| { | |
| "best_metric": 0.49546384811401367, | |
| "best_model_checkpoint": "./vit-lr-poly/checkpoint-500", | |
| "epoch": 4.672897196261682, | |
| "eval_steps": 100, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.947947025299072, | |
| "learning_rate": 9.997199065420561e-05, | |
| "loss": 1.348, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.561226844787598, | |
| "learning_rate": 9.994086915887852e-05, | |
| "loss": 0.7531, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.3946990966796875, | |
| "learning_rate": 9.99097476635514e-05, | |
| "loss": 0.828, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 8.365641593933105, | |
| "learning_rate": 9.987862616822431e-05, | |
| "loss": 0.9812, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.398233413696289, | |
| "learning_rate": 9.98475046728972e-05, | |
| "loss": 0.777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 4.742531776428223, | |
| "learning_rate": 9.98163831775701e-05, | |
| "loss": 0.8676, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 3.7602603435516357, | |
| "learning_rate": 9.9785261682243e-05, | |
| "loss": 0.8795, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.804478406906128, | |
| "learning_rate": 9.975414018691589e-05, | |
| "loss": 0.6726, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 4.296436309814453, | |
| "learning_rate": 9.97230186915888e-05, | |
| "loss": 0.6083, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 4.51690149307251, | |
| "learning_rate": 9.969189719626168e-05, | |
| "loss": 0.629, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_accuracy": 0.7669902912621359, | |
| "eval_f1": 0.756703510675833, | |
| "eval_loss": 0.6392337083816528, | |
| "eval_precision": 0.7767663445122107, | |
| "eval_recall": 0.7669902912621359, | |
| "eval_runtime": 38.5968, | |
| "eval_samples_per_second": 74.721, | |
| "eval_steps_per_second": 9.353, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 4.135130405426025, | |
| "learning_rate": 9.966077570093459e-05, | |
| "loss": 0.6787, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.923845291137695, | |
| "learning_rate": 9.962965420560748e-05, | |
| "loss": 0.6674, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.906153678894043, | |
| "learning_rate": 9.959853271028038e-05, | |
| "loss": 0.6661, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.34134840965271, | |
| "learning_rate": 9.956741121495327e-05, | |
| "loss": 0.5255, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 3.7155816555023193, | |
| "learning_rate": 9.953628971962617e-05, | |
| "loss": 0.5715, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.553337574005127, | |
| "learning_rate": 9.950516822429908e-05, | |
| "loss": 0.5912, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 3.132343292236328, | |
| "learning_rate": 9.947404672897196e-05, | |
| "loss": 0.5385, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.8141937255859375, | |
| "learning_rate": 9.944292523364487e-05, | |
| "loss": 0.6023, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.6527509689331055, | |
| "learning_rate": 9.941180373831776e-05, | |
| "loss": 0.6019, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 5.273530960083008, | |
| "learning_rate": 9.938068224299066e-05, | |
| "loss": 0.5273, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "eval_accuracy": 0.7534674063800277, | |
| "eval_f1": 0.6993989038210382, | |
| "eval_loss": 0.7336179614067078, | |
| "eval_precision": 0.757074834598822, | |
| "eval_recall": 0.7534674063800277, | |
| "eval_runtime": 39.1699, | |
| "eval_samples_per_second": 73.628, | |
| "eval_steps_per_second": 9.216, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 5.3881072998046875, | |
| "learning_rate": 9.934956074766355e-05, | |
| "loss": 0.6564, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.86968994140625, | |
| "learning_rate": 9.931843925233646e-05, | |
| "loss": 0.7071, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.7438251972198486, | |
| "learning_rate": 9.928731775700934e-05, | |
| "loss": 0.4032, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.021817922592163, | |
| "learning_rate": 9.925619626168225e-05, | |
| "loss": 0.4928, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.439582347869873, | |
| "learning_rate": 9.922507476635515e-05, | |
| "loss": 0.5924, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.8290791511535645, | |
| "learning_rate": 9.919395327102804e-05, | |
| "loss": 0.5184, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.504387617111206, | |
| "learning_rate": 9.916283177570094e-05, | |
| "loss": 0.5346, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.36114501953125, | |
| "learning_rate": 9.913171028037383e-05, | |
| "loss": 0.5238, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 6.9440789222717285, | |
| "learning_rate": 9.910058878504674e-05, | |
| "loss": 0.6285, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 7.255244255065918, | |
| "learning_rate": 9.906946728971962e-05, | |
| "loss": 0.7441, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_accuracy": 0.7877947295423023, | |
| "eval_f1": 0.7930213802890876, | |
| "eval_loss": 0.6108404994010925, | |
| "eval_precision": 0.8256955400591147, | |
| "eval_recall": 0.7877947295423023, | |
| "eval_runtime": 39.4096, | |
| "eval_samples_per_second": 73.18, | |
| "eval_steps_per_second": 9.16, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.329797744750977, | |
| "learning_rate": 9.903834579439253e-05, | |
| "loss": 0.5261, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.1623640060424805, | |
| "learning_rate": 9.900722429906543e-05, | |
| "loss": 0.4803, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 5.53623628616333, | |
| "learning_rate": 9.897610280373832e-05, | |
| "loss": 0.4676, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 3.928919792175293, | |
| "learning_rate": 9.894498130841122e-05, | |
| "loss": 0.3781, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 4.361560344696045, | |
| "learning_rate": 9.891385981308411e-05, | |
| "loss": 0.4148, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 3.3351240158081055, | |
| "learning_rate": 9.888273831775702e-05, | |
| "loss": 0.3451, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 3.686657428741455, | |
| "learning_rate": 9.88516168224299e-05, | |
| "loss": 0.4263, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 6.783417701721191, | |
| "learning_rate": 9.882049532710281e-05, | |
| "loss": 0.3546, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 4.0723042488098145, | |
| "learning_rate": 9.878937383177571e-05, | |
| "loss": 0.4317, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.952225923538208, | |
| "learning_rate": 9.87582523364486e-05, | |
| "loss": 0.361, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_accuracy": 0.7916088765603329, | |
| "eval_f1": 0.7985905987510713, | |
| "eval_loss": 0.5592741370201111, | |
| "eval_precision": 0.8100907761912617, | |
| "eval_recall": 0.7916088765603329, | |
| "eval_runtime": 38.6146, | |
| "eval_samples_per_second": 74.687, | |
| "eval_steps_per_second": 9.349, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 3.673111915588379, | |
| "learning_rate": 9.87271308411215e-05, | |
| "loss": 0.4086, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 4.770867824554443, | |
| "learning_rate": 9.86960093457944e-05, | |
| "loss": 0.401, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 5.641880512237549, | |
| "learning_rate": 9.866488785046729e-05, | |
| "loss": 0.5471, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 4.864037036895752, | |
| "learning_rate": 9.863376635514018e-05, | |
| "loss": 0.3733, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 5.797776699066162, | |
| "learning_rate": 9.860264485981309e-05, | |
| "loss": 0.4213, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 4.184650421142578, | |
| "learning_rate": 9.857152336448599e-05, | |
| "loss": 0.3889, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 3.752979278564453, | |
| "learning_rate": 9.854040186915888e-05, | |
| "loss": 0.3001, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 6.5187506675720215, | |
| "learning_rate": 9.850928037383178e-05, | |
| "loss": 0.4106, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 3.3088672161102295, | |
| "learning_rate": 9.847815887850469e-05, | |
| "loss": 0.3579, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.724492073059082, | |
| "learning_rate": 9.844703738317757e-05, | |
| "loss": 0.3435, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "eval_accuracy": 0.8335644937586685, | |
| "eval_f1": 0.8177554455408812, | |
| "eval_loss": 0.49546384811401367, | |
| "eval_precision": 0.8250505389448817, | |
| "eval_recall": 0.8335644937586685, | |
| "eval_runtime": 38.6588, | |
| "eval_samples_per_second": 74.601, | |
| "eval_steps_per_second": 9.338, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 3.606675386428833, | |
| "learning_rate": 9.841591588785048e-05, | |
| "loss": 0.4253, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 4.300064563751221, | |
| "learning_rate": 9.838479439252337e-05, | |
| "loss": 0.4158, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 6.155355453491211, | |
| "learning_rate": 9.835367289719627e-05, | |
| "loss": 0.4208, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 3.8082869052886963, | |
| "learning_rate": 9.832255140186916e-05, | |
| "loss": 0.3003, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 5.345461368560791, | |
| "learning_rate": 9.829142990654206e-05, | |
| "loss": 0.447, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 4.557251930236816, | |
| "learning_rate": 9.826030841121497e-05, | |
| "loss": 0.4203, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 2.788506507873535, | |
| "learning_rate": 9.822918691588785e-05, | |
| "loss": 0.3189, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 3.37888765335083, | |
| "learning_rate": 9.819806542056076e-05, | |
| "loss": 0.3213, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 4.868982791900635, | |
| "learning_rate": 9.816694392523365e-05, | |
| "loss": 0.3754, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 3.80576491355896, | |
| "learning_rate": 9.813582242990655e-05, | |
| "loss": 0.2947, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.800624133148405, | |
| "eval_f1": 0.8016421673345813, | |
| "eval_loss": 0.5583807229995728, | |
| "eval_precision": 0.8392905691478968, | |
| "eval_recall": 0.800624133148405, | |
| "eval_runtime": 39.2558, | |
| "eval_samples_per_second": 73.467, | |
| "eval_steps_per_second": 9.196, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 6.513406753540039, | |
| "learning_rate": 9.810470093457944e-05, | |
| "loss": 0.4176, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 3.1807796955108643, | |
| "learning_rate": 9.807357943925234e-05, | |
| "loss": 0.4386, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.8767467737197876, | |
| "learning_rate": 9.804245794392523e-05, | |
| "loss": 0.2326, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 3.7638049125671387, | |
| "learning_rate": 9.801133644859813e-05, | |
| "loss": 0.3639, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.9639334082603455, | |
| "learning_rate": 9.798021495327104e-05, | |
| "loss": 0.2136, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 3.025496006011963, | |
| "learning_rate": 9.794909345794393e-05, | |
| "loss": 0.191, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 4.878714084625244, | |
| "learning_rate": 9.791797196261683e-05, | |
| "loss": 0.2566, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 4.764608860015869, | |
| "learning_rate": 9.788685046728972e-05, | |
| "loss": 0.1857, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.8398491144180298, | |
| "learning_rate": 9.785572897196263e-05, | |
| "loss": 0.1564, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 5.432285785675049, | |
| "learning_rate": 9.782460747663551e-05, | |
| "loss": 0.1172, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "eval_accuracy": 0.8110263522884882, | |
| "eval_f1": 0.8158669905337602, | |
| "eval_loss": 0.6123086810112, | |
| "eval_precision": 0.826380253298927, | |
| "eval_recall": 0.8110263522884882, | |
| "eval_runtime": 38.9234, | |
| "eval_samples_per_second": 74.094, | |
| "eval_steps_per_second": 9.275, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 5.6711225509643555, | |
| "learning_rate": 9.779348598130842e-05, | |
| "loss": 0.189, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 5.6579060554504395, | |
| "learning_rate": 9.776236448598132e-05, | |
| "loss": 0.3117, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 3.8705556392669678, | |
| "learning_rate": 9.77312429906542e-05, | |
| "loss": 0.367, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 2.6670608520507812, | |
| "learning_rate": 9.770012149532711e-05, | |
| "loss": 0.1847, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 4.425388813018799, | |
| "learning_rate": 9.7669e-05, | |
| "loss": 0.117, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 7.775908470153809, | |
| "learning_rate": 9.763787850467291e-05, | |
| "loss": 0.1914, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 9.93289566040039, | |
| "learning_rate": 9.76067570093458e-05, | |
| "loss": 0.2768, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 4.8882222175598145, | |
| "learning_rate": 9.75756355140187e-05, | |
| "loss": 0.2216, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 4.0092034339904785, | |
| "learning_rate": 9.75445140186916e-05, | |
| "loss": 0.3954, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 3.865039587020874, | |
| "learning_rate": 9.751339252336449e-05, | |
| "loss": 0.234, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "eval_accuracy": 0.8304438280166435, | |
| "eval_f1": 0.8275562687216345, | |
| "eval_loss": 0.5097168684005737, | |
| "eval_precision": 0.8316099910356197, | |
| "eval_recall": 0.8304438280166435, | |
| "eval_runtime": 38.9623, | |
| "eval_samples_per_second": 74.02, | |
| "eval_steps_per_second": 9.265, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 3.957510471343994, | |
| "learning_rate": 9.748227102803739e-05, | |
| "loss": 0.3797, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 9.32695198059082, | |
| "learning_rate": 9.745114953271028e-05, | |
| "loss": 0.2094, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 2.3742001056671143, | |
| "learning_rate": 9.742002803738318e-05, | |
| "loss": 0.224, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 10.046113967895508, | |
| "learning_rate": 9.738890654205607e-05, | |
| "loss": 0.1771, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 3.123710870742798, | |
| "learning_rate": 9.735778504672898e-05, | |
| "loss": 0.1547, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 5.135457515716553, | |
| "learning_rate": 9.732666355140188e-05, | |
| "loss": 0.2084, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 5.636462211608887, | |
| "learning_rate": 9.729554205607477e-05, | |
| "loss": 0.283, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 4.4640092849731445, | |
| "learning_rate": 9.726442056074767e-05, | |
| "loss": 0.2871, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 0.6387704014778137, | |
| "learning_rate": 9.723329906542056e-05, | |
| "loss": 0.2594, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 5.557553768157959, | |
| "learning_rate": 9.720217757009346e-05, | |
| "loss": 0.1469, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_accuracy": 0.8498613037447988, | |
| "eval_f1": 0.8503456387739566, | |
| "eval_loss": 0.4982721507549286, | |
| "eval_precision": 0.8558896203464617, | |
| "eval_recall": 0.8498613037447988, | |
| "eval_runtime": 38.9879, | |
| "eval_samples_per_second": 73.972, | |
| "eval_steps_per_second": 9.259, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.5195624828338623, | |
| "learning_rate": 9.717105607476635e-05, | |
| "loss": 0.317, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 6.157825946807861, | |
| "learning_rate": 9.713993457943926e-05, | |
| "loss": 0.3257, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.9414964318275452, | |
| "learning_rate": 9.710881308411214e-05, | |
| "loss": 0.2491, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 7.8883748054504395, | |
| "learning_rate": 9.707769158878505e-05, | |
| "loss": 0.2412, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 4.660026550292969, | |
| "learning_rate": 9.704657009345795e-05, | |
| "loss": 0.2342, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 7.779811859130859, | |
| "learning_rate": 9.701544859813086e-05, | |
| "loss": 0.2307, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 1.2840813398361206, | |
| "learning_rate": 9.698432710280374e-05, | |
| "loss": 0.1261, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.9387753009796143, | |
| "learning_rate": 9.695320560747665e-05, | |
| "loss": 0.0882, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 1.9628775119781494, | |
| "learning_rate": 9.692208411214954e-05, | |
| "loss": 0.0512, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 4.132315635681152, | |
| "learning_rate": 9.689096261682244e-05, | |
| "loss": 0.0183, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "eval_accuracy": 0.8432732316227461, | |
| "eval_f1": 0.8351311035032215, | |
| "eval_loss": 0.6637601852416992, | |
| "eval_precision": 0.8438934125461472, | |
| "eval_recall": 0.8432732316227461, | |
| "eval_runtime": 38.9722, | |
| "eval_samples_per_second": 74.002, | |
| "eval_steps_per_second": 9.263, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.1714770644903183, | |
| "learning_rate": 9.685984112149533e-05, | |
| "loss": 0.0634, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.29727110266685486, | |
| "learning_rate": 9.682871962616823e-05, | |
| "loss": 0.1281, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 2.162074089050293, | |
| "learning_rate": 9.679759813084112e-05, | |
| "loss": 0.0457, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 1.9378799200057983, | |
| "learning_rate": 9.676647663551402e-05, | |
| "loss": 0.065, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 3.3368661403656006, | |
| "learning_rate": 9.673535514018693e-05, | |
| "loss": 0.0914, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.2931816577911377, | |
| "learning_rate": 9.670423364485982e-05, | |
| "loss": 0.1102, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 3.579129934310913, | |
| "learning_rate": 9.667311214953272e-05, | |
| "loss": 0.1033, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 3.30364990234375, | |
| "learning_rate": 9.664199065420561e-05, | |
| "loss": 0.1055, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 2.374694585800171, | |
| "learning_rate": 9.661086915887851e-05, | |
| "loss": 0.0484, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 6.944613933563232, | |
| "learning_rate": 9.65797476635514e-05, | |
| "loss": 0.0978, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "eval_accuracy": 0.8425797503467406, | |
| "eval_f1": 0.8435126471008092, | |
| "eval_loss": 0.6589659452438354, | |
| "eval_precision": 0.8554481873448666, | |
| "eval_recall": 0.8425797503467406, | |
| "eval_runtime": 39.1512, | |
| "eval_samples_per_second": 73.663, | |
| "eval_steps_per_second": 9.221, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 5.054389953613281, | |
| "learning_rate": 9.65486261682243e-05, | |
| "loss": 0.0369, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 8.978492736816406, | |
| "learning_rate": 9.65175046728972e-05, | |
| "loss": 0.1993, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 2.9282114505767822, | |
| "learning_rate": 9.648638317757009e-05, | |
| "loss": 0.0738, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 8.38040542602539, | |
| "learning_rate": 9.6455261682243e-05, | |
| "loss": 0.1215, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 2.709808111190796, | |
| "learning_rate": 9.642414018691589e-05, | |
| "loss": 0.1876, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 7.647227764129639, | |
| "learning_rate": 9.63930186915888e-05, | |
| "loss": 0.2752, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.319821298122406, | |
| "learning_rate": 9.636189719626168e-05, | |
| "loss": 0.1502, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.14174489676952362, | |
| "learning_rate": 9.633077570093459e-05, | |
| "loss": 0.1044, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 5.115747451782227, | |
| "learning_rate": 9.629965420560749e-05, | |
| "loss": 0.0813, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 4.306972980499268, | |
| "learning_rate": 9.626853271028037e-05, | |
| "loss": 0.0953, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "eval_accuracy": 0.8370319001386962, | |
| "eval_f1": 0.8233052481689616, | |
| "eval_loss": 0.7363972067832947, | |
| "eval_precision": 0.8353574343391107, | |
| "eval_recall": 0.8370319001386962, | |
| "eval_runtime": 38.857, | |
| "eval_samples_per_second": 74.221, | |
| "eval_steps_per_second": 9.29, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 2.1228132247924805, | |
| "learning_rate": 9.623741121495328e-05, | |
| "loss": 0.1308, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 8.311467170715332, | |
| "learning_rate": 9.620628971962617e-05, | |
| "loss": 0.3034, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 1.324053406715393, | |
| "learning_rate": 9.617516822429907e-05, | |
| "loss": 0.1106, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 6.312448501586914, | |
| "learning_rate": 9.614404672897196e-05, | |
| "loss": 0.1456, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 1.1473090648651123, | |
| "learning_rate": 9.611292523364487e-05, | |
| "loss": 0.1003, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 5.944868564605713, | |
| "learning_rate": 9.608180373831777e-05, | |
| "loss": 0.1514, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 5.833784103393555, | |
| "learning_rate": 9.605068224299066e-05, | |
| "loss": 0.181, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 5.8773651123046875, | |
| "learning_rate": 9.601956074766356e-05, | |
| "loss": 0.134, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.10139019042253494, | |
| "learning_rate": 9.598843925233645e-05, | |
| "loss": 0.0281, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 7.28130578994751, | |
| "learning_rate": 9.595731775700935e-05, | |
| "loss": 0.0949, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "eval_accuracy": 0.8432732316227461, | |
| "eval_f1": 0.8314756389957325, | |
| "eval_loss": 0.6495711803436279, | |
| "eval_precision": 0.8405555314046315, | |
| "eval_recall": 0.8432732316227461, | |
| "eval_runtime": 38.8063, | |
| "eval_samples_per_second": 74.318, | |
| "eval_steps_per_second": 9.303, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.12207405269145966, | |
| "learning_rate": 9.592619626168224e-05, | |
| "loss": 0.0667, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 0.08908121287822723, | |
| "learning_rate": 9.589507476635515e-05, | |
| "loss": 0.0737, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.17971865832805634, | |
| "learning_rate": 9.586395327102805e-05, | |
| "loss": 0.0402, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 1.6679821014404297, | |
| "learning_rate": 9.583283177570094e-05, | |
| "loss": 0.0372, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 0.08169826865196228, | |
| "learning_rate": 9.580171028037384e-05, | |
| "loss": 0.1155, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.12340045720338821, | |
| "learning_rate": 9.577058878504673e-05, | |
| "loss": 0.0612, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.4197162389755249, | |
| "learning_rate": 9.573946728971963e-05, | |
| "loss": 0.0637, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 7.830495834350586, | |
| "learning_rate": 9.570834579439252e-05, | |
| "loss": 0.1285, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 0.09970329701900482, | |
| "learning_rate": 9.567722429906543e-05, | |
| "loss": 0.077, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 6.296285152435303, | |
| "learning_rate": 9.564610280373831e-05, | |
| "loss": 0.0416, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "eval_accuracy": 0.8488210818307905, | |
| "eval_f1": 0.8438647450714564, | |
| "eval_loss": 0.6833219528198242, | |
| "eval_precision": 0.8472844358014768, | |
| "eval_recall": 0.8488210818307905, | |
| "eval_runtime": 39.0336, | |
| "eval_samples_per_second": 73.885, | |
| "eval_steps_per_second": 9.248, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 5.955536842346191, | |
| "learning_rate": 9.561498130841122e-05, | |
| "loss": 0.0398, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 9.86475944519043, | |
| "learning_rate": 9.558385981308412e-05, | |
| "loss": 0.0837, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.0481799840927124, | |
| "learning_rate": 9.555273831775701e-05, | |
| "loss": 0.0128, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 0.00958944670855999, | |
| "learning_rate": 9.552161682242991e-05, | |
| "loss": 0.0175, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 1.5484673976898193, | |
| "learning_rate": 9.549049532710282e-05, | |
| "loss": 0.0854, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 0.02997567690908909, | |
| "learning_rate": 9.545937383177571e-05, | |
| "loss": 0.0169, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 1.262243390083313, | |
| "learning_rate": 9.54282523364486e-05, | |
| "loss": 0.0829, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 0.29552406072616577, | |
| "learning_rate": 9.53971308411215e-05, | |
| "loss": 0.0509, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 2.321814775466919, | |
| "learning_rate": 9.53660093457944e-05, | |
| "loss": 0.0871, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 2.159813404083252, | |
| "learning_rate": 9.533488785046729e-05, | |
| "loss": 0.0609, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_accuracy": 0.8554091539528432, | |
| "eval_f1": 0.8522386861505039, | |
| "eval_loss": 0.721152126789093, | |
| "eval_precision": 0.8555126000385453, | |
| "eval_recall": 0.8554091539528432, | |
| "eval_runtime": 39.1072, | |
| "eval_samples_per_second": 73.746, | |
| "eval_steps_per_second": 9.231, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "step": 1500, | |
| "total_flos": 1.8574112402997903e+18, | |
| "train_loss": 0.3071586556136608, | |
| "train_runtime": 1184.6117, | |
| "train_samples_per_second": 432.884, | |
| "train_steps_per_second": 27.097 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 100, | |
| "total_flos": 1.8574112402997903e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |