| { | |
| "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": 6.565009122628392, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0572, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.328717341696224, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0421, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.369275352431423, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.039, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.791532961085991, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0369, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.31775495796551, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.0266, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.736241973015712, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9508, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.5373656626337975, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.941, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.270490602661965, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9695, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.126778852450753, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9473, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.099039686781263, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9185, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.7156654481956903, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8988, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.573784880902445, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8829, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.708299932289841, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8685, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.5631685693376616, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8177, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.4301560338104442, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8102, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.050572738186419, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8021, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.2475668973295932, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7948, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.0074687922044097, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7774, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.0708941905916214, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7666, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.089841591889625, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7577, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9016606318186525, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7445, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8370254181084101, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7328, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.872883023357333, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7471, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.0297613923760245, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7694, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.039143177579734, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7284, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.802143908266341, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7082, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.8976432776800529, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7572, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.2644089427815146, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.723, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.766966129910527, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7346, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.0074038581881464, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.7246, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.6159126408382422, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.7072, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.817555030377531, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7202, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.9321366426212342, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6949, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.263705949782929, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6928, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.9860425684266011, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.7122, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.9568491905399428, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.7105, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.828584087261236, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.688, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.7325923052627724, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6979, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.2223985702397389, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6993, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.2393321588796675, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.699, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.9135362953424209, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6774, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.3636846427338403, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6912, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.149690425488308, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6726, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.2229859057607912, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6774, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.062736107596653, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6788, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.3614326518186355, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6913, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.8104639856438913, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6786, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.8072580270137925, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6798, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.0033262906656475, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6796, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.4898516561328707, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6696, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.9906598638823944, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6633, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.3692128036248665, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6705, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.2215272873346554, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6797, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.996981124774163, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6766, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.9447774230243922, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6697, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 1.0514899180166413, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6692, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.0836540336612241, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6595, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8898405997626494, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6714, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7935568259590772, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6533, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7615242090700829, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.662, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8696153086935214, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6539, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.7497202958546909, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6556, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.0754082945796624, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6507, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.6711297971818809, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6521, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.8340952465667271, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6571, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6481740808064326, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6538, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.5811634835620311, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6505, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6264377586581338, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6803, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6716639828353739, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6433, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6319441327353119, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6584, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.8207243162076922, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6415, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6804878323767257, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6471, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.887068819035412, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.654, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5222276620127227, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6426, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5856867311262312, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6701, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5281735785983478, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6554, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.6094649536525901, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6561, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5519119087702667, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6516, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.5205966588608025, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.6034, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6574821412870239, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.591, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5733536308117289, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5922, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7664634915479503, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5818, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 1.1590081671811712, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5992, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8193985492476491, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5781, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.595023053872972, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5833, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.8009409514722192, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5893, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.9062427005342203, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.6066, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.874502571469157, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.6056, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6640094806089545, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5887, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5527203553338561, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 1.0925521699511622, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5889, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 1.2079594584123021, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5753, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.602303394160385, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5767, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.7891997091366915, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5762, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.8330340597789887, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5904, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.820556401652943, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5829, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 1.0002873658596667, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5783, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.8050270031844088, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5842, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6624353683165365, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5837, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6723433366088991, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5811, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.7503293813106361, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5809, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7528979284936383, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5723, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.4699643102289398, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5708, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.7281498552959331, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5977, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.8062322636778657, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.572, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5626472296953341, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5714, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5240776231703065, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5746, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5822151305815075, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5651, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6326598579204256, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5753, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.7633646276299505, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5739, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.6498163430839372, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5711, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.40911707139968, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5685, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.6886095626477527, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.575, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.49539578922619476, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5777, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.4990587123354689, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5896, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5929657354788382, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5814, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5712927313848153, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5985, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5432391631828835, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5802, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4972774832115366, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5837, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.4572612942269152, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.586, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.4583153950969069, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5735, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4969900946020237, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5837, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.46452451391377214, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5841, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4648087428934368, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5743, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.530850788663003, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.58, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5059821856691323, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.585, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4066399796611767, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5726, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.4633166978365293, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5729, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.46057649659288696, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5813, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5820755063642323, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.567, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.6665575035761632, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5773, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5245997644761258, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5901, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.6081516252004494, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5664, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.7334970884288781, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5731, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.6781461970263889, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5867, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6368601573127258, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.57, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.6946775052435568, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5643, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5330301235950428, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5744, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.6728816573959958, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5746, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.6251002818182534, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5771, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.4431115539932792, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5841, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.5780356756650002, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5835, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.6352777814915348, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5798, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4653192733449068, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5778, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5158581424011935, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5678, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.45532669414323507, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5713, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4241449604630491, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.571, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.3997299375021256, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5598, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.41530672444646866, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5646, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4993662285855839, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5907, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.458107196385476, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5822, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5168425273240373, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5737, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.48248950168386684, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5706, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4220011184212505, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5724, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.511448721721425, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.57, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.42132363271837303, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5697, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.6142426994565252, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.5115, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.5014976654495472, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4883, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.8823072249894934, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.5081, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7139311320088461, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.5047, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6706539614646906, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4976, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5145410000120706, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4915, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.7033508690631125, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.5044, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5707517384942773, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.4909, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6372457619978193, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.4936, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5296029079371827, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.4866, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5662524931664337, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.4827, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5759158940995658, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.5099, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.47709033704149917, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.492, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5924729724712532, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.4878, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.4814407615771033, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.4994, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5915995037884118, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4798, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5016624858144002, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4849, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5059539338427179, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.4823, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4971945449775704, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4755, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.5824144784008408, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.4943, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3733267709243646, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.4809, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.6595636221676086, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4903, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.42190032774353986, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.4993, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.584966635534989, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.5059, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5504202244717419, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4929, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.3957334663665894, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4935, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.6442894500415227, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.4978, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.3830351172997605, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.4936, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4949223105507442, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.5062, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3567662851433714, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.4863, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.438962504361558, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.4865, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.45629573280582075, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4831, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.39707826875574237, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.4898, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.45700430824064886, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.4629, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.3947675683884105, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.4811, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.37453478488652464, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.4823, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.35684910985000506, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.4855, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.3868702459219989, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.4924, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.3202953153639153, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.4876, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.35810716930341874, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4916, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.354691271661972, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.4802, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.34644499546731927, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4979, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.3090839468218357, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.4963, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.34346648729385676, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4967, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3122824400944379, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.4947, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.347053361016479, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.4781, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.39304852748142405, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4894, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.34255027592191745, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.5015, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.35413409510237354, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.5049, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.37256988191755247, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.4646, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.32386964944184815, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.4854, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.42238676563897576, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.4923, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.310955505585705, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.476, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.39756080964896534, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.4761, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3350360577277234, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.4894, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.32984414811702134, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4861, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.38722341920712156, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4854, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.3375991236255547, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.5024, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.37202740704336557, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4902, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3618138057745732, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.4843, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4199055191320438, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.4939, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.35949604076491015, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.5, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4169544842405204, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.4903, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.35203246670737876, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.4914, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3989120150789621, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.488, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3656300269810547, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.4951, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.3792220880483876, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4906, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.4187057185591477, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.4905, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.32291249082023826, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.4678, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3424493949660347, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.4959, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3571033552209353, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.4955, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.31224876115153327, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.4795, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3152340809488724, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.4812, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.34447242751779383, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.4844, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.2908348763529779, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4863, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.33021136854474725, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.4954, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.27219624410114757, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.511, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.33010507081980045, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.4873, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.44292402950556065, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4636, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4032619197662792, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4117, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.7349953892845466, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.417, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5256382505272393, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.4022, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5018330680848114, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.411, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.5994128151952431, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.415, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.43513093999021074, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.4098, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.4975296643957784, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.4113, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4962538067783444, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4182, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.41665613602648904, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.415, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.48381708293718184, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.4123, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4297526791120439, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.4059, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.425476282116733, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.421, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.459519720788456, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.4126, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.40387854570124976, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.4232, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.4013861526222876, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.4127, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.36806870110585893, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.3953, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.3970517399821991, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.4011, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3631552016797807, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.4153, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3352780833944525, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.4154, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3586986848069525, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.4158, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3092500156480622, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4232, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.3473404924546456, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.4037, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3228963172839725, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.4036, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3159771803178001, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.4098, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.28878006210182217, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.4129, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.30651159761391367, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.4205, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.29340341483634563, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.4128, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.30254746277192424, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.4169, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3149305212023689, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.4047, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.29660190740344644, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4273, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.30569924253660125, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.4088, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.27780161522625435, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.4038, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2932940274796318, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.4169, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.2951334689862599, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.4249, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.2679168413320147, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.4138, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2981614965959627, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.4118, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.2664954166777559, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.4106, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.26173131587898135, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.414, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.270082835510258, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.4011, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.27451004289574743, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.4138, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.24984044955393064, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.4216, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.27003356502126874, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.4023, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2616940875521014, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.4041, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.2594582376597173, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4019, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.26512794010436286, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.4324, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.2655340139773929, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.4177, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.281660643217019, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.3984, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.2762031505228174, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.4216, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.254428338469852, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.4122, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.25483393322270115, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.4134, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2846321487765592, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.3889, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.251860998472327, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.4205, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.25350583402161075, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.4166, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.2753056911762651, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.4086, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.23481470705131982, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.4047, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2604667699892976, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.4049, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.26829490226075686, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.4017, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2704332461994144, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.4118, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.2581195024546522, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.4279, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.2587248929973286, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.4122, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2601130808106393, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.4246, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.2526845534865597, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.4194, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.2520916014893489, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.4161, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.2575916107190483, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.4253, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.26218788917196567, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.416, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.2554225008333403, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.4064, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.261375791397792, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.422, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.26302941724503986, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.4196, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.26470302202954, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.42, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.24748246954051706, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.4171, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.24953316511627668, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.4022, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2601979355861529, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.4064, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.2524676576433421, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.4081, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.24238215053792866, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.4214, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.2464913312393768, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.4185, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.24822568666485353, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.4179, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.24529949705957585, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.4091, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.296633441786206, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3988, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.4069234245847686, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.3676, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.310340568610311, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3598, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.27575381010445493, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3455, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.46876629205685755, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.3694, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.4363495408319517, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.3666, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2874338614777583, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.3756, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.3147412085040237, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.3615, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.34011474102677786, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.378, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3054267997394002, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.3715, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.2979149770295045, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.3832, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.29336949522458605, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3678, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2813486621051938, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.3653, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3057423797294473, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.3656, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2871341651633347, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.3702, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.3154162237883921, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3811, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2757683881514584, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.3665, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.27130781533104875, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3713, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2644094825511369, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.3788, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.27378888572394394, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.3584, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.2771846290194857, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.3774, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.2577572321891543, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.3571, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2603582270036936, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3574, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2561384186356945, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3558, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.23904364943476877, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3698, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.24416704176733103, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.37, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.24165689488576905, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3622, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.23484861545405772, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.3696, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.25380339590098017, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.3685, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2461591671323599, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3601, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.2360831843465452, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3851, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.22631980019879241, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3304, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.23784343408067637, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3596, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.24819284770589498, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.354, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.23725479289843976, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.3521, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.22934032370639418, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.3582, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.23510552388910463, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3663, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.22910597870149355, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.351, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.22255875369510628, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3832, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2282434736658511, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.3525, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.22089862420815698, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.373, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.22344336799549067, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.3668, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.22130520355602268, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3638, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.21030477829777675, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3666, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.22008977218598913, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.3638, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.21632111190565897, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.3764, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2160827870024147, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3685, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2074504863241659, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.3801, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2290627906221354, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3631, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.21414805943606877, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.3762, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2168214660823632, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.3625, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.21424457738510186, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.3457, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.22231443511856258, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.3483, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.21512065069961056, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.3643, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.21497656832670586, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3603, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.22210609780459875, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.3654, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2111903486970716, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.3782, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.21418593871091693, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3671, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.206607381683724, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3566, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.20661471531141848, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.3597, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.21058957822719954, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.3502, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2182495817936195, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.3505, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.21605890650209503, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3631, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.21458921700775935, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3533, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.21879093114638698, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.3619, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.21012454603136083, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3595, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.21308811481658155, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.3474, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2065420215109182, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3813, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.237634585613284, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3637, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.207793303734774, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.36, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.20790574057432382, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3607, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2150837927783755, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.3585, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.21813306234711938, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.357, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2047754174501529, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3678, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2166913784740888, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.352, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.20746966246330922, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.3598, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.21186116735657218, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.3592, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.21902294616854368, | |
| "learning_rate": 0.0, | |
| "loss": 0.3552, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.9541851989375713e+18, | |
| "train_loss": 0.5176007389258116, | |
| "train_runtime": 65078.6174, | |
| "train_samples_per_second": 0.768, | |
| "train_steps_per_second": 0.006 | |
| } | |
| ], | |
| "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": 1.9541851989375713e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |