{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 38.46153846153846, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.038461538461538464, "grad_norm": 0.8129774928092957, "learning_rate": 2.7e-07, "loss": 1.3508, "step": 10 }, { "epoch": 0.07692307692307693, "grad_norm": 0.6433653235435486, "learning_rate": 5.7e-07, "loss": 1.3552, "step": 20 }, { "epoch": 0.11538461538461539, "grad_norm": 0.6810219883918762, "learning_rate": 8.7e-07, "loss": 1.3478, "step": 30 }, { "epoch": 0.15384615384615385, "grad_norm": 0.6231173276901245, "learning_rate": 1.17e-06, "loss": 1.3472, "step": 40 }, { "epoch": 0.19230769230769232, "grad_norm": 0.829264223575592, "learning_rate": 1.4700000000000001e-06, "loss": 1.3416, "step": 50 }, { "epoch": 0.23076923076923078, "grad_norm": 0.9087441563606262, "learning_rate": 1.77e-06, "loss": 1.3256, "step": 60 }, { "epoch": 0.2692307692307692, "grad_norm": 1.0764439105987549, "learning_rate": 2.07e-06, "loss": 1.3159, "step": 70 }, { "epoch": 0.3076923076923077, "grad_norm": 1.0241045951843262, "learning_rate": 2.37e-06, "loss": 1.3131, "step": 80 }, { "epoch": 0.34615384615384615, "grad_norm": 1.0223745107650757, "learning_rate": 2.67e-06, "loss": 1.2953, "step": 90 }, { "epoch": 0.38461538461538464, "grad_norm": 1.3292489051818848, "learning_rate": 2.9700000000000004e-06, "loss": 1.2714, "step": 100 }, { "epoch": 0.4230769230769231, "grad_norm": 1.2786927223205566, "learning_rate": 3.27e-06, "loss": 1.2593, "step": 110 }, { "epoch": 0.46153846153846156, "grad_norm": 1.0608361959457397, "learning_rate": 3.57e-06, "loss": 1.2464, "step": 120 }, { "epoch": 0.5, "grad_norm": 0.9192166328430176, "learning_rate": 3.87e-06, "loss": 1.2222, "step": 130 }, { "epoch": 0.5384615384615384, "grad_norm": 0.8084186315536499, "learning_rate": 4.170000000000001e-06, "loss": 1.2077, "step": 140 }, { "epoch": 0.5769230769230769, "grad_norm": 0.8090039491653442, "learning_rate": 4.4699999999999996e-06, "loss": 1.1963, "step": 150 }, { "epoch": 0.6153846153846154, "grad_norm": 0.7544081807136536, "learning_rate": 4.77e-06, "loss": 1.1736, "step": 160 }, { "epoch": 0.6538461538461539, "grad_norm": 0.7331106662750244, "learning_rate": 5.070000000000001e-06, "loss": 1.1634, "step": 170 }, { "epoch": 0.6923076923076923, "grad_norm": 0.8381819128990173, "learning_rate": 5.37e-06, "loss": 1.154, "step": 180 }, { "epoch": 0.7307692307692307, "grad_norm": 0.8138782382011414, "learning_rate": 5.67e-06, "loss": 1.1325, "step": 190 }, { "epoch": 0.7692307692307693, "grad_norm": 0.7330286502838135, "learning_rate": 5.9700000000000004e-06, "loss": 1.1344, "step": 200 }, { "epoch": 0.8076923076923077, "grad_norm": 0.8166984915733337, "learning_rate": 6.27e-06, "loss": 1.1302, "step": 210 }, { "epoch": 0.8461538461538461, "grad_norm": 0.7504987120628357, "learning_rate": 6.57e-06, "loss": 1.1089, "step": 220 }, { "epoch": 0.8846153846153846, "grad_norm": 0.7398339509963989, "learning_rate": 6.87e-06, "loss": 1.0961, "step": 230 }, { "epoch": 0.9230769230769231, "grad_norm": 0.6914257407188416, "learning_rate": 7.17e-06, "loss": 1.0939, "step": 240 }, { "epoch": 0.9615384615384616, "grad_norm": 0.753593921661377, "learning_rate": 7.4700000000000005e-06, "loss": 1.0844, "step": 250 }, { "epoch": 1.0, "grad_norm": 0.775452733039856, "learning_rate": 7.77e-06, "loss": 1.0757, "step": 260 }, { "epoch": 1.0384615384615385, "grad_norm": 0.8942003846168518, "learning_rate": 8.07e-06, "loss": 1.0727, "step": 270 }, { "epoch": 1.0769230769230769, "grad_norm": 0.8823027610778809, "learning_rate": 8.370000000000001e-06, "loss": 1.073, "step": 280 }, { "epoch": 1.1153846153846154, "grad_norm": 0.7825962901115417, "learning_rate": 8.67e-06, "loss": 1.0681, "step": 290 }, { "epoch": 1.1538461538461537, "grad_norm": 1.042375087738037, "learning_rate": 8.97e-06, "loss": 1.0604, "step": 300 }, { "epoch": 1.1923076923076923, "grad_norm": 0.9922687411308289, "learning_rate": 9.27e-06, "loss": 1.0616, "step": 310 }, { "epoch": 1.2307692307692308, "grad_norm": 0.8586735129356384, "learning_rate": 9.57e-06, "loss": 1.0532, "step": 320 }, { "epoch": 1.2692307692307692, "grad_norm": 0.9278848171234131, "learning_rate": 9.87e-06, "loss": 1.0478, "step": 330 }, { "epoch": 1.3076923076923077, "grad_norm": 0.8375242948532104, "learning_rate": 1.0170000000000001e-05, "loss": 1.0513, "step": 340 }, { "epoch": 1.3461538461538463, "grad_norm": 0.7845088839530945, "learning_rate": 1.047e-05, "loss": 1.0438, "step": 350 }, { "epoch": 1.3846153846153846, "grad_norm": 0.9647091031074524, "learning_rate": 1.077e-05, "loss": 1.0285, "step": 360 }, { "epoch": 1.4230769230769231, "grad_norm": 1.1311838626861572, "learning_rate": 1.107e-05, "loss": 1.023, "step": 370 }, { "epoch": 1.4615384615384617, "grad_norm": 1.1281001567840576, "learning_rate": 1.137e-05, "loss": 1.0075, "step": 380 }, { "epoch": 1.5, "grad_norm": 1.461525321006775, "learning_rate": 1.167e-05, "loss": 0.9779, "step": 390 }, { "epoch": 1.5384615384615383, "grad_norm": 1.534185767173767, "learning_rate": 1.197e-05, "loss": 0.9208, "step": 400 }, { "epoch": 1.5769230769230769, "grad_norm": 1.9967076778411865, "learning_rate": 1.227e-05, "loss": 0.8734, "step": 410 }, { "epoch": 1.6153846153846154, "grad_norm": 1.6220159530639648, "learning_rate": 1.257e-05, "loss": 0.8197, "step": 420 }, { "epoch": 1.6538461538461537, "grad_norm": 1.832636833190918, "learning_rate": 1.287e-05, "loss": 0.771, "step": 430 }, { "epoch": 1.6923076923076923, "grad_norm": 2.2742176055908203, "learning_rate": 1.3170000000000001e-05, "loss": 0.7211, "step": 440 }, { "epoch": 1.7307692307692308, "grad_norm": 1.9380594491958618, "learning_rate": 1.3470000000000001e-05, "loss": 0.6881, "step": 450 }, { "epoch": 1.7692307692307692, "grad_norm": 2.589594602584839, "learning_rate": 1.377e-05, "loss": 0.6551, "step": 460 }, { "epoch": 1.8076923076923077, "grad_norm": 2.73272442817688, "learning_rate": 1.4069999999999999e-05, "loss": 0.64, "step": 470 }, { "epoch": 1.8461538461538463, "grad_norm": 2.789524555206299, "learning_rate": 1.437e-05, "loss": 0.6142, "step": 480 }, { "epoch": 1.8846153846153846, "grad_norm": 2.5425634384155273, "learning_rate": 1.467e-05, "loss": 0.5954, "step": 490 }, { "epoch": 1.9230769230769231, "grad_norm": 2.916409969329834, "learning_rate": 1.497e-05, "loss": 0.574, "step": 500 }, { "epoch": 1.9615384615384617, "grad_norm": 2.5941741466522217, "learning_rate": 1.527e-05, "loss": 0.5612, "step": 510 }, { "epoch": 2.0, "grad_norm": 3.0018510818481445, "learning_rate": 1.5570000000000002e-05, "loss": 0.545, "step": 520 }, { "epoch": 2.0384615384615383, "grad_norm": 2.7710626125335693, "learning_rate": 1.5870000000000002e-05, "loss": 0.5413, "step": 530 }, { "epoch": 2.076923076923077, "grad_norm": 3.6336352825164795, "learning_rate": 1.6170000000000003e-05, "loss": 0.5231, "step": 540 }, { "epoch": 2.1153846153846154, "grad_norm": 3.880500555038452, "learning_rate": 1.6470000000000003e-05, "loss": 0.5153, "step": 550 }, { "epoch": 2.1538461538461537, "grad_norm": 3.505535840988159, "learning_rate": 1.677e-05, "loss": 0.5065, "step": 560 }, { "epoch": 2.1923076923076925, "grad_norm": 3.137913465499878, "learning_rate": 1.7069999999999998e-05, "loss": 0.4962, "step": 570 }, { "epoch": 2.230769230769231, "grad_norm": 3.880664110183716, "learning_rate": 1.7369999999999998e-05, "loss": 0.4966, "step": 580 }, { "epoch": 2.269230769230769, "grad_norm": 3.5915846824645996, "learning_rate": 1.767e-05, "loss": 0.498, "step": 590 }, { "epoch": 2.3076923076923075, "grad_norm": 3.1810109615325928, "learning_rate": 1.797e-05, "loss": 0.4988, "step": 600 }, { "epoch": 2.3461538461538463, "grad_norm": 3.6019811630249023, "learning_rate": 1.827e-05, "loss": 0.4974, "step": 610 }, { "epoch": 2.3846153846153846, "grad_norm": 3.5532901287078857, "learning_rate": 1.857e-05, "loss": 0.4819, "step": 620 }, { "epoch": 2.423076923076923, "grad_norm": 3.921673536300659, "learning_rate": 1.887e-05, "loss": 0.4798, "step": 630 }, { "epoch": 2.4615384615384617, "grad_norm": 3.2884128093719482, "learning_rate": 1.917e-05, "loss": 0.4826, "step": 640 }, { "epoch": 2.5, "grad_norm": 3.231353759765625, "learning_rate": 1.947e-05, "loss": 0.4825, "step": 650 }, { "epoch": 2.5384615384615383, "grad_norm": 4.012421131134033, "learning_rate": 1.9770000000000002e-05, "loss": 0.4782, "step": 660 }, { "epoch": 2.5769230769230766, "grad_norm": 3.7715535163879395, "learning_rate": 2.0070000000000003e-05, "loss": 0.4854, "step": 670 }, { "epoch": 2.6153846153846154, "grad_norm": 3.2601428031921387, "learning_rate": 2.0370000000000003e-05, "loss": 0.4762, "step": 680 }, { "epoch": 2.6538461538461537, "grad_norm": 3.3098268508911133, "learning_rate": 2.067e-05, "loss": 0.4699, "step": 690 }, { "epoch": 2.6923076923076925, "grad_norm": 3.7813925743103027, "learning_rate": 2.097e-05, "loss": 0.4739, "step": 700 }, { "epoch": 2.730769230769231, "grad_norm": 3.242644786834717, "learning_rate": 2.1269999999999998e-05, "loss": 0.476, "step": 710 }, { "epoch": 2.769230769230769, "grad_norm": 3.092524766921997, "learning_rate": 2.157e-05, "loss": 0.4687, "step": 720 }, { "epoch": 2.8076923076923075, "grad_norm": 4.034473896026611, "learning_rate": 2.187e-05, "loss": 0.4655, "step": 730 }, { "epoch": 2.8461538461538463, "grad_norm": 3.64107084274292, "learning_rate": 2.217e-05, "loss": 0.4637, "step": 740 }, { "epoch": 2.8846153846153846, "grad_norm": 4.293366432189941, "learning_rate": 2.247e-05, "loss": 0.4662, "step": 750 }, { "epoch": 2.9230769230769234, "grad_norm": 3.5577127933502197, "learning_rate": 2.277e-05, "loss": 0.4571, "step": 760 }, { "epoch": 2.9615384615384617, "grad_norm": 3.0843167304992676, "learning_rate": 2.307e-05, "loss": 0.4625, "step": 770 }, { "epoch": 3.0, "grad_norm": 3.5912671089172363, "learning_rate": 2.337e-05, "loss": 0.457, "step": 780 }, { "epoch": 3.0384615384615383, "grad_norm": 4.415438175201416, "learning_rate": 2.3670000000000002e-05, "loss": 0.4594, "step": 790 }, { "epoch": 3.076923076923077, "grad_norm": 3.8929693698883057, "learning_rate": 2.3970000000000003e-05, "loss": 0.4613, "step": 800 }, { "epoch": 3.1153846153846154, "grad_norm": 4.137643814086914, "learning_rate": 2.4270000000000003e-05, "loss": 0.4628, "step": 810 }, { "epoch": 3.1538461538461537, "grad_norm": 3.0388357639312744, "learning_rate": 2.457e-05, "loss": 0.457, "step": 820 }, { "epoch": 3.1923076923076925, "grad_norm": 3.6971359252929688, "learning_rate": 2.487e-05, "loss": 0.4546, "step": 830 }, { "epoch": 3.230769230769231, "grad_norm": 3.4712138175964355, "learning_rate": 2.517e-05, "loss": 0.4511, "step": 840 }, { "epoch": 3.269230769230769, "grad_norm": 3.3456554412841797, "learning_rate": 2.547e-05, "loss": 0.4505, "step": 850 }, { "epoch": 3.3076923076923075, "grad_norm": 3.5414817333221436, "learning_rate": 2.577e-05, "loss": 0.4545, "step": 860 }, { "epoch": 3.3461538461538463, "grad_norm": 3.727144241333008, "learning_rate": 2.607e-05, "loss": 0.4537, "step": 870 }, { "epoch": 3.3846153846153846, "grad_norm": 3.6448793411254883, "learning_rate": 2.637e-05, "loss": 0.4434, "step": 880 }, { "epoch": 3.423076923076923, "grad_norm": 3.6482105255126953, "learning_rate": 2.667e-05, "loss": 0.4438, "step": 890 }, { "epoch": 3.4615384615384617, "grad_norm": 3.7402617931365967, "learning_rate": 2.697e-05, "loss": 0.4499, "step": 900 }, { "epoch": 3.5, "grad_norm": 3.1410586833953857, "learning_rate": 2.727e-05, "loss": 0.4423, "step": 910 }, { "epoch": 3.5384615384615383, "grad_norm": 4.063307762145996, "learning_rate": 2.7570000000000002e-05, "loss": 0.4476, "step": 920 }, { "epoch": 3.5769230769230766, "grad_norm": 3.781724214553833, "learning_rate": 2.7870000000000003e-05, "loss": 0.4445, "step": 930 }, { "epoch": 3.6153846153846154, "grad_norm": 4.0924787521362305, "learning_rate": 2.817e-05, "loss": 0.4449, "step": 940 }, { "epoch": 3.6538461538461537, "grad_norm": 3.8442327976226807, "learning_rate": 2.847e-05, "loss": 0.452, "step": 950 }, { "epoch": 3.6923076923076925, "grad_norm": 3.5104334354400635, "learning_rate": 2.877e-05, "loss": 0.4363, "step": 960 }, { "epoch": 3.730769230769231, "grad_norm": 4.416885852813721, "learning_rate": 2.907e-05, "loss": 0.4423, "step": 970 }, { "epoch": 3.769230769230769, "grad_norm": 3.5241246223449707, "learning_rate": 2.9370000000000002e-05, "loss": 0.4426, "step": 980 }, { "epoch": 3.8076923076923075, "grad_norm": 3.1697614192962646, "learning_rate": 2.967e-05, "loss": 0.4437, "step": 990 }, { "epoch": 3.8461538461538463, "grad_norm": 3.7319610118865967, "learning_rate": 2.997e-05, "loss": 0.4419, "step": 1000 }, { "epoch": 3.8846153846153846, "grad_norm": 4.392916679382324, "learning_rate": 2.9999983391181253e-05, "loss": 0.4371, "step": 1010 }, { "epoch": 3.9230769230769234, "grad_norm": 3.606473207473755, "learning_rate": 2.9999925978027876e-05, "loss": 0.4403, "step": 1020 }, { "epoch": 3.9615384615384617, "grad_norm": 3.2601609230041504, "learning_rate": 2.9999827555649637e-05, "loss": 0.4357, "step": 1030 }, { "epoch": 4.0, "grad_norm": 3.616961717605591, "learning_rate": 2.999968812431563e-05, "loss": 0.4394, "step": 1040 }, { "epoch": 4.038461538461538, "grad_norm": 4.0653204917907715, "learning_rate": 2.999950768440706e-05, "loss": 0.435, "step": 1050 }, { "epoch": 4.076923076923077, "grad_norm": 4.050017833709717, "learning_rate": 2.999928623641723e-05, "loss": 0.4448, "step": 1060 }, { "epoch": 4.115384615384615, "grad_norm": 3.190156936645508, "learning_rate": 2.9999023780951575e-05, "loss": 0.4356, "step": 1070 }, { "epoch": 4.153846153846154, "grad_norm": 3.323148250579834, "learning_rate": 2.999872031872764e-05, "loss": 0.4384, "step": 1080 }, { "epoch": 4.1923076923076925, "grad_norm": 3.506814479827881, "learning_rate": 2.999837585057508e-05, "loss": 0.4287, "step": 1090 }, { "epoch": 4.230769230769231, "grad_norm": 4.119077205657959, "learning_rate": 2.999799037743565e-05, "loss": 0.427, "step": 1100 }, { "epoch": 4.269230769230769, "grad_norm": 3.2304561138153076, "learning_rate": 2.999756390036323e-05, "loss": 0.4326, "step": 1110 }, { "epoch": 4.3076923076923075, "grad_norm": 3.1747395992279053, "learning_rate": 2.9997096420523788e-05, "loss": 0.4325, "step": 1120 }, { "epoch": 4.346153846153846, "grad_norm": 3.7520697116851807, "learning_rate": 2.9996587939195395e-05, "loss": 0.4372, "step": 1130 }, { "epoch": 4.384615384615385, "grad_norm": 3.395843744277954, "learning_rate": 2.999603845776822e-05, "loss": 0.4344, "step": 1140 }, { "epoch": 4.423076923076923, "grad_norm": 4.427037715911865, "learning_rate": 2.999544797774452e-05, "loss": 0.4342, "step": 1150 }, { "epoch": 4.461538461538462, "grad_norm": 3.310183525085449, "learning_rate": 2.9994816500738648e-05, "loss": 0.433, "step": 1160 }, { "epoch": 4.5, "grad_norm": 3.460261583328247, "learning_rate": 2.999414402847704e-05, "loss": 0.4279, "step": 1170 }, { "epoch": 4.538461538461538, "grad_norm": 3.523247718811035, "learning_rate": 2.999343056279821e-05, "loss": 0.4281, "step": 1180 }, { "epoch": 4.576923076923077, "grad_norm": 3.0562286376953125, "learning_rate": 2.9992676105652746e-05, "loss": 0.4281, "step": 1190 }, { "epoch": 4.615384615384615, "grad_norm": 3.179258108139038, "learning_rate": 2.9991880659103298e-05, "loss": 0.4321, "step": 1200 }, { "epoch": 4.653846153846154, "grad_norm": 3.0872421264648438, "learning_rate": 2.9991044225324593e-05, "loss": 0.4255, "step": 1210 }, { "epoch": 4.6923076923076925, "grad_norm": 3.896038293838501, "learning_rate": 2.9990166806603407e-05, "loss": 0.4221, "step": 1220 }, { "epoch": 4.730769230769231, "grad_norm": 3.345339775085449, "learning_rate": 2.9989248405338573e-05, "loss": 0.4236, "step": 1230 }, { "epoch": 4.769230769230769, "grad_norm": 3.4517908096313477, "learning_rate": 2.9988289024040962e-05, "loss": 0.419, "step": 1240 }, { "epoch": 4.8076923076923075, "grad_norm": 3.4892146587371826, "learning_rate": 2.998728866533348e-05, "loss": 0.4235, "step": 1250 }, { "epoch": 4.846153846153846, "grad_norm": 3.3247246742248535, "learning_rate": 2.9986247331951083e-05, "loss": 0.4143, "step": 1260 }, { "epoch": 4.884615384615385, "grad_norm": 3.2198266983032227, "learning_rate": 2.998516502674072e-05, "loss": 0.4167, "step": 1270 }, { "epoch": 4.923076923076923, "grad_norm": 3.2690086364746094, "learning_rate": 2.9984041752661386e-05, "loss": 0.4142, "step": 1280 }, { "epoch": 4.961538461538462, "grad_norm": 3.7173590660095215, "learning_rate": 2.9982877512784067e-05, "loss": 0.4141, "step": 1290 }, { "epoch": 5.0, "grad_norm": 4.059051990509033, "learning_rate": 2.998167231029174e-05, "loss": 0.4073, "step": 1300 }, { "epoch": 5.038461538461538, "grad_norm": 3.8667192459106445, "learning_rate": 2.99804261484794e-05, "loss": 0.4082, "step": 1310 }, { "epoch": 5.076923076923077, "grad_norm": 3.3809549808502197, "learning_rate": 2.997913903075399e-05, "loss": 0.408, "step": 1320 }, { "epoch": 5.115384615384615, "grad_norm": 3.0045559406280518, "learning_rate": 2.997781096063445e-05, "loss": 0.4067, "step": 1330 }, { "epoch": 5.153846153846154, "grad_norm": 3.27519154548645, "learning_rate": 2.9976441941751663e-05, "loss": 0.4023, "step": 1340 }, { "epoch": 5.1923076923076925, "grad_norm": 3.353156089782715, "learning_rate": 2.997503197784849e-05, "loss": 0.4034, "step": 1350 }, { "epoch": 5.230769230769231, "grad_norm": 3.568828821182251, "learning_rate": 2.9973581072779702e-05, "loss": 0.4025, "step": 1360 }, { "epoch": 5.269230769230769, "grad_norm": 3.06388258934021, "learning_rate": 2.9972089230512035e-05, "loss": 0.3948, "step": 1370 }, { "epoch": 5.3076923076923075, "grad_norm": 3.4603991508483887, "learning_rate": 2.997055645512411e-05, "loss": 0.4045, "step": 1380 }, { "epoch": 5.346153846153846, "grad_norm": 2.7319891452789307, "learning_rate": 2.9968982750806492e-05, "loss": 0.3956, "step": 1390 }, { "epoch": 5.384615384615385, "grad_norm": 3.0891315937042236, "learning_rate": 2.9967368121861623e-05, "loss": 0.4019, "step": 1400 }, { "epoch": 5.423076923076923, "grad_norm": 3.29032039642334, "learning_rate": 2.9965712572703834e-05, "loss": 0.3962, "step": 1410 }, { "epoch": 5.461538461538462, "grad_norm": 3.2606821060180664, "learning_rate": 2.996401610785934e-05, "loss": 0.4016, "step": 1420 }, { "epoch": 5.5, "grad_norm": 3.013518810272217, "learning_rate": 2.99622787319662e-05, "loss": 0.398, "step": 1430 }, { "epoch": 5.538461538461538, "grad_norm": 2.9983019828796387, "learning_rate": 2.9960500449774338e-05, "loss": 0.4002, "step": 1440 }, { "epoch": 5.576923076923077, "grad_norm": 3.4363009929656982, "learning_rate": 2.9958681266145517e-05, "loss": 0.4017, "step": 1450 }, { "epoch": 5.615384615384615, "grad_norm": 3.917628049850464, "learning_rate": 2.995682118605331e-05, "loss": 0.3938, "step": 1460 }, { "epoch": 5.653846153846154, "grad_norm": 2.4763236045837402, "learning_rate": 2.9954920214583107e-05, "loss": 0.3976, "step": 1470 }, { "epoch": 5.6923076923076925, "grad_norm": 3.2113523483276367, "learning_rate": 2.9952978356932084e-05, "loss": 0.3984, "step": 1480 }, { "epoch": 5.730769230769231, "grad_norm": 3.3912851810455322, "learning_rate": 2.9950995618409215e-05, "loss": 0.3925, "step": 1490 }, { "epoch": 5.769230769230769, "grad_norm": 2.927689790725708, "learning_rate": 2.9948972004435228e-05, "loss": 0.4036, "step": 1500 }, { "epoch": 5.8076923076923075, "grad_norm": 3.0390853881835938, "learning_rate": 2.9946907520542602e-05, "loss": 0.3954, "step": 1510 }, { "epoch": 5.846153846153846, "grad_norm": 3.3175177574157715, "learning_rate": 2.9944802172375566e-05, "loss": 0.3984, "step": 1520 }, { "epoch": 5.884615384615385, "grad_norm": 3.478698968887329, "learning_rate": 2.9942655965690053e-05, "loss": 0.3971, "step": 1530 }, { "epoch": 5.923076923076923, "grad_norm": 2.9315621852874756, "learning_rate": 2.9940468906353712e-05, "loss": 0.3989, "step": 1540 }, { "epoch": 5.961538461538462, "grad_norm": 3.185990810394287, "learning_rate": 2.9938241000345887e-05, "loss": 0.3948, "step": 1550 }, { "epoch": 6.0, "grad_norm": 4.063650608062744, "learning_rate": 2.993597225375758e-05, "loss": 0.3937, "step": 1560 }, { "epoch": 6.038461538461538, "grad_norm": 4.225824356079102, "learning_rate": 2.993366267279146e-05, "loss": 0.3931, "step": 1570 }, { "epoch": 6.076923076923077, "grad_norm": 3.3525872230529785, "learning_rate": 2.993131226376183e-05, "loss": 0.3918, "step": 1580 }, { "epoch": 6.115384615384615, "grad_norm": 3.352668523788452, "learning_rate": 2.9928921033094626e-05, "loss": 0.3901, "step": 1590 }, { "epoch": 6.153846153846154, "grad_norm": 3.500871181488037, "learning_rate": 2.9926488987327376e-05, "loss": 0.3905, "step": 1600 }, { "epoch": 6.1923076923076925, "grad_norm": 3.547381639480591, "learning_rate": 2.99240161331092e-05, "loss": 0.3847, "step": 1610 }, { "epoch": 6.230769230769231, "grad_norm": 3.7568917274475098, "learning_rate": 2.992150247720079e-05, "loss": 0.383, "step": 1620 }, { "epoch": 6.269230769230769, "grad_norm": 2.754397392272949, "learning_rate": 2.991894802647438e-05, "loss": 0.3833, "step": 1630 }, { "epoch": 6.3076923076923075, "grad_norm": 2.8265631198883057, "learning_rate": 2.9916352787913746e-05, "loss": 0.3875, "step": 1640 }, { "epoch": 6.346153846153846, "grad_norm": 3.522829532623291, "learning_rate": 2.991371676861417e-05, "loss": 0.3867, "step": 1650 }, { "epoch": 6.384615384615385, "grad_norm": 4.345142841339111, "learning_rate": 2.991103997578243e-05, "loss": 0.3754, "step": 1660 }, { "epoch": 6.423076923076923, "grad_norm": 3.212238073348999, "learning_rate": 2.9908322416736767e-05, "loss": 0.382, "step": 1670 }, { "epoch": 6.461538461538462, "grad_norm": 3.8595285415649414, "learning_rate": 2.990556409890689e-05, "loss": 0.3831, "step": 1680 }, { "epoch": 6.5, "grad_norm": 3.5283541679382324, "learning_rate": 2.990276502983394e-05, "loss": 0.3785, "step": 1690 }, { "epoch": 6.538461538461538, "grad_norm": 2.779311180114746, "learning_rate": 2.9899925217170455e-05, "loss": 0.3722, "step": 1700 }, { "epoch": 6.576923076923077, "grad_norm": 3.0596842765808105, "learning_rate": 2.989704466868038e-05, "loss": 0.373, "step": 1710 }, { "epoch": 6.615384615384615, "grad_norm": 3.0654571056365967, "learning_rate": 2.9894123392239018e-05, "loss": 0.3717, "step": 1720 }, { "epoch": 6.653846153846154, "grad_norm": 3.738708734512329, "learning_rate": 2.9891161395833037e-05, "loss": 0.3652, "step": 1730 }, { "epoch": 6.6923076923076925, "grad_norm": 3.2960824966430664, "learning_rate": 2.988815868756042e-05, "loss": 0.3606, "step": 1740 }, { "epoch": 6.730769230769231, "grad_norm": 3.8682546615600586, "learning_rate": 2.9885115275630447e-05, "loss": 0.3616, "step": 1750 }, { "epoch": 6.769230769230769, "grad_norm": 5.847049236297607, "learning_rate": 2.9882031168363703e-05, "loss": 0.3594, "step": 1760 }, { "epoch": 6.8076923076923075, "grad_norm": 4.252906322479248, "learning_rate": 2.9878906374192013e-05, "loss": 0.3599, "step": 1770 }, { "epoch": 6.846153846153846, "grad_norm": 3.371454954147339, "learning_rate": 2.9875740901658446e-05, "loss": 0.3522, "step": 1780 }, { "epoch": 6.884615384615385, "grad_norm": 3.9505934715270996, "learning_rate": 2.987253475941728e-05, "loss": 0.3518, "step": 1790 }, { "epoch": 6.923076923076923, "grad_norm": 4.2967047691345215, "learning_rate": 2.9869287956233986e-05, "loss": 0.3453, "step": 1800 }, { "epoch": 6.961538461538462, "grad_norm": 3.76397705078125, "learning_rate": 2.9866000500985207e-05, "loss": 0.3433, "step": 1810 }, { "epoch": 7.0, "grad_norm": 3.6324551105499268, "learning_rate": 2.9862672402658712e-05, "loss": 0.3322, "step": 1820 }, { "epoch": 7.038461538461538, "grad_norm": 4.136635780334473, "learning_rate": 2.98593036703534e-05, "loss": 0.3367, "step": 1830 }, { "epoch": 7.076923076923077, "grad_norm": 4.801270008087158, "learning_rate": 2.9855894313279256e-05, "loss": 0.3327, "step": 1840 }, { "epoch": 7.115384615384615, "grad_norm": 3.6734988689422607, "learning_rate": 2.9852444340757326e-05, "loss": 0.33, "step": 1850 }, { "epoch": 7.153846153846154, "grad_norm": 3.806365966796875, "learning_rate": 2.9848953762219707e-05, "loss": 0.3315, "step": 1860 }, { "epoch": 7.1923076923076925, "grad_norm": 4.397286891937256, "learning_rate": 2.984542258720951e-05, "loss": 0.3221, "step": 1870 }, { "epoch": 7.230769230769231, "grad_norm": 4.844172954559326, "learning_rate": 2.984185082538083e-05, "loss": 0.3104, "step": 1880 }, { "epoch": 7.269230769230769, "grad_norm": 3.5059876441955566, "learning_rate": 2.983823848649873e-05, "loss": 0.3044, "step": 1890 }, { "epoch": 7.3076923076923075, "grad_norm": 3.5109610557556152, "learning_rate": 2.9834585580439203e-05, "loss": 0.3145, "step": 1900 }, { "epoch": 7.346153846153846, "grad_norm": 3.935485601425171, "learning_rate": 2.9830892117189157e-05, "loss": 0.3047, "step": 1910 }, { "epoch": 7.384615384615385, "grad_norm": 4.239902019500732, "learning_rate": 2.982715810684638e-05, "loss": 0.3004, "step": 1920 }, { "epoch": 7.423076923076923, "grad_norm": 3.725754737854004, "learning_rate": 2.982338355961951e-05, "loss": 0.2974, "step": 1930 }, { "epoch": 7.461538461538462, "grad_norm": 4.275974750518799, "learning_rate": 2.981956848582802e-05, "loss": 0.2943, "step": 1940 }, { "epoch": 7.5, "grad_norm": 4.536103248596191, "learning_rate": 2.981571289590217e-05, "loss": 0.2879, "step": 1950 }, { "epoch": 7.538461538461538, "grad_norm": 5.503205299377441, "learning_rate": 2.9811816800383003e-05, "loss": 0.2769, "step": 1960 }, { "epoch": 7.576923076923077, "grad_norm": 4.398034572601318, "learning_rate": 2.9807880209922288e-05, "loss": 0.2841, "step": 1970 }, { "epoch": 7.615384615384615, "grad_norm": 4.218934059143066, "learning_rate": 2.9803903135282518e-05, "loss": 0.2787, "step": 1980 }, { "epoch": 7.653846153846154, "grad_norm": 3.5821478366851807, "learning_rate": 2.9799885587336862e-05, "loss": 0.2669, "step": 1990 }, { "epoch": 7.6923076923076925, "grad_norm": 4.140866279602051, "learning_rate": 2.9795827577069145e-05, "loss": 0.2558, "step": 2000 }, { "epoch": 7.730769230769231, "grad_norm": 3.9685962200164795, "learning_rate": 2.9791729115573808e-05, "loss": 0.2493, "step": 2010 }, { "epoch": 7.769230769230769, "grad_norm": 6.242632865905762, "learning_rate": 2.9787590214055887e-05, "loss": 0.2538, "step": 2020 }, { "epoch": 7.8076923076923075, "grad_norm": 5.438572406768799, "learning_rate": 2.9783410883830983e-05, "loss": 0.2555, "step": 2030 }, { "epoch": 7.846153846153846, "grad_norm": 4.109585762023926, "learning_rate": 2.9779191136325233e-05, "loss": 0.2533, "step": 2040 }, { "epoch": 7.884615384615385, "grad_norm": 4.627237319946289, "learning_rate": 2.977493098307525e-05, "loss": 0.2371, "step": 2050 }, { "epoch": 7.923076923076923, "grad_norm": 4.566162109375, "learning_rate": 2.9770630435728142e-05, "loss": 0.2214, "step": 2060 }, { "epoch": 7.961538461538462, "grad_norm": 5.047729969024658, "learning_rate": 2.976628950604144e-05, "loss": 0.222, "step": 2070 }, { "epoch": 8.0, "grad_norm": 5.952073574066162, "learning_rate": 2.9761908205883073e-05, "loss": 0.2164, "step": 2080 }, { "epoch": 8.038461538461538, "grad_norm": 6.023576736450195, "learning_rate": 2.9757486547231357e-05, "loss": 0.2171, "step": 2090 }, { "epoch": 8.076923076923077, "grad_norm": 5.403975486755371, "learning_rate": 2.9753024542174934e-05, "loss": 0.2039, "step": 2100 }, { "epoch": 8.115384615384615, "grad_norm": 6.422732830047607, "learning_rate": 2.9748522202912755e-05, "loss": 0.2025, "step": 2110 }, { "epoch": 8.153846153846153, "grad_norm": 4.933073043823242, "learning_rate": 2.974397954175404e-05, "loss": 0.1943, "step": 2120 }, { "epoch": 8.192307692307692, "grad_norm": 3.9752209186553955, "learning_rate": 2.973939657111826e-05, "loss": 0.1947, "step": 2130 }, { "epoch": 8.23076923076923, "grad_norm": 5.760082721710205, "learning_rate": 2.9734773303535078e-05, "loss": 0.1982, "step": 2140 }, { "epoch": 8.26923076923077, "grad_norm": 4.883352279663086, "learning_rate": 2.9730109751644325e-05, "loss": 0.1932, "step": 2150 }, { "epoch": 8.307692307692308, "grad_norm": 5.062183856964111, "learning_rate": 2.9725405928195985e-05, "loss": 0.1808, "step": 2160 }, { "epoch": 8.346153846153847, "grad_norm": 5.349424362182617, "learning_rate": 2.9720661846050123e-05, "loss": 0.1648, "step": 2170 }, { "epoch": 8.384615384615385, "grad_norm": 4.347816467285156, "learning_rate": 2.971587751817688e-05, "loss": 0.1733, "step": 2180 }, { "epoch": 8.423076923076923, "grad_norm": 6.218843460083008, "learning_rate": 2.9711052957656425e-05, "loss": 0.1568, "step": 2190 }, { "epoch": 8.461538461538462, "grad_norm": 4.51317024230957, "learning_rate": 2.9706188177678924e-05, "loss": 0.1544, "step": 2200 }, { "epoch": 8.5, "grad_norm": 6.40587854385376, "learning_rate": 2.97012831915445e-05, "loss": 0.1501, "step": 2210 }, { "epoch": 8.538461538461538, "grad_norm": 3.9102859497070312, "learning_rate": 2.96963380126632e-05, "loss": 0.1443, "step": 2220 }, { "epoch": 8.576923076923077, "grad_norm": 5.443079948425293, "learning_rate": 2.9691352654554953e-05, "loss": 0.1472, "step": 2230 }, { "epoch": 8.615384615384615, "grad_norm": 6.355679988861084, "learning_rate": 2.9686327130849536e-05, "loss": 0.1383, "step": 2240 }, { "epoch": 8.653846153846153, "grad_norm": 6.52142333984375, "learning_rate": 2.9681261455286538e-05, "loss": 0.1482, "step": 2250 }, { "epoch": 8.692307692307692, "grad_norm": 5.2175703048706055, "learning_rate": 2.9676155641715318e-05, "loss": 0.1405, "step": 2260 }, { "epoch": 8.73076923076923, "grad_norm": 5.178348541259766, "learning_rate": 2.9671009704094988e-05, "loss": 0.1307, "step": 2270 }, { "epoch": 8.76923076923077, "grad_norm": 5.276549816131592, "learning_rate": 2.9665823656494335e-05, "loss": 0.1402, "step": 2280 }, { "epoch": 8.807692307692308, "grad_norm": 4.72429084777832, "learning_rate": 2.9660597513091824e-05, "loss": 0.129, "step": 2290 }, { "epoch": 8.846153846153847, "grad_norm": 4.0139312744140625, "learning_rate": 2.965533128817552e-05, "loss": 0.1181, "step": 2300 }, { "epoch": 8.884615384615385, "grad_norm": 4.549901485443115, "learning_rate": 2.9650024996143084e-05, "loss": 0.1293, "step": 2310 }, { "epoch": 8.923076923076923, "grad_norm": 4.0546183586120605, "learning_rate": 2.964467865150172e-05, "loss": 0.1098, "step": 2320 }, { "epoch": 8.961538461538462, "grad_norm": 5.741022109985352, "learning_rate": 2.9639292268868133e-05, "loss": 0.1255, "step": 2330 }, { "epoch": 9.0, "grad_norm": 5.321122169494629, "learning_rate": 2.9633865862968478e-05, "loss": 0.112, "step": 2340 }, { "epoch": 9.038461538461538, "grad_norm": 4.528491020202637, "learning_rate": 2.9628399448638352e-05, "loss": 0.1198, "step": 2350 }, { "epoch": 9.076923076923077, "grad_norm": 3.827465295791626, "learning_rate": 2.9622893040822714e-05, "loss": 0.1039, "step": 2360 }, { "epoch": 9.115384615384615, "grad_norm": 4.378634452819824, "learning_rate": 2.9617346654575875e-05, "loss": 0.1049, "step": 2370 }, { "epoch": 9.153846153846153, "grad_norm": 4.433044910430908, "learning_rate": 2.9611760305061447e-05, "loss": 0.099, "step": 2380 }, { "epoch": 9.192307692307692, "grad_norm": 4.418703556060791, "learning_rate": 2.9606134007552292e-05, "loss": 0.1097, "step": 2390 }, { "epoch": 9.23076923076923, "grad_norm": 4.773401737213135, "learning_rate": 2.9600467777430497e-05, "loss": 0.1039, "step": 2400 }, { "epoch": 9.26923076923077, "grad_norm": 5.771136283874512, "learning_rate": 2.9594761630187312e-05, "loss": 0.0954, "step": 2410 }, { "epoch": 9.307692307692308, "grad_norm": 4.874659061431885, "learning_rate": 2.9589015581423132e-05, "loss": 0.0973, "step": 2420 }, { "epoch": 9.346153846153847, "grad_norm": 5.661646366119385, "learning_rate": 2.958322964684743e-05, "loss": 0.0958, "step": 2430 }, { "epoch": 9.384615384615385, "grad_norm": 5.317569732666016, "learning_rate": 2.9577403842278735e-05, "loss": 0.1032, "step": 2440 }, { "epoch": 9.423076923076923, "grad_norm": 4.762622356414795, "learning_rate": 2.957153818364457e-05, "loss": 0.0972, "step": 2450 }, { "epoch": 9.461538461538462, "grad_norm": 5.307681083679199, "learning_rate": 2.9565632686981428e-05, "loss": 0.1038, "step": 2460 }, { "epoch": 9.5, "grad_norm": 3.2064321041107178, "learning_rate": 2.9559687368434702e-05, "loss": 0.0902, "step": 2470 }, { "epoch": 9.538461538461538, "grad_norm": 4.020598888397217, "learning_rate": 2.9553702244258674e-05, "loss": 0.0886, "step": 2480 }, { "epoch": 9.576923076923077, "grad_norm": 5.292928218841553, "learning_rate": 2.954767733081644e-05, "loss": 0.0839, "step": 2490 }, { "epoch": 9.615384615384615, "grad_norm": 4.200097560882568, "learning_rate": 2.9541612644579887e-05, "loss": 0.0943, "step": 2500 }, { "epoch": 9.653846153846153, "grad_norm": 4.739330291748047, "learning_rate": 2.9535508202129634e-05, "loss": 0.0903, "step": 2510 }, { "epoch": 9.692307692307692, "grad_norm": 4.430630683898926, "learning_rate": 2.9529364020154994e-05, "loss": 0.0819, "step": 2520 }, { "epoch": 9.73076923076923, "grad_norm": 4.4350481033325195, "learning_rate": 2.9523180115453922e-05, "loss": 0.0798, "step": 2530 }, { "epoch": 9.76923076923077, "grad_norm": 4.632157325744629, "learning_rate": 2.9516956504932984e-05, "loss": 0.0917, "step": 2540 }, { "epoch": 9.807692307692308, "grad_norm": 3.786306619644165, "learning_rate": 2.9510693205607286e-05, "loss": 0.0978, "step": 2550 }, { "epoch": 9.846153846153847, "grad_norm": 4.0172929763793945, "learning_rate": 2.9504390234600456e-05, "loss": 0.0884, "step": 2560 }, { "epoch": 9.884615384615385, "grad_norm": 4.171654224395752, "learning_rate": 2.9498047609144577e-05, "loss": 0.0821, "step": 2570 }, { "epoch": 9.923076923076923, "grad_norm": 5.381162643432617, "learning_rate": 2.9491665346580134e-05, "loss": 0.0872, "step": 2580 }, { "epoch": 9.961538461538462, "grad_norm": 3.150331497192383, "learning_rate": 2.9485243464356e-05, "loss": 0.0859, "step": 2590 }, { "epoch": 10.0, "grad_norm": 4.065011978149414, "learning_rate": 2.9478781980029352e-05, "loss": 0.0823, "step": 2600 }, { "epoch": 10.038461538461538, "grad_norm": 3.450819492340088, "learning_rate": 2.9472280911265642e-05, "loss": 0.079, "step": 2610 }, { "epoch": 10.076923076923077, "grad_norm": 3.7067441940307617, "learning_rate": 2.9465740275838543e-05, "loss": 0.0802, "step": 2620 }, { "epoch": 10.115384615384615, "grad_norm": 3.485081434249878, "learning_rate": 2.94591600916299e-05, "loss": 0.0772, "step": 2630 }, { "epoch": 10.153846153846153, "grad_norm": 2.740015983581543, "learning_rate": 2.9452540376629692e-05, "loss": 0.0712, "step": 2640 }, { "epoch": 10.192307692307692, "grad_norm": 3.398298501968384, "learning_rate": 2.944588114893596e-05, "loss": 0.0826, "step": 2650 }, { "epoch": 10.23076923076923, "grad_norm": 3.7506258487701416, "learning_rate": 2.9439182426754784e-05, "loss": 0.0896, "step": 2660 }, { "epoch": 10.26923076923077, "grad_norm": 3.0388355255126953, "learning_rate": 2.9432444228400208e-05, "loss": 0.0794, "step": 2670 }, { "epoch": 10.307692307692308, "grad_norm": 2.9595770835876465, "learning_rate": 2.9425666572294218e-05, "loss": 0.0823, "step": 2680 }, { "epoch": 10.346153846153847, "grad_norm": 3.7701125144958496, "learning_rate": 2.941884947696666e-05, "loss": 0.0777, "step": 2690 }, { "epoch": 10.384615384615385, "grad_norm": 3.5903477668762207, "learning_rate": 2.9411992961055214e-05, "loss": 0.0864, "step": 2700 }, { "epoch": 10.423076923076923, "grad_norm": 3.5986275672912598, "learning_rate": 2.9405097043305334e-05, "loss": 0.0809, "step": 2710 }, { "epoch": 10.461538461538462, "grad_norm": 3.5043976306915283, "learning_rate": 2.9398161742570196e-05, "loss": 0.0787, "step": 2720 }, { "epoch": 10.5, "grad_norm": 3.9481887817382812, "learning_rate": 2.9391187077810644e-05, "loss": 0.0828, "step": 2730 }, { "epoch": 10.538461538461538, "grad_norm": 3.953305721282959, "learning_rate": 2.9384173068095145e-05, "loss": 0.079, "step": 2740 }, { "epoch": 10.576923076923077, "grad_norm": 4.111423492431641, "learning_rate": 2.937711973259974e-05, "loss": 0.0702, "step": 2750 }, { "epoch": 10.615384615384615, "grad_norm": 3.0147547721862793, "learning_rate": 2.9370027090607974e-05, "loss": 0.0767, "step": 2760 }, { "epoch": 10.653846153846153, "grad_norm": 3.451683759689331, "learning_rate": 2.936289516151086e-05, "loss": 0.0724, "step": 2770 }, { "epoch": 10.692307692307692, "grad_norm": 3.6806628704071045, "learning_rate": 2.935572396480682e-05, "loss": 0.0753, "step": 2780 }, { "epoch": 10.73076923076923, "grad_norm": 3.9452641010284424, "learning_rate": 2.9348513520101636e-05, "loss": 0.0692, "step": 2790 }, { "epoch": 10.76923076923077, "grad_norm": 3.2497544288635254, "learning_rate": 2.9341263847108383e-05, "loss": 0.0745, "step": 2800 }, { "epoch": 10.807692307692308, "grad_norm": 3.549978494644165, "learning_rate": 2.933397496564739e-05, "loss": 0.0749, "step": 2810 }, { "epoch": 10.846153846153847, "grad_norm": 3.3335323333740234, "learning_rate": 2.9326646895646178e-05, "loss": 0.0692, "step": 2820 }, { "epoch": 10.884615384615385, "grad_norm": 3.405339479446411, "learning_rate": 2.931927965713942e-05, "loss": 0.0746, "step": 2830 }, { "epoch": 10.923076923076923, "grad_norm": 3.82893443107605, "learning_rate": 2.931187327026886e-05, "loss": 0.0721, "step": 2840 }, { "epoch": 10.961538461538462, "grad_norm": 3.7267136573791504, "learning_rate": 2.9304427755283278e-05, "loss": 0.0734, "step": 2850 }, { "epoch": 11.0, "grad_norm": 3.892670154571533, "learning_rate": 2.9296943132538425e-05, "loss": 0.0695, "step": 2860 }, { "epoch": 11.038461538461538, "grad_norm": 3.6918320655822754, "learning_rate": 2.928941942249697e-05, "loss": 0.0715, "step": 2870 }, { "epoch": 11.076923076923077, "grad_norm": 3.9371860027313232, "learning_rate": 2.928185664572846e-05, "loss": 0.0698, "step": 2880 }, { "epoch": 11.115384615384615, "grad_norm": 3.618164539337158, "learning_rate": 2.927425482290923e-05, "loss": 0.0722, "step": 2890 }, { "epoch": 11.153846153846153, "grad_norm": 3.7436001300811768, "learning_rate": 2.926661397482238e-05, "loss": 0.0763, "step": 2900 }, { "epoch": 11.192307692307692, "grad_norm": 3.660889148712158, "learning_rate": 2.9258934122357685e-05, "loss": 0.0917, "step": 2910 }, { "epoch": 11.23076923076923, "grad_norm": 3.8613357543945312, "learning_rate": 2.9251215286511574e-05, "loss": 0.072, "step": 2920 }, { "epoch": 11.26923076923077, "grad_norm": 3.738790273666382, "learning_rate": 2.924345748838706e-05, "loss": 0.0709, "step": 2930 }, { "epoch": 11.307692307692308, "grad_norm": 3.2637104988098145, "learning_rate": 2.923566074919365e-05, "loss": 0.0707, "step": 2940 }, { "epoch": 11.346153846153847, "grad_norm": 2.518782615661621, "learning_rate": 2.9227825090247346e-05, "loss": 0.0679, "step": 2950 }, { "epoch": 11.384615384615385, "grad_norm": 3.1399145126342773, "learning_rate": 2.9219950532970526e-05, "loss": 0.0806, "step": 2960 }, { "epoch": 11.423076923076923, "grad_norm": 3.190922498703003, "learning_rate": 2.921203709889194e-05, "loss": 0.0766, "step": 2970 }, { "epoch": 11.461538461538462, "grad_norm": 3.0159084796905518, "learning_rate": 2.9204084809646607e-05, "loss": 0.0671, "step": 2980 }, { "epoch": 11.5, "grad_norm": 3.547037124633789, "learning_rate": 2.9196093686975793e-05, "loss": 0.0661, "step": 2990 }, { "epoch": 11.538461538461538, "grad_norm": 3.7780942916870117, "learning_rate": 2.918806375272691e-05, "loss": 0.0772, "step": 3000 }, { "epoch": 11.576923076923077, "grad_norm": 3.440263509750366, "learning_rate": 2.9179995028853498e-05, "loss": 0.0697, "step": 3010 }, { "epoch": 11.615384615384615, "grad_norm": 3.568605661392212, "learning_rate": 2.917188753741514e-05, "loss": 0.0779, "step": 3020 }, { "epoch": 11.653846153846153, "grad_norm": 3.969608783721924, "learning_rate": 2.916374130057741e-05, "loss": 0.0684, "step": 3030 }, { "epoch": 11.692307692307692, "grad_norm": 3.102283000946045, "learning_rate": 2.91555563406118e-05, "loss": 0.0624, "step": 3040 }, { "epoch": 11.73076923076923, "grad_norm": 2.713280200958252, "learning_rate": 2.9147332679895683e-05, "loss": 0.0723, "step": 3050 }, { "epoch": 11.76923076923077, "grad_norm": 3.4983530044555664, "learning_rate": 2.9139070340912236e-05, "loss": 0.0714, "step": 3060 }, { "epoch": 11.807692307692308, "grad_norm": 3.3425889015197754, "learning_rate": 2.9130769346250376e-05, "loss": 0.0776, "step": 3070 }, { "epoch": 11.846153846153847, "grad_norm": 3.1763546466827393, "learning_rate": 2.9122429718604704e-05, "loss": 0.0705, "step": 3080 }, { "epoch": 11.884615384615385, "grad_norm": 2.6937384605407715, "learning_rate": 2.911405148077545e-05, "loss": 0.0691, "step": 3090 }, { "epoch": 11.923076923076923, "grad_norm": 2.8746800422668457, "learning_rate": 2.9105634655668385e-05, "loss": 0.0715, "step": 3100 }, { "epoch": 11.961538461538462, "grad_norm": 3.535343647003174, "learning_rate": 2.9097179266294794e-05, "loss": 0.0754, "step": 3110 }, { "epoch": 12.0, "grad_norm": 3.433488607406616, "learning_rate": 2.9088685335771396e-05, "loss": 0.0619, "step": 3120 }, { "epoch": 12.038461538461538, "grad_norm": 3.0864665508270264, "learning_rate": 2.9080152887320255e-05, "loss": 0.0732, "step": 3130 }, { "epoch": 12.076923076923077, "grad_norm": 3.4718246459960938, "learning_rate": 2.9071581944268778e-05, "loss": 0.0702, "step": 3140 }, { "epoch": 12.115384615384615, "grad_norm": 3.164069175720215, "learning_rate": 2.906297253004958e-05, "loss": 0.0692, "step": 3150 }, { "epoch": 12.153846153846153, "grad_norm": 2.8084299564361572, "learning_rate": 2.9054324668200483e-05, "loss": 0.0681, "step": 3160 }, { "epoch": 12.192307692307692, "grad_norm": 2.749694347381592, "learning_rate": 2.9045638382364404e-05, "loss": 0.0712, "step": 3170 }, { "epoch": 12.23076923076923, "grad_norm": 2.85776948928833, "learning_rate": 2.9036913696289318e-05, "loss": 0.0653, "step": 3180 }, { "epoch": 12.26923076923077, "grad_norm": 2.447458505630493, "learning_rate": 2.9028150633828186e-05, "loss": 0.068, "step": 3190 }, { "epoch": 12.307692307692308, "grad_norm": 2.5854172706604004, "learning_rate": 2.9019349218938887e-05, "loss": 0.0759, "step": 3200 }, { "epoch": 12.346153846153847, "grad_norm": 2.299642324447632, "learning_rate": 2.9010509475684146e-05, "loss": 0.0663, "step": 3210 }, { "epoch": 12.384615384615385, "grad_norm": 2.9513206481933594, "learning_rate": 2.900163142823149e-05, "loss": 0.0608, "step": 3220 }, { "epoch": 12.423076923076923, "grad_norm": 3.0884900093078613, "learning_rate": 2.8992715100853166e-05, "loss": 0.0658, "step": 3230 }, { "epoch": 12.461538461538462, "grad_norm": 3.2073280811309814, "learning_rate": 2.898376051792606e-05, "loss": 0.0596, "step": 3240 }, { "epoch": 12.5, "grad_norm": 2.7891628742218018, "learning_rate": 2.897476770393167e-05, "loss": 0.0592, "step": 3250 }, { "epoch": 12.538461538461538, "grad_norm": 3.262420892715454, "learning_rate": 2.8965736683456e-05, "loss": 0.0762, "step": 3260 }, { "epoch": 12.576923076923077, "grad_norm": 3.5874862670898438, "learning_rate": 2.895666748118952e-05, "loss": 0.0682, "step": 3270 }, { "epoch": 12.615384615384615, "grad_norm": 3.6109778881073, "learning_rate": 2.8947560121927077e-05, "loss": 0.0615, "step": 3280 }, { "epoch": 12.653846153846153, "grad_norm": 3.210784435272217, "learning_rate": 2.8938414630567852e-05, "loss": 0.0658, "step": 3290 }, { "epoch": 12.692307692307692, "grad_norm": 3.263617992401123, "learning_rate": 2.892923103211526e-05, "loss": 0.0607, "step": 3300 }, { "epoch": 12.73076923076923, "grad_norm": 3.226895332336426, "learning_rate": 2.892000935167691e-05, "loss": 0.058, "step": 3310 }, { "epoch": 12.76923076923077, "grad_norm": 3.4264261722564697, "learning_rate": 2.8910749614464536e-05, "loss": 0.0587, "step": 3320 }, { "epoch": 12.807692307692308, "grad_norm": 3.2235159873962402, "learning_rate": 2.890145184579389e-05, "loss": 0.068, "step": 3330 }, { "epoch": 12.846153846153847, "grad_norm": 3.570420503616333, "learning_rate": 2.8892116071084727e-05, "loss": 0.0631, "step": 3340 }, { "epoch": 12.884615384615385, "grad_norm": 3.176832914352417, "learning_rate": 2.8882742315860692e-05, "loss": 0.0683, "step": 3350 }, { "epoch": 12.923076923076923, "grad_norm": 3.430499792098999, "learning_rate": 2.8873330605749275e-05, "loss": 0.0616, "step": 3360 }, { "epoch": 12.961538461538462, "grad_norm": 3.4460554122924805, "learning_rate": 2.886388096648174e-05, "loss": 0.0755, "step": 3370 }, { "epoch": 13.0, "grad_norm": 3.3051247596740723, "learning_rate": 2.8854393423893024e-05, "loss": 0.0663, "step": 3380 }, { "epoch": 13.038461538461538, "grad_norm": 3.3149800300598145, "learning_rate": 2.8844868003921723e-05, "loss": 0.0687, "step": 3390 }, { "epoch": 13.076923076923077, "grad_norm": 3.245965003967285, "learning_rate": 2.8835304732609962e-05, "loss": 0.0726, "step": 3400 }, { "epoch": 13.115384615384615, "grad_norm": 2.479734420776367, "learning_rate": 2.882570363610336e-05, "loss": 0.064, "step": 3410 }, { "epoch": 13.153846153846153, "grad_norm": 3.601377487182617, "learning_rate": 2.8816064740650954e-05, "loss": 0.0675, "step": 3420 }, { "epoch": 13.192307692307692, "grad_norm": 3.4560110569000244, "learning_rate": 2.880638807260511e-05, "loss": 0.0607, "step": 3430 }, { "epoch": 13.23076923076923, "grad_norm": 2.8063180446624756, "learning_rate": 2.8796673658421472e-05, "loss": 0.0647, "step": 3440 }, { "epoch": 13.26923076923077, "grad_norm": 2.8318352699279785, "learning_rate": 2.8786921524658877e-05, "loss": 0.057, "step": 3450 }, { "epoch": 13.307692307692308, "grad_norm": 2.682603597640991, "learning_rate": 2.8777131697979283e-05, "loss": 0.0701, "step": 3460 }, { "epoch": 13.346153846153847, "grad_norm": 2.213459014892578, "learning_rate": 2.876730420514771e-05, "loss": 0.0572, "step": 3470 }, { "epoch": 13.384615384615385, "grad_norm": 3.3933420181274414, "learning_rate": 2.8757439073032136e-05, "loss": 0.0645, "step": 3480 }, { "epoch": 13.423076923076923, "grad_norm": 2.8533976078033447, "learning_rate": 2.874753632860347e-05, "loss": 0.064, "step": 3490 }, { "epoch": 13.461538461538462, "grad_norm": 2.714489459991455, "learning_rate": 2.873759599893543e-05, "loss": 0.0729, "step": 3500 }, { "epoch": 13.5, "grad_norm": 2.61195969581604, "learning_rate": 2.8727618111204494e-05, "loss": 0.0665, "step": 3510 }, { "epoch": 13.538461538461538, "grad_norm": 2.846728801727295, "learning_rate": 2.871760269268983e-05, "loss": 0.0595, "step": 3520 }, { "epoch": 13.576923076923077, "grad_norm": 2.8665387630462646, "learning_rate": 2.870754977077321e-05, "loss": 0.0626, "step": 3530 }, { "epoch": 13.615384615384615, "grad_norm": 2.974710464477539, "learning_rate": 2.869745937293894e-05, "loss": 0.0685, "step": 3540 }, { "epoch": 13.653846153846153, "grad_norm": 2.7472543716430664, "learning_rate": 2.8687331526773775e-05, "loss": 0.0639, "step": 3550 }, { "epoch": 13.692307692307692, "grad_norm": 2.984053611755371, "learning_rate": 2.867716625996687e-05, "loss": 0.0642, "step": 3560 }, { "epoch": 13.73076923076923, "grad_norm": 2.6401329040527344, "learning_rate": 2.8666963600309672e-05, "loss": 0.0698, "step": 3570 }, { "epoch": 13.76923076923077, "grad_norm": 3.483025550842285, "learning_rate": 2.8656723575695862e-05, "loss": 0.0633, "step": 3580 }, { "epoch": 13.807692307692308, "grad_norm": 3.1316492557525635, "learning_rate": 2.8646446214121276e-05, "loss": 0.0599, "step": 3590 }, { "epoch": 13.846153846153847, "grad_norm": 2.7110755443573, "learning_rate": 2.8636131543683828e-05, "loss": 0.0657, "step": 3600 }, { "epoch": 13.884615384615385, "grad_norm": 2.8898301124572754, "learning_rate": 2.8625779592583436e-05, "loss": 0.0569, "step": 3610 }, { "epoch": 13.923076923076923, "grad_norm": 3.3383960723876953, "learning_rate": 2.861539038912193e-05, "loss": 0.0618, "step": 3620 }, { "epoch": 13.961538461538462, "grad_norm": 2.7277491092681885, "learning_rate": 2.860496396170301e-05, "loss": 0.0626, "step": 3630 }, { "epoch": 14.0, "grad_norm": 3.5116193294525146, "learning_rate": 2.859450033883212e-05, "loss": 0.0632, "step": 3640 }, { "epoch": 14.038461538461538, "grad_norm": 3.129054069519043, "learning_rate": 2.8583999549116413e-05, "loss": 0.065, "step": 3650 }, { "epoch": 14.076923076923077, "grad_norm": 2.848735809326172, "learning_rate": 2.857346162126464e-05, "loss": 0.0536, "step": 3660 }, { "epoch": 14.115384615384615, "grad_norm": 2.7207109928131104, "learning_rate": 2.8562886584087092e-05, "loss": 0.0624, "step": 3670 }, { "epoch": 14.153846153846153, "grad_norm": 2.6145403385162354, "learning_rate": 2.8552274466495525e-05, "loss": 0.0638, "step": 3680 }, { "epoch": 14.192307692307692, "grad_norm": 2.5214338302612305, "learning_rate": 2.8541625297503056e-05, "loss": 0.0645, "step": 3690 }, { "epoch": 14.23076923076923, "grad_norm": 2.453113555908203, "learning_rate": 2.8530939106224106e-05, "loss": 0.059, "step": 3700 }, { "epoch": 14.26923076923077, "grad_norm": 2.9459733963012695, "learning_rate": 2.8520215921874325e-05, "loss": 0.0619, "step": 3710 }, { "epoch": 14.307692307692308, "grad_norm": 3.1279542446136475, "learning_rate": 2.850945577377048e-05, "loss": 0.0652, "step": 3720 }, { "epoch": 14.346153846153847, "grad_norm": 3.1692466735839844, "learning_rate": 2.8498658691330406e-05, "loss": 0.056, "step": 3730 }, { "epoch": 14.384615384615385, "grad_norm": 2.7388417720794678, "learning_rate": 2.8487824704072913e-05, "loss": 0.0641, "step": 3740 }, { "epoch": 14.423076923076923, "grad_norm": 3.1448442935943604, "learning_rate": 2.8476953841617713e-05, "loss": 0.063, "step": 3750 }, { "epoch": 14.461538461538462, "grad_norm": 2.5632693767547607, "learning_rate": 2.846604613368532e-05, "loss": 0.0562, "step": 3760 }, { "epoch": 14.5, "grad_norm": 2.959221601486206, "learning_rate": 2.8455101610097002e-05, "loss": 0.0675, "step": 3770 }, { "epoch": 14.538461538461538, "grad_norm": 2.6775829792022705, "learning_rate": 2.8444120300774666e-05, "loss": 0.0606, "step": 3780 }, { "epoch": 14.576923076923077, "grad_norm": 2.562490940093994, "learning_rate": 2.8433102235740788e-05, "loss": 0.0573, "step": 3790 }, { "epoch": 14.615384615384615, "grad_norm": 2.8279271125793457, "learning_rate": 2.842204744511834e-05, "loss": 0.0595, "step": 3800 }, { "epoch": 14.653846153846153, "grad_norm": 2.3390533924102783, "learning_rate": 2.8410955959130693e-05, "loss": 0.0533, "step": 3810 }, { "epoch": 14.692307692307692, "grad_norm": 2.221428871154785, "learning_rate": 2.8399827808101554e-05, "loss": 0.0623, "step": 3820 }, { "epoch": 14.73076923076923, "grad_norm": 2.7612838745117188, "learning_rate": 2.8388663022454857e-05, "loss": 0.0616, "step": 3830 }, { "epoch": 14.76923076923077, "grad_norm": 2.730405330657959, "learning_rate": 2.83774616327147e-05, "loss": 0.0575, "step": 3840 }, { "epoch": 14.807692307692308, "grad_norm": 2.5589001178741455, "learning_rate": 2.836622366950526e-05, "loss": 0.0576, "step": 3850 }, { "epoch": 14.846153846153847, "grad_norm": 2.984184741973877, "learning_rate": 2.835494916355069e-05, "loss": 0.0635, "step": 3860 }, { "epoch": 14.884615384615385, "grad_norm": 2.6934828758239746, "learning_rate": 2.8343638145675072e-05, "loss": 0.0616, "step": 3870 }, { "epoch": 14.923076923076923, "grad_norm": 2.6913018226623535, "learning_rate": 2.8332290646802282e-05, "loss": 0.0601, "step": 3880 }, { "epoch": 14.961538461538462, "grad_norm": 2.394535779953003, "learning_rate": 2.8320906697955963e-05, "loss": 0.0562, "step": 3890 }, { "epoch": 15.0, "grad_norm": 2.9435136318206787, "learning_rate": 2.8309486330259385e-05, "loss": 0.0626, "step": 3900 }, { "epoch": 15.038461538461538, "grad_norm": 2.9360132217407227, "learning_rate": 2.82980295749354e-05, "loss": 0.0624, "step": 3910 }, { "epoch": 15.076923076923077, "grad_norm": 3.197340965270996, "learning_rate": 2.828653646330634e-05, "loss": 0.0612, "step": 3920 }, { "epoch": 15.115384615384615, "grad_norm": 2.8745882511138916, "learning_rate": 2.8275007026793938e-05, "loss": 0.0597, "step": 3930 }, { "epoch": 15.153846153846153, "grad_norm": 2.903532028198242, "learning_rate": 2.826344129691923e-05, "loss": 0.0632, "step": 3940 }, { "epoch": 15.192307692307692, "grad_norm": 2.7144417762756348, "learning_rate": 2.8251839305302478e-05, "loss": 0.0636, "step": 3950 }, { "epoch": 15.23076923076923, "grad_norm": 2.7780587673187256, "learning_rate": 2.8240201083663088e-05, "loss": 0.0569, "step": 3960 }, { "epoch": 15.26923076923077, "grad_norm": 2.372857093811035, "learning_rate": 2.8228526663819504e-05, "loss": 0.0615, "step": 3970 }, { "epoch": 15.307692307692308, "grad_norm": 2.953294277191162, "learning_rate": 2.8216816077689158e-05, "loss": 0.0569, "step": 3980 }, { "epoch": 15.346153846153847, "grad_norm": 3.1452503204345703, "learning_rate": 2.8205069357288337e-05, "loss": 0.0555, "step": 3990 }, { "epoch": 15.384615384615385, "grad_norm": 2.4111666679382324, "learning_rate": 2.8193286534732128e-05, "loss": 0.0612, "step": 4000 }, { "epoch": 15.423076923076923, "grad_norm": 2.9192652702331543, "learning_rate": 2.8181467642234317e-05, "loss": 0.0578, "step": 4010 }, { "epoch": 15.461538461538462, "grad_norm": 2.613917589187622, "learning_rate": 2.8169612712107306e-05, "loss": 0.0553, "step": 4020 }, { "epoch": 15.5, "grad_norm": 2.7946548461914062, "learning_rate": 2.8157721776762017e-05, "loss": 0.0632, "step": 4030 }, { "epoch": 15.538461538461538, "grad_norm": 2.815171003341675, "learning_rate": 2.814579486870782e-05, "loss": 0.0602, "step": 4040 }, { "epoch": 15.576923076923077, "grad_norm": 2.570457696914673, "learning_rate": 2.813383202055242e-05, "loss": 0.0551, "step": 4050 }, { "epoch": 15.615384615384615, "grad_norm": 2.0170300006866455, "learning_rate": 2.8121833265001792e-05, "loss": 0.0642, "step": 4060 }, { "epoch": 15.653846153846153, "grad_norm": 2.3233728408813477, "learning_rate": 2.8109798634860072e-05, "loss": 0.0588, "step": 4070 }, { "epoch": 15.692307692307692, "grad_norm": 2.1397790908813477, "learning_rate": 2.8097728163029482e-05, "loss": 0.0529, "step": 4080 }, { "epoch": 15.73076923076923, "grad_norm": 2.950807571411133, "learning_rate": 2.8085621882510233e-05, "loss": 0.0569, "step": 4090 }, { "epoch": 15.76923076923077, "grad_norm": 2.839064121246338, "learning_rate": 2.8073479826400425e-05, "loss": 0.0615, "step": 4100 }, { "epoch": 15.807692307692308, "grad_norm": 2.629194736480713, "learning_rate": 2.806130202789598e-05, "loss": 0.0596, "step": 4110 }, { "epoch": 15.846153846153847, "grad_norm": 2.859807252883911, "learning_rate": 2.804908852029054e-05, "loss": 0.0568, "step": 4120 }, { "epoch": 15.884615384615385, "grad_norm": 2.6774954795837402, "learning_rate": 2.8036839336975367e-05, "loss": 0.0573, "step": 4130 }, { "epoch": 15.923076923076923, "grad_norm": 3.763676166534424, "learning_rate": 2.8024554511439253e-05, "loss": 0.0612, "step": 4140 }, { "epoch": 15.961538461538462, "grad_norm": 2.945574998855591, "learning_rate": 2.801223407726844e-05, "loss": 0.0553, "step": 4150 }, { "epoch": 16.0, "grad_norm": 2.359447717666626, "learning_rate": 2.7999878068146537e-05, "loss": 0.049, "step": 4160 }, { "epoch": 16.03846153846154, "grad_norm": 2.7345235347747803, "learning_rate": 2.7987486517854396e-05, "loss": 0.0595, "step": 4170 }, { "epoch": 16.076923076923077, "grad_norm": 2.2802300453186035, "learning_rate": 2.7975059460270037e-05, "loss": 0.0622, "step": 4180 }, { "epoch": 16.115384615384617, "grad_norm": 2.078882932662964, "learning_rate": 2.7962596929368566e-05, "loss": 0.0585, "step": 4190 }, { "epoch": 16.153846153846153, "grad_norm": 2.182929039001465, "learning_rate": 2.795009895922207e-05, "loss": 0.0536, "step": 4200 }, { "epoch": 16.192307692307693, "grad_norm": 3.1578166484832764, "learning_rate": 2.7937565583999513e-05, "loss": 0.0517, "step": 4210 }, { "epoch": 16.23076923076923, "grad_norm": 2.627866268157959, "learning_rate": 2.792499683796667e-05, "loss": 0.0542, "step": 4220 }, { "epoch": 16.26923076923077, "grad_norm": 2.430981159210205, "learning_rate": 2.791239275548601e-05, "loss": 0.0484, "step": 4230 }, { "epoch": 16.307692307692307, "grad_norm": 2.130453586578369, "learning_rate": 2.789975337101662e-05, "loss": 0.0603, "step": 4240 }, { "epoch": 16.346153846153847, "grad_norm": 2.2407262325286865, "learning_rate": 2.788707871911409e-05, "loss": 0.0606, "step": 4250 }, { "epoch": 16.384615384615383, "grad_norm": 2.404421091079712, "learning_rate": 2.7874368834430426e-05, "loss": 0.0561, "step": 4260 }, { "epoch": 16.423076923076923, "grad_norm": 2.6223599910736084, "learning_rate": 2.7861623751713982e-05, "loss": 0.053, "step": 4270 }, { "epoch": 16.46153846153846, "grad_norm": 2.6188464164733887, "learning_rate": 2.7848843505809317e-05, "loss": 0.0559, "step": 4280 }, { "epoch": 16.5, "grad_norm": 2.1133105754852295, "learning_rate": 2.7836028131657142e-05, "loss": 0.0524, "step": 4290 }, { "epoch": 16.53846153846154, "grad_norm": 2.4061028957366943, "learning_rate": 2.7823177664294197e-05, "loss": 0.0545, "step": 4300 }, { "epoch": 16.576923076923077, "grad_norm": 2.635202169418335, "learning_rate": 2.7810292138853168e-05, "loss": 0.0526, "step": 4310 }, { "epoch": 16.615384615384617, "grad_norm": 2.4512650966644287, "learning_rate": 2.779737159056259e-05, "loss": 0.0527, "step": 4320 }, { "epoch": 16.653846153846153, "grad_norm": 2.650878429412842, "learning_rate": 2.7784416054746753e-05, "loss": 0.0518, "step": 4330 }, { "epoch": 16.692307692307693, "grad_norm": 2.4242632389068604, "learning_rate": 2.7771425566825593e-05, "loss": 0.0527, "step": 4340 }, { "epoch": 16.73076923076923, "grad_norm": 2.464691162109375, "learning_rate": 2.7758400162314605e-05, "loss": 0.0547, "step": 4350 }, { "epoch": 16.76923076923077, "grad_norm": 2.6380252838134766, "learning_rate": 2.7745339876824756e-05, "loss": 0.0503, "step": 4360 }, { "epoch": 16.807692307692307, "grad_norm": 2.551723003387451, "learning_rate": 2.7732244746062363e-05, "loss": 0.0537, "step": 4370 }, { "epoch": 16.846153846153847, "grad_norm": 2.8268020153045654, "learning_rate": 2.7719114805829015e-05, "loss": 0.0516, "step": 4380 }, { "epoch": 16.884615384615383, "grad_norm": 2.501936435699463, "learning_rate": 2.7705950092021465e-05, "loss": 0.0583, "step": 4390 }, { "epoch": 16.923076923076923, "grad_norm": 2.710240602493286, "learning_rate": 2.7692750640631533e-05, "loss": 0.0658, "step": 4400 }, { "epoch": 16.96153846153846, "grad_norm": 2.378331422805786, "learning_rate": 2.767951648774603e-05, "loss": 0.0548, "step": 4410 }, { "epoch": 17.0, "grad_norm": 2.894601345062256, "learning_rate": 2.766624766954661e-05, "loss": 0.0567, "step": 4420 }, { "epoch": 17.03846153846154, "grad_norm": 2.357337236404419, "learning_rate": 2.7652944222309727e-05, "loss": 0.0538, "step": 4430 }, { "epoch": 17.076923076923077, "grad_norm": 2.6987788677215576, "learning_rate": 2.7639606182406484e-05, "loss": 0.0617, "step": 4440 }, { "epoch": 17.115384615384617, "grad_norm": 1.861840009689331, "learning_rate": 2.7626233586302583e-05, "loss": 0.0487, "step": 4450 }, { "epoch": 17.153846153846153, "grad_norm": 1.9388928413391113, "learning_rate": 2.7612826470558192e-05, "loss": 0.0568, "step": 4460 }, { "epoch": 17.192307692307693, "grad_norm": 2.5828299522399902, "learning_rate": 2.7599384871827846e-05, "loss": 0.0535, "step": 4470 }, { "epoch": 17.23076923076923, "grad_norm": 2.6794981956481934, "learning_rate": 2.7585908826860368e-05, "loss": 0.0528, "step": 4480 }, { "epoch": 17.26923076923077, "grad_norm": 2.3347725868225098, "learning_rate": 2.757239837249875e-05, "loss": 0.0551, "step": 4490 }, { "epoch": 17.307692307692307, "grad_norm": 3.0247833728790283, "learning_rate": 2.7558853545680057e-05, "loss": 0.0524, "step": 4500 }, { "epoch": 17.346153846153847, "grad_norm": 2.5298354625701904, "learning_rate": 2.754527438343533e-05, "loss": 0.0624, "step": 4510 }, { "epoch": 17.384615384615383, "grad_norm": 2.508488893508911, "learning_rate": 2.7531660922889477e-05, "loss": 0.0548, "step": 4520 }, { "epoch": 17.423076923076923, "grad_norm": 2.869568347930908, "learning_rate": 2.751801320126118e-05, "loss": 0.0567, "step": 4530 }, { "epoch": 17.46153846153846, "grad_norm": 3.067972421646118, "learning_rate": 2.750433125586279e-05, "loss": 0.0527, "step": 4540 }, { "epoch": 17.5, "grad_norm": 2.691254138946533, "learning_rate": 2.7490615124100225e-05, "loss": 0.049, "step": 4550 }, { "epoch": 17.53846153846154, "grad_norm": 2.420832395553589, "learning_rate": 2.747686484347286e-05, "loss": 0.0488, "step": 4560 }, { "epoch": 17.576923076923077, "grad_norm": 2.2831807136535645, "learning_rate": 2.7463080451573447e-05, "loss": 0.0547, "step": 4570 }, { "epoch": 17.615384615384617, "grad_norm": 2.23234486579895, "learning_rate": 2.744926198608798e-05, "loss": 0.0521, "step": 4580 }, { "epoch": 17.653846153846153, "grad_norm": 2.116222620010376, "learning_rate": 2.743540948479561e-05, "loss": 0.0526, "step": 4590 }, { "epoch": 17.692307692307693, "grad_norm": 2.2791247367858887, "learning_rate": 2.7421522985568562e-05, "loss": 0.054, "step": 4600 }, { "epoch": 17.73076923076923, "grad_norm": 2.1204259395599365, "learning_rate": 2.7407602526371983e-05, "loss": 0.0544, "step": 4610 }, { "epoch": 17.76923076923077, "grad_norm": 2.725435733795166, "learning_rate": 2.7393648145263873e-05, "loss": 0.0545, "step": 4620 }, { "epoch": 17.807692307692307, "grad_norm": 2.7043561935424805, "learning_rate": 2.7379659880394996e-05, "loss": 0.0554, "step": 4630 }, { "epoch": 17.846153846153847, "grad_norm": 3.0216851234436035, "learning_rate": 2.7365637770008717e-05, "loss": 0.0551, "step": 4640 }, { "epoch": 17.884615384615383, "grad_norm": 2.6223690509796143, "learning_rate": 2.7351581852440953e-05, "loss": 0.0545, "step": 4650 }, { "epoch": 17.923076923076923, "grad_norm": 2.883230209350586, "learning_rate": 2.7337492166120053e-05, "loss": 0.0632, "step": 4660 }, { "epoch": 17.96153846153846, "grad_norm": 2.9789953231811523, "learning_rate": 2.732336874956667e-05, "loss": 0.0558, "step": 4670 }, { "epoch": 18.0, "grad_norm": 2.189687967300415, "learning_rate": 2.7309211641393696e-05, "loss": 0.0439, "step": 4680 }, { "epoch": 18.03846153846154, "grad_norm": 2.24808931350708, "learning_rate": 2.7295020880306123e-05, "loss": 0.0567, "step": 4690 }, { "epoch": 18.076923076923077, "grad_norm": 2.401685953140259, "learning_rate": 2.7280796505100946e-05, "loss": 0.0578, "step": 4700 }, { "epoch": 18.115384615384617, "grad_norm": 1.9197814464569092, "learning_rate": 2.7266538554667065e-05, "loss": 0.0525, "step": 4710 }, { "epoch": 18.153846153846153, "grad_norm": 2.022991180419922, "learning_rate": 2.725224706798517e-05, "loss": 0.0567, "step": 4720 }, { "epoch": 18.192307692307693, "grad_norm": 2.132225513458252, "learning_rate": 2.7237922084127643e-05, "loss": 0.0524, "step": 4730 }, { "epoch": 18.23076923076923, "grad_norm": 2.158095598220825, "learning_rate": 2.7223563642258446e-05, "loss": 0.0496, "step": 4740 }, { "epoch": 18.26923076923077, "grad_norm": 2.268658399581909, "learning_rate": 2.7209171781633e-05, "loss": 0.0536, "step": 4750 }, { "epoch": 18.307692307692307, "grad_norm": 2.788698673248291, "learning_rate": 2.7194746541598113e-05, "loss": 0.0527, "step": 4760 }, { "epoch": 18.346153846153847, "grad_norm": 2.46181321144104, "learning_rate": 2.7180287961591835e-05, "loss": 0.0547, "step": 4770 }, { "epoch": 18.384615384615383, "grad_norm": 2.1883716583251953, "learning_rate": 2.7165796081143377e-05, "loss": 0.0554, "step": 4780 }, { "epoch": 18.423076923076923, "grad_norm": 2.5119552612304688, "learning_rate": 2.715127093987298e-05, "loss": 0.0502, "step": 4790 }, { "epoch": 18.46153846153846, "grad_norm": 2.3125741481781006, "learning_rate": 2.713671257749183e-05, "loss": 0.0521, "step": 4800 }, { "epoch": 18.5, "grad_norm": 2.01935076713562, "learning_rate": 2.712212103380193e-05, "loss": 0.049, "step": 4810 }, { "epoch": 18.53846153846154, "grad_norm": 2.192258834838867, "learning_rate": 2.7107496348696004e-05, "loss": 0.0468, "step": 4820 }, { "epoch": 18.576923076923077, "grad_norm": 2.0872395038604736, "learning_rate": 2.7092838562157386e-05, "loss": 0.0608, "step": 4830 }, { "epoch": 18.615384615384617, "grad_norm": 2.0910723209381104, "learning_rate": 2.7078147714259905e-05, "loss": 0.0597, "step": 4840 }, { "epoch": 18.653846153846153, "grad_norm": 2.1819539070129395, "learning_rate": 2.7063423845167773e-05, "loss": 0.0514, "step": 4850 }, { "epoch": 18.692307692307693, "grad_norm": 2.541072368621826, "learning_rate": 2.7048666995135494e-05, "loss": 0.0518, "step": 4860 }, { "epoch": 18.73076923076923, "grad_norm": 2.253460168838501, "learning_rate": 2.7033877204507722e-05, "loss": 0.0519, "step": 4870 }, { "epoch": 18.76923076923077, "grad_norm": 2.454606294631958, "learning_rate": 2.701905451371919e-05, "loss": 0.0559, "step": 4880 }, { "epoch": 18.807692307692307, "grad_norm": 2.596813917160034, "learning_rate": 2.7004198963294558e-05, "loss": 0.0557, "step": 4890 }, { "epoch": 18.846153846153847, "grad_norm": 2.3953702449798584, "learning_rate": 2.6989310593848345e-05, "loss": 0.0553, "step": 4900 }, { "epoch": 18.884615384615383, "grad_norm": 2.1888365745544434, "learning_rate": 2.6974389446084776e-05, "loss": 0.0488, "step": 4910 }, { "epoch": 18.923076923076923, "grad_norm": 2.3826677799224854, "learning_rate": 2.6959435560797706e-05, "loss": 0.0495, "step": 4920 }, { "epoch": 18.96153846153846, "grad_norm": 2.1377201080322266, "learning_rate": 2.6944448978870478e-05, "loss": 0.0566, "step": 4930 }, { "epoch": 19.0, "grad_norm": 2.3483967781066895, "learning_rate": 2.6929429741275845e-05, "loss": 0.0488, "step": 4940 }, { "epoch": 19.03846153846154, "grad_norm": 1.9696838855743408, "learning_rate": 2.691437788907582e-05, "loss": 0.051, "step": 4950 }, { "epoch": 19.076923076923077, "grad_norm": 2.692110776901245, "learning_rate": 2.689929346342159e-05, "loss": 0.0461, "step": 4960 }, { "epoch": 19.115384615384617, "grad_norm": 2.2084898948669434, "learning_rate": 2.688417650555341e-05, "loss": 0.0599, "step": 4970 }, { "epoch": 19.153846153846153, "grad_norm": 2.121561288833618, "learning_rate": 2.686902705680046e-05, "loss": 0.0578, "step": 4980 }, { "epoch": 19.192307692307693, "grad_norm": 2.3033506870269775, "learning_rate": 2.6853845158580756e-05, "loss": 0.0519, "step": 4990 }, { "epoch": 19.23076923076923, "grad_norm": 2.0581743717193604, "learning_rate": 2.6838630852401028e-05, "loss": 0.0538, "step": 5000 }, { "epoch": 19.26923076923077, "grad_norm": 2.1499931812286377, "learning_rate": 2.6823384179856602e-05, "loss": 0.0515, "step": 5010 }, { "epoch": 19.307692307692307, "grad_norm": 2.4993667602539062, "learning_rate": 2.6808105182631303e-05, "loss": 0.0537, "step": 5020 }, { "epoch": 19.346153846153847, "grad_norm": 2.7367970943450928, "learning_rate": 2.6792793902497328e-05, "loss": 0.0507, "step": 5030 }, { "epoch": 19.384615384615383, "grad_norm": 2.3506581783294678, "learning_rate": 2.6777450381315133e-05, "loss": 0.0501, "step": 5040 }, { "epoch": 19.423076923076923, "grad_norm": 1.9951107501983643, "learning_rate": 2.676207466103331e-05, "loss": 0.0519, "step": 5050 }, { "epoch": 19.46153846153846, "grad_norm": 2.0051987171173096, "learning_rate": 2.6746666783688503e-05, "loss": 0.0505, "step": 5060 }, { "epoch": 19.5, "grad_norm": 1.8392553329467773, "learning_rate": 2.673122679140525e-05, "loss": 0.0421, "step": 5070 }, { "epoch": 19.53846153846154, "grad_norm": 1.9367773532867432, "learning_rate": 2.671575472639591e-05, "loss": 0.0514, "step": 5080 }, { "epoch": 19.576923076923077, "grad_norm": 2.1854703426361084, "learning_rate": 2.6700250630960506e-05, "loss": 0.0535, "step": 5090 }, { "epoch": 19.615384615384617, "grad_norm": 2.0213778018951416, "learning_rate": 2.6684714547486654e-05, "loss": 0.0546, "step": 5100 }, { "epoch": 19.653846153846153, "grad_norm": 2.09897780418396, "learning_rate": 2.6669146518449407e-05, "loss": 0.0477, "step": 5110 }, { "epoch": 19.692307692307693, "grad_norm": 2.4380099773406982, "learning_rate": 2.665354658641117e-05, "loss": 0.0611, "step": 5120 }, { "epoch": 19.73076923076923, "grad_norm": 2.0932047367095947, "learning_rate": 2.6637914794021552e-05, "loss": 0.0462, "step": 5130 }, { "epoch": 19.76923076923077, "grad_norm": 2.186757802963257, "learning_rate": 2.6622251184017274e-05, "loss": 0.0487, "step": 5140 }, { "epoch": 19.807692307692307, "grad_norm": 1.9822323322296143, "learning_rate": 2.660655579922206e-05, "loss": 0.0498, "step": 5150 }, { "epoch": 19.846153846153847, "grad_norm": 2.4814295768737793, "learning_rate": 2.6590828682546487e-05, "loss": 0.0541, "step": 5160 }, { "epoch": 19.884615384615383, "grad_norm": 2.3453330993652344, "learning_rate": 2.657506987698789e-05, "loss": 0.0403, "step": 5170 }, { "epoch": 19.923076923076923, "grad_norm": 2.3614869117736816, "learning_rate": 2.655927942563024e-05, "loss": 0.0491, "step": 5180 }, { "epoch": 19.96153846153846, "grad_norm": 2.508418083190918, "learning_rate": 2.6543457371644027e-05, "loss": 0.0506, "step": 5190 }, { "epoch": 20.0, "grad_norm": 1.809973955154419, "learning_rate": 2.652760375828615e-05, "loss": 0.0553, "step": 5200 }, { "epoch": 20.03846153846154, "grad_norm": 2.3840510845184326, "learning_rate": 2.651171862889978e-05, "loss": 0.0511, "step": 5210 }, { "epoch": 20.076923076923077, "grad_norm": 2.154362440109253, "learning_rate": 2.649580202691425e-05, "loss": 0.0522, "step": 5220 }, { "epoch": 20.115384615384617, "grad_norm": 2.4772872924804688, "learning_rate": 2.6479853995844942e-05, "loss": 0.0524, "step": 5230 }, { "epoch": 20.153846153846153, "grad_norm": 2.3529508113861084, "learning_rate": 2.646387457929317e-05, "loss": 0.0494, "step": 5240 }, { "epoch": 20.192307692307693, "grad_norm": 2.08630633354187, "learning_rate": 2.6447863820946047e-05, "loss": 0.0462, "step": 5250 }, { "epoch": 20.23076923076923, "grad_norm": 1.6905722618103027, "learning_rate": 2.6431821764576367e-05, "loss": 0.0512, "step": 5260 }, { "epoch": 20.26923076923077, "grad_norm": 1.8629543781280518, "learning_rate": 2.641574845404251e-05, "loss": 0.0502, "step": 5270 }, { "epoch": 20.307692307692307, "grad_norm": 2.0543575286865234, "learning_rate": 2.639964393328829e-05, "loss": 0.0545, "step": 5280 }, { "epoch": 20.346153846153847, "grad_norm": 2.485801935195923, "learning_rate": 2.6383508246342844e-05, "loss": 0.0514, "step": 5290 }, { "epoch": 20.384615384615383, "grad_norm": 1.7313271760940552, "learning_rate": 2.636734143732054e-05, "loss": 0.0488, "step": 5300 }, { "epoch": 20.423076923076923, "grad_norm": 1.8892971277236938, "learning_rate": 2.63511435504208e-05, "loss": 0.05, "step": 5310 }, { "epoch": 20.46153846153846, "grad_norm": 1.8958619832992554, "learning_rate": 2.633491462992804e-05, "loss": 0.0509, "step": 5320 }, { "epoch": 20.5, "grad_norm": 2.0657594203948975, "learning_rate": 2.63186547202115e-05, "loss": 0.0457, "step": 5330 }, { "epoch": 20.53846153846154, "grad_norm": 2.0821566581726074, "learning_rate": 2.6302363865725158e-05, "loss": 0.0515, "step": 5340 }, { "epoch": 20.576923076923077, "grad_norm": 1.938544750213623, "learning_rate": 2.628604211100759e-05, "loss": 0.0505, "step": 5350 }, { "epoch": 20.615384615384617, "grad_norm": 2.1884799003601074, "learning_rate": 2.6269689500681846e-05, "loss": 0.0514, "step": 5360 }, { "epoch": 20.653846153846153, "grad_norm": 2.3804664611816406, "learning_rate": 2.6253306079455337e-05, "loss": 0.0541, "step": 5370 }, { "epoch": 20.692307692307693, "grad_norm": 2.18882155418396, "learning_rate": 2.6236891892119713e-05, "loss": 0.0492, "step": 5380 }, { "epoch": 20.73076923076923, "grad_norm": 2.1746103763580322, "learning_rate": 2.6220446983550738e-05, "loss": 0.0487, "step": 5390 }, { "epoch": 20.76923076923077, "grad_norm": 1.9990772008895874, "learning_rate": 2.6203971398708162e-05, "loss": 0.0492, "step": 5400 }, { "epoch": 20.807692307692307, "grad_norm": 2.243596315383911, "learning_rate": 2.6187465182635598e-05, "loss": 0.0521, "step": 5410 }, { "epoch": 20.846153846153847, "grad_norm": 2.5800552368164062, "learning_rate": 2.6170928380460424e-05, "loss": 0.0468, "step": 5420 }, { "epoch": 20.884615384615383, "grad_norm": 2.2908005714416504, "learning_rate": 2.615436103739362e-05, "loss": 0.051, "step": 5430 }, { "epoch": 20.923076923076923, "grad_norm": 2.4502663612365723, "learning_rate": 2.6137763198729665e-05, "loss": 0.0476, "step": 5440 }, { "epoch": 20.96153846153846, "grad_norm": 2.216867208480835, "learning_rate": 2.6121134909846416e-05, "loss": 0.0553, "step": 5450 }, { "epoch": 21.0, "grad_norm": 2.4995601177215576, "learning_rate": 2.6104476216204985e-05, "loss": 0.0453, "step": 5460 }, { "epoch": 21.03846153846154, "grad_norm": 1.9124618768692017, "learning_rate": 2.6087787163349605e-05, "loss": 0.0479, "step": 5470 }, { "epoch": 21.076923076923077, "grad_norm": 1.9513589143753052, "learning_rate": 2.60710677969075e-05, "loss": 0.0476, "step": 5480 }, { "epoch": 21.115384615384617, "grad_norm": 1.7153360843658447, "learning_rate": 2.6054318162588792e-05, "loss": 0.044, "step": 5490 }, { "epoch": 21.153846153846153, "grad_norm": 2.320129871368408, "learning_rate": 2.6037538306186337e-05, "loss": 0.0456, "step": 5500 }, { "epoch": 21.192307692307693, "grad_norm": 1.6431825160980225, "learning_rate": 2.602072827357562e-05, "loss": 0.0483, "step": 5510 }, { "epoch": 21.23076923076923, "grad_norm": 2.124098539352417, "learning_rate": 2.6003888110714624e-05, "loss": 0.0402, "step": 5520 }, { "epoch": 21.26923076923077, "grad_norm": 2.669579267501831, "learning_rate": 2.5987017863643714e-05, "loss": 0.0547, "step": 5530 }, { "epoch": 21.307692307692307, "grad_norm": 2.1754443645477295, "learning_rate": 2.5970117578485506e-05, "loss": 0.0481, "step": 5540 }, { "epoch": 21.346153846153847, "grad_norm": 2.0452888011932373, "learning_rate": 2.5953187301444733e-05, "loss": 0.0436, "step": 5550 }, { "epoch": 21.384615384615383, "grad_norm": 2.2750635147094727, "learning_rate": 2.5936227078808123e-05, "loss": 0.0498, "step": 5560 }, { "epoch": 21.423076923076923, "grad_norm": 1.7035436630249023, "learning_rate": 2.5919236956944277e-05, "loss": 0.0455, "step": 5570 }, { "epoch": 21.46153846153846, "grad_norm": 2.014969825744629, "learning_rate": 2.5902216982303544e-05, "loss": 0.0502, "step": 5580 }, { "epoch": 21.5, "grad_norm": 2.2059478759765625, "learning_rate": 2.588516720141788e-05, "loss": 0.0466, "step": 5590 }, { "epoch": 21.53846153846154, "grad_norm": 1.8128308057785034, "learning_rate": 2.5868087660900735e-05, "loss": 0.0494, "step": 5600 }, { "epoch": 21.576923076923077, "grad_norm": 2.221346139907837, "learning_rate": 2.5850978407446924e-05, "loss": 0.0532, "step": 5610 }, { "epoch": 21.615384615384617, "grad_norm": 1.9685431718826294, "learning_rate": 2.5833839487832488e-05, "loss": 0.0506, "step": 5620 }, { "epoch": 21.653846153846153, "grad_norm": 1.538164734840393, "learning_rate": 2.5816670948914583e-05, "loss": 0.0446, "step": 5630 }, { "epoch": 21.692307692307693, "grad_norm": 1.8389371633529663, "learning_rate": 2.5799472837631338e-05, "loss": 0.053, "step": 5640 }, { "epoch": 21.73076923076923, "grad_norm": 2.2778820991516113, "learning_rate": 2.578224520100173e-05, "loss": 0.0423, "step": 5650 }, { "epoch": 21.76923076923077, "grad_norm": 2.5849103927612305, "learning_rate": 2.576498808612546e-05, "loss": 0.0489, "step": 5660 }, { "epoch": 21.807692307692307, "grad_norm": 2.3590989112854004, "learning_rate": 2.5747701540182825e-05, "loss": 0.0514, "step": 5670 }, { "epoch": 21.846153846153847, "grad_norm": 1.8261209726333618, "learning_rate": 2.573038561043458e-05, "loss": 0.0459, "step": 5680 }, { "epoch": 21.884615384615383, "grad_norm": 2.492357015609741, "learning_rate": 2.5713040344221815e-05, "loss": 0.053, "step": 5690 }, { "epoch": 21.923076923076923, "grad_norm": 2.3282649517059326, "learning_rate": 2.5695665788965823e-05, "loss": 0.0422, "step": 5700 }, { "epoch": 21.96153846153846, "grad_norm": 2.3369557857513428, "learning_rate": 2.5678261992167978e-05, "loss": 0.0542, "step": 5710 }, { "epoch": 22.0, "grad_norm": 2.2342748641967773, "learning_rate": 2.5660829001409594e-05, "loss": 0.0553, "step": 5720 }, { "epoch": 22.03846153846154, "grad_norm": 1.9400802850723267, "learning_rate": 2.5643366864351806e-05, "loss": 0.0554, "step": 5730 }, { "epoch": 22.076923076923077, "grad_norm": 2.1915154457092285, "learning_rate": 2.5625875628735423e-05, "loss": 0.0463, "step": 5740 }, { "epoch": 22.115384615384617, "grad_norm": 1.7244971990585327, "learning_rate": 2.560835534238082e-05, "loss": 0.0466, "step": 5750 }, { "epoch": 22.153846153846153, "grad_norm": 2.1806528568267822, "learning_rate": 2.5590806053187793e-05, "loss": 0.0483, "step": 5760 }, { "epoch": 22.192307692307693, "grad_norm": 1.9255647659301758, "learning_rate": 2.557322780913542e-05, "loss": 0.0489, "step": 5770 }, { "epoch": 22.23076923076923, "grad_norm": 2.2142374515533447, "learning_rate": 2.555562065828196e-05, "loss": 0.0527, "step": 5780 }, { "epoch": 22.26923076923077, "grad_norm": 2.4573400020599365, "learning_rate": 2.5537984648764684e-05, "loss": 0.0451, "step": 5790 }, { "epoch": 22.307692307692307, "grad_norm": 2.1143500804901123, "learning_rate": 2.5520319828799766e-05, "loss": 0.0479, "step": 5800 }, { "epoch": 22.346153846153847, "grad_norm": 1.8646248579025269, "learning_rate": 2.550262624668216e-05, "loss": 0.0435, "step": 5810 }, { "epoch": 22.384615384615383, "grad_norm": 1.5074269771575928, "learning_rate": 2.5484903950785432e-05, "loss": 0.0446, "step": 5820 }, { "epoch": 22.423076923076923, "grad_norm": 2.103839874267578, "learning_rate": 2.546715298956167e-05, "loss": 0.0527, "step": 5830 }, { "epoch": 22.46153846153846, "grad_norm": 1.9880903959274292, "learning_rate": 2.5449373411541322e-05, "loss": 0.0492, "step": 5840 }, { "epoch": 22.5, "grad_norm": 1.8882006406784058, "learning_rate": 2.5431565265333074e-05, "loss": 0.0552, "step": 5850 }, { "epoch": 22.53846153846154, "grad_norm": 2.3421194553375244, "learning_rate": 2.541372859962372e-05, "loss": 0.0487, "step": 5860 }, { "epoch": 22.576923076923077, "grad_norm": 2.0303971767425537, "learning_rate": 2.5395863463178023e-05, "loss": 0.0467, "step": 5870 }, { "epoch": 22.615384615384617, "grad_norm": 1.9958126544952393, "learning_rate": 2.537796990483858e-05, "loss": 0.0593, "step": 5880 }, { "epoch": 22.653846153846153, "grad_norm": 2.40570068359375, "learning_rate": 2.53600479735257e-05, "loss": 0.046, "step": 5890 }, { "epoch": 22.692307692307693, "grad_norm": 1.7980605363845825, "learning_rate": 2.5342097718237262e-05, "loss": 0.0496, "step": 5900 }, { "epoch": 22.73076923076923, "grad_norm": 2.364152431488037, "learning_rate": 2.5324119188048567e-05, "loss": 0.0485, "step": 5910 }, { "epoch": 22.76923076923077, "grad_norm": 2.065056085586548, "learning_rate": 2.530611243211224e-05, "loss": 0.0483, "step": 5920 }, { "epoch": 22.807692307692307, "grad_norm": 1.8822139501571655, "learning_rate": 2.5288077499658064e-05, "loss": 0.0445, "step": 5930 }, { "epoch": 22.846153846153847, "grad_norm": 1.9192858934402466, "learning_rate": 2.527001443999285e-05, "loss": 0.0462, "step": 5940 }, { "epoch": 22.884615384615383, "grad_norm": 2.175851821899414, "learning_rate": 2.5251923302500318e-05, "loss": 0.0448, "step": 5950 }, { "epoch": 22.923076923076923, "grad_norm": 2.1367673873901367, "learning_rate": 2.523380413664095e-05, "loss": 0.0461, "step": 5960 }, { "epoch": 22.96153846153846, "grad_norm": 2.0076870918273926, "learning_rate": 2.5215656991951844e-05, "loss": 0.0429, "step": 5970 }, { "epoch": 23.0, "grad_norm": 2.2983851432800293, "learning_rate": 2.5197481918046606e-05, "loss": 0.0473, "step": 5980 }, { "epoch": 23.03846153846154, "grad_norm": 1.547244906425476, "learning_rate": 2.5179278964615192e-05, "loss": 0.0441, "step": 5990 }, { "epoch": 23.076923076923077, "grad_norm": 1.8851597309112549, "learning_rate": 2.516104818142379e-05, "loss": 0.0494, "step": 6000 }, { "epoch": 23.115384615384617, "grad_norm": 1.918543815612793, "learning_rate": 2.5142789618314654e-05, "loss": 0.0417, "step": 6010 }, { "epoch": 23.153846153846153, "grad_norm": 1.7837193012237549, "learning_rate": 2.5124503325206006e-05, "loss": 0.0469, "step": 6020 }, { "epoch": 23.192307692307693, "grad_norm": 1.9671348333358765, "learning_rate": 2.5106189352091867e-05, "loss": 0.0429, "step": 6030 }, { "epoch": 23.23076923076923, "grad_norm": 1.8569093942642212, "learning_rate": 2.5087847749041944e-05, "loss": 0.0498, "step": 6040 }, { "epoch": 23.26923076923077, "grad_norm": 2.0214736461639404, "learning_rate": 2.506947856620148e-05, "loss": 0.0395, "step": 6050 }, { "epoch": 23.307692307692307, "grad_norm": 1.9467337131500244, "learning_rate": 2.505108185379111e-05, "loss": 0.046, "step": 6060 }, { "epoch": 23.346153846153847, "grad_norm": 1.8502717018127441, "learning_rate": 2.503265766210676e-05, "loss": 0.044, "step": 6070 }, { "epoch": 23.384615384615383, "grad_norm": 2.310283660888672, "learning_rate": 2.5014206041519456e-05, "loss": 0.0521, "step": 6080 }, { "epoch": 23.423076923076923, "grad_norm": 1.6830577850341797, "learning_rate": 2.499572704247523e-05, "loss": 0.0458, "step": 6090 }, { "epoch": 23.46153846153846, "grad_norm": 1.6736561059951782, "learning_rate": 2.497722071549495e-05, "loss": 0.0422, "step": 6100 }, { "epoch": 23.5, "grad_norm": 1.5868667364120483, "learning_rate": 2.4958687111174216e-05, "loss": 0.0432, "step": 6110 }, { "epoch": 23.53846153846154, "grad_norm": 2.2007408142089844, "learning_rate": 2.494012628018319e-05, "loss": 0.0495, "step": 6120 }, { "epoch": 23.576923076923077, "grad_norm": 1.7047786712646484, "learning_rate": 2.4921538273266475e-05, "loss": 0.0491, "step": 6130 }, { "epoch": 23.615384615384617, "grad_norm": 2.0366289615631104, "learning_rate": 2.490292314124298e-05, "loss": 0.0413, "step": 6140 }, { "epoch": 23.653846153846153, "grad_norm": 2.104093551635742, "learning_rate": 2.4884280935005755e-05, "loss": 0.0395, "step": 6150 }, { "epoch": 23.692307692307693, "grad_norm": 2.0423264503479004, "learning_rate": 2.486561170552188e-05, "loss": 0.051, "step": 6160 }, { "epoch": 23.73076923076923, "grad_norm": 2.2798447608947754, "learning_rate": 2.4846915503832326e-05, "loss": 0.0504, "step": 6170 }, { "epoch": 23.76923076923077, "grad_norm": 2.168039083480835, "learning_rate": 2.4828192381051787e-05, "loss": 0.0494, "step": 6180 }, { "epoch": 23.807692307692307, "grad_norm": 2.1671855449676514, "learning_rate": 2.480944238836857e-05, "loss": 0.0532, "step": 6190 }, { "epoch": 23.846153846153847, "grad_norm": 1.7856274843215942, "learning_rate": 2.4790665577044428e-05, "loss": 0.046, "step": 6200 }, { "epoch": 23.884615384615383, "grad_norm": 2.0235636234283447, "learning_rate": 2.4771861998414458e-05, "loss": 0.0465, "step": 6210 }, { "epoch": 23.923076923076923, "grad_norm": 2.179847478866577, "learning_rate": 2.475303170388692e-05, "loss": 0.0447, "step": 6220 }, { "epoch": 23.96153846153846, "grad_norm": 1.6464632749557495, "learning_rate": 2.4734174744943122e-05, "loss": 0.0449, "step": 6230 }, { "epoch": 24.0, "grad_norm": 1.7687687873840332, "learning_rate": 2.471529117313727e-05, "loss": 0.0447, "step": 6240 }, { "epoch": 24.03846153846154, "grad_norm": 1.8370416164398193, "learning_rate": 2.4696381040096335e-05, "loss": 0.0476, "step": 6250 }, { "epoch": 24.076923076923077, "grad_norm": 1.8568719625473022, "learning_rate": 2.4677444397519883e-05, "loss": 0.0478, "step": 6260 }, { "epoch": 24.115384615384617, "grad_norm": 1.9673206806182861, "learning_rate": 2.4658481297179987e-05, "loss": 0.0466, "step": 6270 }, { "epoch": 24.153846153846153, "grad_norm": 2.064671516418457, "learning_rate": 2.4639491790921028e-05, "loss": 0.0482, "step": 6280 }, { "epoch": 24.192307692307693, "grad_norm": 2.0179781913757324, "learning_rate": 2.4620475930659596e-05, "loss": 0.0491, "step": 6290 }, { "epoch": 24.23076923076923, "grad_norm": 1.8786386251449585, "learning_rate": 2.4601433768384327e-05, "loss": 0.042, "step": 6300 }, { "epoch": 24.26923076923077, "grad_norm": 2.0720064640045166, "learning_rate": 2.4582365356155766e-05, "loss": 0.0452, "step": 6310 }, { "epoch": 24.307692307692307, "grad_norm": 2.04176926612854, "learning_rate": 2.4563270746106224e-05, "loss": 0.0432, "step": 6320 }, { "epoch": 24.346153846153847, "grad_norm": 1.7975056171417236, "learning_rate": 2.4544149990439632e-05, "loss": 0.0431, "step": 6330 }, { "epoch": 24.384615384615383, "grad_norm": 1.9778639078140259, "learning_rate": 2.4525003141431413e-05, "loss": 0.0431, "step": 6340 }, { "epoch": 24.423076923076923, "grad_norm": 1.931883454322815, "learning_rate": 2.450583025142831e-05, "loss": 0.0411, "step": 6350 }, { "epoch": 24.46153846153846, "grad_norm": 2.43259596824646, "learning_rate": 2.4486631372848286e-05, "loss": 0.0464, "step": 6360 }, { "epoch": 24.5, "grad_norm": 2.0310606956481934, "learning_rate": 2.4467406558180328e-05, "loss": 0.0453, "step": 6370 }, { "epoch": 24.53846153846154, "grad_norm": 1.961578130722046, "learning_rate": 2.4448155859984357e-05, "loss": 0.0425, "step": 6380 }, { "epoch": 24.576923076923077, "grad_norm": 1.9241677522659302, "learning_rate": 2.442887933089104e-05, "loss": 0.0489, "step": 6390 }, { "epoch": 24.615384615384617, "grad_norm": 1.6629705429077148, "learning_rate": 2.440957702360167e-05, "loss": 0.0449, "step": 6400 }, { "epoch": 24.653846153846153, "grad_norm": 1.9020706415176392, "learning_rate": 2.4390248990888026e-05, "loss": 0.0468, "step": 6410 }, { "epoch": 24.692307692307693, "grad_norm": 1.9919699430465698, "learning_rate": 2.4370895285592202e-05, "loss": 0.047, "step": 6420 }, { "epoch": 24.73076923076923, "grad_norm": 1.7475947141647339, "learning_rate": 2.43515159606265e-05, "loss": 0.041, "step": 6430 }, { "epoch": 24.76923076923077, "grad_norm": 1.8714221715927124, "learning_rate": 2.4332111068973243e-05, "loss": 0.045, "step": 6440 }, { "epoch": 24.807692307692307, "grad_norm": 1.8813892602920532, "learning_rate": 2.4312680663684674e-05, "loss": 0.0416, "step": 6450 }, { "epoch": 24.846153846153847, "grad_norm": 1.9131197929382324, "learning_rate": 2.429322479788277e-05, "loss": 0.0423, "step": 6460 }, { "epoch": 24.884615384615383, "grad_norm": 1.9985467195510864, "learning_rate": 2.4273743524759132e-05, "loss": 0.0496, "step": 6470 }, { "epoch": 24.923076923076923, "grad_norm": 2.1156630516052246, "learning_rate": 2.4254236897574818e-05, "loss": 0.0425, "step": 6480 }, { "epoch": 24.96153846153846, "grad_norm": 2.0164928436279297, "learning_rate": 2.4234704969660192e-05, "loss": 0.0421, "step": 6490 }, { "epoch": 25.0, "grad_norm": 1.9131778478622437, "learning_rate": 2.4215147794414806e-05, "loss": 0.0431, "step": 6500 }, { "epoch": 25.03846153846154, "grad_norm": 2.0394511222839355, "learning_rate": 2.419556542530723e-05, "loss": 0.0462, "step": 6510 }, { "epoch": 25.076923076923077, "grad_norm": 1.879845142364502, "learning_rate": 2.4175957915874916e-05, "loss": 0.0489, "step": 6520 }, { "epoch": 25.115384615384617, "grad_norm": 1.9313360452651978, "learning_rate": 2.4156325319724037e-05, "loss": 0.0473, "step": 6530 }, { "epoch": 25.153846153846153, "grad_norm": 1.9558639526367188, "learning_rate": 2.4136667690529372e-05, "loss": 0.0432, "step": 6540 }, { "epoch": 25.192307692307693, "grad_norm": 1.932648777961731, "learning_rate": 2.4116985082034126e-05, "loss": 0.0441, "step": 6550 }, { "epoch": 25.23076923076923, "grad_norm": 1.8341952562332153, "learning_rate": 2.409727754804979e-05, "loss": 0.0484, "step": 6560 }, { "epoch": 25.26923076923077, "grad_norm": 2.142963171005249, "learning_rate": 2.4077545142456025e-05, "loss": 0.0491, "step": 6570 }, { "epoch": 25.307692307692307, "grad_norm": 2.0177972316741943, "learning_rate": 2.405778791920046e-05, "loss": 0.0581, "step": 6580 }, { "epoch": 25.346153846153847, "grad_norm": 2.0351908206939697, "learning_rate": 2.4038005932298594e-05, "loss": 0.0448, "step": 6590 }, { "epoch": 25.384615384615383, "grad_norm": 1.8626478910446167, "learning_rate": 2.4018199235833624e-05, "loss": 0.0428, "step": 6600 }, { "epoch": 25.423076923076923, "grad_norm": 1.7484854459762573, "learning_rate": 2.3998367883956306e-05, "loss": 0.0462, "step": 6610 }, { "epoch": 25.46153846153846, "grad_norm": 1.729616403579712, "learning_rate": 2.3978511930884795e-05, "loss": 0.0421, "step": 6620 }, { "epoch": 25.5, "grad_norm": 1.5235844850540161, "learning_rate": 2.3958631430904504e-05, "loss": 0.0489, "step": 6630 }, { "epoch": 25.53846153846154, "grad_norm": 1.9413248300552368, "learning_rate": 2.393872643836797e-05, "loss": 0.0464, "step": 6640 }, { "epoch": 25.576923076923077, "grad_norm": 1.6310465335845947, "learning_rate": 2.3918797007694675e-05, "loss": 0.0412, "step": 6650 }, { "epoch": 25.615384615384617, "grad_norm": 2.049593925476074, "learning_rate": 2.3898843193370923e-05, "loss": 0.0451, "step": 6660 }, { "epoch": 25.653846153846153, "grad_norm": 1.554084062576294, "learning_rate": 2.387886504994969e-05, "loss": 0.0434, "step": 6670 }, { "epoch": 25.692307692307693, "grad_norm": 2.2045257091522217, "learning_rate": 2.385886263205044e-05, "loss": 0.0458, "step": 6680 }, { "epoch": 25.73076923076923, "grad_norm": 2.1516737937927246, "learning_rate": 2.3838835994359036e-05, "loss": 0.0445, "step": 6690 }, { "epoch": 25.76923076923077, "grad_norm": 1.9684396982192993, "learning_rate": 2.3818785191627525e-05, "loss": 0.0442, "step": 6700 }, { "epoch": 25.807692307692307, "grad_norm": 1.7377302646636963, "learning_rate": 2.379871027867405e-05, "loss": 0.0443, "step": 6710 }, { "epoch": 25.846153846153847, "grad_norm": 1.860579490661621, "learning_rate": 2.3778611310382653e-05, "loss": 0.043, "step": 6720 }, { "epoch": 25.884615384615383, "grad_norm": 2.0663163661956787, "learning_rate": 2.3758488341703137e-05, "loss": 0.0444, "step": 6730 }, { "epoch": 25.923076923076923, "grad_norm": 1.8501702547073364, "learning_rate": 2.3738341427650945e-05, "loss": 0.0497, "step": 6740 }, { "epoch": 25.96153846153846, "grad_norm": 1.7028001546859741, "learning_rate": 2.3718170623306955e-05, "loss": 0.0418, "step": 6750 }, { "epoch": 26.0, "grad_norm": 1.7601794004440308, "learning_rate": 2.369797598381739e-05, "loss": 0.0413, "step": 6760 }, { "epoch": 26.03846153846154, "grad_norm": 1.8938238620758057, "learning_rate": 2.3677757564393612e-05, "loss": 0.0397, "step": 6770 }, { "epoch": 26.076923076923077, "grad_norm": 2.0491912364959717, "learning_rate": 2.3657515420312015e-05, "loss": 0.0414, "step": 6780 }, { "epoch": 26.115384615384617, "grad_norm": 2.1569526195526123, "learning_rate": 2.3637249606913847e-05, "loss": 0.045, "step": 6790 }, { "epoch": 26.153846153846153, "grad_norm": 1.8800678253173828, "learning_rate": 2.3616960179605064e-05, "loss": 0.0532, "step": 6800 }, { "epoch": 26.192307692307693, "grad_norm": 1.7340847253799438, "learning_rate": 2.3596647193856188e-05, "loss": 0.0409, "step": 6810 }, { "epoch": 26.23076923076923, "grad_norm": 2.0529165267944336, "learning_rate": 2.3576310705202143e-05, "loss": 0.0409, "step": 6820 }, { "epoch": 26.26923076923077, "grad_norm": 1.351607322692871, "learning_rate": 2.3555950769242122e-05, "loss": 0.0378, "step": 6830 }, { "epoch": 26.307692307692307, "grad_norm": 1.7300976514816284, "learning_rate": 2.3535567441639396e-05, "loss": 0.0416, "step": 6840 }, { "epoch": 26.346153846153847, "grad_norm": 1.606675148010254, "learning_rate": 2.351516077812122e-05, "loss": 0.0385, "step": 6850 }, { "epoch": 26.384615384615383, "grad_norm": 2.1588430404663086, "learning_rate": 2.349473083447863e-05, "loss": 0.0479, "step": 6860 }, { "epoch": 26.423076923076923, "grad_norm": 1.7339332103729248, "learning_rate": 2.3474277666566307e-05, "loss": 0.0499, "step": 6870 }, { "epoch": 26.46153846153846, "grad_norm": 1.6982123851776123, "learning_rate": 2.345380133030243e-05, "loss": 0.0507, "step": 6880 }, { "epoch": 26.5, "grad_norm": 1.991782307624817, "learning_rate": 2.343330188166853e-05, "loss": 0.0455, "step": 6890 }, { "epoch": 26.53846153846154, "grad_norm": 1.6237735748291016, "learning_rate": 2.3412779376709304e-05, "loss": 0.043, "step": 6900 }, { "epoch": 26.576923076923077, "grad_norm": 1.8890820741653442, "learning_rate": 2.3392233871532504e-05, "loss": 0.0465, "step": 6910 }, { "epoch": 26.615384615384617, "grad_norm": 1.9966788291931152, "learning_rate": 2.337166542230876e-05, "loss": 0.0509, "step": 6920 }, { "epoch": 26.653846153846153, "grad_norm": 2.028984546661377, "learning_rate": 2.335107408527142e-05, "loss": 0.0391, "step": 6930 }, { "epoch": 26.692307692307693, "grad_norm": 1.8344041109085083, "learning_rate": 2.3330459916716417e-05, "loss": 0.0387, "step": 6940 }, { "epoch": 26.73076923076923, "grad_norm": 1.8030184507369995, "learning_rate": 2.3309822973002097e-05, "loss": 0.0476, "step": 6950 }, { "epoch": 26.76923076923077, "grad_norm": 2.0031795501708984, "learning_rate": 2.328916331054908e-05, "loss": 0.0437, "step": 6960 }, { "epoch": 26.807692307692307, "grad_norm": 1.6738231182098389, "learning_rate": 2.3268480985840093e-05, "loss": 0.0423, "step": 6970 }, { "epoch": 26.846153846153847, "grad_norm": 1.5337883234024048, "learning_rate": 2.3247776055419826e-05, "loss": 0.0417, "step": 6980 }, { "epoch": 26.884615384615383, "grad_norm": 1.6017706394195557, "learning_rate": 2.3227048575894758e-05, "loss": 0.0451, "step": 6990 }, { "epoch": 26.923076923076923, "grad_norm": 1.8597776889801025, "learning_rate": 2.3206298603933037e-05, "loss": 0.04, "step": 7000 }, { "epoch": 26.96153846153846, "grad_norm": 1.6399952173233032, "learning_rate": 2.3185526196264288e-05, "loss": 0.0434, "step": 7010 }, { "epoch": 27.0, "grad_norm": 1.3050568103790283, "learning_rate": 2.3164731409679476e-05, "loss": 0.0364, "step": 7020 }, { "epoch": 27.03846153846154, "grad_norm": 1.4773322343826294, "learning_rate": 2.3143914301030765e-05, "loss": 0.0443, "step": 7030 }, { "epoch": 27.076923076923077, "grad_norm": 1.4294204711914062, "learning_rate": 2.3123074927231332e-05, "loss": 0.0451, "step": 7040 }, { "epoch": 27.115384615384617, "grad_norm": 1.8464568853378296, "learning_rate": 2.310221334525522e-05, "loss": 0.0509, "step": 7050 }, { "epoch": 27.153846153846153, "grad_norm": 1.5100023746490479, "learning_rate": 2.3081329612137207e-05, "loss": 0.041, "step": 7060 }, { "epoch": 27.192307692307693, "grad_norm": 1.688292145729065, "learning_rate": 2.3060423784972625e-05, "loss": 0.045, "step": 7070 }, { "epoch": 27.23076923076923, "grad_norm": 1.5028406381607056, "learning_rate": 2.3039495920917193e-05, "loss": 0.0398, "step": 7080 }, { "epoch": 27.26923076923077, "grad_norm": 1.6817067861557007, "learning_rate": 2.301854607718691e-05, "loss": 0.0399, "step": 7090 }, { "epoch": 27.307692307692307, "grad_norm": 1.7924225330352783, "learning_rate": 2.299757431105783e-05, "loss": 0.0435, "step": 7100 }, { "epoch": 27.346153846153847, "grad_norm": 1.8683700561523438, "learning_rate": 2.2976580679865972e-05, "loss": 0.0438, "step": 7110 }, { "epoch": 27.384615384615383, "grad_norm": 1.6464298963546753, "learning_rate": 2.2955565241007123e-05, "loss": 0.0462, "step": 7120 }, { "epoch": 27.423076923076923, "grad_norm": 1.7487648725509644, "learning_rate": 2.293452805193669e-05, "loss": 0.0407, "step": 7130 }, { "epoch": 27.46153846153846, "grad_norm": 1.8740178346633911, "learning_rate": 2.291346917016954e-05, "loss": 0.046, "step": 7140 }, { "epoch": 27.5, "grad_norm": 1.5780128240585327, "learning_rate": 2.289238865327985e-05, "loss": 0.0451, "step": 7150 }, { "epoch": 27.53846153846154, "grad_norm": 1.7798570394515991, "learning_rate": 2.2871286558900956e-05, "loss": 0.0461, "step": 7160 }, { "epoch": 27.576923076923077, "grad_norm": 2.028824806213379, "learning_rate": 2.285016294472517e-05, "loss": 0.0448, "step": 7170 }, { "epoch": 27.615384615384617, "grad_norm": 1.669351577758789, "learning_rate": 2.2829017868503658e-05, "loss": 0.042, "step": 7180 }, { "epoch": 27.653846153846153, "grad_norm": 2.1708717346191406, "learning_rate": 2.280785138804624e-05, "loss": 0.0513, "step": 7190 }, { "epoch": 27.692307692307693, "grad_norm": 1.5223692655563354, "learning_rate": 2.2786663561221265e-05, "loss": 0.0407, "step": 7200 }, { "epoch": 27.73076923076923, "grad_norm": 1.670336365699768, "learning_rate": 2.2765454445955452e-05, "loss": 0.0369, "step": 7210 }, { "epoch": 27.76923076923077, "grad_norm": 1.4955790042877197, "learning_rate": 2.2744224100233705e-05, "loss": 0.0479, "step": 7220 }, { "epoch": 27.807692307692307, "grad_norm": 1.7686519622802734, "learning_rate": 2.2722972582098984e-05, "loss": 0.0408, "step": 7230 }, { "epoch": 27.846153846153847, "grad_norm": 2.064610004425049, "learning_rate": 2.2701699949652118e-05, "loss": 0.0415, "step": 7240 }, { "epoch": 27.884615384615383, "grad_norm": 2.065558433532715, "learning_rate": 2.2680406261051685e-05, "loss": 0.0386, "step": 7250 }, { "epoch": 27.923076923076923, "grad_norm": 1.6095936298370361, "learning_rate": 2.2659091574513805e-05, "loss": 0.0433, "step": 7260 }, { "epoch": 27.96153846153846, "grad_norm": 1.7477508783340454, "learning_rate": 2.263775594831202e-05, "loss": 0.0431, "step": 7270 }, { "epoch": 28.0, "grad_norm": 1.827488899230957, "learning_rate": 2.2616399440777128e-05, "loss": 0.0389, "step": 7280 }, { "epoch": 28.03846153846154, "grad_norm": 1.7892285585403442, "learning_rate": 2.2595022110296988e-05, "loss": 0.042, "step": 7290 }, { "epoch": 28.076923076923077, "grad_norm": 2.0402352809906006, "learning_rate": 2.2573624015316418e-05, "loss": 0.0459, "step": 7300 }, { "epoch": 28.115384615384617, "grad_norm": 1.7068613767623901, "learning_rate": 2.2552205214336986e-05, "loss": 0.0445, "step": 7310 }, { "epoch": 28.153846153846153, "grad_norm": 1.5043061971664429, "learning_rate": 2.253076576591688e-05, "loss": 0.041, "step": 7320 }, { "epoch": 28.192307692307693, "grad_norm": 1.7505626678466797, "learning_rate": 2.2509305728670733e-05, "loss": 0.0458, "step": 7330 }, { "epoch": 28.23076923076923, "grad_norm": 1.7237753868103027, "learning_rate": 2.2487825161269463e-05, "loss": 0.0417, "step": 7340 }, { "epoch": 28.26923076923077, "grad_norm": 1.7082198858261108, "learning_rate": 2.2466324122440125e-05, "loss": 0.0397, "step": 7350 }, { "epoch": 28.307692307692307, "grad_norm": 1.798220157623291, "learning_rate": 2.2444802670965732e-05, "loss": 0.0403, "step": 7360 }, { "epoch": 28.346153846153847, "grad_norm": 2.100522994995117, "learning_rate": 2.2423260865685124e-05, "loss": 0.0401, "step": 7370 }, { "epoch": 28.384615384615383, "grad_norm": 2.088228702545166, "learning_rate": 2.2401698765492762e-05, "loss": 0.0438, "step": 7380 }, { "epoch": 28.423076923076923, "grad_norm": 1.6894526481628418, "learning_rate": 2.2380116429338612e-05, "loss": 0.0445, "step": 7390 }, { "epoch": 28.46153846153846, "grad_norm": 1.8538814783096313, "learning_rate": 2.2358513916227945e-05, "loss": 0.0418, "step": 7400 }, { "epoch": 28.5, "grad_norm": 1.6178901195526123, "learning_rate": 2.233689128522122e-05, "loss": 0.0372, "step": 7410 }, { "epoch": 28.53846153846154, "grad_norm": 1.6388591527938843, "learning_rate": 2.2315248595433883e-05, "loss": 0.0455, "step": 7420 }, { "epoch": 28.576923076923077, "grad_norm": 1.7331197261810303, "learning_rate": 2.2293585906036214e-05, "loss": 0.0421, "step": 7430 }, { "epoch": 28.615384615384617, "grad_norm": 1.5612764358520508, "learning_rate": 2.2271903276253183e-05, "loss": 0.0426, "step": 7440 }, { "epoch": 28.653846153846153, "grad_norm": 1.5970007181167603, "learning_rate": 2.2250200765364273e-05, "loss": 0.0393, "step": 7450 }, { "epoch": 28.692307692307693, "grad_norm": 1.7833430767059326, "learning_rate": 2.2228478432703317e-05, "loss": 0.043, "step": 7460 }, { "epoch": 28.73076923076923, "grad_norm": 1.7582666873931885, "learning_rate": 2.2206736337658348e-05, "loss": 0.0426, "step": 7470 }, { "epoch": 28.76923076923077, "grad_norm": 1.8394339084625244, "learning_rate": 2.2184974539671417e-05, "loss": 0.0421, "step": 7480 }, { "epoch": 28.807692307692307, "grad_norm": 2.0509567260742188, "learning_rate": 2.2163193098238453e-05, "loss": 0.0468, "step": 7490 }, { "epoch": 28.846153846153847, "grad_norm": 1.9507832527160645, "learning_rate": 2.2141392072909082e-05, "loss": 0.0427, "step": 7500 }, { "epoch": 28.884615384615383, "grad_norm": 2.1092071533203125, "learning_rate": 2.2119571523286484e-05, "loss": 0.037, "step": 7510 }, { "epoch": 28.923076923076923, "grad_norm": 1.6048542261123657, "learning_rate": 2.2097731509027196e-05, "loss": 0.045, "step": 7520 }, { "epoch": 28.96153846153846, "grad_norm": 1.5821419954299927, "learning_rate": 2.207587208984099e-05, "loss": 0.0443, "step": 7530 }, { "epoch": 29.0, "grad_norm": 1.5150803327560425, "learning_rate": 2.205399332549068e-05, "loss": 0.0385, "step": 7540 }, { "epoch": 29.03846153846154, "grad_norm": 1.6221213340759277, "learning_rate": 2.2032095275791974e-05, "loss": 0.0407, "step": 7550 }, { "epoch": 29.076923076923077, "grad_norm": 1.737960934638977, "learning_rate": 2.2010178000613307e-05, "loss": 0.0407, "step": 7560 }, { "epoch": 29.115384615384617, "grad_norm": 1.7430760860443115, "learning_rate": 2.1988241559875666e-05, "loss": 0.0417, "step": 7570 }, { "epoch": 29.153846153846153, "grad_norm": 1.7095967531204224, "learning_rate": 2.1966286013552448e-05, "loss": 0.0466, "step": 7580 }, { "epoch": 29.192307692307693, "grad_norm": 1.701668381690979, "learning_rate": 2.1944311421669274e-05, "loss": 0.0432, "step": 7590 }, { "epoch": 29.23076923076923, "grad_norm": 1.6135324239730835, "learning_rate": 2.1922317844303846e-05, "loss": 0.045, "step": 7600 }, { "epoch": 29.26923076923077, "grad_norm": 1.5123496055603027, "learning_rate": 2.1900305341585756e-05, "loss": 0.037, "step": 7610 }, { "epoch": 29.307692307692307, "grad_norm": 1.4518818855285645, "learning_rate": 2.187827397369635e-05, "loss": 0.0387, "step": 7620 }, { "epoch": 29.346153846153847, "grad_norm": 1.8243811130523682, "learning_rate": 2.1856223800868542e-05, "loss": 0.0408, "step": 7630 }, { "epoch": 29.384615384615383, "grad_norm": 1.6765333414077759, "learning_rate": 2.183415488338667e-05, "loss": 0.0421, "step": 7640 }, { "epoch": 29.423076923076923, "grad_norm": 1.7374074459075928, "learning_rate": 2.1812067281586312e-05, "loss": 0.0449, "step": 7650 }, { "epoch": 29.46153846153846, "grad_norm": 1.3828516006469727, "learning_rate": 2.178996105585412e-05, "loss": 0.0416, "step": 7660 }, { "epoch": 29.5, "grad_norm": 1.41438627243042, "learning_rate": 2.1767836266627676e-05, "loss": 0.0416, "step": 7670 }, { "epoch": 29.53846153846154, "grad_norm": 1.4291772842407227, "learning_rate": 2.174569297439531e-05, "loss": 0.0381, "step": 7680 }, { "epoch": 29.576923076923077, "grad_norm": 1.4206595420837402, "learning_rate": 2.1723531239695932e-05, "loss": 0.0411, "step": 7690 }, { "epoch": 29.615384615384617, "grad_norm": 1.4999078512191772, "learning_rate": 2.1701351123118886e-05, "loss": 0.0406, "step": 7700 }, { "epoch": 29.653846153846153, "grad_norm": 1.9684511423110962, "learning_rate": 2.167915268530376e-05, "loss": 0.0438, "step": 7710 }, { "epoch": 29.692307692307693, "grad_norm": 1.7212599515914917, "learning_rate": 2.165693598694023e-05, "loss": 0.0397, "step": 7720 }, { "epoch": 29.73076923076923, "grad_norm": 1.5021146535873413, "learning_rate": 2.163470108876791e-05, "loss": 0.0478, "step": 7730 }, { "epoch": 29.76923076923077, "grad_norm": 1.5587773323059082, "learning_rate": 2.161244805157616e-05, "loss": 0.0357, "step": 7740 }, { "epoch": 29.807692307692307, "grad_norm": 1.5270205736160278, "learning_rate": 2.159017693620393e-05, "loss": 0.0415, "step": 7750 }, { "epoch": 29.846153846153847, "grad_norm": 1.5055110454559326, "learning_rate": 2.15678878035396e-05, "loss": 0.04, "step": 7760 }, { "epoch": 29.884615384615383, "grad_norm": 1.9528456926345825, "learning_rate": 2.1545580714520817e-05, "loss": 0.0351, "step": 7770 }, { "epoch": 29.923076923076923, "grad_norm": 1.9646743535995483, "learning_rate": 2.1523255730134294e-05, "loss": 0.0434, "step": 7780 }, { "epoch": 29.96153846153846, "grad_norm": 1.6051669120788574, "learning_rate": 2.15009129114157e-05, "loss": 0.0373, "step": 7790 }, { "epoch": 30.0, "grad_norm": 1.7047405242919922, "learning_rate": 2.1478552319449443e-05, "loss": 0.0408, "step": 7800 }, { "epoch": 30.03846153846154, "grad_norm": 1.537502646446228, "learning_rate": 2.1456174015368527e-05, "loss": 0.0426, "step": 7810 }, { "epoch": 30.076923076923077, "grad_norm": 1.4524402618408203, "learning_rate": 2.1433778060354375e-05, "loss": 0.0464, "step": 7820 }, { "epoch": 30.115384615384617, "grad_norm": 1.7234159708023071, "learning_rate": 2.1411364515636685e-05, "loss": 0.0512, "step": 7830 }, { "epoch": 30.153846153846153, "grad_norm": 1.512885570526123, "learning_rate": 2.1388933442493232e-05, "loss": 0.0438, "step": 7840 }, { "epoch": 30.192307692307693, "grad_norm": 1.5967854261398315, "learning_rate": 2.13664849022497e-05, "loss": 0.039, "step": 7850 }, { "epoch": 30.23076923076923, "grad_norm": 2.0385942459106445, "learning_rate": 2.1344018956279547e-05, "loss": 0.0369, "step": 7860 }, { "epoch": 30.26923076923077, "grad_norm": 1.9069523811340332, "learning_rate": 2.1321535666003817e-05, "loss": 0.0392, "step": 7870 }, { "epoch": 30.307692307692307, "grad_norm": 1.7070564031600952, "learning_rate": 2.1299035092890966e-05, "loss": 0.0447, "step": 7880 }, { "epoch": 30.346153846153847, "grad_norm": 1.4670286178588867, "learning_rate": 2.12765172984567e-05, "loss": 0.0416, "step": 7890 }, { "epoch": 30.384615384615383, "grad_norm": 1.7784276008605957, "learning_rate": 2.1253982344263803e-05, "loss": 0.0378, "step": 7900 }, { "epoch": 30.423076923076923, "grad_norm": 1.4911285638809204, "learning_rate": 2.1231430291921987e-05, "loss": 0.0459, "step": 7910 }, { "epoch": 30.46153846153846, "grad_norm": 1.4901173114776611, "learning_rate": 2.1208861203087695e-05, "loss": 0.0404, "step": 7920 }, { "epoch": 30.5, "grad_norm": 1.8902019262313843, "learning_rate": 2.1186275139463967e-05, "loss": 0.0421, "step": 7930 }, { "epoch": 30.53846153846154, "grad_norm": 1.5284932851791382, "learning_rate": 2.1163672162800222e-05, "loss": 0.0405, "step": 7940 }, { "epoch": 30.576923076923077, "grad_norm": 1.6077547073364258, "learning_rate": 2.114105233489215e-05, "loss": 0.0505, "step": 7950 }, { "epoch": 30.615384615384617, "grad_norm": 1.7635397911071777, "learning_rate": 2.1118415717581487e-05, "loss": 0.0509, "step": 7960 }, { "epoch": 30.653846153846153, "grad_norm": 1.7803516387939453, "learning_rate": 2.1095762372755885e-05, "loss": 0.0428, "step": 7970 }, { "epoch": 30.692307692307693, "grad_norm": 1.7161511182785034, "learning_rate": 2.1073092362348716e-05, "loss": 0.0496, "step": 7980 }, { "epoch": 30.73076923076923, "grad_norm": 1.8986924886703491, "learning_rate": 2.1050405748338933e-05, "loss": 0.0425, "step": 7990 }, { "epoch": 30.76923076923077, "grad_norm": 1.8261290788650513, "learning_rate": 2.102770259275087e-05, "loss": 0.0395, "step": 8000 }, { "epoch": 30.807692307692307, "grad_norm": 1.6952645778656006, "learning_rate": 2.100498295765408e-05, "loss": 0.0454, "step": 8010 }, { "epoch": 30.846153846153847, "grad_norm": 1.8140101432800293, "learning_rate": 2.098224690516319e-05, "loss": 0.0411, "step": 8020 }, { "epoch": 30.884615384615383, "grad_norm": 1.8083970546722412, "learning_rate": 2.0959494497437688e-05, "loss": 0.0427, "step": 8030 }, { "epoch": 30.923076923076923, "grad_norm": 1.270798921585083, "learning_rate": 2.0936725796681796e-05, "loss": 0.0379, "step": 8040 }, { "epoch": 30.96153846153846, "grad_norm": 1.6783995628356934, "learning_rate": 2.0913940865144266e-05, "loss": 0.0372, "step": 8050 }, { "epoch": 31.0, "grad_norm": 1.6331335306167603, "learning_rate": 2.0891139765118235e-05, "loss": 0.0415, "step": 8060 }, { "epoch": 31.03846153846154, "grad_norm": 1.3328851461410522, "learning_rate": 2.086832255894104e-05, "loss": 0.0386, "step": 8070 }, { "epoch": 31.076923076923077, "grad_norm": 1.5463865995407104, "learning_rate": 2.084548930899405e-05, "loss": 0.0392, "step": 8080 }, { "epoch": 31.115384615384617, "grad_norm": 1.4242238998413086, "learning_rate": 2.08226400777025e-05, "loss": 0.0371, "step": 8090 }, { "epoch": 31.153846153846153, "grad_norm": 1.5513811111450195, "learning_rate": 2.0799774927535313e-05, "loss": 0.0368, "step": 8100 }, { "epoch": 31.192307692307693, "grad_norm": 1.7626824378967285, "learning_rate": 2.0776893921004936e-05, "loss": 0.0394, "step": 8110 }, { "epoch": 31.23076923076923, "grad_norm": 1.5672152042388916, "learning_rate": 2.0753997120667172e-05, "loss": 0.0431, "step": 8120 }, { "epoch": 31.26923076923077, "grad_norm": 1.6233484745025635, "learning_rate": 2.0731084589120995e-05, "loss": 0.0358, "step": 8130 }, { "epoch": 31.307692307692307, "grad_norm": 1.7482249736785889, "learning_rate": 2.070815638900839e-05, "loss": 0.0446, "step": 8140 }, { "epoch": 31.346153846153847, "grad_norm": 1.4374932050704956, "learning_rate": 2.0685212583014186e-05, "loss": 0.0407, "step": 8150 }, { "epoch": 31.384615384615383, "grad_norm": 1.583405613899231, "learning_rate": 2.0662253233865866e-05, "loss": 0.0395, "step": 8160 }, { "epoch": 31.423076923076923, "grad_norm": 1.5504549741744995, "learning_rate": 2.063927840433342e-05, "loss": 0.0396, "step": 8170 }, { "epoch": 31.46153846153846, "grad_norm": 1.6861332654953003, "learning_rate": 2.0616288157229154e-05, "loss": 0.0427, "step": 8180 }, { "epoch": 31.5, "grad_norm": 1.3032209873199463, "learning_rate": 2.0593282555407522e-05, "loss": 0.0409, "step": 8190 }, { "epoch": 31.53846153846154, "grad_norm": 1.734984040260315, "learning_rate": 2.057026166176496e-05, "loss": 0.0405, "step": 8200 }, { "epoch": 31.576923076923077, "grad_norm": 1.5121088027954102, "learning_rate": 2.0547225539239715e-05, "loss": 0.04, "step": 8210 }, { "epoch": 31.615384615384617, "grad_norm": 1.6952282190322876, "learning_rate": 2.0524174250811665e-05, "loss": 0.0442, "step": 8220 }, { "epoch": 31.653846153846153, "grad_norm": 1.8165638446807861, "learning_rate": 2.050110785950216e-05, "loss": 0.0415, "step": 8230 }, { "epoch": 31.692307692307693, "grad_norm": 1.6110750436782837, "learning_rate": 2.047802642837382e-05, "loss": 0.0377, "step": 8240 }, { "epoch": 31.73076923076923, "grad_norm": 1.6655876636505127, "learning_rate": 2.0454930020530403e-05, "loss": 0.0385, "step": 8250 }, { "epoch": 31.76923076923077, "grad_norm": 1.491156816482544, "learning_rate": 2.0431818699116606e-05, "loss": 0.0375, "step": 8260 }, { "epoch": 31.807692307692307, "grad_norm": 1.4836500883102417, "learning_rate": 2.04086925273179e-05, "loss": 0.0521, "step": 8270 }, { "epoch": 31.846153846153847, "grad_norm": 1.317879319190979, "learning_rate": 2.0385551568360357e-05, "loss": 0.0393, "step": 8280 }, { "epoch": 31.884615384615383, "grad_norm": 1.4740430116653442, "learning_rate": 2.036239588551047e-05, "loss": 0.0401, "step": 8290 }, { "epoch": 31.923076923076923, "grad_norm": 1.7209266424179077, "learning_rate": 2.0339225542074996e-05, "loss": 0.0352, "step": 8300 }, { "epoch": 31.96153846153846, "grad_norm": 1.6838961839675903, "learning_rate": 2.0316040601400765e-05, "loss": 0.0433, "step": 8310 }, { "epoch": 32.0, "grad_norm": 1.6667869091033936, "learning_rate": 2.029284112687453e-05, "loss": 0.0417, "step": 8320 }, { "epoch": 32.03846153846154, "grad_norm": 1.8445954322814941, "learning_rate": 2.0269627181922752e-05, "loss": 0.0414, "step": 8330 }, { "epoch": 32.07692307692308, "grad_norm": 1.4822769165039062, "learning_rate": 2.0246398830011482e-05, "loss": 0.0382, "step": 8340 }, { "epoch": 32.11538461538461, "grad_norm": 1.6866445541381836, "learning_rate": 2.0223156134646142e-05, "loss": 0.042, "step": 8350 }, { "epoch": 32.15384615384615, "grad_norm": 1.7119044065475464, "learning_rate": 2.019989915937138e-05, "loss": 0.0403, "step": 8360 }, { "epoch": 32.19230769230769, "grad_norm": 2.0892767906188965, "learning_rate": 2.0176627967770873e-05, "loss": 0.035, "step": 8370 }, { "epoch": 32.23076923076923, "grad_norm": 1.5972812175750732, "learning_rate": 2.015334262346717e-05, "loss": 0.0409, "step": 8380 }, { "epoch": 32.26923076923077, "grad_norm": 1.5595967769622803, "learning_rate": 2.0130043190121515e-05, "loss": 0.04, "step": 8390 }, { "epoch": 32.30769230769231, "grad_norm": 1.349557876586914, "learning_rate": 2.0106729731433663e-05, "loss": 0.0337, "step": 8400 }, { "epoch": 32.34615384615385, "grad_norm": 1.450696587562561, "learning_rate": 2.008340231114173e-05, "loss": 0.0388, "step": 8410 }, { "epoch": 32.38461538461539, "grad_norm": 1.7912949323654175, "learning_rate": 2.006006099302199e-05, "loss": 0.0378, "step": 8420 }, { "epoch": 32.42307692307692, "grad_norm": 1.6425917148590088, "learning_rate": 2.003670584088871e-05, "loss": 0.0384, "step": 8430 }, { "epoch": 32.46153846153846, "grad_norm": 1.5332478284835815, "learning_rate": 2.001333691859399e-05, "loss": 0.0348, "step": 8440 }, { "epoch": 32.5, "grad_norm": 1.5672633647918701, "learning_rate": 1.9989954290027565e-05, "loss": 0.0379, "step": 8450 }, { "epoch": 32.53846153846154, "grad_norm": 1.715303897857666, "learning_rate": 1.9966558019116654e-05, "loss": 0.0344, "step": 8460 }, { "epoch": 32.57692307692308, "grad_norm": 1.3653415441513062, "learning_rate": 1.9943148169825766e-05, "loss": 0.0391, "step": 8470 }, { "epoch": 32.61538461538461, "grad_norm": 1.5256359577178955, "learning_rate": 1.991972480615653e-05, "loss": 0.046, "step": 8480 }, { "epoch": 32.65384615384615, "grad_norm": 1.432858943939209, "learning_rate": 1.989628799214754e-05, "loss": 0.0408, "step": 8490 }, { "epoch": 32.69230769230769, "grad_norm": 1.5813000202178955, "learning_rate": 1.987283779187414e-05, "loss": 0.0359, "step": 8500 }, { "epoch": 32.73076923076923, "grad_norm": 1.5625938177108765, "learning_rate": 1.9849374269448288e-05, "loss": 0.0407, "step": 8510 }, { "epoch": 32.76923076923077, "grad_norm": 1.6737438440322876, "learning_rate": 1.982589748901836e-05, "loss": 0.0387, "step": 8520 }, { "epoch": 32.80769230769231, "grad_norm": 1.765540361404419, "learning_rate": 1.9802407514768964e-05, "loss": 0.0369, "step": 8530 }, { "epoch": 32.84615384615385, "grad_norm": 1.6643508672714233, "learning_rate": 1.9778904410920808e-05, "loss": 0.036, "step": 8540 }, { "epoch": 32.88461538461539, "grad_norm": 1.470862865447998, "learning_rate": 1.9755388241730475e-05, "loss": 0.0414, "step": 8550 }, { "epoch": 32.92307692307692, "grad_norm": 1.748119831085205, "learning_rate": 1.973185907149027e-05, "loss": 0.0362, "step": 8560 }, { "epoch": 32.96153846153846, "grad_norm": 1.780103325843811, "learning_rate": 1.970831696452805e-05, "loss": 0.0421, "step": 8570 }, { "epoch": 33.0, "grad_norm": 1.5416651964187622, "learning_rate": 1.9684761985207038e-05, "loss": 0.0369, "step": 8580 }, { "epoch": 33.03846153846154, "grad_norm": 1.6990591287612915, "learning_rate": 1.9661194197925644e-05, "loss": 0.0356, "step": 8590 }, { "epoch": 33.07692307692308, "grad_norm": 1.5676331520080566, "learning_rate": 1.9637613667117303e-05, "loss": 0.0437, "step": 8600 }, { "epoch": 33.11538461538461, "grad_norm": 1.9157164096832275, "learning_rate": 1.961402045725028e-05, "loss": 0.0408, "step": 8610 }, { "epoch": 33.15384615384615, "grad_norm": 1.409117341041565, "learning_rate": 1.9590414632827513e-05, "loss": 0.0414, "step": 8620 }, { "epoch": 33.19230769230769, "grad_norm": 1.700792670249939, "learning_rate": 1.9566796258386424e-05, "loss": 0.037, "step": 8630 }, { "epoch": 33.23076923076923, "grad_norm": 1.9453880786895752, "learning_rate": 1.9543165398498743e-05, "loss": 0.0385, "step": 8640 }, { "epoch": 33.26923076923077, "grad_norm": 1.4440393447875977, "learning_rate": 1.9519522117770355e-05, "loss": 0.0409, "step": 8650 }, { "epoch": 33.30769230769231, "grad_norm": 1.8256205320358276, "learning_rate": 1.9495866480841063e-05, "loss": 0.038, "step": 8660 }, { "epoch": 33.34615384615385, "grad_norm": 2.067312479019165, "learning_rate": 1.9472198552384494e-05, "loss": 0.0378, "step": 8670 }, { "epoch": 33.38461538461539, "grad_norm": 1.8080998659133911, "learning_rate": 1.9448518397107848e-05, "loss": 0.0446, "step": 8680 }, { "epoch": 33.42307692307692, "grad_norm": 1.553263545036316, "learning_rate": 1.942482607975177e-05, "loss": 0.0385, "step": 8690 }, { "epoch": 33.46153846153846, "grad_norm": 1.4473556280136108, "learning_rate": 1.940112166509016e-05, "loss": 0.0382, "step": 8700 }, { "epoch": 33.5, "grad_norm": 1.6522272825241089, "learning_rate": 1.937740521792996e-05, "loss": 0.0339, "step": 8710 }, { "epoch": 33.53846153846154, "grad_norm": 1.5314304828643799, "learning_rate": 1.935367680311106e-05, "loss": 0.0381, "step": 8720 }, { "epoch": 33.57692307692308, "grad_norm": 1.5628886222839355, "learning_rate": 1.9329936485506012e-05, "loss": 0.0403, "step": 8730 }, { "epoch": 33.61538461538461, "grad_norm": 1.291195034980774, "learning_rate": 1.930618433001996e-05, "loss": 0.0386, "step": 8740 }, { "epoch": 33.65384615384615, "grad_norm": 1.1823079586029053, "learning_rate": 1.9282420401590377e-05, "loss": 0.0402, "step": 8750 }, { "epoch": 33.69230769230769, "grad_norm": 1.515400767326355, "learning_rate": 1.925864476518694e-05, "loss": 0.0431, "step": 8760 }, { "epoch": 33.73076923076923, "grad_norm": 1.39093017578125, "learning_rate": 1.9234857485811336e-05, "loss": 0.0366, "step": 8770 }, { "epoch": 33.76923076923077, "grad_norm": 1.5924241542816162, "learning_rate": 1.9211058628497066e-05, "loss": 0.0412, "step": 8780 }, { "epoch": 33.80769230769231, "grad_norm": 1.3143709897994995, "learning_rate": 1.918724825830931e-05, "loss": 0.0447, "step": 8790 }, { "epoch": 33.84615384615385, "grad_norm": 1.6090748310089111, "learning_rate": 1.9163426440344702e-05, "loss": 0.0398, "step": 8800 }, { "epoch": 33.88461538461539, "grad_norm": 1.4839627742767334, "learning_rate": 1.913959323973119e-05, "loss": 0.0405, "step": 8810 }, { "epoch": 33.92307692307692, "grad_norm": 1.691476583480835, "learning_rate": 1.9115748721627827e-05, "loss": 0.041, "step": 8820 }, { "epoch": 33.96153846153846, "grad_norm": 1.4554405212402344, "learning_rate": 1.9091892951224614e-05, "loss": 0.0394, "step": 8830 }, { "epoch": 34.0, "grad_norm": 1.4189049005508423, "learning_rate": 1.906802599374233e-05, "loss": 0.0357, "step": 8840 }, { "epoch": 34.03846153846154, "grad_norm": 1.9339512586593628, "learning_rate": 1.904414791443231e-05, "loss": 0.0432, "step": 8850 }, { "epoch": 34.07692307692308, "grad_norm": 1.4830939769744873, "learning_rate": 1.9020258778576324e-05, "loss": 0.042, "step": 8860 }, { "epoch": 34.11538461538461, "grad_norm": 1.5076172351837158, "learning_rate": 1.8996358651486347e-05, "loss": 0.0365, "step": 8870 }, { "epoch": 34.15384615384615, "grad_norm": 1.5338919162750244, "learning_rate": 1.8972447598504417e-05, "loss": 0.0406, "step": 8880 }, { "epoch": 34.19230769230769, "grad_norm": 1.9823853969573975, "learning_rate": 1.8948525685002438e-05, "loss": 0.0349, "step": 8890 }, { "epoch": 34.23076923076923, "grad_norm": 1.429854393005371, "learning_rate": 1.892459297638201e-05, "loss": 0.0427, "step": 8900 }, { "epoch": 34.26923076923077, "grad_norm": 1.8108021020889282, "learning_rate": 1.890064953807425e-05, "loss": 0.0341, "step": 8910 }, { "epoch": 34.30769230769231, "grad_norm": 1.07712721824646, "learning_rate": 1.8876695435539596e-05, "loss": 0.0371, "step": 8920 }, { "epoch": 34.34615384615385, "grad_norm": 1.3517080545425415, "learning_rate": 1.8852730734267653e-05, "loss": 0.0348, "step": 8930 }, { "epoch": 34.38461538461539, "grad_norm": 1.369992733001709, "learning_rate": 1.8828755499776997e-05, "loss": 0.0379, "step": 8940 }, { "epoch": 34.42307692307692, "grad_norm": 1.4494092464447021, "learning_rate": 1.8804769797615007e-05, "loss": 0.0386, "step": 8950 }, { "epoch": 34.46153846153846, "grad_norm": 1.806056022644043, "learning_rate": 1.8780773693357675e-05, "loss": 0.0416, "step": 8960 }, { "epoch": 34.5, "grad_norm": 1.4004135131835938, "learning_rate": 1.8756767252609433e-05, "loss": 0.034, "step": 8970 }, { "epoch": 34.53846153846154, "grad_norm": 1.3696749210357666, "learning_rate": 1.8732750541002974e-05, "loss": 0.0393, "step": 8980 }, { "epoch": 34.57692307692308, "grad_norm": 1.2756446599960327, "learning_rate": 1.870872362419907e-05, "loss": 0.0398, "step": 8990 }, { "epoch": 34.61538461538461, "grad_norm": 1.3034799098968506, "learning_rate": 1.8684686567886398e-05, "loss": 0.0435, "step": 9000 }, { "epoch": 34.65384615384615, "grad_norm": 1.5392398834228516, "learning_rate": 1.8660639437781344e-05, "loss": 0.0349, "step": 9010 }, { "epoch": 34.69230769230769, "grad_norm": 1.5113469362258911, "learning_rate": 1.8636582299627854e-05, "loss": 0.0359, "step": 9020 }, { "epoch": 34.73076923076923, "grad_norm": 1.1920346021652222, "learning_rate": 1.8612515219197215e-05, "loss": 0.0361, "step": 9030 }, { "epoch": 34.76923076923077, "grad_norm": 1.304843783378601, "learning_rate": 1.858843826228791e-05, "loss": 0.0357, "step": 9040 }, { "epoch": 34.80769230769231, "grad_norm": 1.7710132598876953, "learning_rate": 1.8564351494725423e-05, "loss": 0.0373, "step": 9050 }, { "epoch": 34.84615384615385, "grad_norm": 1.4856317043304443, "learning_rate": 1.8540254982362053e-05, "loss": 0.0371, "step": 9060 }, { "epoch": 34.88461538461539, "grad_norm": 1.8330937623977661, "learning_rate": 1.8516148791076743e-05, "loss": 0.0389, "step": 9070 }, { "epoch": 34.92307692307692, "grad_norm": 1.504800796508789, "learning_rate": 1.8492032986774904e-05, "loss": 0.0359, "step": 9080 }, { "epoch": 34.96153846153846, "grad_norm": 1.2242687940597534, "learning_rate": 1.8467907635388225e-05, "loss": 0.0371, "step": 9090 }, { "epoch": 35.0, "grad_norm": 1.3216480016708374, "learning_rate": 1.844377280287449e-05, "loss": 0.0386, "step": 9100 }, { "epoch": 35.03846153846154, "grad_norm": 1.0509378910064697, "learning_rate": 1.8419628555217407e-05, "loss": 0.0405, "step": 9110 }, { "epoch": 35.07692307692308, "grad_norm": 1.1122572422027588, "learning_rate": 1.839547495842644e-05, "loss": 0.039, "step": 9120 }, { "epoch": 35.11538461538461, "grad_norm": 1.334999680519104, "learning_rate": 1.8371312078536587e-05, "loss": 0.0316, "step": 9130 }, { "epoch": 35.15384615384615, "grad_norm": 1.2729421854019165, "learning_rate": 1.834713998160825e-05, "loss": 0.039, "step": 9140 }, { "epoch": 35.19230769230769, "grad_norm": 1.4823824167251587, "learning_rate": 1.832295873372701e-05, "loss": 0.0428, "step": 9150 }, { "epoch": 35.23076923076923, "grad_norm": 1.5749338865280151, "learning_rate": 1.8298768401003477e-05, "loss": 0.0386, "step": 9160 }, { "epoch": 35.26923076923077, "grad_norm": 1.4943759441375732, "learning_rate": 1.8274569049573103e-05, "loss": 0.0409, "step": 9170 }, { "epoch": 35.30769230769231, "grad_norm": 1.4229377508163452, "learning_rate": 1.8250360745595983e-05, "loss": 0.0398, "step": 9180 }, { "epoch": 35.34615384615385, "grad_norm": 1.192239761352539, "learning_rate": 1.8226143555256703e-05, "loss": 0.0348, "step": 9190 }, { "epoch": 35.38461538461539, "grad_norm": 1.7158582210540771, "learning_rate": 1.820191754476413e-05, "loss": 0.04, "step": 9200 }, { "epoch": 35.42307692307692, "grad_norm": 1.2287282943725586, "learning_rate": 1.8177682780351256e-05, "loss": 0.0381, "step": 9210 }, { "epoch": 35.46153846153846, "grad_norm": 1.578681468963623, "learning_rate": 1.8153439328275e-05, "loss": 0.0434, "step": 9220 }, { "epoch": 35.5, "grad_norm": 1.7834547758102417, "learning_rate": 1.8129187254816035e-05, "loss": 0.0393, "step": 9230 }, { "epoch": 35.53846153846154, "grad_norm": 1.6021229028701782, "learning_rate": 1.81049266262786e-05, "loss": 0.0372, "step": 9240 }, { "epoch": 35.57692307692308, "grad_norm": 1.3831838369369507, "learning_rate": 1.808065750899033e-05, "loss": 0.0409, "step": 9250 }, { "epoch": 35.61538461538461, "grad_norm": 1.727046012878418, "learning_rate": 1.8056379969302066e-05, "loss": 0.0377, "step": 9260 }, { "epoch": 35.65384615384615, "grad_norm": 1.710699439048767, "learning_rate": 1.8032094073587675e-05, "loss": 0.0376, "step": 9270 }, { "epoch": 35.69230769230769, "grad_norm": 1.6480785608291626, "learning_rate": 1.800779988824387e-05, "loss": 0.0343, "step": 9280 }, { "epoch": 35.73076923076923, "grad_norm": 1.4451653957366943, "learning_rate": 1.7983497479690018e-05, "loss": 0.0347, "step": 9290 }, { "epoch": 35.76923076923077, "grad_norm": 1.520155668258667, "learning_rate": 1.795918691436798e-05, "loss": 0.039, "step": 9300 }, { "epoch": 35.80769230769231, "grad_norm": 1.372151494026184, "learning_rate": 1.7934868258741917e-05, "loss": 0.0366, "step": 9310 }, { "epoch": 35.84615384615385, "grad_norm": 1.39862859249115, "learning_rate": 1.79105415792981e-05, "loss": 0.0377, "step": 9320 }, { "epoch": 35.88461538461539, "grad_norm": 1.4081544876098633, "learning_rate": 1.788620694254475e-05, "loss": 0.0412, "step": 9330 }, { "epoch": 35.92307692307692, "grad_norm": 1.39034903049469, "learning_rate": 1.7861864415011827e-05, "loss": 0.0411, "step": 9340 }, { "epoch": 35.96153846153846, "grad_norm": 1.650275468826294, "learning_rate": 1.783751406325087e-05, "loss": 0.0336, "step": 9350 }, { "epoch": 36.0, "grad_norm": 1.45979905128479, "learning_rate": 1.7813155953834814e-05, "loss": 0.0402, "step": 9360 }, { "epoch": 36.03846153846154, "grad_norm": 1.596880555152893, "learning_rate": 1.7788790153357803e-05, "loss": 0.0439, "step": 9370 }, { "epoch": 36.07692307692308, "grad_norm": 1.4605010747909546, "learning_rate": 1.7764416728435e-05, "loss": 0.0429, "step": 9380 }, { "epoch": 36.11538461538461, "grad_norm": 1.4628403186798096, "learning_rate": 1.774003574570242e-05, "loss": 0.0416, "step": 9390 }, { "epoch": 36.15384615384615, "grad_norm": 1.8021140098571777, "learning_rate": 1.7715647271816744e-05, "loss": 0.0346, "step": 9400 }, { "epoch": 36.19230769230769, "grad_norm": 1.4157359600067139, "learning_rate": 1.769125137345512e-05, "loss": 0.0357, "step": 9410 }, { "epoch": 36.23076923076923, "grad_norm": 1.421094536781311, "learning_rate": 1.7666848117315008e-05, "loss": 0.0392, "step": 9420 }, { "epoch": 36.26923076923077, "grad_norm": 1.7203149795532227, "learning_rate": 1.7642437570113974e-05, "loss": 0.0358, "step": 9430 }, { "epoch": 36.30769230769231, "grad_norm": 1.436547040939331, "learning_rate": 1.7618019798589525e-05, "loss": 0.043, "step": 9440 }, { "epoch": 36.34615384615385, "grad_norm": 1.5459340810775757, "learning_rate": 1.7593594869498915e-05, "loss": 0.035, "step": 9450 }, { "epoch": 36.38461538461539, "grad_norm": 1.5073521137237549, "learning_rate": 1.7569162849618966e-05, "loss": 0.0393, "step": 9460 }, { "epoch": 36.42307692307692, "grad_norm": 1.3095002174377441, "learning_rate": 1.75447238057459e-05, "loss": 0.0383, "step": 9470 }, { "epoch": 36.46153846153846, "grad_norm": 1.308945655822754, "learning_rate": 1.752027780469511e-05, "loss": 0.0364, "step": 9480 }, { "epoch": 36.5, "grad_norm": 1.4354068040847778, "learning_rate": 1.7495824913301043e-05, "loss": 0.0417, "step": 9490 }, { "epoch": 36.53846153846154, "grad_norm": 1.756820559501648, "learning_rate": 1.7471365198416957e-05, "loss": 0.0393, "step": 9500 }, { "epoch": 36.57692307692308, "grad_norm": 1.3326730728149414, "learning_rate": 1.7446898726914797e-05, "loss": 0.0369, "step": 9510 }, { "epoch": 36.61538461538461, "grad_norm": 1.3120198249816895, "learning_rate": 1.742242556568495e-05, "loss": 0.0426, "step": 9520 }, { "epoch": 36.65384615384615, "grad_norm": 1.121603012084961, "learning_rate": 1.73979457816361e-05, "loss": 0.0387, "step": 9530 }, { "epoch": 36.69230769230769, "grad_norm": 1.3100221157073975, "learning_rate": 1.7373459441695058e-05, "loss": 0.0423, "step": 9540 }, { "epoch": 36.73076923076923, "grad_norm": 1.3542190790176392, "learning_rate": 1.7348966612806524e-05, "loss": 0.0421, "step": 9550 }, { "epoch": 36.76923076923077, "grad_norm": 1.59927499294281, "learning_rate": 1.7324467361932973e-05, "loss": 0.0397, "step": 9560 }, { "epoch": 36.80769230769231, "grad_norm": 1.6307555437088013, "learning_rate": 1.729996175605441e-05, "loss": 0.0411, "step": 9570 }, { "epoch": 36.84615384615385, "grad_norm": 1.3856720924377441, "learning_rate": 1.7275449862168235e-05, "loss": 0.0401, "step": 9580 }, { "epoch": 36.88461538461539, "grad_norm": 1.2768700122833252, "learning_rate": 1.725093174728902e-05, "loss": 0.0324, "step": 9590 }, { "epoch": 36.92307692307692, "grad_norm": 1.9689397811889648, "learning_rate": 1.7226407478448357e-05, "loss": 0.038, "step": 9600 }, { "epoch": 36.96153846153846, "grad_norm": 1.4385662078857422, "learning_rate": 1.7201877122694666e-05, "loss": 0.0423, "step": 9610 }, { "epoch": 37.0, "grad_norm": 1.3675845861434937, "learning_rate": 1.7177340747093e-05, "loss": 0.0397, "step": 9620 }, { "epoch": 37.03846153846154, "grad_norm": 1.733004093170166, "learning_rate": 1.7152798418724873e-05, "loss": 0.0417, "step": 9630 }, { "epoch": 37.07692307692308, "grad_norm": 1.5616117715835571, "learning_rate": 1.712825020468807e-05, "loss": 0.0359, "step": 9640 }, { "epoch": 37.11538461538461, "grad_norm": 1.513261079788208, "learning_rate": 1.710369617209648e-05, "loss": 0.0414, "step": 9650 }, { "epoch": 37.15384615384615, "grad_norm": 1.6468145847320557, "learning_rate": 1.7079136388079884e-05, "loss": 0.038, "step": 9660 }, { "epoch": 37.19230769230769, "grad_norm": 1.6188808679580688, "learning_rate": 1.7054570919783796e-05, "loss": 0.0382, "step": 9670 }, { "epoch": 37.23076923076923, "grad_norm": 1.556585669517517, "learning_rate": 1.7029999834369264e-05, "loss": 0.0387, "step": 9680 }, { "epoch": 37.26923076923077, "grad_norm": 1.5919781923294067, "learning_rate": 1.7005423199012696e-05, "loss": 0.034, "step": 9690 }, { "epoch": 37.30769230769231, "grad_norm": 1.565858244895935, "learning_rate": 1.6980841080905687e-05, "loss": 0.0344, "step": 9700 }, { "epoch": 37.34615384615385, "grad_norm": 1.190491795539856, "learning_rate": 1.6956253547254798e-05, "loss": 0.0405, "step": 9710 }, { "epoch": 37.38461538461539, "grad_norm": 1.2053945064544678, "learning_rate": 1.693166066528141e-05, "loss": 0.0366, "step": 9720 }, { "epoch": 37.42307692307692, "grad_norm": 1.7797147035598755, "learning_rate": 1.690706250222152e-05, "loss": 0.0404, "step": 9730 }, { "epoch": 37.46153846153846, "grad_norm": 1.3981484174728394, "learning_rate": 1.6882459125325573e-05, "loss": 0.0372, "step": 9740 }, { "epoch": 37.5, "grad_norm": 1.3930727243423462, "learning_rate": 1.685785060185826e-05, "loss": 0.036, "step": 9750 }, { "epoch": 37.53846153846154, "grad_norm": 1.2622334957122803, "learning_rate": 1.683323699909834e-05, "loss": 0.036, "step": 9760 }, { "epoch": 37.57692307692308, "grad_norm": 1.3782292604446411, "learning_rate": 1.6808618384338472e-05, "loss": 0.0375, "step": 9770 }, { "epoch": 37.61538461538461, "grad_norm": 1.3075623512268066, "learning_rate": 1.6783994824885e-05, "loss": 0.0357, "step": 9780 }, { "epoch": 37.65384615384615, "grad_norm": 1.449067234992981, "learning_rate": 1.6759366388057795e-05, "loss": 0.04, "step": 9790 }, { "epoch": 37.69230769230769, "grad_norm": 1.53331458568573, "learning_rate": 1.6734733141190073e-05, "loss": 0.0322, "step": 9800 }, { "epoch": 37.73076923076923, "grad_norm": 1.3725186586380005, "learning_rate": 1.6710095151628182e-05, "loss": 0.0394, "step": 9810 }, { "epoch": 37.76923076923077, "grad_norm": 1.6133525371551514, "learning_rate": 1.668545248673144e-05, "loss": 0.0329, "step": 9820 }, { "epoch": 37.80769230769231, "grad_norm": 1.2607682943344116, "learning_rate": 1.6660805213871962e-05, "loss": 0.0432, "step": 9830 }, { "epoch": 37.84615384615385, "grad_norm": 1.4924263954162598, "learning_rate": 1.663615340043445e-05, "loss": 0.0392, "step": 9840 }, { "epoch": 37.88461538461539, "grad_norm": 1.3537297248840332, "learning_rate": 1.6611497113816014e-05, "loss": 0.0426, "step": 9850 }, { "epoch": 37.92307692307692, "grad_norm": 1.1831531524658203, "learning_rate": 1.6586836421426007e-05, "loss": 0.0383, "step": 9860 }, { "epoch": 37.96153846153846, "grad_norm": 1.317609429359436, "learning_rate": 1.6562171390685815e-05, "loss": 0.033, "step": 9870 }, { "epoch": 38.0, "grad_norm": 1.4860436916351318, "learning_rate": 1.653750208902869e-05, "loss": 0.0377, "step": 9880 }, { "epoch": 38.03846153846154, "grad_norm": 1.7705858945846558, "learning_rate": 1.6512828583899562e-05, "loss": 0.0428, "step": 9890 }, { "epoch": 38.07692307692308, "grad_norm": 1.4256800413131714, "learning_rate": 1.648815094275486e-05, "loss": 0.0374, "step": 9900 }, { "epoch": 38.11538461538461, "grad_norm": 1.430487871170044, "learning_rate": 1.6463469233062302e-05, "loss": 0.0382, "step": 9910 }, { "epoch": 38.15384615384615, "grad_norm": 1.273598074913025, "learning_rate": 1.6438783522300742e-05, "loss": 0.0403, "step": 9920 }, { "epoch": 38.19230769230769, "grad_norm": 1.6827222108840942, "learning_rate": 1.641409387795997e-05, "loss": 0.0318, "step": 9930 }, { "epoch": 38.23076923076923, "grad_norm": 1.5564827919006348, "learning_rate": 1.6389400367540534e-05, "loss": 0.0443, "step": 9940 }, { "epoch": 38.26923076923077, "grad_norm": 1.5248053073883057, "learning_rate": 1.6364703058553552e-05, "loss": 0.0408, "step": 9950 }, { "epoch": 38.30769230769231, "grad_norm": 1.2614673376083374, "learning_rate": 1.6340002018520512e-05, "loss": 0.0372, "step": 9960 }, { "epoch": 38.34615384615385, "grad_norm": 1.43790602684021, "learning_rate": 1.6315297314973126e-05, "loss": 0.0354, "step": 9970 }, { "epoch": 38.38461538461539, "grad_norm": 1.3772172927856445, "learning_rate": 1.6290589015453102e-05, "loss": 0.0362, "step": 9980 }, { "epoch": 38.42307692307692, "grad_norm": 1.5918662548065186, "learning_rate": 1.6265877187511993e-05, "loss": 0.0378, "step": 9990 }, { "epoch": 38.46153846153846, "grad_norm": 1.3942394256591797, "learning_rate": 1.6241161898710993e-05, "loss": 0.0408, "step": 10000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 77, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }