{ "best_metric": 0.16577637195587158, "best_model_checkpoint": "frostsolutions/frost-vision-v2-google_vit-base-patch16-224-v2024-11-11/checkpoint-500", "epoch": 30.0, "eval_steps": 100, "global_step": 2130, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.14084507042253522, "grad_norm": 2.1837456226348877, "learning_rate": 2.3474178403755868e-06, "loss": 0.7828, "step": 10 }, { "epoch": 0.28169014084507044, "grad_norm": 2.2785279750823975, "learning_rate": 4.6948356807511736e-06, "loss": 0.7397, "step": 20 }, { "epoch": 0.4225352112676056, "grad_norm": 1.786941409111023, "learning_rate": 7.042253521126762e-06, "loss": 0.6827, "step": 30 }, { "epoch": 0.5633802816901409, "grad_norm": 1.804363489151001, "learning_rate": 9.389671361502347e-06, "loss": 0.6092, "step": 40 }, { "epoch": 0.704225352112676, "grad_norm": 1.2980760335922241, "learning_rate": 1.1737089201877934e-05, "loss": 0.5265, "step": 50 }, { "epoch": 0.8450704225352113, "grad_norm": 1.0985078811645508, "learning_rate": 1.4084507042253523e-05, "loss": 0.461, "step": 60 }, { "epoch": 0.9859154929577465, "grad_norm": 0.966576099395752, "learning_rate": 1.643192488262911e-05, "loss": 0.4112, "step": 70 }, { "epoch": 1.1267605633802817, "grad_norm": 1.0513771772384644, "learning_rate": 1.8779342723004694e-05, "loss": 0.3653, "step": 80 }, { "epoch": 1.267605633802817, "grad_norm": 0.9769287109375, "learning_rate": 2.112676056338028e-05, "loss": 0.3547, "step": 90 }, { "epoch": 1.408450704225352, "grad_norm": 0.9427245259284973, "learning_rate": 2.347417840375587e-05, "loss": 0.3127, "step": 100 }, { "epoch": 1.408450704225352, "eval_accuracy": 0.8940140845070422, "eval_f1": 0.6724700761697497, "eval_loss": 0.29324474930763245, "eval_precision": 0.8153034300791556, "eval_recall": 0.5722222222222222, "eval_runtime": 5.8994, "eval_samples_per_second": 48.14, "eval_steps_per_second": 6.102, "step": 100 }, { "epoch": 1.5492957746478875, "grad_norm": 0.8860771059989929, "learning_rate": 2.5821596244131456e-05, "loss": 0.2867, "step": 110 }, { "epoch": 1.6901408450704225, "grad_norm": 0.816124439239502, "learning_rate": 2.8169014084507046e-05, "loss": 0.2808, "step": 120 }, { "epoch": 1.8309859154929577, "grad_norm": 0.8954053521156311, "learning_rate": 3.0516431924882634e-05, "loss": 0.2948, "step": 130 }, { "epoch": 1.971830985915493, "grad_norm": 1.4889060258865356, "learning_rate": 3.286384976525822e-05, "loss": 0.2359, "step": 140 }, { "epoch": 2.112676056338028, "grad_norm": 1.1245497465133667, "learning_rate": 3.5211267605633805e-05, "loss": 0.2217, "step": 150 }, { "epoch": 2.2535211267605635, "grad_norm": 0.6540855765342712, "learning_rate": 3.755868544600939e-05, "loss": 0.228, "step": 160 }, { "epoch": 2.3943661971830985, "grad_norm": 0.9159438014030457, "learning_rate": 3.990610328638498e-05, "loss": 0.2061, "step": 170 }, { "epoch": 2.535211267605634, "grad_norm": 1.1491577625274658, "learning_rate": 4.225352112676056e-05, "loss": 0.246, "step": 180 }, { "epoch": 2.676056338028169, "grad_norm": 1.0928665399551392, "learning_rate": 4.4600938967136154e-05, "loss": 0.2159, "step": 190 }, { "epoch": 2.816901408450704, "grad_norm": 0.8002581000328064, "learning_rate": 4.694835680751174e-05, "loss": 0.193, "step": 200 }, { "epoch": 2.816901408450704, "eval_accuracy": 0.9190140845070423, "eval_f1": 0.783427495291902, "eval_loss": 0.21363618969917297, "eval_precision": 0.7969348659003831, "eval_recall": 0.7703703703703704, "eval_runtime": 2.8854, "eval_samples_per_second": 98.426, "eval_steps_per_second": 12.477, "step": 200 }, { "epoch": 2.9577464788732395, "grad_norm": 0.8501047492027283, "learning_rate": 4.929577464788733e-05, "loss": 0.1954, "step": 210 }, { "epoch": 3.0985915492957745, "grad_norm": 1.074354648590088, "learning_rate": 4.981742305685968e-05, "loss": 0.1855, "step": 220 }, { "epoch": 3.23943661971831, "grad_norm": 0.6893348097801208, "learning_rate": 4.9556598852373504e-05, "loss": 0.166, "step": 230 }, { "epoch": 3.380281690140845, "grad_norm": 1.0960265398025513, "learning_rate": 4.929577464788733e-05, "loss": 0.1862, "step": 240 }, { "epoch": 3.52112676056338, "grad_norm": 1.657496690750122, "learning_rate": 4.903495044340115e-05, "loss": 0.1844, "step": 250 }, { "epoch": 3.6619718309859155, "grad_norm": 0.9198503494262695, "learning_rate": 4.8774126238914976e-05, "loss": 0.1841, "step": 260 }, { "epoch": 3.802816901408451, "grad_norm": 0.6908796429634094, "learning_rate": 4.85133020344288e-05, "loss": 0.1796, "step": 270 }, { "epoch": 3.943661971830986, "grad_norm": 1.0879693031311035, "learning_rate": 4.8252477829942624e-05, "loss": 0.1699, "step": 280 }, { "epoch": 4.084507042253521, "grad_norm": 0.7113975286483765, "learning_rate": 4.799165362545645e-05, "loss": 0.135, "step": 290 }, { "epoch": 4.225352112676056, "grad_norm": 0.956768274307251, "learning_rate": 4.773082942097027e-05, "loss": 0.1503, "step": 300 }, { "epoch": 4.225352112676056, "eval_accuracy": 0.9278169014084507, "eval_f1": 0.8100092678405931, "eval_loss": 0.18146201968193054, "eval_precision": 0.8107606679035251, "eval_recall": 0.8092592592592592, "eval_runtime": 1.9894, "eval_samples_per_second": 142.754, "eval_steps_per_second": 18.096, "step": 300 }, { "epoch": 4.366197183098592, "grad_norm": 1.119613528251648, "learning_rate": 4.747000521648409e-05, "loss": 0.1639, "step": 310 }, { "epoch": 4.507042253521127, "grad_norm": 1.3519386053085327, "learning_rate": 4.720918101199791e-05, "loss": 0.1769, "step": 320 }, { "epoch": 4.647887323943662, "grad_norm": 1.2561410665512085, "learning_rate": 4.694835680751174e-05, "loss": 0.1495, "step": 330 }, { "epoch": 4.788732394366197, "grad_norm": 0.8973093032836914, "learning_rate": 4.668753260302556e-05, "loss": 0.1603, "step": 340 }, { "epoch": 4.929577464788732, "grad_norm": 0.5152555108070374, "learning_rate": 4.6426708398539385e-05, "loss": 0.148, "step": 350 }, { "epoch": 5.070422535211268, "grad_norm": 0.6774557828903198, "learning_rate": 4.616588419405321e-05, "loss": 0.1369, "step": 360 }, { "epoch": 5.211267605633803, "grad_norm": 0.7763332724571228, "learning_rate": 4.5905059989567033e-05, "loss": 0.119, "step": 370 }, { "epoch": 5.352112676056338, "grad_norm": 0.9512709975242615, "learning_rate": 4.564423578508086e-05, "loss": 0.1448, "step": 380 }, { "epoch": 5.492957746478873, "grad_norm": 1.2080267667770386, "learning_rate": 4.538341158059468e-05, "loss": 0.1278, "step": 390 }, { "epoch": 5.633802816901408, "grad_norm": 0.7148259282112122, "learning_rate": 4.5122587376108505e-05, "loss": 0.1313, "step": 400 }, { "epoch": 5.633802816901408, "eval_accuracy": 0.9327464788732395, "eval_f1": 0.8182683158896289, "eval_loss": 0.16230623424053192, "eval_precision": 0.8414872798434442, "eval_recall": 0.7962962962962963, "eval_runtime": 2.7042, "eval_samples_per_second": 105.022, "eval_steps_per_second": 13.313, "step": 400 }, { "epoch": 5.774647887323944, "grad_norm": 0.9253571033477783, "learning_rate": 4.486176317162233e-05, "loss": 0.1489, "step": 410 }, { "epoch": 5.915492957746479, "grad_norm": 1.0267525911331177, "learning_rate": 4.4600938967136154e-05, "loss": 0.1426, "step": 420 }, { "epoch": 6.056338028169014, "grad_norm": 0.9201334118843079, "learning_rate": 4.434011476264998e-05, "loss": 0.1458, "step": 430 }, { "epoch": 6.197183098591549, "grad_norm": 0.6784873604774475, "learning_rate": 4.40792905581638e-05, "loss": 0.1387, "step": 440 }, { "epoch": 6.338028169014084, "grad_norm": 0.7604418992996216, "learning_rate": 4.3818466353677626e-05, "loss": 0.1087, "step": 450 }, { "epoch": 6.47887323943662, "grad_norm": 0.8017091751098633, "learning_rate": 4.355764214919145e-05, "loss": 0.1071, "step": 460 }, { "epoch": 6.619718309859155, "grad_norm": 0.8712729215621948, "learning_rate": 4.329681794470527e-05, "loss": 0.1084, "step": 470 }, { "epoch": 6.76056338028169, "grad_norm": 0.8843014240264893, "learning_rate": 4.303599374021909e-05, "loss": 0.1176, "step": 480 }, { "epoch": 6.901408450704225, "grad_norm": 0.7635127305984497, "learning_rate": 4.2775169535732915e-05, "loss": 0.1213, "step": 490 }, { "epoch": 7.042253521126761, "grad_norm": 0.6069650053977966, "learning_rate": 4.251434533124674e-05, "loss": 0.1166, "step": 500 }, { "epoch": 7.042253521126761, "eval_accuracy": 0.9320422535211268, "eval_f1": 0.8224471021159153, "eval_loss": 0.16577637195587158, "eval_precision": 0.8171846435100548, "eval_recall": 0.8277777777777777, "eval_runtime": 2.043, "eval_samples_per_second": 139.008, "eval_steps_per_second": 17.621, "step": 500 }, { "epoch": 7.183098591549296, "grad_norm": 0.9951217770576477, "learning_rate": 4.225352112676056e-05, "loss": 0.1125, "step": 510 }, { "epoch": 7.323943661971831, "grad_norm": 0.4932935833930969, "learning_rate": 4.199269692227439e-05, "loss": 0.1221, "step": 520 }, { "epoch": 7.464788732394366, "grad_norm": 0.9160825610160828, "learning_rate": 4.173187271778821e-05, "loss": 0.1058, "step": 530 }, { "epoch": 7.605633802816901, "grad_norm": 0.8885666131973267, "learning_rate": 4.1471048513302035e-05, "loss": 0.1033, "step": 540 }, { "epoch": 7.746478873239437, "grad_norm": 0.49906080961227417, "learning_rate": 4.121022430881586e-05, "loss": 0.0983, "step": 550 }, { "epoch": 7.887323943661972, "grad_norm": 0.825618326663971, "learning_rate": 4.094940010432968e-05, "loss": 0.1093, "step": 560 }, { "epoch": 8.028169014084508, "grad_norm": 0.8402373194694519, "learning_rate": 4.068857589984351e-05, "loss": 0.112, "step": 570 }, { "epoch": 8.169014084507042, "grad_norm": 0.5090510249137878, "learning_rate": 4.042775169535733e-05, "loss": 0.0828, "step": 580 }, { "epoch": 8.309859154929578, "grad_norm": 0.9439256191253662, "learning_rate": 4.0166927490871155e-05, "loss": 0.1069, "step": 590 }, { "epoch": 8.450704225352112, "grad_norm": 0.47647300362586975, "learning_rate": 3.990610328638498e-05, "loss": 0.093, "step": 600 }, { "epoch": 8.450704225352112, "eval_accuracy": 0.9383802816901409, "eval_f1": 0.8404740200546946, "eval_loss": 0.1605876386165619, "eval_precision": 0.8276481149012568, "eval_recall": 0.8537037037037037, "eval_runtime": 2.9556, "eval_samples_per_second": 96.09, "eval_steps_per_second": 12.18, "step": 600 }, { "epoch": 8.591549295774648, "grad_norm": 0.6321811676025391, "learning_rate": 3.96452790818988e-05, "loss": 0.0942, "step": 610 }, { "epoch": 8.732394366197184, "grad_norm": 0.7512757778167725, "learning_rate": 3.938445487741263e-05, "loss": 0.0991, "step": 620 }, { "epoch": 8.873239436619718, "grad_norm": 0.5704978108406067, "learning_rate": 3.912363067292645e-05, "loss": 0.0988, "step": 630 }, { "epoch": 9.014084507042254, "grad_norm": 1.4220664501190186, "learning_rate": 3.886280646844027e-05, "loss": 0.0947, "step": 640 }, { "epoch": 9.154929577464788, "grad_norm": 1.9923653602600098, "learning_rate": 3.860198226395409e-05, "loss": 0.0987, "step": 650 }, { "epoch": 9.295774647887324, "grad_norm": 1.0009934902191162, "learning_rate": 3.8341158059467916e-05, "loss": 0.0826, "step": 660 }, { "epoch": 9.43661971830986, "grad_norm": 0.5631663799285889, "learning_rate": 3.808033385498174e-05, "loss": 0.1084, "step": 670 }, { "epoch": 9.577464788732394, "grad_norm": 0.6266153454780579, "learning_rate": 3.7819509650495564e-05, "loss": 0.0808, "step": 680 }, { "epoch": 9.71830985915493, "grad_norm": 0.9451386332511902, "learning_rate": 3.755868544600939e-05, "loss": 0.0759, "step": 690 }, { "epoch": 9.859154929577464, "grad_norm": 1.2534326314926147, "learning_rate": 3.729786124152321e-05, "loss": 0.0931, "step": 700 }, { "epoch": 9.859154929577464, "eval_accuracy": 0.9366197183098591, "eval_f1": 0.8369565217391305, "eval_loss": 0.1625019609928131, "eval_precision": 0.8191489361702128, "eval_recall": 0.8555555555555555, "eval_runtime": 2.0664, "eval_samples_per_second": 137.435, "eval_steps_per_second": 17.421, "step": 700 }, { "epoch": 10.0, "grad_norm": 0.4895138144493103, "learning_rate": 3.7037037037037037e-05, "loss": 0.0883, "step": 710 }, { "epoch": 10.140845070422536, "grad_norm": 0.7509373426437378, "learning_rate": 3.677621283255086e-05, "loss": 0.0816, "step": 720 }, { "epoch": 10.28169014084507, "grad_norm": 0.6509613990783691, "learning_rate": 3.6515388628064685e-05, "loss": 0.075, "step": 730 }, { "epoch": 10.422535211267606, "grad_norm": 0.655375599861145, "learning_rate": 3.625456442357851e-05, "loss": 0.0703, "step": 740 }, { "epoch": 10.56338028169014, "grad_norm": 0.4507924020290375, "learning_rate": 3.599374021909233e-05, "loss": 0.0828, "step": 750 }, { "epoch": 10.704225352112676, "grad_norm": 0.9039227366447449, "learning_rate": 3.573291601460616e-05, "loss": 0.0732, "step": 760 }, { "epoch": 10.845070422535212, "grad_norm": 0.5821270942687988, "learning_rate": 3.547209181011998e-05, "loss": 0.0747, "step": 770 }, { "epoch": 10.985915492957746, "grad_norm": 0.9370685815811157, "learning_rate": 3.5211267605633805e-05, "loss": 0.072, "step": 780 }, { "epoch": 11.126760563380282, "grad_norm": 0.729614794254303, "learning_rate": 3.495044340114763e-05, "loss": 0.0733, "step": 790 }, { "epoch": 11.267605633802816, "grad_norm": 0.5709458589553833, "learning_rate": 3.4689619196661446e-05, "loss": 0.0733, "step": 800 }, { "epoch": 11.267605633802816, "eval_accuracy": 0.9355633802816902, "eval_f1": 0.8310249307479224, "eval_loss": 0.17139478027820587, "eval_precision": 0.8287292817679558, "eval_recall": 0.8333333333333334, "eval_runtime": 2.0599, "eval_samples_per_second": 137.87, "eval_steps_per_second": 17.476, "step": 800 }, { "epoch": 11.408450704225352, "grad_norm": 1.1286197900772095, "learning_rate": 3.442879499217527e-05, "loss": 0.0666, "step": 810 }, { "epoch": 11.549295774647888, "grad_norm": 0.9596506953239441, "learning_rate": 3.4167970787689094e-05, "loss": 0.0849, "step": 820 }, { "epoch": 11.690140845070422, "grad_norm": 0.8838114142417908, "learning_rate": 3.3907146583202925e-05, "loss": 0.0693, "step": 830 }, { "epoch": 11.830985915492958, "grad_norm": 0.573143482208252, "learning_rate": 3.364632237871675e-05, "loss": 0.0688, "step": 840 }, { "epoch": 11.971830985915492, "grad_norm": 0.4819594919681549, "learning_rate": 3.338549817423057e-05, "loss": 0.0708, "step": 850 }, { "epoch": 12.112676056338028, "grad_norm": 0.672797679901123, "learning_rate": 3.31246739697444e-05, "loss": 0.0741, "step": 860 }, { "epoch": 12.253521126760564, "grad_norm": 0.3752357065677643, "learning_rate": 3.286384976525822e-05, "loss": 0.0794, "step": 870 }, { "epoch": 12.394366197183098, "grad_norm": 0.7455435991287231, "learning_rate": 3.2603025560772045e-05, "loss": 0.0719, "step": 880 }, { "epoch": 12.535211267605634, "grad_norm": 0.6346479654312134, "learning_rate": 3.234220135628587e-05, "loss": 0.062, "step": 890 }, { "epoch": 12.676056338028168, "grad_norm": 1.10377836227417, "learning_rate": 3.208137715179969e-05, "loss": 0.0693, "step": 900 }, { "epoch": 12.676056338028168, "eval_accuracy": 0.9397887323943662, "eval_f1": 0.8403361344537815, "eval_loss": 0.1568107008934021, "eval_precision": 0.847457627118644, "eval_recall": 0.8333333333333334, "eval_runtime": 2.0263, "eval_samples_per_second": 140.158, "eval_steps_per_second": 17.767, "step": 900 }, { "epoch": 12.816901408450704, "grad_norm": 0.4753211736679077, "learning_rate": 3.182055294731352e-05, "loss": 0.0577, "step": 910 }, { "epoch": 12.95774647887324, "grad_norm": 0.6100496053695679, "learning_rate": 3.155972874282734e-05, "loss": 0.0715, "step": 920 }, { "epoch": 13.098591549295774, "grad_norm": 0.3752472698688507, "learning_rate": 3.1298904538341165e-05, "loss": 0.0683, "step": 930 }, { "epoch": 13.23943661971831, "grad_norm": 1.3385846614837646, "learning_rate": 3.103808033385498e-05, "loss": 0.0578, "step": 940 }, { "epoch": 13.380281690140846, "grad_norm": 0.5425249934196472, "learning_rate": 3.0777256129368806e-05, "loss": 0.0601, "step": 950 }, { "epoch": 13.52112676056338, "grad_norm": 0.8623884320259094, "learning_rate": 3.0516431924882634e-05, "loss": 0.0673, "step": 960 }, { "epoch": 13.661971830985916, "grad_norm": 0.7758656144142151, "learning_rate": 3.0255607720396458e-05, "loss": 0.0707, "step": 970 }, { "epoch": 13.80281690140845, "grad_norm": 0.6038694381713867, "learning_rate": 2.999478351591028e-05, "loss": 0.0631, "step": 980 }, { "epoch": 13.943661971830986, "grad_norm": 0.7183906435966492, "learning_rate": 2.9733959311424102e-05, "loss": 0.0647, "step": 990 }, { "epoch": 14.084507042253522, "grad_norm": 0.6927624940872192, "learning_rate": 2.9473135106937926e-05, "loss": 0.0615, "step": 1000 }, { "epoch": 14.084507042253522, "eval_accuracy": 0.9341549295774648, "eval_f1": 0.8270120259019427, "eval_loss": 0.1666344702243805, "eval_precision": 0.8262476894639557, "eval_recall": 0.8277777777777777, "eval_runtime": 2.4957, "eval_samples_per_second": 113.795, "eval_steps_per_second": 14.425, "step": 1000 }, { "epoch": 14.225352112676056, "grad_norm": 0.937294065952301, "learning_rate": 2.921231090245175e-05, "loss": 0.0545, "step": 1010 }, { "epoch": 14.366197183098592, "grad_norm": 0.9548676013946533, "learning_rate": 2.8951486697965574e-05, "loss": 0.0601, "step": 1020 }, { "epoch": 14.507042253521126, "grad_norm": 0.5984657406806946, "learning_rate": 2.86906624934794e-05, "loss": 0.0522, "step": 1030 }, { "epoch": 14.647887323943662, "grad_norm": 0.6830617785453796, "learning_rate": 2.8429838288993222e-05, "loss": 0.0689, "step": 1040 }, { "epoch": 14.788732394366198, "grad_norm": 1.328912615776062, "learning_rate": 2.8169014084507046e-05, "loss": 0.0753, "step": 1050 }, { "epoch": 14.929577464788732, "grad_norm": 0.7891097664833069, "learning_rate": 2.7908189880020867e-05, "loss": 0.0696, "step": 1060 }, { "epoch": 15.070422535211268, "grad_norm": 0.7651441693305969, "learning_rate": 2.764736567553469e-05, "loss": 0.0546, "step": 1070 }, { "epoch": 15.211267605633802, "grad_norm": 0.9666036367416382, "learning_rate": 2.7386541471048515e-05, "loss": 0.0493, "step": 1080 }, { "epoch": 15.352112676056338, "grad_norm": 0.4921013414859772, "learning_rate": 2.712571726656234e-05, "loss": 0.0478, "step": 1090 }, { "epoch": 15.492957746478874, "grad_norm": 0.6497583389282227, "learning_rate": 2.6864893062076163e-05, "loss": 0.0562, "step": 1100 }, { "epoch": 15.492957746478874, "eval_accuracy": 0.9394366197183098, "eval_f1": 0.8404452690166976, "eval_loss": 0.16357889771461487, "eval_precision": 0.8420074349442379, "eval_recall": 0.8388888888888889, "eval_runtime": 2.0468, "eval_samples_per_second": 138.756, "eval_steps_per_second": 17.589, "step": 1100 }, { "epoch": 15.633802816901408, "grad_norm": 2.279696226119995, "learning_rate": 2.6604068857589987e-05, "loss": 0.0591, "step": 1110 }, { "epoch": 15.774647887323944, "grad_norm": 0.450811505317688, "learning_rate": 2.634324465310381e-05, "loss": 0.0502, "step": 1120 }, { "epoch": 15.915492957746478, "grad_norm": 0.4582817554473877, "learning_rate": 2.6082420448617635e-05, "loss": 0.0576, "step": 1130 }, { "epoch": 16.056338028169016, "grad_norm": 0.46688681840896606, "learning_rate": 2.5821596244131456e-05, "loss": 0.0461, "step": 1140 }, { "epoch": 16.197183098591548, "grad_norm": 0.4218389093875885, "learning_rate": 2.556077203964528e-05, "loss": 0.0516, "step": 1150 }, { "epoch": 16.338028169014084, "grad_norm": 0.9011832475662231, "learning_rate": 2.5299947835159104e-05, "loss": 0.0486, "step": 1160 }, { "epoch": 16.47887323943662, "grad_norm": 0.542711079120636, "learning_rate": 2.5039123630672928e-05, "loss": 0.052, "step": 1170 }, { "epoch": 16.619718309859156, "grad_norm": 0.45541244745254517, "learning_rate": 2.4778299426186752e-05, "loss": 0.046, "step": 1180 }, { "epoch": 16.760563380281692, "grad_norm": 0.6403104662895203, "learning_rate": 2.4517475221700576e-05, "loss": 0.055, "step": 1190 }, { "epoch": 16.901408450704224, "grad_norm": 0.8532978296279907, "learning_rate": 2.42566510172144e-05, "loss": 0.0507, "step": 1200 }, { "epoch": 16.901408450704224, "eval_accuracy": 0.9401408450704225, "eval_f1": 0.8434622467771639, "eval_loss": 0.1612667739391327, "eval_precision": 0.8388278388278388, "eval_recall": 0.8481481481481481, "eval_runtime": 1.994, "eval_samples_per_second": 142.429, "eval_steps_per_second": 18.054, "step": 1200 }, { "epoch": 17.04225352112676, "grad_norm": 1.2943941354751587, "learning_rate": 2.3995826812728224e-05, "loss": 0.058, "step": 1210 }, { "epoch": 17.183098591549296, "grad_norm": 0.5414674282073975, "learning_rate": 2.3735002608242045e-05, "loss": 0.0462, "step": 1220 }, { "epoch": 17.323943661971832, "grad_norm": 0.7418174743652344, "learning_rate": 2.347417840375587e-05, "loss": 0.0408, "step": 1230 }, { "epoch": 17.464788732394368, "grad_norm": 0.339821457862854, "learning_rate": 2.3213354199269693e-05, "loss": 0.0407, "step": 1240 }, { "epoch": 17.6056338028169, "grad_norm": 0.779461145401001, "learning_rate": 2.2952529994783517e-05, "loss": 0.058, "step": 1250 }, { "epoch": 17.746478873239436, "grad_norm": 0.719028651714325, "learning_rate": 2.269170579029734e-05, "loss": 0.048, "step": 1260 }, { "epoch": 17.887323943661972, "grad_norm": 0.6600283980369568, "learning_rate": 2.2430881585811165e-05, "loss": 0.0573, "step": 1270 }, { "epoch": 18.028169014084508, "grad_norm": 0.8104175925254822, "learning_rate": 2.217005738132499e-05, "loss": 0.0494, "step": 1280 }, { "epoch": 18.169014084507044, "grad_norm": 0.7185409665107727, "learning_rate": 2.1909233176838813e-05, "loss": 0.0571, "step": 1290 }, { "epoch": 18.309859154929576, "grad_norm": 0.32650864124298096, "learning_rate": 2.1648408972352633e-05, "loss": 0.0552, "step": 1300 }, { "epoch": 18.309859154929576, "eval_accuracy": 0.9411971830985916, "eval_f1": 0.845513413506013, "eval_loss": 0.15904362499713898, "eval_precision": 0.844731977818854, "eval_recall": 0.8462962962962963, "eval_runtime": 2.6258, "eval_samples_per_second": 108.157, "eval_steps_per_second": 13.71, "step": 1300 }, { "epoch": 18.450704225352112, "grad_norm": 1.3154155015945435, "learning_rate": 2.1387584767866457e-05, "loss": 0.0454, "step": 1310 }, { "epoch": 18.591549295774648, "grad_norm": 0.4282239079475403, "learning_rate": 2.112676056338028e-05, "loss": 0.0431, "step": 1320 }, { "epoch": 18.732394366197184, "grad_norm": 1.1309629678726196, "learning_rate": 2.0865936358894105e-05, "loss": 0.0395, "step": 1330 }, { "epoch": 18.87323943661972, "grad_norm": 0.5871407985687256, "learning_rate": 2.060511215440793e-05, "loss": 0.0477, "step": 1340 }, { "epoch": 19.014084507042252, "grad_norm": 0.30267587304115295, "learning_rate": 2.0344287949921754e-05, "loss": 0.0349, "step": 1350 }, { "epoch": 19.154929577464788, "grad_norm": 1.5481466054916382, "learning_rate": 2.0083463745435578e-05, "loss": 0.0436, "step": 1360 }, { "epoch": 19.295774647887324, "grad_norm": 0.6972795724868774, "learning_rate": 1.98226395409494e-05, "loss": 0.0494, "step": 1370 }, { "epoch": 19.43661971830986, "grad_norm": 0.332928866147995, "learning_rate": 1.9561815336463226e-05, "loss": 0.0367, "step": 1380 }, { "epoch": 19.577464788732396, "grad_norm": 0.5983466506004333, "learning_rate": 1.9300991131977046e-05, "loss": 0.0389, "step": 1390 }, { "epoch": 19.718309859154928, "grad_norm": 0.734262228012085, "learning_rate": 1.904016692749087e-05, "loss": 0.0439, "step": 1400 }, { "epoch": 19.718309859154928, "eval_accuracy": 0.9394366197183098, "eval_f1": 0.8424908424908425, "eval_loss": 0.17035430669784546, "eval_precision": 0.8333333333333334, "eval_recall": 0.8518518518518519, "eval_runtime": 2.448, "eval_samples_per_second": 116.015, "eval_steps_per_second": 14.706, "step": 1400 }, { "epoch": 19.859154929577464, "grad_norm": 0.6112886667251587, "learning_rate": 1.8779342723004694e-05, "loss": 0.0385, "step": 1410 }, { "epoch": 20.0, "grad_norm": 1.010438323020935, "learning_rate": 1.8518518518518518e-05, "loss": 0.0456, "step": 1420 }, { "epoch": 20.140845070422536, "grad_norm": 0.5202921032905579, "learning_rate": 1.8257694314032342e-05, "loss": 0.0397, "step": 1430 }, { "epoch": 20.281690140845072, "grad_norm": 0.9659579992294312, "learning_rate": 1.7996870109546166e-05, "loss": 0.0465, "step": 1440 }, { "epoch": 20.422535211267604, "grad_norm": 0.5208175182342529, "learning_rate": 1.773604590505999e-05, "loss": 0.0435, "step": 1450 }, { "epoch": 20.56338028169014, "grad_norm": 0.5702980756759644, "learning_rate": 1.7475221700573814e-05, "loss": 0.0486, "step": 1460 }, { "epoch": 20.704225352112676, "grad_norm": 0.5662500262260437, "learning_rate": 1.7214397496087635e-05, "loss": 0.0358, "step": 1470 }, { "epoch": 20.845070422535212, "grad_norm": 0.901360034942627, "learning_rate": 1.6953573291601462e-05, "loss": 0.0339, "step": 1480 }, { "epoch": 20.985915492957748, "grad_norm": 0.8144014477729797, "learning_rate": 1.6692749087115286e-05, "loss": 0.0483, "step": 1490 }, { "epoch": 21.12676056338028, "grad_norm": 0.357433021068573, "learning_rate": 1.643192488262911e-05, "loss": 0.0367, "step": 1500 }, { "epoch": 21.12676056338028, "eval_accuracy": 0.9426056338028169, "eval_f1": 0.8483720930232558, "eval_loss": 0.170222669839859, "eval_precision": 0.8523364485981308, "eval_recall": 0.8444444444444444, "eval_runtime": 2.014, "eval_samples_per_second": 141.013, "eval_steps_per_second": 17.875, "step": 1500 }, { "epoch": 21.267605633802816, "grad_norm": 0.521035373210907, "learning_rate": 1.6171100678142934e-05, "loss": 0.0342, "step": 1510 }, { "epoch": 21.408450704225352, "grad_norm": 0.47409939765930176, "learning_rate": 1.591027647365676e-05, "loss": 0.0474, "step": 1520 }, { "epoch": 21.549295774647888, "grad_norm": 0.3274288773536682, "learning_rate": 1.5649452269170582e-05, "loss": 0.0358, "step": 1530 }, { "epoch": 21.690140845070424, "grad_norm": 0.5142138004302979, "learning_rate": 1.5388628064684403e-05, "loss": 0.038, "step": 1540 }, { "epoch": 21.830985915492956, "grad_norm": 0.5675768256187439, "learning_rate": 1.5127803860198229e-05, "loss": 0.0343, "step": 1550 }, { "epoch": 21.971830985915492, "grad_norm": 0.46324819326400757, "learning_rate": 1.4866979655712051e-05, "loss": 0.0393, "step": 1560 }, { "epoch": 22.112676056338028, "grad_norm": 0.3789016604423523, "learning_rate": 1.4606155451225875e-05, "loss": 0.0355, "step": 1570 }, { "epoch": 22.253521126760564, "grad_norm": 0.2880994379520416, "learning_rate": 1.43453312467397e-05, "loss": 0.0284, "step": 1580 }, { "epoch": 22.3943661971831, "grad_norm": 0.6509861350059509, "learning_rate": 1.4084507042253523e-05, "loss": 0.0367, "step": 1590 }, { "epoch": 22.535211267605632, "grad_norm": 1.1531139612197876, "learning_rate": 1.3823682837767346e-05, "loss": 0.0424, "step": 1600 }, { "epoch": 22.535211267605632, "eval_accuracy": 0.9394366197183098, "eval_f1": 0.8419117647058824, "eval_loss": 0.1684611290693283, "eval_precision": 0.8357664233576643, "eval_recall": 0.8481481481481481, "eval_runtime": 2.1171, "eval_samples_per_second": 134.146, "eval_steps_per_second": 17.004, "step": 1600 }, { "epoch": 22.676056338028168, "grad_norm": 1.80601966381073, "learning_rate": 1.356285863328117e-05, "loss": 0.0327, "step": 1610 }, { "epoch": 22.816901408450704, "grad_norm": 0.5846266746520996, "learning_rate": 1.3302034428794994e-05, "loss": 0.0371, "step": 1620 }, { "epoch": 22.95774647887324, "grad_norm": 0.2390224188566208, "learning_rate": 1.3041210224308818e-05, "loss": 0.042, "step": 1630 }, { "epoch": 23.098591549295776, "grad_norm": 0.26173052191734314, "learning_rate": 1.278038601982264e-05, "loss": 0.038, "step": 1640 }, { "epoch": 23.239436619718308, "grad_norm": 0.9115623831748962, "learning_rate": 1.2519561815336464e-05, "loss": 0.0323, "step": 1650 }, { "epoch": 23.380281690140844, "grad_norm": 0.5974897742271423, "learning_rate": 1.2258737610850288e-05, "loss": 0.0347, "step": 1660 }, { "epoch": 23.52112676056338, "grad_norm": 0.9727242588996887, "learning_rate": 1.1997913406364112e-05, "loss": 0.0461, "step": 1670 }, { "epoch": 23.661971830985916, "grad_norm": 0.8638906478881836, "learning_rate": 1.1737089201877934e-05, "loss": 0.0375, "step": 1680 }, { "epoch": 23.802816901408452, "grad_norm": 0.8562279343605042, "learning_rate": 1.1476264997391758e-05, "loss": 0.0312, "step": 1690 }, { "epoch": 23.943661971830984, "grad_norm": 0.6405536532402039, "learning_rate": 1.1215440792905582e-05, "loss": 0.0306, "step": 1700 }, { "epoch": 23.943661971830984, "eval_accuracy": 0.9380281690140845, "eval_f1": 0.8397085610200364, "eval_loss": 0.1770738810300827, "eval_precision": 0.8261648745519713, "eval_recall": 0.8537037037037037, "eval_runtime": 2.9524, "eval_samples_per_second": 96.192, "eval_steps_per_second": 12.193, "step": 1700 }, { "epoch": 24.08450704225352, "grad_norm": 0.9799217581748962, "learning_rate": 1.0954616588419406e-05, "loss": 0.033, "step": 1710 }, { "epoch": 24.225352112676056, "grad_norm": 0.4004919230937958, "learning_rate": 1.0693792383933229e-05, "loss": 0.0334, "step": 1720 }, { "epoch": 24.366197183098592, "grad_norm": 1.033111333847046, "learning_rate": 1.0432968179447053e-05, "loss": 0.0364, "step": 1730 }, { "epoch": 24.507042253521128, "grad_norm": 0.3790343999862671, "learning_rate": 1.0172143974960877e-05, "loss": 0.0369, "step": 1740 }, { "epoch": 24.647887323943664, "grad_norm": 0.20175768435001373, "learning_rate": 9.9113197704747e-06, "loss": 0.0303, "step": 1750 }, { "epoch": 24.788732394366196, "grad_norm": 0.7406692504882812, "learning_rate": 9.650495565988523e-06, "loss": 0.0329, "step": 1760 }, { "epoch": 24.929577464788732, "grad_norm": 0.3440076410770416, "learning_rate": 9.389671361502347e-06, "loss": 0.0395, "step": 1770 }, { "epoch": 25.070422535211268, "grad_norm": 0.2520686686038971, "learning_rate": 9.128847157016171e-06, "loss": 0.0269, "step": 1780 }, { "epoch": 25.211267605633804, "grad_norm": 0.40644267201423645, "learning_rate": 8.868022952529995e-06, "loss": 0.0336, "step": 1790 }, { "epoch": 25.352112676056336, "grad_norm": 0.2972894012928009, "learning_rate": 8.607198748043817e-06, "loss": 0.0352, "step": 1800 }, { "epoch": 25.352112676056336, "eval_accuracy": 0.9401408450704225, "eval_f1": 0.8440366972477065, "eval_loss": 0.1691148281097412, "eval_precision": 0.8363636363636363, "eval_recall": 0.8518518518518519, "eval_runtime": 2.183, "eval_samples_per_second": 130.097, "eval_steps_per_second": 16.491, "step": 1800 }, { "epoch": 25.492957746478872, "grad_norm": 0.7162390947341919, "learning_rate": 8.346374543557643e-06, "loss": 0.0267, "step": 1810 }, { "epoch": 25.633802816901408, "grad_norm": 0.49857279658317566, "learning_rate": 8.085550339071467e-06, "loss": 0.0459, "step": 1820 }, { "epoch": 25.774647887323944, "grad_norm": 0.3226544260978699, "learning_rate": 7.824726134585291e-06, "loss": 0.0383, "step": 1830 }, { "epoch": 25.91549295774648, "grad_norm": 0.10473772883415222, "learning_rate": 7.563901930099114e-06, "loss": 0.0296, "step": 1840 }, { "epoch": 26.056338028169016, "grad_norm": 0.4618949592113495, "learning_rate": 7.303077725612938e-06, "loss": 0.0257, "step": 1850 }, { "epoch": 26.197183098591548, "grad_norm": 0.15693806111812592, "learning_rate": 7.042253521126762e-06, "loss": 0.0274, "step": 1860 }, { "epoch": 26.338028169014084, "grad_norm": 0.3008931279182434, "learning_rate": 6.781429316640585e-06, "loss": 0.0307, "step": 1870 }, { "epoch": 26.47887323943662, "grad_norm": 0.7514234185218811, "learning_rate": 6.520605112154409e-06, "loss": 0.0348, "step": 1880 }, { "epoch": 26.619718309859156, "grad_norm": 1.1500775814056396, "learning_rate": 6.259780907668232e-06, "loss": 0.0304, "step": 1890 }, { "epoch": 26.760563380281692, "grad_norm": 0.8598619103431702, "learning_rate": 5.998956703182056e-06, "loss": 0.0323, "step": 1900 }, { "epoch": 26.760563380281692, "eval_accuracy": 0.9426056338028169, "eval_f1": 0.8508691674290942, "eval_loss": 0.1686810702085495, "eval_precision": 0.840867992766727, "eval_recall": 0.8611111111111112, "eval_runtime": 2.0406, "eval_samples_per_second": 139.175, "eval_steps_per_second": 17.642, "step": 1900 }, { "epoch": 26.901408450704224, "grad_norm": 0.2144249826669693, "learning_rate": 5.738132498695879e-06, "loss": 0.0279, "step": 1910 }, { "epoch": 27.04225352112676, "grad_norm": 0.05483362078666687, "learning_rate": 5.477308294209703e-06, "loss": 0.0273, "step": 1920 }, { "epoch": 27.183098591549296, "grad_norm": 0.3904362618923187, "learning_rate": 5.216484089723526e-06, "loss": 0.0249, "step": 1930 }, { "epoch": 27.323943661971832, "grad_norm": 0.2276570200920105, "learning_rate": 4.95565988523735e-06, "loss": 0.0284, "step": 1940 }, { "epoch": 27.464788732394368, "grad_norm": 0.49757641553878784, "learning_rate": 4.6948356807511736e-06, "loss": 0.0406, "step": 1950 }, { "epoch": 27.6056338028169, "grad_norm": 0.6939036846160889, "learning_rate": 4.434011476264998e-06, "loss": 0.0344, "step": 1960 }, { "epoch": 27.746478873239436, "grad_norm": 0.35933974385261536, "learning_rate": 4.173187271778822e-06, "loss": 0.0246, "step": 1970 }, { "epoch": 27.887323943661972, "grad_norm": 0.7418074011802673, "learning_rate": 3.912363067292646e-06, "loss": 0.0319, "step": 1980 }, { "epoch": 28.028169014084508, "grad_norm": 0.47155845165252686, "learning_rate": 3.651538862806469e-06, "loss": 0.0309, "step": 1990 }, { "epoch": 28.169014084507044, "grad_norm": 0.5308616161346436, "learning_rate": 3.3907146583202924e-06, "loss": 0.0297, "step": 2000 }, { "epoch": 28.169014084507044, "eval_accuracy": 0.9401408450704225, "eval_f1": 0.8454545454545455, "eval_loss": 0.17321611940860748, "eval_precision": 0.8303571428571429, "eval_recall": 0.8611111111111112, "eval_runtime": 2.0215, "eval_samples_per_second": 140.487, "eval_steps_per_second": 17.808, "step": 2000 }, { "epoch": 28.309859154929576, "grad_norm": 0.38366690278053284, "learning_rate": 3.129890453834116e-06, "loss": 0.0315, "step": 2010 }, { "epoch": 28.450704225352112, "grad_norm": 0.33385714888572693, "learning_rate": 2.8690662493479396e-06, "loss": 0.0315, "step": 2020 }, { "epoch": 28.591549295774648, "grad_norm": 0.5694302320480347, "learning_rate": 2.608242044861763e-06, "loss": 0.0267, "step": 2030 }, { "epoch": 28.732394366197184, "grad_norm": 0.7156753540039062, "learning_rate": 2.3474178403755868e-06, "loss": 0.0238, "step": 2040 }, { "epoch": 28.87323943661972, "grad_norm": 0.290353000164032, "learning_rate": 2.086593635889411e-06, "loss": 0.021, "step": 2050 }, { "epoch": 29.014084507042252, "grad_norm": 0.37377166748046875, "learning_rate": 1.8257694314032344e-06, "loss": 0.0225, "step": 2060 }, { "epoch": 29.154929577464788, "grad_norm": 0.4136624038219452, "learning_rate": 1.564945226917058e-06, "loss": 0.0326, "step": 2070 }, { "epoch": 29.295774647887324, "grad_norm": 0.2563035786151886, "learning_rate": 1.3041210224308816e-06, "loss": 0.0338, "step": 2080 }, { "epoch": 29.43661971830986, "grad_norm": 0.5630273222923279, "learning_rate": 1.0432968179447054e-06, "loss": 0.0285, "step": 2090 }, { "epoch": 29.577464788732396, "grad_norm": 0.2827240824699402, "learning_rate": 7.82472613458529e-07, "loss": 0.0229, "step": 2100 }, { "epoch": 29.577464788732396, "eval_accuracy": 0.9411971830985916, "eval_f1": 0.8474885844748858, "eval_loss": 0.17117364704608917, "eval_precision": 0.836036036036036, "eval_recall": 0.8592592592592593, "eval_runtime": 2.8598, "eval_samples_per_second": 99.308, "eval_steps_per_second": 12.588, "step": 2100 }, { "epoch": 29.718309859154928, "grad_norm": 0.1679154634475708, "learning_rate": 5.216484089723527e-07, "loss": 0.0263, "step": 2110 }, { "epoch": 29.859154929577464, "grad_norm": 0.21034198999404907, "learning_rate": 2.6082420448617635e-07, "loss": 0.0236, "step": 2120 }, { "epoch": 30.0, "grad_norm": 0.19389575719833374, "learning_rate": 0.0, "loss": 0.0303, "step": 2130 }, { "epoch": 30.0, "step": 2130, "total_flos": 2.6364665116584346e+18, "train_loss": 0.09721689578932775, "train_runtime": 614.0464, "train_samples_per_second": 55.403, "train_steps_per_second": 3.469 } ], "logging_steps": 10, "max_steps": 2130, "num_input_tokens_seen": 0, "num_train_epochs": 30, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.6364665116584346e+18, "train_batch_size": 16, "trial_name": null, "trial_params": null }