diff --git "a/checkpoint-10000/trainer_state.json" "b/checkpoint-10000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-10000/trainer_state.json" @@ -0,0 +1,6034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 2.793004274368286, + "learning_rate": 1.8e-06, + "loss": 1.1874, + "step": 10 + }, + { + "grad_norm": 1.3802913427352905, + "learning_rate": 3.8e-06, + "loss": 1.1592, + "step": 20 + }, + { + "grad_norm": 0.8943536281585693, + "learning_rate": 5.8e-06, + "loss": 1.1116, + "step": 30 + }, + { + "grad_norm": 0.8201987147331238, + "learning_rate": 7.8e-06, + "loss": 1.0838, + "step": 40 + }, + { + "grad_norm": 0.795853853225708, + "learning_rate": 9.800000000000001e-06, + "loss": 1.0989, + "step": 50 + }, + { + "grad_norm": 0.49384215474128723, + "learning_rate": 1.18e-05, + "loss": 1.1029, + "step": 60 + }, + { + "grad_norm": 0.5673120021820068, + "learning_rate": 1.3800000000000002e-05, + "loss": 1.0939, + "step": 70 + }, + { + "grad_norm": 0.635345995426178, + "learning_rate": 1.58e-05, + "loss": 1.0837, + "step": 80 + }, + { + "grad_norm": 0.8145999312400818, + "learning_rate": 1.78e-05, + "loss": 1.0618, + "step": 90 + }, + { + "grad_norm": 0.4601838290691376, + "learning_rate": 1.9800000000000004e-05, + "loss": 1.0657, + "step": 100 + }, + { + "grad_norm": 0.5672536492347717, + "learning_rate": 2.18e-05, + "loss": 1.0458, + "step": 110 + }, + { + "grad_norm": 0.6255260109901428, + "learning_rate": 2.38e-05, + "loss": 1.0278, + "step": 120 + }, + { + "grad_norm": 1.23146390914917, + "learning_rate": 2.58e-05, + "loss": 1.012, + "step": 130 + }, + { + "grad_norm": 1.8485451936721802, + "learning_rate": 2.7800000000000005e-05, + "loss": 0.9944, + "step": 140 + }, + { + "grad_norm": 0.9352023005485535, + "learning_rate": 2.98e-05, + "loss": 0.9332, + "step": 150 + }, + { + "grad_norm": 0.943343997001648, + "learning_rate": 3.18e-05, + "loss": 0.8947, + "step": 160 + }, + { + "grad_norm": 4.302311420440674, + "learning_rate": 3.38e-05, + "loss": 0.8313, + "step": 170 + }, + { + "grad_norm": 2.1125075817108154, + "learning_rate": 3.58e-05, + "loss": 0.7607, + "step": 180 + }, + { + "grad_norm": 1.6946742534637451, + "learning_rate": 3.7800000000000004e-05, + "loss": 0.6263, + "step": 190 + }, + { + "grad_norm": 2.1044507026672363, + "learning_rate": 3.9800000000000005e-05, + "loss": 0.5075, + "step": 200 + }, + { + "grad_norm": 1.8176668882369995, + "learning_rate": 4.18e-05, + "loss": 0.3628, + "step": 210 + }, + { + "grad_norm": 1.7188639640808105, + "learning_rate": 4.38e-05, + "loss": 0.2611, + "step": 220 + }, + { + "grad_norm": 2.0323312282562256, + "learning_rate": 4.58e-05, + "loss": 0.2257, + "step": 230 + }, + { + "grad_norm": 1.5667383670806885, + "learning_rate": 4.78e-05, + "loss": 0.1943, + "step": 240 + }, + { + "grad_norm": 1.8114680051803589, + "learning_rate": 4.9800000000000004e-05, + "loss": 0.1697, + "step": 250 + }, + { + "grad_norm": 1.5046446323394775, + "learning_rate": 5.1800000000000005e-05, + "loss": 0.1458, + "step": 260 + }, + { + "grad_norm": 1.5899959802627563, + "learning_rate": 5.380000000000001e-05, + "loss": 0.1517, + "step": 270 + }, + { + "grad_norm": 1.264258623123169, + "learning_rate": 5.580000000000001e-05, + "loss": 0.1455, + "step": 280 + }, + { + "grad_norm": 1.3830788135528564, + "learning_rate": 5.7799999999999995e-05, + "loss": 0.1593, + "step": 290 + }, + { + "grad_norm": 1.6816647052764893, + "learning_rate": 5.9800000000000003e-05, + "loss": 0.1357, + "step": 300 + }, + { + "grad_norm": 1.107353925704956, + "learning_rate": 6.18e-05, + "loss": 0.1435, + "step": 310 + }, + { + "grad_norm": 1.0500719547271729, + "learning_rate": 6.38e-05, + "loss": 0.1246, + "step": 320 + }, + { + "grad_norm": 1.0716999769210815, + "learning_rate": 6.58e-05, + "loss": 0.1222, + "step": 330 + }, + { + "grad_norm": 1.2342218160629272, + "learning_rate": 6.780000000000001e-05, + "loss": 0.1206, + "step": 340 + }, + { + "grad_norm": 1.0673514604568481, + "learning_rate": 6.98e-05, + "loss": 0.1081, + "step": 350 + }, + { + "grad_norm": 1.1057475805282593, + "learning_rate": 7.18e-05, + "loss": 0.1142, + "step": 360 + }, + { + "grad_norm": 0.7929712533950806, + "learning_rate": 7.38e-05, + "loss": 0.1126, + "step": 370 + }, + { + "grad_norm": 1.0456504821777344, + "learning_rate": 7.58e-05, + "loss": 0.1046, + "step": 380 + }, + { + "grad_norm": 0.7815558910369873, + "learning_rate": 7.780000000000001e-05, + "loss": 0.1109, + "step": 390 + }, + { + "grad_norm": 0.9830052852630615, + "learning_rate": 7.98e-05, + "loss": 0.116, + "step": 400 + }, + { + "grad_norm": 0.9349966049194336, + "learning_rate": 8.18e-05, + "loss": 0.1172, + "step": 410 + }, + { + "grad_norm": 0.9116413593292236, + "learning_rate": 8.38e-05, + "loss": 0.1059, + "step": 420 + }, + { + "grad_norm": 0.7345888614654541, + "learning_rate": 8.58e-05, + "loss": 0.1003, + "step": 430 + }, + { + "grad_norm": 0.9472398161888123, + "learning_rate": 8.78e-05, + "loss": 0.1063, + "step": 440 + }, + { + "grad_norm": 0.9572857618331909, + "learning_rate": 8.98e-05, + "loss": 0.1073, + "step": 450 + }, + { + "grad_norm": 0.7120158076286316, + "learning_rate": 9.180000000000001e-05, + "loss": 0.109, + "step": 460 + }, + { + "grad_norm": 0.6262388229370117, + "learning_rate": 9.38e-05, + "loss": 0.1081, + "step": 470 + }, + { + "grad_norm": 0.8021785616874695, + "learning_rate": 9.58e-05, + "loss": 0.1056, + "step": 480 + }, + { + "grad_norm": 0.7647976875305176, + "learning_rate": 9.78e-05, + "loss": 0.104, + "step": 490 + }, + { + "grad_norm": 0.7514287829399109, + "learning_rate": 9.98e-05, + "loss": 0.0945, + "step": 500 + }, + { + "grad_norm": 0.82851243019104, + "learning_rate": 9.9999778549206e-05, + "loss": 0.0983, + "step": 510 + }, + { + "grad_norm": 0.660470724105835, + "learning_rate": 9.999901304280685e-05, + "loss": 0.0929, + "step": 520 + }, + { + "grad_norm": 0.6284200549125671, + "learning_rate": 9.999770075521164e-05, + "loss": 0.0968, + "step": 530 + }, + { + "grad_norm": 0.5111472606658936, + "learning_rate": 9.99958417007713e-05, + "loss": 0.0829, + "step": 540 + }, + { + "grad_norm": 0.6126238703727722, + "learning_rate": 9.999343589981615e-05, + "loss": 0.088, + "step": 550 + }, + { + "grad_norm": 0.5818606615066528, + "learning_rate": 9.999048337865568e-05, + "loss": 0.0861, + "step": 560 + }, + { + "grad_norm": 0.5772903561592102, + "learning_rate": 9.998698416957815e-05, + "loss": 0.0795, + "step": 570 + }, + { + "grad_norm": 0.7526791095733643, + "learning_rate": 9.998293831085037e-05, + "loss": 0.0852, + "step": 580 + }, + { + "grad_norm": 0.5732459425926208, + "learning_rate": 9.997834584671719e-05, + "loss": 0.085, + "step": 590 + }, + { + "grad_norm": 0.6462765336036682, + "learning_rate": 9.997320682740107e-05, + "loss": 0.0846, + "step": 600 + }, + { + "grad_norm": 0.6795440316200256, + "learning_rate": 9.996752130910149e-05, + "loss": 0.1051, + "step": 610 + }, + { + "grad_norm": 0.5026015043258667, + "learning_rate": 9.99612893539944e-05, + "loss": 0.083, + "step": 620 + }, + { + "grad_norm": 0.7280054688453674, + "learning_rate": 9.995451103023144e-05, + "loss": 0.0693, + "step": 630 + }, + { + "grad_norm": 0.5001598000526428, + "learning_rate": 9.994718641193928e-05, + "loss": 0.0783, + "step": 640 + }, + { + "grad_norm": 0.48168277740478516, + "learning_rate": 9.993931557921874e-05, + "loss": 0.0759, + "step": 650 + }, + { + "grad_norm": 0.6505987048149109, + "learning_rate": 9.993089861814402e-05, + "loss": 0.0834, + "step": 660 + }, + { + "grad_norm": 0.6558431386947632, + "learning_rate": 9.992193562076166e-05, + "loss": 0.0916, + "step": 670 + }, + { + "grad_norm": 0.566896915435791, + "learning_rate": 9.991242668508954e-05, + "loss": 0.0875, + "step": 680 + }, + { + "grad_norm": 0.6067936420440674, + "learning_rate": 9.990237191511587e-05, + "loss": 0.0745, + "step": 690 + }, + { + "grad_norm": 0.5476846694946289, + "learning_rate": 9.989177142079802e-05, + "loss": 0.0759, + "step": 700 + }, + { + "grad_norm": 0.7813402414321899, + "learning_rate": 9.988062531806126e-05, + "loss": 0.0894, + "step": 710 + }, + { + "grad_norm": 0.37198156118392944, + "learning_rate": 9.986893372879762e-05, + "loss": 0.0737, + "step": 720 + }, + { + "grad_norm": 0.7360274791717529, + "learning_rate": 9.985669678086443e-05, + "loss": 0.0718, + "step": 730 + }, + { + "grad_norm": 0.5045596361160278, + "learning_rate": 9.984391460808298e-05, + "loss": 0.0665, + "step": 740 + }, + { + "grad_norm": 0.37261122465133667, + "learning_rate": 9.983058735023709e-05, + "loss": 0.0757, + "step": 750 + }, + { + "grad_norm": 0.4317455291748047, + "learning_rate": 9.98167151530715e-05, + "loss": 0.0643, + "step": 760 + }, + { + "grad_norm": 0.5593804121017456, + "learning_rate": 9.980229816829034e-05, + "loss": 0.0718, + "step": 770 + }, + { + "grad_norm": 0.51378333568573, + "learning_rate": 9.978733655355544e-05, + "loss": 0.0759, + "step": 780 + }, + { + "grad_norm": 0.6022909283638, + "learning_rate": 9.977183047248464e-05, + "loss": 0.07, + "step": 790 + }, + { + "grad_norm": 0.6415618062019348, + "learning_rate": 9.975578009464992e-05, + "loss": 0.0729, + "step": 800 + }, + { + "grad_norm": 0.3651258051395416, + "learning_rate": 9.97391855955757e-05, + "loss": 0.0727, + "step": 810 + }, + { + "grad_norm": 0.4647364914417267, + "learning_rate": 9.972204715673669e-05, + "loss": 0.0687, + "step": 820 + }, + { + "grad_norm": 0.4876095950603485, + "learning_rate": 9.970436496555617e-05, + "loss": 0.073, + "step": 830 + }, + { + "grad_norm": 0.45116764307022095, + "learning_rate": 9.968613921540373e-05, + "loss": 0.0662, + "step": 840 + }, + { + "grad_norm": 0.41085413098335266, + "learning_rate": 9.966737010559326e-05, + "loss": 0.0815, + "step": 850 + }, + { + "grad_norm": 0.5202544331550598, + "learning_rate": 9.964805784138072e-05, + "loss": 0.0707, + "step": 860 + }, + { + "grad_norm": 0.45897090435028076, + "learning_rate": 9.962820263396195e-05, + "loss": 0.0756, + "step": 870 + }, + { + "grad_norm": 0.5861125588417053, + "learning_rate": 9.960780470047033e-05, + "loss": 0.0693, + "step": 880 + }, + { + "grad_norm": 0.4968416392803192, + "learning_rate": 9.958686426397437e-05, + "loss": 0.0668, + "step": 890 + }, + { + "grad_norm": 0.6051953434944153, + "learning_rate": 9.956538155347534e-05, + "loss": 0.0764, + "step": 900 + }, + { + "grad_norm": 0.6431021094322205, + "learning_rate": 9.95433568039047e-05, + "loss": 0.0645, + "step": 910 + }, + { + "grad_norm": 0.4096144139766693, + "learning_rate": 9.952079025612162e-05, + "loss": 0.0706, + "step": 920 + }, + { + "grad_norm": 0.479817271232605, + "learning_rate": 9.949768215691022e-05, + "loss": 0.0668, + "step": 930 + }, + { + "grad_norm": 0.29437652230262756, + "learning_rate": 9.9474032758977e-05, + "loss": 0.0662, + "step": 940 + }, + { + "grad_norm": 0.357378751039505, + "learning_rate": 9.944984232094794e-05, + "loss": 0.07, + "step": 950 + }, + { + "grad_norm": 0.40683600306510925, + "learning_rate": 9.942511110736584e-05, + "loss": 0.0577, + "step": 960 + }, + { + "grad_norm": 0.6319900751113892, + "learning_rate": 9.939983938868726e-05, + "loss": 0.0603, + "step": 970 + }, + { + "grad_norm": 0.500520646572113, + "learning_rate": 9.93740274412797e-05, + "loss": 0.0703, + "step": 980 + }, + { + "grad_norm": 0.4544166922569275, + "learning_rate": 9.934767554741846e-05, + "loss": 0.065, + "step": 990 + }, + { + "grad_norm": 0.6188819408416748, + "learning_rate": 9.932078399528361e-05, + "loss": 0.0649, + "step": 1000 + }, + { + "grad_norm": 0.379477322101593, + "learning_rate": 9.929335307895689e-05, + "loss": 0.0643, + "step": 1010 + }, + { + "grad_norm": 0.3923026919364929, + "learning_rate": 9.926538309841839e-05, + "loss": 0.0618, + "step": 1020 + }, + { + "grad_norm": 0.4505191743373871, + "learning_rate": 9.923687435954334e-05, + "loss": 0.0624, + "step": 1030 + }, + { + "grad_norm": 0.3937477767467499, + "learning_rate": 9.920782717409873e-05, + "loss": 0.0672, + "step": 1040 + }, + { + "grad_norm": 0.3451329469680786, + "learning_rate": 9.917824185973994e-05, + "loss": 0.0567, + "step": 1050 + }, + { + "grad_norm": 0.34786710143089294, + "learning_rate": 9.914811874000723e-05, + "loss": 0.0676, + "step": 1060 + }, + { + "grad_norm": 0.4594043791294098, + "learning_rate": 9.911745814432218e-05, + "loss": 0.0666, + "step": 1070 + }, + { + "grad_norm": 0.41680940985679626, + "learning_rate": 9.90862604079842e-05, + "loss": 0.0679, + "step": 1080 + }, + { + "grad_norm": 0.39599141478538513, + "learning_rate": 9.90545258721667e-05, + "loss": 0.0492, + "step": 1090 + }, + { + "grad_norm": 0.6308373212814331, + "learning_rate": 9.90222548839135e-05, + "loss": 0.0588, + "step": 1100 + }, + { + "grad_norm": 0.47390222549438477, + "learning_rate": 9.898944779613495e-05, + "loss": 0.0616, + "step": 1110 + }, + { + "grad_norm": 0.3503871262073517, + "learning_rate": 9.89561049676041e-05, + "loss": 0.0621, + "step": 1120 + }, + { + "grad_norm": 0.45727768540382385, + "learning_rate": 9.89222267629528e-05, + "loss": 0.066, + "step": 1130 + }, + { + "grad_norm": 0.37270107865333557, + "learning_rate": 9.888781355266763e-05, + "loss": 0.0696, + "step": 1140 + }, + { + "grad_norm": 0.4304162561893463, + "learning_rate": 9.885286571308598e-05, + "loss": 0.0631, + "step": 1150 + }, + { + "grad_norm": 0.4124678373336792, + "learning_rate": 9.881738362639182e-05, + "loss": 0.0628, + "step": 1160 + }, + { + "grad_norm": 0.31958577036857605, + "learning_rate": 9.878136768061154e-05, + "loss": 0.0586, + "step": 1170 + }, + { + "grad_norm": 0.44271573424339294, + "learning_rate": 9.874481826960979e-05, + "loss": 0.0587, + "step": 1180 + }, + { + "grad_norm": 0.4161085784435272, + "learning_rate": 9.870773579308503e-05, + "loss": 0.0665, + "step": 1190 + }, + { + "grad_norm": 0.470930278301239, + "learning_rate": 9.867012065656533e-05, + "loss": 0.0666, + "step": 1200 + }, + { + "grad_norm": 0.3212951719760895, + "learning_rate": 9.863197327140376e-05, + "loss": 0.0587, + "step": 1210 + }, + { + "grad_norm": 0.46022161841392517, + "learning_rate": 9.859329405477403e-05, + "loss": 0.0522, + "step": 1220 + }, + { + "grad_norm": 0.3353862166404724, + "learning_rate": 9.855408342966585e-05, + "loss": 0.0585, + "step": 1230 + }, + { + "grad_norm": 0.4352401793003082, + "learning_rate": 9.851434182488033e-05, + "loss": 0.0599, + "step": 1240 + }, + { + "grad_norm": 0.44494494795799255, + "learning_rate": 9.84740696750253e-05, + "loss": 0.0627, + "step": 1250 + }, + { + "grad_norm": 0.47765183448791504, + "learning_rate": 9.843326742051055e-05, + "loss": 0.0588, + "step": 1260 + }, + { + "grad_norm": 0.2517593801021576, + "learning_rate": 9.839193550754297e-05, + "loss": 0.0623, + "step": 1270 + }, + { + "grad_norm": 0.4564324915409088, + "learning_rate": 9.835007438812177e-05, + "loss": 0.0643, + "step": 1280 + }, + { + "grad_norm": 0.39434176683425903, + "learning_rate": 9.830768452003341e-05, + "loss": 0.0678, + "step": 1290 + }, + { + "grad_norm": 0.4088485538959503, + "learning_rate": 9.826476636684671e-05, + "loss": 0.0627, + "step": 1300 + }, + { + "grad_norm": 0.3764931559562683, + "learning_rate": 9.822132039790773e-05, + "loss": 0.0586, + "step": 1310 + }, + { + "grad_norm": 0.37533169984817505, + "learning_rate": 9.817734708833461e-05, + "loss": 0.0627, + "step": 1320 + }, + { + "grad_norm": 0.27827706933021545, + "learning_rate": 9.813284691901243e-05, + "loss": 0.0527, + "step": 1330 + }, + { + "grad_norm": 0.6446675658226013, + "learning_rate": 9.808782037658792e-05, + "loss": 0.057, + "step": 1340 + }, + { + "grad_norm": 0.3713286221027374, + "learning_rate": 9.804226795346411e-05, + "loss": 0.0525, + "step": 1350 + }, + { + "grad_norm": 0.5087715983390808, + "learning_rate": 9.799619014779503e-05, + "loss": 0.057, + "step": 1360 + }, + { + "grad_norm": 0.32558947801589966, + "learning_rate": 9.794958746348013e-05, + "loss": 0.0482, + "step": 1370 + }, + { + "grad_norm": 0.4127570688724518, + "learning_rate": 9.790246041015896e-05, + "loss": 0.058, + "step": 1380 + }, + { + "grad_norm": 0.42279231548309326, + "learning_rate": 9.785480950320538e-05, + "loss": 0.05, + "step": 1390 + }, + { + "grad_norm": 0.4614427387714386, + "learning_rate": 9.78066352637221e-05, + "loss": 0.0601, + "step": 1400 + }, + { + "grad_norm": 0.42145198583602905, + "learning_rate": 9.775793821853488e-05, + "loss": 0.052, + "step": 1410 + }, + { + "grad_norm": 0.3475460708141327, + "learning_rate": 9.77087189001868e-05, + "loss": 0.0599, + "step": 1420 + }, + { + "grad_norm": 0.39030691981315613, + "learning_rate": 9.765897784693243e-05, + "loss": 0.054, + "step": 1430 + }, + { + "grad_norm": 0.28959161043167114, + "learning_rate": 9.760871560273197e-05, + "loss": 0.0562, + "step": 1440 + }, + { + "grad_norm": 0.32825151085853577, + "learning_rate": 9.755793271724526e-05, + "loss": 0.0559, + "step": 1450 + }, + { + "grad_norm": 0.38214606046676636, + "learning_rate": 9.750662974582584e-05, + "loss": 0.0504, + "step": 1460 + }, + { + "grad_norm": 0.550482451915741, + "learning_rate": 9.745480724951473e-05, + "loss": 0.0508, + "step": 1470 + }, + { + "grad_norm": 0.34260162711143494, + "learning_rate": 9.740246579503447e-05, + "loss": 0.0544, + "step": 1480 + }, + { + "grad_norm": 0.33075544238090515, + "learning_rate": 9.734960595478284e-05, + "loss": 0.0497, + "step": 1490 + }, + { + "grad_norm": 0.34045588970184326, + "learning_rate": 9.729622830682657e-05, + "loss": 0.0616, + "step": 1500 + }, + { + "grad_norm": 0.29704034328460693, + "learning_rate": 9.724233343489504e-05, + "loss": 0.0513, + "step": 1510 + }, + { + "grad_norm": 0.3109928369522095, + "learning_rate": 9.718792192837396e-05, + "loss": 0.0559, + "step": 1520 + }, + { + "grad_norm": 0.4239391088485718, + "learning_rate": 9.713299438229886e-05, + "loss": 0.0629, + "step": 1530 + }, + { + "grad_norm": 0.4465237259864807, + "learning_rate": 9.707755139734855e-05, + "loss": 0.0612, + "step": 1540 + }, + { + "grad_norm": 0.3261890411376953, + "learning_rate": 9.702159357983866e-05, + "loss": 0.0547, + "step": 1550 + }, + { + "grad_norm": 0.33906134963035583, + "learning_rate": 9.696512154171492e-05, + "loss": 0.0543, + "step": 1560 + }, + { + "grad_norm": 0.4858160614967346, + "learning_rate": 9.690813590054645e-05, + "loss": 0.0536, + "step": 1570 + }, + { + "grad_norm": 0.251074880361557, + "learning_rate": 9.685063727951914e-05, + "loss": 0.0562, + "step": 1580 + }, + { + "grad_norm": 0.29576700925827026, + "learning_rate": 9.679262630742865e-05, + "loss": 0.0552, + "step": 1590 + }, + { + "grad_norm": 0.45864546298980713, + "learning_rate": 9.673410361867373e-05, + "loss": 0.0491, + "step": 1600 + }, + { + "grad_norm": 0.3349587321281433, + "learning_rate": 9.667506985324909e-05, + "loss": 0.0482, + "step": 1610 + }, + { + "grad_norm": 0.3018699586391449, + "learning_rate": 9.661552565673855e-05, + "loss": 0.0442, + "step": 1620 + }, + { + "grad_norm": 0.33417651057243347, + "learning_rate": 9.655547168030789e-05, + "loss": 0.0493, + "step": 1630 + }, + { + "grad_norm": 0.39800384640693665, + "learning_rate": 9.649490858069777e-05, + "loss": 0.0548, + "step": 1640 + }, + { + "grad_norm": 0.3595905005931854, + "learning_rate": 9.643383702021658e-05, + "loss": 0.0582, + "step": 1650 + }, + { + "grad_norm": 0.3571525812149048, + "learning_rate": 9.637225766673307e-05, + "loss": 0.0446, + "step": 1660 + }, + { + "grad_norm": 0.3590371012687683, + "learning_rate": 9.631017119366922e-05, + "loss": 0.0509, + "step": 1670 + }, + { + "grad_norm": 0.3707994818687439, + "learning_rate": 9.624757827999273e-05, + "loss": 0.0511, + "step": 1680 + }, + { + "grad_norm": 0.3809451162815094, + "learning_rate": 9.618447961020971e-05, + "loss": 0.0562, + "step": 1690 + }, + { + "grad_norm": 0.5052781701087952, + "learning_rate": 9.612087587435707e-05, + "loss": 0.0534, + "step": 1700 + }, + { + "grad_norm": 0.2926018536090851, + "learning_rate": 9.605676776799508e-05, + "loss": 0.0501, + "step": 1710 + }, + { + "grad_norm": 0.4085676670074463, + "learning_rate": 9.599215599219973e-05, + "loss": 0.0522, + "step": 1720 + }, + { + "grad_norm": 0.5422471761703491, + "learning_rate": 9.592704125355505e-05, + "loss": 0.0624, + "step": 1730 + }, + { + "grad_norm": 0.4074013829231262, + "learning_rate": 9.586142426414538e-05, + "loss": 0.0531, + "step": 1740 + }, + { + "grad_norm": 0.2795093059539795, + "learning_rate": 9.57953057415476e-05, + "loss": 0.0544, + "step": 1750 + }, + { + "grad_norm": 0.4418717324733734, + "learning_rate": 9.572868640882328e-05, + "loss": 0.0603, + "step": 1760 + }, + { + "grad_norm": 0.27232369780540466, + "learning_rate": 9.56615669945108e-05, + "loss": 0.0615, + "step": 1770 + }, + { + "grad_norm": 0.22374026477336884, + "learning_rate": 9.55939482326173e-05, + "loss": 0.0563, + "step": 1780 + }, + { + "grad_norm": 0.43078160285949707, + "learning_rate": 9.552583086261069e-05, + "loss": 0.0627, + "step": 1790 + }, + { + "grad_norm": 0.252181738615036, + "learning_rate": 9.545721562941168e-05, + "loss": 0.056, + "step": 1800 + }, + { + "grad_norm": 0.3137652277946472, + "learning_rate": 9.538810328338543e-05, + "loss": 0.0528, + "step": 1810 + }, + { + "grad_norm": 0.35691484808921814, + "learning_rate": 9.531849458033349e-05, + "loss": 0.0501, + "step": 1820 + }, + { + "grad_norm": 0.28482335805892944, + "learning_rate": 9.524839028148547e-05, + "loss": 0.0504, + "step": 1830 + }, + { + "grad_norm": 0.37036216259002686, + "learning_rate": 9.517779115349077e-05, + "loss": 0.0592, + "step": 1840 + }, + { + "grad_norm": 0.2693071663379669, + "learning_rate": 9.510669796841014e-05, + "loss": 0.0511, + "step": 1850 + }, + { + "grad_norm": 0.3534975051879883, + "learning_rate": 9.503511150370727e-05, + "loss": 0.0547, + "step": 1860 + }, + { + "grad_norm": 0.3484165072441101, + "learning_rate": 9.496303254224024e-05, + "loss": 0.0552, + "step": 1870 + }, + { + "grad_norm": 0.2709268033504486, + "learning_rate": 9.489046187225306e-05, + "loss": 0.049, + "step": 1880 + }, + { + "grad_norm": 0.3716231882572174, + "learning_rate": 9.481740028736692e-05, + "loss": 0.0429, + "step": 1890 + }, + { + "grad_norm": 0.2585739195346832, + "learning_rate": 9.474384858657164e-05, + "loss": 0.0621, + "step": 1900 + }, + { + "grad_norm": 0.37147170305252075, + "learning_rate": 9.466980757421679e-05, + "loss": 0.0558, + "step": 1910 + }, + { + "grad_norm": 0.2653484344482422, + "learning_rate": 9.459527806000305e-05, + "loss": 0.0531, + "step": 1920 + }, + { + "grad_norm": 0.38023388385772705, + "learning_rate": 9.452026085897325e-05, + "loss": 0.059, + "step": 1930 + }, + { + "grad_norm": 0.2839908003807068, + "learning_rate": 9.444475679150348e-05, + "loss": 0.0542, + "step": 1940 + }, + { + "grad_norm": 0.2802390456199646, + "learning_rate": 9.436876668329411e-05, + "loss": 0.0533, + "step": 1950 + }, + { + "grad_norm": 0.3265979290008545, + "learning_rate": 9.429229136536079e-05, + "loss": 0.0472, + "step": 1960 + }, + { + "grad_norm": 0.3409467339515686, + "learning_rate": 9.421533167402534e-05, + "loss": 0.0553, + "step": 1970 + }, + { + "grad_norm": 0.4405045807361603, + "learning_rate": 9.413788845090666e-05, + "loss": 0.0605, + "step": 1980 + }, + { + "grad_norm": 0.2831200957298279, + "learning_rate": 9.405996254291136e-05, + "loss": 0.0479, + "step": 1990 + }, + { + "grad_norm": 0.3534575402736664, + "learning_rate": 9.398155480222474e-05, + "loss": 0.0473, + "step": 2000 + }, + { + "grad_norm": 0.37762966752052307, + "learning_rate": 9.390266608630128e-05, + "loss": 0.0525, + "step": 2010 + }, + { + "grad_norm": 0.37968921661376953, + "learning_rate": 9.38232972578553e-05, + "loss": 0.0549, + "step": 2020 + }, + { + "grad_norm": 0.3052346706390381, + "learning_rate": 9.374344918485164e-05, + "loss": 0.0584, + "step": 2030 + }, + { + "grad_norm": 0.2952876687049866, + "learning_rate": 9.366312274049602e-05, + "loss": 0.0557, + "step": 2040 + }, + { + "grad_norm": 0.25244060158729553, + "learning_rate": 9.358231880322554e-05, + "loss": 0.052, + "step": 2050 + }, + { + "grad_norm": 0.2958115041255951, + "learning_rate": 9.350103825669916e-05, + "loss": 0.0475, + "step": 2060 + }, + { + "grad_norm": 0.3789975643157959, + "learning_rate": 9.341928198978787e-05, + "loss": 0.0492, + "step": 2070 + }, + { + "grad_norm": 0.3502075672149658, + "learning_rate": 9.333705089656512e-05, + "loss": 0.045, + "step": 2080 + }, + { + "grad_norm": 0.3400934338569641, + "learning_rate": 9.325434587629698e-05, + "loss": 0.047, + "step": 2090 + }, + { + "grad_norm": 0.34106630086898804, + "learning_rate": 9.31711678334323e-05, + "loss": 0.0437, + "step": 2100 + }, + { + "grad_norm": 0.283089280128479, + "learning_rate": 9.308751767759282e-05, + "loss": 0.0427, + "step": 2110 + }, + { + "grad_norm": 0.31602615118026733, + "learning_rate": 9.300339632356325e-05, + "loss": 0.0457, + "step": 2120 + }, + { + "grad_norm": 0.37002506852149963, + "learning_rate": 9.291880469128124e-05, + "loss": 0.0486, + "step": 2130 + }, + { + "grad_norm": 0.3134841322898865, + "learning_rate": 9.283374370582732e-05, + "loss": 0.0514, + "step": 2140 + }, + { + "grad_norm": 0.24214021861553192, + "learning_rate": 9.274821429741482e-05, + "loss": 0.0532, + "step": 2150 + }, + { + "grad_norm": 0.3677908480167389, + "learning_rate": 9.266221740137961e-05, + "loss": 0.0483, + "step": 2160 + }, + { + "grad_norm": 0.2927781343460083, + "learning_rate": 9.257575395817001e-05, + "loss": 0.046, + "step": 2170 + }, + { + "grad_norm": 0.3491074740886688, + "learning_rate": 9.248882491333637e-05, + "loss": 0.048, + "step": 2180 + }, + { + "grad_norm": 0.2807201147079468, + "learning_rate": 9.240143121752076e-05, + "loss": 0.0413, + "step": 2190 + }, + { + "grad_norm": 0.34429672360420227, + "learning_rate": 9.23135738264467e-05, + "loss": 0.0439, + "step": 2200 + }, + { + "grad_norm": 0.23641304671764374, + "learning_rate": 9.222525370090849e-05, + "loss": 0.0445, + "step": 2210 + }, + { + "grad_norm": 0.6537847518920898, + "learning_rate": 9.213647180676088e-05, + "loss": 0.0488, + "step": 2220 + }, + { + "grad_norm": 0.39291369915008545, + "learning_rate": 9.204722911490846e-05, + "loss": 0.0487, + "step": 2230 + }, + { + "grad_norm": 0.246474489569664, + "learning_rate": 9.1957526601295e-05, + "loss": 0.0422, + "step": 2240 + }, + { + "grad_norm": 0.23755821585655212, + "learning_rate": 9.186736524689281e-05, + "loss": 0.0511, + "step": 2250 + }, + { + "grad_norm": 0.4442860186100006, + "learning_rate": 9.177674603769204e-05, + "loss": 0.0494, + "step": 2260 + }, + { + "grad_norm": 0.3126569390296936, + "learning_rate": 9.168566996468983e-05, + "loss": 0.0465, + "step": 2270 + }, + { + "grad_norm": 0.2502683997154236, + "learning_rate": 9.159413802387951e-05, + "loss": 0.0508, + "step": 2280 + }, + { + "grad_norm": 0.27260729670524597, + "learning_rate": 9.150215121623974e-05, + "loss": 0.0466, + "step": 2290 + }, + { + "grad_norm": 0.26124727725982666, + "learning_rate": 9.140971054772349e-05, + "loss": 0.0397, + "step": 2300 + }, + { + "grad_norm": 0.42889511585235596, + "learning_rate": 9.131681702924713e-05, + "loss": 0.0529, + "step": 2310 + }, + { + "grad_norm": 0.3609920144081116, + "learning_rate": 9.122347167667926e-05, + "loss": 0.0455, + "step": 2320 + }, + { + "grad_norm": 0.37415170669555664, + "learning_rate": 9.112967551082973e-05, + "loss": 0.058, + "step": 2330 + }, + { + "grad_norm": 0.2700253129005432, + "learning_rate": 9.103542955743835e-05, + "loss": 0.0487, + "step": 2340 + }, + { + "grad_norm": 0.32367241382598877, + "learning_rate": 9.094073484716381e-05, + "loss": 0.0572, + "step": 2350 + }, + { + "grad_norm": 0.37198108434677124, + "learning_rate": 9.084559241557226e-05, + "loss": 0.0482, + "step": 2360 + }, + { + "grad_norm": 0.26771798729896545, + "learning_rate": 9.075000330312608e-05, + "loss": 0.0487, + "step": 2370 + }, + { + "grad_norm": 0.315613716840744, + "learning_rate": 9.065396855517253e-05, + "loss": 0.0443, + "step": 2380 + }, + { + "grad_norm": 0.1932615041732788, + "learning_rate": 9.055748922193219e-05, + "loss": 0.0486, + "step": 2390 + }, + { + "grad_norm": 0.38552233576774597, + "learning_rate": 9.046056635848761e-05, + "loss": 0.0501, + "step": 2400 + }, + { + "grad_norm": 0.30132681131362915, + "learning_rate": 9.036320102477169e-05, + "loss": 0.045, + "step": 2410 + }, + { + "grad_norm": 0.3855850398540497, + "learning_rate": 9.02653942855561e-05, + "loss": 0.0481, + "step": 2420 + }, + { + "grad_norm": 0.3420015573501587, + "learning_rate": 9.016714721043971e-05, + "loss": 0.046, + "step": 2430 + }, + { + "grad_norm": 0.3513767719268799, + "learning_rate": 9.006846087383675e-05, + "loss": 0.0437, + "step": 2440 + }, + { + "grad_norm": 0.37945640087127686, + "learning_rate": 8.996933635496523e-05, + "loss": 0.045, + "step": 2450 + }, + { + "grad_norm": 0.4292179346084595, + "learning_rate": 8.986977473783498e-05, + "loss": 0.0537, + "step": 2460 + }, + { + "grad_norm": 0.3670051097869873, + "learning_rate": 8.97697771112359e-05, + "loss": 0.0505, + "step": 2470 + }, + { + "grad_norm": 0.3247128427028656, + "learning_rate": 8.966934456872602e-05, + "loss": 0.0472, + "step": 2480 + }, + { + "grad_norm": 0.2907409369945526, + "learning_rate": 8.95684782086195e-05, + "loss": 0.0414, + "step": 2490 + }, + { + "grad_norm": 0.3595443367958069, + "learning_rate": 8.946717913397476e-05, + "loss": 0.0493, + "step": 2500 + }, + { + "grad_norm": 0.36934828758239746, + "learning_rate": 8.93654484525822e-05, + "loss": 0.0468, + "step": 2510 + }, + { + "grad_norm": 0.2803897261619568, + "learning_rate": 8.926328727695226e-05, + "loss": 0.0433, + "step": 2520 + }, + { + "grad_norm": 0.3432396650314331, + "learning_rate": 8.916069672430319e-05, + "loss": 0.0438, + "step": 2530 + }, + { + "grad_norm": 0.2776085436344147, + "learning_rate": 8.905767791654884e-05, + "loss": 0.0531, + "step": 2540 + }, + { + "grad_norm": 0.3068399131298065, + "learning_rate": 8.895423198028638e-05, + "loss": 0.0517, + "step": 2550 + }, + { + "grad_norm": 0.2772804796695709, + "learning_rate": 8.885036004678402e-05, + "loss": 0.0485, + "step": 2560 + }, + { + "grad_norm": 0.38457199931144714, + "learning_rate": 8.874606325196857e-05, + "loss": 0.0459, + "step": 2570 + }, + { + "grad_norm": 0.33462652564048767, + "learning_rate": 8.864134273641304e-05, + "loss": 0.052, + "step": 2580 + }, + { + "grad_norm": 0.2695263922214508, + "learning_rate": 8.853619964532427e-05, + "loss": 0.0497, + "step": 2590 + }, + { + "grad_norm": 0.3235803544521332, + "learning_rate": 8.843063512853019e-05, + "loss": 0.0428, + "step": 2600 + }, + { + "grad_norm": 0.3698258697986603, + "learning_rate": 8.832465034046749e-05, + "loss": 0.0424, + "step": 2610 + }, + { + "grad_norm": 0.29338857531547546, + "learning_rate": 8.821824644016882e-05, + "loss": 0.0462, + "step": 2620 + }, + { + "grad_norm": 0.25523799657821655, + "learning_rate": 8.811142459125019e-05, + "loss": 0.0514, + "step": 2630 + }, + { + "grad_norm": 0.45708855986595154, + "learning_rate": 8.800418596189822e-05, + "loss": 0.0452, + "step": 2640 + }, + { + "grad_norm": 0.2722897529602051, + "learning_rate": 8.789653172485737e-05, + "loss": 0.0465, + "step": 2650 + }, + { + "grad_norm": 0.32770639657974243, + "learning_rate": 8.778846305741715e-05, + "loss": 0.0522, + "step": 2660 + }, + { + "grad_norm": 0.2978685200214386, + "learning_rate": 8.767998114139918e-05, + "loss": 0.0463, + "step": 2670 + }, + { + "grad_norm": 0.4170987010002136, + "learning_rate": 8.757108716314429e-05, + "loss": 0.0509, + "step": 2680 + }, + { + "grad_norm": 0.29006829857826233, + "learning_rate": 8.746178231349962e-05, + "loss": 0.0493, + "step": 2690 + }, + { + "grad_norm": 0.27099934220314026, + "learning_rate": 8.735206778780549e-05, + "loss": 0.0494, + "step": 2700 + }, + { + "grad_norm": 0.38449302315711975, + "learning_rate": 8.724194478588234e-05, + "loss": 0.0444, + "step": 2710 + }, + { + "grad_norm": 0.2451976239681244, + "learning_rate": 8.713141451201772e-05, + "loss": 0.0455, + "step": 2720 + }, + { + "grad_norm": 0.27138909697532654, + "learning_rate": 8.702047817495295e-05, + "loss": 0.0457, + "step": 2730 + }, + { + "grad_norm": 0.2519432008266449, + "learning_rate": 8.69091369878701e-05, + "loss": 0.0454, + "step": 2740 + }, + { + "grad_norm": 0.2632405161857605, + "learning_rate": 8.679739216837849e-05, + "loss": 0.0405, + "step": 2750 + }, + { + "grad_norm": 0.3164638876914978, + "learning_rate": 8.66852449385016e-05, + "loss": 0.0424, + "step": 2760 + }, + { + "grad_norm": 0.3306692838668823, + "learning_rate": 8.657269652466356e-05, + "loss": 0.0421, + "step": 2770 + }, + { + "grad_norm": 0.26113489270210266, + "learning_rate": 8.645974815767577e-05, + "loss": 0.0404, + "step": 2780 + }, + { + "grad_norm": 0.2969265282154083, + "learning_rate": 8.634640107272351e-05, + "loss": 0.0438, + "step": 2790 + }, + { + "grad_norm": 0.3031178116798401, + "learning_rate": 8.623265650935234e-05, + "loss": 0.0445, + "step": 2800 + }, + { + "grad_norm": 0.3494732677936554, + "learning_rate": 8.611851571145456e-05, + "loss": 0.0435, + "step": 2810 + }, + { + "grad_norm": 0.297902375459671, + "learning_rate": 8.600397992725566e-05, + "loss": 0.0395, + "step": 2820 + }, + { + "grad_norm": 0.33495083451271057, + "learning_rate": 8.588905040930061e-05, + "loss": 0.0424, + "step": 2830 + }, + { + "grad_norm": 0.1906622052192688, + "learning_rate": 8.577372841444022e-05, + "loss": 0.0453, + "step": 2840 + }, + { + "grad_norm": 0.3432774245738983, + "learning_rate": 8.565801520381736e-05, + "loss": 0.0418, + "step": 2850 + }, + { + "grad_norm": 0.26721107959747314, + "learning_rate": 8.554191204285313e-05, + "loss": 0.0409, + "step": 2860 + }, + { + "grad_norm": 0.303368479013443, + "learning_rate": 8.542542020123315e-05, + "loss": 0.0421, + "step": 2870 + }, + { + "grad_norm": 0.2342439889907837, + "learning_rate": 8.530854095289347e-05, + "loss": 0.0401, + "step": 2880 + }, + { + "grad_norm": 0.25688156485557556, + "learning_rate": 8.519127557600688e-05, + "loss": 0.04, + "step": 2890 + }, + { + "grad_norm": 0.34693002700805664, + "learning_rate": 8.507362535296871e-05, + "loss": 0.0466, + "step": 2900 + }, + { + "grad_norm": 0.26837608218193054, + "learning_rate": 8.495559157038299e-05, + "loss": 0.0537, + "step": 2910 + }, + { + "grad_norm": 0.2099880874156952, + "learning_rate": 8.483717551904823e-05, + "loss": 0.0418, + "step": 2920 + }, + { + "grad_norm": 0.31592756509780884, + "learning_rate": 8.47183784939434e-05, + "loss": 0.0419, + "step": 2930 + }, + { + "grad_norm": 0.2944737672805786, + "learning_rate": 8.459920179421374e-05, + "loss": 0.0376, + "step": 2940 + }, + { + "grad_norm": 0.2332516759634018, + "learning_rate": 8.447964672315656e-05, + "loss": 0.036, + "step": 2950 + }, + { + "grad_norm": 0.3053232729434967, + "learning_rate": 8.435971458820692e-05, + "loss": 0.0507, + "step": 2960 + }, + { + "grad_norm": 0.2614811658859253, + "learning_rate": 8.423940670092345e-05, + "loss": 0.0482, + "step": 2970 + }, + { + "grad_norm": 0.2904261648654938, + "learning_rate": 8.411872437697394e-05, + "loss": 0.0423, + "step": 2980 + }, + { + "grad_norm": 0.3207901120185852, + "learning_rate": 8.399766893612096e-05, + "loss": 0.0385, + "step": 2990 + }, + { + "grad_norm": 0.2984755337238312, + "learning_rate": 8.38762417022074e-05, + "loss": 0.0436, + "step": 3000 + }, + { + "grad_norm": 0.34906458854675293, + "learning_rate": 8.375444400314204e-05, + "loss": 0.0395, + "step": 3010 + }, + { + "grad_norm": 0.25167936086654663, + "learning_rate": 8.3632277170885e-05, + "loss": 0.0491, + "step": 3020 + }, + { + "grad_norm": 0.36431029438972473, + "learning_rate": 8.350974254143318e-05, + "loss": 0.0403, + "step": 3030 + }, + { + "grad_norm": 0.230340376496315, + "learning_rate": 8.338684145480566e-05, + "loss": 0.0401, + "step": 3040 + }, + { + "grad_norm": 0.29917851090431213, + "learning_rate": 8.326357525502904e-05, + "loss": 0.0464, + "step": 3050 + }, + { + "grad_norm": 0.24232117831707, + "learning_rate": 8.313994529012273e-05, + "loss": 0.0451, + "step": 3060 + }, + { + "grad_norm": 0.3512433171272278, + "learning_rate": 8.301595291208422e-05, + "loss": 0.0472, + "step": 3070 + }, + { + "grad_norm": 0.19280123710632324, + "learning_rate": 8.289159947687427e-05, + "loss": 0.0434, + "step": 3080 + }, + { + "grad_norm": 0.3865027129650116, + "learning_rate": 8.276688634440216e-05, + "loss": 0.0469, + "step": 3090 + }, + { + "grad_norm": 0.25543034076690674, + "learning_rate": 8.26418148785107e-05, + "loss": 0.047, + "step": 3100 + }, + { + "grad_norm": 0.2599276900291443, + "learning_rate": 8.251638644696141e-05, + "loss": 0.0495, + "step": 3110 + }, + { + "grad_norm": 0.31160637736320496, + "learning_rate": 8.23906024214195e-05, + "loss": 0.0442, + "step": 3120 + }, + { + "grad_norm": 0.35604894161224365, + "learning_rate": 8.226446417743897e-05, + "loss": 0.0476, + "step": 3130 + }, + { + "grad_norm": 0.15958856046199799, + "learning_rate": 8.213797309444742e-05, + "loss": 0.0454, + "step": 3140 + }, + { + "grad_norm": 0.3273874819278717, + "learning_rate": 8.201113055573105e-05, + "loss": 0.0434, + "step": 3150 + }, + { + "grad_norm": 0.4467885196208954, + "learning_rate": 8.188393794841958e-05, + "loss": 0.0413, + "step": 3160 + }, + { + "grad_norm": 0.3001309633255005, + "learning_rate": 8.175639666347094e-05, + "loss": 0.0448, + "step": 3170 + }, + { + "grad_norm": 0.21124257147312164, + "learning_rate": 8.162850809565623e-05, + "loss": 0.04, + "step": 3180 + }, + { + "grad_norm": 0.378926157951355, + "learning_rate": 8.150027364354431e-05, + "loss": 0.0495, + "step": 3190 + }, + { + "grad_norm": 0.22007741034030914, + "learning_rate": 8.137169470948662e-05, + "loss": 0.0481, + "step": 3200 + }, + { + "grad_norm": 0.23186221718788147, + "learning_rate": 8.124277269960179e-05, + "loss": 0.0435, + "step": 3210 + }, + { + "grad_norm": 0.30731669068336487, + "learning_rate": 8.111350902376023e-05, + "loss": 0.0422, + "step": 3220 + }, + { + "grad_norm": 0.23267821967601776, + "learning_rate": 8.098390509556883e-05, + "loss": 0.0384, + "step": 3230 + }, + { + "grad_norm": 0.40911155939102173, + "learning_rate": 8.085396233235536e-05, + "loss": 0.0413, + "step": 3240 + }, + { + "grad_norm": 0.3080264627933502, + "learning_rate": 8.072368215515306e-05, + "loss": 0.0382, + "step": 3250 + }, + { + "grad_norm": 0.3228119909763336, + "learning_rate": 8.059306598868506e-05, + "loss": 0.0477, + "step": 3260 + }, + { + "grad_norm": 0.3576311469078064, + "learning_rate": 8.046211526134888e-05, + "loss": 0.0417, + "step": 3270 + }, + { + "grad_norm": 0.3398083448410034, + "learning_rate": 8.033083140520065e-05, + "loss": 0.0407, + "step": 3280 + }, + { + "grad_norm": 0.30744755268096924, + "learning_rate": 8.019921585593962e-05, + "loss": 0.0459, + "step": 3290 + }, + { + "grad_norm": 0.2791251838207245, + "learning_rate": 8.006727005289232e-05, + "loss": 0.0388, + "step": 3300 + }, + { + "grad_norm": 0.26894330978393555, + "learning_rate": 7.993499543899692e-05, + "loss": 0.0399, + "step": 3310 + }, + { + "grad_norm": 0.3140675723552704, + "learning_rate": 7.980239346078742e-05, + "loss": 0.0391, + "step": 3320 + }, + { + "grad_norm": 0.2300155609846115, + "learning_rate": 7.966946556837778e-05, + "loss": 0.041, + "step": 3330 + }, + { + "grad_norm": 0.31205397844314575, + "learning_rate": 7.953621321544616e-05, + "loss": 0.0405, + "step": 3340 + }, + { + "grad_norm": 0.34008124470710754, + "learning_rate": 7.940263785921896e-05, + "loss": 0.0445, + "step": 3350 + }, + { + "grad_norm": 0.3060169219970703, + "learning_rate": 7.926874096045482e-05, + "loss": 0.0406, + "step": 3360 + }, + { + "grad_norm": 0.31090641021728516, + "learning_rate": 7.913452398342881e-05, + "loss": 0.0488, + "step": 3370 + }, + { + "grad_norm": 0.3354012966156006, + "learning_rate": 7.89999883959163e-05, + "loss": 0.0431, + "step": 3380 + }, + { + "grad_norm": 0.2870286703109741, + "learning_rate": 7.886513566917687e-05, + "loss": 0.0472, + "step": 3390 + }, + { + "grad_norm": 0.18670085072517395, + "learning_rate": 7.872996727793838e-05, + "loss": 0.0412, + "step": 3400 + }, + { + "grad_norm": 0.28902074694633484, + "learning_rate": 7.859448470038069e-05, + "loss": 0.0428, + "step": 3410 + }, + { + "grad_norm": 0.3219638168811798, + "learning_rate": 7.845868941811956e-05, + "loss": 0.0429, + "step": 3420 + }, + { + "grad_norm": 0.3541117012500763, + "learning_rate": 7.832258291619043e-05, + "loss": 0.0425, + "step": 3430 + }, + { + "grad_norm": 0.23555654287338257, + "learning_rate": 7.81861666830322e-05, + "loss": 0.0401, + "step": 3440 + }, + { + "grad_norm": 0.16962042450904846, + "learning_rate": 7.804944221047097e-05, + "loss": 0.038, + "step": 3450 + }, + { + "grad_norm": 0.2654859721660614, + "learning_rate": 7.791241099370364e-05, + "loss": 0.042, + "step": 3460 + }, + { + "grad_norm": 0.277908593416214, + "learning_rate": 7.777507453128163e-05, + "loss": 0.0362, + "step": 3470 + }, + { + "grad_norm": 0.24589474499225616, + "learning_rate": 7.763743432509451e-05, + "loss": 0.0415, + "step": 3480 + }, + { + "grad_norm": 0.3107842206954956, + "learning_rate": 7.749949188035353e-05, + "loss": 0.0379, + "step": 3490 + }, + { + "grad_norm": 0.20410962402820587, + "learning_rate": 7.736124870557516e-05, + "loss": 0.045, + "step": 3500 + }, + { + "grad_norm": 0.23238706588745117, + "learning_rate": 7.722270631256459e-05, + "loss": 0.037, + "step": 3510 + }, + { + "grad_norm": 0.3061673045158386, + "learning_rate": 7.708386621639925e-05, + "loss": 0.0351, + "step": 3520 + }, + { + "grad_norm": 0.4029422998428345, + "learning_rate": 7.694472993541219e-05, + "loss": 0.0367, + "step": 3530 + }, + { + "grad_norm": 0.30271047353744507, + "learning_rate": 7.680529899117547e-05, + "loss": 0.0451, + "step": 3540 + }, + { + "grad_norm": 0.223578080534935, + "learning_rate": 7.666557490848358e-05, + "loss": 0.0387, + "step": 3550 + }, + { + "grad_norm": 0.3357184827327728, + "learning_rate": 7.65255592153367e-05, + "loss": 0.0418, + "step": 3560 + }, + { + "grad_norm": 0.2655865550041199, + "learning_rate": 7.638525344292402e-05, + "loss": 0.0421, + "step": 3570 + }, + { + "grad_norm": 0.2709006667137146, + "learning_rate": 7.624465912560697e-05, + "loss": 0.0408, + "step": 3580 + }, + { + "grad_norm": 0.23622557520866394, + "learning_rate": 7.610377780090249e-05, + "loss": 0.0357, + "step": 3590 + }, + { + "grad_norm": 0.2015763223171234, + "learning_rate": 7.596261100946618e-05, + "loss": 0.0403, + "step": 3600 + }, + { + "grad_norm": 0.19690196216106415, + "learning_rate": 7.582116029507542e-05, + "loss": 0.0345, + "step": 3610 + }, + { + "grad_norm": 0.29017189145088196, + "learning_rate": 7.56794272046126e-05, + "loss": 0.0392, + "step": 3620 + }, + { + "grad_norm": 0.35128021240234375, + "learning_rate": 7.55374132880481e-05, + "loss": 0.0436, + "step": 3630 + }, + { + "grad_norm": 0.3749406039714813, + "learning_rate": 7.539512009842333e-05, + "loss": 0.0444, + "step": 3640 + }, + { + "grad_norm": 0.213886559009552, + "learning_rate": 7.525254919183382e-05, + "loss": 0.0394, + "step": 3650 + }, + { + "grad_norm": 0.388240784406662, + "learning_rate": 7.510970212741215e-05, + "loss": 0.0349, + "step": 3660 + }, + { + "grad_norm": 0.23668386042118073, + "learning_rate": 7.496658046731096e-05, + "loss": 0.0412, + "step": 3670 + }, + { + "grad_norm": 0.267797589302063, + "learning_rate": 7.482318577668578e-05, + "loss": 0.0398, + "step": 3680 + }, + { + "grad_norm": 0.3219735622406006, + "learning_rate": 7.467951962367796e-05, + "loss": 0.0413, + "step": 3690 + }, + { + "grad_norm": 0.2474355846643448, + "learning_rate": 7.453558357939755e-05, + "loss": 0.0411, + "step": 3700 + }, + { + "grad_norm": 0.2815854251384735, + "learning_rate": 7.439137921790606e-05, + "loss": 0.0425, + "step": 3710 + }, + { + "grad_norm": 0.27210357785224915, + "learning_rate": 7.42469081161993e-05, + "loss": 0.0412, + "step": 3720 + }, + { + "grad_norm": 0.3003098666667938, + "learning_rate": 7.410217185419006e-05, + "loss": 0.0368, + "step": 3730 + }, + { + "grad_norm": 0.36017704010009766, + "learning_rate": 7.395717201469095e-05, + "loss": 0.0442, + "step": 3740 + }, + { + "grad_norm": 0.21263277530670166, + "learning_rate": 7.381191018339696e-05, + "loss": 0.0385, + "step": 3750 + }, + { + "grad_norm": 0.35366928577423096, + "learning_rate": 7.36663879488682e-05, + "loss": 0.0363, + "step": 3760 + }, + { + "grad_norm": 0.24734224379062653, + "learning_rate": 7.352060690251254e-05, + "loss": 0.0447, + "step": 3770 + }, + { + "grad_norm": 0.27061885595321655, + "learning_rate": 7.337456863856811e-05, + "loss": 0.0418, + "step": 3780 + }, + { + "grad_norm": 0.27420246601104736, + "learning_rate": 7.3228274754086e-05, + "loss": 0.0393, + "step": 3790 + }, + { + "grad_norm": 0.26580116152763367, + "learning_rate": 7.308172684891267e-05, + "loss": 0.0409, + "step": 3800 + }, + { + "grad_norm": 0.2976500391960144, + "learning_rate": 7.293492652567255e-05, + "loss": 0.0388, + "step": 3810 + }, + { + "grad_norm": 0.289000928401947, + "learning_rate": 7.278787538975043e-05, + "loss": 0.0383, + "step": 3820 + }, + { + "grad_norm": 0.24806618690490723, + "learning_rate": 7.2640575049274e-05, + "loss": 0.0388, + "step": 3830 + }, + { + "grad_norm": 0.25186148285865784, + "learning_rate": 7.249302711509616e-05, + "loss": 0.0372, + "step": 3840 + }, + { + "grad_norm": 0.20848169922828674, + "learning_rate": 7.23452332007775e-05, + "loss": 0.0363, + "step": 3850 + }, + { + "grad_norm": 0.21893630921840668, + "learning_rate": 7.219719492256858e-05, + "loss": 0.0453, + "step": 3860 + }, + { + "grad_norm": 0.282888263463974, + "learning_rate": 7.20489138993923e-05, + "loss": 0.0392, + "step": 3870 + }, + { + "grad_norm": 0.2675892114639282, + "learning_rate": 7.190039175282614e-05, + "loss": 0.0346, + "step": 3880 + }, + { + "grad_norm": 0.2734836935997009, + "learning_rate": 7.175163010708455e-05, + "loss": 0.038, + "step": 3890 + }, + { + "grad_norm": 0.2199399769306183, + "learning_rate": 7.1602630589001e-05, + "loss": 0.0418, + "step": 3900 + }, + { + "grad_norm": 0.3131192624568939, + "learning_rate": 7.14533948280104e-05, + "loss": 0.0407, + "step": 3910 + }, + { + "grad_norm": 0.2398889809846878, + "learning_rate": 7.130392445613109e-05, + "loss": 0.0391, + "step": 3920 + }, + { + "grad_norm": 0.2632913887500763, + "learning_rate": 7.115422110794711e-05, + "loss": 0.0436, + "step": 3930 + }, + { + "grad_norm": 0.20034050941467285, + "learning_rate": 7.100428642059033e-05, + "loss": 0.0423, + "step": 3940 + }, + { + "grad_norm": 0.2591931223869324, + "learning_rate": 7.08541220337224e-05, + "loss": 0.041, + "step": 3950 + }, + { + "grad_norm": 0.24841272830963135, + "learning_rate": 7.070372958951706e-05, + "loss": 0.0396, + "step": 3960 + }, + { + "grad_norm": 0.2559775114059448, + "learning_rate": 7.055311073264194e-05, + "loss": 0.0413, + "step": 3970 + }, + { + "grad_norm": 0.27920398116111755, + "learning_rate": 7.040226711024077e-05, + "loss": 0.034, + "step": 3980 + }, + { + "grad_norm": 0.2228671908378601, + "learning_rate": 7.02512003719152e-05, + "loss": 0.0335, + "step": 3990 + }, + { + "grad_norm": 0.29955047369003296, + "learning_rate": 7.00999121697069e-05, + "loss": 0.0418, + "step": 4000 + }, + { + "grad_norm": 0.22248847782611847, + "learning_rate": 6.99484041580794e-05, + "loss": 0.0363, + "step": 4010 + }, + { + "grad_norm": 0.22784554958343506, + "learning_rate": 6.979667799390004e-05, + "loss": 0.0356, + "step": 4020 + }, + { + "grad_norm": 0.3282862603664398, + "learning_rate": 6.964473533642185e-05, + "loss": 0.0376, + "step": 4030 + }, + { + "grad_norm": 0.20865017175674438, + "learning_rate": 6.949257784726539e-05, + "loss": 0.0382, + "step": 4040 + }, + { + "grad_norm": 0.24124769866466522, + "learning_rate": 6.934020719040056e-05, + "loss": 0.0369, + "step": 4050 + }, + { + "grad_norm": 0.2524951100349426, + "learning_rate": 6.918762503212848e-05, + "loss": 0.0461, + "step": 4060 + }, + { + "grad_norm": 0.25436753034591675, + "learning_rate": 6.903483304106319e-05, + "loss": 0.0381, + "step": 4070 + }, + { + "grad_norm": 0.21857470273971558, + "learning_rate": 6.888183288811341e-05, + "loss": 0.0385, + "step": 4080 + }, + { + "grad_norm": 0.2981796860694885, + "learning_rate": 6.87286262464643e-05, + "loss": 0.0395, + "step": 4090 + }, + { + "grad_norm": 0.2975570857524872, + "learning_rate": 6.857521479155915e-05, + "loss": 0.0381, + "step": 4100 + }, + { + "grad_norm": 0.3103868067264557, + "learning_rate": 6.842160020108104e-05, + "loss": 0.0391, + "step": 4110 + }, + { + "grad_norm": 0.23188205063343048, + "learning_rate": 6.826778415493455e-05, + "loss": 0.039, + "step": 4120 + }, + { + "grad_norm": 0.3271331787109375, + "learning_rate": 6.811376833522729e-05, + "loss": 0.0418, + "step": 4130 + }, + { + "grad_norm": 0.26595860719680786, + "learning_rate": 6.795955442625159e-05, + "loss": 0.038, + "step": 4140 + }, + { + "grad_norm": 0.23055477440357208, + "learning_rate": 6.780514411446608e-05, + "loss": 0.0372, + "step": 4150 + }, + { + "grad_norm": 0.3004310429096222, + "learning_rate": 6.765053908847716e-05, + "loss": 0.0333, + "step": 4160 + }, + { + "grad_norm": 0.23769573867321014, + "learning_rate": 6.749574103902064e-05, + "loss": 0.0415, + "step": 4170 + }, + { + "grad_norm": 0.3259679973125458, + "learning_rate": 6.734075165894317e-05, + "loss": 0.0404, + "step": 4180 + }, + { + "grad_norm": 0.274360716342926, + "learning_rate": 6.71855726431838e-05, + "loss": 0.0388, + "step": 4190 + }, + { + "grad_norm": 0.25235307216644287, + "learning_rate": 6.703020568875538e-05, + "loss": 0.0383, + "step": 4200 + }, + { + "grad_norm": 0.19202640652656555, + "learning_rate": 6.687465249472603e-05, + "loss": 0.0325, + "step": 4210 + }, + { + "grad_norm": 0.2002495974302292, + "learning_rate": 6.671891476220055e-05, + "loss": 0.0331, + "step": 4220 + }, + { + "grad_norm": 0.26579415798187256, + "learning_rate": 6.656299419430183e-05, + "loss": 0.0345, + "step": 4230 + }, + { + "grad_norm": 0.20920924842357635, + "learning_rate": 6.640689249615223e-05, + "loss": 0.0346, + "step": 4240 + }, + { + "grad_norm": 0.2713984251022339, + "learning_rate": 6.625061137485491e-05, + "loss": 0.0358, + "step": 4250 + }, + { + "grad_norm": 0.1806119829416275, + "learning_rate": 6.609415253947517e-05, + "loss": 0.0357, + "step": 4260 + }, + { + "grad_norm": 0.2060016244649887, + "learning_rate": 6.593751770102178e-05, + "loss": 0.0356, + "step": 4270 + }, + { + "grad_norm": 0.23546946048736572, + "learning_rate": 6.578070857242823e-05, + "loss": 0.0398, + "step": 4280 + }, + { + "grad_norm": 0.2152469903230667, + "learning_rate": 6.562372686853402e-05, + "loss": 0.0353, + "step": 4290 + }, + { + "grad_norm": 0.34248486161231995, + "learning_rate": 6.546657430606593e-05, + "loss": 0.0368, + "step": 4300 + }, + { + "grad_norm": 0.24777285754680634, + "learning_rate": 6.530925260361918e-05, + "loss": 0.0352, + "step": 4310 + }, + { + "grad_norm": 0.2766440808773041, + "learning_rate": 6.515176348163871e-05, + "loss": 0.0382, + "step": 4320 + }, + { + "grad_norm": 0.22699858248233795, + "learning_rate": 6.499410866240032e-05, + "loss": 0.0318, + "step": 4330 + }, + { + "grad_norm": 0.20788000524044037, + "learning_rate": 6.48362898699919e-05, + "loss": 0.0376, + "step": 4340 + }, + { + "grad_norm": 0.3406324088573456, + "learning_rate": 6.467830883029443e-05, + "loss": 0.0358, + "step": 4350 + }, + { + "grad_norm": 0.2882991135120392, + "learning_rate": 6.452016727096326e-05, + "loss": 0.0347, + "step": 4360 + }, + { + "grad_norm": 0.24940992891788483, + "learning_rate": 6.436186692140916e-05, + "loss": 0.0348, + "step": 4370 + }, + { + "grad_norm": 0.19990091025829315, + "learning_rate": 6.420340951277938e-05, + "loss": 0.0378, + "step": 4380 + }, + { + "grad_norm": 0.22545957565307617, + "learning_rate": 6.404479677793874e-05, + "loss": 0.039, + "step": 4390 + }, + { + "grad_norm": 0.24417118728160858, + "learning_rate": 6.388603045145075e-05, + "loss": 0.0427, + "step": 4400 + }, + { + "grad_norm": 0.27402248978614807, + "learning_rate": 6.372711226955843e-05, + "loss": 0.0362, + "step": 4410 + }, + { + "grad_norm": 0.296137273311615, + "learning_rate": 6.356804397016564e-05, + "loss": 0.0392, + "step": 4420 + }, + { + "grad_norm": 0.18187758326530457, + "learning_rate": 6.340882729281779e-05, + "loss": 0.0379, + "step": 4430 + }, + { + "grad_norm": 0.23053023219108582, + "learning_rate": 6.324946397868294e-05, + "loss": 0.0395, + "step": 4440 + }, + { + "grad_norm": 0.20781144499778748, + "learning_rate": 6.308995577053276e-05, + "loss": 0.0343, + "step": 4450 + }, + { + "grad_norm": 0.2324056178331375, + "learning_rate": 6.293030441272347e-05, + "loss": 0.0385, + "step": 4460 + }, + { + "grad_norm": 0.23749063909053802, + "learning_rate": 6.277051165117677e-05, + "loss": 0.0343, + "step": 4470 + }, + { + "grad_norm": 0.22948290407657623, + "learning_rate": 6.261057923336064e-05, + "loss": 0.0372, + "step": 4480 + }, + { + "grad_norm": 0.26891934871673584, + "learning_rate": 6.245050890827042e-05, + "loss": 0.0376, + "step": 4490 + }, + { + "grad_norm": 0.16613225638866425, + "learning_rate": 6.229030242640952e-05, + "loss": 0.0345, + "step": 4500 + }, + { + "grad_norm": 0.2771390378475189, + "learning_rate": 6.212996153977037e-05, + "loss": 0.0402, + "step": 4510 + }, + { + "grad_norm": 0.21101072430610657, + "learning_rate": 6.196948800181523e-05, + "loss": 0.0339, + "step": 4520 + }, + { + "grad_norm": 0.16528122127056122, + "learning_rate": 6.180888356745695e-05, + "loss": 0.0383, + "step": 4530 + }, + { + "grad_norm": 0.23599044978618622, + "learning_rate": 6.164814999303995e-05, + "loss": 0.0341, + "step": 4540 + }, + { + "grad_norm": 0.32073840498924255, + "learning_rate": 6.148728903632081e-05, + "loss": 0.0308, + "step": 4550 + }, + { + "grad_norm": 0.21075880527496338, + "learning_rate": 6.132630245644921e-05, + "loss": 0.0378, + "step": 4560 + }, + { + "grad_norm": 0.22400715947151184, + "learning_rate": 6.116519201394857e-05, + "loss": 0.0388, + "step": 4570 + }, + { + "grad_norm": 0.2810167372226715, + "learning_rate": 6.10039594706969e-05, + "loss": 0.034, + "step": 4580 + }, + { + "grad_norm": 0.2516627907752991, + "learning_rate": 6.084260658990744e-05, + "loss": 0.0334, + "step": 4590 + }, + { + "grad_norm": 0.2997600734233856, + "learning_rate": 6.068113513610943e-05, + "loss": 0.0321, + "step": 4600 + }, + { + "grad_norm": 0.19859614968299866, + "learning_rate": 6.0519546875128876e-05, + "loss": 0.0336, + "step": 4610 + }, + { + "grad_norm": 0.1953340768814087, + "learning_rate": 6.035784357406906e-05, + "loss": 0.0366, + "step": 4620 + }, + { + "grad_norm": 0.22977666556835175, + "learning_rate": 6.01960270012914e-05, + "loss": 0.0337, + "step": 4630 + }, + { + "grad_norm": 0.21568147838115692, + "learning_rate": 6.003409892639599e-05, + "loss": 0.0371, + "step": 4640 + }, + { + "grad_norm": 0.2854381501674652, + "learning_rate": 5.9872061120202336e-05, + "loss": 0.0361, + "step": 4650 + }, + { + "grad_norm": 0.2356075942516327, + "learning_rate": 5.9709915354729914e-05, + "loss": 0.0431, + "step": 4660 + }, + { + "grad_norm": 0.2051800787448883, + "learning_rate": 5.9547663403178824e-05, + "loss": 0.0371, + "step": 4670 + }, + { + "grad_norm": 0.20173141360282898, + "learning_rate": 5.9385307039910445e-05, + "loss": 0.0328, + "step": 4680 + }, + { + "grad_norm": 0.2628331482410431, + "learning_rate": 5.922284804042792e-05, + "loss": 0.0401, + "step": 4690 + }, + { + "grad_norm": 0.244556725025177, + "learning_rate": 5.906028818135687e-05, + "loss": 0.038, + "step": 4700 + }, + { + "grad_norm": 0.21201924979686737, + "learning_rate": 5.889762924042585e-05, + "loss": 0.0341, + "step": 4710 + }, + { + "grad_norm": 0.18467576801776886, + "learning_rate": 5.873487299644699e-05, + "loss": 0.0348, + "step": 4720 + }, + { + "grad_norm": 0.2353990375995636, + "learning_rate": 5.857202122929649e-05, + "loss": 0.0362, + "step": 4730 + }, + { + "grad_norm": 0.27192428708076477, + "learning_rate": 5.840907571989518e-05, + "loss": 0.0332, + "step": 4740 + }, + { + "grad_norm": 0.26230132579803467, + "learning_rate": 5.824603825018904e-05, + "loss": 0.0322, + "step": 4750 + }, + { + "grad_norm": 0.29458925127983093, + "learning_rate": 5.808291060312975e-05, + "loss": 0.0314, + "step": 4760 + }, + { + "grad_norm": 0.2697901129722595, + "learning_rate": 5.7919694562655083e-05, + "loss": 0.0356, + "step": 4770 + }, + { + "grad_norm": 0.2372838407754898, + "learning_rate": 5.775639191366954e-05, + "loss": 0.0375, + "step": 4780 + }, + { + "grad_norm": 0.2567492425441742, + "learning_rate": 5.75930044420247e-05, + "loss": 0.0344, + "step": 4790 + }, + { + "grad_norm": 0.22764883935451508, + "learning_rate": 5.74295339344998e-05, + "loss": 0.033, + "step": 4800 + }, + { + "grad_norm": 0.2397421896457672, + "learning_rate": 5.726598217878211e-05, + "loss": 0.0309, + "step": 4810 + }, + { + "grad_norm": 0.269631564617157, + "learning_rate": 5.71023509634474e-05, + "loss": 0.0328, + "step": 4820 + }, + { + "grad_norm": 0.23010453581809998, + "learning_rate": 5.693864207794049e-05, + "loss": 0.0388, + "step": 4830 + }, + { + "grad_norm": 0.26805612444877625, + "learning_rate": 5.677485731255545e-05, + "loss": 0.044, + "step": 4840 + }, + { + "grad_norm": 0.2965623140335083, + "learning_rate": 5.6610998458416296e-05, + "loss": 0.036, + "step": 4850 + }, + { + "grad_norm": 0.3152885437011719, + "learning_rate": 5.644706730745716e-05, + "loss": 0.0342, + "step": 4860 + }, + { + "grad_norm": 0.306333988904953, + "learning_rate": 5.628306565240287e-05, + "loss": 0.032, + "step": 4870 + }, + { + "grad_norm": 0.25830549001693726, + "learning_rate": 5.611899528674923e-05, + "loss": 0.0387, + "step": 4880 + }, + { + "grad_norm": 0.25811290740966797, + "learning_rate": 5.595485800474349e-05, + "loss": 0.034, + "step": 4890 + }, + { + "grad_norm": 0.25292131304740906, + "learning_rate": 5.579065560136467e-05, + "loss": 0.0354, + "step": 4900 + }, + { + "grad_norm": 0.1542666107416153, + "learning_rate": 5.562638987230392e-05, + "loss": 0.0308, + "step": 4910 + }, + { + "grad_norm": 0.28586792945861816, + "learning_rate": 5.546206261394498e-05, + "loss": 0.0297, + "step": 4920 + }, + { + "grad_norm": 0.2532980144023895, + "learning_rate": 5.529767562334437e-05, + "loss": 0.0357, + "step": 4930 + }, + { + "grad_norm": 0.28376898169517517, + "learning_rate": 5.5133230698211926e-05, + "loss": 0.0338, + "step": 4940 + }, + { + "grad_norm": 0.21100687980651855, + "learning_rate": 5.496872963689096e-05, + "loss": 0.0375, + "step": 4950 + }, + { + "grad_norm": 0.21206516027450562, + "learning_rate": 5.4804174238338756e-05, + "loss": 0.0316, + "step": 4960 + }, + { + "grad_norm": 0.25067418813705444, + "learning_rate": 5.463956630210678e-05, + "loss": 0.0358, + "step": 4970 + }, + { + "grad_norm": 0.5143365263938904, + "learning_rate": 5.4474907628321046e-05, + "loss": 0.0383, + "step": 4980 + }, + { + "grad_norm": 0.23279064893722534, + "learning_rate": 5.431020001766244e-05, + "loss": 0.0305, + "step": 4990 + }, + { + "grad_norm": 0.20432204008102417, + "learning_rate": 5.4145445271346986e-05, + "loss": 0.0303, + "step": 5000 + }, + { + "grad_norm": 0.2530362904071808, + "learning_rate": 5.398064519110622e-05, + "loss": 0.0331, + "step": 5010 + }, + { + "grad_norm": 0.20965495705604553, + "learning_rate": 5.3815801579167394e-05, + "loss": 0.0323, + "step": 5020 + }, + { + "grad_norm": 0.1938072144985199, + "learning_rate": 5.365091623823382e-05, + "loss": 0.0347, + "step": 5030 + }, + { + "grad_norm": 0.1815754622220993, + "learning_rate": 5.348599097146521e-05, + "loss": 0.0313, + "step": 5040 + }, + { + "grad_norm": 0.22825242578983307, + "learning_rate": 5.3321027582457836e-05, + "loss": 0.0358, + "step": 5050 + }, + { + "grad_norm": 0.2713790535926819, + "learning_rate": 5.315602787522491e-05, + "loss": 0.0353, + "step": 5060 + }, + { + "grad_norm": 0.20216213166713715, + "learning_rate": 5.299099365417678e-05, + "loss": 0.0271, + "step": 5070 + }, + { + "grad_norm": 0.22530780732631683, + "learning_rate": 5.2825926724101236e-05, + "loss": 0.0327, + "step": 5080 + }, + { + "grad_norm": 0.2990708649158478, + "learning_rate": 5.26608288901438e-05, + "loss": 0.0353, + "step": 5090 + }, + { + "grad_norm": 0.2576608657836914, + "learning_rate": 5.24957019577879e-05, + "loss": 0.0384, + "step": 5100 + }, + { + "grad_norm": 0.24573999643325806, + "learning_rate": 5.2330547732835266e-05, + "loss": 0.0328, + "step": 5110 + }, + { + "grad_norm": 0.3096584975719452, + "learning_rate": 5.2165368021385996e-05, + "loss": 0.0405, + "step": 5120 + }, + { + "grad_norm": 0.16731584072113037, + "learning_rate": 5.200016462981897e-05, + "loss": 0.031, + "step": 5130 + }, + { + "grad_norm": 0.22490406036376953, + "learning_rate": 5.1834939364772015e-05, + "loss": 0.0299, + "step": 5140 + }, + { + "grad_norm": 0.20373359322547913, + "learning_rate": 5.166969403312214e-05, + "loss": 0.0327, + "step": 5150 + }, + { + "grad_norm": 0.25911012291908264, + "learning_rate": 5.1504430441965844e-05, + "loss": 0.0335, + "step": 5160 + }, + { + "grad_norm": 0.1642448455095291, + "learning_rate": 5.133915039859923e-05, + "loss": 0.0258, + "step": 5170 + }, + { + "grad_norm": 0.29947248101234436, + "learning_rate": 5.1173855710498444e-05, + "loss": 0.0325, + "step": 5180 + }, + { + "grad_norm": 0.22944921255111694, + "learning_rate": 5.100854818529967e-05, + "loss": 0.0361, + "step": 5190 + }, + { + "grad_norm": 0.24382486939430237, + "learning_rate": 5.084322963077951e-05, + "loss": 0.0379, + "step": 5200 + }, + { + "grad_norm": 0.23937542736530304, + "learning_rate": 5.067790185483522e-05, + "loss": 0.0338, + "step": 5210 + }, + { + "grad_norm": 0.15932729840278625, + "learning_rate": 5.0512566665464844e-05, + "loss": 0.0311, + "step": 5220 + }, + { + "grad_norm": 0.2194763571023941, + "learning_rate": 5.034722587074755e-05, + "loss": 0.0312, + "step": 5230 + }, + { + "grad_norm": 0.226780503988266, + "learning_rate": 5.018188127882375e-05, + "loss": 0.0309, + "step": 5240 + }, + { + "grad_norm": 0.1667359173297882, + "learning_rate": 5.0016534697875417e-05, + "loss": 0.0287, + "step": 5250 + }, + { + "grad_norm": 0.1375981569290161, + "learning_rate": 4.9851187936106294e-05, + "loss": 0.035, + "step": 5260 + }, + { + "grad_norm": 0.28929659724235535, + "learning_rate": 4.968584280172206e-05, + "loss": 0.033, + "step": 5270 + }, + { + "grad_norm": 0.1564723402261734, + "learning_rate": 4.95205011029106e-05, + "loss": 0.0342, + "step": 5280 + }, + { + "grad_norm": 0.21241077780723572, + "learning_rate": 4.935516464782227e-05, + "loss": 0.0285, + "step": 5290 + }, + { + "grad_norm": 0.20711779594421387, + "learning_rate": 4.918983524455003e-05, + "loss": 0.0316, + "step": 5300 + }, + { + "grad_norm": 0.2070685178041458, + "learning_rate": 4.9024514701109766e-05, + "loss": 0.029, + "step": 5310 + }, + { + "grad_norm": 0.23496617376804352, + "learning_rate": 4.885920482542043e-05, + "loss": 0.0349, + "step": 5320 + }, + { + "grad_norm": 0.4176364839076996, + "learning_rate": 4.869390742528438e-05, + "loss": 0.0325, + "step": 5330 + }, + { + "grad_norm": 0.27644309401512146, + "learning_rate": 4.852862430836744e-05, + "loss": 0.0367, + "step": 5340 + }, + { + "grad_norm": 0.25068041682243347, + "learning_rate": 4.836335728217933e-05, + "loss": 0.0302, + "step": 5350 + }, + { + "grad_norm": 0.18797720968723297, + "learning_rate": 4.819810815405379e-05, + "loss": 0.0317, + "step": 5360 + }, + { + "grad_norm": 0.27008476853370667, + "learning_rate": 4.803287873112877e-05, + "loss": 0.033, + "step": 5370 + }, + { + "grad_norm": 0.3013102114200592, + "learning_rate": 4.786767082032681e-05, + "loss": 0.0318, + "step": 5380 + }, + { + "grad_norm": 0.2835000157356262, + "learning_rate": 4.77024862283351e-05, + "loss": 0.0295, + "step": 5390 + }, + { + "grad_norm": 0.22792765498161316, + "learning_rate": 4.753732676158593e-05, + "loss": 0.0303, + "step": 5400 + }, + { + "grad_norm": 0.21493951976299286, + "learning_rate": 4.737219422623672e-05, + "loss": 0.031, + "step": 5410 + }, + { + "grad_norm": 0.25162848830223083, + "learning_rate": 4.720709042815044e-05, + "loss": 0.0279, + "step": 5420 + }, + { + "grad_norm": 0.29204848408699036, + "learning_rate": 4.704201717287578e-05, + "loss": 0.0317, + "step": 5430 + }, + { + "grad_norm": 0.21430036425590515, + "learning_rate": 4.6876976265627404e-05, + "loss": 0.0294, + "step": 5440 + }, + { + "grad_norm": 0.2776707410812378, + "learning_rate": 4.671196951126626e-05, + "loss": 0.0369, + "step": 5450 + }, + { + "grad_norm": 0.2198633849620819, + "learning_rate": 4.654699871427971e-05, + "loss": 0.029, + "step": 5460 + }, + { + "grad_norm": 0.22923733294010162, + "learning_rate": 4.6382065678762034e-05, + "loss": 0.0298, + "step": 5470 + }, + { + "grad_norm": 0.18683795630931854, + "learning_rate": 4.6217172208394424e-05, + "loss": 0.0306, + "step": 5480 + }, + { + "grad_norm": 0.29420194029808044, + "learning_rate": 4.605232010642549e-05, + "loss": 0.0259, + "step": 5490 + }, + { + "grad_norm": 0.2075209766626358, + "learning_rate": 4.588751117565142e-05, + "loss": 0.0279, + "step": 5500 + }, + { + "grad_norm": 0.16531507670879364, + "learning_rate": 4.5722747218396214e-05, + "loss": 0.0285, + "step": 5510 + }, + { + "grad_norm": 0.2374963015317917, + "learning_rate": 4.5558030036492194e-05, + "loss": 0.0268, + "step": 5520 + }, + { + "grad_norm": 0.2805657982826233, + "learning_rate": 4.539336143125999e-05, + "loss": 0.0402, + "step": 5530 + }, + { + "grad_norm": 0.23020616173744202, + "learning_rate": 4.522874320348916e-05, + "loss": 0.0314, + "step": 5540 + }, + { + "grad_norm": 0.31708887219429016, + "learning_rate": 4.506417715341821e-05, + "loss": 0.0291, + "step": 5550 + }, + { + "grad_norm": 0.3270540237426758, + "learning_rate": 4.489966508071511e-05, + "loss": 0.0296, + "step": 5560 + }, + { + "grad_norm": 0.30636781454086304, + "learning_rate": 4.4735208784457575e-05, + "loss": 0.0306, + "step": 5570 + }, + { + "grad_norm": 0.15036705136299133, + "learning_rate": 4.457081006311325e-05, + "loss": 0.0279, + "step": 5580 + }, + { + "grad_norm": 0.20207194983959198, + "learning_rate": 4.440647071452027e-05, + "loss": 0.0298, + "step": 5590 + }, + { + "grad_norm": 0.3105098605155945, + "learning_rate": 4.424219253586737e-05, + "loss": 0.0316, + "step": 5600 + }, + { + "grad_norm": 0.26028376817703247, + "learning_rate": 4.407797732367443e-05, + "loss": 0.0332, + "step": 5610 + }, + { + "grad_norm": 0.2690197229385376, + "learning_rate": 4.391382687377268e-05, + "loss": 0.0311, + "step": 5620 + }, + { + "grad_norm": 0.17695750296115875, + "learning_rate": 4.374974298128512e-05, + "loss": 0.0295, + "step": 5630 + }, + { + "grad_norm": 0.23683609068393707, + "learning_rate": 4.358572744060699e-05, + "loss": 0.034, + "step": 5640 + }, + { + "grad_norm": 0.20435236394405365, + "learning_rate": 4.342178204538588e-05, + "loss": 0.0244, + "step": 5650 + }, + { + "grad_norm": 0.2131694257259369, + "learning_rate": 4.325790858850241e-05, + "loss": 0.0309, + "step": 5660 + }, + { + "grad_norm": 0.1823224425315857, + "learning_rate": 4.309410886205043e-05, + "loss": 0.0309, + "step": 5670 + }, + { + "grad_norm": 0.23058725893497467, + "learning_rate": 4.293038465731752e-05, + "loss": 0.0292, + "step": 5680 + }, + { + "grad_norm": 0.16288995742797852, + "learning_rate": 4.276673776476533e-05, + "loss": 0.0316, + "step": 5690 + }, + { + "grad_norm": 0.15214325487613678, + "learning_rate": 4.260316997401007e-05, + "loss": 0.0236, + "step": 5700 + }, + { + "grad_norm": 0.27146801352500916, + "learning_rate": 4.243968307380293e-05, + "loss": 0.031, + "step": 5710 + }, + { + "grad_norm": 0.26338663697242737, + "learning_rate": 4.22762788520104e-05, + "loss": 0.0326, + "step": 5720 + }, + { + "grad_norm": 0.1726326048374176, + "learning_rate": 4.211295909559491e-05, + "loss": 0.0337, + "step": 5730 + }, + { + "grad_norm": 0.2653937041759491, + "learning_rate": 4.194972559059511e-05, + "loss": 0.029, + "step": 5740 + }, + { + "grad_norm": 0.2020241618156433, + "learning_rate": 4.178658012210651e-05, + "loss": 0.03, + "step": 5750 + }, + { + "grad_norm": 0.20574520528316498, + "learning_rate": 4.162352447426177e-05, + "loss": 0.0293, + "step": 5760 + }, + { + "grad_norm": 0.2579389810562134, + "learning_rate": 4.146056043021135e-05, + "loss": 0.0273, + "step": 5770 + }, + { + "grad_norm": 0.17306716740131378, + "learning_rate": 4.1297689772103944e-05, + "loss": 0.0332, + "step": 5780 + }, + { + "grad_norm": 0.14769788086414337, + "learning_rate": 4.113491428106694e-05, + "loss": 0.0338, + "step": 5790 + }, + { + "grad_norm": 0.19954118132591248, + "learning_rate": 4.0972235737187055e-05, + "loss": 0.0284, + "step": 5800 + }, + { + "grad_norm": 0.1594008505344391, + "learning_rate": 4.080965591949076e-05, + "loss": 0.0282, + "step": 5810 + }, + { + "grad_norm": 0.18839597702026367, + "learning_rate": 4.0647176605924924e-05, + "loss": 0.0259, + "step": 5820 + }, + { + "grad_norm": 0.18666644394397736, + "learning_rate": 4.0484799573337255e-05, + "loss": 0.0315, + "step": 5830 + }, + { + "grad_norm": 0.2885925769805908, + "learning_rate": 4.032252659745699e-05, + "loss": 0.03, + "step": 5840 + }, + { + "grad_norm": 0.21594734489917755, + "learning_rate": 4.016035945287539e-05, + "loss": 0.0328, + "step": 5850 + }, + { + "grad_norm": 0.18419182300567627, + "learning_rate": 3.999829991302635e-05, + "loss": 0.0287, + "step": 5860 + }, + { + "grad_norm": 0.20551085472106934, + "learning_rate": 3.983634975016707e-05, + "loss": 0.0297, + "step": 5870 + }, + { + "grad_norm": 0.19250456988811493, + "learning_rate": 3.967451073535854e-05, + "loss": 0.0368, + "step": 5880 + }, + { + "grad_norm": 0.20462240278720856, + "learning_rate": 3.951278463844633e-05, + "loss": 0.0302, + "step": 5890 + }, + { + "grad_norm": 0.17498831450939178, + "learning_rate": 3.935117322804111e-05, + "loss": 0.0337, + "step": 5900 + }, + { + "grad_norm": 0.13612870872020721, + "learning_rate": 3.918967827149938e-05, + "loss": 0.0298, + "step": 5910 + }, + { + "grad_norm": 0.21322321891784668, + "learning_rate": 3.9028301534904094e-05, + "loss": 0.0317, + "step": 5920 + }, + { + "grad_norm": 0.17184172570705414, + "learning_rate": 3.88670447830454e-05, + "loss": 0.0271, + "step": 5930 + }, + { + "grad_norm": 0.17313244938850403, + "learning_rate": 3.870590977940132e-05, + "loss": 0.0275, + "step": 5940 + }, + { + "grad_norm": 0.21457086503505707, + "learning_rate": 3.8544898286118404e-05, + "loss": 0.0295, + "step": 5950 + }, + { + "grad_norm": 0.20838065445423126, + "learning_rate": 3.838401206399257e-05, + "loss": 0.028, + "step": 5960 + }, + { + "grad_norm": 0.16620972752571106, + "learning_rate": 3.822325287244975e-05, + "loss": 0.0248, + "step": 5970 + }, + { + "grad_norm": 0.3082650899887085, + "learning_rate": 3.8062622469526725e-05, + "loss": 0.03, + "step": 5980 + }, + { + "grad_norm": 0.3031490743160248, + "learning_rate": 3.790212261185183e-05, + "loss": 0.0277, + "step": 5990 + }, + { + "grad_norm": 0.20429182052612305, + "learning_rate": 3.7741755054625794e-05, + "loss": 0.028, + "step": 6000 + }, + { + "grad_norm": 0.28596487641334534, + "learning_rate": 3.758152155160255e-05, + "loss": 0.0271, + "step": 6010 + }, + { + "grad_norm": 0.19709128141403198, + "learning_rate": 3.742142385506999e-05, + "loss": 0.0267, + "step": 6020 + }, + { + "grad_norm": 0.25257137417793274, + "learning_rate": 3.72614637158309e-05, + "loss": 0.0239, + "step": 6030 + }, + { + "grad_norm": 0.16534167528152466, + "learning_rate": 3.710164288318371e-05, + "loss": 0.0293, + "step": 6040 + }, + { + "grad_norm": 0.18057389557361603, + "learning_rate": 3.694196310490345e-05, + "loss": 0.0246, + "step": 6050 + }, + { + "grad_norm": 0.2280239760875702, + "learning_rate": 3.678242612722259e-05, + "loss": 0.0266, + "step": 6060 + }, + { + "grad_norm": 0.2778986990451813, + "learning_rate": 3.6623033694811953e-05, + "loss": 0.0282, + "step": 6070 + }, + { + "grad_norm": 0.19239820539951324, + "learning_rate": 3.6463787550761665e-05, + "loss": 0.031, + "step": 6080 + }, + { + "grad_norm": 0.20039702951908112, + "learning_rate": 3.630468943656202e-05, + "loss": 0.024, + "step": 6090 + }, + { + "grad_norm": 0.24246586859226227, + "learning_rate": 3.6145741092084523e-05, + "loss": 0.024, + "step": 6100 + }, + { + "grad_norm": 0.17191071808338165, + "learning_rate": 3.598694425556278e-05, + "loss": 0.0268, + "step": 6110 + }, + { + "grad_norm": 0.3668980002403259, + "learning_rate": 3.58283006635736e-05, + "loss": 0.028, + "step": 6120 + }, + { + "grad_norm": 0.2849811315536499, + "learning_rate": 3.566981205101781e-05, + "loss": 0.03, + "step": 6130 + }, + { + "grad_norm": 0.20126973092556, + "learning_rate": 3.5511480151101556e-05, + "loss": 0.0261, + "step": 6140 + }, + { + "grad_norm": 0.19636747241020203, + "learning_rate": 3.5353306695317104e-05, + "loss": 0.0242, + "step": 6150 + }, + { + "grad_norm": 0.15131956338882446, + "learning_rate": 3.519529341342402e-05, + "loss": 0.0263, + "step": 6160 + }, + { + "grad_norm": 0.23002402484416962, + "learning_rate": 3.503744203343026e-05, + "loss": 0.0271, + "step": 6170 + }, + { + "grad_norm": 0.2139393538236618, + "learning_rate": 3.487975428157318e-05, + "loss": 0.0282, + "step": 6180 + }, + { + "grad_norm": 0.20546658337116241, + "learning_rate": 3.472223188230083e-05, + "loss": 0.0309, + "step": 6190 + }, + { + "grad_norm": 0.18871566653251648, + "learning_rate": 3.4564876558252866e-05, + "loss": 0.0277, + "step": 6200 + }, + { + "grad_norm": 0.20693008601665497, + "learning_rate": 3.440769003024195e-05, + "loss": 0.0277, + "step": 6210 + }, + { + "grad_norm": 0.18188361823558807, + "learning_rate": 3.425067401723477e-05, + "loss": 0.0263, + "step": 6220 + }, + { + "grad_norm": 0.1726355403661728, + "learning_rate": 3.409383023633325e-05, + "loss": 0.0304, + "step": 6230 + }, + { + "grad_norm": 0.261430561542511, + "learning_rate": 3.3937160402755894e-05, + "loss": 0.0279, + "step": 6240 + }, + { + "grad_norm": 0.19474512338638306, + "learning_rate": 3.378066622981885e-05, + "loss": 0.0257, + "step": 6250 + }, + { + "grad_norm": 0.2604807913303375, + "learning_rate": 3.362434942891738e-05, + "loss": 0.0268, + "step": 6260 + }, + { + "grad_norm": 0.18818962574005127, + "learning_rate": 3.346821170950693e-05, + "loss": 0.0258, + "step": 6270 + }, + { + "grad_norm": 0.2013111263513565, + "learning_rate": 3.3312254779084585e-05, + "loss": 0.0287, + "step": 6280 + }, + { + "grad_norm": 0.22038617730140686, + "learning_rate": 3.315648034317039e-05, + "loss": 0.0307, + "step": 6290 + }, + { + "grad_norm": 0.18479305505752563, + "learning_rate": 3.3000890105288564e-05, + "loss": 0.0265, + "step": 6300 + }, + { + "grad_norm": 0.19435778260231018, + "learning_rate": 3.284548576694908e-05, + "loss": 0.0273, + "step": 6310 + }, + { + "grad_norm": 0.2135755717754364, + "learning_rate": 3.2690269027628815e-05, + "loss": 0.0273, + "step": 6320 + }, + { + "grad_norm": 0.18725253641605377, + "learning_rate": 3.253524158475324e-05, + "loss": 0.027, + "step": 6330 + }, + { + "grad_norm": 0.20736408233642578, + "learning_rate": 3.238040513367757e-05, + "loss": 0.0264, + "step": 6340 + }, + { + "grad_norm": 0.2280939519405365, + "learning_rate": 3.222576136766843e-05, + "loss": 0.0238, + "step": 6350 + }, + { + "grad_norm": 0.15856464207172394, + "learning_rate": 3.2071311977885324e-05, + "loss": 0.0279, + "step": 6360 + }, + { + "grad_norm": 0.18191306293010712, + "learning_rate": 3.191705865336197e-05, + "loss": 0.0293, + "step": 6370 + }, + { + "grad_norm": 0.2248312085866928, + "learning_rate": 3.1763003080988075e-05, + "loss": 0.0284, + "step": 6380 + }, + { + "grad_norm": 0.197534441947937, + "learning_rate": 3.160914694549063e-05, + "loss": 0.0304, + "step": 6390 + }, + { + "grad_norm": 0.23441855609416962, + "learning_rate": 3.145549192941573e-05, + "loss": 0.027, + "step": 6400 + }, + { + "grad_norm": 0.18199796974658966, + "learning_rate": 3.130203971310999e-05, + "loss": 0.0297, + "step": 6410 + }, + { + "grad_norm": 0.20147624611854553, + "learning_rate": 3.114879197470225e-05, + "loss": 0.0264, + "step": 6420 + }, + { + "grad_norm": 0.1779860258102417, + "learning_rate": 3.0995750390085285e-05, + "loss": 0.0234, + "step": 6430 + }, + { + "grad_norm": 0.21999907493591309, + "learning_rate": 3.084291663289728e-05, + "loss": 0.0249, + "step": 6440 + }, + { + "grad_norm": 0.21762172877788544, + "learning_rate": 3.069029237450375e-05, + "loss": 0.0257, + "step": 6450 + }, + { + "grad_norm": 0.191232830286026, + "learning_rate": 3.053787928397911e-05, + "loss": 0.0249, + "step": 6460 + }, + { + "grad_norm": 0.2265884429216385, + "learning_rate": 3.0385679028088526e-05, + "loss": 0.0282, + "step": 6470 + }, + { + "grad_norm": 0.24025024473667145, + "learning_rate": 3.023369327126959e-05, + "loss": 0.0224, + "step": 6480 + }, + { + "grad_norm": 0.1672016680240631, + "learning_rate": 3.0081923675614198e-05, + "loss": 0.0265, + "step": 6490 + }, + { + "grad_norm": 0.20879749953746796, + "learning_rate": 2.993037190085034e-05, + "loss": 0.0239, + "step": 6500 + }, + { + "grad_norm": 0.43519845604896545, + "learning_rate": 2.977903960432392e-05, + "loss": 0.0276, + "step": 6510 + }, + { + "grad_norm": 0.2978155314922333, + "learning_rate": 2.9627928440980722e-05, + "loss": 0.0264, + "step": 6520 + }, + { + "grad_norm": 0.2873488664627075, + "learning_rate": 2.9477040063348183e-05, + "loss": 0.0292, + "step": 6530 + }, + { + "grad_norm": 0.22392164170742035, + "learning_rate": 2.9326376121517456e-05, + "loss": 0.0276, + "step": 6540 + }, + { + "grad_norm": 0.2012801617383957, + "learning_rate": 2.9175938263125236e-05, + "loss": 0.0262, + "step": 6550 + }, + { + "grad_norm": 0.18279588222503662, + "learning_rate": 2.9025728133335873e-05, + "loss": 0.0265, + "step": 6560 + }, + { + "grad_norm": 0.1400076150894165, + "learning_rate": 2.8875747374823288e-05, + "loss": 0.0281, + "step": 6570 + }, + { + "grad_norm": 0.23712073266506195, + "learning_rate": 2.872599762775298e-05, + "loss": 0.0329, + "step": 6580 + }, + { + "grad_norm": 0.21293634176254272, + "learning_rate": 2.857648052976425e-05, + "loss": 0.0225, + "step": 6590 + }, + { + "grad_norm": 0.18780572712421417, + "learning_rate": 2.8427197715952047e-05, + "loss": 0.0229, + "step": 6600 + }, + { + "grad_norm": 0.2058478444814682, + "learning_rate": 2.8278150818849393e-05, + "loss": 0.0284, + "step": 6610 + }, + { + "grad_norm": 0.18974906206130981, + "learning_rate": 2.812934146840922e-05, + "loss": 0.0224, + "step": 6620 + }, + { + "grad_norm": 0.2508644461631775, + "learning_rate": 2.7980771291986764e-05, + "loss": 0.0282, + "step": 6630 + }, + { + "grad_norm": 0.20421051979064941, + "learning_rate": 2.783244191432167e-05, + "loss": 0.0272, + "step": 6640 + }, + { + "grad_norm": 0.25093674659729004, + "learning_rate": 2.768435495752022e-05, + "loss": 0.0245, + "step": 6650 + }, + { + "grad_norm": 0.19735698401927948, + "learning_rate": 2.753651204103771e-05, + "loss": 0.0244, + "step": 6660 + }, + { + "grad_norm": 0.19064104557037354, + "learning_rate": 2.7388914781660523e-05, + "loss": 0.0238, + "step": 6670 + }, + { + "grad_norm": 0.21755774319171906, + "learning_rate": 2.7241564793488693e-05, + "loss": 0.0259, + "step": 6680 + }, + { + "grad_norm": 0.22687704861164093, + "learning_rate": 2.7094463687918037e-05, + "loss": 0.025, + "step": 6690 + }, + { + "grad_norm": 0.23115026950836182, + "learning_rate": 2.694761307362268e-05, + "loss": 0.023, + "step": 6700 + }, + { + "grad_norm": 0.23495085537433624, + "learning_rate": 2.6801014556537467e-05, + "loss": 0.0251, + "step": 6710 + }, + { + "grad_norm": 0.23663659393787384, + "learning_rate": 2.6654669739840243e-05, + "loss": 0.0198, + "step": 6720 + }, + { + "grad_norm": 0.22000780701637268, + "learning_rate": 2.650858022393451e-05, + "loss": 0.0264, + "step": 6730 + }, + { + "grad_norm": 0.18597948551177979, + "learning_rate": 2.6362747606431747e-05, + "loss": 0.0266, + "step": 6740 + }, + { + "grad_norm": 0.12833118438720703, + "learning_rate": 2.6217173482134172e-05, + "loss": 0.0234, + "step": 6750 + }, + { + "grad_norm": 0.16999608278274536, + "learning_rate": 2.6071859443017044e-05, + "loss": 0.0227, + "step": 6760 + }, + { + "grad_norm": 0.2000948190689087, + "learning_rate": 2.5926807078211414e-05, + "loss": 0.0223, + "step": 6770 + }, + { + "grad_norm": 0.20711363852024078, + "learning_rate": 2.5782017973986728e-05, + "loss": 0.0247, + "step": 6780 + }, + { + "grad_norm": 0.1407475471496582, + "learning_rate": 2.5637493713733374e-05, + "loss": 0.0212, + "step": 6790 + }, + { + "grad_norm": 0.21618331968784332, + "learning_rate": 2.549323587794559e-05, + "loss": 0.0299, + "step": 6800 + }, + { + "grad_norm": 0.19166132807731628, + "learning_rate": 2.5349246044203895e-05, + "loss": 0.0212, + "step": 6810 + }, + { + "grad_norm": 0.2313280701637268, + "learning_rate": 2.520552578715808e-05, + "loss": 0.0252, + "step": 6820 + }, + { + "grad_norm": 0.16042548418045044, + "learning_rate": 2.506207667850981e-05, + "loss": 0.02, + "step": 6830 + }, + { + "grad_norm": 0.20389500260353088, + "learning_rate": 2.4918900286995555e-05, + "loss": 0.0257, + "step": 6840 + }, + { + "grad_norm": 0.12122675031423569, + "learning_rate": 2.4775998178369458e-05, + "loss": 0.0238, + "step": 6850 + }, + { + "grad_norm": 0.1844255030155182, + "learning_rate": 2.4633371915386017e-05, + "loss": 0.028, + "step": 6860 + }, + { + "grad_norm": 0.14271092414855957, + "learning_rate": 2.4491023057783235e-05, + "loss": 0.0257, + "step": 6870 + }, + { + "grad_norm": 0.20435892045497894, + "learning_rate": 2.4348953162265375e-05, + "loss": 0.0253, + "step": 6880 + }, + { + "grad_norm": 0.14910104870796204, + "learning_rate": 2.420716378248607e-05, + "loss": 0.0244, + "step": 6890 + }, + { + "grad_norm": 0.17666257917881012, + "learning_rate": 2.4065656469031266e-05, + "loss": 0.0241, + "step": 6900 + }, + { + "grad_norm": 0.16489280760288239, + "learning_rate": 2.3924432769402268e-05, + "loss": 0.025, + "step": 6910 + }, + { + "grad_norm": 0.1911742240190506, + "learning_rate": 2.3783494227998844e-05, + "loss": 0.0309, + "step": 6920 + }, + { + "grad_norm": 0.20780859887599945, + "learning_rate": 2.3642842386102264e-05, + "loss": 0.0204, + "step": 6930 + }, + { + "grad_norm": 0.1516815423965454, + "learning_rate": 2.3502478781858567e-05, + "loss": 0.0219, + "step": 6940 + }, + { + "grad_norm": 0.1857873797416687, + "learning_rate": 2.3362404950261628e-05, + "loss": 0.0264, + "step": 6950 + }, + { + "grad_norm": 0.20639120042324066, + "learning_rate": 2.3222622423136458e-05, + "loss": 0.0223, + "step": 6960 + }, + { + "grad_norm": 0.206916943192482, + "learning_rate": 2.3083132729122332e-05, + "loss": 0.0223, + "step": 6970 + }, + { + "grad_norm": 0.20637910068035126, + "learning_rate": 2.294393739365621e-05, + "loss": 0.0228, + "step": 6980 + }, + { + "grad_norm": 0.21050766110420227, + "learning_rate": 2.2805037938956e-05, + "loss": 0.0243, + "step": 6990 + }, + { + "grad_norm": 0.1792493313550949, + "learning_rate": 2.266643588400386e-05, + "loss": 0.0233, + "step": 7000 + }, + { + "grad_norm": 0.1626683622598648, + "learning_rate": 2.252813274452969e-05, + "loss": 0.0251, + "step": 7010 + }, + { + "grad_norm": 0.19085238873958588, + "learning_rate": 2.2390130032994427e-05, + "loss": 0.0199, + "step": 7020 + }, + { + "grad_norm": 0.21508406102657318, + "learning_rate": 2.2252429258573633e-05, + "loss": 0.0244, + "step": 7030 + }, + { + "grad_norm": 0.20514574646949768, + "learning_rate": 2.2115031927140904e-05, + "loss": 0.0265, + "step": 7040 + }, + { + "grad_norm": 0.30970498919487, + "learning_rate": 2.1977939541251463e-05, + "loss": 0.0285, + "step": 7050 + }, + { + "grad_norm": 0.23108956217765808, + "learning_rate": 2.1841153600125684e-05, + "loss": 0.0225, + "step": 7060 + }, + { + "grad_norm": 0.17514687776565552, + "learning_rate": 2.170467559963267e-05, + "loss": 0.024, + "step": 7070 + }, + { + "grad_norm": 0.14796708524227142, + "learning_rate": 2.1568507032273982e-05, + "loss": 0.0229, + "step": 7080 + }, + { + "grad_norm": 0.17860688269138336, + "learning_rate": 2.1432649387167264e-05, + "loss": 0.0227, + "step": 7090 + }, + { + "grad_norm": 0.2551751434803009, + "learning_rate": 2.1297104150029973e-05, + "loss": 0.0236, + "step": 7100 + }, + { + "grad_norm": 0.243229478597641, + "learning_rate": 2.116187280316307e-05, + "loss": 0.0224, + "step": 7110 + }, + { + "grad_norm": 0.15843313932418823, + "learning_rate": 2.1026956825434908e-05, + "loss": 0.0233, + "step": 7120 + }, + { + "grad_norm": 0.16808034479618073, + "learning_rate": 2.0892357692265017e-05, + "loss": 0.0214, + "step": 7130 + }, + { + "grad_norm": 0.16679708659648895, + "learning_rate": 2.0758076875607947e-05, + "loss": 0.0211, + "step": 7140 + }, + { + "grad_norm": 0.2173215001821518, + "learning_rate": 2.0624115843937207e-05, + "loss": 0.0235, + "step": 7150 + }, + { + "grad_norm": 0.2244049459695816, + "learning_rate": 2.0490476062229157e-05, + "loss": 0.0242, + "step": 7160 + }, + { + "grad_norm": 0.1374778151512146, + "learning_rate": 2.035715899194704e-05, + "loss": 0.0245, + "step": 7170 + }, + { + "grad_norm": 0.18641945719718933, + "learning_rate": 2.022416609102499e-05, + "loss": 0.0233, + "step": 7180 + }, + { + "grad_norm": 0.17437012493610382, + "learning_rate": 2.009149881385205e-05, + "loss": 0.0293, + "step": 7190 + }, + { + "grad_norm": 0.1621212661266327, + "learning_rate": 1.995915861125634e-05, + "loss": 0.0213, + "step": 7200 + }, + { + "grad_norm": 0.25304773449897766, + "learning_rate": 1.9827146930489065e-05, + "loss": 0.02, + "step": 7210 + }, + { + "grad_norm": 0.1868961602449417, + "learning_rate": 1.9695465215208848e-05, + "loss": 0.0272, + "step": 7220 + }, + { + "grad_norm": 0.14957502484321594, + "learning_rate": 1.9564114905465813e-05, + "loss": 0.0231, + "step": 7230 + }, + { + "grad_norm": 0.12837262451648712, + "learning_rate": 1.9433097437685936e-05, + "loss": 0.0225, + "step": 7240 + }, + { + "grad_norm": 0.2177591770887375, + "learning_rate": 1.930241424465521e-05, + "loss": 0.0219, + "step": 7250 + }, + { + "grad_norm": 0.2072312831878662, + "learning_rate": 1.9172066755504115e-05, + "loss": 0.0217, + "step": 7260 + }, + { + "grad_norm": 0.16259168088436127, + "learning_rate": 1.9042056395691914e-05, + "loss": 0.0249, + "step": 7270 + }, + { + "grad_norm": 0.19199766218662262, + "learning_rate": 1.8912384586991066e-05, + "loss": 0.0202, + "step": 7280 + }, + { + "grad_norm": 0.24740499258041382, + "learning_rate": 1.8783052747471717e-05, + "loss": 0.0223, + "step": 7290 + }, + { + "grad_norm": 0.19696103036403656, + "learning_rate": 1.865406229148611e-05, + "loss": 0.0235, + "step": 7300 + }, + { + "grad_norm": 0.10264314711093903, + "learning_rate": 1.8525414629653233e-05, + "loss": 0.0248, + "step": 7310 + }, + { + "grad_norm": 0.19011767208576202, + "learning_rate": 1.8397111168843255e-05, + "loss": 0.024, + "step": 7320 + }, + { + "grad_norm": 0.15908634662628174, + "learning_rate": 1.8269153312162323e-05, + "loss": 0.0241, + "step": 7330 + }, + { + "grad_norm": 0.2369306981563568, + "learning_rate": 1.8141542458937054e-05, + "loss": 0.0235, + "step": 7340 + }, + { + "grad_norm": 0.2199007123708725, + "learning_rate": 1.8014280004699268e-05, + "loss": 0.0206, + "step": 7350 + }, + { + "grad_norm": 0.1905750334262848, + "learning_rate": 1.788736734117078e-05, + "loss": 0.0213, + "step": 7360 + }, + { + "grad_norm": 0.15303325653076172, + "learning_rate": 1.7760805856248152e-05, + "loss": 0.0237, + "step": 7370 + }, + { + "grad_norm": 0.2621123492717743, + "learning_rate": 1.7634596933987518e-05, + "loss": 0.024, + "step": 7380 + }, + { + "grad_norm": 0.14430855214595795, + "learning_rate": 1.7508741954589404e-05, + "loss": 0.024, + "step": 7390 + }, + { + "grad_norm": 0.15998873114585876, + "learning_rate": 1.7383242294383717e-05, + "loss": 0.0224, + "step": 7400 + }, + { + "grad_norm": 0.18560044467449188, + "learning_rate": 1.7258099325814632e-05, + "loss": 0.021, + "step": 7410 + }, + { + "grad_norm": 0.2068602740764618, + "learning_rate": 1.7133314417425594e-05, + "loss": 0.027, + "step": 7420 + }, + { + "grad_norm": 0.19259780645370483, + "learning_rate": 1.7008888933844408e-05, + "loss": 0.0233, + "step": 7430 + }, + { + "grad_norm": 0.20057493448257446, + "learning_rate": 1.6884824235768172e-05, + "loss": 0.0225, + "step": 7440 + }, + { + "grad_norm": 0.16601672768592834, + "learning_rate": 1.6761121679948592e-05, + "loss": 0.0197, + "step": 7450 + }, + { + "grad_norm": 0.15567000210285187, + "learning_rate": 1.663778261917695e-05, + "loss": 0.0236, + "step": 7460 + }, + { + "grad_norm": 0.22523567080497742, + "learning_rate": 1.651480840226952e-05, + "loss": 0.0225, + "step": 7470 + }, + { + "grad_norm": 0.25100064277648926, + "learning_rate": 1.639220037405258e-05, + "loss": 0.0242, + "step": 7480 + }, + { + "grad_norm": 0.2267504334449768, + "learning_rate": 1.6269959875347906e-05, + "loss": 0.0251, + "step": 7490 + }, + { + "grad_norm": 0.23382706940174103, + "learning_rate": 1.614808824295802e-05, + "loss": 0.021, + "step": 7500 + }, + { + "grad_norm": 0.1817302703857422, + "learning_rate": 1.602658680965152e-05, + "loss": 0.0207, + "step": 7510 + }, + { + "grad_norm": 0.1941787600517273, + "learning_rate": 1.5905456904148686e-05, + "loss": 0.0242, + "step": 7520 + }, + { + "grad_norm": 0.1802794188261032, + "learning_rate": 1.57846998511067e-05, + "loss": 0.0276, + "step": 7530 + }, + { + "grad_norm": 0.18045306205749512, + "learning_rate": 1.566431697110538e-05, + "loss": 0.0187, + "step": 7540 + }, + { + "grad_norm": 0.1808602511882782, + "learning_rate": 1.554430958063259e-05, + "loss": 0.0203, + "step": 7550 + }, + { + "grad_norm": 0.17564177513122559, + "learning_rate": 1.5424678992069912e-05, + "loss": 0.0198, + "step": 7560 + }, + { + "grad_norm": 0.25102999806404114, + "learning_rate": 1.5305426513678362e-05, + "loss": 0.0235, + "step": 7570 + }, + { + "grad_norm": 0.26221200823783875, + "learning_rate": 1.518655344958388e-05, + "loss": 0.0258, + "step": 7580 + }, + { + "grad_norm": 0.17686818540096283, + "learning_rate": 1.5068061099763275e-05, + "loss": 0.0216, + "step": 7590 + }, + { + "grad_norm": 0.16580519080162048, + "learning_rate": 1.494995076002988e-05, + "loss": 0.0216, + "step": 7600 + }, + { + "grad_norm": 0.1977156549692154, + "learning_rate": 1.4832223722019456e-05, + "loss": 0.0197, + "step": 7610 + }, + { + "grad_norm": 0.15229500830173492, + "learning_rate": 1.4714881273176035e-05, + "loss": 0.0203, + "step": 7620 + }, + { + "grad_norm": 0.19441953301429749, + "learning_rate": 1.4597924696737835e-05, + "loss": 0.0206, + "step": 7630 + }, + { + "grad_norm": 0.15327852964401245, + "learning_rate": 1.4481355271723252e-05, + "loss": 0.0205, + "step": 7640 + }, + { + "grad_norm": 0.1510356217622757, + "learning_rate": 1.4365174272916809e-05, + "loss": 0.0249, + "step": 7650 + }, + { + "grad_norm": 0.35722121596336365, + "learning_rate": 1.4249382970855319e-05, + "loss": 0.0209, + "step": 7660 + }, + { + "grad_norm": 0.17828625440597534, + "learning_rate": 1.4133982631813903e-05, + "loss": 0.021, + "step": 7670 + }, + { + "grad_norm": 0.14859116077423096, + "learning_rate": 1.4018974517792194e-05, + "loss": 0.0223, + "step": 7680 + }, + { + "grad_norm": 0.21912069618701935, + "learning_rate": 1.390435988650048e-05, + "loss": 0.0203, + "step": 7690 + }, + { + "grad_norm": 0.19409886002540588, + "learning_rate": 1.3790139991346006e-05, + "loss": 0.025, + "step": 7700 + }, + { + "grad_norm": 0.20902155339717865, + "learning_rate": 1.367631608141926e-05, + "loss": 0.0233, + "step": 7710 + }, + { + "grad_norm": 0.19436250627040863, + "learning_rate": 1.3562889401480278e-05, + "loss": 0.0169, + "step": 7720 + }, + { + "grad_norm": 0.13195233047008514, + "learning_rate": 1.3449861191945074e-05, + "loss": 0.0242, + "step": 7730 + }, + { + "grad_norm": 0.20837698876857758, + "learning_rate": 1.3337232688872009e-05, + "loss": 0.0206, + "step": 7740 + }, + { + "grad_norm": 0.24595220386981964, + "learning_rate": 1.3225005123948364e-05, + "loss": 0.0231, + "step": 7750 + }, + { + "grad_norm": 0.2161119133234024, + "learning_rate": 1.311317972447681e-05, + "loss": 0.0192, + "step": 7760 + }, + { + "grad_norm": 0.1619511842727661, + "learning_rate": 1.3001757713361996e-05, + "loss": 0.0197, + "step": 7770 + }, + { + "grad_norm": 0.2564446032047272, + "learning_rate": 1.2890740309097204e-05, + "loss": 0.0202, + "step": 7780 + }, + { + "grad_norm": 0.23809245228767395, + "learning_rate": 1.2780128725750944e-05, + "loss": 0.0204, + "step": 7790 + }, + { + "grad_norm": 0.19956621527671814, + "learning_rate": 1.266992417295379e-05, + "loss": 0.0206, + "step": 7800 + }, + { + "grad_norm": 0.16629883646965027, + "learning_rate": 1.2560127855885073e-05, + "loss": 0.0237, + "step": 7810 + }, + { + "grad_norm": 0.21347004175186157, + "learning_rate": 1.2450740975259745e-05, + "loss": 0.0244, + "step": 7820 + }, + { + "grad_norm": 0.15153144299983978, + "learning_rate": 1.234176472731517e-05, + "loss": 0.0289, + "step": 7830 + }, + { + "grad_norm": 0.28545379638671875, + "learning_rate": 1.2233200303798158e-05, + "loss": 0.0229, + "step": 7840 + }, + { + "grad_norm": 0.18485994637012482, + "learning_rate": 1.2125048891951846e-05, + "loss": 0.0218, + "step": 7850 + }, + { + "grad_norm": 0.17887774109840393, + "learning_rate": 1.2017311674502745e-05, + "loss": 0.0189, + "step": 7860 + }, + { + "grad_norm": 0.17393815517425537, + "learning_rate": 1.1909989829647822e-05, + "loss": 0.0196, + "step": 7870 + }, + { + "grad_norm": 0.18791592121124268, + "learning_rate": 1.1803084531041553e-05, + "loss": 0.0212, + "step": 7880 + }, + { + "grad_norm": 0.4066464602947235, + "learning_rate": 1.1696596947783162e-05, + "loss": 0.0217, + "step": 7890 + }, + { + "grad_norm": 0.21090605854988098, + "learning_rate": 1.1590528244403803e-05, + "loss": 0.0195, + "step": 7900 + }, + { + "grad_norm": 0.22359126806259155, + "learning_rate": 1.148487958085382e-05, + "loss": 0.0255, + "step": 7910 + }, + { + "grad_norm": 0.13211634755134583, + "learning_rate": 1.1379652112490086e-05, + "loss": 0.0202, + "step": 7920 + }, + { + "grad_norm": 0.22465965151786804, + "learning_rate": 1.1274846990063315e-05, + "loss": 0.0245, + "step": 7930 + }, + { + "grad_norm": 0.1914447695016861, + "learning_rate": 1.117046535970554e-05, + "loss": 0.0224, + "step": 7940 + }, + { + "grad_norm": 0.1983439177274704, + "learning_rate": 1.106650836291755e-05, + "loss": 0.0257, + "step": 7950 + }, + { + "grad_norm": 0.2157418131828308, + "learning_rate": 1.0962977136556418e-05, + "loss": 0.0269, + "step": 7960 + }, + { + "grad_norm": 0.23345190286636353, + "learning_rate": 1.0859872812823024e-05, + "loss": 0.0232, + "step": 7970 + }, + { + "grad_norm": 0.2735104262828827, + "learning_rate": 1.0757196519249747e-05, + "loss": 0.0286, + "step": 7980 + }, + { + "grad_norm": 0.14943794906139374, + "learning_rate": 1.0654949378688077e-05, + "loss": 0.0207, + "step": 7990 + }, + { + "grad_norm": 0.12805604934692383, + "learning_rate": 1.0553132509296376e-05, + "loss": 0.0203, + "step": 8000 + }, + { + "grad_norm": 0.14370910823345184, + "learning_rate": 1.0451747024527613e-05, + "loss": 0.0205, + "step": 8010 + }, + { + "grad_norm": 0.15158973634243011, + "learning_rate": 1.0350794033117189e-05, + "loss": 0.0209, + "step": 8020 + }, + { + "grad_norm": 0.20000551640987396, + "learning_rate": 1.0250274639070856e-05, + "loss": 0.0224, + "step": 8030 + }, + { + "grad_norm": 0.19087548553943634, + "learning_rate": 1.0150189941652599e-05, + "loss": 0.0182, + "step": 8040 + }, + { + "grad_norm": 0.15682381391525269, + "learning_rate": 1.0050541035372635e-05, + "loss": 0.0183, + "step": 8050 + }, + { + "grad_norm": 0.14305630326271057, + "learning_rate": 9.951329009975458e-06, + "loss": 0.0198, + "step": 8060 + }, + { + "grad_norm": 0.21597491204738617, + "learning_rate": 9.852554950427845e-06, + "loss": 0.0196, + "step": 8070 + }, + { + "grad_norm": 0.1693221926689148, + "learning_rate": 9.754219936907105e-06, + "loss": 0.0204, + "step": 8080 + }, + { + "grad_norm": 0.21354550123214722, + "learning_rate": 9.656325044789194e-06, + "loss": 0.0191, + "step": 8090 + }, + { + "grad_norm": 0.24106918275356293, + "learning_rate": 9.55887134463697e-06, + "loss": 0.0179, + "step": 8100 + }, + { + "grad_norm": 0.2606845498085022, + "learning_rate": 9.461859902188475e-06, + "loss": 0.02, + "step": 8110 + }, + { + "grad_norm": 0.16718357801437378, + "learning_rate": 9.365291778345303e-06, + "loss": 0.0245, + "step": 8120 + }, + { + "grad_norm": 0.12630660831928253, + "learning_rate": 9.269168029160991e-06, + "loss": 0.0188, + "step": 8130 + }, + { + "grad_norm": 0.186232328414917, + "learning_rate": 9.173489705829447e-06, + "loss": 0.021, + "step": 8140 + }, + { + "grad_norm": 0.20905108749866486, + "learning_rate": 9.078257854673516e-06, + "loss": 0.0222, + "step": 8150 + }, + { + "grad_norm": 0.09392578154802322, + "learning_rate": 8.983473517133429e-06, + "loss": 0.0214, + "step": 8160 + }, + { + "grad_norm": 0.1933511346578598, + "learning_rate": 8.889137729755537e-06, + "loss": 0.0209, + "step": 8170 + }, + { + "grad_norm": 0.12343988567590714, + "learning_rate": 8.79525152418087e-06, + "loss": 0.0212, + "step": 8180 + }, + { + "grad_norm": 0.12143982201814651, + "learning_rate": 8.701815927133961e-06, + "loss": 0.0206, + "step": 8190 + }, + { + "grad_norm": 0.1415945291519165, + "learning_rate": 8.608831960411534e-06, + "loss": 0.0185, + "step": 8200 + }, + { + "grad_norm": 0.15870840847492218, + "learning_rate": 8.516300640871321e-06, + "loss": 0.0186, + "step": 8210 + }, + { + "grad_norm": 0.33147114515304565, + "learning_rate": 8.424222980421038e-06, + "loss": 0.0204, + "step": 8220 + }, + { + "grad_norm": 0.1807570904493332, + "learning_rate": 8.332599986007184e-06, + "loss": 0.0212, + "step": 8230 + }, + { + "grad_norm": 0.11691354215145111, + "learning_rate": 8.241432659604203e-06, + "loss": 0.0166, + "step": 8240 + }, + { + "grad_norm": 0.12723970413208008, + "learning_rate": 8.150721998203331e-06, + "loss": 0.0173, + "step": 8250 + }, + { + "grad_norm": 0.13822627067565918, + "learning_rate": 8.06046899380184e-06, + "loss": 0.0161, + "step": 8260 + }, + { + "grad_norm": 0.28834235668182373, + "learning_rate": 7.970674633392133e-06, + "loss": 0.0246, + "step": 8270 + }, + { + "grad_norm": 0.23663489520549774, + "learning_rate": 7.881339898950924e-06, + "loss": 0.0183, + "step": 8280 + }, + { + "grad_norm": 0.17509469389915466, + "learning_rate": 7.792465767428597e-06, + "loss": 0.0204, + "step": 8290 + }, + { + "grad_norm": 0.13312990963459015, + "learning_rate": 7.704053210738376e-06, + "loss": 0.018, + "step": 8300 + }, + { + "grad_norm": 0.17818228900432587, + "learning_rate": 7.6161031957458494e-06, + "loss": 0.0246, + "step": 8310 + }, + { + "grad_norm": 0.17211821675300598, + "learning_rate": 7.5286166842582605e-06, + "loss": 0.0217, + "step": 8320 + }, + { + "grad_norm": 0.1497829705476761, + "learning_rate": 7.4415946330140814e-06, + "loss": 0.0214, + "step": 8330 + }, + { + "grad_norm": 0.19737355411052704, + "learning_rate": 7.3550379936725644e-06, + "loss": 0.0194, + "step": 8340 + }, + { + "grad_norm": 0.1559825837612152, + "learning_rate": 7.2689477128032035e-06, + "loss": 0.0203, + "step": 8350 + }, + { + "grad_norm": 0.21592740714550018, + "learning_rate": 7.183324731875551e-06, + "loss": 0.0184, + "step": 8360 + }, + { + "grad_norm": 0.1490253061056137, + "learning_rate": 7.098169987248782e-06, + "loss": 0.016, + "step": 8370 + }, + { + "grad_norm": 0.11406192183494568, + "learning_rate": 7.013484410161553e-06, + "loss": 0.0196, + "step": 8380 + }, + { + "grad_norm": 0.15339259803295135, + "learning_rate": 6.92926892672176e-06, + "loss": 0.0205, + "step": 8390 + }, + { + "grad_norm": 0.1071595773100853, + "learning_rate": 6.845524457896446e-06, + "loss": 0.0144, + "step": 8400 + }, + { + "grad_norm": 0.16246089339256287, + "learning_rate": 6.7622519195017165e-06, + "loss": 0.0194, + "step": 8410 + }, + { + "grad_norm": 0.15160700678825378, + "learning_rate": 6.679452222192684e-06, + "loss": 0.0207, + "step": 8420 + }, + { + "grad_norm": 0.4935346841812134, + "learning_rate": 6.597126271453579e-06, + "loss": 0.0185, + "step": 8430 + }, + { + "grad_norm": 0.21958622336387634, + "learning_rate": 6.51527496758782e-06, + "loss": 0.0164, + "step": 8440 + }, + { + "grad_norm": 0.1885659247636795, + "learning_rate": 6.433899205708155e-06, + "loss": 0.0179, + "step": 8450 + }, + { + "grad_norm": 0.15334294736385345, + "learning_rate": 6.352999875726856e-06, + "loss": 0.0149, + "step": 8460 + }, + { + "grad_norm": 0.16802644729614258, + "learning_rate": 6.272577862346052e-06, + "loss": 0.0196, + "step": 8470 + }, + { + "grad_norm": 0.21121948957443237, + "learning_rate": 6.192634045047996e-06, + "loss": 0.0226, + "step": 8480 + }, + { + "grad_norm": 0.15672186017036438, + "learning_rate": 6.113169298085458e-06, + "loss": 0.023, + "step": 8490 + }, + { + "grad_norm": 0.1584971398115158, + "learning_rate": 6.034184490472195e-06, + "loss": 0.0186, + "step": 8500 + }, + { + "grad_norm": 0.16784633696079254, + "learning_rate": 5.955680485973386e-06, + "loss": 0.0213, + "step": 8510 + }, + { + "grad_norm": 0.25108593702316284, + "learning_rate": 5.877658143096265e-06, + "loss": 0.0191, + "step": 8520 + }, + { + "grad_norm": 0.17944753170013428, + "learning_rate": 5.800118315080661e-06, + "loss": 0.0188, + "step": 8530 + }, + { + "grad_norm": 0.15853747725486755, + "learning_rate": 5.723061849889716e-06, + "loss": 0.0176, + "step": 8540 + }, + { + "grad_norm": 0.21179339289665222, + "learning_rate": 5.646489590200604e-06, + "loss": 0.0192, + "step": 8550 + }, + { + "grad_norm": 0.16561046242713928, + "learning_rate": 5.570402373395256e-06, + "loss": 0.0174, + "step": 8560 + }, + { + "grad_norm": 0.16707658767700195, + "learning_rate": 5.494801031551305e-06, + "loss": 0.0186, + "step": 8570 + }, + { + "grad_norm": 0.16819314658641815, + "learning_rate": 5.41968639143291e-06, + "loss": 0.0163, + "step": 8580 + }, + { + "grad_norm": 0.1450139284133911, + "learning_rate": 5.345059274481751e-06, + "loss": 0.0161, + "step": 8590 + }, + { + "grad_norm": 0.1834927350282669, + "learning_rate": 5.270920496808002e-06, + "loss": 0.0228, + "step": 8600 + }, + { + "grad_norm": 0.21522456407546997, + "learning_rate": 5.1972708691814695e-06, + "loss": 0.0199, + "step": 8610 + }, + { + "grad_norm": 0.1714378148317337, + "learning_rate": 5.124111197022674e-06, + "loss": 0.0207, + "step": 8620 + }, + { + "grad_norm": 0.13858215510845184, + "learning_rate": 5.051442280394081e-06, + "loss": 0.0182, + "step": 8630 + }, + { + "grad_norm": 0.1516513079404831, + "learning_rate": 4.979264913991322e-06, + "loss": 0.017, + "step": 8640 + }, + { + "grad_norm": 0.1680678427219391, + "learning_rate": 4.907579887134489e-06, + "loss": 0.0208, + "step": 8650 + }, + { + "grad_norm": 0.16671350598335266, + "learning_rate": 4.836387983759572e-06, + "loss": 0.0171, + "step": 8660 + }, + { + "grad_norm": 0.32937777042388916, + "learning_rate": 4.765689982409816e-06, + "loss": 0.0191, + "step": 8670 + }, + { + "grad_norm": 0.21558895707130432, + "learning_rate": 4.695486656227233e-06, + "loss": 0.0198, + "step": 8680 + }, + { + "grad_norm": 0.27218809723854065, + "learning_rate": 4.625778772944156e-06, + "loss": 0.0187, + "step": 8690 + }, + { + "grad_norm": 0.1461278349161148, + "learning_rate": 4.556567094874825e-06, + "loss": 0.0187, + "step": 8700 + }, + { + "grad_norm": 0.19476385414600372, + "learning_rate": 4.487852378907059e-06, + "loss": 0.0184, + "step": 8710 + }, + { + "grad_norm": 0.13335484266281128, + "learning_rate": 4.419635376493986e-06, + "loss": 0.0199, + "step": 8720 + }, + { + "grad_norm": 0.1359914392232895, + "learning_rate": 4.351916833645825e-06, + "loss": 0.0214, + "step": 8730 + }, + { + "grad_norm": 0.24102504551410675, + "learning_rate": 4.284697490921691e-06, + "loss": 0.0187, + "step": 8740 + }, + { + "grad_norm": 0.23733040690422058, + "learning_rate": 4.2179780834215585e-06, + "loss": 0.0197, + "step": 8750 + }, + { + "grad_norm": 0.21047444641590118, + "learning_rate": 4.151759340778178e-06, + "loss": 0.0231, + "step": 8760 + }, + { + "grad_norm": 0.15853050351142883, + "learning_rate": 4.086041987149109e-06, + "loss": 0.0179, + "step": 8770 + }, + { + "grad_norm": 0.2990325689315796, + "learning_rate": 4.020826741208811e-06, + "loss": 0.0218, + "step": 8780 + }, + { + "grad_norm": 0.1852436363697052, + "learning_rate": 3.956114316140746e-06, + "loss": 0.0221, + "step": 8790 + }, + { + "grad_norm": 0.2054145634174347, + "learning_rate": 3.891905419629643e-06, + "loss": 0.0208, + "step": 8800 + }, + { + "grad_norm": 0.18285734951496124, + "learning_rate": 3.8282007538536946e-06, + "loss": 0.0207, + "step": 8810 + }, + { + "grad_norm": 0.2325831651687622, + "learning_rate": 3.7650010154769265e-06, + "loss": 0.0208, + "step": 8820 + }, + { + "grad_norm": 0.20530511438846588, + "learning_rate": 3.7023068956415608e-06, + "loss": 0.0208, + "step": 8830 + }, + { + "grad_norm": 0.15905103087425232, + "learning_rate": 3.6401190799604303e-06, + "loss": 0.0184, + "step": 8840 + }, + { + "grad_norm": 0.2593521773815155, + "learning_rate": 3.578438248509536e-06, + "loss": 0.0198, + "step": 8850 + }, + { + "grad_norm": 0.19246645271778107, + "learning_rate": 3.5172650758205583e-06, + "loss": 0.0205, + "step": 8860 + }, + { + "grad_norm": 0.2052496373653412, + "learning_rate": 3.45660023087353e-06, + "loss": 0.017, + "step": 8870 + }, + { + "grad_norm": 0.1621578484773636, + "learning_rate": 3.3964443770894528e-06, + "loss": 0.0204, + "step": 8880 + }, + { + "grad_norm": 0.19428662955760956, + "learning_rate": 3.3367981723231245e-06, + "loss": 0.0162, + "step": 8890 + }, + { + "grad_norm": 0.25268009305000305, + "learning_rate": 3.2776622688558746e-06, + "loss": 0.0204, + "step": 8900 + }, + { + "grad_norm": 0.17166587710380554, + "learning_rate": 3.2190373133884677e-06, + "loss": 0.0174, + "step": 8910 + }, + { + "grad_norm": 0.24295078217983246, + "learning_rate": 3.1609239470340446e-06, + "loss": 0.0167, + "step": 8920 + }, + { + "grad_norm": 0.17703662812709808, + "learning_rate": 3.1033228053110373e-06, + "loss": 0.0168, + "step": 8930 + }, + { + "grad_norm": 0.14562709629535675, + "learning_rate": 3.0462345181363314e-06, + "loss": 0.0187, + "step": 8940 + }, + { + "grad_norm": 0.27955362200737, + "learning_rate": 2.9896597098182654e-06, + "loss": 0.0177, + "step": 8950 + }, + { + "grad_norm": 0.17375405132770538, + "learning_rate": 2.933598999049891e-06, + "loss": 0.0213, + "step": 8960 + }, + { + "grad_norm": 0.38965892791748047, + "learning_rate": 2.8780529989021697e-06, + "loss": 0.0172, + "step": 8970 + }, + { + "grad_norm": 0.1740860790014267, + "learning_rate": 2.823022316817242e-06, + "loss": 0.0206, + "step": 8980 + }, + { + "grad_norm": 0.27768465876579285, + "learning_rate": 2.7685075546018456e-06, + "loss": 0.0187, + "step": 8990 + }, + { + "grad_norm": 0.1763262301683426, + "learning_rate": 2.7145093084206598e-06, + "loss": 0.0217, + "step": 9000 + }, + { + "grad_norm": 0.25405117869377136, + "learning_rate": 2.661028168789892e-06, + "loss": 0.0161, + "step": 9010 + }, + { + "grad_norm": 0.7108403444290161, + "learning_rate": 2.6080647205706855e-06, + "loss": 0.0167, + "step": 9020 + }, + { + "grad_norm": 0.2510545551776886, + "learning_rate": 2.555619542962834e-06, + "loss": 0.0178, + "step": 9030 + }, + { + "grad_norm": 0.16971340775489807, + "learning_rate": 2.503693209498409e-06, + "loss": 0.0151, + "step": 9040 + }, + { + "grad_norm": 0.07835154980421066, + "learning_rate": 2.452286288035449e-06, + "loss": 0.0178, + "step": 9050 + }, + { + "grad_norm": 0.13570959866046906, + "learning_rate": 2.4013993407518363e-06, + "loss": 0.0191, + "step": 9060 + }, + { + "grad_norm": 0.24155135452747345, + "learning_rate": 2.351032924139063e-06, + "loss": 0.018, + "step": 9070 + }, + { + "grad_norm": 0.14278733730316162, + "learning_rate": 2.30118758899619e-06, + "loss": 0.0186, + "step": 9080 + }, + { + "grad_norm": 0.12462281435728073, + "learning_rate": 2.2518638804238157e-06, + "loss": 0.0203, + "step": 9090 + }, + { + "grad_norm": 0.15641923248767853, + "learning_rate": 2.203062337818118e-06, + "loss": 0.0176, + "step": 9100 + }, + { + "grad_norm": 0.12422550469636917, + "learning_rate": 2.1547834948649483e-06, + "loss": 0.0209, + "step": 9110 + }, + { + "grad_norm": 0.19502922892570496, + "learning_rate": 2.1070278795340017e-06, + "loss": 0.0204, + "step": 9120 + }, + { + "grad_norm": 0.14859560132026672, + "learning_rate": 2.059796014073029e-06, + "loss": 0.0135, + "step": 9130 + }, + { + "grad_norm": 0.1611856073141098, + "learning_rate": 2.01308841500214e-06, + "loss": 0.0199, + "step": 9140 + }, + { + "grad_norm": 0.14454419910907745, + "learning_rate": 1.9669055931081704e-06, + "loss": 0.0217, + "step": 9150 + }, + { + "grad_norm": 0.19837559759616852, + "learning_rate": 1.9212480534390507e-06, + "loss": 0.0181, + "step": 9160 + }, + { + "grad_norm": 0.2739919126033783, + "learning_rate": 1.8761162952983246e-06, + "loss": 0.0236, + "step": 9170 + }, + { + "grad_norm": 0.152645081281662, + "learning_rate": 1.8315108122396618e-06, + "loss": 0.0197, + "step": 9180 + }, + { + "grad_norm": 0.2050704061985016, + "learning_rate": 1.787432092061475e-06, + "loss": 0.0147, + "step": 9190 + }, + { + "grad_norm": 0.12005405873060226, + "learning_rate": 1.743880616801602e-06, + "loss": 0.0201, + "step": 9200 + }, + { + "grad_norm": 0.19026225805282593, + "learning_rate": 1.7008568627319865e-06, + "loss": 0.0191, + "step": 9210 + }, + { + "grad_norm": 0.20590677857398987, + "learning_rate": 1.6583613003535226e-06, + "loss": 0.0236, + "step": 9220 + }, + { + "grad_norm": 0.1922249048948288, + "learning_rate": 1.6163943943908522e-06, + "loss": 0.0204, + "step": 9230 + }, + { + "grad_norm": 0.15301360189914703, + "learning_rate": 1.5749566037873476e-06, + "loss": 0.0185, + "step": 9240 + }, + { + "grad_norm": 0.456008642911911, + "learning_rate": 1.5340483817000428e-06, + "loss": 0.0194, + "step": 9250 + }, + { + "grad_norm": 0.7787569165229797, + "learning_rate": 1.4936701754947101e-06, + "loss": 0.021, + "step": 9260 + }, + { + "grad_norm": 0.1533495932817459, + "learning_rate": 1.4538224267409361e-06, + "loss": 0.0184, + "step": 9270 + }, + { + "grad_norm": 0.160642609000206, + "learning_rate": 1.414505571207314e-06, + "loss": 0.0172, + "step": 9280 + }, + { + "grad_norm": 0.12171051651239395, + "learning_rate": 1.3757200388566816e-06, + "loss": 0.0176, + "step": 9290 + }, + { + "grad_norm": 0.14637814462184906, + "learning_rate": 1.3374662538414074e-06, + "loss": 0.0185, + "step": 9300 + }, + { + "grad_norm": 0.19669920206069946, + "learning_rate": 1.2997446344987617e-06, + "loss": 0.0197, + "step": 9310 + }, + { + "grad_norm": 0.20806103944778442, + "learning_rate": 1.262555593346315e-06, + "loss": 0.0214, + "step": 9320 + }, + { + "grad_norm": 0.20998172461986542, + "learning_rate": 1.2258995370774685e-06, + "loss": 0.0212, + "step": 9330 + }, + { + "grad_norm": 0.1567491590976715, + "learning_rate": 1.1897768665569798e-06, + "loss": 0.0182, + "step": 9340 + }, + { + "grad_norm": 0.13502928614616394, + "learning_rate": 1.1541879768165954e-06, + "loss": 0.0146, + "step": 9350 + }, + { + "grad_norm": 0.3173781633377075, + "learning_rate": 1.1191332570507085e-06, + "loss": 0.0217, + "step": 9360 + }, + { + "grad_norm": 0.15872761607170105, + "learning_rate": 1.0846130906121132e-06, + "loss": 0.0208, + "step": 9370 + }, + { + "grad_norm": 0.1684407889842987, + "learning_rate": 1.0506278550078131e-06, + "loss": 0.0176, + "step": 9380 + }, + { + "grad_norm": 0.20592917501926422, + "learning_rate": 1.0171779218949185e-06, + "loss": 0.0206, + "step": 9390 + }, + { + "grad_norm": 0.1955174207687378, + "learning_rate": 9.842636570765174e-07, + "loss": 0.0155, + "step": 9400 + }, + { + "grad_norm": 0.1864374428987503, + "learning_rate": 9.518854204977612e-07, + "loss": 0.018, + "step": 9410 + }, + { + "grad_norm": 0.155269056558609, + "learning_rate": 9.200435662418349e-07, + "loss": 0.0182, + "step": 9420 + }, + { + "grad_norm": 0.16822904348373413, + "learning_rate": 8.887384425261658e-07, + "loss": 0.0177, + "step": 9430 + }, + { + "grad_norm": 0.1613638550043106, + "learning_rate": 8.579703916985648e-07, + "loss": 0.0208, + "step": 9440 + }, + { + "grad_norm": 0.09026357531547546, + "learning_rate": 8.277397502335194e-07, + "loss": 0.0154, + "step": 9450 + }, + { + "grad_norm": 0.18123410642147064, + "learning_rate": 7.980468487284675e-07, + "loss": 0.0187, + "step": 9460 + }, + { + "grad_norm": 0.1521826535463333, + "learning_rate": 7.688920119002297e-07, + "loss": 0.0182, + "step": 9470 + }, + { + "grad_norm": 0.15285071730613708, + "learning_rate": 7.402755585814269e-07, + "loss": 0.0188, + "step": 9480 + }, + { + "grad_norm": 0.1563393771648407, + "learning_rate": 7.121978017170073e-07, + "loss": 0.0217, + "step": 9490 + }, + { + "grad_norm": 0.2133265733718872, + "learning_rate": 6.846590483608306e-07, + "loss": 0.0211, + "step": 9500 + }, + { + "grad_norm": 0.20891901850700378, + "learning_rate": 6.576595996722834e-07, + "loss": 0.0184, + "step": 9510 + }, + { + "grad_norm": 0.18856456875801086, + "learning_rate": 6.311997509130141e-07, + "loss": 0.0194, + "step": 9520 + }, + { + "grad_norm": 0.16375388205051422, + "learning_rate": 6.052797914436803e-07, + "loss": 0.0185, + "step": 9530 + }, + { + "grad_norm": 0.1643153876066208, + "learning_rate": 5.799000047208181e-07, + "loss": 0.0186, + "step": 9540 + }, + { + "grad_norm": 0.322010338306427, + "learning_rate": 5.550606682937054e-07, + "loss": 0.0209, + "step": 9550 + }, + { + "grad_norm": 0.16154605150222778, + "learning_rate": 5.307620538013481e-07, + "loss": 0.019, + "step": 9560 + }, + { + "grad_norm": 0.1823592334985733, + "learning_rate": 5.070044269694874e-07, + "loss": 0.0185, + "step": 9570 + }, + { + "grad_norm": 0.14982075989246368, + "learning_rate": 4.837880476077417e-07, + "loss": 0.0222, + "step": 9580 + }, + { + "grad_norm": 0.2312108278274536, + "learning_rate": 4.6111316960670835e-07, + "loss": 0.0173, + "step": 9590 + }, + { + "grad_norm": 0.11508405208587646, + "learning_rate": 4.389800409352218e-07, + "loss": 0.019, + "step": 9600 + }, + { + "grad_norm": 0.21661221981048584, + "learning_rate": 4.173889036376277e-07, + "loss": 0.0237, + "step": 9610 + }, + { + "grad_norm": 0.15002961456775665, + "learning_rate": 3.963399938311463e-07, + "loss": 0.0199, + "step": 9620 + }, + { + "grad_norm": 0.18903791904449463, + "learning_rate": 3.7583354170328545e-07, + "loss": 0.0209, + "step": 9630 + }, + { + "grad_norm": 0.10356654226779938, + "learning_rate": 3.558697715093207e-07, + "loss": 0.0196, + "step": 9640 + }, + { + "grad_norm": 0.3069995939731598, + "learning_rate": 3.3644890156983576e-07, + "loss": 0.0176, + "step": 9650 + }, + { + "grad_norm": 0.14816462993621826, + "learning_rate": 3.175711442683638e-07, + "loss": 0.0168, + "step": 9660 + }, + { + "grad_norm": 0.1555994600057602, + "learning_rate": 2.9923670604902197e-07, + "loss": 0.0201, + "step": 9670 + }, + { + "grad_norm": 0.1686890423297882, + "learning_rate": 2.814457874143028e-07, + "loss": 0.0179, + "step": 9680 + }, + { + "grad_norm": 0.13360834121704102, + "learning_rate": 2.641985829228366e-07, + "loss": 0.0152, + "step": 9690 + }, + { + "grad_norm": 0.10458117723464966, + "learning_rate": 2.474952811872877e-07, + "loss": 0.0177, + "step": 9700 + }, + { + "grad_norm": 0.1069650799036026, + "learning_rate": 2.3133606487228397e-07, + "loss": 0.0209, + "step": 9710 + }, + { + "grad_norm": 0.13212023675441742, + "learning_rate": 2.157211106924295e-07, + "loss": 0.0217, + "step": 9720 + }, + { + "grad_norm": 0.15600892901420593, + "learning_rate": 2.006505894103672e-07, + "loss": 0.0197, + "step": 9730 + }, + { + "grad_norm": 0.1441756784915924, + "learning_rate": 1.8612466583489696e-07, + "loss": 0.0197, + "step": 9740 + }, + { + "grad_norm": 0.1801835596561432, + "learning_rate": 1.7214349881918834e-07, + "loss": 0.0178, + "step": 9750 + }, + { + "grad_norm": 0.13342534005641937, + "learning_rate": 1.5870724125904845e-07, + "loss": 0.0193, + "step": 9760 + }, + { + "grad_norm": 0.15874849259853363, + "learning_rate": 1.4581604009124006e-07, + "loss": 0.016, + "step": 9770 + }, + { + "grad_norm": 0.14151382446289062, + "learning_rate": 1.334700362918717e-07, + "loss": 0.0179, + "step": 9780 + }, + { + "grad_norm": 0.15286359190940857, + "learning_rate": 1.2166936487486015e-07, + "loss": 0.0188, + "step": 9790 + }, + { + "grad_norm": 0.21923774480819702, + "learning_rate": 1.1041415489045914e-07, + "loss": 0.0163, + "step": 9800 + }, + { + "grad_norm": 0.16885928809642792, + "learning_rate": 9.970452942384412e-08, + "loss": 0.0191, + "step": 9810 + }, + { + "grad_norm": 0.22894100844860077, + "learning_rate": 8.954060559375754e-08, + "loss": 0.0173, + "step": 9820 + }, + { + "grad_norm": 0.1346498727798462, + "learning_rate": 7.99224945512489e-08, + "loss": 0.0206, + "step": 9830 + }, + { + "grad_norm": 0.5627322196960449, + "learning_rate": 7.085030147843675e-08, + "loss": 0.0216, + "step": 9840 + }, + { + "grad_norm": 0.22307898104190826, + "learning_rate": 6.232412558736523e-08, + "loss": 0.02, + "step": 9850 + }, + { + "grad_norm": 0.1825118213891983, + "learning_rate": 5.434406011893822e-08, + "loss": 0.0208, + "step": 9860 + }, + { + "grad_norm": 0.18380069732666016, + "learning_rate": 4.6910192341864664e-08, + "loss": 0.0163, + "step": 9870 + }, + { + "grad_norm": 0.19817106425762177, + "learning_rate": 4.0022603551737035e-08, + "loss": 0.0194, + "step": 9880 + }, + { + "grad_norm": 0.25512945652008057, + "learning_rate": 3.3681369070120985e-08, + "loss": 0.0182, + "step": 9890 + }, + { + "grad_norm": 0.17368295788764954, + "learning_rate": 2.7886558243744866e-08, + "loss": 0.0183, + "step": 9900 + }, + { + "grad_norm": 0.18219643831253052, + "learning_rate": 2.2638234443722596e-08, + "loss": 0.0192, + "step": 9910 + }, + { + "grad_norm": 0.13428278267383575, + "learning_rate": 1.7936455064887504e-08, + "loss": 0.0193, + "step": 9920 + }, + { + "grad_norm": 0.13242749869823456, + "learning_rate": 1.378127152514841e-08, + "loss": 0.015, + "step": 9930 + }, + { + "grad_norm": 0.17854952812194824, + "learning_rate": 1.0172729264917857e-08, + "loss": 0.0213, + "step": 9940 + }, + { + "grad_norm": 0.26226091384887695, + "learning_rate": 7.1108677466458215e-09, + "loss": 0.0215, + "step": 9950 + }, + { + "grad_norm": 0.1486857384443283, + "learning_rate": 4.595720454353414e-09, + "loss": 0.0164, + "step": 9960 + }, + { + "grad_norm": 0.17997819185256958, + "learning_rate": 2.627314893294264e-09, + "loss": 0.0177, + "step": 9970 + }, + { + "grad_norm": 0.1811828762292862, + "learning_rate": 1.2056725896270048e-09, + "loss": 0.0186, + "step": 9980 + }, + { + "grad_norm": 0.3201776444911957, + "learning_rate": 3.308090902098826e-10, + "loss": 0.0184, + "step": 9990 + }, + { + "grad_norm": 0.26741084456443787, + "learning_rate": 2.7339624120159555e-12, + "loss": 0.0198, + "step": 10000 + } + ], + "logging_steps": 10, + "max_steps": 10000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}