": 172 + }, + { + "epoch": 0.06624545280490139, + "grad_norm": 0.18214775621891022, + "learning_rate": 4.998900671146326e-06, + "loss": 0.7282, + "step": 173 + }, + { + "epoch": 0.06662837449741528, + "grad_norm": 0.18141110241413116, + "learning_rate": 4.998870348539797e-06, + "loss": 0.7435, + "step": 174 + }, + { + "epoch": 0.06701129618992915, + "grad_norm": 0.18263347446918488, + "learning_rate": 4.998839613505932e-06, + "loss": 0.7287, + "step": 175 + }, + { + "epoch": 0.06739421788244304, + "grad_norm": 0.1850353181362152, + "learning_rate": 4.998808466049803e-06, + "loss": 0.7257, + "step": 176 + }, + { + "epoch": 0.06777713957495692, + "grad_norm": 0.18254730105400085, + "learning_rate": 4.998776906176553e-06, + "loss": 0.7411, + "step": 177 + }, + { + "epoch": 0.0681600612674708, + "grad_norm": 0.18580451607704163, + "learning_rate": 4.99874493389139e-06, + "loss": 0.757, + "step": 178 + }, + { + "epoch": 0.06854298295998468, + "grad_norm": 0.18148383498191833, + "learning_rate": 4.998712549199589e-06, + "loss": 0.7234, + "step": 179 + }, + { + "epoch": 0.06892590465249857, + "grad_norm": 0.18495115637779236, + "learning_rate": 4.998679752106496e-06, + "loss": 0.7551, + "step": 180 + }, + { + "epoch": 0.06930882634501244, + "grad_norm": 0.18330596387386322, + "learning_rate": 4.998646542617525e-06, + "loss": 0.7452, + "step": 181 + }, + { + "epoch": 0.06969174803752633, + "grad_norm": 0.17899492383003235, + "learning_rate": 4.9986129207381555e-06, + "loss": 0.7054, + "step": 182 + }, + { + "epoch": 0.0700746697300402, + "grad_norm": 0.1845427006483078, + "learning_rate": 4.9985788864739375e-06, + "loss": 0.7393, + "step": 183 + }, + { + "epoch": 0.07045759142255409, + "grad_norm": 0.18768930435180664, + "learning_rate": 4.998544439830488e-06, + "loss": 0.7473, + "step": 184 + }, + { + "epoch": 0.07084051311506796, + "grad_norm": 0.18853271007537842, + "learning_rate": 4.998509580813493e-06, + "loss": 0.7192, + "step": 185 + }, + { + "epoch": 0.07122343480758185, + "grad_norm": 0.19321995973587036, + "learning_rate": 4.998474309428705e-06, + "loss": 0.7518, + "step": 186 + }, + { + "epoch": 0.07160635650009572, + "grad_norm": 0.18516433238983154, + "learning_rate": 4.9984386256819455e-06, + "loss": 0.7253, + "step": 187 + }, + { + "epoch": 0.07198927819260961, + "grad_norm": 0.18672487139701843, + "learning_rate": 4.998402529579104e-06, + "loss": 0.75, + "step": 188 + }, + { + "epoch": 0.07237219988512349, + "grad_norm": 0.19649949669837952, + "learning_rate": 4.998366021126139e-06, + "loss": 0.7248, + "step": 189 + }, + { + "epoch": 0.07275512157763737, + "grad_norm": 0.1804923266172409, + "learning_rate": 4.998329100329073e-06, + "loss": 0.7222, + "step": 190 + }, + { + "epoch": 0.07313804327015125, + "grad_norm": 0.19032728672027588, + "learning_rate": 4.998291767194003e-06, + "loss": 0.7319, + "step": 191 + }, + { + "epoch": 0.07352096496266514, + "grad_norm": 0.1883210986852646, + "learning_rate": 4.998254021727089e-06, + "loss": 0.7417, + "step": 192 + }, + { + "epoch": 0.07390388665517901, + "grad_norm": 0.17950430512428284, + "learning_rate": 4.998215863934561e-06, + "loss": 0.7145, + "step": 193 + }, + { + "epoch": 0.0742868083476929, + "grad_norm": 0.1965007483959198, + "learning_rate": 4.998177293822718e-06, + "loss": 0.7482, + "step": 194 + }, + { + "epoch": 0.07466973004020677, + "grad_norm": 0.18340559303760529, + "learning_rate": 4.998138311397924e-06, + "loss": 0.7101, + "step": 195 + }, + { + "epoch": 0.07505265173272066, + "grad_norm": 0.18560108542442322, + "learning_rate": 4.998098916666613e-06, + "loss": 0.7384, + "step": 196 + }, + { + "epoch": 0.07543557342523453, + "grad_norm": 0.20068591833114624, + "learning_rate": 4.9980591096352885e-06, + "loss": 0.7537, + "step": 197 + }, + { + "epoch": 0.07581849511774842, + "grad_norm": 0.18525315821170807, + "learning_rate": 4.998018890310518e-06, + "loss": 0.7359, + "step": 198 + }, + { + "epoch": 0.0762014168102623, + "grad_norm": 0.18367959558963776, + "learning_rate": 4.997978258698942e-06, + "loss": 0.6975, + "step": 199 + }, + { + "epoch": 0.07658433850277618, + "grad_norm": 0.18198728561401367, + "learning_rate": 4.997937214807265e-06, + "loss": 0.7338, + "step": 200 + }, + { + "epoch": 0.07696726019529006, + "grad_norm": 0.18427884578704834, + "learning_rate": 4.997895758642262e-06, + "loss": 0.7367, + "step": 201 + }, + { + "epoch": 0.07735018188780395, + "grad_norm": 0.18311342597007751, + "learning_rate": 4.997853890210775e-06, + "loss": 0.7326, + "step": 202 + }, + { + "epoch": 0.07773310358031782, + "grad_norm": 0.18621565401554108, + "learning_rate": 4.997811609519714e-06, + "loss": 0.737, + "step": 203 + }, + { + "epoch": 0.07811602527283171, + "grad_norm": 0.1909012496471405, + "learning_rate": 4.997768916576057e-06, + "loss": 0.7481, + "step": 204 + }, + { + "epoch": 0.07849894696534558, + "grad_norm": 0.1837858110666275, + "learning_rate": 4.99772581138685e-06, + "loss": 0.7266, + "step": 205 + }, + { + "epoch": 0.07888186865785947, + "grad_norm": 0.1804484724998474, + "learning_rate": 4.9976822939592085e-06, + "loss": 0.7187, + "step": 206 + }, + { + "epoch": 0.07926479035037334, + "grad_norm": 0.1876893788576126, + "learning_rate": 4.997638364300314e-06, + "loss": 0.7199, + "step": 207 + }, + { + "epoch": 0.07964771204288723, + "grad_norm": 0.1820950210094452, + "learning_rate": 4.997594022417417e-06, + "loss": 0.7137, + "step": 208 + }, + { + "epoch": 0.0800306337354011, + "grad_norm": 0.18648962676525116, + "learning_rate": 4.997549268317836e-06, + "loss": 0.7309, + "step": 209 + }, + { + "epoch": 0.08041355542791499, + "grad_norm": 0.200729101896286, + "learning_rate": 4.997504102008957e-06, + "loss": 0.7374, + "step": 210 + }, + { + "epoch": 0.08079647712042887, + "grad_norm": 0.18724213540554047, + "learning_rate": 4.997458523498236e-06, + "loss": 0.7415, + "step": 211 + }, + { + "epoch": 0.08117939881294275, + "grad_norm": 0.1862158477306366, + "learning_rate": 4.997412532793195e-06, + "loss": 0.7286, + "step": 212 + }, + { + "epoch": 0.08156232050545663, + "grad_norm": 0.206449493765831, + "learning_rate": 4.997366129901422e-06, + "loss": 0.7608, + "step": 213 + }, + { + "epoch": 0.08194524219797052, + "grad_norm": 0.18451204895973206, + "learning_rate": 4.9973193148305784e-06, + "loss": 0.7334, + "step": 214 + }, + { + "epoch": 0.08232816389048439, + "grad_norm": 0.1887936145067215, + "learning_rate": 4.997272087588391e-06, + "loss": 0.7387, + "step": 215 + }, + { + "epoch": 0.08271108558299828, + "grad_norm": 0.1905495524406433, + "learning_rate": 4.9972244481826524e-06, + "loss": 0.7284, + "step": 216 + }, + { + "epoch": 0.08309400727551215, + "grad_norm": 0.18892931938171387, + "learning_rate": 4.997176396621227e-06, + "loss": 0.7268, + "step": 217 + }, + { + "epoch": 0.08347692896802604, + "grad_norm": 0.1919863075017929, + "learning_rate": 4.9971279329120445e-06, + "loss": 0.7433, + "step": 218 + }, + { + "epoch": 0.08385985066053991, + "grad_norm": 0.192040354013443, + "learning_rate": 4.9970790570631045e-06, + "loss": 0.7529, + "step": 219 + }, + { + "epoch": 0.0842427723530538, + "grad_norm": 0.19068577885627747, + "learning_rate": 4.9970297690824735e-06, + "loss": 0.74, + "step": 220 + }, + { + "epoch": 0.08462569404556768, + "grad_norm": 0.19420921802520752, + "learning_rate": 4.996980068978285e-06, + "loss": 0.7388, + "step": 221 + }, + { + "epoch": 0.08500861573808156, + "grad_norm": 0.19175250828266144, + "learning_rate": 4.9969299567587435e-06, + "loss": 0.7227, + "step": 222 + }, + { + "epoch": 0.08539153743059544, + "grad_norm": 0.19421349465847015, + "learning_rate": 4.996879432432118e-06, + "loss": 0.7348, + "step": 223 + }, + { + "epoch": 0.08577445912310933, + "grad_norm": 0.18917426466941833, + "learning_rate": 4.996828496006749e-06, + "loss": 0.7273, + "step": 224 + }, + { + "epoch": 0.0861573808156232, + "grad_norm": 0.1845076084136963, + "learning_rate": 4.996777147491043e-06, + "loss": 0.7105, + "step": 225 + }, + { + "epoch": 0.08654030250813709, + "grad_norm": 0.18641623854637146, + "learning_rate": 4.996725386893475e-06, + "loss": 0.7244, + "step": 226 + }, + { + "epoch": 0.08692322420065096, + "grad_norm": 0.19008800387382507, + "learning_rate": 4.996673214222586e-06, + "loss": 0.7281, + "step": 227 + }, + { + "epoch": 0.08730614589316485, + "grad_norm": 0.19295676052570343, + "learning_rate": 4.99662062948699e-06, + "loss": 0.7342, + "step": 228 + }, + { + "epoch": 0.08768906758567872, + "grad_norm": 0.1977388709783554, + "learning_rate": 4.9965676326953624e-06, + "loss": 0.7277, + "step": 229 + }, + { + "epoch": 0.08807198927819261, + "grad_norm": 0.18937358260154724, + "learning_rate": 4.996514223856453e-06, + "loss": 0.7153, + "step": 230 + }, + { + "epoch": 0.08845491097070649, + "grad_norm": 0.19146166741847992, + "learning_rate": 4.996460402979076e-06, + "loss": 0.7255, + "step": 231 + }, + { + "epoch": 0.08883783266322037, + "grad_norm": 0.20314456522464752, + "learning_rate": 4.9964061700721124e-06, + "loss": 0.7245, + "step": 232 + }, + { + "epoch": 0.08922075435573425, + "grad_norm": 0.19764558970928192, + "learning_rate": 4.996351525144515e-06, + "loss": 0.7496, + "step": 233 + }, + { + "epoch": 0.08960367604824813, + "grad_norm": 0.19225457310676575, + "learning_rate": 4.9962964682053015e-06, + "loss": 0.7286, + "step": 234 + }, + { + "epoch": 0.08998659774076201, + "grad_norm": 0.19454413652420044, + "learning_rate": 4.996240999263561e-06, + "loss": 0.7274, + "step": 235 + }, + { + "epoch": 0.0903695194332759, + "grad_norm": 0.19086669385433197, + "learning_rate": 4.996185118328446e-06, + "loss": 0.7282, + "step": 236 + }, + { + "epoch": 0.09075244112578977, + "grad_norm": 0.19874230027198792, + "learning_rate": 4.99612882540918e-06, + "loss": 0.7324, + "step": 237 + }, + { + "epoch": 0.09113536281830366, + "grad_norm": 0.19608773291110992, + "learning_rate": 4.9960721205150545e-06, + "loss": 0.7302, + "step": 238 + }, + { + "epoch": 0.09151828451081753, + "grad_norm": 0.19033324718475342, + "learning_rate": 4.996015003655429e-06, + "loss": 0.7208, + "step": 239 + }, + { + "epoch": 0.09190120620333142, + "grad_norm": 0.19876620173454285, + "learning_rate": 4.995957474839728e-06, + "loss": 0.7351, + "step": 240 + }, + { + "epoch": 0.0922841278958453, + "grad_norm": 0.19782209396362305, + "learning_rate": 4.9958995340774486e-06, + "loss": 0.7312, + "step": 241 + }, + { + "epoch": 0.09266704958835918, + "grad_norm": 0.19374790787696838, + "learning_rate": 4.995841181378153e-06, + "loss": 0.7439, + "step": 242 + }, + { + "epoch": 0.09304997128087306, + "grad_norm": 0.19944606721401215, + "learning_rate": 4.995782416751472e-06, + "loss": 0.7373, + "step": 243 + }, + { + "epoch": 0.09343289297338694, + "grad_norm": 0.19650287926197052, + "learning_rate": 4.9957232402071045e-06, + "loss": 0.71, + "step": 244 + }, + { + "epoch": 0.09381581466590082, + "grad_norm": 0.19653984904289246, + "learning_rate": 4.995663651754816e-06, + "loss": 0.7128, + "step": 245 + }, + { + "epoch": 0.0941987363584147, + "grad_norm": 0.19898127019405365, + "learning_rate": 4.995603651404444e-06, + "loss": 0.7288, + "step": 246 + }, + { + "epoch": 0.09458165805092858, + "grad_norm": 0.19377675652503967, + "learning_rate": 4.9955432391658886e-06, + "loss": 0.7256, + "step": 247 + }, + { + "epoch": 0.09496457974344247, + "grad_norm": 0.2016996592283249, + "learning_rate": 4.995482415049123e-06, + "loss": 0.7465, + "step": 248 + }, + { + "epoch": 0.09534750143595634, + "grad_norm": 0.19421467185020447, + "learning_rate": 4.995421179064183e-06, + "loss": 0.7215, + "step": 249 + }, + { + "epoch": 0.09573042312847023, + "grad_norm": 0.19672785699367523, + "learning_rate": 4.995359531221179e-06, + "loss": 0.7218, + "step": 250 + }, + { + "epoch": 0.0961133448209841, + "grad_norm": 0.20065009593963623, + "learning_rate": 4.995297471530283e-06, + "loss": 0.7325, + "step": 251 + }, + { + "epoch": 0.09649626651349799, + "grad_norm": 0.1908031553030014, + "learning_rate": 4.995235000001739e-06, + "loss": 0.7283, + "step": 252 + }, + { + "epoch": 0.09687918820601187, + "grad_norm": 0.19145245850086212, + "learning_rate": 4.995172116645856e-06, + "loss": 0.7255, + "step": 253 + }, + { + "epoch": 0.09726210989852575, + "grad_norm": 0.1942753791809082, + "learning_rate": 4.995108821473014e-06, + "loss": 0.7038, + "step": 254 + }, + { + "epoch": 0.09764503159103963, + "grad_norm": 0.19058293104171753, + "learning_rate": 4.99504511449366e-06, + "loss": 0.7134, + "step": 255 + }, + { + "epoch": 0.09802795328355352, + "grad_norm": 0.19463270902633667, + "learning_rate": 4.994980995718306e-06, + "loss": 0.7274, + "step": 256 + }, + { + "epoch": 0.09841087497606739, + "grad_norm": 0.19273309409618378, + "learning_rate": 4.994916465157538e-06, + "loss": 0.7141, + "step": 257 + }, + { + "epoch": 0.09879379666858128, + "grad_norm": 0.18812306225299835, + "learning_rate": 4.994851522822004e-06, + "loss": 0.7033, + "step": 258 + }, + { + "epoch": 0.09917671836109515, + "grad_norm": 0.1985722929239273, + "learning_rate": 4.994786168722424e-06, + "loss": 0.7086, + "step": 259 + }, + { + "epoch": 0.09955964005360904, + "grad_norm": 0.2013116478919983, + "learning_rate": 4.994720402869583e-06, + "loss": 0.7115, + "step": 260 + }, + { + "epoch": 0.09994256174612291, + "grad_norm": 0.20156681537628174, + "learning_rate": 4.994654225274337e-06, + "loss": 0.7357, + "step": 261 + }, + { + "epoch": 0.1003254834386368, + "grad_norm": 0.20074698328971863, + "learning_rate": 4.994587635947607e-06, + "loss": 0.7185, + "step": 262 + }, + { + "epoch": 0.10070840513115067, + "grad_norm": 0.20263032615184784, + "learning_rate": 4.994520634900383e-06, + "loss": 0.7151, + "step": 263 + }, + { + "epoch": 0.10109132682366456, + "grad_norm": 0.2002752721309662, + "learning_rate": 4.994453222143724e-06, + "loss": 0.7333, + "step": 264 + }, + { + "epoch": 0.10147424851617844, + "grad_norm": 0.19758231937885284, + "learning_rate": 4.994385397688756e-06, + "loss": 0.7337, + "step": 265 + }, + { + "epoch": 0.10185717020869232, + "grad_norm": 0.19643495976924896, + "learning_rate": 4.994317161546672e-06, + "loss": 0.756, + "step": 266 + }, + { + "epoch": 0.1022400919012062, + "grad_norm": 0.19787892699241638, + "learning_rate": 4.994248513728737e-06, + "loss": 0.7251, + "step": 267 + }, + { + "epoch": 0.10262301359372009, + "grad_norm": 0.2025986909866333, + "learning_rate": 4.994179454246279e-06, + "loss": 0.7428, + "step": 268 + }, + { + "epoch": 0.10300593528623396, + "grad_norm": 0.1986268311738968, + "learning_rate": 4.994109983110695e-06, + "loss": 0.7141, + "step": 269 + }, + { + "epoch": 0.10338885697874785, + "grad_norm": 0.1972547024488449, + "learning_rate": 4.9940401003334525e-06, + "loss": 0.717, + "step": 270 + }, + { + "epoch": 0.10377177867126172, + "grad_norm": 0.20430248975753784, + "learning_rate": 4.993969805926085e-06, + "loss": 0.7417, + "step": 271 + }, + { + "epoch": 0.10415470036377561, + "grad_norm": 0.198611319065094, + "learning_rate": 4.993899099900194e-06, + "loss": 0.7337, + "step": 272 + }, + { + "epoch": 0.10453762205628948, + "grad_norm": 0.2080761045217514, + "learning_rate": 4.993827982267451e-06, + "loss": 0.6972, + "step": 273 + }, + { + "epoch": 0.10492054374880337, + "grad_norm": 0.2025095522403717, + "learning_rate": 4.99375645303959e-06, + "loss": 0.7337, + "step": 274 + }, + { + "epoch": 0.10530346544131725, + "grad_norm": 0.20767749845981598, + "learning_rate": 4.993684512228419e-06, + "loss": 0.708, + "step": 275 + }, + { + "epoch": 0.10568638713383113, + "grad_norm": 0.20261400938034058, + "learning_rate": 4.993612159845812e-06, + "loss": 0.7095, + "step": 276 + }, + { + "epoch": 0.10606930882634501, + "grad_norm": 0.211074098944664, + "learning_rate": 4.993539395903709e-06, + "loss": 0.7349, + "step": 277 + }, + { + "epoch": 0.1064522305188589, + "grad_norm": 0.20446090400218964, + "learning_rate": 4.99346622041412e-06, + "loss": 0.7142, + "step": 278 + }, + { + "epoch": 0.10683515221137277, + "grad_norm": 0.1996392160654068, + "learning_rate": 4.993392633389123e-06, + "loss": 0.7298, + "step": 279 + }, + { + "epoch": 0.10721807390388666, + "grad_norm": 0.20801635086536407, + "learning_rate": 4.99331863484086e-06, + "loss": 0.7335, + "step": 280 + }, + { + "epoch": 0.10760099559640053, + "grad_norm": 0.20142178237438202, + "learning_rate": 4.993244224781549e-06, + "loss": 0.7336, + "step": 281 + }, + { + "epoch": 0.10798391728891442, + "grad_norm": 0.20546960830688477, + "learning_rate": 4.993169403223468e-06, + "loss": 0.7472, + "step": 282 + }, + { + "epoch": 0.10836683898142829, + "grad_norm": 0.20113351941108704, + "learning_rate": 4.993094170178966e-06, + "loss": 0.7222, + "step": 283 + }, + { + "epoch": 0.10874976067394218, + "grad_norm": 0.20306220650672913, + "learning_rate": 4.99301852566046e-06, + "loss": 0.7059, + "step": 284 + }, + { + "epoch": 0.10913268236645605, + "grad_norm": 0.2000206559896469, + "learning_rate": 4.992942469680437e-06, + "loss": 0.7042, + "step": 285 + }, + { + "epoch": 0.10951560405896994, + "grad_norm": 0.20378772914409637, + "learning_rate": 4.992866002251446e-06, + "loss": 0.7338, + "step": 286 + }, + { + "epoch": 0.10989852575148382, + "grad_norm": 0.19586919248104095, + "learning_rate": 4.99278912338611e-06, + "loss": 0.7108, + "step": 287 + }, + { + "epoch": 0.1102814474439977, + "grad_norm": 0.19886094331741333, + "learning_rate": 4.992711833097117e-06, + "loss": 0.7137, + "step": 288 + }, + { + "epoch": 0.11066436913651158, + "grad_norm": 0.20048004388809204, + "learning_rate": 4.992634131397223e-06, + "loss": 0.7281, + "step": 289 + }, + { + "epoch": 0.11104729082902547, + "grad_norm": 0.2020483762025833, + "learning_rate": 4.992556018299255e-06, + "loss": 0.7294, + "step": 290 + }, + { + "epoch": 0.11143021252153934, + "grad_norm": 0.1963844746351242, + "learning_rate": 4.9924774938161015e-06, + "loss": 0.7267, + "step": 291 + }, + { + "epoch": 0.11181313421405323, + "grad_norm": 0.1980431079864502, + "learning_rate": 4.992398557960724e-06, + "loss": 0.7172, + "step": 292 + }, + { + "epoch": 0.1121960559065671, + "grad_norm": 0.19689813256263733, + "learning_rate": 4.99231921074615e-06, + "loss": 0.6962, + "step": 293 + }, + { + "epoch": 0.11257897759908099, + "grad_norm": 0.20420525968074799, + "learning_rate": 4.992239452185476e-06, + "loss": 0.7157, + "step": 294 + }, + { + "epoch": 0.11296189929159486, + "grad_norm": 0.2101951539516449, + "learning_rate": 4.992159282291868e-06, + "loss": 0.7417, + "step": 295 + }, + { + "epoch": 0.11334482098410875, + "grad_norm": 0.2151687741279602, + "learning_rate": 4.9920787010785536e-06, + "loss": 0.7279, + "step": 296 + }, + { + "epoch": 0.11372774267662263, + "grad_norm": 0.20385341346263885, + "learning_rate": 4.991997708558835e-06, + "loss": 0.7436, + "step": 297 + }, + { + "epoch": 0.11411066436913651, + "grad_norm": 0.21158473193645477, + "learning_rate": 4.991916304746077e-06, + "loss": 0.7322, + "step": 298 + }, + { + "epoch": 0.11449358606165039, + "grad_norm": 0.20609347522258759, + "learning_rate": 4.991834489653718e-06, + "loss": 0.734, + "step": 299 + }, + { + "epoch": 0.11487650775416428, + "grad_norm": 0.2112123966217041, + "learning_rate": 4.991752263295261e-06, + "loss": 0.7063, + "step": 300 + }, + { + "epoch": 0.11525942944667815, + "grad_norm": 0.2099299132823944, + "learning_rate": 4.991669625684276e-06, + "loss": 0.7216, + "step": 301 + }, + { + "epoch": 0.11564235113919204, + "grad_norm": 0.20389284193515778, + "learning_rate": 4.9915865768344005e-06, + "loss": 0.7186, + "step": 302 + }, + { + "epoch": 0.11602527283170591, + "grad_norm": 0.2019105851650238, + "learning_rate": 4.991503116759345e-06, + "loss": 0.703, + "step": 303 + }, + { + "epoch": 0.1164081945242198, + "grad_norm": 0.20238110423088074, + "learning_rate": 4.99141924547288e-06, + "loss": 0.7024, + "step": 304 + }, + { + "epoch": 0.11679111621673367, + "grad_norm": 0.2012929469347, + "learning_rate": 4.991334962988852e-06, + "loss": 0.7171, + "step": 305 + }, + { + "epoch": 0.11717403790924756, + "grad_norm": 0.21306589245796204, + "learning_rate": 4.991250269321169e-06, + "loss": 0.714, + "step": 306 + }, + { + "epoch": 0.11755695960176143, + "grad_norm": 0.2003168761730194, + "learning_rate": 4.99116516448381e-06, + "loss": 0.7159, + "step": 307 + }, + { + "epoch": 0.11793988129427532, + "grad_norm": 0.20672893524169922, + "learning_rate": 4.991079648490821e-06, + "loss": 0.7351, + "step": 308 + }, + { + "epoch": 0.1183228029867892, + "grad_norm": 0.2076590359210968, + "learning_rate": 4.990993721356317e-06, + "loss": 0.7316, + "step": 309 + }, + { + "epoch": 0.11870572467930308, + "grad_norm": 0.2050575613975525, + "learning_rate": 4.990907383094478e-06, + "loss": 0.7133, + "step": 310 + }, + { + "epoch": 0.11908864637181696, + "grad_norm": 0.20803672075271606, + "learning_rate": 4.990820633719556e-06, + "loss": 0.7173, + "step": 311 + }, + { + "epoch": 0.11947156806433085, + "grad_norm": 0.20378495752811432, + "learning_rate": 4.990733473245866e-06, + "loss": 0.721, + "step": 312 + }, + { + "epoch": 0.11985448975684472, + "grad_norm": 0.20986348390579224, + "learning_rate": 4.9906459016877955e-06, + "loss": 0.7366, + "step": 313 + }, + { + "epoch": 0.12023741144935861, + "grad_norm": 0.202503502368927, + "learning_rate": 4.990557919059798e-06, + "loss": 0.7066, + "step": 314 + }, + { + "epoch": 0.12062033314187248, + "grad_norm": 0.21007058024406433, + "learning_rate": 4.990469525376392e-06, + "loss": 0.7503, + "step": 315 + }, + { + "epoch": 0.12100325483438637, + "grad_norm": 0.21536998450756073, + "learning_rate": 4.99038072065217e-06, + "loss": 0.7204, + "step": 316 + }, + { + "epoch": 0.12138617652690024, + "grad_norm": 0.20570923388004303, + "learning_rate": 4.990291504901785e-06, + "loss": 0.7302, + "step": 317 + }, + { + "epoch": 0.12176909821941413, + "grad_norm": 0.20915095508098602, + "learning_rate": 4.990201878139966e-06, + "loss": 0.7294, + "step": 318 + }, + { + "epoch": 0.122152019911928, + "grad_norm": 0.205043226480484, + "learning_rate": 4.9901118403815016e-06, + "loss": 0.7169, + "step": 319 + }, + { + "epoch": 0.1225349416044419, + "grad_norm": 0.20671123266220093, + "learning_rate": 4.990021391641255e-06, + "loss": 0.7065, + "step": 320 + }, + { + "epoch": 0.12291786329695577, + "grad_norm": 0.21355882287025452, + "learning_rate": 4.989930531934152e-06, + "loss": 0.7336, + "step": 321 + }, + { + "epoch": 0.12330078498946966, + "grad_norm": 0.21913346648216248, + "learning_rate": 4.989839261275191e-06, + "loss": 0.7168, + "step": 322 + }, + { + "epoch": 0.12368370668198353, + "grad_norm": 0.20357254147529602, + "learning_rate": 4.989747579679434e-06, + "loss": 0.6956, + "step": 323 + }, + { + "epoch": 0.12406662837449742, + "grad_norm": 0.20099161565303802, + "learning_rate": 4.989655487162013e-06, + "loss": 0.7097, + "step": 324 + }, + { + "epoch": 0.12444955006701129, + "grad_norm": 0.20629535615444183, + "learning_rate": 4.989562983738128e-06, + "loss": 0.7182, + "step": 325 + }, + { + "epoch": 0.12483247175952518, + "grad_norm": 0.20825421810150146, + "learning_rate": 4.989470069423045e-06, + "loss": 0.7101, + "step": 326 + }, + { + "epoch": 0.12521539345203905, + "grad_norm": 0.2155250906944275, + "learning_rate": 4.989376744232101e-06, + "loss": 0.714, + "step": 327 + }, + { + "epoch": 0.12559831514455294, + "grad_norm": 0.2070736289024353, + "learning_rate": 4.989283008180697e-06, + "loss": 0.7107, + "step": 328 + }, + { + "epoch": 0.12598123683706683, + "grad_norm": 0.20970770716667175, + "learning_rate": 4.989188861284305e-06, + "loss": 0.7054, + "step": 329 + }, + { + "epoch": 0.1263641585295807, + "grad_norm": 0.2124643325805664, + "learning_rate": 4.989094303558463e-06, + "loss": 0.7188, + "step": 330 + }, + { + "epoch": 0.12674708022209458, + "grad_norm": 0.2089206874370575, + "learning_rate": 4.988999335018777e-06, + "loss": 0.704, + "step": 331 + }, + { + "epoch": 0.12713000191460846, + "grad_norm": 0.2029915302991867, + "learning_rate": 4.988903955680923e-06, + "loss": 0.7125, + "step": 332 + }, + { + "epoch": 0.12751292360712235, + "grad_norm": 0.20648616552352905, + "learning_rate": 4.98880816556064e-06, + "loss": 0.7043, + "step": 333 + }, + { + "epoch": 0.1278958452996362, + "grad_norm": 0.20721794664859772, + "learning_rate": 4.98871196467374e-06, + "loss": 0.714, + "step": 334 + }, + { + "epoch": 0.1282787669921501, + "grad_norm": 0.2074366956949234, + "learning_rate": 4.988615353036101e-06, + "loss": 0.7005, + "step": 335 + }, + { + "epoch": 0.128661688684664, + "grad_norm": 0.2147573083639145, + "learning_rate": 4.988518330663665e-06, + "loss": 0.7176, + "step": 336 + }, + { + "epoch": 0.12904461037717788, + "grad_norm": 0.22254632413387299, + "learning_rate": 4.988420897572448e-06, + "loss": 0.7276, + "step": 337 + }, + { + "epoch": 0.12942753206969174, + "grad_norm": 0.21455398201942444, + "learning_rate": 4.98832305377853e-06, + "loss": 0.6962, + "step": 338 + }, + { + "epoch": 0.12981045376220562, + "grad_norm": 0.214791938662529, + "learning_rate": 4.98822479929806e-06, + "loss": 0.7168, + "step": 339 + }, + { + "epoch": 0.1301933754547195, + "grad_norm": 0.2137991189956665, + "learning_rate": 4.988126134147254e-06, + "loss": 0.7236, + "step": 340 + }, + { + "epoch": 0.1305762971472334, + "grad_norm": 0.21773158013820648, + "learning_rate": 4.988027058342396e-06, + "loss": 0.7225, + "step": 341 + }, + { + "epoch": 0.13095921883974726, + "grad_norm": 0.21584519743919373, + "learning_rate": 4.98792757189984e-06, + "loss": 0.7368, + "step": 342 + }, + { + "epoch": 0.13134214053226115, + "grad_norm": 0.21187016367912292, + "learning_rate": 4.987827674836003e-06, + "loss": 0.7163, + "step": 343 + }, + { + "epoch": 0.13172506222477504, + "grad_norm": 0.21678388118743896, + "learning_rate": 4.987727367167374e-06, + "loss": 0.721, + "step": 344 + }, + { + "epoch": 0.13210798391728892, + "grad_norm": 0.2196904569864273, + "learning_rate": 4.987626648910508e-06, + "loss": 0.7357, + "step": 345 + }, + { + "epoch": 0.13249090560980278, + "grad_norm": 0.22697031497955322, + "learning_rate": 4.9875255200820296e-06, + "loss": 0.7419, + "step": 346 + }, + { + "epoch": 0.13287382730231667, + "grad_norm": 0.21749354898929596, + "learning_rate": 4.987423980698627e-06, + "loss": 0.724, + "step": 347 + }, + { + "epoch": 0.13325674899483056, + "grad_norm": 0.20972779393196106, + "learning_rate": 4.9873220307770606e-06, + "loss": 0.7208, + "step": 348 + }, + { + "epoch": 0.13363967068734445, + "grad_norm": 0.2108878791332245, + "learning_rate": 4.987219670334156e-06, + "loss": 0.705, + "step": 349 + }, + { + "epoch": 0.1340225923798583, + "grad_norm": 0.21607382595539093, + "learning_rate": 4.987116899386808e-06, + "loss": 0.7033, + "step": 350 + }, + { + "epoch": 0.1344055140723722, + "grad_norm": 0.22202174365520477, + "learning_rate": 4.987013717951978e-06, + "loss": 0.7083, + "step": 351 + }, + { + "epoch": 0.13478843576488608, + "grad_norm": 0.21639248728752136, + "learning_rate": 4.986910126046695e-06, + "loss": 0.7111, + "step": 352 + }, + { + "epoch": 0.13517135745739997, + "grad_norm": 0.2141568511724472, + "learning_rate": 4.986806123688059e-06, + "loss": 0.7111, + "step": 353 + }, + { + "epoch": 0.13555427914991383, + "grad_norm": 0.21743625402450562, + "learning_rate": 4.986701710893232e-06, + "loss": 0.7262, + "step": 354 + }, + { + "epoch": 0.13593720084242772, + "grad_norm": 0.21381889283657074, + "learning_rate": 4.9865968876794476e-06, + "loss": 0.7068, + "step": 355 + }, + { + "epoch": 0.1363201225349416, + "grad_norm": 0.2151729315519333, + "learning_rate": 4.986491654064009e-06, + "loss": 0.7174, + "step": 356 + }, + { + "epoch": 0.1367030442274555, + "grad_norm": 0.21748758852481842, + "learning_rate": 4.986386010064281e-06, + "loss": 0.7201, + "step": 357 + }, + { + "epoch": 0.13708596591996935, + "grad_norm": 0.21864524483680725, + "learning_rate": 4.986279955697701e-06, + "loss": 0.7119, + "step": 358 + }, + { + "epoch": 0.13746888761248324, + "grad_norm": 0.21429306268692017, + "learning_rate": 4.986173490981773e-06, + "loss": 0.7173, + "step": 359 + }, + { + "epoch": 0.13785180930499713, + "grad_norm": 0.21339617669582367, + "learning_rate": 4.986066615934069e-06, + "loss": 0.7323, + "step": 360 + }, + { + "epoch": 0.13823473099751102, + "grad_norm": 0.21178525686264038, + "learning_rate": 4.985959330572227e-06, + "loss": 0.7143, + "step": 361 + }, + { + "epoch": 0.13861765269002488, + "grad_norm": 0.21623355150222778, + "learning_rate": 4.985851634913955e-06, + "loss": 0.6892, + "step": 362 + }, + { + "epoch": 0.13900057438253877, + "grad_norm": 0.21810880303382874, + "learning_rate": 4.985743528977029e-06, + "loss": 0.7135, + "step": 363 + }, + { + "epoch": 0.13938349607505265, + "grad_norm": 0.2134767770767212, + "learning_rate": 4.985635012779288e-06, + "loss": 0.7034, + "step": 364 + }, + { + "epoch": 0.13976641776756654, + "grad_norm": 0.21651697158813477, + "learning_rate": 4.985526086338644e-06, + "loss": 0.7017, + "step": 365 + }, + { + "epoch": 0.1401493394600804, + "grad_norm": 0.21995492279529572, + "learning_rate": 4.985416749673075e-06, + "loss": 0.7278, + "step": 366 + }, + { + "epoch": 0.1405322611525943, + "grad_norm": 0.2165203094482422, + "learning_rate": 4.985307002800626e-06, + "loss": 0.7089, + "step": 367 + }, + { + "epoch": 0.14091518284510818, + "grad_norm": 0.21220526099205017, + "learning_rate": 4.985196845739409e-06, + "loss": 0.7092, + "step": 368 + }, + { + "epoch": 0.14129810453762207, + "grad_norm": 0.22091609239578247, + "learning_rate": 4.9850862785076085e-06, + "loss": 0.7389, + "step": 369 + }, + { + "epoch": 0.14168102623013593, + "grad_norm": 0.21460163593292236, + "learning_rate": 4.98497530112347e-06, + "loss": 0.6937, + "step": 370 + }, + { + "epoch": 0.1420639479226498, + "grad_norm": 0.21762219071388245, + "learning_rate": 4.984863913605312e-06, + "loss": 0.7161, + "step": 371 + }, + { + "epoch": 0.1424468696151637, + "grad_norm": 0.21648332476615906, + "learning_rate": 4.9847521159715165e-06, + "loss": 0.7175, + "step": 372 + }, + { + "epoch": 0.1428297913076776, + "grad_norm": 0.21821698546409607, + "learning_rate": 4.984639908240536e-06, + "loss": 0.7006, + "step": 373 + }, + { + "epoch": 0.14321271300019145, + "grad_norm": 0.22117845714092255, + "learning_rate": 4.984527290430889e-06, + "loss": 0.7221, + "step": 374 + }, + { + "epoch": 0.14359563469270534, + "grad_norm": 0.22102953493595123, + "learning_rate": 4.984414262561166e-06, + "loss": 0.7028, + "step": 375 + }, + { + "epoch": 0.14397855638521923, + "grad_norm": 0.21901941299438477, + "learning_rate": 4.984300824650017e-06, + "loss": 0.7126, + "step": 376 + }, + { + "epoch": 0.1443614780777331, + "grad_norm": 0.22398801147937775, + "learning_rate": 4.984186976716167e-06, + "loss": 0.7104, + "step": 377 + }, + { + "epoch": 0.14474439977024697, + "grad_norm": 0.21393615007400513, + "learning_rate": 4.984072718778406e-06, + "loss": 0.7071, + "step": 378 + }, + { + "epoch": 0.14512732146276086, + "grad_norm": 0.219665989279747, + "learning_rate": 4.9839580508555915e-06, + "loss": 0.7232, + "step": 379 + }, + { + "epoch": 0.14551024315527475, + "grad_norm": 0.21342062950134277, + "learning_rate": 4.983842972966649e-06, + "loss": 0.6934, + "step": 380 + }, + { + "epoch": 0.14589316484778864, + "grad_norm": 0.21543823182582855, + "learning_rate": 4.983727485130572e-06, + "loss": 0.7114, + "step": 381 + }, + { + "epoch": 0.1462760865403025, + "grad_norm": 0.2211451679468155, + "learning_rate": 4.9836115873664195e-06, + "loss": 0.6972, + "step": 382 + }, + { + "epoch": 0.14665900823281638, + "grad_norm": 0.21137243509292603, + "learning_rate": 4.983495279693322e-06, + "loss": 0.7089, + "step": 383 + }, + { + "epoch": 0.14704192992533027, + "grad_norm": 0.22880102694034576, + "learning_rate": 4.983378562130474e-06, + "loss": 0.7372, + "step": 384 + }, + { + "epoch": 0.14742485161784416, + "grad_norm": 0.2202749252319336, + "learning_rate": 4.983261434697141e-06, + "loss": 0.6957, + "step": 385 + }, + { + "epoch": 0.14780777331035802, + "grad_norm": 0.22220788896083832, + "learning_rate": 4.9831438974126525e-06, + "loss": 0.7208, + "step": 386 + }, + { + "epoch": 0.1481906950028719, + "grad_norm": 0.2227896898984909, + "learning_rate": 4.983025950296408e-06, + "loss": 0.7037, + "step": 387 + }, + { + "epoch": 0.1485736166953858, + "grad_norm": 0.2159932553768158, + "learning_rate": 4.982907593367876e-06, + "loss": 0.7111, + "step": 388 + }, + { + "epoch": 0.14895653838789968, + "grad_norm": 0.220118910074234, + "learning_rate": 4.9827888266465876e-06, + "loss": 0.7127, + "step": 389 + }, + { + "epoch": 0.14933946008041354, + "grad_norm": 0.22082901000976562, + "learning_rate": 4.982669650152148e-06, + "loss": 0.6958, + "step": 390 + }, + { + "epoch": 0.14972238177292743, + "grad_norm": 0.22335520386695862, + "learning_rate": 4.982550063904223e-06, + "loss": 0.7102, + "step": 391 + }, + { + "epoch": 0.15010530346544132, + "grad_norm": 0.22340475022792816, + "learning_rate": 4.982430067922553e-06, + "loss": 0.7178, + "step": 392 + }, + { + "epoch": 0.1504882251579552, + "grad_norm": 0.22858190536499023, + "learning_rate": 4.9823096622269415e-06, + "loss": 0.7196, + "step": 393 + }, + { + "epoch": 0.15087114685046907, + "grad_norm": 0.21896734833717346, + "learning_rate": 4.9821888468372615e-06, + "loss": 0.6883, + "step": 394 + }, + { + "epoch": 0.15125406854298296, + "grad_norm": 0.23210509121418, + "learning_rate": 4.982067621773452e-06, + "loss": 0.7139, + "step": 395 + }, + { + "epoch": 0.15163699023549684, + "grad_norm": 0.22944042086601257, + "learning_rate": 4.981945987055521e-06, + "loss": 0.7194, + "step": 396 + }, + { + "epoch": 0.15201991192801073, + "grad_norm": 0.2184358388185501, + "learning_rate": 4.981823942703544e-06, + "loss": 0.7202, + "step": 397 + }, + { + "epoch": 0.1524028336205246, + "grad_norm": 0.22763623297214508, + "learning_rate": 4.9817014887376645e-06, + "loss": 0.7241, + "step": 398 + }, + { + "epoch": 0.15278575531303848, + "grad_norm": 0.22321753203868866, + "learning_rate": 4.981578625178093e-06, + "loss": 0.7001, + "step": 399 + }, + { + "epoch": 0.15316867700555237, + "grad_norm": 0.23028865456581116, + "learning_rate": 4.981455352045106e-06, + "loss": 0.7091, + "step": 400 + }, + { + "epoch": 0.15355159869806626, + "grad_norm": 0.22734788060188293, + "learning_rate": 4.981331669359049e-06, + "loss": 0.7291, + "step": 401 + }, + { + "epoch": 0.15393452039058012, + "grad_norm": 0.2209152728319168, + "learning_rate": 4.981207577140338e-06, + "loss": 0.722, + "step": 402 + }, + { + "epoch": 0.154317442083094, + "grad_norm": 0.216696634888649, + "learning_rate": 4.981083075409452e-06, + "loss": 0.7181, + "step": 403 + }, + { + "epoch": 0.1547003637756079, + "grad_norm": 0.2153381109237671, + "learning_rate": 4.98095816418694e-06, + "loss": 0.7228, + "step": 404 + }, + { + "epoch": 0.15508328546812178, + "grad_norm": 0.2197411209344864, + "learning_rate": 4.980832843493417e-06, + "loss": 0.7071, + "step": 405 + }, + { + "epoch": 0.15546620716063564, + "grad_norm": 0.21317681670188904, + "learning_rate": 4.980707113349567e-06, + "loss": 0.6979, + "step": 406 + }, + { + "epoch": 0.15584912885314953, + "grad_norm": 0.21619658172130585, + "learning_rate": 4.980580973776143e-06, + "loss": 0.7002, + "step": 407 + }, + { + "epoch": 0.15623205054566341, + "grad_norm": 0.21794861555099487, + "learning_rate": 4.980454424793961e-06, + "loss": 0.716, + "step": 408 + }, + { + "epoch": 0.1566149722381773, + "grad_norm": 0.2209479808807373, + "learning_rate": 4.980327466423909e-06, + "loss": 0.7106, + "step": 409 + }, + { + "epoch": 0.15699789393069116, + "grad_norm": 0.2195109724998474, + "learning_rate": 4.9802000986869415e-06, + "loss": 0.7041, + "step": 410 + }, + { + "epoch": 0.15738081562320505, + "grad_norm": 0.21919915080070496, + "learning_rate": 4.980072321604078e-06, + "loss": 0.7141, + "step": 411 + }, + { + "epoch": 0.15776373731571894, + "grad_norm": 0.21942178905010223, + "learning_rate": 4.979944135196409e-06, + "loss": 0.7016, + "step": 412 + }, + { + "epoch": 0.15814665900823283, + "grad_norm": 0.22026552259922028, + "learning_rate": 4.97981553948509e-06, + "loss": 0.7176, + "step": 413 + }, + { + "epoch": 0.1585295807007467, + "grad_norm": 0.22086086869239807, + "learning_rate": 4.979686534491347e-06, + "loss": 0.7022, + "step": 414 + }, + { + "epoch": 0.15891250239326057, + "grad_norm": 0.22359570860862732, + "learning_rate": 4.9795571202364705e-06, + "loss": 0.7176, + "step": 415 + }, + { + "epoch": 0.15929542408577446, + "grad_norm": 0.22810761630535126, + "learning_rate": 4.979427296741818e-06, + "loss": 0.7182, + "step": 416 + }, + { + "epoch": 0.15967834577828835, + "grad_norm": 0.2195035070180893, + "learning_rate": 4.97929706402882e-06, + "loss": 0.7047, + "step": 417 + }, + { + "epoch": 0.1600612674708022, + "grad_norm": 0.21832919120788574, + "learning_rate": 4.9791664221189675e-06, + "loss": 0.7105, + "step": 418 + }, + { + "epoch": 0.1604441891633161, + "grad_norm": 0.2244974821805954, + "learning_rate": 4.979035371033824e-06, + "loss": 0.7345, + "step": 419 + }, + { + "epoch": 0.16082711085582999, + "grad_norm": 0.2344287633895874, + "learning_rate": 4.97890391079502e-06, + "loss": 0.7368, + "step": 420 + }, + { + "epoch": 0.16121003254834387, + "grad_norm": 0.2236819863319397, + "learning_rate": 4.978772041424249e-06, + "loss": 0.7133, + "step": 421 + }, + { + "epoch": 0.16159295424085773, + "grad_norm": 0.22271811962127686, + "learning_rate": 4.978639762943279e-06, + "loss": 0.7214, + "step": 422 + }, + { + "epoch": 0.16197587593337162, + "grad_norm": 0.2247050553560257, + "learning_rate": 4.97850707537394e-06, + "loss": 0.6983, + "step": 423 + }, + { + "epoch": 0.1623587976258855, + "grad_norm": 0.22323517501354218, + "learning_rate": 4.978373978738132e-06, + "loss": 0.7194, + "step": 424 + }, + { + "epoch": 0.1627417193183994, + "grad_norm": 0.2268957793712616, + "learning_rate": 4.9782404730578225e-06, + "loss": 0.719, + "step": 425 + }, + { + "epoch": 0.16312464101091326, + "grad_norm": 0.22057485580444336, + "learning_rate": 4.978106558355044e-06, + "loss": 0.6976, + "step": 426 + }, + { + "epoch": 0.16350756270342715, + "grad_norm": 0.2155829668045044, + "learning_rate": 4.9779722346519014e-06, + "loss": 0.7051, + "step": 427 + }, + { + "epoch": 0.16389048439594103, + "grad_norm": 0.21671557426452637, + "learning_rate": 4.977837501970562e-06, + "loss": 0.69, + "step": 428 + }, + { + "epoch": 0.16427340608845492, + "grad_norm": 0.22859551012516022, + "learning_rate": 4.977702360333264e-06, + "loss": 0.6963, + "step": 429 + }, + { + "epoch": 0.16465632778096878, + "grad_norm": 0.2260478436946869, + "learning_rate": 4.977566809762312e-06, + "loss": 0.6866, + "step": 430 + }, + { + "epoch": 0.16503924947348267, + "grad_norm": 0.22191403806209564, + "learning_rate": 4.977430850280076e-06, + "loss": 0.7066, + "step": 431 + }, + { + "epoch": 0.16542217116599656, + "grad_norm": 0.22449170053005219, + "learning_rate": 4.977294481908999e-06, + "loss": 0.7283, + "step": 432 + }, + { + "epoch": 0.16580509285851044, + "grad_norm": 0.2276216596364975, + "learning_rate": 4.977157704671585e-06, + "loss": 0.7271, + "step": 433 + }, + { + "epoch": 0.1661880145510243, + "grad_norm": 0.23174940049648285, + "learning_rate": 4.97702051859041e-06, + "loss": 0.7095, + "step": 434 + }, + { + "epoch": 0.1665709362435382, + "grad_norm": 0.23013226687908173, + "learning_rate": 4.976882923688114e-06, + "loss": 0.7124, + "step": 435 + }, + { + "epoch": 0.16695385793605208, + "grad_norm": 0.23638951778411865, + "learning_rate": 4.976744919987409e-06, + "loss": 0.7311, + "step": 436 + }, + { + "epoch": 0.16733677962856597, + "grad_norm": 0.2298298478126526, + "learning_rate": 4.976606507511072e-06, + "loss": 0.6905, + "step": 437 + }, + { + "epoch": 0.16771970132107983, + "grad_norm": 0.23514661192893982, + "learning_rate": 4.976467686281945e-06, + "loss": 0.7382, + "step": 438 + }, + { + "epoch": 0.16810262301359372, + "grad_norm": 0.2360498607158661, + "learning_rate": 4.976328456322941e-06, + "loss": 0.708, + "step": 439 + }, + { + "epoch": 0.1684855447061076, + "grad_norm": 0.23474329710006714, + "learning_rate": 4.976188817657039e-06, + "loss": 0.717, + "step": 440 + }, + { + "epoch": 0.1688684663986215, + "grad_norm": 0.21756932139396667, + "learning_rate": 4.976048770307286e-06, + "loss": 0.6842, + "step": 441 + }, + { + "epoch": 0.16925138809113535, + "grad_norm": 0.22361315786838531, + "learning_rate": 4.9759083142967965e-06, + "loss": 0.7273, + "step": 442 + }, + { + "epoch": 0.16963430978364924, + "grad_norm": 0.22651907801628113, + "learning_rate": 4.975767449648751e-06, + "loss": 0.7157, + "step": 443 + }, + { + "epoch": 0.17001723147616313, + "grad_norm": 0.22782133519649506, + "learning_rate": 4.9756261763864e-06, + "loss": 0.7122, + "step": 444 + }, + { + "epoch": 0.17040015316867702, + "grad_norm": 0.2352352738380432, + "learning_rate": 4.97548449453306e-06, + "loss": 0.7224, + "step": 445 + }, + { + "epoch": 0.17078307486119088, + "grad_norm": 0.22195972502231598, + "learning_rate": 4.975342404112114e-06, + "loss": 0.7092, + "step": 446 + }, + { + "epoch": 0.17116599655370476, + "grad_norm": 0.21784667670726776, + "learning_rate": 4.975199905147014e-06, + "loss": 0.6812, + "step": 447 + }, + { + "epoch": 0.17154891824621865, + "grad_norm": 0.21954038739204407, + "learning_rate": 4.9750569976612776e-06, + "loss": 0.6989, + "step": 448 + }, + { + "epoch": 0.17193183993873254, + "grad_norm": 0.2203446477651596, + "learning_rate": 4.9749136816784924e-06, + "loss": 0.7066, + "step": 449 + }, + { + "epoch": 0.1723147616312464, + "grad_norm": 0.22727733850479126, + "learning_rate": 4.974769957222312e-06, + "loss": 0.6922, + "step": 450 + }, + { + "epoch": 0.1726976833237603, + "grad_norm": 0.22717690467834473, + "learning_rate": 4.974625824316456e-06, + "loss": 0.6975, + "step": 451 + }, + { + "epoch": 0.17308060501627417, + "grad_norm": 0.2288667857646942, + "learning_rate": 4.974481282984716e-06, + "loss": 0.7036, + "step": 452 + }, + { + "epoch": 0.17346352670878806, + "grad_norm": 0.22407589852809906, + "learning_rate": 4.974336333250944e-06, + "loss": 0.6969, + "step": 453 + }, + { + "epoch": 0.17384644840130192, + "grad_norm": 0.2228279709815979, + "learning_rate": 4.9741909751390664e-06, + "loss": 0.7024, + "step": 454 + }, + { + "epoch": 0.1742293700938158, + "grad_norm": 0.2224264144897461, + "learning_rate": 4.9740452086730725e-06, + "loss": 0.6967, + "step": 455 + }, + { + "epoch": 0.1746122917863297, + "grad_norm": 0.22775305807590485, + "learning_rate": 4.97389903387702e-06, + "loss": 0.6995, + "step": 456 + }, + { + "epoch": 0.1749952134788436, + "grad_norm": 0.22632873058319092, + "learning_rate": 4.973752450775036e-06, + "loss": 0.6902, + "step": 457 + }, + { + "epoch": 0.17537813517135745, + "grad_norm": 0.22588998079299927, + "learning_rate": 4.973605459391312e-06, + "loss": 0.7133, + "step": 458 + }, + { + "epoch": 0.17576105686387133, + "grad_norm": 0.22500504553318024, + "learning_rate": 4.973458059750108e-06, + "loss": 0.712, + "step": 459 + }, + { + "epoch": 0.17614397855638522, + "grad_norm": 0.22925838828086853, + "learning_rate": 4.973310251875753e-06, + "loss": 0.696, + "step": 460 + }, + { + "epoch": 0.1765269002488991, + "grad_norm": 0.22367170453071594, + "learning_rate": 4.973162035792641e-06, + "loss": 0.681, + "step": 461 + }, + { + "epoch": 0.17690982194141297, + "grad_norm": 0.22455467283725739, + "learning_rate": 4.973013411525235e-06, + "loss": 0.7047, + "step": 462 + }, + { + "epoch": 0.17729274363392686, + "grad_norm": 0.23515203595161438, + "learning_rate": 4.972864379098064e-06, + "loss": 0.7032, + "step": 463 + }, + { + "epoch": 0.17767566532644075, + "grad_norm": 0.230363667011261, + "learning_rate": 4.972714938535727e-06, + "loss": 0.698, + "step": 464 + }, + { + "epoch": 0.17805858701895463, + "grad_norm": 0.2247180938720703, + "learning_rate": 4.972565089862886e-06, + "loss": 0.7066, + "step": 465 + }, + { + "epoch": 0.1784415087114685, + "grad_norm": 0.23525093495845795, + "learning_rate": 4.972414833104274e-06, + "loss": 0.7219, + "step": 466 + }, + { + "epoch": 0.17882443040398238, + "grad_norm": 0.22985999286174774, + "learning_rate": 4.9722641682846914e-06, + "loss": 0.7186, + "step": 467 + }, + { + "epoch": 0.17920735209649627, + "grad_norm": 0.22825634479522705, + "learning_rate": 4.972113095429003e-06, + "loss": 0.7191, + "step": 468 + }, + { + "epoch": 0.17959027378901016, + "grad_norm": 0.22180423140525818, + "learning_rate": 4.971961614562143e-06, + "loss": 0.6829, + "step": 469 + }, + { + "epoch": 0.17997319548152402, + "grad_norm": 0.22781048715114594, + "learning_rate": 4.971809725709112e-06, + "loss": 0.6898, + "step": 470 + }, + { + "epoch": 0.1803561171740379, + "grad_norm": 0.23069560527801514, + "learning_rate": 4.9716574288949804e-06, + "loss": 0.7253, + "step": 471 + }, + { + "epoch": 0.1807390388665518, + "grad_norm": 0.22806665301322937, + "learning_rate": 4.971504724144884e-06, + "loss": 0.6897, + "step": 472 + }, + { + "epoch": 0.18112196055906568, + "grad_norm": 0.23605108261108398, + "learning_rate": 4.971351611484023e-06, + "loss": 0.7297, + "step": 473 + }, + { + "epoch": 0.18150488225157954, + "grad_norm": 0.23285695910453796, + "learning_rate": 4.971198090937671e-06, + "loss": 0.7113, + "step": 474 + }, + { + "epoch": 0.18188780394409343, + "grad_norm": 0.23309025168418884, + "learning_rate": 4.971044162531165e-06, + "loss": 0.7037, + "step": 475 + }, + { + "epoch": 0.18227072563660732, + "grad_norm": 0.23858731985092163, + "learning_rate": 4.970889826289912e-06, + "loss": 0.7117, + "step": 476 + }, + { + "epoch": 0.1826536473291212, + "grad_norm": 0.23126953840255737, + "learning_rate": 4.9707350822393805e-06, + "loss": 0.7114, + "step": 477 + }, + { + "epoch": 0.18303656902163506, + "grad_norm": 0.23670174181461334, + "learning_rate": 4.970579930405114e-06, + "loss": 0.7077, + "step": 478 + }, + { + "epoch": 0.18341949071414895, + "grad_norm": 0.2424033135175705, + "learning_rate": 4.970424370812718e-06, + "loss": 0.6897, + "step": 479 + }, + { + "epoch": 0.18380241240666284, + "grad_norm": 0.2186141163110733, + "learning_rate": 4.970268403487866e-06, + "loss": 0.7031, + "step": 480 + }, + { + "epoch": 0.18418533409917673, + "grad_norm": 0.22593295574188232, + "learning_rate": 4.970112028456301e-06, + "loss": 0.6798, + "step": 481 + }, + { + "epoch": 0.1845682557916906, + "grad_norm": 0.23915334045886993, + "learning_rate": 4.969955245743832e-06, + "loss": 0.6964, + "step": 482 + }, + { + "epoch": 0.18495117748420448, + "grad_norm": 0.23232129216194153, + "learning_rate": 4.9697980553763356e-06, + "loss": 0.7196, + "step": 483 + }, + { + "epoch": 0.18533409917671836, + "grad_norm": 0.22518081963062286, + "learning_rate": 4.969640457379754e-06, + "loss": 0.6879, + "step": 484 + }, + { + "epoch": 0.18571702086923225, + "grad_norm": 0.22891272604465485, + "learning_rate": 4.969482451780099e-06, + "loss": 0.6841, + "step": 485 + }, + { + "epoch": 0.1860999425617461, + "grad_norm": 0.23179848492145538, + "learning_rate": 4.969324038603449e-06, + "loss": 0.7019, + "step": 486 + }, + { + "epoch": 0.18648286425426, + "grad_norm": 0.23586530983448029, + "learning_rate": 4.969165217875948e-06, + "loss": 0.6975, + "step": 487 + }, + { + "epoch": 0.1868657859467739, + "grad_norm": 0.23328320682048798, + "learning_rate": 4.969005989623812e-06, + "loss": 0.7116, + "step": 488 + }, + { + "epoch": 0.18724870763928778, + "grad_norm": 0.22832241654396057, + "learning_rate": 4.968846353873316e-06, + "loss": 0.7058, + "step": 489 + }, + { + "epoch": 0.18763162933180164, + "grad_norm": 0.23435759544372559, + "learning_rate": 4.968686310650811e-06, + "loss": 0.7225, + "step": 490 + }, + { + "epoch": 0.18801455102431552, + "grad_norm": 0.23468558490276337, + "learning_rate": 4.968525859982709e-06, + "loss": 0.6828, + "step": 491 + }, + { + "epoch": 0.1883974727168294, + "grad_norm": 0.2350325882434845, + "learning_rate": 4.9683650018954935e-06, + "loss": 0.6831, + "step": 492 + }, + { + "epoch": 0.1887803944093433, + "grad_norm": 0.22212567925453186, + "learning_rate": 4.968203736415713e-06, + "loss": 0.6908, + "step": 493 + }, + { + "epoch": 0.18916331610185716, + "grad_norm": 0.2288699448108673, + "learning_rate": 4.968042063569982e-06, + "loss": 0.6914, + "step": 494 + }, + { + "epoch": 0.18954623779437105, + "grad_norm": 0.24049511551856995, + "learning_rate": 4.967879983384987e-06, + "loss": 0.7103, + "step": 495 + }, + { + "epoch": 0.18992915948688494, + "grad_norm": 0.23875749111175537, + "learning_rate": 4.967717495887476e-06, + "loss": 0.6892, + "step": 496 + }, + { + "epoch": 0.19031208117939882, + "grad_norm": 0.230393648147583, + "learning_rate": 4.967554601104268e-06, + "loss": 0.6968, + "step": 497 + }, + { + "epoch": 0.19069500287191268, + "grad_norm": 0.22483550012111664, + "learning_rate": 4.967391299062247e-06, + "loss": 0.683, + "step": 498 + }, + { + "epoch": 0.19107792456442657, + "grad_norm": 0.2363535761833191, + "learning_rate": 4.967227589788365e-06, + "loss": 0.7108, + "step": 499 + }, + { + "epoch": 0.19146084625694046, + "grad_norm": 0.23204447329044342, + "learning_rate": 4.967063473309644e-06, + "loss": 0.6833, + "step": 500 + }, + { + "epoch": 0.19184376794945435, + "grad_norm": 0.23585835099220276, + "learning_rate": 4.9668989496531684e-06, + "loss": 0.6713, + "step": 501 + }, + { + "epoch": 0.1922266896419682, + "grad_norm": 0.24668626487255096, + "learning_rate": 4.966734018846093e-06, + "loss": 0.6959, + "step": 502 + }, + { + "epoch": 0.1926096113344821, + "grad_norm": 0.24632133543491364, + "learning_rate": 4.966568680915639e-06, + "loss": 0.7179, + "step": 503 + }, + { + "epoch": 0.19299253302699598, + "grad_norm": 0.2375093400478363, + "learning_rate": 4.966402935889094e-06, + "loss": 0.6956, + "step": 504 + }, + { + "epoch": 0.19337545471950987, + "grad_norm": 0.2281908541917801, + "learning_rate": 4.966236783793814e-06, + "loss": 0.7169, + "step": 505 + }, + { + "epoch": 0.19375837641202373, + "grad_norm": 0.23612891137599945, + "learning_rate": 4.9660702246572215e-06, + "loss": 0.6921, + "step": 506 + }, + { + "epoch": 0.19414129810453762, + "grad_norm": 0.2388295829296112, + "learning_rate": 4.965903258506806e-06, + "loss": 0.7058, + "step": 507 + }, + { + "epoch": 0.1945242197970515, + "grad_norm": 0.23923897743225098, + "learning_rate": 4.965735885370126e-06, + "loss": 0.6878, + "step": 508 + }, + { + "epoch": 0.1949071414895654, + "grad_norm": 0.24728970229625702, + "learning_rate": 4.965568105274803e-06, + "loss": 0.7118, + "step": 509 + }, + { + "epoch": 0.19529006318207925, + "grad_norm": 0.24075351655483246, + "learning_rate": 4.96539991824853e-06, + "loss": 0.7101, + "step": 510 + }, + { + "epoch": 0.19567298487459314, + "grad_norm": 0.23857763409614563, + "learning_rate": 4.965231324319067e-06, + "loss": 0.7122, + "step": 511 + }, + { + "epoch": 0.19605590656710703, + "grad_norm": 0.24270926415920258, + "learning_rate": 4.965062323514237e-06, + "loss": 0.6893, + "step": 512 + }, + { + "epoch": 0.19643882825962092, + "grad_norm": 0.23443444073200226, + "learning_rate": 4.964892915861934e-06, + "loss": 0.6913, + "step": 513 + }, + { + "epoch": 0.19682174995213478, + "grad_norm": 0.23783984780311584, + "learning_rate": 4.964723101390118e-06, + "loss": 0.7059, + "step": 514 + }, + { + "epoch": 0.19720467164464867, + "grad_norm": 0.23205767571926117, + "learning_rate": 4.964552880126816e-06, + "loss": 0.6983, + "step": 515 + }, + { + "epoch": 0.19758759333716255, + "grad_norm": 0.22749286890029907, + "learning_rate": 4.964382252100122e-06, + "loss": 0.7025, + "step": 516 + }, + { + "epoch": 0.19797051502967644, + "grad_norm": 0.23700077831745148, + "learning_rate": 4.9642112173381976e-06, + "loss": 0.6903, + "step": 517 + }, + { + "epoch": 0.1983534367221903, + "grad_norm": 0.2351311892271042, + "learning_rate": 4.964039775869271e-06, + "loss": 0.6945, + "step": 518 + }, + { + "epoch": 0.1987363584147042, + "grad_norm": 0.24296842515468597, + "learning_rate": 4.963867927721639e-06, + "loss": 0.7023, + "step": 519 + }, + { + "epoch": 0.19911928010721808, + "grad_norm": 0.2292814999818802, + "learning_rate": 4.963695672923663e-06, + "loss": 0.7018, + "step": 520 + }, + { + "epoch": 0.19950220179973197, + "grad_norm": 0.23721252381801605, + "learning_rate": 4.963523011503775e-06, + "loss": 0.6805, + "step": 521 + }, + { + "epoch": 0.19988512349224583, + "grad_norm": 0.24078230559825897, + "learning_rate": 4.9633499434904685e-06, + "loss": 0.6988, + "step": 522 + }, + { + "epoch": 0.2002680451847597, + "grad_norm": 0.2441691756248474, + "learning_rate": 4.96317646891231e-06, + "loss": 0.6765, + "step": 523 + }, + { + "epoch": 0.2006509668772736, + "grad_norm": 0.23363593220710754, + "learning_rate": 4.96300258779793e-06, + "loss": 0.6955, + "step": 524 + }, + { + "epoch": 0.2010338885697875, + "grad_norm": 0.24009943008422852, + "learning_rate": 4.962828300176028e-06, + "loss": 0.7068, + "step": 525 + }, + { + "epoch": 0.20141681026230135, + "grad_norm": 0.236362487077713, + "learning_rate": 4.9626536060753686e-06, + "loss": 0.6873, + "step": 526 + }, + { + "epoch": 0.20179973195481524, + "grad_norm": 0.2265264391899109, + "learning_rate": 4.962478505524783e-06, + "loss": 0.6931, + "step": 527 + }, + { + "epoch": 0.20218265364732912, + "grad_norm": 0.23028504848480225, + "learning_rate": 4.962302998553173e-06, + "loss": 0.6955, + "step": 528 + }, + { + "epoch": 0.202565575339843, + "grad_norm": 0.2343558669090271, + "learning_rate": 4.9621270851895035e-06, + "loss": 0.7112, + "step": 529 + }, + { + "epoch": 0.20294849703235687, + "grad_norm": 0.2429768443107605, + "learning_rate": 4.961950765462809e-06, + "loss": 0.7068, + "step": 530 + }, + { + "epoch": 0.20333141872487076, + "grad_norm": 0.23821046948432922, + "learning_rate": 4.961774039402189e-06, + "loss": 0.7064, + "step": 531 + }, + { + "epoch": 0.20371434041738465, + "grad_norm": 0.23160751163959503, + "learning_rate": 4.961596907036813e-06, + "loss": 0.7006, + "step": 532 + }, + { + "epoch": 0.20409726210989854, + "grad_norm": 0.2379685938358307, + "learning_rate": 4.961419368395916e-06, + "loss": 0.6992, + "step": 533 + }, + { + "epoch": 0.2044801838024124, + "grad_norm": 0.23077930510044098, + "learning_rate": 4.961241423508799e-06, + "loss": 0.6935, + "step": 534 + }, + { + "epoch": 0.20486310549492628, + "grad_norm": 0.23924565315246582, + "learning_rate": 4.9610630724048315e-06, + "loss": 0.6875, + "step": 535 + }, + { + "epoch": 0.20524602718744017, + "grad_norm": 0.24769453704357147, + "learning_rate": 4.960884315113449e-06, + "loss": 0.6924, + "step": 536 + }, + { + "epoch": 0.20562894887995406, + "grad_norm": 0.23660869896411896, + "learning_rate": 4.960705151664155e-06, + "loss": 0.6824, + "step": 537 + }, + { + "epoch": 0.20601187057246792, + "grad_norm": 0.24358759820461273, + "learning_rate": 4.960525582086519e-06, + "loss": 0.7247, + "step": 538 + }, + { + "epoch": 0.2063947922649818, + "grad_norm": 0.23302821815013885, + "learning_rate": 4.960345606410179e-06, + "loss": 0.7004, + "step": 539 + }, + { + "epoch": 0.2067777139574957, + "grad_norm": 0.2414388209581375, + "learning_rate": 4.96016522466484e-06, + "loss": 0.708, + "step": 540 + }, + { + "epoch": 0.20716063565000958, + "grad_norm": 0.23602741956710815, + "learning_rate": 4.959984436880271e-06, + "loss": 0.6991, + "step": 541 + }, + { + "epoch": 0.20754355734252344, + "grad_norm": 0.23501873016357422, + "learning_rate": 4.959803243086312e-06, + "loss": 0.6956, + "step": 542 + }, + { + "epoch": 0.20792647903503733, + "grad_norm": 0.2384674996137619, + "learning_rate": 4.959621643312867e-06, + "loss": 0.691, + "step": 543 + }, + { + "epoch": 0.20830940072755122, + "grad_norm": 0.24334029853343964, + "learning_rate": 4.959439637589909e-06, + "loss": 0.7004, + "step": 544 + }, + { + "epoch": 0.2086923224200651, + "grad_norm": 0.2337769865989685, + "learning_rate": 4.959257225947477e-06, + "loss": 0.7001, + "step": 545 + }, + { + "epoch": 0.20907524411257897, + "grad_norm": 0.24959228932857513, + "learning_rate": 4.959074408415676e-06, + "loss": 0.6854, + "step": 546 + }, + { + "epoch": 0.20945816580509286, + "grad_norm": 0.23717932403087616, + "learning_rate": 4.958891185024682e-06, + "loss": 0.6883, + "step": 547 + }, + { + "epoch": 0.20984108749760674, + "grad_norm": 0.22805754840373993, + "learning_rate": 4.958707555804734e-06, + "loss": 0.6875, + "step": 548 + }, + { + "epoch": 0.21022400919012063, + "grad_norm": 0.23835137486457825, + "learning_rate": 4.958523520786137e-06, + "loss": 0.6849, + "step": 549 + }, + { + "epoch": 0.2106069308826345, + "grad_norm": 0.2451052963733673, + "learning_rate": 4.958339079999268e-06, + "loss": 0.6947, + "step": 550 + }, + { + "epoch": 0.21098985257514838, + "grad_norm": 0.23878198862075806, + "learning_rate": 4.958154233474567e-06, + "loss": 0.6897, + "step": 551 + }, + { + "epoch": 0.21137277426766227, + "grad_norm": 0.243367537856102, + "learning_rate": 4.957968981242543e-06, + "loss": 0.7118, + "step": 552 + }, + { + "epoch": 0.21175569596017615, + "grad_norm": 0.2516103684902191, + "learning_rate": 4.957783323333768e-06, + "loss": 0.7118, + "step": 553 + }, + { + "epoch": 0.21213861765269001, + "grad_norm": 0.24193982779979706, + "learning_rate": 4.9575972597788875e-06, + "loss": 0.6816, + "step": 554 + }, + { + "epoch": 0.2125215393452039, + "grad_norm": 0.2456333190202713, + "learning_rate": 4.95741079060861e-06, + "loss": 0.7218, + "step": 555 + }, + { + "epoch": 0.2129044610377178, + "grad_norm": 0.23211970925331116, + "learning_rate": 4.9572239158537095e-06, + "loss": 0.6877, + "step": 556 + }, + { + "epoch": 0.21328738273023168, + "grad_norm": 0.2383391559123993, + "learning_rate": 4.95703663554503e-06, + "loss": 0.6862, + "step": 557 + }, + { + "epoch": 0.21367030442274554, + "grad_norm": 0.2499326914548874, + "learning_rate": 4.956848949713481e-06, + "loss": 0.7077, + "step": 558 + }, + { + "epoch": 0.21405322611525943, + "grad_norm": 0.24644067883491516, + "learning_rate": 4.95666085839004e-06, + "loss": 0.7048, + "step": 559 + }, + { + "epoch": 0.21443614780777331, + "grad_norm": 0.24420374631881714, + "learning_rate": 4.95647236160575e-06, + "loss": 0.6893, + "step": 560 + }, + { + "epoch": 0.2148190695002872, + "grad_norm": 0.23663920164108276, + "learning_rate": 4.956283459391721e-06, + "loss": 0.6912, + "step": 561 + }, + { + "epoch": 0.21520199119280106, + "grad_norm": 0.2413814216852188, + "learning_rate": 4.956094151779132e-06, + "loss": 0.6999, + "step": 562 + }, + { + "epoch": 0.21558491288531495, + "grad_norm": 0.25189071893692017, + "learning_rate": 4.955904438799225e-06, + "loss": 0.7002, + "step": 563 + }, + { + "epoch": 0.21596783457782884, + "grad_norm": 0.24884158372879028, + "learning_rate": 4.955714320483314e-06, + "loss": 0.698, + "step": 564 + }, + { + "epoch": 0.21635075627034273, + "grad_norm": 0.23854660987854004, + "learning_rate": 4.955523796862775e-06, + "loss": 0.6886, + "step": 565 + }, + { + "epoch": 0.21673367796285659, + "grad_norm": 0.24080811440944672, + "learning_rate": 4.9553328679690545e-06, + "loss": 0.704, + "step": 566 + }, + { + "epoch": 0.21711659965537047, + "grad_norm": 0.24664145708084106, + "learning_rate": 4.955141533833663e-06, + "loss": 0.6978, + "step": 567 + }, + { + "epoch": 0.21749952134788436, + "grad_norm": 0.240121990442276, + "learning_rate": 4.954949794488181e-06, + "loss": 0.7007, + "step": 568 + }, + { + "epoch": 0.21788244304039825, + "grad_norm": 0.2344406694173813, + "learning_rate": 4.954757649964253e-06, + "loss": 0.7014, + "step": 569 + }, + { + "epoch": 0.2182653647329121, + "grad_norm": 0.23647718131542206, + "learning_rate": 4.954565100293593e-06, + "loss": 0.6952, + "step": 570 + }, + { + "epoch": 0.218648286425426, + "grad_norm": 0.25824204087257385, + "learning_rate": 4.954372145507979e-06, + "loss": 0.7084, + "step": 571 + }, + { + "epoch": 0.21903120811793989, + "grad_norm": 0.23864921927452087, + "learning_rate": 4.954178785639257e-06, + "loss": 0.7012, + "step": 572 + }, + { + "epoch": 0.21941412981045377, + "grad_norm": 0.23659467697143555, + "learning_rate": 4.953985020719342e-06, + "loss": 0.7097, + "step": 573 + }, + { + "epoch": 0.21979705150296763, + "grad_norm": 0.23632408678531647, + "learning_rate": 4.953790850780213e-06, + "loss": 0.7185, + "step": 574 + }, + { + "epoch": 0.22017997319548152, + "grad_norm": 0.23713268339633942, + "learning_rate": 4.9535962758539155e-06, + "loss": 0.702, + "step": 575 + }, + { + "epoch": 0.2205628948879954, + "grad_norm": 0.23817777633666992, + "learning_rate": 4.953401295972565e-06, + "loss": 0.6957, + "step": 576 + }, + { + "epoch": 0.2209458165805093, + "grad_norm": 0.24862009286880493, + "learning_rate": 4.953205911168343e-06, + "loss": 0.6936, + "step": 577 + }, + { + "epoch": 0.22132873827302316, + "grad_norm": 0.24382250010967255, + "learning_rate": 4.953010121473495e-06, + "loss": 0.7066, + "step": 578 + }, + { + "epoch": 0.22171165996553704, + "grad_norm": 0.23864789307117462, + "learning_rate": 4.952813926920335e-06, + "loss": 0.6881, + "step": 579 + }, + { + "epoch": 0.22209458165805093, + "grad_norm": 0.24444547295570374, + "learning_rate": 4.952617327541246e-06, + "loss": 0.7082, + "step": 580 + }, + { + "epoch": 0.22247750335056482, + "grad_norm": 0.24924702942371368, + "learning_rate": 4.952420323368673e-06, + "loss": 0.732, + "step": 581 + }, + { + "epoch": 0.22286042504307868, + "grad_norm": 0.24768982827663422, + "learning_rate": 4.952222914435134e-06, + "loss": 0.7057, + "step": 582 + }, + { + "epoch": 0.22324334673559257, + "grad_norm": 0.23804764449596405, + "learning_rate": 4.952025100773208e-06, + "loss": 0.7024, + "step": 583 + }, + { + "epoch": 0.22362626842810646, + "grad_norm": 0.23817269504070282, + "learning_rate": 4.951826882415544e-06, + "loss": 0.7052, + "step": 584 + }, + { + "epoch": 0.22400919012062034, + "grad_norm": 0.2470194399356842, + "learning_rate": 4.9516282593948566e-06, + "loss": 0.6916, + "step": 585 + }, + { + "epoch": 0.2243921118131342, + "grad_norm": 0.252437949180603, + "learning_rate": 4.951429231743929e-06, + "loss": 0.6999, + "step": 586 + }, + { + "epoch": 0.2247750335056481, + "grad_norm": 0.23925770819187164, + "learning_rate": 4.951229799495609e-06, + "loss": 0.6975, + "step": 587 + }, + { + "epoch": 0.22515795519816198, + "grad_norm": 0.23760266602039337, + "learning_rate": 4.951029962682811e-06, + "loss": 0.6877, + "step": 588 + }, + { + "epoch": 0.22554087689067587, + "grad_norm": 0.25361642241477966, + "learning_rate": 4.950829721338518e-06, + "loss": 0.6848, + "step": 589 + }, + { + "epoch": 0.22592379858318973, + "grad_norm": 0.25293734669685364, + "learning_rate": 4.95062907549578e-06, + "loss": 0.6876, + "step": 590 + }, + { + "epoch": 0.22630672027570362, + "grad_norm": 0.24178287386894226, + "learning_rate": 4.9504280251877105e-06, + "loss": 0.6681, + "step": 591 + }, + { + "epoch": 0.2266896419682175, + "grad_norm": 0.2571343779563904, + "learning_rate": 4.9502265704474935e-06, + "loss": 0.6889, + "step": 592 + }, + { + "epoch": 0.2270725636607314, + "grad_norm": 0.244108647108078, + "learning_rate": 4.950024711308379e-06, + "loss": 0.6934, + "step": 593 + }, + { + "epoch": 0.22745548535324525, + "grad_norm": 0.2604058086872101, + "learning_rate": 4.949822447803681e-06, + "loss": 0.7133, + "step": 594 + }, + { + "epoch": 0.22783840704575914, + "grad_norm": 0.24538995325565338, + "learning_rate": 4.949619779966783e-06, + "loss": 0.6937, + "step": 595 + }, + { + "epoch": 0.22822132873827303, + "grad_norm": 0.2400553971529007, + "learning_rate": 4.949416707831134e-06, + "loss": 0.6809, + "step": 596 + }, + { + "epoch": 0.22860425043078691, + "grad_norm": 0.2515188157558441, + "learning_rate": 4.9492132314302515e-06, + "loss": 0.6999, + "step": 597 + }, + { + "epoch": 0.22898717212330078, + "grad_norm": 0.23625919222831726, + "learning_rate": 4.949009350797717e-06, + "loss": 0.6963, + "step": 598 + }, + { + "epoch": 0.22937009381581466, + "grad_norm": 0.24061603844165802, + "learning_rate": 4.94880506596718e-06, + "loss": 0.6925, + "step": 599 + }, + { + "epoch": 0.22975301550832855, + "grad_norm": 0.24639979004859924, + "learning_rate": 4.9486003769723574e-06, + "loss": 0.6729, + "step": 600 + }, + { + "epoch": 0.23013593720084244, + "grad_norm": 0.2444503903388977, + "learning_rate": 4.948395283847033e-06, + "loss": 0.7179, + "step": 601 + }, + { + "epoch": 0.2305188588933563, + "grad_norm": 0.2438957691192627, + "learning_rate": 4.948189786625055e-06, + "loss": 0.6953, + "step": 602 + }, + { + "epoch": 0.2309017805858702, + "grad_norm": 0.24474462866783142, + "learning_rate": 4.947983885340341e-06, + "loss": 0.6761, + "step": 603 + }, + { + "epoch": 0.23128470227838407, + "grad_norm": 0.23572909832000732, + "learning_rate": 4.947777580026872e-06, + "loss": 0.7164, + "step": 604 + }, + { + "epoch": 0.23166762397089796, + "grad_norm": 0.24651245772838593, + "learning_rate": 4.947570870718701e-06, + "loss": 0.6898, + "step": 605 + }, + { + "epoch": 0.23205054566341182, + "grad_norm": 0.24207252264022827, + "learning_rate": 4.947363757449941e-06, + "loss": 0.6962, + "step": 606 + }, + { + "epoch": 0.2324334673559257, + "grad_norm": 0.2437017560005188, + "learning_rate": 4.947156240254778e-06, + "loss": 0.6916, + "step": 607 + }, + { + "epoch": 0.2328163890484396, + "grad_norm": 0.23833702504634857, + "learning_rate": 4.946948319167461e-06, + "loss": 0.6734, + "step": 608 + }, + { + "epoch": 0.23319931074095349, + "grad_norm": 0.2502235770225525, + "learning_rate": 4.946739994222305e-06, + "loss": 0.6793, + "step": 609 + }, + { + "epoch": 0.23358223243346735, + "grad_norm": 0.24891985952854156, + "learning_rate": 4.946531265453694e-06, + "loss": 0.6727, + "step": 610 + }, + { + "epoch": 0.23396515412598123, + "grad_norm": 0.24262681603431702, + "learning_rate": 4.9463221328960785e-06, + "loss": 0.6662, + "step": 611 + }, + { + "epoch": 0.23434807581849512, + "grad_norm": 0.2385813146829605, + "learning_rate": 4.946112596583974e-06, + "loss": 0.6895, + "step": 612 + }, + { + "epoch": 0.234730997511009, + "grad_norm": 0.24113699793815613, + "learning_rate": 4.945902656551964e-06, + "loss": 0.7058, + "step": 613 + }, + { + "epoch": 0.23511391920352287, + "grad_norm": 0.24542860686779022, + "learning_rate": 4.945692312834698e-06, + "loss": 0.6796, + "step": 614 + }, + { + "epoch": 0.23549684089603676, + "grad_norm": 0.23808449506759644, + "learning_rate": 4.945481565466891e-06, + "loss": 0.7038, + "step": 615 + }, + { + "epoch": 0.23587976258855065, + "grad_norm": 0.23834019899368286, + "learning_rate": 4.945270414483328e-06, + "loss": 0.676, + "step": 616 + }, + { + "epoch": 0.23626268428106453, + "grad_norm": 0.23723448812961578, + "learning_rate": 4.945058859918858e-06, + "loss": 0.6925, + "step": 617 + }, + { + "epoch": 0.2366456059735784, + "grad_norm": 0.24539196491241455, + "learning_rate": 4.944846901808397e-06, + "loss": 0.6928, + "step": 618 + }, + { + "epoch": 0.23702852766609228, + "grad_norm": 0.24122965335845947, + "learning_rate": 4.944634540186927e-06, + "loss": 0.6637, + "step": 619 + }, + { + "epoch": 0.23741144935860617, + "grad_norm": 0.2460906058549881, + "learning_rate": 4.944421775089497e-06, + "loss": 0.7034, + "step": 620 + }, + { + "epoch": 0.23779437105112006, + "grad_norm": 0.2400168925523758, + "learning_rate": 4.9442086065512254e-06, + "loss": 0.688, + "step": 621 + }, + { + "epoch": 0.23817729274363392, + "grad_norm": 0.2557614743709564, + "learning_rate": 4.943995034607292e-06, + "loss": 0.7188, + "step": 622 + }, + { + "epoch": 0.2385602144361478, + "grad_norm": 0.2443343997001648, + "learning_rate": 4.943781059292948e-06, + "loss": 0.6951, + "step": 623 + }, + { + "epoch": 0.2389431361286617, + "grad_norm": 0.2577515244483948, + "learning_rate": 4.943566680643508e-06, + "loss": 0.7086, + "step": 624 + }, + { + "epoch": 0.23932605782117558, + "grad_norm": 0.2645318806171417, + "learning_rate": 4.943351898694353e-06, + "loss": 0.6923, + "step": 625 + }, + { + "epoch": 0.23970897951368944, + "grad_norm": 0.24452121555805206, + "learning_rate": 4.943136713480934e-06, + "loss": 0.6806, + "step": 626 + }, + { + "epoch": 0.24009190120620333, + "grad_norm": 0.2464938759803772, + "learning_rate": 4.942921125038765e-06, + "loss": 0.7121, + "step": 627 + }, + { + "epoch": 0.24047482289871722, + "grad_norm": 0.24740079045295715, + "learning_rate": 4.942705133403429e-06, + "loss": 0.6819, + "step": 628 + }, + { + "epoch": 0.2408577445912311, + "grad_norm": 0.24974393844604492, + "learning_rate": 4.942488738610574e-06, + "loss": 0.6604, + "step": 629 + }, + { + "epoch": 0.24124066628374496, + "grad_norm": 0.24305953085422516, + "learning_rate": 4.942271940695913e-06, + "loss": 0.6738, + "step": 630 + }, + { + "epoch": 0.24162358797625885, + "grad_norm": 0.2633149027824402, + "learning_rate": 4.942054739695231e-06, + "loss": 0.6646, + "step": 631 + }, + { + "epoch": 0.24200650966877274, + "grad_norm": 0.24853935837745667, + "learning_rate": 4.941837135644374e-06, + "loss": 0.687, + "step": 632 + }, + { + "epoch": 0.24238943136128663, + "grad_norm": 0.24881082773208618, + "learning_rate": 4.941619128579257e-06, + "loss": 0.6794, + "step": 633 + }, + { + "epoch": 0.2427723530538005, + "grad_norm": 0.26114174723625183, + "learning_rate": 4.94140071853586e-06, + "loss": 0.708, + "step": 634 + }, + { + "epoch": 0.24315527474631438, + "grad_norm": 0.2673856317996979, + "learning_rate": 4.941181905550233e-06, + "loss": 0.6966, + "step": 635 + }, + { + "epoch": 0.24353819643882826, + "grad_norm": 0.25460389256477356, + "learning_rate": 4.940962689658488e-06, + "loss": 0.7009, + "step": 636 + }, + { + "epoch": 0.24392111813134215, + "grad_norm": 0.25717636942863464, + "learning_rate": 4.940743070896806e-06, + "loss": 0.7, + "step": 637 + }, + { + "epoch": 0.244304039823856, + "grad_norm": 0.25204774737358093, + "learning_rate": 4.940523049301436e-06, + "loss": 0.6667, + "step": 638 + }, + { + "epoch": 0.2446869615163699, + "grad_norm": 0.24111990630626678, + "learning_rate": 4.940302624908689e-06, + "loss": 0.6826, + "step": 639 + }, + { + "epoch": 0.2450698832088838, + "grad_norm": 0.2586113512516022, + "learning_rate": 4.9400817977549455e-06, + "loss": 0.691, + "step": 640 + }, + { + "epoch": 0.24545280490139768, + "grad_norm": 0.24710243940353394, + "learning_rate": 4.939860567876652e-06, + "loss": 0.7184, + "step": 641 + }, + { + "epoch": 0.24583572659391154, + "grad_norm": 0.2547800838947296, + "learning_rate": 4.939638935310324e-06, + "loss": 0.6763, + "step": 642 + }, + { + "epoch": 0.24621864828642542, + "grad_norm": 0.25336694717407227, + "learning_rate": 4.939416900092537e-06, + "loss": 0.6925, + "step": 643 + }, + { + "epoch": 0.2466015699789393, + "grad_norm": 0.24745817482471466, + "learning_rate": 4.9391944622599405e-06, + "loss": 0.6996, + "step": 644 + }, + { + "epoch": 0.2469844916714532, + "grad_norm": 0.2559024691581726, + "learning_rate": 4.9389716218492445e-06, + "loss": 0.6987, + "step": 645 + }, + { + "epoch": 0.24736741336396706, + "grad_norm": 0.2510485053062439, + "learning_rate": 4.938748378897229e-06, + "loss": 0.6981, + "step": 646 + }, + { + "epoch": 0.24775033505648095, + "grad_norm": 0.24602986872196198, + "learning_rate": 4.938524733440738e-06, + "loss": 0.6833, + "step": 647 + }, + { + "epoch": 0.24813325674899483, + "grad_norm": 0.2407660037279129, + "learning_rate": 4.938300685516686e-06, + "loss": 0.6933, + "step": 648 + }, + { + "epoch": 0.24851617844150872, + "grad_norm": 0.2540413439273834, + "learning_rate": 4.938076235162047e-06, + "loss": 0.6903, + "step": 649 + }, + { + "epoch": 0.24889910013402258, + "grad_norm": 0.26849615573883057, + "learning_rate": 4.937851382413869e-06, + "loss": 0.6988, + "step": 650 + }, + { + "epoch": 0.24928202182653647, + "grad_norm": 0.248337984085083, + "learning_rate": 4.9376261273092614e-06, + "loss": 0.6965, + "step": 651 + }, + { + "epoch": 0.24966494351905036, + "grad_norm": 0.24852517247200012, + "learning_rate": 4.937400469885401e-06, + "loss": 0.6975, + "step": 652 + }, + { + "epoch": 0.25004786521156425, + "grad_norm": 0.24815134704113007, + "learning_rate": 4.937174410179534e-06, + "loss": 0.7148, + "step": 653 + }, + { + "epoch": 0.2504307869040781, + "grad_norm": 0.242036372423172, + "learning_rate": 4.936947948228967e-06, + "loss": 0.6911, + "step": 654 + }, + { + "epoch": 0.250813708596592, + "grad_norm": 0.24787138402462006, + "learning_rate": 4.936721084071079e-06, + "loss": 0.708, + "step": 655 + }, + { + "epoch": 0.2511966302891059, + "grad_norm": 0.2459169179201126, + "learning_rate": 4.936493817743313e-06, + "loss": 0.6897, + "step": 656 + }, + { + "epoch": 0.25157955198161974, + "grad_norm": 0.2538677453994751, + "learning_rate": 4.936266149283176e-06, + "loss": 0.7044, + "step": 657 + }, + { + "epoch": 0.25196247367413366, + "grad_norm": 0.2510700821876526, + "learning_rate": 4.936038078728247e-06, + "loss": 0.6953, + "step": 658 + }, + { + "epoch": 0.2523453953666475, + "grad_norm": 0.24351082742214203, + "learning_rate": 4.935809606116165e-06, + "loss": 0.6808, + "step": 659 + }, + { + "epoch": 0.2527283170591614, + "grad_norm": 0.25081127882003784, + "learning_rate": 4.93558073148464e-06, + "loss": 0.702, + "step": 660 + }, + { + "epoch": 0.2531112387516753, + "grad_norm": 0.2425825446844101, + "learning_rate": 4.935351454871447e-06, + "loss": 0.7019, + "step": 661 + }, + { + "epoch": 0.25349416044418915, + "grad_norm": 0.24126264452934265, + "learning_rate": 4.935121776314425e-06, + "loss": 0.721, + "step": 662 + }, + { + "epoch": 0.25387708213670307, + "grad_norm": 0.2510846257209778, + "learning_rate": 4.934891695851483e-06, + "loss": 0.6832, + "step": 663 + }, + { + "epoch": 0.25426000382921693, + "grad_norm": 0.2648461163043976, + "learning_rate": 4.934661213520596e-06, + "loss": 0.693, + "step": 664 + }, + { + "epoch": 0.2546429255217308, + "grad_norm": 0.24153569340705872, + "learning_rate": 4.934430329359803e-06, + "loss": 0.6797, + "step": 665 + }, + { + "epoch": 0.2550258472142447, + "grad_norm": 0.25208204984664917, + "learning_rate": 4.934199043407209e-06, + "loss": 0.6902, + "step": 666 + }, + { + "epoch": 0.25540876890675857, + "grad_norm": 0.2638026177883148, + "learning_rate": 4.9339673557009875e-06, + "loss": 0.686, + "step": 667 + }, + { + "epoch": 0.2557916905992724, + "grad_norm": 0.2548025846481323, + "learning_rate": 4.933735266279377e-06, + "loss": 0.6804, + "step": 668 + }, + { + "epoch": 0.25617461229178634, + "grad_norm": 0.2605096101760864, + "learning_rate": 4.9335027751806855e-06, + "loss": 0.7069, + "step": 669 + }, + { + "epoch": 0.2565575339843002, + "grad_norm": 0.24284419417381287, + "learning_rate": 4.933269882443281e-06, + "loss": 0.6813, + "step": 670 + }, + { + "epoch": 0.2569404556768141, + "grad_norm": 0.2466583251953125, + "learning_rate": 4.933036588105604e-06, + "loss": 0.7137, + "step": 671 + }, + { + "epoch": 0.257323377369328, + "grad_norm": 0.2555830478668213, + "learning_rate": 4.932802892206158e-06, + "loss": 0.6957, + "step": 672 + }, + { + "epoch": 0.25770629906184184, + "grad_norm": 0.24138735234737396, + "learning_rate": 4.932568794783513e-06, + "loss": 0.6662, + "step": 673 + }, + { + "epoch": 0.25808922075435575, + "grad_norm": 0.2518264651298523, + "learning_rate": 4.932334295876306e-06, + "loss": 0.6965, + "step": 674 + }, + { + "epoch": 0.2584721424468696, + "grad_norm": 0.2566470801830292, + "learning_rate": 4.9320993955232396e-06, + "loss": 0.7018, + "step": 675 + }, + { + "epoch": 0.2588550641393835, + "grad_norm": 0.24778035283088684, + "learning_rate": 4.931864093763085e-06, + "loss": 0.6782, + "step": 676 + }, + { + "epoch": 0.2592379858318974, + "grad_norm": 0.24779430031776428, + "learning_rate": 4.931628390634674e-06, + "loss": 0.674, + "step": 677 + }, + { + "epoch": 0.25962090752441125, + "grad_norm": 0.2474948763847351, + "learning_rate": 4.931392286176911e-06, + "loss": 0.6987, + "step": 678 + }, + { + "epoch": 0.26000382921692516, + "grad_norm": 0.24874965846538544, + "learning_rate": 4.931155780428764e-06, + "loss": 0.7017, + "step": 679 + }, + { + "epoch": 0.260386750909439, + "grad_norm": 0.2501877248287201, + "learning_rate": 4.930918873429268e-06, + "loss": 0.6959, + "step": 680 + }, + { + "epoch": 0.2607696726019529, + "grad_norm": 0.24506975710391998, + "learning_rate": 4.93068156521752e-06, + "loss": 0.6941, + "step": 681 + }, + { + "epoch": 0.2611525942944668, + "grad_norm": 0.24200353026390076, + "learning_rate": 4.930443855832689e-06, + "loss": 0.6911, + "step": 682 + }, + { + "epoch": 0.26153551598698066, + "grad_norm": 0.24502404034137726, + "learning_rate": 4.9302057453140085e-06, + "loss": 0.6985, + "step": 683 + }, + { + "epoch": 0.2619184376794945, + "grad_norm": 0.24491210281848907, + "learning_rate": 4.929967233700776e-06, + "loss": 0.6831, + "step": 684 + }, + { + "epoch": 0.26230135937200844, + "grad_norm": 0.24369360506534576, + "learning_rate": 4.929728321032358e-06, + "loss": 0.6796, + "step": 685 + }, + { + "epoch": 0.2626842810645223, + "grad_norm": 0.24483689665794373, + "learning_rate": 4.929489007348185e-06, + "loss": 0.6918, + "step": 686 + }, + { + "epoch": 0.2630672027570362, + "grad_norm": 0.24488209187984467, + "learning_rate": 4.929249292687756e-06, + "loss": 0.6828, + "step": 687 + }, + { + "epoch": 0.26345012444955007, + "grad_norm": 0.2576925456523895, + "learning_rate": 4.929009177090632e-06, + "loss": 0.6968, + "step": 688 + }, + { + "epoch": 0.26383304614206393, + "grad_norm": 0.25123924016952515, + "learning_rate": 4.928768660596446e-06, + "loss": 0.6976, + "step": 689 + }, + { + "epoch": 0.26421596783457785, + "grad_norm": 0.2571817636489868, + "learning_rate": 4.928527743244893e-06, + "loss": 0.6797, + "step": 690 + }, + { + "epoch": 0.2645988895270917, + "grad_norm": 0.25800999999046326, + "learning_rate": 4.928286425075736e-06, + "loss": 0.6964, + "step": 691 + }, + { + "epoch": 0.26498181121960557, + "grad_norm": 0.2471861094236374, + "learning_rate": 4.928044706128803e-06, + "loss": 0.6927, + "step": 692 + }, + { + "epoch": 0.2653647329121195, + "grad_norm": 0.24648526310920715, + "learning_rate": 4.927802586443987e-06, + "loss": 0.6979, + "step": 693 + }, + { + "epoch": 0.26574765460463334, + "grad_norm": 0.25215795636177063, + "learning_rate": 4.927560066061251e-06, + "loss": 0.7, + "step": 694 + }, + { + "epoch": 0.26613057629714726, + "grad_norm": 0.2663182020187378, + "learning_rate": 4.927317145020622e-06, + "loss": 0.693, + "step": 695 + }, + { + "epoch": 0.2665134979896611, + "grad_norm": 0.25089675188064575, + "learning_rate": 4.9270738233621915e-06, + "loss": 0.6702, + "step": 696 + }, + { + "epoch": 0.266896419682175, + "grad_norm": 0.24895228445529938, + "learning_rate": 4.92683010112612e-06, + "loss": 0.6962, + "step": 697 + }, + { + "epoch": 0.2672793413746889, + "grad_norm": 0.2562544345855713, + "learning_rate": 4.926585978352632e-06, + "loss": 0.7017, + "step": 698 + }, + { + "epoch": 0.26766226306720275, + "grad_norm": 0.25019386410713196, + "learning_rate": 4.926341455082018e-06, + "loss": 0.6772, + "step": 699 + }, + { + "epoch": 0.2680451847597166, + "grad_norm": 0.2618144750595093, + "learning_rate": 4.926096531354637e-06, + "loss": 0.6831, + "step": 700 + }, + { + "epoch": 0.26842810645223053, + "grad_norm": 0.2517690062522888, + "learning_rate": 4.925851207210912e-06, + "loss": 0.675, + "step": 701 + }, + { + "epoch": 0.2688110281447444, + "grad_norm": 0.2592850923538208, + "learning_rate": 4.925605482691332e-06, + "loss": 0.6971, + "step": 702 + }, + { + "epoch": 0.2691939498372583, + "grad_norm": 0.26684650778770447, + "learning_rate": 4.9253593578364545e-06, + "loss": 0.701, + "step": 703 + }, + { + "epoch": 0.26957687152977217, + "grad_norm": 0.25434690713882446, + "learning_rate": 4.9251128326869e-06, + "loss": 0.6915, + "step": 704 + }, + { + "epoch": 0.269959793222286, + "grad_norm": 0.2535107731819153, + "learning_rate": 4.924865907283356e-06, + "loss": 0.6853, + "step": 705 + }, + { + "epoch": 0.27034271491479994, + "grad_norm": 0.25544658303260803, + "learning_rate": 4.924618581666578e-06, + "loss": 0.6827, + "step": 706 + }, + { + "epoch": 0.2707256366073138, + "grad_norm": 0.26514726877212524, + "learning_rate": 4.924370855877384e-06, + "loss": 0.7171, + "step": 707 + }, + { + "epoch": 0.27110855829982766, + "grad_norm": 0.25380080938339233, + "learning_rate": 4.924122729956662e-06, + "loss": 0.6958, + "step": 708 + }, + { + "epoch": 0.2714914799923416, + "grad_norm": 0.25360482931137085, + "learning_rate": 4.9238742039453626e-06, + "loss": 0.7015, + "step": 709 + }, + { + "epoch": 0.27187440168485544, + "grad_norm": 0.25626882910728455, + "learning_rate": 4.923625277884504e-06, + "loss": 0.6868, + "step": 710 + }, + { + "epoch": 0.27225732337736935, + "grad_norm": 0.26761749386787415, + "learning_rate": 4.923375951815171e-06, + "loss": 0.6919, + "step": 711 + }, + { + "epoch": 0.2726402450698832, + "grad_norm": 0.263866126537323, + "learning_rate": 4.923126225778513e-06, + "loss": 0.7029, + "step": 712 + }, + { + "epoch": 0.2730231667623971, + "grad_norm": 0.2864058315753937, + "learning_rate": 4.9228760998157475e-06, + "loss": 0.6957, + "step": 713 + }, + { + "epoch": 0.273406088454911, + "grad_norm": 0.2648506760597229, + "learning_rate": 4.922625573968155e-06, + "loss": 0.6834, + "step": 714 + }, + { + "epoch": 0.27378901014742485, + "grad_norm": 0.2676790952682495, + "learning_rate": 4.922374648277085e-06, + "loss": 0.7012, + "step": 715 + }, + { + "epoch": 0.2741719318399387, + "grad_norm": 0.25565361976623535, + "learning_rate": 4.922123322783951e-06, + "loss": 0.7191, + "step": 716 + }, + { + "epoch": 0.2745548535324526, + "grad_norm": 0.2526341676712036, + "learning_rate": 4.921871597530234e-06, + "loss": 0.6957, + "step": 717 + }, + { + "epoch": 0.2749377752249665, + "grad_norm": 0.2454516589641571, + "learning_rate": 4.921619472557478e-06, + "loss": 0.6863, + "step": 718 + }, + { + "epoch": 0.2753206969174804, + "grad_norm": 0.2534882426261902, + "learning_rate": 4.9213669479072964e-06, + "loss": 0.6813, + "step": 719 + }, + { + "epoch": 0.27570361860999426, + "grad_norm": 0.2647958993911743, + "learning_rate": 4.921114023621368e-06, + "loss": 0.6723, + "step": 720 + }, + { + "epoch": 0.2760865403025081, + "grad_norm": 0.2482769936323166, + "learning_rate": 4.920860699741436e-06, + "loss": 0.6816, + "step": 721 + }, + { + "epoch": 0.27646946199502204, + "grad_norm": 0.25155022740364075, + "learning_rate": 4.920606976309311e-06, + "loss": 0.6829, + "step": 722 + }, + { + "epoch": 0.2768523836875359, + "grad_norm": 0.25965702533721924, + "learning_rate": 4.920352853366869e-06, + "loss": 0.6881, + "step": 723 + }, + { + "epoch": 0.27723530538004976, + "grad_norm": 0.26698043942451477, + "learning_rate": 4.9200983309560505e-06, + "loss": 0.6947, + "step": 724 + }, + { + "epoch": 0.2776182270725637, + "grad_norm": 0.2601257562637329, + "learning_rate": 4.919843409118863e-06, + "loss": 0.6816, + "step": 725 + }, + { + "epoch": 0.27800114876507753, + "grad_norm": 0.24915724992752075, + "learning_rate": 4.919588087897383e-06, + "loss": 0.6667, + "step": 726 + }, + { + "epoch": 0.27838407045759145, + "grad_norm": 0.2661370635032654, + "learning_rate": 4.919332367333748e-06, + "loss": 0.7011, + "step": 727 + }, + { + "epoch": 0.2787669921501053, + "grad_norm": 0.24617598950862885, + "learning_rate": 4.919076247470163e-06, + "loss": 0.6789, + "step": 728 + }, + { + "epoch": 0.27914991384261917, + "grad_norm": 0.259954571723938, + "learning_rate": 4.918819728348901e-06, + "loss": 0.6858, + "step": 729 + }, + { + "epoch": 0.2795328355351331, + "grad_norm": 0.2742285430431366, + "learning_rate": 4.918562810012299e-06, + "loss": 0.6576, + "step": 730 + }, + { + "epoch": 0.27991575722764694, + "grad_norm": 0.2507971525192261, + "learning_rate": 4.91830549250276e-06, + "loss": 0.7046, + "step": 731 + }, + { + "epoch": 0.2802986789201608, + "grad_norm": 0.25771722197532654, + "learning_rate": 4.918047775862752e-06, + "loss": 0.6869, + "step": 732 + }, + { + "epoch": 0.2806816006126747, + "grad_norm": 0.2926652431488037, + "learning_rate": 4.917789660134812e-06, + "loss": 0.6852, + "step": 733 + }, + { + "epoch": 0.2810645223051886, + "grad_norm": 0.2552712559700012, + "learning_rate": 4.91753114536154e-06, + "loss": 0.6691, + "step": 734 + }, + { + "epoch": 0.2814474439977025, + "grad_norm": 0.2440652698278427, + "learning_rate": 4.917272231585601e-06, + "loss": 0.665, + "step": 735 + }, + { + "epoch": 0.28183036569021636, + "grad_norm": 0.286358505487442, + "learning_rate": 4.91701291884973e-06, + "loss": 0.693, + "step": 736 + }, + { + "epoch": 0.2822132873827302, + "grad_norm": 0.2688429653644562, + "learning_rate": 4.916753207196725e-06, + "loss": 0.6837, + "step": 737 + }, + { + "epoch": 0.28259620907524413, + "grad_norm": 0.25828817486763, + "learning_rate": 4.916493096669449e-06, + "loss": 0.6796, + "step": 738 + }, + { + "epoch": 0.282979130767758, + "grad_norm": 0.2633633017539978, + "learning_rate": 4.916232587310833e-06, + "loss": 0.7175, + "step": 739 + }, + { + "epoch": 0.28336205246027185, + "grad_norm": 0.26460349559783936, + "learning_rate": 4.915971679163873e-06, + "loss": 0.6979, + "step": 740 + }, + { + "epoch": 0.28374497415278577, + "grad_norm": 0.25029146671295166, + "learning_rate": 4.91571037227163e-06, + "loss": 0.6978, + "step": 741 + }, + { + "epoch": 0.2841278958452996, + "grad_norm": 0.2536710202693939, + "learning_rate": 4.915448666677233e-06, + "loss": 0.6759, + "step": 742 + }, + { + "epoch": 0.28451081753781354, + "grad_norm": 0.252708375453949, + "learning_rate": 4.915186562423872e-06, + "loss": 0.7197, + "step": 743 + }, + { + "epoch": 0.2848937392303274, + "grad_norm": 0.2493618130683899, + "learning_rate": 4.914924059554811e-06, + "loss": 0.6846, + "step": 744 + }, + { + "epoch": 0.28527666092284126, + "grad_norm": 0.25273728370666504, + "learning_rate": 4.914661158113372e-06, + "loss": 0.708, + "step": 745 + }, + { + "epoch": 0.2856595826153552, + "grad_norm": 0.25201109051704407, + "learning_rate": 4.914397858142945e-06, + "loss": 0.6955, + "step": 746 + }, + { + "epoch": 0.28604250430786904, + "grad_norm": 0.2539828419685364, + "learning_rate": 4.914134159686988e-06, + "loss": 0.6937, + "step": 747 + }, + { + "epoch": 0.2864254260003829, + "grad_norm": 0.24846617877483368, + "learning_rate": 4.913870062789023e-06, + "loss": 0.672, + "step": 748 + }, + { + "epoch": 0.2868083476928968, + "grad_norm": 0.25717365741729736, + "learning_rate": 4.913605567492636e-06, + "loss": 0.6838, + "step": 749 + }, + { + "epoch": 0.2871912693854107, + "grad_norm": 0.2502661347389221, + "learning_rate": 4.9133406738414845e-06, + "loss": 0.6803, + "step": 750 + }, + { + "epoch": 0.2875741910779246, + "grad_norm": 0.25706034898757935, + "learning_rate": 4.913075381879285e-06, + "loss": 0.6741, + "step": 751 + }, + { + "epoch": 0.28795711277043845, + "grad_norm": 0.26075178384780884, + "learning_rate": 4.9128096916498245e-06, + "loss": 0.6923, + "step": 752 + }, + { + "epoch": 0.2883400344629523, + "grad_norm": 0.2518405318260193, + "learning_rate": 4.912543603196954e-06, + "loss": 0.6902, + "step": 753 + }, + { + "epoch": 0.2887229561554662, + "grad_norm": 0.26775062084198, + "learning_rate": 4.912277116564588e-06, + "loss": 0.6924, + "step": 754 + }, + { + "epoch": 0.2891058778479801, + "grad_norm": 0.254680871963501, + "learning_rate": 4.912010231796711e-06, + "loss": 0.6906, + "step": 755 + }, + { + "epoch": 0.28948879954049395, + "grad_norm": 0.2532293200492859, + "learning_rate": 4.911742948937371e-06, + "loss": 0.6899, + "step": 756 + }, + { + "epoch": 0.28987172123300786, + "grad_norm": 0.2509908676147461, + "learning_rate": 4.911475268030681e-06, + "loss": 0.6785, + "step": 757 + }, + { + "epoch": 0.2902546429255217, + "grad_norm": 0.2599368989467621, + "learning_rate": 4.91120718912082e-06, + "loss": 0.7066, + "step": 758 + }, + { + "epoch": 0.29063756461803564, + "grad_norm": 0.24985715746879578, + "learning_rate": 4.910938712252036e-06, + "loss": 0.6531, + "step": 759 + }, + { + "epoch": 0.2910204863105495, + "grad_norm": 0.2594451904296875, + "learning_rate": 4.910669837468637e-06, + "loss": 0.6768, + "step": 760 + }, + { + "epoch": 0.29140340800306336, + "grad_norm": 0.24278171360492706, + "learning_rate": 4.910400564815e-06, + "loss": 0.6756, + "step": 761 + }, + { + "epoch": 0.2917863296955773, + "grad_norm": 0.26272347569465637, + "learning_rate": 4.9101308943355685e-06, + "loss": 0.682, + "step": 762 + }, + { + "epoch": 0.29216925138809113, + "grad_norm": 0.25561460852622986, + "learning_rate": 4.90986082607485e-06, + "loss": 0.7127, + "step": 763 + }, + { + "epoch": 0.292552173080605, + "grad_norm": 0.26740762591362, + "learning_rate": 4.909590360077417e-06, + "loss": 0.6828, + "step": 764 + }, + { + "epoch": 0.2929350947731189, + "grad_norm": 0.25563833117485046, + "learning_rate": 4.909319496387911e-06, + "loss": 0.699, + "step": 765 + }, + { + "epoch": 0.29331801646563277, + "grad_norm": 0.25876274704933167, + "learning_rate": 4.909048235051033e-06, + "loss": 0.6922, + "step": 766 + }, + { + "epoch": 0.2937009381581467, + "grad_norm": 0.2549036741256714, + "learning_rate": 4.908776576111558e-06, + "loss": 0.6704, + "step": 767 + }, + { + "epoch": 0.29408385985066055, + "grad_norm": 0.25628861784935, + "learning_rate": 4.908504519614319e-06, + "loss": 0.6954, + "step": 768 + }, + { + "epoch": 0.2944667815431744, + "grad_norm": 0.2500770688056946, + "learning_rate": 4.908232065604219e-06, + "loss": 0.6795, + "step": 769 + }, + { + "epoch": 0.2948497032356883, + "grad_norm": 0.2657643258571625, + "learning_rate": 4.907959214126225e-06, + "loss": 0.712, + "step": 770 + }, + { + "epoch": 0.2952326249282022, + "grad_norm": 0.26558786630630493, + "learning_rate": 4.907685965225369e-06, + "loss": 0.7003, + "step": 771 + }, + { + "epoch": 0.29561554662071604, + "grad_norm": 0.2577509582042694, + "learning_rate": 4.9074123189467506e-06, + "loss": 0.6976, + "step": 772 + }, + { + "epoch": 0.29599846831322996, + "grad_norm": 0.2677637040615082, + "learning_rate": 4.9071382753355335e-06, + "loss": 0.6844, + "step": 773 + }, + { + "epoch": 0.2963813900057438, + "grad_norm": 0.2563629448413849, + "learning_rate": 4.906863834436948e-06, + "loss": 0.6913, + "step": 774 + }, + { + "epoch": 0.29676431169825773, + "grad_norm": 0.26639705896377563, + "learning_rate": 4.906588996296289e-06, + "loss": 0.6813, + "step": 775 + }, + { + "epoch": 0.2971472333907716, + "grad_norm": 0.26096269488334656, + "learning_rate": 4.9063137609589166e-06, + "loss": 0.6939, + "step": 776 + }, + { + "epoch": 0.29753015508328545, + "grad_norm": 0.250601202249527, + "learning_rate": 4.906038128470258e-06, + "loss": 0.6461, + "step": 777 + }, + { + "epoch": 0.29791307677579937, + "grad_norm": 0.25932541489601135, + "learning_rate": 4.905762098875804e-06, + "loss": 0.6873, + "step": 778 + }, + { + "epoch": 0.29829599846831323, + "grad_norm": 0.256270170211792, + "learning_rate": 4.905485672221113e-06, + "loss": 0.6746, + "step": 779 + }, + { + "epoch": 0.2986789201608271, + "grad_norm": 0.2550566494464874, + "learning_rate": 4.905208848551809e-06, + "loss": 0.6768, + "step": 780 + }, + { + "epoch": 0.299061841853341, + "grad_norm": 0.24824075400829315, + "learning_rate": 4.904931627913578e-06, + "loss": 0.6764, + "step": 781 + }, + { + "epoch": 0.29944476354585486, + "grad_norm": 0.2658909261226654, + "learning_rate": 4.9046540103521765e-06, + "loss": 0.6801, + "step": 782 + }, + { + "epoch": 0.2998276852383688, + "grad_norm": 0.25569969415664673, + "learning_rate": 4.904375995913422e-06, + "loss": 0.696, + "step": 783 + }, + { + "epoch": 0.30021060693088264, + "grad_norm": 0.2601824104785919, + "learning_rate": 4.904097584643201e-06, + "loss": 0.707, + "step": 784 + }, + { + "epoch": 0.3005935286233965, + "grad_norm": 0.2594698667526245, + "learning_rate": 4.903818776587464e-06, + "loss": 0.705, + "step": 785 + }, + { + "epoch": 0.3009764503159104, + "grad_norm": 0.24596643447875977, + "learning_rate": 4.903539571792224e-06, + "loss": 0.6957, + "step": 786 + }, + { + "epoch": 0.3013593720084243, + "grad_norm": 0.25608620047569275, + "learning_rate": 4.903259970303567e-06, + "loss": 0.6978, + "step": 787 + }, + { + "epoch": 0.30174229370093814, + "grad_norm": 0.25500133633613586, + "learning_rate": 4.902979972167637e-06, + "loss": 0.6831, + "step": 788 + }, + { + "epoch": 0.30212521539345205, + "grad_norm": 0.2699365019798279, + "learning_rate": 4.902699577430649e-06, + "loss": 0.6903, + "step": 789 + }, + { + "epoch": 0.3025081370859659, + "grad_norm": 0.25621193647384644, + "learning_rate": 4.902418786138878e-06, + "loss": 0.6923, + "step": 790 + }, + { + "epoch": 0.3028910587784798, + "grad_norm": 0.2602829039096832, + "learning_rate": 4.902137598338669e-06, + "loss": 0.6556, + "step": 791 + }, + { + "epoch": 0.3032739804709937, + "grad_norm": 0.25946536660194397, + "learning_rate": 4.901856014076429e-06, + "loss": 0.6696, + "step": 792 + }, + { + "epoch": 0.30365690216350755, + "grad_norm": 0.2651613652706146, + "learning_rate": 4.901574033398635e-06, + "loss": 0.6842, + "step": 793 + }, + { + "epoch": 0.30403982385602146, + "grad_norm": 0.2593366801738739, + "learning_rate": 4.9012916563518255e-06, + "loss": 0.6854, + "step": 794 + }, + { + "epoch": 0.3044227455485353, + "grad_norm": 0.26688313484191895, + "learning_rate": 4.901008882982605e-06, + "loss": 0.6859, + "step": 795 + }, + { + "epoch": 0.3048056672410492, + "grad_norm": 0.254645973443985, + "learning_rate": 4.900725713337643e-06, + "loss": 0.672, + "step": 796 + }, + { + "epoch": 0.3051885889335631, + "grad_norm": 0.27753981947898865, + "learning_rate": 4.900442147463679e-06, + "loss": 0.6849, + "step": 797 + }, + { + "epoch": 0.30557151062607696, + "grad_norm": 0.25524258613586426, + "learning_rate": 4.90015818540751e-06, + "loss": 0.6982, + "step": 798 + }, + { + "epoch": 0.3059544323185909, + "grad_norm": 0.2585286796092987, + "learning_rate": 4.899873827216005e-06, + "loss": 0.6935, + "step": 799 + }, + { + "epoch": 0.30633735401110473, + "grad_norm": 0.2776353061199188, + "learning_rate": 4.899589072936097e-06, + "loss": 0.7047, + "step": 800 + }, + { + "epoch": 0.3067202757036186, + "grad_norm": 0.26528456807136536, + "learning_rate": 4.89930392261478e-06, + "loss": 0.716, + "step": 801 + }, + { + "epoch": 0.3071031973961325, + "grad_norm": 0.2629239559173584, + "learning_rate": 4.89901837629912e-06, + "loss": 0.6856, + "step": 802 + }, + { + "epoch": 0.30748611908864637, + "grad_norm": 0.25472939014434814, + "learning_rate": 4.8987324340362445e-06, + "loss": 0.6658, + "step": 803 + }, + { + "epoch": 0.30786904078116023, + "grad_norm": 0.25384870171546936, + "learning_rate": 4.898446095873345e-06, + "loss": 0.6897, + "step": 804 + }, + { + "epoch": 0.30825196247367415, + "grad_norm": 0.26003652811050415, + "learning_rate": 4.898159361857683e-06, + "loss": 0.6742, + "step": 805 + }, + { + "epoch": 0.308634884166188, + "grad_norm": 0.26269611716270447, + "learning_rate": 4.897872232036581e-06, + "loss": 0.685, + "step": 806 + }, + { + "epoch": 0.3090178058587019, + "grad_norm": 0.2580546736717224, + "learning_rate": 4.897584706457429e-06, + "loss": 0.6812, + "step": 807 + }, + { + "epoch": 0.3094007275512158, + "grad_norm": 0.25622835755348206, + "learning_rate": 4.897296785167681e-06, + "loss": 0.6828, + "step": 808 + }, + { + "epoch": 0.30978364924372964, + "grad_norm": 0.2555811107158661, + "learning_rate": 4.897008468214859e-06, + "loss": 0.6909, + "step": 809 + }, + { + "epoch": 0.31016657093624356, + "grad_norm": 0.2533404231071472, + "learning_rate": 4.896719755646548e-06, + "loss": 0.6968, + "step": 810 + }, + { + "epoch": 0.3105494926287574, + "grad_norm": 0.27069544792175293, + "learning_rate": 4.896430647510396e-06, + "loss": 0.6924, + "step": 811 + }, + { + "epoch": 0.3109324143212713, + "grad_norm": 0.2559230923652649, + "learning_rate": 4.896141143854122e-06, + "loss": 0.6983, + "step": 812 + }, + { + "epoch": 0.3113153360137852, + "grad_norm": 0.24792268872261047, + "learning_rate": 4.8958512447255055e-06, + "loss": 0.6721, + "step": 813 + }, + { + "epoch": 0.31169825770629905, + "grad_norm": 0.27516400814056396, + "learning_rate": 4.895560950172395e-06, + "loss": 0.6672, + "step": 814 + }, + { + "epoch": 0.31208117939881297, + "grad_norm": 0.25589725375175476, + "learning_rate": 4.895270260242701e-06, + "loss": 0.6861, + "step": 815 + }, + { + "epoch": 0.31246410109132683, + "grad_norm": 0.2611221373081207, + "learning_rate": 4.8949791749844e-06, + "loss": 0.6716, + "step": 816 + }, + { + "epoch": 0.3128470227838407, + "grad_norm": 0.26100388169288635, + "learning_rate": 4.8946876944455355e-06, + "loss": 0.6676, + "step": 817 + }, + { + "epoch": 0.3132299444763546, + "grad_norm": 0.27761179208755493, + "learning_rate": 4.8943958186742145e-06, + "loss": 0.6845, + "step": 818 + }, + { + "epoch": 0.31361286616886846, + "grad_norm": 0.26875218749046326, + "learning_rate": 4.894103547718611e-06, + "loss": 0.6902, + "step": 819 + }, + { + "epoch": 0.3139957878613823, + "grad_norm": 0.26297008991241455, + "learning_rate": 4.893810881626961e-06, + "loss": 0.7205, + "step": 820 + }, + { + "epoch": 0.31437870955389624, + "grad_norm": 0.2671011686325073, + "learning_rate": 4.893517820447568e-06, + "loss": 0.6782, + "step": 821 + }, + { + "epoch": 0.3147616312464101, + "grad_norm": 0.271348774433136, + "learning_rate": 4.8932243642288e-06, + "loss": 0.6993, + "step": 822 + }, + { + "epoch": 0.315144552938924, + "grad_norm": 0.271992564201355, + "learning_rate": 4.892930513019094e-06, + "loss": 0.682, + "step": 823 + }, + { + "epoch": 0.3155274746314379, + "grad_norm": 0.2696378827095032, + "learning_rate": 4.8926362668669454e-06, + "loss": 0.6805, + "step": 824 + }, + { + "epoch": 0.31591039632395174, + "grad_norm": 0.25821250677108765, + "learning_rate": 4.89234162582092e-06, + "loss": 0.6712, + "step": 825 + }, + { + "epoch": 0.31629331801646565, + "grad_norm": 0.2636469602584839, + "learning_rate": 4.892046589929645e-06, + "loss": 0.6956, + "step": 826 + }, + { + "epoch": 0.3166762397089795, + "grad_norm": 0.2635691165924072, + "learning_rate": 4.891751159241817e-06, + "loss": 0.6714, + "step": 827 + }, + { + "epoch": 0.3170591614014934, + "grad_norm": 0.26654869318008423, + "learning_rate": 4.891455333806194e-06, + "loss": 0.6818, + "step": 828 + }, + { + "epoch": 0.3174420830940073, + "grad_norm": 0.26420891284942627, + "learning_rate": 4.8911591136716015e-06, + "loss": 0.6848, + "step": 829 + }, + { + "epoch": 0.31782500478652115, + "grad_norm": 0.27182644605636597, + "learning_rate": 4.89086249888693e-06, + "loss": 0.7063, + "step": 830 + }, + { + "epoch": 0.31820792647903506, + "grad_norm": 0.26235124468803406, + "learning_rate": 4.8905654895011325e-06, + "loss": 0.6772, + "step": 831 + }, + { + "epoch": 0.3185908481715489, + "grad_norm": 0.27484792470932007, + "learning_rate": 4.89026808556323e-06, + "loss": 0.6824, + "step": 832 + }, + { + "epoch": 0.3189737698640628, + "grad_norm": 0.26560086011886597, + "learning_rate": 4.889970287122307e-06, + "loss": 0.6896, + "step": 833 + }, + { + "epoch": 0.3193566915565767, + "grad_norm": 0.2708161771297455, + "learning_rate": 4.889672094227515e-06, + "loss": 0.6785, + "step": 834 + }, + { + "epoch": 0.31973961324909056, + "grad_norm": 0.2682739794254303, + "learning_rate": 4.88937350692807e-06, + "loss": 0.7022, + "step": 835 + }, + { + "epoch": 0.3201225349416044, + "grad_norm": 0.26009342074394226, + "learning_rate": 4.8890745252732504e-06, + "loss": 0.6895, + "step": 836 + }, + { + "epoch": 0.32050545663411834, + "grad_norm": 0.2639565169811249, + "learning_rate": 4.888775149312403e-06, + "loss": 0.6939, + "step": 837 + }, + { + "epoch": 0.3208883783266322, + "grad_norm": 0.2597087025642395, + "learning_rate": 4.888475379094938e-06, + "loss": 0.6914, + "step": 838 + }, + { + "epoch": 0.3212713000191461, + "grad_norm": 0.2735402584075928, + "learning_rate": 4.888175214670331e-06, + "loss": 0.6842, + "step": 839 + }, + { + "epoch": 0.32165422171165997, + "grad_norm": 0.24896885454654694, + "learning_rate": 4.887874656088124e-06, + "loss": 0.6972, + "step": 840 + }, + { + "epoch": 0.32203714340417383, + "grad_norm": 0.25625696778297424, + "learning_rate": 4.887573703397921e-06, + "loss": 0.6811, + "step": 841 + }, + { + "epoch": 0.32242006509668775, + "grad_norm": 0.2628985047340393, + "learning_rate": 4.887272356649395e-06, + "loss": 0.6847, + "step": 842 + }, + { + "epoch": 0.3228029867892016, + "grad_norm": 0.26313072443008423, + "learning_rate": 4.88697061589228e-06, + "loss": 0.6871, + "step": 843 + }, + { + "epoch": 0.32318590848171547, + "grad_norm": 0.27291008830070496, + "learning_rate": 4.886668481176377e-06, + "loss": 0.6791, + "step": 844 + }, + { + "epoch": 0.3235688301742294, + "grad_norm": 0.26743587851524353, + "learning_rate": 4.886365952551555e-06, + "loss": 0.6861, + "step": 845 + }, + { + "epoch": 0.32395175186674324, + "grad_norm": 0.25605612993240356, + "learning_rate": 4.886063030067741e-06, + "loss": 0.6695, + "step": 846 + }, + { + "epoch": 0.32433467355925716, + "grad_norm": 0.2629465162754059, + "learning_rate": 4.885759713774933e-06, + "loss": 0.6802, + "step": 847 + }, + { + "epoch": 0.324717595251771, + "grad_norm": 0.26355117559432983, + "learning_rate": 4.885456003723192e-06, + "loss": 0.657, + "step": 848 + }, + { + "epoch": 0.3251005169442849, + "grad_norm": 0.258645623922348, + "learning_rate": 4.885151899962644e-06, + "loss": 0.6602, + "step": 849 + }, + { + "epoch": 0.3254834386367988, + "grad_norm": 0.2570325434207916, + "learning_rate": 4.884847402543479e-06, + "loss": 0.6907, + "step": 850 + }, + { + "epoch": 0.32586636032931265, + "grad_norm": 0.2724635899066925, + "learning_rate": 4.884542511515954e-06, + "loss": 0.6648, + "step": 851 + }, + { + "epoch": 0.3262492820218265, + "grad_norm": 0.2683566212654114, + "learning_rate": 4.884237226930389e-06, + "loss": 0.6936, + "step": 852 + }, + { + "epoch": 0.32663220371434043, + "grad_norm": 0.27550438046455383, + "learning_rate": 4.883931548837172e-06, + "loss": 0.6915, + "step": 853 + }, + { + "epoch": 0.3270151254068543, + "grad_norm": 0.26245197653770447, + "learning_rate": 4.883625477286751e-06, + "loss": 0.6805, + "step": 854 + }, + { + "epoch": 0.3273980470993682, + "grad_norm": 0.26423540711402893, + "learning_rate": 4.883319012329643e-06, + "loss": 0.6811, + "step": 855 + }, + { + "epoch": 0.32778096879188207, + "grad_norm": 0.28096291422843933, + "learning_rate": 4.883012154016428e-06, + "loss": 0.6846, + "step": 856 + }, + { + "epoch": 0.3281638904843959, + "grad_norm": 0.2643588185310364, + "learning_rate": 4.882704902397752e-06, + "loss": 0.7193, + "step": 857 + }, + { + "epoch": 0.32854681217690984, + "grad_norm": 0.2602258026599884, + "learning_rate": 4.882397257524326e-06, + "loss": 0.6753, + "step": 858 + }, + { + "epoch": 0.3289297338694237, + "grad_norm": 0.27310651540756226, + "learning_rate": 4.882089219446925e-06, + "loss": 0.6663, + "step": 859 + }, + { + "epoch": 0.32931265556193756, + "grad_norm": 0.26811903715133667, + "learning_rate": 4.8817807882163904e-06, + "loss": 0.6825, + "step": 860 + }, + { + "epoch": 0.3296955772544515, + "grad_norm": 0.266361266374588, + "learning_rate": 4.881471963883625e-06, + "loss": 0.694, + "step": 861 + }, + { + "epoch": 0.33007849894696534, + "grad_norm": 0.27039119601249695, + "learning_rate": 4.8811627464996005e-06, + "loss": 0.6773, + "step": 862 + }, + { + "epoch": 0.33046142063947925, + "grad_norm": 0.28884178400039673, + "learning_rate": 4.880853136115352e-06, + "loss": 0.6791, + "step": 863 + }, + { + "epoch": 0.3308443423319931, + "grad_norm": 0.2610810399055481, + "learning_rate": 4.8805431327819786e-06, + "loss": 0.6783, + "step": 864 + }, + { + "epoch": 0.331227264024507, + "grad_norm": 0.27807381749153137, + "learning_rate": 4.8802327365506456e-06, + "loss": 0.6904, + "step": 865 + }, + { + "epoch": 0.3316101857170209, + "grad_norm": 0.2709735333919525, + "learning_rate": 4.8799219474725814e-06, + "loss": 0.6842, + "step": 866 + }, + { + "epoch": 0.33199310740953475, + "grad_norm": 0.26760631799697876, + "learning_rate": 4.879610765599081e-06, + "loss": 0.6987, + "step": 867 + }, + { + "epoch": 0.3323760291020486, + "grad_norm": 0.26859745383262634, + "learning_rate": 4.8792991909815035e-06, + "loss": 0.6839, + "step": 868 + }, + { + "epoch": 0.3327589507945625, + "grad_norm": 0.27805396914482117, + "learning_rate": 4.878987223671274e-06, + "loss": 0.6883, + "step": 869 + }, + { + "epoch": 0.3331418724870764, + "grad_norm": 0.2685054540634155, + "learning_rate": 4.878674863719879e-06, + "loss": 0.6723, + "step": 870 + }, + { + "epoch": 0.3335247941795903, + "grad_norm": 0.2590447664260864, + "learning_rate": 4.878362111178875e-06, + "loss": 0.6792, + "step": 871 + }, + { + "epoch": 0.33390771587210416, + "grad_norm": 0.2719431519508362, + "learning_rate": 4.8780489660998776e-06, + "loss": 0.6908, + "step": 872 + }, + { + "epoch": 0.334290637564618, + "grad_norm": 0.27068865299224854, + "learning_rate": 4.877735428534572e-06, + "loss": 0.6869, + "step": 873 + }, + { + "epoch": 0.33467355925713194, + "grad_norm": 0.263995885848999, + "learning_rate": 4.877421498534705e-06, + "loss": 0.6813, + "step": 874 + }, + { + "epoch": 0.3350564809496458, + "grad_norm": 0.27139508724212646, + "learning_rate": 4.877107176152089e-06, + "loss": 0.6733, + "step": 875 + }, + { + "epoch": 0.33543940264215966, + "grad_norm": 0.2608059346675873, + "learning_rate": 4.876792461438604e-06, + "loss": 0.6744, + "step": 876 + }, + { + "epoch": 0.33582232433467357, + "grad_norm": 0.26705706119537354, + "learning_rate": 4.8764773544461895e-06, + "loss": 0.6919, + "step": 877 + }, + { + "epoch": 0.33620524602718743, + "grad_norm": 0.2600170373916626, + "learning_rate": 4.8761618552268534e-06, + "loss": 0.6843, + "step": 878 + }, + { + "epoch": 0.33658816771970135, + "grad_norm": 0.2667454481124878, + "learning_rate": 4.875845963832667e-06, + "loss": 0.6843, + "step": 879 + }, + { + "epoch": 0.3369710894122152, + "grad_norm": 0.2666393220424652, + "learning_rate": 4.875529680315768e-06, + "loss": 0.6843, + "step": 880 + }, + { + "epoch": 0.33735401110472907, + "grad_norm": 0.2725430727005005, + "learning_rate": 4.8752130047283565e-06, + "loss": 0.678, + "step": 881 + }, + { + "epoch": 0.337736932797243, + "grad_norm": 0.2890147864818573, + "learning_rate": 4.874895937122699e-06, + "loss": 0.696, + "step": 882 + }, + { + "epoch": 0.33811985448975684, + "grad_norm": 0.2604501247406006, + "learning_rate": 4.874578477551126e-06, + "loss": 0.6901, + "step": 883 + }, + { + "epoch": 0.3385027761822707, + "grad_norm": 0.26774969696998596, + "learning_rate": 4.8742606260660315e-06, + "loss": 0.6895, + "step": 884 + }, + { + "epoch": 0.3388856978747846, + "grad_norm": 0.2798639237880707, + "learning_rate": 4.873942382719878e-06, + "loss": 0.6654, + "step": 885 + }, + { + "epoch": 0.3392686195672985, + "grad_norm": 0.27278244495391846, + "learning_rate": 4.873623747565188e-06, + "loss": 0.6967, + "step": 886 + }, + { + "epoch": 0.3396515412598124, + "grad_norm": 0.25627878308296204, + "learning_rate": 4.873304720654551e-06, + "loss": 0.6756, + "step": 887 + }, + { + "epoch": 0.34003446295232626, + "grad_norm": 0.26743650436401367, + "learning_rate": 4.872985302040622e-06, + "loss": 0.7111, + "step": 888 + }, + { + "epoch": 0.3404173846448401, + "grad_norm": 0.27230358123779297, + "learning_rate": 4.872665491776119e-06, + "loss": 0.6911, + "step": 889 + }, + { + "epoch": 0.34080030633735403, + "grad_norm": 0.2792832553386688, + "learning_rate": 4.872345289913826e-06, + "loss": 0.6947, + "step": 890 + }, + { + "epoch": 0.3411832280298679, + "grad_norm": 0.27925628423690796, + "learning_rate": 4.872024696506588e-06, + "loss": 0.6942, + "step": 891 + }, + { + "epoch": 0.34156614972238175, + "grad_norm": 0.2760670781135559, + "learning_rate": 4.871703711607322e-06, + "loss": 0.6865, + "step": 892 + }, + { + "epoch": 0.34194907141489567, + "grad_norm": 0.27710941433906555, + "learning_rate": 4.871382335269001e-06, + "loss": 0.6918, + "step": 893 + }, + { + "epoch": 0.3423319931074095, + "grad_norm": 0.2548944056034088, + "learning_rate": 4.871060567544671e-06, + "loss": 0.6917, + "step": 894 + }, + { + "epoch": 0.34271491479992344, + "grad_norm": 0.26869043707847595, + "learning_rate": 4.870738408487433e-06, + "loss": 0.699, + "step": 895 + }, + { + "epoch": 0.3430978364924373, + "grad_norm": 0.2725992500782013, + "learning_rate": 4.870415858150462e-06, + "loss": 0.6977, + "step": 896 + }, + { + "epoch": 0.34348075818495116, + "grad_norm": 0.2812182605266571, + "learning_rate": 4.8700929165869925e-06, + "loss": 0.6617, + "step": 897 + }, + { + "epoch": 0.3438636798774651, + "grad_norm": 0.257763534784317, + "learning_rate": 4.869769583850324e-06, + "loss": 0.7007, + "step": 898 + }, + { + "epoch": 0.34424660156997894, + "grad_norm": 0.26617327332496643, + "learning_rate": 4.869445859993822e-06, + "loss": 0.6779, + "step": 899 + }, + { + "epoch": 0.3446295232624928, + "grad_norm": 0.2741871178150177, + "learning_rate": 4.8691217450709145e-06, + "loss": 0.6832, + "step": 900 + }, + { + "epoch": 0.3450124449550067, + "grad_norm": 0.2675487697124481, + "learning_rate": 4.8687972391350955e-06, + "loss": 0.6533, + "step": 901 + }, + { + "epoch": 0.3453953666475206, + "grad_norm": 0.2638150453567505, + "learning_rate": 4.868472342239924e-06, + "loss": 0.6685, + "step": 902 + }, + { + "epoch": 0.3457782883400345, + "grad_norm": 0.29722779989242554, + "learning_rate": 4.868147054439022e-06, + "loss": 0.687, + "step": 903 + }, + { + "epoch": 0.34616121003254835, + "grad_norm": 0.2719815969467163, + "learning_rate": 4.8678213757860774e-06, + "loss": 0.6888, + "step": 904 + }, + { + "epoch": 0.3465441317250622, + "grad_norm": 0.27905941009521484, + "learning_rate": 4.8674953063348415e-06, + "loss": 0.7021, + "step": 905 + }, + { + "epoch": 0.3469270534175761, + "grad_norm": 0.28070658445358276, + "learning_rate": 4.867168846139131e-06, + "loss": 0.6817, + "step": 906 + }, + { + "epoch": 0.34730997511009, + "grad_norm": 0.2584090828895569, + "learning_rate": 4.866841995252826e-06, + "loss": 0.6798, + "step": 907 + }, + { + "epoch": 0.34769289680260385, + "grad_norm": 0.2600197196006775, + "learning_rate": 4.8665147537298726e-06, + "loss": 0.6869, + "step": 908 + }, + { + "epoch": 0.34807581849511776, + "grad_norm": 0.2670934796333313, + "learning_rate": 4.8661871216242805e-06, + "loss": 0.6887, + "step": 909 + }, + { + "epoch": 0.3484587401876316, + "grad_norm": 0.2869112193584442, + "learning_rate": 4.865859098990123e-06, + "loss": 0.675, + "step": 910 + }, + { + "epoch": 0.34884166188014554, + "grad_norm": 0.300727903842926, + "learning_rate": 4.865530685881541e-06, + "loss": 0.6701, + "step": 911 + }, + { + "epoch": 0.3492245835726594, + "grad_norm": 0.2582904100418091, + "learning_rate": 4.8652018823527345e-06, + "loss": 0.6827, + "step": 912 + }, + { + "epoch": 0.34960750526517326, + "grad_norm": 0.27412885427474976, + "learning_rate": 4.864872688457972e-06, + "loss": 0.6772, + "step": 913 + }, + { + "epoch": 0.3499904269576872, + "grad_norm": 0.323688268661499, + "learning_rate": 4.864543104251587e-06, + "loss": 0.6837, + "step": 914 + }, + { + "epoch": 0.35037334865020103, + "grad_norm": 0.27548015117645264, + "learning_rate": 4.864213129787975e-06, + "loss": 0.6885, + "step": 915 + }, + { + "epoch": 0.3507562703427149, + "grad_norm": 0.2752622067928314, + "learning_rate": 4.863882765121596e-06, + "loss": 0.7021, + "step": 916 + }, + { + "epoch": 0.3511391920352288, + "grad_norm": 0.2829137444496155, + "learning_rate": 4.863552010306976e-06, + "loss": 0.7023, + "step": 917 + }, + { + "epoch": 0.35152211372774267, + "grad_norm": 0.30739355087280273, + "learning_rate": 4.863220865398705e-06, + "loss": 0.6655, + "step": 918 + }, + { + "epoch": 0.3519050354202566, + "grad_norm": 0.26492998003959656, + "learning_rate": 4.862889330451437e-06, + "loss": 0.6986, + "step": 919 + }, + { + "epoch": 0.35228795711277044, + "grad_norm": 0.3324214518070221, + "learning_rate": 4.862557405519889e-06, + "loss": 0.6975, + "step": 920 + }, + { + "epoch": 0.3526708788052843, + "grad_norm": 0.3313480615615845, + "learning_rate": 4.862225090658844e-06, + "loss": 0.6782, + "step": 921 + }, + { + "epoch": 0.3530538004977982, + "grad_norm": 0.26132631301879883, + "learning_rate": 4.861892385923151e-06, + "loss": 0.6698, + "step": 922 + }, + { + "epoch": 0.3534367221903121, + "grad_norm": 0.2920131981372833, + "learning_rate": 4.86155929136772e-06, + "loss": 0.6999, + "step": 923 + }, + { + "epoch": 0.35381964388282594, + "grad_norm": 0.30363237857818604, + "learning_rate": 4.861225807047526e-06, + "loss": 0.6962, + "step": 924 + }, + { + "epoch": 0.35420256557533986, + "grad_norm": 0.28039583563804626, + "learning_rate": 4.86089193301761e-06, + "loss": 0.6997, + "step": 925 + }, + { + "epoch": 0.3545854872678537, + "grad_norm": 0.2662922739982605, + "learning_rate": 4.860557669333076e-06, + "loss": 0.6465, + "step": 926 + }, + { + "epoch": 0.35496840896036763, + "grad_norm": 0.26688340306282043, + "learning_rate": 4.860223016049094e-06, + "loss": 0.6846, + "step": 927 + }, + { + "epoch": 0.3553513306528815, + "grad_norm": 0.2974063754081726, + "learning_rate": 4.859887973220896e-06, + "loss": 0.6984, + "step": 928 + }, + { + "epoch": 0.35573425234539535, + "grad_norm": 0.2696240246295929, + "learning_rate": 4.859552540903779e-06, + "loss": 0.6646, + "step": 929 + }, + { + "epoch": 0.35611717403790927, + "grad_norm": 0.26638904213905334, + "learning_rate": 4.859216719153105e-06, + "loss": 0.6651, + "step": 930 + }, + { + "epoch": 0.3565000957304231, + "grad_norm": 0.2707373797893524, + "learning_rate": 4.8588805080243e-06, + "loss": 0.7, + "step": 931 + }, + { + "epoch": 0.356883017422937, + "grad_norm": 0.2757275700569153, + "learning_rate": 4.858543907572854e-06, + "loss": 0.6929, + "step": 932 + }, + { + "epoch": 0.3572659391154509, + "grad_norm": 0.26656290888786316, + "learning_rate": 4.858206917854322e-06, + "loss": 0.6706, + "step": 933 + }, + { + "epoch": 0.35764886080796476, + "grad_norm": 0.2593183219432831, + "learning_rate": 4.85786953892432e-06, + "loss": 0.6753, + "step": 934 + }, + { + "epoch": 0.3580317825004787, + "grad_norm": 0.26421022415161133, + "learning_rate": 4.857531770838534e-06, + "loss": 0.6742, + "step": 935 + }, + { + "epoch": 0.35841470419299254, + "grad_norm": 0.2740086317062378, + "learning_rate": 4.857193613652711e-06, + "loss": 0.6766, + "step": 936 + }, + { + "epoch": 0.3587976258855064, + "grad_norm": 0.270923912525177, + "learning_rate": 4.856855067422661e-06, + "loss": 0.6986, + "step": 937 + }, + { + "epoch": 0.3591805475780203, + "grad_norm": 0.2739553153514862, + "learning_rate": 4.856516132204258e-06, + "loss": 0.6595, + "step": 938 + }, + { + "epoch": 0.3595634692705342, + "grad_norm": 0.2649962902069092, + "learning_rate": 4.856176808053446e-06, + "loss": 0.696, + "step": 939 + }, + { + "epoch": 0.35994639096304804, + "grad_norm": 0.27668675780296326, + "learning_rate": 4.8558370950262256e-06, + "loss": 0.6849, + "step": 940 + }, + { + "epoch": 0.36032931265556195, + "grad_norm": 0.2732677757740021, + "learning_rate": 4.855496993178666e-06, + "loss": 0.6809, + "step": 941 + }, + { + "epoch": 0.3607122343480758, + "grad_norm": 0.27247822284698486, + "learning_rate": 4.855156502566899e-06, + "loss": 0.6912, + "step": 942 + }, + { + "epoch": 0.3610951560405897, + "grad_norm": 0.2725856900215149, + "learning_rate": 4.854815623247122e-06, + "loss": 0.6715, + "step": 943 + }, + { + "epoch": 0.3614780777331036, + "grad_norm": 0.25335705280303955, + "learning_rate": 4.854474355275595e-06, + "loss": 0.6646, + "step": 944 + }, + { + "epoch": 0.36186099942561745, + "grad_norm": 0.2773241400718689, + "learning_rate": 4.854132698708642e-06, + "loss": 0.6806, + "step": 945 + }, + { + "epoch": 0.36224392111813136, + "grad_norm": 0.26226091384887695, + "learning_rate": 4.853790653602653e-06, + "loss": 0.6695, + "step": 946 + }, + { + "epoch": 0.3626268428106452, + "grad_norm": 0.27693572640419006, + "learning_rate": 4.85344822001408e-06, + "loss": 0.6796, + "step": 947 + }, + { + "epoch": 0.3630097645031591, + "grad_norm": 0.26777103543281555, + "learning_rate": 4.853105397999442e-06, + "loss": 0.6876, + "step": 948 + }, + { + "epoch": 0.363392686195673, + "grad_norm": 0.2739417850971222, + "learning_rate": 4.852762187615318e-06, + "loss": 0.6838, + "step": 949 + }, + { + "epoch": 0.36377560788818686, + "grad_norm": 0.2750835418701172, + "learning_rate": 4.852418588918355e-06, + "loss": 0.6644, + "step": 950 + }, + { + "epoch": 0.3641585295807008, + "grad_norm": 0.2623859941959381, + "learning_rate": 4.852074601965261e-06, + "loss": 0.6554, + "step": 951 + }, + { + "epoch": 0.36454145127321463, + "grad_norm": 0.26994043588638306, + "learning_rate": 4.851730226812811e-06, + "loss": 0.6794, + "step": 952 + }, + { + "epoch": 0.3649243729657285, + "grad_norm": 0.25890594720840454, + "learning_rate": 4.851385463517841e-06, + "loss": 0.6914, + "step": 953 + }, + { + "epoch": 0.3653072946582424, + "grad_norm": 0.26775312423706055, + "learning_rate": 4.851040312137254e-06, + "loss": 0.6853, + "step": 954 + }, + { + "epoch": 0.36569021635075627, + "grad_norm": 0.26391908526420593, + "learning_rate": 4.850694772728015e-06, + "loss": 0.679, + "step": 955 + }, + { + "epoch": 0.36607313804327013, + "grad_norm": 0.25619781017303467, + "learning_rate": 4.850348845347155e-06, + "loss": 0.6732, + "step": 956 + }, + { + "epoch": 0.36645605973578405, + "grad_norm": 0.26470351219177246, + "learning_rate": 4.850002530051765e-06, + "loss": 0.6927, + "step": 957 + }, + { + "epoch": 0.3668389814282979, + "grad_norm": 0.2638048529624939, + "learning_rate": 4.849655826899005e-06, + "loss": 0.6904, + "step": 958 + }, + { + "epoch": 0.3672219031208118, + "grad_norm": 0.25979623198509216, + "learning_rate": 4.849308735946096e-06, + "loss": 0.6736, + "step": 959 + }, + { + "epoch": 0.3676048248133257, + "grad_norm": 0.2671094834804535, + "learning_rate": 4.8489612572503245e-06, + "loss": 0.662, + "step": 960 + }, + { + "epoch": 0.36798774650583954, + "grad_norm": 0.26025474071502686, + "learning_rate": 4.84861339086904e-06, + "loss": 0.6571, + "step": 961 + }, + { + "epoch": 0.36837066819835346, + "grad_norm": 0.26459765434265137, + "learning_rate": 4.848265136859656e-06, + "loss": 0.6853, + "step": 962 + }, + { + "epoch": 0.3687535898908673, + "grad_norm": 0.2781910300254822, + "learning_rate": 4.84791649527965e-06, + "loss": 0.6968, + "step": 963 + }, + { + "epoch": 0.3691365115833812, + "grad_norm": 0.2741437256336212, + "learning_rate": 4.8475674661865636e-06, + "loss": 0.6897, + "step": 964 + }, + { + "epoch": 0.3695194332758951, + "grad_norm": 0.2687261402606964, + "learning_rate": 4.8472180496380035e-06, + "loss": 0.6884, + "step": 965 + }, + { + "epoch": 0.36990235496840895, + "grad_norm": 0.2683657705783844, + "learning_rate": 4.846868245691639e-06, + "loss": 0.6919, + "step": 966 + }, + { + "epoch": 0.37028527666092287, + "grad_norm": 0.28037914633750916, + "learning_rate": 4.846518054405203e-06, + "loss": 0.6939, + "step": 967 + }, + { + "epoch": 0.37066819835343673, + "grad_norm": 0.2678395211696625, + "learning_rate": 4.846167475836494e-06, + "loss": 0.6784, + "step": 968 + }, + { + "epoch": 0.3710511200459506, + "grad_norm": 0.263564795255661, + "learning_rate": 4.8458165100433725e-06, + "loss": 0.6776, + "step": 969 + }, + { + "epoch": 0.3714340417384645, + "grad_norm": 0.26803597807884216, + "learning_rate": 4.845465157083765e-06, + "loss": 0.6867, + "step": 970 + }, + { + "epoch": 0.37181696343097836, + "grad_norm": 0.27506449818611145, + "learning_rate": 4.8451134170156585e-06, + "loss": 0.6862, + "step": 971 + }, + { + "epoch": 0.3721998851234922, + "grad_norm": 0.26313772797584534, + "learning_rate": 4.844761289897109e-06, + "loss": 0.6739, + "step": 972 + }, + { + "epoch": 0.37258280681600614, + "grad_norm": 0.26042506098747253, + "learning_rate": 4.844408775786231e-06, + "loss": 0.686, + "step": 973 + }, + { + "epoch": 0.37296572850852, + "grad_norm": 0.2822672426700592, + "learning_rate": 4.844055874741208e-06, + "loss": 0.6853, + "step": 974 + }, + { + "epoch": 0.3733486502010339, + "grad_norm": 0.26979514956474304, + "learning_rate": 4.843702586820282e-06, + "loss": 0.6936, + "step": 975 + }, + { + "epoch": 0.3737315718935478, + "grad_norm": 0.27148738503456116, + "learning_rate": 4.843348912081763e-06, + "loss": 0.6817, + "step": 976 + }, + { + "epoch": 0.37411449358606164, + "grad_norm": 0.2711876332759857, + "learning_rate": 4.842994850584023e-06, + "loss": 0.6781, + "step": 977 + }, + { + "epoch": 0.37449741527857555, + "grad_norm": 0.25822556018829346, + "learning_rate": 4.842640402385499e-06, + "loss": 0.6595, + "step": 978 + }, + { + "epoch": 0.3748803369710894, + "grad_norm": 0.27487432956695557, + "learning_rate": 4.84228556754469e-06, + "loss": 0.6633, + "step": 979 + }, + { + "epoch": 0.37526325866360327, + "grad_norm": 0.27101749181747437, + "learning_rate": 4.841930346120161e-06, + "loss": 0.6827, + "step": 980 + }, + { + "epoch": 0.3756461803561172, + "grad_norm": 0.26800671219825745, + "learning_rate": 4.841574738170538e-06, + "loss": 0.6703, + "step": 981 + }, + { + "epoch": 0.37602910204863105, + "grad_norm": 0.2811592221260071, + "learning_rate": 4.841218743754515e-06, + "loss": 0.7082, + "step": 982 + }, + { + "epoch": 0.37641202374114496, + "grad_norm": 0.2691783010959625, + "learning_rate": 4.840862362930845e-06, + "loss": 0.7007, + "step": 983 + }, + { + "epoch": 0.3767949454336588, + "grad_norm": 0.2679787576198578, + "learning_rate": 4.840505595758348e-06, + "loss": 0.6866, + "step": 984 + }, + { + "epoch": 0.3771778671261727, + "grad_norm": 0.2782638669013977, + "learning_rate": 4.840148442295907e-06, + "loss": 0.6829, + "step": 985 + }, + { + "epoch": 0.3775607888186866, + "grad_norm": 0.25745078921318054, + "learning_rate": 4.839790902602469e-06, + "loss": 0.6601, + "step": 986 + }, + { + "epoch": 0.37794371051120046, + "grad_norm": 0.2637532353401184, + "learning_rate": 4.839432976737043e-06, + "loss": 0.6565, + "step": 987 + }, + { + "epoch": 0.3783266322037143, + "grad_norm": 0.269976943731308, + "learning_rate": 4.839074664758705e-06, + "loss": 0.6715, + "step": 988 + }, + { + "epoch": 0.37870955389622823, + "grad_norm": 0.2653595209121704, + "learning_rate": 4.83871596672659e-06, + "loss": 0.6675, + "step": 989 + }, + { + "epoch": 0.3790924755887421, + "grad_norm": 0.26211804151535034, + "learning_rate": 4.838356882699902e-06, + "loss": 0.6545, + "step": 990 + }, + { + "epoch": 0.379475397281256, + "grad_norm": 0.2690889537334442, + "learning_rate": 4.837997412737905e-06, + "loss": 0.6998, + "step": 991 + }, + { + "epoch": 0.37985831897376987, + "grad_norm": 0.26827001571655273, + "learning_rate": 4.837637556899928e-06, + "loss": 0.6807, + "step": 992 + }, + { + "epoch": 0.38024124066628373, + "grad_norm": 0.2662839889526367, + "learning_rate": 4.837277315245364e-06, + "loss": 0.6842, + "step": 993 + }, + { + "epoch": 0.38062416235879765, + "grad_norm": 0.26887455582618713, + "learning_rate": 4.836916687833668e-06, + "loss": 0.6807, + "step": 994 + }, + { + "epoch": 0.3810070840513115, + "grad_norm": 0.2758978009223938, + "learning_rate": 4.836555674724362e-06, + "loss": 0.6982, + "step": 995 + }, + { + "epoch": 0.38139000574382537, + "grad_norm": 0.2752774953842163, + "learning_rate": 4.836194275977027e-06, + "loss": 0.6831, + "step": 996 + }, + { + "epoch": 0.3817729274363393, + "grad_norm": 0.3027510344982147, + "learning_rate": 4.835832491651314e-06, + "loss": 0.7033, + "step": 997 + }, + { + "epoch": 0.38215584912885314, + "grad_norm": 0.26609593629837036, + "learning_rate": 4.83547032180693e-06, + "loss": 0.6718, + "step": 998 + }, + { + "epoch": 0.38253877082136706, + "grad_norm": 0.2735961675643921, + "learning_rate": 4.835107766503652e-06, + "loss": 0.6791, + "step": 999 + }, + { + "epoch": 0.3829216925138809, + "grad_norm": 0.26820945739746094, + "learning_rate": 4.834744825801318e-06, + "loss": 0.6805, + "step": 1000 + }, + { + "epoch": 0.3833046142063948, + "grad_norm": 0.2665892541408539, + "learning_rate": 4.834381499759827e-06, + "loss": 0.6672, + "step": 1001 + }, + { + "epoch": 0.3836875358989087, + "grad_norm": 0.2774675786495209, + "learning_rate": 4.834017788439148e-06, + "loss": 0.6774, + "step": 1002 + }, + { + "epoch": 0.38407045759142255, + "grad_norm": 0.27913185954093933, + "learning_rate": 4.833653691899308e-06, + "loss": 0.6656, + "step": 1003 + }, + { + "epoch": 0.3844533792839364, + "grad_norm": 0.2730454206466675, + "learning_rate": 4.833289210200399e-06, + "loss": 0.691, + "step": 1004 + }, + { + "epoch": 0.38483630097645033, + "grad_norm": 0.27499866485595703, + "learning_rate": 4.8329243434025785e-06, + "loss": 0.688, + "step": 1005 + }, + { + "epoch": 0.3852192226689642, + "grad_norm": 0.26650702953338623, + "learning_rate": 4.832559091566066e-06, + "loss": 0.6836, + "step": 1006 + }, + { + "epoch": 0.3856021443614781, + "grad_norm": 0.2753099501132965, + "learning_rate": 4.832193454751143e-06, + "loss": 0.6721, + "step": 1007 + }, + { + "epoch": 0.38598506605399197, + "grad_norm": 0.275810569524765, + "learning_rate": 4.831827433018158e-06, + "loss": 0.6722, + "step": 1008 + }, + { + "epoch": 0.3863679877465058, + "grad_norm": 0.27451977133750916, + "learning_rate": 4.8314610264275205e-06, + "loss": 0.6783, + "step": 1009 + }, + { + "epoch": 0.38675090943901974, + "grad_norm": 0.26779070496559143, + "learning_rate": 4.831094235039704e-06, + "loss": 0.6762, + "step": 1010 + }, + { + "epoch": 0.3871338311315336, + "grad_norm": 0.2588435113430023, + "learning_rate": 4.830727058915247e-06, + "loss": 0.6666, + "step": 1011 + }, + { + "epoch": 0.38751675282404746, + "grad_norm": 0.2793857157230377, + "learning_rate": 4.830359498114749e-06, + "loss": 0.6658, + "step": 1012 + }, + { + "epoch": 0.3878996745165614, + "grad_norm": 0.2771088182926178, + "learning_rate": 4.829991552698875e-06, + "loss": 0.674, + "step": 1013 + }, + { + "epoch": 0.38828259620907524, + "grad_norm": 0.2702658772468567, + "learning_rate": 4.829623222728352e-06, + "loss": 0.6715, + "step": 1014 + }, + { + "epoch": 0.38866551790158915, + "grad_norm": 0.28128373622894287, + "learning_rate": 4.829254508263971e-06, + "loss": 0.6675, + "step": 1015 + }, + { + "epoch": 0.389048439594103, + "grad_norm": 0.26931461691856384, + "learning_rate": 4.828885409366589e-06, + "loss": 0.6567, + "step": 1016 + }, + { + "epoch": 0.3894313612866169, + "grad_norm": 0.26983270049095154, + "learning_rate": 4.828515926097121e-06, + "loss": 0.6727, + "step": 1017 + }, + { + "epoch": 0.3898142829791308, + "grad_norm": 0.2659534215927124, + "learning_rate": 4.82814605851655e-06, + "loss": 0.6587, + "step": 1018 + }, + { + "epoch": 0.39019720467164465, + "grad_norm": 0.26131471991539, + "learning_rate": 4.827775806685922e-06, + "loss": 0.6813, + "step": 1019 + }, + { + "epoch": 0.3905801263641585, + "grad_norm": 0.28075239062309265, + "learning_rate": 4.8274051706663435e-06, + "loss": 0.672, + "step": 1020 + }, + { + "epoch": 0.3909630480566724, + "grad_norm": 0.2587306499481201, + "learning_rate": 4.827034150518987e-06, + "loss": 0.6643, + "step": 1021 + }, + { + "epoch": 0.3913459697491863, + "grad_norm": 0.2753465473651886, + "learning_rate": 4.826662746305089e-06, + "loss": 0.656, + "step": 1022 + }, + { + "epoch": 0.3917288914417002, + "grad_norm": 0.27708113193511963, + "learning_rate": 4.826290958085945e-06, + "loss": 0.6798, + "step": 1023 + }, + { + "epoch": 0.39211181313421406, + "grad_norm": 0.2657153308391571, + "learning_rate": 4.825918785922921e-06, + "loss": 0.6881, + "step": 1024 + }, + { + "epoch": 0.3924947348267279, + "grad_norm": 0.26796260476112366, + "learning_rate": 4.825546229877439e-06, + "loss": 0.6856, + "step": 1025 + }, + { + "epoch": 0.39287765651924184, + "grad_norm": 0.2698305547237396, + "learning_rate": 4.82517329001099e-06, + "loss": 0.6748, + "step": 1026 + }, + { + "epoch": 0.3932605782117557, + "grad_norm": 0.2884180545806885, + "learning_rate": 4.824799966385124e-06, + "loss": 0.683, + "step": 1027 + }, + { + "epoch": 0.39364349990426956, + "grad_norm": 0.269398033618927, + "learning_rate": 4.824426259061458e-06, + "loss": 0.6904, + "step": 1028 + }, + { + "epoch": 0.39402642159678347, + "grad_norm": 0.2623305320739746, + "learning_rate": 4.82405216810167e-06, + "loss": 0.6731, + "step": 1029 + }, + { + "epoch": 0.39440934328929733, + "grad_norm": 0.2715912461280823, + "learning_rate": 4.823677693567502e-06, + "loss": 0.6661, + "step": 1030 + }, + { + "epoch": 0.39479226498181125, + "grad_norm": 0.27043530344963074, + "learning_rate": 4.82330283552076e-06, + "loss": 0.6994, + "step": 1031 + }, + { + "epoch": 0.3951751866743251, + "grad_norm": 0.27699190378189087, + "learning_rate": 4.822927594023312e-06, + "loss": 0.6645, + "step": 1032 + }, + { + "epoch": 0.39555810836683897, + "grad_norm": 0.27182725071907043, + "learning_rate": 4.8225519691370905e-06, + "loss": 0.679, + "step": 1033 + }, + { + "epoch": 0.3959410300593529, + "grad_norm": 0.27347299456596375, + "learning_rate": 4.8221759609240894e-06, + "loss": 0.678, + "step": 1034 + }, + { + "epoch": 0.39632395175186674, + "grad_norm": 0.2720860242843628, + "learning_rate": 4.821799569446368e-06, + "loss": 0.7019, + "step": 1035 + }, + { + "epoch": 0.3967068734443806, + "grad_norm": 0.27360793948173523, + "learning_rate": 4.821422794766049e-06, + "loss": 0.6719, + "step": 1036 + }, + { + "epoch": 0.3970897951368945, + "grad_norm": 0.2907519042491913, + "learning_rate": 4.821045636945316e-06, + "loss": 0.6768, + "step": 1037 + }, + { + "epoch": 0.3974727168294084, + "grad_norm": 0.2736169397830963, + "learning_rate": 4.820668096046418e-06, + "loss": 0.6717, + "step": 1038 + }, + { + "epoch": 0.3978556385219223, + "grad_norm": 0.270683616399765, + "learning_rate": 4.820290172131667e-06, + "loss": 0.6808, + "step": 1039 + }, + { + "epoch": 0.39823856021443615, + "grad_norm": 0.2760835587978363, + "learning_rate": 4.819911865263434e-06, + "loss": 0.6861, + "step": 1040 + }, + { + "epoch": 0.39862148190695, + "grad_norm": 0.2689175605773926, + "learning_rate": 4.8195331755041616e-06, + "loss": 0.6642, + "step": 1041 + }, + { + "epoch": 0.39900440359946393, + "grad_norm": 0.2810250222682953, + "learning_rate": 4.819154102916349e-06, + "loss": 0.6601, + "step": 1042 + }, + { + "epoch": 0.3993873252919778, + "grad_norm": 0.2765415608882904, + "learning_rate": 4.81877464756256e-06, + "loss": 0.6694, + "step": 1043 + }, + { + "epoch": 0.39977024698449165, + "grad_norm": 0.2762261927127838, + "learning_rate": 4.818394809505421e-06, + "loss": 0.6962, + "step": 1044 + }, + { + "epoch": 0.40015316867700557, + "grad_norm": 0.28600603342056274, + "learning_rate": 4.818014588807625e-06, + "loss": 0.6734, + "step": 1045 + }, + { + "epoch": 0.4005360903695194, + "grad_norm": 0.26933541893959045, + "learning_rate": 4.817633985531924e-06, + "loss": 0.6775, + "step": 1046 + }, + { + "epoch": 0.4009190120620333, + "grad_norm": 0.2707444131374359, + "learning_rate": 4.817252999741137e-06, + "loss": 0.6924, + "step": 1047 + }, + { + "epoch": 0.4013019337545472, + "grad_norm": 0.27176135778427124, + "learning_rate": 4.816871631498141e-06, + "loss": 0.6689, + "step": 1048 + }, + { + "epoch": 0.40168485544706106, + "grad_norm": 0.28387269377708435, + "learning_rate": 4.81648988086588e-06, + "loss": 0.6879, + "step": 1049 + }, + { + "epoch": 0.402067777139575, + "grad_norm": 0.2803856432437897, + "learning_rate": 4.816107747907362e-06, + "loss": 0.6779, + "step": 1050 + }, + { + "epoch": 0.40245069883208884, + "grad_norm": 0.26472827792167664, + "learning_rate": 4.815725232685654e-06, + "loss": 0.6671, + "step": 1051 + }, + { + "epoch": 0.4028336205246027, + "grad_norm": 0.2867729663848877, + "learning_rate": 4.815342335263891e-06, + "loss": 0.6849, + "step": 1052 + }, + { + "epoch": 0.4032165422171166, + "grad_norm": 0.2667827904224396, + "learning_rate": 4.814959055705266e-06, + "loss": 0.6759, + "step": 1053 + }, + { + "epoch": 0.4035994639096305, + "grad_norm": 0.2698298394680023, + "learning_rate": 4.814575394073039e-06, + "loss": 0.6777, + "step": 1054 + }, + { + "epoch": 0.40398238560214433, + "grad_norm": 0.27925172448158264, + "learning_rate": 4.8141913504305305e-06, + "loss": 0.6872, + "step": 1055 + }, + { + "epoch": 0.40436530729465825, + "grad_norm": 0.26620960235595703, + "learning_rate": 4.813806924841125e-06, + "loss": 0.6848, + "step": 1056 + }, + { + "epoch": 0.4047482289871721, + "grad_norm": 0.2703430950641632, + "learning_rate": 4.813422117368273e-06, + "loss": 0.6508, + "step": 1057 + }, + { + "epoch": 0.405131150679686, + "grad_norm": 0.2778773903846741, + "learning_rate": 4.813036928075483e-06, + "loss": 0.6802, + "step": 1058 + }, + { + "epoch": 0.4055140723721999, + "grad_norm": 0.268136590719223, + "learning_rate": 4.812651357026329e-06, + "loss": 0.6894, + "step": 1059 + }, + { + "epoch": 0.40589699406471375, + "grad_norm": 0.2591809332370758, + "learning_rate": 4.812265404284447e-06, + "loss": 0.6751, + "step": 1060 + }, + { + "epoch": 0.40627991575722766, + "grad_norm": 0.26853811740875244, + "learning_rate": 4.811879069913539e-06, + "loss": 0.7125, + "step": 1061 + }, + { + "epoch": 0.4066628374497415, + "grad_norm": 0.2730932831764221, + "learning_rate": 4.811492353977366e-06, + "loss": 0.6852, + "step": 1062 + }, + { + "epoch": 0.4070457591422554, + "grad_norm": 0.26328590512275696, + "learning_rate": 4.811105256539754e-06, + "loss": 0.6585, + "step": 1063 + }, + { + "epoch": 0.4074286808347693, + "grad_norm": 0.27260056138038635, + "learning_rate": 4.810717777664592e-06, + "loss": 0.677, + "step": 1064 + }, + { + "epoch": 0.40781160252728316, + "grad_norm": 0.2737332880496979, + "learning_rate": 4.810329917415832e-06, + "loss": 0.6636, + "step": 1065 + }, + { + "epoch": 0.4081945242197971, + "grad_norm": 0.2796006500720978, + "learning_rate": 4.809941675857487e-06, + "loss": 0.6968, + "step": 1066 + }, + { + "epoch": 0.40857744591231093, + "grad_norm": 0.2870083153247833, + "learning_rate": 4.809553053053635e-06, + "loss": 0.6814, + "step": 1067 + }, + { + "epoch": 0.4089603676048248, + "grad_norm": 0.2642041742801666, + "learning_rate": 4.809164049068419e-06, + "loss": 0.6851, + "step": 1068 + }, + { + "epoch": 0.4093432892973387, + "grad_norm": 0.27570584416389465, + "learning_rate": 4.808774663966039e-06, + "loss": 0.6857, + "step": 1069 + }, + { + "epoch": 0.40972621098985257, + "grad_norm": 0.2671034038066864, + "learning_rate": 4.808384897810763e-06, + "loss": 0.663, + "step": 1070 + }, + { + "epoch": 0.41010913268236643, + "grad_norm": 0.2757396996021271, + "learning_rate": 4.80799475066692e-06, + "loss": 0.6729, + "step": 1071 + }, + { + "epoch": 0.41049205437488034, + "grad_norm": 0.28543686866760254, + "learning_rate": 4.807604222598902e-06, + "loss": 0.7029, + "step": 1072 + }, + { + "epoch": 0.4108749760673942, + "grad_norm": 0.28299906849861145, + "learning_rate": 4.807213313671163e-06, + "loss": 0.6813, + "step": 1073 + }, + { + "epoch": 0.4112578977599081, + "grad_norm": 0.27811309695243835, + "learning_rate": 4.806822023948223e-06, + "loss": 0.6832, + "step": 1074 + }, + { + "epoch": 0.411640819452422, + "grad_norm": 0.2794390320777893, + "learning_rate": 4.80643035349466e-06, + "loss": 0.6993, + "step": 1075 + }, + { + "epoch": 0.41202374114493584, + "grad_norm": 0.2820933163166046, + "learning_rate": 4.806038302375119e-06, + "loss": 0.6751, + "step": 1076 + }, + { + "epoch": 0.41240666283744976, + "grad_norm": 0.28078168630599976, + "learning_rate": 4.805645870654306e-06, + "loss": 0.6661, + "step": 1077 + }, + { + "epoch": 0.4127895845299636, + "grad_norm": 0.2717399001121521, + "learning_rate": 4.805253058396989e-06, + "loss": 0.6955, + "step": 1078 + }, + { + "epoch": 0.4131725062224775, + "grad_norm": 0.2717928886413574, + "learning_rate": 4.804859865668002e-06, + "loss": 0.6648, + "step": 1079 + }, + { + "epoch": 0.4135554279149914, + "grad_norm": 0.28002336621284485, + "learning_rate": 4.804466292532239e-06, + "loss": 0.6931, + "step": 1080 + }, + { + "epoch": 0.41393834960750525, + "grad_norm": 0.28147804737091064, + "learning_rate": 4.804072339054656e-06, + "loss": 0.6844, + "step": 1081 + }, + { + "epoch": 0.41432127130001917, + "grad_norm": 0.2660662829875946, + "learning_rate": 4.803678005300274e-06, + "loss": 0.6701, + "step": 1082 + }, + { + "epoch": 0.414704192992533, + "grad_norm": 0.26961514353752136, + "learning_rate": 4.803283291334177e-06, + "loss": 0.6874, + "step": 1083 + }, + { + "epoch": 0.4150871146850469, + "grad_norm": 0.2822124660015106, + "learning_rate": 4.802888197221511e-06, + "loss": 0.6702, + "step": 1084 + }, + { + "epoch": 0.4154700363775608, + "grad_norm": 0.264700710773468, + "learning_rate": 4.802492723027482e-06, + "loss": 0.6755, + "step": 1085 + }, + { + "epoch": 0.41585295807007466, + "grad_norm": 0.2882467210292816, + "learning_rate": 4.802096868817363e-06, + "loss": 0.6995, + "step": 1086 + }, + { + "epoch": 0.4162358797625885, + "grad_norm": 0.2640404999256134, + "learning_rate": 4.801700634656488e-06, + "loss": 0.6657, + "step": 1087 + }, + { + "epoch": 0.41661880145510244, + "grad_norm": 0.2711586654186249, + "learning_rate": 4.801304020610255e-06, + "loss": 0.6831, + "step": 1088 + }, + { + "epoch": 0.4170017231476163, + "grad_norm": 0.27783191204071045, + "learning_rate": 4.80090702674412e-06, + "loss": 0.6816, + "step": 1089 + }, + { + "epoch": 0.4173846448401302, + "grad_norm": 0.28974947333335876, + "learning_rate": 4.800509653123607e-06, + "loss": 0.6523, + "step": 1090 + }, + { + "epoch": 0.4177675665326441, + "grad_norm": 0.27213558554649353, + "learning_rate": 4.800111899814301e-06, + "loss": 0.6873, + "step": 1091 + }, + { + "epoch": 0.41815048822515793, + "grad_norm": 0.26818469166755676, + "learning_rate": 4.7997137668818495e-06, + "loss": 0.6833, + "step": 1092 + }, + { + "epoch": 0.41853340991767185, + "grad_norm": 0.2849678099155426, + "learning_rate": 4.7993152543919615e-06, + "loss": 0.676, + "step": 1093 + }, + { + "epoch": 0.4189163316101857, + "grad_norm": 0.2663874328136444, + "learning_rate": 4.798916362410411e-06, + "loss": 0.6421, + "step": 1094 + }, + { + "epoch": 0.41929925330269957, + "grad_norm": 0.28957879543304443, + "learning_rate": 4.798517091003032e-06, + "loss": 0.691, + "step": 1095 + }, + { + "epoch": 0.4196821749952135, + "grad_norm": 0.2755984663963318, + "learning_rate": 4.798117440235723e-06, + "loss": 0.6774, + "step": 1096 + }, + { + "epoch": 0.42006509668772735, + "grad_norm": 0.2807273268699646, + "learning_rate": 4.797717410174445e-06, + "loss": 0.6792, + "step": 1097 + }, + { + "epoch": 0.42044801838024126, + "grad_norm": 0.2845119833946228, + "learning_rate": 4.79731700088522e-06, + "loss": 0.6551, + "step": 1098 + }, + { + "epoch": 0.4208309400727551, + "grad_norm": 0.28206029534339905, + "learning_rate": 4.7969162124341354e-06, + "loss": 0.6965, + "step": 1099 + }, + { + "epoch": 0.421213861765269, + "grad_norm": 0.276937872171402, + "learning_rate": 4.796515044887338e-06, + "loss": 0.6652, + "step": 1100 + }, + { + "epoch": 0.4215967834577829, + "grad_norm": 0.2648085653781891, + "learning_rate": 4.79611349831104e-06, + "loss": 0.6611, + "step": 1101 + }, + { + "epoch": 0.42197970515029676, + "grad_norm": 0.28414374589920044, + "learning_rate": 4.795711572771515e-06, + "loss": 0.6802, + "step": 1102 + }, + { + "epoch": 0.4223626268428106, + "grad_norm": 0.27629321813583374, + "learning_rate": 4.795309268335096e-06, + "loss": 0.6681, + "step": 1103 + }, + { + "epoch": 0.42274554853532453, + "grad_norm": 0.27043506503105164, + "learning_rate": 4.794906585068185e-06, + "loss": 0.6861, + "step": 1104 + }, + { + "epoch": 0.4231284702278384, + "grad_norm": 0.2690626382827759, + "learning_rate": 4.794503523037242e-06, + "loss": 0.6826, + "step": 1105 + }, + { + "epoch": 0.4235113919203523, + "grad_norm": 0.2778787910938263, + "learning_rate": 4.794100082308791e-06, + "loss": 0.6767, + "step": 1106 + }, + { + "epoch": 0.42389431361286617, + "grad_norm": 0.27799922227859497, + "learning_rate": 4.793696262949417e-06, + "loss": 0.6542, + "step": 1107 + }, + { + "epoch": 0.42427723530538003, + "grad_norm": 0.2779891788959503, + "learning_rate": 4.793292065025768e-06, + "loss": 0.6926, + "step": 1108 + }, + { + "epoch": 0.42466015699789394, + "grad_norm": 0.2670944631099701, + "learning_rate": 4.792887488604558e-06, + "loss": 0.6685, + "step": 1109 + }, + { + "epoch": 0.4250430786904078, + "grad_norm": 0.2732831537723541, + "learning_rate": 4.792482533752557e-06, + "loss": 0.6636, + "step": 1110 + }, + { + "epoch": 0.42542600038292167, + "grad_norm": 0.2689459025859833, + "learning_rate": 4.7920772005366035e-06, + "loss": 0.6914, + "step": 1111 + }, + { + "epoch": 0.4258089220754356, + "grad_norm": 0.27064675092697144, + "learning_rate": 4.791671489023595e-06, + "loss": 0.6674, + "step": 1112 + }, + { + "epoch": 0.42619184376794944, + "grad_norm": 0.2776367962360382, + "learning_rate": 4.791265399280493e-06, + "loss": 0.6828, + "step": 1113 + }, + { + "epoch": 0.42657476546046336, + "grad_norm": 0.2685474157333374, + "learning_rate": 4.7908589313743195e-06, + "loss": 0.6682, + "step": 1114 + }, + { + "epoch": 0.4269576871529772, + "grad_norm": 0.2742334306240082, + "learning_rate": 4.7904520853721616e-06, + "loss": 0.6943, + "step": 1115 + }, + { + "epoch": 0.4273406088454911, + "grad_norm": 0.2855285704135895, + "learning_rate": 4.790044861341166e-06, + "loss": 0.6797, + "step": 1116 + }, + { + "epoch": 0.427723530538005, + "grad_norm": 0.2690899074077606, + "learning_rate": 4.789637259348544e-06, + "loss": 0.6776, + "step": 1117 + }, + { + "epoch": 0.42810645223051885, + "grad_norm": 0.2653801143169403, + "learning_rate": 4.789229279461569e-06, + "loss": 0.6502, + "step": 1118 + }, + { + "epoch": 0.4284893739230327, + "grad_norm": 0.2711450755596161, + "learning_rate": 4.788820921747576e-06, + "loss": 0.6854, + "step": 1119 + }, + { + "epoch": 0.42887229561554663, + "grad_norm": 0.27033019065856934, + "learning_rate": 4.788412186273961e-06, + "loss": 0.6678, + "step": 1120 + }, + { + "epoch": 0.4292552173080605, + "grad_norm": 0.26320865750312805, + "learning_rate": 4.788003073108186e-06, + "loss": 0.6555, + "step": 1121 + }, + { + "epoch": 0.4296381390005744, + "grad_norm": 0.2914315462112427, + "learning_rate": 4.787593582317772e-06, + "loss": 0.6776, + "step": 1122 + }, + { + "epoch": 0.43002106069308826, + "grad_norm": 0.269633412361145, + "learning_rate": 4.787183713970304e-06, + "loss": 0.6617, + "step": 1123 + }, + { + "epoch": 0.4304039823856021, + "grad_norm": 0.2707017660140991, + "learning_rate": 4.786773468133431e-06, + "loss": 0.6623, + "step": 1124 + }, + { + "epoch": 0.43078690407811604, + "grad_norm": 0.2661648392677307, + "learning_rate": 4.786362844874859e-06, + "loss": 0.6749, + "step": 1125 + }, + { + "epoch": 0.4311698257706299, + "grad_norm": 0.2712387144565582, + "learning_rate": 4.785951844262361e-06, + "loss": 0.6721, + "step": 1126 + }, + { + "epoch": 0.43155274746314376, + "grad_norm": 0.2739746868610382, + "learning_rate": 4.7855404663637695e-06, + "loss": 0.6413, + "step": 1127 + }, + { + "epoch": 0.4319356691556577, + "grad_norm": 0.27242210507392883, + "learning_rate": 4.785128711246983e-06, + "loss": 0.6718, + "step": 1128 + }, + { + "epoch": 0.43231859084817154, + "grad_norm": 0.2739618718624115, + "learning_rate": 4.7847165789799586e-06, + "loss": 0.6632, + "step": 1129 + }, + { + "epoch": 0.43270151254068545, + "grad_norm": 0.272318571805954, + "learning_rate": 4.784304069630716e-06, + "loss": 0.6604, + "step": 1130 + }, + { + "epoch": 0.4330844342331993, + "grad_norm": 0.2719273865222931, + "learning_rate": 4.78389118326734e-06, + "loss": 0.6635, + "step": 1131 + }, + { + "epoch": 0.43346735592571317, + "grad_norm": 0.28921982645988464, + "learning_rate": 4.783477919957973e-06, + "loss": 0.6716, + "step": 1132 + }, + { + "epoch": 0.4338502776182271, + "grad_norm": 0.26451462507247925, + "learning_rate": 4.783064279770825e-06, + "loss": 0.6699, + "step": 1133 + }, + { + "epoch": 0.43423319931074095, + "grad_norm": 0.2656538784503937, + "learning_rate": 4.782650262774164e-06, + "loss": 0.6748, + "step": 1134 + }, + { + "epoch": 0.4346161210032548, + "grad_norm": 0.27099308371543884, + "learning_rate": 4.782235869036321e-06, + "loss": 0.6716, + "step": 1135 + }, + { + "epoch": 0.4349990426957687, + "grad_norm": 0.28783848881721497, + "learning_rate": 4.781821098625691e-06, + "loss": 0.673, + "step": 1136 + }, + { + "epoch": 0.4353819643882826, + "grad_norm": 0.27346324920654297, + "learning_rate": 4.78140595161073e-06, + "loss": 0.6527, + "step": 1137 + }, + { + "epoch": 0.4357648860807965, + "grad_norm": 0.27585262060165405, + "learning_rate": 4.7809904280599564e-06, + "loss": 0.6528, + "step": 1138 + }, + { + "epoch": 0.43614780777331036, + "grad_norm": 0.28196394443511963, + "learning_rate": 4.780574528041949e-06, + "loss": 0.6911, + "step": 1139 + }, + { + "epoch": 0.4365307294658242, + "grad_norm": 0.27955520153045654, + "learning_rate": 4.780158251625351e-06, + "loss": 0.6759, + "step": 1140 + }, + { + "epoch": 0.43691365115833813, + "grad_norm": 0.2854354679584503, + "learning_rate": 4.7797415988788675e-06, + "loss": 0.6759, + "step": 1141 + }, + { + "epoch": 0.437296572850852, + "grad_norm": 0.2818031311035156, + "learning_rate": 4.7793245698712655e-06, + "loss": 0.677, + "step": 1142 + }, + { + "epoch": 0.43767949454336585, + "grad_norm": 0.2737523913383484, + "learning_rate": 4.778907164671372e-06, + "loss": 0.7044, + "step": 1143 + }, + { + "epoch": 0.43806241623587977, + "grad_norm": 0.27465736865997314, + "learning_rate": 4.778489383348079e-06, + "loss": 0.6752, + "step": 1144 + }, + { + "epoch": 0.43844533792839363, + "grad_norm": 0.27898678183555603, + "learning_rate": 4.77807122597034e-06, + "loss": 0.6794, + "step": 1145 + }, + { + "epoch": 0.43882825962090755, + "grad_norm": 0.27347368001937866, + "learning_rate": 4.777652692607168e-06, + "loss": 0.6559, + "step": 1146 + }, + { + "epoch": 0.4392111813134214, + "grad_norm": 0.2761583924293518, + "learning_rate": 4.7772337833276425e-06, + "loss": 0.679, + "step": 1147 + }, + { + "epoch": 0.43959410300593527, + "grad_norm": 0.285239577293396, + "learning_rate": 4.776814498200901e-06, + "loss": 0.6658, + "step": 1148 + }, + { + "epoch": 0.4399770246984492, + "grad_norm": 0.2704262435436249, + "learning_rate": 4.776394837296145e-06, + "loss": 0.6773, + "step": 1149 + }, + { + "epoch": 0.44035994639096304, + "grad_norm": 0.2713160514831543, + "learning_rate": 4.775974800682639e-06, + "loss": 0.6722, + "step": 1150 + }, + { + "epoch": 0.4407428680834769, + "grad_norm": 0.2755696177482605, + "learning_rate": 4.775554388429706e-06, + "loss": 0.6703, + "step": 1151 + }, + { + "epoch": 0.4411257897759908, + "grad_norm": 0.2795194685459137, + "learning_rate": 4.775133600606734e-06, + "loss": 0.6668, + "step": 1152 + }, + { + "epoch": 0.4415087114685047, + "grad_norm": 0.2720242440700531, + "learning_rate": 4.774712437283172e-06, + "loss": 0.6758, + "step": 1153 + }, + { + "epoch": 0.4418916331610186, + "grad_norm": 0.27648648619651794, + "learning_rate": 4.774290898528532e-06, + "loss": 0.6724, + "step": 1154 + }, + { + "epoch": 0.44227455485353245, + "grad_norm": 0.29117751121520996, + "learning_rate": 4.773868984412386e-06, + "loss": 0.6599, + "step": 1155 + }, + { + "epoch": 0.4426574765460463, + "grad_norm": 0.2748560607433319, + "learning_rate": 4.7734466950043686e-06, + "loss": 0.6716, + "step": 1156 + }, + { + "epoch": 0.44304039823856023, + "grad_norm": 0.29462385177612305, + "learning_rate": 4.773024030374179e-06, + "loss": 0.6465, + "step": 1157 + }, + { + "epoch": 0.4434233199310741, + "grad_norm": 0.27808573842048645, + "learning_rate": 4.772600990591574e-06, + "loss": 0.6567, + "step": 1158 + }, + { + "epoch": 0.44380624162358795, + "grad_norm": 0.2768598198890686, + "learning_rate": 4.772177575726375e-06, + "loss": 0.6814, + "step": 1159 + }, + { + "epoch": 0.44418916331610186, + "grad_norm": 0.2841338813304901, + "learning_rate": 4.7717537858484654e-06, + "loss": 0.666, + "step": 1160 + }, + { + "epoch": 0.4445720850086157, + "grad_norm": 0.2649598717689514, + "learning_rate": 4.77132962102779e-06, + "loss": 0.6514, + "step": 1161 + }, + { + "epoch": 0.44495500670112964, + "grad_norm": 0.27885758876800537, + "learning_rate": 4.770905081334352e-06, + "loss": 0.6653, + "step": 1162 + }, + { + "epoch": 0.4453379283936435, + "grad_norm": 0.2814213037490845, + "learning_rate": 4.770480166838224e-06, + "loss": 0.6548, + "step": 1163 + }, + { + "epoch": 0.44572085008615736, + "grad_norm": 0.2747483551502228, + "learning_rate": 4.770054877609534e-06, + "loss": 0.6671, + "step": 1164 + }, + { + "epoch": 0.4461037717786713, + "grad_norm": 0.26971474289894104, + "learning_rate": 4.769629213718476e-06, + "loss": 0.6826, + "step": 1165 + }, + { + "epoch": 0.44648669347118514, + "grad_norm": 0.27830854058265686, + "learning_rate": 4.769203175235302e-06, + "loss": 0.6828, + "step": 1166 + }, + { + "epoch": 0.446869615163699, + "grad_norm": 0.27223125100135803, + "learning_rate": 4.768776762230328e-06, + "loss": 0.6711, + "step": 1167 + }, + { + "epoch": 0.4472525368562129, + "grad_norm": 0.2749093174934387, + "learning_rate": 4.768349974773931e-06, + "loss": 0.6877, + "step": 1168 + }, + { + "epoch": 0.4476354585487268, + "grad_norm": 0.2665577530860901, + "learning_rate": 4.767922812936552e-06, + "loss": 0.6861, + "step": 1169 + }, + { + "epoch": 0.4480183802412407, + "grad_norm": 0.2772512137889862, + "learning_rate": 4.767495276788692e-06, + "loss": 0.6654, + "step": 1170 + }, + { + "epoch": 0.44840130193375455, + "grad_norm": 0.2787632942199707, + "learning_rate": 4.767067366400913e-06, + "loss": 0.6702, + "step": 1171 + }, + { + "epoch": 0.4487842236262684, + "grad_norm": 0.2907279431819916, + "learning_rate": 4.766639081843839e-06, + "loss": 0.6818, + "step": 1172 + }, + { + "epoch": 0.4491671453187823, + "grad_norm": 0.27912425994873047, + "learning_rate": 4.766210423188158e-06, + "loss": 0.6535, + "step": 1173 + }, + { + "epoch": 0.4495500670112962, + "grad_norm": 0.29031383991241455, + "learning_rate": 4.765781390504617e-06, + "loss": 0.6915, + "step": 1174 + }, + { + "epoch": 0.44993298870381004, + "grad_norm": 0.2793537676334381, + "learning_rate": 4.765351983864027e-06, + "loss": 0.6703, + "step": 1175 + }, + { + "epoch": 0.45031591039632396, + "grad_norm": 0.28281429409980774, + "learning_rate": 4.76492220333726e-06, + "loss": 0.6725, + "step": 1176 + }, + { + "epoch": 0.4506988320888378, + "grad_norm": 0.2810090482234955, + "learning_rate": 4.7644920489952474e-06, + "loss": 0.6859, + "step": 1177 + }, + { + "epoch": 0.45108175378135174, + "grad_norm": 0.28927022218704224, + "learning_rate": 4.764061520908986e-06, + "loss": 0.6828, + "step": 1178 + }, + { + "epoch": 0.4514646754738656, + "grad_norm": 0.29566314816474915, + "learning_rate": 4.763630619149531e-06, + "loss": 0.6731, + "step": 1179 + }, + { + "epoch": 0.45184759716637946, + "grad_norm": 0.2778233289718628, + "learning_rate": 4.763199343788002e-06, + "loss": 0.6833, + "step": 1180 + }, + { + "epoch": 0.45223051885889337, + "grad_norm": 0.27039292454719543, + "learning_rate": 4.762767694895579e-06, + "loss": 0.661, + "step": 1181 + }, + { + "epoch": 0.45261344055140723, + "grad_norm": 0.2723767161369324, + "learning_rate": 4.762335672543503e-06, + "loss": 0.6646, + "step": 1182 + }, + { + "epoch": 0.4529963622439211, + "grad_norm": 0.2879144847393036, + "learning_rate": 4.761903276803078e-06, + "loss": 0.6699, + "step": 1183 + }, + { + "epoch": 0.453379283936435, + "grad_norm": 0.2802518904209137, + "learning_rate": 4.761470507745669e-06, + "loss": 0.6644, + "step": 1184 + }, + { + "epoch": 0.45376220562894887, + "grad_norm": 0.2758473753929138, + "learning_rate": 4.761037365442703e-06, + "loss": 0.6555, + "step": 1185 + }, + { + "epoch": 0.4541451273214628, + "grad_norm": 0.282187819480896, + "learning_rate": 4.760603849965666e-06, + "loss": 0.6723, + "step": 1186 + }, + { + "epoch": 0.45452804901397664, + "grad_norm": 0.275897741317749, + "learning_rate": 4.760169961386111e-06, + "loss": 0.6669, + "step": 1187 + }, + { + "epoch": 0.4549109707064905, + "grad_norm": 0.2768770754337311, + "learning_rate": 4.759735699775647e-06, + "loss": 0.6624, + "step": 1188 + }, + { + "epoch": 0.4552938923990044, + "grad_norm": 0.26994645595550537, + "learning_rate": 4.759301065205947e-06, + "loss": 0.667, + "step": 1189 + }, + { + "epoch": 0.4556768140915183, + "grad_norm": 0.27651163935661316, + "learning_rate": 4.758866057748747e-06, + "loss": 0.6694, + "step": 1190 + }, + { + "epoch": 0.45605973578403214, + "grad_norm": 0.2905432879924774, + "learning_rate": 4.758430677475841e-06, + "loss": 0.6865, + "step": 1191 + }, + { + "epoch": 0.45644265747654605, + "grad_norm": 0.27474573254585266, + "learning_rate": 4.757994924459089e-06, + "loss": 0.6821, + "step": 1192 + }, + { + "epoch": 0.4568255791690599, + "grad_norm": 0.27210354804992676, + "learning_rate": 4.7575587987704085e-06, + "loss": 0.6592, + "step": 1193 + }, + { + "epoch": 0.45720850086157383, + "grad_norm": 0.26745250821113586, + "learning_rate": 4.75712230048178e-06, + "loss": 0.6641, + "step": 1194 + }, + { + "epoch": 0.4575914225540877, + "grad_norm": 0.271131306886673, + "learning_rate": 4.756685429665246e-06, + "loss": 0.6684, + "step": 1195 + }, + { + "epoch": 0.45797434424660155, + "grad_norm": 0.2732897102832794, + "learning_rate": 4.75624818639291e-06, + "loss": 0.6835, + "step": 1196 + }, + { + "epoch": 0.45835726593911547, + "grad_norm": 0.2688729465007782, + "learning_rate": 4.755810570736938e-06, + "loss": 0.6834, + "step": 1197 + }, + { + "epoch": 0.4587401876316293, + "grad_norm": 0.2814828157424927, + "learning_rate": 4.755372582769554e-06, + "loss": 0.6643, + "step": 1198 + }, + { + "epoch": 0.4591231093241432, + "grad_norm": 0.2694145441055298, + "learning_rate": 4.754934222563048e-06, + "loss": 0.6753, + "step": 1199 + }, + { + "epoch": 0.4595060310166571, + "grad_norm": 0.2745696008205414, + "learning_rate": 4.754495490189769e-06, + "loss": 0.6885, + "step": 1200 + }, + { + "epoch": 0.45988895270917096, + "grad_norm": 0.2827391028404236, + "learning_rate": 4.754056385722128e-06, + "loss": 0.6745, + "step": 1201 + }, + { + "epoch": 0.4602718744016849, + "grad_norm": 0.2724725306034088, + "learning_rate": 4.753616909232597e-06, + "loss": 0.6589, + "step": 1202 + }, + { + "epoch": 0.46065479609419874, + "grad_norm": 0.2880653142929077, + "learning_rate": 4.753177060793709e-06, + "loss": 0.6796, + "step": 1203 + }, + { + "epoch": 0.4610377177867126, + "grad_norm": 0.27427205443382263, + "learning_rate": 4.75273684047806e-06, + "loss": 0.6961, + "step": 1204 + }, + { + "epoch": 0.4614206394792265, + "grad_norm": 0.2762588858604431, + "learning_rate": 4.752296248358306e-06, + "loss": 0.6542, + "step": 1205 + }, + { + "epoch": 0.4618035611717404, + "grad_norm": 0.2718510925769806, + "learning_rate": 4.7518552845071646e-06, + "loss": 0.676, + "step": 1206 + }, + { + "epoch": 0.46218648286425423, + "grad_norm": 0.27809596061706543, + "learning_rate": 4.7514139489974145e-06, + "loss": 0.6526, + "step": 1207 + }, + { + "epoch": 0.46256940455676815, + "grad_norm": 0.2677210867404938, + "learning_rate": 4.750972241901897e-06, + "loss": 0.6587, + "step": 1208 + }, + { + "epoch": 0.462952326249282, + "grad_norm": 0.2862596809864044, + "learning_rate": 4.750530163293514e-06, + "loss": 0.6818, + "step": 1209 + }, + { + "epoch": 0.4633352479417959, + "grad_norm": 0.27850595116615295, + "learning_rate": 4.750087713245227e-06, + "loss": 0.6662, + "step": 1210 + }, + { + "epoch": 0.4637181696343098, + "grad_norm": 0.27618399262428284, + "learning_rate": 4.749644891830063e-06, + "loss": 0.6614, + "step": 1211 + }, + { + "epoch": 0.46410109132682364, + "grad_norm": 0.2669837176799774, + "learning_rate": 4.749201699121105e-06, + "loss": 0.6749, + "step": 1212 + }, + { + "epoch": 0.46448401301933756, + "grad_norm": 0.27776703238487244, + "learning_rate": 4.748758135191502e-06, + "loss": 0.6695, + "step": 1213 + }, + { + "epoch": 0.4648669347118514, + "grad_norm": 0.26955899596214294, + "learning_rate": 4.74831420011446e-06, + "loss": 0.6766, + "step": 1214 + }, + { + "epoch": 0.4652498564043653, + "grad_norm": 0.2800334393978119, + "learning_rate": 4.747869893963252e-06, + "loss": 0.6707, + "step": 1215 + }, + { + "epoch": 0.4656327780968792, + "grad_norm": 0.3023186922073364, + "learning_rate": 4.747425216811205e-06, + "loss": 0.669, + "step": 1216 + }, + { + "epoch": 0.46601569978939306, + "grad_norm": 0.2796546518802643, + "learning_rate": 4.746980168731713e-06, + "loss": 0.6911, + "step": 1217 + }, + { + "epoch": 0.46639862148190697, + "grad_norm": 0.26911649107933044, + "learning_rate": 4.7465347497982286e-06, + "loss": 0.6709, + "step": 1218 + }, + { + "epoch": 0.46678154317442083, + "grad_norm": 0.2805267870426178, + "learning_rate": 4.746088960084267e-06, + "loss": 0.6772, + "step": 1219 + }, + { + "epoch": 0.4671644648669347, + "grad_norm": 0.26144716143608093, + "learning_rate": 4.745642799663403e-06, + "loss": 0.6947, + "step": 1220 + }, + { + "epoch": 0.4675473865594486, + "grad_norm": 0.28579649329185486, + "learning_rate": 4.745196268609274e-06, + "loss": 0.66, + "step": 1221 + }, + { + "epoch": 0.46793030825196247, + "grad_norm": 0.27812519669532776, + "learning_rate": 4.744749366995576e-06, + "loss": 0.6628, + "step": 1222 + }, + { + "epoch": 0.46831322994447633, + "grad_norm": 0.26414886116981506, + "learning_rate": 4.744302094896071e-06, + "loss": 0.653, + "step": 1223 + }, + { + "epoch": 0.46869615163699024, + "grad_norm": 0.2737218737602234, + "learning_rate": 4.743854452384577e-06, + "loss": 0.6817, + "step": 1224 + }, + { + "epoch": 0.4690790733295041, + "grad_norm": 0.2919729948043823, + "learning_rate": 4.743406439534976e-06, + "loss": 0.6666, + "step": 1225 + }, + { + "epoch": 0.469461995022018, + "grad_norm": 0.2719593644142151, + "learning_rate": 4.742958056421211e-06, + "loss": 0.674, + "step": 1226 + }, + { + "epoch": 0.4698449167145319, + "grad_norm": 0.27731117606163025, + "learning_rate": 4.742509303117286e-06, + "loss": 0.6695, + "step": 1227 + }, + { + "epoch": 0.47022783840704574, + "grad_norm": 0.2854015529155731, + "learning_rate": 4.742060179697264e-06, + "loss": 0.6684, + "step": 1228 + }, + { + "epoch": 0.47061076009955966, + "grad_norm": 0.275790810585022, + "learning_rate": 4.741610686235273e-06, + "loss": 0.673, + "step": 1229 + }, + { + "epoch": 0.4709936817920735, + "grad_norm": 0.27415552735328674, + "learning_rate": 4.741160822805499e-06, + "loss": 0.6658, + "step": 1230 + }, + { + "epoch": 0.4713766034845874, + "grad_norm": 0.2823668122291565, + "learning_rate": 4.740710589482189e-06, + "loss": 0.681, + "step": 1231 + }, + { + "epoch": 0.4717595251771013, + "grad_norm": 0.27480632066726685, + "learning_rate": 4.740259986339652e-06, + "loss": 0.6796, + "step": 1232 + }, + { + "epoch": 0.47214244686961515, + "grad_norm": 0.28374215960502625, + "learning_rate": 4.73980901345226e-06, + "loss": 0.6785, + "step": 1233 + }, + { + "epoch": 0.47252536856212907, + "grad_norm": 0.28118008375167847, + "learning_rate": 4.739357670894442e-06, + "loss": 0.6686, + "step": 1234 + }, + { + "epoch": 0.4729082902546429, + "grad_norm": 0.28157249093055725, + "learning_rate": 4.738905958740692e-06, + "loss": 0.6779, + "step": 1235 + }, + { + "epoch": 0.4732912119471568, + "grad_norm": 0.2744462192058563, + "learning_rate": 4.738453877065561e-06, + "loss": 0.6666, + "step": 1236 + }, + { + "epoch": 0.4736741336396707, + "grad_norm": 0.28920602798461914, + "learning_rate": 4.738001425943664e-06, + "loss": 0.7009, + "step": 1237 + }, + { + "epoch": 0.47405705533218456, + "grad_norm": 0.261415034532547, + "learning_rate": 4.737548605449675e-06, + "loss": 0.6634, + "step": 1238 + }, + { + "epoch": 0.4744399770246984, + "grad_norm": 0.2748316526412964, + "learning_rate": 4.7370954156583326e-06, + "loss": 0.6804, + "step": 1239 + }, + { + "epoch": 0.47482289871721234, + "grad_norm": 0.2750059962272644, + "learning_rate": 4.736641856644431e-06, + "loss": 0.6934, + "step": 1240 + }, + { + "epoch": 0.4752058204097262, + "grad_norm": 0.28661346435546875, + "learning_rate": 4.73618792848283e-06, + "loss": 0.698, + "step": 1241 + }, + { + "epoch": 0.4755887421022401, + "grad_norm": 0.2804570496082306, + "learning_rate": 4.7357336312484466e-06, + "loss": 0.6702, + "step": 1242 + }, + { + "epoch": 0.475971663794754, + "grad_norm": 0.2865023612976074, + "learning_rate": 4.735278965016262e-06, + "loss": 0.6707, + "step": 1243 + }, + { + "epoch": 0.47635458548726783, + "grad_norm": 0.30556708574295044, + "learning_rate": 4.734823929861317e-06, + "loss": 0.6755, + "step": 1244 + }, + { + "epoch": 0.47673750717978175, + "grad_norm": 0.28410694003105164, + "learning_rate": 4.734368525858713e-06, + "loss": 0.6764, + "step": 1245 + }, + { + "epoch": 0.4771204288722956, + "grad_norm": 0.2751966416835785, + "learning_rate": 4.7339127530836095e-06, + "loss": 0.6598, + "step": 1246 + }, + { + "epoch": 0.47750335056480947, + "grad_norm": 0.2905697226524353, + "learning_rate": 4.733456611611233e-06, + "loss": 0.6876, + "step": 1247 + }, + { + "epoch": 0.4778862722573234, + "grad_norm": 0.30000394582748413, + "learning_rate": 4.733000101516868e-06, + "loss": 0.6747, + "step": 1248 + }, + { + "epoch": 0.47826919394983725, + "grad_norm": 0.26689612865448, + "learning_rate": 4.732543222875857e-06, + "loss": 0.6729, + "step": 1249 + }, + { + "epoch": 0.47865211564235116, + "grad_norm": 0.2899976372718811, + "learning_rate": 4.7320859757636074e-06, + "loss": 0.6758, + "step": 1250 + }, + { + "epoch": 0.479035037334865, + "grad_norm": 0.3092990815639496, + "learning_rate": 4.731628360255585e-06, + "loss": 0.6826, + "step": 1251 + }, + { + "epoch": 0.4794179590273789, + "grad_norm": 0.2779189348220825, + "learning_rate": 4.731170376427317e-06, + "loss": 0.667, + "step": 1252 + }, + { + "epoch": 0.4798008807198928, + "grad_norm": 0.29295408725738525, + "learning_rate": 4.7307120243543925e-06, + "loss": 0.6593, + "step": 1253 + }, + { + "epoch": 0.48018380241240666, + "grad_norm": 0.29263874888420105, + "learning_rate": 4.73025330411246e-06, + "loss": 0.6545, + "step": 1254 + }, + { + "epoch": 0.4805667241049205, + "grad_norm": 0.26484522223472595, + "learning_rate": 4.729794215777229e-06, + "loss": 0.6459, + "step": 1255 + }, + { + "epoch": 0.48094964579743443, + "grad_norm": 0.277542382478714, + "learning_rate": 4.729334759424471e-06, + "loss": 0.6722, + "step": 1256 + }, + { + "epoch": 0.4813325674899483, + "grad_norm": 0.31448233127593994, + "learning_rate": 4.728874935130015e-06, + "loss": 0.689, + "step": 1257 + }, + { + "epoch": 0.4817154891824622, + "grad_norm": 0.2727276384830475, + "learning_rate": 4.728414742969756e-06, + "loss": 0.6703, + "step": 1258 + }, + { + "epoch": 0.48209841087497607, + "grad_norm": 0.29821106791496277, + "learning_rate": 4.727954183019644e-06, + "loss": 0.6801, + "step": 1259 + }, + { + "epoch": 0.48248133256748993, + "grad_norm": 0.28618431091308594, + "learning_rate": 4.727493255355693e-06, + "loss": 0.6812, + "step": 1260 + }, + { + "epoch": 0.48286425426000384, + "grad_norm": 0.26692572236061096, + "learning_rate": 4.727031960053978e-06, + "loss": 0.6776, + "step": 1261 + }, + { + "epoch": 0.4832471759525177, + "grad_norm": 0.27408280968666077, + "learning_rate": 4.726570297190632e-06, + "loss": 0.6467, + "step": 1262 + }, + { + "epoch": 0.48363009764503156, + "grad_norm": 0.3020334839820862, + "learning_rate": 4.726108266841854e-06, + "loss": 0.657, + "step": 1263 + }, + { + "epoch": 0.4840130193375455, + "grad_norm": 0.2789212465286255, + "learning_rate": 4.725645869083895e-06, + "loss": 0.6625, + "step": 1264 + }, + { + "epoch": 0.48439594103005934, + "grad_norm": 0.27989277243614197, + "learning_rate": 4.725183103993075e-06, + "loss": 0.6611, + "step": 1265 + }, + { + "epoch": 0.48477886272257326, + "grad_norm": 0.2788684368133545, + "learning_rate": 4.724719971645771e-06, + "loss": 0.6794, + "step": 1266 + }, + { + "epoch": 0.4851617844150871, + "grad_norm": 0.2791091203689575, + "learning_rate": 4.724256472118419e-06, + "loss": 0.6688, + "step": 1267 + }, + { + "epoch": 0.485544706107601, + "grad_norm": 0.27072206139564514, + "learning_rate": 4.7237926054875206e-06, + "loss": 0.6753, + "step": 1268 + }, + { + "epoch": 0.4859276278001149, + "grad_norm": 0.27124497294425964, + "learning_rate": 4.723328371829632e-06, + "loss": 0.6501, + "step": 1269 + }, + { + "epoch": 0.48631054949262875, + "grad_norm": 0.27444323897361755, + "learning_rate": 4.7228637712213756e-06, + "loss": 0.6664, + "step": 1270 + }, + { + "epoch": 0.4866934711851426, + "grad_norm": 0.2776550054550171, + "learning_rate": 4.722398803739428e-06, + "loss": 0.6751, + "step": 1271 + }, + { + "epoch": 0.4870763928776565, + "grad_norm": 0.27999234199523926, + "learning_rate": 4.721933469460534e-06, + "loss": 0.6773, + "step": 1272 + }, + { + "epoch": 0.4874593145701704, + "grad_norm": 0.28608107566833496, + "learning_rate": 4.721467768461493e-06, + "loss": 0.6839, + "step": 1273 + }, + { + "epoch": 0.4878422362626843, + "grad_norm": 0.2772696614265442, + "learning_rate": 4.721001700819166e-06, + "loss": 0.6797, + "step": 1274 + }, + { + "epoch": 0.48822515795519816, + "grad_norm": 0.27554333209991455, + "learning_rate": 4.720535266610477e-06, + "loss": 0.7028, + "step": 1275 + }, + { + "epoch": 0.488608079647712, + "grad_norm": 0.2786296606063843, + "learning_rate": 4.720068465912407e-06, + "loss": 0.672, + "step": 1276 + }, + { + "epoch": 0.48899100134022594, + "grad_norm": 0.28190577030181885, + "learning_rate": 4.719601298802002e-06, + "loss": 0.6731, + "step": 1277 + }, + { + "epoch": 0.4893739230327398, + "grad_norm": 0.27765437960624695, + "learning_rate": 4.7191337653563635e-06, + "loss": 0.6686, + "step": 1278 + }, + { + "epoch": 0.48975684472525366, + "grad_norm": 0.28349748253822327, + "learning_rate": 4.718665865652657e-06, + "loss": 0.6928, + "step": 1279 + }, + { + "epoch": 0.4901397664177676, + "grad_norm": 0.2807316482067108, + "learning_rate": 4.7181975997681075e-06, + "loss": 0.6791, + "step": 1280 + }, + { + "epoch": 0.49052268811028144, + "grad_norm": 0.27356651425361633, + "learning_rate": 4.717728967779999e-06, + "loss": 0.6482, + "step": 1281 + }, + { + "epoch": 0.49090560980279535, + "grad_norm": 0.2867351472377777, + "learning_rate": 4.717259969765678e-06, + "loss": 0.6438, + "step": 1282 + }, + { + "epoch": 0.4912885314953092, + "grad_norm": 0.2828361392021179, + "learning_rate": 4.716790605802549e-06, + "loss": 0.6748, + "step": 1283 + }, + { + "epoch": 0.49167145318782307, + "grad_norm": 0.28004953265190125, + "learning_rate": 4.716320875968081e-06, + "loss": 0.6665, + "step": 1284 + }, + { + "epoch": 0.492054374880337, + "grad_norm": 0.27554139494895935, + "learning_rate": 4.715850780339799e-06, + "loss": 0.6589, + "step": 1285 + }, + { + "epoch": 0.49243729657285085, + "grad_norm": 0.2844911217689514, + "learning_rate": 4.7153803189952906e-06, + "loss": 0.6761, + "step": 1286 + }, + { + "epoch": 0.4928202182653647, + "grad_norm": 0.2917320132255554, + "learning_rate": 4.714909492012204e-06, + "loss": 0.6466, + "step": 1287 + }, + { + "epoch": 0.4932031399578786, + "grad_norm": 0.28606316447257996, + "learning_rate": 4.714438299468245e-06, + "loss": 0.6543, + "step": 1288 + }, + { + "epoch": 0.4935860616503925, + "grad_norm": 0.27791330218315125, + "learning_rate": 4.713966741441185e-06, + "loss": 0.6776, + "step": 1289 + }, + { + "epoch": 0.4939689833429064, + "grad_norm": 0.279191792011261, + "learning_rate": 4.713494818008849e-06, + "loss": 0.6945, + "step": 1290 + }, + { + "epoch": 0.49435190503542026, + "grad_norm": 0.28241732716560364, + "learning_rate": 4.71302252924913e-06, + "loss": 0.6591, + "step": 1291 + }, + { + "epoch": 0.4947348267279341, + "grad_norm": 0.2747727334499359, + "learning_rate": 4.712549875239972e-06, + "loss": 0.6614, + "step": 1292 + }, + { + "epoch": 0.49511774842044803, + "grad_norm": 0.2691855728626251, + "learning_rate": 4.712076856059389e-06, + "loss": 0.6235, + "step": 1293 + }, + { + "epoch": 0.4955006701129619, + "grad_norm": 0.2748763859272003, + "learning_rate": 4.711603471785449e-06, + "loss": 0.6372, + "step": 1294 + }, + { + "epoch": 0.49588359180547575, + "grad_norm": 0.28962263464927673, + "learning_rate": 4.7111297224962805e-06, + "loss": 0.648, + "step": 1295 + }, + { + "epoch": 0.49626651349798967, + "grad_norm": 0.27451568841934204, + "learning_rate": 4.710655608270076e-06, + "loss": 0.6611, + "step": 1296 + }, + { + "epoch": 0.49664943519050353, + "grad_norm": 0.28200188279151917, + "learning_rate": 4.710181129185085e-06, + "loss": 0.6583, + "step": 1297 + }, + { + "epoch": 0.49703235688301745, + "grad_norm": 0.26564377546310425, + "learning_rate": 4.7097062853196175e-06, + "loss": 0.6768, + "step": 1298 + }, + { + "epoch": 0.4974152785755313, + "grad_norm": 0.28046175837516785, + "learning_rate": 4.709231076752045e-06, + "loss": 0.6771, + "step": 1299 + }, + { + "epoch": 0.49779820026804517, + "grad_norm": 0.27810144424438477, + "learning_rate": 4.708755503560799e-06, + "loss": 0.6817, + "step": 1300 + }, + { + "epoch": 0.4981811219605591, + "grad_norm": 0.28252771496772766, + "learning_rate": 4.70827956582437e-06, + "loss": 0.6763, + "step": 1301 + }, + { + "epoch": 0.49856404365307294, + "grad_norm": 0.2781440019607544, + "learning_rate": 4.70780326362131e-06, + "loss": 0.6608, + "step": 1302 + }, + { + "epoch": 0.4989469653455868, + "grad_norm": 0.2670237123966217, + "learning_rate": 4.707326597030229e-06, + "loss": 0.6565, + "step": 1303 + }, + { + "epoch": 0.4993298870381007, + "grad_norm": 0.2824726104736328, + "learning_rate": 4.706849566129801e-06, + "loss": 0.6763, + "step": 1304 + }, + { + "epoch": 0.4997128087306146, + "grad_norm": 0.28362149000167847, + "learning_rate": 4.706372170998756e-06, + "loss": 0.668, + "step": 1305 + }, + { + "epoch": 0.5000957304231285, + "grad_norm": 0.27518826723098755, + "learning_rate": 4.705894411715887e-06, + "loss": 0.677, + "step": 1306 + }, + { + "epoch": 0.5004786521156424, + "grad_norm": 0.2708577513694763, + "learning_rate": 4.705416288360045e-06, + "loss": 0.6512, + "step": 1307 + }, + { + "epoch": 0.5008615738081562, + "grad_norm": 0.27491506934165955, + "learning_rate": 4.704937801010144e-06, + "loss": 0.6766, + "step": 1308 + }, + { + "epoch": 0.5012444955006701, + "grad_norm": 0.3019718527793884, + "learning_rate": 4.704458949745154e-06, + "loss": 0.6728, + "step": 1309 + }, + { + "epoch": 0.501627417193184, + "grad_norm": 0.2824283838272095, + "learning_rate": 4.703979734644107e-06, + "loss": 0.6695, + "step": 1310 + }, + { + "epoch": 0.5020103388856979, + "grad_norm": 0.2763907015323639, + "learning_rate": 4.703500155786098e-06, + "loss": 0.6738, + "step": 1311 + }, + { + "epoch": 0.5023932605782118, + "grad_norm": 0.27073514461517334, + "learning_rate": 4.703020213250278e-06, + "loss": 0.6664, + "step": 1312 + }, + { + "epoch": 0.5027761822707256, + "grad_norm": 0.2841322720050812, + "learning_rate": 4.7025399071158575e-06, + "loss": 0.6603, + "step": 1313 + }, + { + "epoch": 0.5031591039632395, + "grad_norm": 0.28222954273223877, + "learning_rate": 4.702059237462112e-06, + "loss": 0.6768, + "step": 1314 + }, + { + "epoch": 0.5035420256557533, + "grad_norm": 0.28418296575546265, + "learning_rate": 4.701578204368371e-06, + "loss": 0.677, + "step": 1315 + }, + { + "epoch": 0.5039249473482673, + "grad_norm": 0.2761109173297882, + "learning_rate": 4.70109680791403e-06, + "loss": 0.6829, + "step": 1316 + }, + { + "epoch": 0.5043078690407812, + "grad_norm": 0.2744296193122864, + "learning_rate": 4.700615048178538e-06, + "loss": 0.6865, + "step": 1317 + }, + { + "epoch": 0.504690790733295, + "grad_norm": 0.2845577299594879, + "learning_rate": 4.7001329252414095e-06, + "loss": 0.6632, + "step": 1318 + }, + { + "epoch": 0.5050737124258089, + "grad_norm": 0.28562062978744507, + "learning_rate": 4.699650439182216e-06, + "loss": 0.6737, + "step": 1319 + }, + { + "epoch": 0.5054566341183228, + "grad_norm": 0.28627490997314453, + "learning_rate": 4.69916759008059e-06, + "loss": 0.655, + "step": 1320 + }, + { + "epoch": 0.5058395558108367, + "grad_norm": 0.27140292525291443, + "learning_rate": 4.698684378016223e-06, + "loss": 0.6518, + "step": 1321 + }, + { + "epoch": 0.5062224775033506, + "grad_norm": 0.2792279124259949, + "learning_rate": 4.698200803068868e-06, + "loss": 0.6855, + "step": 1322 + }, + { + "epoch": 0.5066053991958644, + "grad_norm": 0.2705386281013489, + "learning_rate": 4.697716865318335e-06, + "loss": 0.6755, + "step": 1323 + }, + { + "epoch": 0.5069883208883783, + "grad_norm": 0.2843645215034485, + "learning_rate": 4.6972325648444974e-06, + "loss": 0.6692, + "step": 1324 + }, + { + "epoch": 0.5073712425808922, + "grad_norm": 0.28340429067611694, + "learning_rate": 4.6967479017272865e-06, + "loss": 0.6628, + "step": 1325 + }, + { + "epoch": 0.5077541642734061, + "grad_norm": 0.2885495722293854, + "learning_rate": 4.696262876046693e-06, + "loss": 0.6676, + "step": 1326 + }, + { + "epoch": 0.50813708596592, + "grad_norm": 0.2888467609882355, + "learning_rate": 4.6957774878827694e-06, + "loss": 0.678, + "step": 1327 + }, + { + "epoch": 0.5085200076584339, + "grad_norm": 0.28116628527641296, + "learning_rate": 4.695291737315626e-06, + "loss": 0.6849, + "step": 1328 + }, + { + "epoch": 0.5089029293509477, + "grad_norm": 0.29132142663002014, + "learning_rate": 4.694805624425435e-06, + "loss": 0.6769, + "step": 1329 + }, + { + "epoch": 0.5092858510434616, + "grad_norm": 0.2741115391254425, + "learning_rate": 4.6943191492924246e-06, + "loss": 0.6757, + "step": 1330 + }, + { + "epoch": 0.5096687727359754, + "grad_norm": 0.28962674736976624, + "learning_rate": 4.693832311996887e-06, + "loss": 0.6501, + "step": 1331 + }, + { + "epoch": 0.5100516944284894, + "grad_norm": 0.2865285873413086, + "learning_rate": 4.693345112619173e-06, + "loss": 0.6683, + "step": 1332 + }, + { + "epoch": 0.5104346161210033, + "grad_norm": 0.2814137935638428, + "learning_rate": 4.692857551239692e-06, + "loss": 0.6767, + "step": 1333 + }, + { + "epoch": 0.5108175378135171, + "grad_norm": 0.2837572991847992, + "learning_rate": 4.692369627938914e-06, + "loss": 0.6554, + "step": 1334 + }, + { + "epoch": 0.511200459506031, + "grad_norm": 0.2752167284488678, + "learning_rate": 4.691881342797368e-06, + "loss": 0.6835, + "step": 1335 + }, + { + "epoch": 0.5115833811985449, + "grad_norm": 0.2947768270969391, + "learning_rate": 4.691392695895645e-06, + "loss": 0.6759, + "step": 1336 + }, + { + "epoch": 0.5119663028910588, + "grad_norm": 0.30000224709510803, + "learning_rate": 4.690903687314391e-06, + "loss": 0.6697, + "step": 1337 + }, + { + "epoch": 0.5123492245835727, + "grad_norm": 0.2816091477870941, + "learning_rate": 4.690414317134318e-06, + "loss": 0.6646, + "step": 1338 + }, + { + "epoch": 0.5127321462760865, + "grad_norm": 0.2815537750720978, + "learning_rate": 4.689924585436192e-06, + "loss": 0.6631, + "step": 1339 + }, + { + "epoch": 0.5131150679686004, + "grad_norm": 0.29559898376464844, + "learning_rate": 4.689434492300841e-06, + "loss": 0.669, + "step": 1340 + }, + { + "epoch": 0.5134979896611143, + "grad_norm": 0.28763526678085327, + "learning_rate": 4.688944037809154e-06, + "loss": 0.6675, + "step": 1341 + }, + { + "epoch": 0.5138809113536282, + "grad_norm": 0.2929317355155945, + "learning_rate": 4.688453222042077e-06, + "loss": 0.7009, + "step": 1342 + }, + { + "epoch": 0.5142638330461421, + "grad_norm": 0.2719643712043762, + "learning_rate": 4.687962045080619e-06, + "loss": 0.6697, + "step": 1343 + }, + { + "epoch": 0.514646754738656, + "grad_norm": 0.299638569355011, + "learning_rate": 4.687470507005845e-06, + "loss": 0.6644, + "step": 1344 + }, + { + "epoch": 0.5150296764311698, + "grad_norm": 0.2654574513435364, + "learning_rate": 4.686978607898881e-06, + "loss": 0.6745, + "step": 1345 + }, + { + "epoch": 0.5154125981236837, + "grad_norm": 0.2910325527191162, + "learning_rate": 4.686486347840913e-06, + "loss": 0.6711, + "step": 1346 + }, + { + "epoch": 0.5157955198161975, + "grad_norm": 0.3001216650009155, + "learning_rate": 4.685993726913186e-06, + "loss": 0.6611, + "step": 1347 + }, + { + "epoch": 0.5161784415087115, + "grad_norm": 0.31864020228385925, + "learning_rate": 4.685500745197006e-06, + "loss": 0.6557, + "step": 1348 + }, + { + "epoch": 0.5165613632012254, + "grad_norm": 0.31896722316741943, + "learning_rate": 4.685007402773737e-06, + "loss": 0.6679, + "step": 1349 + }, + { + "epoch": 0.5169442848937392, + "grad_norm": 0.28124380111694336, + "learning_rate": 4.684513699724802e-06, + "loss": 0.6614, + "step": 1350 + }, + { + "epoch": 0.5173272065862531, + "grad_norm": 0.28782278299331665, + "learning_rate": 4.684019636131685e-06, + "loss": 0.6586, + "step": 1351 + }, + { + "epoch": 0.517710128278767, + "grad_norm": 0.2993348240852356, + "learning_rate": 4.6835252120759286e-06, + "loss": 0.662, + "step": 1352 + }, + { + "epoch": 0.5180930499712809, + "grad_norm": 0.2770722806453705, + "learning_rate": 4.683030427639137e-06, + "loss": 0.652, + "step": 1353 + }, + { + "epoch": 0.5184759716637948, + "grad_norm": 0.2981843054294586, + "learning_rate": 4.6825352829029705e-06, + "loss": 0.6851, + "step": 1354 + }, + { + "epoch": 0.5188588933563086, + "grad_norm": 0.2957566976547241, + "learning_rate": 4.682039777949151e-06, + "loss": 0.6664, + "step": 1355 + }, + { + "epoch": 0.5192418150488225, + "grad_norm": 0.293414831161499, + "learning_rate": 4.68154391285946e-06, + "loss": 0.684, + "step": 1356 + }, + { + "epoch": 0.5196247367413364, + "grad_norm": 0.2782313823699951, + "learning_rate": 4.6810476877157365e-06, + "loss": 0.6713, + "step": 1357 + }, + { + "epoch": 0.5200076584338503, + "grad_norm": 0.286119282245636, + "learning_rate": 4.680551102599881e-06, + "loss": 0.6933, + "step": 1358 + }, + { + "epoch": 0.5203905801263642, + "grad_norm": 0.3102959096431732, + "learning_rate": 4.680054157593852e-06, + "loss": 0.6777, + "step": 1359 + }, + { + "epoch": 0.520773501818878, + "grad_norm": 0.2751920521259308, + "learning_rate": 4.679556852779668e-06, + "loss": 0.6723, + "step": 1360 + }, + { + "epoch": 0.5211564235113919, + "grad_norm": 0.28185367584228516, + "learning_rate": 4.679059188239408e-06, + "loss": 0.6609, + "step": 1361 + }, + { + "epoch": 0.5215393452039058, + "grad_norm": 0.31509310007095337, + "learning_rate": 4.67856116405521e-06, + "loss": 0.6618, + "step": 1362 + }, + { + "epoch": 0.5219222668964196, + "grad_norm": 0.27555981278419495, + "learning_rate": 4.678062780309268e-06, + "loss": 0.6678, + "step": 1363 + }, + { + "epoch": 0.5223051885889336, + "grad_norm": 0.2945083677768707, + "learning_rate": 4.677564037083841e-06, + "loss": 0.6553, + "step": 1364 + }, + { + "epoch": 0.5226881102814475, + "grad_norm": 0.28949201107025146, + "learning_rate": 4.677064934461241e-06, + "loss": 0.6602, + "step": 1365 + }, + { + "epoch": 0.5230710319739613, + "grad_norm": 0.3095218241214752, + "learning_rate": 4.676565472523845e-06, + "loss": 0.6551, + "step": 1366 + }, + { + "epoch": 0.5234539536664752, + "grad_norm": 0.2740354537963867, + "learning_rate": 4.676065651354087e-06, + "loss": 0.6821, + "step": 1367 + }, + { + "epoch": 0.523836875358989, + "grad_norm": 0.29293930530548096, + "learning_rate": 4.6755654710344585e-06, + "loss": 0.6524, + "step": 1368 + }, + { + "epoch": 0.524219797051503, + "grad_norm": 0.289955198764801, + "learning_rate": 4.675064931647514e-06, + "loss": 0.6827, + "step": 1369 + }, + { + "epoch": 0.5246027187440169, + "grad_norm": 0.29977336525917053, + "learning_rate": 4.674564033275864e-06, + "loss": 0.6796, + "step": 1370 + }, + { + "epoch": 0.5249856404365307, + "grad_norm": 0.2801453471183777, + "learning_rate": 4.674062776002181e-06, + "loss": 0.6512, + "step": 1371 + }, + { + "epoch": 0.5253685621290446, + "grad_norm": 0.3363115191459656, + "learning_rate": 4.673561159909192e-06, + "loss": 0.6616, + "step": 1372 + }, + { + "epoch": 0.5257514838215585, + "grad_norm": 0.2854152321815491, + "learning_rate": 4.67305918507969e-06, + "loss": 0.6748, + "step": 1373 + }, + { + "epoch": 0.5261344055140724, + "grad_norm": 0.2665503919124603, + "learning_rate": 4.672556851596522e-06, + "loss": 0.6601, + "step": 1374 + }, + { + "epoch": 0.5265173272065863, + "grad_norm": 0.31339603662490845, + "learning_rate": 4.672054159542597e-06, + "loss": 0.661, + "step": 1375 + }, + { + "epoch": 0.5269002488991001, + "grad_norm": 0.2875712513923645, + "learning_rate": 4.671551109000882e-06, + "loss": 0.6817, + "step": 1376 + }, + { + "epoch": 0.527283170591614, + "grad_norm": 0.27537474036216736, + "learning_rate": 4.671047700054401e-06, + "loss": 0.6596, + "step": 1377 + }, + { + "epoch": 0.5276660922841279, + "grad_norm": 0.29016685485839844, + "learning_rate": 4.6705439327862415e-06, + "loss": 0.6781, + "step": 1378 + }, + { + "epoch": 0.5280490139766417, + "grad_norm": 0.3029150664806366, + "learning_rate": 4.670039807279547e-06, + "loss": 0.6677, + "step": 1379 + }, + { + "epoch": 0.5284319356691557, + "grad_norm": 0.28253310918807983, + "learning_rate": 4.669535323617522e-06, + "loss": 0.6905, + "step": 1380 + }, + { + "epoch": 0.5288148573616696, + "grad_norm": 0.29214736819267273, + "learning_rate": 4.669030481883429e-06, + "loss": 0.6606, + "step": 1381 + }, + { + "epoch": 0.5291977790541834, + "grad_norm": 0.28440895676612854, + "learning_rate": 4.66852528216059e-06, + "loss": 0.6699, + "step": 1382 + }, + { + "epoch": 0.5295807007466973, + "grad_norm": 0.295693963766098, + "learning_rate": 4.668019724532385e-06, + "loss": 0.6803, + "step": 1383 + }, + { + "epoch": 0.5299636224392111, + "grad_norm": 0.28351619839668274, + "learning_rate": 4.667513809082255e-06, + "loss": 0.6652, + "step": 1384 + }, + { + "epoch": 0.5303465441317251, + "grad_norm": 0.2787420153617859, + "learning_rate": 4.667007535893698e-06, + "loss": 0.6753, + "step": 1385 + }, + { + "epoch": 0.530729465824239, + "grad_norm": 0.2919510006904602, + "learning_rate": 4.666500905050274e-06, + "loss": 0.6764, + "step": 1386 + }, + { + "epoch": 0.5311123875167528, + "grad_norm": 0.2776666283607483, + "learning_rate": 4.665993916635599e-06, + "loss": 0.6651, + "step": 1387 + }, + { + "epoch": 0.5314953092092667, + "grad_norm": 0.2779915928840637, + "learning_rate": 4.66548657073335e-06, + "loss": 0.6683, + "step": 1388 + }, + { + "epoch": 0.5318782309017805, + "grad_norm": 0.28630954027175903, + "learning_rate": 4.66497886742726e-06, + "loss": 0.6743, + "step": 1389 + }, + { + "epoch": 0.5322611525942945, + "grad_norm": 0.2745286822319031, + "learning_rate": 4.664470806801125e-06, + "loss": 0.6704, + "step": 1390 + }, + { + "epoch": 0.5326440742868084, + "grad_norm": 0.28013500571250916, + "learning_rate": 4.663962388938797e-06, + "loss": 0.6716, + "step": 1391 + }, + { + "epoch": 0.5330269959793222, + "grad_norm": 0.2824057936668396, + "learning_rate": 4.66345361392419e-06, + "loss": 0.6646, + "step": 1392 + }, + { + "epoch": 0.5334099176718361, + "grad_norm": 0.2962542474269867, + "learning_rate": 4.662944481841273e-06, + "loss": 0.6611, + "step": 1393 + }, + { + "epoch": 0.53379283936435, + "grad_norm": 0.28005772829055786, + "learning_rate": 4.662434992774077e-06, + "loss": 0.6657, + "step": 1394 + }, + { + "epoch": 0.5341757610568638, + "grad_norm": 0.2697499394416809, + "learning_rate": 4.66192514680669e-06, + "loss": 0.6613, + "step": 1395 + }, + { + "epoch": 0.5345586827493778, + "grad_norm": 0.29074200987815857, + "learning_rate": 4.661414944023261e-06, + "loss": 0.6743, + "step": 1396 + }, + { + "epoch": 0.5349416044418916, + "grad_norm": 0.28916990756988525, + "learning_rate": 4.660904384507996e-06, + "loss": 0.6538, + "step": 1397 + }, + { + "epoch": 0.5353245261344055, + "grad_norm": 0.27587875723838806, + "learning_rate": 4.66039346834516e-06, + "loss": 0.6739, + "step": 1398 + }, + { + "epoch": 0.5357074478269194, + "grad_norm": 0.27750566601753235, + "learning_rate": 4.659882195619079e-06, + "loss": 0.6638, + "step": 1399 + }, + { + "epoch": 0.5360903695194332, + "grad_norm": 0.2798866927623749, + "learning_rate": 4.6593705664141356e-06, + "loss": 0.6735, + "step": 1400 + }, + { + "epoch": 0.5364732912119472, + "grad_norm": 0.27637436985969543, + "learning_rate": 4.658858580814772e-06, + "loss": 0.6794, + "step": 1401 + }, + { + "epoch": 0.5368562129044611, + "grad_norm": 0.27356451749801636, + "learning_rate": 4.658346238905487e-06, + "loss": 0.6762, + "step": 1402 + }, + { + "epoch": 0.5372391345969749, + "grad_norm": 0.2700641453266144, + "learning_rate": 4.657833540770843e-06, + "loss": 0.6445, + "step": 1403 + }, + { + "epoch": 0.5376220562894888, + "grad_norm": 0.286549836397171, + "learning_rate": 4.657320486495458e-06, + "loss": 0.6895, + "step": 1404 + }, + { + "epoch": 0.5380049779820026, + "grad_norm": 0.29119592905044556, + "learning_rate": 4.656807076164008e-06, + "loss": 0.6788, + "step": 1405 + }, + { + "epoch": 0.5383878996745166, + "grad_norm": 0.2834751307964325, + "learning_rate": 4.656293309861231e-06, + "loss": 0.6681, + "step": 1406 + }, + { + "epoch": 0.5387708213670305, + "grad_norm": 0.27905791997909546, + "learning_rate": 4.65577918767192e-06, + "loss": 0.6561, + "step": 1407 + }, + { + "epoch": 0.5391537430595443, + "grad_norm": 0.26963064074516296, + "learning_rate": 4.65526470968093e-06, + "loss": 0.6637, + "step": 1408 + }, + { + "epoch": 0.5395366647520582, + "grad_norm": 0.29435136914253235, + "learning_rate": 4.6547498759731725e-06, + "loss": 0.6698, + "step": 1409 + }, + { + "epoch": 0.539919586444572, + "grad_norm": 0.2793678045272827, + "learning_rate": 4.654234686633619e-06, + "loss": 0.6593, + "step": 1410 + }, + { + "epoch": 0.5403025081370859, + "grad_norm": 0.29543066024780273, + "learning_rate": 4.653719141747297e-06, + "loss": 0.6729, + "step": 1411 + }, + { + "epoch": 0.5406854298295999, + "grad_norm": 0.2844732105731964, + "learning_rate": 4.653203241399299e-06, + "loss": 0.6711, + "step": 1412 + }, + { + "epoch": 0.5410683515221137, + "grad_norm": 0.29418832063674927, + "learning_rate": 4.652686985674769e-06, + "loss": 0.6463, + "step": 1413 + }, + { + "epoch": 0.5414512732146276, + "grad_norm": 0.2700132131576538, + "learning_rate": 4.652170374658913e-06, + "loss": 0.6678, + "step": 1414 + }, + { + "epoch": 0.5418341949071415, + "grad_norm": 0.2833193242549896, + "learning_rate": 4.651653408436996e-06, + "loss": 0.662, + "step": 1415 + }, + { + "epoch": 0.5422171165996553, + "grad_norm": 0.28621354699134827, + "learning_rate": 4.651136087094341e-06, + "loss": 0.6552, + "step": 1416 + }, + { + "epoch": 0.5426000382921693, + "grad_norm": 0.28285256028175354, + "learning_rate": 4.65061841071633e-06, + "loss": 0.6707, + "step": 1417 + }, + { + "epoch": 0.5429829599846832, + "grad_norm": 0.29332616925239563, + "learning_rate": 4.650100379388401e-06, + "loss": 0.6639, + "step": 1418 + }, + { + "epoch": 0.543365881677197, + "grad_norm": 0.27814438939094543, + "learning_rate": 4.649581993196055e-06, + "loss": 0.6575, + "step": 1419 + }, + { + "epoch": 0.5437488033697109, + "grad_norm": 0.2910020649433136, + "learning_rate": 4.6490632522248485e-06, + "loss": 0.6731, + "step": 1420 + }, + { + "epoch": 0.5441317250622247, + "grad_norm": 0.2723335325717926, + "learning_rate": 4.648544156560397e-06, + "loss": 0.6646, + "step": 1421 + }, + { + "epoch": 0.5445146467547387, + "grad_norm": 0.2737279534339905, + "learning_rate": 4.648024706288376e-06, + "loss": 0.6533, + "step": 1422 + }, + { + "epoch": 0.5448975684472526, + "grad_norm": 0.2696785032749176, + "learning_rate": 4.647504901494517e-06, + "loss": 0.6515, + "step": 1423 + }, + { + "epoch": 0.5452804901397664, + "grad_norm": 0.280367374420166, + "learning_rate": 4.646984742264613e-06, + "loss": 0.6497, + "step": 1424 + }, + { + "epoch": 0.5456634118322803, + "grad_norm": 0.2739644944667816, + "learning_rate": 4.646464228684512e-06, + "loss": 0.6475, + "step": 1425 + }, + { + "epoch": 0.5460463335247941, + "grad_norm": 0.27324730157852173, + "learning_rate": 4.645943360840124e-06, + "loss": 0.6874, + "step": 1426 + }, + { + "epoch": 0.546429255217308, + "grad_norm": 0.28439703583717346, + "learning_rate": 4.6454221388174165e-06, + "loss": 0.6622, + "step": 1427 + }, + { + "epoch": 0.546812176909822, + "grad_norm": 0.2706713378429413, + "learning_rate": 4.6449005627024126e-06, + "loss": 0.6481, + "step": 1428 + }, + { + "epoch": 0.5471950986023358, + "grad_norm": 0.28453972935676575, + "learning_rate": 4.644378632581198e-06, + "loss": 0.668, + "step": 1429 + }, + { + "epoch": 0.5475780202948497, + "grad_norm": 0.28511837124824524, + "learning_rate": 4.643856348539913e-06, + "loss": 0.6736, + "step": 1430 + }, + { + "epoch": 0.5479609419873636, + "grad_norm": 0.2903471291065216, + "learning_rate": 4.643333710664761e-06, + "loss": 0.659, + "step": 1431 + }, + { + "epoch": 0.5483438636798774, + "grad_norm": 0.28712940216064453, + "learning_rate": 4.642810719041999e-06, + "loss": 0.6664, + "step": 1432 + }, + { + "epoch": 0.5487267853723914, + "grad_norm": 0.27602365612983704, + "learning_rate": 4.642287373757946e-06, + "loss": 0.675, + "step": 1433 + }, + { + "epoch": 0.5491097070649053, + "grad_norm": 0.29832765460014343, + "learning_rate": 4.641763674898975e-06, + "loss": 0.6763, + "step": 1434 + }, + { + "epoch": 0.5494926287574191, + "grad_norm": 0.2785443663597107, + "learning_rate": 4.641239622551522e-06, + "loss": 0.6493, + "step": 1435 + }, + { + "epoch": 0.549875550449933, + "grad_norm": 0.28630906343460083, + "learning_rate": 4.64071521680208e-06, + "loss": 0.6475, + "step": 1436 + }, + { + "epoch": 0.5502584721424468, + "grad_norm": 0.27429166436195374, + "learning_rate": 4.6401904577371995e-06, + "loss": 0.6491, + "step": 1437 + }, + { + "epoch": 0.5506413938349608, + "grad_norm": 0.2856921851634979, + "learning_rate": 4.6396653454434884e-06, + "loss": 0.6831, + "step": 1438 + }, + { + "epoch": 0.5510243155274747, + "grad_norm": 0.2848499119281769, + "learning_rate": 4.639139880007615e-06, + "loss": 0.6703, + "step": 1439 + }, + { + "epoch": 0.5514072372199885, + "grad_norm": 0.2933283746242523, + "learning_rate": 4.6386140615163056e-06, + "loss": 0.6472, + "step": 1440 + }, + { + "epoch": 0.5517901589125024, + "grad_norm": 0.28132501244544983, + "learning_rate": 4.638087890056343e-06, + "loss": 0.6666, + "step": 1441 + }, + { + "epoch": 0.5521730806050162, + "grad_norm": 0.29337993264198303, + "learning_rate": 4.637561365714571e-06, + "loss": 0.6824, + "step": 1442 + }, + { + "epoch": 0.5525560022975301, + "grad_norm": 0.2802111804485321, + "learning_rate": 4.637034488577889e-06, + "loss": 0.6486, + "step": 1443 + }, + { + "epoch": 0.5529389239900441, + "grad_norm": 0.27604779601097107, + "learning_rate": 4.636507258733255e-06, + "loss": 0.655, + "step": 1444 + }, + { + "epoch": 0.5533218456825579, + "grad_norm": 0.2832767069339752, + "learning_rate": 4.6359796762676875e-06, + "loss": 0.6719, + "step": 1445 + }, + { + "epoch": 0.5537047673750718, + "grad_norm": 0.2880399227142334, + "learning_rate": 4.6354517412682605e-06, + "loss": 0.6646, + "step": 1446 + }, + { + "epoch": 0.5540876890675857, + "grad_norm": 0.2932654619216919, + "learning_rate": 4.634923453822108e-06, + "loss": 0.6746, + "step": 1447 + }, + { + "epoch": 0.5544706107600995, + "grad_norm": 0.2653713822364807, + "learning_rate": 4.634394814016421e-06, + "loss": 0.6555, + "step": 1448 + }, + { + "epoch": 0.5548535324526135, + "grad_norm": 0.2792087495326996, + "learning_rate": 4.633865821938449e-06, + "loss": 0.6799, + "step": 1449 + }, + { + "epoch": 0.5552364541451273, + "grad_norm": 0.28352558612823486, + "learning_rate": 4.6333364776755006e-06, + "loss": 0.6632, + "step": 1450 + }, + { + "epoch": 0.5556193758376412, + "grad_norm": 0.2973845303058624, + "learning_rate": 4.632806781314941e-06, + "loss": 0.664, + "step": 1451 + }, + { + "epoch": 0.5560022975301551, + "grad_norm": 0.2806834578514099, + "learning_rate": 4.632276732944194e-06, + "loss": 0.6655, + "step": 1452 + }, + { + "epoch": 0.5563852192226689, + "grad_norm": 0.2828749120235443, + "learning_rate": 4.631746332650743e-06, + "loss": 0.6776, + "step": 1453 + }, + { + "epoch": 0.5567681409151829, + "grad_norm": 0.28335845470428467, + "learning_rate": 4.631215580522126e-06, + "loss": 0.6798, + "step": 1454 + }, + { + "epoch": 0.5571510626076968, + "grad_norm": 0.2798022925853729, + "learning_rate": 4.6306844766459435e-06, + "loss": 0.6655, + "step": 1455 + }, + { + "epoch": 0.5575339843002106, + "grad_norm": 0.284993976354599, + "learning_rate": 4.63015302110985e-06, + "loss": 0.668, + "step": 1456 + }, + { + "epoch": 0.5579169059927245, + "grad_norm": 0.2809392213821411, + "learning_rate": 4.629621214001562e-06, + "loss": 0.6799, + "step": 1457 + }, + { + "epoch": 0.5582998276852383, + "grad_norm": 0.2810289263725281, + "learning_rate": 4.6290890554088485e-06, + "loss": 0.6751, + "step": 1458 + }, + { + "epoch": 0.5586827493777522, + "grad_norm": 0.2804025709629059, + "learning_rate": 4.628556545419543e-06, + "loss": 0.6704, + "step": 1459 + }, + { + "epoch": 0.5590656710702662, + "grad_norm": 0.2831740379333496, + "learning_rate": 4.6280236841215335e-06, + "loss": 0.6615, + "step": 1460 + }, + { + "epoch": 0.55944859276278, + "grad_norm": 0.2875085175037384, + "learning_rate": 4.627490471602765e-06, + "loss": 0.6644, + "step": 1461 + }, + { + "epoch": 0.5598315144552939, + "grad_norm": 0.27246806025505066, + "learning_rate": 4.626956907951242e-06, + "loss": 0.642, + "step": 1462 + }, + { + "epoch": 0.5602144361478077, + "grad_norm": 0.2839241325855255, + "learning_rate": 4.626422993255028e-06, + "loss": 0.6532, + "step": 1463 + }, + { + "epoch": 0.5605973578403216, + "grad_norm": 0.276871919631958, + "learning_rate": 4.6258887276022425e-06, + "loss": 0.663, + "step": 1464 + }, + { + "epoch": 0.5609802795328356, + "grad_norm": 0.2767009139060974, + "learning_rate": 4.6253541110810635e-06, + "loss": 0.6735, + "step": 1465 + }, + { + "epoch": 0.5613632012253494, + "grad_norm": 0.28140896558761597, + "learning_rate": 4.6248191437797275e-06, + "loss": 0.6337, + "step": 1466 + }, + { + "epoch": 0.5617461229178633, + "grad_norm": 0.2746315896511078, + "learning_rate": 4.624283825786528e-06, + "loss": 0.6402, + "step": 1467 + }, + { + "epoch": 0.5621290446103772, + "grad_norm": 0.2852254807949066, + "learning_rate": 4.623748157189817e-06, + "loss": 0.6781, + "step": 1468 + }, + { + "epoch": 0.562511966302891, + "grad_norm": 0.2737531363964081, + "learning_rate": 4.623212138078004e-06, + "loss": 0.6339, + "step": 1469 + }, + { + "epoch": 0.562894887995405, + "grad_norm": 0.2868412137031555, + "learning_rate": 4.622675768539557e-06, + "loss": 0.6779, + "step": 1470 + }, + { + "epoch": 0.5632778096879189, + "grad_norm": 0.2785370647907257, + "learning_rate": 4.622139048663001e-06, + "loss": 0.676, + "step": 1471 + }, + { + "epoch": 0.5636607313804327, + "grad_norm": 0.29825419187545776, + "learning_rate": 4.621601978536919e-06, + "loss": 0.679, + "step": 1472 + }, + { + "epoch": 0.5640436530729466, + "grad_norm": 0.2933095693588257, + "learning_rate": 4.6210645582499536e-06, + "loss": 0.651, + "step": 1473 + }, + { + "epoch": 0.5644265747654604, + "grad_norm": 0.27432429790496826, + "learning_rate": 4.620526787890801e-06, + "loss": 0.6734, + "step": 1474 + }, + { + "epoch": 0.5648094964579743, + "grad_norm": 0.29858604073524475, + "learning_rate": 4.6199886675482195e-06, + "loss": 0.6762, + "step": 1475 + }, + { + "epoch": 0.5651924181504883, + "grad_norm": 0.2854273021221161, + "learning_rate": 4.619450197311024e-06, + "loss": 0.6666, + "step": 1476 + }, + { + "epoch": 0.5655753398430021, + "grad_norm": 0.2858935296535492, + "learning_rate": 4.618911377268085e-06, + "loss": 0.656, + "step": 1477 + }, + { + "epoch": 0.565958261535516, + "grad_norm": 0.2800264358520508, + "learning_rate": 4.618372207508333e-06, + "loss": 0.6727, + "step": 1478 + }, + { + "epoch": 0.5663411832280298, + "grad_norm": 0.2903977334499359, + "learning_rate": 4.617832688120756e-06, + "loss": 0.6712, + "step": 1479 + }, + { + "epoch": 0.5667241049205437, + "grad_norm": 0.30259647965431213, + "learning_rate": 4.617292819194398e-06, + "loss": 0.6776, + "step": 1480 + }, + { + "epoch": 0.5671070266130577, + "grad_norm": 0.2795424461364746, + "learning_rate": 4.616752600818363e-06, + "loss": 0.6726, + "step": 1481 + }, + { + "epoch": 0.5674899483055715, + "grad_norm": 0.2867182195186615, + "learning_rate": 4.616212033081811e-06, + "loss": 0.6604, + "step": 1482 + }, + { + "epoch": 0.5678728699980854, + "grad_norm": 0.2813325822353363, + "learning_rate": 4.615671116073962e-06, + "loss": 0.6661, + "step": 1483 + }, + { + "epoch": 0.5682557916905993, + "grad_norm": 0.29366227984428406, + "learning_rate": 4.61512984988409e-06, + "loss": 0.6698, + "step": 1484 + }, + { + "epoch": 0.5686387133831131, + "grad_norm": 0.271113783121109, + "learning_rate": 4.614588234601529e-06, + "loss": 0.6711, + "step": 1485 + }, + { + "epoch": 0.5690216350756271, + "grad_norm": 0.2789374887943268, + "learning_rate": 4.614046270315671e-06, + "loss": 0.6735, + "step": 1486 + }, + { + "epoch": 0.569404556768141, + "grad_norm": 0.2933051884174347, + "learning_rate": 4.613503957115963e-06, + "loss": 0.6819, + "step": 1487 + }, + { + "epoch": 0.5697874784606548, + "grad_norm": 0.2969438135623932, + "learning_rate": 4.612961295091913e-06, + "loss": 0.6719, + "step": 1488 + }, + { + "epoch": 0.5701704001531687, + "grad_norm": 0.2947516143321991, + "learning_rate": 4.612418284333085e-06, + "loss": 0.6706, + "step": 1489 + }, + { + "epoch": 0.5705533218456825, + "grad_norm": 0.30186280608177185, + "learning_rate": 4.6118749249290985e-06, + "loss": 0.6443, + "step": 1490 + }, + { + "epoch": 0.5709362435381964, + "grad_norm": 0.2903065085411072, + "learning_rate": 4.611331216969634e-06, + "loss": 0.6672, + "step": 1491 + }, + { + "epoch": 0.5713191652307104, + "grad_norm": 0.2867312431335449, + "learning_rate": 4.610787160544429e-06, + "loss": 0.6658, + "step": 1492 + }, + { + "epoch": 0.5717020869232242, + "grad_norm": 0.2932186424732208, + "learning_rate": 4.610242755743276e-06, + "loss": 0.6628, + "step": 1493 + }, + { + "epoch": 0.5720850086157381, + "grad_norm": 0.2865498960018158, + "learning_rate": 4.609698002656027e-06, + "loss": 0.6605, + "step": 1494 + }, + { + "epoch": 0.5724679303082519, + "grad_norm": 0.2915932238101959, + "learning_rate": 4.609152901372592e-06, + "loss": 0.6669, + "step": 1495 + }, + { + "epoch": 0.5728508520007658, + "grad_norm": 0.291251540184021, + "learning_rate": 4.608607451982936e-06, + "loss": 0.6642, + "step": 1496 + }, + { + "epoch": 0.5732337736932798, + "grad_norm": 0.28860780596733093, + "learning_rate": 4.608061654577084e-06, + "loss": 0.6716, + "step": 1497 + }, + { + "epoch": 0.5736166953857936, + "grad_norm": 0.2796918749809265, + "learning_rate": 4.607515509245116e-06, + "loss": 0.6678, + "step": 1498 + }, + { + "epoch": 0.5739996170783075, + "grad_norm": 0.2800701856613159, + "learning_rate": 4.606969016077173e-06, + "loss": 0.66, + "step": 1499 + }, + { + "epoch": 0.5743825387708213, + "grad_norm": 0.2912375032901764, + "learning_rate": 4.6064221751634495e-06, + "loss": 0.657, + "step": 1500 + }, + { + "epoch": 0.5747654604633352, + "grad_norm": 0.29719606041908264, + "learning_rate": 4.6058749865942e-06, + "loss": 0.6662, + "step": 1501 + }, + { + "epoch": 0.5751483821558492, + "grad_norm": 0.2698865532875061, + "learning_rate": 4.605327450459736e-06, + "loss": 0.6733, + "step": 1502 + }, + { + "epoch": 0.575531303848363, + "grad_norm": 0.29630282521247864, + "learning_rate": 4.604779566850425e-06, + "loss": 0.6667, + "step": 1503 + }, + { + "epoch": 0.5759142255408769, + "grad_norm": 0.291897714138031, + "learning_rate": 4.604231335856692e-06, + "loss": 0.6604, + "step": 1504 + }, + { + "epoch": 0.5762971472333908, + "grad_norm": 0.278290718793869, + "learning_rate": 4.603682757569021e-06, + "loss": 0.6448, + "step": 1505 + }, + { + "epoch": 0.5766800689259046, + "grad_norm": 0.2939246594905853, + "learning_rate": 4.603133832077953e-06, + "loss": 0.6592, + "step": 1506 + }, + { + "epoch": 0.5770629906184185, + "grad_norm": 0.2818027138710022, + "learning_rate": 4.602584559474085e-06, + "loss": 0.6665, + "step": 1507 + }, + { + "epoch": 0.5774459123109325, + "grad_norm": 0.2802014946937561, + "learning_rate": 4.602034939848072e-06, + "loss": 0.6741, + "step": 1508 + }, + { + "epoch": 0.5778288340034463, + "grad_norm": 0.2793828547000885, + "learning_rate": 4.601484973290627e-06, + "loss": 0.6701, + "step": 1509 + }, + { + "epoch": 0.5782117556959602, + "grad_norm": 0.29030466079711914, + "learning_rate": 4.600934659892518e-06, + "loss": 0.6631, + "step": 1510 + }, + { + "epoch": 0.578594677388474, + "grad_norm": 0.2780533730983734, + "learning_rate": 4.600383999744572e-06, + "loss": 0.6601, + "step": 1511 + }, + { + "epoch": 0.5789775990809879, + "grad_norm": 0.2862532436847687, + "learning_rate": 4.599832992937674e-06, + "loss": 0.6734, + "step": 1512 + }, + { + "epoch": 0.5793605207735019, + "grad_norm": 0.2872373163700104, + "learning_rate": 4.599281639562765e-06, + "loss": 0.6543, + "step": 1513 + }, + { + "epoch": 0.5797434424660157, + "grad_norm": 0.2874769866466522, + "learning_rate": 4.598729939710843e-06, + "loss": 0.6712, + "step": 1514 + }, + { + "epoch": 0.5801263641585296, + "grad_norm": 0.2845470607280731, + "learning_rate": 4.598177893472963e-06, + "loss": 0.6732, + "step": 1515 + }, + { + "epoch": 0.5805092858510434, + "grad_norm": 0.2877166271209717, + "learning_rate": 4.59762550094024e-06, + "loss": 0.6772, + "step": 1516 + }, + { + "epoch": 0.5808922075435573, + "grad_norm": 0.2804192900657654, + "learning_rate": 4.5970727622038415e-06, + "loss": 0.6493, + "step": 1517 + }, + { + "epoch": 0.5812751292360713, + "grad_norm": 0.29686498641967773, + "learning_rate": 4.596519677354995e-06, + "loss": 0.6717, + "step": 1518 + }, + { + "epoch": 0.5816580509285851, + "grad_norm": 0.2753806710243225, + "learning_rate": 4.595966246484986e-06, + "loss": 0.6572, + "step": 1519 + }, + { + "epoch": 0.582040972621099, + "grad_norm": 0.2850971221923828, + "learning_rate": 4.595412469685154e-06, + "loss": 0.647, + "step": 1520 + }, + { + "epoch": 0.5824238943136129, + "grad_norm": 0.3031977713108063, + "learning_rate": 4.594858347046899e-06, + "loss": 0.6634, + "step": 1521 + }, + { + "epoch": 0.5828068160061267, + "grad_norm": 0.2768973112106323, + "learning_rate": 4.594303878661676e-06, + "loss": 0.6569, + "step": 1522 + }, + { + "epoch": 0.5831897376986406, + "grad_norm": 0.29982253909111023, + "learning_rate": 4.593749064620996e-06, + "loss": 0.6503, + "step": 1523 + }, + { + "epoch": 0.5835726593911545, + "grad_norm": 0.3226918876171112, + "learning_rate": 4.593193905016431e-06, + "loss": 0.6593, + "step": 1524 + }, + { + "epoch": 0.5839555810836684, + "grad_norm": 0.28621339797973633, + "learning_rate": 4.592638399939606e-06, + "loss": 0.6677, + "step": 1525 + }, + { + "epoch": 0.5843385027761823, + "grad_norm": 0.31656476855278015, + "learning_rate": 4.592082549482205e-06, + "loss": 0.6422, + "step": 1526 + }, + { + "epoch": 0.5847214244686961, + "grad_norm": 0.31038275361061096, + "learning_rate": 4.5915263537359685e-06, + "loss": 0.6521, + "step": 1527 + }, + { + "epoch": 0.58510434616121, + "grad_norm": 0.2947365641593933, + "learning_rate": 4.590969812792694e-06, + "loss": 0.6655, + "step": 1528 + }, + { + "epoch": 0.585487267853724, + "grad_norm": 0.3024202883243561, + "learning_rate": 4.590412926744237e-06, + "loss": 0.6577, + "step": 1529 + }, + { + "epoch": 0.5858701895462378, + "grad_norm": 0.2764289081096649, + "learning_rate": 4.589855695682508e-06, + "loss": 0.64, + "step": 1530 + }, + { + "epoch": 0.5862531112387517, + "grad_norm": 0.28834182024002075, + "learning_rate": 4.5892981196994755e-06, + "loss": 0.6613, + "step": 1531 + }, + { + "epoch": 0.5866360329312655, + "grad_norm": 0.29963499307632446, + "learning_rate": 4.588740198887166e-06, + "loss": 0.6848, + "step": 1532 + }, + { + "epoch": 0.5870189546237794, + "grad_norm": 0.2901003360748291, + "learning_rate": 4.588181933337661e-06, + "loss": 0.6728, + "step": 1533 + }, + { + "epoch": 0.5874018763162934, + "grad_norm": 0.28126630187034607, + "learning_rate": 4.587623323143099e-06, + "loss": 0.6958, + "step": 1534 + }, + { + "epoch": 0.5877847980088072, + "grad_norm": 0.2929370105266571, + "learning_rate": 4.587064368395679e-06, + "loss": 0.6556, + "step": 1535 + }, + { + "epoch": 0.5881677197013211, + "grad_norm": 0.2871021330356598, + "learning_rate": 4.58650506918765e-06, + "loss": 0.6726, + "step": 1536 + }, + { + "epoch": 0.588550641393835, + "grad_norm": 0.280502587556839, + "learning_rate": 4.5859454256113235e-06, + "loss": 0.686, + "step": 1537 + }, + { + "epoch": 0.5889335630863488, + "grad_norm": 0.28914710879325867, + "learning_rate": 4.5853854377590675e-06, + "loss": 0.6656, + "step": 1538 + }, + { + "epoch": 0.5893164847788627, + "grad_norm": 0.2821902930736542, + "learning_rate": 4.584825105723304e-06, + "loss": 0.6686, + "step": 1539 + }, + { + "epoch": 0.5896994064713766, + "grad_norm": 0.29660072922706604, + "learning_rate": 4.5842644295965136e-06, + "loss": 0.6291, + "step": 1540 + }, + { + "epoch": 0.5900823281638905, + "grad_norm": 0.28437337279319763, + "learning_rate": 4.583703409471234e-06, + "loss": 0.6716, + "step": 1541 + }, + { + "epoch": 0.5904652498564044, + "grad_norm": 0.27621564269065857, + "learning_rate": 4.5831420454400575e-06, + "loss": 0.653, + "step": 1542 + }, + { + "epoch": 0.5908481715489182, + "grad_norm": 0.2817482054233551, + "learning_rate": 4.582580337595636e-06, + "loss": 0.6565, + "step": 1543 + }, + { + "epoch": 0.5912310932414321, + "grad_norm": 0.289776474237442, + "learning_rate": 4.582018286030677e-06, + "loss": 0.6541, + "step": 1544 + }, + { + "epoch": 0.591614014933946, + "grad_norm": 0.2802707850933075, + "learning_rate": 4.581455890837943e-06, + "loss": 0.6412, + "step": 1545 + }, + { + "epoch": 0.5919969366264599, + "grad_norm": 0.28509002923965454, + "learning_rate": 4.5808931521102585e-06, + "loss": 0.669, + "step": 1546 + }, + { + "epoch": 0.5923798583189738, + "grad_norm": 0.2864404022693634, + "learning_rate": 4.580330069940496e-06, + "loss": 0.6604, + "step": 1547 + }, + { + "epoch": 0.5927627800114876, + "grad_norm": 0.28585776686668396, + "learning_rate": 4.579766644421593e-06, + "loss": 0.6552, + "step": 1548 + }, + { + "epoch": 0.5931457017040015, + "grad_norm": 0.29642269015312195, + "learning_rate": 4.57920287564654e-06, + "loss": 0.6626, + "step": 1549 + }, + { + "epoch": 0.5935286233965155, + "grad_norm": 0.3006981313228607, + "learning_rate": 4.578638763708384e-06, + "loss": 0.6478, + "step": 1550 + }, + { + "epoch": 0.5939115450890293, + "grad_norm": 0.28076109290122986, + "learning_rate": 4.578074308700228e-06, + "loss": 0.6551, + "step": 1551 + }, + { + "epoch": 0.5942944667815432, + "grad_norm": 0.2876197397708893, + "learning_rate": 4.577509510715234e-06, + "loss": 0.6683, + "step": 1552 + }, + { + "epoch": 0.594677388474057, + "grad_norm": 0.285412460565567, + "learning_rate": 4.576944369846621e-06, + "loss": 0.6697, + "step": 1553 + }, + { + "epoch": 0.5950603101665709, + "grad_norm": 0.2955072820186615, + "learning_rate": 4.57637888618766e-06, + "loss": 0.6512, + "step": 1554 + }, + { + "epoch": 0.5954432318590848, + "grad_norm": 0.2828025221824646, + "learning_rate": 4.575813059831683e-06, + "loss": 0.6519, + "step": 1555 + }, + { + "epoch": 0.5958261535515987, + "grad_norm": 0.29677820205688477, + "learning_rate": 4.575246890872077e-06, + "loss": 0.6749, + "step": 1556 + }, + { + "epoch": 0.5962090752441126, + "grad_norm": 0.28829675912857056, + "learning_rate": 4.574680379402285e-06, + "loss": 0.663, + "step": 1557 + }, + { + "epoch": 0.5965919969366265, + "grad_norm": 0.28253594040870667, + "learning_rate": 4.574113525515809e-06, + "loss": 0.6689, + "step": 1558 + }, + { + "epoch": 0.5969749186291403, + "grad_norm": 0.29308921098709106, + "learning_rate": 4.5735463293062026e-06, + "loss": 0.6837, + "step": 1559 + }, + { + "epoch": 0.5973578403216542, + "grad_norm": 0.3044282793998718, + "learning_rate": 4.572978790867082e-06, + "loss": 0.6541, + "step": 1560 + }, + { + "epoch": 0.5977407620141681, + "grad_norm": 0.2985263764858246, + "learning_rate": 4.572410910292116e-06, + "loss": 0.6708, + "step": 1561 + }, + { + "epoch": 0.598123683706682, + "grad_norm": 0.3009042739868164, + "learning_rate": 4.571842687675029e-06, + "loss": 0.6658, + "step": 1562 + }, + { + "epoch": 0.5985066053991959, + "grad_norm": 0.2898600101470947, + "learning_rate": 4.571274123109606e-06, + "loss": 0.6702, + "step": 1563 + }, + { + "epoch": 0.5988895270917097, + "grad_norm": 0.28798240423202515, + "learning_rate": 4.570705216689685e-06, + "loss": 0.668, + "step": 1564 + }, + { + "epoch": 0.5992724487842236, + "grad_norm": 0.2966676354408264, + "learning_rate": 4.5701359685091605e-06, + "loss": 0.6587, + "step": 1565 + }, + { + "epoch": 0.5996553704767376, + "grad_norm": 0.28760483860969543, + "learning_rate": 4.569566378661987e-06, + "loss": 0.6412, + "step": 1566 + }, + { + "epoch": 0.6000382921692514, + "grad_norm": 0.290778785943985, + "learning_rate": 4.56899644724217e-06, + "loss": 0.6698, + "step": 1567 + }, + { + "epoch": 0.6004212138617653, + "grad_norm": 0.284491628408432, + "learning_rate": 4.5684261743437756e-06, + "loss": 0.6863, + "step": 1568 + }, + { + "epoch": 0.6008041355542791, + "grad_norm": 0.28828680515289307, + "learning_rate": 4.567855560060923e-06, + "loss": 0.656, + "step": 1569 + }, + { + "epoch": 0.601187057246793, + "grad_norm": 0.2792324721813202, + "learning_rate": 4.567284604487793e-06, + "loss": 0.6665, + "step": 1570 + }, + { + "epoch": 0.6015699789393069, + "grad_norm": 0.2900835871696472, + "learning_rate": 4.566713307718616e-06, + "loss": 0.6789, + "step": 1571 + }, + { + "epoch": 0.6019529006318208, + "grad_norm": 0.2878018915653229, + "learning_rate": 4.566141669847683e-06, + "loss": 0.6603, + "step": 1572 + }, + { + "epoch": 0.6023358223243347, + "grad_norm": 0.28778883814811707, + "learning_rate": 4.565569690969341e-06, + "loss": 0.6343, + "step": 1573 + }, + { + "epoch": 0.6027187440168486, + "grad_norm": 0.2889214754104614, + "learning_rate": 4.564997371177992e-06, + "loss": 0.6672, + "step": 1574 + }, + { + "epoch": 0.6031016657093624, + "grad_norm": 0.2788236737251282, + "learning_rate": 4.564424710568095e-06, + "loss": 0.6515, + "step": 1575 + }, + { + "epoch": 0.6034845874018763, + "grad_norm": 0.27206557989120483, + "learning_rate": 4.563851709234164e-06, + "loss": 0.6483, + "step": 1576 + }, + { + "epoch": 0.6038675090943902, + "grad_norm": 0.28735655546188354, + "learning_rate": 4.563278367270772e-06, + "loss": 0.6597, + "step": 1577 + }, + { + "epoch": 0.6042504307869041, + "grad_norm": 0.29162538051605225, + "learning_rate": 4.562704684772545e-06, + "loss": 0.6685, + "step": 1578 + }, + { + "epoch": 0.604633352479418, + "grad_norm": 0.2774061858654022, + "learning_rate": 4.5621306618341674e-06, + "loss": 0.658, + "step": 1579 + }, + { + "epoch": 0.6050162741719318, + "grad_norm": 0.2785862386226654, + "learning_rate": 4.561556298550379e-06, + "loss": 0.6621, + "step": 1580 + }, + { + "epoch": 0.6053991958644457, + "grad_norm": 0.29351580142974854, + "learning_rate": 4.560981595015976e-06, + "loss": 0.6577, + "step": 1581 + }, + { + "epoch": 0.6057821175569597, + "grad_norm": 0.2703281044960022, + "learning_rate": 4.560406551325811e-06, + "loss": 0.6528, + "step": 1582 + }, + { + "epoch": 0.6061650392494735, + "grad_norm": 0.28498125076293945, + "learning_rate": 4.559831167574792e-06, + "loss": 0.6598, + "step": 1583 + }, + { + "epoch": 0.6065479609419874, + "grad_norm": 0.27280595898628235, + "learning_rate": 4.5592554438578825e-06, + "loss": 0.662, + "step": 1584 + }, + { + "epoch": 0.6069308826345012, + "grad_norm": 0.2835381031036377, + "learning_rate": 4.5586793802701055e-06, + "loss": 0.6476, + "step": 1585 + }, + { + "epoch": 0.6073138043270151, + "grad_norm": 0.2844344675540924, + "learning_rate": 4.5581029769065356e-06, + "loss": 0.6461, + "step": 1586 + }, + { + "epoch": 0.607696726019529, + "grad_norm": 0.29839304089546204, + "learning_rate": 4.557526233862306e-06, + "loss": 0.6476, + "step": 1587 + }, + { + "epoch": 0.6080796477120429, + "grad_norm": 0.288681298494339, + "learning_rate": 4.556949151232606e-06, + "loss": 0.6656, + "step": 1588 + }, + { + "epoch": 0.6084625694045568, + "grad_norm": 0.28268155455589294, + "learning_rate": 4.556371729112681e-06, + "loss": 0.6607, + "step": 1589 + }, + { + "epoch": 0.6088454910970706, + "grad_norm": 0.28057920932769775, + "learning_rate": 4.55579396759783e-06, + "loss": 0.666, + "step": 1590 + }, + { + "epoch": 0.6092284127895845, + "grad_norm": 0.2781616747379303, + "learning_rate": 4.555215866783412e-06, + "loss": 0.6482, + "step": 1591 + }, + { + "epoch": 0.6096113344820984, + "grad_norm": 0.27866655588150024, + "learning_rate": 4.55463742676484e-06, + "loss": 0.6714, + "step": 1592 + }, + { + "epoch": 0.6099942561746123, + "grad_norm": 0.2757575213909149, + "learning_rate": 4.55405864763758e-06, + "loss": 0.6692, + "step": 1593 + }, + { + "epoch": 0.6103771778671262, + "grad_norm": 0.2932291030883789, + "learning_rate": 4.553479529497161e-06, + "loss": 0.6728, + "step": 1594 + }, + { + "epoch": 0.6107600995596401, + "grad_norm": 0.2893361747264862, + "learning_rate": 4.552900072439161e-06, + "loss": 0.6583, + "step": 1595 + }, + { + "epoch": 0.6111430212521539, + "grad_norm": 0.28445225954055786, + "learning_rate": 4.552320276559218e-06, + "loss": 0.651, + "step": 1596 + }, + { + "epoch": 0.6115259429446678, + "grad_norm": 0.2812953591346741, + "learning_rate": 4.551740141953024e-06, + "loss": 0.6565, + "step": 1597 + }, + { + "epoch": 0.6119088646371817, + "grad_norm": 0.2781602442264557, + "learning_rate": 4.551159668716328e-06, + "loss": 0.6742, + "step": 1598 + }, + { + "epoch": 0.6122917863296956, + "grad_norm": 0.29508906602859497, + "learning_rate": 4.550578856944934e-06, + "loss": 0.664, + "step": 1599 + }, + { + "epoch": 0.6126747080222095, + "grad_norm": 0.27833086252212524, + "learning_rate": 4.549997706734703e-06, + "loss": 0.6435, + "step": 1600 + }, + { + "epoch": 0.6130576297147233, + "grad_norm": 0.2886672616004944, + "learning_rate": 4.5494162181815515e-06, + "loss": 0.6747, + "step": 1601 + }, + { + "epoch": 0.6134405514072372, + "grad_norm": 0.2867382764816284, + "learning_rate": 4.5488343913814514e-06, + "loss": 0.6662, + "step": 1602 + }, + { + "epoch": 0.613823473099751, + "grad_norm": 0.2865198850631714, + "learning_rate": 4.548252226430429e-06, + "loss": 0.6753, + "step": 1603 + }, + { + "epoch": 0.614206394792265, + "grad_norm": 0.27338707447052, + "learning_rate": 4.547669723424571e-06, + "loss": 0.6728, + "step": 1604 + }, + { + "epoch": 0.6145893164847789, + "grad_norm": 0.2886753976345062, + "learning_rate": 4.547086882460015e-06, + "loss": 0.6351, + "step": 1605 + }, + { + "epoch": 0.6149722381772927, + "grad_norm": 0.2924022674560547, + "learning_rate": 4.5465037036329555e-06, + "loss": 0.6796, + "step": 1606 + }, + { + "epoch": 0.6153551598698066, + "grad_norm": 0.29398876428604126, + "learning_rate": 4.545920187039645e-06, + "loss": 0.6784, + "step": 1607 + }, + { + "epoch": 0.6157380815623205, + "grad_norm": 0.29421985149383545, + "learning_rate": 4.545336332776391e-06, + "loss": 0.6648, + "step": 1608 + }, + { + "epoch": 0.6161210032548344, + "grad_norm": 0.268839031457901, + "learning_rate": 4.544752140939553e-06, + "loss": 0.6481, + "step": 1609 + }, + { + "epoch": 0.6165039249473483, + "grad_norm": 0.2773914337158203, + "learning_rate": 4.5441676116255524e-06, + "loss": 0.6589, + "step": 1610 + }, + { + "epoch": 0.6168868466398622, + "grad_norm": 0.3153362572193146, + "learning_rate": 4.543582744930861e-06, + "loss": 0.6506, + "step": 1611 + }, + { + "epoch": 0.617269768332376, + "grad_norm": 0.28729212284088135, + "learning_rate": 4.54299754095201e-06, + "loss": 0.6676, + "step": 1612 + }, + { + "epoch": 0.6176526900248899, + "grad_norm": 0.2928646206855774, + "learning_rate": 4.542411999785584e-06, + "loss": 0.6375, + "step": 1613 + }, + { + "epoch": 0.6180356117174038, + "grad_norm": 0.28571853041648865, + "learning_rate": 4.541826121528225e-06, + "loss": 0.6607, + "step": 1614 + }, + { + "epoch": 0.6184185334099177, + "grad_norm": 0.2773991525173187, + "learning_rate": 4.541239906276627e-06, + "loss": 0.6351, + "step": 1615 + }, + { + "epoch": 0.6188014551024316, + "grad_norm": 0.28088879585266113, + "learning_rate": 4.540653354127544e-06, + "loss": 0.6465, + "step": 1616 + }, + { + "epoch": 0.6191843767949454, + "grad_norm": 0.288987934589386, + "learning_rate": 4.5400664651777835e-06, + "loss": 0.6516, + "step": 1617 + }, + { + "epoch": 0.6195672984874593, + "grad_norm": 0.28885790705680847, + "learning_rate": 4.539479239524209e-06, + "loss": 0.6507, + "step": 1618 + }, + { + "epoch": 0.6199502201799731, + "grad_norm": 0.3078453540802002, + "learning_rate": 4.538891677263739e-06, + "loss": 0.6939, + "step": 1619 + }, + { + "epoch": 0.6203331418724871, + "grad_norm": 0.28978171944618225, + "learning_rate": 4.5383037784933494e-06, + "loss": 0.6672, + "step": 1620 + }, + { + "epoch": 0.620716063565001, + "grad_norm": 0.28613409399986267, + "learning_rate": 4.537715543310068e-06, + "loss": 0.6395, + "step": 1621 + }, + { + "epoch": 0.6210989852575148, + "grad_norm": 0.27480873465538025, + "learning_rate": 4.537126971810982e-06, + "loss": 0.6531, + "step": 1622 + }, + { + "epoch": 0.6214819069500287, + "grad_norm": 0.29575228691101074, + "learning_rate": 4.5365380640932335e-06, + "loss": 0.6687, + "step": 1623 + }, + { + "epoch": 0.6218648286425426, + "grad_norm": 0.29107657074928284, + "learning_rate": 4.5359488202540156e-06, + "loss": 0.6475, + "step": 1624 + }, + { + "epoch": 0.6222477503350565, + "grad_norm": 0.2802061438560486, + "learning_rate": 4.535359240390584e-06, + "loss": 0.6574, + "step": 1625 + }, + { + "epoch": 0.6226306720275704, + "grad_norm": 0.30288583040237427, + "learning_rate": 4.534769324600243e-06, + "loss": 0.6612, + "step": 1626 + }, + { + "epoch": 0.6230135937200842, + "grad_norm": 0.27718931436538696, + "learning_rate": 4.534179072980359e-06, + "loss": 0.6539, + "step": 1627 + }, + { + "epoch": 0.6233965154125981, + "grad_norm": 0.2963017225265503, + "learning_rate": 4.533588485628347e-06, + "loss": 0.6723, + "step": 1628 + }, + { + "epoch": 0.623779437105112, + "grad_norm": 0.29171425104141235, + "learning_rate": 4.532997562641683e-06, + "loss": 0.6616, + "step": 1629 + }, + { + "epoch": 0.6241623587976259, + "grad_norm": 0.2799478769302368, + "learning_rate": 4.532406304117896e-06, + "loss": 0.6597, + "step": 1630 + }, + { + "epoch": 0.6245452804901398, + "grad_norm": 0.28800514340400696, + "learning_rate": 4.53181471015457e-06, + "loss": 0.6743, + "step": 1631 + }, + { + "epoch": 0.6249282021826537, + "grad_norm": 0.284427672624588, + "learning_rate": 4.5312227808493445e-06, + "loss": 0.6608, + "step": 1632 + }, + { + "epoch": 0.6253111238751675, + "grad_norm": 0.2893275022506714, + "learning_rate": 4.530630516299915e-06, + "loss": 0.668, + "step": 1633 + }, + { + "epoch": 0.6256940455676814, + "grad_norm": 0.2894516885280609, + "learning_rate": 4.530037916604033e-06, + "loss": 0.6694, + "step": 1634 + }, + { + "epoch": 0.6260769672601952, + "grad_norm": 0.2836352288722992, + "learning_rate": 4.5294449818595045e-06, + "loss": 0.6578, + "step": 1635 + }, + { + "epoch": 0.6264598889527092, + "grad_norm": 0.30977001786231995, + "learning_rate": 4.52885171216419e-06, + "loss": 0.6582, + "step": 1636 + }, + { + "epoch": 0.6268428106452231, + "grad_norm": 0.2887343764305115, + "learning_rate": 4.528258107616005e-06, + "loss": 0.6636, + "step": 1637 + }, + { + "epoch": 0.6272257323377369, + "grad_norm": 0.2849438190460205, + "learning_rate": 4.527664168312923e-06, + "loss": 0.642, + "step": 1638 + }, + { + "epoch": 0.6276086540302508, + "grad_norm": 0.28767022490501404, + "learning_rate": 4.52706989435297e-06, + "loss": 0.6772, + "step": 1639 + }, + { + "epoch": 0.6279915757227647, + "grad_norm": 0.30522409081459045, + "learning_rate": 4.526475285834229e-06, + "loss": 0.6664, + "step": 1640 + }, + { + "epoch": 0.6283744974152786, + "grad_norm": 0.2825944423675537, + "learning_rate": 4.525880342854836e-06, + "loss": 0.6559, + "step": 1641 + }, + { + "epoch": 0.6287574191077925, + "grad_norm": 0.289946049451828, + "learning_rate": 4.525285065512986e-06, + "loss": 0.6728, + "step": 1642 + }, + { + "epoch": 0.6291403408003063, + "grad_norm": 0.2922952473163605, + "learning_rate": 4.524689453906926e-06, + "loss": 0.6641, + "step": 1643 + }, + { + "epoch": 0.6295232624928202, + "grad_norm": 0.2942145764827728, + "learning_rate": 4.524093508134957e-06, + "loss": 0.6946, + "step": 1644 + }, + { + "epoch": 0.6299061841853341, + "grad_norm": 0.271501362323761, + "learning_rate": 4.52349722829544e-06, + "loss": 0.6785, + "step": 1645 + }, + { + "epoch": 0.630289105877848, + "grad_norm": 0.27138328552246094, + "learning_rate": 4.522900614486786e-06, + "loss": 0.653, + "step": 1646 + }, + { + "epoch": 0.6306720275703619, + "grad_norm": 0.2834060788154602, + "learning_rate": 4.522303666807465e-06, + "loss": 0.6615, + "step": 1647 + }, + { + "epoch": 0.6310549492628758, + "grad_norm": 0.28810012340545654, + "learning_rate": 4.521706385356e-06, + "loss": 0.6805, + "step": 1648 + }, + { + "epoch": 0.6314378709553896, + "grad_norm": 0.3068544864654541, + "learning_rate": 4.52110877023097e-06, + "loss": 0.6554, + "step": 1649 + }, + { + "epoch": 0.6318207926479035, + "grad_norm": 0.2793084979057312, + "learning_rate": 4.520510821531008e-06, + "loss": 0.6598, + "step": 1650 + }, + { + "epoch": 0.6322037143404173, + "grad_norm": 0.28801456093788147, + "learning_rate": 4.519912539354803e-06, + "loss": 0.655, + "step": 1651 + }, + { + "epoch": 0.6325866360329313, + "grad_norm": 0.28550705313682556, + "learning_rate": 4.5193139238011e-06, + "loss": 0.6442, + "step": 1652 + }, + { + "epoch": 0.6329695577254452, + "grad_norm": 0.2896331548690796, + "learning_rate": 4.518714974968696e-06, + "loss": 0.654, + "step": 1653 + }, + { + "epoch": 0.633352479417959, + "grad_norm": 0.2918282747268677, + "learning_rate": 4.518115692956445e-06, + "loss": 0.6489, + "step": 1654 + }, + { + "epoch": 0.6337354011104729, + "grad_norm": 0.2840275168418884, + "learning_rate": 4.517516077863256e-06, + "loss": 0.6819, + "step": 1655 + }, + { + "epoch": 0.6341183228029867, + "grad_norm": 0.28906315565109253, + "learning_rate": 4.516916129788095e-06, + "loss": 0.6513, + "step": 1656 + }, + { + "epoch": 0.6345012444955007, + "grad_norm": 0.281761109828949, + "learning_rate": 4.516315848829977e-06, + "loss": 0.6519, + "step": 1657 + }, + { + "epoch": 0.6348841661880146, + "grad_norm": 0.28358712792396545, + "learning_rate": 4.515715235087979e-06, + "loss": 0.6615, + "step": 1658 + }, + { + "epoch": 0.6352670878805284, + "grad_norm": 0.27831947803497314, + "learning_rate": 4.515114288661227e-06, + "loss": 0.6573, + "step": 1659 + }, + { + "epoch": 0.6356500095730423, + "grad_norm": 0.29940974712371826, + "learning_rate": 4.514513009648907e-06, + "loss": 0.6503, + "step": 1660 + }, + { + "epoch": 0.6360329312655562, + "grad_norm": 0.28754907846450806, + "learning_rate": 4.5139113981502545e-06, + "loss": 0.6643, + "step": 1661 + }, + { + "epoch": 0.6364158529580701, + "grad_norm": 0.2905157506465912, + "learning_rate": 4.513309454264566e-06, + "loss": 0.6751, + "step": 1662 + }, + { + "epoch": 0.636798774650584, + "grad_norm": 0.29055503010749817, + "learning_rate": 4.512707178091187e-06, + "loss": 0.6474, + "step": 1663 + }, + { + "epoch": 0.6371816963430978, + "grad_norm": 0.2892800569534302, + "learning_rate": 4.512104569729521e-06, + "loss": 0.6609, + "step": 1664 + }, + { + "epoch": 0.6375646180356117, + "grad_norm": 0.28348493576049805, + "learning_rate": 4.511501629279027e-06, + "loss": 0.6618, + "step": 1665 + }, + { + "epoch": 0.6379475397281256, + "grad_norm": 0.27873194217681885, + "learning_rate": 4.510898356839216e-06, + "loss": 0.6448, + "step": 1666 + }, + { + "epoch": 0.6383304614206394, + "grad_norm": 0.28761720657348633, + "learning_rate": 4.5102947525096575e-06, + "loss": 0.6746, + "step": 1667 + }, + { + "epoch": 0.6387133831131534, + "grad_norm": 0.3056875169277191, + "learning_rate": 4.509690816389971e-06, + "loss": 0.6671, + "step": 1668 + }, + { + "epoch": 0.6390963048056673, + "grad_norm": 0.27585890889167786, + "learning_rate": 4.509086548579836e-06, + "loss": 0.6602, + "step": 1669 + }, + { + "epoch": 0.6394792264981811, + "grad_norm": 0.2888655364513397, + "learning_rate": 4.508481949178983e-06, + "loss": 0.6606, + "step": 1670 + }, + { + "epoch": 0.639862148190695, + "grad_norm": 0.2705753445625305, + "learning_rate": 4.507877018287199e-06, + "loss": 0.6507, + "step": 1671 + }, + { + "epoch": 0.6402450698832088, + "grad_norm": 0.2835851013660431, + "learning_rate": 4.507271756004323e-06, + "loss": 0.6638, + "step": 1672 + }, + { + "epoch": 0.6406279915757228, + "grad_norm": 0.28667914867401123, + "learning_rate": 4.506666162430254e-06, + "loss": 0.6861, + "step": 1673 + }, + { + "epoch": 0.6410109132682367, + "grad_norm": 0.2852957844734192, + "learning_rate": 4.506060237664941e-06, + "loss": 0.6582, + "step": 1674 + }, + { + "epoch": 0.6413938349607505, + "grad_norm": 0.287737101316452, + "learning_rate": 4.505453981808389e-06, + "loss": 0.6758, + "step": 1675 + }, + { + "epoch": 0.6417767566532644, + "grad_norm": 0.2849019765853882, + "learning_rate": 4.504847394960657e-06, + "loss": 0.6753, + "step": 1676 + }, + { + "epoch": 0.6421596783457783, + "grad_norm": 0.28787752985954285, + "learning_rate": 4.504240477221861e-06, + "loss": 0.6547, + "step": 1677 + }, + { + "epoch": 0.6425426000382922, + "grad_norm": 0.2829058766365051, + "learning_rate": 4.503633228692169e-06, + "loss": 0.643, + "step": 1678 + }, + { + "epoch": 0.6429255217308061, + "grad_norm": 0.2887209951877594, + "learning_rate": 4.503025649471806e-06, + "loss": 0.6785, + "step": 1679 + }, + { + "epoch": 0.6433084434233199, + "grad_norm": 0.28281310200691223, + "learning_rate": 4.502417739661049e-06, + "loss": 0.6697, + "step": 1680 + }, + { + "epoch": 0.6436913651158338, + "grad_norm": 0.277035117149353, + "learning_rate": 4.50180949936023e-06, + "loss": 0.6516, + "step": 1681 + }, + { + "epoch": 0.6440742868083477, + "grad_norm": 0.28803032636642456, + "learning_rate": 4.501200928669739e-06, + "loss": 0.6869, + "step": 1682 + }, + { + "epoch": 0.6444572085008615, + "grad_norm": 0.27366191148757935, + "learning_rate": 4.500592027690015e-06, + "loss": 0.6537, + "step": 1683 + }, + { + "epoch": 0.6448401301933755, + "grad_norm": 0.2732759416103363, + "learning_rate": 4.499982796521556e-06, + "loss": 0.6628, + "step": 1684 + }, + { + "epoch": 0.6452230518858894, + "grad_norm": 0.28367921710014343, + "learning_rate": 4.499373235264913e-06, + "loss": 0.7002, + "step": 1685 + }, + { + "epoch": 0.6456059735784032, + "grad_norm": 0.2750464379787445, + "learning_rate": 4.4987633440206905e-06, + "loss": 0.6538, + "step": 1686 + }, + { + "epoch": 0.6459888952709171, + "grad_norm": 0.286153644323349, + "learning_rate": 4.498153122889549e-06, + "loss": 0.6687, + "step": 1687 + }, + { + "epoch": 0.6463718169634309, + "grad_norm": 0.28421804308891296, + "learning_rate": 4.4975425719722025e-06, + "loss": 0.659, + "step": 1688 + }, + { + "epoch": 0.6467547386559449, + "grad_norm": 0.2871510684490204, + "learning_rate": 4.496931691369419e-06, + "loss": 0.6614, + "step": 1689 + }, + { + "epoch": 0.6471376603484588, + "grad_norm": 0.2869611978530884, + "learning_rate": 4.496320481182023e-06, + "loss": 0.6705, + "step": 1690 + }, + { + "epoch": 0.6475205820409726, + "grad_norm": 0.28083720803260803, + "learning_rate": 4.49570894151089e-06, + "loss": 0.662, + "step": 1691 + }, + { + "epoch": 0.6479035037334865, + "grad_norm": 0.27947762608528137, + "learning_rate": 4.495097072456954e-06, + "loss": 0.6655, + "step": 1692 + }, + { + "epoch": 0.6482864254260003, + "grad_norm": 0.28797227144241333, + "learning_rate": 4.4944848741212e-06, + "loss": 0.659, + "step": 1693 + }, + { + "epoch": 0.6486693471185143, + "grad_norm": 0.2906720042228699, + "learning_rate": 4.4938723466046694e-06, + "loss": 0.6401, + "step": 1694 + }, + { + "epoch": 0.6490522688110282, + "grad_norm": 0.2948593199253082, + "learning_rate": 4.493259490008457e-06, + "loss": 0.6663, + "step": 1695 + }, + { + "epoch": 0.649435190503542, + "grad_norm": 0.29138094186782837, + "learning_rate": 4.492646304433711e-06, + "loss": 0.6919, + "step": 1696 + }, + { + "epoch": 0.6498181121960559, + "grad_norm": 0.28296101093292236, + "learning_rate": 4.4920327899816356e-06, + "loss": 0.6407, + "step": 1697 + }, + { + "epoch": 0.6502010338885698, + "grad_norm": 0.2952142357826233, + "learning_rate": 4.491418946753489e-06, + "loss": 0.6761, + "step": 1698 + }, + { + "epoch": 0.6505839555810836, + "grad_norm": 0.2904972434043884, + "learning_rate": 4.490804774850582e-06, + "loss": 0.6652, + "step": 1699 + }, + { + "epoch": 0.6509668772735976, + "grad_norm": 0.29449400305747986, + "learning_rate": 4.4901902743742826e-06, + "loss": 0.6747, + "step": 1700 + }, + { + "epoch": 0.6513497989661114, + "grad_norm": 0.2727620303630829, + "learning_rate": 4.48957544542601e-06, + "loss": 0.6389, + "step": 1701 + }, + { + "epoch": 0.6517327206586253, + "grad_norm": 0.29242339730262756, + "learning_rate": 4.48896028810724e-06, + "loss": 0.6538, + "step": 1702 + }, + { + "epoch": 0.6521156423511392, + "grad_norm": 0.2878548502922058, + "learning_rate": 4.488344802519501e-06, + "loss": 0.6608, + "step": 1703 + }, + { + "epoch": 0.652498564043653, + "grad_norm": 0.2867409884929657, + "learning_rate": 4.487728988764377e-06, + "loss": 0.6515, + "step": 1704 + }, + { + "epoch": 0.652881485736167, + "grad_norm": 0.28430137038230896, + "learning_rate": 4.487112846943503e-06, + "loss": 0.6466, + "step": 1705 + }, + { + "epoch": 0.6532644074286809, + "grad_norm": 0.28649213910102844, + "learning_rate": 4.486496377158572e-06, + "loss": 0.6495, + "step": 1706 + }, + { + "epoch": 0.6536473291211947, + "grad_norm": 0.2940782308578491, + "learning_rate": 4.485879579511331e-06, + "loss": 0.6508, + "step": 1707 + }, + { + "epoch": 0.6540302508137086, + "grad_norm": 0.28731769323349, + "learning_rate": 4.485262454103577e-06, + "loss": 0.6755, + "step": 1708 + }, + { + "epoch": 0.6544131725062224, + "grad_norm": 0.28093746304512024, + "learning_rate": 4.4846450010371656e-06, + "loss": 0.6552, + "step": 1709 + }, + { + "epoch": 0.6547960941987364, + "grad_norm": 0.2853204905986786, + "learning_rate": 4.484027220414003e-06, + "loss": 0.6542, + "step": 1710 + }, + { + "epoch": 0.6551790158912503, + "grad_norm": 0.28559935092926025, + "learning_rate": 4.483409112336054e-06, + "loss": 0.6569, + "step": 1711 + }, + { + "epoch": 0.6555619375837641, + "grad_norm": 0.2773648500442505, + "learning_rate": 4.482790676905331e-06, + "loss": 0.6666, + "step": 1712 + }, + { + "epoch": 0.655944859276278, + "grad_norm": 0.28084635734558105, + "learning_rate": 4.482171914223906e-06, + "loss": 0.665, + "step": 1713 + }, + { + "epoch": 0.6563277809687919, + "grad_norm": 0.28929758071899414, + "learning_rate": 4.481552824393903e-06, + "loss": 0.6717, + "step": 1714 + }, + { + "epoch": 0.6567107026613057, + "grad_norm": 0.28908833861351013, + "learning_rate": 4.4809334075175e-06, + "loss": 0.6682, + "step": 1715 + }, + { + "epoch": 0.6570936243538197, + "grad_norm": 0.28981640934944153, + "learning_rate": 4.480313663696928e-06, + "loss": 0.6601, + "step": 1716 + }, + { + "epoch": 0.6574765460463335, + "grad_norm": 0.30396929383277893, + "learning_rate": 4.479693593034475e-06, + "loss": 0.6716, + "step": 1717 + }, + { + "epoch": 0.6578594677388474, + "grad_norm": 0.27771109342575073, + "learning_rate": 4.479073195632478e-06, + "loss": 0.642, + "step": 1718 + }, + { + "epoch": 0.6582423894313613, + "grad_norm": 0.305408239364624, + "learning_rate": 4.478452471593333e-06, + "loss": 0.652, + "step": 1719 + }, + { + "epoch": 0.6586253111238751, + "grad_norm": 0.28755393624305725, + "learning_rate": 4.477831421019486e-06, + "loss": 0.6782, + "step": 1720 + }, + { + "epoch": 0.6590082328163891, + "grad_norm": 0.2777462303638458, + "learning_rate": 4.477210044013441e-06, + "loss": 0.6638, + "step": 1721 + }, + { + "epoch": 0.659391154508903, + "grad_norm": 0.2919204533100128, + "learning_rate": 4.476588340677751e-06, + "loss": 0.6705, + "step": 1722 + }, + { + "epoch": 0.6597740762014168, + "grad_norm": 0.2931578457355499, + "learning_rate": 4.475966311115027e-06, + "loss": 0.663, + "step": 1723 + }, + { + "epoch": 0.6601569978939307, + "grad_norm": 0.28356921672821045, + "learning_rate": 4.475343955427932e-06, + "loss": 0.6698, + "step": 1724 + }, + { + "epoch": 0.6605399195864445, + "grad_norm": 0.2742938995361328, + "learning_rate": 4.474721273719183e-06, + "loss": 0.6678, + "step": 1725 + }, + { + "epoch": 0.6609228412789585, + "grad_norm": 0.2960493862628937, + "learning_rate": 4.4740982660915485e-06, + "loss": 0.6545, + "step": 1726 + }, + { + "epoch": 0.6613057629714724, + "grad_norm": 0.2709237337112427, + "learning_rate": 4.473474932647856e-06, + "loss": 0.663, + "step": 1727 + }, + { + "epoch": 0.6616886846639862, + "grad_norm": 0.2795345187187195, + "learning_rate": 4.472851273490985e-06, + "loss": 0.6517, + "step": 1728 + }, + { + "epoch": 0.6620716063565001, + "grad_norm": 0.2873730957508087, + "learning_rate": 4.472227288723863e-06, + "loss": 0.6635, + "step": 1729 + }, + { + "epoch": 0.662454528049014, + "grad_norm": 0.2867303788661957, + "learning_rate": 4.47160297844948e-06, + "loss": 0.6524, + "step": 1730 + }, + { + "epoch": 0.6628374497415278, + "grad_norm": 0.2772071361541748, + "learning_rate": 4.470978342770873e-06, + "loss": 0.6543, + "step": 1731 + }, + { + "epoch": 0.6632203714340418, + "grad_norm": 0.2774072587490082, + "learning_rate": 4.470353381791138e-06, + "loss": 0.6521, + "step": 1732 + }, + { + "epoch": 0.6636032931265556, + "grad_norm": 0.285992294549942, + "learning_rate": 4.469728095613419e-06, + "loss": 0.666, + "step": 1733 + }, + { + "epoch": 0.6639862148190695, + "grad_norm": 0.2970382571220398, + "learning_rate": 4.469102484340919e-06, + "loss": 0.6812, + "step": 1734 + }, + { + "epoch": 0.6643691365115834, + "grad_norm": 0.28140783309936523, + "learning_rate": 4.468476548076891e-06, + "loss": 0.6624, + "step": 1735 + }, + { + "epoch": 0.6647520582040972, + "grad_norm": 0.28314921259880066, + "learning_rate": 4.4678502869246435e-06, + "loss": 0.649, + "step": 1736 + }, + { + "epoch": 0.6651349798966112, + "grad_norm": 0.2802729308605194, + "learning_rate": 4.467223700987538e-06, + "loss": 0.6639, + "step": 1737 + }, + { + "epoch": 0.665517901589125, + "grad_norm": 0.2866319417953491, + "learning_rate": 4.46659679036899e-06, + "loss": 0.6548, + "step": 1738 + }, + { + "epoch": 0.6659008232816389, + "grad_norm": 0.2863485515117645, + "learning_rate": 4.465969555172468e-06, + "loss": 0.6637, + "step": 1739 + }, + { + "epoch": 0.6662837449741528, + "grad_norm": 0.2826560437679291, + "learning_rate": 4.465341995501493e-06, + "loss": 0.6571, + "step": 1740 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.2898571789264679, + "learning_rate": 4.464714111459644e-06, + "loss": 0.6683, + "step": 1741 + }, + { + "epoch": 0.6670495883591806, + "grad_norm": 0.29692378640174866, + "learning_rate": 4.464085903150548e-06, + "loss": 0.6669, + "step": 1742 + }, + { + "epoch": 0.6674325100516945, + "grad_norm": 0.2766851782798767, + "learning_rate": 4.463457370677889e-06, + "loss": 0.6526, + "step": 1743 + }, + { + "epoch": 0.6678154317442083, + "grad_norm": 0.2955440580844879, + "learning_rate": 4.462828514145402e-06, + "loss": 0.6662, + "step": 1744 + }, + { + "epoch": 0.6681983534367222, + "grad_norm": 0.3091248571872711, + "learning_rate": 4.462199333656879e-06, + "loss": 0.6773, + "step": 1745 + }, + { + "epoch": 0.668581275129236, + "grad_norm": 0.29684820771217346, + "learning_rate": 4.461569829316163e-06, + "loss": 0.6738, + "step": 1746 + }, + { + "epoch": 0.6689641968217499, + "grad_norm": 0.28915321826934814, + "learning_rate": 4.46094000122715e-06, + "loss": 0.6582, + "step": 1747 + }, + { + "epoch": 0.6693471185142639, + "grad_norm": 0.2869669198989868, + "learning_rate": 4.4603098494937905e-06, + "loss": 0.6486, + "step": 1748 + }, + { + "epoch": 0.6697300402067777, + "grad_norm": 0.28078657388687134, + "learning_rate": 4.4596793742200886e-06, + "loss": 0.6572, + "step": 1749 + }, + { + "epoch": 0.6701129618992916, + "grad_norm": 0.275255024433136, + "learning_rate": 4.4590485755101e-06, + "loss": 0.6426, + "step": 1750 + }, + { + "epoch": 0.6704958835918055, + "grad_norm": 0.2922937273979187, + "learning_rate": 4.4584174534679375e-06, + "loss": 0.6681, + "step": 1751 + }, + { + "epoch": 0.6708788052843193, + "grad_norm": 0.2934999167919159, + "learning_rate": 4.4577860081977635e-06, + "loss": 0.6581, + "step": 1752 + }, + { + "epoch": 0.6712617269768333, + "grad_norm": 0.29109132289886475, + "learning_rate": 4.457154239803796e-06, + "loss": 0.6644, + "step": 1753 + }, + { + "epoch": 0.6716446486693471, + "grad_norm": 0.28353652358055115, + "learning_rate": 4.456522148390304e-06, + "loss": 0.6681, + "step": 1754 + }, + { + "epoch": 0.672027570361861, + "grad_norm": 0.28220492601394653, + "learning_rate": 4.455889734061613e-06, + "loss": 0.6493, + "step": 1755 + }, + { + "epoch": 0.6724104920543749, + "grad_norm": 0.2920376658439636, + "learning_rate": 4.4552569969221e-06, + "loss": 0.6581, + "step": 1756 + }, + { + "epoch": 0.6727934137468887, + "grad_norm": 0.3047402799129486, + "learning_rate": 4.454623937076195e-06, + "loss": 0.6537, + "step": 1757 + }, + { + "epoch": 0.6731763354394027, + "grad_norm": 0.29887130856513977, + "learning_rate": 4.453990554628382e-06, + "loss": 0.6724, + "step": 1758 + }, + { + "epoch": 0.6735592571319166, + "grad_norm": 0.29401645064353943, + "learning_rate": 4.4533568496831965e-06, + "loss": 0.6733, + "step": 1759 + }, + { + "epoch": 0.6739421788244304, + "grad_norm": 0.2885599136352539, + "learning_rate": 4.452722822345231e-06, + "loss": 0.6693, + "step": 1760 + }, + { + "epoch": 0.6743251005169443, + "grad_norm": 0.28618887066841125, + "learning_rate": 4.452088472719127e-06, + "loss": 0.6704, + "step": 1761 + }, + { + "epoch": 0.6747080222094581, + "grad_norm": 0.2804790437221527, + "learning_rate": 4.451453800909582e-06, + "loss": 0.6591, + "step": 1762 + }, + { + "epoch": 0.675090943901972, + "grad_norm": 0.29983004927635193, + "learning_rate": 4.450818807021345e-06, + "loss": 0.6763, + "step": 1763 + }, + { + "epoch": 0.675473865594486, + "grad_norm": 0.2811990976333618, + "learning_rate": 4.45018349115922e-06, + "loss": 0.6881, + "step": 1764 + }, + { + "epoch": 0.6758567872869998, + "grad_norm": 0.2975471317768097, + "learning_rate": 4.449547853428061e-06, + "loss": 0.6758, + "step": 1765 + }, + { + "epoch": 0.6762397089795137, + "grad_norm": 0.2989833652973175, + "learning_rate": 4.4489118939327795e-06, + "loss": 0.6724, + "step": 1766 + }, + { + "epoch": 0.6766226306720275, + "grad_norm": 0.2850721776485443, + "learning_rate": 4.448275612778336e-06, + "loss": 0.6687, + "step": 1767 + }, + { + "epoch": 0.6770055523645414, + "grad_norm": 0.2916548550128937, + "learning_rate": 4.447639010069748e-06, + "loss": 0.6719, + "step": 1768 + }, + { + "epoch": 0.6773884740570554, + "grad_norm": 0.28292280435562134, + "learning_rate": 4.447002085912082e-06, + "loss": 0.6601, + "step": 1769 + }, + { + "epoch": 0.6777713957495692, + "grad_norm": 0.2910638153553009, + "learning_rate": 4.44636484041046e-06, + "loss": 0.6704, + "step": 1770 + }, + { + "epoch": 0.6781543174420831, + "grad_norm": 0.2982378900051117, + "learning_rate": 4.445727273670057e-06, + "loss": 0.6322, + "step": 1771 + }, + { + "epoch": 0.678537239134597, + "grad_norm": 0.2990197539329529, + "learning_rate": 4.445089385796099e-06, + "loss": 0.6653, + "step": 1772 + }, + { + "epoch": 0.6789201608271108, + "grad_norm": 0.2974478304386139, + "learning_rate": 4.444451176893868e-06, + "loss": 0.6725, + "step": 1773 + }, + { + "epoch": 0.6793030825196248, + "grad_norm": 0.2938443720340729, + "learning_rate": 4.443812647068697e-06, + "loss": 0.65, + "step": 1774 + }, + { + "epoch": 0.6796860042121387, + "grad_norm": 0.28330492973327637, + "learning_rate": 4.443173796425973e-06, + "loss": 0.6685, + "step": 1775 + }, + { + "epoch": 0.6800689259046525, + "grad_norm": 0.28686612844467163, + "learning_rate": 4.4425346250711355e-06, + "loss": 0.647, + "step": 1776 + }, + { + "epoch": 0.6804518475971664, + "grad_norm": 0.28971579670906067, + "learning_rate": 4.441895133109675e-06, + "loss": 0.669, + "step": 1777 + }, + { + "epoch": 0.6808347692896802, + "grad_norm": 0.2774415910243988, + "learning_rate": 4.441255320647139e-06, + "loss": 0.6576, + "step": 1778 + }, + { + "epoch": 0.6812176909821941, + "grad_norm": 0.2871474027633667, + "learning_rate": 4.440615187789124e-06, + "loss": 0.6644, + "step": 1779 + }, + { + "epoch": 0.6816006126747081, + "grad_norm": 0.28074896335601807, + "learning_rate": 4.4399747346412834e-06, + "loss": 0.6573, + "step": 1780 + }, + { + "epoch": 0.6819835343672219, + "grad_norm": 0.2852190434932709, + "learning_rate": 4.439333961309319e-06, + "loss": 0.6534, + "step": 1781 + }, + { + "epoch": 0.6823664560597358, + "grad_norm": 0.3023935556411743, + "learning_rate": 4.438692867898988e-06, + "loss": 0.6626, + "step": 1782 + }, + { + "epoch": 0.6827493777522496, + "grad_norm": 0.29996219277381897, + "learning_rate": 4.4380514545161e-06, + "loss": 0.6467, + "step": 1783 + }, + { + "epoch": 0.6831322994447635, + "grad_norm": 0.2838921546936035, + "learning_rate": 4.437409721266518e-06, + "loss": 0.665, + "step": 1784 + }, + { + "epoch": 0.6835152211372775, + "grad_norm": 0.2941323220729828, + "learning_rate": 4.436767668256157e-06, + "loss": 0.6541, + "step": 1785 + }, + { + "epoch": 0.6838981428297913, + "grad_norm": 0.2942572832107544, + "learning_rate": 4.436125295590984e-06, + "loss": 0.6671, + "step": 1786 + }, + { + "epoch": 0.6842810645223052, + "grad_norm": 0.28804951906204224, + "learning_rate": 4.43548260337702e-06, + "loss": 0.6837, + "step": 1787 + }, + { + "epoch": 0.684663986214819, + "grad_norm": 0.28608110547065735, + "learning_rate": 4.43483959172034e-06, + "loss": 0.6569, + "step": 1788 + }, + { + "epoch": 0.6850469079073329, + "grad_norm": 0.2884553074836731, + "learning_rate": 4.434196260727069e-06, + "loss": 0.6325, + "step": 1789 + }, + { + "epoch": 0.6854298295998469, + "grad_norm": 0.28959155082702637, + "learning_rate": 4.433552610503386e-06, + "loss": 0.6596, + "step": 1790 + }, + { + "epoch": 0.6858127512923607, + "grad_norm": 0.28355011343955994, + "learning_rate": 4.432908641155522e-06, + "loss": 0.6754, + "step": 1791 + }, + { + "epoch": 0.6861956729848746, + "grad_norm": 0.30101996660232544, + "learning_rate": 4.432264352789763e-06, + "loss": 0.6523, + "step": 1792 + }, + { + "epoch": 0.6865785946773885, + "grad_norm": 0.2937653064727783, + "learning_rate": 4.431619745512443e-06, + "loss": 0.6745, + "step": 1793 + }, + { + "epoch": 0.6869615163699023, + "grad_norm": 0.2968035340309143, + "learning_rate": 4.430974819429954e-06, + "loss": 0.6867, + "step": 1794 + }, + { + "epoch": 0.6873444380624162, + "grad_norm": 0.28310543298721313, + "learning_rate": 4.430329574648737e-06, + "loss": 0.6451, + "step": 1795 + }, + { + "epoch": 0.6877273597549302, + "grad_norm": 0.2911510467529297, + "learning_rate": 4.429684011275288e-06, + "loss": 0.669, + "step": 1796 + }, + { + "epoch": 0.688110281447444, + "grad_norm": 0.29060813784599304, + "learning_rate": 4.429038129416153e-06, + "loss": 0.6443, + "step": 1797 + }, + { + "epoch": 0.6884932031399579, + "grad_norm": 0.2903493642807007, + "learning_rate": 4.428391929177933e-06, + "loss": 0.6735, + "step": 1798 + }, + { + "epoch": 0.6888761248324717, + "grad_norm": 0.2872551381587982, + "learning_rate": 4.42774541066728e-06, + "loss": 0.6419, + "step": 1799 + }, + { + "epoch": 0.6892590465249856, + "grad_norm": 0.27953702211380005, + "learning_rate": 4.427098573990897e-06, + "loss": 0.6544, + "step": 1800 + }, + { + "epoch": 0.6896419682174996, + "grad_norm": 0.2751811444759369, + "learning_rate": 4.4264514192555456e-06, + "loss": 0.6413, + "step": 1801 + }, + { + "epoch": 0.6900248899100134, + "grad_norm": 0.280831515789032, + "learning_rate": 4.425803946568033e-06, + "loss": 0.6744, + "step": 1802 + }, + { + "epoch": 0.6904078116025273, + "grad_norm": 0.2886102497577667, + "learning_rate": 4.4251561560352215e-06, + "loss": 0.6643, + "step": 1803 + }, + { + "epoch": 0.6907907332950411, + "grad_norm": 0.27799156308174133, + "learning_rate": 4.424508047764027e-06, + "loss": 0.6584, + "step": 1804 + }, + { + "epoch": 0.691173654987555, + "grad_norm": 0.28802868723869324, + "learning_rate": 4.423859621861417e-06, + "loss": 0.6397, + "step": 1805 + }, + { + "epoch": 0.691556576680069, + "grad_norm": 0.2986770570278168, + "learning_rate": 4.42321087843441e-06, + "loss": 0.6442, + "step": 1806 + }, + { + "epoch": 0.6919394983725828, + "grad_norm": 0.2985818088054657, + "learning_rate": 4.42256181759008e-06, + "loss": 0.6433, + "step": 1807 + }, + { + "epoch": 0.6923224200650967, + "grad_norm": 0.2903055250644684, + "learning_rate": 4.421912439435549e-06, + "loss": 0.6487, + "step": 1808 + }, + { + "epoch": 0.6927053417576106, + "grad_norm": 0.28338488936424255, + "learning_rate": 4.421262744077997e-06, + "loss": 0.6602, + "step": 1809 + }, + { + "epoch": 0.6930882634501244, + "grad_norm": 0.2961269021034241, + "learning_rate": 4.420612731624652e-06, + "loss": 0.6797, + "step": 1810 + }, + { + "epoch": 0.6934711851426383, + "grad_norm": 0.2834930121898651, + "learning_rate": 4.4199624021827954e-06, + "loss": 0.6852, + "step": 1811 + }, + { + "epoch": 0.6938541068351523, + "grad_norm": 0.2932717204093933, + "learning_rate": 4.419311755859761e-06, + "loss": 0.6547, + "step": 1812 + }, + { + "epoch": 0.6942370285276661, + "grad_norm": 0.2781160771846771, + "learning_rate": 4.418660792762935e-06, + "loss": 0.6625, + "step": 1813 + }, + { + "epoch": 0.69461995022018, + "grad_norm": 0.2849212884902954, + "learning_rate": 4.418009512999757e-06, + "loss": 0.6531, + "step": 1814 + }, + { + "epoch": 0.6950028719126938, + "grad_norm": 0.2946004867553711, + "learning_rate": 4.417357916677716e-06, + "loss": 0.6722, + "step": 1815 + }, + { + "epoch": 0.6953857936052077, + "grad_norm": 0.2866877317428589, + "learning_rate": 4.416706003904357e-06, + "loss": 0.6451, + "step": 1816 + }, + { + "epoch": 0.6957687152977217, + "grad_norm": 0.29456812143325806, + "learning_rate": 4.416053774787274e-06, + "loss": 0.6667, + "step": 1817 + }, + { + "epoch": 0.6961516369902355, + "grad_norm": 0.30627280473709106, + "learning_rate": 4.415401229434115e-06, + "loss": 0.6727, + "step": 1818 + }, + { + "epoch": 0.6965345586827494, + "grad_norm": 0.2828829288482666, + "learning_rate": 4.414748367952579e-06, + "loss": 0.6611, + "step": 1819 + }, + { + "epoch": 0.6969174803752632, + "grad_norm": 0.28239619731903076, + "learning_rate": 4.4140951904504194e-06, + "loss": 0.6568, + "step": 1820 + }, + { + "epoch": 0.6973004020677771, + "grad_norm": 0.32551002502441406, + "learning_rate": 4.413441697035439e-06, + "loss": 0.6598, + "step": 1821 + }, + { + "epoch": 0.6976833237602911, + "grad_norm": 0.281393438577652, + "learning_rate": 4.412787887815494e-06, + "loss": 0.6662, + "step": 1822 + }, + { + "epoch": 0.6980662454528049, + "grad_norm": 0.288491427898407, + "learning_rate": 4.412133762898493e-06, + "loss": 0.6675, + "step": 1823 + }, + { + "epoch": 0.6984491671453188, + "grad_norm": 0.289437472820282, + "learning_rate": 4.411479322392396e-06, + "loss": 0.6629, + "step": 1824 + }, + { + "epoch": 0.6988320888378327, + "grad_norm": 0.2796823978424072, + "learning_rate": 4.410824566405215e-06, + "loss": 0.6668, + "step": 1825 + }, + { + "epoch": 0.6992150105303465, + "grad_norm": 0.29910945892333984, + "learning_rate": 4.410169495045017e-06, + "loss": 0.6625, + "step": 1826 + }, + { + "epoch": 0.6995979322228604, + "grad_norm": 0.30837470293045044, + "learning_rate": 4.4095141084199155e-06, + "loss": 0.6756, + "step": 1827 + }, + { + "epoch": 0.6999808539153743, + "grad_norm": 0.29551079869270325, + "learning_rate": 4.4088584066380804e-06, + "loss": 0.654, + "step": 1828 + }, + { + "epoch": 0.7003637756078882, + "grad_norm": 0.293451726436615, + "learning_rate": 4.408202389807733e-06, + "loss": 0.6499, + "step": 1829 + }, + { + "epoch": 0.7007466973004021, + "grad_norm": 0.3022894263267517, + "learning_rate": 4.4075460580371455e-06, + "loss": 0.6477, + "step": 1830 + }, + { + "epoch": 0.7011296189929159, + "grad_norm": 0.29170361161231995, + "learning_rate": 4.406889411434641e-06, + "loss": 0.6666, + "step": 1831 + }, + { + "epoch": 0.7015125406854298, + "grad_norm": 0.27581146359443665, + "learning_rate": 4.4062324501086e-06, + "loss": 0.6501, + "step": 1832 + }, + { + "epoch": 0.7018954623779438, + "grad_norm": 0.3292374014854431, + "learning_rate": 4.4055751741674454e-06, + "loss": 0.6692, + "step": 1833 + }, + { + "epoch": 0.7022783840704576, + "grad_norm": 0.28566980361938477, + "learning_rate": 4.404917583719663e-06, + "loss": 0.6558, + "step": 1834 + }, + { + "epoch": 0.7026613057629715, + "grad_norm": 0.2983189523220062, + "learning_rate": 4.404259678873782e-06, + "loss": 0.6573, + "step": 1835 + }, + { + "epoch": 0.7030442274554853, + "grad_norm": 0.28240758180618286, + "learning_rate": 4.403601459738387e-06, + "loss": 0.6597, + "step": 1836 + }, + { + "epoch": 0.7034271491479992, + "grad_norm": 0.2784506380558014, + "learning_rate": 4.402942926422115e-06, + "loss": 0.6627, + "step": 1837 + }, + { + "epoch": 0.7038100708405132, + "grad_norm": 0.28692924976348877, + "learning_rate": 4.402284079033654e-06, + "loss": 0.6611, + "step": 1838 + }, + { + "epoch": 0.704192992533027, + "grad_norm": 0.28517448902130127, + "learning_rate": 4.401624917681743e-06, + "loss": 0.6547, + "step": 1839 + }, + { + "epoch": 0.7045759142255409, + "grad_norm": 0.28843966126441956, + "learning_rate": 4.400965442475173e-06, + "loss": 0.6786, + "step": 1840 + }, + { + "epoch": 0.7049588359180547, + "grad_norm": 0.2878416180610657, + "learning_rate": 4.4003056535227905e-06, + "loss": 0.6551, + "step": 1841 + }, + { + "epoch": 0.7053417576105686, + "grad_norm": 0.2929476797580719, + "learning_rate": 4.3996455509334875e-06, + "loss": 0.6737, + "step": 1842 + }, + { + "epoch": 0.7057246793030825, + "grad_norm": 0.317920058965683, + "learning_rate": 4.398985134816211e-06, + "loss": 0.6606, + "step": 1843 + }, + { + "epoch": 0.7061076009955964, + "grad_norm": 0.2867056131362915, + "learning_rate": 4.398324405279963e-06, + "loss": 0.6522, + "step": 1844 + }, + { + "epoch": 0.7064905226881103, + "grad_norm": 0.3179545998573303, + "learning_rate": 4.39766336243379e-06, + "loss": 0.6614, + "step": 1845 + }, + { + "epoch": 0.7068734443806242, + "grad_norm": 0.27060508728027344, + "learning_rate": 4.397002006386797e-06, + "loss": 0.6424, + "step": 1846 + }, + { + "epoch": 0.707256366073138, + "grad_norm": 0.2941526770591736, + "learning_rate": 4.3963403372481364e-06, + "loss": 0.6826, + "step": 1847 + }, + { + "epoch": 0.7076392877656519, + "grad_norm": 0.28471407294273376, + "learning_rate": 4.395678355127013e-06, + "loss": 0.6682, + "step": 1848 + }, + { + "epoch": 0.7080222094581659, + "grad_norm": 0.2784045338630676, + "learning_rate": 4.3950160601326865e-06, + "loss": 0.6473, + "step": 1849 + }, + { + "epoch": 0.7084051311506797, + "grad_norm": 0.29154446721076965, + "learning_rate": 4.394353452374464e-06, + "loss": 0.6697, + "step": 1850 + }, + { + "epoch": 0.7087880528431936, + "grad_norm": 0.3087870180606842, + "learning_rate": 4.3936905319617056e-06, + "loss": 0.6606, + "step": 1851 + }, + { + "epoch": 0.7091709745357074, + "grad_norm": 0.290182888507843, + "learning_rate": 4.3930272990038245e-06, + "loss": 0.6679, + "step": 1852 + }, + { + "epoch": 0.7095538962282213, + "grad_norm": 0.2789461612701416, + "learning_rate": 4.392363753610284e-06, + "loss": 0.6768, + "step": 1853 + }, + { + "epoch": 0.7099368179207353, + "grad_norm": 0.2986856698989868, + "learning_rate": 4.3916998958905975e-06, + "loss": 0.6807, + "step": 1854 + }, + { + "epoch": 0.7103197396132491, + "grad_norm": 0.31243690848350525, + "learning_rate": 4.391035725954335e-06, + "loss": 0.6561, + "step": 1855 + }, + { + "epoch": 0.710702661305763, + "grad_norm": 0.28528085350990295, + "learning_rate": 4.390371243911112e-06, + "loss": 0.6574, + "step": 1856 + }, + { + "epoch": 0.7110855829982768, + "grad_norm": 0.29143473505973816, + "learning_rate": 4.3897064498706e-06, + "loss": 0.6511, + "step": 1857 + }, + { + "epoch": 0.7114685046907907, + "grad_norm": 0.29057714343070984, + "learning_rate": 4.389041343942519e-06, + "loss": 0.6707, + "step": 1858 + }, + { + "epoch": 0.7118514263833046, + "grad_norm": 0.2966049611568451, + "learning_rate": 4.3883759262366435e-06, + "loss": 0.6547, + "step": 1859 + }, + { + "epoch": 0.7122343480758185, + "grad_norm": 0.283877432346344, + "learning_rate": 4.387710196862796e-06, + "loss": 0.6563, + "step": 1860 + }, + { + "epoch": 0.7126172697683324, + "grad_norm": 0.28457963466644287, + "learning_rate": 4.387044155930852e-06, + "loss": 0.6565, + "step": 1861 + }, + { + "epoch": 0.7130001914608463, + "grad_norm": 0.3141806721687317, + "learning_rate": 4.38637780355074e-06, + "loss": 0.6428, + "step": 1862 + }, + { + "epoch": 0.7133831131533601, + "grad_norm": 0.2856174409389496, + "learning_rate": 4.385711139832437e-06, + "loss": 0.6543, + "step": 1863 + }, + { + "epoch": 0.713766034845874, + "grad_norm": 0.2827475666999817, + "learning_rate": 4.3850441648859745e-06, + "loss": 0.6714, + "step": 1864 + }, + { + "epoch": 0.714148956538388, + "grad_norm": 0.32685548067092896, + "learning_rate": 4.384376878821432e-06, + "loss": 0.6511, + "step": 1865 + }, + { + "epoch": 0.7145318782309018, + "grad_norm": 0.3012218177318573, + "learning_rate": 4.383709281748943e-06, + "loss": 0.6505, + "step": 1866 + }, + { + "epoch": 0.7149147999234157, + "grad_norm": 0.28398776054382324, + "learning_rate": 4.383041373778691e-06, + "loss": 0.6643, + "step": 1867 + }, + { + "epoch": 0.7152977216159295, + "grad_norm": 0.30807650089263916, + "learning_rate": 4.382373155020913e-06, + "loss": 0.6628, + "step": 1868 + }, + { + "epoch": 0.7156806433084434, + "grad_norm": 0.30395251512527466, + "learning_rate": 4.381704625585893e-06, + "loss": 0.6634, + "step": 1869 + }, + { + "epoch": 0.7160635650009574, + "grad_norm": 0.2886289060115814, + "learning_rate": 4.3810357855839705e-06, + "loss": 0.647, + "step": 1870 + }, + { + "epoch": 0.7164464866934712, + "grad_norm": 0.2849857211112976, + "learning_rate": 4.380366635125533e-06, + "loss": 0.6356, + "step": 1871 + }, + { + "epoch": 0.7168294083859851, + "grad_norm": 0.3216630816459656, + "learning_rate": 4.379697174321021e-06, + "loss": 0.6569, + "step": 1872 + }, + { + "epoch": 0.7172123300784989, + "grad_norm": 0.3093796372413635, + "learning_rate": 4.379027403280927e-06, + "loss": 0.6558, + "step": 1873 + }, + { + "epoch": 0.7175952517710128, + "grad_norm": 0.30553147196769714, + "learning_rate": 4.378357322115795e-06, + "loss": 0.6695, + "step": 1874 + }, + { + "epoch": 0.7179781734635267, + "grad_norm": 0.3528289496898651, + "learning_rate": 4.377686930936216e-06, + "loss": 0.6471, + "step": 1875 + }, + { + "epoch": 0.7183610951560406, + "grad_norm": 0.3141665756702423, + "learning_rate": 4.377016229852836e-06, + "loss": 0.6499, + "step": 1876 + }, + { + "epoch": 0.7187440168485545, + "grad_norm": 0.30015069246292114, + "learning_rate": 4.3763452189763514e-06, + "loss": 0.6564, + "step": 1877 + }, + { + "epoch": 0.7191269385410684, + "grad_norm": 0.3349354565143585, + "learning_rate": 4.3756738984175104e-06, + "loss": 0.6503, + "step": 1878 + }, + { + "epoch": 0.7195098602335822, + "grad_norm": 0.338704913854599, + "learning_rate": 4.375002268287109e-06, + "loss": 0.6371, + "step": 1879 + }, + { + "epoch": 0.7198927819260961, + "grad_norm": 0.3038701117038727, + "learning_rate": 4.374330328696001e-06, + "loss": 0.6472, + "step": 1880 + }, + { + "epoch": 0.72027570361861, + "grad_norm": 0.30665767192840576, + "learning_rate": 4.373658079755082e-06, + "loss": 0.6557, + "step": 1881 + }, + { + "epoch": 0.7206586253111239, + "grad_norm": 0.36826714873313904, + "learning_rate": 4.372985521575307e-06, + "loss": 0.6397, + "step": 1882 + }, + { + "epoch": 0.7210415470036378, + "grad_norm": 0.2825036942958832, + "learning_rate": 4.372312654267678e-06, + "loss": 0.6817, + "step": 1883 + }, + { + "epoch": 0.7214244686961516, + "grad_norm": 0.3002195656299591, + "learning_rate": 4.371639477943248e-06, + "loss": 0.6813, + "step": 1884 + }, + { + "epoch": 0.7218073903886655, + "grad_norm": 0.315328449010849, + "learning_rate": 4.370965992713123e-06, + "loss": 0.668, + "step": 1885 + }, + { + "epoch": 0.7221903120811795, + "grad_norm": 0.326954185962677, + "learning_rate": 4.3702921986884576e-06, + "loss": 0.6608, + "step": 1886 + }, + { + "epoch": 0.7225732337736933, + "grad_norm": 0.30041590332984924, + "learning_rate": 4.369618095980458e-06, + "loss": 0.6568, + "step": 1887 + }, + { + "epoch": 0.7229561554662072, + "grad_norm": 0.28943854570388794, + "learning_rate": 4.368943684700384e-06, + "loss": 0.6685, + "step": 1888 + }, + { + "epoch": 0.723339077158721, + "grad_norm": 0.3098449110984802, + "learning_rate": 4.368268964959542e-06, + "loss": 0.6878, + "step": 1889 + }, + { + "epoch": 0.7237219988512349, + "grad_norm": 0.34474292397499084, + "learning_rate": 4.3675939368692925e-06, + "loss": 0.6672, + "step": 1890 + }, + { + "epoch": 0.7241049205437488, + "grad_norm": 0.28440067172050476, + "learning_rate": 4.366918600541045e-06, + "loss": 0.666, + "step": 1891 + }, + { + "epoch": 0.7244878422362627, + "grad_norm": 0.29537713527679443, + "learning_rate": 4.366242956086262e-06, + "loss": 0.6772, + "step": 1892 + }, + { + "epoch": 0.7248707639287766, + "grad_norm": 0.3174888491630554, + "learning_rate": 4.365567003616455e-06, + "loss": 0.6373, + "step": 1893 + }, + { + "epoch": 0.7252536856212904, + "grad_norm": 0.31100913882255554, + "learning_rate": 4.364890743243187e-06, + "loss": 0.6606, + "step": 1894 + }, + { + "epoch": 0.7256366073138043, + "grad_norm": 0.29556766152381897, + "learning_rate": 4.364214175078071e-06, + "loss": 0.6574, + "step": 1895 + }, + { + "epoch": 0.7260195290063182, + "grad_norm": 0.2948823869228363, + "learning_rate": 4.363537299232773e-06, + "loss": 0.6535, + "step": 1896 + }, + { + "epoch": 0.7264024506988321, + "grad_norm": 0.3056114912033081, + "learning_rate": 4.362860115819007e-06, + "loss": 0.6618, + "step": 1897 + }, + { + "epoch": 0.726785372391346, + "grad_norm": 0.30672791600227356, + "learning_rate": 4.36218262494854e-06, + "loss": 0.6685, + "step": 1898 + }, + { + "epoch": 0.7271682940838599, + "grad_norm": 0.2834531366825104, + "learning_rate": 4.3615048267331875e-06, + "loss": 0.6522, + "step": 1899 + }, + { + "epoch": 0.7275512157763737, + "grad_norm": 0.28408098220825195, + "learning_rate": 4.360826721284818e-06, + "loss": 0.624, + "step": 1900 + }, + { + "epoch": 0.7279341374688876, + "grad_norm": 0.2917988896369934, + "learning_rate": 4.360148308715349e-06, + "loss": 0.6591, + "step": 1901 + }, + { + "epoch": 0.7283170591614015, + "grad_norm": 0.2883460223674774, + "learning_rate": 4.359469589136752e-06, + "loss": 0.6486, + "step": 1902 + }, + { + "epoch": 0.7286999808539154, + "grad_norm": 0.284496933221817, + "learning_rate": 4.358790562661044e-06, + "loss": 0.6544, + "step": 1903 + }, + { + "epoch": 0.7290829025464293, + "grad_norm": 0.29910603165626526, + "learning_rate": 4.358111229400296e-06, + "loss": 0.65, + "step": 1904 + }, + { + "epoch": 0.7294658242389431, + "grad_norm": 0.29271793365478516, + "learning_rate": 4.357431589466629e-06, + "loss": 0.6547, + "step": 1905 + }, + { + "epoch": 0.729848745931457, + "grad_norm": 0.2830300033092499, + "learning_rate": 4.356751642972214e-06, + "loss": 0.6682, + "step": 1906 + }, + { + "epoch": 0.7302316676239708, + "grad_norm": 0.2894238829612732, + "learning_rate": 4.356071390029275e-06, + "loss": 0.6732, + "step": 1907 + }, + { + "epoch": 0.7306145893164848, + "grad_norm": 0.2848636209964752, + "learning_rate": 4.355390830750081e-06, + "loss": 0.6644, + "step": 1908 + }, + { + "epoch": 0.7309975110089987, + "grad_norm": 0.27900969982147217, + "learning_rate": 4.35470996524696e-06, + "loss": 0.6324, + "step": 1909 + }, + { + "epoch": 0.7313804327015125, + "grad_norm": 0.2817847430706024, + "learning_rate": 4.354028793632281e-06, + "loss": 0.6676, + "step": 1910 + }, + { + "epoch": 0.7317633543940264, + "grad_norm": 0.28212547302246094, + "learning_rate": 4.3533473160184725e-06, + "loss": 0.6704, + "step": 1911 + }, + { + "epoch": 0.7321462760865403, + "grad_norm": 0.28725746273994446, + "learning_rate": 4.352665532518007e-06, + "loss": 0.6529, + "step": 1912 + }, + { + "epoch": 0.7325291977790542, + "grad_norm": 0.297559916973114, + "learning_rate": 4.3519834432434095e-06, + "loss": 0.6707, + "step": 1913 + }, + { + "epoch": 0.7329121194715681, + "grad_norm": 0.2973473072052002, + "learning_rate": 4.351301048307257e-06, + "loss": 0.6776, + "step": 1914 + }, + { + "epoch": 0.733295041164082, + "grad_norm": 0.2814374566078186, + "learning_rate": 4.350618347822175e-06, + "loss": 0.6393, + "step": 1915 + }, + { + "epoch": 0.7336779628565958, + "grad_norm": 0.2936658263206482, + "learning_rate": 4.349935341900841e-06, + "loss": 0.6767, + "step": 1916 + }, + { + "epoch": 0.7340608845491097, + "grad_norm": 0.29175832867622375, + "learning_rate": 4.349252030655982e-06, + "loss": 0.6656, + "step": 1917 + }, + { + "epoch": 0.7344438062416236, + "grad_norm": 0.2962329387664795, + "learning_rate": 4.3485684142003734e-06, + "loss": 0.6584, + "step": 1918 + }, + { + "epoch": 0.7348267279341375, + "grad_norm": 0.2833195924758911, + "learning_rate": 4.347884492646846e-06, + "loss": 0.6468, + "step": 1919 + }, + { + "epoch": 0.7352096496266514, + "grad_norm": 0.2913295030593872, + "learning_rate": 4.347200266108275e-06, + "loss": 0.6416, + "step": 1920 + }, + { + "epoch": 0.7355925713191652, + "grad_norm": 0.29444143176078796, + "learning_rate": 4.346515734697592e-06, + "loss": 0.6737, + "step": 1921 + }, + { + "epoch": 0.7359754930116791, + "grad_norm": 0.2952899932861328, + "learning_rate": 4.345830898527773e-06, + "loss": 0.656, + "step": 1922 + }, + { + "epoch": 0.7363584147041929, + "grad_norm": 0.2749316990375519, + "learning_rate": 4.345145757711849e-06, + "loss": 0.629, + "step": 1923 + }, + { + "epoch": 0.7367413363967069, + "grad_norm": 0.2893662750720978, + "learning_rate": 4.344460312362899e-06, + "loss": 0.6492, + "step": 1924 + }, + { + "epoch": 0.7371242580892208, + "grad_norm": 0.28754866123199463, + "learning_rate": 4.343774562594053e-06, + "loss": 0.6572, + "step": 1925 + }, + { + "epoch": 0.7375071797817346, + "grad_norm": 0.2948099970817566, + "learning_rate": 4.343088508518489e-06, + "loss": 0.6643, + "step": 1926 + }, + { + "epoch": 0.7378901014742485, + "grad_norm": 0.28945931792259216, + "learning_rate": 4.34240215024944e-06, + "loss": 0.6538, + "step": 1927 + }, + { + "epoch": 0.7382730231667624, + "grad_norm": 0.29446089267730713, + "learning_rate": 4.3417154879001845e-06, + "loss": 0.6541, + "step": 1928 + }, + { + "epoch": 0.7386559448592763, + "grad_norm": 0.28317776322364807, + "learning_rate": 4.341028521584053e-06, + "loss": 0.6502, + "step": 1929 + }, + { + "epoch": 0.7390388665517902, + "grad_norm": 0.28600841760635376, + "learning_rate": 4.340341251414427e-06, + "loss": 0.6562, + "step": 1930 + }, + { + "epoch": 0.739421788244304, + "grad_norm": 0.2847815155982971, + "learning_rate": 4.339653677504737e-06, + "loss": 0.6656, + "step": 1931 + }, + { + "epoch": 0.7398047099368179, + "grad_norm": 0.29913073778152466, + "learning_rate": 4.338965799968465e-06, + "loss": 0.6739, + "step": 1932 + }, + { + "epoch": 0.7401876316293318, + "grad_norm": 0.276906818151474, + "learning_rate": 4.33827761891914e-06, + "loss": 0.6686, + "step": 1933 + }, + { + "epoch": 0.7405705533218457, + "grad_norm": 0.27966615557670593, + "learning_rate": 4.337589134470345e-06, + "loss": 0.6497, + "step": 1934 + }, + { + "epoch": 0.7409534750143596, + "grad_norm": 0.28722241520881653, + "learning_rate": 4.33690034673571e-06, + "loss": 0.6552, + "step": 1935 + }, + { + "epoch": 0.7413363967068735, + "grad_norm": 0.28933677077293396, + "learning_rate": 4.336211255828917e-06, + "loss": 0.6609, + "step": 1936 + }, + { + "epoch": 0.7417193183993873, + "grad_norm": 0.2806028723716736, + "learning_rate": 4.335521861863698e-06, + "loss": 0.646, + "step": 1937 + }, + { + "epoch": 0.7421022400919012, + "grad_norm": 0.2815571129322052, + "learning_rate": 4.334832164953833e-06, + "loss": 0.6587, + "step": 1938 + }, + { + "epoch": 0.742485161784415, + "grad_norm": 0.279063880443573, + "learning_rate": 4.334142165213154e-06, + "loss": 0.6603, + "step": 1939 + }, + { + "epoch": 0.742868083476929, + "grad_norm": 0.28541961312294006, + "learning_rate": 4.333451862755543e-06, + "loss": 0.6857, + "step": 1940 + }, + { + "epoch": 0.7432510051694429, + "grad_norm": 0.2963019907474518, + "learning_rate": 4.332761257694931e-06, + "loss": 0.6595, + "step": 1941 + }, + { + "epoch": 0.7436339268619567, + "grad_norm": 0.27836063504219055, + "learning_rate": 4.332070350145297e-06, + "loss": 0.6561, + "step": 1942 + }, + { + "epoch": 0.7440168485544706, + "grad_norm": 0.28990232944488525, + "learning_rate": 4.3313791402206765e-06, + "loss": 0.6476, + "step": 1943 + }, + { + "epoch": 0.7443997702469844, + "grad_norm": 0.29161038994789124, + "learning_rate": 4.330687628035147e-06, + "loss": 0.6857, + "step": 1944 + }, + { + "epoch": 0.7447826919394984, + "grad_norm": 0.291381299495697, + "learning_rate": 4.329995813702842e-06, + "loss": 0.6229, + "step": 1945 + }, + { + "epoch": 0.7451656136320123, + "grad_norm": 0.28754791617393494, + "learning_rate": 4.32930369733794e-06, + "loss": 0.6511, + "step": 1946 + }, + { + "epoch": 0.7455485353245261, + "grad_norm": 0.28677695989608765, + "learning_rate": 4.328611279054673e-06, + "loss": 0.6583, + "step": 1947 + }, + { + "epoch": 0.74593145701704, + "grad_norm": 0.29700377583503723, + "learning_rate": 4.327918558967321e-06, + "loss": 0.6356, + "step": 1948 + }, + { + "epoch": 0.7463143787095539, + "grad_norm": 0.2861420214176178, + "learning_rate": 4.327225537190215e-06, + "loss": 0.6599, + "step": 1949 + }, + { + "epoch": 0.7466973004020678, + "grad_norm": 0.29210418462753296, + "learning_rate": 4.326532213837735e-06, + "loss": 0.6394, + "step": 1950 + }, + { + "epoch": 0.7470802220945817, + "grad_norm": 0.28130602836608887, + "learning_rate": 4.325838589024312e-06, + "loss": 0.6691, + "step": 1951 + }, + { + "epoch": 0.7474631437870956, + "grad_norm": 0.28725385665893555, + "learning_rate": 4.3251446628644224e-06, + "loss": 0.6825, + "step": 1952 + }, + { + "epoch": 0.7478460654796094, + "grad_norm": 0.29483655095100403, + "learning_rate": 4.324450435472598e-06, + "loss": 0.661, + "step": 1953 + }, + { + "epoch": 0.7482289871721233, + "grad_norm": 0.2880934476852417, + "learning_rate": 4.323755906963419e-06, + "loss": 0.6563, + "step": 1954 + }, + { + "epoch": 0.7486119088646371, + "grad_norm": 0.29493528604507446, + "learning_rate": 4.32306107745151e-06, + "loss": 0.6526, + "step": 1955 + }, + { + "epoch": 0.7489948305571511, + "grad_norm": 0.2936811149120331, + "learning_rate": 4.322365947051554e-06, + "loss": 0.6359, + "step": 1956 + }, + { + "epoch": 0.749377752249665, + "grad_norm": 0.282199889421463, + "learning_rate": 4.321670515878277e-06, + "loss": 0.6715, + "step": 1957 + }, + { + "epoch": 0.7497606739421788, + "grad_norm": 0.2862069606781006, + "learning_rate": 4.320974784046458e-06, + "loss": 0.6765, + "step": 1958 + }, + { + "epoch": 0.7501435956346927, + "grad_norm": 0.2906152904033661, + "learning_rate": 4.320278751670922e-06, + "loss": 0.6715, + "step": 1959 + }, + { + "epoch": 0.7505265173272065, + "grad_norm": 0.2825862169265747, + "learning_rate": 4.319582418866549e-06, + "loss": 0.6532, + "step": 1960 + }, + { + "epoch": 0.7509094390197205, + "grad_norm": 0.28862181305885315, + "learning_rate": 4.318885785748264e-06, + "loss": 0.6526, + "step": 1961 + }, + { + "epoch": 0.7512923607122344, + "grad_norm": 0.28897467255592346, + "learning_rate": 4.318188852431043e-06, + "loss": 0.648, + "step": 1962 + }, + { + "epoch": 0.7516752824047482, + "grad_norm": 0.28821879625320435, + "learning_rate": 4.317491619029911e-06, + "loss": 0.6303, + "step": 1963 + }, + { + "epoch": 0.7520582040972621, + "grad_norm": 0.290291428565979, + "learning_rate": 4.316794085659946e-06, + "loss": 0.6506, + "step": 1964 + }, + { + "epoch": 0.752441125789776, + "grad_norm": 0.29138660430908203, + "learning_rate": 4.316096252436271e-06, + "loss": 0.6614, + "step": 1965 + }, + { + "epoch": 0.7528240474822899, + "grad_norm": 0.3001135587692261, + "learning_rate": 4.315398119474059e-06, + "loss": 0.653, + "step": 1966 + }, + { + "epoch": 0.7532069691748038, + "grad_norm": 0.29022935032844543, + "learning_rate": 4.314699686888536e-06, + "loss": 0.6664, + "step": 1967 + }, + { + "epoch": 0.7535898908673176, + "grad_norm": 0.30858826637268066, + "learning_rate": 4.314000954794974e-06, + "loss": 0.6643, + "step": 1968 + }, + { + "epoch": 0.7539728125598315, + "grad_norm": 0.29346123337745667, + "learning_rate": 4.313301923308696e-06, + "loss": 0.6671, + "step": 1969 + }, + { + "epoch": 0.7543557342523454, + "grad_norm": 0.28771576285362244, + "learning_rate": 4.312602592545073e-06, + "loss": 0.6453, + "step": 1970 + }, + { + "epoch": 0.7547386559448592, + "grad_norm": 0.28616222739219666, + "learning_rate": 4.311902962619529e-06, + "loss": 0.6666, + "step": 1971 + }, + { + "epoch": 0.7551215776373732, + "grad_norm": 0.2930017113685608, + "learning_rate": 4.311203033647532e-06, + "loss": 0.6458, + "step": 1972 + }, + { + "epoch": 0.7555044993298871, + "grad_norm": 0.2986257076263428, + "learning_rate": 4.310502805744603e-06, + "loss": 0.6402, + "step": 1973 + }, + { + "epoch": 0.7558874210224009, + "grad_norm": 0.29714515805244446, + "learning_rate": 4.3098022790263115e-06, + "loss": 0.6673, + "step": 1974 + }, + { + "epoch": 0.7562703427149148, + "grad_norm": 0.2817387580871582, + "learning_rate": 4.309101453608276e-06, + "loss": 0.6604, + "step": 1975 + }, + { + "epoch": 0.7566532644074286, + "grad_norm": 0.29210686683654785, + "learning_rate": 4.3084003296061655e-06, + "loss": 0.6518, + "step": 1976 + }, + { + "epoch": 0.7570361860999426, + "grad_norm": 0.29889750480651855, + "learning_rate": 4.307698907135697e-06, + "loss": 0.6293, + "step": 1977 + }, + { + "epoch": 0.7574191077924565, + "grad_norm": 0.28968730568885803, + "learning_rate": 4.306997186312637e-06, + "loss": 0.662, + "step": 1978 + }, + { + "epoch": 0.7578020294849703, + "grad_norm": 0.29846274852752686, + "learning_rate": 4.306295167252801e-06, + "loss": 0.6515, + "step": 1979 + }, + { + "epoch": 0.7581849511774842, + "grad_norm": 0.3046862483024597, + "learning_rate": 4.3055928500720546e-06, + "loss": 0.656, + "step": 1980 + }, + { + "epoch": 0.758567872869998, + "grad_norm": 0.2888549566268921, + "learning_rate": 4.3048902348863116e-06, + "loss": 0.6594, + "step": 1981 + }, + { + "epoch": 0.758950794562512, + "grad_norm": 0.29400166869163513, + "learning_rate": 4.304187321811536e-06, + "loss": 0.6515, + "step": 1982 + }, + { + "epoch": 0.7593337162550259, + "grad_norm": 0.28630971908569336, + "learning_rate": 4.30348411096374e-06, + "loss": 0.6469, + "step": 1983 + }, + { + "epoch": 0.7597166379475397, + "grad_norm": 0.2840445935726166, + "learning_rate": 4.302780602458987e-06, + "loss": 0.6764, + "step": 1984 + }, + { + "epoch": 0.7600995596400536, + "grad_norm": 0.2886156141757965, + "learning_rate": 4.302076796413385e-06, + "loss": 0.6644, + "step": 1985 + }, + { + "epoch": 0.7604824813325675, + "grad_norm": 0.2923414409160614, + "learning_rate": 4.301372692943096e-06, + "loss": 0.6501, + "step": 1986 + }, + { + "epoch": 0.7608654030250813, + "grad_norm": 0.2898506820201874, + "learning_rate": 4.300668292164329e-06, + "loss": 0.6712, + "step": 1987 + }, + { + "epoch": 0.7612483247175953, + "grad_norm": 0.2883935868740082, + "learning_rate": 4.299963594193342e-06, + "loss": 0.6675, + "step": 1988 + }, + { + "epoch": 0.7616312464101092, + "grad_norm": 0.2911122739315033, + "learning_rate": 4.2992585991464434e-06, + "loss": 0.666, + "step": 1989 + }, + { + "epoch": 0.762014168102623, + "grad_norm": 0.28685012459754944, + "learning_rate": 4.2985533071399865e-06, + "loss": 0.6582, + "step": 1990 + }, + { + "epoch": 0.7623970897951369, + "grad_norm": 0.28739434480667114, + "learning_rate": 4.297847718290378e-06, + "loss": 0.6596, + "step": 1991 + }, + { + "epoch": 0.7627800114876507, + "grad_norm": 0.2994686961174011, + "learning_rate": 4.297141832714073e-06, + "loss": 0.644, + "step": 1992 + }, + { + "epoch": 0.7631629331801647, + "grad_norm": 0.31029248237609863, + "learning_rate": 4.296435650527575e-06, + "loss": 0.6405, + "step": 1993 + }, + { + "epoch": 0.7635458548726786, + "grad_norm": 0.2902706563472748, + "learning_rate": 4.2957291718474356e-06, + "loss": 0.6474, + "step": 1994 + }, + { + "epoch": 0.7639287765651924, + "grad_norm": 0.2979045808315277, + "learning_rate": 4.2950223967902545e-06, + "loss": 0.6488, + "step": 1995 + }, + { + "epoch": 0.7643116982577063, + "grad_norm": 0.3013862371444702, + "learning_rate": 4.2943153254726835e-06, + "loss": 0.6656, + "step": 1996 + }, + { + "epoch": 0.7646946199502201, + "grad_norm": 0.30337440967559814, + "learning_rate": 4.2936079580114215e-06, + "loss": 0.6534, + "step": 1997 + }, + { + "epoch": 0.7650775416427341, + "grad_norm": 0.3107244074344635, + "learning_rate": 4.292900294523216e-06, + "loss": 0.6734, + "step": 1998 + }, + { + "epoch": 0.765460463335248, + "grad_norm": 0.29626262187957764, + "learning_rate": 4.292192335124863e-06, + "loss": 0.6505, + "step": 1999 + }, + { + "epoch": 0.7658433850277618, + "grad_norm": 0.28887301683425903, + "learning_rate": 4.291484079933208e-06, + "loss": 0.6537, + "step": 2000 + }, + { + "epoch": 0.7662263067202757, + "grad_norm": 0.28143438696861267, + "learning_rate": 4.290775529065146e-06, + "loss": 0.6745, + "step": 2001 + }, + { + "epoch": 0.7666092284127896, + "grad_norm": 0.2881353199481964, + "learning_rate": 4.290066682637622e-06, + "loss": 0.6453, + "step": 2002 + }, + { + "epoch": 0.7669921501053034, + "grad_norm": 0.2832963168621063, + "learning_rate": 4.289357540767625e-06, + "loss": 0.652, + "step": 2003 + }, + { + "epoch": 0.7673750717978174, + "grad_norm": 0.29034432768821716, + "learning_rate": 4.288648103572196e-06, + "loss": 0.6567, + "step": 2004 + }, + { + "epoch": 0.7677579934903312, + "grad_norm": 0.28968292474746704, + "learning_rate": 4.287938371168425e-06, + "loss": 0.6488, + "step": 2005 + }, + { + "epoch": 0.7681409151828451, + "grad_norm": 0.3006720244884491, + "learning_rate": 4.287228343673451e-06, + "loss": 0.6778, + "step": 2006 + }, + { + "epoch": 0.768523836875359, + "grad_norm": 0.28195542097091675, + "learning_rate": 4.28651802120446e-06, + "loss": 0.6434, + "step": 2007 + }, + { + "epoch": 0.7689067585678728, + "grad_norm": 0.2804199457168579, + "learning_rate": 4.285807403878687e-06, + "loss": 0.6645, + "step": 2008 + }, + { + "epoch": 0.7692896802603868, + "grad_norm": 0.2874583899974823, + "learning_rate": 4.2850964918134175e-06, + "loss": 0.6535, + "step": 2009 + }, + { + "epoch": 0.7696726019529007, + "grad_norm": 0.29446691274642944, + "learning_rate": 4.284385285125983e-06, + "loss": 0.6552, + "step": 2010 + }, + { + "epoch": 0.7700555236454145, + "grad_norm": 0.28751301765441895, + "learning_rate": 4.283673783933767e-06, + "loss": 0.6768, + "step": 2011 + }, + { + "epoch": 0.7704384453379284, + "grad_norm": 0.30472174286842346, + "learning_rate": 4.282961988354197e-06, + "loss": 0.6275, + "step": 2012 + }, + { + "epoch": 0.7708213670304422, + "grad_norm": 0.2874471843242645, + "learning_rate": 4.2822498985047535e-06, + "loss": 0.6547, + "step": 2013 + }, + { + "epoch": 0.7712042887229562, + "grad_norm": 0.2831214666366577, + "learning_rate": 4.281537514502962e-06, + "loss": 0.6268, + "step": 2014 + }, + { + "epoch": 0.7715872104154701, + "grad_norm": 0.28203102946281433, + "learning_rate": 4.280824836466401e-06, + "loss": 0.6709, + "step": 2015 + }, + { + "epoch": 0.7719701321079839, + "grad_norm": 0.2814365029335022, + "learning_rate": 4.280111864512693e-06, + "loss": 0.6597, + "step": 2016 + }, + { + "epoch": 0.7723530538004978, + "grad_norm": 0.29994678497314453, + "learning_rate": 4.279398598759511e-06, + "loss": 0.6542, + "step": 2017 + }, + { + "epoch": 0.7727359754930117, + "grad_norm": 0.2828892767429352, + "learning_rate": 4.278685039324576e-06, + "loss": 0.644, + "step": 2018 + }, + { + "epoch": 0.7731188971855255, + "grad_norm": 0.2878169119358063, + "learning_rate": 4.277971186325658e-06, + "loss": 0.6564, + "step": 2019 + }, + { + "epoch": 0.7735018188780395, + "grad_norm": 0.28668612241744995, + "learning_rate": 4.277257039880576e-06, + "loss": 0.6468, + "step": 2020 + }, + { + "epoch": 0.7738847405705533, + "grad_norm": 0.2916262149810791, + "learning_rate": 4.276542600107195e-06, + "loss": 0.6558, + "step": 2021 + }, + { + "epoch": 0.7742676622630672, + "grad_norm": 0.29401564598083496, + "learning_rate": 4.2758278671234315e-06, + "loss": 0.6591, + "step": 2022 + }, + { + "epoch": 0.7746505839555811, + "grad_norm": 0.3051813244819641, + "learning_rate": 4.275112841047249e-06, + "loss": 0.6466, + "step": 2023 + }, + { + "epoch": 0.7750335056480949, + "grad_norm": 0.2905217707157135, + "learning_rate": 4.274397521996658e-06, + "loss": 0.6694, + "step": 2024 + }, + { + "epoch": 0.7754164273406089, + "grad_norm": 0.2871151864528656, + "learning_rate": 4.273681910089721e-06, + "loss": 0.6376, + "step": 2025 + }, + { + "epoch": 0.7757993490331228, + "grad_norm": 0.28331542015075684, + "learning_rate": 4.272966005444545e-06, + "loss": 0.6481, + "step": 2026 + }, + { + "epoch": 0.7761822707256366, + "grad_norm": 0.28714999556541443, + "learning_rate": 4.272249808179286e-06, + "loss": 0.6762, + "step": 2027 + }, + { + "epoch": 0.7765651924181505, + "grad_norm": 0.29239681363105774, + "learning_rate": 4.271533318412151e-06, + "loss": 0.6519, + "step": 2028 + }, + { + "epoch": 0.7769481141106643, + "grad_norm": 0.28685858845710754, + "learning_rate": 4.2708165362613915e-06, + "loss": 0.6599, + "step": 2029 + }, + { + "epoch": 0.7773310358031783, + "grad_norm": 0.30236196517944336, + "learning_rate": 4.270099461845311e-06, + "loss": 0.6368, + "step": 2030 + }, + { + "epoch": 0.7777139574956922, + "grad_norm": 0.2790868282318115, + "learning_rate": 4.2693820952822594e-06, + "loss": 0.6438, + "step": 2031 + }, + { + "epoch": 0.778096879188206, + "grad_norm": 0.2869567275047302, + "learning_rate": 4.268664436690633e-06, + "loss": 0.6599, + "step": 2032 + }, + { + "epoch": 0.7784798008807199, + "grad_norm": 0.29027190804481506, + "learning_rate": 4.2679464861888796e-06, + "loss": 0.6568, + "step": 2033 + }, + { + "epoch": 0.7788627225732337, + "grad_norm": 0.28916051983833313, + "learning_rate": 4.267228243895494e-06, + "loss": 0.6513, + "step": 2034 + }, + { + "epoch": 0.7792456442657476, + "grad_norm": 0.2791849672794342, + "learning_rate": 4.266509709929017e-06, + "loss": 0.6332, + "step": 2035 + }, + { + "epoch": 0.7796285659582616, + "grad_norm": 0.2916945219039917, + "learning_rate": 4.265790884408042e-06, + "loss": 0.6468, + "step": 2036 + }, + { + "epoch": 0.7800114876507754, + "grad_norm": 0.30323725938796997, + "learning_rate": 4.265071767451206e-06, + "loss": 0.6513, + "step": 2037 + }, + { + "epoch": 0.7803944093432893, + "grad_norm": 0.2838743031024933, + "learning_rate": 4.264352359177196e-06, + "loss": 0.6575, + "step": 2038 + }, + { + "epoch": 0.7807773310358032, + "grad_norm": 0.2878491282463074, + "learning_rate": 4.263632659704748e-06, + "loss": 0.6424, + "step": 2039 + }, + { + "epoch": 0.781160252728317, + "grad_norm": 0.29993826150894165, + "learning_rate": 4.262912669152644e-06, + "loss": 0.6447, + "step": 2040 + }, + { + "epoch": 0.781543174420831, + "grad_norm": 0.29164037108421326, + "learning_rate": 4.2621923876397175e-06, + "loss": 0.6704, + "step": 2041 + }, + { + "epoch": 0.7819260961133448, + "grad_norm": 0.27584871649742126, + "learning_rate": 4.261471815284846e-06, + "loss": 0.6518, + "step": 2042 + }, + { + "epoch": 0.7823090178058587, + "grad_norm": 0.2810426354408264, + "learning_rate": 4.260750952206957e-06, + "loss": 0.6738, + "step": 2043 + }, + { + "epoch": 0.7826919394983726, + "grad_norm": 0.2827095687389374, + "learning_rate": 4.260029798525025e-06, + "loss": 0.6357, + "step": 2044 + }, + { + "epoch": 0.7830748611908864, + "grad_norm": 0.3064935505390167, + "learning_rate": 4.2593083543580735e-06, + "loss": 0.6453, + "step": 2045 + }, + { + "epoch": 0.7834577828834004, + "grad_norm": 0.2863493263721466, + "learning_rate": 4.258586619825176e-06, + "loss": 0.66, + "step": 2046 + }, + { + "epoch": 0.7838407045759143, + "grad_norm": 0.294557124376297, + "learning_rate": 4.2578645950454475e-06, + "loss": 0.6747, + "step": 2047 + }, + { + "epoch": 0.7842236262684281, + "grad_norm": 0.2933705747127533, + "learning_rate": 4.257142280138058e-06, + "loss": 0.6598, + "step": 2048 + }, + { + "epoch": 0.784606547960942, + "grad_norm": 0.2848224639892578, + "learning_rate": 4.256419675222222e-06, + "loss": 0.66, + "step": 2049 + }, + { + "epoch": 0.7849894696534558, + "grad_norm": 0.28834375739097595, + "learning_rate": 4.255696780417201e-06, + "loss": 0.6373, + "step": 2050 + }, + { + "epoch": 0.7853723913459697, + "grad_norm": 0.28696176409721375, + "learning_rate": 4.2549735958423065e-06, + "loss": 0.6574, + "step": 2051 + }, + { + "epoch": 0.7857553130384837, + "grad_norm": 0.2905980944633484, + "learning_rate": 4.254250121616895e-06, + "loss": 0.635, + "step": 2052 + }, + { + "epoch": 0.7861382347309975, + "grad_norm": 0.2829233407974243, + "learning_rate": 4.253526357860376e-06, + "loss": 0.6393, + "step": 2053 + }, + { + "epoch": 0.7865211564235114, + "grad_norm": 0.2907569408416748, + "learning_rate": 4.2528023046922005e-06, + "loss": 0.6564, + "step": 2054 + }, + { + "epoch": 0.7869040781160253, + "grad_norm": 0.27671337127685547, + "learning_rate": 4.252077962231872e-06, + "loss": 0.6582, + "step": 2055 + }, + { + "epoch": 0.7872869998085391, + "grad_norm": 0.2857098877429962, + "learning_rate": 4.251353330598938e-06, + "loss": 0.6529, + "step": 2056 + }, + { + "epoch": 0.7876699215010531, + "grad_norm": 0.28680211305618286, + "learning_rate": 4.250628409912998e-06, + "loss": 0.6568, + "step": 2057 + }, + { + "epoch": 0.7880528431935669, + "grad_norm": 0.28168386220932007, + "learning_rate": 4.2499032002936955e-06, + "loss": 0.6642, + "step": 2058 + }, + { + "epoch": 0.7884357648860808, + "grad_norm": 0.27457278966903687, + "learning_rate": 4.249177701860724e-06, + "loss": 0.6267, + "step": 2059 + }, + { + "epoch": 0.7888186865785947, + "grad_norm": 0.2845628261566162, + "learning_rate": 4.2484519147338235e-06, + "loss": 0.6384, + "step": 2060 + }, + { + "epoch": 0.7892016082711085, + "grad_norm": 0.28178882598876953, + "learning_rate": 4.247725839032781e-06, + "loss": 0.6455, + "step": 2061 + }, + { + "epoch": 0.7895845299636225, + "grad_norm": 0.29785293340682983, + "learning_rate": 4.246999474877433e-06, + "loss": 0.6516, + "step": 2062 + }, + { + "epoch": 0.7899674516561364, + "grad_norm": 0.30443844199180603, + "learning_rate": 4.246272822387662e-06, + "loss": 0.6749, + "step": 2063 + }, + { + "epoch": 0.7903503733486502, + "grad_norm": 0.2878495454788208, + "learning_rate": 4.245545881683399e-06, + "loss": 0.6285, + "step": 2064 + }, + { + "epoch": 0.7907332950411641, + "grad_norm": 0.280362069606781, + "learning_rate": 4.244818652884623e-06, + "loss": 0.6376, + "step": 2065 + }, + { + "epoch": 0.7911162167336779, + "grad_norm": 0.2950364351272583, + "learning_rate": 4.244091136111359e-06, + "loss": 0.6311, + "step": 2066 + }, + { + "epoch": 0.7914991384261918, + "grad_norm": 0.2914651036262512, + "learning_rate": 4.243363331483681e-06, + "loss": 0.6459, + "step": 2067 + }, + { + "epoch": 0.7918820601187058, + "grad_norm": 0.28748568892478943, + "learning_rate": 4.242635239121709e-06, + "loss": 0.6404, + "step": 2068 + }, + { + "epoch": 0.7922649818112196, + "grad_norm": 0.29014062881469727, + "learning_rate": 4.241906859145611e-06, + "loss": 0.6436, + "step": 2069 + }, + { + "epoch": 0.7926479035037335, + "grad_norm": 0.29581183195114136, + "learning_rate": 4.241178191675606e-06, + "loss": 0.6611, + "step": 2070 + }, + { + "epoch": 0.7930308251962473, + "grad_norm": 0.285098671913147, + "learning_rate": 4.240449236831954e-06, + "loss": 0.6412, + "step": 2071 + }, + { + "epoch": 0.7934137468887612, + "grad_norm": 0.2723717987537384, + "learning_rate": 4.239719994734965e-06, + "loss": 0.6238, + "step": 2072 + }, + { + "epoch": 0.7937966685812752, + "grad_norm": 0.2964610159397125, + "learning_rate": 4.238990465505001e-06, + "loss": 0.6535, + "step": 2073 + }, + { + "epoch": 0.794179590273789, + "grad_norm": 0.2931128740310669, + "learning_rate": 4.238260649262465e-06, + "loss": 0.6777, + "step": 2074 + }, + { + "epoch": 0.7945625119663029, + "grad_norm": 0.2829414904117584, + "learning_rate": 4.237530546127809e-06, + "loss": 0.6667, + "step": 2075 + }, + { + "epoch": 0.7949454336588168, + "grad_norm": 0.29371336102485657, + "learning_rate": 4.236800156221536e-06, + "loss": 0.6654, + "step": 2076 + }, + { + "epoch": 0.7953283553513306, + "grad_norm": 0.29698020219802856, + "learning_rate": 4.2360694796641915e-06, + "loss": 0.6351, + "step": 2077 + }, + { + "epoch": 0.7957112770438446, + "grad_norm": 0.29009485244750977, + "learning_rate": 4.2353385165763715e-06, + "loss": 0.6565, + "step": 2078 + }, + { + "epoch": 0.7960941987363584, + "grad_norm": 0.2830968201160431, + "learning_rate": 4.234607267078718e-06, + "loss": 0.6514, + "step": 2079 + }, + { + "epoch": 0.7964771204288723, + "grad_norm": 0.2878909409046173, + "learning_rate": 4.2338757312919184e-06, + "loss": 0.6565, + "step": 2080 + }, + { + "epoch": 0.7968600421213862, + "grad_norm": 0.28284013271331787, + "learning_rate": 4.233143909336712e-06, + "loss": 0.637, + "step": 2081 + }, + { + "epoch": 0.7972429638139, + "grad_norm": 0.2860836386680603, + "learning_rate": 4.232411801333882e-06, + "loss": 0.6633, + "step": 2082 + }, + { + "epoch": 0.7976258855064139, + "grad_norm": 0.2868812680244446, + "learning_rate": 4.23167940740426e-06, + "loss": 0.654, + "step": 2083 + }, + { + "epoch": 0.7980088071989279, + "grad_norm": 0.2891739308834076, + "learning_rate": 4.230946727668723e-06, + "loss": 0.6667, + "step": 2084 + }, + { + "epoch": 0.7983917288914417, + "grad_norm": 0.3002691864967346, + "learning_rate": 4.230213762248198e-06, + "loss": 0.6595, + "step": 2085 + }, + { + "epoch": 0.7987746505839556, + "grad_norm": 0.2895083725452423, + "learning_rate": 4.229480511263655e-06, + "loss": 0.6366, + "step": 2086 + }, + { + "epoch": 0.7991575722764694, + "grad_norm": 0.29998913407325745, + "learning_rate": 4.2287469748361174e-06, + "loss": 0.6609, + "step": 2087 + }, + { + "epoch": 0.7995404939689833, + "grad_norm": 0.30361121892929077, + "learning_rate": 4.2280131530866495e-06, + "loss": 0.6884, + "step": 2088 + }, + { + "epoch": 0.7999234156614973, + "grad_norm": 0.2897234857082367, + "learning_rate": 4.227279046136367e-06, + "loss": 0.6719, + "step": 2089 + }, + { + "epoch": 0.8003063373540111, + "grad_norm": 0.28361034393310547, + "learning_rate": 4.226544654106429e-06, + "loss": 0.6422, + "step": 2090 + }, + { + "epoch": 0.800689259046525, + "grad_norm": 0.2946014702320099, + "learning_rate": 4.225809977118046e-06, + "loss": 0.639, + "step": 2091 + }, + { + "epoch": 0.8010721807390389, + "grad_norm": 0.2925739288330078, + "learning_rate": 4.225075015292471e-06, + "loss": 0.646, + "step": 2092 + }, + { + "epoch": 0.8014551024315527, + "grad_norm": 0.296834796667099, + "learning_rate": 4.224339768751008e-06, + "loss": 0.6659, + "step": 2093 + }, + { + "epoch": 0.8018380241240666, + "grad_norm": 0.30825576186180115, + "learning_rate": 4.223604237615004e-06, + "loss": 0.6568, + "step": 2094 + }, + { + "epoch": 0.8022209458165805, + "grad_norm": 0.30276602506637573, + "learning_rate": 4.222868422005856e-06, + "loss": 0.6322, + "step": 2095 + }, + { + "epoch": 0.8026038675090944, + "grad_norm": 0.2918483018875122, + "learning_rate": 4.222132322045009e-06, + "loss": 0.6462, + "step": 2096 + }, + { + "epoch": 0.8029867892016083, + "grad_norm": 0.2914141118526459, + "learning_rate": 4.221395937853949e-06, + "loss": 0.6463, + "step": 2097 + }, + { + "epoch": 0.8033697108941221, + "grad_norm": 0.2930675148963928, + "learning_rate": 4.220659269554217e-06, + "loss": 0.6689, + "step": 2098 + }, + { + "epoch": 0.803752632586636, + "grad_norm": 0.2935783565044403, + "learning_rate": 4.219922317267395e-06, + "loss": 0.6522, + "step": 2099 + }, + { + "epoch": 0.80413555427915, + "grad_norm": 0.28854966163635254, + "learning_rate": 4.219185081115113e-06, + "loss": 0.6435, + "step": 2100 + }, + { + "epoch": 0.8045184759716638, + "grad_norm": 0.29615625739097595, + "learning_rate": 4.218447561219049e-06, + "loss": 0.6402, + "step": 2101 + }, + { + "epoch": 0.8049013976641777, + "grad_norm": 0.2831020951271057, + "learning_rate": 4.217709757700928e-06, + "loss": 0.6549, + "step": 2102 + }, + { + "epoch": 0.8052843193566915, + "grad_norm": 0.3040040135383606, + "learning_rate": 4.216971670682522e-06, + "loss": 0.6589, + "step": 2103 + }, + { + "epoch": 0.8056672410492054, + "grad_norm": 0.2909078299999237, + "learning_rate": 4.216233300285645e-06, + "loss": 0.6438, + "step": 2104 + }, + { + "epoch": 0.8060501627417194, + "grad_norm": 0.2976897954940796, + "learning_rate": 4.215494646632165e-06, + "loss": 0.6488, + "step": 2105 + }, + { + "epoch": 0.8064330844342332, + "grad_norm": 0.2799261212348938, + "learning_rate": 4.214755709843995e-06, + "loss": 0.6438, + "step": 2106 + }, + { + "epoch": 0.8068160061267471, + "grad_norm": 0.3053889274597168, + "learning_rate": 4.214016490043089e-06, + "loss": 0.6533, + "step": 2107 + }, + { + "epoch": 0.807198927819261, + "grad_norm": 0.30414000153541565, + "learning_rate": 4.213276987351455e-06, + "loss": 0.6334, + "step": 2108 + }, + { + "epoch": 0.8075818495117748, + "grad_norm": 0.3047749996185303, + "learning_rate": 4.212537201891143e-06, + "loss": 0.6575, + "step": 2109 + }, + { + "epoch": 0.8079647712042887, + "grad_norm": 0.29599082469940186, + "learning_rate": 4.211797133784252e-06, + "loss": 0.6378, + "step": 2110 + }, + { + "epoch": 0.8083476928968026, + "grad_norm": 0.31033197045326233, + "learning_rate": 4.211056783152927e-06, + "loss": 0.6599, + "step": 2111 + }, + { + "epoch": 0.8087306145893165, + "grad_norm": 0.3023338317871094, + "learning_rate": 4.210316150119359e-06, + "loss": 0.6323, + "step": 2112 + }, + { + "epoch": 0.8091135362818304, + "grad_norm": 0.29808109998703003, + "learning_rate": 4.209575234805786e-06, + "loss": 0.6549, + "step": 2113 + }, + { + "epoch": 0.8094964579743442, + "grad_norm": 0.28547993302345276, + "learning_rate": 4.208834037334494e-06, + "loss": 0.6489, + "step": 2114 + }, + { + "epoch": 0.8098793796668581, + "grad_norm": 0.2818564176559448, + "learning_rate": 4.208092557827813e-06, + "loss": 0.6689, + "step": 2115 + }, + { + "epoch": 0.810262301359372, + "grad_norm": 0.28998851776123047, + "learning_rate": 4.207350796408122e-06, + "loss": 0.6408, + "step": 2116 + }, + { + "epoch": 0.8106452230518859, + "grad_norm": 0.28835082054138184, + "learning_rate": 4.2066087531978444e-06, + "loss": 0.6675, + "step": 2117 + }, + { + "epoch": 0.8110281447443998, + "grad_norm": 0.30105793476104736, + "learning_rate": 4.205866428319451e-06, + "loss": 0.6485, + "step": 2118 + }, + { + "epoch": 0.8114110664369136, + "grad_norm": 0.2862924337387085, + "learning_rate": 4.20512382189546e-06, + "loss": 0.6579, + "step": 2119 + }, + { + "epoch": 0.8117939881294275, + "grad_norm": 0.29366791248321533, + "learning_rate": 4.2043809340484355e-06, + "loss": 0.6453, + "step": 2120 + }, + { + "epoch": 0.8121769098219415, + "grad_norm": 0.2950849235057831, + "learning_rate": 4.203637764900987e-06, + "loss": 0.6704, + "step": 2121 + }, + { + "epoch": 0.8125598315144553, + "grad_norm": 0.2939912974834442, + "learning_rate": 4.202894314575771e-06, + "loss": 0.6549, + "step": 2122 + }, + { + "epoch": 0.8129427532069692, + "grad_norm": 0.2888442873954773, + "learning_rate": 4.202150583195491e-06, + "loss": 0.6425, + "step": 2123 + }, + { + "epoch": 0.813325674899483, + "grad_norm": 0.29482021927833557, + "learning_rate": 4.201406570882898e-06, + "loss": 0.6466, + "step": 2124 + }, + { + "epoch": 0.8137085965919969, + "grad_norm": 0.2820267379283905, + "learning_rate": 4.200662277760785e-06, + "loss": 0.6533, + "step": 2125 + }, + { + "epoch": 0.8140915182845108, + "grad_norm": 0.29276299476623535, + "learning_rate": 4.199917703951997e-06, + "loss": 0.6764, + "step": 2126 + }, + { + "epoch": 0.8144744399770247, + "grad_norm": 0.2967347502708435, + "learning_rate": 4.19917284957942e-06, + "loss": 0.6613, + "step": 2127 + }, + { + "epoch": 0.8148573616695386, + "grad_norm": 0.2874256670475006, + "learning_rate": 4.198427714765991e-06, + "loss": 0.6733, + "step": 2128 + }, + { + "epoch": 0.8152402833620525, + "grad_norm": 0.2876199185848236, + "learning_rate": 4.19768229963469e-06, + "loss": 0.667, + "step": 2129 + }, + { + "epoch": 0.8156232050545663, + "grad_norm": 0.293805867433548, + "learning_rate": 4.196936604308546e-06, + "loss": 0.6846, + "step": 2130 + }, + { + "epoch": 0.8160061267470802, + "grad_norm": 0.29738301038742065, + "learning_rate": 4.19619062891063e-06, + "loss": 0.6382, + "step": 2131 + }, + { + "epoch": 0.8163890484395941, + "grad_norm": 0.2883049249649048, + "learning_rate": 4.195444373564063e-06, + "loss": 0.6522, + "step": 2132 + }, + { + "epoch": 0.816771970132108, + "grad_norm": 0.2838800251483917, + "learning_rate": 4.194697838392013e-06, + "loss": 0.6569, + "step": 2133 + }, + { + "epoch": 0.8171548918246219, + "grad_norm": 0.2773458659648895, + "learning_rate": 4.1939510235176895e-06, + "loss": 0.6588, + "step": 2134 + }, + { + "epoch": 0.8175378135171357, + "grad_norm": 0.2932501435279846, + "learning_rate": 4.1932039290643534e-06, + "loss": 0.6636, + "step": 2135 + }, + { + "epoch": 0.8179207352096496, + "grad_norm": 0.29236772656440735, + "learning_rate": 4.192456555155306e-06, + "loss": 0.6298, + "step": 2136 + }, + { + "epoch": 0.8183036569021636, + "grad_norm": 0.29943132400512695, + "learning_rate": 4.191708901913902e-06, + "loss": 0.6605, + "step": 2137 + }, + { + "epoch": 0.8186865785946774, + "grad_norm": 0.2825752794742584, + "learning_rate": 4.190960969463536e-06, + "loss": 0.6612, + "step": 2138 + }, + { + "epoch": 0.8190695002871913, + "grad_norm": 0.29188987612724304, + "learning_rate": 4.190212757927651e-06, + "loss": 0.651, + "step": 2139 + }, + { + "epoch": 0.8194524219797051, + "grad_norm": 0.2977637052536011, + "learning_rate": 4.189464267429736e-06, + "loss": 0.6714, + "step": 2140 + }, + { + "epoch": 0.819835343672219, + "grad_norm": 0.2999652922153473, + "learning_rate": 4.188715498093326e-06, + "loss": 0.6558, + "step": 2141 + }, + { + "epoch": 0.8202182653647329, + "grad_norm": 0.28152018785476685, + "learning_rate": 4.1879664500420014e-06, + "loss": 0.6545, + "step": 2142 + }, + { + "epoch": 0.8206011870572468, + "grad_norm": 0.2833508253097534, + "learning_rate": 4.187217123399391e-06, + "loss": 0.667, + "step": 2143 + }, + { + "epoch": 0.8209841087497607, + "grad_norm": 0.2977939248085022, + "learning_rate": 4.186467518289166e-06, + "loss": 0.6519, + "step": 2144 + }, + { + "epoch": 0.8213670304422745, + "grad_norm": 0.29274484515190125, + "learning_rate": 4.185717634835047e-06, + "loss": 0.6487, + "step": 2145 + }, + { + "epoch": 0.8217499521347884, + "grad_norm": 0.2835085988044739, + "learning_rate": 4.184967473160797e-06, + "loss": 0.6472, + "step": 2146 + }, + { + "epoch": 0.8221328738273023, + "grad_norm": 0.29186347126960754, + "learning_rate": 4.184217033390229e-06, + "loss": 0.6438, + "step": 2147 + }, + { + "epoch": 0.8225157955198162, + "grad_norm": 0.2994496524333954, + "learning_rate": 4.183466315647198e-06, + "loss": 0.6501, + "step": 2148 + }, + { + "epoch": 0.8228987172123301, + "grad_norm": 0.298163503408432, + "learning_rate": 4.182715320055607e-06, + "loss": 0.6686, + "step": 2149 + }, + { + "epoch": 0.823281638904844, + "grad_norm": 0.29106858372688293, + "learning_rate": 4.181964046739404e-06, + "loss": 0.6421, + "step": 2150 + }, + { + "epoch": 0.8236645605973578, + "grad_norm": 0.3015703856945038, + "learning_rate": 4.181212495822585e-06, + "loss": 0.6518, + "step": 2151 + }, + { + "epoch": 0.8240474822898717, + "grad_norm": 0.28766122460365295, + "learning_rate": 4.180460667429188e-06, + "loss": 0.6584, + "step": 2152 + }, + { + "epoch": 0.8244304039823857, + "grad_norm": 0.28938916325569153, + "learning_rate": 4.1797085616833015e-06, + "loss": 0.6476, + "step": 2153 + }, + { + "epoch": 0.8248133256748995, + "grad_norm": 0.2908332049846649, + "learning_rate": 4.178956178709054e-06, + "loss": 0.6345, + "step": 2154 + }, + { + "epoch": 0.8251962473674134, + "grad_norm": 0.3065876066684723, + "learning_rate": 4.178203518630625e-06, + "loss": 0.6461, + "step": 2155 + }, + { + "epoch": 0.8255791690599272, + "grad_norm": 0.2874744236469269, + "learning_rate": 4.177450581572236e-06, + "loss": 0.6561, + "step": 2156 + }, + { + "epoch": 0.8259620907524411, + "grad_norm": 0.3040878176689148, + "learning_rate": 4.176697367658159e-06, + "loss": 0.6527, + "step": 2157 + }, + { + "epoch": 0.826345012444955, + "grad_norm": 0.313347190618515, + "learning_rate": 4.175943877012706e-06, + "loss": 0.6602, + "step": 2158 + }, + { + "epoch": 0.8267279341374689, + "grad_norm": 0.28679871559143066, + "learning_rate": 4.175190109760237e-06, + "loss": 0.6439, + "step": 2159 + }, + { + "epoch": 0.8271108558299828, + "grad_norm": 0.3119087517261505, + "learning_rate": 4.17443606602516e-06, + "loss": 0.639, + "step": 2160 + }, + { + "epoch": 0.8274937775224966, + "grad_norm": 0.2842283844947815, + "learning_rate": 4.173681745931925e-06, + "loss": 0.6531, + "step": 2161 + }, + { + "epoch": 0.8278766992150105, + "grad_norm": 0.28919950127601624, + "learning_rate": 4.1729271496050284e-06, + "loss": 0.6349, + "step": 2162 + }, + { + "epoch": 0.8282596209075244, + "grad_norm": 0.29278960824012756, + "learning_rate": 4.172172277169015e-06, + "loss": 0.6223, + "step": 2163 + }, + { + "epoch": 0.8286425426000383, + "grad_norm": 0.291665643453598, + "learning_rate": 4.171417128748472e-06, + "loss": 0.6451, + "step": 2164 + }, + { + "epoch": 0.8290254642925522, + "grad_norm": 0.2924218773841858, + "learning_rate": 4.170661704468033e-06, + "loss": 0.6545, + "step": 2165 + }, + { + "epoch": 0.829408385985066, + "grad_norm": 0.2932129204273224, + "learning_rate": 4.169906004452378e-06, + "loss": 0.64, + "step": 2166 + }, + { + "epoch": 0.8297913076775799, + "grad_norm": 0.30480051040649414, + "learning_rate": 4.169150028826232e-06, + "loss": 0.6519, + "step": 2167 + }, + { + "epoch": 0.8301742293700938, + "grad_norm": 0.2939715087413788, + "learning_rate": 4.168393777714364e-06, + "loss": 0.66, + "step": 2168 + }, + { + "epoch": 0.8305571510626077, + "grad_norm": 0.2925496995449066, + "learning_rate": 4.167637251241592e-06, + "loss": 0.6571, + "step": 2169 + }, + { + "epoch": 0.8309400727551216, + "grad_norm": 0.3134564757347107, + "learning_rate": 4.166880449532776e-06, + "loss": 0.6606, + "step": 2170 + }, + { + "epoch": 0.8313229944476355, + "grad_norm": 0.29317769408226013, + "learning_rate": 4.166123372712823e-06, + "loss": 0.6538, + "step": 2171 + }, + { + "epoch": 0.8317059161401493, + "grad_norm": 0.28427040576934814, + "learning_rate": 4.1653660209066835e-06, + "loss": 0.6629, + "step": 2172 + }, + { + "epoch": 0.8320888378326632, + "grad_norm": 0.30414286255836487, + "learning_rate": 4.164608394239358e-06, + "loss": 0.6428, + "step": 2173 + }, + { + "epoch": 0.832471759525177, + "grad_norm": 0.2859039902687073, + "learning_rate": 4.163850492835888e-06, + "loss": 0.6483, + "step": 2174 + }, + { + "epoch": 0.832854681217691, + "grad_norm": 0.30278804898262024, + "learning_rate": 4.163092316821361e-06, + "loss": 0.6487, + "step": 2175 + }, + { + "epoch": 0.8332376029102049, + "grad_norm": 0.3010164201259613, + "learning_rate": 4.162333866320911e-06, + "loss": 0.6645, + "step": 2176 + }, + { + "epoch": 0.8336205246027187, + "grad_norm": 0.3262551426887512, + "learning_rate": 4.161575141459718e-06, + "loss": 0.6419, + "step": 2177 + }, + { + "epoch": 0.8340034462952326, + "grad_norm": 0.3013157248497009, + "learning_rate": 4.160816142363005e-06, + "loss": 0.6385, + "step": 2178 + }, + { + "epoch": 0.8343863679877465, + "grad_norm": 0.28078892827033997, + "learning_rate": 4.160056869156041e-06, + "loss": 0.6637, + "step": 2179 + }, + { + "epoch": 0.8347692896802604, + "grad_norm": 0.34636321663856506, + "learning_rate": 4.159297321964141e-06, + "loss": 0.6199, + "step": 2180 + }, + { + "epoch": 0.8351522113727743, + "grad_norm": 0.30984798073768616, + "learning_rate": 4.158537500912666e-06, + "loss": 0.6431, + "step": 2181 + }, + { + "epoch": 0.8355351330652881, + "grad_norm": 0.29280543327331543, + "learning_rate": 4.15777740612702e-06, + "loss": 0.6798, + "step": 2182 + }, + { + "epoch": 0.835918054757802, + "grad_norm": 0.3050188422203064, + "learning_rate": 4.157017037732652e-06, + "loss": 0.6618, + "step": 2183 + }, + { + "epoch": 0.8363009764503159, + "grad_norm": 0.32322168350219727, + "learning_rate": 4.156256395855061e-06, + "loss": 0.6472, + "step": 2184 + }, + { + "epoch": 0.8366838981428298, + "grad_norm": 0.29877689480781555, + "learning_rate": 4.155495480619784e-06, + "loss": 0.6285, + "step": 2185 + }, + { + "epoch": 0.8370668198353437, + "grad_norm": 0.29478585720062256, + "learning_rate": 4.154734292152408e-06, + "loss": 0.6351, + "step": 2186 + }, + { + "epoch": 0.8374497415278576, + "grad_norm": 0.28708431124687195, + "learning_rate": 4.153972830578563e-06, + "loss": 0.6482, + "step": 2187 + }, + { + "epoch": 0.8378326632203714, + "grad_norm": 0.3226141035556793, + "learning_rate": 4.153211096023927e-06, + "loss": 0.6394, + "step": 2188 + }, + { + "epoch": 0.8382155849128853, + "grad_norm": 0.3001896142959595, + "learning_rate": 4.152449088614218e-06, + "loss": 0.6491, + "step": 2189 + }, + { + "epoch": 0.8385985066053991, + "grad_norm": 0.2948077917098999, + "learning_rate": 4.151686808475204e-06, + "loss": 0.6552, + "step": 2190 + }, + { + "epoch": 0.8389814282979131, + "grad_norm": 0.30342164635658264, + "learning_rate": 4.150924255732695e-06, + "loss": 0.6501, + "step": 2191 + }, + { + "epoch": 0.839364349990427, + "grad_norm": 0.2866464853286743, + "learning_rate": 4.150161430512547e-06, + "loss": 0.6526, + "step": 2192 + }, + { + "epoch": 0.8397472716829408, + "grad_norm": 0.2831853926181793, + "learning_rate": 4.1493983329406605e-06, + "loss": 0.6462, + "step": 2193 + }, + { + "epoch": 0.8401301933754547, + "grad_norm": 0.3006346523761749, + "learning_rate": 4.148634963142982e-06, + "loss": 0.6494, + "step": 2194 + }, + { + "epoch": 0.8405131150679686, + "grad_norm": 0.30146533250808716, + "learning_rate": 4.147871321245503e-06, + "loss": 0.6524, + "step": 2195 + }, + { + "epoch": 0.8408960367604825, + "grad_norm": 0.30787044763565063, + "learning_rate": 4.147107407374258e-06, + "loss": 0.6353, + "step": 2196 + }, + { + "epoch": 0.8412789584529964, + "grad_norm": 0.30071893334388733, + "learning_rate": 4.146343221655327e-06, + "loss": 0.6527, + "step": 2197 + }, + { + "epoch": 0.8416618801455102, + "grad_norm": 0.2882702648639679, + "learning_rate": 4.145578764214836e-06, + "loss": 0.6407, + "step": 2198 + }, + { + "epoch": 0.8420448018380241, + "grad_norm": 0.2802189886569977, + "learning_rate": 4.144814035178956e-06, + "loss": 0.6425, + "step": 2199 + }, + { + "epoch": 0.842427723530538, + "grad_norm": 0.29071342945098877, + "learning_rate": 4.144049034673901e-06, + "loss": 0.6593, + "step": 2200 + }, + { + "epoch": 0.8428106452230519, + "grad_norm": 0.2915160357952118, + "learning_rate": 4.1432837628259325e-06, + "loss": 0.6523, + "step": 2201 + }, + { + "epoch": 0.8431935669155658, + "grad_norm": 0.2865479290485382, + "learning_rate": 4.142518219761354e-06, + "loss": 0.639, + "step": 2202 + }, + { + "epoch": 0.8435764886080797, + "grad_norm": 0.289639949798584, + "learning_rate": 4.141752405606514e-06, + "loss": 0.6538, + "step": 2203 + }, + { + "epoch": 0.8439594103005935, + "grad_norm": 0.29012227058410645, + "learning_rate": 4.140986320487809e-06, + "loss": 0.6522, + "step": 2204 + }, + { + "epoch": 0.8443423319931074, + "grad_norm": 0.2918872833251953, + "learning_rate": 4.1402199645316756e-06, + "loss": 0.6591, + "step": 2205 + }, + { + "epoch": 0.8447252536856212, + "grad_norm": 0.28388962149620056, + "learning_rate": 4.139453337864599e-06, + "loss": 0.653, + "step": 2206 + }, + { + "epoch": 0.8451081753781352, + "grad_norm": 0.293449342250824, + "learning_rate": 4.138686440613107e-06, + "loss": 0.6742, + "step": 2207 + }, + { + "epoch": 0.8454910970706491, + "grad_norm": 0.29294872283935547, + "learning_rate": 4.137919272903772e-06, + "loss": 0.6322, + "step": 2208 + }, + { + "epoch": 0.8458740187631629, + "grad_norm": 0.29742491245269775, + "learning_rate": 4.137151834863213e-06, + "loss": 0.6551, + "step": 2209 + }, + { + "epoch": 0.8462569404556768, + "grad_norm": 0.2894884943962097, + "learning_rate": 4.136384126618091e-06, + "loss": 0.6372, + "step": 2210 + }, + { + "epoch": 0.8466398621481906, + "grad_norm": 0.2900755703449249, + "learning_rate": 4.135616148295113e-06, + "loss": 0.6525, + "step": 2211 + }, + { + "epoch": 0.8470227838407046, + "grad_norm": 0.28828489780426025, + "learning_rate": 4.1348479000210305e-06, + "loss": 0.6503, + "step": 2212 + }, + { + "epoch": 0.8474057055332185, + "grad_norm": 0.2918212115764618, + "learning_rate": 4.1340793819226395e-06, + "loss": 0.6751, + "step": 2213 + }, + { + "epoch": 0.8477886272257323, + "grad_norm": 0.29055020213127136, + "learning_rate": 4.1333105941267805e-06, + "loss": 0.6618, + "step": 2214 + }, + { + "epoch": 0.8481715489182462, + "grad_norm": 0.2884266674518585, + "learning_rate": 4.132541536760338e-06, + "loss": 0.6556, + "step": 2215 + }, + { + "epoch": 0.8485544706107601, + "grad_norm": 0.2972971498966217, + "learning_rate": 4.1317722099502425e-06, + "loss": 0.6481, + "step": 2216 + }, + { + "epoch": 0.848937392303274, + "grad_norm": 0.2842160761356354, + "learning_rate": 4.131002613823467e-06, + "loss": 0.6648, + "step": 2217 + }, + { + "epoch": 0.8493203139957879, + "grad_norm": 0.2807370126247406, + "learning_rate": 4.13023274850703e-06, + "loss": 0.6519, + "step": 2218 + }, + { + "epoch": 0.8497032356883018, + "grad_norm": 0.282664954662323, + "learning_rate": 4.1294626141279944e-06, + "loss": 0.662, + "step": 2219 + }, + { + "epoch": 0.8500861573808156, + "grad_norm": 0.2957598567008972, + "learning_rate": 4.128692210813468e-06, + "loss": 0.672, + "step": 2220 + }, + { + "epoch": 0.8504690790733295, + "grad_norm": 0.2845847010612488, + "learning_rate": 4.127921538690601e-06, + "loss": 0.6555, + "step": 2221 + }, + { + "epoch": 0.8508520007658433, + "grad_norm": 0.286517471075058, + "learning_rate": 4.127150597886591e-06, + "loss": 0.6611, + "step": 2222 + }, + { + "epoch": 0.8512349224583573, + "grad_norm": 0.29605650901794434, + "learning_rate": 4.1263793885286775e-06, + "loss": 0.6256, + "step": 2223 + }, + { + "epoch": 0.8516178441508712, + "grad_norm": 0.2793962061405182, + "learning_rate": 4.125607910744146e-06, + "loss": 0.6548, + "step": 2224 + }, + { + "epoch": 0.852000765843385, + "grad_norm": 0.2938767075538635, + "learning_rate": 4.124836164660323e-06, + "loss": 0.672, + "step": 2225 + }, + { + "epoch": 0.8523836875358989, + "grad_norm": 0.30065855383872986, + "learning_rate": 4.124064150404584e-06, + "loss": 0.6595, + "step": 2226 + }, + { + "epoch": 0.8527666092284127, + "grad_norm": 0.2833341360092163, + "learning_rate": 4.1232918681043465e-06, + "loss": 0.6237, + "step": 2227 + }, + { + "epoch": 0.8531495309209267, + "grad_norm": 0.2910335659980774, + "learning_rate": 4.122519317887072e-06, + "loss": 0.6479, + "step": 2228 + }, + { + "epoch": 0.8535324526134406, + "grad_norm": 0.2925763726234436, + "learning_rate": 4.121746499880265e-06, + "loss": 0.6394, + "step": 2229 + }, + { + "epoch": 0.8539153743059544, + "grad_norm": 0.29192501306533813, + "learning_rate": 4.120973414211477e-06, + "loss": 0.6658, + "step": 2230 + }, + { + "epoch": 0.8542982959984683, + "grad_norm": 0.2963842749595642, + "learning_rate": 4.120200061008303e-06, + "loss": 0.6319, + "step": 2231 + }, + { + "epoch": 0.8546812176909822, + "grad_norm": 0.3151523768901825, + "learning_rate": 4.119426440398378e-06, + "loss": 0.6558, + "step": 2232 + }, + { + "epoch": 0.8550641393834961, + "grad_norm": 0.29208821058273315, + "learning_rate": 4.11865255250939e-06, + "loss": 0.6442, + "step": 2233 + }, + { + "epoch": 0.85544706107601, + "grad_norm": 0.296401709318161, + "learning_rate": 4.117878397469062e-06, + "loss": 0.6463, + "step": 2234 + }, + { + "epoch": 0.8558299827685238, + "grad_norm": 0.2935371398925781, + "learning_rate": 4.117103975405165e-06, + "loss": 0.6702, + "step": 2235 + }, + { + "epoch": 0.8562129044610377, + "grad_norm": 0.30976298451423645, + "learning_rate": 4.116329286445515e-06, + "loss": 0.6891, + "step": 2236 + }, + { + "epoch": 0.8565958261535516, + "grad_norm": 0.29204466938972473, + "learning_rate": 4.11555433071797e-06, + "loss": 0.6513, + "step": 2237 + }, + { + "epoch": 0.8569787478460654, + "grad_norm": 0.2978476583957672, + "learning_rate": 4.114779108350434e-06, + "loss": 0.6617, + "step": 2238 + }, + { + "epoch": 0.8573616695385794, + "grad_norm": 0.32826146483421326, + "learning_rate": 4.114003619470853e-06, + "loss": 0.6763, + "step": 2239 + }, + { + "epoch": 0.8577445912310933, + "grad_norm": 0.30818793177604675, + "learning_rate": 4.113227864207219e-06, + "loss": 0.6575, + "step": 2240 + }, + { + "epoch": 0.8581275129236071, + "grad_norm": 0.29063844680786133, + "learning_rate": 4.112451842687567e-06, + "loss": 0.6495, + "step": 2241 + }, + { + "epoch": 0.858510434616121, + "grad_norm": 0.29370319843292236, + "learning_rate": 4.1116755550399735e-06, + "loss": 0.6713, + "step": 2242 + }, + { + "epoch": 0.8588933563086348, + "grad_norm": 0.3022923767566681, + "learning_rate": 4.110899001392563e-06, + "loss": 0.6339, + "step": 2243 + }, + { + "epoch": 0.8592762780011488, + "grad_norm": 0.28490132093429565, + "learning_rate": 4.110122181873504e-06, + "loss": 0.6433, + "step": 2244 + }, + { + "epoch": 0.8596591996936627, + "grad_norm": 0.2953353822231293, + "learning_rate": 4.109345096611004e-06, + "loss": 0.6681, + "step": 2245 + }, + { + "epoch": 0.8600421213861765, + "grad_norm": 0.3136865198612213, + "learning_rate": 4.108567745733318e-06, + "loss": 0.649, + "step": 2246 + }, + { + "epoch": 0.8604250430786904, + "grad_norm": 0.31371670961380005, + "learning_rate": 4.1077901293687464e-06, + "loss": 0.6384, + "step": 2247 + }, + { + "epoch": 0.8608079647712042, + "grad_norm": 0.2912449538707733, + "learning_rate": 4.107012247645629e-06, + "loss": 0.6313, + "step": 2248 + }, + { + "epoch": 0.8611908864637182, + "grad_norm": 0.3179062306880951, + "learning_rate": 4.106234100692352e-06, + "loss": 0.656, + "step": 2249 + }, + { + "epoch": 0.8615738081562321, + "grad_norm": 0.3314395844936371, + "learning_rate": 4.105455688637347e-06, + "loss": 0.6606, + "step": 2250 + }, + { + "epoch": 0.8619567298487459, + "grad_norm": 0.29691487550735474, + "learning_rate": 4.104677011609085e-06, + "loss": 0.6442, + "step": 2251 + }, + { + "epoch": 0.8623396515412598, + "grad_norm": 0.3266996145248413, + "learning_rate": 4.103898069736084e-06, + "loss": 0.6441, + "step": 2252 + }, + { + "epoch": 0.8627225732337737, + "grad_norm": 0.3615022599697113, + "learning_rate": 4.103118863146905e-06, + "loss": 0.6529, + "step": 2253 + }, + { + "epoch": 0.8631054949262875, + "grad_norm": 0.28917160630226135, + "learning_rate": 4.102339391970153e-06, + "loss": 0.6371, + "step": 2254 + }, + { + "epoch": 0.8634884166188015, + "grad_norm": 0.31351974606513977, + "learning_rate": 4.101559656334476e-06, + "loss": 0.634, + "step": 2255 + }, + { + "epoch": 0.8638713383113154, + "grad_norm": 0.3009597063064575, + "learning_rate": 4.100779656368566e-06, + "loss": 0.6495, + "step": 2256 + }, + { + "epoch": 0.8642542600038292, + "grad_norm": 0.2938322424888611, + "learning_rate": 4.099999392201157e-06, + "loss": 0.6509, + "step": 2257 + }, + { + "epoch": 0.8646371816963431, + "grad_norm": 0.2903949022293091, + "learning_rate": 4.09921886396103e-06, + "loss": 0.6459, + "step": 2258 + }, + { + "epoch": 0.8650201033888569, + "grad_norm": 0.30466920137405396, + "learning_rate": 4.098438071777006e-06, + "loss": 0.6632, + "step": 2259 + }, + { + "epoch": 0.8654030250813709, + "grad_norm": 0.29187896847724915, + "learning_rate": 4.097657015777954e-06, + "loss": 0.6704, + "step": 2260 + }, + { + "epoch": 0.8657859467738848, + "grad_norm": 0.298870325088501, + "learning_rate": 4.09687569609278e-06, + "loss": 0.644, + "step": 2261 + }, + { + "epoch": 0.8661688684663986, + "grad_norm": 0.29359111189842224, + "learning_rate": 4.096094112850441e-06, + "loss": 0.6666, + "step": 2262 + }, + { + "epoch": 0.8665517901589125, + "grad_norm": 0.3013385832309723, + "learning_rate": 4.095312266179931e-06, + "loss": 0.6528, + "step": 2263 + }, + { + "epoch": 0.8669347118514263, + "grad_norm": 0.29920855164527893, + "learning_rate": 4.094530156210292e-06, + "loss": 0.6637, + "step": 2264 + }, + { + "epoch": 0.8673176335439403, + "grad_norm": 0.2906420826911926, + "learning_rate": 4.093747783070607e-06, + "loss": 0.6572, + "step": 2265 + }, + { + "epoch": 0.8677005552364542, + "grad_norm": 0.28690263628959656, + "learning_rate": 4.092965146890002e-06, + "loss": 0.6644, + "step": 2266 + }, + { + "epoch": 0.868083476928968, + "grad_norm": 0.30393892526626587, + "learning_rate": 4.09218224779765e-06, + "loss": 0.651, + "step": 2267 + }, + { + "epoch": 0.8684663986214819, + "grad_norm": 0.29441913962364197, + "learning_rate": 4.091399085922764e-06, + "loss": 0.6748, + "step": 2268 + }, + { + "epoch": 0.8688493203139958, + "grad_norm": 0.28890660405158997, + "learning_rate": 4.090615661394601e-06, + "loss": 0.6493, + "step": 2269 + }, + { + "epoch": 0.8692322420065096, + "grad_norm": 0.28217312693595886, + "learning_rate": 4.089831974342462e-06, + "loss": 0.6621, + "step": 2270 + }, + { + "epoch": 0.8696151636990236, + "grad_norm": 0.28720107674598694, + "learning_rate": 4.08904802489569e-06, + "loss": 0.629, + "step": 2271 + }, + { + "epoch": 0.8699980853915374, + "grad_norm": 0.29145902395248413, + "learning_rate": 4.088263813183674e-06, + "loss": 0.6307, + "step": 2272 + }, + { + "epoch": 0.8703810070840513, + "grad_norm": 0.2882422208786011, + "learning_rate": 4.087479339335844e-06, + "loss": 0.6531, + "step": 2273 + }, + { + "epoch": 0.8707639287765652, + "grad_norm": 0.27108249068260193, + "learning_rate": 4.086694603481673e-06, + "loss": 0.6418, + "step": 2274 + }, + { + "epoch": 0.871146850469079, + "grad_norm": 0.28425681591033936, + "learning_rate": 4.0859096057506805e-06, + "loss": 0.6432, + "step": 2275 + }, + { + "epoch": 0.871529772161593, + "grad_norm": 0.29180005192756653, + "learning_rate": 4.0851243462724245e-06, + "loss": 0.6494, + "step": 2276 + }, + { + "epoch": 0.8719126938541069, + "grad_norm": 0.2760147452354431, + "learning_rate": 4.084338825176509e-06, + "loss": 0.6551, + "step": 2277 + }, + { + "epoch": 0.8722956155466207, + "grad_norm": 0.27937087416648865, + "learning_rate": 4.083553042592581e-06, + "loss": 0.6519, + "step": 2278 + }, + { + "epoch": 0.8726785372391346, + "grad_norm": 0.28661325573921204, + "learning_rate": 4.082766998650331e-06, + "loss": 0.6527, + "step": 2279 + }, + { + "epoch": 0.8730614589316484, + "grad_norm": 0.29728376865386963, + "learning_rate": 4.08198069347949e-06, + "loss": 0.618, + "step": 2280 + }, + { + "epoch": 0.8734443806241624, + "grad_norm": 0.3050868809223175, + "learning_rate": 4.081194127209837e-06, + "loss": 0.6618, + "step": 2281 + }, + { + "epoch": 0.8738273023166763, + "grad_norm": 0.2857801914215088, + "learning_rate": 4.08040729997119e-06, + "loss": 0.6464, + "step": 2282 + }, + { + "epoch": 0.8742102240091901, + "grad_norm": 0.29985737800598145, + "learning_rate": 4.07962021189341e-06, + "loss": 0.6489, + "step": 2283 + }, + { + "epoch": 0.874593145701704, + "grad_norm": 0.2908843457698822, + "learning_rate": 4.078832863106404e-06, + "loss": 0.657, + "step": 2284 + }, + { + "epoch": 0.8749760673942178, + "grad_norm": 0.2829788625240326, + "learning_rate": 4.078045253740121e-06, + "loss": 0.6667, + "step": 2285 + }, + { + "epoch": 0.8753589890867317, + "grad_norm": 0.28448620438575745, + "learning_rate": 4.07725738392455e-06, + "loss": 0.6482, + "step": 2286 + }, + { + "epoch": 0.8757419107792457, + "grad_norm": 0.2935240864753723, + "learning_rate": 4.076469253789727e-06, + "loss": 0.6479, + "step": 2287 + }, + { + "epoch": 0.8761248324717595, + "grad_norm": 0.28987401723861694, + "learning_rate": 4.075680863465729e-06, + "loss": 0.659, + "step": 2288 + }, + { + "epoch": 0.8765077541642734, + "grad_norm": 0.2843879461288452, + "learning_rate": 4.074892213082676e-06, + "loss": 0.6618, + "step": 2289 + }, + { + "epoch": 0.8768906758567873, + "grad_norm": 0.2928538918495178, + "learning_rate": 4.074103302770731e-06, + "loss": 0.6552, + "step": 2290 + }, + { + "epoch": 0.8772735975493011, + "grad_norm": 0.28846338391304016, + "learning_rate": 4.073314132660102e-06, + "loss": 0.6462, + "step": 2291 + }, + { + "epoch": 0.8776565192418151, + "grad_norm": 0.2852533757686615, + "learning_rate": 4.072524702881036e-06, + "loss": 0.6578, + "step": 2292 + }, + { + "epoch": 0.878039440934329, + "grad_norm": 0.28771287202835083, + "learning_rate": 4.071735013563824e-06, + "loss": 0.6648, + "step": 2293 + }, + { + "epoch": 0.8784223626268428, + "grad_norm": 0.28862497210502625, + "learning_rate": 4.0709450648388034e-06, + "loss": 0.6423, + "step": 2294 + }, + { + "epoch": 0.8788052843193567, + "grad_norm": 0.2900797724723816, + "learning_rate": 4.070154856836351e-06, + "loss": 0.6323, + "step": 2295 + }, + { + "epoch": 0.8791882060118705, + "grad_norm": 0.30567964911460876, + "learning_rate": 4.069364389686886e-06, + "loss": 0.6732, + "step": 2296 + }, + { + "epoch": 0.8795711277043845, + "grad_norm": 0.297863632440567, + "learning_rate": 4.068573663520871e-06, + "loss": 0.6706, + "step": 2297 + }, + { + "epoch": 0.8799540493968984, + "grad_norm": 0.29229220747947693, + "learning_rate": 4.0677826784688136e-06, + "loss": 0.6686, + "step": 2298 + }, + { + "epoch": 0.8803369710894122, + "grad_norm": 0.28343459963798523, + "learning_rate": 4.066991434661262e-06, + "loss": 0.6478, + "step": 2299 + }, + { + "epoch": 0.8807198927819261, + "grad_norm": 0.29238826036453247, + "learning_rate": 4.066199932228805e-06, + "loss": 0.658, + "step": 2300 + }, + { + "epoch": 0.8811028144744399, + "grad_norm": 0.3182104527950287, + "learning_rate": 4.06540817130208e-06, + "loss": 0.6373, + "step": 2301 + }, + { + "epoch": 0.8814857361669538, + "grad_norm": 0.2971557080745697, + "learning_rate": 4.064616152011761e-06, + "loss": 0.644, + "step": 2302 + }, + { + "epoch": 0.8818686578594678, + "grad_norm": 0.3003726303577423, + "learning_rate": 4.063823874488569e-06, + "loss": 0.6362, + "step": 2303 + }, + { + "epoch": 0.8822515795519816, + "grad_norm": 0.28735241293907166, + "learning_rate": 4.0630313388632645e-06, + "loss": 0.6592, + "step": 2304 + }, + { + "epoch": 0.8826345012444955, + "grad_norm": 0.2918720841407776, + "learning_rate": 4.062238545266654e-06, + "loss": 0.6378, + "step": 2305 + }, + { + "epoch": 0.8830174229370094, + "grad_norm": 0.30211561918258667, + "learning_rate": 4.061445493829582e-06, + "loss": 0.6392, + "step": 2306 + }, + { + "epoch": 0.8834003446295232, + "grad_norm": 0.2915424108505249, + "learning_rate": 4.060652184682938e-06, + "loss": 0.6617, + "step": 2307 + }, + { + "epoch": 0.8837832663220372, + "grad_norm": 0.29661399126052856, + "learning_rate": 4.059858617957657e-06, + "loss": 0.6405, + "step": 2308 + }, + { + "epoch": 0.884166188014551, + "grad_norm": 0.28728193044662476, + "learning_rate": 4.059064793784711e-06, + "loss": 0.64, + "step": 2309 + }, + { + "epoch": 0.8845491097070649, + "grad_norm": 0.3088046610355377, + "learning_rate": 4.058270712295118e-06, + "loss": 0.6753, + "step": 2310 + }, + { + "epoch": 0.8849320313995788, + "grad_norm": 0.28920331597328186, + "learning_rate": 4.057476373619938e-06, + "loss": 0.6485, + "step": 2311 + }, + { + "epoch": 0.8853149530920926, + "grad_norm": 0.29318201541900635, + "learning_rate": 4.056681777890273e-06, + "loss": 0.6602, + "step": 2312 + }, + { + "epoch": 0.8856978747846066, + "grad_norm": 0.2955826222896576, + "learning_rate": 4.055886925237266e-06, + "loss": 0.6645, + "step": 2313 + }, + { + "epoch": 0.8860807964771205, + "grad_norm": 0.2962397634983063, + "learning_rate": 4.055091815792106e-06, + "loss": 0.6318, + "step": 2314 + }, + { + "epoch": 0.8864637181696343, + "grad_norm": 0.2946590483188629, + "learning_rate": 4.05429644968602e-06, + "loss": 0.6649, + "step": 2315 + }, + { + "epoch": 0.8868466398621482, + "grad_norm": 0.291528582572937, + "learning_rate": 4.0535008270502815e-06, + "loss": 0.6452, + "step": 2316 + }, + { + "epoch": 0.887229561554662, + "grad_norm": 0.28559184074401855, + "learning_rate": 4.052704948016204e-06, + "loss": 0.6632, + "step": 2317 + }, + { + "epoch": 0.8876124832471759, + "grad_norm": 0.3063846230506897, + "learning_rate": 4.051908812715143e-06, + "loss": 0.6577, + "step": 2318 + }, + { + "epoch": 0.8879954049396899, + "grad_norm": 0.29996639490127563, + "learning_rate": 4.051112421278498e-06, + "loss": 0.661, + "step": 2319 + }, + { + "epoch": 0.8883783266322037, + "grad_norm": 0.2953406572341919, + "learning_rate": 4.050315773837708e-06, + "loss": 0.646, + "step": 2320 + }, + { + "epoch": 0.8887612483247176, + "grad_norm": 0.3012860119342804, + "learning_rate": 4.049518870524259e-06, + "loss": 0.6583, + "step": 2321 + }, + { + "epoch": 0.8891441700172314, + "grad_norm": 0.2990375757217407, + "learning_rate": 4.048721711469675e-06, + "loss": 0.6394, + "step": 2322 + }, + { + "epoch": 0.8895270917097453, + "grad_norm": 0.2893383800983429, + "learning_rate": 4.047924296805522e-06, + "loss": 0.6623, + "step": 2323 + }, + { + "epoch": 0.8899100134022593, + "grad_norm": 0.29396453499794006, + "learning_rate": 4.047126626663413e-06, + "loss": 0.6473, + "step": 2324 + }, + { + "epoch": 0.8902929350947731, + "grad_norm": 0.28445231914520264, + "learning_rate": 4.046328701174995e-06, + "loss": 0.6367, + "step": 2325 + }, + { + "epoch": 0.890675856787287, + "grad_norm": 0.29325002431869507, + "learning_rate": 4.045530520471968e-06, + "loss": 0.6418, + "step": 2326 + }, + { + "epoch": 0.8910587784798009, + "grad_norm": 0.28435030579566956, + "learning_rate": 4.044732084686062e-06, + "loss": 0.6629, + "step": 2327 + }, + { + "epoch": 0.8914417001723147, + "grad_norm": 0.2926534116268158, + "learning_rate": 4.0439333939490604e-06, + "loss": 0.6488, + "step": 2328 + }, + { + "epoch": 0.8918246218648287, + "grad_norm": 0.2846304476261139, + "learning_rate": 4.043134448392782e-06, + "loss": 0.6469, + "step": 2329 + }, + { + "epoch": 0.8922075435573426, + "grad_norm": 0.29454225301742554, + "learning_rate": 4.0423352481490865e-06, + "loss": 0.663, + "step": 2330 + }, + { + "epoch": 0.8925904652498564, + "grad_norm": 0.30551332235336304, + "learning_rate": 4.041535793349881e-06, + "loss": 0.6511, + "step": 2331 + }, + { + "epoch": 0.8929733869423703, + "grad_norm": 0.30226585268974304, + "learning_rate": 4.040736084127113e-06, + "loss": 0.6418, + "step": 2332 + }, + { + "epoch": 0.8933563086348841, + "grad_norm": 0.29474323987960815, + "learning_rate": 4.039936120612767e-06, + "loss": 0.6412, + "step": 2333 + }, + { + "epoch": 0.893739230327398, + "grad_norm": 0.30319032073020935, + "learning_rate": 4.039135902938877e-06, + "loss": 0.6713, + "step": 2334 + }, + { + "epoch": 0.894122152019912, + "grad_norm": 0.3011849522590637, + "learning_rate": 4.038335431237513e-06, + "loss": 0.648, + "step": 2335 + }, + { + "epoch": 0.8945050737124258, + "grad_norm": 0.2926468551158905, + "learning_rate": 4.037534705640792e-06, + "loss": 0.65, + "step": 2336 + }, + { + "epoch": 0.8948879954049397, + "grad_norm": 0.2955300211906433, + "learning_rate": 4.036733726280867e-06, + "loss": 0.6275, + "step": 2337 + }, + { + "epoch": 0.8952709170974535, + "grad_norm": 0.2946912348270416, + "learning_rate": 4.035932493289938e-06, + "loss": 0.6608, + "step": 2338 + }, + { + "epoch": 0.8956538387899674, + "grad_norm": 0.29266324639320374, + "learning_rate": 4.0351310068002434e-06, + "loss": 0.6578, + "step": 2339 + }, + { + "epoch": 0.8960367604824814, + "grad_norm": 0.2917732298374176, + "learning_rate": 4.034329266944068e-06, + "loss": 0.651, + "step": 2340 + }, + { + "epoch": 0.8964196821749952, + "grad_norm": 0.2940225303173065, + "learning_rate": 4.033527273853731e-06, + "loss": 0.647, + "step": 2341 + }, + { + "epoch": 0.8968026038675091, + "grad_norm": 0.29845407605171204, + "learning_rate": 4.032725027661601e-06, + "loss": 0.6519, + "step": 2342 + }, + { + "epoch": 0.897185525560023, + "grad_norm": 0.29687100648880005, + "learning_rate": 4.031922528500085e-06, + "loss": 0.6442, + "step": 2343 + }, + { + "epoch": 0.8975684472525368, + "grad_norm": 0.29440659284591675, + "learning_rate": 4.03111977650163e-06, + "loss": 0.6609, + "step": 2344 + }, + { + "epoch": 0.8979513689450508, + "grad_norm": 0.2876416742801666, + "learning_rate": 4.030316771798728e-06, + "loss": 0.6568, + "step": 2345 + }, + { + "epoch": 0.8983342906375646, + "grad_norm": 0.2890982925891876, + "learning_rate": 4.029513514523911e-06, + "loss": 0.6433, + "step": 2346 + }, + { + "epoch": 0.8987172123300785, + "grad_norm": 0.2900437116622925, + "learning_rate": 4.028710004809753e-06, + "loss": 0.6384, + "step": 2347 + }, + { + "epoch": 0.8991001340225924, + "grad_norm": 0.2865522801876068, + "learning_rate": 4.027906242788869e-06, + "loss": 0.662, + "step": 2348 + }, + { + "epoch": 0.8994830557151062, + "grad_norm": 0.2898208498954773, + "learning_rate": 4.0271022285939175e-06, + "loss": 0.648, + "step": 2349 + }, + { + "epoch": 0.8998659774076201, + "grad_norm": 0.2918628454208374, + "learning_rate": 4.026297962357598e-06, + "loss": 0.6657, + "step": 2350 + }, + { + "epoch": 0.9002488991001341, + "grad_norm": 0.28812891244888306, + "learning_rate": 4.0254934442126495e-06, + "loss": 0.672, + "step": 2351 + }, + { + "epoch": 0.9006318207926479, + "grad_norm": 0.2973209023475647, + "learning_rate": 4.024688674291854e-06, + "loss": 0.6431, + "step": 2352 + }, + { + "epoch": 0.9010147424851618, + "grad_norm": 0.3002658486366272, + "learning_rate": 4.023883652728036e-06, + "loss": 0.6493, + "step": 2353 + }, + { + "epoch": 0.9013976641776756, + "grad_norm": 0.2987782061100006, + "learning_rate": 4.0230783796540595e-06, + "loss": 0.6589, + "step": 2354 + }, + { + "epoch": 0.9017805858701895, + "grad_norm": 0.3019258975982666, + "learning_rate": 4.022272855202833e-06, + "loss": 0.6543, + "step": 2355 + }, + { + "epoch": 0.9021635075627035, + "grad_norm": 0.30396032333374023, + "learning_rate": 4.021467079507303e-06, + "loss": 0.6438, + "step": 2356 + }, + { + "epoch": 0.9025464292552173, + "grad_norm": 0.2919063866138458, + "learning_rate": 4.020661052700462e-06, + "loss": 0.6235, + "step": 2357 + }, + { + "epoch": 0.9029293509477312, + "grad_norm": 0.2995392382144928, + "learning_rate": 4.0198547749153376e-06, + "loss": 0.6538, + "step": 2358 + }, + { + "epoch": 0.903312272640245, + "grad_norm": 0.2881810963153839, + "learning_rate": 4.019048246285004e-06, + "loss": 0.6459, + "step": 2359 + }, + { + "epoch": 0.9036951943327589, + "grad_norm": 0.27683988213539124, + "learning_rate": 4.018241466942575e-06, + "loss": 0.6679, + "step": 2360 + }, + { + "epoch": 0.9040781160252729, + "grad_norm": 0.3057282865047455, + "learning_rate": 4.017434437021206e-06, + "loss": 0.6717, + "step": 2361 + }, + { + "epoch": 0.9044610377177867, + "grad_norm": 0.2937275767326355, + "learning_rate": 4.016627156654094e-06, + "loss": 0.6344, + "step": 2362 + }, + { + "epoch": 0.9048439594103006, + "grad_norm": 0.28433698415756226, + "learning_rate": 4.015819625974475e-06, + "loss": 0.6257, + "step": 2363 + }, + { + "epoch": 0.9052268811028145, + "grad_norm": 0.30163976550102234, + "learning_rate": 4.015011845115631e-06, + "loss": 0.6455, + "step": 2364 + }, + { + "epoch": 0.9056098027953283, + "grad_norm": 0.2907462418079376, + "learning_rate": 4.014203814210882e-06, + "loss": 0.6772, + "step": 2365 + }, + { + "epoch": 0.9059927244878422, + "grad_norm": 0.30592411756515503, + "learning_rate": 4.0133955333935884e-06, + "loss": 0.6606, + "step": 2366 + }, + { + "epoch": 0.9063756461803562, + "grad_norm": 0.2915680706501007, + "learning_rate": 4.012587002797154e-06, + "loss": 0.6497, + "step": 2367 + }, + { + "epoch": 0.90675856787287, + "grad_norm": 0.31376343965530396, + "learning_rate": 4.0117782225550245e-06, + "loss": 0.6459, + "step": 2368 + }, + { + "epoch": 0.9071414895653839, + "grad_norm": 0.29236075282096863, + "learning_rate": 4.010969192800684e-06, + "loss": 0.6547, + "step": 2369 + }, + { + "epoch": 0.9075244112578977, + "grad_norm": 0.2886476516723633, + "learning_rate": 4.010159913667659e-06, + "loss": 0.664, + "step": 2370 + }, + { + "epoch": 0.9079073329504116, + "grad_norm": 0.30533117055892944, + "learning_rate": 4.0093503852895176e-06, + "loss": 0.6572, + "step": 2371 + }, + { + "epoch": 0.9082902546429256, + "grad_norm": 0.3040546178817749, + "learning_rate": 4.00854060779987e-06, + "loss": 0.6462, + "step": 2372 + }, + { + "epoch": 0.9086731763354394, + "grad_norm": 0.28597450256347656, + "learning_rate": 4.007730581332366e-06, + "loss": 0.6433, + "step": 2373 + }, + { + "epoch": 0.9090560980279533, + "grad_norm": 0.27774664759635925, + "learning_rate": 4.006920306020696e-06, + "loss": 0.6539, + "step": 2374 + }, + { + "epoch": 0.9094390197204671, + "grad_norm": 0.29287946224212646, + "learning_rate": 4.006109781998593e-06, + "loss": 0.662, + "step": 2375 + }, + { + "epoch": 0.909821941412981, + "grad_norm": 0.3047502636909485, + "learning_rate": 4.005299009399829e-06, + "loss": 0.6461, + "step": 2376 + }, + { + "epoch": 0.910204863105495, + "grad_norm": 0.28814053535461426, + "learning_rate": 4.00448798835822e-06, + "loss": 0.6602, + "step": 2377 + }, + { + "epoch": 0.9105877847980088, + "grad_norm": 0.28405508399009705, + "learning_rate": 4.00367671900762e-06, + "loss": 0.6353, + "step": 2378 + }, + { + "epoch": 0.9109707064905227, + "grad_norm": 0.29485368728637695, + "learning_rate": 4.0028652014819276e-06, + "loss": 0.6547, + "step": 2379 + }, + { + "epoch": 0.9113536281830366, + "grad_norm": 0.28571370244026184, + "learning_rate": 4.002053435915078e-06, + "loss": 0.6387, + "step": 2380 + }, + { + "epoch": 0.9117365498755504, + "grad_norm": 0.2909572124481201, + "learning_rate": 4.00124142244105e-06, + "loss": 0.6237, + "step": 2381 + }, + { + "epoch": 0.9121194715680643, + "grad_norm": 0.29019856452941895, + "learning_rate": 4.000429161193863e-06, + "loss": 0.6357, + "step": 2382 + }, + { + "epoch": 0.9125023932605782, + "grad_norm": 0.30379518866539, + "learning_rate": 3.999616652307576e-06, + "loss": 0.6513, + "step": 2383 + }, + { + "epoch": 0.9128853149530921, + "grad_norm": 0.279227077960968, + "learning_rate": 3.998803895916292e-06, + "loss": 0.6447, + "step": 2384 + }, + { + "epoch": 0.913268236645606, + "grad_norm": 0.2874331772327423, + "learning_rate": 3.9979908921541525e-06, + "loss": 0.6471, + "step": 2385 + }, + { + "epoch": 0.9136511583381198, + "grad_norm": 0.298424631357193, + "learning_rate": 3.997177641155338e-06, + "loss": 0.6438, + "step": 2386 + }, + { + "epoch": 0.9140340800306337, + "grad_norm": 0.291726678609848, + "learning_rate": 3.996364143054074e-06, + "loss": 0.6654, + "step": 2387 + }, + { + "epoch": 0.9144170017231477, + "grad_norm": 0.29213789105415344, + "learning_rate": 3.995550397984624e-06, + "loss": 0.6605, + "step": 2388 + }, + { + "epoch": 0.9147999234156615, + "grad_norm": 0.290400892496109, + "learning_rate": 3.9947364060812925e-06, + "loss": 0.6615, + "step": 2389 + }, + { + "epoch": 0.9151828451081754, + "grad_norm": 0.3002352714538574, + "learning_rate": 3.993922167478426e-06, + "loss": 0.6687, + "step": 2390 + }, + { + "epoch": 0.9155657668006892, + "grad_norm": 0.2931500971317291, + "learning_rate": 3.993107682310411e-06, + "loss": 0.6399, + "step": 2391 + }, + { + "epoch": 0.9159486884932031, + "grad_norm": 0.30621516704559326, + "learning_rate": 3.992292950711674e-06, + "loss": 0.6578, + "step": 2392 + }, + { + "epoch": 0.9163316101857171, + "grad_norm": 0.304832398891449, + "learning_rate": 3.991477972816682e-06, + "loss": 0.6457, + "step": 2393 + }, + { + "epoch": 0.9167145318782309, + "grad_norm": 0.2896204888820648, + "learning_rate": 3.990662748759946e-06, + "loss": 0.665, + "step": 2394 + }, + { + "epoch": 0.9170974535707448, + "grad_norm": 0.3004479706287384, + "learning_rate": 3.989847278676012e-06, + "loss": 0.6534, + "step": 2395 + }, + { + "epoch": 0.9174803752632587, + "grad_norm": 0.2964184284210205, + "learning_rate": 3.989031562699472e-06, + "loss": 0.6652, + "step": 2396 + }, + { + "epoch": 0.9178632969557725, + "grad_norm": 0.30814141035079956, + "learning_rate": 3.988215600964956e-06, + "loss": 0.6453, + "step": 2397 + }, + { + "epoch": 0.9182462186482864, + "grad_norm": 0.297821968793869, + "learning_rate": 3.987399393607134e-06, + "loss": 0.6634, + "step": 2398 + }, + { + "epoch": 0.9186291403408003, + "grad_norm": 0.307809978723526, + "learning_rate": 3.986582940760717e-06, + "loss": 0.6391, + "step": 2399 + }, + { + "epoch": 0.9190120620333142, + "grad_norm": 0.2985922396183014, + "learning_rate": 3.985766242560458e-06, + "loss": 0.6593, + "step": 2400 + }, + { + "epoch": 0.9193949837258281, + "grad_norm": 0.2958591878414154, + "learning_rate": 3.984949299141149e-06, + "loss": 0.6563, + "step": 2401 + }, + { + "epoch": 0.9197779054183419, + "grad_norm": 0.29857268929481506, + "learning_rate": 3.984132110637622e-06, + "loss": 0.6599, + "step": 2402 + }, + { + "epoch": 0.9201608271108558, + "grad_norm": 0.297341525554657, + "learning_rate": 3.983314677184752e-06, + "loss": 0.6493, + "step": 2403 + }, + { + "epoch": 0.9205437488033698, + "grad_norm": 0.30079755187034607, + "learning_rate": 3.982496998917451e-06, + "loss": 0.6528, + "step": 2404 + }, + { + "epoch": 0.9209266704958836, + "grad_norm": 0.29261571168899536, + "learning_rate": 3.981679075970675e-06, + "loss": 0.6587, + "step": 2405 + }, + { + "epoch": 0.9213095921883975, + "grad_norm": 0.2840799391269684, + "learning_rate": 3.980860908479417e-06, + "loss": 0.6337, + "step": 2406 + }, + { + "epoch": 0.9216925138809113, + "grad_norm": 0.29934579133987427, + "learning_rate": 3.9800424965787115e-06, + "loss": 0.6602, + "step": 2407 + }, + { + "epoch": 0.9220754355734252, + "grad_norm": 0.2924461364746094, + "learning_rate": 3.979223840403636e-06, + "loss": 0.6252, + "step": 2408 + }, + { + "epoch": 0.9224583572659392, + "grad_norm": 0.2996419668197632, + "learning_rate": 3.978404940089302e-06, + "loss": 0.6351, + "step": 2409 + }, + { + "epoch": 0.922841278958453, + "grad_norm": 0.28820133209228516, + "learning_rate": 3.9775857957708695e-06, + "loss": 0.6527, + "step": 2410 + }, + { + "epoch": 0.9232242006509669, + "grad_norm": 0.2895001471042633, + "learning_rate": 3.976766407583533e-06, + "loss": 0.6568, + "step": 2411 + }, + { + "epoch": 0.9236071223434807, + "grad_norm": 0.2845630943775177, + "learning_rate": 3.975946775662528e-06, + "loss": 0.6405, + "step": 2412 + }, + { + "epoch": 0.9239900440359946, + "grad_norm": 0.2985185384750366, + "learning_rate": 3.975126900143133e-06, + "loss": 0.6695, + "step": 2413 + }, + { + "epoch": 0.9243729657285085, + "grad_norm": 0.29201918840408325, + "learning_rate": 3.974306781160664e-06, + "loss": 0.6376, + "step": 2414 + }, + { + "epoch": 0.9247558874210224, + "grad_norm": 0.2998596429824829, + "learning_rate": 3.973486418850476e-06, + "loss": 0.6619, + "step": 2415 + }, + { + "epoch": 0.9251388091135363, + "grad_norm": 0.2949308156967163, + "learning_rate": 3.9726658133479676e-06, + "loss": 0.6558, + "step": 2416 + }, + { + "epoch": 0.9255217308060502, + "grad_norm": 0.28473496437072754, + "learning_rate": 3.971844964788577e-06, + "loss": 0.6446, + "step": 2417 + }, + { + "epoch": 0.925904652498564, + "grad_norm": 0.29481175541877747, + "learning_rate": 3.971023873307781e-06, + "loss": 0.66, + "step": 2418 + }, + { + "epoch": 0.9262875741910779, + "grad_norm": 0.30271342396736145, + "learning_rate": 3.970202539041096e-06, + "loss": 0.6628, + "step": 2419 + }, + { + "epoch": 0.9266704958835918, + "grad_norm": 0.30491283535957336, + "learning_rate": 3.969380962124081e-06, + "loss": 0.6553, + "step": 2420 + }, + { + "epoch": 0.9270534175761057, + "grad_norm": 0.2750245928764343, + "learning_rate": 3.968559142692333e-06, + "loss": 0.6516, + "step": 2421 + }, + { + "epoch": 0.9274363392686196, + "grad_norm": 0.2851044535636902, + "learning_rate": 3.9677370808814885e-06, + "loss": 0.6303, + "step": 2422 + }, + { + "epoch": 0.9278192609611334, + "grad_norm": 0.29597732424736023, + "learning_rate": 3.966914776827227e-06, + "loss": 0.6486, + "step": 2423 + }, + { + "epoch": 0.9282021826536473, + "grad_norm": 0.29492276906967163, + "learning_rate": 3.966092230665265e-06, + "loss": 0.6677, + "step": 2424 + }, + { + "epoch": 0.9285851043461613, + "grad_norm": 0.2950708866119385, + "learning_rate": 3.9652694425313606e-06, + "loss": 0.6562, + "step": 2425 + }, + { + "epoch": 0.9289680260386751, + "grad_norm": 0.29869431257247925, + "learning_rate": 3.96444641256131e-06, + "loss": 0.6403, + "step": 2426 + }, + { + "epoch": 0.929350947731189, + "grad_norm": 0.29527124762535095, + "learning_rate": 3.963623140890952e-06, + "loss": 0.6691, + "step": 2427 + }, + { + "epoch": 0.9297338694237028, + "grad_norm": 0.2984407842159271, + "learning_rate": 3.962799627656163e-06, + "loss": 0.6479, + "step": 2428 + }, + { + "epoch": 0.9301167911162167, + "grad_norm": 0.3068842887878418, + "learning_rate": 3.9619758729928624e-06, + "loss": 0.6341, + "step": 2429 + }, + { + "epoch": 0.9304997128087306, + "grad_norm": 0.2958669364452362, + "learning_rate": 3.961151877037004e-06, + "loss": 0.6526, + "step": 2430 + }, + { + "epoch": 0.9308826345012445, + "grad_norm": 0.30777838826179504, + "learning_rate": 3.9603276399245864e-06, + "loss": 0.6617, + "step": 2431 + }, + { + "epoch": 0.9312655561937584, + "grad_norm": 0.29739493131637573, + "learning_rate": 3.959503161791645e-06, + "loss": 0.6562, + "step": 2432 + }, + { + "epoch": 0.9316484778862723, + "grad_norm": 0.29062512516975403, + "learning_rate": 3.958678442774259e-06, + "loss": 0.6568, + "step": 2433 + }, + { + "epoch": 0.9320313995787861, + "grad_norm": 0.2967102825641632, + "learning_rate": 3.957853483008541e-06, + "loss": 0.6532, + "step": 2434 + }, + { + "epoch": 0.9324143212713, + "grad_norm": 0.3010478615760803, + "learning_rate": 3.957028282630649e-06, + "loss": 0.6368, + "step": 2435 + }, + { + "epoch": 0.9327972429638139, + "grad_norm": 0.3022819459438324, + "learning_rate": 3.956202841776779e-06, + "loss": 0.6658, + "step": 2436 + }, + { + "epoch": 0.9331801646563278, + "grad_norm": 0.29249346256256104, + "learning_rate": 3.955377160583165e-06, + "loss": 0.647, + "step": 2437 + }, + { + "epoch": 0.9335630863488417, + "grad_norm": 0.29361385107040405, + "learning_rate": 3.954551239186082e-06, + "loss": 0.6237, + "step": 2438 + }, + { + "epoch": 0.9339460080413555, + "grad_norm": 0.2925169765949249, + "learning_rate": 3.953725077721846e-06, + "loss": 0.6477, + "step": 2439 + }, + { + "epoch": 0.9343289297338694, + "grad_norm": 0.2952122688293457, + "learning_rate": 3.95289867632681e-06, + "loss": 0.639, + "step": 2440 + }, + { + "epoch": 0.9347118514263834, + "grad_norm": 0.2906058728694916, + "learning_rate": 3.952072035137368e-06, + "loss": 0.6342, + "step": 2441 + }, + { + "epoch": 0.9350947731188972, + "grad_norm": 0.28912121057510376, + "learning_rate": 3.9512451542899534e-06, + "loss": 0.6512, + "step": 2442 + }, + { + "epoch": 0.9354776948114111, + "grad_norm": 0.28959429264068604, + "learning_rate": 3.95041803392104e-06, + "loss": 0.6387, + "step": 2443 + }, + { + "epoch": 0.9358606165039249, + "grad_norm": 0.2835792899131775, + "learning_rate": 3.94959067416714e-06, + "loss": 0.6477, + "step": 2444 + }, + { + "epoch": 0.9362435381964388, + "grad_norm": 0.3021864593029022, + "learning_rate": 3.948763075164804e-06, + "loss": 0.6412, + "step": 2445 + }, + { + "epoch": 0.9366264598889527, + "grad_norm": 0.3023189902305603, + "learning_rate": 3.947935237050625e-06, + "loss": 0.6636, + "step": 2446 + }, + { + "epoch": 0.9370093815814666, + "grad_norm": 0.299590140581131, + "learning_rate": 3.947107159961234e-06, + "loss": 0.641, + "step": 2447 + }, + { + "epoch": 0.9373923032739805, + "grad_norm": 0.2920341193675995, + "learning_rate": 3.946278844033301e-06, + "loss": 0.6586, + "step": 2448 + }, + { + "epoch": 0.9377752249664943, + "grad_norm": 0.3043041527271271, + "learning_rate": 3.945450289403535e-06, + "loss": 0.6569, + "step": 2449 + }, + { + "epoch": 0.9381581466590082, + "grad_norm": 0.3054685592651367, + "learning_rate": 3.944621496208687e-06, + "loss": 0.6608, + "step": 2450 + }, + { + "epoch": 0.9385410683515221, + "grad_norm": 0.2984093725681305, + "learning_rate": 3.943792464585544e-06, + "loss": 0.643, + "step": 2451 + }, + { + "epoch": 0.938923990044036, + "grad_norm": 0.2824691832065582, + "learning_rate": 3.942963194670934e-06, + "loss": 0.6291, + "step": 2452 + }, + { + "epoch": 0.9393069117365499, + "grad_norm": 0.3064553439617157, + "learning_rate": 3.942133686601727e-06, + "loss": 0.6465, + "step": 2453 + }, + { + "epoch": 0.9396898334290638, + "grad_norm": 0.299063503742218, + "learning_rate": 3.941303940514826e-06, + "loss": 0.6309, + "step": 2454 + }, + { + "epoch": 0.9400727551215776, + "grad_norm": 0.29257938265800476, + "learning_rate": 3.940473956547177e-06, + "loss": 0.627, + "step": 2455 + }, + { + "epoch": 0.9404556768140915, + "grad_norm": 0.2964414656162262, + "learning_rate": 3.939643734835768e-06, + "loss": 0.649, + "step": 2456 + }, + { + "epoch": 0.9408385985066055, + "grad_norm": 0.28775110840797424, + "learning_rate": 3.938813275517622e-06, + "loss": 0.6679, + "step": 2457 + }, + { + "epoch": 0.9412215201991193, + "grad_norm": 0.28418123722076416, + "learning_rate": 3.937982578729802e-06, + "loss": 0.659, + "step": 2458 + }, + { + "epoch": 0.9416044418916332, + "grad_norm": 0.2941417694091797, + "learning_rate": 3.937151644609412e-06, + "loss": 0.6513, + "step": 2459 + }, + { + "epoch": 0.941987363584147, + "grad_norm": 0.29704391956329346, + "learning_rate": 3.936320473293592e-06, + "loss": 0.6446, + "step": 2460 + }, + { + "epoch": 0.9423702852766609, + "grad_norm": 0.2998020350933075, + "learning_rate": 3.9354890649195246e-06, + "loss": 0.658, + "step": 2461 + }, + { + "epoch": 0.9427532069691748, + "grad_norm": 0.28896981477737427, + "learning_rate": 3.934657419624429e-06, + "loss": 0.6569, + "step": 2462 + }, + { + "epoch": 0.9431361286616887, + "grad_norm": 0.2883463203907013, + "learning_rate": 3.933825537545566e-06, + "loss": 0.6553, + "step": 2463 + }, + { + "epoch": 0.9435190503542026, + "grad_norm": 0.29482680559158325, + "learning_rate": 3.932993418820233e-06, + "loss": 0.6549, + "step": 2464 + }, + { + "epoch": 0.9439019720467164, + "grad_norm": 0.2945811450481415, + "learning_rate": 3.9321610635857675e-06, + "loss": 0.6426, + "step": 2465 + }, + { + "epoch": 0.9442848937392303, + "grad_norm": 0.29945310950279236, + "learning_rate": 3.931328471979545e-06, + "loss": 0.6511, + "step": 2466 + }, + { + "epoch": 0.9446678154317442, + "grad_norm": 0.30127647519111633, + "learning_rate": 3.930495644138984e-06, + "loss": 0.6652, + "step": 2467 + }, + { + "epoch": 0.9450507371242581, + "grad_norm": 0.29690733551979065, + "learning_rate": 3.929662580201536e-06, + "loss": 0.6691, + "step": 2468 + }, + { + "epoch": 0.945433658816772, + "grad_norm": 0.2941737473011017, + "learning_rate": 3.9288292803046954e-06, + "loss": 0.6687, + "step": 2469 + }, + { + "epoch": 0.9458165805092859, + "grad_norm": 0.29528361558914185, + "learning_rate": 3.927995744585995e-06, + "loss": 0.6419, + "step": 2470 + }, + { + "epoch": 0.9461995022017997, + "grad_norm": 0.2993655502796173, + "learning_rate": 3.927161973183007e-06, + "loss": 0.6709, + "step": 2471 + }, + { + "epoch": 0.9465824238943136, + "grad_norm": 0.2861931025981903, + "learning_rate": 3.92632796623334e-06, + "loss": 0.6465, + "step": 2472 + }, + { + "epoch": 0.9469653455868275, + "grad_norm": 0.2920815944671631, + "learning_rate": 3.925493723874643e-06, + "loss": 0.6495, + "step": 2473 + }, + { + "epoch": 0.9473482672793414, + "grad_norm": 0.28802213072776794, + "learning_rate": 3.924659246244604e-06, + "loss": 0.6422, + "step": 2474 + }, + { + "epoch": 0.9477311889718553, + "grad_norm": 0.28447744250297546, + "learning_rate": 3.9238245334809525e-06, + "loss": 0.6422, + "step": 2475 + }, + { + "epoch": 0.9481141106643691, + "grad_norm": 0.3059292137622833, + "learning_rate": 3.92298958572145e-06, + "loss": 0.6537, + "step": 2476 + }, + { + "epoch": 0.948497032356883, + "grad_norm": 0.2921289801597595, + "learning_rate": 3.922154403103904e-06, + "loss": 0.6334, + "step": 2477 + }, + { + "epoch": 0.9488799540493968, + "grad_norm": 0.2936064600944519, + "learning_rate": 3.921318985766156e-06, + "loss": 0.6615, + "step": 2478 + }, + { + "epoch": 0.9492628757419108, + "grad_norm": 0.285410076379776, + "learning_rate": 3.920483333846089e-06, + "loss": 0.63, + "step": 2479 + }, + { + "epoch": 0.9496457974344247, + "grad_norm": 0.2900061011314392, + "learning_rate": 3.9196474474816225e-06, + "loss": 0.6462, + "step": 2480 + }, + { + "epoch": 0.9500287191269385, + "grad_norm": 0.2814337909221649, + "learning_rate": 3.918811326810717e-06, + "loss": 0.6745, + "step": 2481 + }, + { + "epoch": 0.9504116408194524, + "grad_norm": 0.29532554745674133, + "learning_rate": 3.917974971971369e-06, + "loss": 0.6406, + "step": 2482 + }, + { + "epoch": 0.9507945625119663, + "grad_norm": 0.2889082729816437, + "learning_rate": 3.917138383101616e-06, + "loss": 0.6555, + "step": 2483 + }, + { + "epoch": 0.9511774842044802, + "grad_norm": 0.29341015219688416, + "learning_rate": 3.916301560339533e-06, + "loss": 0.6534, + "step": 2484 + }, + { + "epoch": 0.9515604058969941, + "grad_norm": 0.2957211136817932, + "learning_rate": 3.915464503823234e-06, + "loss": 0.6656, + "step": 2485 + }, + { + "epoch": 0.951943327589508, + "grad_norm": 0.30043575167655945, + "learning_rate": 3.9146272136908715e-06, + "loss": 0.6657, + "step": 2486 + }, + { + "epoch": 0.9523262492820218, + "grad_norm": 0.2976532578468323, + "learning_rate": 3.913789690080636e-06, + "loss": 0.6578, + "step": 2487 + }, + { + "epoch": 0.9527091709745357, + "grad_norm": 0.27795541286468506, + "learning_rate": 3.912951933130758e-06, + "loss": 0.6348, + "step": 2488 + }, + { + "epoch": 0.9530920926670496, + "grad_norm": 0.29307544231414795, + "learning_rate": 3.912113942979504e-06, + "loss": 0.6661, + "step": 2489 + }, + { + "epoch": 0.9534750143595635, + "grad_norm": 0.2948361933231354, + "learning_rate": 3.911275719765182e-06, + "loss": 0.6552, + "step": 2490 + }, + { + "epoch": 0.9538579360520774, + "grad_norm": 0.29154619574546814, + "learning_rate": 3.910437263626137e-06, + "loss": 0.635, + "step": 2491 + }, + { + "epoch": 0.9542408577445912, + "grad_norm": 0.2864062488079071, + "learning_rate": 3.909598574700751e-06, + "loss": 0.6311, + "step": 2492 + }, + { + "epoch": 0.9546237794371051, + "grad_norm": 0.3037509322166443, + "learning_rate": 3.908759653127446e-06, + "loss": 0.6497, + "step": 2493 + }, + { + "epoch": 0.9550067011296189, + "grad_norm": 0.3685162663459778, + "learning_rate": 3.907920499044684e-06, + "loss": 0.6597, + "step": 2494 + }, + { + "epoch": 0.9553896228221329, + "grad_norm": 0.3015262186527252, + "learning_rate": 3.907081112590962e-06, + "loss": 0.6539, + "step": 2495 + }, + { + "epoch": 0.9557725445146468, + "grad_norm": 0.29515382647514343, + "learning_rate": 3.906241493904818e-06, + "loss": 0.6462, + "step": 2496 + }, + { + "epoch": 0.9561554662071606, + "grad_norm": 0.29523947834968567, + "learning_rate": 3.905401643124826e-06, + "loss": 0.6477, + "step": 2497 + }, + { + "epoch": 0.9565383878996745, + "grad_norm": 0.3039727509021759, + "learning_rate": 3.904561560389601e-06, + "loss": 0.6656, + "step": 2498 + }, + { + "epoch": 0.9569213095921884, + "grad_norm": 0.29031190276145935, + "learning_rate": 3.903721245837794e-06, + "loss": 0.6619, + "step": 2499 + }, + { + "epoch": 0.9573042312847023, + "grad_norm": 0.28732046484947205, + "learning_rate": 3.902880699608095e-06, + "loss": 0.6615, + "step": 2500 + }, + { + "epoch": 0.9576871529772162, + "grad_norm": 0.292327344417572, + "learning_rate": 3.902039921839233e-06, + "loss": 0.6526, + "step": 2501 + }, + { + "epoch": 0.95807007466973, + "grad_norm": 0.30777719616889954, + "learning_rate": 3.9011989126699746e-06, + "loss": 0.6652, + "step": 2502 + }, + { + "epoch": 0.9584529963622439, + "grad_norm": 0.286577969789505, + "learning_rate": 3.900357672239125e-06, + "loss": 0.6277, + "step": 2503 + }, + { + "epoch": 0.9588359180547578, + "grad_norm": 0.30643603205680847, + "learning_rate": 3.899516200685525e-06, + "loss": 0.6704, + "step": 2504 + }, + { + "epoch": 0.9592188397472717, + "grad_norm": 0.3036949932575226, + "learning_rate": 3.898674498148058e-06, + "loss": 0.6674, + "step": 2505 + }, + { + "epoch": 0.9596017614397856, + "grad_norm": 0.29391995072364807, + "learning_rate": 3.8978325647656426e-06, + "loss": 0.6631, + "step": 2506 + }, + { + "epoch": 0.9599846831322995, + "grad_norm": 0.2952329218387604, + "learning_rate": 3.896990400677235e-06, + "loss": 0.6674, + "step": 2507 + }, + { + "epoch": 0.9603676048248133, + "grad_norm": 0.3082561492919922, + "learning_rate": 3.896148006021833e-06, + "loss": 0.6497, + "step": 2508 + }, + { + "epoch": 0.9607505265173272, + "grad_norm": 0.2941219210624695, + "learning_rate": 3.895305380938468e-06, + "loss": 0.6625, + "step": 2509 + }, + { + "epoch": 0.961133448209841, + "grad_norm": 0.2878792881965637, + "learning_rate": 3.894462525566212e-06, + "loss": 0.6405, + "step": 2510 + }, + { + "epoch": 0.961516369902355, + "grad_norm": 0.2885337471961975, + "learning_rate": 3.893619440044174e-06, + "loss": 0.6448, + "step": 2511 + }, + { + "epoch": 0.9618992915948689, + "grad_norm": 0.29196929931640625, + "learning_rate": 3.892776124511503e-06, + "loss": 0.6434, + "step": 2512 + }, + { + "epoch": 0.9622822132873827, + "grad_norm": 0.3003067970275879, + "learning_rate": 3.891932579107384e-06, + "loss": 0.6387, + "step": 2513 + }, + { + "epoch": 0.9626651349798966, + "grad_norm": 0.2982931435108185, + "learning_rate": 3.8910888039710395e-06, + "loss": 0.6637, + "step": 2514 + }, + { + "epoch": 0.9630480566724104, + "grad_norm": 0.3193811774253845, + "learning_rate": 3.8902447992417305e-06, + "loss": 0.6677, + "step": 2515 + }, + { + "epoch": 0.9634309783649244, + "grad_norm": 0.29214462637901306, + "learning_rate": 3.889400565058758e-06, + "loss": 0.6428, + "step": 2516 + }, + { + "epoch": 0.9638139000574383, + "grad_norm": 0.3077375590801239, + "learning_rate": 3.888556101561457e-06, + "loss": 0.6399, + "step": 2517 + }, + { + "epoch": 0.9641968217499521, + "grad_norm": 0.30510959029197693, + "learning_rate": 3.887711408889204e-06, + "loss": 0.643, + "step": 2518 + }, + { + "epoch": 0.964579743442466, + "grad_norm": 0.2839040160179138, + "learning_rate": 3.886866487181411e-06, + "loss": 0.6301, + "step": 2519 + }, + { + "epoch": 0.9649626651349799, + "grad_norm": 0.28605926036834717, + "learning_rate": 3.886021336577528e-06, + "loss": 0.655, + "step": 2520 + }, + { + "epoch": 0.9653455868274938, + "grad_norm": 0.3065052628517151, + "learning_rate": 3.885175957217045e-06, + "loss": 0.6483, + "step": 2521 + }, + { + "epoch": 0.9657285085200077, + "grad_norm": 0.282900333404541, + "learning_rate": 3.884330349239486e-06, + "loss": 0.6417, + "step": 2522 + }, + { + "epoch": 0.9661114302125215, + "grad_norm": 0.3028445541858673, + "learning_rate": 3.883484512784417e-06, + "loss": 0.6474, + "step": 2523 + }, + { + "epoch": 0.9664943519050354, + "grad_norm": 0.2901262044906616, + "learning_rate": 3.882638447991438e-06, + "loss": 0.6533, + "step": 2524 + }, + { + "epoch": 0.9668772735975493, + "grad_norm": 0.3081863820552826, + "learning_rate": 3.881792155000188e-06, + "loss": 0.6544, + "step": 2525 + }, + { + "epoch": 0.9672601952900631, + "grad_norm": 0.2975146174430847, + "learning_rate": 3.880945633950345e-06, + "loss": 0.6488, + "step": 2526 + }, + { + "epoch": 0.9676431169825771, + "grad_norm": 0.3060862720012665, + "learning_rate": 3.880098884981622e-06, + "loss": 0.645, + "step": 2527 + }, + { + "epoch": 0.968026038675091, + "grad_norm": 0.2798968255519867, + "learning_rate": 3.879251908233773e-06, + "loss": 0.6332, + "step": 2528 + }, + { + "epoch": 0.9684089603676048, + "grad_norm": 0.30223652720451355, + "learning_rate": 3.878404703846586e-06, + "loss": 0.6519, + "step": 2529 + }, + { + "epoch": 0.9687918820601187, + "grad_norm": 0.30503448843955994, + "learning_rate": 3.8775572719598896e-06, + "loss": 0.6669, + "step": 2530 + }, + { + "epoch": 0.9691748037526325, + "grad_norm": 0.29106512665748596, + "learning_rate": 3.876709612713548e-06, + "loss": 0.6663, + "step": 2531 + }, + { + "epoch": 0.9695577254451465, + "grad_norm": 0.28600332140922546, + "learning_rate": 3.875861726247464e-06, + "loss": 0.6219, + "step": 2532 + }, + { + "epoch": 0.9699406471376604, + "grad_norm": 0.29538947343826294, + "learning_rate": 3.875013612701577e-06, + "loss": 0.6668, + "step": 2533 + }, + { + "epoch": 0.9703235688301742, + "grad_norm": 0.2874445617198944, + "learning_rate": 3.874165272215863e-06, + "loss": 0.6328, + "step": 2534 + }, + { + "epoch": 0.9707064905226881, + "grad_norm": 0.29347294569015503, + "learning_rate": 3.87331670493034e-06, + "loss": 0.6546, + "step": 2535 + }, + { + "epoch": 0.971089412215202, + "grad_norm": 0.29179924726486206, + "learning_rate": 3.8724679109850574e-06, + "loss": 0.6458, + "step": 2536 + }, + { + "epoch": 0.9714723339077159, + "grad_norm": 0.2879701256752014, + "learning_rate": 3.871618890520105e-06, + "loss": 0.6203, + "step": 2537 + }, + { + "epoch": 0.9718552556002298, + "grad_norm": 0.2921457886695862, + "learning_rate": 3.8707696436756125e-06, + "loss": 0.6372, + "step": 2538 + }, + { + "epoch": 0.9722381772927436, + "grad_norm": 0.294516921043396, + "learning_rate": 3.8699201705917416e-06, + "loss": 0.6376, + "step": 2539 + }, + { + "epoch": 0.9726210989852575, + "grad_norm": 0.28862977027893066, + "learning_rate": 3.869070471408694e-06, + "loss": 0.6449, + "step": 2540 + }, + { + "epoch": 0.9730040206777714, + "grad_norm": 0.308726966381073, + "learning_rate": 3.8682205462667115e-06, + "loss": 0.6462, + "step": 2541 + }, + { + "epoch": 0.9733869423702852, + "grad_norm": 0.28768783807754517, + "learning_rate": 3.8673703953060685e-06, + "loss": 0.6473, + "step": 2542 + }, + { + "epoch": 0.9737698640627992, + "grad_norm": 0.29856985807418823, + "learning_rate": 3.866520018667077e-06, + "loss": 0.6494, + "step": 2543 + }, + { + "epoch": 0.974152785755313, + "grad_norm": 0.295842707157135, + "learning_rate": 3.86566941649009e-06, + "loss": 0.6598, + "step": 2544 + }, + { + "epoch": 0.9745357074478269, + "grad_norm": 0.2955799102783203, + "learning_rate": 3.864818588915495e-06, + "loss": 0.6566, + "step": 2545 + }, + { + "epoch": 0.9749186291403408, + "grad_norm": 0.29329386353492737, + "learning_rate": 3.863967536083718e-06, + "loss": 0.6511, + "step": 2546 + }, + { + "epoch": 0.9753015508328546, + "grad_norm": 0.28482767939567566, + "learning_rate": 3.86311625813522e-06, + "loss": 0.6419, + "step": 2547 + }, + { + "epoch": 0.9756844725253686, + "grad_norm": 0.2978568971157074, + "learning_rate": 3.862264755210501e-06, + "loss": 0.6295, + "step": 2548 + }, + { + "epoch": 0.9760673942178825, + "grad_norm": 0.2871286869049072, + "learning_rate": 3.8614130274501e-06, + "loss": 0.6611, + "step": 2549 + }, + { + "epoch": 0.9764503159103963, + "grad_norm": 0.2937026917934418, + "learning_rate": 3.860561074994587e-06, + "loss": 0.6614, + "step": 2550 + }, + { + "epoch": 0.9768332376029102, + "grad_norm": 0.295292466878891, + "learning_rate": 3.859708897984575e-06, + "loss": 0.6334, + "step": 2551 + }, + { + "epoch": 0.977216159295424, + "grad_norm": 0.2879801094532013, + "learning_rate": 3.858856496560713e-06, + "loss": 0.6299, + "step": 2552 + }, + { + "epoch": 0.977599080987938, + "grad_norm": 0.2882961928844452, + "learning_rate": 3.858003870863684e-06, + "loss": 0.644, + "step": 2553 + }, + { + "epoch": 0.9779820026804519, + "grad_norm": 0.2908332943916321, + "learning_rate": 3.857151021034211e-06, + "loss": 0.63, + "step": 2554 + }, + { + "epoch": 0.9783649243729657, + "grad_norm": 0.29038068652153015, + "learning_rate": 3.8562979472130535e-06, + "loss": 0.6295, + "step": 2555 + }, + { + "epoch": 0.9787478460654796, + "grad_norm": 0.2989266812801361, + "learning_rate": 3.855444649541007e-06, + "loss": 0.6527, + "step": 2556 + }, + { + "epoch": 0.9791307677579935, + "grad_norm": 0.3030693829059601, + "learning_rate": 3.8545911281589046e-06, + "loss": 0.6557, + "step": 2557 + }, + { + "epoch": 0.9795136894505073, + "grad_norm": 0.2984217405319214, + "learning_rate": 3.853737383207615e-06, + "loss": 0.6572, + "step": 2558 + }, + { + "epoch": 0.9798966111430213, + "grad_norm": 0.3122209906578064, + "learning_rate": 3.852883414828046e-06, + "loss": 0.6373, + "step": 2559 + }, + { + "epoch": 0.9802795328355352, + "grad_norm": 0.2848958671092987, + "learning_rate": 3.8520292231611425e-06, + "loss": 0.6531, + "step": 2560 + }, + { + "epoch": 0.980662454528049, + "grad_norm": 0.3094291090965271, + "learning_rate": 3.851174808347883e-06, + "loss": 0.651, + "step": 2561 + }, + { + "epoch": 0.9810453762205629, + "grad_norm": 0.28635185956954956, + "learning_rate": 3.850320170529285e-06, + "loss": 0.63, + "step": 2562 + }, + { + "epoch": 0.9814282979130767, + "grad_norm": 0.28737953305244446, + "learning_rate": 3.849465309846404e-06, + "loss": 0.644, + "step": 2563 + }, + { + "epoch": 0.9818112196055907, + "grad_norm": 0.31178387999534607, + "learning_rate": 3.84861022644033e-06, + "loss": 0.6624, + "step": 2564 + }, + { + "epoch": 0.9821941412981046, + "grad_norm": 0.2955792248249054, + "learning_rate": 3.847754920452191e-06, + "loss": 0.6539, + "step": 2565 + }, + { + "epoch": 0.9825770629906184, + "grad_norm": 0.2956996560096741, + "learning_rate": 3.846899392023151e-06, + "loss": 0.659, + "step": 2566 + }, + { + "epoch": 0.9829599846831323, + "grad_norm": 0.3170894384384155, + "learning_rate": 3.846043641294411e-06, + "loss": 0.6562, + "step": 2567 + }, + { + "epoch": 0.9833429063756461, + "grad_norm": 0.3065207302570343, + "learning_rate": 3.84518766840721e-06, + "loss": 0.658, + "step": 2568 + }, + { + "epoch": 0.9837258280681601, + "grad_norm": 0.3012702465057373, + "learning_rate": 3.844331473502821e-06, + "loss": 0.6451, + "step": 2569 + }, + { + "epoch": 0.984108749760674, + "grad_norm": 0.3071759045124054, + "learning_rate": 3.843475056722555e-06, + "loss": 0.6721, + "step": 2570 + }, + { + "epoch": 0.9844916714531878, + "grad_norm": 0.29533374309539795, + "learning_rate": 3.842618418207762e-06, + "loss": 0.6434, + "step": 2571 + }, + { + "epoch": 0.9848745931457017, + "grad_norm": 0.29023638367652893, + "learning_rate": 3.841761558099824e-06, + "loss": 0.6353, + "step": 2572 + }, + { + "epoch": 0.9852575148382156, + "grad_norm": 0.3099059760570526, + "learning_rate": 3.840904476540164e-06, + "loss": 0.6499, + "step": 2573 + }, + { + "epoch": 0.9856404365307294, + "grad_norm": 0.305209219455719, + "learning_rate": 3.840047173670236e-06, + "loss": 0.6497, + "step": 2574 + }, + { + "epoch": 0.9860233582232434, + "grad_norm": 0.307308167219162, + "learning_rate": 3.839189649631538e-06, + "loss": 0.658, + "step": 2575 + }, + { + "epoch": 0.9864062799157572, + "grad_norm": 0.31121161580085754, + "learning_rate": 3.8383319045655985e-06, + "loss": 0.6529, + "step": 2576 + }, + { + "epoch": 0.9867892016082711, + "grad_norm": 0.2935795187950134, + "learning_rate": 3.837473938613984e-06, + "loss": 0.6491, + "step": 2577 + }, + { + "epoch": 0.987172123300785, + "grad_norm": 0.2984502911567688, + "learning_rate": 3.8366157519183e-06, + "loss": 0.636, + "step": 2578 + }, + { + "epoch": 0.9875550449932988, + "grad_norm": 0.3134182393550873, + "learning_rate": 3.835757344620183e-06, + "loss": 0.6476, + "step": 2579 + }, + { + "epoch": 0.9879379666858128, + "grad_norm": 0.3170115649700165, + "learning_rate": 3.8348987168613126e-06, + "loss": 0.6368, + "step": 2580 + }, + { + "epoch": 0.9883208883783267, + "grad_norm": 0.3063246011734009, + "learning_rate": 3.8340398687834e-06, + "loss": 0.6595, + "step": 2581 + }, + { + "epoch": 0.9887038100708405, + "grad_norm": 0.294207900762558, + "learning_rate": 3.8331808005281935e-06, + "loss": 0.6539, + "step": 2582 + }, + { + "epoch": 0.9890867317633544, + "grad_norm": 0.3054911494255066, + "learning_rate": 3.83232151223748e-06, + "loss": 0.6492, + "step": 2583 + }, + { + "epoch": 0.9894696534558682, + "grad_norm": 0.29723116755485535, + "learning_rate": 3.8314620040530805e-06, + "loss": 0.6324, + "step": 2584 + }, + { + "epoch": 0.9898525751483822, + "grad_norm": 0.2975267767906189, + "learning_rate": 3.830602276116853e-06, + "loss": 0.6455, + "step": 2585 + }, + { + "epoch": 0.9902354968408961, + "grad_norm": 0.31881144642829895, + "learning_rate": 3.829742328570691e-06, + "loss": 0.6548, + "step": 2586 + }, + { + "epoch": 0.9906184185334099, + "grad_norm": 0.30828386545181274, + "learning_rate": 3.828882161556527e-06, + "loss": 0.6653, + "step": 2587 + }, + { + "epoch": 0.9910013402259238, + "grad_norm": 0.30030110478401184, + "learning_rate": 3.828021775216325e-06, + "loss": 0.6333, + "step": 2588 + }, + { + "epoch": 0.9913842619184376, + "grad_norm": 0.298301100730896, + "learning_rate": 3.82716116969209e-06, + "loss": 0.6538, + "step": 2589 + }, + { + "epoch": 0.9917671836109515, + "grad_norm": 0.2987930476665497, + "learning_rate": 3.82630034512586e-06, + "loss": 0.6413, + "step": 2590 + }, + { + "epoch": 0.9921501053034655, + "grad_norm": 0.2933206856250763, + "learning_rate": 3.825439301659711e-06, + "loss": 0.6419, + "step": 2591 + }, + { + "epoch": 0.9925330269959793, + "grad_norm": 0.3178972005844116, + "learning_rate": 3.824578039435754e-06, + "loss": 0.6697, + "step": 2592 + }, + { + "epoch": 0.9929159486884932, + "grad_norm": 0.2987993657588959, + "learning_rate": 3.8237165585961355e-06, + "loss": 0.6468, + "step": 2593 + }, + { + "epoch": 0.9932988703810071, + "grad_norm": 0.31328508257865906, + "learning_rate": 3.82285485928304e-06, + "loss": 0.6411, + "step": 2594 + }, + { + "epoch": 0.9936817920735209, + "grad_norm": 0.306226909160614, + "learning_rate": 3.821992941638688e-06, + "loss": 0.6516, + "step": 2595 + }, + { + "epoch": 0.9940647137660349, + "grad_norm": 0.29100140929222107, + "learning_rate": 3.821130805805334e-06, + "loss": 0.6319, + "step": 2596 + }, + { + "epoch": 0.9944476354585488, + "grad_norm": 0.30655136704444885, + "learning_rate": 3.82026845192527e-06, + "loss": 0.6527, + "step": 2597 + }, + { + "epoch": 0.9948305571510626, + "grad_norm": 0.2994900345802307, + "learning_rate": 3.819405880140823e-06, + "loss": 0.6568, + "step": 2598 + }, + { + "epoch": 0.9952134788435765, + "grad_norm": 0.2904807925224304, + "learning_rate": 3.8185430905943564e-06, + "loss": 0.6687, + "step": 2599 + }, + { + "epoch": 0.9955964005360903, + "grad_norm": 0.2897905111312866, + "learning_rate": 3.8176800834282715e-06, + "loss": 0.6394, + "step": 2600 + }, + { + "epoch": 0.9959793222286043, + "grad_norm": 0.30152979493141174, + "learning_rate": 3.816816858785002e-06, + "loss": 0.6519, + "step": 2601 + }, + { + "epoch": 0.9963622439211182, + "grad_norm": 0.28953397274017334, + "learning_rate": 3.815953416807021e-06, + "loss": 0.6339, + "step": 2602 + }, + { + "epoch": 0.996745165613632, + "grad_norm": 0.2889082729816437, + "learning_rate": 3.815089757636834e-06, + "loss": 0.6486, + "step": 2603 + }, + { + "epoch": 0.9971280873061459, + "grad_norm": 0.29194390773773193, + "learning_rate": 3.814225881416984e-06, + "loss": 0.6397, + "step": 2604 + }, + { + "epoch": 0.9975110089986597, + "grad_norm": 0.28397029638290405, + "learning_rate": 3.813361788290052e-06, + "loss": 0.6517, + "step": 2605 + }, + { + "epoch": 0.9978939306911736, + "grad_norm": 0.2952829897403717, + "learning_rate": 3.812497478398651e-06, + "loss": 0.6401, + "step": 2606 + }, + { + "epoch": 0.9982768523836876, + "grad_norm": 0.2944144904613495, + "learning_rate": 3.8116329518854313e-06, + "loss": 0.6656, + "step": 2607 + }, + { + "epoch": 0.9986597740762014, + "grad_norm": 0.29345735907554626, + "learning_rate": 3.8107682088930797e-06, + "loss": 0.6343, + "step": 2608 + }, + { + "epoch": 0.9990426957687153, + "grad_norm": 0.2927607297897339, + "learning_rate": 3.809903249564318e-06, + "loss": 0.6371, + "step": 2609 + }, + { + "epoch": 0.9994256174612292, + "grad_norm": 0.288296639919281, + "learning_rate": 3.809038074041904e-06, + "loss": 0.6506, + "step": 2610 + }, + { + "epoch": 0.999808539153743, + "grad_norm": 0.29800692200660706, + "learning_rate": 3.8081726824686303e-06, + "loss": 0.6114, + "step": 2611 + } + ], + "logging_steps": 1, + "max_steps": 7833, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 2611, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.339605353386423e+19, + "train_batch_size": 12, + "trial_name": null, + "trial_params": null +}