DylanJHJ's picture
add fidmrg
7b27ed3
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.2597631645250693,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06,
"learning_rate": 9.5e-05,
"loss": 37.0618,
"step": 500
},
{
"epoch": 0.06,
"eval_loss": 27.398916244506836,
"eval_runtime": 3.1349,
"eval_samples_per_second": 31.899,
"eval_steps_per_second": 4.147,
"step": 500
},
{
"epoch": 0.13,
"learning_rate": 9e-05,
"loss": 34.5311,
"step": 1000
},
{
"epoch": 0.13,
"eval_loss": 36.926673889160156,
"eval_runtime": 3.1612,
"eval_samples_per_second": 31.633,
"eval_steps_per_second": 4.112,
"step": 1000
},
{
"epoch": 0.19,
"learning_rate": 8.5e-05,
"loss": 33.5869,
"step": 1500
},
{
"epoch": 0.19,
"eval_loss": 31.76376724243164,
"eval_runtime": 3.1795,
"eval_samples_per_second": 31.452,
"eval_steps_per_second": 4.089,
"step": 1500
},
{
"epoch": 0.25,
"learning_rate": 8e-05,
"loss": 34.2002,
"step": 2000
},
{
"epoch": 0.25,
"eval_loss": 33.623844146728516,
"eval_runtime": 3.1768,
"eval_samples_per_second": 31.479,
"eval_steps_per_second": 4.092,
"step": 2000
},
{
"epoch": 0.31,
"learning_rate": 7.500000000000001e-05,
"loss": 34.0832,
"step": 2500
},
{
"epoch": 0.31,
"eval_loss": 27.767799377441406,
"eval_runtime": 3.166,
"eval_samples_per_second": 31.586,
"eval_steps_per_second": 4.106,
"step": 2500
},
{
"epoch": 0.38,
"learning_rate": 7e-05,
"loss": 32.413,
"step": 3000
},
{
"epoch": 0.38,
"eval_loss": 34.710330963134766,
"eval_runtime": 3.1812,
"eval_samples_per_second": 31.434,
"eval_steps_per_second": 4.086,
"step": 3000
},
{
"epoch": 0.44,
"learning_rate": 6.500000000000001e-05,
"loss": 33.3394,
"step": 3500
},
{
"epoch": 0.44,
"eval_loss": 29.091354370117188,
"eval_runtime": 3.1601,
"eval_samples_per_second": 31.645,
"eval_steps_per_second": 4.114,
"step": 3500
},
{
"epoch": 0.5,
"learning_rate": 6e-05,
"loss": 32.6119,
"step": 4000
},
{
"epoch": 0.5,
"eval_loss": 30.3073673248291,
"eval_runtime": 3.1555,
"eval_samples_per_second": 31.691,
"eval_steps_per_second": 4.12,
"step": 4000
},
{
"epoch": 0.57,
"learning_rate": 5.500000000000001e-05,
"loss": 33.4702,
"step": 4500
},
{
"epoch": 0.57,
"eval_loss": 30.2010555267334,
"eval_runtime": 3.1576,
"eval_samples_per_second": 31.67,
"eval_steps_per_second": 4.117,
"step": 4500
},
{
"epoch": 0.63,
"learning_rate": 5e-05,
"loss": 32.9516,
"step": 5000
},
{
"epoch": 0.63,
"eval_loss": 28.628616333007812,
"eval_runtime": 3.1678,
"eval_samples_per_second": 31.568,
"eval_steps_per_second": 4.104,
"step": 5000
},
{
"epoch": 0.69,
"learning_rate": 4.5e-05,
"loss": 32.9256,
"step": 5500
},
{
"epoch": 0.69,
"eval_loss": 29.695850372314453,
"eval_runtime": 3.1616,
"eval_samples_per_second": 31.63,
"eval_steps_per_second": 4.112,
"step": 5500
},
{
"epoch": 0.76,
"learning_rate": 4e-05,
"loss": 31.5253,
"step": 6000
},
{
"epoch": 0.76,
"eval_loss": 24.25625228881836,
"eval_runtime": 3.1405,
"eval_samples_per_second": 31.842,
"eval_steps_per_second": 4.14,
"step": 6000
},
{
"epoch": 0.82,
"learning_rate": 3.5e-05,
"loss": 34.0972,
"step": 6500
},
{
"epoch": 0.82,
"eval_loss": 30.526784896850586,
"eval_runtime": 3.1519,
"eval_samples_per_second": 31.727,
"eval_steps_per_second": 4.125,
"step": 6500
},
{
"epoch": 0.88,
"learning_rate": 3e-05,
"loss": 32.1422,
"step": 7000
},
{
"epoch": 0.88,
"eval_loss": 29.38189697265625,
"eval_runtime": 3.164,
"eval_samples_per_second": 31.606,
"eval_steps_per_second": 4.109,
"step": 7000
},
{
"epoch": 0.94,
"learning_rate": 2.5e-05,
"loss": 32.7259,
"step": 7500
},
{
"epoch": 0.94,
"eval_loss": 31.17078399658203,
"eval_runtime": 3.1747,
"eval_samples_per_second": 31.499,
"eval_steps_per_second": 4.095,
"step": 7500
},
{
"epoch": 1.01,
"learning_rate": 2e-05,
"loss": 31.0833,
"step": 8000
},
{
"epoch": 1.01,
"eval_loss": 28.586397171020508,
"eval_runtime": 3.1531,
"eval_samples_per_second": 31.715,
"eval_steps_per_second": 4.123,
"step": 8000
},
{
"epoch": 1.07,
"learning_rate": 1.5e-05,
"loss": 31.1692,
"step": 8500
},
{
"epoch": 1.07,
"eval_loss": 32.3286247253418,
"eval_runtime": 3.1659,
"eval_samples_per_second": 31.586,
"eval_steps_per_second": 4.106,
"step": 8500
},
{
"epoch": 1.13,
"learning_rate": 1e-05,
"loss": 31.318,
"step": 9000
},
{
"epoch": 1.13,
"eval_loss": 30.532176971435547,
"eval_runtime": 3.1657,
"eval_samples_per_second": 31.588,
"eval_steps_per_second": 4.106,
"step": 9000
},
{
"epoch": 1.2,
"learning_rate": 5e-06,
"loss": 30.6785,
"step": 9500
},
{
"epoch": 1.2,
"eval_loss": 29.415210723876953,
"eval_runtime": 3.1501,
"eval_samples_per_second": 31.745,
"eval_steps_per_second": 4.127,
"step": 9500
},
{
"epoch": 1.26,
"learning_rate": 0.0,
"loss": 31.4601,
"step": 10000
},
{
"epoch": 1.26,
"eval_loss": 31.739580154418945,
"eval_runtime": 3.1633,
"eval_samples_per_second": 31.612,
"eval_steps_per_second": 4.11,
"step": 10000
}
],
"logging_steps": 500,
"max_steps": 10000,
"num_train_epochs": 2,
"save_steps": 5000,
"total_flos": 1.3694633480945664e+17,
"trial_name": null,
"trial_params": null
}