trainer_state.json

{
 "best_metric": 1.1342445611953735,
 "best_model_checkpoint": "data/Mistral-7B_task-1_60-samples_config-2/checkpoint-17",
 "epoch": 12.869565217391305,
 "eval_steps": 500,
 "global_step": 37,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.34782608695652173,
 "grad_norm": 5.7090020179748535,
 "learning_rate": 1e-05,
 "loss": 2.2744,
 "step": 1
 },
 {
 "epoch": 0.6956521739130435,
 "grad_norm": 5.940366744995117,
 "learning_rate": 2e-05,
 "loss": 2.4067,
 "step": 2
 },
 {
 "epoch": 0.6956521739130435,
 "eval_loss": 2.121497392654419,
 "eval_runtime": 4.8521,
 "eval_samples_per_second": 2.473,
 "eval_steps_per_second": 2.473,
 "step": 2
 },
 {
 "epoch": 1.391304347826087,
 "grad_norm": 5.272120475769043,
 "learning_rate": 4e-05,
 "loss": 2.2389,
 "step": 4
 },
 {
 "epoch": 1.7391304347826086,
 "eval_loss": 1.6696044206619263,
 "eval_runtime": 4.8505,
 "eval_samples_per_second": 2.474,
 "eval_steps_per_second": 2.474,
 "step": 5
 },
 {
 "epoch": 2.0869565217391304,
 "grad_norm": 3.463669538497925,
 "learning_rate": 6e-05,
 "loss": 1.7683,
 "step": 6
 },
 {
 "epoch": 2.782608695652174,
 "grad_norm": 1.782424807548523,
 "learning_rate": 8e-05,
 "loss": 1.4281,
 "step": 8
 },
 {
 "epoch": 2.782608695652174,
 "eval_loss": 1.3318464756011963,
 "eval_runtime": 4.8402,
 "eval_samples_per_second": 2.479,
 "eval_steps_per_second": 2.479,
 "step": 8
 },
 {
 "epoch": 3.4782608695652173,
 "grad_norm": 1.1576231718063354,
 "learning_rate": 0.0001,
 "loss": 1.1834,
 "step": 10
 },
 {
 "epoch": 3.8260869565217392,
 "eval_loss": 1.26274836063385,
 "eval_runtime": 4.848,
 "eval_samples_per_second": 2.475,
 "eval_steps_per_second": 2.475,
 "step": 11
 },
 {
 "epoch": 4.173913043478261,
 "grad_norm": 1.503166675567627,
 "learning_rate": 9.987820251299122e-05,
 "loss": 1.1743,
 "step": 12
 },
 {
 "epoch": 4.869565217391305,
 "grad_norm": 1.343213677406311,
 "learning_rate": 9.951340343707852e-05,
 "loss": 1.0416,
 "step": 14
 },
 {
 "epoch": 4.869565217391305,
 "eval_loss": 1.1743342876434326,
 "eval_runtime": 4.8537,
 "eval_samples_per_second": 2.472,
 "eval_steps_per_second": 2.472,
 "step": 14
 },
 {
 "epoch": 5.565217391304348,
 "grad_norm": 1.2101751565933228,
 "learning_rate": 9.890738003669029e-05,
 "loss": 0.9159,
 "step": 16
 },
 {
 "epoch": 5.913043478260869,
 "eval_loss": 1.1342445611953735,
 "eval_runtime": 4.8437,
 "eval_samples_per_second": 2.477,
 "eval_steps_per_second": 2.477,
 "step": 17
 },
 {
 "epoch": 6.260869565217392,
 "grad_norm": 0.9623075723648071,
 "learning_rate": 9.806308479691595e-05,
 "loss": 0.7135,
 "step": 18
 },
 {
 "epoch": 6.956521739130435,
 "grad_norm": 1.0475118160247803,
 "learning_rate": 9.698463103929542e-05,
 "loss": 0.6449,
 "step": 20
 },
 {
 "epoch": 6.956521739130435,
 "eval_loss": 1.2086139917373657,
 "eval_runtime": 4.8507,
 "eval_samples_per_second": 2.474,
 "eval_steps_per_second": 2.474,
 "step": 20
 },
 {
 "epoch": 7.6521739130434785,
 "grad_norm": 1.033207654953003,
 "learning_rate": 9.567727288213005e-05,
 "loss": 0.5118,
 "step": 22
 },
 {
 "epoch": 8.0,
 "eval_loss": 1.4013780355453491,
 "eval_runtime": 4.8413,
 "eval_samples_per_second": 2.479,
 "eval_steps_per_second": 2.479,
 "step": 23
 },
 {
 "epoch": 8.347826086956522,
 "grad_norm": 1.3158878087997437,
 "learning_rate": 9.414737964294636e-05,
 "loss": 0.3817,
 "step": 24
 },
 {
 "epoch": 8.695652173913043,
 "eval_loss": 1.588330626487732,
 "eval_runtime": 4.8467,
 "eval_samples_per_second": 2.476,
 "eval_steps_per_second": 2.476,
 "step": 25
 },
 {
 "epoch": 9.043478260869565,
 "grad_norm": 1.5955661535263062,
 "learning_rate": 9.24024048078213e-05,
 "loss": 0.252,
 "step": 26
 },
 {
 "epoch": 9.73913043478261,
 "grad_norm": 1.8562427759170532,
 "learning_rate": 9.045084971874738e-05,
 "loss": 0.1478,
 "step": 28
 },
 {
 "epoch": 9.73913043478261,
 "eval_loss": 1.8900376558303833,
 "eval_runtime": 4.8472,
 "eval_samples_per_second": 2.476,
 "eval_steps_per_second": 2.476,
 "step": 28
 },
 {
 "epoch": 10.434782608695652,
 "grad_norm": 1.258542537689209,
 "learning_rate": 8.83022221559489e-05,
 "loss": 0.0974,
 "step": 30
 },
 {
 "epoch": 10.782608695652174,
 "eval_loss": 2.1700103282928467,
 "eval_runtime": 4.8533,
 "eval_samples_per_second": 2.473,
 "eval_steps_per_second": 2.473,
 "step": 31
 },
 {
 "epoch": 11.130434782608695,
 "grad_norm": 2.9292681217193604,
 "learning_rate": 8.596699001693255e-05,
 "loss": 0.0797,
 "step": 32
 },
 {
 "epoch": 11.826086956521738,
 "grad_norm": 2.1792283058166504,
 "learning_rate": 8.345653031794292e-05,
 "loss": 0.0395,
 "step": 34
 },
 {
 "epoch": 11.826086956521738,
 "eval_loss": 2.3129453659057617,
 "eval_runtime": 4.8304,
 "eval_samples_per_second": 2.484,
 "eval_steps_per_second": 2.484,
 "step": 34
 },
 {
 "epoch": 12.521739130434783,
 "grad_norm": 0.9826315641403198,
 "learning_rate": 8.07830737662829e-05,
 "loss": 0.0325,
 "step": 36
 },
 {
 "epoch": 12.869565217391305,
 "eval_loss": 2.348360061645508,
 "eval_runtime": 4.8265,
 "eval_samples_per_second": 2.486,
 "eval_steps_per_second": 2.486,
 "step": 37
 },
 {
 "epoch": 12.869565217391305,
 "step": 37,
 "total_flos": 8750092630622208.0,
 "train_loss": 0.8113765451755073,
 "train_runtime": 700.6388,
 "train_samples_per_second": 3.283,
 "train_steps_per_second": 0.143
 }
 ],
 "logging_steps": 2,
 "max_steps": 100,
 "num_input_tokens_seen": 0,
 "num_train_epochs": 50,
 "save_steps": 25,
 "stateful_callbacks": {
 "EarlyStoppingCallback": {
 "args": {
 "early_stopping_patience": 7,
 "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": true
 },
 "attributes": {}
 }
 },
 "total_flos": 8750092630622208.0,
 "train_batch_size": 1,
 "trial_name": null,
 "trial_params": null
}