bert-qasper / trainer_state.json
popos
train results without checkpoints
e4ce187
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"global_step": 7378,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14,
"learning_rate": 2.796692870696666e-05,
"loss": 2.2318,
"step": 500
},
{
"epoch": 0.27,
"learning_rate": 2.5933857413933314e-05,
"loss": 1.4035,
"step": 1000
},
{
"epoch": 0.41,
"learning_rate": 2.3900786120899972e-05,
"loss": 1.2463,
"step": 1500
},
{
"epoch": 0.54,
"learning_rate": 2.186771482786663e-05,
"loss": 1.1831,
"step": 2000
},
{
"epoch": 0.68,
"learning_rate": 1.983464353483329e-05,
"loss": 1.1009,
"step": 2500
},
{
"epoch": 0.81,
"learning_rate": 1.7801572241799947e-05,
"loss": 1.1077,
"step": 3000
},
{
"epoch": 0.95,
"learning_rate": 1.5768500948766605e-05,
"loss": 1.0565,
"step": 3500
},
{
"epoch": 1.08,
"learning_rate": 1.3735429655733262e-05,
"loss": 0.8975,
"step": 4000
},
{
"epoch": 1.22,
"learning_rate": 1.1702358362699918e-05,
"loss": 0.7853,
"step": 4500
},
{
"epoch": 1.36,
"learning_rate": 9.669287069666576e-06,
"loss": 0.7936,
"step": 5000
},
{
"epoch": 1.49,
"learning_rate": 7.636215776633235e-06,
"loss": 0.7883,
"step": 5500
},
{
"epoch": 1.63,
"learning_rate": 5.603144483599892e-06,
"loss": 0.7723,
"step": 6000
},
{
"epoch": 1.76,
"learning_rate": 3.5700731905665493e-06,
"loss": 0.7556,
"step": 6500
},
{
"epoch": 1.9,
"learning_rate": 1.5370018975332069e-06,
"loss": 0.754,
"step": 7000
},
{
"epoch": 2.0,
"step": 7378,
"total_flos": 3.4696551139946496e+16,
"train_loss": 1.046734738265896,
"train_runtime": 47684.7355,
"train_samples_per_second": 3.713,
"train_steps_per_second": 0.155
}
],
"max_steps": 7378,
"num_train_epochs": 2,
"total_flos": 3.4696551139946496e+16,
"trial_name": null,
"trial_params": null
}