| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0, |
| "eval_steps": 500, |
| "global_step": 34, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.2623544037342072, |
| "epoch": 0.06153846153846154, |
| "grad_norm": 0.19557932019233704, |
| "learning_rate": 0.0, |
| "loss": 1.363303303718567, |
| "mean_token_accuracy": 0.6967771947383881, |
| "num_tokens": 8983.0, |
| "step": 1 |
| }, |
| { |
| "entropy": 1.2696461975574493, |
| "epoch": 0.12307692307692308, |
| "grad_norm": 0.2623090445995331, |
| "learning_rate": 0.0001, |
| "loss": 1.3970271348953247, |
| "mean_token_accuracy": 0.7032225728034973, |
| "num_tokens": 15469.0, |
| "step": 2 |
| }, |
| { |
| "entropy": 1.351563960313797, |
| "epoch": 0.18461538461538463, |
| "grad_norm": 0.2595767080783844, |
| "learning_rate": 0.0002, |
| "loss": 1.438194990158081, |
| "mean_token_accuracy": 0.6762479990720749, |
| "num_tokens": 23136.0, |
| "step": 3 |
| }, |
| { |
| "entropy": 1.156097874045372, |
| "epoch": 0.24615384615384617, |
| "grad_norm": 0.9452286958694458, |
| "learning_rate": 0.00019979453927503364, |
| "loss": 1.191060185432434, |
| "mean_token_accuracy": 0.7312490344047546, |
| "num_tokens": 31431.0, |
| "step": 4 |
| }, |
| { |
| "entropy": 1.2470719814300537, |
| "epoch": 0.3076923076923077, |
| "grad_norm": 0.29575249552726746, |
| "learning_rate": 0.0001991790013823246, |
| "loss": 1.2078289985656738, |
| "mean_token_accuracy": 0.7162232249975204, |
| "num_tokens": 39311.0, |
| "step": 5 |
| }, |
| { |
| "entropy": 1.3020514845848083, |
| "epoch": 0.36923076923076925, |
| "grad_norm": 0.12131688743829727, |
| "learning_rate": 0.00019815591569910654, |
| "loss": 1.236517310142517, |
| "mean_token_accuracy": 0.7079942375421524, |
| "num_tokens": 46157.0, |
| "step": 6 |
| }, |
| { |
| "entropy": 1.1494204998016357, |
| "epoch": 0.4307692307692308, |
| "grad_norm": 0.07860170304775238, |
| "learning_rate": 0.00019672948630390294, |
| "loss": 1.0905468463897705, |
| "mean_token_accuracy": 0.74454265832901, |
| "num_tokens": 55535.0, |
| "step": 7 |
| }, |
| { |
| "entropy": 1.0420133769512177, |
| "epoch": 0.49230769230769234, |
| "grad_norm": 0.09150254726409912, |
| "learning_rate": 0.00019490557470106686, |
| "loss": 1.0026195049285889, |
| "mean_token_accuracy": 0.7577853351831436, |
| "num_tokens": 63422.0, |
| "step": 8 |
| }, |
| { |
| "entropy": 1.0447807163000107, |
| "epoch": 0.5538461538461539, |
| "grad_norm": 0.10369911789894104, |
| "learning_rate": 0.0001926916757346022, |
| "loss": 1.0252422094345093, |
| "mean_token_accuracy": 0.7560650110244751, |
| "num_tokens": 70555.0, |
| "step": 9 |
| }, |
| { |
| "entropy": 1.0285034626722336, |
| "epoch": 0.6153846153846154, |
| "grad_norm": 0.11572559177875519, |
| "learning_rate": 0.0001900968867902419, |
| "loss": 1.052419900894165, |
| "mean_token_accuracy": 0.7432190179824829, |
| "num_tokens": 78003.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 1.0061500668525696, |
| "epoch": 0.676923076923077, |
| "grad_norm": 0.10965128242969513, |
| "learning_rate": 0.00018713187041233896, |
| "loss": 0.9987847805023193, |
| "mean_token_accuracy": 0.7628708183765411, |
| "num_tokens": 86065.0, |
| "step": 11 |
| }, |
| { |
| "entropy": 1.0449193269014359, |
| "epoch": 0.7384615384615385, |
| "grad_norm": 0.10073477029800415, |
| "learning_rate": 0.00018380881048918405, |
| "loss": 1.0622799396514893, |
| "mean_token_accuracy": 0.746578574180603, |
| "num_tokens": 93482.0, |
| "step": 12 |
| }, |
| { |
| "entropy": 0.8957007825374603, |
| "epoch": 0.8, |
| "grad_norm": 0.10005328059196472, |
| "learning_rate": 0.00018014136218679567, |
| "loss": 0.8889887928962708, |
| "mean_token_accuracy": 0.77289679646492, |
| "num_tokens": 100654.0, |
| "step": 13 |
| }, |
| { |
| "entropy": 0.9818997383117676, |
| "epoch": 0.8615384615384616, |
| "grad_norm": 0.12628009915351868, |
| "learning_rate": 0.00017614459583691346, |
| "loss": 0.976428210735321, |
| "mean_token_accuracy": 0.7596915662288666, |
| "num_tokens": 108091.0, |
| "step": 14 |
| }, |
| { |
| "entropy": 0.9654281437397003, |
| "epoch": 0.9230769230769231, |
| "grad_norm": 0.11806830018758774, |
| "learning_rate": 0.00017183493500977278, |
| "loss": 0.9671178460121155, |
| "mean_token_accuracy": 0.7701253145933151, |
| "num_tokens": 114707.0, |
| "step": 15 |
| }, |
| { |
| "entropy": 0.9159971326589584, |
| "epoch": 0.9846153846153847, |
| "grad_norm": 0.09751465171575546, |
| "learning_rate": 0.0001672300890261317, |
| "loss": 0.902397632598877, |
| "mean_token_accuracy": 0.7794631719589233, |
| "num_tokens": 121769.0, |
| "step": 16 |
| }, |
| { |
| "entropy": 1.1713899374008179, |
| "epoch": 1.0, |
| "grad_norm": 0.5108214020729065, |
| "learning_rate": 0.00016234898018587337, |
| "loss": 1.1008179187774658, |
| "mean_token_accuracy": 0.7128027677536011, |
| "num_tokens": 122059.0, |
| "step": 17 |
| }, |
| { |
| "entropy": 0.9143469482660294, |
| "epoch": 1.0615384615384615, |
| "grad_norm": 0.08762291073799133, |
| "learning_rate": 0.00015721166601221698, |
| "loss": 0.8315602540969849, |
| "mean_token_accuracy": 0.793931856751442, |
| "num_tokens": 128993.0, |
| "step": 18 |
| }, |
| { |
| "entropy": 0.8749453127384186, |
| "epoch": 1.123076923076923, |
| "grad_norm": 0.08239830285310745, |
| "learning_rate": 0.00015183925683105254, |
| "loss": 0.8309834003448486, |
| "mean_token_accuracy": 0.7881838083267212, |
| "num_tokens": 137231.0, |
| "step": 19 |
| }, |
| { |
| "entropy": 0.8598180264234543, |
| "epoch": 1.1846153846153846, |
| "grad_norm": 0.08933497220277786, |
| "learning_rate": 0.00014625382902408356, |
| "loss": 0.8326812982559204, |
| "mean_token_accuracy": 0.789072722196579, |
| "num_tokens": 145190.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 0.9529785364866257, |
| "epoch": 1.2461538461538462, |
| "grad_norm": 0.10007654130458832, |
| "learning_rate": 0.00014047833431223938, |
| "loss": 0.9316624999046326, |
| "mean_token_accuracy": 0.7664911448955536, |
| "num_tokens": 152202.0, |
| "step": 21 |
| }, |
| { |
| "entropy": 0.7975367158651352, |
| "epoch": 1.3076923076923077, |
| "grad_norm": 0.097578264772892, |
| "learning_rate": 0.00013453650544213076, |
| "loss": 0.7583404779434204, |
| "mean_token_accuracy": 0.8191335052251816, |
| "num_tokens": 158331.0, |
| "step": 22 |
| }, |
| { |
| "entropy": 0.8262939751148224, |
| "epoch": 1.3692307692307693, |
| "grad_norm": 0.08703760802745819, |
| "learning_rate": 0.00012845275866310324, |
| "loss": 0.8296029567718506, |
| "mean_token_accuracy": 0.7906320244073868, |
| "num_tokens": 166002.0, |
| "step": 23 |
| }, |
| { |
| "entropy": 0.9157162606716156, |
| "epoch": 1.4307692307692308, |
| "grad_norm": 0.09159163385629654, |
| "learning_rate": 0.00012225209339563145, |
| "loss": 0.9092485308647156, |
| "mean_token_accuracy": 0.7690109759569168, |
| "num_tokens": 175547.0, |
| "step": 24 |
| }, |
| { |
| "entropy": 0.7814479619264603, |
| "epoch": 1.4923076923076923, |
| "grad_norm": 0.0804533138871193, |
| "learning_rate": 0.00011595998950333793, |
| "loss": 0.7761488556861877, |
| "mean_token_accuracy": 0.7984755784273148, |
| "num_tokens": 184577.0, |
| "step": 25 |
| }, |
| { |
| "entropy": 0.7985701262950897, |
| "epoch": 1.5538461538461539, |
| "grad_norm": 0.0917685404419899, |
| "learning_rate": 0.00010960230259076818, |
| "loss": 0.7678935527801514, |
| "mean_token_accuracy": 0.8032983243465424, |
| "num_tokens": 191871.0, |
| "step": 26 |
| }, |
| { |
| "entropy": 0.9452711939811707, |
| "epoch": 1.6153846153846154, |
| "grad_norm": 0.10309547930955887, |
| "learning_rate": 0.00010320515775716555, |
| "loss": 0.9123054146766663, |
| "mean_token_accuracy": 0.7641141563653946, |
| "num_tokens": 199434.0, |
| "step": 27 |
| }, |
| { |
| "entropy": 0.9179778099060059, |
| "epoch": 1.676923076923077, |
| "grad_norm": 0.0981035828590393, |
| "learning_rate": 9.679484224283449e-05, |
| "loss": 0.8683165311813354, |
| "mean_token_accuracy": 0.7796975672245026, |
| "num_tokens": 206254.0, |
| "step": 28 |
| }, |
| { |
| "entropy": 0.9250332564115524, |
| "epoch": 1.7384615384615385, |
| "grad_norm": 0.08907375484704971, |
| "learning_rate": 9.039769740923183e-05, |
| "loss": 0.9016717672348022, |
| "mean_token_accuracy": 0.7748660296201706, |
| "num_tokens": 213583.0, |
| "step": 29 |
| }, |
| { |
| "entropy": 0.8822590410709381, |
| "epoch": 1.8, |
| "grad_norm": 0.09369569271802902, |
| "learning_rate": 8.404001049666211e-05, |
| "loss": 0.8634146451950073, |
| "mean_token_accuracy": 0.7857015281915665, |
| "num_tokens": 220995.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.8900642246007919, |
| "epoch": 1.8615384615384616, |
| "grad_norm": 0.1084684282541275, |
| "learning_rate": 7.774790660436858e-05, |
| "loss": 0.9049892425537109, |
| "mean_token_accuracy": 0.7676405757665634, |
| "num_tokens": 228050.0, |
| "step": 31 |
| }, |
| { |
| "entropy": 0.8679154813289642, |
| "epoch": 1.9230769230769231, |
| "grad_norm": 0.09301979839801788, |
| "learning_rate": 7.154724133689677e-05, |
| "loss": 0.8593922257423401, |
| "mean_token_accuracy": 0.7785892337560654, |
| "num_tokens": 235526.0, |
| "step": 32 |
| }, |
| { |
| "entropy": 0.8450447171926498, |
| "epoch": 1.9846153846153847, |
| "grad_norm": 0.0842856764793396, |
| "learning_rate": 6.546349455786926e-05, |
| "loss": 0.8173251152038574, |
| "mean_token_accuracy": 0.7973018437623978, |
| "num_tokens": 243742.0, |
| "step": 33 |
| }, |
| { |
| "entropy": 0.8048986196517944, |
| "epoch": 2.0, |
| "grad_norm": 0.3837374150753021, |
| "learning_rate": 5.952166568776062e-05, |
| "loss": 0.6818726658821106, |
| "mean_token_accuracy": 0.8026666641235352, |
| "num_tokens": 244118.0, |
| "step": 34 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 51, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2955225949673472.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
