dummy / checkpoint-70 /trainer_state.json
abhinavp's picture
Training in progress, step 70, checkpoint
b22ce0e
raw
history blame
2.71 kB
{
"best_metric": 8.279220581054688,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/2023-11-20/14-54-26/checkpoints/checkpoint-70",
"epoch": 0.56,
"eval_steps": 10,
"global_step": 70,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.08,
"learning_rate": 4.600000000000001e-05,
"loss": 10.8138,
"step": 10
},
{
"epoch": 0.08,
"eval_loss": 10.804353713989258,
"eval_runtime": 2.6966,
"eval_samples_per_second": 370.843,
"eval_steps_per_second": 46.355,
"step": 10
},
{
"epoch": 0.16,
"learning_rate": 4.2e-05,
"loss": 10.7944,
"step": 20
},
{
"epoch": 0.16,
"eval_loss": 10.780308723449707,
"eval_runtime": 2.9168,
"eval_samples_per_second": 342.842,
"eval_steps_per_second": 42.855,
"step": 20
},
{
"epoch": 0.24,
"learning_rate": 3.8e-05,
"loss": 10.7582,
"step": 30
},
{
"epoch": 0.24,
"eval_loss": 10.715201377868652,
"eval_runtime": 2.6781,
"eval_samples_per_second": 373.4,
"eval_steps_per_second": 46.675,
"step": 30
},
{
"epoch": 0.32,
"learning_rate": 3.4000000000000007e-05,
"loss": 10.5902,
"step": 40
},
{
"epoch": 0.32,
"eval_loss": 10.216275215148926,
"eval_runtime": 2.8449,
"eval_samples_per_second": 351.507,
"eval_steps_per_second": 43.938,
"step": 40
},
{
"epoch": 0.4,
"learning_rate": 3e-05,
"loss": 9.8089,
"step": 50
},
{
"epoch": 0.4,
"eval_loss": 9.195159912109375,
"eval_runtime": 2.6916,
"eval_samples_per_second": 371.531,
"eval_steps_per_second": 46.441,
"step": 50
},
{
"epoch": 0.48,
"learning_rate": 2.6000000000000002e-05,
"loss": 9.0218,
"step": 60
},
{
"epoch": 0.48,
"eval_loss": 8.602910041809082,
"eval_runtime": 2.6711,
"eval_samples_per_second": 374.379,
"eval_steps_per_second": 46.797,
"step": 60
},
{
"epoch": 0.56,
"learning_rate": 2.2000000000000003e-05,
"loss": 8.5416,
"step": 70
},
{
"epoch": 0.56,
"eval_loss": 8.279220581054688,
"eval_runtime": 2.7623,
"eval_samples_per_second": 362.019,
"eval_steps_per_second": 45.252,
"step": 70
}
],
"logging_steps": 10,
"max_steps": 125,
"num_train_epochs": 1,
"save_steps": 10,
"total_flos": 4923721736640.0,
"trial_name": null,
"trial_params": null
}