| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 6.233467267119553, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.8108, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 5.904196326876289, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.7847, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.23851753064735, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8192, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.415128171631446, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.7829, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.04380559994215, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7687, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2587637623659984, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7336, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.046066080232856, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7165, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.069511156688382, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7287, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.2611185393960005, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.7252, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.260307749363621, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6916, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.5860528236126874, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6714, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.0408678402576386, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6509, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.7405635844975456, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6462, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.089378409802929, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6075, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.5136568318392694, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.6198, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.1263674006123536, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.5743, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0303264886226746, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5916, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.9294082822832469, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5856, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.947170872331032, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5385, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.8518591702550199, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.568, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.8725288325536481, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5631, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.0592322309688051, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.5431, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7594966772209739, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5403, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.0147100967231533, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5346, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8092853991109913, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5391, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.7480207959415263, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5183, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6725540446498556, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5211, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.752524063992389, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5116, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8714792547605933, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.5252, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.587686864084582, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.5117, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7359163335855232, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.5165, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.683074670763304, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.5302, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7899607579501703, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.5131, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7203007550703167, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.4912, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.0497807505193113, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.5064, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.1723249764639616, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.5085, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.8733264133015174, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.5093, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.2145299440192203, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.5151, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.8295361766624695, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4846, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.8655831204630615, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.4902, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.7831772024924701, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.499, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.084114214652019, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4961, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9639426639425154, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.5058, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.6394336440537755, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4806, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7819635417728371, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.5032, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7955295880993887, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.5021, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.6671700101502559, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4857, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.5514514988706577, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.4747, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.6556176731195519, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4726, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6240651998663801, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4884, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.5551386581113544, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4906, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.5904946318868938, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4821, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.538919779997332, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4823, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5333470534988501, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.4719, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.4850010563636148, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4665, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.44539126503300186, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4743, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5032408043020593, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4723, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.46974810278602536, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4726, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.4608092351714356, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4671, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.45516143398106607, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.466, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.4744623675372948, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4741, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.43856795287418615, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4601, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.3761545034457208, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4706, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.41810263359022676, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.467, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.40583434815588343, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4695, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.3914490774493594, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4748, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.389130816543567, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4537, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.4581445439882817, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4586, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5126383880525731, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4724, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5181747207489492, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4674, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.4295166893762927, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4508, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.4868403059192456, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4655, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5989861771380917, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4661, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5656347544570441, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4803, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.46441753698075355, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4639, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.510393828467784, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4708, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.48942431176760787, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.4524, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.4968874160759365, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4643, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.8861501634736185, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.8012, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.231652143097992, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4398, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6712625292151638, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4121, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.1356345596864996, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4285, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.6662175711028467, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4184, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8930266388840874, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4219, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.5713183757359342, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.4266, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6446930790537315, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4306, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5449251995753388, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4363, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.49840540333271255, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4198, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.636524520280604, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.427, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5763577377012762, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.434, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5518205698347075, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.4081, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5685776152816718, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4277, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6757526217251246, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4176, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.536729337543259, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4153, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.42544799934531624, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4172, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.42010888741086944, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4102, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.43678400955846786, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.4161, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5082670907410142, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4197, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.357137321779293, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4048, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.40817714129576205, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.419, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.47645094237883096, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.4417, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.39257867572254557, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4215, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.4206670690315583, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4201, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.3865293036754881, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4347, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.39367421741174213, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4206, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.4656673614605536, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.3881, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.35335946805147883, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4118, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.4628742760548775, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.416, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.39568236935918416, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4135, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5466276296066325, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.417, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5090028901174583, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4111, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.48855888787545754, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4036, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.45101035460617356, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4207, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5737035543815443, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4089, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.39146919766674915, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4164, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.3630072800119681, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.3981, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4547222351106011, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4196, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.37216593927043445, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4107, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4050307805559563, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4122, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.38269243213429416, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4073, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.36260299781848226, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4252, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4438690588669624, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4122, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.41544202097631083, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.416, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.641936060850495, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4205, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5487268480206667, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4185, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.3876577101511856, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4103, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.42818180937132194, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4205, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.2680370619157418, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.4256, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.39573261758619904, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4168, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.3543678918298844, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.411, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.3717876519061496, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4166, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.41594670318211513, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4185, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.39133110888303313, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4139, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.43836287131270635, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4159, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.3757188844435322, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4196, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.34971023678161894, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.3985, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.3512090580148879, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4073, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.34629713258117534, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4201, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.46977338380502903, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4087, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.40101157445187413, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4147, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.47691626196033376, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4004, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.3435761452724062, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4126, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.3645128134974153, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4007, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.35491499023320583, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4087, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.3734869903646821, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.419, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.30082037154900365, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4138, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4590029162532056, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4081, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.29447660882935095, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4009, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.48573583205938736, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4258, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3631988369037435, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4203, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.3332268388037149, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4006, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.35727991875553283, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4094, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3559631224614753, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4213, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.37804917682629285, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4199, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.37950577893585047, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.3961, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.32608515028658464, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4055, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.7669011198805715, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.7441, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6652887388218975, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3507, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.7385081279732356, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3736, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7320172922150582, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3587, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.49807265418424446, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3623, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.48163598394041085, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3659, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.48731905853762586, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3476, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.42827838103728594, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3553, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.34579240618250967, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.36, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.38008557988183295, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3481, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.3376533317787802, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3645, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.348488923971927, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3635, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.3599933955892379, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3542, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.333690464884174, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3556, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.3350844064785929, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3647, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.37148148955239746, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3554, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.3280010701875189, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3532, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.332491142609001, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3544, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.2881390860119507, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3455, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.32116573700586765, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3621, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.30659625060436463, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3419, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.30519135259940783, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3524, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.2980472772642158, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3493, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.32253352509389444, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3587, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.27321091859832713, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.35, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.2985443161344337, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3582, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.323527392124704, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3559, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.2948657008807223, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3568, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.2991813048805828, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3532, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.2783668607947485, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3458, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.28042445346441125, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3421, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.3471962042525829, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3453, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.2823227915730697, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3432, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.3530570165142917, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3548, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.3056427441596844, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3613, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3086741723859052, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3417, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.31428784464812703, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3581, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.2859896069564287, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3474, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.27059096851688597, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3442, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.3326884769262199, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3683, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.29505923783406746, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3477, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.32073943104242425, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3595, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.2840222270592631, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3473, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3418714380147271, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3521, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.2797908282100119, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3543, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.36420961265518004, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3533, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.314568754851869, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3575, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3342111687945849, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.346, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.3064734258766498, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3671, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.29854230408342164, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3424, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.2827054055912993, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3526, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.28275958845846305, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3441, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.312064605803253, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3598, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.28452488535848225, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3595, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.2744945291582551, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3514, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.2851263975583895, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3578, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.24934062786332797, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3658, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.32531564637396143, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3475, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.24195021899743993, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3509, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.33103816433572836, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3541, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.2269071264266118, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3602, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.29798422862653967, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3433, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.25616892838087085, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3547, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.2569631346633302, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3678, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.32778475734128903, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3523, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.2369726285253331, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3449, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.2810512494548086, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3631, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.24745825973149008, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3526, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.2586808546029739, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3428, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.23360130898404668, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.345, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.240900500337055, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3487, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.23747073767970822, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.356, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.2201875141981078, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3578, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.23449074731345043, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3619, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.23953740851380273, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3587, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.20232912490254967, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3447, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.25646246387237664, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.359, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.22729675252261325, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3487, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.6215353715874271, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6241, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.3696909872798802, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.3034, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5375575773965741, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.3032, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.38516960983552195, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.3061, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.4742991871870546, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2937, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4135969453525089, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.3138, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.42172855815907073, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.308, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.3771894608374879, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.3076, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4163462974642452, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.3071, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3704830350749835, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.3056, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.36115694049544883, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.3047, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.3358650426314368, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2925, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3744792395559403, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2965, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.3361916360656167, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.3127, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.3091988714486053, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.3035, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3277119477509228, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.299, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.2873917613244393, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2928, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.28358641106288984, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.291, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.27138723040994367, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2888, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.2753898544881945, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2953, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.2647672143999282, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2912, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.2879829562601937, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.3111, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.2685828458948742, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2899, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.2618038733902564, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2916, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.22753375801417833, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.305, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.2930244026853143, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.3039, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.22977159381404028, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2878, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.24270939761725097, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2949, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.24447085983026712, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2993, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.23152423854583742, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2944, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.22985233893166343, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2955, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.23372953797080312, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.288, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.21690130609713476, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2904, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2246463430594481, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2989, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.21332209794451956, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2975, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.24970279498059386, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.3064, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.21261729326140266, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2862, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.22206764450175856, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.3092, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.24047571831325013, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2971, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.21849137758843346, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2984, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.2229834366517268, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2961, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.22968975608492073, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.3033, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.22151889834022945, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.3039, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2128895075658792, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2943, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.21164754242909364, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2919, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.20884461582582384, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2997, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.21541996842202368, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.3002, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.20772455990874986, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2921, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.21297651398483633, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.311, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2190112768874125, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2919, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.21673928683300617, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2841, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.21487488035766797, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.3014, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.21578163765129071, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.3019, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.224427277921113, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2976, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.22042655864613764, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2989, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.20374940342407785, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2923, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2114696586781579, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2969, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.20549511788083236, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.3061, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.19979906515975762, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2823, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.19229453116296086, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2997, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.20712539472061123, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2959, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2015468139758988, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2979, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.20097170185132207, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2926, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.19852055662210327, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2913, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.18906994406437755, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.3006, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2021153809962757, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2923, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.1999899974233003, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2915, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.19630347526748365, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2866, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.1920809980534528, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.3032, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.1901195320725009, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2909, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.21032225123095197, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.3084, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.1868686917904852, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2905, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.18502790597756422, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2937, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.1928945785679494, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2989, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.19560615341315463, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.3126, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.18840138055047603, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2934, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.19588915112265848, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2905, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.19730007784376413, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.3044, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.572586285609381, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.5472, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.2949768631584251, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2704, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.2574596792309237, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2549, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.21477368558781973, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2621, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.2832750553273873, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2838, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.31022611032250574, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2622, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2628883778841394, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.2482, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.25960167507018217, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2637, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.23079097623497014, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2583, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.23734069583525863, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2636, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.24231979125793324, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2677, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.222245031459651, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.2611, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.20312084213998194, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.2585, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.197752485913063, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.259, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2026654928471488, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2514, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.2226046433033049, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2717, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.19738358880511364, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.2568, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.20518193085549377, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.2593, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2010784131419505, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2672, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.2040581583418926, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.2672, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.2051153804496881, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2577, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.19036151835844245, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2608, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.1878364416043157, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2523, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.19031269408103865, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.2545, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.18967066362788515, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2589, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.18796101285613187, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.264, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.18728077492824852, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.275, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.1920736215287993, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2668, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.18710726206783124, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.2626, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.18070835403401606, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.2539, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.18040022199428507, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.2683, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.18024941762685995, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2629, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.17679299731428164, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2574, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.17535840405973557, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.2583, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.17209557575763554, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2538, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.17269823781217156, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2573, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.17569630903609573, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2636, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.17222091170740247, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.2564, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.17986994282527735, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2658, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.17749306159579448, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2642, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.17369779306753141, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2585, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.1762932829589361, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2676, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.17826711105619464, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.2707, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.17362813852251957, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2622, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.17946540544791956, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.262, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.17382409916390584, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2547, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.17625896763947824, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2623, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.1699245760301564, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2694, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.16895611272526004, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2575, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.1693241796677865, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2459, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.17362449206981356, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2716, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.17025045249936588, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.2534, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.16875385878273286, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2586, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.17389976665994425, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2645, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.17023072304351422, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2569, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.17186663723394052, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2643, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.1680634909255728, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2623, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.17120199576002865, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2613, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.16412195443224825, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.2589, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.17379655511927136, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2632, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.1656832409615132, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2478, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.16951075627616047, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.2516, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.17537557692366856, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.2676, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.17374885403540272, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2605, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.16888552813586052, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.2718, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.17354897645874484, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.2706, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.1700437193077696, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.2622, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.16925795308526098, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2532, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.17169956156014887, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2543, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.16991884248389066, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.254, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.17543705217318195, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2678, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.17309881142676284, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2528, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.16669028171466665, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.2568, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.16908323842167242, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2625, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.16654126250710174, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2552, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.16486951649258705, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.2553, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.16746724674913815, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2546, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.17103013097653133, | |
| "learning_rate": 0.0, | |
| "loss": 0.268, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.8773457441390592e+18, | |
| "train_loss": 0.3770592579474816, | |
| "train_runtime": 19616.6723, | |
| "train_samples_per_second": 2.549, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 1.8773457441390592e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |