| { | |
| "best_metric": 2.4503767490386963, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.11286681715575621, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007524454477050414, | |
| "grad_norm": 4.9338603019714355, | |
| "learning_rate": 1.018e-05, | |
| "loss": 4.511, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007524454477050414, | |
| "eval_loss": 2.8468708992004395, | |
| "eval_runtime": 37.6411, | |
| "eval_samples_per_second": 14.877, | |
| "eval_steps_per_second": 3.719, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0015048908954100827, | |
| "grad_norm": 5.828423500061035, | |
| "learning_rate": 2.036e-05, | |
| "loss": 3.8614, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002257336343115124, | |
| "grad_norm": 7.995044708251953, | |
| "learning_rate": 3.0539999999999996e-05, | |
| "loss": 4.6689, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0030097817908201654, | |
| "grad_norm": 6.929337024688721, | |
| "learning_rate": 4.072e-05, | |
| "loss": 4.2038, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003762227238525207, | |
| "grad_norm": 6.514157772064209, | |
| "learning_rate": 5.09e-05, | |
| "loss": 4.6263, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004514672686230248, | |
| "grad_norm": 6.524606227874756, | |
| "learning_rate": 6.107999999999999e-05, | |
| "loss": 4.8234, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005267118133935289, | |
| "grad_norm": 9.753814697265625, | |
| "learning_rate": 7.125999999999999e-05, | |
| "loss": 4.544, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006019563581640331, | |
| "grad_norm": 6.565380096435547, | |
| "learning_rate": 8.144e-05, | |
| "loss": 4.2002, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.006772009029345372, | |
| "grad_norm": 7.157611846923828, | |
| "learning_rate": 9.162e-05, | |
| "loss": 3.7329, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.007524454477050414, | |
| "grad_norm": 8.59189510345459, | |
| "learning_rate": 0.0001018, | |
| "loss": 5.0335, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008276899924755455, | |
| "grad_norm": 10.278609275817871, | |
| "learning_rate": 0.00010126421052631578, | |
| "loss": 5.018, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009029345372460496, | |
| "grad_norm": 11.457633018493652, | |
| "learning_rate": 0.00010072842105263156, | |
| "loss": 5.185, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.009781790820165538, | |
| "grad_norm": 8.473273277282715, | |
| "learning_rate": 0.00010019263157894736, | |
| "loss": 4.661, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.010534236267870579, | |
| "grad_norm": 9.23606014251709, | |
| "learning_rate": 9.965684210526316e-05, | |
| "loss": 5.1647, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.011286681715575621, | |
| "grad_norm": 10.785460472106934, | |
| "learning_rate": 9.912105263157895e-05, | |
| "loss": 5.2782, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.012039127163280662, | |
| "grad_norm": 9.376898765563965, | |
| "learning_rate": 9.858526315789473e-05, | |
| "loss": 4.4393, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.012791572610985704, | |
| "grad_norm": 10.466320037841797, | |
| "learning_rate": 9.804947368421052e-05, | |
| "loss": 5.3007, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.013544018058690745, | |
| "grad_norm": 8.497576713562012, | |
| "learning_rate": 9.75136842105263e-05, | |
| "loss": 4.941, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.014296463506395787, | |
| "grad_norm": 9.376507759094238, | |
| "learning_rate": 9.69778947368421e-05, | |
| "loss": 5.1504, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.015048908954100828, | |
| "grad_norm": 9.785571098327637, | |
| "learning_rate": 9.644210526315789e-05, | |
| "loss": 4.6566, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01580135440180587, | |
| "grad_norm": 16.524185180664062, | |
| "learning_rate": 9.590631578947369e-05, | |
| "loss": 5.2649, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01655379984951091, | |
| "grad_norm": 9.190619468688965, | |
| "learning_rate": 9.537052631578947e-05, | |
| "loss": 4.9649, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01730624529721595, | |
| "grad_norm": 15.659117698669434, | |
| "learning_rate": 9.483473684210526e-05, | |
| "loss": 6.6098, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01805869074492099, | |
| "grad_norm": 9.867484092712402, | |
| "learning_rate": 9.429894736842104e-05, | |
| "loss": 5.5985, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.018811136192626036, | |
| "grad_norm": 8.590432167053223, | |
| "learning_rate": 9.376315789473684e-05, | |
| "loss": 4.1838, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.019563581640331076, | |
| "grad_norm": 8.883073806762695, | |
| "learning_rate": 9.322736842105262e-05, | |
| "loss": 4.8669, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.020316027088036117, | |
| "grad_norm": 8.349895477294922, | |
| "learning_rate": 9.269157894736842e-05, | |
| "loss": 5.1803, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.021068472535741158, | |
| "grad_norm": 11.773541450500488, | |
| "learning_rate": 9.215578947368421e-05, | |
| "loss": 5.0873, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0218209179834462, | |
| "grad_norm": 13.537124633789062, | |
| "learning_rate": 9.162e-05, | |
| "loss": 4.6728, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.022573363431151242, | |
| "grad_norm": 12.410065650939941, | |
| "learning_rate": 9.108421052631578e-05, | |
| "loss": 5.1454, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.023325808878856283, | |
| "grad_norm": 11.18578052520752, | |
| "learning_rate": 9.054842105263158e-05, | |
| "loss": 4.2279, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.024078254326561323, | |
| "grad_norm": 38.089595794677734, | |
| "learning_rate": 9.001263157894736e-05, | |
| "loss": 5.0972, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.024830699774266364, | |
| "grad_norm": 20.254100799560547, | |
| "learning_rate": 8.947684210526315e-05, | |
| "loss": 4.9312, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.025583145221971408, | |
| "grad_norm": 32.83479309082031, | |
| "learning_rate": 8.894105263157895e-05, | |
| "loss": 5.4251, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02633559066967645, | |
| "grad_norm": 14.390934944152832, | |
| "learning_rate": 8.840526315789473e-05, | |
| "loss": 5.1836, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02708803611738149, | |
| "grad_norm": 17.68592071533203, | |
| "learning_rate": 8.786947368421052e-05, | |
| "loss": 5.4996, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.02784048156508653, | |
| "grad_norm": 15.407584190368652, | |
| "learning_rate": 8.733368421052632e-05, | |
| "loss": 5.5683, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.028592927012791574, | |
| "grad_norm": 13.758222579956055, | |
| "learning_rate": 8.67978947368421e-05, | |
| "loss": 5.6531, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.029345372460496615, | |
| "grad_norm": 15.094158172607422, | |
| "learning_rate": 8.626210526315789e-05, | |
| "loss": 4.6223, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.030097817908201655, | |
| "grad_norm": 14.733675003051758, | |
| "learning_rate": 8.572631578947367e-05, | |
| "loss": 5.6096, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.030850263355906696, | |
| "grad_norm": 11.245537757873535, | |
| "learning_rate": 8.519052631578947e-05, | |
| "loss": 4.9552, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03160270880361174, | |
| "grad_norm": 15.303187370300293, | |
| "learning_rate": 8.465473684210527e-05, | |
| "loss": 6.245, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03235515425131678, | |
| "grad_norm": 12.705514907836914, | |
| "learning_rate": 8.411894736842105e-05, | |
| "loss": 5.5335, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.03310759969902182, | |
| "grad_norm": 12.836231231689453, | |
| "learning_rate": 8.358315789473684e-05, | |
| "loss": 5.8837, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.033860045146726865, | |
| "grad_norm": 12.260278701782227, | |
| "learning_rate": 8.304736842105262e-05, | |
| "loss": 4.22, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0346124905944319, | |
| "grad_norm": 15.98351764678955, | |
| "learning_rate": 8.251157894736841e-05, | |
| "loss": 4.8508, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.035364936042136946, | |
| "grad_norm": 18.42877197265625, | |
| "learning_rate": 8.197578947368421e-05, | |
| "loss": 5.7646, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03611738148984198, | |
| "grad_norm": 15.042816162109375, | |
| "learning_rate": 8.144e-05, | |
| "loss": 6.2102, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03686982693754703, | |
| "grad_norm": 20.197011947631836, | |
| "learning_rate": 8.090421052631579e-05, | |
| "loss": 7.2784, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03762227238525207, | |
| "grad_norm": 22.858545303344727, | |
| "learning_rate": 8.036842105263158e-05, | |
| "loss": 6.3548, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03762227238525207, | |
| "eval_loss": 2.573653221130371, | |
| "eval_runtime": 37.6194, | |
| "eval_samples_per_second": 14.886, | |
| "eval_steps_per_second": 3.721, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03837471783295711, | |
| "grad_norm": 6.201414108276367, | |
| "learning_rate": 7.983263157894736e-05, | |
| "loss": 4.6951, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03912716328066215, | |
| "grad_norm": 4.830835342407227, | |
| "learning_rate": 7.929684210526315e-05, | |
| "loss": 5.02, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0398796087283672, | |
| "grad_norm": 7.732100009918213, | |
| "learning_rate": 7.876105263157895e-05, | |
| "loss": 4.9632, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.040632054176072234, | |
| "grad_norm": 7.422882556915283, | |
| "learning_rate": 7.822526315789473e-05, | |
| "loss": 4.2894, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04138449962377728, | |
| "grad_norm": 6.019810676574707, | |
| "learning_rate": 7.768947368421053e-05, | |
| "loss": 4.5985, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.042136945071482315, | |
| "grad_norm": 7.448675632476807, | |
| "learning_rate": 7.715368421052631e-05, | |
| "loss": 5.6458, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.04288939051918736, | |
| "grad_norm": 6.525016784667969, | |
| "learning_rate": 7.66178947368421e-05, | |
| "loss": 4.4928, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0436418359668924, | |
| "grad_norm": 5.862019062042236, | |
| "learning_rate": 7.608210526315788e-05, | |
| "loss": 5.0442, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04439428141459744, | |
| "grad_norm": 6.698094844818115, | |
| "learning_rate": 7.554631578947368e-05, | |
| "loss": 5.1862, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.045146726862302484, | |
| "grad_norm": 5.901148796081543, | |
| "learning_rate": 7.501052631578947e-05, | |
| "loss": 4.6401, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04589917231000752, | |
| "grad_norm": 8.506747245788574, | |
| "learning_rate": 7.447473684210527e-05, | |
| "loss": 5.6167, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.046651617757712566, | |
| "grad_norm": 8.143284797668457, | |
| "learning_rate": 7.393894736842105e-05, | |
| "loss": 4.5655, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04740406320541761, | |
| "grad_norm": 5.302389621734619, | |
| "learning_rate": 7.340315789473684e-05, | |
| "loss": 4.4897, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04815650865312265, | |
| "grad_norm": 5.529751300811768, | |
| "learning_rate": 7.286736842105262e-05, | |
| "loss": 4.2722, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04890895410082769, | |
| "grad_norm": 8.762489318847656, | |
| "learning_rate": 7.233157894736842e-05, | |
| "loss": 5.1436, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04966139954853273, | |
| "grad_norm": 11.295607566833496, | |
| "learning_rate": 7.179578947368421e-05, | |
| "loss": 4.9496, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05041384499623777, | |
| "grad_norm": 9.971809387207031, | |
| "learning_rate": 7.125999999999999e-05, | |
| "loss": 5.4893, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.051166290443942816, | |
| "grad_norm": 7.344180583953857, | |
| "learning_rate": 7.072421052631579e-05, | |
| "loss": 4.6742, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05191873589164785, | |
| "grad_norm": 9.844782829284668, | |
| "learning_rate": 7.018842105263158e-05, | |
| "loss": 5.1691, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0526711813393529, | |
| "grad_norm": 8.95771598815918, | |
| "learning_rate": 6.965263157894736e-05, | |
| "loss": 4.879, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05342362678705794, | |
| "grad_norm": 8.04973030090332, | |
| "learning_rate": 6.911684210526316e-05, | |
| "loss": 5.0598, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05417607223476298, | |
| "grad_norm": 8.579294204711914, | |
| "learning_rate": 6.858105263157894e-05, | |
| "loss": 4.8055, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05492851768246802, | |
| "grad_norm": 7.308038234710693, | |
| "learning_rate": 6.804526315789473e-05, | |
| "loss": 4.7293, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05568096313017306, | |
| "grad_norm": 7.7795233726501465, | |
| "learning_rate": 6.750947368421052e-05, | |
| "loss": 5.0379, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.056433408577878104, | |
| "grad_norm": 8.343374252319336, | |
| "learning_rate": 6.697368421052631e-05, | |
| "loss": 4.4904, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05718585402558315, | |
| "grad_norm": 7.44524621963501, | |
| "learning_rate": 6.64378947368421e-05, | |
| "loss": 4.7685, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.057938299473288185, | |
| "grad_norm": 10.718270301818848, | |
| "learning_rate": 6.59021052631579e-05, | |
| "loss": 5.2094, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.05869074492099323, | |
| "grad_norm": 9.407214164733887, | |
| "learning_rate": 6.536631578947368e-05, | |
| "loss": 5.0866, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.059443190368698266, | |
| "grad_norm": 8.373135566711426, | |
| "learning_rate": 6.483052631578947e-05, | |
| "loss": 4.8998, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.06019563581640331, | |
| "grad_norm": 8.654524803161621, | |
| "learning_rate": 6.429473684210525e-05, | |
| "loss": 4.6364, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.060948081264108354, | |
| "grad_norm": 9.621380805969238, | |
| "learning_rate": 6.375894736842104e-05, | |
| "loss": 5.1948, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06170052671181339, | |
| "grad_norm": 13.150786399841309, | |
| "learning_rate": 6.322315789473684e-05, | |
| "loss": 5.4827, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.062452972159518436, | |
| "grad_norm": 10.533498764038086, | |
| "learning_rate": 6.268736842105264e-05, | |
| "loss": 4.9195, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.06320541760722348, | |
| "grad_norm": 10.708104133605957, | |
| "learning_rate": 6.215157894736842e-05, | |
| "loss": 4.8833, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.06395786305492852, | |
| "grad_norm": 7.892517566680908, | |
| "learning_rate": 6.16157894736842e-05, | |
| "loss": 5.0272, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06471030850263355, | |
| "grad_norm": 13.007377624511719, | |
| "learning_rate": 6.107999999999999e-05, | |
| "loss": 4.8137, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.0654627539503386, | |
| "grad_norm": 11.001848220825195, | |
| "learning_rate": 6.054421052631578e-05, | |
| "loss": 5.8105, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.06621519939804364, | |
| "grad_norm": 13.360245704650879, | |
| "learning_rate": 6.000842105263157e-05, | |
| "loss": 5.2308, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.06696764484574869, | |
| "grad_norm": 8.78776741027832, | |
| "learning_rate": 5.947263157894737e-05, | |
| "loss": 4.3293, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.06772009029345373, | |
| "grad_norm": 11.788161277770996, | |
| "learning_rate": 5.893684210526316e-05, | |
| "loss": 5.5094, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06847253574115876, | |
| "grad_norm": 13.817206382751465, | |
| "learning_rate": 5.8401052631578944e-05, | |
| "loss": 5.6535, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.0692249811888638, | |
| "grad_norm": 10.35663890838623, | |
| "learning_rate": 5.7865263157894736e-05, | |
| "loss": 4.9656, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.06997742663656885, | |
| "grad_norm": 12.754554748535156, | |
| "learning_rate": 5.732947368421052e-05, | |
| "loss": 6.037, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07072987208427389, | |
| "grad_norm": 13.788698196411133, | |
| "learning_rate": 5.6793684210526306e-05, | |
| "loss": 5.4732, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.07148231753197894, | |
| "grad_norm": 10.369476318359375, | |
| "learning_rate": 5.6257894736842105e-05, | |
| "loss": 4.9698, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07223476297968397, | |
| "grad_norm": 11.039383888244629, | |
| "learning_rate": 5.57221052631579e-05, | |
| "loss": 5.0991, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07298720842738901, | |
| "grad_norm": 18.217975616455078, | |
| "learning_rate": 5.518631578947368e-05, | |
| "loss": 5.7764, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.07373965387509406, | |
| "grad_norm": 13.361612319946289, | |
| "learning_rate": 5.4650526315789474e-05, | |
| "loss": 4.8924, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.0744920993227991, | |
| "grad_norm": 24.20296287536621, | |
| "learning_rate": 5.411473684210526e-05, | |
| "loss": 6.2003, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.07524454477050414, | |
| "grad_norm": 16.34416389465332, | |
| "learning_rate": 5.3578947368421044e-05, | |
| "loss": 5.3467, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07524454477050414, | |
| "eval_loss": 2.492374897003174, | |
| "eval_runtime": 37.6751, | |
| "eval_samples_per_second": 14.864, | |
| "eval_steps_per_second": 3.716, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07599699021820917, | |
| "grad_norm": 4.632258415222168, | |
| "learning_rate": 5.3043157894736836e-05, | |
| "loss": 4.8204, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.07674943566591422, | |
| "grad_norm": 5.53971004486084, | |
| "learning_rate": 5.2507368421052635e-05, | |
| "loss": 5.2553, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.07750188111361926, | |
| "grad_norm": 3.8668930530548096, | |
| "learning_rate": 5.197157894736842e-05, | |
| "loss": 4.3132, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.0782543265613243, | |
| "grad_norm": 5.11984920501709, | |
| "learning_rate": 5.143578947368421e-05, | |
| "loss": 4.5635, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.07900677200902935, | |
| "grad_norm": 4.6450347900390625, | |
| "learning_rate": 5.09e-05, | |
| "loss": 3.4141, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0797592174567344, | |
| "grad_norm": 5.824936389923096, | |
| "learning_rate": 5.036421052631578e-05, | |
| "loss": 4.7563, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08051166290443942, | |
| "grad_norm": 4.639711380004883, | |
| "learning_rate": 4.982842105263158e-05, | |
| "loss": 4.3026, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08126410835214447, | |
| "grad_norm": 8.702680587768555, | |
| "learning_rate": 4.9292631578947366e-05, | |
| "loss": 5.7537, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08201655379984951, | |
| "grad_norm": 5.901224613189697, | |
| "learning_rate": 4.875684210526315e-05, | |
| "loss": 5.0042, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.08276899924755456, | |
| "grad_norm": 8.818628311157227, | |
| "learning_rate": 4.822105263157894e-05, | |
| "loss": 5.3414, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0835214446952596, | |
| "grad_norm": 7.103747367858887, | |
| "learning_rate": 4.7685263157894735e-05, | |
| "loss": 5.2836, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.08427389014296463, | |
| "grad_norm": 6.621494770050049, | |
| "learning_rate": 4.714947368421052e-05, | |
| "loss": 4.9538, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.08502633559066967, | |
| "grad_norm": 8.953717231750488, | |
| "learning_rate": 4.661368421052631e-05, | |
| "loss": 5.1009, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.08577878103837472, | |
| "grad_norm": 8.522113800048828, | |
| "learning_rate": 4.6077894736842104e-05, | |
| "loss": 5.0785, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.08653122648607976, | |
| "grad_norm": 6.302427291870117, | |
| "learning_rate": 4.554210526315789e-05, | |
| "loss": 4.6044, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0872836719337848, | |
| "grad_norm": 8.132070541381836, | |
| "learning_rate": 4.500631578947368e-05, | |
| "loss": 4.9639, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.08803611738148984, | |
| "grad_norm": 7.749171733856201, | |
| "learning_rate": 4.447052631578947e-05, | |
| "loss": 5.2865, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.08878856282919488, | |
| "grad_norm": 5.969038963317871, | |
| "learning_rate": 4.393473684210526e-05, | |
| "loss": 4.7642, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.08954100827689992, | |
| "grad_norm": 6.512506484985352, | |
| "learning_rate": 4.339894736842105e-05, | |
| "loss": 2.9121, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.09029345372460497, | |
| "grad_norm": 6.616455554962158, | |
| "learning_rate": 4.2863157894736835e-05, | |
| "loss": 5.1259, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09104589917231001, | |
| "grad_norm": 8.678909301757812, | |
| "learning_rate": 4.2327368421052634e-05, | |
| "loss": 5.4049, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.09179834462001504, | |
| "grad_norm": 7.853146553039551, | |
| "learning_rate": 4.179157894736842e-05, | |
| "loss": 5.1167, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.09255079006772009, | |
| "grad_norm": 6.326202392578125, | |
| "learning_rate": 4.1255789473684204e-05, | |
| "loss": 5.1124, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.09330323551542513, | |
| "grad_norm": 6.513983726501465, | |
| "learning_rate": 4.072e-05, | |
| "loss": 4.8156, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.09405568096313018, | |
| "grad_norm": 7.685911178588867, | |
| "learning_rate": 4.018421052631579e-05, | |
| "loss": 4.3464, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.09480812641083522, | |
| "grad_norm": 8.669236183166504, | |
| "learning_rate": 3.9648421052631573e-05, | |
| "loss": 4.7814, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.09556057185854025, | |
| "grad_norm": 7.881282806396484, | |
| "learning_rate": 3.9112631578947365e-05, | |
| "loss": 5.0522, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0963130173062453, | |
| "grad_norm": 6.99576473236084, | |
| "learning_rate": 3.857684210526316e-05, | |
| "loss": 4.4357, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.09706546275395034, | |
| "grad_norm": 8.76285171508789, | |
| "learning_rate": 3.804105263157894e-05, | |
| "loss": 5.5137, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.09781790820165538, | |
| "grad_norm": 7.629359245300293, | |
| "learning_rate": 3.7505263157894734e-05, | |
| "loss": 4.3788, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09857035364936043, | |
| "grad_norm": 9.962780952453613, | |
| "learning_rate": 3.6969473684210526e-05, | |
| "loss": 4.5262, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.09932279909706546, | |
| "grad_norm": 7.68848180770874, | |
| "learning_rate": 3.643368421052631e-05, | |
| "loss": 4.803, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.1000752445447705, | |
| "grad_norm": 12.457582473754883, | |
| "learning_rate": 3.5897894736842103e-05, | |
| "loss": 5.286, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.10082768999247554, | |
| "grad_norm": 10.64576244354248, | |
| "learning_rate": 3.5362105263157895e-05, | |
| "loss": 5.5489, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.10158013544018059, | |
| "grad_norm": 8.363515853881836, | |
| "learning_rate": 3.482631578947368e-05, | |
| "loss": 4.724, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.10233258088788563, | |
| "grad_norm": 9.08327579498291, | |
| "learning_rate": 3.429052631578947e-05, | |
| "loss": 4.5964, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.10308502633559068, | |
| "grad_norm": 10.975964546203613, | |
| "learning_rate": 3.375473684210526e-05, | |
| "loss": 4.7091, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1038374717832957, | |
| "grad_norm": 8.273202896118164, | |
| "learning_rate": 3.321894736842105e-05, | |
| "loss": 3.8967, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.10458991723100075, | |
| "grad_norm": 10.234407424926758, | |
| "learning_rate": 3.268315789473684e-05, | |
| "loss": 5.5976, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1053423626787058, | |
| "grad_norm": 8.687202453613281, | |
| "learning_rate": 3.2147368421052627e-05, | |
| "loss": 4.7047, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.10609480812641084, | |
| "grad_norm": 9.25235652923584, | |
| "learning_rate": 3.161157894736842e-05, | |
| "loss": 4.7048, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.10684725357411588, | |
| "grad_norm": 10.904390335083008, | |
| "learning_rate": 3.107578947368421e-05, | |
| "loss": 5.0487, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.10759969902182091, | |
| "grad_norm": 12.776407241821289, | |
| "learning_rate": 3.0539999999999996e-05, | |
| "loss": 5.6051, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.10835214446952596, | |
| "grad_norm": 10.124897003173828, | |
| "learning_rate": 3.0004210526315784e-05, | |
| "loss": 5.6051, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.109104589917231, | |
| "grad_norm": 10.322992324829102, | |
| "learning_rate": 2.946842105263158e-05, | |
| "loss": 4.5447, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.10985703536493605, | |
| "grad_norm": 17.68702507019043, | |
| "learning_rate": 2.8932631578947368e-05, | |
| "loss": 5.1524, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.11060948081264109, | |
| "grad_norm": 14.997350692749023, | |
| "learning_rate": 2.8396842105263153e-05, | |
| "loss": 5.7192, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.11136192626034612, | |
| "grad_norm": 12.031723022460938, | |
| "learning_rate": 2.786105263157895e-05, | |
| "loss": 5.2456, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.11211437170805116, | |
| "grad_norm": 17.6466007232666, | |
| "learning_rate": 2.7325263157894737e-05, | |
| "loss": 6.2655, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.11286681715575621, | |
| "grad_norm": 18.988000869750977, | |
| "learning_rate": 2.6789473684210522e-05, | |
| "loss": 7.0167, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11286681715575621, | |
| "eval_loss": 2.4503767490386963, | |
| "eval_runtime": 37.6662, | |
| "eval_samples_per_second": 14.867, | |
| "eval_steps_per_second": 3.717, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.437792434153062e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |