3-allnli-p-a-p-a-p-a-64-128-3e-5 / trainer_state.json
Zhuoxu Huang
init
4d22cf8
{
"best_metric": 0.795758744982025,
"best_model_checkpoint": "3-allnli-p-a-p-a-p-a-64-128-3e-5-2600/checkpoint-2400",
"epoch": 1.0586319218241043,
"global_step": 2600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 1.5e-05,
"loss": 0.172,
"step": 100
},
{
"epoch": 0.08,
"learning_rate": 3e-05,
"loss": 0.0551,
"step": 200
},
{
"epoch": 0.12,
"learning_rate": 2.875e-05,
"loss": 0.0433,
"step": 300
},
{
"epoch": 0.16,
"learning_rate": 2.75e-05,
"loss": 0.0512,
"step": 400
},
{
"epoch": 0.2,
"learning_rate": 2.625e-05,
"loss": 0.0533,
"step": 500
},
{
"epoch": 0.24,
"learning_rate": 2.5e-05,
"loss": 0.0523,
"step": 600
},
{
"epoch": 0.29,
"learning_rate": 2.3749999999999998e-05,
"loss": 0.0643,
"step": 700
},
{
"epoch": 0.33,
"learning_rate": 2.25e-05,
"loss": 0.056,
"step": 800
},
{
"epoch": 0.37,
"learning_rate": 2.125e-05,
"loss": 0.0626,
"step": 900
},
{
"epoch": 0.41,
"learning_rate": 1.9999999999999998e-05,
"loss": 0.0658,
"step": 1000
},
{
"epoch": 0.45,
"learning_rate": 1.8750000000000002e-05,
"loss": 0.0658,
"step": 1100
},
{
"epoch": 0.49,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.0716,
"step": 1200
},
{
"epoch": 0.53,
"learning_rate": 1.625e-05,
"loss": 0.0779,
"step": 1300
},
{
"epoch": 0.57,
"learning_rate": 1.5e-05,
"loss": 0.0878,
"step": 1400
},
{
"epoch": 0.61,
"learning_rate": 1.375e-05,
"loss": 0.0931,
"step": 1500
},
{
"epoch": 0.65,
"learning_rate": 1.25e-05,
"loss": 0.0982,
"step": 1600
},
{
"epoch": 0.69,
"learning_rate": 1.125e-05,
"loss": 0.1042,
"step": 1700
},
{
"epoch": 0.73,
"learning_rate": 9.999999999999999e-06,
"loss": 0.1096,
"step": 1800
},
{
"epoch": 0.77,
"learning_rate": 8.750000000000001e-06,
"loss": 0.1249,
"step": 1900
},
{
"epoch": 0.81,
"learning_rate": 7.5e-06,
"loss": 0.1448,
"step": 2000
},
{
"epoch": 0.86,
"learning_rate": 6.25e-06,
"loss": 0.1516,
"step": 2100
},
{
"epoch": 0.9,
"learning_rate": 4.9999999999999996e-06,
"loss": 0.168,
"step": 2200
},
{
"epoch": 0.94,
"learning_rate": 3.75e-06,
"loss": 0.178,
"step": 2300
},
{
"epoch": 0.98,
"learning_rate": 2.4999999999999998e-06,
"loss": 0.1992,
"step": 2400
},
{
"epoch": 1.02,
"learning_rate": 1.2499999999999999e-06,
"loss": 0.1685,
"step": 2500
},
{
"epoch": 1.06,
"learning_rate": 0.0,
"loss": 0.1461,
"step": 2600
},
{
"epoch": 1.06,
"step": 2600,
"total_flos": 172275249862656.0,
"train_loss": 0.10250109305748573,
"train_runtime": 1493.1893,
"train_samples_per_second": 222.879,
"train_steps_per_second": 1.741
}
],
"max_steps": 2600,
"num_train_epochs": 2,
"total_flos": 172275249862656.0,
"trial_name": null,
"trial_params": null
}