NeuroSkeptic / trainer_state.json
mipatov's picture
model
d1d9db8
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"global_step": 15,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"learning_rate": 9.333333333333334e-06,
"loss": 3.6363,
"step": 3
},
{
"epoch": 1.0,
"eval_accuracy": 0.40818656229615136,
"eval_loss": 3.208984375,
"eval_runtime": 9.2632,
"eval_samples_per_second": 2.591,
"eval_steps_per_second": 0.108,
"step": 3
},
{
"epoch": 2.0,
"learning_rate": 7.333333333333333e-06,
"loss": 2.8168,
"step": 6
},
{
"epoch": 2.0,
"eval_accuracy": 0.4873613829093281,
"eval_loss": 2.48046875,
"eval_runtime": 7.2903,
"eval_samples_per_second": 3.292,
"eval_steps_per_second": 0.137,
"step": 6
},
{
"epoch": 3.0,
"learning_rate": 5.333333333333334e-06,
"loss": 2.3529,
"step": 9
},
{
"epoch": 3.0,
"eval_accuracy": 0.49151989562948467,
"eval_loss": 2.421875,
"eval_runtime": 7.8557,
"eval_samples_per_second": 3.055,
"eval_steps_per_second": 0.127,
"step": 9
},
{
"epoch": 4.0,
"learning_rate": 3.3333333333333333e-06,
"loss": 2.1842,
"step": 12
},
{
"epoch": 4.0,
"eval_accuracy": 0.4991030658838878,
"eval_loss": 2.40234375,
"eval_runtime": 7.3477,
"eval_samples_per_second": 3.266,
"eval_steps_per_second": 0.136,
"step": 12
},
{
"epoch": 5.0,
"learning_rate": 1.3333333333333334e-06,
"loss": 2.0765,
"step": 15
},
{
"epoch": 5.0,
"eval_accuracy": 0.5020384866275277,
"eval_loss": 2.396484375,
"eval_runtime": 7.3064,
"eval_samples_per_second": 3.285,
"eval_steps_per_second": 0.137,
"step": 15
},
{
"before_init_mem_cpu": 19409301504,
"before_init_mem_gpu": 514785792,
"epoch": 5.0,
"init_mem_cpu_alloc_delta": 20480,
"init_mem_cpu_peaked_delta": 0,
"init_mem_gpu_alloc_delta": 0,
"init_mem_gpu_peaked_delta": 0,
"step": 15,
"total_flos": 6182393610240.0,
"train_loss": 2.6133464177449546,
"train_mem_cpu_alloc_delta": 87850864640,
"train_mem_cpu_peaked_delta": 4164583424,
"train_mem_gpu_alloc_delta": -456804352,
"train_mem_gpu_peaked_delta": 13894116352,
"train_runtime": 736.506,
"train_samples_per_second": 1.12,
"train_steps_per_second": 0.02
}
],
"max_steps": 15,
"num_train_epochs": 5,
"total_flos": 6182393610240.0,
"trial_name": null,
"trial_params": null
}