mbert-xdm-en-it / trainer_state.json
Nikita Moghe Moghe
First model version
15e9699
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 10.0,
"global_step": 125000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.2,
"learning_rate": 4.9e-05,
"loss": 2.4248,
"step": 2500
},
{
"epoch": 0.4,
"learning_rate": 4.8e-05,
"loss": 2.2834,
"step": 5000
},
{
"epoch": 0.6,
"learning_rate": 4.7e-05,
"loss": 2.2008,
"step": 7500
},
{
"epoch": 0.8,
"learning_rate": 4.600000000000001e-05,
"loss": 2.1521,
"step": 10000
},
{
"epoch": 1.0,
"learning_rate": 4.5e-05,
"loss": 2.1051,
"step": 12500
},
{
"epoch": 1.2,
"learning_rate": 4.4000000000000006e-05,
"loss": 2.035,
"step": 15000
},
{
"epoch": 1.4,
"learning_rate": 4.3e-05,
"loss": 2.0125,
"step": 17500
},
{
"epoch": 1.6,
"learning_rate": 4.2e-05,
"loss": 1.9866,
"step": 20000
},
{
"epoch": 1.8,
"learning_rate": 4.1e-05,
"loss": 1.9653,
"step": 22500
},
{
"epoch": 2.0,
"learning_rate": 4e-05,
"loss": 1.938,
"step": 25000
},
{
"epoch": 2.2,
"learning_rate": 3.9000000000000006e-05,
"loss": 1.8827,
"step": 27500
},
{
"epoch": 2.4,
"learning_rate": 3.8e-05,
"loss": 1.8802,
"step": 30000
},
{
"epoch": 2.6,
"learning_rate": 3.7e-05,
"loss": 1.856,
"step": 32500
},
{
"epoch": 2.8,
"learning_rate": 3.6e-05,
"loss": 1.8387,
"step": 35000
},
{
"epoch": 3.0,
"learning_rate": 3.5e-05,
"loss": 1.8186,
"step": 37500
},
{
"epoch": 3.2,
"learning_rate": 3.4000000000000007e-05,
"loss": 1.7706,
"step": 40000
},
{
"epoch": 3.4,
"learning_rate": 3.3e-05,
"loss": 1.7607,
"step": 42500
},
{
"epoch": 3.6,
"learning_rate": 3.2000000000000005e-05,
"loss": 1.7582,
"step": 45000
},
{
"epoch": 3.8,
"learning_rate": 3.1e-05,
"loss": 1.7462,
"step": 47500
},
{
"epoch": 4.0,
"learning_rate": 3e-05,
"loss": 1.7304,
"step": 50000
},
{
"epoch": 4.2,
"learning_rate": 2.9e-05,
"loss": 1.6883,
"step": 52500
},
{
"epoch": 4.4,
"learning_rate": 2.8000000000000003e-05,
"loss": 1.6699,
"step": 55000
},
{
"epoch": 4.6,
"learning_rate": 2.7000000000000002e-05,
"loss": 1.6767,
"step": 57500
},
{
"epoch": 4.8,
"learning_rate": 2.6000000000000002e-05,
"loss": 1.6543,
"step": 60000
},
{
"epoch": 5.0,
"learning_rate": 2.5e-05,
"loss": 1.6477,
"step": 62500
},
{
"epoch": 5.2,
"learning_rate": 2.4e-05,
"loss": 1.616,
"step": 65000
},
{
"epoch": 5.4,
"learning_rate": 2.3000000000000003e-05,
"loss": 1.6056,
"step": 67500
},
{
"epoch": 5.6,
"learning_rate": 2.2000000000000003e-05,
"loss": 1.6012,
"step": 70000
},
{
"epoch": 5.8,
"learning_rate": 2.1e-05,
"loss": 1.5866,
"step": 72500
},
{
"epoch": 6.0,
"learning_rate": 2e-05,
"loss": 1.582,
"step": 75000
},
{
"epoch": 6.2,
"learning_rate": 1.9e-05,
"loss": 1.5482,
"step": 77500
},
{
"epoch": 6.4,
"learning_rate": 1.8e-05,
"loss": 1.5416,
"step": 80000
},
{
"epoch": 6.6,
"learning_rate": 1.7000000000000003e-05,
"loss": 1.5267,
"step": 82500
},
{
"epoch": 6.8,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.5236,
"step": 85000
},
{
"epoch": 7.0,
"learning_rate": 1.5e-05,
"loss": 1.5171,
"step": 87500
},
{
"epoch": 7.2,
"learning_rate": 1.4000000000000001e-05,
"loss": 1.4837,
"step": 90000
},
{
"epoch": 7.4,
"learning_rate": 1.3000000000000001e-05,
"loss": 1.4763,
"step": 92500
},
{
"epoch": 7.6,
"learning_rate": 1.2e-05,
"loss": 1.4712,
"step": 95000
},
{
"epoch": 7.8,
"learning_rate": 1.1000000000000001e-05,
"loss": 1.4608,
"step": 97500
},
{
"epoch": 8.0,
"learning_rate": 1e-05,
"loss": 1.4534,
"step": 100000
},
{
"epoch": 8.2,
"learning_rate": 9e-06,
"loss": 1.432,
"step": 102500
},
{
"epoch": 8.4,
"learning_rate": 8.000000000000001e-06,
"loss": 1.4253,
"step": 105000
},
{
"epoch": 8.6,
"learning_rate": 7.000000000000001e-06,
"loss": 1.4196,
"step": 107500
},
{
"epoch": 8.8,
"learning_rate": 6e-06,
"loss": 1.4144,
"step": 110000
},
{
"epoch": 9.0,
"learning_rate": 5e-06,
"loss": 1.4121,
"step": 112500
},
{
"epoch": 9.2,
"learning_rate": 4.000000000000001e-06,
"loss": 1.3911,
"step": 115000
},
{
"epoch": 9.4,
"learning_rate": 3e-06,
"loss": 1.3983,
"step": 117500
},
{
"epoch": 9.6,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.3846,
"step": 120000
},
{
"epoch": 9.8,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.384,
"step": 122500
},
{
"epoch": 10.0,
"learning_rate": 0.0,
"loss": 1.3779,
"step": 125000
},
{
"epoch": 10.0,
"step": 125000,
"train_runtime": 57584.0925,
"train_samples_per_second": 2.171
}
],
"max_steps": 125000,
"num_train_epochs": 10,
"total_flos": 152219401526604744,
"trial_name": null,
"trial_params": null
}