| { | |
| "best_global_step": 8800, | |
| "best_metric": 0.8558670305135202, | |
| "best_model_checkpoint": "./distilbert-feedback/checkpoint-8800", | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 9070, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011025358324145534, | |
| "grad_norm": 6.280033111572266, | |
| "learning_rate": 4.972987872105844e-05, | |
| "loss": 1.7001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.022050716648291068, | |
| "grad_norm": 6.537757873535156, | |
| "learning_rate": 4.94542447629548e-05, | |
| "loss": 1.3668, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03307607497243661, | |
| "grad_norm": 6.568482875823975, | |
| "learning_rate": 4.917861080485116e-05, | |
| "loss": 1.2012, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.044101433296582136, | |
| "grad_norm": 7.553868293762207, | |
| "learning_rate": 4.890297684674752e-05, | |
| "loss": 1.2455, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.044101433296582136, | |
| "eval_accuracy": 0.5194740759116845, | |
| "eval_loss": 1.2424163818359375, | |
| "eval_runtime": 44.855, | |
| "eval_samples_per_second": 89.867, | |
| "eval_steps_per_second": 11.236, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05512679162072767, | |
| "grad_norm": 9.827457427978516, | |
| "learning_rate": 4.8627342888643884e-05, | |
| "loss": 1.1816, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06615214994487321, | |
| "grad_norm": 9.297719955444336, | |
| "learning_rate": 4.835170893054025e-05, | |
| "loss": 1.1326, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07717750826901874, | |
| "grad_norm": 7.382741451263428, | |
| "learning_rate": 4.807607497243661e-05, | |
| "loss": 1.061, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08820286659316427, | |
| "grad_norm": 7.543405532836914, | |
| "learning_rate": 4.780044101433297e-05, | |
| "loss": 1.0238, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.08820286659316427, | |
| "eval_accuracy": 0.6350781443810469, | |
| "eval_loss": 0.9708635210990906, | |
| "eval_runtime": 45.4434, | |
| "eval_samples_per_second": 88.704, | |
| "eval_steps_per_second": 11.091, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.09922822491730982, | |
| "grad_norm": 21.24835968017578, | |
| "learning_rate": 4.752480705622933e-05, | |
| "loss": 0.9726, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11025358324145534, | |
| "grad_norm": 11.24039077758789, | |
| "learning_rate": 4.724917309812569e-05, | |
| "loss": 1.0019, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12127894156560089, | |
| "grad_norm": 14.267502784729004, | |
| "learning_rate": 4.6973539140022054e-05, | |
| "loss": 0.9928, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.13230429988974643, | |
| "grad_norm": 11.773629188537598, | |
| "learning_rate": 4.669790518191842e-05, | |
| "loss": 0.8424, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.13230429988974643, | |
| "eval_accuracy": 0.6789878442073927, | |
| "eval_loss": 0.8846617937088013, | |
| "eval_runtime": 45.2393, | |
| "eval_samples_per_second": 89.104, | |
| "eval_steps_per_second": 11.141, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.14332965821389196, | |
| "grad_norm": 4.492119312286377, | |
| "learning_rate": 4.642227122381478e-05, | |
| "loss": 0.9987, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.1543550165380375, | |
| "grad_norm": 8.530481338500977, | |
| "learning_rate": 4.614663726571114e-05, | |
| "loss": 0.8802, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.16538037486218302, | |
| "grad_norm": 15.022302627563477, | |
| "learning_rate": 4.58710033076075e-05, | |
| "loss": 0.8428, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.17640573318632854, | |
| "grad_norm": 11.992727279663086, | |
| "learning_rate": 4.559536934950386e-05, | |
| "loss": 0.9022, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.17640573318632854, | |
| "eval_accuracy": 0.6566608781939965, | |
| "eval_loss": 0.9172375202178955, | |
| "eval_runtime": 46.9407, | |
| "eval_samples_per_second": 85.874, | |
| "eval_steps_per_second": 10.737, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.1874310915104741, | |
| "grad_norm": 10.07384204864502, | |
| "learning_rate": 4.5319735391400225e-05, | |
| "loss": 0.8949, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.19845644983461963, | |
| "grad_norm": 7.739764213562012, | |
| "learning_rate": 4.504410143329658e-05, | |
| "loss": 0.8179, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.20948180815876516, | |
| "grad_norm": 13.591370582580566, | |
| "learning_rate": 4.4768467475192944e-05, | |
| "loss": 0.8201, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2205071664829107, | |
| "grad_norm": 10.648216247558594, | |
| "learning_rate": 4.449283351708931e-05, | |
| "loss": 0.8128, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2205071664829107, | |
| "eval_accuracy": 0.7142148350285289, | |
| "eval_loss": 0.7965238690376282, | |
| "eval_runtime": 48.533, | |
| "eval_samples_per_second": 83.057, | |
| "eval_steps_per_second": 10.385, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23153252480705622, | |
| "grad_norm": 8.068305015563965, | |
| "learning_rate": 4.421719955898567e-05, | |
| "loss": 0.8249, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.24255788313120177, | |
| "grad_norm": 6.480292797088623, | |
| "learning_rate": 4.3941565600882026e-05, | |
| "loss": 0.7703, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2535832414553473, | |
| "grad_norm": 7.40410041809082, | |
| "learning_rate": 4.366593164277839e-05, | |
| "loss": 0.8165, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.26460859977949286, | |
| "grad_norm": 13.612632751464844, | |
| "learning_rate": 4.339029768467475e-05, | |
| "loss": 0.8101, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.26460859977949286, | |
| "eval_accuracy": 0.6998263458198958, | |
| "eval_loss": 0.7793557643890381, | |
| "eval_runtime": 47.2524, | |
| "eval_samples_per_second": 85.308, | |
| "eval_steps_per_second": 10.666, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.2756339581036384, | |
| "grad_norm": 11.233166694641113, | |
| "learning_rate": 4.3114663726571114e-05, | |
| "loss": 0.7948, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.2866593164277839, | |
| "grad_norm": 12.69255256652832, | |
| "learning_rate": 4.283902976846748e-05, | |
| "loss": 0.7261, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.29768467475192945, | |
| "grad_norm": 8.976383209228516, | |
| "learning_rate": 4.256339581036384e-05, | |
| "loss": 0.7946, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.308710033076075, | |
| "grad_norm": 8.328288078308105, | |
| "learning_rate": 4.2287761852260196e-05, | |
| "loss": 0.7378, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.308710033076075, | |
| "eval_accuracy": 0.7050359712230215, | |
| "eval_loss": 0.7756134867668152, | |
| "eval_runtime": 48.2156, | |
| "eval_samples_per_second": 83.604, | |
| "eval_steps_per_second": 10.453, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3197353914002205, | |
| "grad_norm": 9.693521499633789, | |
| "learning_rate": 4.201212789415656e-05, | |
| "loss": 0.7439, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.33076074972436603, | |
| "grad_norm": 9.462974548339844, | |
| "learning_rate": 4.173649393605292e-05, | |
| "loss": 0.7106, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.34178610804851156, | |
| "grad_norm": 11.551827430725098, | |
| "learning_rate": 4.1460859977949285e-05, | |
| "loss": 0.6444, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.3528114663726571, | |
| "grad_norm": 8.0066499710083, | |
| "learning_rate": 4.118522601984565e-05, | |
| "loss": 0.7226, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.3528114663726571, | |
| "eval_accuracy": 0.7271148598362689, | |
| "eval_loss": 0.7424366474151611, | |
| "eval_runtime": 49.1199, | |
| "eval_samples_per_second": 82.064, | |
| "eval_steps_per_second": 10.261, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.3638368246968026, | |
| "grad_norm": 11.164033889770508, | |
| "learning_rate": 4.090959206174201e-05, | |
| "loss": 0.6636, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.3748621830209482, | |
| "grad_norm": 9.30005168914795, | |
| "learning_rate": 4.063395810363837e-05, | |
| "loss": 0.7728, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.38588754134509373, | |
| "grad_norm": 8.798018455505371, | |
| "learning_rate": 4.035832414553473e-05, | |
| "loss": 0.755, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.39691289966923926, | |
| "grad_norm": 17.514118194580078, | |
| "learning_rate": 4.008269018743109e-05, | |
| "loss": 0.7246, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.39691289966923926, | |
| "eval_accuracy": 0.7472091292483255, | |
| "eval_loss": 0.6997360587120056, | |
| "eval_runtime": 50.8367, | |
| "eval_samples_per_second": 79.293, | |
| "eval_steps_per_second": 9.914, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.4079382579933848, | |
| "grad_norm": 19.774938583374023, | |
| "learning_rate": 3.9807056229327455e-05, | |
| "loss": 0.6325, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4189636163175303, | |
| "grad_norm": 11.262466430664062, | |
| "learning_rate": 3.953142227122382e-05, | |
| "loss": 0.7494, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.42998897464167585, | |
| "grad_norm": 6.449489593505859, | |
| "learning_rate": 3.925578831312018e-05, | |
| "loss": 0.6959, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.4410143329658214, | |
| "grad_norm": 11.40916633605957, | |
| "learning_rate": 3.898015435501654e-05, | |
| "loss": 0.7348, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4410143329658214, | |
| "eval_accuracy": 0.7591168444554701, | |
| "eval_loss": 0.6443637013435364, | |
| "eval_runtime": 49.1885, | |
| "eval_samples_per_second": 81.95, | |
| "eval_steps_per_second": 10.246, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4520396912899669, | |
| "grad_norm": 8.649637222290039, | |
| "learning_rate": 3.87045203969129e-05, | |
| "loss": 0.7338, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.46306504961411243, | |
| "grad_norm": 10.41679573059082, | |
| "learning_rate": 3.842888643880926e-05, | |
| "loss": 0.6482, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.474090407938258, | |
| "grad_norm": 12.902926445007324, | |
| "learning_rate": 3.8153252480705626e-05, | |
| "loss": 0.6849, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.48511576626240355, | |
| "grad_norm": 10.325972557067871, | |
| "learning_rate": 3.787761852260199e-05, | |
| "loss": 0.7373, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.48511576626240355, | |
| "eval_accuracy": 0.7578764574547259, | |
| "eval_loss": 0.6530741453170776, | |
| "eval_runtime": 50.2869, | |
| "eval_samples_per_second": 80.16, | |
| "eval_steps_per_second": 10.022, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.4961411245865491, | |
| "grad_norm": 7.638971328735352, | |
| "learning_rate": 3.760198456449835e-05, | |
| "loss": 0.5985, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5071664829106945, | |
| "grad_norm": 6.052102565765381, | |
| "learning_rate": 3.7326350606394714e-05, | |
| "loss": 0.6742, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5181918412348401, | |
| "grad_norm": 19.927108764648438, | |
| "learning_rate": 3.705071664829107e-05, | |
| "loss": 0.5846, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.5292171995589857, | |
| "grad_norm": 11.63871955871582, | |
| "learning_rate": 3.677508269018743e-05, | |
| "loss": 0.6943, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.5292171995589857, | |
| "eval_accuracy": 0.7568841478541305, | |
| "eval_loss": 0.6658030152320862, | |
| "eval_runtime": 51.4998, | |
| "eval_samples_per_second": 78.272, | |
| "eval_steps_per_second": 9.786, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.5402425578831312, | |
| "grad_norm": 10.362872123718262, | |
| "learning_rate": 3.6499448732083796e-05, | |
| "loss": 0.6303, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.5512679162072768, | |
| "grad_norm": 6.9453606605529785, | |
| "learning_rate": 3.622381477398016e-05, | |
| "loss": 0.6494, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5622932745314223, | |
| "grad_norm": 22.353281021118164, | |
| "learning_rate": 3.594818081587652e-05, | |
| "loss": 0.6271, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.5733186328555678, | |
| "grad_norm": 4.380390644073486, | |
| "learning_rate": 3.5672546857772885e-05, | |
| "loss": 0.6887, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.5733186328555678, | |
| "eval_accuracy": 0.7412552716447531, | |
| "eval_loss": 0.7230509519577026, | |
| "eval_runtime": 51.9788, | |
| "eval_samples_per_second": 77.551, | |
| "eval_steps_per_second": 9.696, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.5843439911797134, | |
| "grad_norm": 15.384754180908203, | |
| "learning_rate": 3.539691289966924e-05, | |
| "loss": 0.71, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.5953693495038589, | |
| "grad_norm": 7.5070390701293945, | |
| "learning_rate": 3.5121278941565604e-05, | |
| "loss": 0.6521, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6063947078280044, | |
| "grad_norm": 9.219125747680664, | |
| "learning_rate": 3.484564498346197e-05, | |
| "loss": 0.6781, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.61742006615215, | |
| "grad_norm": 9.403863906860352, | |
| "learning_rate": 3.457001102535832e-05, | |
| "loss": 0.6017, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.61742006615215, | |
| "eval_accuracy": 0.759364921855619, | |
| "eval_loss": 0.6455864906311035, | |
| "eval_runtime": 51.1726, | |
| "eval_samples_per_second": 78.773, | |
| "eval_steps_per_second": 9.849, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.6284454244762955, | |
| "grad_norm": 12.217251777648926, | |
| "learning_rate": 3.4294377067254686e-05, | |
| "loss": 0.6596, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.639470782800441, | |
| "grad_norm": 12.6351957321167, | |
| "learning_rate": 3.401874310915105e-05, | |
| "loss": 0.7351, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.6504961411245865, | |
| "grad_norm": 16.1507568359375, | |
| "learning_rate": 3.374310915104741e-05, | |
| "loss": 0.6256, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.6615214994487321, | |
| "grad_norm": 12.999460220336914, | |
| "learning_rate": 3.346747519294377e-05, | |
| "loss": 0.6082, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6615214994487321, | |
| "eval_accuracy": 0.7749937980649962, | |
| "eval_loss": 0.6113554835319519, | |
| "eval_runtime": 51.8574, | |
| "eval_samples_per_second": 77.732, | |
| "eval_steps_per_second": 9.719, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6725468577728776, | |
| "grad_norm": 12.123536109924316, | |
| "learning_rate": 3.319184123484013e-05, | |
| "loss": 0.5792, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.6835722160970231, | |
| "grad_norm": 13.215846061706543, | |
| "learning_rate": 3.291620727673649e-05, | |
| "loss": 0.6383, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.6945975744211687, | |
| "grad_norm": 18.97852325439453, | |
| "learning_rate": 3.2640573318632856e-05, | |
| "loss": 0.616, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.7056229327453142, | |
| "grad_norm": 13.575655937194824, | |
| "learning_rate": 3.236493936052922e-05, | |
| "loss": 0.4946, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.7056229327453142, | |
| "eval_accuracy": 0.7752418754651451, | |
| "eval_loss": 0.6550247669219971, | |
| "eval_runtime": 51.001, | |
| "eval_samples_per_second": 79.038, | |
| "eval_steps_per_second": 9.882, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.7166482910694597, | |
| "grad_norm": 13.980043411254883, | |
| "learning_rate": 3.208930540242558e-05, | |
| "loss": 0.7087, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.7276736493936052, | |
| "grad_norm": 16.2875919342041, | |
| "learning_rate": 3.181367144432194e-05, | |
| "loss": 0.6054, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.7386990077177509, | |
| "grad_norm": 4.533209323883057, | |
| "learning_rate": 3.15380374862183e-05, | |
| "loss": 0.5812, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.7497243660418964, | |
| "grad_norm": 9.643322944641113, | |
| "learning_rate": 3.1262403528114664e-05, | |
| "loss": 0.6084, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.7497243660418964, | |
| "eval_accuracy": 0.759364921855619, | |
| "eval_loss": 0.6849791407585144, | |
| "eval_runtime": 51.3096, | |
| "eval_samples_per_second": 78.562, | |
| "eval_steps_per_second": 9.823, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.7607497243660419, | |
| "grad_norm": 12.94963550567627, | |
| "learning_rate": 3.0986769570011027e-05, | |
| "loss": 0.5558, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.7717750826901875, | |
| "grad_norm": 11.617074966430664, | |
| "learning_rate": 3.071113561190739e-05, | |
| "loss": 0.6267, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.782800441014333, | |
| "grad_norm": 11.406350135803223, | |
| "learning_rate": 3.043550165380375e-05, | |
| "loss": 0.5174, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.7938257993384785, | |
| "grad_norm": 6.776385307312012, | |
| "learning_rate": 3.0159867695700112e-05, | |
| "loss": 0.71, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.7938257993384785, | |
| "eval_accuracy": 0.7928553708757132, | |
| "eval_loss": 0.578441858291626, | |
| "eval_runtime": 53.7733, | |
| "eval_samples_per_second": 74.963, | |
| "eval_steps_per_second": 9.373, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.804851157662624, | |
| "grad_norm": 21.484195709228516, | |
| "learning_rate": 2.9884233737596475e-05, | |
| "loss": 0.5633, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.8158765159867696, | |
| "grad_norm": 19.93082046508789, | |
| "learning_rate": 2.9608599779492834e-05, | |
| "loss": 0.5926, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.8269018743109151, | |
| "grad_norm": 13.265941619873047, | |
| "learning_rate": 2.9332965821389197e-05, | |
| "loss": 0.6305, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.8379272326350606, | |
| "grad_norm": 42.45393371582031, | |
| "learning_rate": 2.905733186328556e-05, | |
| "loss": 0.5575, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.8379272326350606, | |
| "eval_accuracy": 0.8020342346812206, | |
| "eval_loss": 0.5739887356758118, | |
| "eval_runtime": 55.181, | |
| "eval_samples_per_second": 73.051, | |
| "eval_steps_per_second": 9.134, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.8489525909592062, | |
| "grad_norm": 10.487258911132812, | |
| "learning_rate": 2.878169790518192e-05, | |
| "loss": 0.6158, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.8599779492833517, | |
| "grad_norm": 8.130553245544434, | |
| "learning_rate": 2.8506063947078282e-05, | |
| "loss": 0.6652, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.8710033076074972, | |
| "grad_norm": 13.992973327636719, | |
| "learning_rate": 2.8230429988974645e-05, | |
| "loss": 0.5247, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.8820286659316428, | |
| "grad_norm": 15.441995620727539, | |
| "learning_rate": 2.7954796030871005e-05, | |
| "loss": 0.6478, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8820286659316428, | |
| "eval_accuracy": 0.8144381046886628, | |
| "eval_loss": 0.524382472038269, | |
| "eval_runtime": 52.1097, | |
| "eval_samples_per_second": 77.356, | |
| "eval_steps_per_second": 9.672, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8930540242557883, | |
| "grad_norm": 18.06987762451172, | |
| "learning_rate": 2.7679162072767367e-05, | |
| "loss": 0.5071, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.9040793825799338, | |
| "grad_norm": 7.0000505447387695, | |
| "learning_rate": 2.740352811466373e-05, | |
| "loss": 0.5079, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.9151047409040793, | |
| "grad_norm": 11.618234634399414, | |
| "learning_rate": 2.712789415656009e-05, | |
| "loss": 0.5333, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.9261300992282249, | |
| "grad_norm": 29.748830795288086, | |
| "learning_rate": 2.6852260198456453e-05, | |
| "loss": 0.5765, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.9261300992282249, | |
| "eval_accuracy": 0.8196477300917886, | |
| "eval_loss": 0.537652850151062, | |
| "eval_runtime": 53.3275, | |
| "eval_samples_per_second": 75.59, | |
| "eval_steps_per_second": 9.451, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.9371554575523704, | |
| "grad_norm": 35.87691116333008, | |
| "learning_rate": 2.6576626240352816e-05, | |
| "loss": 0.6954, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.948180815876516, | |
| "grad_norm": 9.353912353515625, | |
| "learning_rate": 2.6300992282249175e-05, | |
| "loss": 0.5576, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.9592061742006616, | |
| "grad_norm": 26.52071189880371, | |
| "learning_rate": 2.6025358324145538e-05, | |
| "loss": 0.6409, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.9702315325248071, | |
| "grad_norm": 15.028057098388672, | |
| "learning_rate": 2.57497243660419e-05, | |
| "loss": 0.5363, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.9702315325248071, | |
| "eval_accuracy": 0.8022823120813694, | |
| "eval_loss": 0.5548905730247498, | |
| "eval_runtime": 51.9595, | |
| "eval_samples_per_second": 77.58, | |
| "eval_steps_per_second": 9.7, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.9812568908489526, | |
| "grad_norm": 12.458972930908203, | |
| "learning_rate": 2.547409040793826e-05, | |
| "loss": 0.551, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.9922822491730982, | |
| "grad_norm": 13.601612091064453, | |
| "learning_rate": 2.5198456449834623e-05, | |
| "loss": 0.4871, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.0033076074972436, | |
| "grad_norm": 4.1361870765686035, | |
| "learning_rate": 2.4922822491730983e-05, | |
| "loss": 0.4908, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.014332965821389, | |
| "grad_norm": 7.244948863983154, | |
| "learning_rate": 2.4647188533627346e-05, | |
| "loss": 0.4169, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.014332965821389, | |
| "eval_accuracy": 0.8005457702803275, | |
| "eval_loss": 0.5810744166374207, | |
| "eval_runtime": 44.6894, | |
| "eval_samples_per_second": 90.2, | |
| "eval_steps_per_second": 11.278, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.0253583241455346, | |
| "grad_norm": 1.200239896774292, | |
| "learning_rate": 2.4371554575523705e-05, | |
| "loss": 0.4883, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.0363836824696802, | |
| "grad_norm": 6.8026251792907715, | |
| "learning_rate": 2.4095920617420068e-05, | |
| "loss": 0.443, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.0474090407938257, | |
| "grad_norm": 12.317420959472656, | |
| "learning_rate": 2.382028665931643e-05, | |
| "loss": 0.5633, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 1.0584343991179714, | |
| "grad_norm": 12.111879348754883, | |
| "learning_rate": 2.354465270121279e-05, | |
| "loss": 0.4637, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.0584343991179714, | |
| "eval_accuracy": 0.820640039692384, | |
| "eval_loss": 0.530491292476654, | |
| "eval_runtime": 44.799, | |
| "eval_samples_per_second": 89.98, | |
| "eval_steps_per_second": 11.25, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.069459757442117, | |
| "grad_norm": 7.548985004425049, | |
| "learning_rate": 2.3269018743109153e-05, | |
| "loss": 0.4133, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 1.0804851157662625, | |
| "grad_norm": 7.3979997634887695, | |
| "learning_rate": 2.2993384785005516e-05, | |
| "loss": 0.4151, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.091510474090408, | |
| "grad_norm": 11.04232120513916, | |
| "learning_rate": 2.2717750826901875e-05, | |
| "loss": 0.4044, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 1.1025358324145536, | |
| "grad_norm": 9.042499542236328, | |
| "learning_rate": 2.244211686879824e-05, | |
| "loss": 0.5072, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.1025358324145536, | |
| "eval_accuracy": 0.7878938228727363, | |
| "eval_loss": 0.6346498727798462, | |
| "eval_runtime": 46.0611, | |
| "eval_samples_per_second": 87.514, | |
| "eval_steps_per_second": 10.942, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.113561190738699, | |
| "grad_norm": 11.084900856018066, | |
| "learning_rate": 2.21664829106946e-05, | |
| "loss": 0.415, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 1.1245865490628446, | |
| "grad_norm": 12.728482246398926, | |
| "learning_rate": 2.189084895259096e-05, | |
| "loss": 0.4814, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.1356119073869901, | |
| "grad_norm": 4.183909893035889, | |
| "learning_rate": 2.161521499448732e-05, | |
| "loss": 0.417, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 1.1466372657111357, | |
| "grad_norm": 4.626537322998047, | |
| "learning_rate": 2.1339581036383683e-05, | |
| "loss": 0.5391, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.1466372657111357, | |
| "eval_accuracy": 0.8196477300917886, | |
| "eval_loss": 0.5200665593147278, | |
| "eval_runtime": 51.4326, | |
| "eval_samples_per_second": 78.374, | |
| "eval_steps_per_second": 9.799, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.1576626240352812, | |
| "grad_norm": 2.050447463989258, | |
| "learning_rate": 2.1063947078280043e-05, | |
| "loss": 0.4601, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 1.1686879823594267, | |
| "grad_norm": 4.204873561859131, | |
| "learning_rate": 2.0788313120176405e-05, | |
| "loss": 0.3647, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.1797133406835723, | |
| "grad_norm": 4.201039791107178, | |
| "learning_rate": 2.051267916207277e-05, | |
| "loss": 0.4608, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 1.1907386990077178, | |
| "grad_norm": 26.340166091918945, | |
| "learning_rate": 2.0237045203969128e-05, | |
| "loss": 0.4807, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.1907386990077178, | |
| "eval_accuracy": 0.8007938476804763, | |
| "eval_loss": 0.6092058420181274, | |
| "eval_runtime": 48.0907, | |
| "eval_samples_per_second": 83.821, | |
| "eval_steps_per_second": 10.48, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.2017640573318633, | |
| "grad_norm": 3.7308127880096436, | |
| "learning_rate": 1.996141124586549e-05, | |
| "loss": 0.4879, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.2127894156560088, | |
| "grad_norm": 2.054511785507202, | |
| "learning_rate": 1.9685777287761854e-05, | |
| "loss": 0.4749, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.2238147739801544, | |
| "grad_norm": 21.566608428955078, | |
| "learning_rate": 1.9410143329658213e-05, | |
| "loss": 0.4921, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.2348401323043, | |
| "grad_norm": 12.240718841552734, | |
| "learning_rate": 1.9134509371554576e-05, | |
| "loss": 0.3721, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.2348401323043, | |
| "eval_accuracy": 0.8248573554949145, | |
| "eval_loss": 0.5377492904663086, | |
| "eval_runtime": 50.1111, | |
| "eval_samples_per_second": 80.441, | |
| "eval_steps_per_second": 10.058, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.2458654906284454, | |
| "grad_norm": 25.768829345703125, | |
| "learning_rate": 1.885887541345094e-05, | |
| "loss": 0.4781, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.256890848952591, | |
| "grad_norm": 34.470481872558594, | |
| "learning_rate": 1.85832414553473e-05, | |
| "loss": 0.5108, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.2679162072767365, | |
| "grad_norm": 12.565007209777832, | |
| "learning_rate": 1.830760749724366e-05, | |
| "loss": 0.4817, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.278941565600882, | |
| "grad_norm": 15.793746948242188, | |
| "learning_rate": 1.8031973539140024e-05, | |
| "loss": 0.4601, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.278941565600882, | |
| "eval_accuracy": 0.8226246588935748, | |
| "eval_loss": 0.5374011993408203, | |
| "eval_runtime": 55.1619, | |
| "eval_samples_per_second": 73.076, | |
| "eval_steps_per_second": 9.137, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.2899669239250275, | |
| "grad_norm": 7.197420597076416, | |
| "learning_rate": 1.7756339581036387e-05, | |
| "loss": 0.3642, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.300992282249173, | |
| "grad_norm": 16.694839477539062, | |
| "learning_rate": 1.7480705622932746e-05, | |
| "loss": 0.4722, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.3120176405733186, | |
| "grad_norm": 4.159826278686523, | |
| "learning_rate": 1.720507166482911e-05, | |
| "loss": 0.3647, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.3230429988974641, | |
| "grad_norm": 3.6812713146209717, | |
| "learning_rate": 1.6929437706725472e-05, | |
| "loss": 0.4906, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.3230429988974641, | |
| "eval_accuracy": 0.8322996774993798, | |
| "eval_loss": 0.5462606549263, | |
| "eval_runtime": 50.2107, | |
| "eval_samples_per_second": 80.282, | |
| "eval_steps_per_second": 10.038, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.3340683572216097, | |
| "grad_norm": 16.1009521484375, | |
| "learning_rate": 1.665380374862183e-05, | |
| "loss": 0.5013, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.3450937155457552, | |
| "grad_norm": 9.130171775817871, | |
| "learning_rate": 1.637816979051819e-05, | |
| "loss": 0.5111, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.3561190738699007, | |
| "grad_norm": 7.678464412689209, | |
| "learning_rate": 1.6102535832414554e-05, | |
| "loss": 0.4683, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.3671444321940462, | |
| "grad_norm": 14.63408088684082, | |
| "learning_rate": 1.5826901874310913e-05, | |
| "loss": 0.4328, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.3671444321940462, | |
| "eval_accuracy": 0.8322996774993798, | |
| "eval_loss": 0.5245828032493591, | |
| "eval_runtime": 51.9955, | |
| "eval_samples_per_second": 77.526, | |
| "eval_steps_per_second": 9.693, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.3781697905181918, | |
| "grad_norm": 35.199092864990234, | |
| "learning_rate": 1.5551267916207276e-05, | |
| "loss": 0.4871, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.3891951488423373, | |
| "grad_norm": 12.516600608825684, | |
| "learning_rate": 1.527563395810364e-05, | |
| "loss": 0.4812, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.4002205071664828, | |
| "grad_norm": 12.948488235473633, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.4149, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.4112458654906286, | |
| "grad_norm": 17.36246681213379, | |
| "learning_rate": 1.4724366041896362e-05, | |
| "loss": 0.3668, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.4112458654906286, | |
| "eval_accuracy": 0.8419746961051848, | |
| "eval_loss": 0.5143093466758728, | |
| "eval_runtime": 51.9669, | |
| "eval_samples_per_second": 77.569, | |
| "eval_steps_per_second": 9.698, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.422271223814774, | |
| "grad_norm": 22.98056411743164, | |
| "learning_rate": 1.4448732083792724e-05, | |
| "loss": 0.4214, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.4332965821389196, | |
| "grad_norm": 23.85386848449707, | |
| "learning_rate": 1.4173098125689086e-05, | |
| "loss": 0.5135, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.4443219404630652, | |
| "grad_norm": 15.967286109924316, | |
| "learning_rate": 1.3897464167585447e-05, | |
| "loss": 0.4906, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.4553472987872107, | |
| "grad_norm": 15.373555183410645, | |
| "learning_rate": 1.362183020948181e-05, | |
| "loss": 0.4426, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.4553472987872107, | |
| "eval_accuracy": 0.8347804515008683, | |
| "eval_loss": 0.5102077126502991, | |
| "eval_runtime": 56.3926, | |
| "eval_samples_per_second": 71.481, | |
| "eval_steps_per_second": 8.937, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.4663726571113562, | |
| "grad_norm": 14.563850402832031, | |
| "learning_rate": 1.334619625137817e-05, | |
| "loss": 0.3955, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.4773980154355018, | |
| "grad_norm": 15.834396362304688, | |
| "learning_rate": 1.3070562293274532e-05, | |
| "loss": 0.3435, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.4884233737596473, | |
| "grad_norm": 9.47835636138916, | |
| "learning_rate": 1.2794928335170895e-05, | |
| "loss": 0.3906, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.4994487320837928, | |
| "grad_norm": 20.611888885498047, | |
| "learning_rate": 1.2519294377067256e-05, | |
| "loss": 0.374, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.4994487320837928, | |
| "eval_accuracy": 0.8397419995038452, | |
| "eval_loss": 0.5277830958366394, | |
| "eval_runtime": 59.4632, | |
| "eval_samples_per_second": 67.79, | |
| "eval_steps_per_second": 8.476, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.5104740904079383, | |
| "grad_norm": 9.410921096801758, | |
| "learning_rate": 1.2243660418963617e-05, | |
| "loss": 0.4374, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 1.5214994487320839, | |
| "grad_norm": 2.510164737701416, | |
| "learning_rate": 1.1968026460859978e-05, | |
| "loss": 0.4131, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 1.5325248070562294, | |
| "grad_norm": 7.07089376449585, | |
| "learning_rate": 1.169239250275634e-05, | |
| "loss": 0.4937, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 1.543550165380375, | |
| "grad_norm": 12.722711563110352, | |
| "learning_rate": 1.1416758544652702e-05, | |
| "loss": 0.391, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.543550165380375, | |
| "eval_accuracy": 0.8429670057057802, | |
| "eval_loss": 0.5098406076431274, | |
| "eval_runtime": 58.5831, | |
| "eval_samples_per_second": 68.808, | |
| "eval_steps_per_second": 8.603, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.5545755237045205, | |
| "grad_norm": 5.899824619293213, | |
| "learning_rate": 1.1141124586549064e-05, | |
| "loss": 0.4584, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 1.565600882028666, | |
| "grad_norm": 20.70318031311035, | |
| "learning_rate": 1.0865490628445425e-05, | |
| "loss": 0.4132, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 1.5766262403528115, | |
| "grad_norm": 16.140867233276367, | |
| "learning_rate": 1.0589856670341788e-05, | |
| "loss": 0.3984, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 1.587651598676957, | |
| "grad_norm": 20.221525192260742, | |
| "learning_rate": 1.0314222712238149e-05, | |
| "loss": 0.4062, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.587651598676957, | |
| "eval_accuracy": 0.8509054825105433, | |
| "eval_loss": 0.5056183934211731, | |
| "eval_runtime": 57.4166, | |
| "eval_samples_per_second": 70.206, | |
| "eval_steps_per_second": 8.778, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.5986769570011026, | |
| "grad_norm": 26.3918514251709, | |
| "learning_rate": 1.0038588754134508e-05, | |
| "loss": 0.4098, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.609702315325248, | |
| "grad_norm": 25.124940872192383, | |
| "learning_rate": 9.762954796030871e-06, | |
| "loss": 0.3674, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.6207276736493936, | |
| "grad_norm": 18.994909286499023, | |
| "learning_rate": 9.487320837927232e-06, | |
| "loss": 0.4137, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.6317530319735392, | |
| "grad_norm": 0.34515783190727234, | |
| "learning_rate": 9.211686879823594e-06, | |
| "loss": 0.3067, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.6317530319735392, | |
| "eval_accuracy": 0.8447035475068221, | |
| "eval_loss": 0.5277951955795288, | |
| "eval_runtime": 62.2601, | |
| "eval_samples_per_second": 64.745, | |
| "eval_steps_per_second": 8.095, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.6427783902976847, | |
| "grad_norm": 20.627864837646484, | |
| "learning_rate": 8.936052921719956e-06, | |
| "loss": 0.3772, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.6538037486218302, | |
| "grad_norm": 12.762669563293457, | |
| "learning_rate": 8.660418963616318e-06, | |
| "loss": 0.4413, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.6648291069459757, | |
| "grad_norm": 1.8872555494308472, | |
| "learning_rate": 8.38478500551268e-06, | |
| "loss": 0.414, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.6758544652701213, | |
| "grad_norm": 5.339422225952148, | |
| "learning_rate": 8.109151047409042e-06, | |
| "loss": 0.4419, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.6758544652701213, | |
| "eval_accuracy": 0.8486727859092037, | |
| "eval_loss": 0.508654773235321, | |
| "eval_runtime": 60.7174, | |
| "eval_samples_per_second": 66.39, | |
| "eval_steps_per_second": 8.301, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.6868798235942668, | |
| "grad_norm": 26.65837287902832, | |
| "learning_rate": 7.833517089305403e-06, | |
| "loss": 0.4067, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.6979051819184123, | |
| "grad_norm": 1.583815574645996, | |
| "learning_rate": 7.557883131201765e-06, | |
| "loss": 0.3626, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.7089305402425579, | |
| "grad_norm": 7.412702560424805, | |
| "learning_rate": 7.282249173098125e-06, | |
| "loss": 0.3608, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.7199558985667034, | |
| "grad_norm": 18.560007095336914, | |
| "learning_rate": 7.006615214994487e-06, | |
| "loss": 0.4403, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.7199558985667034, | |
| "eval_accuracy": 0.8521458695112876, | |
| "eval_loss": 0.498369038105011, | |
| "eval_runtime": 61.2014, | |
| "eval_samples_per_second": 65.865, | |
| "eval_steps_per_second": 8.235, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.730981256890849, | |
| "grad_norm": 0.4578012228012085, | |
| "learning_rate": 6.730981256890849e-06, | |
| "loss": 0.2849, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.7420066152149944, | |
| "grad_norm": 22.82770347595215, | |
| "learning_rate": 6.4553472987872105e-06, | |
| "loss": 0.4316, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.75303197353914, | |
| "grad_norm": 2.745412588119507, | |
| "learning_rate": 6.1797133406835725e-06, | |
| "loss": 0.4044, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.7640573318632855, | |
| "grad_norm": 2.724490165710449, | |
| "learning_rate": 5.9040793825799346e-06, | |
| "loss": 0.4577, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.7640573318632855, | |
| "eval_accuracy": 0.843463160506078, | |
| "eval_loss": 0.5229588150978088, | |
| "eval_runtime": 61.6171, | |
| "eval_samples_per_second": 65.42, | |
| "eval_steps_per_second": 8.18, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.775082690187431, | |
| "grad_norm": 14.0263671875, | |
| "learning_rate": 5.628445424476296e-06, | |
| "loss": 0.4339, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.7861080485115766, | |
| "grad_norm": 10.231954574584961, | |
| "learning_rate": 5.352811466372657e-06, | |
| "loss": 0.366, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.797133406835722, | |
| "grad_norm": 7.47442102432251, | |
| "learning_rate": 5.077177508269019e-06, | |
| "loss": 0.3847, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.8081587651598676, | |
| "grad_norm": 34.1908073425293, | |
| "learning_rate": 4.801543550165381e-06, | |
| "loss": 0.3816, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.8081587651598676, | |
| "eval_accuracy": 0.8516497147109898, | |
| "eval_loss": 0.5057588219642639, | |
| "eval_runtime": 62.4381, | |
| "eval_samples_per_second": 64.56, | |
| "eval_steps_per_second": 8.072, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.8191841234840131, | |
| "grad_norm": 6.812930583953857, | |
| "learning_rate": 4.525909592061742e-06, | |
| "loss": 0.3876, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.8302094818081587, | |
| "grad_norm": 6.9469313621521, | |
| "learning_rate": 4.250275633958104e-06, | |
| "loss": 0.3791, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.8412348401323042, | |
| "grad_norm": 1.059395432472229, | |
| "learning_rate": 3.974641675854465e-06, | |
| "loss": 0.3403, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.8522601984564497, | |
| "grad_norm": 14.005857467651367, | |
| "learning_rate": 3.699007717750827e-06, | |
| "loss": 0.3351, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.8522601984564497, | |
| "eval_accuracy": 0.8536343339121806, | |
| "eval_loss": 0.49633243680000305, | |
| "eval_runtime": 60.6204, | |
| "eval_samples_per_second": 66.496, | |
| "eval_steps_per_second": 8.314, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.8632855567805953, | |
| "grad_norm": 10.774523735046387, | |
| "learning_rate": 3.4233737596471886e-06, | |
| "loss": 0.3543, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.8743109151047408, | |
| "grad_norm": 17.46559715270996, | |
| "learning_rate": 3.1477398015435506e-06, | |
| "loss": 0.4039, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.8853362734288863, | |
| "grad_norm": 8.823070526123047, | |
| "learning_rate": 2.8721058434399118e-06, | |
| "loss": 0.4881, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.8963616317530319, | |
| "grad_norm": 12.282500267028809, | |
| "learning_rate": 2.596471885336274e-06, | |
| "loss": 0.3894, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.8963616317530319, | |
| "eval_accuracy": 0.8551227983130737, | |
| "eval_loss": 0.48355239629745483, | |
| "eval_runtime": 61.0535, | |
| "eval_samples_per_second": 66.024, | |
| "eval_steps_per_second": 8.255, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.9073869900771774, | |
| "grad_norm": 8.131113052368164, | |
| "learning_rate": 2.320837927232635e-06, | |
| "loss": 0.4065, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.918412348401323, | |
| "grad_norm": 0.6533594727516174, | |
| "learning_rate": 2.045203969128997e-06, | |
| "loss": 0.3384, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.9294377067254684, | |
| "grad_norm": 8.164648056030273, | |
| "learning_rate": 1.7695700110253584e-06, | |
| "loss": 0.3509, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.940463065049614, | |
| "grad_norm": 4.368854522705078, | |
| "learning_rate": 1.49393605292172e-06, | |
| "loss": 0.3611, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.940463065049614, | |
| "eval_accuracy": 0.8558670305135202, | |
| "eval_loss": 0.47785690426826477, | |
| "eval_runtime": 61.2031, | |
| "eval_samples_per_second": 65.863, | |
| "eval_steps_per_second": 8.235, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.9514884233737595, | |
| "grad_norm": 7.265945911407471, | |
| "learning_rate": 1.2183020948180816e-06, | |
| "loss": 0.4892, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.962513781697905, | |
| "grad_norm": 2.180917978286743, | |
| "learning_rate": 9.426681367144433e-07, | |
| "loss": 0.4405, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.9735391400220506, | |
| "grad_norm": 11.077986717224121, | |
| "learning_rate": 6.670341786108048e-07, | |
| "loss": 0.3206, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.9845644983461963, | |
| "grad_norm": 20.9344482421875, | |
| "learning_rate": 3.914002205071665e-07, | |
| "loss": 0.3568, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.9845644983461963, | |
| "eval_accuracy": 0.8558670305135202, | |
| "eval_loss": 0.474208801984787, | |
| "eval_runtime": 64.6572, | |
| "eval_samples_per_second": 62.344, | |
| "eval_steps_per_second": 7.795, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.9955898566703418, | |
| "grad_norm": 11.30726432800293, | |
| "learning_rate": 1.1576626240352813e-07, | |
| "loss": 0.4508, | |
| "step": 9050 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 9070, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2402970222720000.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |