qa3k / trainer_state.json
shrinath-suresh's picture
End of training
4125d38
raw
history blame
6.5 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 100.0,
"global_step": 24100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.07,
"learning_rate": 2.937759336099585e-05,
"loss": 0.3885,
"step": 500
},
{
"epoch": 4.15,
"learning_rate": 2.8755186721991704e-05,
"loss": 0.2791,
"step": 1000
},
{
"epoch": 6.22,
"learning_rate": 2.8132780082987553e-05,
"loss": 0.2363,
"step": 1500
},
{
"epoch": 8.3,
"learning_rate": 2.75103734439834e-05,
"loss": 0.2137,
"step": 2000
},
{
"epoch": 10.37,
"learning_rate": 2.6887966804979253e-05,
"loss": 0.1969,
"step": 2500
},
{
"epoch": 12.45,
"learning_rate": 2.6265560165975105e-05,
"loss": 0.1854,
"step": 3000
},
{
"epoch": 14.52,
"learning_rate": 2.5643153526970954e-05,
"loss": 0.1694,
"step": 3500
},
{
"epoch": 16.6,
"learning_rate": 2.5020746887966805e-05,
"loss": 0.1722,
"step": 4000
},
{
"epoch": 18.67,
"learning_rate": 2.4398340248962657e-05,
"loss": 0.1514,
"step": 4500
},
{
"epoch": 20.75,
"learning_rate": 2.3775933609958506e-05,
"loss": 0.1499,
"step": 5000
},
{
"epoch": 22.82,
"learning_rate": 2.3153526970954358e-05,
"loss": 0.1464,
"step": 5500
},
{
"epoch": 24.9,
"learning_rate": 2.253112033195021e-05,
"loss": 0.1343,
"step": 6000
},
{
"epoch": 26.97,
"learning_rate": 2.1908713692946058e-05,
"loss": 0.1304,
"step": 6500
},
{
"epoch": 29.05,
"learning_rate": 2.128630705394191e-05,
"loss": 0.1243,
"step": 7000
},
{
"epoch": 31.12,
"learning_rate": 2.0663900414937758e-05,
"loss": 0.1165,
"step": 7500
},
{
"epoch": 33.2,
"learning_rate": 2.004149377593361e-05,
"loss": 0.119,
"step": 8000
},
{
"epoch": 35.27,
"learning_rate": 1.9419087136929462e-05,
"loss": 0.112,
"step": 8500
},
{
"epoch": 37.34,
"learning_rate": 1.879668049792531e-05,
"loss": 0.1061,
"step": 9000
},
{
"epoch": 39.42,
"learning_rate": 1.8174273858921162e-05,
"loss": 0.1048,
"step": 9500
},
{
"epoch": 41.49,
"learning_rate": 1.7551867219917014e-05,
"loss": 0.1008,
"step": 10000
},
{
"epoch": 43.57,
"learning_rate": 1.6929460580912863e-05,
"loss": 0.0924,
"step": 10500
},
{
"epoch": 45.64,
"learning_rate": 1.630705394190871e-05,
"loss": 0.0912,
"step": 11000
},
{
"epoch": 47.72,
"learning_rate": 1.5684647302904566e-05,
"loss": 0.0942,
"step": 11500
},
{
"epoch": 49.79,
"learning_rate": 1.5062240663900415e-05,
"loss": 0.0855,
"step": 12000
},
{
"epoch": 51.87,
"learning_rate": 1.4439834024896267e-05,
"loss": 0.0855,
"step": 12500
},
{
"epoch": 53.94,
"learning_rate": 1.3817427385892115e-05,
"loss": 0.0801,
"step": 13000
},
{
"epoch": 56.02,
"learning_rate": 1.3195020746887967e-05,
"loss": 0.0787,
"step": 13500
},
{
"epoch": 58.09,
"learning_rate": 1.2572614107883817e-05,
"loss": 0.0777,
"step": 14000
},
{
"epoch": 60.17,
"learning_rate": 1.1950207468879667e-05,
"loss": 0.0733,
"step": 14500
},
{
"epoch": 62.24,
"learning_rate": 1.132780082987552e-05,
"loss": 0.0722,
"step": 15000
},
{
"epoch": 64.32,
"learning_rate": 1.070539419087137e-05,
"loss": 0.0713,
"step": 15500
},
{
"epoch": 66.39,
"learning_rate": 1.0082987551867221e-05,
"loss": 0.0679,
"step": 16000
},
{
"epoch": 68.46,
"learning_rate": 9.46058091286307e-06,
"loss": 0.0685,
"step": 16500
},
{
"epoch": 70.54,
"learning_rate": 8.838174273858922e-06,
"loss": 0.0656,
"step": 17000
},
{
"epoch": 72.61,
"learning_rate": 8.215767634854773e-06,
"loss": 0.0657,
"step": 17500
},
{
"epoch": 74.69,
"learning_rate": 7.593360995850622e-06,
"loss": 0.0611,
"step": 18000
},
{
"epoch": 76.76,
"learning_rate": 6.970954356846474e-06,
"loss": 0.061,
"step": 18500
},
{
"epoch": 78.84,
"learning_rate": 6.348547717842324e-06,
"loss": 0.0605,
"step": 19000
},
{
"epoch": 80.91,
"learning_rate": 5.726141078838174e-06,
"loss": 0.0602,
"step": 19500
},
{
"epoch": 82.99,
"learning_rate": 5.103734439834025e-06,
"loss": 0.0585,
"step": 20000
},
{
"epoch": 85.06,
"learning_rate": 4.481327800829875e-06,
"loss": 0.0583,
"step": 20500
},
{
"epoch": 87.14,
"learning_rate": 3.858921161825726e-06,
"loss": 0.056,
"step": 21000
},
{
"epoch": 89.21,
"learning_rate": 3.236514522821577e-06,
"loss": 0.055,
"step": 21500
},
{
"epoch": 91.29,
"learning_rate": 2.6141078838174274e-06,
"loss": 0.0582,
"step": 22000
},
{
"epoch": 93.36,
"learning_rate": 1.991701244813278e-06,
"loss": 0.0556,
"step": 22500
},
{
"epoch": 95.44,
"learning_rate": 1.3692946058091286e-06,
"loss": 0.0545,
"step": 23000
},
{
"epoch": 97.51,
"learning_rate": 7.468879668049792e-07,
"loss": 0.0554,
"step": 23500
},
{
"epoch": 99.59,
"learning_rate": 1.2448132780082988e-07,
"loss": 0.0548,
"step": 24000
},
{
"epoch": 100.0,
"step": 24100,
"total_flos": 3.90325755445248e+16,
"train_loss": 0.1092665815749109,
"train_runtime": 5451.3155,
"train_samples_per_second": 52.905,
"train_steps_per_second": 4.421
}
],
"max_steps": 24100,
"num_train_epochs": 100,
"total_flos": 3.90325755445248e+16,
"trial_name": null,
"trial_params": null
}