| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 10, | |
| "global_step": 154, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006493506493506494, | |
| "grad_norm": 5312.028745324425, | |
| "learning_rate": 0.0, | |
| "loss": 12.6644, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012987012987012988, | |
| "grad_norm": 3464.512608575103, | |
| "learning_rate": 1.25e-06, | |
| "loss": 13.386, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01948051948051948, | |
| "grad_norm": 31521.52866951872, | |
| "learning_rate": 2.5e-06, | |
| "loss": 12.124, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 12714.37327874382, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 9.26, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032467532467532464, | |
| "grad_norm": 2957.000475133963, | |
| "learning_rate": 5e-06, | |
| "loss": 7.0851, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03896103896103896, | |
| "grad_norm": 1856.4304411247954, | |
| "learning_rate": 6.25e-06, | |
| "loss": 4.3426, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 380.9114379987856, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 3.9004, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 249.7480254394976, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 4.2357, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05844155844155844, | |
| "grad_norm": 163.12129308225673, | |
| "learning_rate": 1e-05, | |
| "loss": 3.4659, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 306.28794651548174, | |
| "learning_rate": 1.125e-05, | |
| "loss": 3.2504, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 458.0129175616477, | |
| "learning_rate": 1.25e-05, | |
| "loss": 6.6892, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 220.92021209863208, | |
| "learning_rate": 1.375e-05, | |
| "loss": 3.3174, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08441558441558442, | |
| "grad_norm": 140.9175310968349, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 3.8063, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 100.53464591827579, | |
| "learning_rate": 1.6250000000000002e-05, | |
| "loss": 2.7457, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09740259740259741, | |
| "grad_norm": 81.16333748640615, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 2.7346, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 94.60838239841561, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 3.6905, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11038961038961038, | |
| "grad_norm": 96.27886786384646, | |
| "learning_rate": 2e-05, | |
| "loss": 2.7146, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11688311688311688, | |
| "grad_norm": 638.8030278028655, | |
| "learning_rate": 1.9997408848413494e-05, | |
| "loss": 2.9735, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12337662337662338, | |
| "grad_norm": 101.7239129398942, | |
| "learning_rate": 1.9989636736467278e-05, | |
| "loss": 3.0153, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 67.71433029957166, | |
| "learning_rate": 1.9976687691905394e-05, | |
| "loss": 2.0634, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 61.33799178204917, | |
| "learning_rate": 1.9958568425315316e-05, | |
| "loss": 1.7994, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 77.67967542573722, | |
| "learning_rate": 1.9935288326650314e-05, | |
| "loss": 1.8646, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.14935064935064934, | |
| "grad_norm": 66.7541786942052, | |
| "learning_rate": 1.9906859460363307e-05, | |
| "loss": 1.6082, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 66.42183035318644, | |
| "learning_rate": 1.98732965591547e-05, | |
| "loss": 1.4584, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16233766233766234, | |
| "grad_norm": 59.277341273127604, | |
| "learning_rate": 1.9834617016337424e-05, | |
| "loss": 1.2276, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16883116883116883, | |
| "grad_norm": 60.09621841530829, | |
| "learning_rate": 1.979084087682323e-05, | |
| "loss": 1.1038, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17532467532467533, | |
| "grad_norm": 53.451286572243966, | |
| "learning_rate": 1.9741990826734793e-05, | |
| "loss": 0.9448, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 52.91840992849332, | |
| "learning_rate": 1.9688092181649065e-05, | |
| "loss": 0.8475, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18831168831168832, | |
| "grad_norm": 42.915320217692674, | |
| "learning_rate": 1.9629172873477995e-05, | |
| "loss": 0.6988, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 36.403813181510394, | |
| "learning_rate": 1.956526343599335e-05, | |
| "loss": 0.591, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2012987012987013, | |
| "grad_norm": 27.894365664450408, | |
| "learning_rate": 1.9496396989003195e-05, | |
| "loss": 0.504, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 33.58287541986749, | |
| "learning_rate": 1.9422609221188208e-05, | |
| "loss": 0.4665, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 33.5121042744897, | |
| "learning_rate": 1.9343938371606714e-05, | |
| "loss": 0.4967, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22077922077922077, | |
| "grad_norm": 27.982938599625648, | |
| "learning_rate": 1.9260425209878052e-05, | |
| "loss": 0.4135, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 51.93090738473159, | |
| "learning_rate": 1.917211301505453e-05, | |
| "loss": 0.5827, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.23376623376623376, | |
| "grad_norm": 57.25489042342769, | |
| "learning_rate": 1.907904755319289e-05, | |
| "loss": 0.6328, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24025974025974026, | |
| "grad_norm": 11.875972697569978, | |
| "learning_rate": 1.8981277053636963e-05, | |
| "loss": 0.3192, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24675324675324675, | |
| "grad_norm": 27.883526944555015, | |
| "learning_rate": 1.8878852184023754e-05, | |
| "loss": 0.437, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2532467532467532, | |
| "grad_norm": 43.732501540378045, | |
| "learning_rate": 1.8771826024025944e-05, | |
| "loss": 0.7421, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 28.473023146202923, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.4279, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2662337662337662, | |
| "grad_norm": 13.106421492490322, | |
| "learning_rate": 1.8544194045464888e-05, | |
| "loss": 0.3088, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 5.104577671097258, | |
| "learning_rate": 1.8423706192694118e-05, | |
| "loss": 0.2931, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2792207792207792, | |
| "grad_norm": 52.64196123136193, | |
| "learning_rate": 1.8298852919990254e-05, | |
| "loss": 0.4912, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 31.65156603514557, | |
| "learning_rate": 1.816969893010442e-05, | |
| "loss": 0.3692, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2922077922077922, | |
| "grad_norm": 17.666169589586605, | |
| "learning_rate": 1.8036311154549783e-05, | |
| "loss": 0.2944, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2987012987012987, | |
| "grad_norm": 38.35439518358878, | |
| "learning_rate": 1.789875871891559e-05, | |
| "loss": 0.5252, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3051948051948052, | |
| "grad_norm": 39.96972433176561, | |
| "learning_rate": 1.77571129070442e-05, | |
| "loss": 0.5614, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 19.4739183485063, | |
| "learning_rate": 1.761144712408965e-05, | |
| "loss": 0.3101, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 2122.177114683787, | |
| "learning_rate": 1.7461836858476858e-05, | |
| "loss": 8.3024, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 53.72514322304979, | |
| "learning_rate": 1.730835964278124e-05, | |
| "loss": 0.5558, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.33116883116883117, | |
| "grad_norm": 53.70022188889065, | |
| "learning_rate": 1.7151095013548996e-05, | |
| "loss": 0.5592, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33766233766233766, | |
| "grad_norm": 33.4569293851714, | |
| "learning_rate": 1.699012447007882e-05, | |
| "loss": 0.3671, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.34415584415584416, | |
| "grad_norm": 17.137073429830238, | |
| "learning_rate": 1.6825531432186545e-05, | |
| "loss": 0.3092, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.35064935064935066, | |
| "grad_norm": 22.60309242557496, | |
| "learning_rate": 1.6657401196974405e-05, | |
| "loss": 0.3291, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 8.833069765914521, | |
| "learning_rate": 1.648582089462756e-05, | |
| "loss": 0.2884, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 42.48152192562097, | |
| "learning_rate": 1.631087944326053e-05, | |
| "loss": 0.4151, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.37012987012987014, | |
| "grad_norm": 51.1875826358593, | |
| "learning_rate": 1.6132667502837164e-05, | |
| "loss": 0.4914, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.37662337662337664, | |
| "grad_norm": 29.57316461617743, | |
| "learning_rate": 1.59512774281879e-05, | |
| "loss": 0.3513, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.38311688311688313, | |
| "grad_norm": 1.6690431301081146, | |
| "learning_rate": 1.5766803221148676e-05, | |
| "loss": 0.279, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 17.11791002555302, | |
| "learning_rate": 1.5579340481846338e-05, | |
| "loss": 0.3062, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3961038961038961, | |
| "grad_norm": 2.9548663461162916, | |
| "learning_rate": 1.538898635915576e-05, | |
| "loss": 0.2837, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4025974025974026, | |
| "grad_norm": 30.32462862806025, | |
| "learning_rate": 1.5195839500354337e-05, | |
| "loss": 0.3502, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 27.26334803927827, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3433, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 25.006274002753365, | |
| "learning_rate": 1.4801569348059158e-05, | |
| "loss": 0.3237, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.42207792207792205, | |
| "grad_norm": 9.0889223900474, | |
| "learning_rate": 1.4600650377311523e-05, | |
| "loss": 0.2881, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 17.44533004191656, | |
| "learning_rate": 1.4397347210059059e-05, | |
| "loss": 0.3027, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.43506493506493504, | |
| "grad_norm": 19.97489878606359, | |
| "learning_rate": 1.4191765204166643e-05, | |
| "loss": 0.3013, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.44155844155844154, | |
| "grad_norm": 16.933568982459366, | |
| "learning_rate": 1.3984010898462417e-05, | |
| "loss": 0.3018, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.44805194805194803, | |
| "grad_norm": 29.645129941175437, | |
| "learning_rate": 1.3774191957526144e-05, | |
| "loss": 0.3391, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 9.80114973140799, | |
| "learning_rate": 1.356241711589417e-05, | |
| "loss": 0.2963, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.461038961038961, | |
| "grad_norm": 4.338277899349995, | |
| "learning_rate": 1.3348796121709862e-05, | |
| "loss": 0.2732, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 22.540725434050554, | |
| "learning_rate": 1.3133439679848824e-05, | |
| "loss": 0.3253, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.474025974025974, | |
| "grad_norm": 6.180079243988936, | |
| "learning_rate": 1.291645939454825e-05, | |
| "loss": 0.2838, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4805194805194805, | |
| "grad_norm": 9.944164489572458, | |
| "learning_rate": 1.2697967711570243e-05, | |
| "loss": 0.2912, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.487012987012987, | |
| "grad_norm": 22.006231698484978, | |
| "learning_rate": 1.2478077859929e-05, | |
| "loss": 0.3109, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4935064935064935, | |
| "grad_norm": 0.9993376662904194, | |
| "learning_rate": 1.2256903793212107e-05, | |
| "loss": 0.2769, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 6.736485919129092, | |
| "learning_rate": 1.2034560130526341e-05, | |
| "loss": 0.2776, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5064935064935064, | |
| "grad_norm": 1.5505988858577997, | |
| "learning_rate": 1.1811162097098559e-05, | |
| "loss": 0.259, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.512987012987013, | |
| "grad_norm": 19.317012183505398, | |
| "learning_rate": 1.1586825464562515e-05, | |
| "loss": 0.3168, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 24.98335565690831, | |
| "learning_rate": 1.1361666490962468e-05, | |
| "loss": 0.361, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.525974025974026, | |
| "grad_norm": 15.391559490756684, | |
| "learning_rate": 1.113580186050475e-05, | |
| "loss": 0.3048, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5324675324675324, | |
| "grad_norm": 4.355627253302395, | |
| "learning_rate": 1.0909348623088472e-05, | |
| "loss": 0.28, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.538961038961039, | |
| "grad_norm": 17.405555596408444, | |
| "learning_rate": 1.0682424133646712e-05, | |
| "loss": 0.2902, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 13.407085055385325, | |
| "learning_rate": 1.0455145991329639e-05, | |
| "loss": 0.2765, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.551948051948052, | |
| "grad_norm": 2.6139015199788034, | |
| "learning_rate": 1.0227631978561057e-05, | |
| "loss": 0.2785, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5584415584415584, | |
| "grad_norm": 12.810695157873214, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2845, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.564935064935065, | |
| "grad_norm": 8.816846652405896, | |
| "learning_rate": 9.772368021438943e-06, | |
| "loss": 0.2827, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 21.911718407785227, | |
| "learning_rate": 9.544854008670366e-06, | |
| "loss": 0.3217, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.577922077922078, | |
| "grad_norm": 12.550189539530098, | |
| "learning_rate": 9.317575866353293e-06, | |
| "loss": 0.292, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 15.167023999667723, | |
| "learning_rate": 9.090651376911532e-06, | |
| "loss": 0.3001, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5909090909090909, | |
| "grad_norm": 7.973065612261872, | |
| "learning_rate": 8.86419813949525e-06, | |
| "loss": 0.2886, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5974025974025974, | |
| "grad_norm": 16.15242808963043, | |
| "learning_rate": 8.638333509037537e-06, | |
| "loss": 0.299, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6038961038961039, | |
| "grad_norm": 18.179063200373662, | |
| "learning_rate": 8.413174535437486e-06, | |
| "loss": 0.3178, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6103896103896104, | |
| "grad_norm": 17.88073233317887, | |
| "learning_rate": 8.188837902901441e-06, | |
| "loss": 0.3026, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6168831168831169, | |
| "grad_norm": 17.894047190834744, | |
| "learning_rate": 7.965439869473664e-06, | |
| "loss": 0.3063, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.9912752221363975, | |
| "learning_rate": 7.743096206787894e-06, | |
| "loss": 0.2861, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6298701298701299, | |
| "grad_norm": 12.59974964084945, | |
| "learning_rate": 7.521922140071003e-06, | |
| "loss": 0.2873, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 5.247248714404641, | |
| "learning_rate": 7.3020322884297565e-06, | |
| "loss": 0.2772, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 15.713529017302493, | |
| "learning_rate": 7.0835406054517505e-06, | |
| "loss": 0.2954, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 14.670526789277048, | |
| "learning_rate": 6.866560320151179e-06, | |
| "loss": 0.2928, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6558441558441559, | |
| "grad_norm": 1.4479761532269853, | |
| "learning_rate": 6.651203878290139e-06, | |
| "loss": 0.2747, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6623376623376623, | |
| "grad_norm": 8.546550504285172, | |
| "learning_rate": 6.437582884105835e-06, | |
| "loss": 0.2796, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6688311688311688, | |
| "grad_norm": 5.205487389970545, | |
| "learning_rate": 6.225808042473857e-06, | |
| "loss": 0.2645, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 10.186260022708211, | |
| "learning_rate": 6.015989101537586e-06, | |
| "loss": 0.2813, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 8.851239912787552, | |
| "learning_rate": 5.8082347958333625e-06, | |
| "loss": 0.2801, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6883116883116883, | |
| "grad_norm": 0.47334586479886903, | |
| "learning_rate": 5.602652789940941e-06, | |
| "loss": 0.2814, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6948051948051948, | |
| "grad_norm": 10.137248782728967, | |
| "learning_rate": 5.399349622688479e-06, | |
| "loss": 0.2771, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7012987012987013, | |
| "grad_norm": 14.23493444448787, | |
| "learning_rate": 5.198430651940846e-06, | |
| "loss": 0.3031, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7077922077922078, | |
| "grad_norm": 25.77828835781949, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.3317, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 4.983254834642921, | |
| "learning_rate": 4.804160499645667e-06, | |
| "loss": 0.284, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7207792207792207, | |
| "grad_norm": 3.210921791733905, | |
| "learning_rate": 4.611013640844245e-06, | |
| "loss": 0.2656, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 12.882476468987955, | |
| "learning_rate": 4.420659518153667e-06, | |
| "loss": 0.2952, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7337662337662337, | |
| "grad_norm": 8.496092897870577, | |
| "learning_rate": 4.2331967788513295e-06, | |
| "loss": 0.2894, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7402597402597403, | |
| "grad_norm": 14.782117212880996, | |
| "learning_rate": 4.048722571812105e-06, | |
| "loss": 0.2943, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7467532467532467, | |
| "grad_norm": 8.221163363530726, | |
| "learning_rate": 3.867332497162836e-06, | |
| "loss": 0.2717, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7532467532467533, | |
| "grad_norm": 3.319249243011192, | |
| "learning_rate": 3.689120556739475e-06, | |
| "loss": 0.2675, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7597402597402597, | |
| "grad_norm": 5.8378264660336665, | |
| "learning_rate": 3.5141791053724405e-06, | |
| "loss": 0.2891, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7662337662337663, | |
| "grad_norm": 7.759679103047117, | |
| "learning_rate": 3.342598803025595e-06, | |
| "loss": 0.2742, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7727272727272727, | |
| "grad_norm": 3.391758679327134, | |
| "learning_rate": 3.174468567813461e-06, | |
| "loss": 0.2828, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 4.212349972665664, | |
| "learning_rate": 3.009875529921181e-06, | |
| "loss": 0.2702, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 16.260452235360823, | |
| "learning_rate": 2.8489049864510053e-06, | |
| "loss": 0.2843, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7922077922077922, | |
| "grad_norm": 12.65271768504945, | |
| "learning_rate": 2.691640357218759e-06, | |
| "loss": 0.2767, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7987012987012987, | |
| "grad_norm": 7.533838996842269, | |
| "learning_rate": 2.5381631415231455e-06, | |
| "loss": 0.2646, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8051948051948052, | |
| "grad_norm": 2.97864671142113, | |
| "learning_rate": 2.388552875910354e-06, | |
| "loss": 0.2922, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8116883116883117, | |
| "grad_norm": 0.3722989784490546, | |
| "learning_rate": 2.2428870929558012e-06, | |
| "loss": 0.2836, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.7184053225638433, | |
| "learning_rate": 2.101241281084416e-06, | |
| "loss": 0.2809, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8246753246753247, | |
| "grad_norm": 7.043657965270735, | |
| "learning_rate": 1.963688845450218e-06, | |
| "loss": 0.2745, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 10.025849594465186, | |
| "learning_rate": 1.8303010698955803e-06, | |
| "loss": 0.2933, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8376623376623377, | |
| "grad_norm": 3.4255662736765844, | |
| "learning_rate": 1.7011470800097496e-06, | |
| "loss": 0.279, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 9.520856580468687, | |
| "learning_rate": 1.5762938073058853e-06, | |
| "loss": 0.2927, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8506493506493507, | |
| "grad_norm": 4.373148084541858, | |
| "learning_rate": 1.4558059545351144e-06, | |
| "loss": 0.2723, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 5.775208392633576, | |
| "learning_rate": 1.339745962155613e-06, | |
| "loss": 0.2787, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8636363636363636, | |
| "grad_norm": 2.1662144441766635, | |
| "learning_rate": 1.2281739759740575e-06, | |
| "loss": 0.2775, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8701298701298701, | |
| "grad_norm": 4.505056971511744, | |
| "learning_rate": 1.121147815976248e-06, | |
| "loss": 0.2818, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8766233766233766, | |
| "grad_norm": 5.227972464338057, | |
| "learning_rate": 1.01872294636304e-06, | |
| "loss": 0.2704, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 1.8227797095460878, | |
| "learning_rate": 9.209524468071096e-07, | |
| "loss": 0.2765, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8896103896103896, | |
| "grad_norm": 8.904887866554748, | |
| "learning_rate": 8.278869849454718e-07, | |
| "loss": 0.2752, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8961038961038961, | |
| "grad_norm": 3.2675223808297766, | |
| "learning_rate": 7.395747901219474e-07, | |
| "loss": 0.2721, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9025974025974026, | |
| "grad_norm": 1.0782893352353118, | |
| "learning_rate": 6.560616283932897e-07, | |
| "loss": 0.2666, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 10.237618635319794, | |
| "learning_rate": 5.77390778811796e-07, | |
| "loss": 0.279, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9155844155844156, | |
| "grad_norm": 7.524756342899654, | |
| "learning_rate": 5.036030109968082e-07, | |
| "loss": 0.2798, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.922077922077922, | |
| "grad_norm": 6.865627899284416, | |
| "learning_rate": 4.3473656400665256e-07, | |
| "loss": 0.2697, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.7216520448297371, | |
| "learning_rate": 3.708271265220087e-07, | |
| "loss": 0.2745, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 1.7017798820527004, | |
| "learning_rate": 3.119078183509372e-07, | |
| "loss": 0.2682, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9415584415584416, | |
| "grad_norm": 8.223255165977816, | |
| "learning_rate": 2.5800917326521013e-07, | |
| "loss": 0.2781, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.948051948051948, | |
| "grad_norm": 9.206762539208919, | |
| "learning_rate": 2.091591231767709e-07, | |
| "loss": 0.2708, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9545454545454546, | |
| "grad_norm": 4.056391637422524, | |
| "learning_rate": 1.6538298366257975e-07, | |
| "loss": 0.2666, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.961038961038961, | |
| "grad_norm": 4.623116122211878, | |
| "learning_rate": 1.2670344084530384e-07, | |
| "loss": 0.2809, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9675324675324676, | |
| "grad_norm": 0.5191836726417989, | |
| "learning_rate": 9.314053963669245e-08, | |
| "loss": 0.2679, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 3.9295463971723485, | |
| "learning_rate": 6.471167334968887e-08, | |
| "loss": 0.2691, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9805194805194806, | |
| "grad_norm": 1.6238761795878893, | |
| "learning_rate": 4.143157468468717e-08, | |
| "loss": 0.276, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 8.527985399139007, | |
| "learning_rate": 2.3312308094607382e-08, | |
| "loss": 0.2796, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9935064935064936, | |
| "grad_norm": 10.628431594177156, | |
| "learning_rate": 1.0363263532724433e-08, | |
| "loss": 0.2645, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.0009395147341795, | |
| "learning_rate": 2.591151586508467e-09, | |
| "loss": 0.2856, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 154, | |
| "total_flos": 98781319004160.0, | |
| "train_loss": 1.0868141991751534, | |
| "train_runtime": 5317.4783, | |
| "train_samples_per_second": 1.852, | |
| "train_steps_per_second": 0.029 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 154, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 98781319004160.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |