groot-6000-bidex-v4 / trainer_state.json
ArnieRamesh's picture
Upload folder using huggingface_hub
71e531b verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 19.54397394136808,
"eval_steps": 500,
"global_step": 6000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03257328990228013,
"grad_norm": 0.646706223487854,
"learning_rate": 9e-07,
"loss": 1.1814,
"step": 10
},
{
"epoch": 0.06514657980456026,
"grad_norm": 0.4603618383407593,
"learning_rate": 1.9e-06,
"loss": 1.1738,
"step": 20
},
{
"epoch": 0.09771986970684039,
"grad_norm": 0.4963187277317047,
"learning_rate": 2.9e-06,
"loss": 1.1665,
"step": 30
},
{
"epoch": 0.13029315960912052,
"grad_norm": 0.4428938329219818,
"learning_rate": 3.9e-06,
"loss": 1.1512,
"step": 40
},
{
"epoch": 0.16286644951140064,
"grad_norm": 0.36607691645622253,
"learning_rate": 4.9000000000000005e-06,
"loss": 1.1333,
"step": 50
},
{
"epoch": 0.19543973941368079,
"grad_norm": 0.30768465995788574,
"learning_rate": 5.9e-06,
"loss": 1.1202,
"step": 60
},
{
"epoch": 0.2280130293159609,
"grad_norm": 0.23890309035778046,
"learning_rate": 6.900000000000001e-06,
"loss": 1.1081,
"step": 70
},
{
"epoch": 0.26058631921824105,
"grad_norm": 0.25988394021987915,
"learning_rate": 7.9e-06,
"loss": 1.0964,
"step": 80
},
{
"epoch": 0.2931596091205212,
"grad_norm": 0.2948506772518158,
"learning_rate": 8.9e-06,
"loss": 1.0804,
"step": 90
},
{
"epoch": 0.3257328990228013,
"grad_norm": 0.29228347539901733,
"learning_rate": 9.900000000000002e-06,
"loss": 1.0646,
"step": 100
},
{
"epoch": 0.3583061889250814,
"grad_norm": 0.24975909292697906,
"learning_rate": 1.09e-05,
"loss": 1.0531,
"step": 110
},
{
"epoch": 0.39087947882736157,
"grad_norm": 0.30400097370147705,
"learning_rate": 1.19e-05,
"loss": 1.0387,
"step": 120
},
{
"epoch": 0.4234527687296417,
"grad_norm": 0.2392139881849289,
"learning_rate": 1.29e-05,
"loss": 1.0307,
"step": 130
},
{
"epoch": 0.4560260586319218,
"grad_norm": 0.2712036371231079,
"learning_rate": 1.3900000000000002e-05,
"loss": 1.0198,
"step": 140
},
{
"epoch": 0.48859934853420195,
"grad_norm": 0.23245830833911896,
"learning_rate": 1.49e-05,
"loss": 1.0154,
"step": 150
},
{
"epoch": 0.5211726384364821,
"grad_norm": 0.2727172374725342,
"learning_rate": 1.59e-05,
"loss": 1.0088,
"step": 160
},
{
"epoch": 0.5537459283387622,
"grad_norm": 0.30211466550827026,
"learning_rate": 1.69e-05,
"loss": 0.9998,
"step": 170
},
{
"epoch": 0.5863192182410424,
"grad_norm": 0.36023062467575073,
"learning_rate": 1.79e-05,
"loss": 0.9862,
"step": 180
},
{
"epoch": 0.6188925081433225,
"grad_norm": 0.6321305632591248,
"learning_rate": 1.8900000000000002e-05,
"loss": 0.9607,
"step": 190
},
{
"epoch": 0.6514657980456026,
"grad_norm": 0.7605375647544861,
"learning_rate": 1.9900000000000003e-05,
"loss": 0.9141,
"step": 200
},
{
"epoch": 0.6840390879478827,
"grad_norm": 1.6783995628356934,
"learning_rate": 2.09e-05,
"loss": 0.8218,
"step": 210
},
{
"epoch": 0.7166123778501629,
"grad_norm": 1.3917943239212036,
"learning_rate": 2.19e-05,
"loss": 0.726,
"step": 220
},
{
"epoch": 0.749185667752443,
"grad_norm": 1.6959741115570068,
"learning_rate": 2.29e-05,
"loss": 0.6331,
"step": 230
},
{
"epoch": 0.7817589576547231,
"grad_norm": 1.6433072090148926,
"learning_rate": 2.39e-05,
"loss": 0.5575,
"step": 240
},
{
"epoch": 0.8143322475570033,
"grad_norm": 2.0065226554870605,
"learning_rate": 2.4900000000000002e-05,
"loss": 0.4904,
"step": 250
},
{
"epoch": 0.8469055374592834,
"grad_norm": 1.7093048095703125,
"learning_rate": 2.5900000000000003e-05,
"loss": 0.4331,
"step": 260
},
{
"epoch": 0.8794788273615635,
"grad_norm": 2.0112454891204834,
"learning_rate": 2.6900000000000003e-05,
"loss": 0.385,
"step": 270
},
{
"epoch": 0.9120521172638436,
"grad_norm": 2.178093910217285,
"learning_rate": 2.7900000000000004e-05,
"loss": 0.3391,
"step": 280
},
{
"epoch": 0.9446254071661238,
"grad_norm": 2.12821888923645,
"learning_rate": 2.8899999999999998e-05,
"loss": 0.3148,
"step": 290
},
{
"epoch": 0.9771986970684039,
"grad_norm": 2.353641986846924,
"learning_rate": 2.9900000000000002e-05,
"loss": 0.2889,
"step": 300
},
{
"epoch": 1.009771986970684,
"grad_norm": 2.5022237300872803,
"learning_rate": 3.09e-05,
"loss": 0.2709,
"step": 310
},
{
"epoch": 1.0423452768729642,
"grad_norm": 2.0082218647003174,
"learning_rate": 3.19e-05,
"loss": 0.2614,
"step": 320
},
{
"epoch": 1.0749185667752443,
"grad_norm": 2.1790828704833984,
"learning_rate": 3.29e-05,
"loss": 0.2539,
"step": 330
},
{
"epoch": 1.1074918566775245,
"grad_norm": 2.4001519680023193,
"learning_rate": 3.3900000000000004e-05,
"loss": 0.2386,
"step": 340
},
{
"epoch": 1.1400651465798046,
"grad_norm": 2.5760984420776367,
"learning_rate": 3.49e-05,
"loss": 0.2327,
"step": 350
},
{
"epoch": 1.1726384364820848,
"grad_norm": 2.5301952362060547,
"learning_rate": 3.59e-05,
"loss": 0.2292,
"step": 360
},
{
"epoch": 1.205211726384365,
"grad_norm": 2.474808692932129,
"learning_rate": 3.69e-05,
"loss": 0.2293,
"step": 370
},
{
"epoch": 1.237785016286645,
"grad_norm": 2.524728298187256,
"learning_rate": 3.79e-05,
"loss": 0.2253,
"step": 380
},
{
"epoch": 1.2703583061889252,
"grad_norm": 2.349524974822998,
"learning_rate": 3.8900000000000004e-05,
"loss": 0.2259,
"step": 390
},
{
"epoch": 1.3029315960912053,
"grad_norm": 1.8449956178665161,
"learning_rate": 3.99e-05,
"loss": 0.2201,
"step": 400
},
{
"epoch": 1.3355048859934853,
"grad_norm": 2.82806134223938,
"learning_rate": 4.09e-05,
"loss": 0.2162,
"step": 410
},
{
"epoch": 1.3680781758957654,
"grad_norm": 2.695847988128662,
"learning_rate": 4.19e-05,
"loss": 0.2181,
"step": 420
},
{
"epoch": 1.4006514657980456,
"grad_norm": 2.953904628753662,
"learning_rate": 4.29e-05,
"loss": 0.2181,
"step": 430
},
{
"epoch": 1.4332247557003257,
"grad_norm": 2.7019174098968506,
"learning_rate": 4.39e-05,
"loss": 0.2166,
"step": 440
},
{
"epoch": 1.4657980456026058,
"grad_norm": 2.1432275772094727,
"learning_rate": 4.49e-05,
"loss": 0.2164,
"step": 450
},
{
"epoch": 1.498371335504886,
"grad_norm": 3.0038726329803467,
"learning_rate": 4.5900000000000004e-05,
"loss": 0.2136,
"step": 460
},
{
"epoch": 1.5309446254071661,
"grad_norm": 3.1258187294006348,
"learning_rate": 4.69e-05,
"loss": 0.2132,
"step": 470
},
{
"epoch": 1.5635179153094463,
"grad_norm": 2.207747459411621,
"learning_rate": 4.79e-05,
"loss": 0.2112,
"step": 480
},
{
"epoch": 1.5960912052117264,
"grad_norm": 2.581730604171753,
"learning_rate": 4.89e-05,
"loss": 0.2099,
"step": 490
},
{
"epoch": 1.6286644951140063,
"grad_norm": 3.099320411682129,
"learning_rate": 4.99e-05,
"loss": 0.2097,
"step": 500
},
{
"epoch": 1.6612377850162865,
"grad_norm": 2.6373846530914307,
"learning_rate": 5.0900000000000004e-05,
"loss": 0.2079,
"step": 510
},
{
"epoch": 1.6938110749185666,
"grad_norm": 2.9202771186828613,
"learning_rate": 5.19e-05,
"loss": 0.2043,
"step": 520
},
{
"epoch": 1.7263843648208468,
"grad_norm": 2.989102363586426,
"learning_rate": 5.2900000000000005e-05,
"loss": 0.2033,
"step": 530
},
{
"epoch": 1.758957654723127,
"grad_norm": 2.375894784927368,
"learning_rate": 5.390000000000001e-05,
"loss": 0.2059,
"step": 540
},
{
"epoch": 1.791530944625407,
"grad_norm": 2.6863505840301514,
"learning_rate": 5.4900000000000006e-05,
"loss": 0.2032,
"step": 550
},
{
"epoch": 1.8241042345276872,
"grad_norm": 2.326268434524536,
"learning_rate": 5.590000000000001e-05,
"loss": 0.2012,
"step": 560
},
{
"epoch": 1.8566775244299674,
"grad_norm": 2.779634952545166,
"learning_rate": 5.69e-05,
"loss": 0.2027,
"step": 570
},
{
"epoch": 1.8892508143322475,
"grad_norm": 2.6916491985321045,
"learning_rate": 5.79e-05,
"loss": 0.1983,
"step": 580
},
{
"epoch": 1.9218241042345277,
"grad_norm": 2.6536519527435303,
"learning_rate": 5.89e-05,
"loss": 0.2007,
"step": 590
},
{
"epoch": 1.9543973941368078,
"grad_norm": 2.6414382457733154,
"learning_rate": 5.99e-05,
"loss": 0.1943,
"step": 600
},
{
"epoch": 1.986970684039088,
"grad_norm": 3.05230712890625,
"learning_rate": 6.09e-05,
"loss": 0.1898,
"step": 610
},
{
"epoch": 2.019543973941368,
"grad_norm": 3.180234909057617,
"learning_rate": 6.19e-05,
"loss": 0.1873,
"step": 620
},
{
"epoch": 2.0521172638436482,
"grad_norm": 2.8422200679779053,
"learning_rate": 6.29e-05,
"loss": 0.1811,
"step": 630
},
{
"epoch": 2.0846905537459284,
"grad_norm": 3.6337952613830566,
"learning_rate": 6.390000000000001e-05,
"loss": 0.182,
"step": 640
},
{
"epoch": 2.1172638436482085,
"grad_norm": 5.311783313751221,
"learning_rate": 6.49e-05,
"loss": 0.1789,
"step": 650
},
{
"epoch": 2.1498371335504887,
"grad_norm": 2.879094123840332,
"learning_rate": 6.59e-05,
"loss": 0.1757,
"step": 660
},
{
"epoch": 2.182410423452769,
"grad_norm": 2.7306885719299316,
"learning_rate": 6.690000000000001e-05,
"loss": 0.1756,
"step": 670
},
{
"epoch": 2.214983713355049,
"grad_norm": 2.2274434566497803,
"learning_rate": 6.790000000000001e-05,
"loss": 0.1696,
"step": 680
},
{
"epoch": 2.247557003257329,
"grad_norm": 2.76662540435791,
"learning_rate": 6.89e-05,
"loss": 0.1633,
"step": 690
},
{
"epoch": 2.2801302931596092,
"grad_norm": 2.6583666801452637,
"learning_rate": 6.99e-05,
"loss": 0.1603,
"step": 700
},
{
"epoch": 2.3127035830618894,
"grad_norm": 2.7752110958099365,
"learning_rate": 7.09e-05,
"loss": 0.1555,
"step": 710
},
{
"epoch": 2.3452768729641695,
"grad_norm": 2.919973611831665,
"learning_rate": 7.19e-05,
"loss": 0.1509,
"step": 720
},
{
"epoch": 2.3778501628664497,
"grad_norm": 2.369131565093994,
"learning_rate": 7.29e-05,
"loss": 0.1453,
"step": 730
},
{
"epoch": 2.41042345276873,
"grad_norm": 6.763181686401367,
"learning_rate": 7.390000000000001e-05,
"loss": 0.1469,
"step": 740
},
{
"epoch": 2.44299674267101,
"grad_norm": 5.001865386962891,
"learning_rate": 7.49e-05,
"loss": 0.1568,
"step": 750
},
{
"epoch": 2.47557003257329,
"grad_norm": 3.4782662391662598,
"learning_rate": 7.59e-05,
"loss": 0.1372,
"step": 760
},
{
"epoch": 2.5081433224755703,
"grad_norm": 5.221865177154541,
"learning_rate": 7.69e-05,
"loss": 0.1373,
"step": 770
},
{
"epoch": 2.5407166123778504,
"grad_norm": 3.3979134559631348,
"learning_rate": 7.790000000000001e-05,
"loss": 0.131,
"step": 780
},
{
"epoch": 2.5732899022801305,
"grad_norm": 3.932870626449585,
"learning_rate": 7.890000000000001e-05,
"loss": 0.1176,
"step": 790
},
{
"epoch": 2.6058631921824107,
"grad_norm": 4.280229091644287,
"learning_rate": 7.99e-05,
"loss": 0.1133,
"step": 800
},
{
"epoch": 2.6384364820846904,
"grad_norm": 2.9509708881378174,
"learning_rate": 8.090000000000001e-05,
"loss": 0.1046,
"step": 810
},
{
"epoch": 2.6710097719869705,
"grad_norm": 2.813812255859375,
"learning_rate": 8.19e-05,
"loss": 0.0959,
"step": 820
},
{
"epoch": 2.7035830618892507,
"grad_norm": 3.498904228210449,
"learning_rate": 8.29e-05,
"loss": 0.0842,
"step": 830
},
{
"epoch": 2.736156351791531,
"grad_norm": 2.793267250061035,
"learning_rate": 8.39e-05,
"loss": 0.0821,
"step": 840
},
{
"epoch": 2.768729641693811,
"grad_norm": 3.1398348808288574,
"learning_rate": 8.49e-05,
"loss": 0.0783,
"step": 850
},
{
"epoch": 2.801302931596091,
"grad_norm": 3.613935947418213,
"learning_rate": 8.59e-05,
"loss": 0.0679,
"step": 860
},
{
"epoch": 2.8338762214983713,
"grad_norm": 4.147997856140137,
"learning_rate": 8.69e-05,
"loss": 0.0601,
"step": 870
},
{
"epoch": 2.8664495114006514,
"grad_norm": 4.906704425811768,
"learning_rate": 8.790000000000001e-05,
"loss": 0.0589,
"step": 880
},
{
"epoch": 2.8990228013029316,
"grad_norm": 4.920569896697998,
"learning_rate": 8.89e-05,
"loss": 0.0602,
"step": 890
},
{
"epoch": 2.9315960912052117,
"grad_norm": 3.0054733753204346,
"learning_rate": 8.99e-05,
"loss": 0.0567,
"step": 900
},
{
"epoch": 2.964169381107492,
"grad_norm": 3.0714895725250244,
"learning_rate": 9.090000000000001e-05,
"loss": 0.0478,
"step": 910
},
{
"epoch": 2.996742671009772,
"grad_norm": 2.0885047912597656,
"learning_rate": 9.190000000000001e-05,
"loss": 0.0417,
"step": 920
},
{
"epoch": 3.029315960912052,
"grad_norm": 3.2938196659088135,
"learning_rate": 9.290000000000001e-05,
"loss": 0.0504,
"step": 930
},
{
"epoch": 3.0618892508143323,
"grad_norm": 3.084390640258789,
"learning_rate": 9.39e-05,
"loss": 0.0482,
"step": 940
},
{
"epoch": 3.0944625407166124,
"grad_norm": 3.5595154762268066,
"learning_rate": 9.49e-05,
"loss": 0.0464,
"step": 950
},
{
"epoch": 3.1270358306188926,
"grad_norm": 3.5651934146881104,
"learning_rate": 9.59e-05,
"loss": 0.0461,
"step": 960
},
{
"epoch": 3.1596091205211727,
"grad_norm": 2.7966909408569336,
"learning_rate": 9.69e-05,
"loss": 0.0455,
"step": 970
},
{
"epoch": 3.192182410423453,
"grad_norm": 2.4500396251678467,
"learning_rate": 9.790000000000001e-05,
"loss": 0.0395,
"step": 980
},
{
"epoch": 3.224755700325733,
"grad_norm": 3.2736151218414307,
"learning_rate": 9.89e-05,
"loss": 0.0403,
"step": 990
},
{
"epoch": 3.257328990228013,
"grad_norm": 2.5784664154052734,
"learning_rate": 9.99e-05,
"loss": 0.0382,
"step": 1000
},
{
"epoch": 3.2899022801302933,
"grad_norm": 2.018731117248535,
"learning_rate": 9.999994463727085e-05,
"loss": 0.0376,
"step": 1010
},
{
"epoch": 3.3224755700325734,
"grad_norm": 2.5587472915649414,
"learning_rate": 9.999975326009292e-05,
"loss": 0.0385,
"step": 1020
},
{
"epoch": 3.3550488599348536,
"grad_norm": 2.2966997623443604,
"learning_rate": 9.999942518549879e-05,
"loss": 0.0364,
"step": 1030
},
{
"epoch": 3.3876221498371337,
"grad_norm": 2.336106061935425,
"learning_rate": 9.999896041438544e-05,
"loss": 0.036,
"step": 1040
},
{
"epoch": 3.420195439739414,
"grad_norm": 3.2040939331054688,
"learning_rate": 9.999835894802353e-05,
"loss": 0.0393,
"step": 1050
},
{
"epoch": 3.4527687296416936,
"grad_norm": 2.786553382873535,
"learning_rate": 9.999762078805743e-05,
"loss": 0.0345,
"step": 1060
},
{
"epoch": 3.4853420195439737,
"grad_norm": 1.6308751106262207,
"learning_rate": 9.999674593650526e-05,
"loss": 0.037,
"step": 1070
},
{
"epoch": 3.517915309446254,
"grad_norm": 1.8265235424041748,
"learning_rate": 9.99957343957588e-05,
"loss": 0.0344,
"step": 1080
},
{
"epoch": 3.550488599348534,
"grad_norm": 2.4011759757995605,
"learning_rate": 9.99945861685836e-05,
"loss": 0.0304,
"step": 1090
},
{
"epoch": 3.583061889250814,
"grad_norm": 2.0200209617614746,
"learning_rate": 9.999330125811884e-05,
"loss": 0.0328,
"step": 1100
},
{
"epoch": 3.6156351791530943,
"grad_norm": 2.044693946838379,
"learning_rate": 9.999187966787744e-05,
"loss": 0.0355,
"step": 1110
},
{
"epoch": 3.6482084690553744,
"grad_norm": 2.1637792587280273,
"learning_rate": 9.999032140174595e-05,
"loss": 0.0342,
"step": 1120
},
{
"epoch": 3.6807817589576546,
"grad_norm": 1.9781519174575806,
"learning_rate": 9.998862646398464e-05,
"loss": 0.0349,
"step": 1130
},
{
"epoch": 3.7133550488599347,
"grad_norm": 2.601254940032959,
"learning_rate": 9.998679485922739e-05,
"loss": 0.0311,
"step": 1140
},
{
"epoch": 3.745928338762215,
"grad_norm": 1.977401614189148,
"learning_rate": 9.998482659248174e-05,
"loss": 0.0317,
"step": 1150
},
{
"epoch": 3.778501628664495,
"grad_norm": 1.9435837268829346,
"learning_rate": 9.998272166912883e-05,
"loss": 0.0325,
"step": 1160
},
{
"epoch": 3.811074918566775,
"grad_norm": 2.054821014404297,
"learning_rate": 9.998048009492347e-05,
"loss": 0.0352,
"step": 1170
},
{
"epoch": 3.8436482084690553,
"grad_norm": 1.950892448425293,
"learning_rate": 9.997810187599403e-05,
"loss": 0.0325,
"step": 1180
},
{
"epoch": 3.8762214983713354,
"grad_norm": 2.4616150856018066,
"learning_rate": 9.997558701884249e-05,
"loss": 0.0316,
"step": 1190
},
{
"epoch": 3.9087947882736156,
"grad_norm": 1.8379050493240356,
"learning_rate": 9.997293553034433e-05,
"loss": 0.0307,
"step": 1200
},
{
"epoch": 3.9413680781758957,
"grad_norm": 1.9157487154006958,
"learning_rate": 9.997014741774866e-05,
"loss": 0.0319,
"step": 1210
},
{
"epoch": 3.973941368078176,
"grad_norm": 1.8840038776397705,
"learning_rate": 9.996722268867803e-05,
"loss": 0.0315,
"step": 1220
},
{
"epoch": 4.006514657980456,
"grad_norm": 1.825744867324829,
"learning_rate": 9.996416135112858e-05,
"loss": 0.0297,
"step": 1230
},
{
"epoch": 4.039087947882736,
"grad_norm": 1.5622966289520264,
"learning_rate": 9.996096341346988e-05,
"loss": 0.0291,
"step": 1240
},
{
"epoch": 4.071661237785016,
"grad_norm": 1.652907371520996,
"learning_rate": 9.995762888444495e-05,
"loss": 0.0298,
"step": 1250
},
{
"epoch": 4.1042345276872965,
"grad_norm": 2.1362810134887695,
"learning_rate": 9.995415777317027e-05,
"loss": 0.0304,
"step": 1260
},
{
"epoch": 4.136807817589577,
"grad_norm": 1.7935142517089844,
"learning_rate": 9.995055008913574e-05,
"loss": 0.028,
"step": 1270
},
{
"epoch": 4.169381107491857,
"grad_norm": 2.094823122024536,
"learning_rate": 9.994680584220463e-05,
"loss": 0.0294,
"step": 1280
},
{
"epoch": 4.201954397394137,
"grad_norm": 1.5280042886734009,
"learning_rate": 9.994292504261355e-05,
"loss": 0.0289,
"step": 1290
},
{
"epoch": 4.234527687296417,
"grad_norm": 1.7017203569412231,
"learning_rate": 9.993890770097247e-05,
"loss": 0.0267,
"step": 1300
},
{
"epoch": 4.267100977198697,
"grad_norm": 1.8569612503051758,
"learning_rate": 9.993475382826467e-05,
"loss": 0.0282,
"step": 1310
},
{
"epoch": 4.299674267100977,
"grad_norm": 1.792932152748108,
"learning_rate": 9.993046343584664e-05,
"loss": 0.0278,
"step": 1320
},
{
"epoch": 4.3322475570032575,
"grad_norm": 1.6156980991363525,
"learning_rate": 9.992603653544816e-05,
"loss": 0.0276,
"step": 1330
},
{
"epoch": 4.364820846905538,
"grad_norm": 1.6737850904464722,
"learning_rate": 9.992147313917222e-05,
"loss": 0.0276,
"step": 1340
},
{
"epoch": 4.397394136807818,
"grad_norm": 1.6165825128555298,
"learning_rate": 9.991677325949497e-05,
"loss": 0.0277,
"step": 1350
},
{
"epoch": 4.429967426710098,
"grad_norm": 1.5214238166809082,
"learning_rate": 9.991193690926568e-05,
"loss": 0.0258,
"step": 1360
},
{
"epoch": 4.462540716612378,
"grad_norm": 1.648809790611267,
"learning_rate": 9.990696410170678e-05,
"loss": 0.0274,
"step": 1370
},
{
"epoch": 4.495114006514658,
"grad_norm": 1.5280410051345825,
"learning_rate": 9.990185485041371e-05,
"loss": 0.0247,
"step": 1380
},
{
"epoch": 4.527687296416938,
"grad_norm": 1.4952828884124756,
"learning_rate": 9.989660916935498e-05,
"loss": 0.0251,
"step": 1390
},
{
"epoch": 4.5602605863192185,
"grad_norm": 1.3932770490646362,
"learning_rate": 9.989122707287208e-05,
"loss": 0.0256,
"step": 1400
},
{
"epoch": 4.592833876221499,
"grad_norm": 1.6718034744262695,
"learning_rate": 9.988570857567945e-05,
"loss": 0.0272,
"step": 1410
},
{
"epoch": 4.625407166123779,
"grad_norm": 1.5743407011032104,
"learning_rate": 9.988005369286446e-05,
"loss": 0.0261,
"step": 1420
},
{
"epoch": 4.657980456026059,
"grad_norm": 1.6346157789230347,
"learning_rate": 9.987426243988734e-05,
"loss": 0.0276,
"step": 1430
},
{
"epoch": 4.690553745928339,
"grad_norm": 1.754012942314148,
"learning_rate": 9.986833483258114e-05,
"loss": 0.026,
"step": 1440
},
{
"epoch": 4.723127035830619,
"grad_norm": 1.5971155166625977,
"learning_rate": 9.986227088715173e-05,
"loss": 0.0258,
"step": 1450
},
{
"epoch": 4.755700325732899,
"grad_norm": 1.5895884037017822,
"learning_rate": 9.98560706201777e-05,
"loss": 0.0255,
"step": 1460
},
{
"epoch": 4.7882736156351795,
"grad_norm": 1.799141764640808,
"learning_rate": 9.984973404861036e-05,
"loss": 0.028,
"step": 1470
},
{
"epoch": 4.82084690553746,
"grad_norm": 1.5767462253570557,
"learning_rate": 9.984326118977361e-05,
"loss": 0.0277,
"step": 1480
},
{
"epoch": 4.85342019543974,
"grad_norm": 1.547109842300415,
"learning_rate": 9.983665206136406e-05,
"loss": 0.0239,
"step": 1490
},
{
"epoch": 4.88599348534202,
"grad_norm": 1.6011266708374023,
"learning_rate": 9.982990668145075e-05,
"loss": 0.023,
"step": 1500
},
{
"epoch": 4.918566775244299,
"grad_norm": 1.5277559757232666,
"learning_rate": 9.982302506847534e-05,
"loss": 0.0232,
"step": 1510
},
{
"epoch": 4.95114006514658,
"grad_norm": 1.3489938974380493,
"learning_rate": 9.981600724125189e-05,
"loss": 0.0231,
"step": 1520
},
{
"epoch": 4.9837133550488595,
"grad_norm": 1.2097440958023071,
"learning_rate": 9.980885321896685e-05,
"loss": 0.0226,
"step": 1530
},
{
"epoch": 5.01628664495114,
"grad_norm": 1.5271075963974,
"learning_rate": 9.980156302117905e-05,
"loss": 0.0217,
"step": 1540
},
{
"epoch": 5.04885993485342,
"grad_norm": 1.4403456449508667,
"learning_rate": 9.979413666781963e-05,
"loss": 0.0259,
"step": 1550
},
{
"epoch": 5.0814332247557,
"grad_norm": 1.3812097311019897,
"learning_rate": 9.978657417919193e-05,
"loss": 0.023,
"step": 1560
},
{
"epoch": 5.11400651465798,
"grad_norm": 1.5333428382873535,
"learning_rate": 9.977887557597153e-05,
"loss": 0.0242,
"step": 1570
},
{
"epoch": 5.14657980456026,
"grad_norm": 1.1747639179229736,
"learning_rate": 9.97710408792061e-05,
"loss": 0.0232,
"step": 1580
},
{
"epoch": 5.17915309446254,
"grad_norm": 1.4856892824172974,
"learning_rate": 9.976307011031542e-05,
"loss": 0.0207,
"step": 1590
},
{
"epoch": 5.2117263843648205,
"grad_norm": 1.1282317638397217,
"learning_rate": 9.975496329109126e-05,
"loss": 0.0233,
"step": 1600
},
{
"epoch": 5.244299674267101,
"grad_norm": 1.4065935611724854,
"learning_rate": 9.974672044369732e-05,
"loss": 0.0218,
"step": 1610
},
{
"epoch": 5.276872964169381,
"grad_norm": 1.1849043369293213,
"learning_rate": 9.97383415906693e-05,
"loss": 0.0221,
"step": 1620
},
{
"epoch": 5.309446254071661,
"grad_norm": 1.325810194015503,
"learning_rate": 9.97298267549146e-05,
"loss": 0.0213,
"step": 1630
},
{
"epoch": 5.342019543973941,
"grad_norm": 1.113312840461731,
"learning_rate": 9.972117595971249e-05,
"loss": 0.0228,
"step": 1640
},
{
"epoch": 5.374592833876221,
"grad_norm": 1.230494499206543,
"learning_rate": 9.971238922871391e-05,
"loss": 0.0253,
"step": 1650
},
{
"epoch": 5.407166123778501,
"grad_norm": 1.4597903490066528,
"learning_rate": 9.970346658594142e-05,
"loss": 0.0224,
"step": 1660
},
{
"epoch": 5.4397394136807815,
"grad_norm": 1.2914201021194458,
"learning_rate": 9.969440805578923e-05,
"loss": 0.0224,
"step": 1670
},
{
"epoch": 5.472312703583062,
"grad_norm": 1.2426233291625977,
"learning_rate": 9.968521366302298e-05,
"loss": 0.0233,
"step": 1680
},
{
"epoch": 5.504885993485342,
"grad_norm": 1.2101874351501465,
"learning_rate": 9.967588343277981e-05,
"loss": 0.0233,
"step": 1690
},
{
"epoch": 5.537459283387622,
"grad_norm": 1.576130747795105,
"learning_rate": 9.966641739056818e-05,
"loss": 0.0204,
"step": 1700
},
{
"epoch": 5.570032573289902,
"grad_norm": 1.4965754747390747,
"learning_rate": 9.965681556226793e-05,
"loss": 0.021,
"step": 1710
},
{
"epoch": 5.602605863192182,
"grad_norm": 1.2106349468231201,
"learning_rate": 9.964707797413006e-05,
"loss": 0.0217,
"step": 1720
},
{
"epoch": 5.635179153094462,
"grad_norm": 1.2536280155181885,
"learning_rate": 9.963720465277679e-05,
"loss": 0.0229,
"step": 1730
},
{
"epoch": 5.6677524429967425,
"grad_norm": 1.2137837409973145,
"learning_rate": 9.96271956252014e-05,
"loss": 0.0227,
"step": 1740
},
{
"epoch": 5.700325732899023,
"grad_norm": 1.4352983236312866,
"learning_rate": 9.961705091876816e-05,
"loss": 0.0217,
"step": 1750
},
{
"epoch": 5.732899022801303,
"grad_norm": 1.0887129306793213,
"learning_rate": 9.960677056121235e-05,
"loss": 0.02,
"step": 1760
},
{
"epoch": 5.765472312703583,
"grad_norm": 1.2834426164627075,
"learning_rate": 9.959635458064005e-05,
"loss": 0.0232,
"step": 1770
},
{
"epoch": 5.798045602605863,
"grad_norm": 1.3449937105178833,
"learning_rate": 9.958580300552815e-05,
"loss": 0.0205,
"step": 1780
},
{
"epoch": 5.830618892508143,
"grad_norm": 1.375016689300537,
"learning_rate": 9.957511586472426e-05,
"loss": 0.0231,
"step": 1790
},
{
"epoch": 5.863192182410423,
"grad_norm": 1.2283987998962402,
"learning_rate": 9.956429318744662e-05,
"loss": 0.0229,
"step": 1800
},
{
"epoch": 5.8957654723127035,
"grad_norm": 1.226679801940918,
"learning_rate": 9.955333500328404e-05,
"loss": 0.0201,
"step": 1810
},
{
"epoch": 5.928338762214984,
"grad_norm": 1.6469484567642212,
"learning_rate": 9.95422413421957e-05,
"loss": 0.0211,
"step": 1820
},
{
"epoch": 5.960912052117264,
"grad_norm": 1.1678240299224854,
"learning_rate": 9.953101223451133e-05,
"loss": 0.0214,
"step": 1830
},
{
"epoch": 5.993485342019544,
"grad_norm": 1.2131609916687012,
"learning_rate": 9.951964771093085e-05,
"loss": 0.0202,
"step": 1840
},
{
"epoch": 6.026058631921824,
"grad_norm": 1.3158583641052246,
"learning_rate": 9.950814780252442e-05,
"loss": 0.0204,
"step": 1850
},
{
"epoch": 6.058631921824104,
"grad_norm": 1.2186533212661743,
"learning_rate": 9.949651254073236e-05,
"loss": 0.0192,
"step": 1860
},
{
"epoch": 6.091205211726384,
"grad_norm": 1.2744988203048706,
"learning_rate": 9.948474195736504e-05,
"loss": 0.0209,
"step": 1870
},
{
"epoch": 6.1237785016286646,
"grad_norm": 1.1233289241790771,
"learning_rate": 9.947283608460277e-05,
"loss": 0.02,
"step": 1880
},
{
"epoch": 6.156351791530945,
"grad_norm": 1.092411994934082,
"learning_rate": 9.946079495499577e-05,
"loss": 0.0202,
"step": 1890
},
{
"epoch": 6.188925081433225,
"grad_norm": 1.3971725702285767,
"learning_rate": 9.944861860146401e-05,
"loss": 0.021,
"step": 1900
},
{
"epoch": 6.221498371335505,
"grad_norm": 1.7205326557159424,
"learning_rate": 9.943630705729719e-05,
"loss": 0.0205,
"step": 1910
},
{
"epoch": 6.254071661237785,
"grad_norm": 1.2655634880065918,
"learning_rate": 9.942386035615459e-05,
"loss": 0.0222,
"step": 1920
},
{
"epoch": 6.286644951140065,
"grad_norm": 0.9892485737800598,
"learning_rate": 9.941127853206503e-05,
"loss": 0.0191,
"step": 1930
},
{
"epoch": 6.319218241042345,
"grad_norm": 1.2320622205734253,
"learning_rate": 9.939856161942673e-05,
"loss": 0.0196,
"step": 1940
},
{
"epoch": 6.351791530944626,
"grad_norm": 1.315510630607605,
"learning_rate": 9.938570965300724e-05,
"loss": 0.0184,
"step": 1950
},
{
"epoch": 6.384364820846906,
"grad_norm": 0.9320159554481506,
"learning_rate": 9.937272266794335e-05,
"loss": 0.0231,
"step": 1960
},
{
"epoch": 6.416938110749186,
"grad_norm": 0.9621773958206177,
"learning_rate": 9.935960069974096e-05,
"loss": 0.0178,
"step": 1970
},
{
"epoch": 6.449511400651466,
"grad_norm": 1.2875672578811646,
"learning_rate": 9.934634378427506e-05,
"loss": 0.0204,
"step": 1980
},
{
"epoch": 6.482084690553746,
"grad_norm": 1.1186437606811523,
"learning_rate": 9.933295195778954e-05,
"loss": 0.0211,
"step": 1990
},
{
"epoch": 6.514657980456026,
"grad_norm": 1.0017237663269043,
"learning_rate": 9.931942525689715e-05,
"loss": 0.023,
"step": 2000
},
{
"epoch": 6.547231270358306,
"grad_norm": 0.9961879253387451,
"learning_rate": 9.930576371857936e-05,
"loss": 0.0205,
"step": 2010
},
{
"epoch": 6.579804560260587,
"grad_norm": 1.1414068937301636,
"learning_rate": 9.929196738018629e-05,
"loss": 0.0182,
"step": 2020
},
{
"epoch": 6.612377850162867,
"grad_norm": 1.139420509338379,
"learning_rate": 9.927803627943662e-05,
"loss": 0.0175,
"step": 2030
},
{
"epoch": 6.644951140065147,
"grad_norm": 1.1123589277267456,
"learning_rate": 9.926397045441744e-05,
"loss": 0.0195,
"step": 2040
},
{
"epoch": 6.677524429967427,
"grad_norm": 1.2396354675292969,
"learning_rate": 9.924976994358417e-05,
"loss": 0.0199,
"step": 2050
},
{
"epoch": 6.710097719869707,
"grad_norm": 1.2193408012390137,
"learning_rate": 9.923543478576048e-05,
"loss": 0.0197,
"step": 2060
},
{
"epoch": 6.742671009771987,
"grad_norm": 1.1543775796890259,
"learning_rate": 9.922096502013813e-05,
"loss": 0.0192,
"step": 2070
},
{
"epoch": 6.7752442996742674,
"grad_norm": 1.1310466527938843,
"learning_rate": 9.92063606862769e-05,
"loss": 0.0196,
"step": 2080
},
{
"epoch": 6.807817589576548,
"grad_norm": 1.2145164012908936,
"learning_rate": 9.919162182410453e-05,
"loss": 0.0196,
"step": 2090
},
{
"epoch": 6.840390879478828,
"grad_norm": 1.0499675273895264,
"learning_rate": 9.917674847391645e-05,
"loss": 0.0179,
"step": 2100
},
{
"epoch": 6.872964169381108,
"grad_norm": 0.8699261546134949,
"learning_rate": 9.916174067637584e-05,
"loss": 0.0168,
"step": 2110
},
{
"epoch": 6.905537459283387,
"grad_norm": 0.8845049142837524,
"learning_rate": 9.914659847251348e-05,
"loss": 0.0163,
"step": 2120
},
{
"epoch": 6.938110749185668,
"grad_norm": 0.8768344521522522,
"learning_rate": 9.913132190372753e-05,
"loss": 0.0189,
"step": 2130
},
{
"epoch": 6.970684039087947,
"grad_norm": 1.0072346925735474,
"learning_rate": 9.911591101178359e-05,
"loss": 0.017,
"step": 2140
},
{
"epoch": 7.003257328990228,
"grad_norm": 0.9805561304092407,
"learning_rate": 9.910036583881443e-05,
"loss": 0.0168,
"step": 2150
},
{
"epoch": 7.035830618892508,
"grad_norm": 0.9882418513298035,
"learning_rate": 9.908468642731995e-05,
"loss": 0.019,
"step": 2160
},
{
"epoch": 7.068403908794788,
"grad_norm": 1.0311851501464844,
"learning_rate": 9.906887282016707e-05,
"loss": 0.018,
"step": 2170
},
{
"epoch": 7.100977198697068,
"grad_norm": 1.3635246753692627,
"learning_rate": 9.90529250605896e-05,
"loss": 0.019,
"step": 2180
},
{
"epoch": 7.133550488599348,
"grad_norm": 0.94489586353302,
"learning_rate": 9.903684319218809e-05,
"loss": 0.0186,
"step": 2190
},
{
"epoch": 7.166123778501628,
"grad_norm": 1.1365646123886108,
"learning_rate": 9.902062725892976e-05,
"loss": 0.0195,
"step": 2200
},
{
"epoch": 7.198697068403908,
"grad_norm": 1.1778250932693481,
"learning_rate": 9.900427730514834e-05,
"loss": 0.0196,
"step": 2210
},
{
"epoch": 7.231270358306189,
"grad_norm": 1.0461801290512085,
"learning_rate": 9.8987793375544e-05,
"loss": 0.0183,
"step": 2220
},
{
"epoch": 7.263843648208469,
"grad_norm": 1.068851351737976,
"learning_rate": 9.897117551518318e-05,
"loss": 0.0172,
"step": 2230
},
{
"epoch": 7.296416938110749,
"grad_norm": 0.906645655632019,
"learning_rate": 9.895442376949844e-05,
"loss": 0.016,
"step": 2240
},
{
"epoch": 7.328990228013029,
"grad_norm": 0.9077069759368896,
"learning_rate": 9.893753818428845e-05,
"loss": 0.0178,
"step": 2250
},
{
"epoch": 7.361563517915309,
"grad_norm": 0.8832686543464661,
"learning_rate": 9.892051880571773e-05,
"loss": 0.017,
"step": 2260
},
{
"epoch": 7.394136807817589,
"grad_norm": 0.9024226665496826,
"learning_rate": 9.890336568031663e-05,
"loss": 0.0154,
"step": 2270
},
{
"epoch": 7.4267100977198695,
"grad_norm": 0.8265984058380127,
"learning_rate": 9.888607885498113e-05,
"loss": 0.0166,
"step": 2280
},
{
"epoch": 7.45928338762215,
"grad_norm": 0.9777111411094666,
"learning_rate": 9.886865837697275e-05,
"loss": 0.0172,
"step": 2290
},
{
"epoch": 7.49185667752443,
"grad_norm": 0.9763800501823425,
"learning_rate": 9.88511042939184e-05,
"loss": 0.0188,
"step": 2300
},
{
"epoch": 7.52442996742671,
"grad_norm": 1.0408334732055664,
"learning_rate": 9.883341665381028e-05,
"loss": 0.0187,
"step": 2310
},
{
"epoch": 7.55700325732899,
"grad_norm": 0.6989148259162903,
"learning_rate": 9.881559550500575e-05,
"loss": 0.0174,
"step": 2320
},
{
"epoch": 7.58957654723127,
"grad_norm": 1.2441219091415405,
"learning_rate": 9.879764089622712e-05,
"loss": 0.0175,
"step": 2330
},
{
"epoch": 7.62214983713355,
"grad_norm": 0.9775832891464233,
"learning_rate": 9.87795528765616e-05,
"loss": 0.0174,
"step": 2340
},
{
"epoch": 7.6547231270358305,
"grad_norm": 1.1409343481063843,
"learning_rate": 9.876133149546118e-05,
"loss": 0.0198,
"step": 2350
},
{
"epoch": 7.687296416938111,
"grad_norm": 0.9530286192893982,
"learning_rate": 9.874297680274238e-05,
"loss": 0.0174,
"step": 2360
},
{
"epoch": 7.719869706840391,
"grad_norm": 0.9352869391441345,
"learning_rate": 9.872448884858624e-05,
"loss": 0.0175,
"step": 2370
},
{
"epoch": 7.752442996742671,
"grad_norm": 0.9207677245140076,
"learning_rate": 9.870586768353815e-05,
"loss": 0.0183,
"step": 2380
},
{
"epoch": 7.785016286644951,
"grad_norm": 0.8910165429115295,
"learning_rate": 9.868711335850764e-05,
"loss": 0.0189,
"step": 2390
},
{
"epoch": 7.817589576547231,
"grad_norm": 1.199159026145935,
"learning_rate": 9.866822592476833e-05,
"loss": 0.0181,
"step": 2400
},
{
"epoch": 7.850162866449511,
"grad_norm": 1.0165252685546875,
"learning_rate": 9.86492054339577e-05,
"loss": 0.0193,
"step": 2410
},
{
"epoch": 7.8827361563517915,
"grad_norm": 0.8919486999511719,
"learning_rate": 9.863005193807711e-05,
"loss": 0.0173,
"step": 2420
},
{
"epoch": 7.915309446254072,
"grad_norm": 0.7932745218276978,
"learning_rate": 9.861076548949143e-05,
"loss": 0.017,
"step": 2430
},
{
"epoch": 7.947882736156352,
"grad_norm": 0.8434972763061523,
"learning_rate": 9.859134614092912e-05,
"loss": 0.0177,
"step": 2440
},
{
"epoch": 7.980456026058632,
"grad_norm": 1.005811333656311,
"learning_rate": 9.857179394548191e-05,
"loss": 0.0156,
"step": 2450
},
{
"epoch": 8.013029315960912,
"grad_norm": 0.9450655579566956,
"learning_rate": 9.855210895660477e-05,
"loss": 0.0179,
"step": 2460
},
{
"epoch": 8.045602605863191,
"grad_norm": 1.0200849771499634,
"learning_rate": 9.853229122811568e-05,
"loss": 0.0182,
"step": 2470
},
{
"epoch": 8.078175895765472,
"grad_norm": 0.7630968689918518,
"learning_rate": 9.851234081419559e-05,
"loss": 0.0175,
"step": 2480
},
{
"epoch": 8.110749185667752,
"grad_norm": 0.7456977963447571,
"learning_rate": 9.849225776938814e-05,
"loss": 0.0182,
"step": 2490
},
{
"epoch": 8.143322475570033,
"grad_norm": 0.7413007020950317,
"learning_rate": 9.847204214859964e-05,
"loss": 0.0169,
"step": 2500
},
{
"epoch": 8.175895765472312,
"grad_norm": 0.704645037651062,
"learning_rate": 9.845169400709879e-05,
"loss": 0.0169,
"step": 2510
},
{
"epoch": 8.208469055374593,
"grad_norm": 0.9419644474983215,
"learning_rate": 9.843121340051664e-05,
"loss": 0.0164,
"step": 2520
},
{
"epoch": 8.241042345276872,
"grad_norm": 1.1566686630249023,
"learning_rate": 9.841060038484641e-05,
"loss": 0.0159,
"step": 2530
},
{
"epoch": 8.273615635179153,
"grad_norm": 0.862409770488739,
"learning_rate": 9.838985501644328e-05,
"loss": 0.016,
"step": 2540
},
{
"epoch": 8.306188925081432,
"grad_norm": 0.8754869103431702,
"learning_rate": 9.83689773520243e-05,
"loss": 0.016,
"step": 2550
},
{
"epoch": 8.338762214983714,
"grad_norm": 1.02450692653656,
"learning_rate": 9.834796744866819e-05,
"loss": 0.0164,
"step": 2560
},
{
"epoch": 8.371335504885993,
"grad_norm": 0.7608270645141602,
"learning_rate": 9.832682536381525e-05,
"loss": 0.0155,
"step": 2570
},
{
"epoch": 8.403908794788274,
"grad_norm": 0.7060204744338989,
"learning_rate": 9.830555115526711e-05,
"loss": 0.0164,
"step": 2580
},
{
"epoch": 8.436482084690553,
"grad_norm": 0.6277144551277161,
"learning_rate": 9.828414488118667e-05,
"loss": 0.0153,
"step": 2590
},
{
"epoch": 8.469055374592834,
"grad_norm": 0.7759043574333191,
"learning_rate": 9.826260660009785e-05,
"loss": 0.0147,
"step": 2600
},
{
"epoch": 8.501628664495113,
"grad_norm": 0.8001388311386108,
"learning_rate": 9.824093637088547e-05,
"loss": 0.0151,
"step": 2610
},
{
"epoch": 8.534201954397394,
"grad_norm": 0.7903552055358887,
"learning_rate": 9.821913425279514e-05,
"loss": 0.016,
"step": 2620
},
{
"epoch": 8.566775244299674,
"grad_norm": 0.840638279914856,
"learning_rate": 9.8197200305433e-05,
"loss": 0.0155,
"step": 2630
},
{
"epoch": 8.599348534201955,
"grad_norm": 0.9356780648231506,
"learning_rate": 9.817513458876564e-05,
"loss": 0.0176,
"step": 2640
},
{
"epoch": 8.631921824104234,
"grad_norm": 0.8237717151641846,
"learning_rate": 9.815293716311987e-05,
"loss": 0.0157,
"step": 2650
},
{
"epoch": 8.664495114006515,
"grad_norm": 1.1324107646942139,
"learning_rate": 9.813060808918262e-05,
"loss": 0.0156,
"step": 2660
},
{
"epoch": 8.697068403908794,
"grad_norm": 0.7818335294723511,
"learning_rate": 9.810814742800069e-05,
"loss": 0.0156,
"step": 2670
},
{
"epoch": 8.729641693811075,
"grad_norm": 0.8096736669540405,
"learning_rate": 9.808555524098074e-05,
"loss": 0.0148,
"step": 2680
},
{
"epoch": 8.762214983713354,
"grad_norm": 1.0488592386245728,
"learning_rate": 9.806283158988887e-05,
"loss": 0.0151,
"step": 2690
},
{
"epoch": 8.794788273615636,
"grad_norm": 0.8972393870353699,
"learning_rate": 9.803997653685072e-05,
"loss": 0.0153,
"step": 2700
},
{
"epoch": 8.827361563517915,
"grad_norm": 0.8795937299728394,
"learning_rate": 9.801699014435112e-05,
"loss": 0.0159,
"step": 2710
},
{
"epoch": 8.859934853420196,
"grad_norm": 0.7514339685440063,
"learning_rate": 9.799387247523398e-05,
"loss": 0.0153,
"step": 2720
},
{
"epoch": 8.892508143322475,
"grad_norm": 0.8946635127067566,
"learning_rate": 9.797062359270215e-05,
"loss": 0.0136,
"step": 2730
},
{
"epoch": 8.925081433224756,
"grad_norm": 0.8136707544326782,
"learning_rate": 9.794724356031715e-05,
"loss": 0.0143,
"step": 2740
},
{
"epoch": 8.957654723127035,
"grad_norm": 0.6647421717643738,
"learning_rate": 9.792373244199913e-05,
"loss": 0.0142,
"step": 2750
},
{
"epoch": 8.990228013029316,
"grad_norm": 0.8160433769226074,
"learning_rate": 9.790009030202658e-05,
"loss": 0.0158,
"step": 2760
},
{
"epoch": 9.022801302931596,
"grad_norm": 0.6846116185188293,
"learning_rate": 9.78763172050362e-05,
"loss": 0.0133,
"step": 2770
},
{
"epoch": 9.055374592833877,
"grad_norm": 0.806671142578125,
"learning_rate": 9.785241321602274e-05,
"loss": 0.0163,
"step": 2780
},
{
"epoch": 9.087947882736156,
"grad_norm": 0.7517706751823425,
"learning_rate": 9.782837840033879e-05,
"loss": 0.0163,
"step": 2790
},
{
"epoch": 9.120521172638437,
"grad_norm": 0.866486668586731,
"learning_rate": 9.780421282369461e-05,
"loss": 0.017,
"step": 2800
},
{
"epoch": 9.153094462540716,
"grad_norm": 0.7923924326896667,
"learning_rate": 9.777991655215797e-05,
"loss": 0.0187,
"step": 2810
},
{
"epoch": 9.185667752442997,
"grad_norm": 0.8920378684997559,
"learning_rate": 9.775548965215394e-05,
"loss": 0.0161,
"step": 2820
},
{
"epoch": 9.218241042345277,
"grad_norm": 0.7856971621513367,
"learning_rate": 9.773093219046474e-05,
"loss": 0.0154,
"step": 2830
},
{
"epoch": 9.250814332247558,
"grad_norm": 1.1112109422683716,
"learning_rate": 9.770624423422954e-05,
"loss": 0.0164,
"step": 2840
},
{
"epoch": 9.283387622149837,
"grad_norm": 0.8900606632232666,
"learning_rate": 9.768142585094426e-05,
"loss": 0.0148,
"step": 2850
},
{
"epoch": 9.315960912052118,
"grad_norm": 0.808012068271637,
"learning_rate": 9.765647710846142e-05,
"loss": 0.015,
"step": 2860
},
{
"epoch": 9.348534201954397,
"grad_norm": 0.7280744910240173,
"learning_rate": 9.763139807498991e-05,
"loss": 0.0168,
"step": 2870
},
{
"epoch": 9.381107491856678,
"grad_norm": 0.7034526467323303,
"learning_rate": 9.760618881909487e-05,
"loss": 0.0142,
"step": 2880
},
{
"epoch": 9.413680781758957,
"grad_norm": 0.6753115653991699,
"learning_rate": 9.758084940969744e-05,
"loss": 0.0147,
"step": 2890
},
{
"epoch": 9.446254071661238,
"grad_norm": 0.7895611524581909,
"learning_rate": 9.755537991607459e-05,
"loss": 0.0136,
"step": 2900
},
{
"epoch": 9.478827361563518,
"grad_norm": 0.7646326422691345,
"learning_rate": 9.752978040785895e-05,
"loss": 0.0153,
"step": 2910
},
{
"epoch": 9.511400651465799,
"grad_norm": 0.784275233745575,
"learning_rate": 9.750405095503859e-05,
"loss": 0.0136,
"step": 2920
},
{
"epoch": 9.543973941368078,
"grad_norm": 0.6947634816169739,
"learning_rate": 9.747819162795686e-05,
"loss": 0.0156,
"step": 2930
},
{
"epoch": 9.576547231270359,
"grad_norm": 0.7826408743858337,
"learning_rate": 9.745220249731217e-05,
"loss": 0.0134,
"step": 2940
},
{
"epoch": 9.609120521172638,
"grad_norm": 0.8155382871627808,
"learning_rate": 9.742608363415781e-05,
"loss": 0.0131,
"step": 2950
},
{
"epoch": 9.64169381107492,
"grad_norm": 0.7903912663459778,
"learning_rate": 9.739983510990176e-05,
"loss": 0.0151,
"step": 2960
},
{
"epoch": 9.674267100977199,
"grad_norm": 0.6831985712051392,
"learning_rate": 9.737345699630647e-05,
"loss": 0.0142,
"step": 2970
},
{
"epoch": 9.70684039087948,
"grad_norm": 0.7938959002494812,
"learning_rate": 9.734694936548869e-05,
"loss": 0.0157,
"step": 2980
},
{
"epoch": 9.739413680781759,
"grad_norm": 0.8478028774261475,
"learning_rate": 9.732031228991932e-05,
"loss": 0.0135,
"step": 2990
},
{
"epoch": 9.77198697068404,
"grad_norm": 0.8881486058235168,
"learning_rate": 9.729354584242302e-05,
"loss": 0.0145,
"step": 3000
},
{
"epoch": 9.80456026058632,
"grad_norm": 0.8107954859733582,
"learning_rate": 9.726665009617832e-05,
"loss": 0.0147,
"step": 3010
},
{
"epoch": 9.8371335504886,
"grad_norm": 0.593552827835083,
"learning_rate": 9.723962512471714e-05,
"loss": 0.0151,
"step": 3020
},
{
"epoch": 9.86970684039088,
"grad_norm": 0.7691318392753601,
"learning_rate": 9.72124710019247e-05,
"loss": 0.0133,
"step": 3030
},
{
"epoch": 9.90228013029316,
"grad_norm": 0.8209340572357178,
"learning_rate": 9.718518780203934e-05,
"loss": 0.0143,
"step": 3040
},
{
"epoch": 9.93485342019544,
"grad_norm": 0.786946713924408,
"learning_rate": 9.715777559965228e-05,
"loss": 0.016,
"step": 3050
},
{
"epoch": 9.967426710097719,
"grad_norm": 0.833220899105072,
"learning_rate": 9.713023446970746e-05,
"loss": 0.0142,
"step": 3060
},
{
"epoch": 10.0,
"grad_norm": 0.8491772413253784,
"learning_rate": 9.710256448750126e-05,
"loss": 0.0155,
"step": 3070
},
{
"epoch": 10.03257328990228,
"grad_norm": 0.7305663228034973,
"learning_rate": 9.707476572868235e-05,
"loss": 0.0135,
"step": 3080
},
{
"epoch": 10.06514657980456,
"grad_norm": 0.7415571808815002,
"learning_rate": 9.704683826925149e-05,
"loss": 0.0127,
"step": 3090
},
{
"epoch": 10.09771986970684,
"grad_norm": 0.723970890045166,
"learning_rate": 9.701878218556129e-05,
"loss": 0.0139,
"step": 3100
},
{
"epoch": 10.13029315960912,
"grad_norm": 0.7780591249465942,
"learning_rate": 9.699059755431598e-05,
"loss": 0.015,
"step": 3110
},
{
"epoch": 10.1628664495114,
"grad_norm": 0.8553524017333984,
"learning_rate": 9.696228445257132e-05,
"loss": 0.0144,
"step": 3120
},
{
"epoch": 10.19543973941368,
"grad_norm": 0.6987228393554688,
"learning_rate": 9.693384295773419e-05,
"loss": 0.0143,
"step": 3130
},
{
"epoch": 10.22801302931596,
"grad_norm": 0.6953451633453369,
"learning_rate": 9.690527314756259e-05,
"loss": 0.0126,
"step": 3140
},
{
"epoch": 10.260586319218241,
"grad_norm": 0.7900610566139221,
"learning_rate": 9.687657510016527e-05,
"loss": 0.0142,
"step": 3150
},
{
"epoch": 10.29315960912052,
"grad_norm": 0.7502635717391968,
"learning_rate": 9.684774889400161e-05,
"loss": 0.0137,
"step": 3160
},
{
"epoch": 10.325732899022801,
"grad_norm": 0.8514730334281921,
"learning_rate": 9.681879460788135e-05,
"loss": 0.0145,
"step": 3170
},
{
"epoch": 10.35830618892508,
"grad_norm": 0.6699235439300537,
"learning_rate": 9.67897123209644e-05,
"loss": 0.016,
"step": 3180
},
{
"epoch": 10.390879478827362,
"grad_norm": 0.9717312455177307,
"learning_rate": 9.676050211276062e-05,
"loss": 0.0139,
"step": 3190
},
{
"epoch": 10.423452768729641,
"grad_norm": 0.8204063177108765,
"learning_rate": 9.673116406312962e-05,
"loss": 0.0149,
"step": 3200
},
{
"epoch": 10.456026058631922,
"grad_norm": 0.6988447308540344,
"learning_rate": 9.67016982522805e-05,
"loss": 0.0169,
"step": 3210
},
{
"epoch": 10.488599348534201,
"grad_norm": 0.7764069437980652,
"learning_rate": 9.667210476077164e-05,
"loss": 0.0134,
"step": 3220
},
{
"epoch": 10.521172638436482,
"grad_norm": 0.8181342482566833,
"learning_rate": 9.664238366951055e-05,
"loss": 0.014,
"step": 3230
},
{
"epoch": 10.553745928338762,
"grad_norm": 0.8816471099853516,
"learning_rate": 9.661253505975355e-05,
"loss": 0.0132,
"step": 3240
},
{
"epoch": 10.586319218241043,
"grad_norm": 0.6358229517936707,
"learning_rate": 9.658255901310557e-05,
"loss": 0.0141,
"step": 3250
},
{
"epoch": 10.618892508143322,
"grad_norm": 0.8587698340415955,
"learning_rate": 9.655245561152e-05,
"loss": 0.0147,
"step": 3260
},
{
"epoch": 10.651465798045603,
"grad_norm": 0.7497543692588806,
"learning_rate": 9.65222249372984e-05,
"loss": 0.0138,
"step": 3270
},
{
"epoch": 10.684039087947882,
"grad_norm": 0.7951953411102295,
"learning_rate": 9.649186707309026e-05,
"loss": 0.0124,
"step": 3280
},
{
"epoch": 10.716612377850163,
"grad_norm": 0.7938198447227478,
"learning_rate": 9.646138210189283e-05,
"loss": 0.0144,
"step": 3290
},
{
"epoch": 10.749185667752442,
"grad_norm": 0.7425335049629211,
"learning_rate": 9.643077010705087e-05,
"loss": 0.0124,
"step": 3300
},
{
"epoch": 10.781758957654723,
"grad_norm": 0.6020338535308838,
"learning_rate": 9.640003117225637e-05,
"loss": 0.014,
"step": 3310
},
{
"epoch": 10.814332247557003,
"grad_norm": 0.7467803359031677,
"learning_rate": 9.636916538154846e-05,
"loss": 0.016,
"step": 3320
},
{
"epoch": 10.846905537459284,
"grad_norm": 0.7082899212837219,
"learning_rate": 9.633817281931296e-05,
"loss": 0.0168,
"step": 3330
},
{
"epoch": 10.879478827361563,
"grad_norm": 0.7912280559539795,
"learning_rate": 9.630705357028242e-05,
"loss": 0.0136,
"step": 3340
},
{
"epoch": 10.912052117263844,
"grad_norm": 0.6572038531303406,
"learning_rate": 9.627580771953563e-05,
"loss": 0.0129,
"step": 3350
},
{
"epoch": 10.944625407166123,
"grad_norm": 0.8878430724143982,
"learning_rate": 9.624443535249759e-05,
"loss": 0.0134,
"step": 3360
},
{
"epoch": 10.977198697068404,
"grad_norm": 0.6283133625984192,
"learning_rate": 9.621293655493913e-05,
"loss": 0.0159,
"step": 3370
},
{
"epoch": 11.009771986970684,
"grad_norm": 0.7855213284492493,
"learning_rate": 9.618131141297675e-05,
"loss": 0.0151,
"step": 3380
},
{
"epoch": 11.042345276872965,
"grad_norm": 0.6657193303108215,
"learning_rate": 9.614956001307242e-05,
"loss": 0.0143,
"step": 3390
},
{
"epoch": 11.074918566775244,
"grad_norm": 0.99737948179245,
"learning_rate": 9.611768244203321e-05,
"loss": 0.0149,
"step": 3400
},
{
"epoch": 11.107491856677525,
"grad_norm": 0.5985683798789978,
"learning_rate": 9.60856787870112e-05,
"loss": 0.0142,
"step": 3410
},
{
"epoch": 11.140065146579804,
"grad_norm": 0.7139508724212646,
"learning_rate": 9.605354913550318e-05,
"loss": 0.0149,
"step": 3420
},
{
"epoch": 11.172638436482085,
"grad_norm": 0.6530821919441223,
"learning_rate": 9.602129357535037e-05,
"loss": 0.0135,
"step": 3430
},
{
"epoch": 11.205211726384364,
"grad_norm": 0.6087831854820251,
"learning_rate": 9.598891219473825e-05,
"loss": 0.0132,
"step": 3440
},
{
"epoch": 11.237785016286646,
"grad_norm": 0.631524384021759,
"learning_rate": 9.595640508219625e-05,
"loss": 0.0136,
"step": 3450
},
{
"epoch": 11.270358306188925,
"grad_norm": 0.7701088786125183,
"learning_rate": 9.592377232659761e-05,
"loss": 0.0146,
"step": 3460
},
{
"epoch": 11.302931596091206,
"grad_norm": 0.6873641610145569,
"learning_rate": 9.589101401715904e-05,
"loss": 0.0139,
"step": 3470
},
{
"epoch": 11.335504885993485,
"grad_norm": 0.7974376678466797,
"learning_rate": 9.585813024344045e-05,
"loss": 0.0128,
"step": 3480
},
{
"epoch": 11.368078175895766,
"grad_norm": 0.7017616033554077,
"learning_rate": 9.58251210953449e-05,
"loss": 0.0138,
"step": 3490
},
{
"epoch": 11.400651465798045,
"grad_norm": 0.7259907722473145,
"learning_rate": 9.579198666311809e-05,
"loss": 0.013,
"step": 3500
},
{
"epoch": 11.433224755700326,
"grad_norm": 0.6088874340057373,
"learning_rate": 9.575872703734832e-05,
"loss": 0.0126,
"step": 3510
},
{
"epoch": 11.465798045602606,
"grad_norm": 0.5169392824172974,
"learning_rate": 9.572534230896611e-05,
"loss": 0.0116,
"step": 3520
},
{
"epoch": 11.498371335504887,
"grad_norm": 0.8139016628265381,
"learning_rate": 9.569183256924403e-05,
"loss": 0.0126,
"step": 3530
},
{
"epoch": 11.530944625407166,
"grad_norm": 0.7374736070632935,
"learning_rate": 9.565819790979646e-05,
"loss": 0.0131,
"step": 3540
},
{
"epoch": 11.563517915309447,
"grad_norm": 0.6468128561973572,
"learning_rate": 9.562443842257925e-05,
"loss": 0.0129,
"step": 3550
},
{
"epoch": 11.596091205211726,
"grad_norm": 0.6643866896629333,
"learning_rate": 9.559055419988956e-05,
"loss": 0.0133,
"step": 3560
},
{
"epoch": 11.628664495114007,
"grad_norm": 0.681177020072937,
"learning_rate": 9.555654533436557e-05,
"loss": 0.0135,
"step": 3570
},
{
"epoch": 11.661237785016286,
"grad_norm": 0.7795754075050354,
"learning_rate": 9.552241191898621e-05,
"loss": 0.0114,
"step": 3580
},
{
"epoch": 11.693811074918568,
"grad_norm": 0.6136661171913147,
"learning_rate": 9.548815404707092e-05,
"loss": 0.0139,
"step": 3590
},
{
"epoch": 11.726384364820847,
"grad_norm": 0.6191911697387695,
"learning_rate": 9.545377181227942e-05,
"loss": 0.0147,
"step": 3600
},
{
"epoch": 11.758957654723128,
"grad_norm": 0.67829829454422,
"learning_rate": 9.541926530861145e-05,
"loss": 0.013,
"step": 3610
},
{
"epoch": 11.791530944625407,
"grad_norm": 0.7790681719779968,
"learning_rate": 9.538463463040645e-05,
"loss": 0.0131,
"step": 3620
},
{
"epoch": 11.824104234527688,
"grad_norm": 0.694274365901947,
"learning_rate": 9.534987987234337e-05,
"loss": 0.0138,
"step": 3630
},
{
"epoch": 11.856677524429967,
"grad_norm": 0.5702154636383057,
"learning_rate": 9.53150011294404e-05,
"loss": 0.0121,
"step": 3640
},
{
"epoch": 11.889250814332247,
"grad_norm": 0.5818066000938416,
"learning_rate": 9.527999849705471e-05,
"loss": 0.0132,
"step": 3650
},
{
"epoch": 11.921824104234528,
"grad_norm": 0.6952799558639526,
"learning_rate": 9.524487207088213e-05,
"loss": 0.0117,
"step": 3660
},
{
"epoch": 11.954397394136809,
"grad_norm": 0.5978224277496338,
"learning_rate": 9.520962194695698e-05,
"loss": 0.0137,
"step": 3670
},
{
"epoch": 11.986970684039088,
"grad_norm": 0.6350916028022766,
"learning_rate": 9.517424822165175e-05,
"loss": 0.0116,
"step": 3680
},
{
"epoch": 12.019543973941367,
"grad_norm": 0.651490330696106,
"learning_rate": 9.513875099167685e-05,
"loss": 0.0164,
"step": 3690
},
{
"epoch": 12.052117263843648,
"grad_norm": 0.7070964574813843,
"learning_rate": 9.510313035408035e-05,
"loss": 0.0142,
"step": 3700
},
{
"epoch": 12.084690553745927,
"grad_norm": 0.8875077962875366,
"learning_rate": 9.506738640624775e-05,
"loss": 0.0121,
"step": 3710
},
{
"epoch": 12.117263843648209,
"grad_norm": 0.8056143522262573,
"learning_rate": 9.50315192459016e-05,
"loss": 0.013,
"step": 3720
},
{
"epoch": 12.149837133550488,
"grad_norm": 0.7540460228919983,
"learning_rate": 9.499552897110136e-05,
"loss": 0.0126,
"step": 3730
},
{
"epoch": 12.182410423452769,
"grad_norm": 0.7277035713195801,
"learning_rate": 9.495941568024304e-05,
"loss": 0.014,
"step": 3740
},
{
"epoch": 12.214983713355048,
"grad_norm": 0.7212410569190979,
"learning_rate": 9.492317947205904e-05,
"loss": 0.0116,
"step": 3750
},
{
"epoch": 12.247557003257329,
"grad_norm": 0.6804754137992859,
"learning_rate": 9.488682044561775e-05,
"loss": 0.0124,
"step": 3760
},
{
"epoch": 12.280130293159608,
"grad_norm": 0.7375156879425049,
"learning_rate": 9.485033870032335e-05,
"loss": 0.0132,
"step": 3770
},
{
"epoch": 12.31270358306189,
"grad_norm": 0.8549667000770569,
"learning_rate": 9.481373433591556e-05,
"loss": 0.0123,
"step": 3780
},
{
"epoch": 12.345276872964169,
"grad_norm": 0.6877515912055969,
"learning_rate": 9.47770074524693e-05,
"loss": 0.0149,
"step": 3790
},
{
"epoch": 12.37785016286645,
"grad_norm": 0.7423529624938965,
"learning_rate": 9.474015815039446e-05,
"loss": 0.0115,
"step": 3800
},
{
"epoch": 12.410423452768729,
"grad_norm": 0.7478250861167908,
"learning_rate": 9.470318653043565e-05,
"loss": 0.0125,
"step": 3810
},
{
"epoch": 12.44299674267101,
"grad_norm": 0.6396955847740173,
"learning_rate": 9.466609269367185e-05,
"loss": 0.0128,
"step": 3820
},
{
"epoch": 12.47557003257329,
"grad_norm": 0.6483830809593201,
"learning_rate": 9.46288767415162e-05,
"loss": 0.0137,
"step": 3830
},
{
"epoch": 12.50814332247557,
"grad_norm": 0.7733943462371826,
"learning_rate": 9.459153877571567e-05,
"loss": 0.015,
"step": 3840
},
{
"epoch": 12.54071661237785,
"grad_norm": 0.6195006370544434,
"learning_rate": 9.455407889835087e-05,
"loss": 0.012,
"step": 3850
},
{
"epoch": 12.57328990228013,
"grad_norm": 0.6243380308151245,
"learning_rate": 9.451649721183564e-05,
"loss": 0.0149,
"step": 3860
},
{
"epoch": 12.60586319218241,
"grad_norm": 0.6876091361045837,
"learning_rate": 9.447879381891692e-05,
"loss": 0.0118,
"step": 3870
},
{
"epoch": 12.63843648208469,
"grad_norm": 0.7486156225204468,
"learning_rate": 9.444096882267428e-05,
"loss": 0.0125,
"step": 3880
},
{
"epoch": 12.67100977198697,
"grad_norm": 0.6939566731452942,
"learning_rate": 9.440302232651988e-05,
"loss": 0.0136,
"step": 3890
},
{
"epoch": 12.703583061889251,
"grad_norm": 0.6740759015083313,
"learning_rate": 9.436495443419795e-05,
"loss": 0.0121,
"step": 3900
},
{
"epoch": 12.73615635179153,
"grad_norm": 0.7866451144218445,
"learning_rate": 9.432676524978466e-05,
"loss": 0.0121,
"step": 3910
},
{
"epoch": 12.768729641693811,
"grad_norm": 0.5861152410507202,
"learning_rate": 9.42884548776878e-05,
"loss": 0.0111,
"step": 3920
},
{
"epoch": 12.80130293159609,
"grad_norm": 0.5740252137184143,
"learning_rate": 9.425002342264646e-05,
"loss": 0.0119,
"step": 3930
},
{
"epoch": 12.833876221498372,
"grad_norm": 0.6962918043136597,
"learning_rate": 9.421147098973077e-05,
"loss": 0.0118,
"step": 3940
},
{
"epoch": 12.866449511400651,
"grad_norm": 0.4894341826438904,
"learning_rate": 9.41727976843416e-05,
"loss": 0.0123,
"step": 3950
},
{
"epoch": 12.899022801302932,
"grad_norm": 0.6954609155654907,
"learning_rate": 9.413400361221029e-05,
"loss": 0.0131,
"step": 3960
},
{
"epoch": 12.931596091205211,
"grad_norm": 0.576310932636261,
"learning_rate": 9.409508887939835e-05,
"loss": 0.0121,
"step": 3970
},
{
"epoch": 12.964169381107492,
"grad_norm": 0.4614481031894684,
"learning_rate": 9.40560535922972e-05,
"loss": 0.0132,
"step": 3980
},
{
"epoch": 12.996742671009772,
"grad_norm": 0.47341176867485046,
"learning_rate": 9.40168978576278e-05,
"loss": 0.0137,
"step": 3990
},
{
"epoch": 13.029315960912053,
"grad_norm": 0.5697060227394104,
"learning_rate": 9.397762178244043e-05,
"loss": 0.0133,
"step": 4000
},
{
"epoch": 13.061889250814332,
"grad_norm": 0.612964928150177,
"learning_rate": 9.393822547411439e-05,
"loss": 0.0109,
"step": 4010
},
{
"epoch": 13.094462540716613,
"grad_norm": 0.6282061338424683,
"learning_rate": 9.389870904035769e-05,
"loss": 0.0116,
"step": 4020
},
{
"epoch": 13.127035830618892,
"grad_norm": 0.5586996078491211,
"learning_rate": 9.385907258920672e-05,
"loss": 0.0123,
"step": 4030
},
{
"epoch": 13.159609120521173,
"grad_norm": 0.5667091012001038,
"learning_rate": 9.381931622902607e-05,
"loss": 0.0119,
"step": 4040
},
{
"epoch": 13.192182410423452,
"grad_norm": 0.6335421800613403,
"learning_rate": 9.377944006850807e-05,
"loss": 0.0133,
"step": 4050
},
{
"epoch": 13.224755700325733,
"grad_norm": 0.8205068111419678,
"learning_rate": 9.373944421667265e-05,
"loss": 0.0122,
"step": 4060
},
{
"epoch": 13.257328990228013,
"grad_norm": 0.676856279373169,
"learning_rate": 9.369932878286691e-05,
"loss": 0.0126,
"step": 4070
},
{
"epoch": 13.289902280130294,
"grad_norm": 0.6334593296051025,
"learning_rate": 9.365909387676494e-05,
"loss": 0.0126,
"step": 4080
},
{
"epoch": 13.322475570032573,
"grad_norm": 0.6614803671836853,
"learning_rate": 9.361873960836744e-05,
"loss": 0.0126,
"step": 4090
},
{
"epoch": 13.355048859934854,
"grad_norm": 0.6303547620773315,
"learning_rate": 9.357826608800142e-05,
"loss": 0.0121,
"step": 4100
},
{
"epoch": 13.387622149837133,
"grad_norm": 0.5380171537399292,
"learning_rate": 9.353767342631994e-05,
"loss": 0.0121,
"step": 4110
},
{
"epoch": 13.420195439739414,
"grad_norm": 0.5387251377105713,
"learning_rate": 9.34969617343018e-05,
"loss": 0.0102,
"step": 4120
},
{
"epoch": 13.452768729641694,
"grad_norm": 0.5303720235824585,
"learning_rate": 9.345613112325122e-05,
"loss": 0.0126,
"step": 4130
},
{
"epoch": 13.485342019543975,
"grad_norm": 0.5876774787902832,
"learning_rate": 9.34151817047975e-05,
"loss": 0.0111,
"step": 4140
},
{
"epoch": 13.517915309446254,
"grad_norm": 0.6997594237327576,
"learning_rate": 9.33741135908948e-05,
"loss": 0.0119,
"step": 4150
},
{
"epoch": 13.550488599348535,
"grad_norm": 0.602103054523468,
"learning_rate": 9.33329268938218e-05,
"loss": 0.0118,
"step": 4160
},
{
"epoch": 13.583061889250814,
"grad_norm": 0.6966025829315186,
"learning_rate": 9.329162172618132e-05,
"loss": 0.0127,
"step": 4170
},
{
"epoch": 13.615635179153095,
"grad_norm": 0.5898999571800232,
"learning_rate": 9.325019820090013e-05,
"loss": 0.0113,
"step": 4180
},
{
"epoch": 13.648208469055374,
"grad_norm": 0.6419972777366638,
"learning_rate": 9.320865643122855e-05,
"loss": 0.0122,
"step": 4190
},
{
"epoch": 13.680781758957655,
"grad_norm": 0.48004379868507385,
"learning_rate": 9.316699653074023e-05,
"loss": 0.0115,
"step": 4200
},
{
"epoch": 13.713355048859935,
"grad_norm": 0.6654482483863831,
"learning_rate": 9.312521861333172e-05,
"loss": 0.0128,
"step": 4210
},
{
"epoch": 13.745928338762216,
"grad_norm": 0.5661275386810303,
"learning_rate": 9.308332279322224e-05,
"loss": 0.0114,
"step": 4220
},
{
"epoch": 13.778501628664495,
"grad_norm": 0.5483682155609131,
"learning_rate": 9.304130918495338e-05,
"loss": 0.0133,
"step": 4230
},
{
"epoch": 13.811074918566776,
"grad_norm": 0.7615593075752258,
"learning_rate": 9.299917790338874e-05,
"loss": 0.0119,
"step": 4240
},
{
"epoch": 13.843648208469055,
"grad_norm": 0.8771687150001526,
"learning_rate": 9.295692906371363e-05,
"loss": 0.013,
"step": 4250
},
{
"epoch": 13.876221498371336,
"grad_norm": 0.7540078163146973,
"learning_rate": 9.291456278143476e-05,
"loss": 0.0138,
"step": 4260
},
{
"epoch": 13.908794788273616,
"grad_norm": 0.711463212966919,
"learning_rate": 9.287207917237994e-05,
"loss": 0.0115,
"step": 4270
},
{
"epoch": 13.941368078175895,
"grad_norm": 0.6685953736305237,
"learning_rate": 9.282947835269773e-05,
"loss": 0.0128,
"step": 4280
},
{
"epoch": 13.973941368078176,
"grad_norm": 0.6462761163711548,
"learning_rate": 9.278676043885715e-05,
"loss": 0.0142,
"step": 4290
},
{
"epoch": 14.006514657980455,
"grad_norm": 0.577616274356842,
"learning_rate": 9.274392554764733e-05,
"loss": 0.0114,
"step": 4300
},
{
"epoch": 14.039087947882736,
"grad_norm": 0.6425755620002747,
"learning_rate": 9.270097379617723e-05,
"loss": 0.0124,
"step": 4310
},
{
"epoch": 14.071661237785015,
"grad_norm": 0.5634329319000244,
"learning_rate": 9.26579053018753e-05,
"loss": 0.0125,
"step": 4320
},
{
"epoch": 14.104234527687296,
"grad_norm": 0.5592512488365173,
"learning_rate": 9.261472018248918e-05,
"loss": 0.0131,
"step": 4330
},
{
"epoch": 14.136807817589576,
"grad_norm": 0.7388976216316223,
"learning_rate": 9.25714185560853e-05,
"loss": 0.013,
"step": 4340
},
{
"epoch": 14.169381107491857,
"grad_norm": 0.5503047108650208,
"learning_rate": 9.252800054104868e-05,
"loss": 0.0122,
"step": 4350
},
{
"epoch": 14.201954397394136,
"grad_norm": 0.4611617922782898,
"learning_rate": 9.248446625608252e-05,
"loss": 0.0111,
"step": 4360
},
{
"epoch": 14.234527687296417,
"grad_norm": 0.5560048222541809,
"learning_rate": 9.244081582020789e-05,
"loss": 0.0124,
"step": 4370
},
{
"epoch": 14.267100977198696,
"grad_norm": 0.6567198634147644,
"learning_rate": 9.239704935276339e-05,
"loss": 0.0119,
"step": 4380
},
{
"epoch": 14.299674267100977,
"grad_norm": 0.5934193134307861,
"learning_rate": 9.235316697340489e-05,
"loss": 0.0108,
"step": 4390
},
{
"epoch": 14.332247557003257,
"grad_norm": 0.7520450353622437,
"learning_rate": 9.230916880210512e-05,
"loss": 0.0141,
"step": 4400
},
{
"epoch": 14.364820846905538,
"grad_norm": 0.6309568285942078,
"learning_rate": 9.226505495915342e-05,
"loss": 0.0106,
"step": 4410
},
{
"epoch": 14.397394136807817,
"grad_norm": 0.5420092344284058,
"learning_rate": 9.222082556515536e-05,
"loss": 0.0115,
"step": 4420
},
{
"epoch": 14.429967426710098,
"grad_norm": 0.5959463715553284,
"learning_rate": 9.217648074103242e-05,
"loss": 0.0129,
"step": 4430
},
{
"epoch": 14.462540716612377,
"grad_norm": 0.5967207551002502,
"learning_rate": 9.213202060802161e-05,
"loss": 0.0127,
"step": 4440
},
{
"epoch": 14.495114006514658,
"grad_norm": 0.4098617732524872,
"learning_rate": 9.208744528767528e-05,
"loss": 0.0119,
"step": 4450
},
{
"epoch": 14.527687296416937,
"grad_norm": 0.6198092699050903,
"learning_rate": 9.204275490186064e-05,
"loss": 0.0125,
"step": 4460
},
{
"epoch": 14.560260586319218,
"grad_norm": 0.5683520436286926,
"learning_rate": 9.199794957275949e-05,
"loss": 0.014,
"step": 4470
},
{
"epoch": 14.592833876221498,
"grad_norm": 0.4835983216762543,
"learning_rate": 9.19530294228679e-05,
"loss": 0.011,
"step": 4480
},
{
"epoch": 14.625407166123779,
"grad_norm": 0.5733904242515564,
"learning_rate": 9.190799457499583e-05,
"loss": 0.0131,
"step": 4490
},
{
"epoch": 14.657980456026058,
"grad_norm": 0.4585092067718506,
"learning_rate": 9.186284515226686e-05,
"loss": 0.0128,
"step": 4500
},
{
"epoch": 14.690553745928339,
"grad_norm": 0.4910542070865631,
"learning_rate": 9.181758127811777e-05,
"loss": 0.0135,
"step": 4510
},
{
"epoch": 14.723127035830618,
"grad_norm": 0.5943559408187866,
"learning_rate": 9.177220307629825e-05,
"loss": 0.0128,
"step": 4520
},
{
"epoch": 14.7557003257329,
"grad_norm": 0.6429430246353149,
"learning_rate": 9.172671067087059e-05,
"loss": 0.012,
"step": 4530
},
{
"epoch": 14.788273615635179,
"grad_norm": 0.5697653889656067,
"learning_rate": 9.16811041862093e-05,
"loss": 0.0107,
"step": 4540
},
{
"epoch": 14.82084690553746,
"grad_norm": 0.5715034008026123,
"learning_rate": 9.163538374700076e-05,
"loss": 0.0118,
"step": 4550
},
{
"epoch": 14.853420195439739,
"grad_norm": 0.638119101524353,
"learning_rate": 9.158954947824287e-05,
"loss": 0.011,
"step": 4560
},
{
"epoch": 14.88599348534202,
"grad_norm": 0.5536983609199524,
"learning_rate": 9.154360150524482e-05,
"loss": 0.013,
"step": 4570
},
{
"epoch": 14.9185667752443,
"grad_norm": 0.6488440632820129,
"learning_rate": 9.14975399536266e-05,
"loss": 0.0108,
"step": 4580
},
{
"epoch": 14.95114006514658,
"grad_norm": 0.5250594019889832,
"learning_rate": 9.14513649493187e-05,
"loss": 0.0112,
"step": 4590
},
{
"epoch": 14.98371335504886,
"grad_norm": 0.53516685962677,
"learning_rate": 9.140507661856187e-05,
"loss": 0.0117,
"step": 4600
},
{
"epoch": 15.01628664495114,
"grad_norm": 0.722775936126709,
"learning_rate": 9.135867508790661e-05,
"loss": 0.0133,
"step": 4610
},
{
"epoch": 15.04885993485342,
"grad_norm": 0.5718010067939758,
"learning_rate": 9.131216048421291e-05,
"loss": 0.0121,
"step": 4620
},
{
"epoch": 15.0814332247557,
"grad_norm": 0.705929696559906,
"learning_rate": 9.126553293464998e-05,
"loss": 0.0113,
"step": 4630
},
{
"epoch": 15.11400651465798,
"grad_norm": 0.5129148960113525,
"learning_rate": 9.121879256669572e-05,
"loss": 0.013,
"step": 4640
},
{
"epoch": 15.146579804560261,
"grad_norm": 0.5438552498817444,
"learning_rate": 9.117193950813652e-05,
"loss": 0.0137,
"step": 4650
},
{
"epoch": 15.17915309446254,
"grad_norm": 0.6388025283813477,
"learning_rate": 9.112497388706685e-05,
"loss": 0.0127,
"step": 4660
},
{
"epoch": 15.211726384364821,
"grad_norm": 0.6098414659500122,
"learning_rate": 9.10778958318889e-05,
"loss": 0.0101,
"step": 4670
},
{
"epoch": 15.2442996742671,
"grad_norm": 0.6415327191352844,
"learning_rate": 9.103070547131232e-05,
"loss": 0.0142,
"step": 4680
},
{
"epoch": 15.276872964169382,
"grad_norm": 0.6037464141845703,
"learning_rate": 9.098340293435375e-05,
"loss": 0.0117,
"step": 4690
},
{
"epoch": 15.309446254071661,
"grad_norm": 0.5115535855293274,
"learning_rate": 9.093598835033649e-05,
"loss": 0.0105,
"step": 4700
},
{
"epoch": 15.342019543973942,
"grad_norm": 0.5303134918212891,
"learning_rate": 9.088846184889021e-05,
"loss": 0.0118,
"step": 4710
},
{
"epoch": 15.374592833876221,
"grad_norm": 0.5767205357551575,
"learning_rate": 9.084082355995057e-05,
"loss": 0.0105,
"step": 4720
},
{
"epoch": 15.407166123778502,
"grad_norm": 0.5441441535949707,
"learning_rate": 9.079307361375882e-05,
"loss": 0.0105,
"step": 4730
},
{
"epoch": 15.439739413680782,
"grad_norm": 0.5901165008544922,
"learning_rate": 9.074521214086149e-05,
"loss": 0.0123,
"step": 4740
},
{
"epoch": 15.472312703583063,
"grad_norm": 0.4205402731895447,
"learning_rate": 9.069723927211001e-05,
"loss": 0.0103,
"step": 4750
},
{
"epoch": 15.504885993485342,
"grad_norm": 0.6429307460784912,
"learning_rate": 9.064915513866037e-05,
"loss": 0.0093,
"step": 4760
},
{
"epoch": 15.537459283387623,
"grad_norm": 0.47192737460136414,
"learning_rate": 9.060095987197279e-05,
"loss": 0.0109,
"step": 4770
},
{
"epoch": 15.570032573289902,
"grad_norm": 0.5266391634941101,
"learning_rate": 9.055265360381126e-05,
"loss": 0.0097,
"step": 4780
},
{
"epoch": 15.602605863192183,
"grad_norm": 0.485973984003067,
"learning_rate": 9.050423646624326e-05,
"loss": 0.0101,
"step": 4790
},
{
"epoch": 15.635179153094462,
"grad_norm": 0.5286204218864441,
"learning_rate": 9.045570859163943e-05,
"loss": 0.0117,
"step": 4800
},
{
"epoch": 15.667752442996743,
"grad_norm": 0.6801819205284119,
"learning_rate": 9.04070701126731e-05,
"loss": 0.0106,
"step": 4810
},
{
"epoch": 15.700325732899023,
"grad_norm": 0.5797834396362305,
"learning_rate": 9.035832116232001e-05,
"loss": 0.0115,
"step": 4820
},
{
"epoch": 15.732899022801304,
"grad_norm": 0.7416790723800659,
"learning_rate": 9.030946187385796e-05,
"loss": 0.0112,
"step": 4830
},
{
"epoch": 15.765472312703583,
"grad_norm": 0.5395382046699524,
"learning_rate": 9.026049238086635e-05,
"loss": 0.0101,
"step": 4840
},
{
"epoch": 15.798045602605864,
"grad_norm": 0.667809784412384,
"learning_rate": 9.021141281722591e-05,
"loss": 0.0112,
"step": 4850
},
{
"epoch": 15.830618892508143,
"grad_norm": 0.543626606464386,
"learning_rate": 9.01622233171183e-05,
"loss": 0.0117,
"step": 4860
},
{
"epoch": 15.863192182410424,
"grad_norm": 0.5138244032859802,
"learning_rate": 9.011292401502574e-05,
"loss": 0.0106,
"step": 4870
},
{
"epoch": 15.895765472312704,
"grad_norm": 0.491041898727417,
"learning_rate": 9.006351504573063e-05,
"loss": 0.0126,
"step": 4880
},
{
"epoch": 15.928338762214985,
"grad_norm": 0.3895576298236847,
"learning_rate": 9.001399654431519e-05,
"loss": 0.0108,
"step": 4890
},
{
"epoch": 15.960912052117264,
"grad_norm": 0.6080408692359924,
"learning_rate": 8.996436864616116e-05,
"loss": 0.013,
"step": 4900
},
{
"epoch": 15.993485342019543,
"grad_norm": 0.6155601143836975,
"learning_rate": 8.991463148694925e-05,
"loss": 0.0096,
"step": 4910
},
{
"epoch": 16.026058631921824,
"grad_norm": 0.6588138341903687,
"learning_rate": 8.986478520265902e-05,
"loss": 0.0104,
"step": 4920
},
{
"epoch": 16.058631921824105,
"grad_norm": 0.6062794327735901,
"learning_rate": 8.981482992956827e-05,
"loss": 0.0107,
"step": 4930
},
{
"epoch": 16.091205211726383,
"grad_norm": 0.6590397357940674,
"learning_rate": 8.976476580425282e-05,
"loss": 0.0131,
"step": 4940
},
{
"epoch": 16.123778501628664,
"grad_norm": 0.5601251125335693,
"learning_rate": 8.971459296358606e-05,
"loss": 0.0121,
"step": 4950
},
{
"epoch": 16.156351791530945,
"grad_norm": 0.5221347212791443,
"learning_rate": 8.966431154473864e-05,
"loss": 0.0121,
"step": 4960
},
{
"epoch": 16.188925081433226,
"grad_norm": 0.57920902967453,
"learning_rate": 8.961392168517803e-05,
"loss": 0.0124,
"step": 4970
},
{
"epoch": 16.221498371335503,
"grad_norm": 0.5354865789413452,
"learning_rate": 8.956342352266821e-05,
"loss": 0.012,
"step": 4980
},
{
"epoch": 16.254071661237784,
"grad_norm": 0.618729293346405,
"learning_rate": 8.95128171952692e-05,
"loss": 0.0104,
"step": 4990
},
{
"epoch": 16.286644951140065,
"grad_norm": 0.42464056611061096,
"learning_rate": 8.946210284133676e-05,
"loss": 0.0114,
"step": 5000
},
{
"epoch": 16.319218241042346,
"grad_norm": 0.4815506041049957,
"learning_rate": 8.941128059952201e-05,
"loss": 0.01,
"step": 5010
},
{
"epoch": 16.351791530944624,
"grad_norm": 0.49649620056152344,
"learning_rate": 8.936035060877102e-05,
"loss": 0.0106,
"step": 5020
},
{
"epoch": 16.384364820846905,
"grad_norm": 0.5027061700820923,
"learning_rate": 8.930931300832443e-05,
"loss": 0.0104,
"step": 5030
},
{
"epoch": 16.416938110749186,
"grad_norm": 0.5375000834465027,
"learning_rate": 8.925816793771711e-05,
"loss": 0.0095,
"step": 5040
},
{
"epoch": 16.449511400651467,
"grad_norm": 0.4966764450073242,
"learning_rate": 8.92069155367777e-05,
"loss": 0.0114,
"step": 5050
},
{
"epoch": 16.482084690553744,
"grad_norm": 0.7109740376472473,
"learning_rate": 8.915555594562834e-05,
"loss": 0.0129,
"step": 5060
},
{
"epoch": 16.514657980456025,
"grad_norm": 0.5642775297164917,
"learning_rate": 8.910408930468416e-05,
"loss": 0.0102,
"step": 5070
},
{
"epoch": 16.547231270358306,
"grad_norm": 0.5686492323875427,
"learning_rate": 8.905251575465303e-05,
"loss": 0.0102,
"step": 5080
},
{
"epoch": 16.579804560260587,
"grad_norm": 0.5111770033836365,
"learning_rate": 8.900083543653502e-05,
"loss": 0.0101,
"step": 5090
},
{
"epoch": 16.612377850162865,
"grad_norm": 0.48311617970466614,
"learning_rate": 8.894904849162218e-05,
"loss": 0.0113,
"step": 5100
},
{
"epoch": 16.644951140065146,
"grad_norm": 0.4612133502960205,
"learning_rate": 8.889715506149802e-05,
"loss": 0.0097,
"step": 5110
},
{
"epoch": 16.677524429967427,
"grad_norm": 0.6265704035758972,
"learning_rate": 8.884515528803722e-05,
"loss": 0.0101,
"step": 5120
},
{
"epoch": 16.710097719869708,
"grad_norm": 0.5671060681343079,
"learning_rate": 8.879304931340517e-05,
"loss": 0.0098,
"step": 5130
},
{
"epoch": 16.742671009771986,
"grad_norm": 0.4591500461101532,
"learning_rate": 8.874083728005759e-05,
"loss": 0.0106,
"step": 5140
},
{
"epoch": 16.775244299674267,
"grad_norm": 0.7012014389038086,
"learning_rate": 8.868851933074021e-05,
"loss": 0.0115,
"step": 5150
},
{
"epoch": 16.807817589576548,
"grad_norm": 0.6138771176338196,
"learning_rate": 8.863609560848829e-05,
"loss": 0.0117,
"step": 5160
},
{
"epoch": 16.84039087947883,
"grad_norm": 0.5631691813468933,
"learning_rate": 8.85835662566263e-05,
"loss": 0.0096,
"step": 5170
},
{
"epoch": 16.872964169381106,
"grad_norm": 0.6066752076148987,
"learning_rate": 8.853093141876747e-05,
"loss": 0.0118,
"step": 5180
},
{
"epoch": 16.905537459283387,
"grad_norm": 0.3970053791999817,
"learning_rate": 8.847819123881343e-05,
"loss": 0.0103,
"step": 5190
},
{
"epoch": 16.938110749185668,
"grad_norm": 0.5872803330421448,
"learning_rate": 8.842534586095383e-05,
"loss": 0.0093,
"step": 5200
},
{
"epoch": 16.97068403908795,
"grad_norm": 0.5278239846229553,
"learning_rate": 8.837239542966593e-05,
"loss": 0.0116,
"step": 5210
},
{
"epoch": 17.003257328990227,
"grad_norm": 0.5052501559257507,
"learning_rate": 8.831934008971417e-05,
"loss": 0.0102,
"step": 5220
},
{
"epoch": 17.035830618892508,
"grad_norm": 0.5707674026489258,
"learning_rate": 8.826617998614982e-05,
"loss": 0.0085,
"step": 5230
},
{
"epoch": 17.06840390879479,
"grad_norm": 0.5154997706413269,
"learning_rate": 8.821291526431056e-05,
"loss": 0.0113,
"step": 5240
},
{
"epoch": 17.10097719869707,
"grad_norm": 0.4334968328475952,
"learning_rate": 8.815954606982015e-05,
"loss": 0.0112,
"step": 5250
},
{
"epoch": 17.133550488599347,
"grad_norm": 0.6048435568809509,
"learning_rate": 8.810607254858789e-05,
"loss": 0.0117,
"step": 5260
},
{
"epoch": 17.16612377850163,
"grad_norm": 0.4900098145008087,
"learning_rate": 8.805249484680838e-05,
"loss": 0.0098,
"step": 5270
},
{
"epoch": 17.19869706840391,
"grad_norm": 0.43989741802215576,
"learning_rate": 8.799881311096096e-05,
"loss": 0.0119,
"step": 5280
},
{
"epoch": 17.23127035830619,
"grad_norm": 0.4381119906902313,
"learning_rate": 8.794502748780949e-05,
"loss": 0.0098,
"step": 5290
},
{
"epoch": 17.263843648208468,
"grad_norm": 0.5834327340126038,
"learning_rate": 8.78911381244018e-05,
"loss": 0.0115,
"step": 5300
},
{
"epoch": 17.29641693811075,
"grad_norm": 0.43369826674461365,
"learning_rate": 8.783714516806933e-05,
"loss": 0.0093,
"step": 5310
},
{
"epoch": 17.32899022801303,
"grad_norm": 0.5850085020065308,
"learning_rate": 8.77830487664268e-05,
"loss": 0.0112,
"step": 5320
},
{
"epoch": 17.36156351791531,
"grad_norm": 0.5535557270050049,
"learning_rate": 8.772884906737167e-05,
"loss": 0.0117,
"step": 5330
},
{
"epoch": 17.39413680781759,
"grad_norm": 0.5345462560653687,
"learning_rate": 8.767454621908387e-05,
"loss": 0.0112,
"step": 5340
},
{
"epoch": 17.42671009771987,
"grad_norm": 0.5913922190666199,
"learning_rate": 8.76201403700253e-05,
"loss": 0.0118,
"step": 5350
},
{
"epoch": 17.45928338762215,
"grad_norm": 0.5394867658615112,
"learning_rate": 8.756563166893949e-05,
"loss": 0.0103,
"step": 5360
},
{
"epoch": 17.49185667752443,
"grad_norm": 0.637417197227478,
"learning_rate": 8.751102026485113e-05,
"loss": 0.0121,
"step": 5370
},
{
"epoch": 17.52442996742671,
"grad_norm": 0.5312429666519165,
"learning_rate": 8.745630630706571e-05,
"loss": 0.0112,
"step": 5380
},
{
"epoch": 17.55700325732899,
"grad_norm": 0.5353013873100281,
"learning_rate": 8.740148994516912e-05,
"loss": 0.0104,
"step": 5390
},
{
"epoch": 17.58957654723127,
"grad_norm": 0.5983322858810425,
"learning_rate": 8.73465713290272e-05,
"loss": 0.0091,
"step": 5400
},
{
"epoch": 17.622149837133552,
"grad_norm": 0.6481793522834778,
"learning_rate": 8.729155060878533e-05,
"loss": 0.0109,
"step": 5410
},
{
"epoch": 17.65472312703583,
"grad_norm": 0.51454097032547,
"learning_rate": 8.723642793486809e-05,
"loss": 0.0104,
"step": 5420
},
{
"epoch": 17.68729641693811,
"grad_norm": 0.6376858949661255,
"learning_rate": 8.718120345797873e-05,
"loss": 0.01,
"step": 5430
},
{
"epoch": 17.71986970684039,
"grad_norm": 0.5232966542243958,
"learning_rate": 8.712587732909889e-05,
"loss": 0.0104,
"step": 5440
},
{
"epoch": 17.752442996742673,
"grad_norm": 0.40592941641807556,
"learning_rate": 8.707044969948806e-05,
"loss": 0.0103,
"step": 5450
},
{
"epoch": 17.78501628664495,
"grad_norm": 0.5707107782363892,
"learning_rate": 8.701492072068329e-05,
"loss": 0.0097,
"step": 5460
},
{
"epoch": 17.81758957654723,
"grad_norm": 0.5730787515640259,
"learning_rate": 8.695929054449869e-05,
"loss": 0.0109,
"step": 5470
},
{
"epoch": 17.850162866449512,
"grad_norm": 0.5907419323921204,
"learning_rate": 8.690355932302501e-05,
"loss": 0.0106,
"step": 5480
},
{
"epoch": 17.88273615635179,
"grad_norm": 0.5448195934295654,
"learning_rate": 8.684772720862931e-05,
"loss": 0.0102,
"step": 5490
},
{
"epoch": 17.91530944625407,
"grad_norm": 0.5052299499511719,
"learning_rate": 8.679179435395446e-05,
"loss": 0.0102,
"step": 5500
},
{
"epoch": 17.94788273615635,
"grad_norm": 0.57944256067276,
"learning_rate": 8.673576091191874e-05,
"loss": 0.0115,
"step": 5510
},
{
"epoch": 17.980456026058633,
"grad_norm": 0.5408090949058533,
"learning_rate": 8.667962703571541e-05,
"loss": 0.0105,
"step": 5520
},
{
"epoch": 18.01302931596091,
"grad_norm": 0.5815545320510864,
"learning_rate": 8.662339287881238e-05,
"loss": 0.0098,
"step": 5530
},
{
"epoch": 18.04560260586319,
"grad_norm": 0.5216434001922607,
"learning_rate": 8.656705859495169e-05,
"loss": 0.0108,
"step": 5540
},
{
"epoch": 18.078175895765472,
"grad_norm": 0.41256457567214966,
"learning_rate": 8.651062433814912e-05,
"loss": 0.0096,
"step": 5550
},
{
"epoch": 18.110749185667753,
"grad_norm": 0.4612780809402466,
"learning_rate": 8.645409026269375e-05,
"loss": 0.0098,
"step": 5560
},
{
"epoch": 18.14332247557003,
"grad_norm": 0.6036962866783142,
"learning_rate": 8.639745652314759e-05,
"loss": 0.0109,
"step": 5570
},
{
"epoch": 18.175895765472312,
"grad_norm": 0.45757317543029785,
"learning_rate": 8.634072327434515e-05,
"loss": 0.0104,
"step": 5580
},
{
"epoch": 18.208469055374593,
"grad_norm": 0.5633235573768616,
"learning_rate": 8.628389067139294e-05,
"loss": 0.0105,
"step": 5590
},
{
"epoch": 18.241042345276874,
"grad_norm": 0.48329007625579834,
"learning_rate": 8.622695886966911e-05,
"loss": 0.01,
"step": 5600
},
{
"epoch": 18.27361563517915,
"grad_norm": 0.4269546568393707,
"learning_rate": 8.616992802482308e-05,
"loss": 0.0123,
"step": 5610
},
{
"epoch": 18.306188925081432,
"grad_norm": 0.4761641323566437,
"learning_rate": 8.611279829277496e-05,
"loss": 0.0112,
"step": 5620
},
{
"epoch": 18.338762214983714,
"grad_norm": 0.5688640475273132,
"learning_rate": 8.605556982971528e-05,
"loss": 0.011,
"step": 5630
},
{
"epoch": 18.371335504885995,
"grad_norm": 0.5268420577049255,
"learning_rate": 8.599824279210447e-05,
"loss": 0.0111,
"step": 5640
},
{
"epoch": 18.403908794788272,
"grad_norm": 0.5499677658081055,
"learning_rate": 8.594081733667243e-05,
"loss": 0.0087,
"step": 5650
},
{
"epoch": 18.436482084690553,
"grad_norm": 0.592589259147644,
"learning_rate": 8.58832936204182e-05,
"loss": 0.0095,
"step": 5660
},
{
"epoch": 18.469055374592834,
"grad_norm": 0.4770977795124054,
"learning_rate": 8.582567180060942e-05,
"loss": 0.0099,
"step": 5670
},
{
"epoch": 18.501628664495115,
"grad_norm": 0.5618078708648682,
"learning_rate": 8.576795203478194e-05,
"loss": 0.0107,
"step": 5680
},
{
"epoch": 18.534201954397393,
"grad_norm": 0.46116217970848083,
"learning_rate": 8.571013448073939e-05,
"loss": 0.0114,
"step": 5690
},
{
"epoch": 18.566775244299674,
"grad_norm": 0.4571772515773773,
"learning_rate": 8.565221929655275e-05,
"loss": 0.0102,
"step": 5700
},
{
"epoch": 18.599348534201955,
"grad_norm": 0.4506760537624359,
"learning_rate": 8.559420664055992e-05,
"loss": 0.0114,
"step": 5710
},
{
"epoch": 18.631921824104236,
"grad_norm": 0.487032413482666,
"learning_rate": 8.553609667136532e-05,
"loss": 0.0098,
"step": 5720
},
{
"epoch": 18.664495114006513,
"grad_norm": 0.40095847845077515,
"learning_rate": 8.547788954783936e-05,
"loss": 0.0093,
"step": 5730
},
{
"epoch": 18.697068403908794,
"grad_norm": 0.43920227885246277,
"learning_rate": 8.541958542911808e-05,
"loss": 0.0108,
"step": 5740
},
{
"epoch": 18.729641693811075,
"grad_norm": 0.45471495389938354,
"learning_rate": 8.536118447460275e-05,
"loss": 0.0085,
"step": 5750
},
{
"epoch": 18.762214983713356,
"grad_norm": 0.3742855191230774,
"learning_rate": 8.530268684395932e-05,
"loss": 0.0098,
"step": 5760
},
{
"epoch": 18.794788273615634,
"grad_norm": 0.5401120781898499,
"learning_rate": 8.524409269711807e-05,
"loss": 0.011,
"step": 5770
},
{
"epoch": 18.827361563517915,
"grad_norm": 0.4368399679660797,
"learning_rate": 8.51854021942732e-05,
"loss": 0.0117,
"step": 5780
},
{
"epoch": 18.859934853420196,
"grad_norm": 0.5863309502601624,
"learning_rate": 8.512661549588227e-05,
"loss": 0.0108,
"step": 5790
},
{
"epoch": 18.892508143322477,
"grad_norm": 0.4833745062351227,
"learning_rate": 8.506773276266588e-05,
"loss": 0.011,
"step": 5800
},
{
"epoch": 18.925081433224754,
"grad_norm": 0.5938926339149475,
"learning_rate": 8.500875415560721e-05,
"loss": 0.0107,
"step": 5810
},
{
"epoch": 18.957654723127035,
"grad_norm": 0.47496920824050903,
"learning_rate": 8.494967983595144e-05,
"loss": 0.0128,
"step": 5820
},
{
"epoch": 18.990228013029316,
"grad_norm": 0.43583986163139343,
"learning_rate": 8.489050996520558e-05,
"loss": 0.0116,
"step": 5830
},
{
"epoch": 19.022801302931597,
"grad_norm": 0.4981800317764282,
"learning_rate": 8.483124470513775e-05,
"loss": 0.0097,
"step": 5840
},
{
"epoch": 19.055374592833875,
"grad_norm": 0.5075828433036804,
"learning_rate": 8.477188421777692e-05,
"loss": 0.0094,
"step": 5850
},
{
"epoch": 19.087947882736156,
"grad_norm": 0.49366408586502075,
"learning_rate": 8.47124286654124e-05,
"loss": 0.0113,
"step": 5860
},
{
"epoch": 19.120521172638437,
"grad_norm": 0.6152781844139099,
"learning_rate": 8.465287821059341e-05,
"loss": 0.01,
"step": 5870
},
{
"epoch": 19.153094462540718,
"grad_norm": 0.5412830710411072,
"learning_rate": 8.45932330161286e-05,
"loss": 0.0107,
"step": 5880
},
{
"epoch": 19.185667752442995,
"grad_norm": 0.46815627813339233,
"learning_rate": 8.453349324508567e-05,
"loss": 0.0104,
"step": 5890
},
{
"epoch": 19.218241042345277,
"grad_norm": 0.43253058195114136,
"learning_rate": 8.447365906079088e-05,
"loss": 0.0093,
"step": 5900
},
{
"epoch": 19.250814332247558,
"grad_norm": 0.40782538056373596,
"learning_rate": 8.441373062682856e-05,
"loss": 0.0104,
"step": 5910
},
{
"epoch": 19.28338762214984,
"grad_norm": 0.4392983317375183,
"learning_rate": 8.43537081070408e-05,
"loss": 0.0103,
"step": 5920
},
{
"epoch": 19.315960912052116,
"grad_norm": 0.47564175724983215,
"learning_rate": 8.429359166552689e-05,
"loss": 0.011,
"step": 5930
},
{
"epoch": 19.348534201954397,
"grad_norm": 0.4908657670021057,
"learning_rate": 8.423338146664284e-05,
"loss": 0.0112,
"step": 5940
},
{
"epoch": 19.381107491856678,
"grad_norm": 0.4445395767688751,
"learning_rate": 8.417307767500107e-05,
"loss": 0.0104,
"step": 5950
},
{
"epoch": 19.41368078175896,
"grad_norm": 0.38990992307662964,
"learning_rate": 8.411268045546983e-05,
"loss": 0.0104,
"step": 5960
},
{
"epoch": 19.446254071661237,
"grad_norm": 0.5295190811157227,
"learning_rate": 8.405218997317281e-05,
"loss": 0.0113,
"step": 5970
},
{
"epoch": 19.478827361563518,
"grad_norm": 0.5957738757133484,
"learning_rate": 8.399160639348869e-05,
"loss": 0.0136,
"step": 5980
},
{
"epoch": 19.5114006514658,
"grad_norm": 0.5141053199768066,
"learning_rate": 8.393092988205065e-05,
"loss": 0.0107,
"step": 5990
},
{
"epoch": 19.54397394136808,
"grad_norm": 0.5684329271316528,
"learning_rate": 8.387016060474597e-05,
"loss": 0.0105,
"step": 6000
}
],
"logging_steps": 10,
"max_steps": 20000,
"num_input_tokens_seen": 0,
"num_train_epochs": 66,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}