| { | |
| "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": 8.025663346068319, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.1663, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 8.010365036638017, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.1727, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 7.843541941375245, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.167, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 7.421590803800207, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.1321, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 5.805442949718198, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.0654, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.2603146548909483, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.0231, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.4566327965356973, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9756, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 5.131605177002356, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 1.0153, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.906714594857795, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9494, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 5.191276044454782, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9364, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.859610945094299, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.9595, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.611776180080131, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8851, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 3.2071393973457853, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8609, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 3.3164237243549466, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8659, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.8241206428341763, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8637, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 2.5429042446588745, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8177, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 2.233797254080317, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.8268, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6997937530369438, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.8072, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.699841364037304, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7964, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.5492702146359056, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.8058, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.2836274496193323, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7692, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.3773948260948878, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.798, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.9779024320556644, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7898, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.2856224063212611, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7798, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.167576837058794, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7678, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.1982452842844535, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7704, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 1.4953948330166138, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7707, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.0238727274473236, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7768, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 1.1534610250721853, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7743, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.9055874647009523, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.7258, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.3165475640221869, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.7585, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9866053582540805, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7522, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.0543336745444751, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.7478, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.0227139967912582, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.7716, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.024288689780223, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.7303, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.9339962453032254, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.7317, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.0869869233699212, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.7695, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.0873092741588706, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.7452, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.031333961155868, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7233, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.0777106186591243, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.7303, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.012613449746541, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.7525, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.015287100824931, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.7431, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9547666584275336, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.7319, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.125422016781764, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.7368, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.0370344242184166, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.7734, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.9362145505319984, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.7214, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.8967046364088168, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.713, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.9460244206042299, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.7246, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.0085919950975422, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.7273, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8839138156298016, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.7378, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8111663747625724, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.7249, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.0259051345509014, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.7434, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.8301599918624412, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.7256, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9154124540951556, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.735, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7513948251693194, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.7343, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8351370147873389, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.7308, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.7560411621872986, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6877, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8386013917599205, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.7273, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7909573435703522, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.7151, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.048222420941934, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.7068, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8656447604390025, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.7468, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.9225813872947002, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.7259, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.0576195176304044, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.7256, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.9518126408204165, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6897, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.851324477460364, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.7522, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6716149073553244, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.7013, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.7482771712729901, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.7222, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6357576574196572, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6922, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.0644856785033119, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.723, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6405157264180997, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.7169, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.8419706953240476, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.7358, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6542849794196405, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.7195, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.8634302833607698, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.7203, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6567808913482877, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.7118, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.7812778861933127, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.7201, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7326117314540337, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.7116, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.6834658442763747, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6688, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.7736960730883061, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.7422, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.8791421881042536, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.6103, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.7675433065867431, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5395, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.9250645931902065, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5512, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.9096027266656481, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5287, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7763008273177653, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5327, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.9327758899718949, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5667, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.8127513945945763, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5238, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.7251658165407489, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5025, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 1.0593447852460571, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5418, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7502126366792881, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5436, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.9025217984099986, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5632, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.7752528081018466, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5698, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.8801407549146449, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5438, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.7496356441331226, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5192, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.8631634999670239, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5209, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.8499904547567155, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5391, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.8005857162187027, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5316, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.7725253831065256, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.532, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.8111181981066931, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5233, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.9389753171466569, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5075, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.644270337116294, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4967, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.9122151730709385, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5285, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.8354629879773563, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5268, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6812178009255956, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5123, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.7042512191630624, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5129, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.7299803805994697, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5343, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6871381008164456, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5059, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.6771983066014872, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5135, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.6556281104458144, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5435, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.7112832435326318, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5285, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6998462790858951, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5258, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.6339926050593321, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5046, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.8460983667609926, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5018, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.6412627311526433, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.548, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.8449945191687394, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4819, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.7266016409490449, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5382, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.8366889568978516, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5364, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.6935555814080242, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5231, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.8605730069954024, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4948, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.6444476399667671, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5073, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.793199587534131, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5191, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.822697091265982, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5037, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.6553802937542286, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.524, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.8048151807443502, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4966, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.6723206384765505, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5171, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.686453936800869, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5169, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.6470781206840747, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5159, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.6909611931420626, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5097, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.6666646530696388, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5226, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.6640563394076078, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5604, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.6512711394321913, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5032, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.6205754298591117, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5266, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.6072899721655831, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5188, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6149441351391679, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5357, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.7240301884950392, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5402, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.6201070184569217, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5015, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.7091061099254196, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5408, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.624733686451166, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5203, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.5983194782889787, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5119, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.6052490460873315, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5244, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.631071576498964, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5078, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.6333833828469582, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5029, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.6765084985059329, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5409, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.6874631191014754, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5275, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.646534216918847, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4959, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.622434488996995, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5297, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.6870636440715688, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5419, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.7008181875957741, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5061, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.6695818972080918, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5212, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.6249566453914562, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5178, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.6001095980182269, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5055, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.6343835646364906, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5164, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.5818556366158101, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5121, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.6685596112450095, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5255, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.6398404450965353, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5476, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.562220294385078, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5247, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.6575778041043532, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5159, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5998901587688095, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5145, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.8178696802765414, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.3839, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6287005638269078, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3389, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.2342897436763312, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3668, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.8044570182081753, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3334, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.733943874642229, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3329, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6995215819465724, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3387, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5957877635339981, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3182, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.6911074879178541, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3369, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6148675957794936, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3162, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.7359940694844278, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3371, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.543831434662558, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3161, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.7158141974018504, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3483, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.639906701994446, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3294, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.654613598085333, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3283, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5665845652546264, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.319, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.568410708254759, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3269, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5954621395011873, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3427, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5586111590619681, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3107, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5376766169086076, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3005, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.528351139194909, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3126, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.546595127743902, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3231, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5122337622446388, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.323, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5263518084262206, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3164, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5430640335364739, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3085, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5082204751148293, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3088, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.541141604264627, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3067, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.5233973076722868, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3234, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.503039465860446, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3473, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.5398717298058211, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3188, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.5278276809965626, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3357, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.5384100052646946, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3074, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5085981338026319, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3224, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5189683817102482, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3257, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5569733322985922, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.333, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.5033112925952238, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3125, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.5241847984037186, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3014, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.565705979756712, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3157, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.5041792244538641, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3196, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.5516283301790813, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3076, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.5439002732552327, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3295, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.48448477685253244, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3302, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.5335115783744465, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3069, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.47126856730409306, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3219, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.5261114215080503, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3171, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.5218592685513823, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3285, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.49004562394385515, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3363, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.5702026561037369, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.322, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.4794169168636671, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3193, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.4988347779020925, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3165, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.5093981810238166, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3399, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4721884611441708, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3229, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.49887073567592627, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3175, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.5045437946272339, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3005, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.49903024953277764, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3474, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.5314703321453352, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3122, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.5092684458014048, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3233, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.47352836758947026, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.301, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.4860543590228337, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3151, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.5257123347462322, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3193, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4943979366346464, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.2958, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.44952496033451056, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3068, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.4758452886198691, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3148, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4682473906979366, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3706, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.4791242451892492, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3288, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.5085817724900927, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3044, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.48647263186682693, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3289, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.5046420474541058, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3205, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.5586278799937978, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3203, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.46467805164179493, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.313, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.5133947729747058, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3354, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.4940637792212559, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.314, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.46383168322908513, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3496, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.48213425867862153, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3217, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4901278166547155, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3014, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.4864530534987615, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3244, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.4971270379983246, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.2798, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.48406441007747025, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3223, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.4860868077756716, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3202, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.506416799803235, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.277, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.6012281247858366, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2251, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.3811428602824129, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2041, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5784667293009617, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2162, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.705302440808711, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.1914, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4942964544263069, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2047, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.4421046816159049, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2131, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5079935773124865, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.1727, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4901753704796098, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.1913, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.42921694755148015, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.1899, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4176592012336345, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2108, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4555936330018958, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.1944, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.439819633982298, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.1793, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4507068322369287, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.1987, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4267563221021692, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.1786, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3876171615365498, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.1966, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.4057833864281153, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.1958, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.4087899210587149, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.1852, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.39407281907345965, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.1827, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.35935266163921825, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.1989, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.37536457589473, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2041, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.39428821774784384, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.196, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.4006580445730096, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.181, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.40910842845823364, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.207, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.39974974632699173, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2047, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.38363020981213264, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.1784, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3854610036805951, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2061, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.4164002914164664, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2037, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3602169675364507, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.1904, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3645159691692519, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.1852, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.4029660604014823, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.214, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.38159597795606703, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.189, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.38275208914530784, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2009, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3723632786915106, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.1869, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3678856931649929, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2174, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3502192302775391, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.1934, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3598509379361589, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2024, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3837829509093238, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.1778, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.37321603082177557, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2028, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3646624135896105, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.1795, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3619279737435615, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.1959, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.4045915856656408, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2002, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.36699856081232585, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.1768, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3593940759056696, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.1831, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.38036589475024113, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.1947, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3854399777626176, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.1927, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.35303194527568804, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.184, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3484026536415648, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.1848, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.35254870824694956, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2174, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.35010223875661445, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.1808, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.33589842070128134, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.1849, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3417155292449215, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.1932, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.337080699159681, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.1792, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.35890218507105415, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2147, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.36129251338506446, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3959496367146204, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2047, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3804537691867654, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2011, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3567951138023965, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.1977, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.34862334322445726, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.1472, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.3444326428697489, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2068, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.34315467198982963, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.1801, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.36053148046849726, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.1797, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3446001058185602, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.1782, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3572721398571936, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.1722, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3524926329338016, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.1868, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.3392133753693257, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.1923, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.37883269375786016, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.1846, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.35621682181716785, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.1991, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.34699611669774555, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.1756, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.331837247579926, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2133, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3358679464482438, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2214, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.35504197065657817, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.1761, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3311346372703272, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.198, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.36682363964880216, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.187, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.32597937356367246, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2129, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.3380380052817791, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2023, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.33938858828486923, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.1951, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.3488126726253992, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.1892, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.3487996176824251, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.1691, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3736447589429214, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.144, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.3384671060788562, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1548, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.29592263318995493, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1324, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.25723587257926206, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1354, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.29994660788305383, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1374, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.34946760400650806, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1503, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.38995283037498696, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1595, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.366017265083761, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1502, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.36389466953807864, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.15, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3170091114058245, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1404, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.26593367619599734, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1187, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2718128922765933, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1432, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.2954806585268095, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1103, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2814929538727662, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.157, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.2696282300167148, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1246, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.26906920340312424, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1198, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.24892417900423874, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1283, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.27222961061443396, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1648, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.23693650059319968, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1231, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.24598374781410315, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1239, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.26142026108083005, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1518, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.25507881255355563, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1558, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2637840647987734, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1419, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.2628127561012397, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1175, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2679837646112785, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1366, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2632869027854267, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.133, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2574616441403653, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1473, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2567180790770113, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1529, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.25759651968280495, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1389, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.2694492179857689, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1676, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2542976978117736, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1264, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2587661015213375, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1754, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2577619521286932, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1308, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.2482204520963082, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1363, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.26815500281534765, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1551, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.25071743077202685, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1564, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.2527811681338482, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1429, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2472715666298151, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1276, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2658156327118546, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.141, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.245573731147634, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1394, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2426493496398299, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1603, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.24716411463307858, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1437, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2487733898003392, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1317, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2316370048474447, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1242, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2448921595410395, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1468, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.23495875452719822, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1215, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.24476071444081385, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1299, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.25342925971274427, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1381, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.24807086311102455, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1371, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.24989429343702707, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1551, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2478877254023866, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1214, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.24910977380194788, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1266, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.25184436317148107, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1231, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.26081502123563133, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1411, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.253071429048111, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1375, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2540425002906153, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.136, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.2574206686510939, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1307, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2609227567583143, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1346, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.254691465480472, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.158, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.2581011350666695, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1263, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.24198525732995088, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1193, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.25164331206812923, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1266, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2492077783545167, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.137, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2436272770874148, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1353, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2502879339931626, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1159, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2442634210698441, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1375, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.24280458947882716, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1151, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2558531188851725, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1403, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.25245664987849675, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.11, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2545467999410838, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.147, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.23852493560647273, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.133, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2578644326341971, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1375, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.24082567476482564, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1476, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2581628432525251, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1213, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.25077241087237473, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1705, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2294943734104996, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1098, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.25551200611550673, | |
| "learning_rate": 0.0, | |
| "loss": 0.1474, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 5.514668037678039e+17, | |
| "train_loss": 0.39462805023560155, | |
| "train_runtime": 20897.0448, | |
| "train_samples_per_second": 2.393, | |
| "train_steps_per_second": 0.019 | |
| } | |
| ], | |
| "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": 5.514668037678039e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |