| { | |
| "best_metric": 1.6946364641189575, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.0047846889952154, | |
| "eval_steps": 50, | |
| "global_step": 314, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009569377990430622, | |
| "grad_norm": 0.34759339690208435, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3538, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009569377990430622, | |
| "eval_loss": 1.8290536403656006, | |
| "eval_runtime": 5.5952, | |
| "eval_samples_per_second": 31.634, | |
| "eval_steps_per_second": 8.043, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019138755980861243, | |
| "grad_norm": 0.4182339608669281, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6102, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.028708133971291867, | |
| "grad_norm": 0.44165006279945374, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6496, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03827751196172249, | |
| "grad_norm": 0.4331439733505249, | |
| "learning_rate": 4e-05, | |
| "loss": 1.469, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04784688995215311, | |
| "grad_norm": 0.4460451006889343, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5172, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05741626794258373, | |
| "grad_norm": 0.4264964759349823, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5255, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06698564593301436, | |
| "grad_norm": 0.4508237838745117, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5265, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07655502392344497, | |
| "grad_norm": 0.4574127495288849, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4239, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0861244019138756, | |
| "grad_norm": 0.49037742614746094, | |
| "learning_rate": 9e-05, | |
| "loss": 1.6395, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09569377990430622, | |
| "grad_norm": 0.5026654005050659, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4849, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 0.5212116837501526, | |
| "learning_rate": 9.999733013856754e-05, | |
| "loss": 1.5304, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11483253588516747, | |
| "grad_norm": 0.5327346920967102, | |
| "learning_rate": 9.998932083939656e-05, | |
| "loss": 1.5773, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12440191387559808, | |
| "grad_norm": 0.5605044364929199, | |
| "learning_rate": 9.99759729578358e-05, | |
| "loss": 1.6525, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1339712918660287, | |
| "grad_norm": 0.5749955177307129, | |
| "learning_rate": 9.995728791936504e-05, | |
| "loss": 1.752, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14354066985645933, | |
| "grad_norm": 0.6115134954452515, | |
| "learning_rate": 9.993326771944284e-05, | |
| "loss": 1.7045, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15311004784688995, | |
| "grad_norm": 0.6606630086898804, | |
| "learning_rate": 9.990391492329341e-05, | |
| "loss": 1.9934, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16267942583732056, | |
| "grad_norm": 0.6611381769180298, | |
| "learning_rate": 9.986923266563267e-05, | |
| "loss": 1.7449, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1722488038277512, | |
| "grad_norm": 0.673966109752655, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 1.9033, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.7516447305679321, | |
| "learning_rate": 9.97838951500302e-05, | |
| "loss": 1.9845, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19138755980861244, | |
| "grad_norm": 0.7882755398750305, | |
| "learning_rate": 9.973324900566213e-05, | |
| "loss": 1.8804, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.20095693779904306, | |
| "grad_norm": 0.8274745345115662, | |
| "learning_rate": 9.967729162595682e-05, | |
| "loss": 1.9417, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.9981335401535034, | |
| "learning_rate": 9.961602898685226e-05, | |
| "loss": 1.9608, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22009569377990432, | |
| "grad_norm": 1.044714093208313, | |
| "learning_rate": 9.954946763085872e-05, | |
| "loss": 2.0085, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22966507177033493, | |
| "grad_norm": 1.3464337587356567, | |
| "learning_rate": 9.947761466636014e-05, | |
| "loss": 2.1601, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23923444976076555, | |
| "grad_norm": 1.9024715423583984, | |
| "learning_rate": 9.940047776685484e-05, | |
| "loss": 1.9602, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24880382775119617, | |
| "grad_norm": 4.2014479637146, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 2.2396, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2583732057416268, | |
| "grad_norm": 0.4245186746120453, | |
| "learning_rate": 9.923038567741256e-05, | |
| "loss": 1.6088, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2679425837320574, | |
| "grad_norm": 0.5188934206962585, | |
| "learning_rate": 9.913744865236798e-05, | |
| "loss": 1.6053, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.27751196172248804, | |
| "grad_norm": 0.45742854475975037, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 1.5371, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28708133971291866, | |
| "grad_norm": 0.42776361107826233, | |
| "learning_rate": 9.893584226636772e-05, | |
| "loss": 1.5269, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2966507177033493, | |
| "grad_norm": 0.4135408401489258, | |
| "learning_rate": 9.882719443585664e-05, | |
| "loss": 1.5842, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3062200956937799, | |
| "grad_norm": 0.3922288715839386, | |
| "learning_rate": 9.871333213161438e-05, | |
| "loss": 1.634, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 0.37230274081230164, | |
| "learning_rate": 9.859426751350391e-05, | |
| "loss": 1.4989, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3253588516746411, | |
| "grad_norm": 0.39755305647850037, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 1.476, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3349282296650718, | |
| "grad_norm": 0.4258255362510681, | |
| "learning_rate": 9.834058275166383e-05, | |
| "loss": 1.4965, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3444976076555024, | |
| "grad_norm": 0.43615198135375977, | |
| "learning_rate": 9.820598970006069e-05, | |
| "loss": 1.4774, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.35406698564593303, | |
| "grad_norm": 0.4417138397693634, | |
| "learning_rate": 9.806624851594898e-05, | |
| "loss": 1.473, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.4854167699813843, | |
| "learning_rate": 9.792137412291265e-05, | |
| "loss": 1.5505, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.37320574162679426, | |
| "grad_norm": 0.4760205149650574, | |
| "learning_rate": 9.777138199273384e-05, | |
| "loss": 1.5259, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3827751196172249, | |
| "grad_norm": 0.4862656891345978, | |
| "learning_rate": 9.761628814374073e-05, | |
| "loss": 1.5189, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3923444976076555, | |
| "grad_norm": 0.5357599258422852, | |
| "learning_rate": 9.745610913909674e-05, | |
| "loss": 1.6591, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4019138755980861, | |
| "grad_norm": 0.5464658141136169, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 1.7461, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41148325358851673, | |
| "grad_norm": 0.5899918079376221, | |
| "learning_rate": 9.712056462901519e-05, | |
| "loss": 1.6656, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.6547144651412964, | |
| "learning_rate": 9.694523495787149e-05, | |
| "loss": 1.8081, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.430622009569378, | |
| "grad_norm": 0.6419932842254639, | |
| "learning_rate": 9.67648917958377e-05, | |
| "loss": 1.9222, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44019138755980863, | |
| "grad_norm": 0.6938351988792419, | |
| "learning_rate": 9.657955440256395e-05, | |
| "loss": 1.9533, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.44976076555023925, | |
| "grad_norm": 0.7407806515693665, | |
| "learning_rate": 9.638924257105659e-05, | |
| "loss": 1.8787, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.45933014354066987, | |
| "grad_norm": 0.838914155960083, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 1.9171, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4688995215311005, | |
| "grad_norm": 0.9300668239593506, | |
| "learning_rate": 9.599377741940792e-05, | |
| "loss": 2.0435, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4784688995215311, | |
| "grad_norm": 1.1782912015914917, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 2.1426, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4784688995215311, | |
| "eval_loss": 1.7335811853408813, | |
| "eval_runtime": 5.6074, | |
| "eval_samples_per_second": 31.566, | |
| "eval_steps_per_second": 8.025, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4880382775119617, | |
| "grad_norm": 1.5746591091156006, | |
| "learning_rate": 9.55786652703264e-05, | |
| "loss": 1.946, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.49760765550239233, | |
| "grad_norm": 2.7460060119628906, | |
| "learning_rate": 9.5363796659078e-05, | |
| "loss": 1.6779, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.507177033492823, | |
| "grad_norm": 0.29309824109077454, | |
| "learning_rate": 9.514408344578439e-05, | |
| "loss": 1.4202, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5167464114832536, | |
| "grad_norm": 0.35270825028419495, | |
| "learning_rate": 9.491954909459895e-05, | |
| "loss": 1.682, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.3822132647037506, | |
| "learning_rate": 9.469021758454586e-05, | |
| "loss": 1.5133, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5358851674641149, | |
| "grad_norm": 0.3833141326904297, | |
| "learning_rate": 9.445611340695926e-05, | |
| "loss": 1.4092, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.35923632979393005, | |
| "learning_rate": 9.421726156286777e-05, | |
| "loss": 1.4326, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5550239234449761, | |
| "grad_norm": 0.36976897716522217, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 1.5209, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5645933014354066, | |
| "grad_norm": 0.37268730998039246, | |
| "learning_rate": 9.372541741168272e-05, | |
| "loss": 1.5378, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5741626794258373, | |
| "grad_norm": 0.3804684579372406, | |
| "learning_rate": 9.347247763081835e-05, | |
| "loss": 1.488, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.583732057416268, | |
| "grad_norm": 0.38005149364471436, | |
| "learning_rate": 9.321489523029797e-05, | |
| "loss": 1.4471, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5933014354066986, | |
| "grad_norm": 0.41117990016937256, | |
| "learning_rate": 9.295269771849427e-05, | |
| "loss": 1.4504, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6028708133971292, | |
| "grad_norm": 0.42418965697288513, | |
| "learning_rate": 9.26859130966482e-05, | |
| "loss": 1.6096, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6124401913875598, | |
| "grad_norm": 0.4689091145992279, | |
| "learning_rate": 9.241456985587868e-05, | |
| "loss": 1.5569, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6220095693779905, | |
| "grad_norm": 0.4588874876499176, | |
| "learning_rate": 9.213869697413986e-05, | |
| "loss": 1.595, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.4907477796077728, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 1.6135, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6411483253588517, | |
| "grad_norm": 0.5110955834388733, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 1.6928, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6507177033492823, | |
| "grad_norm": 0.5802223682403564, | |
| "learning_rate": 9.12841974998278e-05, | |
| "loss": 1.8763, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6602870813397129, | |
| "grad_norm": 0.5587326884269714, | |
| "learning_rate": 9.099050546106133e-05, | |
| "loss": 1.6938, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6698564593301436, | |
| "grad_norm": 0.5897432565689087, | |
| "learning_rate": 9.069243586350975e-05, | |
| "loss": 1.7649, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6794258373205742, | |
| "grad_norm": 0.6415350437164307, | |
| "learning_rate": 9.039002053935398e-05, | |
| "loss": 1.7284, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6889952153110048, | |
| "grad_norm": 0.6885668039321899, | |
| "learning_rate": 9.008329178487442e-05, | |
| "loss": 1.9107, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6985645933014354, | |
| "grad_norm": 0.7731958627700806, | |
| "learning_rate": 8.977228235700196e-05, | |
| "loss": 1.8644, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7081339712918661, | |
| "grad_norm": 0.896846354007721, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 1.9361, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7177033492822966, | |
| "grad_norm": 0.9463577270507812, | |
| "learning_rate": 8.913755479101574e-05, | |
| "loss": 2.0266, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.2584304809570312, | |
| "learning_rate": 8.881390443828787e-05, | |
| "loss": 2.1377, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 1.4642155170440674, | |
| "learning_rate": 8.84861089756999e-05, | |
| "loss": 1.7037, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7464114832535885, | |
| "grad_norm": 1.769221544265747, | |
| "learning_rate": 8.815420340999033e-05, | |
| "loss": 1.4371, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7559808612440191, | |
| "grad_norm": 0.2741526663303375, | |
| "learning_rate": 8.781822318683394e-05, | |
| "loss": 1.4535, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7655502392344498, | |
| "grad_norm": 0.30794793367385864, | |
| "learning_rate": 8.74782041870563e-05, | |
| "loss": 1.3062, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7751196172248804, | |
| "grad_norm": 0.33890411257743835, | |
| "learning_rate": 8.7134182722802e-05, | |
| "loss": 1.4661, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.784688995215311, | |
| "grad_norm": 0.34827500581741333, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 1.4192, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7942583732057417, | |
| "grad_norm": 0.38004156947135925, | |
| "learning_rate": 8.643427978272306e-05, | |
| "loss": 1.5631, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8038277511961722, | |
| "grad_norm": 0.3742336630821228, | |
| "learning_rate": 8.60784730526531e-05, | |
| "loss": 1.5425, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8133971291866029, | |
| "grad_norm": 0.39842689037323, | |
| "learning_rate": 8.571881334163334e-05, | |
| "loss": 1.523, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8229665071770335, | |
| "grad_norm": 0.3779315948486328, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 1.4504, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8325358851674641, | |
| "grad_norm": 0.4018203616142273, | |
| "learning_rate": 8.498808902277398e-05, | |
| "loss": 1.5239, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.4032476246356964, | |
| "learning_rate": 8.461710245224148e-05, | |
| "loss": 1.4557, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8516746411483254, | |
| "grad_norm": 0.4354485273361206, | |
| "learning_rate": 8.424241896703936e-05, | |
| "loss": 1.5673, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.861244019138756, | |
| "grad_norm": 0.44848817586898804, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 1.6307, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8708133971291866, | |
| "grad_norm": 0.4750164747238159, | |
| "learning_rate": 8.348212169964076e-05, | |
| "loss": 1.5991, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8803827751196173, | |
| "grad_norm": 0.4986872673034668, | |
| "learning_rate": 8.309658911297834e-05, | |
| "loss": 1.6703, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8899521531100478, | |
| "grad_norm": 0.5188276171684265, | |
| "learning_rate": 8.270752199404316e-05, | |
| "loss": 1.5482, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8995215311004785, | |
| "grad_norm": 0.5662826895713806, | |
| "learning_rate": 8.231496189304704e-05, | |
| "loss": 1.7155, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.5895065069198608, | |
| "learning_rate": 8.191895073323294e-05, | |
| "loss": 1.6824, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9186602870813397, | |
| "grad_norm": 0.6644408702850342, | |
| "learning_rate": 8.151953080639775e-05, | |
| "loss": 1.8169, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9282296650717703, | |
| "grad_norm": 0.7228156924247742, | |
| "learning_rate": 8.111674476837581e-05, | |
| "loss": 1.9145, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.937799043062201, | |
| "grad_norm": 0.8052177429199219, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 1.9175, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.7819249033927917, | |
| "learning_rate": 8.030124677492512e-05, | |
| "loss": 1.8468, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "grad_norm": 0.874369204044342, | |
| "learning_rate": 7.988862191016205e-05, | |
| "loss": 1.8646, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "eval_loss": 1.6946364641189575, | |
| "eval_runtime": 5.7261, | |
| "eval_samples_per_second": 30.911, | |
| "eval_steps_per_second": 7.859, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9665071770334929, | |
| "grad_norm": 1.0219553709030151, | |
| "learning_rate": 7.947280510624267e-05, | |
| "loss": 1.8337, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9760765550239234, | |
| "grad_norm": 1.1334084272384644, | |
| "learning_rate": 7.905384077009693e-05, | |
| "loss": 1.842, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9856459330143541, | |
| "grad_norm": 1.5165338516235352, | |
| "learning_rate": 7.863177364479368e-05, | |
| "loss": 1.86, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9952153110047847, | |
| "grad_norm": 1.9935665130615234, | |
| "learning_rate": 7.820664880476256e-05, | |
| "loss": 1.5559, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0047846889952152, | |
| "grad_norm": 0.6355700492858887, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 2.3976, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.014354066985646, | |
| "grad_norm": 0.30273714661598206, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 1.2784, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0239234449760766, | |
| "grad_norm": 0.33532121777534485, | |
| "learning_rate": 7.691338360967673e-05, | |
| "loss": 1.569, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0334928229665072, | |
| "grad_norm": 0.3462691605091095, | |
| "learning_rate": 7.647648511303544e-05, | |
| "loss": 1.3755, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0430622009569377, | |
| "grad_norm": 0.3888528645038605, | |
| "learning_rate": 7.603675907453534e-05, | |
| "loss": 1.4853, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.4027070999145508, | |
| "learning_rate": 7.559425245448006e-05, | |
| "loss": 1.426, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.062200956937799, | |
| "grad_norm": 0.4104244112968445, | |
| "learning_rate": 7.514901251012394e-05, | |
| "loss": 1.4623, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0717703349282297, | |
| "grad_norm": 0.3884302079677582, | |
| "learning_rate": 7.470108679062521e-05, | |
| "loss": 1.3809, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0813397129186604, | |
| "grad_norm": 0.37219735980033875, | |
| "learning_rate": 7.425052313196799e-05, | |
| "loss": 1.2797, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.39946213364601135, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 1.5812, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1004784688995215, | |
| "grad_norm": 0.3919571042060852, | |
| "learning_rate": 7.334167474456227e-05, | |
| "loss": 1.3676, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1100478468899522, | |
| "grad_norm": 0.3989256024360657, | |
| "learning_rate": 7.288348707578408e-05, | |
| "loss": 1.4576, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1196172248803828, | |
| "grad_norm": 0.4338986575603485, | |
| "learning_rate": 7.242285557742254e-05, | |
| "loss": 1.4479, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1291866028708135, | |
| "grad_norm": 0.4401372969150543, | |
| "learning_rate": 7.195982944236851e-05, | |
| "loss": 1.4468, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.138755980861244, | |
| "grad_norm": 0.46051037311553955, | |
| "learning_rate": 7.149445811924685e-05, | |
| "loss": 1.4306, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1483253588516746, | |
| "grad_norm": 0.48773014545440674, | |
| "learning_rate": 7.102679130713537e-05, | |
| "loss": 1.4961, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.5366123914718628, | |
| "learning_rate": 7.055687895025755e-05, | |
| "loss": 1.6909, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.167464114832536, | |
| "grad_norm": 0.5703873038291931, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 1.4873, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1770334928229664, | |
| "grad_norm": 0.5885592699050903, | |
| "learning_rate": 6.961051857279565e-05, | |
| "loss": 1.4342, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.186602870813397, | |
| "grad_norm": 0.6451343894004822, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 1.689, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1961722488038278, | |
| "grad_norm": 0.6953871846199036, | |
| "learning_rate": 6.865578124023952e-05, | |
| "loss": 1.5967, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2057416267942584, | |
| "grad_norm": 0.7246585488319397, | |
| "learning_rate": 6.817539852819149e-05, | |
| "loss": 1.5335, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.215311004784689, | |
| "grad_norm": 0.8152251243591309, | |
| "learning_rate": 6.769307478432148e-05, | |
| "loss": 1.5731, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2248803827751196, | |
| "grad_norm": 0.9615043997764587, | |
| "learning_rate": 6.720886151813194e-05, | |
| "loss": 1.6312, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2344497607655502, | |
| "grad_norm": 1.104920506477356, | |
| "learning_rate": 6.672281044091582e-05, | |
| "loss": 1.3792, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.244019138755981, | |
| "grad_norm": 1.6358625888824463, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 1.0759, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2535885167464116, | |
| "grad_norm": 1.1198513507843018, | |
| "learning_rate": 6.57454026743726e-05, | |
| "loss": 1.2198, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.4242468774318695, | |
| "learning_rate": 6.525415036677744e-05, | |
| "loss": 1.4268, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.4336591958999634, | |
| "learning_rate": 6.476126900047234e-05, | |
| "loss": 1.4419, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2822966507177034, | |
| "grad_norm": 0.42183172702789307, | |
| "learning_rate": 6.426681121245527e-05, | |
| "loss": 1.3825, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.291866028708134, | |
| "grad_norm": 0.44051334261894226, | |
| "learning_rate": 6.37708298080774e-05, | |
| "loss": 1.3154, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3014354066985647, | |
| "grad_norm": 0.44545966386795044, | |
| "learning_rate": 6.327337775540362e-05, | |
| "loss": 1.4548, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3110047846889952, | |
| "grad_norm": 0.45855438709259033, | |
| "learning_rate": 6.277450817955594e-05, | |
| "loss": 1.3711, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3205741626794258, | |
| "grad_norm": 0.44808974862098694, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 1.2933, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3301435406698565, | |
| "grad_norm": 0.4726501405239105, | |
| "learning_rate": 6.177272971005528e-05, | |
| "loss": 1.3949, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.339712918660287, | |
| "grad_norm": 0.46715983748435974, | |
| "learning_rate": 6.126992780079031e-05, | |
| "loss": 1.2692, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3492822966507176, | |
| "grad_norm": 0.4883113503456116, | |
| "learning_rate": 6.0765922325702065e-05, | |
| "loss": 1.3629, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3588516746411483, | |
| "grad_norm": 0.504198431968689, | |
| "learning_rate": 6.026076710978171e-05, | |
| "loss": 1.5393, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.5321323871612549, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 1.4842, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3779904306220097, | |
| "grad_norm": 0.5718645453453064, | |
| "learning_rate": 5.924722336357793e-05, | |
| "loss": 1.6818, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.38755980861244, | |
| "grad_norm": 0.5473811030387878, | |
| "learning_rate": 5.873894307414879e-05, | |
| "loss": 1.473, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3971291866028708, | |
| "grad_norm": 0.578420877456665, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 1.5664, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4066985645933014, | |
| "grad_norm": 0.6025746464729309, | |
| "learning_rate": 5.7719637064427444e-05, | |
| "loss": 1.4072, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4162679425837321, | |
| "grad_norm": 0.6686691641807556, | |
| "learning_rate": 5.720872020036734e-05, | |
| "loss": 1.665, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4258373205741628, | |
| "grad_norm": 0.6871073842048645, | |
| "learning_rate": 5.669703348494564e-05, | |
| "loss": 1.6128, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4354066985645932, | |
| "grad_norm": 0.751925528049469, | |
| "learning_rate": 5.618463156346739e-05, | |
| "loss": 1.6338, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4354066985645932, | |
| "eval_loss": 1.7151299715042114, | |
| "eval_runtime": 5.6112, | |
| "eval_samples_per_second": 31.544, | |
| "eval_steps_per_second": 8.02, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.444976076555024, | |
| "grad_norm": 0.8360044360160828, | |
| "learning_rate": 5.567156915761773e-05, | |
| "loss": 1.5875, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.829786479473114, | |
| "learning_rate": 5.515790105961786e-05, | |
| "loss": 1.4418, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.464114832535885, | |
| "grad_norm": 1.025622010231018, | |
| "learning_rate": 5.4643682126373505e-05, | |
| "loss": 1.5733, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 1.1344419717788696, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 1.5007, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4832535885167464, | |
| "grad_norm": 1.346632719039917, | |
| "learning_rate": 5.361381147004053e-05, | |
| "loss": 1.3065, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.492822966507177, | |
| "grad_norm": 2.3821861743927, | |
| "learning_rate": 5.3098269731429736e-05, | |
| "loss": 1.4532, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5023923444976077, | |
| "grad_norm": 1.320799469947815, | |
| "learning_rate": 5.258239711478441e-05, | |
| "loss": 1.0567, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5119617224880382, | |
| "grad_norm": 0.3712046146392822, | |
| "learning_rate": 5.2066248712440656e-05, | |
| "loss": 1.3992, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5215311004784688, | |
| "grad_norm": 0.39161741733551025, | |
| "learning_rate": 5.154987964618704e-05, | |
| "loss": 1.3432, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5311004784688995, | |
| "grad_norm": 0.4063500761985779, | |
| "learning_rate": 5.103334506137772e-05, | |
| "loss": 1.4145, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5406698564593302, | |
| "grad_norm": 0.41844049096107483, | |
| "learning_rate": 5.051670012104337e-05, | |
| "loss": 1.402, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5502392344497609, | |
| "grad_norm": 0.4229384958744049, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3833, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5598086124401913, | |
| "grad_norm": 0.43778741359710693, | |
| "learning_rate": 4.948329987895665e-05, | |
| "loss": 1.4334, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.569377990430622, | |
| "grad_norm": 0.43413645029067993, | |
| "learning_rate": 4.8966654938622295e-05, | |
| "loss": 1.3416, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.45575451850891113, | |
| "learning_rate": 4.845012035381298e-05, | |
| "loss": 1.2998, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.588516746411483, | |
| "grad_norm": 0.47130417823791504, | |
| "learning_rate": 4.7933751287559335e-05, | |
| "loss": 1.4124, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.598086124401914, | |
| "grad_norm": 0.4700482487678528, | |
| "learning_rate": 4.7417602885215604e-05, | |
| "loss": 1.2738, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6076555023923444, | |
| "grad_norm": 0.4811384677886963, | |
| "learning_rate": 4.6901730268570275e-05, | |
| "loss": 1.3465, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6172248803827751, | |
| "grad_norm": 0.49769333004951477, | |
| "learning_rate": 4.638618852995947e-05, | |
| "loss": 1.2767, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6267942583732058, | |
| "grad_norm": 0.5342416167259216, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 1.407, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.54058438539505, | |
| "learning_rate": 4.53563178736265e-05, | |
| "loss": 1.3725, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6459330143540671, | |
| "grad_norm": 0.5859925150871277, | |
| "learning_rate": 4.4842098940382155e-05, | |
| "loss": 1.4315, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6555023923444976, | |
| "grad_norm": 0.6444745659828186, | |
| "learning_rate": 4.4328430842382285e-05, | |
| "loss": 1.5215, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6650717703349283, | |
| "grad_norm": 0.6744863390922546, | |
| "learning_rate": 4.381536843653262e-05, | |
| "loss": 1.5074, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.674641148325359, | |
| "grad_norm": 0.7167531251907349, | |
| "learning_rate": 4.330296651505438e-05, | |
| "loss": 1.5645, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.7303672432899475, | |
| "learning_rate": 4.2791279799632666e-05, | |
| "loss": 1.5725, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.69377990430622, | |
| "grad_norm": 0.7596977949142456, | |
| "learning_rate": 4.2280362935572575e-05, | |
| "loss": 1.5644, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7033492822966507, | |
| "grad_norm": 0.7862472534179688, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 1.4374, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7129186602870812, | |
| "grad_norm": 0.9110777378082275, | |
| "learning_rate": 4.126105692585121e-05, | |
| "loss": 1.5906, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.722488038277512, | |
| "grad_norm": 1.032729148864746, | |
| "learning_rate": 4.075277663642208e-05, | |
| "loss": 1.4351, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7320574162679425, | |
| "grad_norm": 1.3209054470062256, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 1.4821, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7416267942583732, | |
| "grad_norm": 1.7023085355758667, | |
| "learning_rate": 3.973923289021829e-05, | |
| "loss": 1.1474, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7511961722488039, | |
| "grad_norm": 1.3560224771499634, | |
| "learning_rate": 3.9234077674297953e-05, | |
| "loss": 1.2007, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7607655502392343, | |
| "grad_norm": 0.3711628019809723, | |
| "learning_rate": 3.87300721992097e-05, | |
| "loss": 1.4431, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7703349282296652, | |
| "grad_norm": 0.41066452860832214, | |
| "learning_rate": 3.822727028994471e-05, | |
| "loss": 1.4489, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7799043062200957, | |
| "grad_norm": 0.4065649211406708, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 1.34, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.4324609637260437, | |
| "learning_rate": 3.7225491820444075e-05, | |
| "loss": 1.4823, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.799043062200957, | |
| "grad_norm": 0.4210042655467987, | |
| "learning_rate": 3.67266222445964e-05, | |
| "loss": 1.2545, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8086124401913874, | |
| "grad_norm": 0.4477483034133911, | |
| "learning_rate": 3.622917019192261e-05, | |
| "loss": 1.4699, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.4530695378780365, | |
| "learning_rate": 3.5733188787544745e-05, | |
| "loss": 1.3271, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8277511961722488, | |
| "grad_norm": 0.4583020508289337, | |
| "learning_rate": 3.523873099952768e-05, | |
| "loss": 1.3542, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8373205741626795, | |
| "grad_norm": 0.4657497704029083, | |
| "learning_rate": 3.474584963322257e-05, | |
| "loss": 1.321, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8468899521531101, | |
| "grad_norm": 0.504671573638916, | |
| "learning_rate": 3.425459732562741e-05, | |
| "loss": 1.4282, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8564593301435406, | |
| "grad_norm": 0.5105240941047668, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 1.3156, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8660287081339713, | |
| "grad_norm": 0.5363412499427795, | |
| "learning_rate": 3.327718955908419e-05, | |
| "loss": 1.4376, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.875598086124402, | |
| "grad_norm": 0.5492069721221924, | |
| "learning_rate": 3.279113848186808e-05, | |
| "loss": 1.4962, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8851674641148324, | |
| "grad_norm": 0.5526624917984009, | |
| "learning_rate": 3.230692521567852e-05, | |
| "loss": 1.4229, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.5729860067367554, | |
| "learning_rate": 3.18246014718085e-05, | |
| "loss": 1.4785, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9043062200956937, | |
| "grad_norm": 0.5993075966835022, | |
| "learning_rate": 3.13442187597605e-05, | |
| "loss": 1.4457, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9138755980861244, | |
| "grad_norm": 0.6616885662078857, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 1.5659, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9138755980861244, | |
| "eval_loss": 1.6985137462615967, | |
| "eval_runtime": 5.501, | |
| "eval_samples_per_second": 32.176, | |
| "eval_steps_per_second": 8.18, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.923444976076555, | |
| "grad_norm": 0.719092845916748, | |
| "learning_rate": 3.038948142720437e-05, | |
| "loss": 1.7551, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9330143540669855, | |
| "grad_norm": 0.7387653589248657, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 1.3894, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9425837320574164, | |
| "grad_norm": 0.8578170537948608, | |
| "learning_rate": 2.9443121049742474e-05, | |
| "loss": 1.611, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9521531100478469, | |
| "grad_norm": 0.8904649615287781, | |
| "learning_rate": 2.8973208692864624e-05, | |
| "loss": 1.5407, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9617224880382775, | |
| "grad_norm": 0.9571223855018616, | |
| "learning_rate": 2.8505541880753172e-05, | |
| "loss": 1.5899, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9712918660287082, | |
| "grad_norm": 1.179518461227417, | |
| "learning_rate": 2.804017055763149e-05, | |
| "loss": 1.4807, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9808612440191387, | |
| "grad_norm": 1.447110652923584, | |
| "learning_rate": 2.757714442257747e-05, | |
| "loss": 1.3747, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9904306220095693, | |
| "grad_norm": 1.6941592693328857, | |
| "learning_rate": 2.711651292421593e-05, | |
| "loss": 1.1666, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1691313982009888, | |
| "learning_rate": 2.6658325255437744e-05, | |
| "loss": 1.705, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.0095693779904304, | |
| "grad_norm": 0.32746973633766174, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 1.3439, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0191387559808613, | |
| "grad_norm": 0.3683173954486847, | |
| "learning_rate": 2.5749476868032025e-05, | |
| "loss": 1.2786, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.028708133971292, | |
| "grad_norm": 0.37325266003608704, | |
| "learning_rate": 2.529891320937481e-05, | |
| "loss": 1.2246, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0382775119617227, | |
| "grad_norm": 0.4059229791164398, | |
| "learning_rate": 2.485098748987608e-05, | |
| "loss": 1.3188, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.047846889952153, | |
| "grad_norm": 0.4075366258621216, | |
| "learning_rate": 2.4405747545519963e-05, | |
| "loss": 1.3766, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0574162679425836, | |
| "grad_norm": 0.4230879247188568, | |
| "learning_rate": 2.3963240925464685e-05, | |
| "loss": 1.4019, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0669856459330145, | |
| "grad_norm": 0.44408464431762695, | |
| "learning_rate": 2.352351488696457e-05, | |
| "loss": 1.2217, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.076555023923445, | |
| "grad_norm": 0.45052194595336914, | |
| "learning_rate": 2.3086616390323286e-05, | |
| "loss": 1.2761, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0861244019138754, | |
| "grad_norm": 0.4570644795894623, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 1.2865, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.0956937799043063, | |
| "grad_norm": 0.4676749110221863, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 1.3524, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.4937705099582672, | |
| "learning_rate": 2.179335119523745e-05, | |
| "loss": 1.3469, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1148325358851676, | |
| "grad_norm": 0.5289208292961121, | |
| "learning_rate": 2.1368226355206323e-05, | |
| "loss": 1.3204, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.124401913875598, | |
| "grad_norm": 0.5206736922264099, | |
| "learning_rate": 2.094615922990309e-05, | |
| "loss": 1.3027, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1339712918660285, | |
| "grad_norm": 0.5759319067001343, | |
| "learning_rate": 2.0527194893757323e-05, | |
| "loss": 1.4209, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1435406698564594, | |
| "grad_norm": 0.5884088277816772, | |
| "learning_rate": 2.0111378089837956e-05, | |
| "loss": 1.4384, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.15311004784689, | |
| "grad_norm": 0.6209349632263184, | |
| "learning_rate": 1.9698753225074886e-05, | |
| "loss": 1.3493, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1626794258373208, | |
| "grad_norm": 0.6421443819999695, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 1.3352, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.172248803827751, | |
| "grad_norm": 0.6960612535476685, | |
| "learning_rate": 1.8883255231624215e-05, | |
| "loss": 1.4377, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.7550749182701111, | |
| "learning_rate": 1.848046919360225e-05, | |
| "loss": 1.4201, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1913875598086126, | |
| "grad_norm": 0.7307442426681519, | |
| "learning_rate": 1.8081049266767046e-05, | |
| "loss": 1.2216, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.200956937799043, | |
| "grad_norm": 0.7626149654388428, | |
| "learning_rate": 1.768503810695295e-05, | |
| "loss": 1.2509, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.8778466582298279, | |
| "learning_rate": 1.7292478005956848e-05, | |
| "loss": 1.2469, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2200956937799043, | |
| "grad_norm": 0.9473854303359985, | |
| "learning_rate": 1.6903410887021676e-05, | |
| "loss": 1.1896, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.229665071770335, | |
| "grad_norm": 1.163466453552246, | |
| "learning_rate": 1.651787830035926e-05, | |
| "loss": 1.197, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2392344497607657, | |
| "grad_norm": 1.2001004219055176, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.744, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.248803827751196, | |
| "grad_norm": 1.8168264627456665, | |
| "learning_rate": 1.5757581032960638e-05, | |
| "loss": 0.7809, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.258373205741627, | |
| "grad_norm": 0.40678074955940247, | |
| "learning_rate": 1.5382897547758514e-05, | |
| "loss": 1.4282, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2679425837320575, | |
| "grad_norm": 0.4623485207557678, | |
| "learning_rate": 1.5011910977226018e-05, | |
| "loss": 1.4699, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.277511961722488, | |
| "grad_norm": 0.4686884880065918, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.3035, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.287081339712919, | |
| "grad_norm": 0.48231029510498047, | |
| "learning_rate": 1.428118665836668e-05, | |
| "loss": 1.3419, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.2966507177033493, | |
| "grad_norm": 0.49849793314933777, | |
| "learning_rate": 1.3921526947346902e-05, | |
| "loss": 1.2716, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3062200956937797, | |
| "grad_norm": 0.5077965259552002, | |
| "learning_rate": 1.3565720217276939e-05, | |
| "loss": 1.2874, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.511222779750824, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 1.3424, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.325358851674641, | |
| "grad_norm": 0.5203294157981873, | |
| "learning_rate": 1.2865817277198006e-05, | |
| "loss": 1.0946, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.334928229665072, | |
| "grad_norm": 0.5559858083724976, | |
| "learning_rate": 1.2521795812943704e-05, | |
| "loss": 1.3866, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3444976076555024, | |
| "grad_norm": 0.5409491658210754, | |
| "learning_rate": 1.218177681316608e-05, | |
| "loss": 1.2148, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.354066985645933, | |
| "grad_norm": 0.5826281309127808, | |
| "learning_rate": 1.1845796590009683e-05, | |
| "loss": 1.2908, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.570479691028595, | |
| "learning_rate": 1.1513891024300122e-05, | |
| "loss": 1.2363, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.373205741626794, | |
| "grad_norm": 0.5943195223808289, | |
| "learning_rate": 1.118609556171213e-05, | |
| "loss": 1.297, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.382775119617225, | |
| "grad_norm": 0.6288456916809082, | |
| "learning_rate": 1.086244520898428e-05, | |
| "loss": 1.4145, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3923444976076556, | |
| "grad_norm": 0.6577967405319214, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 1.4022, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.3923444976076556, | |
| "eval_loss": 1.7232935428619385, | |
| "eval_runtime": 5.6194, | |
| "eval_samples_per_second": 31.498, | |
| "eval_steps_per_second": 8.008, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.401913875598086, | |
| "grad_norm": 0.6639474630355835, | |
| "learning_rate": 1.0227717642998037e-05, | |
| "loss": 1.3842, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.411483253588517, | |
| "grad_norm": 0.689608097076416, | |
| "learning_rate": 9.916708215125587e-06, | |
| "loss": 1.3595, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.7789566516876221, | |
| "learning_rate": 9.60997946064604e-06, | |
| "loss": 1.3465, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.430622009569378, | |
| "grad_norm": 0.7904567122459412, | |
| "learning_rate": 9.307564136490254e-06, | |
| "loss": 1.4182, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4401913875598087, | |
| "grad_norm": 0.8855068683624268, | |
| "learning_rate": 9.009494538938685e-06, | |
| "loss": 1.4275, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.449760765550239, | |
| "grad_norm": 0.874548614025116, | |
| "learning_rate": 8.715802500172216e-06, | |
| "loss": 1.3804, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.45933014354067, | |
| "grad_norm": 0.9674124121665955, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 1.2389, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4688995215311005, | |
| "grad_norm": 0.9652609825134277, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 1.2455, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.478468899521531, | |
| "grad_norm": 1.2640658617019653, | |
| "learning_rate": 7.861303025860135e-06, | |
| "loss": 1.1927, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.488038277511962, | |
| "grad_norm": 1.4929181337356567, | |
| "learning_rate": 7.585430144121319e-06, | |
| "loss": 0.9064, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.4976076555023923, | |
| "grad_norm": 1.8298393487930298, | |
| "learning_rate": 7.314086903351797e-06, | |
| "loss": 0.8691, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.507177033492823, | |
| "grad_norm": 0.38617315888404846, | |
| "learning_rate": 7.047302281505736e-06, | |
| "loss": 1.3576, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5167464114832536, | |
| "grad_norm": 0.40857091546058655, | |
| "learning_rate": 6.785104769702033e-06, | |
| "loss": 1.2592, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.45714735984802246, | |
| "learning_rate": 6.527522369181655e-06, | |
| "loss": 1.4395, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.535885167464115, | |
| "grad_norm": 0.4632229804992676, | |
| "learning_rate": 6.27458258831729e-06, | |
| "loss": 1.3476, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.472298800945282, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 1.3684, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.555023923444976, | |
| "grad_norm": 0.5024515986442566, | |
| "learning_rate": 5.782738437132224e-06, | |
| "loss": 1.4151, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5645933014354068, | |
| "grad_norm": 0.5029109716415405, | |
| "learning_rate": 5.543886593040737e-06, | |
| "loss": 1.4251, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.574162679425837, | |
| "grad_norm": 0.4956536293029785, | |
| "learning_rate": 5.309782415454151e-06, | |
| "loss": 1.2067, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.583732057416268, | |
| "grad_norm": 0.49592724442481995, | |
| "learning_rate": 5.080450905401057e-06, | |
| "loss": 1.2904, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.5933014354066986, | |
| "grad_norm": 0.5410796403884888, | |
| "learning_rate": 4.855916554215617e-06, | |
| "loss": 1.3218, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6028708133971294, | |
| "grad_norm": 0.5492650866508484, | |
| "learning_rate": 4.636203340922008e-06, | |
| "loss": 1.325, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.61244019138756, | |
| "grad_norm": 0.5458071231842041, | |
| "learning_rate": 4.421334729673593e-06, | |
| "loss": 1.1769, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6220095693779903, | |
| "grad_norm": 0.6131075024604797, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 1.3194, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.6115835905075073, | |
| "learning_rate": 4.00622258059209e-06, | |
| "loss": 1.3663, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6411483253588517, | |
| "grad_norm": 0.6199522614479065, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 1.3035, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.650717703349282, | |
| "grad_norm": 0.6394442915916443, | |
| "learning_rate": 3.6107574289434286e-06, | |
| "loss": 1.4385, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.660287081339713, | |
| "grad_norm": 0.6572586894035339, | |
| "learning_rate": 3.420445597436056e-06, | |
| "loss": 1.2626, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6698564593301435, | |
| "grad_norm": 0.7447577714920044, | |
| "learning_rate": 3.2351082041623125e-06, | |
| "loss": 1.3553, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.679425837320574, | |
| "grad_norm": 0.7671573758125305, | |
| "learning_rate": 3.054765042128521e-06, | |
| "loss": 1.2451, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.688995215311005, | |
| "grad_norm": 0.8798071146011353, | |
| "learning_rate": 2.8794353709848178e-06, | |
| "loss": 1.4746, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.6985645933014353, | |
| "grad_norm": 0.8808631300926208, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 1.3686, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.708133971291866, | |
| "grad_norm": 0.9506293535232544, | |
| "learning_rate": 2.5438908609032698e-06, | |
| "loss": 1.2823, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7177033492822966, | |
| "grad_norm": 0.9930800199508667, | |
| "learning_rate": 2.3837118562592797e-06, | |
| "loss": 1.1185, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.1336544752120972, | |
| "learning_rate": 2.22861800726617e-06, | |
| "loss": 1.1536, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 1.4703437089920044, | |
| "learning_rate": 2.0786258770873647e-06, | |
| "loss": 1.0495, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7464114832535884, | |
| "grad_norm": 1.7416390180587769, | |
| "learning_rate": 1.933751484051027e-06, | |
| "loss": 0.7923, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7559808612440193, | |
| "grad_norm": 0.3726996183395386, | |
| "learning_rate": 1.7940102999393194e-06, | |
| "loss": 1.2954, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7655502392344498, | |
| "grad_norm": 0.4166678786277771, | |
| "learning_rate": 1.6594172483361758e-06, | |
| "loss": 1.3303, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.77511961722488, | |
| "grad_norm": 0.4617297351360321, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 1.4843, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.784688995215311, | |
| "grad_norm": 0.4334274232387543, | |
| "learning_rate": 1.4057324864960975e-06, | |
| "loss": 1.2801, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7942583732057416, | |
| "grad_norm": 0.45249223709106445, | |
| "learning_rate": 1.286667868385627e-06, | |
| "loss": 1.1776, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.803827751196172, | |
| "grad_norm": 0.48716938495635986, | |
| "learning_rate": 1.172805564143359e-06, | |
| "loss": 1.3103, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.813397129186603, | |
| "grad_norm": 0.5110920667648315, | |
| "learning_rate": 1.064157733632276e-06, | |
| "loss": 1.3387, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8229665071770333, | |
| "grad_norm": 0.5020343661308289, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 1.2326, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8325358851674642, | |
| "grad_norm": 0.5177103281021118, | |
| "learning_rate": 8.62551347632029e-07, | |
| "loss": 1.2947, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.5239975452423096, | |
| "learning_rate": 7.696143225874475e-07, | |
| "loss": 1.0978, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8516746411483256, | |
| "grad_norm": 0.5703707933425903, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 1.2674, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.861244019138756, | |
| "grad_norm": 0.5623457431793213, | |
| "learning_rate": 5.995222331451722e-07, | |
| "loss": 1.3601, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8708133971291865, | |
| "grad_norm": 0.5984331369400024, | |
| "learning_rate": 5.223853336398632e-07, | |
| "loss": 1.3409, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8708133971291865, | |
| "eval_loss": 1.7220124006271362, | |
| "eval_runtime": 5.728, | |
| "eval_samples_per_second": 30.901, | |
| "eval_steps_per_second": 7.856, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8803827751196174, | |
| "grad_norm": 0.6223769783973694, | |
| "learning_rate": 4.505323691412711e-07, | |
| "loss": 1.3203, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.889952153110048, | |
| "grad_norm": 0.6258677244186401, | |
| "learning_rate": 3.839710131477492e-07, | |
| "loss": 1.1759, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.8995215311004783, | |
| "grad_norm": 0.6979731321334839, | |
| "learning_rate": 3.2270837404318464e-07, | |
| "loss": 1.541, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.7176746129989624, | |
| "learning_rate": 2.667509943378721e-07, | |
| "loss": 1.3922, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.9186602870813396, | |
| "grad_norm": 0.7628390789031982, | |
| "learning_rate": 2.161048499698115e-07, | |
| "loss": 1.3605, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.92822966507177, | |
| "grad_norm": 0.785723865032196, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 1.3788, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.937799043062201, | |
| "grad_norm": 0.8244166374206543, | |
| "learning_rate": 1.3076733436734322e-07, | |
| "loss": 1.323, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.9065255522727966, | |
| "learning_rate": 9.60850767065924e-08, | |
| "loss": 1.2804, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9569377990430623, | |
| "grad_norm": 1.011901617050171, | |
| "learning_rate": 6.673228055715241e-08, | |
| "loss": 1.2738, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9665071770334928, | |
| "grad_norm": 1.1093995571136475, | |
| "learning_rate": 4.2712080634949024e-08, | |
| "loss": 1.1685, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9760765550239237, | |
| "grad_norm": 1.3170857429504395, | |
| "learning_rate": 2.4027042164198598e-08, | |
| "loss": 0.9393, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.985645933014354, | |
| "grad_norm": 1.4136571884155273, | |
| "learning_rate": 1.0679160603449534e-08, | |
| "loss": 0.8915, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952153110047846, | |
| "grad_norm": 2.444126844406128, | |
| "learning_rate": 2.669861432463838e-09, | |
| "loss": 0.636, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.0047846889952154, | |
| "grad_norm": 0.7383264303207397, | |
| "learning_rate": 0.0, | |
| "loss": 1.9637, | |
| "step": 314 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 314, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 4 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.763060328071168e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |