| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.523809523809524, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 57.57447834029504, | |
| "learning_rate": 6.000000000000001e-08, | |
| "loss": 4.0982, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 57.2163757354146, | |
| "learning_rate": 1.2000000000000002e-07, | |
| "loss": 3.8959, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 54.21668090486469, | |
| "learning_rate": 1.8e-07, | |
| "loss": 4.0292, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 58.33661945400434, | |
| "learning_rate": 2.4000000000000003e-07, | |
| "loss": 4.1284, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 63.01395357482665, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 4.188, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 57.71828103888573, | |
| "learning_rate": 3.6e-07, | |
| "loss": 3.9705, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 51.089655348018326, | |
| "learning_rate": 4.2000000000000006e-07, | |
| "loss": 3.8838, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 58.15717323097846, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 4.1963, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 56.91779340921828, | |
| "learning_rate": 5.4e-07, | |
| "loss": 3.9283, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 51.10490139053562, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 3.8356, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 52.268827098443424, | |
| "learning_rate": 6.6e-07, | |
| "loss": 3.7757, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 51.80062279098391, | |
| "learning_rate": 7.2e-07, | |
| "loss": 3.9636, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 44.40217995730599, | |
| "learning_rate": 7.8e-07, | |
| "loss": 3.6742, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 44.82724671161504, | |
| "learning_rate": 8.400000000000001e-07, | |
| "loss": 3.768, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 42.24923860328258, | |
| "learning_rate": 9e-07, | |
| "loss": 3.8381, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 42.90764521769713, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 3.7314, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 45.24513919191596, | |
| "learning_rate": 1.0200000000000002e-06, | |
| "loss": 3.8718, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 31.63482178736033, | |
| "learning_rate": 1.08e-06, | |
| "loss": 3.3766, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 32.21560942157832, | |
| "learning_rate": 1.14e-06, | |
| "loss": 3.3699, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 32.952289793686546, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 3.4645, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 29.56293796088727, | |
| "learning_rate": 1.26e-06, | |
| "loss": 3.2721, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 28.09762343844428, | |
| "learning_rate": 1.32e-06, | |
| "loss": 3.1558, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 27.621733664600917, | |
| "learning_rate": 1.3800000000000001e-06, | |
| "loss": 3.2164, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 18.800357074717578, | |
| "learning_rate": 1.44e-06, | |
| "loss": 3.0664, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 13.800920246917714, | |
| "learning_rate": 1.5e-06, | |
| "loss": 2.7536, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 13.61642291684508, | |
| "learning_rate": 1.56e-06, | |
| "loss": 2.4849, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 13.43480789920014, | |
| "learning_rate": 1.6200000000000002e-06, | |
| "loss": 2.6918, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 14.095592842222162, | |
| "learning_rate": 1.6800000000000002e-06, | |
| "loss": 2.6273, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 12.620058865170318, | |
| "learning_rate": 1.7399999999999999e-06, | |
| "loss": 2.5829, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 9.873668620990621, | |
| "learning_rate": 1.8e-06, | |
| "loss": 2.6698, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 10.645995976559039, | |
| "learning_rate": 1.86e-06, | |
| "loss": 2.7849, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 2.0317460317460316, | |
| "grad_norm": 10.770458724192382, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 2.6262, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.0952380952380953, | |
| "grad_norm": 9.288284603719239, | |
| "learning_rate": 1.98e-06, | |
| "loss": 2.405, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 2.1587301587301586, | |
| "grad_norm": 11.040656078736276, | |
| "learning_rate": 2.0400000000000004e-06, | |
| "loss": 2.3917, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 9.014452408383601, | |
| "learning_rate": 2.1e-06, | |
| "loss": 2.2996, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 10.977152245763714, | |
| "learning_rate": 2.16e-06, | |
| "loss": 2.4617, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.3492063492063493, | |
| "grad_norm": 11.961573908046798, | |
| "learning_rate": 2.22e-06, | |
| "loss": 2.4845, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 2.4126984126984126, | |
| "grad_norm": 12.506586582376126, | |
| "learning_rate": 2.28e-06, | |
| "loss": 2.2293, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.4761904761904763, | |
| "grad_norm": 15.888921935838896, | |
| "learning_rate": 2.34e-06, | |
| "loss": 2.2445, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 2.5396825396825395, | |
| "grad_norm": 17.724634829521865, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 2.5632, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.6031746031746033, | |
| "grad_norm": 16.577971095615666, | |
| "learning_rate": 2.4599999999999997e-06, | |
| "loss": 1.9954, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 20.04871359977933, | |
| "learning_rate": 2.52e-06, | |
| "loss": 1.8782, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.7301587301587302, | |
| "grad_norm": 16.171304753721177, | |
| "learning_rate": 2.58e-06, | |
| "loss": 1.9382, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 2.7936507936507935, | |
| "grad_norm": 19.412153353563298, | |
| "learning_rate": 2.64e-06, | |
| "loss": 1.752, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 17.206107408812862, | |
| "learning_rate": 2.7e-06, | |
| "loss": 1.9237, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.9206349206349205, | |
| "grad_norm": 20.99711783167552, | |
| "learning_rate": 2.7600000000000003e-06, | |
| "loss": 1.5944, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 2.984126984126984, | |
| "grad_norm": 17.220612862140708, | |
| "learning_rate": 2.82e-06, | |
| "loss": 1.6411, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 3.0476190476190474, | |
| "grad_norm": 26.644954169593106, | |
| "learning_rate": 2.88e-06, | |
| "loss": 1.4512, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 65.10523878873882, | |
| "learning_rate": 2.9400000000000002e-06, | |
| "loss": 1.6913, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 3.1746031746031744, | |
| "grad_norm": 27.458331836448174, | |
| "learning_rate": 3e-06, | |
| "loss": 1.6158, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.238095238095238, | |
| "grad_norm": 17.879988381877414, | |
| "learning_rate": 2.9992598405485977e-06, | |
| "loss": 1.6304, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 3.3015873015873014, | |
| "grad_norm": 20.512906161569692, | |
| "learning_rate": 2.9970400926424076e-06, | |
| "loss": 1.3412, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 3.365079365079365, | |
| "grad_norm": 28.162531827787188, | |
| "learning_rate": 2.99334294690462e-06, | |
| "loss": 1.308, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 20.08257176199191, | |
| "learning_rate": 2.988172051971717e-06, | |
| "loss": 1.2513, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 3.492063492063492, | |
| "grad_norm": 22.419438866388326, | |
| "learning_rate": 2.981532510892707e-06, | |
| "loss": 1.3811, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 18.961751404248563, | |
| "learning_rate": 2.9734308760930334e-06, | |
| "loss": 1.2914, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.619047619047619, | |
| "grad_norm": 37.34845292165996, | |
| "learning_rate": 2.9638751429081213e-06, | |
| "loss": 1.2351, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 3.682539682539683, | |
| "grad_norm": 14.747371601837061, | |
| "learning_rate": 2.9528747416929465e-06, | |
| "loss": 1.2103, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 3.746031746031746, | |
| "grad_norm": 17.04660819603052, | |
| "learning_rate": 2.9404405285154148e-06, | |
| "loss": 1.235, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 3.8095238095238093, | |
| "grad_norm": 19.420464839345144, | |
| "learning_rate": 2.9265847744427307e-06, | |
| "loss": 1.1033, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 3.873015873015873, | |
| "grad_norm": 35.5290042332465, | |
| "learning_rate": 2.9113211534313383e-06, | |
| "loss": 1.2773, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 3.9365079365079367, | |
| "grad_norm": 22.266882915667832, | |
| "learning_rate": 2.894664728832377e-06, | |
| "loss": 1.1921, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 15.491496380357695, | |
| "learning_rate": 2.8766319385259716e-06, | |
| "loss": 1.0918, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 4.063492063492063, | |
| "grad_norm": 14.926456699177113, | |
| "learning_rate": 2.8572405786990296e-06, | |
| "loss": 0.9075, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 4.1269841269841265, | |
| "grad_norm": 13.661102423305078, | |
| "learning_rate": 2.8365097862825516e-06, | |
| "loss": 1.0097, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 4.190476190476191, | |
| "grad_norm": 13.631384875740038, | |
| "learning_rate": 2.814460020065795e-06, | |
| "loss": 0.9625, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 4.253968253968254, | |
| "grad_norm": 14.239925114199261, | |
| "learning_rate": 2.7911130405059155e-06, | |
| "loss": 0.9898, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 4.317460317460317, | |
| "grad_norm": 12.77560062730983, | |
| "learning_rate": 2.7664918882530226e-06, | |
| "loss": 0.978, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.380952380952381, | |
| "grad_norm": 14.003703335394645, | |
| "learning_rate": 2.7406208614118425e-06, | |
| "loss": 0.988, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 13.375052129651001, | |
| "learning_rate": 2.713525491562421e-06, | |
| "loss": 0.8734, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.507936507936508, | |
| "grad_norm": 12.679615580914351, | |
| "learning_rate": 2.685232518563536e-06, | |
| "loss": 0.7357, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 12.990007533232859, | |
| "learning_rate": 2.655769864163684e-06, | |
| "loss": 0.8909, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 4.634920634920634, | |
| "grad_norm": 12.221083897238143, | |
| "learning_rate": 2.6251666044456895e-06, | |
| "loss": 0.78, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 4.698412698412699, | |
| "grad_norm": 15.252264298495929, | |
| "learning_rate": 2.5934529411321173e-06, | |
| "loss": 0.6989, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 4.761904761904762, | |
| "grad_norm": 13.687902392472823, | |
| "learning_rate": 2.5606601717798212e-06, | |
| "loss": 0.9173, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 4.825396825396825, | |
| "grad_norm": 16.092259560486735, | |
| "learning_rate": 2.526820658893033e-06, | |
| "loss": 0.8555, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 11.153841319781899, | |
| "learning_rate": 2.491967797985478e-06, | |
| "loss": 0.768, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 4.9523809523809526, | |
| "grad_norm": 16.24314620878438, | |
| "learning_rate": 2.456135984623035e-06, | |
| "loss": 0.6937, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 5.015873015873016, | |
| "grad_norm": 12.98652461918944, | |
| "learning_rate": 2.419360580479465e-06, | |
| "loss": 0.8355, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 5.079365079365079, | |
| "grad_norm": 8.727731435832583, | |
| "learning_rate": 2.3816778784387097e-06, | |
| "loss": 0.6985, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 11.299395468739695, | |
| "learning_rate": 2.343125066778196e-06, | |
| "loss": 0.6097, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 5.2063492063492065, | |
| "grad_norm": 10.89035476736735, | |
| "learning_rate": 2.303740192468495e-06, | |
| "loss": 0.8029, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 5.26984126984127, | |
| "grad_norm": 8.148154849352979, | |
| "learning_rate": 2.263562123625557e-06, | |
| "loss": 0.5835, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 12.49282469909081, | |
| "learning_rate": 2.222630511152573e-06, | |
| "loss": 0.628, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 5.396825396825397, | |
| "grad_norm": 8.307586947822, | |
| "learning_rate": 2.18098574960932e-06, | |
| "loss": 0.7011, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 5.4603174603174605, | |
| "grad_norm": 8.270887155947076, | |
| "learning_rate": 2.138668937347609e-06, | |
| "loss": 0.5918, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 5.523809523809524, | |
| "grad_norm": 12.098343605915145, | |
| "learning_rate": 2.0957218359521707e-06, | |
| "loss": 0.4842, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 5.587301587301587, | |
| "grad_norm": 10.793266890734438, | |
| "learning_rate": 2.0521868290270174e-06, | |
| "loss": 0.7194, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 5.650793650793651, | |
| "grad_norm": 9.811217896188793, | |
| "learning_rate": 2.0081068803679374e-06, | |
| "loss": 0.5915, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 17.005853804161248, | |
| "learning_rate": 1.963525491562421e-06, | |
| "loss": 0.6159, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 11.50337527021372, | |
| "learning_rate": 1.918486659058844e-06, | |
| "loss": 0.6515, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 5.841269841269841, | |
| "grad_norm": 11.271416372518264, | |
| "learning_rate": 1.8730348307472826e-06, | |
| "loss": 0.4101, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 5.904761904761905, | |
| "grad_norm": 9.732208459925952, | |
| "learning_rate": 1.827214862094814e-06, | |
| "loss": 0.5937, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 5.968253968253968, | |
| "grad_norm": 9.882839220456333, | |
| "learning_rate": 1.7810719718785873e-06, | |
| "loss": 0.5382, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 6.031746031746032, | |
| "grad_norm": 10.962287605399686, | |
| "learning_rate": 1.7346516975603465e-06, | |
| "loss": 0.4762, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 6.095238095238095, | |
| "grad_norm": 8.430198784148699, | |
| "learning_rate": 1.6879998503464564e-06, | |
| "loss": 0.5713, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 6.158730158730159, | |
| "grad_norm": 10.894222790901619, | |
| "learning_rate": 1.6411624699777718e-06, | |
| "loss": 0.5565, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 12.878403484815633, | |
| "learning_rate": 1.5941857792939703e-06, | |
| "loss": 0.5058, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 6.577107386592493, | |
| "learning_rate": 1.5471161386171925e-06, | |
| "loss": 0.4028, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 6.349206349206349, | |
| "grad_norm": 8.332341230736613, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.4107, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.412698412698413, | |
| "grad_norm": 10.897392576231315, | |
| "learning_rate": 1.4528838613828075e-06, | |
| "loss": 0.4851, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 6.476190476190476, | |
| "grad_norm": 10.09022209906221, | |
| "learning_rate": 1.40581422070603e-06, | |
| "loss": 0.4957, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 6.5396825396825395, | |
| "grad_norm": 7.353237258425595, | |
| "learning_rate": 1.3588375300222285e-06, | |
| "loss": 0.4534, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 6.603174603174603, | |
| "grad_norm": 9.004695390709724, | |
| "learning_rate": 1.3120001496535434e-06, | |
| "loss": 0.4615, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 8.971252585562915, | |
| "learning_rate": 1.2653483024396534e-06, | |
| "loss": 0.3708, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 6.73015873015873, | |
| "grad_norm": 7.803315873016153, | |
| "learning_rate": 1.2189280281214128e-06, | |
| "loss": 0.3618, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 6.7936507936507935, | |
| "grad_norm": 10.1946500046286, | |
| "learning_rate": 1.1727851379051866e-06, | |
| "loss": 0.5127, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 8.988718449624637, | |
| "learning_rate": 1.1269651692527181e-06, | |
| "loss": 0.3902, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 6.920634920634921, | |
| "grad_norm": 7.493496463903085, | |
| "learning_rate": 1.0815133409411564e-06, | |
| "loss": 0.4069, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 6.984126984126984, | |
| "grad_norm": 8.348873531323628, | |
| "learning_rate": 1.036474508437579e-06, | |
| "loss": 0.3469, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.0476190476190474, | |
| "grad_norm": 9.038998481408308, | |
| "learning_rate": 9.918931196320629e-07, | |
| "loss": 0.3247, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 7.111111111111111, | |
| "grad_norm": 7.844047940277719, | |
| "learning_rate": 9.478131709729831e-07, | |
| "loss": 0.3509, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 7.174603174603175, | |
| "grad_norm": 7.425090910067537, | |
| "learning_rate": 9.042781640478293e-07, | |
| "loss": 0.3648, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 7.238095238095238, | |
| "grad_norm": 12.359869006545956, | |
| "learning_rate": 8.613310626523911e-07, | |
| "loss": 0.2861, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 7.301587301587301, | |
| "grad_norm": 11.80335907456322, | |
| "learning_rate": 8.190142503906799e-07, | |
| "loss": 0.4778, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 7.365079365079365, | |
| "grad_norm": 12.471156069815597, | |
| "learning_rate": 7.773694888474268e-07, | |
| "loss": 0.3658, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 9.243334578032606, | |
| "learning_rate": 7.36437876374443e-07, | |
| "loss": 0.3291, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 7.492063492063492, | |
| "grad_norm": 8.40873822616543, | |
| "learning_rate": 6.962598075315047e-07, | |
| "loss": 0.2889, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 7.555555555555555, | |
| "grad_norm": 10.437791126518906, | |
| "learning_rate": 6.568749332218045e-07, | |
| "loss": 0.4034, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 7.619047619047619, | |
| "grad_norm": 8.296256384898921, | |
| "learning_rate": 6.183221215612905e-07, | |
| "loss": 0.3604, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 7.682539682539683, | |
| "grad_norm": 8.052915673295223, | |
| "learning_rate": 5.806394195205356e-07, | |
| "loss": 0.3938, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 7.746031746031746, | |
| "grad_norm": 5.792626137273427, | |
| "learning_rate": 5.438640153769653e-07, | |
| "loss": 0.3031, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 7.809523809523809, | |
| "grad_norm": 7.495104357983482, | |
| "learning_rate": 5.080322020145225e-07, | |
| "loss": 0.3826, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 7.8730158730158735, | |
| "grad_norm": 8.81251429634198, | |
| "learning_rate": 4.731793411069669e-07, | |
| "loss": 0.292, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 7.936507936507937, | |
| "grad_norm": 8.034376369007889, | |
| "learning_rate": 4.3933982822017883e-07, | |
| "loss": 0.4526, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.333920405739056, | |
| "learning_rate": 4.06547058867883e-07, | |
| "loss": 0.3632, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 8.063492063492063, | |
| "grad_norm": 6.19175131287288, | |
| "learning_rate": 3.748333955543106e-07, | |
| "loss": 0.2562, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 8.126984126984127, | |
| "grad_norm": 6.481701827503514, | |
| "learning_rate": 3.442301358363163e-07, | |
| "loss": 0.237, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 8.19047619047619, | |
| "grad_norm": 6.479683967653911, | |
| "learning_rate": 3.147674814364644e-07, | |
| "loss": 0.2607, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 8.253968253968253, | |
| "grad_norm": 5.676011576423432, | |
| "learning_rate": 2.86474508437579e-07, | |
| "loss": 0.2384, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 8.317460317460318, | |
| "grad_norm": 5.757077383945261, | |
| "learning_rate": 2.593791385881571e-07, | |
| "loss": 0.3726, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 8.380952380952381, | |
| "grad_norm": 6.871067257863073, | |
| "learning_rate": 2.3350811174697772e-07, | |
| "loss": 0.3818, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 8.444444444444445, | |
| "grad_norm": 7.268914398833981, | |
| "learning_rate": 2.0888695949408471e-07, | |
| "loss": 0.4703, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 8.507936507936508, | |
| "grad_norm": 7.1532015341319575, | |
| "learning_rate": 1.8553997993420495e-07, | |
| "loss": 0.357, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 6.681075937364691, | |
| "learning_rate": 1.634902137174483e-07, | |
| "loss": 0.3128, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 8.634920634920634, | |
| "grad_norm": 7.434604156502297, | |
| "learning_rate": 1.4275942130097098e-07, | |
| "loss": 0.3583, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 8.698412698412698, | |
| "grad_norm": 6.334177324650447, | |
| "learning_rate": 1.2336806147402828e-07, | |
| "loss": 0.3273, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 8.761904761904763, | |
| "grad_norm": 6.460288148240497, | |
| "learning_rate": 1.0533527116762298e-07, | |
| "loss": 0.298, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 8.825396825396826, | |
| "grad_norm": 6.793693864896331, | |
| "learning_rate": 8.867884656866182e-08, | |
| "loss": 0.4737, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 6.087635250584677, | |
| "learning_rate": 7.341522555726971e-08, | |
| "loss": 0.2492, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 8.952380952380953, | |
| "grad_norm": 5.958472793551493, | |
| "learning_rate": 5.9559471484585404e-08, | |
| "loss": 0.2531, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 9.015873015873016, | |
| "grad_norm": 6.181482193201246, | |
| "learning_rate": 4.712525830705339e-08, | |
| "loss": 0.2074, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 9.079365079365079, | |
| "grad_norm": 4.95043049803299, | |
| "learning_rate": 3.6124857091878847e-08, | |
| "loss": 0.1696, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 5.624633357663993, | |
| "learning_rate": 2.6569123906967087e-08, | |
| "loss": 0.3333, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 9.206349206349206, | |
| "grad_norm": 6.112000874321065, | |
| "learning_rate": 1.846748910729351e-08, | |
| "loss": 0.4033, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 9.26984126984127, | |
| "grad_norm": 6.64513673651448, | |
| "learning_rate": 1.1827948028283353e-08, | |
| "loss": 0.3716, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 5.509458009190147, | |
| "learning_rate": 6.657053095380006e-09, | |
| "loss": 0.3588, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 9.396825396825397, | |
| "grad_norm": 5.1740170504268015, | |
| "learning_rate": 2.9599073575926614e-09, | |
| "loss": 0.2379, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 9.46031746031746, | |
| "grad_norm": 5.9471797504961, | |
| "learning_rate": 7.401594514026e-10, | |
| "loss": 0.2427, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 9.523809523809524, | |
| "grad_norm": 5.771578335180772, | |
| "learning_rate": 0.0, | |
| "loss": 0.2553, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 139054587576320.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |