jako_mbartLarge_6p_run1 / trainer_state.json
yesj1234's picture
Upload folder using huggingface_hub
d218160
{
"best_metric": 1.1539206504821777,
"best_model_checkpoint": "./jako_mbartLarge_6p_run1/checkpoint-4000",
"epoch": 3.8396928245740343,
"eval_steps": 1000,
"global_step": 8000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.24,
"learning_rate": 4.951290793960059e-05,
"loss": 1.8861,
"step": 500
},
{
"epoch": 0.48,
"learning_rate": 4.829517778860205e-05,
"loss": 1.4641,
"step": 1000
},
{
"epoch": 0.48,
"eval_bleu": 21.6162,
"eval_gen_len": 19.4434,
"eval_loss": 1.3276299238204956,
"eval_runtime": 299.0357,
"eval_samples_per_second": 13.931,
"eval_steps_per_second": 0.873,
"step": 1000
},
{
"epoch": 0.72,
"learning_rate": 4.707744763760351e-05,
"loss": 1.3282,
"step": 1500
},
{
"epoch": 0.96,
"learning_rate": 4.585971748660497e-05,
"loss": 1.2615,
"step": 2000
},
{
"epoch": 0.96,
"eval_bleu": 24.346,
"eval_gen_len": 19.4734,
"eval_loss": 1.186624526977539,
"eval_runtime": 297.0522,
"eval_samples_per_second": 14.024,
"eval_steps_per_second": 0.879,
"step": 2000
},
{
"epoch": 1.2,
"learning_rate": 4.4641987335606436e-05,
"loss": 1.0805,
"step": 2500
},
{
"epoch": 1.44,
"learning_rate": 4.342425718460789e-05,
"loss": 0.9103,
"step": 3000
},
{
"epoch": 1.44,
"eval_bleu": 25.4249,
"eval_gen_len": 19.0086,
"eval_loss": 1.1637648344039917,
"eval_runtime": 293.4921,
"eval_samples_per_second": 14.195,
"eval_steps_per_second": 0.889,
"step": 3000
},
{
"epoch": 1.68,
"learning_rate": 4.2206527033609356e-05,
"loss": 0.8534,
"step": 3500
},
{
"epoch": 1.92,
"learning_rate": 4.0988796882610817e-05,
"loss": 0.8285,
"step": 4000
},
{
"epoch": 1.92,
"eval_bleu": 26.2658,
"eval_gen_len": 19.3961,
"eval_loss": 1.1539206504821777,
"eval_runtime": 298.2089,
"eval_samples_per_second": 13.97,
"eval_steps_per_second": 0.875,
"step": 4000
},
{
"epoch": 2.16,
"learning_rate": 3.977106673161228e-05,
"loss": 0.7521,
"step": 4500
},
{
"epoch": 2.4,
"learning_rate": 3.855333658061374e-05,
"loss": 0.5977,
"step": 5000
},
{
"epoch": 2.4,
"eval_bleu": 25.5651,
"eval_gen_len": 19.6248,
"eval_loss": 1.1977771520614624,
"eval_runtime": 299.5483,
"eval_samples_per_second": 13.908,
"eval_steps_per_second": 0.871,
"step": 5000
},
{
"epoch": 2.64,
"learning_rate": 3.73356064296152e-05,
"loss": 0.5686,
"step": 5500
},
{
"epoch": 2.88,
"learning_rate": 3.611787627861666e-05,
"loss": 0.5423,
"step": 6000
},
{
"epoch": 2.88,
"eval_bleu": 26.8441,
"eval_gen_len": 19.1349,
"eval_loss": 1.1830259561538696,
"eval_runtime": 285.8007,
"eval_samples_per_second": 14.577,
"eval_steps_per_second": 0.913,
"step": 6000
},
{
"epoch": 3.12,
"learning_rate": 3.4900146127618125e-05,
"loss": 0.5099,
"step": 6500
},
{
"epoch": 3.36,
"learning_rate": 3.368241597661958e-05,
"loss": 0.3816,
"step": 7000
},
{
"epoch": 3.36,
"eval_bleu": 26.1301,
"eval_gen_len": 19.1207,
"eval_loss": 1.266960620880127,
"eval_runtime": 292.1624,
"eval_samples_per_second": 14.259,
"eval_steps_per_second": 0.893,
"step": 7000
},
{
"epoch": 3.6,
"learning_rate": 3.2464685825621045e-05,
"loss": 0.3637,
"step": 7500
},
{
"epoch": 3.84,
"learning_rate": 3.1246955674622506e-05,
"loss": 0.3412,
"step": 8000
},
{
"epoch": 3.84,
"eval_bleu": 26.7783,
"eval_gen_len": 19.2417,
"eval_loss": 1.2869776487350464,
"eval_runtime": 291.2259,
"eval_samples_per_second": 14.305,
"eval_steps_per_second": 0.896,
"step": 8000
},
{
"epoch": 3.84,
"step": 8000,
"total_flos": 2.7749663225623347e+17,
"train_loss": 0.8543571968078614,
"train_runtime": 11541.5643,
"train_samples_per_second": 28.881,
"train_steps_per_second": 1.805
}
],
"logging_steps": 500,
"max_steps": 20830,
"num_train_epochs": 10,
"save_steps": 1000,
"total_flos": 2.7749663225623347e+17,
"trial_name": null,
"trial_params": null
}