| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 2358, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6949152542372882e-06, | |
| "loss": 1.9885, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 3.3898305084745763e-06, | |
| "loss": 2.7383, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5.084745762711865e-06, | |
| "loss": 2.5169, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 6.779661016949153e-06, | |
| "loss": 2.3984, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 8.47457627118644e-06, | |
| "loss": 1.9978, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 1.016949152542373e-05, | |
| "loss": 2.2204, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 1.1864406779661018e-05, | |
| "loss": 1.9677, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.3559322033898305e-05, | |
| "loss": 2.2108, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.5254237288135594e-05, | |
| "loss": 1.5879, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 1.694915254237288e-05, | |
| "loss": 2.2145, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 1.864406779661017e-05, | |
| "loss": 2.0709, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 1.9999960660084782e-05, | |
| "loss": 2.4549, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.9998583795552086e-05, | |
| "loss": 2.0769, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1.9995240244773315e-05, | |
| "loss": 1.801, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 1.9989930665413148e-05, | |
| "loss": 1.9293, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.998265610184716e-05, | |
| "loss": 1.9712, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.9973417984956404e-05, | |
| "loss": 2.4024, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 1.996221813184596e-05, | |
| "loss": 2.1186, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 1.9949058745487524e-05, | |
| "loss": 2.1329, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 1.9933942414286082e-05, | |
| "loss": 1.8044, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.9916872111570782e-05, | |
| "loss": 2.5074, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 1.98978511950101e-05, | |
| "loss": 1.6585, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.9876883405951378e-05, | |
| "loss": 1.9255, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9853972868684942e-05, | |
| "loss": 1.5736, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.982912408963285e-05, | |
| "loss": 1.6555, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.980234195646248e-05, | |
| "loss": 2.1425, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.9773631737125192e-05, | |
| "loss": 1.1852, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.974299907882009e-05, | |
| "loss": 2.1502, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.9710450006883258e-05, | |
| "loss": 1.4109, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.96759909236026e-05, | |
| "loss": 1.7527, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.9639628606958535e-05, | |
| "loss": 1.2232, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.960137020929078e-05, | |
| "loss": 2.0229, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 1.956122325589152e-05, | |
| "loss": 1.5771, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.9519195643525205e-05, | |
| "loss": 1.5255, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.947529563887529e-05, | |
| "loss": 1.6161, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.9429531876918193e-05, | |
| "loss": 1.9704, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.9381913359224844e-05, | |
| "loss": 1.2751, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.9332449452190096e-05, | |
| "loss": 1.6944, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.928114988519039e-05, | |
| "loss": 1.0435, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4892086330935252, | |
| "eval_loss": 1.4798991680145264, | |
| "eval_runtime": 14.722, | |
| "eval_samples_per_second": 9.442, | |
| "eval_steps_per_second": 2.377, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.922802474867003e-05, | |
| "loss": 1.9562, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 1.9173084492156406e-05, | |
| "loss": 1.6383, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 1.911633992220464e-05, | |
| "loss": 1.1736, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 1.9057802200271943e-05, | |
| "loss": 1.0087, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 1.8997482840522218e-05, | |
| "loss": 1.5162, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 1.893539370756125e-05, | |
| "loss": 1.2644, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 1.8871547014103e-05, | |
| "loss": 1.2083, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 1.880595531856738e-05, | |
| "loss": 1.0534, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.8738631522610077e-05, | |
| "loss": 0.939, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 1.8669588868584832e-05, | |
| "loss": 1.4496, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 1.859884093693872e-05, | |
| "loss": 1.2077, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.8526401643540924e-05, | |
| "loss": 1.1323, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 1.845228523694553e-05, | |
| "loss": 1.2509, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 1.8376506295588887e-05, | |
| "loss": 1.2086, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 1.8299079724922087e-05, | |
| "loss": 1.4408, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 1.8220020754479104e-05, | |
| "loss": 0.8179, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 1.8139344934881195e-05, | |
| "loss": 0.831, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 1.8057068134778174e-05, | |
| "loss": 1.8693, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 1.797320653772707e-05, | |
| "loss": 0.7718, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.7887776639008912e-05, | |
| "loss": 1.3469, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 1.7800795242384126e-05, | |
| "loss": 1.297, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 1.7712279456787334e-05, | |
| "loss": 1.0095, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 1.762224669296206e-05, | |
| "loss": 1.8096, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.7530714660036112e-05, | |
| "loss": 1.1618, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 1.743770136203825e-05, | |
| "loss": 1.0257, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 1.7343225094356857e-05, | |
| "loss": 1.644, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 1.724730444014132e-05, | |
| "loss": 1.4658, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 1.7149958266646756e-05, | |
| "loss": 1.1911, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 1.705120572152292e-05, | |
| "loss": 1.1788, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 1.695106622904791e-05, | |
| "loss": 1.56, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 1.6849559486307493e-05, | |
| "loss": 1.5506, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 1.6746705459320746e-05, | |
| "loss": 1.3003, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.664252437911282e-05, | |
| "loss": 1.2845, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 1.6537036737735563e-05, | |
| "loss": 1.9069, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 1.6430263284236803e-05, | |
| "loss": 1.178, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 1.63222250205791e-05, | |
| "loss": 1.3253, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 1.6212943197508706e-05, | |
| "loss": 1.0718, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 1.6102439310375637e-05, | |
| "loss": 0.9187, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 1.5990735094905622e-05, | |
| "loss": 1.1554, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4892086330935252, | |
| "eval_loss": 1.4938154220581055, | |
| "eval_runtime": 14.7051, | |
| "eval_samples_per_second": 9.453, | |
| "eval_steps_per_second": 2.38, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 1.1346, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.576381379803763e-05, | |
| "loss": 1.2287, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 1.564864135126018e-05, | |
| "loss": 1.02, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 1.553235783660735e-05, | |
| "loss": 1.2653, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 1.541498612663726e-05, | |
| "loss": 0.7407, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 1.5296549307952203e-05, | |
| "loss": 1.2226, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 1.5177070676657632e-05, | |
| "loss": 0.9274, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.5056573733779848e-05, | |
| "loss": 0.9529, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.493508218064347e-05, | |
| "loss": 0.8886, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.4812619914209455e-05, | |
| "loss": 1.0895, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.4689211022374645e-05, | |
| "loss": 1.669, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 1.4564879779233773e-05, | |
| "loss": 1.1092, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.4439650640304822e-05, | |
| "loss": 1.1258, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 1.431354823771872e-05, | |
| "loss": 0.8034, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 1.4186597375374283e-05, | |
| "loss": 1.1004, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 1.4058823024059374e-05, | |
| "loss": 0.8297, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.3930250316539237e-05, | |
| "loss": 0.8974, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.380090454261298e-05, | |
| "loss": 1.3636, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 1.3670811144139144e-05, | |
| "loss": 0.9185, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 1.3539995710031391e-05, | |
| "loss": 0.8012, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 1.340848397122525e-05, | |
| "loss": 0.9984, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 1.3276301795616937e-05, | |
| "loss": 0.8015, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 1.3143475182975222e-05, | |
| "loss": 0.9148, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.3010030259827378e-05, | |
| "loss": 1.1759, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 1.2875993274320173e-05, | |
| "loss": 0.606, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 1.2741390591056964e-05, | |
| "loss": 1.253, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 1.2606248685911856e-05, | |
| "loss": 1.1392, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 1.2470594140822008e-05, | |
| "loss": 0.9999, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 1.2334453638559057e-05, | |
| "loss": 1.0181, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 1.2197853957480689e-05, | |
| "loss": 0.9913, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 1.2060821966263465e-05, | |
| "loss": 0.7443, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 1.1923384618617819e-05, | |
| "loss": 0.9462, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 1.1785568947986368e-05, | |
| "loss": 1.6673, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 1.1647402062226534e-05, | |
| "loss": 1.494, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 1.1508911138278516e-05, | |
| "loss": 1.3606, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.1370123416819683e-05, | |
| "loss": 0.5196, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.123106619690643e-05, | |
| "loss": 1.0671, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 1.1091766830604552e-05, | |
| "loss": 1.1892, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 1.0952252717609188e-05, | |
| "loss": 1.5041, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.3597122302158273, | |
| "eval_loss": 2.1702427864074707, | |
| "eval_runtime": 14.7195, | |
| "eval_samples_per_second": 9.443, | |
| "eval_steps_per_second": 2.378, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.0812551299855389e-05, | |
| "loss": 1.1424, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 1.0672690056120398e-05, | |
| "loss": 0.9707, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 1.0532696496618664e-05, | |
| "loss": 0.6855, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 1.0392598157590687e-05, | |
| "loss": 0.362, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 1.0252422595886723e-05, | |
| "loss": 1.2138, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 1.011219738354646e-05, | |
| "loss": 0.4679, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 9.971950102375696e-06, | |
| "loss": 0.4326, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 9.83170833852108e-06, | |
| "loss": 0.4858, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 9.691499677044032e-06, | |
| "loss": 0.6772, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 9.551351696494854e-06, | |
| "loss": 0.5304, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 9.41129196348811e-06, | |
| "loss": 0.6597, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 9.2713480272804e-06, | |
| "loss": 0.8794, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 9.131547414351486e-06, | |
| "loss": 0.4008, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 8.991917622989957e-06, | |
| "loss": 0.9578, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 8.852486117884391e-06, | |
| "loss": 0.8909, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "learning_rate": 8.713280324721163e-06, | |
| "loss": 0.7782, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 8.574327624789904e-06, | |
| "loss": 0.6601, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 8.43565534959769e-06, | |
| "loss": 0.597, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 8.297290775493051e-06, | |
| "loss": 0.6266, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 8.159261118300784e-06, | |
| "loss": 0.4856, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 8.02159352796871e-06, | |
| "loss": 0.5143, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 7.884315083227373e-06, | |
| "loss": 0.4511, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 7.747452786263755e-06, | |
| "loss": 0.7128, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 7.611033557410027e-06, | |
| "loss": 0.6028, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 7.4750842298484205e-06, | |
| "loss": 0.7882, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 7.33963154433325e-06, | |
| "loss": 0.6686, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 7.204702143931072e-06, | |
| "loss": 0.7159, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 7.070322568780111e-06, | |
| "loss": 1.4071, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 6.9365192508698975e-06, | |
| "loss": 0.9269, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 6.8033185088421874e-06, | |
| "loss": 1.0381, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 6.670746542814187e-06, | |
| "loss": 0.9117, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 6.538829429225068e-06, | |
| "loss": 0.6995, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 6.407593115706849e-06, | |
| "loss": 1.5987, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 6.277063415980549e-06, | |
| "loss": 0.7295, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 6.147266004778755e-06, | |
| "loss": 0.1654, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 6.018226412795469e-06, | |
| "loss": 0.7961, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 5.889970021664329e-06, | |
| "loss": 0.6862, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 5.7625220589661136e-06, | |
| "loss": 0.3667, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 5.635907593266578e-06, | |
| "loss": 0.8349, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 5.510151529185554e-06, | |
| "loss": 1.0457, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5683453237410072, | |
| "eval_loss": 1.5413200855255127, | |
| "eval_runtime": 14.7224, | |
| "eval_samples_per_second": 9.441, | |
| "eval_steps_per_second": 2.377, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 5.385278602498288e-06, | |
| "loss": 0.762, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 5.2613133752700145e-06, | |
| "loss": 0.4267, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 5.138280231024676e-06, | |
| "loss": 0.6985, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 5.016203369948774e-06, | |
| "loss": 0.5179, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 4.895106804131274e-06, | |
| "loss": 0.7994, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 4.775014352840512e-06, | |
| "loss": 0.3598, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 4.655949637839042e-06, | |
| "loss": 0.5081, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 4.537936078737291e-06, | |
| "loss": 0.956, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "learning_rate": 4.420996888387033e-06, | |
| "loss": 1.2839, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "learning_rate": 4.305155068315481e-06, | |
| "loss": 0.6122, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 4.190433404200977e-06, | |
| "loss": 1.3103, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "learning_rate": 4.07685446139113e-06, | |
| "loss": 0.6164, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 3.964440580464286e-06, | |
| "loss": 0.9398, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 3.853213872835229e-06, | |
| "loss": 0.1861, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "learning_rate": 3.743196216405922e-06, | |
| "loss": 0.543, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3.6344092512622173e-06, | |
| "loss": 0.6261, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 3.5268743754173174e-06, | |
| "loss": 0.3409, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "learning_rate": 3.4206127406028744e-06, | |
| "loss": 0.4565, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 3.3156452481085056e-06, | |
| "loss": 0.4853, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 3.2119925446705824e-06, | |
| "loss": 0.3187, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "learning_rate": 3.109675018411108e-06, | |
| "loss": 0.5515, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 3.0087127948274264e-06, | |
| "loss": 0.8492, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 2.909125732833612e-06, | |
| "loss": 0.8423, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 2.8109334208542994e-06, | |
| "loss": 0.7927, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "learning_rate": 2.7141551729716974e-06, | |
| "loss": 0.2295, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 2.6188100251265947e-06, | |
| "loss": 0.6877, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "learning_rate": 2.5249167313740307e-06, | |
| "loss": 0.6192, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "learning_rate": 2.432493760194469e-06, | |
| "loss": 0.4079, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 2.3415592908610874e-06, | |
| "loss": 0.1516, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 2.2521312098639914e-06, | |
| "loss": 0.8265, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "learning_rate": 2.164227107392002e-06, | |
| "loss": 0.4464, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 2.077864273872727e-06, | |
| "loss": 0.8132, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 1.9930596965715985e-06, | |
| "loss": 0.6039, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.3869, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 1.8281917238868707e-06, | |
| "loss": 0.2218, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "learning_rate": 1.7481607574533044e-06, | |
| "loss": 0.0394, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 1.6697528987592893e-06, | |
| "loss": 0.5701, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 1.5929835703546992e-06, | |
| "loss": 0.8708, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "learning_rate": 1.517867872496267e-06, | |
| "loss": 0.3315, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.697841726618705, | |
| "eval_loss": 1.0769166946411133, | |
| "eval_runtime": 14.7066, | |
| "eval_samples_per_second": 9.452, | |
| "eval_steps_per_second": 2.38, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "learning_rate": 1.4444205801774204e-06, | |
| "loss": 0.8844, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 1.3726561402220818e-06, | |
| "loss": 0.4769, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "learning_rate": 1.3025886684430467e-06, | |
| "loss": 0.4141, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "learning_rate": 1.2342319468654329e-06, | |
| "loss": 0.0528, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 1.167599421015818e-06, | |
| "loss": 0.3157, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "learning_rate": 1.1027041972775442e-06, | |
| "loss": 0.2718, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "learning_rate": 1.0395590403127487e-06, | |
| "loss": 0.5592, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 9.78176370551588e-07, | |
| "loss": 0.4728, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 9.185682617491865e-07, | |
| "loss": 0.3589, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "learning_rate": 8.6074643861078e-07, | |
| "loss": 0.6848, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 8.047222744854943e-07, | |
| "loss": 0.3668, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 7.50506789129255e-07, | |
| "loss": 0.8621, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "learning_rate": 6.981106465372389e-07, | |
| "loss": 0.6104, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 6.475441528463033e-07, | |
| "loss": 0.4872, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "learning_rate": 5.988172543078097e-07, | |
| "loss": 0.3606, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "learning_rate": 5.519395353312195e-07, | |
| "loss": 0.4107, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "learning_rate": 5.069202165988874e-07, | |
| "loss": 0.3428, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "learning_rate": 4.6376815325237966e-07, | |
| "loss": 0.6439, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "learning_rate": 4.224918331506955e-07, | |
| "loss": 0.2382, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "learning_rate": 3.8309937520074283e-07, | |
| "loss": 0.4123, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "learning_rate": 3.455985277603713e-07, | |
| "loss": 0.4865, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "learning_rate": 3.099966671143018e-07, | |
| "loss": 0.378, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 2.7630079602323447e-07, | |
| "loss": 0.3095, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 2.445175423464341e-07, | |
| "loss": 0.113, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "learning_rate": 2.1465315773804617e-07, | |
| "loss": 0.43, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "learning_rate": 1.867135164174283e-07, | |
| "loss": 0.5848, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "learning_rate": 1.6070411401370335e-07, | |
| "loss": 0.4236, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 1.3663006648479415e-07, | |
| "loss": 0.3211, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "learning_rate": 1.1449610911112741e-07, | |
| "loss": 0.4408, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "learning_rate": 9.43065955642275e-08, | |
| "loss": 0.1455, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "learning_rate": 7.606549705035937e-08, | |
| "loss": 0.055, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 5.977640152940556e-08, | |
| "loss": 0.8437, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "learning_rate": 4.544251300913405e-08, | |
| "loss": 0.7393, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "learning_rate": 3.3066650914970186e-08, | |
| "loss": 0.0791, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "learning_rate": 2.265124953543918e-08, | |
| "loss": 0.2199, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "learning_rate": 1.4198357543336871e-08, | |
| "loss": 0.4947, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "learning_rate": 7.70963759277099e-09, | |
| "loss": 0.2422, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 5.95, | |
| "learning_rate": 3.1863659921216316e-09, | |
| "loss": 0.7831, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 6.294324529942942e-10, | |
| "loss": 0.2178, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6762589928057554, | |
| "eval_loss": 1.233504295349121, | |
| "eval_runtime": 14.7199, | |
| "eval_samples_per_second": 9.443, | |
| "eval_steps_per_second": 2.378, | |
| "step": 2358 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 2358, | |
| "total_flos": 6.779516273836425e+18, | |
| "train_loss": 0.9992087238718637, | |
| "train_runtime": 3103.0559, | |
| "train_samples_per_second": 1.52, | |
| "train_steps_per_second": 0.76 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2358, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 6.779516273836425e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |