| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.140210886292391, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005699629524080934, | |
| "grad_norm": 19.25, | |
| "learning_rate": 9.982896237172178e-06, | |
| "loss": 2.0698, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011399259048161869, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 9.963892056252377e-06, | |
| "loss": 1.5026, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.017098888572242805, | |
| "grad_norm": 6.5, | |
| "learning_rate": 9.944887875332574e-06, | |
| "loss": 1.4607, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.022798518096323737, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 9.925883694412771e-06, | |
| "loss": 1.3855, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.028498147620404674, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 9.90687951349297e-06, | |
| "loss": 1.4154, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03419777714448561, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 9.887875332573167e-06, | |
| "loss": 1.3327, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.039897406668566546, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 9.868871151653364e-06, | |
| "loss": 1.3797, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.045597036192647475, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 9.849866970733563e-06, | |
| "loss": 1.361, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05129666571672841, | |
| "grad_norm": 7.0, | |
| "learning_rate": 9.83086278981376e-06, | |
| "loss": 1.4097, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05699629524080935, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 9.811858608893958e-06, | |
| "loss": 1.3784, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06269592476489028, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 9.792854427974155e-06, | |
| "loss": 1.3886, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06839555428897122, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 9.773850247054353e-06, | |
| "loss": 1.2982, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07409518381305215, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 9.75484606613455e-06, | |
| "loss": 1.3887, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07979481333713309, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 9.735841885214748e-06, | |
| "loss": 1.3336, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08549444286121402, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 9.716837704294946e-06, | |
| "loss": 1.3523, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09119407238529495, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 9.697833523375144e-06, | |
| "loss": 1.3483, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09689370190937589, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 9.678829342455342e-06, | |
| "loss": 1.3502, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.10259333143345682, | |
| "grad_norm": 7.0, | |
| "learning_rate": 9.65982516153554e-06, | |
| "loss": 1.3553, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.10829296095753776, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 9.640820980615736e-06, | |
| "loss": 1.338, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1139925904816187, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 9.621816799695934e-06, | |
| "loss": 1.3868, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11969222000569962, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 9.602812618776132e-06, | |
| "loss": 1.3996, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.12539184952978055, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 9.58380843785633e-06, | |
| "loss": 1.3638, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1310914790538615, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 9.564804256936528e-06, | |
| "loss": 1.2893, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13679110857794244, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 9.545800076016724e-06, | |
| "loss": 1.3408, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.14249073810202337, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 9.526795895096922e-06, | |
| "loss": 1.3522, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1481903676261043, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 9.50779171417712e-06, | |
| "loss": 1.3595, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.15388999715018523, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.488787533257318e-06, | |
| "loss": 1.381, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15958962667426618, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 9.469783352337516e-06, | |
| "loss": 1.3109, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1652892561983471, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.450779171417712e-06, | |
| "loss": 1.3793, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.17098888572242804, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.43177499049791e-06, | |
| "loss": 1.3373, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17668851524650897, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 9.412770809578108e-06, | |
| "loss": 1.3474, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1823881447705899, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 9.393766628658306e-06, | |
| "loss": 1.3076, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.18808777429467086, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 9.374762447738504e-06, | |
| "loss": 1.2929, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.19378740381875179, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 9.3557582668187e-06, | |
| "loss": 1.3652, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.19948703334283271, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 9.336754085898898e-06, | |
| "loss": 1.3535, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.20518666286691364, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 9.317749904979096e-06, | |
| "loss": 1.3388, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.21088629239099457, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 9.298745724059294e-06, | |
| "loss": 1.3083, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.21658592191507553, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.279741543139492e-06, | |
| "loss": 1.3409, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.22228555143915646, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 9.26073736221969e-06, | |
| "loss": 1.3149, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2279851809632374, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 9.241733181299886e-06, | |
| "loss": 1.3525, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23368481048731832, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 9.222729000380084e-06, | |
| "loss": 1.3343, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.23938444001139925, | |
| "grad_norm": 6.875, | |
| "learning_rate": 9.203724819460282e-06, | |
| "loss": 1.3098, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2450840695354802, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 9.18472063854048e-06, | |
| "loss": 1.2878, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2507836990595611, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 9.165716457620676e-06, | |
| "loss": 1.2681, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.2564833285836421, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.146712276700876e-06, | |
| "loss": 1.2974, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.262182958107723, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 9.127708095781072e-06, | |
| "loss": 1.3047, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.26788258763180395, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 9.10870391486127e-06, | |
| "loss": 1.2063, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2735822171558849, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 9.089699733941468e-06, | |
| "loss": 1.3046, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2792818466799658, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 9.070695553021666e-06, | |
| "loss": 1.2633, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.28498147620404674, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 9.051691372101862e-06, | |
| "loss": 1.3394, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29068110572812766, | |
| "grad_norm": 5.75, | |
| "learning_rate": 9.032687191182062e-06, | |
| "loss": 1.2653, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2963807352522086, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 9.013683010262258e-06, | |
| "loss": 1.2359, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3020803647762895, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 8.994678829342456e-06, | |
| "loss": 1.2021, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.30777999430037045, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.975674648422654e-06, | |
| "loss": 1.233, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.31347962382445144, | |
| "grad_norm": 7.5, | |
| "learning_rate": 8.956670467502852e-06, | |
| "loss": 1.3498, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.31917925334853237, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 8.937666286583048e-06, | |
| "loss": 1.2261, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3248788828726133, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.918662105663248e-06, | |
| "loss": 1.3059, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3305785123966942, | |
| "grad_norm": 6.5, | |
| "learning_rate": 8.899657924743444e-06, | |
| "loss": 1.241, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.33627814192077515, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 8.880653743823642e-06, | |
| "loss": 1.3466, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3419777714448561, | |
| "grad_norm": 6.75, | |
| "learning_rate": 8.861649562903838e-06, | |
| "loss": 1.314, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.347677400968937, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.842645381984038e-06, | |
| "loss": 1.2967, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.35337703049301794, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 8.823641201064234e-06, | |
| "loss": 1.3309, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.35907666001709887, | |
| "grad_norm": 7.375, | |
| "learning_rate": 8.804637020144432e-06, | |
| "loss": 1.248, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3647762895411798, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.78563283922463e-06, | |
| "loss": 1.2862, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3704759190652608, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 8.766628658304828e-06, | |
| "loss": 1.2528, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3761755485893417, | |
| "grad_norm": 6.75, | |
| "learning_rate": 8.747624477385024e-06, | |
| "loss": 1.2739, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.38187517811342264, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 8.728620296465224e-06, | |
| "loss": 1.2521, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.38757480763750357, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 8.70961611554542e-06, | |
| "loss": 1.2068, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3932744371615845, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 8.690611934625618e-06, | |
| "loss": 1.3239, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.39897406668566543, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 8.671607753705816e-06, | |
| "loss": 1.2795, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.40467369620974636, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 8.652603572786014e-06, | |
| "loss": 1.2848, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4103733257338273, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 8.63359939186621e-06, | |
| "loss": 1.2491, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.4160729552579082, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 8.61459521094641e-06, | |
| "loss": 1.229, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.42177258478198915, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 8.595591030026606e-06, | |
| "loss": 1.2394, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.42747221430607013, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.576586849106804e-06, | |
| "loss": 1.3423, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.43317184383015106, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 8.557582668187002e-06, | |
| "loss": 1.3051, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.438871473354232, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 8.5385784872672e-06, | |
| "loss": 1.2504, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.4445711028783129, | |
| "grad_norm": 7.5, | |
| "learning_rate": 8.519574306347396e-06, | |
| "loss": 1.2255, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.45027073240239385, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 8.500570125427594e-06, | |
| "loss": 1.2367, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.4559703619264748, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 8.481565944507792e-06, | |
| "loss": 1.2855, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4616699914505557, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.46256176358799e-06, | |
| "loss": 1.2875, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.46736962097463663, | |
| "grad_norm": 6.75, | |
| "learning_rate": 8.443557582668188e-06, | |
| "loss": 1.2946, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.47306925049871756, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 8.424553401748386e-06, | |
| "loss": 1.2672, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4787688800227985, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 8.405549220828583e-06, | |
| "loss": 1.2279, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4844685095468795, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 8.38654503990878e-06, | |
| "loss": 1.2593, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4901681390709604, | |
| "grad_norm": 7.0, | |
| "learning_rate": 8.367540858988978e-06, | |
| "loss": 1.2662, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.49586776859504134, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 8.348536678069176e-06, | |
| "loss": 1.2404, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5015673981191222, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 8.329532497149374e-06, | |
| "loss": 1.3002, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5072670276432032, | |
| "grad_norm": 6.5, | |
| "learning_rate": 8.31052831622957e-06, | |
| "loss": 1.2786, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.5129666571672842, | |
| "grad_norm": 7.125, | |
| "learning_rate": 8.291524135309769e-06, | |
| "loss": 1.2384, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.518666286691365, | |
| "grad_norm": 7.375, | |
| "learning_rate": 8.272519954389967e-06, | |
| "loss": 1.2383, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.524365916215446, | |
| "grad_norm": 7.375, | |
| "learning_rate": 8.253515773470164e-06, | |
| "loss": 1.2136, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5300655457395269, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 8.234511592550362e-06, | |
| "loss": 1.2139, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5357651752636079, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.215507411630559e-06, | |
| "loss": 1.2511, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5414648047876888, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 8.196503230710757e-06, | |
| "loss": 1.2315, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5471644343117698, | |
| "grad_norm": 6.875, | |
| "learning_rate": 8.177499049790955e-06, | |
| "loss": 1.2691, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5528640638358506, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 8.158494868871153e-06, | |
| "loss": 1.2067, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5585636933599316, | |
| "grad_norm": 6.125, | |
| "learning_rate": 8.13949068795135e-06, | |
| "loss": 1.1774, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5642633228840125, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 8.120486507031547e-06, | |
| "loss": 1.2299, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5699629524080935, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 8.101482326111745e-06, | |
| "loss": 1.2525, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5756625819321745, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 8.082478145191943e-06, | |
| "loss": 1.2287, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5813622114562553, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 8.06347396427214e-06, | |
| "loss": 1.2515, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5870618409803363, | |
| "grad_norm": 8.625, | |
| "learning_rate": 8.044469783352339e-06, | |
| "loss": 1.228, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5927614705044172, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 8.025465602432537e-06, | |
| "loss": 1.2277, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5984611000284982, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.006461421512733e-06, | |
| "loss": 1.2442, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.604160729552579, | |
| "grad_norm": 6.875, | |
| "learning_rate": 7.98745724059293e-06, | |
| "loss": 1.177, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.60986035907666, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 7.968453059673129e-06, | |
| "loss": 1.1878, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.6155599886007409, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 7.949448878753327e-06, | |
| "loss": 1.285, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.6212596181248219, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 7.930444697833523e-06, | |
| "loss": 1.2323, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.6269592476489029, | |
| "grad_norm": 6.625, | |
| "learning_rate": 7.911440516913723e-06, | |
| "loss": 1.1727, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6326588771729837, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 7.892436335993919e-06, | |
| "loss": 1.1775, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.6383585066970647, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 7.873432155074117e-06, | |
| "loss": 1.3263, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6440581362211456, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 7.854427974154315e-06, | |
| "loss": 1.1943, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6497577657452266, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 7.835423793234513e-06, | |
| "loss": 1.2537, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6554573952693075, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 7.816419612314709e-06, | |
| "loss": 1.1514, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6611570247933884, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 7.797415431394909e-06, | |
| "loss": 1.1908, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6668566543174693, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 7.778411250475105e-06, | |
| "loss": 1.2328, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6725562838415503, | |
| "grad_norm": 7.375, | |
| "learning_rate": 7.759407069555303e-06, | |
| "loss": 1.2281, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6782559133656312, | |
| "grad_norm": 7.125, | |
| "learning_rate": 7.7404028886355e-06, | |
| "loss": 1.2469, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6839555428897122, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 7.721398707715699e-06, | |
| "loss": 1.2444, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 7.702394526795895e-06, | |
| "loss": 1.3066, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.695354801937874, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 7.683390345876095e-06, | |
| "loss": 1.2125, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.701054431461955, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 7.664386164956291e-06, | |
| "loss": 1.2275, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.7067540609860359, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 7.645381984036489e-06, | |
| "loss": 1.2702, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.7124536905101169, | |
| "grad_norm": 7.375, | |
| "learning_rate": 7.626377803116686e-06, | |
| "loss": 1.2547, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.7181533200341977, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 7.607373622196884e-06, | |
| "loss": 1.2192, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.7238529495582787, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 7.588369441277081e-06, | |
| "loss": 1.1881, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.7295525790823596, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 7.569365260357279e-06, | |
| "loss": 1.2096, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.7352522086064406, | |
| "grad_norm": 6.75, | |
| "learning_rate": 7.550361079437477e-06, | |
| "loss": 1.246, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.7409518381305216, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 7.531356898517674e-06, | |
| "loss": 1.1759, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7466514676546024, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 7.512352717597872e-06, | |
| "loss": 1.1641, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7523510971786834, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 7.49334853667807e-06, | |
| "loss": 1.2776, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7580507267027643, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 7.474344355758267e-06, | |
| "loss": 1.2651, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7637503562268453, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 7.455340174838464e-06, | |
| "loss": 1.2367, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7694499857509262, | |
| "grad_norm": 8.125, | |
| "learning_rate": 7.436335993918663e-06, | |
| "loss": 1.2126, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7751496152750071, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 7.41733181299886e-06, | |
| "loss": 1.1739, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.780849244799088, | |
| "grad_norm": 6.875, | |
| "learning_rate": 7.398327632079057e-06, | |
| "loss": 1.1959, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.786548874323169, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 7.379323451159256e-06, | |
| "loss": 1.1734, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7922485038472499, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 7.360319270239453e-06, | |
| "loss": 1.2139, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.7979481333713309, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 7.34131508931965e-06, | |
| "loss": 1.1497, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8036477628954118, | |
| "grad_norm": 8.75, | |
| "learning_rate": 7.322310908399849e-06, | |
| "loss": 1.2907, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.8093473924194927, | |
| "grad_norm": 7.125, | |
| "learning_rate": 7.303306727480046e-06, | |
| "loss": 1.2595, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.8150470219435737, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 7.284302546560243e-06, | |
| "loss": 1.2449, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.8207466514676546, | |
| "grad_norm": 6.75, | |
| "learning_rate": 7.265298365640442e-06, | |
| "loss": 1.2571, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 7.246294184720639e-06, | |
| "loss": 1.2183, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.8321459105158164, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 7.227290003800836e-06, | |
| "loss": 1.2293, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.8378455400398974, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 7.208285822881035e-06, | |
| "loss": 1.1618, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.8435451695639783, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 7.189281641961232e-06, | |
| "loss": 1.1948, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.8492447990880593, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 7.170277461041429e-06, | |
| "loss": 1.1967, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8549444286121403, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 7.151273280121628e-06, | |
| "loss": 1.2639, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8606440581362211, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 7.132269099201825e-06, | |
| "loss": 1.2751, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8663436876603021, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 7.113264918282022e-06, | |
| "loss": 1.2803, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.872043317184383, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 7.094260737362221e-06, | |
| "loss": 1.1903, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.877742946708464, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 7.075256556442418e-06, | |
| "loss": 1.1932, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8834425762325449, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 7.056252375522615e-06, | |
| "loss": 1.2517, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8891422057566258, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 7.037248194602814e-06, | |
| "loss": 1.2413, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8948418352807067, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.018244013683011e-06, | |
| "loss": 1.1405, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.9005414648047877, | |
| "grad_norm": 7.25, | |
| "learning_rate": 6.999239832763208e-06, | |
| "loss": 1.2416, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.9062410943288686, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 6.9802356518434055e-06, | |
| "loss": 1.195, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.9119407238529496, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 6.961231470923604e-06, | |
| "loss": 1.2723, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9176403533770305, | |
| "grad_norm": 7.25, | |
| "learning_rate": 6.942227290003801e-06, | |
| "loss": 1.2642, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.9233399829011114, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 6.9232231090839985e-06, | |
| "loss": 1.2173, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.9290396124251924, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 6.904218928164197e-06, | |
| "loss": 1.2514, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.9347392419492733, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 6.885214747244394e-06, | |
| "loss": 1.1989, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.9404388714733543, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 6.8662105663245915e-06, | |
| "loss": 1.2291, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.9461385009974351, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 6.8472063854047895e-06, | |
| "loss": 1.2637, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.9518381305215161, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 6.828202204484987e-06, | |
| "loss": 1.1973, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.957537760045597, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 6.8091980235651845e-06, | |
| "loss": 1.2915, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.963237389569678, | |
| "grad_norm": 7.875, | |
| "learning_rate": 6.7901938426453825e-06, | |
| "loss": 1.1392, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.968937019093759, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.7711896617255804e-06, | |
| "loss": 1.1384, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9746366486178398, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 6.7521854808057775e-06, | |
| "loss": 1.1322, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9803362781419208, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 6.7331812998859755e-06, | |
| "loss": 1.2261, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9860359076660017, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 6.7141771189661735e-06, | |
| "loss": 1.1714, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9917355371900827, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 6.6951729380463706e-06, | |
| "loss": 1.1884, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9974351667141635, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.6761687571265685e-06, | |
| "loss": 1.28, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.0034197777144485, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 6.657164576206766e-06, | |
| "loss": 1.3048, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.0091194072385294, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 6.638160395286964e-06, | |
| "loss": 1.1814, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.0148190367626104, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 6.6191562143671615e-06, | |
| "loss": 1.2091, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.0205186662866914, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.600152033447359e-06, | |
| "loss": 1.2139, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.0262182958107724, | |
| "grad_norm": 8.5, | |
| "learning_rate": 6.581147852527557e-06, | |
| "loss": 1.1838, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.0319179253348532, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 6.5621436716077546e-06, | |
| "loss": 1.2291, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.0376175548589341, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.543139490687952e-06, | |
| "loss": 1.1906, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.0433171843830151, | |
| "grad_norm": 6.375, | |
| "learning_rate": 6.52413530976815e-06, | |
| "loss": 1.2225, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.0490168139070961, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 6.5051311288483476e-06, | |
| "loss": 1.1612, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.0547164434311769, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 6.486126947928545e-06, | |
| "loss": 1.1946, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.0604160729552579, | |
| "grad_norm": 7.375, | |
| "learning_rate": 6.467122767008742e-06, | |
| "loss": 1.2219, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0661157024793388, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 6.448118586088941e-06, | |
| "loss": 1.2461, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.0718153320034198, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 6.429114405169138e-06, | |
| "loss": 1.1393, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.0775149615275008, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 6.410110224249335e-06, | |
| "loss": 1.2083, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.0832145910515816, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 6.391106043329534e-06, | |
| "loss": 1.243, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.0889142205756626, | |
| "grad_norm": 6.875, | |
| "learning_rate": 6.372101862409731e-06, | |
| "loss": 1.2199, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.0946138500997435, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 6.353097681489928e-06, | |
| "loss": 1.202, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.1003134796238245, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.334093500570126e-06, | |
| "loss": 1.1115, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.1060131091479053, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 6.315089319650324e-06, | |
| "loss": 1.1845, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.1117127386719863, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 6.296085138730521e-06, | |
| "loss": 1.1988, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.1174123681960673, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 6.277080957810718e-06, | |
| "loss": 1.1937, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.1231119977201482, | |
| "grad_norm": 7.75, | |
| "learning_rate": 6.258076776890917e-06, | |
| "loss": 1.2473, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.1288116272442292, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 6.239072595971114e-06, | |
| "loss": 1.2198, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.13451125676831, | |
| "grad_norm": 9.625, | |
| "learning_rate": 6.220068415051311e-06, | |
| "loss": 1.1505, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.140210886292391, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 6.20106423413151e-06, | |
| "loss": 1.1457, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5262, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.661025051744666e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |