| { | |
| "best_metric": 1.3523932695388794, | |
| "best_model_checkpoint": "checkpoints/sft_2/checkpoint-2739", | |
| "epoch": 7.504109589041096, | |
| "eval_steps": 913, | |
| "global_step": 2739, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10136986301369863, | |
| "grad_norm": 18.251895904541016, | |
| "learning_rate": 5.068493150684932e-07, | |
| "loss": 2.6033, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.20273972602739726, | |
| "grad_norm": 18.156850814819336, | |
| "learning_rate": 1.0136986301369864e-06, | |
| "loss": 2.4493, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3041095890410959, | |
| "grad_norm": 16.538768768310547, | |
| "learning_rate": 1.5205479452054797e-06, | |
| "loss": 2.1375, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4054794520547945, | |
| "grad_norm": 4.952304840087891, | |
| "learning_rate": 2.027397260273973e-06, | |
| "loss": 1.7107, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 3.9061036109924316, | |
| "learning_rate": 2.534246575342466e-06, | |
| "loss": 1.5491, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6082191780821918, | |
| "grad_norm": 4.238989353179932, | |
| "learning_rate": 3.0410958904109593e-06, | |
| "loss": 1.4969, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7095890410958904, | |
| "grad_norm": 3.786147356033325, | |
| "learning_rate": 3.5479452054794523e-06, | |
| "loss": 1.4779, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.810958904109589, | |
| "grad_norm": 3.699124813079834, | |
| "learning_rate": 4.054794520547946e-06, | |
| "loss": 1.4554, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9123287671232877, | |
| "grad_norm": 3.877596855163574, | |
| "learning_rate": 4.561643835616439e-06, | |
| "loss": 1.4529, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 5.006152153015137, | |
| "learning_rate": 4.999971418949206e-06, | |
| "loss": 1.4465, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.115068493150685, | |
| "grad_norm": 3.469430923461914, | |
| "learning_rate": 4.997983588332731e-06, | |
| "loss": 1.4214, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.2164383561643834, | |
| "grad_norm": 3.827889919281006, | |
| "learning_rate": 4.992868406108372e-06, | |
| "loss": 1.4312, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.3178082191780822, | |
| "grad_norm": 3.8270199298858643, | |
| "learning_rate": 4.9846322762306745e-06, | |
| "loss": 1.4214, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.4191780821917808, | |
| "grad_norm": 3.366734266281128, | |
| "learning_rate": 4.973285509925916e-06, | |
| "loss": 1.4377, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.5205479452054793, | |
| "grad_norm": 3.5708305835723877, | |
| "learning_rate": 4.958842312782962e-06, | |
| "loss": 1.4188, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.621917808219178, | |
| "grad_norm": 3.532292366027832, | |
| "learning_rate": 4.94132076696857e-06, | |
| "loss": 1.4017, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.7232876712328768, | |
| "grad_norm": 4.288147449493408, | |
| "learning_rate": 4.920742808589422e-06, | |
| "loss": 1.4026, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.8246575342465754, | |
| "grad_norm": 3.560300350189209, | |
| "learning_rate": 4.897134200229196e-06, | |
| "loss": 1.4065, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.926027397260274, | |
| "grad_norm": 3.666012763977051, | |
| "learning_rate": 4.870524498695093e-06, | |
| "loss": 1.4041, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 4.007699966430664, | |
| "learning_rate": 4.8409470180141825e-06, | |
| "loss": 1.3843, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.128767123287671, | |
| "grad_norm": 3.6797378063201904, | |
| "learning_rate": 4.808438787725889e-06, | |
| "loss": 1.405, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 2.23013698630137, | |
| "grad_norm": 4.038980960845947, | |
| "learning_rate": 4.773040506522845e-06, | |
| "loss": 1.3826, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 2.3315068493150686, | |
| "grad_norm": 3.8665153980255127, | |
| "learning_rate": 4.734796491298143e-06, | |
| "loss": 1.3791, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 2.432876712328767, | |
| "grad_norm": 4.490111827850342, | |
| "learning_rate": 4.693754621662789e-06, | |
| "loss": 1.3699, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 2.5013698630136987, | |
| "eval_loss": 1.387344479560852, | |
| "eval_runtime": 41.7812, | |
| "eval_samples_per_second": 23.934, | |
| "eval_steps_per_second": 2.992, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 2.5342465753424657, | |
| "grad_norm": 4.025441646575928, | |
| "learning_rate": 4.649966280002798e-06, | |
| "loss": 1.3657, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.6356164383561644, | |
| "grad_norm": 4.344162940979004, | |
| "learning_rate": 4.6034862871509954e-06, | |
| "loss": 1.3728, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 2.736986301369863, | |
| "grad_norm": 3.900259494781494, | |
| "learning_rate": 4.5543728337540524e-06, | |
| "loss": 1.3747, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 2.8383561643835615, | |
| "grad_norm": 4.291511058807373, | |
| "learning_rate": 4.502687407420681e-06, | |
| "loss": 1.3742, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 2.9397260273972603, | |
| "grad_norm": 3.9363627433776855, | |
| "learning_rate": 4.4484947157421985e-06, | |
| "loss": 1.3765, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 3.041095890410959, | |
| "grad_norm": 4.16279411315918, | |
| "learning_rate": 4.391862605281827e-06, | |
| "loss": 1.3606, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.1424657534246574, | |
| "grad_norm": 4.134684085845947, | |
| "learning_rate": 4.332861976634164e-06, | |
| "loss": 1.3434, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 3.243835616438356, | |
| "grad_norm": 4.688838958740234, | |
| "learning_rate": 4.27156669566115e-06, | |
| "loss": 1.3553, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 3.345205479452055, | |
| "grad_norm": 4.097309589385986, | |
| "learning_rate": 4.208053501015674e-06, | |
| "loss": 1.3692, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 3.4465753424657533, | |
| "grad_norm": 4.752201080322266, | |
| "learning_rate": 4.142401908068583e-06, | |
| "loss": 1.3528, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 3.547945205479452, | |
| "grad_norm": 4.456752777099609, | |
| "learning_rate": 4.0746941093593815e-06, | |
| "loss": 1.3494, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 3.649315068493151, | |
| "grad_norm": 4.169310092926025, | |
| "learning_rate": 4.005014871695243e-06, | |
| "loss": 1.3572, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 3.750684931506849, | |
| "grad_norm": 4.615598678588867, | |
| "learning_rate": 3.933451430027176e-06, | |
| "loss": 1.3404, | |
| "step": 1369 | |
| }, | |
| { | |
| "epoch": 3.852054794520548, | |
| "grad_norm": 4.378635406494141, | |
| "learning_rate": 3.8600933782361875e-06, | |
| "loss": 1.3494, | |
| "step": 1406 | |
| }, | |
| { | |
| "epoch": 3.9534246575342467, | |
| "grad_norm": 4.689962387084961, | |
| "learning_rate": 3.78503255696618e-06, | |
| "loss": 1.3388, | |
| "step": 1443 | |
| }, | |
| { | |
| "epoch": 4.054794520547945, | |
| "grad_norm": 4.6829071044921875, | |
| "learning_rate": 3.7083629386440304e-06, | |
| "loss": 1.3513, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.156164383561644, | |
| "grad_norm": 4.488067626953125, | |
| "learning_rate": 3.6301805098307614e-06, | |
| "loss": 1.3284, | |
| "step": 1517 | |
| }, | |
| { | |
| "epoch": 4.257534246575342, | |
| "grad_norm": 4.579895496368408, | |
| "learning_rate": 3.5505831510511272e-06, | |
| "loss": 1.3345, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 4.358904109589041, | |
| "grad_norm": 4.769827365875244, | |
| "learning_rate": 3.4696705142520537e-06, | |
| "loss": 1.3316, | |
| "step": 1591 | |
| }, | |
| { | |
| "epoch": 4.46027397260274, | |
| "grad_norm": 4.928876876831055, | |
| "learning_rate": 3.3875438980433367e-06, | |
| "loss": 1.3359, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 4.561643835616438, | |
| "grad_norm": 4.699948310852051, | |
| "learning_rate": 3.3043061208768075e-06, | |
| "loss": 1.3425, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 4.663013698630137, | |
| "grad_norm": 4.816317558288574, | |
| "learning_rate": 3.2200613923227255e-06, | |
| "loss": 1.3236, | |
| "step": 1702 | |
| }, | |
| { | |
| "epoch": 4.764383561643836, | |
| "grad_norm": 5.371675491333008, | |
| "learning_rate": 3.134915182604566e-06, | |
| "loss": 1.3372, | |
| "step": 1739 | |
| }, | |
| { | |
| "epoch": 4.865753424657534, | |
| "grad_norm": 5.1571455001831055, | |
| "learning_rate": 3.0489740905555297e-06, | |
| "loss": 1.3102, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 4.967123287671233, | |
| "grad_norm": 4.910325050354004, | |
| "learning_rate": 2.9623457101620844e-06, | |
| "loss": 1.319, | |
| "step": 1813 | |
| }, | |
| { | |
| "epoch": 5.002739726027397, | |
| "eval_loss": 1.3639509677886963, | |
| "eval_runtime": 41.8029, | |
| "eval_samples_per_second": 23.922, | |
| "eval_steps_per_second": 2.99, | |
| "step": 1826 | |
| }, | |
| { | |
| "epoch": 5.068493150684931, | |
| "grad_norm": 5.355637073516846, | |
| "learning_rate": 2.8751384958616318e-06, | |
| "loss": 1.3268, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 5.16986301369863, | |
| "grad_norm": 5.53813362121582, | |
| "learning_rate": 2.787461626762929e-06, | |
| "loss": 1.2971, | |
| "step": 1887 | |
| }, | |
| { | |
| "epoch": 5.271232876712329, | |
| "grad_norm": 5.353805065155029, | |
| "learning_rate": 2.6994248699592545e-06, | |
| "loss": 1.323, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 5.372602739726028, | |
| "grad_norm": 5.467613220214844, | |
| "learning_rate": 2.611138443105452e-06, | |
| "loss": 1.2996, | |
| "step": 1961 | |
| }, | |
| { | |
| "epoch": 5.473972602739726, | |
| "grad_norm": 5.071502685546875, | |
| "learning_rate": 2.5227128764308887e-06, | |
| "loss": 1.3153, | |
| "step": 1998 | |
| }, | |
| { | |
| "epoch": 5.575342465753424, | |
| "grad_norm": 5.327346324920654, | |
| "learning_rate": 2.4342588743610904e-06, | |
| "loss": 1.3218, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 5.676712328767123, | |
| "grad_norm": 5.512831211090088, | |
| "learning_rate": 2.345887176921286e-06, | |
| "loss": 1.3151, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 5.778082191780822, | |
| "grad_norm": 5.303826332092285, | |
| "learning_rate": 2.257708421095391e-06, | |
| "loss": 1.31, | |
| "step": 2109 | |
| }, | |
| { | |
| "epoch": 5.879452054794521, | |
| "grad_norm": 5.505942344665527, | |
| "learning_rate": 2.1698330023139837e-06, | |
| "loss": 1.3062, | |
| "step": 2146 | |
| }, | |
| { | |
| "epoch": 5.980821917808219, | |
| "grad_norm": 5.343113422393799, | |
| "learning_rate": 2.0823709362447025e-06, | |
| "loss": 1.2966, | |
| "step": 2183 | |
| }, | |
| { | |
| "epoch": 6.082191780821918, | |
| "grad_norm": 5.3861541748046875, | |
| "learning_rate": 1.995431721058082e-06, | |
| "loss": 1.2911, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 6.183561643835616, | |
| "grad_norm": 5.535419464111328, | |
| "learning_rate": 1.909124200341277e-06, | |
| "loss": 1.3043, | |
| "step": 2257 | |
| }, | |
| { | |
| "epoch": 6.284931506849315, | |
| "grad_norm": 6.381133556365967, | |
| "learning_rate": 1.82355642683128e-06, | |
| "loss": 1.2911, | |
| "step": 2294 | |
| }, | |
| { | |
| "epoch": 6.3863013698630136, | |
| "grad_norm": 5.855857849121094, | |
| "learning_rate": 1.7388355271382565e-06, | |
| "loss": 1.2897, | |
| "step": 2331 | |
| }, | |
| { | |
| "epoch": 6.487671232876712, | |
| "grad_norm": 6.03603982925415, | |
| "learning_rate": 1.6550675676283428e-06, | |
| "loss": 1.2915, | |
| "step": 2368 | |
| }, | |
| { | |
| "epoch": 6.589041095890411, | |
| "grad_norm": 6.094645977020264, | |
| "learning_rate": 1.5723574216338066e-06, | |
| "loss": 1.3027, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 6.69041095890411, | |
| "grad_norm": 5.70314359664917, | |
| "learning_rate": 1.4908086381568398e-06, | |
| "loss": 1.3001, | |
| "step": 2442 | |
| }, | |
| { | |
| "epoch": 6.791780821917808, | |
| "grad_norm": 6.103550434112549, | |
| "learning_rate": 1.4105233122313416e-06, | |
| "loss": 1.3084, | |
| "step": 2479 | |
| }, | |
| { | |
| "epoch": 6.8931506849315065, | |
| "grad_norm": 6.549497127532959, | |
| "learning_rate": 1.331601957104995e-06, | |
| "loss": 1.2865, | |
| "step": 2516 | |
| }, | |
| { | |
| "epoch": 6.994520547945205, | |
| "grad_norm": 5.893728256225586, | |
| "learning_rate": 1.2541433784016639e-06, | |
| "loss": 1.2883, | |
| "step": 2553 | |
| }, | |
| { | |
| "epoch": 7.095890410958904, | |
| "grad_norm": 6.355903148651123, | |
| "learning_rate": 1.1782445504216552e-06, | |
| "loss": 1.285, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 7.197260273972603, | |
| "grad_norm": 6.0843729972839355, | |
| "learning_rate": 1.1040004947346974e-06, | |
| "loss": 1.2681, | |
| "step": 2627 | |
| }, | |
| { | |
| "epoch": 7.298630136986302, | |
| "grad_norm": 6.470739841461182, | |
| "learning_rate": 1.0315041612176476e-06, | |
| "loss": 1.281, | |
| "step": 2664 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 5.7129950523376465, | |
| "learning_rate": 9.608463116858544e-07, | |
| "loss": 1.3001, | |
| "step": 2701 | |
| }, | |
| { | |
| "epoch": 7.501369863013698, | |
| "grad_norm": 6.322762489318848, | |
| "learning_rate": 8.921154062638679e-07, | |
| "loss": 1.2816, | |
| "step": 2738 | |
| }, | |
| { | |
| "epoch": 7.504109589041096, | |
| "eval_loss": 1.3523932695388794, | |
| "eval_runtime": 41.8059, | |
| "eval_samples_per_second": 23.92, | |
| "eval_steps_per_second": 2.99, | |
| "step": 2739 | |
| } | |
| ], | |
| "logging_steps": 37, | |
| "max_steps": 3650, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 913, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.607336348896461e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |