mbart-sahitrans_new_data / trainer_state.json
balaramas's picture
Model save
530d025 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 37815,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 4.9338886685177844e-05,
"loss": 3.2991,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 4.8677773370355686e-05,
"loss": 2.7662,
"step": 1000
},
{
"epoch": 0.12,
"learning_rate": 4.801666005553352e-05,
"loss": 2.662,
"step": 1500
},
{
"epoch": 0.16,
"learning_rate": 4.735554674071136e-05,
"loss": 2.5305,
"step": 2000
},
{
"epoch": 0.2,
"learning_rate": 4.66944334258892e-05,
"loss": 2.4368,
"step": 2500
},
{
"epoch": 0.24,
"learning_rate": 4.603332011106704e-05,
"loss": 2.4286,
"step": 3000
},
{
"epoch": 0.28,
"learning_rate": 4.5372206796244874e-05,
"loss": 2.3404,
"step": 3500
},
{
"epoch": 0.32,
"learning_rate": 4.4711093481422716e-05,
"loss": 2.2398,
"step": 4000
},
{
"epoch": 0.36,
"learning_rate": 4.404998016660056e-05,
"loss": 2.2966,
"step": 4500
},
{
"epoch": 0.4,
"learning_rate": 4.33888668517784e-05,
"loss": 2.2255,
"step": 5000
},
{
"epoch": 0.44,
"learning_rate": 4.272775353695624e-05,
"loss": 2.2713,
"step": 5500
},
{
"epoch": 0.48,
"learning_rate": 4.2066640222134076e-05,
"loss": 2.2018,
"step": 6000
},
{
"epoch": 0.52,
"learning_rate": 4.140552690731192e-05,
"loss": 2.1203,
"step": 6500
},
{
"epoch": 0.56,
"learning_rate": 4.074441359248975e-05,
"loss": 2.1148,
"step": 7000
},
{
"epoch": 0.6,
"learning_rate": 4.0083300277667595e-05,
"loss": 2.1149,
"step": 7500
},
{
"epoch": 0.63,
"learning_rate": 3.942218696284543e-05,
"loss": 2.0219,
"step": 8000
},
{
"epoch": 0.67,
"learning_rate": 3.876107364802327e-05,
"loss": 2.0354,
"step": 8500
},
{
"epoch": 0.71,
"learning_rate": 3.809996033320111e-05,
"loss": 2.0341,
"step": 9000
},
{
"epoch": 0.75,
"learning_rate": 3.7438847018378955e-05,
"loss": 1.9972,
"step": 9500
},
{
"epoch": 0.79,
"learning_rate": 3.6777733703556796e-05,
"loss": 1.9654,
"step": 10000
},
{
"epoch": 0.83,
"learning_rate": 3.611662038873463e-05,
"loss": 1.9853,
"step": 10500
},
{
"epoch": 0.87,
"learning_rate": 3.5455507073912466e-05,
"loss": 1.9487,
"step": 11000
},
{
"epoch": 0.91,
"learning_rate": 3.479439375909031e-05,
"loss": 1.9498,
"step": 11500
},
{
"epoch": 0.95,
"learning_rate": 3.413328044426815e-05,
"loss": 1.8963,
"step": 12000
},
{
"epoch": 0.99,
"learning_rate": 3.3472167129445985e-05,
"loss": 1.9259,
"step": 12500
},
{
"epoch": 1.03,
"learning_rate": 3.2811053814623827e-05,
"loss": 1.5878,
"step": 13000
},
{
"epoch": 1.07,
"learning_rate": 3.214994049980167e-05,
"loss": 1.4017,
"step": 13500
},
{
"epoch": 1.11,
"learning_rate": 3.148882718497951e-05,
"loss": 1.4809,
"step": 14000
},
{
"epoch": 1.15,
"learning_rate": 3.082771387015735e-05,
"loss": 1.4646,
"step": 14500
},
{
"epoch": 1.19,
"learning_rate": 3.0166600555335183e-05,
"loss": 1.5017,
"step": 15000
},
{
"epoch": 1.23,
"learning_rate": 2.9505487240513025e-05,
"loss": 1.4745,
"step": 15500
},
{
"epoch": 1.27,
"learning_rate": 2.8844373925690867e-05,
"loss": 1.4496,
"step": 16000
},
{
"epoch": 1.31,
"learning_rate": 2.8183260610868705e-05,
"loss": 1.4599,
"step": 16500
},
{
"epoch": 1.35,
"learning_rate": 2.752214729604654e-05,
"loss": 1.3974,
"step": 17000
},
{
"epoch": 1.39,
"learning_rate": 2.6861033981224382e-05,
"loss": 1.397,
"step": 17500
},
{
"epoch": 1.43,
"learning_rate": 2.6199920666402224e-05,
"loss": 1.436,
"step": 18000
},
{
"epoch": 1.47,
"learning_rate": 2.5538807351580062e-05,
"loss": 1.4359,
"step": 18500
},
{
"epoch": 1.51,
"learning_rate": 2.48776940367579e-05,
"loss": 1.4215,
"step": 19000
},
{
"epoch": 1.55,
"learning_rate": 2.4216580721935742e-05,
"loss": 1.3611,
"step": 19500
},
{
"epoch": 1.59,
"learning_rate": 2.355546740711358e-05,
"loss": 1.4515,
"step": 20000
},
{
"epoch": 1.63,
"learning_rate": 2.289435409229142e-05,
"loss": 1.3923,
"step": 20500
},
{
"epoch": 1.67,
"learning_rate": 2.2233240777469257e-05,
"loss": 1.3968,
"step": 21000
},
{
"epoch": 1.71,
"learning_rate": 2.15721274626471e-05,
"loss": 1.4511,
"step": 21500
},
{
"epoch": 1.75,
"learning_rate": 2.091101414782494e-05,
"loss": 1.3736,
"step": 22000
},
{
"epoch": 1.79,
"learning_rate": 2.024990083300278e-05,
"loss": 1.4196,
"step": 22500
},
{
"epoch": 1.82,
"learning_rate": 1.9588787518180617e-05,
"loss": 1.4012,
"step": 23000
},
{
"epoch": 1.86,
"learning_rate": 1.8927674203358456e-05,
"loss": 1.3995,
"step": 23500
},
{
"epoch": 1.9,
"learning_rate": 1.8266560888536297e-05,
"loss": 1.3706,
"step": 24000
},
{
"epoch": 1.94,
"learning_rate": 1.7605447573714136e-05,
"loss": 1.357,
"step": 24500
},
{
"epoch": 1.98,
"learning_rate": 1.6944334258891974e-05,
"loss": 1.3894,
"step": 25000
},
{
"epoch": 2.02,
"learning_rate": 1.6283220944069812e-05,
"loss": 1.1266,
"step": 25500
},
{
"epoch": 2.06,
"learning_rate": 1.5622107629247654e-05,
"loss": 0.9034,
"step": 26000
},
{
"epoch": 2.1,
"learning_rate": 1.4960994314425494e-05,
"loss": 0.88,
"step": 26500
},
{
"epoch": 2.14,
"learning_rate": 1.4299880999603333e-05,
"loss": 0.9379,
"step": 27000
},
{
"epoch": 2.18,
"learning_rate": 1.3638767684781173e-05,
"loss": 0.9224,
"step": 27500
},
{
"epoch": 2.22,
"learning_rate": 1.2977654369959011e-05,
"loss": 0.9108,
"step": 28000
},
{
"epoch": 2.26,
"learning_rate": 1.2316541055136851e-05,
"loss": 0.9068,
"step": 28500
},
{
"epoch": 2.3,
"learning_rate": 1.1655427740314691e-05,
"loss": 0.9158,
"step": 29000
},
{
"epoch": 2.34,
"learning_rate": 1.099431442549253e-05,
"loss": 0.901,
"step": 29500
},
{
"epoch": 2.38,
"learning_rate": 1.033320111067037e-05,
"loss": 0.8898,
"step": 30000
},
{
"epoch": 2.42,
"learning_rate": 9.67208779584821e-06,
"loss": 0.9295,
"step": 30500
},
{
"epoch": 2.46,
"learning_rate": 9.010974481026048e-06,
"loss": 0.9325,
"step": 31000
},
{
"epoch": 2.5,
"learning_rate": 8.349861166203888e-06,
"loss": 0.9357,
"step": 31500
},
{
"epoch": 2.54,
"learning_rate": 7.688747851381726e-06,
"loss": 0.8832,
"step": 32000
},
{
"epoch": 2.58,
"learning_rate": 7.027634536559567e-06,
"loss": 0.9101,
"step": 32500
},
{
"epoch": 2.62,
"learning_rate": 6.366521221737406e-06,
"loss": 0.9018,
"step": 33000
},
{
"epoch": 2.66,
"learning_rate": 5.7054079069152455e-06,
"loss": 0.8886,
"step": 33500
},
{
"epoch": 2.7,
"learning_rate": 5.044294592093085e-06,
"loss": 0.8771,
"step": 34000
},
{
"epoch": 2.74,
"learning_rate": 4.383181277270925e-06,
"loss": 0.8956,
"step": 34500
},
{
"epoch": 2.78,
"learning_rate": 3.7220679624487635e-06,
"loss": 0.8586,
"step": 35000
},
{
"epoch": 2.82,
"learning_rate": 3.060954647626603e-06,
"loss": 0.9039,
"step": 35500
},
{
"epoch": 2.86,
"learning_rate": 2.3998413328044427e-06,
"loss": 0.8817,
"step": 36000
},
{
"epoch": 2.9,
"learning_rate": 1.7387280179822822e-06,
"loss": 0.8601,
"step": 36500
},
{
"epoch": 2.94,
"learning_rate": 1.0776147031601218e-06,
"loss": 0.8837,
"step": 37000
},
{
"epoch": 2.98,
"learning_rate": 4.165013883379611e-07,
"loss": 0.8894,
"step": 37500
},
{
"epoch": 3.0,
"step": 37815,
"total_flos": 9288563680542720.0,
"train_loss": 1.5174339204652274,
"train_runtime": 9956.4506,
"train_samples_per_second": 15.192,
"train_steps_per_second": 3.798
}
],
"logging_steps": 500,
"max_steps": 37815,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 9288563680542720.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}