17713 lines
399 KiB
JSON
17713 lines
399 KiB
JSON
{
|
|
"best_metric": 1.2748980522155762,
|
|
"best_model_checkpoint": "outputs-ziya-llama-13b-sft-med-v2/checkpoint-63200",
|
|
"epoch": 4.004562159422126,
|
|
"global_step": 63200,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0,
|
|
"learning_rate": 2e-05,
|
|
"loss": 1.9402,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.0,
|
|
"eval_loss": 1.3943634033203125,
|
|
"eval_runtime": 3.7851,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.9993661659377576e-05,
|
|
"loss": 1.7943,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"eval_loss": 1.3783146142959595,
|
|
"eval_runtime": 3.7902,
|
|
"eval_samples_per_second": 13.192,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.998732331875515e-05,
|
|
"loss": 1.7393,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"eval_loss": 1.3769421577453613,
|
|
"eval_runtime": 3.7851,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.9980984978132726e-05,
|
|
"loss": 1.7253,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"eval_loss": 1.3832125663757324,
|
|
"eval_runtime": 3.7862,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.99746466375103e-05,
|
|
"loss": 1.6986,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"eval_loss": 1.3861111402511597,
|
|
"eval_runtime": 3.7837,
|
|
"eval_samples_per_second": 13.215,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.9968308296887876e-05,
|
|
"loss": 1.6863,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"eval_loss": 1.3855526447296143,
|
|
"eval_runtime": 3.7891,
|
|
"eval_samples_per_second": 13.196,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.9961969956265454e-05,
|
|
"loss": 1.651,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"eval_loss": 1.3862779140472412,
|
|
"eval_runtime": 3.7834,
|
|
"eval_samples_per_second": 13.216,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 1.9955631615643025e-05,
|
|
"loss": 1.6424,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"eval_loss": 1.3855341672897339,
|
|
"eval_runtime": 3.7852,
|
|
"eval_samples_per_second": 13.209,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 1.99492932750206e-05,
|
|
"loss": 1.6205,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"eval_loss": 1.3793355226516724,
|
|
"eval_runtime": 3.7851,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 1.9942954934398175e-05,
|
|
"loss": 1.6049,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"eval_loss": 1.3790518045425415,
|
|
"eval_runtime": 3.7866,
|
|
"eval_samples_per_second": 13.205,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 1.993661659377575e-05,
|
|
"loss": 1.5928,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"eval_loss": 1.3790216445922852,
|
|
"eval_runtime": 3.7841,
|
|
"eval_samples_per_second": 13.213,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 1.9930278253153324e-05,
|
|
"loss": 1.6008,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"eval_loss": 1.3779619932174683,
|
|
"eval_runtime": 3.7838,
|
|
"eval_samples_per_second": 13.214,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 1.9923939912530902e-05,
|
|
"loss": 1.5788,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"eval_loss": 1.3782728910446167,
|
|
"eval_runtime": 3.7849,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 1.9917601571908477e-05,
|
|
"loss": 1.5742,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"eval_loss": 1.3791857957839966,
|
|
"eval_runtime": 3.786,
|
|
"eval_samples_per_second": 13.207,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 1.9911263231286052e-05,
|
|
"loss": 1.5578,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"eval_loss": 1.3797205686569214,
|
|
"eval_runtime": 3.7868,
|
|
"eval_samples_per_second": 13.204,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 1.9904924890663627e-05,
|
|
"loss": 1.5477,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"eval_loss": 1.3817371129989624,
|
|
"eval_runtime": 3.7855,
|
|
"eval_samples_per_second": 13.208,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 1.98985865500412e-05,
|
|
"loss": 1.5509,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"eval_loss": 1.3795477151870728,
|
|
"eval_runtime": 3.7863,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.9892248209418776e-05,
|
|
"loss": 1.5638,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"eval_loss": 1.378434181213379,
|
|
"eval_runtime": 3.7822,
|
|
"eval_samples_per_second": 13.22,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.988590986879635e-05,
|
|
"loss": 1.5505,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"eval_loss": 1.3711613416671753,
|
|
"eval_runtime": 3.7882,
|
|
"eval_samples_per_second": 13.199,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.9879571528173926e-05,
|
|
"loss": 1.5678,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"eval_loss": 1.374761939048767,
|
|
"eval_runtime": 3.7863,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.98732331875515e-05,
|
|
"loss": 1.5438,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"eval_loss": 1.3767447471618652,
|
|
"eval_runtime": 3.7836,
|
|
"eval_samples_per_second": 13.215,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.9866894846929075e-05,
|
|
"loss": 1.558,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"eval_loss": 1.3725579977035522,
|
|
"eval_runtime": 3.7848,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.986055650630665e-05,
|
|
"loss": 1.5244,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"eval_loss": 1.3703813552856445,
|
|
"eval_runtime": 3.79,
|
|
"eval_samples_per_second": 13.193,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"learning_rate": 1.9854218165684225e-05,
|
|
"loss": 1.5558,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"eval_loss": 1.3710592985153198,
|
|
"eval_runtime": 3.785,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"learning_rate": 1.98478798250618e-05,
|
|
"loss": 1.5556,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"eval_loss": 1.3695827722549438,
|
|
"eval_runtime": 3.7841,
|
|
"eval_samples_per_second": 13.213,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"learning_rate": 1.9841541484439374e-05,
|
|
"loss": 1.5593,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"eval_loss": 1.3656278848648071,
|
|
"eval_runtime": 3.7829,
|
|
"eval_samples_per_second": 13.217,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.983520314381695e-05,
|
|
"loss": 1.5463,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"eval_loss": 1.3648173809051514,
|
|
"eval_runtime": 3.7879,
|
|
"eval_samples_per_second": 13.2,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.9828864803194524e-05,
|
|
"loss": 1.5396,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"eval_loss": 1.3657265901565552,
|
|
"eval_runtime": 3.785,
|
|
"eval_samples_per_second": 13.21,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.98225264625721e-05,
|
|
"loss": 1.5331,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"eval_loss": 1.365883231163025,
|
|
"eval_runtime": 3.7838,
|
|
"eval_samples_per_second": 13.214,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9816188121949677e-05,
|
|
"loss": 1.5138,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"eval_loss": 1.36909818649292,
|
|
"eval_runtime": 3.7846,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9809849781327252e-05,
|
|
"loss": 1.5243,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"eval_loss": 1.367846965789795,
|
|
"eval_runtime": 3.7854,
|
|
"eval_samples_per_second": 13.209,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9803511440704826e-05,
|
|
"loss": 1.5509,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"eval_loss": 1.3653690814971924,
|
|
"eval_runtime": 3.7873,
|
|
"eval_samples_per_second": 13.202,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.97971731000824e-05,
|
|
"loss": 1.5269,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"eval_loss": 1.3630925416946411,
|
|
"eval_runtime": 3.7825,
|
|
"eval_samples_per_second": 13.219,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.9790834759459976e-05,
|
|
"loss": 1.5274,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"eval_loss": 1.3645691871643066,
|
|
"eval_runtime": 3.7862,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.978449641883755e-05,
|
|
"loss": 1.4974,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"eval_loss": 1.3679264783859253,
|
|
"eval_runtime": 3.7861,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.9778158078215126e-05,
|
|
"loss": 1.5322,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"eval_loss": 1.366579532623291,
|
|
"eval_runtime": 3.7847,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.97718197375927e-05,
|
|
"loss": 1.5226,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"eval_loss": 1.3667091131210327,
|
|
"eval_runtime": 3.7811,
|
|
"eval_samples_per_second": 13.224,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.9765481396970275e-05,
|
|
"loss": 1.5297,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"eval_loss": 1.3697519302368164,
|
|
"eval_runtime": 3.7846,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.975914305634785e-05,
|
|
"loss": 1.5134,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"eval_loss": 1.3626987934112549,
|
|
"eval_runtime": 3.7868,
|
|
"eval_samples_per_second": 13.204,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9752804715725425e-05,
|
|
"loss": 1.5272,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"eval_loss": 1.3688554763793945,
|
|
"eval_runtime": 3.7872,
|
|
"eval_samples_per_second": 13.202,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9746466375103e-05,
|
|
"loss": 1.508,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"eval_loss": 1.3672854900360107,
|
|
"eval_runtime": 3.7828,
|
|
"eval_samples_per_second": 13.218,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9740128034480574e-05,
|
|
"loss": 1.5063,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"eval_loss": 1.3676797151565552,
|
|
"eval_runtime": 3.7834,
|
|
"eval_samples_per_second": 13.216,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.973378969385815e-05,
|
|
"loss": 1.4978,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"eval_loss": 1.3642038106918335,
|
|
"eval_runtime": 3.7847,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.9727451353235724e-05,
|
|
"loss": 1.5138,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"eval_loss": 1.3679579496383667,
|
|
"eval_runtime": 3.7883,
|
|
"eval_samples_per_second": 13.198,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.97211130126133e-05,
|
|
"loss": 1.521,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"eval_loss": 1.3639538288116455,
|
|
"eval_runtime": 3.7838,
|
|
"eval_samples_per_second": 13.214,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.9714774671990873e-05,
|
|
"loss": 1.4849,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"eval_loss": 1.3618545532226562,
|
|
"eval_runtime": 3.7862,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.970843633136845e-05,
|
|
"loss": 1.5282,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"eval_loss": 1.366089940071106,
|
|
"eval_runtime": 3.7896,
|
|
"eval_samples_per_second": 13.194,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.9702097990746026e-05,
|
|
"loss": 1.5095,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"eval_loss": 1.3617390394210815,
|
|
"eval_runtime": 3.7846,
|
|
"eval_samples_per_second": 13.211,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.96957596501236e-05,
|
|
"loss": 1.4784,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"eval_loss": 1.3627303838729858,
|
|
"eval_runtime": 3.7842,
|
|
"eval_samples_per_second": 13.213,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.9689421309501172e-05,
|
|
"loss": 1.48,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"eval_loss": 1.363022804260254,
|
|
"eval_runtime": 3.7839,
|
|
"eval_samples_per_second": 13.214,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.9683082968878747e-05,
|
|
"loss": 1.5182,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"eval_loss": 1.354172706604004,
|
|
"eval_runtime": 3.7881,
|
|
"eval_samples_per_second": 13.199,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.9676744628256322e-05,
|
|
"loss": 1.4869,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"eval_loss": 1.3622251749038696,
|
|
"eval_runtime": 3.7863,
|
|
"eval_samples_per_second": 13.206,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.96704062876339e-05,
|
|
"loss": 1.5135,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"eval_loss": 1.3611241579055786,
|
|
"eval_runtime": 3.781,
|
|
"eval_samples_per_second": 13.224,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.9664067947011475e-05,
|
|
"loss": 1.4981,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"eval_loss": 1.3572229146957397,
|
|
"eval_runtime": 3.7799,
|
|
"eval_samples_per_second": 13.228,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.965772960638905e-05,
|
|
"loss": 1.5045,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"eval_loss": 1.364942193031311,
|
|
"eval_runtime": 3.7761,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.9651391265766625e-05,
|
|
"loss": 1.4927,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"eval_loss": 1.3538453578948975,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.96450529251442e-05,
|
|
"loss": 1.4685,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"eval_loss": 1.3631401062011719,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.9638714584521774e-05,
|
|
"loss": 1.4658,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"eval_loss": 1.3613299131393433,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"learning_rate": 1.9632503010711796e-05,
|
|
"loss": 1.4987,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"eval_loss": 1.3675013780593872,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"learning_rate": 1.962616467008937e-05,
|
|
"loss": 1.459,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"eval_loss": 1.3627259731292725,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"learning_rate": 1.9619826329466945e-05,
|
|
"loss": 1.4893,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"eval_loss": 1.3637642860412598,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.961348798884452e-05,
|
|
"loss": 1.4872,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"eval_loss": 1.3603347539901733,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.96071496482221e-05,
|
|
"loss": 1.4905,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"eval_loss": 1.3588290214538574,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.9600811307599673e-05,
|
|
"loss": 1.4531,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"eval_loss": 1.3601138591766357,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9594472966977248e-05,
|
|
"loss": 1.5118,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"eval_loss": 1.358245849609375,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9588134626354823e-05,
|
|
"loss": 1.5047,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"eval_loss": 1.3589282035827637,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9581796285732397e-05,
|
|
"loss": 1.5123,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"eval_loss": 1.3560609817504883,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.9575457945109972e-05,
|
|
"loss": 1.51,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"eval_loss": 1.354966640472412,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.9569119604487547e-05,
|
|
"loss": 1.4611,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"eval_loss": 1.3565185070037842,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.9562781263865122e-05,
|
|
"loss": 1.4787,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"eval_loss": 1.3581827878952026,
|
|
"eval_runtime": 3.7752,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.9556442923242697e-05,
|
|
"loss": 1.4879,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"eval_loss": 1.3542126417160034,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.955010458262027e-05,
|
|
"loss": 1.4671,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"eval_loss": 1.353346347808838,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.9543766241997846e-05,
|
|
"loss": 1.4757,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"eval_loss": 1.353261113166809,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.953742790137542e-05,
|
|
"loss": 1.4637,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"eval_loss": 1.3608603477478027,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.9531089560752996e-05,
|
|
"loss": 1.479,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"eval_loss": 1.3540550470352173,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.952475122013057e-05,
|
|
"loss": 1.4767,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"eval_loss": 1.3591903448104858,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.9518412879508145e-05,
|
|
"loss": 1.4735,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"eval_loss": 1.3510184288024902,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.951207453888572e-05,
|
|
"loss": 1.4611,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"eval_loss": 1.352201223373413,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.9505736198263295e-05,
|
|
"loss": 1.4716,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"eval_loss": 1.3535534143447876,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.9499397857640873e-05,
|
|
"loss": 1.4831,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"eval_loss": 1.3486146926879883,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.9493059517018448e-05,
|
|
"loss": 1.4663,
|
|
"step": 4050
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"eval_loss": 1.3517810106277466,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 4050
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.9486721176396022e-05,
|
|
"loss": 1.4623,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"eval_loss": 1.3499053716659546,
|
|
"eval_runtime": 3.7758,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.9480382835773597e-05,
|
|
"loss": 1.4923,
|
|
"step": 4150
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"eval_loss": 1.3492207527160645,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4150
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.9474044495151172e-05,
|
|
"loss": 1.477,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"eval_loss": 1.3482263088226318,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.9467706154528743e-05,
|
|
"loss": 1.4831,
|
|
"step": 4250
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"eval_loss": 1.3458422422409058,
|
|
"eval_runtime": 3.7661,
|
|
"eval_samples_per_second": 13.276,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4250
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.946136781390632e-05,
|
|
"loss": 1.4619,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"eval_loss": 1.3500503301620483,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.9455029473283896e-05,
|
|
"loss": 1.4616,
|
|
"step": 4350
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"eval_loss": 1.349271535873413,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4350
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.944869113266147e-05,
|
|
"loss": 1.4647,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"eval_loss": 1.3516141176223755,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.9442352792039046e-05,
|
|
"loss": 1.471,
|
|
"step": 4450
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"eval_loss": 1.3483657836914062,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4450
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.943601445141662e-05,
|
|
"loss": 1.455,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"eval_loss": 1.3461648225784302,
|
|
"eval_runtime": 3.7664,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.9429676110794195e-05,
|
|
"loss": 1.4535,
|
|
"step": 4550
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"eval_loss": 1.3466649055480957,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4550
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.942333777017177e-05,
|
|
"loss": 1.4689,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"eval_loss": 1.3467597961425781,
|
|
"eval_runtime": 3.7675,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.9416999429549345e-05,
|
|
"loss": 1.4769,
|
|
"step": 4650
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"eval_loss": 1.356662631034851,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 4650
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.941066108892692e-05,
|
|
"loss": 1.4739,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"eval_loss": 1.345522165298462,
|
|
"eval_runtime": 3.7676,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.9404322748304495e-05,
|
|
"loss": 1.4629,
|
|
"step": 4750
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"eval_loss": 1.3463581800460815,
|
|
"eval_runtime": 3.7678,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4750
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.9397984407682073e-05,
|
|
"loss": 1.4757,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"eval_loss": 1.3450006246566772,
|
|
"eval_runtime": 3.767,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9391646067059648e-05,
|
|
"loss": 1.4549,
|
|
"step": 4850
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"eval_loss": 1.3511794805526733,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4850
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9385307726437222e-05,
|
|
"loss": 1.4821,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"eval_loss": 1.348404884338379,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9378969385814794e-05,
|
|
"loss": 1.4423,
|
|
"step": 4950
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"eval_loss": 1.350976824760437,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 4950
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.937263104519237e-05,
|
|
"loss": 1.4751,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"eval_loss": 1.349683403968811,
|
|
"eval_runtime": 3.7679,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.9366419471382394e-05,
|
|
"loss": 1.4577,
|
|
"step": 5050
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"eval_loss": 1.3500033617019653,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5050
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.936008113075997e-05,
|
|
"loss": 1.4526,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"eval_loss": 1.3477158546447754,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.9353742790137543e-05,
|
|
"loss": 1.4558,
|
|
"step": 5150
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"eval_loss": 1.3458642959594727,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5150
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.9347404449515118e-05,
|
|
"loss": 1.4791,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"eval_loss": 1.3409136533737183,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.9341066108892693e-05,
|
|
"loss": 1.4328,
|
|
"step": 5250
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"eval_loss": 1.3430328369140625,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5250
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.933472776827027e-05,
|
|
"loss": 1.4402,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"eval_loss": 1.3471790552139282,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.9328389427647842e-05,
|
|
"loss": 1.4252,
|
|
"step": 5350
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"eval_loss": 1.3500759601593018,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5350
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.9322051087025417e-05,
|
|
"loss": 1.48,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"eval_loss": 1.3430708646774292,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9315712746402992e-05,
|
|
"loss": 1.4446,
|
|
"step": 5450
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"eval_loss": 1.345218539237976,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5450
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9309374405780567e-05,
|
|
"loss": 1.4408,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"eval_loss": 1.3471763134002686,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.930303606515814e-05,
|
|
"loss": 1.4622,
|
|
"step": 5550
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"eval_loss": 1.343950629234314,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5550
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9296697724535716e-05,
|
|
"loss": 1.4645,
|
|
"step": 5600
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"eval_loss": 1.341795802116394,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5600
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.9290359383913294e-05,
|
|
"loss": 1.4421,
|
|
"step": 5650
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"eval_loss": 1.3437079191207886,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5650
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.928402104329087e-05,
|
|
"loss": 1.4433,
|
|
"step": 5700
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"eval_loss": 1.3465303182601929,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5700
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.9277682702668444e-05,
|
|
"loss": 1.4672,
|
|
"step": 5750
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"eval_loss": 1.3458932638168335,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5750
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.927134436204602e-05,
|
|
"loss": 1.453,
|
|
"step": 5800
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"eval_loss": 1.3377151489257812,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5800
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.9265006021423593e-05,
|
|
"loss": 1.4525,
|
|
"step": 5850
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"eval_loss": 1.341667652130127,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5850
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.9258667680801168e-05,
|
|
"loss": 1.4261,
|
|
"step": 5900
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"eval_loss": 1.340978741645813,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 5900
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.9252329340178743e-05,
|
|
"loss": 1.4485,
|
|
"step": 5950
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"eval_loss": 1.344750165939331,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 5950
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.9245990999556318e-05,
|
|
"loss": 1.4481,
|
|
"step": 6000
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"eval_loss": 1.341100811958313,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6000
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.9239652658933893e-05,
|
|
"loss": 1.4555,
|
|
"step": 6050
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"eval_loss": 1.339819073677063,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6050
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.9233314318311467e-05,
|
|
"loss": 1.4386,
|
|
"step": 6100
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"eval_loss": 1.3417630195617676,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6100
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.9226975977689042e-05,
|
|
"loss": 1.463,
|
|
"step": 6150
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"eval_loss": 1.3422001600265503,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6150
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.9220637637066617e-05,
|
|
"loss": 1.4492,
|
|
"step": 6200
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"eval_loss": 1.3382301330566406,
|
|
"eval_runtime": 3.7666,
|
|
"eval_samples_per_second": 13.274,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6200
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.921429929644419e-05,
|
|
"loss": 1.4372,
|
|
"step": 6250
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"eval_loss": 1.343968152999878,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6250
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.9207960955821766e-05,
|
|
"loss": 1.4264,
|
|
"step": 6300
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"eval_loss": 1.3375903367996216,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6300
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.920162261519934e-05,
|
|
"loss": 1.4586,
|
|
"step": 6350
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"eval_loss": 1.3373645544052124,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6350
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.9195284274576916e-05,
|
|
"loss": 1.4663,
|
|
"step": 6400
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"eval_loss": 1.3406654596328735,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6400
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.9188945933954494e-05,
|
|
"loss": 1.4568,
|
|
"step": 6450
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"eval_loss": 1.3379381895065308,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6450
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.918260759333207e-05,
|
|
"loss": 1.453,
|
|
"step": 6500
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"eval_loss": 1.341365933418274,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6500
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.9176269252709644e-05,
|
|
"loss": 1.4434,
|
|
"step": 6550
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"eval_loss": 1.3401316404342651,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6550
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.916993091208722e-05,
|
|
"loss": 1.448,
|
|
"step": 6600
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"eval_loss": 1.3429800271987915,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6600
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.9163592571464793e-05,
|
|
"loss": 1.4267,
|
|
"step": 6650
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"eval_loss": 1.3384313583374023,
|
|
"eval_runtime": 3.7654,
|
|
"eval_samples_per_second": 13.279,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6650
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.9157254230842368e-05,
|
|
"loss": 1.4556,
|
|
"step": 6700
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"eval_loss": 1.3433630466461182,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6700
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.915091589021994e-05,
|
|
"loss": 1.4277,
|
|
"step": 6750
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"eval_loss": 1.343968391418457,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6750
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.9144577549597518e-05,
|
|
"loss": 1.438,
|
|
"step": 6800
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"eval_loss": 1.3406360149383545,
|
|
"eval_runtime": 3.7678,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6800
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.9138239208975092e-05,
|
|
"loss": 1.4463,
|
|
"step": 6850
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"eval_loss": 1.3366354703903198,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6850
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.9131900868352667e-05,
|
|
"loss": 1.43,
|
|
"step": 6900
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"eval_loss": 1.3320492506027222,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 6900
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.9125562527730242e-05,
|
|
"loss": 1.4219,
|
|
"step": 6950
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"eval_loss": 1.3377419710159302,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 6950
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.9119224187107817e-05,
|
|
"loss": 1.4439,
|
|
"step": 7000
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"eval_loss": 1.3389599323272705,
|
|
"eval_runtime": 3.7666,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7000
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.911288584648539e-05,
|
|
"loss": 1.4444,
|
|
"step": 7050
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"eval_loss": 1.335353136062622,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7050
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.9106547505862966e-05,
|
|
"loss": 1.4611,
|
|
"step": 7100
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"eval_loss": 1.3384804725646973,
|
|
"eval_runtime": 3.7672,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7100
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.910020916524054e-05,
|
|
"loss": 1.4291,
|
|
"step": 7150
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"eval_loss": 1.337825059890747,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 7150
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.9093870824618116e-05,
|
|
"loss": 1.4106,
|
|
"step": 7200
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"eval_loss": 1.3410215377807617,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 7200
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.908753248399569e-05,
|
|
"loss": 1.4648,
|
|
"step": 7250
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"eval_loss": 1.3364254236221313,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7250
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.908119414337327e-05,
|
|
"loss": 1.4683,
|
|
"step": 7300
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"eval_loss": 1.332759141921997,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7300
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.9074855802750844e-05,
|
|
"loss": 1.4111,
|
|
"step": 7350
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"eval_loss": 1.341154932975769,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7350
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.9068517462128415e-05,
|
|
"loss": 1.4445,
|
|
"step": 7400
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"eval_loss": 1.3391424417495728,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7400
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.906217912150599e-05,
|
|
"loss": 1.4616,
|
|
"step": 7450
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"eval_loss": 1.3362289667129517,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7450
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.9055840780883564e-05,
|
|
"loss": 1.4352,
|
|
"step": 7500
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"eval_loss": 1.335687279701233,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7500
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.904950244026114e-05,
|
|
"loss": 1.4349,
|
|
"step": 7550
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"eval_loss": 1.34110689163208,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 7550
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.9043164099638714e-05,
|
|
"loss": 1.4418,
|
|
"step": 7600
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"eval_loss": 1.333461880683899,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 7600
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.9036825759016292e-05,
|
|
"loss": 1.4197,
|
|
"step": 7650
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"eval_loss": 1.3351945877075195,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 7650
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9030487418393867e-05,
|
|
"loss": 1.4238,
|
|
"step": 7700
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"eval_loss": 1.3362082242965698,
|
|
"eval_runtime": 3.7804,
|
|
"eval_samples_per_second": 13.226,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 7700
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9024149077771442e-05,
|
|
"loss": 1.4396,
|
|
"step": 7750
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"eval_loss": 1.330649733543396,
|
|
"eval_runtime": 3.7878,
|
|
"eval_samples_per_second": 13.2,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 7750
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9017810737149017e-05,
|
|
"loss": 1.4424,
|
|
"step": 7800
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"eval_loss": 1.3354008197784424,
|
|
"eval_runtime": 3.7561,
|
|
"eval_samples_per_second": 13.312,
|
|
"eval_steps_per_second": 0.532,
|
|
"step": 7800
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.901147239652659e-05,
|
|
"loss": 1.4222,
|
|
"step": 7850
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"eval_loss": 1.3342909812927246,
|
|
"eval_runtime": 3.7885,
|
|
"eval_samples_per_second": 13.198,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 7850
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.9005134055904166e-05,
|
|
"loss": 1.4217,
|
|
"step": 7900
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"eval_loss": 1.331030249595642,
|
|
"eval_runtime": 3.7868,
|
|
"eval_samples_per_second": 13.204,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 7900
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.899879571528174e-05,
|
|
"loss": 1.4269,
|
|
"step": 7950
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"eval_loss": 1.3378578424453735,
|
|
"eval_runtime": 3.8143,
|
|
"eval_samples_per_second": 13.109,
|
|
"eval_steps_per_second": 0.524,
|
|
"step": 7950
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"learning_rate": 1.8992457374659316e-05,
|
|
"loss": 1.4221,
|
|
"step": 8000
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"eval_loss": 1.3354885578155518,
|
|
"eval_runtime": 3.7834,
|
|
"eval_samples_per_second": 13.216,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 8000
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"learning_rate": 1.898611903403689e-05,
|
|
"loss": 1.4156,
|
|
"step": 8050
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"eval_loss": 1.3381201028823853,
|
|
"eval_runtime": 3.7869,
|
|
"eval_samples_per_second": 13.203,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 8050
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"learning_rate": 1.8979780693414465e-05,
|
|
"loss": 1.4163,
|
|
"step": 8100
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"eval_loss": 1.3321332931518555,
|
|
"eval_runtime": 3.7853,
|
|
"eval_samples_per_second": 13.209,
|
|
"eval_steps_per_second": 0.528,
|
|
"step": 8100
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.897344235279204e-05,
|
|
"loss": 1.458,
|
|
"step": 8150
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"eval_loss": 1.3282891511917114,
|
|
"eval_runtime": 3.7924,
|
|
"eval_samples_per_second": 13.184,
|
|
"eval_steps_per_second": 0.527,
|
|
"step": 8150
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.8967104012169615e-05,
|
|
"loss": 1.4372,
|
|
"step": 8200
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"eval_loss": 1.330866813659668,
|
|
"eval_runtime": 3.7763,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8200
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.896076567154719e-05,
|
|
"loss": 1.4611,
|
|
"step": 8250
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"eval_loss": 1.3348468542099,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8250
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.8954427330924764e-05,
|
|
"loss": 1.4386,
|
|
"step": 8300
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"eval_loss": 1.3345593214035034,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 8300
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.894808899030234e-05,
|
|
"loss": 1.4426,
|
|
"step": 8350
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"eval_loss": 1.3326621055603027,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8350
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.8941750649679914e-05,
|
|
"loss": 1.4068,
|
|
"step": 8400
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"eval_loss": 1.3300801515579224,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8400
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.8935412309057492e-05,
|
|
"loss": 1.4481,
|
|
"step": 8450
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"eval_loss": 1.3321070671081543,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8450
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.8929073968435067e-05,
|
|
"loss": 1.4407,
|
|
"step": 8500
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"eval_loss": 1.3314056396484375,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8500
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.892273562781264e-05,
|
|
"loss": 1.4324,
|
|
"step": 8550
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"eval_loss": 1.3374255895614624,
|
|
"eval_runtime": 3.7764,
|
|
"eval_samples_per_second": 13.24,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8550
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.8916397287190216e-05,
|
|
"loss": 1.4156,
|
|
"step": 8600
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"eval_loss": 1.3291068077087402,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 8600
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.891005894656779e-05,
|
|
"loss": 1.4407,
|
|
"step": 8650
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"eval_loss": 1.3375924825668335,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8650
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.8903720605945366e-05,
|
|
"loss": 1.4491,
|
|
"step": 8700
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"eval_loss": 1.3330168724060059,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8700
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.889738226532294e-05,
|
|
"loss": 1.4289,
|
|
"step": 8750
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"eval_loss": 1.3330626487731934,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8750
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.8891043924700515e-05,
|
|
"loss": 1.43,
|
|
"step": 8800
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"eval_loss": 1.3307543992996216,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 8800
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.888470558407809e-05,
|
|
"loss": 1.4502,
|
|
"step": 8850
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"eval_loss": 1.3406033515930176,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8850
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.8878367243455665e-05,
|
|
"loss": 1.4483,
|
|
"step": 8900
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"eval_loss": 1.332836627960205,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 8900
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.887202890283324e-05,
|
|
"loss": 1.4222,
|
|
"step": 8950
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"eval_loss": 1.3319545984268188,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 8950
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.8865690562210815e-05,
|
|
"loss": 1.4195,
|
|
"step": 9000
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"eval_loss": 1.3331190347671509,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9000
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.885935222158839e-05,
|
|
"loss": 1.4353,
|
|
"step": 9050
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"eval_loss": 1.3306701183319092,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9050
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.8853013880965964e-05,
|
|
"loss": 1.4219,
|
|
"step": 9100
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"eval_loss": 1.3325066566467285,
|
|
"eval_runtime": 3.7655,
|
|
"eval_samples_per_second": 13.278,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9100
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.884667554034354e-05,
|
|
"loss": 1.4234,
|
|
"step": 9150
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"eval_loss": 1.327727198600769,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9150
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.8840337199721114e-05,
|
|
"loss": 1.4307,
|
|
"step": 9200
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"eval_loss": 1.33103609085083,
|
|
"eval_runtime": 3.7431,
|
|
"eval_samples_per_second": 13.358,
|
|
"eval_steps_per_second": 0.534,
|
|
"step": 9200
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.883399885909869e-05,
|
|
"loss": 1.411,
|
|
"step": 9250
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"eval_loss": 1.3286285400390625,
|
|
"eval_runtime": 3.7753,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9250
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.8827660518476267e-05,
|
|
"loss": 1.4318,
|
|
"step": 9300
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"eval_loss": 1.3292878866195679,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9300
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.882132217785384e-05,
|
|
"loss": 1.4109,
|
|
"step": 9350
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"eval_loss": 1.3338630199432373,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9350
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.8815110604043863e-05,
|
|
"loss": 1.4152,
|
|
"step": 9400
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"eval_loss": 1.3317309617996216,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9400
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.8808772263421438e-05,
|
|
"loss": 1.4265,
|
|
"step": 9450
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"eval_loss": 1.3338027000427246,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9450
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.8802433922799013e-05,
|
|
"loss": 1.4278,
|
|
"step": 9500
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"eval_loss": 1.3360050916671753,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9500
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8796095582176588e-05,
|
|
"loss": 1.4,
|
|
"step": 9550
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"eval_loss": 1.3343561887741089,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9550
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8789757241554162e-05,
|
|
"loss": 1.4118,
|
|
"step": 9600
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"eval_loss": 1.3415179252624512,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9600
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8783418900931737e-05,
|
|
"loss": 1.4549,
|
|
"step": 9650
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"eval_loss": 1.3305364847183228,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9650
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8777080560309312e-05,
|
|
"loss": 1.4233,
|
|
"step": 9700
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"eval_loss": 1.332060694694519,
|
|
"eval_runtime": 3.7675,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9700
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.8770742219686887e-05,
|
|
"loss": 1.4352,
|
|
"step": 9750
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"eval_loss": 1.3307561874389648,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9750
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.8764403879064465e-05,
|
|
"loss": 1.4287,
|
|
"step": 9800
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"eval_loss": 1.3289636373519897,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9800
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.875806553844204e-05,
|
|
"loss": 1.4087,
|
|
"step": 9850
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"eval_loss": 1.3348122835159302,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9850
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.875172719781961e-05,
|
|
"loss": 1.4349,
|
|
"step": 9900
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"eval_loss": 1.3324333429336548,
|
|
"eval_runtime": 3.7647,
|
|
"eval_samples_per_second": 13.281,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 9900
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.8745515624009636e-05,
|
|
"loss": 1.4369,
|
|
"step": 9950
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"eval_loss": 1.3347926139831543,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 9950
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.873917728338721e-05,
|
|
"loss": 1.4248,
|
|
"step": 10000
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"eval_loss": 1.3310209512710571,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10000
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8732965709577233e-05,
|
|
"loss": 1.4384,
|
|
"step": 10050
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"eval_loss": 1.3247287273406982,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10050
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8726627368954807e-05,
|
|
"loss": 1.4419,
|
|
"step": 10100
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"eval_loss": 1.3305790424346924,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10100
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8720289028332382e-05,
|
|
"loss": 1.436,
|
|
"step": 10150
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"eval_loss": 1.3302339315414429,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10150
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.8713950687709957e-05,
|
|
"loss": 1.4144,
|
|
"step": 10200
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"eval_loss": 1.329056978225708,
|
|
"eval_runtime": 3.7763,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10200
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.8707739113899982e-05,
|
|
"loss": 1.433,
|
|
"step": 10250
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"eval_loss": 1.3310526609420776,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10250
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.8701400773277557e-05,
|
|
"loss": 1.425,
|
|
"step": 10300
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"eval_loss": 1.3268311023712158,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10300
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"learning_rate": 1.8695062432655132e-05,
|
|
"loss": 1.441,
|
|
"step": 10350
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"eval_loss": 1.328212857246399,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10350
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"learning_rate": 1.868872409203271e-05,
|
|
"loss": 1.4092,
|
|
"step": 10400
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"eval_loss": 1.3228343725204468,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10400
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"learning_rate": 1.868238575141028e-05,
|
|
"loss": 1.4094,
|
|
"step": 10450
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"eval_loss": 1.3348655700683594,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10450
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.8676047410787856e-05,
|
|
"loss": 1.454,
|
|
"step": 10500
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"eval_loss": 1.3274316787719727,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10500
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.866970907016543e-05,
|
|
"loss": 1.4306,
|
|
"step": 10550
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"eval_loss": 1.3241510391235352,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10550
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.8663370729543006e-05,
|
|
"loss": 1.4208,
|
|
"step": 10600
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"eval_loss": 1.3249012231826782,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10600
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.865703238892058e-05,
|
|
"loss": 1.4105,
|
|
"step": 10650
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"eval_loss": 1.323702096939087,
|
|
"eval_runtime": 3.7673,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10650
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.865069404829816e-05,
|
|
"loss": 1.4334,
|
|
"step": 10700
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"eval_loss": 1.3316683769226074,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10700
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.8644355707675733e-05,
|
|
"loss": 1.3968,
|
|
"step": 10750
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"eval_loss": 1.3293657302856445,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10750
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.8638017367053308e-05,
|
|
"loss": 1.4005,
|
|
"step": 10800
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"eval_loss": 1.329288125038147,
|
|
"eval_runtime": 3.7668,
|
|
"eval_samples_per_second": 13.274,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10800
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.8631679026430883e-05,
|
|
"loss": 1.4094,
|
|
"step": 10850
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"eval_loss": 1.3340065479278564,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10850
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.8625340685808458e-05,
|
|
"loss": 1.3981,
|
|
"step": 10900
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"eval_loss": 1.3305424451828003,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 10900
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.8619002345186032e-05,
|
|
"loss": 1.4298,
|
|
"step": 10950
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"eval_loss": 1.3277804851531982,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 10950
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.8612664004563607e-05,
|
|
"loss": 1.4342,
|
|
"step": 11000
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"eval_loss": 1.3289930820465088,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 11000
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.8606325663941182e-05,
|
|
"loss": 1.4205,
|
|
"step": 11050
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"eval_loss": 1.326532244682312,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11050
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.8599987323318757e-05,
|
|
"loss": 1.4171,
|
|
"step": 11100
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"eval_loss": 1.3211607933044434,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 11100
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.859364898269633e-05,
|
|
"loss": 1.4165,
|
|
"step": 11150
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"eval_loss": 1.3257567882537842,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 11150
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.8587310642073906e-05,
|
|
"loss": 1.4337,
|
|
"step": 11200
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"eval_loss": 1.3250231742858887,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 11200
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.858097230145148e-05,
|
|
"loss": 1.4487,
|
|
"step": 11250
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"eval_loss": 1.3210654258728027,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11250
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"learning_rate": 1.8574633960829056e-05,
|
|
"loss": 1.4071,
|
|
"step": 11300
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"eval_loss": 1.3362988233566284,
|
|
"eval_runtime": 3.767,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 11300
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"learning_rate": 1.856829562020663e-05,
|
|
"loss": 1.3933,
|
|
"step": 11350
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"eval_loss": 1.326880931854248,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11350
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"learning_rate": 1.8561957279584205e-05,
|
|
"loss": 1.3997,
|
|
"step": 11400
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"eval_loss": 1.3318688869476318,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11400
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"learning_rate": 1.855561893896178e-05,
|
|
"loss": 1.4113,
|
|
"step": 11450
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"eval_loss": 1.3329873085021973,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11450
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"learning_rate": 1.8549280598339355e-05,
|
|
"loss": 1.4228,
|
|
"step": 11500
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"eval_loss": 1.3230202198028564,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11500
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"learning_rate": 1.8542942257716933e-05,
|
|
"loss": 1.4155,
|
|
"step": 11550
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"eval_loss": 1.324250340461731,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11550
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.8536603917094508e-05,
|
|
"loss": 1.411,
|
|
"step": 11600
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"eval_loss": 1.3332675695419312,
|
|
"eval_runtime": 3.7821,
|
|
"eval_samples_per_second": 13.22,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11600
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.8530265576472083e-05,
|
|
"loss": 1.4028,
|
|
"step": 11650
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"eval_loss": 1.3285317420959473,
|
|
"eval_runtime": 3.7792,
|
|
"eval_samples_per_second": 13.23,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11650
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.8523927235849657e-05,
|
|
"loss": 1.4285,
|
|
"step": 11700
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"eval_loss": 1.3309657573699951,
|
|
"eval_runtime": 3.7791,
|
|
"eval_samples_per_second": 13.231,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11700
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.8517588895227232e-05,
|
|
"loss": 1.4237,
|
|
"step": 11750
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"eval_loss": 1.325954556465149,
|
|
"eval_runtime": 3.7762,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 11750
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.8511250554604804e-05,
|
|
"loss": 1.4486,
|
|
"step": 11800
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"eval_loss": 1.3282676935195923,
|
|
"eval_runtime": 3.7804,
|
|
"eval_samples_per_second": 13.226,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11800
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.850491221398238e-05,
|
|
"loss": 1.4111,
|
|
"step": 11850
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"eval_loss": 1.3274351358413696,
|
|
"eval_runtime": 3.7811,
|
|
"eval_samples_per_second": 13.224,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11850
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.8498573873359957e-05,
|
|
"loss": 1.435,
|
|
"step": 11900
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"eval_loss": 1.3289445638656616,
|
|
"eval_runtime": 3.7805,
|
|
"eval_samples_per_second": 13.226,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11900
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"learning_rate": 1.849223553273753e-05,
|
|
"loss": 1.4033,
|
|
"step": 11950
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"eval_loss": 1.3252307176589966,
|
|
"eval_runtime": 3.7815,
|
|
"eval_samples_per_second": 13.222,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 11950
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"learning_rate": 1.8485897192115106e-05,
|
|
"loss": 1.4148,
|
|
"step": 12000
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"eval_loss": 1.3234962224960327,
|
|
"eval_runtime": 3.777,
|
|
"eval_samples_per_second": 13.238,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 12000
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"learning_rate": 1.847955885149268e-05,
|
|
"loss": 1.4118,
|
|
"step": 12050
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"eval_loss": 1.330000877380371,
|
|
"eval_runtime": 3.7787,
|
|
"eval_samples_per_second": 13.232,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 12050
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"learning_rate": 1.8473220510870256e-05,
|
|
"loss": 1.4157,
|
|
"step": 12100
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"eval_loss": 1.3198950290679932,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12100
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"learning_rate": 1.846688217024783e-05,
|
|
"loss": 1.4225,
|
|
"step": 12150
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"eval_loss": 1.3209259510040283,
|
|
"eval_runtime": 3.7643,
|
|
"eval_samples_per_second": 13.283,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12150
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"learning_rate": 1.8460543829625405e-05,
|
|
"loss": 1.4071,
|
|
"step": 12200
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"eval_loss": 1.3278964757919312,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 12200
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.845420548900298e-05,
|
|
"loss": 1.4045,
|
|
"step": 12250
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"eval_loss": 1.325210690498352,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 12250
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.8447867148380555e-05,
|
|
"loss": 1.4075,
|
|
"step": 12300
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"eval_loss": 1.3226252794265747,
|
|
"eval_runtime": 3.7668,
|
|
"eval_samples_per_second": 13.274,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12300
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.844152880775813e-05,
|
|
"loss": 1.3924,
|
|
"step": 12350
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"eval_loss": 1.3236082792282104,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12350
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.8435190467135708e-05,
|
|
"loss": 1.4138,
|
|
"step": 12400
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"eval_loss": 1.3255488872528076,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 12400
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.8428852126513283e-05,
|
|
"loss": 1.412,
|
|
"step": 12450
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"eval_loss": 1.321956992149353,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12450
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.8422513785890854e-05,
|
|
"loss": 1.4241,
|
|
"step": 12500
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"eval_loss": 1.3237371444702148,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12500
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.841617544526843e-05,
|
|
"loss": 1.4174,
|
|
"step": 12550
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"eval_loss": 1.3220739364624023,
|
|
"eval_runtime": 3.7664,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12550
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.8409837104646003e-05,
|
|
"loss": 1.4013,
|
|
"step": 12600
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"eval_loss": 1.3229719400405884,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 12600
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.8403498764023578e-05,
|
|
"loss": 1.4309,
|
|
"step": 12650
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"eval_loss": 1.3267501592636108,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12650
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.8397160423401153e-05,
|
|
"loss": 1.4127,
|
|
"step": 12700
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"eval_loss": 1.322455883026123,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12700
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"learning_rate": 1.839082208277873e-05,
|
|
"loss": 1.4143,
|
|
"step": 12750
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"eval_loss": 1.323978304862976,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12750
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"learning_rate": 1.8384483742156306e-05,
|
|
"loss": 1.4402,
|
|
"step": 12800
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"eval_loss": 1.3225884437561035,
|
|
"eval_runtime": 3.7678,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12800
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"learning_rate": 1.837814540153388e-05,
|
|
"loss": 1.402,
|
|
"step": 12850
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"eval_loss": 1.3222764730453491,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12850
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.8371807060911456e-05,
|
|
"loss": 1.4062,
|
|
"step": 12900
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"eval_loss": 1.3183293342590332,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12900
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.836546872028903e-05,
|
|
"loss": 1.4102,
|
|
"step": 12950
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"eval_loss": 1.3232107162475586,
|
|
"eval_runtime": 3.7675,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 12950
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.8359130379666605e-05,
|
|
"loss": 1.4074,
|
|
"step": 13000
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"eval_loss": 1.3166580200195312,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13000
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.835279203904418e-05,
|
|
"loss": 1.4229,
|
|
"step": 13050
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"eval_loss": 1.3245640993118286,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13050
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.8346453698421755e-05,
|
|
"loss": 1.3976,
|
|
"step": 13100
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"eval_loss": 1.3256316184997559,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13100
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.834011535779933e-05,
|
|
"loss": 1.4366,
|
|
"step": 13150
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"eval_loss": 1.321149468421936,
|
|
"eval_runtime": 3.7665,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13150
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"learning_rate": 1.8333777017176904e-05,
|
|
"loss": 1.4223,
|
|
"step": 13200
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"eval_loss": 1.3268109560012817,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13200
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"learning_rate": 1.832743867655448e-05,
|
|
"loss": 1.4311,
|
|
"step": 13250
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"eval_loss": 1.3200267553329468,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13250
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"learning_rate": 1.8321100335932054e-05,
|
|
"loss": 1.4095,
|
|
"step": 13300
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"eval_loss": 1.3248249292373657,
|
|
"eval_runtime": 3.7679,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13300
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.831476199530963e-05,
|
|
"loss": 1.3894,
|
|
"step": 13350
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"eval_loss": 1.320139765739441,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13350
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.8308423654687203e-05,
|
|
"loss": 1.4087,
|
|
"step": 13400
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"eval_loss": 1.323474407196045,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13400
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.8302085314064778e-05,
|
|
"loss": 1.3547,
|
|
"step": 13450
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"eval_loss": 1.324234127998352,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13450
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.8295746973442353e-05,
|
|
"loss": 1.4348,
|
|
"step": 13500
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"eval_loss": 1.3240126371383667,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13500
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.828940863281993e-05,
|
|
"loss": 1.4407,
|
|
"step": 13550
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"eval_loss": 1.3234056234359741,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13550
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.8283070292197506e-05,
|
|
"loss": 1.4077,
|
|
"step": 13600
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"eval_loss": 1.3208906650543213,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13600
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.827673195157508e-05,
|
|
"loss": 1.4193,
|
|
"step": 13650
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"eval_loss": 1.3239349126815796,
|
|
"eval_runtime": 3.7654,
|
|
"eval_samples_per_second": 13.279,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13650
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.8270393610952655e-05,
|
|
"loss": 1.3959,
|
|
"step": 13700
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"eval_loss": 1.3197979927062988,
|
|
"eval_runtime": 3.7671,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13700
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.826405527033023e-05,
|
|
"loss": 1.4251,
|
|
"step": 13750
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"eval_loss": 1.316582202911377,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13750
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.8257716929707805e-05,
|
|
"loss": 1.3933,
|
|
"step": 13800
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"eval_loss": 1.3277032375335693,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13800
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"learning_rate": 1.8251378589085376e-05,
|
|
"loss": 1.4111,
|
|
"step": 13850
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"eval_loss": 1.3177947998046875,
|
|
"eval_runtime": 3.7675,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 13850
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"learning_rate": 1.8245040248462954e-05,
|
|
"loss": 1.4167,
|
|
"step": 13900
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"eval_loss": 1.3203887939453125,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13900
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"learning_rate": 1.823870190784053e-05,
|
|
"loss": 1.4137,
|
|
"step": 13950
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"eval_loss": 1.3194351196289062,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 13950
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.8232363567218104e-05,
|
|
"loss": 1.435,
|
|
"step": 14000
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"eval_loss": 1.319800853729248,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 14000
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.822602522659568e-05,
|
|
"loss": 1.416,
|
|
"step": 14050
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"eval_loss": 1.324612021446228,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14050
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.8219686885973254e-05,
|
|
"loss": 1.4155,
|
|
"step": 14100
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"eval_loss": 1.321866512298584,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 14100
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.821334854535083e-05,
|
|
"loss": 1.3869,
|
|
"step": 14150
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"eval_loss": 1.3211545944213867,
|
|
"eval_runtime": 3.7746,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14150
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.8207010204728403e-05,
|
|
"loss": 1.3951,
|
|
"step": 14200
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"eval_loss": 1.321603536605835,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14200
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.8200671864105978e-05,
|
|
"loss": 1.3797,
|
|
"step": 14250
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"eval_loss": 1.3218741416931152,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 14250
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"learning_rate": 1.8194333523483553e-05,
|
|
"loss": 1.4026,
|
|
"step": 14300
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"eval_loss": 1.3158413171768188,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 14300
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"learning_rate": 1.8187995182861127e-05,
|
|
"loss": 1.4177,
|
|
"step": 14350
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"eval_loss": 1.326074242591858,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14350
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"learning_rate": 1.8181656842238706e-05,
|
|
"loss": 1.4113,
|
|
"step": 14400
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"eval_loss": 1.318434238433838,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14400
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"learning_rate": 1.817531850161628e-05,
|
|
"loss": 1.3909,
|
|
"step": 14450
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"eval_loss": 1.3192695379257202,
|
|
"eval_runtime": 3.7669,
|
|
"eval_samples_per_second": 13.274,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 14450
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"learning_rate": 1.8168980160993855e-05,
|
|
"loss": 1.4022,
|
|
"step": 14500
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"eval_loss": 1.3201017379760742,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14500
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"learning_rate": 1.8162641820371427e-05,
|
|
"loss": 1.3892,
|
|
"step": 14550
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"eval_loss": 1.32220458984375,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14550
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.8156303479749e-05,
|
|
"loss": 1.3939,
|
|
"step": 14600
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"eval_loss": 1.3229151964187622,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14600
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.8149965139126576e-05,
|
|
"loss": 1.4056,
|
|
"step": 14650
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"eval_loss": 1.3187562227249146,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14650
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.814362679850415e-05,
|
|
"loss": 1.3882,
|
|
"step": 14700
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"eval_loss": 1.3208309412002563,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14700
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.813728845788173e-05,
|
|
"loss": 1.3949,
|
|
"step": 14750
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"eval_loss": 1.3208937644958496,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14750
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.8130950117259304e-05,
|
|
"loss": 1.4009,
|
|
"step": 14800
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"eval_loss": 1.3226596117019653,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14800
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.812461177663688e-05,
|
|
"loss": 1.4117,
|
|
"step": 14850
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"eval_loss": 1.3224977254867554,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14850
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.8118273436014453e-05,
|
|
"loss": 1.4055,
|
|
"step": 14900
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"eval_loss": 1.3202691078186035,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14900
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"learning_rate": 1.8111935095392028e-05,
|
|
"loss": 1.4189,
|
|
"step": 14950
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"eval_loss": 1.3199864625930786,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 14950
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"learning_rate": 1.8105596754769603e-05,
|
|
"loss": 1.3899,
|
|
"step": 15000
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"eval_loss": 1.3188636302947998,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15000
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"learning_rate": 1.8099258414147178e-05,
|
|
"loss": 1.4209,
|
|
"step": 15050
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"eval_loss": 1.3177701234817505,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15050
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.8092920073524752e-05,
|
|
"loss": 1.4017,
|
|
"step": 15100
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"eval_loss": 1.3189513683319092,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15100
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.8086581732902327e-05,
|
|
"loss": 1.4006,
|
|
"step": 15150
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"eval_loss": 1.3205195665359497,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15150
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.8080243392279902e-05,
|
|
"loss": 1.4041,
|
|
"step": 15200
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"eval_loss": 1.3222854137420654,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15200
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"learning_rate": 1.8073905051657477e-05,
|
|
"loss": 1.3937,
|
|
"step": 15250
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"eval_loss": 1.3183249235153198,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15250
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"learning_rate": 1.806756671103505e-05,
|
|
"loss": 1.3928,
|
|
"step": 15300
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"eval_loss": 1.3198072910308838,
|
|
"eval_runtime": 3.7755,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15300
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"learning_rate": 1.8061228370412626e-05,
|
|
"loss": 1.3818,
|
|
"step": 15350
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"eval_loss": 1.3190951347351074,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15350
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.80548900297902e-05,
|
|
"loss": 1.3871,
|
|
"step": 15400
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"eval_loss": 1.3201597929000854,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15400
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.8048551689167776e-05,
|
|
"loss": 1.4073,
|
|
"step": 15450
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"eval_loss": 1.3151811361312866,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15450
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.804221334854535e-05,
|
|
"loss": 1.4235,
|
|
"step": 15500
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"eval_loss": 1.3252227306365967,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15500
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.803587500792293e-05,
|
|
"loss": 1.4128,
|
|
"step": 15550
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"eval_loss": 1.318635106086731,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15550
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.8029536667300504e-05,
|
|
"loss": 1.415,
|
|
"step": 15600
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"eval_loss": 1.3195527791976929,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15600
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.802319832667808e-05,
|
|
"loss": 1.41,
|
|
"step": 15650
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"eval_loss": 1.320070505142212,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15650
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.8016859986055653e-05,
|
|
"loss": 1.4014,
|
|
"step": 15700
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"eval_loss": 1.316506028175354,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15700
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"learning_rate": 1.8010521645433228e-05,
|
|
"loss": 1.4115,
|
|
"step": 15750
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_loss": 1.3209404945373535,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15750
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"learning_rate": 1.8004183304810803e-05,
|
|
"loss": 1.4245,
|
|
"step": 15800
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_loss": 1.3161125183105469,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15800
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"learning_rate": 1.7997844964188378e-05,
|
|
"loss": 1.4057,
|
|
"step": 15850
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_loss": 1.3166499137878418,
|
|
"eval_runtime": 3.7676,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15850
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"learning_rate": 1.7991506623565952e-05,
|
|
"loss": 1.3842,
|
|
"step": 15900
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"eval_loss": 1.316170334815979,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 15900
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"learning_rate": 1.7985168282943527e-05,
|
|
"loss": 1.3804,
|
|
"step": 15950
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"eval_loss": 1.3176696300506592,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 15950
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"learning_rate": 1.7978829942321102e-05,
|
|
"loss": 1.3854,
|
|
"step": 16000
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"eval_loss": 1.3182568550109863,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16000
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"learning_rate": 1.7972491601698677e-05,
|
|
"loss": 1.3661,
|
|
"step": 16050
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"eval_loss": 1.3186635971069336,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16050
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"learning_rate": 1.796615326107625e-05,
|
|
"loss": 1.3759,
|
|
"step": 16100
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"eval_loss": 1.31988525390625,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16100
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"learning_rate": 1.7959814920453826e-05,
|
|
"loss": 1.3816,
|
|
"step": 16150
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"eval_loss": 1.3128911256790161,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16150
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"learning_rate": 1.79534765798314e-05,
|
|
"loss": 1.3771,
|
|
"step": 16200
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"eval_loss": 1.3184537887573242,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16200
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"learning_rate": 1.7947138239208976e-05,
|
|
"loss": 1.3847,
|
|
"step": 16250
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"eval_loss": 1.3200515508651733,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16250
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"learning_rate": 1.794079989858655e-05,
|
|
"loss": 1.3918,
|
|
"step": 16300
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"eval_loss": 1.3177796602249146,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16300
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.7934461557964125e-05,
|
|
"loss": 1.3673,
|
|
"step": 16350
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"eval_loss": 1.3130323886871338,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16350
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.7928123217341703e-05,
|
|
"loss": 1.3802,
|
|
"step": 16400
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"eval_loss": 1.318576693534851,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16400
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.7921784876719278e-05,
|
|
"loss": 1.386,
|
|
"step": 16450
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"eval_loss": 1.321928858757019,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16450
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.7915446536096853e-05,
|
|
"loss": 1.35,
|
|
"step": 16500
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"eval_loss": 1.3189457654953003,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16500
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.7909108195474428e-05,
|
|
"loss": 1.3824,
|
|
"step": 16550
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"eval_loss": 1.3211742639541626,
|
|
"eval_runtime": 3.7752,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16550
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.7902769854852e-05,
|
|
"loss": 1.395,
|
|
"step": 16600
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"eval_loss": 1.3174998760223389,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16600
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.7896431514229574e-05,
|
|
"loss": 1.3795,
|
|
"step": 16650
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"eval_loss": 1.314656376838684,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16650
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"learning_rate": 1.789009317360715e-05,
|
|
"loss": 1.3968,
|
|
"step": 16700
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"eval_loss": 1.3165061473846436,
|
|
"eval_runtime": 3.7672,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16700
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"learning_rate": 1.7883754832984727e-05,
|
|
"loss": 1.377,
|
|
"step": 16750
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"eval_loss": 1.310689091682434,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16750
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"learning_rate": 1.78774164923623e-05,
|
|
"loss": 1.3927,
|
|
"step": 16800
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"eval_loss": 1.3171191215515137,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16800
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.7871078151739876e-05,
|
|
"loss": 1.3634,
|
|
"step": 16850
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"eval_loss": 1.3151047229766846,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16850
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.786473981111745e-05,
|
|
"loss": 1.3743,
|
|
"step": 16900
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"eval_loss": 1.3156814575195312,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 16900
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.7858401470495026e-05,
|
|
"loss": 1.4184,
|
|
"step": 16950
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"eval_loss": 1.3098622560501099,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 16950
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"learning_rate": 1.78520631298726e-05,
|
|
"loss": 1.382,
|
|
"step": 17000
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"eval_loss": 1.3162333965301514,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 17000
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"learning_rate": 1.7845724789250176e-05,
|
|
"loss": 1.3924,
|
|
"step": 17050
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"eval_loss": 1.3188899755477905,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17050
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"learning_rate": 1.783938644862775e-05,
|
|
"loss": 1.3574,
|
|
"step": 17100
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"eval_loss": 1.316525936126709,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17100
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"learning_rate": 1.7833048108005325e-05,
|
|
"loss": 1.3725,
|
|
"step": 17150
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"eval_loss": 1.317154049873352,
|
|
"eval_runtime": 3.7772,
|
|
"eval_samples_per_second": 13.237,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 17150
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"learning_rate": 1.78267097673829e-05,
|
|
"loss": 1.4028,
|
|
"step": 17200
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"eval_loss": 1.3122023344039917,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17200
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"learning_rate": 1.7820371426760478e-05,
|
|
"loss": 1.39,
|
|
"step": 17250
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"eval_loss": 1.3124886751174927,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17250
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"learning_rate": 1.781403308613805e-05,
|
|
"loss": 1.3834,
|
|
"step": 17300
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"eval_loss": 1.3173812627792358,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 17300
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"learning_rate": 1.7807694745515624e-05,
|
|
"loss": 1.3844,
|
|
"step": 17350
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"eval_loss": 1.3155839443206787,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17350
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"learning_rate": 1.78013564048932e-05,
|
|
"loss": 1.3753,
|
|
"step": 17400
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"eval_loss": 1.3156344890594482,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17400
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.7795144831083224e-05,
|
|
"loss": 1.3764,
|
|
"step": 17450
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"eval_loss": 1.3103086948394775,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17450
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.77888064904608e-05,
|
|
"loss": 1.3855,
|
|
"step": 17500
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"eval_loss": 1.3111189603805542,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 17500
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.7782468149838374e-05,
|
|
"loss": 1.4001,
|
|
"step": 17550
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"eval_loss": 1.3170082569122314,
|
|
"eval_runtime": 3.776,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17550
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.777612980921595e-05,
|
|
"loss": 1.3889,
|
|
"step": 17600
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"eval_loss": 1.3102355003356934,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17600
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.7769791468593523e-05,
|
|
"loss": 1.3734,
|
|
"step": 17650
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"eval_loss": 1.3260153532028198,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17650
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.7763453127971098e-05,
|
|
"loss": 1.3679,
|
|
"step": 17700
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"eval_loss": 1.3219112157821655,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17700
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.7757114787348673e-05,
|
|
"loss": 1.4034,
|
|
"step": 17750
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"eval_loss": 1.3143190145492554,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17750
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"learning_rate": 1.7750776446726248e-05,
|
|
"loss": 1.3499,
|
|
"step": 17800
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"eval_loss": 1.3147352933883667,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17800
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"learning_rate": 1.7744564872916273e-05,
|
|
"loss": 1.3789,
|
|
"step": 17850
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"eval_loss": 1.3178960084915161,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17850
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"learning_rate": 1.7738226532293848e-05,
|
|
"loss": 1.4097,
|
|
"step": 17900
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"eval_loss": 1.3134623765945435,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 17900
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"learning_rate": 1.7731888191671422e-05,
|
|
"loss": 1.4072,
|
|
"step": 17950
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"eval_loss": 1.315904974937439,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 17950
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"learning_rate": 1.7725549851048997e-05,
|
|
"loss": 1.3676,
|
|
"step": 18000
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"eval_loss": 1.3206485509872437,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18000
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"learning_rate": 1.7719211510426572e-05,
|
|
"loss": 1.4065,
|
|
"step": 18050
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"eval_loss": 1.3188087940216064,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18050
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.7712873169804147e-05,
|
|
"loss": 1.3728,
|
|
"step": 18100
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"eval_loss": 1.3164232969284058,
|
|
"eval_runtime": 4.4184,
|
|
"eval_samples_per_second": 11.316,
|
|
"eval_steps_per_second": 0.453,
|
|
"step": 18100
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.770653482918172e-05,
|
|
"loss": 1.372,
|
|
"step": 18150
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"eval_loss": 1.314164161682129,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18150
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.7700196488559296e-05,
|
|
"loss": 1.377,
|
|
"step": 18200
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"eval_loss": 1.3114633560180664,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18200
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 1.769385814793687e-05,
|
|
"loss": 1.3739,
|
|
"step": 18250
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"eval_loss": 1.3073451519012451,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18250
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 1.7687519807314446e-05,
|
|
"loss": 1.3593,
|
|
"step": 18300
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"eval_loss": 1.3098547458648682,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18300
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 1.768118146669202e-05,
|
|
"loss": 1.3727,
|
|
"step": 18350
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"eval_loss": 1.316469430923462,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18350
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"learning_rate": 1.7674843126069595e-05,
|
|
"loss": 1.4008,
|
|
"step": 18400
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"eval_loss": 1.314163327217102,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18400
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"learning_rate": 1.766850478544717e-05,
|
|
"loss": 1.3854,
|
|
"step": 18450
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"eval_loss": 1.3162171840667725,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18450
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"learning_rate": 1.7662166444824745e-05,
|
|
"loss": 1.4238,
|
|
"step": 18500
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"eval_loss": 1.3114781379699707,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18500
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.7655828104202323e-05,
|
|
"loss": 1.375,
|
|
"step": 18550
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"eval_loss": 1.311274766921997,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18550
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.7649489763579898e-05,
|
|
"loss": 1.3987,
|
|
"step": 18600
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"eval_loss": 1.3154056072235107,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18600
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.7643151422957473e-05,
|
|
"loss": 1.3622,
|
|
"step": 18650
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"eval_loss": 1.3166488409042358,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18650
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.7636813082335047e-05,
|
|
"loss": 1.3634,
|
|
"step": 18700
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"eval_loss": 1.309778094291687,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 18700
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"learning_rate": 1.7630474741712622e-05,
|
|
"loss": 1.3899,
|
|
"step": 18750
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"eval_loss": 1.3120732307434082,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18750
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"learning_rate": 1.7624136401090194e-05,
|
|
"loss": 1.3726,
|
|
"step": 18800
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"eval_loss": 1.3112399578094482,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18800
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"learning_rate": 1.761779806046777e-05,
|
|
"loss": 1.3993,
|
|
"step": 18850
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"eval_loss": 1.3085761070251465,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18850
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"learning_rate": 1.7611459719845346e-05,
|
|
"loss": 1.3778,
|
|
"step": 18900
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"eval_loss": 1.3147207498550415,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18900
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"learning_rate": 1.760512137922292e-05,
|
|
"loss": 1.3733,
|
|
"step": 18950
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"eval_loss": 1.3182721138000488,
|
|
"eval_runtime": 3.7754,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 18950
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"learning_rate": 1.7598783038600496e-05,
|
|
"loss": 1.3861,
|
|
"step": 19000
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"eval_loss": 1.3165748119354248,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19000
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"learning_rate": 1.759244469797807e-05,
|
|
"loss": 1.384,
|
|
"step": 19050
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"eval_loss": 1.311570167541504,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 19050
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"learning_rate": 1.7586106357355646e-05,
|
|
"loss": 1.3816,
|
|
"step": 19100
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"eval_loss": 1.3148012161254883,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 19100
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"learning_rate": 1.757976801673322e-05,
|
|
"loss": 1.3539,
|
|
"step": 19150
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"eval_loss": 1.3159393072128296,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19150
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.7573429676110795e-05,
|
|
"loss": 1.4083,
|
|
"step": 19200
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"eval_loss": 1.3101261854171753,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19200
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.756709133548837e-05,
|
|
"loss": 1.3744,
|
|
"step": 19250
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"eval_loss": 1.313029170036316,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19250
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.7560752994865945e-05,
|
|
"loss": 1.3793,
|
|
"step": 19300
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"eval_loss": 1.3120161294937134,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19300
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"learning_rate": 1.7554414654243523e-05,
|
|
"loss": 1.3693,
|
|
"step": 19350
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"eval_loss": 1.3200759887695312,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19350
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"learning_rate": 1.7548076313621098e-05,
|
|
"loss": 1.4094,
|
|
"step": 19400
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"eval_loss": 1.3178431987762451,
|
|
"eval_runtime": 3.7769,
|
|
"eval_samples_per_second": 13.238,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19400
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"learning_rate": 1.7541737972998672e-05,
|
|
"loss": 1.3821,
|
|
"step": 19450
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"eval_loss": 1.3110003471374512,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19450
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"learning_rate": 1.7535399632376244e-05,
|
|
"loss": 1.3681,
|
|
"step": 19500
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"eval_loss": 1.3211922645568848,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19500
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"learning_rate": 1.752906129175382e-05,
|
|
"loss": 1.3909,
|
|
"step": 19550
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"eval_loss": 1.3123958110809326,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19550
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"learning_rate": 1.7522722951131393e-05,
|
|
"loss": 1.358,
|
|
"step": 19600
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"eval_loss": 1.3105111122131348,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19600
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.7516384610508968e-05,
|
|
"loss": 1.3636,
|
|
"step": 19650
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"eval_loss": 1.3122894763946533,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19650
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.7510046269886546e-05,
|
|
"loss": 1.3695,
|
|
"step": 19700
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"eval_loss": 1.3131426572799683,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19700
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.750370792926412e-05,
|
|
"loss": 1.3873,
|
|
"step": 19750
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"eval_loss": 1.3153001070022583,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19750
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.7497369588641696e-05,
|
|
"loss": 1.375,
|
|
"step": 19800
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"eval_loss": 1.311892032623291,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19800
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"learning_rate": 1.749103124801927e-05,
|
|
"loss": 1.3752,
|
|
"step": 19850
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"eval_loss": 1.3058511018753052,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19850
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"learning_rate": 1.7484692907396845e-05,
|
|
"loss": 1.3709,
|
|
"step": 19900
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"eval_loss": 1.3147565126419067,
|
|
"eval_runtime": 3.779,
|
|
"eval_samples_per_second": 13.231,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 19900
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"learning_rate": 1.747835456677442e-05,
|
|
"loss": 1.4037,
|
|
"step": 19950
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"eval_loss": 1.3101611137390137,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 19950
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"learning_rate": 1.7472016226151995e-05,
|
|
"loss": 1.375,
|
|
"step": 20000
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"eval_loss": 1.3099762201309204,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20000
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"learning_rate": 1.7465804652342017e-05,
|
|
"loss": 1.3588,
|
|
"step": 20050
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"eval_loss": 1.3158316612243652,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20050
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"learning_rate": 1.745946631171959e-05,
|
|
"loss": 1.3767,
|
|
"step": 20100
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"eval_loss": 1.313093900680542,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20100
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"learning_rate": 1.7453127971097166e-05,
|
|
"loss": 1.3444,
|
|
"step": 20150
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"eval_loss": 1.3141725063323975,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20150
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"learning_rate": 1.7446789630474744e-05,
|
|
"loss": 1.3756,
|
|
"step": 20200
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"eval_loss": 1.3108264207839966,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20200
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"learning_rate": 1.744045128985232e-05,
|
|
"loss": 1.3849,
|
|
"step": 20250
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"eval_loss": 1.3127444982528687,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20250
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.7434112949229894e-05,
|
|
"loss": 1.3854,
|
|
"step": 20300
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"eval_loss": 1.3133628368377686,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20300
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.742777460860747e-05,
|
|
"loss": 1.3468,
|
|
"step": 20350
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"eval_loss": 1.3124346733093262,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20350
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.7421436267985044e-05,
|
|
"loss": 1.3677,
|
|
"step": 20400
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"eval_loss": 1.3119783401489258,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20400
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"learning_rate": 1.741509792736262e-05,
|
|
"loss": 1.3636,
|
|
"step": 20450
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"eval_loss": 1.3125382661819458,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20450
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"learning_rate": 1.7408759586740193e-05,
|
|
"loss": 1.3721,
|
|
"step": 20500
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"eval_loss": 1.3122693300247192,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20500
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"learning_rate": 1.7402421246117768e-05,
|
|
"loss": 1.3852,
|
|
"step": 20550
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"eval_loss": 1.3100529909133911,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20550
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 1.7396082905495343e-05,
|
|
"loss": 1.3518,
|
|
"step": 20600
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"eval_loss": 1.3218302726745605,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20600
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 1.7389744564872917e-05,
|
|
"loss": 1.3852,
|
|
"step": 20650
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"eval_loss": 1.3104270696640015,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20650
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 1.7383406224250492e-05,
|
|
"loss": 1.3826,
|
|
"step": 20700
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"eval_loss": 1.3127949237823486,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20700
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 1.7377067883628067e-05,
|
|
"loss": 1.3877,
|
|
"step": 20750
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"eval_loss": 1.3149232864379883,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20750
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"learning_rate": 1.7370729543005642e-05,
|
|
"loss": 1.3652,
|
|
"step": 20800
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"eval_loss": 1.3115489482879639,
|
|
"eval_runtime": 3.7747,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20800
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"learning_rate": 1.7364391202383217e-05,
|
|
"loss": 1.3574,
|
|
"step": 20850
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"eval_loss": 1.3170777559280396,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20850
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"learning_rate": 1.735805286176079e-05,
|
|
"loss": 1.3702,
|
|
"step": 20900
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"eval_loss": 1.3097151517868042,
|
|
"eval_runtime": 3.7748,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 20900
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 1.7351714521138366e-05,
|
|
"loss": 1.3735,
|
|
"step": 20950
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"eval_loss": 1.3063160181045532,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 20950
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 1.7345376180515944e-05,
|
|
"loss": 1.3689,
|
|
"step": 21000
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"eval_loss": 1.311417818069458,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21000
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 1.733903783989352e-05,
|
|
"loss": 1.3907,
|
|
"step": 21050
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"eval_loss": 1.3121113777160645,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21050
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"learning_rate": 1.7332699499271094e-05,
|
|
"loss": 1.3769,
|
|
"step": 21100
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"eval_loss": 1.3109712600708008,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21100
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"learning_rate": 1.732636115864867e-05,
|
|
"loss": 1.389,
|
|
"step": 21150
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"eval_loss": 1.3077359199523926,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21150
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"learning_rate": 1.7320022818026243e-05,
|
|
"loss": 1.3578,
|
|
"step": 21200
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"eval_loss": 1.3106968402862549,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21200
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"learning_rate": 1.7313684477403818e-05,
|
|
"loss": 1.4056,
|
|
"step": 21250
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"eval_loss": 1.3085529804229736,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21250
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"learning_rate": 1.730734613678139e-05,
|
|
"loss": 1.3689,
|
|
"step": 21300
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"eval_loss": 1.314315676689148,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21300
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"learning_rate": 1.7301007796158968e-05,
|
|
"loss": 1.3702,
|
|
"step": 21350
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"eval_loss": 1.3193659782409668,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21350
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 1.7294669455536542e-05,
|
|
"loss": 1.3792,
|
|
"step": 21400
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"eval_loss": 1.313162088394165,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21400
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 1.7288331114914117e-05,
|
|
"loss": 1.3839,
|
|
"step": 21450
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"eval_loss": 1.313085913658142,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21450
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 1.7281992774291692e-05,
|
|
"loss": 1.3622,
|
|
"step": 21500
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"eval_loss": 1.311946988105774,
|
|
"eval_runtime": 3.7754,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21500
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 1.7275654433669267e-05,
|
|
"loss": 1.3506,
|
|
"step": 21550
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"eval_loss": 1.3043593168258667,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21550
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 1.726931609304684e-05,
|
|
"loss": 1.3686,
|
|
"step": 21600
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"eval_loss": 1.314071774482727,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21600
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 1.7262977752424416e-05,
|
|
"loss": 1.3661,
|
|
"step": 21650
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"eval_loss": 1.3092154264450073,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21650
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 1.725663941180199e-05,
|
|
"loss": 1.3656,
|
|
"step": 21700
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"eval_loss": 1.3098061084747314,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21700
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"learning_rate": 1.7250301071179566e-05,
|
|
"loss": 1.3612,
|
|
"step": 21750
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"eval_loss": 1.3110477924346924,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21750
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"learning_rate": 1.724396273055714e-05,
|
|
"loss": 1.3457,
|
|
"step": 21800
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"eval_loss": 1.3121354579925537,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21800
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"learning_rate": 1.723762438993472e-05,
|
|
"loss": 1.3665,
|
|
"step": 21850
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"eval_loss": 1.3136210441589355,
|
|
"eval_runtime": 3.7679,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21850
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"learning_rate": 1.7231286049312294e-05,
|
|
"loss": 1.3802,
|
|
"step": 21900
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"eval_loss": 1.3116146326065063,
|
|
"eval_runtime": 3.7767,
|
|
"eval_samples_per_second": 13.239,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 21900
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"learning_rate": 1.7224947708689865e-05,
|
|
"loss": 1.375,
|
|
"step": 21950
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"eval_loss": 1.308994174003601,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 21950
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"learning_rate": 1.721860936806744e-05,
|
|
"loss": 1.3647,
|
|
"step": 22000
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"eval_loss": 1.3138041496276855,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22000
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"learning_rate": 1.7212271027445015e-05,
|
|
"loss": 1.3526,
|
|
"step": 22050
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"eval_loss": 1.310899257659912,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 22050
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"learning_rate": 1.720593268682259e-05,
|
|
"loss": 1.3776,
|
|
"step": 22100
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"eval_loss": 1.3149839639663696,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22100
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"learning_rate": 1.7199594346200164e-05,
|
|
"loss": 1.3508,
|
|
"step": 22150
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"eval_loss": 1.3105652332305908,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22150
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"learning_rate": 1.7193256005577742e-05,
|
|
"loss": 1.382,
|
|
"step": 22200
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"eval_loss": 1.3149703741073608,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22200
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"learning_rate": 1.7186917664955317e-05,
|
|
"loss": 1.3785,
|
|
"step": 22250
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"eval_loss": 1.3060740232467651,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22250
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"learning_rate": 1.7180579324332892e-05,
|
|
"loss": 1.3838,
|
|
"step": 22300
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"eval_loss": 1.3014811277389526,
|
|
"eval_runtime": 3.7766,
|
|
"eval_samples_per_second": 13.24,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22300
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"learning_rate": 1.7174240983710467e-05,
|
|
"loss": 1.366,
|
|
"step": 22350
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"eval_loss": 1.3100804090499878,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22350
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"learning_rate": 1.716790264308804e-05,
|
|
"loss": 1.3847,
|
|
"step": 22400
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"eval_loss": 1.3137387037277222,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22400
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"learning_rate": 1.7161564302465616e-05,
|
|
"loss": 1.3804,
|
|
"step": 22450
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"eval_loss": 1.3055957555770874,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22450
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"learning_rate": 1.715522596184319e-05,
|
|
"loss": 1.3863,
|
|
"step": 22500
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"eval_loss": 1.3113762140274048,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22500
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"learning_rate": 1.7148887621220766e-05,
|
|
"loss": 1.3623,
|
|
"step": 22550
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"eval_loss": 1.3158456087112427,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22550
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"learning_rate": 1.714254928059834e-05,
|
|
"loss": 1.3742,
|
|
"step": 22600
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"eval_loss": 1.30805242061615,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22600
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.7136210939975915e-05,
|
|
"loss": 1.3907,
|
|
"step": 22650
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"eval_loss": 1.3145793676376343,
|
|
"eval_runtime": 3.7765,
|
|
"eval_samples_per_second": 13.24,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22650
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.712987259935349e-05,
|
|
"loss": 1.373,
|
|
"step": 22700
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"eval_loss": 1.3100571632385254,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22700
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.7123534258731065e-05,
|
|
"loss": 1.3572,
|
|
"step": 22750
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"eval_loss": 1.312801718711853,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22750
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.711719591810864e-05,
|
|
"loss": 1.3852,
|
|
"step": 22800
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"eval_loss": 1.3154374361038208,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22800
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"learning_rate": 1.7110857577486214e-05,
|
|
"loss": 1.3665,
|
|
"step": 22850
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"eval_loss": 1.3039851188659668,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22850
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"learning_rate": 1.710451923686379e-05,
|
|
"loss": 1.3497,
|
|
"step": 22900
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"eval_loss": 1.3083395957946777,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22900
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"learning_rate": 1.7098180896241364e-05,
|
|
"loss": 1.3593,
|
|
"step": 22950
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"eval_loss": 1.303240180015564,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 22950
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"learning_rate": 1.7091842555618942e-05,
|
|
"loss": 1.3722,
|
|
"step": 23000
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"eval_loss": 1.3092082738876343,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 23000
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"learning_rate": 1.7085504214996517e-05,
|
|
"loss": 1.3386,
|
|
"step": 23050
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"eval_loss": 1.311246395111084,
|
|
"eval_runtime": 3.7755,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23050
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"learning_rate": 1.707916587437409e-05,
|
|
"loss": 1.3754,
|
|
"step": 23100
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"eval_loss": 1.3097790479660034,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 23100
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 1.7072827533751666e-05,
|
|
"loss": 1.3886,
|
|
"step": 23150
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"eval_loss": 1.3141131401062012,
|
|
"eval_runtime": 3.7665,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 23150
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 1.706648919312924e-05,
|
|
"loss": 1.4021,
|
|
"step": 23200
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"eval_loss": 1.307629108428955,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 23200
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 1.7060150852506816e-05,
|
|
"loss": 1.3659,
|
|
"step": 23250
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"eval_loss": 1.3088146448135376,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23250
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"learning_rate": 1.705381251188439e-05,
|
|
"loss": 1.3739,
|
|
"step": 23300
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"eval_loss": 1.307511329650879,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23300
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"learning_rate": 1.7047474171261966e-05,
|
|
"loss": 1.3992,
|
|
"step": 23350
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"eval_loss": 1.3039134740829468,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23350
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"learning_rate": 1.704113583063954e-05,
|
|
"loss": 1.388,
|
|
"step": 23400
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"eval_loss": 1.3038339614868164,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23400
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"learning_rate": 1.7034797490017115e-05,
|
|
"loss": 1.3658,
|
|
"step": 23450
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"eval_loss": 1.3082275390625,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23450
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"learning_rate": 1.702845914939469e-05,
|
|
"loss": 1.3879,
|
|
"step": 23500
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"eval_loss": 1.3087126016616821,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23500
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"learning_rate": 1.7022120808772265e-05,
|
|
"loss": 1.3706,
|
|
"step": 23550
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"eval_loss": 1.306036114692688,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23550
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 1.701578246814984e-05,
|
|
"loss": 1.3534,
|
|
"step": 23600
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"eval_loss": 1.311568260192871,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 23600
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 1.7009444127527414e-05,
|
|
"loss": 1.4031,
|
|
"step": 23650
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"eval_loss": 1.3089189529418945,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23650
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 1.700310578690499e-05,
|
|
"loss": 1.3769,
|
|
"step": 23700
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"eval_loss": 1.305688500404358,
|
|
"eval_runtime": 4.1635,
|
|
"eval_samples_per_second": 12.009,
|
|
"eval_steps_per_second": 0.48,
|
|
"step": 23700
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 1.6996767446282564e-05,
|
|
"loss": 1.3467,
|
|
"step": 23750
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"eval_loss": 1.3095543384552002,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23750
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"learning_rate": 1.699042910566014e-05,
|
|
"loss": 1.367,
|
|
"step": 23800
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"eval_loss": 1.3076460361480713,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23800
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"learning_rate": 1.6984090765037717e-05,
|
|
"loss": 1.3616,
|
|
"step": 23850
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"eval_loss": 1.3067145347595215,
|
|
"eval_runtime": 3.7775,
|
|
"eval_samples_per_second": 13.236,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 23850
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"learning_rate": 1.697775242441529e-05,
|
|
"loss": 1.3575,
|
|
"step": 23900
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"eval_loss": 1.3107126951217651,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23900
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"learning_rate": 1.6971414083792866e-05,
|
|
"loss": 1.3616,
|
|
"step": 23950
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"eval_loss": 1.3045284748077393,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 23950
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"learning_rate": 1.6965075743170438e-05,
|
|
"loss": 1.4114,
|
|
"step": 24000
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"eval_loss": 1.3077421188354492,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24000
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"learning_rate": 1.6958737402548012e-05,
|
|
"loss": 1.3635,
|
|
"step": 24050
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"eval_loss": 1.3037108182907104,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24050
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"learning_rate": 1.6952399061925587e-05,
|
|
"loss": 1.371,
|
|
"step": 24100
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"eval_loss": 1.3108770847320557,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24100
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"learning_rate": 1.6946060721303162e-05,
|
|
"loss": 1.3572,
|
|
"step": 24150
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"eval_loss": 1.3085899353027344,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24150
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"learning_rate": 1.6939849147493187e-05,
|
|
"loss": 1.3874,
|
|
"step": 24200
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"eval_loss": 1.305327296257019,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24200
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"learning_rate": 1.6933510806870762e-05,
|
|
"loss": 1.3531,
|
|
"step": 24250
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"eval_loss": 1.3097971677780151,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24250
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"learning_rate": 1.6927172466248337e-05,
|
|
"loss": 1.3532,
|
|
"step": 24300
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"eval_loss": 1.304316759109497,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24300
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"learning_rate": 1.6920834125625915e-05,
|
|
"loss": 1.3541,
|
|
"step": 24350
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"eval_loss": 1.3054053783416748,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24350
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 1.691449578500349e-05,
|
|
"loss": 1.3296,
|
|
"step": 24400
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"eval_loss": 1.302626132965088,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24400
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 1.690815744438106e-05,
|
|
"loss": 1.3758,
|
|
"step": 24450
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"eval_loss": 1.3106575012207031,
|
|
"eval_runtime": 3.7782,
|
|
"eval_samples_per_second": 13.234,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 24450
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 1.6901819103758636e-05,
|
|
"loss": 1.3679,
|
|
"step": 24500
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"eval_loss": 1.3039908409118652,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24500
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"learning_rate": 1.689548076313621e-05,
|
|
"loss": 1.3822,
|
|
"step": 24550
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"eval_loss": 1.3058950901031494,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24550
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"learning_rate": 1.6889142422513785e-05,
|
|
"loss": 1.3612,
|
|
"step": 24600
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"eval_loss": 1.309067726135254,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24600
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"learning_rate": 1.6882804081891364e-05,
|
|
"loss": 1.3772,
|
|
"step": 24650
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"eval_loss": 1.3012826442718506,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24650
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 1.6876465741268938e-05,
|
|
"loss": 1.3703,
|
|
"step": 24700
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"eval_loss": 1.3039789199829102,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24700
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 1.6870127400646513e-05,
|
|
"loss": 1.3777,
|
|
"step": 24750
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"eval_loss": 1.3049794435501099,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24750
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 1.6863789060024088e-05,
|
|
"loss": 1.3586,
|
|
"step": 24800
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"eval_loss": 1.3099775314331055,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24800
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 1.6857450719401663e-05,
|
|
"loss": 1.3348,
|
|
"step": 24850
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"eval_loss": 1.307841181755066,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 24850
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 1.6851112378779237e-05,
|
|
"loss": 1.3668,
|
|
"step": 24900
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"eval_loss": 1.3069860935211182,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 24900
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 1.684490080496926e-05,
|
|
"loss": 1.3574,
|
|
"step": 24950
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"eval_loss": 1.3024590015411377,
|
|
"eval_runtime": 3.7803,
|
|
"eval_samples_per_second": 13.227,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 24950
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 1.6838562464346834e-05,
|
|
"loss": 1.3535,
|
|
"step": 25000
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"eval_loss": 1.3021525144577026,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25000
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"learning_rate": 1.683222412372441e-05,
|
|
"loss": 1.3727,
|
|
"step": 25050
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"eval_loss": 1.3053768873214722,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25050
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"learning_rate": 1.6825885783101984e-05,
|
|
"loss": 1.3698,
|
|
"step": 25100
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"eval_loss": 1.305099606513977,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25100
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"learning_rate": 1.681954744247956e-05,
|
|
"loss": 1.363,
|
|
"step": 25150
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"eval_loss": 1.3059375286102295,
|
|
"eval_runtime": 3.7671,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 25150
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"learning_rate": 1.6813209101857136e-05,
|
|
"loss": 1.3596,
|
|
"step": 25200
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"eval_loss": 1.3022193908691406,
|
|
"eval_runtime": 3.7782,
|
|
"eval_samples_per_second": 13.234,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 25200
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"learning_rate": 1.680687076123471e-05,
|
|
"loss": 1.371,
|
|
"step": 25250
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"eval_loss": 1.3017759323120117,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25250
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"learning_rate": 1.6800532420612286e-05,
|
|
"loss": 1.3608,
|
|
"step": 25300
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"eval_loss": 1.3046982288360596,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25300
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"learning_rate": 1.679419407998986e-05,
|
|
"loss": 1.3938,
|
|
"step": 25350
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"eval_loss": 1.301830768585205,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25350
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"learning_rate": 1.6787855739367436e-05,
|
|
"loss": 1.3648,
|
|
"step": 25400
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"eval_loss": 1.306410789489746,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25400
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"learning_rate": 1.678151739874501e-05,
|
|
"loss": 1.3493,
|
|
"step": 25450
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"eval_loss": 1.3086802959442139,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25450
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"learning_rate": 1.6775179058122585e-05,
|
|
"loss": 1.3439,
|
|
"step": 25500
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"eval_loss": 1.3023231029510498,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25500
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"learning_rate": 1.676884071750016e-05,
|
|
"loss": 1.3811,
|
|
"step": 25550
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"eval_loss": 1.3024687767028809,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25550
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"learning_rate": 1.6762502376877735e-05,
|
|
"loss": 1.3477,
|
|
"step": 25600
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"eval_loss": 1.3032989501953125,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25600
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 1.675616403625531e-05,
|
|
"loss": 1.3487,
|
|
"step": 25650
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"eval_loss": 1.3073408603668213,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25650
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 1.6749825695632884e-05,
|
|
"loss": 1.3728,
|
|
"step": 25700
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"eval_loss": 1.3046859502792358,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25700
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 1.674348735501046e-05,
|
|
"loss": 1.3465,
|
|
"step": 25750
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"eval_loss": 1.3042266368865967,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25750
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 1.6737149014388034e-05,
|
|
"loss": 1.3813,
|
|
"step": 25800
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"eval_loss": 1.301363468170166,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 25800
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"learning_rate": 1.673081067376561e-05,
|
|
"loss": 1.3882,
|
|
"step": 25850
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"eval_loss": 1.3051186800003052,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25850
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"learning_rate": 1.6724472333143183e-05,
|
|
"loss": 1.3215,
|
|
"step": 25900
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"eval_loss": 1.3043999671936035,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25900
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"learning_rate": 1.6718133992520758e-05,
|
|
"loss": 1.3655,
|
|
"step": 25950
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"eval_loss": 1.3121837377548218,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 25950
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 1.6711922418710783e-05,
|
|
"loss": 1.3882,
|
|
"step": 26000
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"eval_loss": 1.3099251985549927,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26000
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 1.6705584078088358e-05,
|
|
"loss": 1.3717,
|
|
"step": 26050
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"eval_loss": 1.3041249513626099,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26050
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 1.6699245737465933e-05,
|
|
"loss": 1.3372,
|
|
"step": 26100
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"eval_loss": 1.3081796169281006,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26100
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"learning_rate": 1.6692907396843508e-05,
|
|
"loss": 1.3834,
|
|
"step": 26150
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"eval_loss": 1.3096734285354614,
|
|
"eval_runtime": 3.7776,
|
|
"eval_samples_per_second": 13.236,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 26150
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"learning_rate": 1.6686569056221082e-05,
|
|
"loss": 1.3696,
|
|
"step": 26200
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"eval_loss": 1.3115854263305664,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26200
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"learning_rate": 1.6680230715598657e-05,
|
|
"loss": 1.3565,
|
|
"step": 26250
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"eval_loss": 1.301089882850647,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26250
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"learning_rate": 1.6673892374976232e-05,
|
|
"loss": 1.3849,
|
|
"step": 26300
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"eval_loss": 1.3010739088058472,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26300
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"learning_rate": 1.6667554034353807e-05,
|
|
"loss": 1.3544,
|
|
"step": 26350
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"eval_loss": 1.3007947206497192,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26350
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"learning_rate": 1.666121569373138e-05,
|
|
"loss": 1.3639,
|
|
"step": 26400
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"eval_loss": 1.2976875305175781,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26400
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"learning_rate": 1.665487735310896e-05,
|
|
"loss": 1.3575,
|
|
"step": 26450
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"eval_loss": 1.3054636716842651,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26450
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"learning_rate": 1.6648539012486534e-05,
|
|
"loss": 1.3649,
|
|
"step": 26500
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"eval_loss": 1.3066623210906982,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26500
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"learning_rate": 1.664220067186411e-05,
|
|
"loss": 1.3842,
|
|
"step": 26550
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"eval_loss": 1.305976152420044,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26550
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 1.6635862331241684e-05,
|
|
"loss": 1.3713,
|
|
"step": 26600
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"eval_loss": 1.3037573099136353,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26600
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 1.6629523990619255e-05,
|
|
"loss": 1.3812,
|
|
"step": 26650
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"eval_loss": 1.302822232246399,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26650
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 1.662318564999683e-05,
|
|
"loss": 1.3599,
|
|
"step": 26700
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"eval_loss": 1.303798794746399,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26700
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 1.6616847309374405e-05,
|
|
"loss": 1.3628,
|
|
"step": 26750
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"eval_loss": 1.3052434921264648,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26750
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"learning_rate": 1.6610508968751983e-05,
|
|
"loss": 1.3746,
|
|
"step": 26800
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"eval_loss": 1.3013827800750732,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26800
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"learning_rate": 1.6604170628129558e-05,
|
|
"loss": 1.3751,
|
|
"step": 26850
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"eval_loss": 1.3035222291946411,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26850
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"learning_rate": 1.6597832287507133e-05,
|
|
"loss": 1.3778,
|
|
"step": 26900
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"eval_loss": 1.3040968179702759,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 26900
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"learning_rate": 1.6591493946884707e-05,
|
|
"loss": 1.3506,
|
|
"step": 26950
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"eval_loss": 1.3041977882385254,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 26950
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"learning_rate": 1.6585155606262282e-05,
|
|
"loss": 1.3554,
|
|
"step": 27000
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"eval_loss": 1.3110145330429077,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27000
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"learning_rate": 1.6578817265639857e-05,
|
|
"loss": 1.352,
|
|
"step": 27050
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"eval_loss": 1.3075376749038696,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27050
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"learning_rate": 1.6572478925017432e-05,
|
|
"loss": 1.3545,
|
|
"step": 27100
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"eval_loss": 1.2995811700820923,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27100
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"learning_rate": 1.6566140584395007e-05,
|
|
"loss": 1.3543,
|
|
"step": 27150
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"eval_loss": 1.3038663864135742,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27150
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"learning_rate": 1.655980224377258e-05,
|
|
"loss": 1.3854,
|
|
"step": 27200
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"eval_loss": 1.2978318929672241,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27200
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"learning_rate": 1.6553463903150156e-05,
|
|
"loss": 1.351,
|
|
"step": 27250
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"eval_loss": 1.3069562911987305,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 27250
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"learning_rate": 1.654712556252773e-05,
|
|
"loss": 1.3662,
|
|
"step": 27300
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"eval_loss": 1.3018920421600342,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27300
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"learning_rate": 1.6540787221905306e-05,
|
|
"loss": 1.3442,
|
|
"step": 27350
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"eval_loss": 1.3017157316207886,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27350
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"learning_rate": 1.653444888128288e-05,
|
|
"loss": 1.3583,
|
|
"step": 27400
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"eval_loss": 1.3102655410766602,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27400
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"learning_rate": 1.6528110540660455e-05,
|
|
"loss": 1.3517,
|
|
"step": 27450
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"eval_loss": 1.3030922412872314,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27450
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"learning_rate": 1.652177220003803e-05,
|
|
"loss": 1.3491,
|
|
"step": 27500
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"eval_loss": 1.3042519092559814,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27500
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"learning_rate": 1.6515433859415605e-05,
|
|
"loss": 1.3722,
|
|
"step": 27550
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"eval_loss": 1.3028026819229126,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27550
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"learning_rate": 1.650909551879318e-05,
|
|
"loss": 1.3524,
|
|
"step": 27600
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"eval_loss": 1.3010423183441162,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27600
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"learning_rate": 1.6502757178170758e-05,
|
|
"loss": 1.3673,
|
|
"step": 27650
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"eval_loss": 1.3038748502731323,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27650
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 1.6496418837548332e-05,
|
|
"loss": 1.3756,
|
|
"step": 27700
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"eval_loss": 1.3028035163879395,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27700
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 1.6490080496925907e-05,
|
|
"loss": 1.3457,
|
|
"step": 27750
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"eval_loss": 1.3020347356796265,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27750
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 1.6483742156303482e-05,
|
|
"loss": 1.3395,
|
|
"step": 27800
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"eval_loss": 1.3019260168075562,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27800
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 1.6477403815681057e-05,
|
|
"loss": 1.3466,
|
|
"step": 27850
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"eval_loss": 1.3060325384140015,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27850
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"learning_rate": 1.647106547505863e-05,
|
|
"loss": 1.3408,
|
|
"step": 27900
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"eval_loss": 1.3008695840835571,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27900
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"learning_rate": 1.6464727134436206e-05,
|
|
"loss": 1.3354,
|
|
"step": 27950
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"eval_loss": 1.3000245094299316,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 27950
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"learning_rate": 1.645838879381378e-05,
|
|
"loss": 1.3522,
|
|
"step": 28000
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"eval_loss": 1.3058466911315918,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28000
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"learning_rate": 1.6452050453191356e-05,
|
|
"loss": 1.3577,
|
|
"step": 28050
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"eval_loss": 1.30289626121521,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28050
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"learning_rate": 1.644571211256893e-05,
|
|
"loss": 1.3627,
|
|
"step": 28100
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"eval_loss": 1.305790662765503,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28100
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"learning_rate": 1.6439373771946505e-05,
|
|
"loss": 1.3648,
|
|
"step": 28150
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"eval_loss": 1.3006936311721802,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28150
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"learning_rate": 1.643303543132408e-05,
|
|
"loss": 1.3505,
|
|
"step": 28200
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"eval_loss": 1.3063627481460571,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28200
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"learning_rate": 1.6426697090701655e-05,
|
|
"loss": 1.3814,
|
|
"step": 28250
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"eval_loss": 1.3062107563018799,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28250
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"learning_rate": 1.642035875007923e-05,
|
|
"loss": 1.3513,
|
|
"step": 28300
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"eval_loss": 1.3031411170959473,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28300
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"learning_rate": 1.6414020409456805e-05,
|
|
"loss": 1.3459,
|
|
"step": 28350
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"eval_loss": 1.3107800483703613,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28350
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"learning_rate": 1.640768206883438e-05,
|
|
"loss": 1.3782,
|
|
"step": 28400
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"eval_loss": 1.3018364906311035,
|
|
"eval_runtime": 3.7759,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28400
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"learning_rate": 1.6401343728211958e-05,
|
|
"loss": 1.3813,
|
|
"step": 28450
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"eval_loss": 1.3011293411254883,
|
|
"eval_runtime": 3.7759,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28450
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"learning_rate": 1.6395005387589532e-05,
|
|
"loss": 1.3655,
|
|
"step": 28500
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"eval_loss": 1.3019990921020508,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28500
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"learning_rate": 1.6388667046967107e-05,
|
|
"loss": 1.368,
|
|
"step": 28550
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"eval_loss": 1.2994332313537598,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28550
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"learning_rate": 1.6382328706344682e-05,
|
|
"loss": 1.3653,
|
|
"step": 28600
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"eval_loss": 1.303462266921997,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28600
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"learning_rate": 1.6375990365722257e-05,
|
|
"loss": 1.3523,
|
|
"step": 28650
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"eval_loss": 1.3095431327819824,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28650
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"learning_rate": 1.6369652025099828e-05,
|
|
"loss": 1.3651,
|
|
"step": 28700
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"eval_loss": 1.308469533920288,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28700
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"learning_rate": 1.6363313684477403e-05,
|
|
"loss": 1.3479,
|
|
"step": 28750
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"eval_loss": 1.3007841110229492,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28750
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"learning_rate": 1.635697534385498e-05,
|
|
"loss": 1.378,
|
|
"step": 28800
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"eval_loss": 1.302858829498291,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28800
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"learning_rate": 1.6350637003232556e-05,
|
|
"loss": 1.365,
|
|
"step": 28850
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"eval_loss": 1.3014613389968872,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 28850
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"learning_rate": 1.634429866261013e-05,
|
|
"loss": 1.3549,
|
|
"step": 28900
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"eval_loss": 1.300676703453064,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 28900
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"learning_rate": 1.6337960321987705e-05,
|
|
"loss": 1.3532,
|
|
"step": 28950
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"eval_loss": 1.299085259437561,
|
|
"eval_runtime": 3.7662,
|
|
"eval_samples_per_second": 13.276,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 28950
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"learning_rate": 1.633162198136528e-05,
|
|
"loss": 1.3653,
|
|
"step": 29000
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"eval_loss": 1.3018462657928467,
|
|
"eval_runtime": 3.7758,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29000
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"learning_rate": 1.6325283640742855e-05,
|
|
"loss": 1.3511,
|
|
"step": 29050
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"eval_loss": 1.2959985733032227,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29050
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"learning_rate": 1.631894530012043e-05,
|
|
"loss": 1.3681,
|
|
"step": 29100
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"eval_loss": 1.2985323667526245,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 29100
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"learning_rate": 1.6312606959498004e-05,
|
|
"loss": 1.3345,
|
|
"step": 29150
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"eval_loss": 1.2978233098983765,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29150
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"learning_rate": 1.630626861887558e-05,
|
|
"loss": 1.3514,
|
|
"step": 29200
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"eval_loss": 1.2977333068847656,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29200
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"learning_rate": 1.6299930278253154e-05,
|
|
"loss": 1.3734,
|
|
"step": 29250
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"eval_loss": 1.2973655462265015,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 29250
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"learning_rate": 1.6293591937630732e-05,
|
|
"loss": 1.362,
|
|
"step": 29300
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"eval_loss": 1.3016226291656494,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29300
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"learning_rate": 1.6287253597008307e-05,
|
|
"loss": 1.3705,
|
|
"step": 29350
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"eval_loss": 1.2990288734436035,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29350
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"learning_rate": 1.6280915256385878e-05,
|
|
"loss": 1.3745,
|
|
"step": 29400
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"eval_loss": 1.3012051582336426,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29400
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"learning_rate": 1.6274576915763453e-05,
|
|
"loss": 1.3363,
|
|
"step": 29450
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"eval_loss": 1.3014848232269287,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29450
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"learning_rate": 1.6268238575141028e-05,
|
|
"loss": 1.3617,
|
|
"step": 29500
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"eval_loss": 1.2997702360153198,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29500
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"learning_rate": 1.6261900234518603e-05,
|
|
"loss": 1.3848,
|
|
"step": 29550
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"eval_loss": 1.3055031299591064,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 29550
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"learning_rate": 1.6255561893896177e-05,
|
|
"loss": 1.3323,
|
|
"step": 29600
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"eval_loss": 1.297582745552063,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 29600
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"learning_rate": 1.6249223553273756e-05,
|
|
"loss": 1.3466,
|
|
"step": 29650
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"eval_loss": 1.2984001636505127,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29650
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"learning_rate": 1.624288521265133e-05,
|
|
"loss": 1.3588,
|
|
"step": 29700
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"eval_loss": 1.3052302598953247,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29700
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"learning_rate": 1.6236546872028905e-05,
|
|
"loss": 1.3532,
|
|
"step": 29750
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"eval_loss": 1.2981939315795898,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29750
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"learning_rate": 1.623020853140648e-05,
|
|
"loss": 1.3335,
|
|
"step": 29800
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"eval_loss": 1.303356647491455,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29800
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"learning_rate": 1.6223870190784055e-05,
|
|
"loss": 1.3695,
|
|
"step": 29850
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"eval_loss": 1.2969903945922852,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29850
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"learning_rate": 1.621753185016163e-05,
|
|
"loss": 1.3723,
|
|
"step": 29900
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"eval_loss": 1.2991429567337036,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29900
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"learning_rate": 1.6211193509539204e-05,
|
|
"loss": 1.354,
|
|
"step": 29950
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"eval_loss": 1.2996094226837158,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 29950
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"learning_rate": 1.620485516891678e-05,
|
|
"loss": 1.3663,
|
|
"step": 30000
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"eval_loss": 1.2983999252319336,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30000
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"learning_rate": 1.6198516828294354e-05,
|
|
"loss": 1.3448,
|
|
"step": 30050
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"eval_loss": 1.3011195659637451,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30050
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"learning_rate": 1.619217848767193e-05,
|
|
"loss": 1.3733,
|
|
"step": 30100
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"eval_loss": 1.2934362888336182,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30100
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"learning_rate": 1.6185966913861954e-05,
|
|
"loss": 1.3324,
|
|
"step": 30150
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"eval_loss": 1.3005836009979248,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30150
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"learning_rate": 1.617962857323953e-05,
|
|
"loss": 1.3719,
|
|
"step": 30200
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"eval_loss": 1.2997490167617798,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30200
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"learning_rate": 1.6173290232617103e-05,
|
|
"loss": 1.36,
|
|
"step": 30250
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"eval_loss": 1.2998112440109253,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 30250
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"learning_rate": 1.6166951891994678e-05,
|
|
"loss": 1.3526,
|
|
"step": 30300
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"eval_loss": 1.2992067337036133,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30300
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"learning_rate": 1.6160613551372253e-05,
|
|
"loss": 1.3635,
|
|
"step": 30350
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"eval_loss": 1.2963122129440308,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30350
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"learning_rate": 1.6154275210749828e-05,
|
|
"loss": 1.3587,
|
|
"step": 30400
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"eval_loss": 1.3038519620895386,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30400
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"learning_rate": 1.6147936870127402e-05,
|
|
"loss": 1.3605,
|
|
"step": 30450
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"eval_loss": 1.3022255897521973,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30450
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"learning_rate": 1.6141598529504977e-05,
|
|
"loss": 1.3308,
|
|
"step": 30500
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"eval_loss": 1.2980443239212036,
|
|
"eval_runtime": 3.7678,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 30500
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"learning_rate": 1.6135260188882552e-05,
|
|
"loss": 1.3461,
|
|
"step": 30550
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"eval_loss": 1.3008222579956055,
|
|
"eval_runtime": 3.7757,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30550
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"learning_rate": 1.6128921848260127e-05,
|
|
"loss": 1.3649,
|
|
"step": 30600
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"eval_loss": 1.2989284992218018,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30600
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"learning_rate": 1.61225835076377e-05,
|
|
"loss": 1.3483,
|
|
"step": 30650
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"eval_loss": 1.3000298738479614,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30650
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 1.6116245167015276e-05,
|
|
"loss": 1.3603,
|
|
"step": 30700
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"eval_loss": 1.2987351417541504,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 30700
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 1.610990682639285e-05,
|
|
"loss": 1.3625,
|
|
"step": 30750
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"eval_loss": 1.303464651107788,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30750
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 1.6103568485770426e-05,
|
|
"loss": 1.3899,
|
|
"step": 30800
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"eval_loss": 1.2980417013168335,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30800
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 1.6097230145148e-05,
|
|
"loss": 1.3662,
|
|
"step": 30850
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"eval_loss": 1.296869158744812,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 30850
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"learning_rate": 1.6090891804525575e-05,
|
|
"loss": 1.3511,
|
|
"step": 30900
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"eval_loss": 1.3018838167190552,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 30900
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"learning_rate": 1.6084553463903154e-05,
|
|
"loss": 1.3401,
|
|
"step": 30950
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"eval_loss": 1.299475073814392,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 30950
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"learning_rate": 1.6078215123280728e-05,
|
|
"loss": 1.3381,
|
|
"step": 31000
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"eval_loss": 1.3049876689910889,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31000
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"learning_rate": 1.6071876782658303e-05,
|
|
"loss": 1.3677,
|
|
"step": 31050
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"eval_loss": 1.2973532676696777,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 31050
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"learning_rate": 1.6065538442035878e-05,
|
|
"loss": 1.3761,
|
|
"step": 31100
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"eval_loss": 1.2960923910140991,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 31100
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"learning_rate": 1.605920010141345e-05,
|
|
"loss": 1.3679,
|
|
"step": 31150
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"eval_loss": 1.3015331029891968,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31150
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"learning_rate": 1.6052861760791024e-05,
|
|
"loss": 1.3642,
|
|
"step": 31200
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"eval_loss": 1.2991764545440674,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31200
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"learning_rate": 1.60465234201686e-05,
|
|
"loss": 1.3392,
|
|
"step": 31250
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"eval_loss": 1.3025933504104614,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31250
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"learning_rate": 1.6040185079546177e-05,
|
|
"loss": 1.3413,
|
|
"step": 31300
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"eval_loss": 1.3018898963928223,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31300
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"learning_rate": 1.6033846738923752e-05,
|
|
"loss": 1.3416,
|
|
"step": 31350
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"eval_loss": 1.3048564195632935,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31350
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"learning_rate": 1.6027508398301327e-05,
|
|
"loss": 1.3574,
|
|
"step": 31400
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"eval_loss": 1.2960922718048096,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31400
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"learning_rate": 1.60211700576789e-05,
|
|
"loss": 1.367,
|
|
"step": 31450
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"eval_loss": 1.3009661436080933,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 31450
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"learning_rate": 1.6014831717056476e-05,
|
|
"loss": 1.3677,
|
|
"step": 31500
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_loss": 1.2992533445358276,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31500
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"learning_rate": 1.600849337643405e-05,
|
|
"loss": 1.3412,
|
|
"step": 31550
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_loss": 1.2949190139770508,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31550
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"learning_rate": 1.6002155035811626e-05,
|
|
"loss": 1.3602,
|
|
"step": 31600
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_loss": 1.297387957572937,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31600
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"learning_rate": 1.59958166951892e-05,
|
|
"loss": 1.3334,
|
|
"step": 31650
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"eval_loss": 1.2971540689468384,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31650
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"learning_rate": 1.5989478354566775e-05,
|
|
"loss": 1.3564,
|
|
"step": 31700
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"eval_loss": 1.3020439147949219,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31700
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"learning_rate": 1.598314001394435e-05,
|
|
"loss": 1.348,
|
|
"step": 31750
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"eval_loss": 1.3007164001464844,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31750
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"learning_rate": 1.5976801673321928e-05,
|
|
"loss": 1.3167,
|
|
"step": 31800
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"eval_loss": 1.2973201274871826,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 31800
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"learning_rate": 1.59704633326995e-05,
|
|
"loss": 1.3173,
|
|
"step": 31850
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"eval_loss": 1.303665280342102,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31850
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"learning_rate": 1.5964124992077074e-05,
|
|
"loss": 1.3595,
|
|
"step": 31900
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"eval_loss": 1.3055217266082764,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 31900
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"learning_rate": 1.595778665145465e-05,
|
|
"loss": 1.3164,
|
|
"step": 31950
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"eval_loss": 1.3072150945663452,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 31950
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"learning_rate": 1.5951448310832224e-05,
|
|
"loss": 1.3191,
|
|
"step": 32000
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"eval_loss": 1.303810715675354,
|
|
"eval_runtime": 3.7686,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 32000
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"learning_rate": 1.59451099702098e-05,
|
|
"loss": 1.3362,
|
|
"step": 32050
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"eval_loss": 1.2971174716949463,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32050
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"learning_rate": 1.5938771629587377e-05,
|
|
"loss": 1.3397,
|
|
"step": 32100
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"eval_loss": 1.3013195991516113,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32100
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"learning_rate": 1.593243328896495e-05,
|
|
"loss": 1.3473,
|
|
"step": 32150
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"eval_loss": 1.2995288372039795,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32150
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"learning_rate": 1.5926094948342526e-05,
|
|
"loss": 1.3496,
|
|
"step": 32200
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"eval_loss": 1.3053009510040283,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 32200
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"learning_rate": 1.59197566077201e-05,
|
|
"loss": 1.3477,
|
|
"step": 32250
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"eval_loss": 1.2995527982711792,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32250
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"learning_rate": 1.5913418267097676e-05,
|
|
"loss": 1.3573,
|
|
"step": 32300
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"eval_loss": 1.3038856983184814,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32300
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"learning_rate": 1.590707992647525e-05,
|
|
"loss": 1.3393,
|
|
"step": 32350
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"eval_loss": 1.3001644611358643,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32350
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"learning_rate": 1.5900741585852825e-05,
|
|
"loss": 1.348,
|
|
"step": 32400
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"eval_loss": 1.296790361404419,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32400
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"learning_rate": 1.58944032452304e-05,
|
|
"loss": 1.3525,
|
|
"step": 32450
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"eval_loss": 1.3004753589630127,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32450
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"learning_rate": 1.5888064904607975e-05,
|
|
"loss": 1.3293,
|
|
"step": 32500
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"eval_loss": 1.301581859588623,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32500
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"learning_rate": 1.588172656398555e-05,
|
|
"loss": 1.298,
|
|
"step": 32550
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"eval_loss": 1.2965809106826782,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32550
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"learning_rate": 1.5875388223363125e-05,
|
|
"loss": 1.3458,
|
|
"step": 32600
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"eval_loss": 1.3038264513015747,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32600
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"learning_rate": 1.58690498827407e-05,
|
|
"loss": 1.3364,
|
|
"step": 32650
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"eval_loss": 1.2998013496398926,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32650
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"learning_rate": 1.5862711542118274e-05,
|
|
"loss": 1.3295,
|
|
"step": 32700
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"eval_loss": 1.3004248142242432,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32700
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"learning_rate": 1.585637320149585e-05,
|
|
"loss": 1.3462,
|
|
"step": 32750
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"eval_loss": 1.2997674942016602,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 32750
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"learning_rate": 1.5850034860873424e-05,
|
|
"loss": 1.3776,
|
|
"step": 32800
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"eval_loss": 1.2985196113586426,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32800
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"learning_rate": 1.5843696520251e-05,
|
|
"loss": 1.3187,
|
|
"step": 32850
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"eval_loss": 1.2970695495605469,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32850
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"learning_rate": 1.5837358179628573e-05,
|
|
"loss": 1.3341,
|
|
"step": 32900
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"eval_loss": 1.2948143482208252,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 32900
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"learning_rate": 1.583101983900615e-05,
|
|
"loss": 1.3457,
|
|
"step": 32950
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"eval_loss": 1.3056460618972778,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 32950
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"learning_rate": 1.5824681498383726e-05,
|
|
"loss": 1.3235,
|
|
"step": 33000
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"eval_loss": 1.2999091148376465,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33000
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"learning_rate": 1.58183431577613e-05,
|
|
"loss": 1.3351,
|
|
"step": 33050
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"eval_loss": 1.3032591342926025,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33050
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"learning_rate": 1.5812004817138876e-05,
|
|
"loss": 1.3321,
|
|
"step": 33100
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"eval_loss": 1.2999855279922485,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33100
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"learning_rate": 1.580566647651645e-05,
|
|
"loss": 1.333,
|
|
"step": 33150
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"eval_loss": 1.294285535812378,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33150
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"learning_rate": 1.5799328135894022e-05,
|
|
"loss": 1.3566,
|
|
"step": 33200
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"eval_loss": 1.296621561050415,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33200
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"learning_rate": 1.5792989795271597e-05,
|
|
"loss": 1.3526,
|
|
"step": 33250
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"eval_loss": 1.297002911567688,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33250
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"learning_rate": 1.5786651454649175e-05,
|
|
"loss": 1.3372,
|
|
"step": 33300
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"eval_loss": 1.304821491241455,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33300
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"learning_rate": 1.578031311402675e-05,
|
|
"loss": 1.3549,
|
|
"step": 33350
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"eval_loss": 1.299644947052002,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 33350
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"learning_rate": 1.5773974773404324e-05,
|
|
"loss": 1.3425,
|
|
"step": 33400
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"eval_loss": 1.296370267868042,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33400
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"learning_rate": 1.57676364327819e-05,
|
|
"loss": 1.3041,
|
|
"step": 33450
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"eval_loss": 1.2937836647033691,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33450
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"learning_rate": 1.5761298092159474e-05,
|
|
"loss": 1.3093,
|
|
"step": 33500
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"eval_loss": 1.295989990234375,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 33500
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"learning_rate": 1.575495975153705e-05,
|
|
"loss": 1.3491,
|
|
"step": 33550
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"eval_loss": 1.3016209602355957,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33550
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"learning_rate": 1.5748621410914623e-05,
|
|
"loss": 1.3325,
|
|
"step": 33600
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"eval_loss": 1.3030569553375244,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33600
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"learning_rate": 1.5742283070292198e-05,
|
|
"loss": 1.3733,
|
|
"step": 33650
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"eval_loss": 1.2962210178375244,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33650
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"learning_rate": 1.5735944729669773e-05,
|
|
"loss": 1.3241,
|
|
"step": 33700
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"eval_loss": 1.297899603843689,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33700
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"learning_rate": 1.5729606389047348e-05,
|
|
"loss": 1.3103,
|
|
"step": 33750
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"eval_loss": 1.2943065166473389,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33750
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"learning_rate": 1.5723268048424926e-05,
|
|
"loss": 1.3702,
|
|
"step": 33800
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"eval_loss": 1.3004677295684814,
|
|
"eval_runtime": 3.7748,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33800
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"learning_rate": 1.57169297078025e-05,
|
|
"loss": 1.3574,
|
|
"step": 33850
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"eval_loss": 1.2956440448760986,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 33850
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"learning_rate": 1.5710591367180072e-05,
|
|
"loss": 1.3482,
|
|
"step": 33900
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"eval_loss": 1.2978112697601318,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 33900
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"learning_rate": 1.5704253026557647e-05,
|
|
"loss": 1.34,
|
|
"step": 33950
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"eval_loss": 1.302622675895691,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 33950
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"learning_rate": 1.569791468593522e-05,
|
|
"loss": 1.3286,
|
|
"step": 34000
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"eval_loss": 1.299667477607727,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34000
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"learning_rate": 1.5691576345312796e-05,
|
|
"loss": 1.3541,
|
|
"step": 34050
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"eval_loss": 1.2928789854049683,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34050
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"learning_rate": 1.5685238004690375e-05,
|
|
"loss": 1.3303,
|
|
"step": 34100
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"eval_loss": 1.2975785732269287,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34100
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"learning_rate": 1.567889966406795e-05,
|
|
"loss": 1.3209,
|
|
"step": 34150
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"eval_loss": 1.3008363246917725,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 34150
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"learning_rate": 1.5672561323445524e-05,
|
|
"loss": 1.3388,
|
|
"step": 34200
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"eval_loss": 1.2967095375061035,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34200
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"learning_rate": 1.56662229828231e-05,
|
|
"loss": 1.3325,
|
|
"step": 34250
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"eval_loss": 1.3002748489379883,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34250
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"learning_rate": 1.5659884642200674e-05,
|
|
"loss": 1.3388,
|
|
"step": 34300
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"eval_loss": 1.29612398147583,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34300
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"learning_rate": 1.565354630157825e-05,
|
|
"loss": 1.3453,
|
|
"step": 34350
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"eval_loss": 1.2992526292800903,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34350
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"learning_rate": 1.5647207960955823e-05,
|
|
"loss": 1.3108,
|
|
"step": 34400
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"eval_loss": 1.2986329793930054,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 34400
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"learning_rate": 1.5640869620333398e-05,
|
|
"loss": 1.3462,
|
|
"step": 34450
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"eval_loss": 1.2992162704467773,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34450
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"learning_rate": 1.563465804652342e-05,
|
|
"loss": 1.3451,
|
|
"step": 34500
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"eval_loss": 1.2982722520828247,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34500
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"learning_rate": 1.5628319705900995e-05,
|
|
"loss": 1.3529,
|
|
"step": 34550
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"eval_loss": 1.2952470779418945,
|
|
"eval_runtime": 3.7761,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34550
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"learning_rate": 1.5621981365278573e-05,
|
|
"loss": 1.3407,
|
|
"step": 34600
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"eval_loss": 1.3065522909164429,
|
|
"eval_runtime": 3.7757,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34600
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"learning_rate": 1.5615643024656148e-05,
|
|
"loss": 1.3363,
|
|
"step": 34650
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"eval_loss": 1.2977278232574463,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 34650
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"learning_rate": 1.5609304684033722e-05,
|
|
"loss": 1.3618,
|
|
"step": 34700
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"eval_loss": 1.3001843690872192,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 34700
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"learning_rate": 1.5602966343411297e-05,
|
|
"loss": 1.3342,
|
|
"step": 34750
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"eval_loss": 1.297072410583496,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34750
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"learning_rate": 1.5596628002788872e-05,
|
|
"loss": 1.3369,
|
|
"step": 34800
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"eval_loss": 1.306389331817627,
|
|
"eval_runtime": 3.782,
|
|
"eval_samples_per_second": 13.221,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 34800
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"learning_rate": 1.5590289662166447e-05,
|
|
"loss": 1.3271,
|
|
"step": 34850
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"eval_loss": 1.301561713218689,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34850
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"learning_rate": 1.558395132154402e-05,
|
|
"loss": 1.346,
|
|
"step": 34900
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"eval_loss": 1.3028095960617065,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 34900
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"learning_rate": 1.5577612980921596e-05,
|
|
"loss": 1.3541,
|
|
"step": 34950
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"eval_loss": 1.2969186305999756,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 34950
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"learning_rate": 1.557127464029917e-05,
|
|
"loss": 1.3241,
|
|
"step": 35000
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"eval_loss": 1.2969589233398438,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35000
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"learning_rate": 1.5564936299676746e-05,
|
|
"loss": 1.3578,
|
|
"step": 35050
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"eval_loss": 1.3046318292617798,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35050
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"learning_rate": 1.555859795905432e-05,
|
|
"loss": 1.3332,
|
|
"step": 35100
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"eval_loss": 1.2990514039993286,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35100
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"learning_rate": 1.5552259618431895e-05,
|
|
"loss": 1.3273,
|
|
"step": 35150
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"eval_loss": 1.2978941202163696,
|
|
"eval_runtime": 3.7747,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35150
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"learning_rate": 1.554592127780947e-05,
|
|
"loss": 1.3343,
|
|
"step": 35200
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"eval_loss": 1.2978546619415283,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35200
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"learning_rate": 1.5539582937187045e-05,
|
|
"loss": 1.3199,
|
|
"step": 35250
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"eval_loss": 1.2972488403320312,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35250
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"learning_rate": 1.553324459656462e-05,
|
|
"loss": 1.3495,
|
|
"step": 35300
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"eval_loss": 1.2972238063812256,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35300
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"learning_rate": 1.5526906255942194e-05,
|
|
"loss": 1.3439,
|
|
"step": 35350
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"eval_loss": 1.2991609573364258,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35350
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"learning_rate": 1.552056791531977e-05,
|
|
"loss": 1.3524,
|
|
"step": 35400
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"eval_loss": 1.294219970703125,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35400
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"learning_rate": 1.5514229574697347e-05,
|
|
"loss": 1.339,
|
|
"step": 35450
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"eval_loss": 1.292604923248291,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35450
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"learning_rate": 1.5507891234074922e-05,
|
|
"loss": 1.3358,
|
|
"step": 35500
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"eval_loss": 1.295400857925415,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35500
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"learning_rate": 1.5501552893452497e-05,
|
|
"loss": 1.3434,
|
|
"step": 35550
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"eval_loss": 1.2985919713974,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35550
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"learning_rate": 1.5495214552830072e-05,
|
|
"loss": 1.3394,
|
|
"step": 35600
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"eval_loss": 1.2987098693847656,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35600
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"learning_rate": 1.5488876212207646e-05,
|
|
"loss": 1.3503,
|
|
"step": 35650
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"eval_loss": 1.3002333641052246,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35650
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"learning_rate": 1.5482537871585218e-05,
|
|
"loss": 1.3232,
|
|
"step": 35700
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"eval_loss": 1.2978768348693848,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35700
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 1.5476199530962796e-05,
|
|
"loss": 1.3123,
|
|
"step": 35750
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"eval_loss": 1.2958645820617676,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35750
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 1.546986119034037e-05,
|
|
"loss": 1.3608,
|
|
"step": 35800
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"eval_loss": 1.2969403266906738,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35800
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 1.5463522849717946e-05,
|
|
"loss": 1.3377,
|
|
"step": 35850
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"eval_loss": 1.2962383031845093,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 35850
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 1.545718450909552e-05,
|
|
"loss": 1.3348,
|
|
"step": 35900
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"eval_loss": 1.2938027381896973,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35900
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"learning_rate": 1.5450846168473095e-05,
|
|
"loss": 1.3528,
|
|
"step": 35950
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"eval_loss": 1.2962663173675537,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 35950
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"learning_rate": 1.544450782785067e-05,
|
|
"loss": 1.3539,
|
|
"step": 36000
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"eval_loss": 1.2986482381820679,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36000
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"learning_rate": 1.5438169487228245e-05,
|
|
"loss": 1.3208,
|
|
"step": 36050
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"eval_loss": 1.3013414144515991,
|
|
"eval_runtime": 3.767,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 36050
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"learning_rate": 1.543183114660582e-05,
|
|
"loss": 1.356,
|
|
"step": 36100
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"eval_loss": 1.2966651916503906,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36100
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"learning_rate": 1.5425492805983394e-05,
|
|
"loss": 1.3602,
|
|
"step": 36150
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"eval_loss": 1.3037729263305664,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36150
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"learning_rate": 1.541915446536097e-05,
|
|
"loss": 1.324,
|
|
"step": 36200
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"eval_loss": 1.2935600280761719,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 36200
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"learning_rate": 1.5412816124738547e-05,
|
|
"loss": 1.3774,
|
|
"step": 36250
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"eval_loss": 1.2977104187011719,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36250
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"learning_rate": 1.5406477784116122e-05,
|
|
"loss": 1.3413,
|
|
"step": 36300
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"eval_loss": 1.2955683469772339,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36300
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"learning_rate": 1.5400139443493697e-05,
|
|
"loss": 1.3193,
|
|
"step": 36350
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"eval_loss": 1.2930635213851929,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 36350
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"learning_rate": 1.5393801102871268e-05,
|
|
"loss": 1.3253,
|
|
"step": 36400
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"eval_loss": 1.3015633821487427,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36400
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"learning_rate": 1.5387462762248843e-05,
|
|
"loss": 1.3222,
|
|
"step": 36450
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"eval_loss": 1.3011013269424438,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36450
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"learning_rate": 1.5381124421626418e-05,
|
|
"loss": 1.3329,
|
|
"step": 36500
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"eval_loss": 1.2986356019973755,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36500
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"learning_rate": 1.5374786081003992e-05,
|
|
"loss": 1.3612,
|
|
"step": 36550
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"eval_loss": 1.2964348793029785,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36550
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"learning_rate": 1.536844774038157e-05,
|
|
"loss": 1.3239,
|
|
"step": 36600
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"eval_loss": 1.2984528541564941,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36600
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"learning_rate": 1.5362109399759145e-05,
|
|
"loss": 1.3487,
|
|
"step": 36650
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"eval_loss": 1.2961621284484863,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36650
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 1.535577105913672e-05,
|
|
"loss": 1.3364,
|
|
"step": 36700
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"eval_loss": 1.2978811264038086,
|
|
"eval_runtime": 3.7676,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 36700
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 1.5349432718514295e-05,
|
|
"loss": 1.3158,
|
|
"step": 36750
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"eval_loss": 1.2946953773498535,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36750
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 1.534309437789187e-05,
|
|
"loss": 1.3134,
|
|
"step": 36800
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"eval_loss": 1.2971347570419312,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36800
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 1.5336756037269445e-05,
|
|
"loss": 1.3193,
|
|
"step": 36850
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"eval_loss": 1.2953553199768066,
|
|
"eval_runtime": 3.7756,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36850
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"learning_rate": 1.533041769664702e-05,
|
|
"loss": 1.3472,
|
|
"step": 36900
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"eval_loss": 1.2913790941238403,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36900
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"learning_rate": 1.5324079356024594e-05,
|
|
"loss": 1.3197,
|
|
"step": 36950
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"eval_loss": 1.2945469617843628,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 36950
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"learning_rate": 1.531774101540217e-05,
|
|
"loss": 1.3428,
|
|
"step": 37000
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"eval_loss": 1.2970346212387085,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37000
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"learning_rate": 1.5311402674779744e-05,
|
|
"loss": 1.3382,
|
|
"step": 37050
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"eval_loss": 1.2981318235397339,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37050
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"learning_rate": 1.530506433415732e-05,
|
|
"loss": 1.3126,
|
|
"step": 37100
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"eval_loss": 1.301912784576416,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37100
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"learning_rate": 1.5298725993534893e-05,
|
|
"loss": 1.3508,
|
|
"step": 37150
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"eval_loss": 1.2953013181686401,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37150
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"learning_rate": 1.5292387652912468e-05,
|
|
"loss": 1.3351,
|
|
"step": 37200
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"eval_loss": 1.295505404472351,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 37200
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"learning_rate": 1.5286049312290043e-05,
|
|
"loss": 1.3399,
|
|
"step": 37250
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"eval_loss": 1.2950987815856934,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37250
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"learning_rate": 1.5279710971667617e-05,
|
|
"loss": 1.3354,
|
|
"step": 37300
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"eval_loss": 1.3015360832214355,
|
|
"eval_runtime": 3.7823,
|
|
"eval_samples_per_second": 13.219,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 37300
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"learning_rate": 1.5273372631045192e-05,
|
|
"loss": 1.3312,
|
|
"step": 37350
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"eval_loss": 1.2964391708374023,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37350
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"learning_rate": 1.5267034290422767e-05,
|
|
"loss": 1.3408,
|
|
"step": 37400
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"eval_loss": 1.2919100522994995,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37400
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"learning_rate": 1.5260695949800345e-05,
|
|
"loss": 1.358,
|
|
"step": 37450
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"eval_loss": 1.2946839332580566,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37450
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"learning_rate": 1.5254357609177918e-05,
|
|
"loss": 1.3596,
|
|
"step": 37500
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"eval_loss": 1.2970671653747559,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37500
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"learning_rate": 1.5248019268555495e-05,
|
|
"loss": 1.3281,
|
|
"step": 37550
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"eval_loss": 1.2969365119934082,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 37550
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"learning_rate": 1.524168092793307e-05,
|
|
"loss": 1.3068,
|
|
"step": 37600
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"eval_loss": 1.2915878295898438,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37600
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"learning_rate": 1.5235342587310644e-05,
|
|
"loss": 1.3382,
|
|
"step": 37650
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"eval_loss": 1.2946220636367798,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37650
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"learning_rate": 1.5229004246688219e-05,
|
|
"loss": 1.3528,
|
|
"step": 37700
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"eval_loss": 1.2945616245269775,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37700
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"learning_rate": 1.5222665906065792e-05,
|
|
"loss": 1.3252,
|
|
"step": 37750
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"eval_loss": 1.2927005290985107,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37750
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"learning_rate": 1.5216454332255817e-05,
|
|
"loss": 1.3318,
|
|
"step": 37800
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"eval_loss": 1.2960560321807861,
|
|
"eval_runtime": 3.778,
|
|
"eval_samples_per_second": 13.235,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 37800
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"learning_rate": 1.5210115991633392e-05,
|
|
"loss": 1.3502,
|
|
"step": 37850
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"eval_loss": 1.2924162149429321,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37850
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"learning_rate": 1.5203777651010967e-05,
|
|
"loss": 1.3279,
|
|
"step": 37900
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"eval_loss": 1.2968086004257202,
|
|
"eval_runtime": 3.7756,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37900
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"learning_rate": 1.5197439310388542e-05,
|
|
"loss": 1.316,
|
|
"step": 37950
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"eval_loss": 1.2995247840881348,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 37950
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"learning_rate": 1.5191100969766116e-05,
|
|
"loss": 1.3257,
|
|
"step": 38000
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"eval_loss": 1.2933276891708374,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38000
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"learning_rate": 1.5184762629143693e-05,
|
|
"loss": 1.3383,
|
|
"step": 38050
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"eval_loss": 1.2948951721191406,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38050
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"learning_rate": 1.5178551055333715e-05,
|
|
"loss": 1.3345,
|
|
"step": 38100
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"eval_loss": 1.293394923210144,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38100
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"learning_rate": 1.517221271471129e-05,
|
|
"loss": 1.3533,
|
|
"step": 38150
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"eval_loss": 1.295020580291748,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38150
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"learning_rate": 1.5165874374088864e-05,
|
|
"loss": 1.3346,
|
|
"step": 38200
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"eval_loss": 1.2933144569396973,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38200
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"learning_rate": 1.5159536033466439e-05,
|
|
"loss": 1.3319,
|
|
"step": 38250
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"eval_loss": 1.2897781133651733,
|
|
"eval_runtime": 3.7672,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 38250
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"learning_rate": 1.5153197692844016e-05,
|
|
"loss": 1.2898,
|
|
"step": 38300
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"eval_loss": 1.294044017791748,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38300
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"learning_rate": 1.514685935222159e-05,
|
|
"loss": 1.3445,
|
|
"step": 38350
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"eval_loss": 1.2895246744155884,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38350
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"learning_rate": 1.5140521011599165e-05,
|
|
"loss": 1.3522,
|
|
"step": 38400
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"eval_loss": 1.2923784255981445,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38400
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"learning_rate": 1.513418267097674e-05,
|
|
"loss": 1.3403,
|
|
"step": 38450
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"eval_loss": 1.291225790977478,
|
|
"eval_runtime": 3.7679,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 38450
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"learning_rate": 1.5127844330354315e-05,
|
|
"loss": 1.3063,
|
|
"step": 38500
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"eval_loss": 1.2989550828933716,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38500
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"learning_rate": 1.5121505989731888e-05,
|
|
"loss": 1.3292,
|
|
"step": 38550
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"eval_loss": 1.2922873497009277,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38550
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"learning_rate": 1.5115167649109464e-05,
|
|
"loss": 1.3316,
|
|
"step": 38600
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"eval_loss": 1.2946301698684692,
|
|
"eval_runtime": 3.766,
|
|
"eval_samples_per_second": 13.277,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 38600
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"learning_rate": 1.5108829308487039e-05,
|
|
"loss": 1.3644,
|
|
"step": 38650
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"eval_loss": 1.2937086820602417,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38650
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"learning_rate": 1.5102490967864614e-05,
|
|
"loss": 1.324,
|
|
"step": 38700
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"eval_loss": 1.300889492034912,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38700
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"learning_rate": 1.5096152627242189e-05,
|
|
"loss": 1.3196,
|
|
"step": 38750
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"eval_loss": 1.2900445461273193,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38750
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"learning_rate": 1.5089814286619763e-05,
|
|
"loss": 1.3167,
|
|
"step": 38800
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"eval_loss": 1.2944896221160889,
|
|
"eval_runtime": 3.7762,
|
|
"eval_samples_per_second": 13.241,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38800
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"learning_rate": 1.508347594599734e-05,
|
|
"loss": 1.339,
|
|
"step": 38850
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"eval_loss": 1.2919236421585083,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38850
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"learning_rate": 1.5077137605374915e-05,
|
|
"loss": 1.3459,
|
|
"step": 38900
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"eval_loss": 1.2921767234802246,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 38900
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"learning_rate": 1.507079926475249e-05,
|
|
"loss": 1.3192,
|
|
"step": 38950
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"eval_loss": 1.2934505939483643,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 38950
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"learning_rate": 1.5064460924130064e-05,
|
|
"loss": 1.3292,
|
|
"step": 39000
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"eval_loss": 1.289178729057312,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39000
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"learning_rate": 1.5058122583507639e-05,
|
|
"loss": 1.3084,
|
|
"step": 39050
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"eval_loss": 1.295685887336731,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39050
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"learning_rate": 1.5051784242885215e-05,
|
|
"loss": 1.3344,
|
|
"step": 39100
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"eval_loss": 1.2958979606628418,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 39100
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"learning_rate": 1.504544590226279e-05,
|
|
"loss": 1.3343,
|
|
"step": 39150
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"eval_loss": 1.294732928276062,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 39150
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"learning_rate": 1.5039107561640365e-05,
|
|
"loss": 1.3323,
|
|
"step": 39200
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"eval_loss": 1.2921279668807983,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39200
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"learning_rate": 1.5032769221017938e-05,
|
|
"loss": 1.3303,
|
|
"step": 39250
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"eval_loss": 1.2920302152633667,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39250
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"learning_rate": 1.5026430880395513e-05,
|
|
"loss": 1.3247,
|
|
"step": 39300
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"eval_loss": 1.294016718864441,
|
|
"eval_runtime": 3.7672,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 39300
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"learning_rate": 1.5020092539773088e-05,
|
|
"loss": 1.3445,
|
|
"step": 39350
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"eval_loss": 1.2947169542312622,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39350
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"learning_rate": 1.5013754199150662e-05,
|
|
"loss": 1.3459,
|
|
"step": 39400
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"eval_loss": 1.2927649021148682,
|
|
"eval_runtime": 3.7752,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39400
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"learning_rate": 1.5007415858528239e-05,
|
|
"loss": 1.3359,
|
|
"step": 39450
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"eval_loss": 1.2903590202331543,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39450
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"learning_rate": 1.5001077517905814e-05,
|
|
"loss": 1.341,
|
|
"step": 39500
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"eval_loss": 1.2914254665374756,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39500
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"learning_rate": 1.4994739177283388e-05,
|
|
"loss": 1.3255,
|
|
"step": 39550
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"eval_loss": 1.2883648872375488,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 39550
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"learning_rate": 1.4988400836660963e-05,
|
|
"loss": 1.3375,
|
|
"step": 39600
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"eval_loss": 1.2902703285217285,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39600
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"learning_rate": 1.4982062496038538e-05,
|
|
"loss": 1.3311,
|
|
"step": 39650
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"eval_loss": 1.2961833477020264,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39650
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"learning_rate": 1.4975724155416114e-05,
|
|
"loss": 1.3141,
|
|
"step": 39700
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"eval_loss": 1.2969005107879639,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 39700
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"learning_rate": 1.4969385814793689e-05,
|
|
"loss": 1.3349,
|
|
"step": 39750
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"eval_loss": 1.2934705018997192,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39750
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"learning_rate": 1.4963047474171264e-05,
|
|
"loss": 1.3158,
|
|
"step": 39800
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"eval_loss": 1.2936718463897705,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39800
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.4956709133548839e-05,
|
|
"loss": 1.3353,
|
|
"step": 39850
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"eval_loss": 1.286584734916687,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39850
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.4950370792926413e-05,
|
|
"loss": 1.3404,
|
|
"step": 39900
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"eval_loss": 1.2973397970199585,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39900
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.4944032452303987e-05,
|
|
"loss": 1.3502,
|
|
"step": 39950
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"eval_loss": 1.2906968593597412,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 39950
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.4937694111681561e-05,
|
|
"loss": 1.3355,
|
|
"step": 40000
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"eval_loss": 1.2949856519699097,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40000
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"learning_rate": 1.4931355771059138e-05,
|
|
"loss": 1.3013,
|
|
"step": 40050
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"eval_loss": 1.2885959148406982,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 40050
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"learning_rate": 1.4925017430436713e-05,
|
|
"loss": 1.3445,
|
|
"step": 40100
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"eval_loss": 1.2899366617202759,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 40100
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"learning_rate": 1.4918679089814287e-05,
|
|
"loss": 1.3411,
|
|
"step": 40150
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"eval_loss": 1.2902860641479492,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40150
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"learning_rate": 1.4912340749191862e-05,
|
|
"loss": 1.3266,
|
|
"step": 40200
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"eval_loss": 1.2935954332351685,
|
|
"eval_runtime": 3.7778,
|
|
"eval_samples_per_second": 13.235,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 40200
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"learning_rate": 1.4906002408569437e-05,
|
|
"loss": 1.3181,
|
|
"step": 40250
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"eval_loss": 1.289138674736023,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40250
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"learning_rate": 1.4899664067947013e-05,
|
|
"loss": 1.3278,
|
|
"step": 40300
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"eval_loss": 1.2971075773239136,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 40300
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"learning_rate": 1.4893325727324588e-05,
|
|
"loss": 1.3257,
|
|
"step": 40350
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"eval_loss": 1.2971751689910889,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 40350
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"learning_rate": 1.4886987386702163e-05,
|
|
"loss": 1.3405,
|
|
"step": 40400
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"eval_loss": 1.291967511177063,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40400
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"learning_rate": 1.4880649046079738e-05,
|
|
"loss": 1.3208,
|
|
"step": 40450
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"eval_loss": 1.2884658575057983,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40450
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"learning_rate": 1.4874310705457312e-05,
|
|
"loss": 1.3416,
|
|
"step": 40500
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"eval_loss": 1.290800929069519,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40500
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"learning_rate": 1.4867972364834889e-05,
|
|
"loss": 1.3449,
|
|
"step": 40550
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"eval_loss": 1.290928840637207,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40550
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"learning_rate": 1.4861634024212464e-05,
|
|
"loss": 1.3242,
|
|
"step": 40600
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"eval_loss": 1.2922505140304565,
|
|
"eval_runtime": 3.776,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40600
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"learning_rate": 1.4855295683590037e-05,
|
|
"loss": 1.3287,
|
|
"step": 40650
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"eval_loss": 1.2913994789123535,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40650
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"learning_rate": 1.4848957342967612e-05,
|
|
"loss": 1.3415,
|
|
"step": 40700
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"eval_loss": 1.2920857667922974,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40700
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"learning_rate": 1.4842619002345186e-05,
|
|
"loss": 1.33,
|
|
"step": 40750
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"eval_loss": 1.2984005212783813,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40750
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"learning_rate": 1.4836280661722761e-05,
|
|
"loss": 1.3453,
|
|
"step": 40800
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"eval_loss": 1.295319676399231,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40800
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"learning_rate": 1.4829942321100338e-05,
|
|
"loss": 1.3299,
|
|
"step": 40850
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"eval_loss": 1.2999267578125,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40850
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"learning_rate": 1.4823603980477912e-05,
|
|
"loss": 1.3207,
|
|
"step": 40900
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"eval_loss": 1.2958366870880127,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40900
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"learning_rate": 1.4817265639855487e-05,
|
|
"loss": 1.3209,
|
|
"step": 40950
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"eval_loss": 1.2899904251098633,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 40950
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"learning_rate": 1.4810927299233062e-05,
|
|
"loss": 1.3285,
|
|
"step": 41000
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"eval_loss": 1.2934271097183228,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41000
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"learning_rate": 1.4804588958610637e-05,
|
|
"loss": 1.3127,
|
|
"step": 41050
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"eval_loss": 1.2928721904754639,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41050
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"learning_rate": 1.4798250617988213e-05,
|
|
"loss": 1.345,
|
|
"step": 41100
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"eval_loss": 1.2898428440093994,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41100
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"learning_rate": 1.4791912277365788e-05,
|
|
"loss": 1.3217,
|
|
"step": 41150
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"eval_loss": 1.2914966344833374,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41150
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"learning_rate": 1.4785573936743363e-05,
|
|
"loss": 1.3429,
|
|
"step": 41200
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"eval_loss": 1.2993860244750977,
|
|
"eval_runtime": 3.7756,
|
|
"eval_samples_per_second": 13.243,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41200
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"learning_rate": 1.4779235596120938e-05,
|
|
"loss": 1.3321,
|
|
"step": 41250
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"eval_loss": 1.2899794578552246,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41250
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"learning_rate": 1.477289725549851e-05,
|
|
"loss": 1.3551,
|
|
"step": 41300
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"eval_loss": 1.2923322916030884,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41300
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"learning_rate": 1.4766558914876085e-05,
|
|
"loss": 1.3608,
|
|
"step": 41350
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"eval_loss": 1.2880319356918335,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41350
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"learning_rate": 1.476022057425366e-05,
|
|
"loss": 1.3499,
|
|
"step": 41400
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"eval_loss": 1.2882798910140991,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41400
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"learning_rate": 1.4753882233631237e-05,
|
|
"loss": 1.3259,
|
|
"step": 41450
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"eval_loss": 1.2902902364730835,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41450
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"learning_rate": 1.4747543893008811e-05,
|
|
"loss": 1.3458,
|
|
"step": 41500
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"eval_loss": 1.2899441719055176,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41500
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"learning_rate": 1.4741205552386386e-05,
|
|
"loss": 1.349,
|
|
"step": 41550
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"eval_loss": 1.289746880531311,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41550
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"learning_rate": 1.4734867211763961e-05,
|
|
"loss": 1.3276,
|
|
"step": 41600
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"eval_loss": 1.288796067237854,
|
|
"eval_runtime": 3.7657,
|
|
"eval_samples_per_second": 13.278,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41600
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"learning_rate": 1.4728528871141536e-05,
|
|
"loss": 1.3398,
|
|
"step": 41650
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"eval_loss": 1.2927296161651611,
|
|
"eval_runtime": 3.7781,
|
|
"eval_samples_per_second": 13.234,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 41650
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"learning_rate": 1.4722190530519112e-05,
|
|
"loss": 1.333,
|
|
"step": 41700
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"eval_loss": 1.287598729133606,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41700
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"learning_rate": 1.4715852189896687e-05,
|
|
"loss": 1.3196,
|
|
"step": 41750
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"eval_loss": 1.2902435064315796,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41750
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"learning_rate": 1.4709513849274262e-05,
|
|
"loss": 1.3139,
|
|
"step": 41800
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"eval_loss": 1.2944111824035645,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41800
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"learning_rate": 1.4703175508651837e-05,
|
|
"loss": 1.3405,
|
|
"step": 41850
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"eval_loss": 1.2928820848464966,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41850
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"learning_rate": 1.4696837168029411e-05,
|
|
"loss": 1.3107,
|
|
"step": 41900
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"eval_loss": 1.2899792194366455,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 41900
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"learning_rate": 1.4690498827406988e-05,
|
|
"loss": 1.3242,
|
|
"step": 41950
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"eval_loss": 1.2877752780914307,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 41950
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"learning_rate": 1.468428725359701e-05,
|
|
"loss": 1.3353,
|
|
"step": 42000
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"eval_loss": 1.2853622436523438,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42000
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"learning_rate": 1.4677948912974584e-05,
|
|
"loss": 1.3475,
|
|
"step": 42050
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"eval_loss": 1.2884793281555176,
|
|
"eval_runtime": 3.7767,
|
|
"eval_samples_per_second": 13.239,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42050
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"learning_rate": 1.4671610572352159e-05,
|
|
"loss": 1.3459,
|
|
"step": 42100
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"eval_loss": 1.2918078899383545,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42100
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"learning_rate": 1.4665272231729734e-05,
|
|
"loss": 1.3342,
|
|
"step": 42150
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"eval_loss": 1.287268877029419,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 42150
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"learning_rate": 1.465893389110731e-05,
|
|
"loss": 1.3283,
|
|
"step": 42200
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"eval_loss": 1.2882168292999268,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42200
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"learning_rate": 1.4652595550484885e-05,
|
|
"loss": 1.3169,
|
|
"step": 42250
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"eval_loss": 1.2856416702270508,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 42250
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"learning_rate": 1.464625720986246e-05,
|
|
"loss": 1.3326,
|
|
"step": 42300
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"eval_loss": 1.2885664701461792,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42300
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"learning_rate": 1.4639918869240035e-05,
|
|
"loss": 1.3238,
|
|
"step": 42350
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"eval_loss": 1.2899413108825684,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 42350
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"learning_rate": 1.4633580528617608e-05,
|
|
"loss": 1.321,
|
|
"step": 42400
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"eval_loss": 1.2910962104797363,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 42400
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"learning_rate": 1.4627242187995183e-05,
|
|
"loss": 1.3181,
|
|
"step": 42450
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"eval_loss": 1.28545343875885,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42450
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"learning_rate": 1.4620903847372759e-05,
|
|
"loss": 1.3195,
|
|
"step": 42500
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"eval_loss": 1.2867140769958496,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42500
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"learning_rate": 1.4614565506750334e-05,
|
|
"loss": 1.3374,
|
|
"step": 42550
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"eval_loss": 1.2835534811019897,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42550
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"learning_rate": 1.4608227166127909e-05,
|
|
"loss": 1.3474,
|
|
"step": 42600
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"eval_loss": 1.287997841835022,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42600
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"learning_rate": 1.4601888825505483e-05,
|
|
"loss": 1.3424,
|
|
"step": 42650
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"eval_loss": 1.2894395589828491,
|
|
"eval_runtime": 3.776,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42650
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"learning_rate": 1.4595550484883058e-05,
|
|
"loss": 1.3353,
|
|
"step": 42700
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"eval_loss": 1.286597728729248,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42700
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"learning_rate": 1.4589212144260635e-05,
|
|
"loss": 1.3324,
|
|
"step": 42750
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"eval_loss": 1.287002682685852,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42750
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"learning_rate": 1.458287380363821e-05,
|
|
"loss": 1.3353,
|
|
"step": 42800
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"eval_loss": 1.2935638427734375,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 42800
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"learning_rate": 1.4576535463015784e-05,
|
|
"loss": 1.3255,
|
|
"step": 42850
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"eval_loss": 1.2929604053497314,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42850
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"learning_rate": 1.4570197122393359e-05,
|
|
"loss": 1.3363,
|
|
"step": 42900
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"eval_loss": 1.2887177467346191,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42900
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"learning_rate": 1.4563858781770934e-05,
|
|
"loss": 1.3123,
|
|
"step": 42950
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"eval_loss": 1.291319727897644,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 42950
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"learning_rate": 1.455752044114851e-05,
|
|
"loss": 1.3097,
|
|
"step": 43000
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"eval_loss": 1.2872183322906494,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43000
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"learning_rate": 1.4551182100526085e-05,
|
|
"loss": 1.3369,
|
|
"step": 43050
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"eval_loss": 1.292192816734314,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43050
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"learning_rate": 1.4544843759903658e-05,
|
|
"loss": 1.3255,
|
|
"step": 43100
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"eval_loss": 1.2942190170288086,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43100
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"learning_rate": 1.4538505419281233e-05,
|
|
"loss": 1.2986,
|
|
"step": 43150
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"eval_loss": 1.2886135578155518,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 43150
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"learning_rate": 1.4532167078658808e-05,
|
|
"loss": 1.3091,
|
|
"step": 43200
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"eval_loss": 1.2886717319488525,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43200
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"learning_rate": 1.4525828738036382e-05,
|
|
"loss": 1.2946,
|
|
"step": 43250
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"eval_loss": 1.2904119491577148,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43250
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"learning_rate": 1.4519490397413957e-05,
|
|
"loss": 1.3179,
|
|
"step": 43300
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"eval_loss": 1.2890146970748901,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43300
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"learning_rate": 1.4513152056791534e-05,
|
|
"loss": 1.3137,
|
|
"step": 43350
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"eval_loss": 1.2953622341156006,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43350
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"learning_rate": 1.4506813716169108e-05,
|
|
"loss": 1.3293,
|
|
"step": 43400
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"eval_loss": 1.2892794609069824,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 43400
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"learning_rate": 1.4500475375546683e-05,
|
|
"loss": 1.3277,
|
|
"step": 43450
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"eval_loss": 1.290398120880127,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43450
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"learning_rate": 1.4494137034924258e-05,
|
|
"loss": 1.3341,
|
|
"step": 43500
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"eval_loss": 1.2935774326324463,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43500
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"learning_rate": 1.4487798694301833e-05,
|
|
"loss": 1.3724,
|
|
"step": 43550
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"eval_loss": 1.2943332195281982,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43550
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"learning_rate": 1.448146035367941e-05,
|
|
"loss": 1.334,
|
|
"step": 43600
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"eval_loss": 1.290490984916687,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43600
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"learning_rate": 1.4475122013056984e-05,
|
|
"loss": 1.3096,
|
|
"step": 43650
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"eval_loss": 1.2859091758728027,
|
|
"eval_runtime": 3.7673,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 43650
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"learning_rate": 1.4468783672434559e-05,
|
|
"loss": 1.2823,
|
|
"step": 43700
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"eval_loss": 1.294083833694458,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 43700
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"learning_rate": 1.4462445331812134e-05,
|
|
"loss": 1.3131,
|
|
"step": 43750
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"eval_loss": 1.2863880395889282,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43750
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 1.4456106991189707e-05,
|
|
"loss": 1.2996,
|
|
"step": 43800
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"eval_loss": 1.2920746803283691,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 43800
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 1.4449768650567281e-05,
|
|
"loss": 1.3179,
|
|
"step": 43850
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"eval_loss": 1.2915823459625244,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43850
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 1.4443430309944856e-05,
|
|
"loss": 1.3186,
|
|
"step": 43900
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"eval_loss": 1.2937850952148438,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43900
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 1.4437091969322433e-05,
|
|
"loss": 1.3089,
|
|
"step": 43950
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"eval_loss": 1.29170560836792,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 43950
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"learning_rate": 1.4430753628700007e-05,
|
|
"loss": 1.3126,
|
|
"step": 44000
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"eval_loss": 1.2858994007110596,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 44000
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"learning_rate": 1.4424415288077582e-05,
|
|
"loss": 1.357,
|
|
"step": 44050
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"eval_loss": 1.2939144372940063,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44050
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"learning_rate": 1.4418076947455157e-05,
|
|
"loss": 1.3093,
|
|
"step": 44100
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"eval_loss": 1.2964445352554321,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44100
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"learning_rate": 1.4411865373645182e-05,
|
|
"loss": 1.3134,
|
|
"step": 44150
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"eval_loss": 1.2847365140914917,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44150
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"learning_rate": 1.4405527033022755e-05,
|
|
"loss": 1.3182,
|
|
"step": 44200
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"eval_loss": 1.2918041944503784,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44200
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"learning_rate": 1.439918869240033e-05,
|
|
"loss": 1.33,
|
|
"step": 44250
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"eval_loss": 1.2843211889266968,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44250
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"learning_rate": 1.4392850351777905e-05,
|
|
"loss": 1.321,
|
|
"step": 44300
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"eval_loss": 1.2892286777496338,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44300
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"learning_rate": 1.438651201115548e-05,
|
|
"loss": 1.3215,
|
|
"step": 44350
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"eval_loss": 1.2926931381225586,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44350
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"learning_rate": 1.4380173670533056e-05,
|
|
"loss": 1.3344,
|
|
"step": 44400
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"eval_loss": 1.290739893913269,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 44400
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"learning_rate": 1.437383532991063e-05,
|
|
"loss": 1.3236,
|
|
"step": 44450
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"eval_loss": 1.2874236106872559,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44450
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"learning_rate": 1.4367496989288206e-05,
|
|
"loss": 1.3352,
|
|
"step": 44500
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"eval_loss": 1.2880526781082153,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44500
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"learning_rate": 1.436115864866578e-05,
|
|
"loss": 1.3263,
|
|
"step": 44550
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"eval_loss": 1.2898814678192139,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44550
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"learning_rate": 1.4354820308043355e-05,
|
|
"loss": 1.328,
|
|
"step": 44600
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"eval_loss": 1.2884628772735596,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44600
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"learning_rate": 1.4348481967420932e-05,
|
|
"loss": 1.3329,
|
|
"step": 44650
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"eval_loss": 1.2876232862472534,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44650
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"learning_rate": 1.4342143626798506e-05,
|
|
"loss": 1.3266,
|
|
"step": 44700
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"eval_loss": 1.2916231155395508,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44700
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"learning_rate": 1.4335805286176081e-05,
|
|
"loss": 1.3366,
|
|
"step": 44750
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"eval_loss": 1.2916532754898071,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44750
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"learning_rate": 1.4329466945553656e-05,
|
|
"loss": 1.329,
|
|
"step": 44800
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"eval_loss": 1.2999464273452759,
|
|
"eval_runtime": 3.7779,
|
|
"eval_samples_per_second": 13.235,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 44800
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"learning_rate": 1.432312860493123e-05,
|
|
"loss": 1.3053,
|
|
"step": 44850
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"eval_loss": 1.2890410423278809,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44850
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.4316790264308804e-05,
|
|
"loss": 1.3248,
|
|
"step": 44900
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"eval_loss": 1.2961732149124146,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44900
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.4310451923686379e-05,
|
|
"loss": 1.353,
|
|
"step": 44950
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"eval_loss": 1.2963011264801025,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 44950
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.4304113583063955e-05,
|
|
"loss": 1.3084,
|
|
"step": 45000
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"eval_loss": 1.291407585144043,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45000
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.429777524244153e-05,
|
|
"loss": 1.3177,
|
|
"step": 45050
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"eval_loss": 1.2891829013824463,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45050
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"learning_rate": 1.4291436901819105e-05,
|
|
"loss": 1.321,
|
|
"step": 45100
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"eval_loss": 1.2884130477905273,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 45100
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"learning_rate": 1.428509856119668e-05,
|
|
"loss": 1.3395,
|
|
"step": 45150
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"eval_loss": 1.2875961065292358,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45150
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"learning_rate": 1.4278760220574254e-05,
|
|
"loss": 1.3071,
|
|
"step": 45200
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"eval_loss": 1.2922155857086182,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45200
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"learning_rate": 1.427242187995183e-05,
|
|
"loss": 1.3075,
|
|
"step": 45250
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"eval_loss": 1.2871181964874268,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45250
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"learning_rate": 1.4266083539329405e-05,
|
|
"loss": 1.3303,
|
|
"step": 45300
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"eval_loss": 1.291816234588623,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 45300
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"learning_rate": 1.425974519870698e-05,
|
|
"loss": 1.3247,
|
|
"step": 45350
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"eval_loss": 1.2857388257980347,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 45350
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"learning_rate": 1.4253406858084555e-05,
|
|
"loss": 1.3453,
|
|
"step": 45400
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"eval_loss": 1.2904013395309448,
|
|
"eval_runtime": 3.7671,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 45400
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"learning_rate": 1.424706851746213e-05,
|
|
"loss": 1.3394,
|
|
"step": 45450
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"eval_loss": 1.29264235496521,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45450
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"learning_rate": 1.4240730176839706e-05,
|
|
"loss": 1.33,
|
|
"step": 45500
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"eval_loss": 1.2884577512741089,
|
|
"eval_runtime": 3.7753,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45500
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"learning_rate": 1.4234518603029728e-05,
|
|
"loss": 1.3274,
|
|
"step": 45550
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"eval_loss": 1.2859132289886475,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45550
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"learning_rate": 1.4228180262407303e-05,
|
|
"loss": 1.3072,
|
|
"step": 45600
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"eval_loss": 1.289865255355835,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45600
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"learning_rate": 1.4221841921784878e-05,
|
|
"loss": 1.3212,
|
|
"step": 45650
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"eval_loss": 1.2887279987335205,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45650
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"learning_rate": 1.4215503581162452e-05,
|
|
"loss": 1.3089,
|
|
"step": 45700
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"eval_loss": 1.2890145778656006,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45700
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"learning_rate": 1.4209165240540029e-05,
|
|
"loss": 1.3301,
|
|
"step": 45750
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"eval_loss": 1.2922050952911377,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45750
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"learning_rate": 1.4202826899917604e-05,
|
|
"loss": 1.3304,
|
|
"step": 45800
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"eval_loss": 1.2859066724777222,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45800
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"learning_rate": 1.4196488559295178e-05,
|
|
"loss": 1.3038,
|
|
"step": 45850
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"eval_loss": 1.2912070751190186,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45850
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"learning_rate": 1.4190150218672753e-05,
|
|
"loss": 1.3349,
|
|
"step": 45900
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"eval_loss": 1.2870663404464722,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45900
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"learning_rate": 1.4183811878050328e-05,
|
|
"loss": 1.3371,
|
|
"step": 45950
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"eval_loss": 1.2932062149047852,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 45950
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"learning_rate": 1.4177473537427901e-05,
|
|
"loss": 1.3442,
|
|
"step": 46000
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"eval_loss": 1.2901263236999512,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46000
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"learning_rate": 1.4171135196805477e-05,
|
|
"loss": 1.3105,
|
|
"step": 46050
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"eval_loss": 1.2861688137054443,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 46050
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"learning_rate": 1.4164796856183052e-05,
|
|
"loss": 1.3002,
|
|
"step": 46100
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"eval_loss": 1.2894814014434814,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46100
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"learning_rate": 1.4158458515560627e-05,
|
|
"loss": 1.3145,
|
|
"step": 46150
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"eval_loss": 1.2939649820327759,
|
|
"eval_runtime": 3.7753,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46150
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"learning_rate": 1.4152120174938202e-05,
|
|
"loss": 1.2912,
|
|
"step": 46200
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"eval_loss": 1.2945733070373535,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46200
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"learning_rate": 1.4145781834315777e-05,
|
|
"loss": 1.3115,
|
|
"step": 46250
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"eval_loss": 1.2894076108932495,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46250
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"learning_rate": 1.4139443493693353e-05,
|
|
"loss": 1.3267,
|
|
"step": 46300
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"eval_loss": 1.2906025648117065,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 46300
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"learning_rate": 1.4133105153070928e-05,
|
|
"loss": 1.314,
|
|
"step": 46350
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"eval_loss": 1.2906670570373535,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46350
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"learning_rate": 1.4126766812448503e-05,
|
|
"loss": 1.2977,
|
|
"step": 46400
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"eval_loss": 1.2908565998077393,
|
|
"eval_runtime": 3.7722,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46400
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"learning_rate": 1.4120428471826077e-05,
|
|
"loss": 1.351,
|
|
"step": 46450
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"eval_loss": 1.2942793369293213,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 46450
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"learning_rate": 1.4114090131203652e-05,
|
|
"loss": 1.3293,
|
|
"step": 46500
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"eval_loss": 1.286929726600647,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46500
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"learning_rate": 1.4107751790581229e-05,
|
|
"loss": 1.2885,
|
|
"step": 46550
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"eval_loss": 1.2930958271026611,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46550
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"learning_rate": 1.4101413449958803e-05,
|
|
"loss": 1.314,
|
|
"step": 46600
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"eval_loss": 1.291534423828125,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46600
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"learning_rate": 1.4095075109336376e-05,
|
|
"loss": 1.3247,
|
|
"step": 46650
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"eval_loss": 1.2937606573104858,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46650
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"learning_rate": 1.4088736768713951e-05,
|
|
"loss": 1.3298,
|
|
"step": 46700
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"eval_loss": 1.2894597053527832,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46700
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"learning_rate": 1.4082398428091526e-05,
|
|
"loss": 1.3324,
|
|
"step": 46750
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"eval_loss": 1.289685845375061,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 46750
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"learning_rate": 1.40760600874691e-05,
|
|
"loss": 1.3248,
|
|
"step": 46800
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"eval_loss": 1.289353370666504,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46800
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"learning_rate": 1.4069721746846676e-05,
|
|
"loss": 1.3391,
|
|
"step": 46850
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"eval_loss": 1.2854732275009155,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46850
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"learning_rate": 1.4063383406224252e-05,
|
|
"loss": 1.3459,
|
|
"step": 46900
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"eval_loss": 1.2839033603668213,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46900
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"learning_rate": 1.4057045065601827e-05,
|
|
"loss": 1.3335,
|
|
"step": 46950
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"eval_loss": 1.2852214574813843,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 46950
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"learning_rate": 1.4050706724979402e-05,
|
|
"loss": 1.315,
|
|
"step": 47000
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"eval_loss": 1.2876884937286377,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47000
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"learning_rate": 1.4044368384356976e-05,
|
|
"loss": 1.3382,
|
|
"step": 47050
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"eval_loss": 1.2895230054855347,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47050
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"learning_rate": 1.4038030043734551e-05,
|
|
"loss": 1.3233,
|
|
"step": 47100
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"eval_loss": 1.289626121520996,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47100
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"learning_rate": 1.4031691703112128e-05,
|
|
"loss": 1.3561,
|
|
"step": 47150
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"eval_loss": 1.2865580320358276,
|
|
"eval_runtime": 3.7736,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47150
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"learning_rate": 1.4025353362489702e-05,
|
|
"loss": 1.3269,
|
|
"step": 47200
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"eval_loss": 1.2933601140975952,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47200
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"learning_rate": 1.4019015021867277e-05,
|
|
"loss": 1.3314,
|
|
"step": 47250
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"eval_loss": 1.2862509489059448,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47250
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"learning_rate": 1.4012676681244852e-05,
|
|
"loss": 1.3201,
|
|
"step": 47300
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_loss": 1.2914092540740967,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47300
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"learning_rate": 1.4006338340622425e-05,
|
|
"loss": 1.3382,
|
|
"step": 47350
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_loss": 1.289807915687561,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47350
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"learning_rate": 1.4e-05,
|
|
"loss": 1.2882,
|
|
"step": 47400
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_loss": 1.2870560884475708,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47400
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"learning_rate": 1.3993661659377575e-05,
|
|
"loss": 1.3226,
|
|
"step": 47450
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"eval_loss": 1.2883172035217285,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47450
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"learning_rate": 1.3987323318755151e-05,
|
|
"loss": 1.2863,
|
|
"step": 47500
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"eval_loss": 1.2840888500213623,
|
|
"eval_runtime": 3.7744,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47500
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"learning_rate": 1.3980984978132726e-05,
|
|
"loss": 1.2923,
|
|
"step": 47550
|
|
},
|
|
{
|
|
"epoch": 3.01,
|
|
"eval_loss": 1.2925286293029785,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47550
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"learning_rate": 1.39746466375103e-05,
|
|
"loss": 1.3157,
|
|
"step": 47600
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"eval_loss": 1.285499095916748,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47600
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"learning_rate": 1.3968308296887875e-05,
|
|
"loss": 1.3056,
|
|
"step": 47650
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"eval_loss": 1.2872999906539917,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47650
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"learning_rate": 1.396196995626545e-05,
|
|
"loss": 1.3286,
|
|
"step": 47700
|
|
},
|
|
{
|
|
"epoch": 3.02,
|
|
"eval_loss": 1.285345196723938,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47700
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"learning_rate": 1.3955631615643027e-05,
|
|
"loss": 1.3033,
|
|
"step": 47750
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"eval_loss": 1.289099097251892,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47750
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"learning_rate": 1.3949293275020601e-05,
|
|
"loss": 1.2941,
|
|
"step": 47800
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"eval_loss": 1.2864618301391602,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47800
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"learning_rate": 1.3942954934398176e-05,
|
|
"loss": 1.3192,
|
|
"step": 47850
|
|
},
|
|
{
|
|
"epoch": 3.03,
|
|
"eval_loss": 1.2831422090530396,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47850
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"learning_rate": 1.3936616593775751e-05,
|
|
"loss": 1.3161,
|
|
"step": 47900
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"eval_loss": 1.2853567600250244,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 47900
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"learning_rate": 1.3930278253153326e-05,
|
|
"loss": 1.3111,
|
|
"step": 47950
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"eval_loss": 1.2888298034667969,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 47950
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"learning_rate": 1.3923939912530902e-05,
|
|
"loss": 1.3247,
|
|
"step": 48000
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"eval_loss": 1.2861146926879883,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48000
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"learning_rate": 1.3917601571908475e-05,
|
|
"loss": 1.3101,
|
|
"step": 48050
|
|
},
|
|
{
|
|
"epoch": 3.04,
|
|
"eval_loss": 1.2856858968734741,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48050
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"learning_rate": 1.391126323128605e-05,
|
|
"loss": 1.3195,
|
|
"step": 48100
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"eval_loss": 1.2888274192810059,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48100
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"learning_rate": 1.3904924890663625e-05,
|
|
"loss": 1.3156,
|
|
"step": 48150
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"eval_loss": 1.2859907150268555,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 48150
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"learning_rate": 1.38985865500412e-05,
|
|
"loss": 1.3043,
|
|
"step": 48200
|
|
},
|
|
{
|
|
"epoch": 3.05,
|
|
"eval_loss": 1.2805801630020142,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48200
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"learning_rate": 1.3892248209418774e-05,
|
|
"loss": 1.3085,
|
|
"step": 48250
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"eval_loss": 1.2856419086456299,
|
|
"eval_runtime": 3.7758,
|
|
"eval_samples_per_second": 13.242,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48250
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"learning_rate": 1.3885909868796351e-05,
|
|
"loss": 1.2988,
|
|
"step": 48300
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"eval_loss": 1.2814656496047974,
|
|
"eval_runtime": 3.7713,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48300
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"learning_rate": 1.3879571528173926e-05,
|
|
"loss": 1.3086,
|
|
"step": 48350
|
|
},
|
|
{
|
|
"epoch": 3.06,
|
|
"eval_loss": 1.284053921699524,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48350
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"learning_rate": 1.38732331875515e-05,
|
|
"loss": 1.283,
|
|
"step": 48400
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"eval_loss": 1.289674997329712,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48400
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"learning_rate": 1.3866894846929075e-05,
|
|
"loss": 1.3117,
|
|
"step": 48450
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"eval_loss": 1.2874929904937744,
|
|
"eval_runtime": 3.775,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48450
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"learning_rate": 1.386055650630665e-05,
|
|
"loss": 1.2987,
|
|
"step": 48500
|
|
},
|
|
{
|
|
"epoch": 3.07,
|
|
"eval_loss": 1.285253643989563,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48500
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"learning_rate": 1.3854218165684226e-05,
|
|
"loss": 1.2914,
|
|
"step": 48550
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"eval_loss": 1.289374828338623,
|
|
"eval_runtime": 3.7778,
|
|
"eval_samples_per_second": 13.235,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 48550
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"learning_rate": 1.3847879825061801e-05,
|
|
"loss": 1.3223,
|
|
"step": 48600
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"eval_loss": 1.2871770858764648,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48600
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"learning_rate": 1.3841541484439376e-05,
|
|
"loss": 1.2799,
|
|
"step": 48650
|
|
},
|
|
{
|
|
"epoch": 3.08,
|
|
"eval_loss": 1.2840865850448608,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 48650
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"learning_rate": 1.383520314381695e-05,
|
|
"loss": 1.304,
|
|
"step": 48700
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"eval_loss": 1.2872166633605957,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48700
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"learning_rate": 1.3828864803194524e-05,
|
|
"loss": 1.3254,
|
|
"step": 48750
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"eval_loss": 1.2906036376953125,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48750
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"learning_rate": 1.3822526462572099e-05,
|
|
"loss": 1.3191,
|
|
"step": 48800
|
|
},
|
|
{
|
|
"epoch": 3.09,
|
|
"eval_loss": 1.2823243141174316,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 48800
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"learning_rate": 1.3816188121949673e-05,
|
|
"loss": 1.3171,
|
|
"step": 48850
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"eval_loss": 1.2939705848693848,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 48850
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"learning_rate": 1.380984978132725e-05,
|
|
"loss": 1.3028,
|
|
"step": 48900
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"eval_loss": 1.2873985767364502,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48900
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"learning_rate": 1.3803511440704825e-05,
|
|
"loss": 1.3287,
|
|
"step": 48950
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"eval_loss": 1.2868144512176514,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 48950
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"learning_rate": 1.37971731000824e-05,
|
|
"loss": 1.3121,
|
|
"step": 49000
|
|
},
|
|
{
|
|
"epoch": 3.1,
|
|
"eval_loss": 1.2848081588745117,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49000
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"learning_rate": 1.3790961526272425e-05,
|
|
"loss": 1.291,
|
|
"step": 49050
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"eval_loss": 1.2847713232040405,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49050
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"learning_rate": 1.378462318565e-05,
|
|
"loss": 1.3218,
|
|
"step": 49100
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"eval_loss": 1.287498950958252,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49100
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"learning_rate": 1.3778284845027572e-05,
|
|
"loss": 1.2983,
|
|
"step": 49150
|
|
},
|
|
{
|
|
"epoch": 3.11,
|
|
"eval_loss": 1.2906882762908936,
|
|
"eval_runtime": 3.777,
|
|
"eval_samples_per_second": 13.238,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49150
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"learning_rate": 1.3771946504405147e-05,
|
|
"loss": 1.3104,
|
|
"step": 49200
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"eval_loss": 1.2896733283996582,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49200
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"learning_rate": 1.3765608163782722e-05,
|
|
"loss": 1.3249,
|
|
"step": 49250
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"eval_loss": 1.2836467027664185,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49250
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"learning_rate": 1.3759269823160297e-05,
|
|
"loss": 1.3095,
|
|
"step": 49300
|
|
},
|
|
{
|
|
"epoch": 3.12,
|
|
"eval_loss": 1.290600299835205,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49300
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"learning_rate": 1.3752931482537872e-05,
|
|
"loss": 1.3063,
|
|
"step": 49350
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"eval_loss": 1.2883328199386597,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49350
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"learning_rate": 1.3746593141915448e-05,
|
|
"loss": 1.3388,
|
|
"step": 49400
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"eval_loss": 1.2841116189956665,
|
|
"eval_runtime": 3.7745,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49400
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"learning_rate": 1.3740254801293023e-05,
|
|
"loss": 1.3168,
|
|
"step": 49450
|
|
},
|
|
{
|
|
"epoch": 3.13,
|
|
"eval_loss": 1.2897932529449463,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49450
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"learning_rate": 1.3733916460670598e-05,
|
|
"loss": 1.3163,
|
|
"step": 49500
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"eval_loss": 1.2857288122177124,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49500
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"learning_rate": 1.3727578120048172e-05,
|
|
"loss": 1.2943,
|
|
"step": 49550
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"eval_loss": 1.2849351167678833,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49550
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"learning_rate": 1.3721239779425747e-05,
|
|
"loss": 1.3069,
|
|
"step": 49600
|
|
},
|
|
{
|
|
"epoch": 3.14,
|
|
"eval_loss": 1.2908416986465454,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49600
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"learning_rate": 1.3714901438803324e-05,
|
|
"loss": 1.3176,
|
|
"step": 49650
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"eval_loss": 1.288088083267212,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49650
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"learning_rate": 1.3708563098180898e-05,
|
|
"loss": 1.2928,
|
|
"step": 49700
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"eval_loss": 1.2876183986663818,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49700
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"learning_rate": 1.3702224757558473e-05,
|
|
"loss": 1.2711,
|
|
"step": 49750
|
|
},
|
|
{
|
|
"epoch": 3.15,
|
|
"eval_loss": 1.290259599685669,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49750
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"learning_rate": 1.3695886416936048e-05,
|
|
"loss": 1.2938,
|
|
"step": 49800
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"eval_loss": 1.2888896465301514,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49800
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"learning_rate": 1.3689548076313621e-05,
|
|
"loss": 1.3214,
|
|
"step": 49850
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"eval_loss": 1.2869081497192383,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49850
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"learning_rate": 1.3683209735691196e-05,
|
|
"loss": 1.2983,
|
|
"step": 49900
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"eval_loss": 1.2879716157913208,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 49900
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"learning_rate": 1.3676871395068772e-05,
|
|
"loss": 1.3165,
|
|
"step": 49950
|
|
},
|
|
{
|
|
"epoch": 3.16,
|
|
"eval_loss": 1.2864917516708374,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 49950
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"learning_rate": 1.3670533054446347e-05,
|
|
"loss": 1.3067,
|
|
"step": 50000
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"eval_loss": 1.2828024625778198,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50000
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"learning_rate": 1.3664194713823922e-05,
|
|
"loss": 1.3017,
|
|
"step": 50050
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"eval_loss": 1.2888184785842896,
|
|
"eval_runtime": 3.7657,
|
|
"eval_samples_per_second": 13.278,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50050
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"learning_rate": 1.3657856373201497e-05,
|
|
"loss": 1.3136,
|
|
"step": 50100
|
|
},
|
|
{
|
|
"epoch": 3.17,
|
|
"eval_loss": 1.2851626873016357,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50100
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"learning_rate": 1.3651518032579071e-05,
|
|
"loss": 1.3045,
|
|
"step": 50150
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"eval_loss": 1.2868449687957764,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50150
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"learning_rate": 1.3645179691956648e-05,
|
|
"loss": 1.3158,
|
|
"step": 50200
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"eval_loss": 1.2862974405288696,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50200
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"learning_rate": 1.3638841351334223e-05,
|
|
"loss": 1.3151,
|
|
"step": 50250
|
|
},
|
|
{
|
|
"epoch": 3.18,
|
|
"eval_loss": 1.2867438793182373,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50250
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"learning_rate": 1.3632503010711797e-05,
|
|
"loss": 1.3132,
|
|
"step": 50300
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"eval_loss": 1.2874759435653687,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50300
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"learning_rate": 1.3626164670089372e-05,
|
|
"loss": 1.2794,
|
|
"step": 50350
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"eval_loss": 1.2847301959991455,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50350
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"learning_rate": 1.3619826329466947e-05,
|
|
"loss": 1.2904,
|
|
"step": 50400
|
|
},
|
|
{
|
|
"epoch": 3.19,
|
|
"eval_loss": 1.2875010967254639,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50400
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"learning_rate": 1.3613487988844523e-05,
|
|
"loss": 1.3222,
|
|
"step": 50450
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"eval_loss": 1.289933443069458,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50450
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"learning_rate": 1.3607149648222095e-05,
|
|
"loss": 1.3291,
|
|
"step": 50500
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"eval_loss": 1.287125825881958,
|
|
"eval_runtime": 3.7704,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50500
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"learning_rate": 1.3600811307599671e-05,
|
|
"loss": 1.3138,
|
|
"step": 50550
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"eval_loss": 1.2834783792495728,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50550
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"learning_rate": 1.3594472966977246e-05,
|
|
"loss": 1.3139,
|
|
"step": 50600
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"eval_loss": 1.283707857131958,
|
|
"eval_runtime": 3.7691,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50600
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"learning_rate": 1.358813462635482e-05,
|
|
"loss": 1.3014,
|
|
"step": 50650
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"eval_loss": 1.295029640197754,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50650
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"learning_rate": 1.3581796285732396e-05,
|
|
"loss": 1.3078,
|
|
"step": 50700
|
|
},
|
|
{
|
|
"epoch": 3.21,
|
|
"eval_loss": 1.2928059101104736,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50700
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"learning_rate": 1.357545794510997e-05,
|
|
"loss": 1.2939,
|
|
"step": 50750
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"eval_loss": 1.2884776592254639,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50750
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"learning_rate": 1.3569119604487547e-05,
|
|
"loss": 1.31,
|
|
"step": 50800
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"eval_loss": 1.28407621383667,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50800
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"learning_rate": 1.3562781263865122e-05,
|
|
"loss": 1.2786,
|
|
"step": 50850
|
|
},
|
|
{
|
|
"epoch": 3.22,
|
|
"eval_loss": 1.2833775281906128,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50850
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"learning_rate": 1.3556442923242696e-05,
|
|
"loss": 1.2798,
|
|
"step": 50900
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"eval_loss": 1.2914105653762817,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 50900
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"learning_rate": 1.3550104582620271e-05,
|
|
"loss": 1.3175,
|
|
"step": 50950
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"eval_loss": 1.288995623588562,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 50950
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"learning_rate": 1.3543766241997846e-05,
|
|
"loss": 1.3186,
|
|
"step": 51000
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"eval_loss": 1.2805770635604858,
|
|
"eval_runtime": 3.7674,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51000
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"learning_rate": 1.3537427901375422e-05,
|
|
"loss": 1.3422,
|
|
"step": 51050
|
|
},
|
|
{
|
|
"epoch": 3.23,
|
|
"eval_loss": 1.2885582447052002,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51050
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"learning_rate": 1.3531089560752997e-05,
|
|
"loss": 1.3137,
|
|
"step": 51100
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"eval_loss": 1.2872198820114136,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51100
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"learning_rate": 1.3524751220130572e-05,
|
|
"loss": 1.3156,
|
|
"step": 51150
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"eval_loss": 1.2876853942871094,
|
|
"eval_runtime": 3.7664,
|
|
"eval_samples_per_second": 13.275,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51150
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"learning_rate": 1.3518412879508145e-05,
|
|
"loss": 1.311,
|
|
"step": 51200
|
|
},
|
|
{
|
|
"epoch": 3.24,
|
|
"eval_loss": 1.2858058214187622,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51200
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"learning_rate": 1.351207453888572e-05,
|
|
"loss": 1.2894,
|
|
"step": 51250
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"eval_loss": 1.2863818407058716,
|
|
"eval_runtime": 3.7771,
|
|
"eval_samples_per_second": 13.237,
|
|
"eval_steps_per_second": 0.529,
|
|
"step": 51250
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"learning_rate": 1.3505736198263295e-05,
|
|
"loss": 1.331,
|
|
"step": 51300
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"eval_loss": 1.286733865737915,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51300
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"learning_rate": 1.349939785764087e-05,
|
|
"loss": 1.2866,
|
|
"step": 51350
|
|
},
|
|
{
|
|
"epoch": 3.25,
|
|
"eval_loss": 1.2881466150283813,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51350
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"learning_rate": 1.3493059517018446e-05,
|
|
"loss": 1.3038,
|
|
"step": 51400
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"eval_loss": 1.28376042842865,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51400
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"learning_rate": 1.348672117639602e-05,
|
|
"loss": 1.3063,
|
|
"step": 51450
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"eval_loss": 1.285447597503662,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51450
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"learning_rate": 1.3480382835773595e-05,
|
|
"loss": 1.3229,
|
|
"step": 51500
|
|
},
|
|
{
|
|
"epoch": 3.26,
|
|
"eval_loss": 1.2888320684432983,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51500
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"learning_rate": 1.347404449515117e-05,
|
|
"loss": 1.3139,
|
|
"step": 51550
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"eval_loss": 1.2825188636779785,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51550
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"learning_rate": 1.3467706154528745e-05,
|
|
"loss": 1.3081,
|
|
"step": 51600
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"eval_loss": 1.2877098321914673,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51600
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"learning_rate": 1.3461367813906321e-05,
|
|
"loss": 1.2963,
|
|
"step": 51650
|
|
},
|
|
{
|
|
"epoch": 3.27,
|
|
"eval_loss": 1.286777377128601,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51650
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"learning_rate": 1.3455029473283896e-05,
|
|
"loss": 1.3126,
|
|
"step": 51700
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"eval_loss": 1.2857211828231812,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51700
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"learning_rate": 1.3448691132661471e-05,
|
|
"loss": 1.3117,
|
|
"step": 51750
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"eval_loss": 1.2868152856826782,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51750
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"learning_rate": 1.3442352792039046e-05,
|
|
"loss": 1.3058,
|
|
"step": 51800
|
|
},
|
|
{
|
|
"epoch": 3.28,
|
|
"eval_loss": 1.282294511795044,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51800
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"learning_rate": 1.343601445141662e-05,
|
|
"loss": 1.3104,
|
|
"step": 51850
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"eval_loss": 1.281558871269226,
|
|
"eval_runtime": 3.7746,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51850
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"learning_rate": 1.3429676110794194e-05,
|
|
"loss": 1.3198,
|
|
"step": 51900
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"eval_loss": 1.2874211072921753,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 51900
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"learning_rate": 1.342333777017177e-05,
|
|
"loss": 1.2938,
|
|
"step": 51950
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"eval_loss": 1.284914255142212,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 51950
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"learning_rate": 1.3416999429549345e-05,
|
|
"loss": 1.3237,
|
|
"step": 52000
|
|
},
|
|
{
|
|
"epoch": 3.29,
|
|
"eval_loss": 1.2883204221725464,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52000
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"learning_rate": 1.341066108892692e-05,
|
|
"loss": 1.3144,
|
|
"step": 52050
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"eval_loss": 1.2854008674621582,
|
|
"eval_runtime": 3.7748,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52050
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"learning_rate": 1.3404322748304494e-05,
|
|
"loss": 1.2908,
|
|
"step": 52100
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"eval_loss": 1.2861278057098389,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52100
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"learning_rate": 1.339798440768207e-05,
|
|
"loss": 1.2833,
|
|
"step": 52150
|
|
},
|
|
{
|
|
"epoch": 3.3,
|
|
"eval_loss": 1.2840849161148071,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52150
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"learning_rate": 1.3391646067059646e-05,
|
|
"loss": 1.2999,
|
|
"step": 52200
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"eval_loss": 1.2868108749389648,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52200
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"learning_rate": 1.338530772643722e-05,
|
|
"loss": 1.3009,
|
|
"step": 52250
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"eval_loss": 1.2854992151260376,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52250
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"learning_rate": 1.3378969385814795e-05,
|
|
"loss": 1.3063,
|
|
"step": 52300
|
|
},
|
|
{
|
|
"epoch": 3.31,
|
|
"eval_loss": 1.285149335861206,
|
|
"eval_runtime": 3.7752,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52300
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"learning_rate": 1.337263104519237e-05,
|
|
"loss": 1.2919,
|
|
"step": 52350
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"eval_loss": 1.2940950393676758,
|
|
"eval_runtime": 3.7671,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 52350
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"learning_rate": 1.3366292704569945e-05,
|
|
"loss": 1.3304,
|
|
"step": 52400
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"eval_loss": 1.2862727642059326,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52400
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"learning_rate": 1.3359954363947521e-05,
|
|
"loss": 1.3085,
|
|
"step": 52450
|
|
},
|
|
{
|
|
"epoch": 3.32,
|
|
"eval_loss": 1.2856841087341309,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52450
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"learning_rate": 1.3353616023325096e-05,
|
|
"loss": 1.2911,
|
|
"step": 52500
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"eval_loss": 1.2933928966522217,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52500
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"learning_rate": 1.3347277682702669e-05,
|
|
"loss": 1.3175,
|
|
"step": 52550
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"eval_loss": 1.2892231941223145,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52550
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"learning_rate": 1.3340939342080244e-05,
|
|
"loss": 1.3153,
|
|
"step": 52600
|
|
},
|
|
{
|
|
"epoch": 3.33,
|
|
"eval_loss": 1.2931110858917236,
|
|
"eval_runtime": 3.7754,
|
|
"eval_samples_per_second": 13.244,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52600
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"learning_rate": 1.3334601001457819e-05,
|
|
"loss": 1.3383,
|
|
"step": 52650
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"eval_loss": 1.282524824142456,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52650
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"learning_rate": 1.3328262660835393e-05,
|
|
"loss": 1.2964,
|
|
"step": 52700
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"eval_loss": 1.2855784893035889,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52700
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"learning_rate": 1.3321924320212968e-05,
|
|
"loss": 1.2984,
|
|
"step": 52750
|
|
},
|
|
{
|
|
"epoch": 3.34,
|
|
"eval_loss": 1.2847896814346313,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52750
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"learning_rate": 1.3315585979590545e-05,
|
|
"loss": 1.3237,
|
|
"step": 52800
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"eval_loss": 1.284273386001587,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52800
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"learning_rate": 1.330924763896812e-05,
|
|
"loss": 1.3503,
|
|
"step": 52850
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"eval_loss": 1.2860875129699707,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52850
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"learning_rate": 1.3302909298345694e-05,
|
|
"loss": 1.2984,
|
|
"step": 52900
|
|
},
|
|
{
|
|
"epoch": 3.35,
|
|
"eval_loss": 1.2833102941513062,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52900
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"learning_rate": 1.3296570957723269e-05,
|
|
"loss": 1.3112,
|
|
"step": 52950
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"eval_loss": 1.282645583152771,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 52950
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"learning_rate": 1.3290232617100844e-05,
|
|
"loss": 1.3042,
|
|
"step": 53000
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"eval_loss": 1.284591555595398,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53000
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"learning_rate": 1.328389427647842e-05,
|
|
"loss": 1.293,
|
|
"step": 53050
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"eval_loss": 1.2798117399215698,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53050
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"learning_rate": 1.3277555935855995e-05,
|
|
"loss": 1.2956,
|
|
"step": 53100
|
|
},
|
|
{
|
|
"epoch": 3.36,
|
|
"eval_loss": 1.2897180318832397,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53100
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"learning_rate": 1.327121759523357e-05,
|
|
"loss": 1.3109,
|
|
"step": 53150
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"eval_loss": 1.2877709865570068,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53150
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"learning_rate": 1.3264879254611145e-05,
|
|
"loss": 1.2974,
|
|
"step": 53200
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"eval_loss": 1.2809820175170898,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53200
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"learning_rate": 1.3258540913988718e-05,
|
|
"loss": 1.2972,
|
|
"step": 53250
|
|
},
|
|
{
|
|
"epoch": 3.37,
|
|
"eval_loss": 1.2851955890655518,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53250
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"learning_rate": 1.3252202573366292e-05,
|
|
"loss": 1.3243,
|
|
"step": 53300
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"eval_loss": 1.2849105596542358,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53300
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"learning_rate": 1.3245864232743867e-05,
|
|
"loss": 1.3232,
|
|
"step": 53350
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"eval_loss": 1.2811602354049683,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53350
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"learning_rate": 1.3239525892121444e-05,
|
|
"loss": 1.2966,
|
|
"step": 53400
|
|
},
|
|
{
|
|
"epoch": 3.38,
|
|
"eval_loss": 1.2851645946502686,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53400
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"learning_rate": 1.3233187551499018e-05,
|
|
"loss": 1.3172,
|
|
"step": 53450
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"eval_loss": 1.2857338190078735,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53450
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"learning_rate": 1.3226975977689042e-05,
|
|
"loss": 1.2838,
|
|
"step": 53500
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"eval_loss": 1.2862354516983032,
|
|
"eval_runtime": 3.7739,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53500
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"learning_rate": 1.3220637637066618e-05,
|
|
"loss": 1.3171,
|
|
"step": 53550
|
|
},
|
|
{
|
|
"epoch": 3.39,
|
|
"eval_loss": 1.2938047647476196,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53550
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"learning_rate": 1.3214299296444193e-05,
|
|
"loss": 1.3062,
|
|
"step": 53600
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"eval_loss": 1.286950707435608,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53600
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"learning_rate": 1.3207960955821766e-05,
|
|
"loss": 1.3226,
|
|
"step": 53650
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"eval_loss": 1.285588264465332,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53650
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"learning_rate": 1.3201622615199341e-05,
|
|
"loss": 1.301,
|
|
"step": 53700
|
|
},
|
|
{
|
|
"epoch": 3.4,
|
|
"eval_loss": 1.284866213798523,
|
|
"eval_runtime": 3.7668,
|
|
"eval_samples_per_second": 13.274,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53700
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"learning_rate": 1.3195284274576916e-05,
|
|
"loss": 1.3439,
|
|
"step": 53750
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"eval_loss": 1.2845818996429443,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53750
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"learning_rate": 1.318894593395449e-05,
|
|
"loss": 1.3259,
|
|
"step": 53800
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"eval_loss": 1.28196120262146,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53800
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"learning_rate": 1.3182607593332067e-05,
|
|
"loss": 1.3055,
|
|
"step": 53850
|
|
},
|
|
{
|
|
"epoch": 3.41,
|
|
"eval_loss": 1.2864174842834473,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 53850
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"learning_rate": 1.3176269252709642e-05,
|
|
"loss": 1.3074,
|
|
"step": 53900
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"eval_loss": 1.2848498821258545,
|
|
"eval_runtime": 3.7612,
|
|
"eval_samples_per_second": 13.294,
|
|
"eval_steps_per_second": 0.532,
|
|
"step": 53900
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"learning_rate": 1.3169930912087217e-05,
|
|
"loss": 1.3177,
|
|
"step": 53950
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"eval_loss": 1.2837496995925903,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 53950
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"learning_rate": 1.3163592571464791e-05,
|
|
"loss": 1.3065,
|
|
"step": 54000
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"eval_loss": 1.2816375494003296,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54000
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"learning_rate": 1.3157254230842366e-05,
|
|
"loss": 1.2973,
|
|
"step": 54050
|
|
},
|
|
{
|
|
"epoch": 3.42,
|
|
"eval_loss": 1.2858256101608276,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54050
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"learning_rate": 1.3150915890219943e-05,
|
|
"loss": 1.2867,
|
|
"step": 54100
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"eval_loss": 1.2842872142791748,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54100
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"learning_rate": 1.3144577549597517e-05,
|
|
"loss": 1.3046,
|
|
"step": 54150
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"eval_loss": 1.2836482524871826,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54150
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"learning_rate": 1.3138239208975092e-05,
|
|
"loss": 1.3051,
|
|
"step": 54200
|
|
},
|
|
{
|
|
"epoch": 3.43,
|
|
"eval_loss": 1.291039228439331,
|
|
"eval_runtime": 3.7771,
|
|
"eval_samples_per_second": 13.238,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54200
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"learning_rate": 1.3131900868352667e-05,
|
|
"loss": 1.3264,
|
|
"step": 54250
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"eval_loss": 1.2793464660644531,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54250
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"learning_rate": 1.3125562527730242e-05,
|
|
"loss": 1.3272,
|
|
"step": 54300
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"eval_loss": 1.2878021001815796,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54300
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"learning_rate": 1.3119224187107815e-05,
|
|
"loss": 1.2976,
|
|
"step": 54350
|
|
},
|
|
{
|
|
"epoch": 3.44,
|
|
"eval_loss": 1.2889869213104248,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54350
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"learning_rate": 1.311288584648539e-05,
|
|
"loss": 1.3002,
|
|
"step": 54400
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"eval_loss": 1.2873438596725464,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54400
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"learning_rate": 1.3106547505862966e-05,
|
|
"loss": 1.3294,
|
|
"step": 54450
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"eval_loss": 1.2851972579956055,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54450
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"learning_rate": 1.3100209165240541e-05,
|
|
"loss": 1.2959,
|
|
"step": 54500
|
|
},
|
|
{
|
|
"epoch": 3.45,
|
|
"eval_loss": 1.2863737344741821,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54500
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"learning_rate": 1.3093870824618116e-05,
|
|
"loss": 1.3207,
|
|
"step": 54550
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"eval_loss": 1.2855852842330933,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54550
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"learning_rate": 1.308753248399569e-05,
|
|
"loss": 1.3384,
|
|
"step": 54600
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"eval_loss": 1.2864645719528198,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54600
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"learning_rate": 1.3081194143373265e-05,
|
|
"loss": 1.313,
|
|
"step": 54650
|
|
},
|
|
{
|
|
"epoch": 3.46,
|
|
"eval_loss": 1.2822331190109253,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54650
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"learning_rate": 1.3074855802750842e-05,
|
|
"loss": 1.283,
|
|
"step": 54700
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"eval_loss": 1.2842830419540405,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54700
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"learning_rate": 1.3068517462128416e-05,
|
|
"loss": 1.3129,
|
|
"step": 54750
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"eval_loss": 1.2846298217773438,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 54750
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"learning_rate": 1.3062179121505991e-05,
|
|
"loss": 1.2923,
|
|
"step": 54800
|
|
},
|
|
{
|
|
"epoch": 3.47,
|
|
"eval_loss": 1.28788423538208,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54800
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"learning_rate": 1.3055840780883566e-05,
|
|
"loss": 1.2953,
|
|
"step": 54850
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"eval_loss": 1.2856794595718384,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54850
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"learning_rate": 1.304950244026114e-05,
|
|
"loss": 1.2977,
|
|
"step": 54900
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"eval_loss": 1.2865501642227173,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54900
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"learning_rate": 1.3043164099638717e-05,
|
|
"loss": 1.3095,
|
|
"step": 54950
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"eval_loss": 1.2839913368225098,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 54950
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"learning_rate": 1.3036825759016292e-05,
|
|
"loss": 1.2997,
|
|
"step": 55000
|
|
},
|
|
{
|
|
"epoch": 3.48,
|
|
"eval_loss": 1.2830027341842651,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55000
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"learning_rate": 1.3030487418393865e-05,
|
|
"loss": 1.3011,
|
|
"step": 55050
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"eval_loss": 1.283743143081665,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55050
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"learning_rate": 1.302414907777144e-05,
|
|
"loss": 1.3028,
|
|
"step": 55100
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"eval_loss": 1.2855929136276245,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55100
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"learning_rate": 1.3017810737149015e-05,
|
|
"loss": 1.3373,
|
|
"step": 55150
|
|
},
|
|
{
|
|
"epoch": 3.49,
|
|
"eval_loss": 1.2842646837234497,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 55150
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"learning_rate": 1.301147239652659e-05,
|
|
"loss": 1.2971,
|
|
"step": 55200
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"eval_loss": 1.2828935384750366,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55200
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"learning_rate": 1.3005134055904164e-05,
|
|
"loss": 1.3404,
|
|
"step": 55250
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"eval_loss": 1.2834892272949219,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55250
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"learning_rate": 1.299879571528174e-05,
|
|
"loss": 1.3173,
|
|
"step": 55300
|
|
},
|
|
{
|
|
"epoch": 3.5,
|
|
"eval_loss": 1.2819851636886597,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 55300
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"learning_rate": 1.2992457374659315e-05,
|
|
"loss": 1.304,
|
|
"step": 55350
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"eval_loss": 1.2823426723480225,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55350
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"learning_rate": 1.298611903403689e-05,
|
|
"loss": 1.3141,
|
|
"step": 55400
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"eval_loss": 1.2874670028686523,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55400
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"learning_rate": 1.2979780693414465e-05,
|
|
"loss": 1.2824,
|
|
"step": 55450
|
|
},
|
|
{
|
|
"epoch": 3.51,
|
|
"eval_loss": 1.2810462713241577,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 55450
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"learning_rate": 1.297344235279204e-05,
|
|
"loss": 1.3069,
|
|
"step": 55500
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"eval_loss": 1.2945303916931152,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55500
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"learning_rate": 1.2967104012169616e-05,
|
|
"loss": 1.3228,
|
|
"step": 55550
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"eval_loss": 1.2789846658706665,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 55550
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"learning_rate": 1.2960892438359638e-05,
|
|
"loss": 1.3066,
|
|
"step": 55600
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"eval_loss": 1.2848808765411377,
|
|
"eval_runtime": 3.7692,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 55600
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"learning_rate": 1.2954554097737213e-05,
|
|
"loss": 1.3076,
|
|
"step": 55650
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"eval_loss": 1.2851313352584839,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55650
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"learning_rate": 1.2948215757114788e-05,
|
|
"loss": 1.3035,
|
|
"step": 55700
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"eval_loss": 1.2848730087280273,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55700
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"learning_rate": 1.2941877416492364e-05,
|
|
"loss": 1.3077,
|
|
"step": 55750
|
|
},
|
|
{
|
|
"epoch": 3.53,
|
|
"eval_loss": 1.2845330238342285,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55750
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"learning_rate": 1.2935539075869939e-05,
|
|
"loss": 1.3134,
|
|
"step": 55800
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"eval_loss": 1.2925485372543335,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55800
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"learning_rate": 1.2929200735247514e-05,
|
|
"loss": 1.3183,
|
|
"step": 55850
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"eval_loss": 1.2867827415466309,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55850
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"learning_rate": 1.2922862394625088e-05,
|
|
"loss": 1.3189,
|
|
"step": 55900
|
|
},
|
|
{
|
|
"epoch": 3.54,
|
|
"eval_loss": 1.2831721305847168,
|
|
"eval_runtime": 3.7723,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55900
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"learning_rate": 1.2916524054002663e-05,
|
|
"loss": 1.3018,
|
|
"step": 55950
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"eval_loss": 1.2842330932617188,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 55950
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"learning_rate": 1.291018571338024e-05,
|
|
"loss": 1.3242,
|
|
"step": 56000
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"eval_loss": 1.282920241355896,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56000
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"learning_rate": 1.2903847372757814e-05,
|
|
"loss": 1.3107,
|
|
"step": 56050
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"eval_loss": 1.28363037109375,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56050
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"learning_rate": 1.289750903213539e-05,
|
|
"loss": 1.3088,
|
|
"step": 56100
|
|
},
|
|
{
|
|
"epoch": 3.55,
|
|
"eval_loss": 1.2792989015579224,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56100
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"learning_rate": 1.2891170691512962e-05,
|
|
"loss": 1.3102,
|
|
"step": 56150
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"eval_loss": 1.283616065979004,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56150
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"learning_rate": 1.2884832350890537e-05,
|
|
"loss": 1.2983,
|
|
"step": 56200
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"eval_loss": 1.2842761278152466,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 56200
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"learning_rate": 1.2878494010268112e-05,
|
|
"loss": 1.2895,
|
|
"step": 56250
|
|
},
|
|
{
|
|
"epoch": 3.56,
|
|
"eval_loss": 1.278899073600769,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56250
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"learning_rate": 1.2872155669645687e-05,
|
|
"loss": 1.3113,
|
|
"step": 56300
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"eval_loss": 1.2829270362854004,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56300
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"learning_rate": 1.2865817329023263e-05,
|
|
"loss": 1.295,
|
|
"step": 56350
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"eval_loss": 1.2824040651321411,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56350
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"learning_rate": 1.2859478988400838e-05,
|
|
"loss": 1.2965,
|
|
"step": 56400
|
|
},
|
|
{
|
|
"epoch": 3.57,
|
|
"eval_loss": 1.2882225513458252,
|
|
"eval_runtime": 3.772,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56400
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"learning_rate": 1.2853140647778413e-05,
|
|
"loss": 1.3091,
|
|
"step": 56450
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"eval_loss": 1.2843596935272217,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56450
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"learning_rate": 1.2846802307155987e-05,
|
|
"loss": 1.3216,
|
|
"step": 56500
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"eval_loss": 1.2819628715515137,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 56500
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"learning_rate": 1.2840463966533562e-05,
|
|
"loss": 1.282,
|
|
"step": 56550
|
|
},
|
|
{
|
|
"epoch": 3.58,
|
|
"eval_loss": 1.283927321434021,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 56550
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"learning_rate": 1.2834125625911139e-05,
|
|
"loss": 1.3211,
|
|
"step": 56600
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"eval_loss": 1.2833044528961182,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56600
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"learning_rate": 1.2827787285288713e-05,
|
|
"loss": 1.3109,
|
|
"step": 56650
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"eval_loss": 1.2844871282577515,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56650
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"learning_rate": 1.2821448944666288e-05,
|
|
"loss": 1.2888,
|
|
"step": 56700
|
|
},
|
|
{
|
|
"epoch": 3.59,
|
|
"eval_loss": 1.2839691638946533,
|
|
"eval_runtime": 3.7738,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56700
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"learning_rate": 1.2815110604043863e-05,
|
|
"loss": 1.2945,
|
|
"step": 56750
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"eval_loss": 1.281978964805603,
|
|
"eval_runtime": 3.7746,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56750
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"learning_rate": 1.2808772263421438e-05,
|
|
"loss": 1.3155,
|
|
"step": 56800
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"eval_loss": 1.2823559045791626,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56800
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"learning_rate": 1.2802433922799011e-05,
|
|
"loss": 1.2862,
|
|
"step": 56850
|
|
},
|
|
{
|
|
"epoch": 3.6,
|
|
"eval_loss": 1.2845180034637451,
|
|
"eval_runtime": 3.7743,
|
|
"eval_samples_per_second": 13.247,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56850
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"learning_rate": 1.2796095582176586e-05,
|
|
"loss": 1.2846,
|
|
"step": 56900
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"eval_loss": 1.2806909084320068,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56900
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"learning_rate": 1.2789757241554162e-05,
|
|
"loss": 1.3102,
|
|
"step": 56950
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"eval_loss": 1.2847956418991089,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 56950
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"learning_rate": 1.2783418900931737e-05,
|
|
"loss": 1.3137,
|
|
"step": 57000
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"eval_loss": 1.281103491783142,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57000
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"learning_rate": 1.2777080560309312e-05,
|
|
"loss": 1.3154,
|
|
"step": 57050
|
|
},
|
|
{
|
|
"epoch": 3.61,
|
|
"eval_loss": 1.2803092002868652,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57050
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"learning_rate": 1.2770742219686886e-05,
|
|
"loss": 1.2961,
|
|
"step": 57100
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"eval_loss": 1.279066801071167,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57100
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"learning_rate": 1.2764403879064461e-05,
|
|
"loss": 1.2927,
|
|
"step": 57150
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"eval_loss": 1.2792245149612427,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57150
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"learning_rate": 1.2758065538442038e-05,
|
|
"loss": 1.2641,
|
|
"step": 57200
|
|
},
|
|
{
|
|
"epoch": 3.62,
|
|
"eval_loss": 1.288253664970398,
|
|
"eval_runtime": 3.7719,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57200
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"learning_rate": 1.2751727197819612e-05,
|
|
"loss": 1.2971,
|
|
"step": 57250
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"eval_loss": 1.2804347276687622,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57250
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"learning_rate": 1.2745388857197187e-05,
|
|
"loss": 1.3137,
|
|
"step": 57300
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"eval_loss": 1.279028296470642,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57300
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"learning_rate": 1.2739050516574762e-05,
|
|
"loss": 1.3054,
|
|
"step": 57350
|
|
},
|
|
{
|
|
"epoch": 3.63,
|
|
"eval_loss": 1.2844449281692505,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57350
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"learning_rate": 1.2732712175952337e-05,
|
|
"loss": 1.2973,
|
|
"step": 57400
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"eval_loss": 1.2878245115280151,
|
|
"eval_runtime": 3.7714,
|
|
"eval_samples_per_second": 13.258,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57400
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"learning_rate": 1.2726373835329913e-05,
|
|
"loss": 1.2966,
|
|
"step": 57450
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"eval_loss": 1.2829391956329346,
|
|
"eval_runtime": 3.7748,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57450
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"learning_rate": 1.2720035494707486e-05,
|
|
"loss": 1.2949,
|
|
"step": 57500
|
|
},
|
|
{
|
|
"epoch": 3.64,
|
|
"eval_loss": 1.2807366847991943,
|
|
"eval_runtime": 3.7735,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57500
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"learning_rate": 1.2713697154085061e-05,
|
|
"loss": 1.3145,
|
|
"step": 57550
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"eval_loss": 1.2856266498565674,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57550
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"learning_rate": 1.2707358813462636e-05,
|
|
"loss": 1.3226,
|
|
"step": 57600
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"eval_loss": 1.2831237316131592,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57600
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"learning_rate": 1.270102047284021e-05,
|
|
"loss": 1.314,
|
|
"step": 57650
|
|
},
|
|
{
|
|
"epoch": 3.65,
|
|
"eval_loss": 1.288843035697937,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57650
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"learning_rate": 1.2694682132217785e-05,
|
|
"loss": 1.2981,
|
|
"step": 57700
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"eval_loss": 1.2854373455047607,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57700
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"learning_rate": 1.2688343791595362e-05,
|
|
"loss": 1.2861,
|
|
"step": 57750
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"eval_loss": 1.2818267345428467,
|
|
"eval_runtime": 3.768,
|
|
"eval_samples_per_second": 13.27,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 57750
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"learning_rate": 1.2682005450972937e-05,
|
|
"loss": 1.3117,
|
|
"step": 57800
|
|
},
|
|
{
|
|
"epoch": 3.66,
|
|
"eval_loss": 1.282399296760559,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57800
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"learning_rate": 1.2675667110350511e-05,
|
|
"loss": 1.2659,
|
|
"step": 57850
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"eval_loss": 1.287675142288208,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57850
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"learning_rate": 1.2669328769728086e-05,
|
|
"loss": 1.2858,
|
|
"step": 57900
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"eval_loss": 1.2851409912109375,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 57900
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"learning_rate": 1.2662990429105661e-05,
|
|
"loss": 1.3305,
|
|
"step": 57950
|
|
},
|
|
{
|
|
"epoch": 3.67,
|
|
"eval_loss": 1.281197190284729,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 57950
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"learning_rate": 1.2656652088483238e-05,
|
|
"loss": 1.318,
|
|
"step": 58000
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"eval_loss": 1.2855117321014404,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 58000
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"learning_rate": 1.2650313747860812e-05,
|
|
"loss": 1.319,
|
|
"step": 58050
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"eval_loss": 1.2804759740829468,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58050
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"learning_rate": 1.2643975407238387e-05,
|
|
"loss": 1.2925,
|
|
"step": 58100
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"eval_loss": 1.2818081378936768,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58100
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"learning_rate": 1.2637637066615962e-05,
|
|
"loss": 1.3082,
|
|
"step": 58150
|
|
},
|
|
{
|
|
"epoch": 3.68,
|
|
"eval_loss": 1.27784264087677,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58150
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"learning_rate": 1.2631298725993535e-05,
|
|
"loss": 1.3022,
|
|
"step": 58200
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"eval_loss": 1.2815487384796143,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 58200
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"learning_rate": 1.262496038537111e-05,
|
|
"loss": 1.3099,
|
|
"step": 58250
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"eval_loss": 1.287879467010498,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58250
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"learning_rate": 1.2618622044748684e-05,
|
|
"loss": 1.3414,
|
|
"step": 58300
|
|
},
|
|
{
|
|
"epoch": 3.69,
|
|
"eval_loss": 1.2876009941101074,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58300
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"learning_rate": 1.2612283704126261e-05,
|
|
"loss": 1.3051,
|
|
"step": 58350
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"eval_loss": 1.2809441089630127,
|
|
"eval_runtime": 3.7708,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58350
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"learning_rate": 1.2605945363503836e-05,
|
|
"loss": 1.2911,
|
|
"step": 58400
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"eval_loss": 1.2829344272613525,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58400
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"learning_rate": 1.259960702288141e-05,
|
|
"loss": 1.3122,
|
|
"step": 58450
|
|
},
|
|
{
|
|
"epoch": 3.7,
|
|
"eval_loss": 1.2806299924850464,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58450
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"learning_rate": 1.2593268682258985e-05,
|
|
"loss": 1.303,
|
|
"step": 58500
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"eval_loss": 1.2824500799179077,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58500
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"learning_rate": 1.258693034163656e-05,
|
|
"loss": 1.3268,
|
|
"step": 58550
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"eval_loss": 1.2826014757156372,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58550
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"learning_rate": 1.2580592001014137e-05,
|
|
"loss": 1.3324,
|
|
"step": 58600
|
|
},
|
|
{
|
|
"epoch": 3.71,
|
|
"eval_loss": 1.2844127416610718,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58600
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"learning_rate": 1.2574253660391711e-05,
|
|
"loss": 1.3069,
|
|
"step": 58650
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"eval_loss": 1.2814210653305054,
|
|
"eval_runtime": 3.7747,
|
|
"eval_samples_per_second": 13.246,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58650
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"learning_rate": 1.2567915319769286e-05,
|
|
"loss": 1.2776,
|
|
"step": 58700
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"eval_loss": 1.2827531099319458,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58700
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"learning_rate": 1.2561576979146861e-05,
|
|
"loss": 1.3115,
|
|
"step": 58750
|
|
},
|
|
{
|
|
"epoch": 3.72,
|
|
"eval_loss": 1.2842662334442139,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58750
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"learning_rate": 1.2555238638524436e-05,
|
|
"loss": 1.3231,
|
|
"step": 58800
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"eval_loss": 1.2820935249328613,
|
|
"eval_runtime": 3.7684,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 58800
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"learning_rate": 1.2548900297902012e-05,
|
|
"loss": 1.3035,
|
|
"step": 58850
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"eval_loss": 1.2800058126449585,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 58850
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"learning_rate": 1.2542561957279584e-05,
|
|
"loss": 1.2988,
|
|
"step": 58900
|
|
},
|
|
{
|
|
"epoch": 3.73,
|
|
"eval_loss": 1.2826249599456787,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 58900
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"learning_rate": 1.253622361665716e-05,
|
|
"loss": 1.279,
|
|
"step": 58950
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"eval_loss": 1.2850234508514404,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 58950
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"learning_rate": 1.2529885276034735e-05,
|
|
"loss": 1.296,
|
|
"step": 59000
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"eval_loss": 1.2853106260299683,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59000
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"learning_rate": 1.252354693541231e-05,
|
|
"loss": 1.3047,
|
|
"step": 59050
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"eval_loss": 1.28371262550354,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59050
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"learning_rate": 1.2517208594789884e-05,
|
|
"loss": 1.3079,
|
|
"step": 59100
|
|
},
|
|
{
|
|
"epoch": 3.74,
|
|
"eval_loss": 1.2867982387542725,
|
|
"eval_runtime": 3.7733,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59100
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"learning_rate": 1.2510870254167459e-05,
|
|
"loss": 1.3034,
|
|
"step": 59150
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"eval_loss": 1.2894738912582397,
|
|
"eval_runtime": 3.7676,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59150
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"learning_rate": 1.2504531913545036e-05,
|
|
"loss": 1.3129,
|
|
"step": 59200
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"eval_loss": 1.2861546277999878,
|
|
"eval_runtime": 3.7711,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59200
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"learning_rate": 1.249819357292261e-05,
|
|
"loss": 1.3051,
|
|
"step": 59250
|
|
},
|
|
{
|
|
"epoch": 3.75,
|
|
"eval_loss": 1.2836936712265015,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59250
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"learning_rate": 1.2491855232300185e-05,
|
|
"loss": 1.318,
|
|
"step": 59300
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"eval_loss": 1.2870465517044067,
|
|
"eval_runtime": 3.7718,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59300
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"learning_rate": 1.248551689167776e-05,
|
|
"loss": 1.3124,
|
|
"step": 59350
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"eval_loss": 1.2857956886291504,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59350
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"learning_rate": 1.2479305317867783e-05,
|
|
"loss": 1.2896,
|
|
"step": 59400
|
|
},
|
|
{
|
|
"epoch": 3.76,
|
|
"eval_loss": 1.2858422994613647,
|
|
"eval_runtime": 3.7671,
|
|
"eval_samples_per_second": 13.273,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59400
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"learning_rate": 1.2472966977245358e-05,
|
|
"loss": 1.3217,
|
|
"step": 59450
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"eval_loss": 1.2826424837112427,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59450
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"learning_rate": 1.2466628636622933e-05,
|
|
"loss": 1.2801,
|
|
"step": 59500
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"eval_loss": 1.285921335220337,
|
|
"eval_runtime": 3.7751,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59500
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"learning_rate": 1.2460290296000508e-05,
|
|
"loss": 1.3021,
|
|
"step": 59550
|
|
},
|
|
{
|
|
"epoch": 3.77,
|
|
"eval_loss": 1.2805360555648804,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59550
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"learning_rate": 1.2453951955378082e-05,
|
|
"loss": 1.3048,
|
|
"step": 59600
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"eval_loss": 1.2933862209320068,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59600
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"learning_rate": 1.2447613614755659e-05,
|
|
"loss": 1.302,
|
|
"step": 59650
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"eval_loss": 1.2820707559585571,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59650
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"learning_rate": 1.2441275274133234e-05,
|
|
"loss": 1.3158,
|
|
"step": 59700
|
|
},
|
|
{
|
|
"epoch": 3.78,
|
|
"eval_loss": 1.2801158428192139,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59700
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"learning_rate": 1.2434936933510808e-05,
|
|
"loss": 1.3255,
|
|
"step": 59750
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"eval_loss": 1.288580060005188,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59750
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"learning_rate": 1.2428598592888383e-05,
|
|
"loss": 1.3001,
|
|
"step": 59800
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"eval_loss": 1.2897282838821411,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 59800
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"learning_rate": 1.2422260252265958e-05,
|
|
"loss": 1.2946,
|
|
"step": 59850
|
|
},
|
|
{
|
|
"epoch": 3.79,
|
|
"eval_loss": 1.285589337348938,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59850
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"learning_rate": 1.2415921911643533e-05,
|
|
"loss": 1.2937,
|
|
"step": 59900
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"eval_loss": 1.286169409751892,
|
|
"eval_runtime": 3.7767,
|
|
"eval_samples_per_second": 13.239,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59900
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"learning_rate": 1.240958357102111e-05,
|
|
"loss": 1.3012,
|
|
"step": 59950
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"eval_loss": 1.2858511209487915,
|
|
"eval_runtime": 3.7749,
|
|
"eval_samples_per_second": 13.245,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 59950
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"learning_rate": 1.2403245230398682e-05,
|
|
"loss": 1.2999,
|
|
"step": 60000
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"eval_loss": 1.280913233757019,
|
|
"eval_runtime": 3.7675,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60000
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"learning_rate": 1.2396906889776257e-05,
|
|
"loss": 1.309,
|
|
"step": 60050
|
|
},
|
|
{
|
|
"epoch": 3.8,
|
|
"eval_loss": 1.2815738916397095,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60050
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"learning_rate": 1.2390568549153832e-05,
|
|
"loss": 1.285,
|
|
"step": 60100
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"eval_loss": 1.2818210124969482,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60100
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"learning_rate": 1.2384230208531407e-05,
|
|
"loss": 1.283,
|
|
"step": 60150
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"eval_loss": 1.279291033744812,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60150
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"learning_rate": 1.2377891867908981e-05,
|
|
"loss": 1.3035,
|
|
"step": 60200
|
|
},
|
|
{
|
|
"epoch": 3.81,
|
|
"eval_loss": 1.2826550006866455,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60200
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"learning_rate": 1.2371553527286558e-05,
|
|
"loss": 1.3121,
|
|
"step": 60250
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"eval_loss": 1.2821593284606934,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60250
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"learning_rate": 1.2365215186664133e-05,
|
|
"loss": 1.2973,
|
|
"step": 60300
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"eval_loss": 1.2816184759140015,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60300
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"learning_rate": 1.2358876846041708e-05,
|
|
"loss": 1.3021,
|
|
"step": 60350
|
|
},
|
|
{
|
|
"epoch": 3.82,
|
|
"eval_loss": 1.2808351516723633,
|
|
"eval_runtime": 3.774,
|
|
"eval_samples_per_second": 13.249,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60350
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"learning_rate": 1.2352538505419282e-05,
|
|
"loss": 1.3164,
|
|
"step": 60400
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"eval_loss": 1.2792418003082275,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60400
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"learning_rate": 1.2346200164796857e-05,
|
|
"loss": 1.271,
|
|
"step": 60450
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"eval_loss": 1.285904884338379,
|
|
"eval_runtime": 3.7656,
|
|
"eval_samples_per_second": 13.278,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60450
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"learning_rate": 1.2339861824174434e-05,
|
|
"loss": 1.3119,
|
|
"step": 60500
|
|
},
|
|
{
|
|
"epoch": 3.83,
|
|
"eval_loss": 1.2828646898269653,
|
|
"eval_runtime": 3.765,
|
|
"eval_samples_per_second": 13.28,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60500
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"learning_rate": 1.2333523483552008e-05,
|
|
"loss": 1.2884,
|
|
"step": 60550
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"eval_loss": 1.2857215404510498,
|
|
"eval_runtime": 3.7729,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60550
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"learning_rate": 1.2327185142929583e-05,
|
|
"loss": 1.295,
|
|
"step": 60600
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"eval_loss": 1.2787235975265503,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60600
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"learning_rate": 1.2320846802307156e-05,
|
|
"loss": 1.2898,
|
|
"step": 60650
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"eval_loss": 1.2830379009246826,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60650
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"learning_rate": 1.2314508461684731e-05,
|
|
"loss": 1.2879,
|
|
"step": 60700
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"eval_loss": 1.2803202867507935,
|
|
"eval_runtime": 3.7701,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60700
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"learning_rate": 1.2308170121062306e-05,
|
|
"loss": 1.2844,
|
|
"step": 60750
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"eval_loss": 1.2833034992218018,
|
|
"eval_runtime": 3.7709,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60750
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"learning_rate": 1.230183178043988e-05,
|
|
"loss": 1.3159,
|
|
"step": 60800
|
|
},
|
|
{
|
|
"epoch": 3.85,
|
|
"eval_loss": 1.2802387475967407,
|
|
"eval_runtime": 3.7685,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60800
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"learning_rate": 1.2295493439817457e-05,
|
|
"loss": 1.3153,
|
|
"step": 60850
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"eval_loss": 1.28177809715271,
|
|
"eval_runtime": 3.7696,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60850
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"learning_rate": 1.2289155099195032e-05,
|
|
"loss": 1.3184,
|
|
"step": 60900
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"eval_loss": 1.2795437574386597,
|
|
"eval_runtime": 3.7721,
|
|
"eval_samples_per_second": 13.255,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 60900
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"learning_rate": 1.2282816758572607e-05,
|
|
"loss": 1.2905,
|
|
"step": 60950
|
|
},
|
|
{
|
|
"epoch": 3.86,
|
|
"eval_loss": 1.2796874046325684,
|
|
"eval_runtime": 3.7697,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 60950
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"learning_rate": 1.2276478417950181e-05,
|
|
"loss": 1.2935,
|
|
"step": 61000
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"eval_loss": 1.2820053100585938,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61000
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"learning_rate": 1.2270140077327756e-05,
|
|
"loss": 1.2909,
|
|
"step": 61050
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"eval_loss": 1.283310890197754,
|
|
"eval_runtime": 3.7706,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61050
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"learning_rate": 1.2263801736705333e-05,
|
|
"loss": 1.2766,
|
|
"step": 61100
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"eval_loss": 1.2836205959320068,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 61100
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"learning_rate": 1.2257463396082907e-05,
|
|
"loss": 1.3095,
|
|
"step": 61150
|
|
},
|
|
{
|
|
"epoch": 3.87,
|
|
"eval_loss": 1.2817699909210205,
|
|
"eval_runtime": 3.7694,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 61150
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"learning_rate": 1.2251125055460482e-05,
|
|
"loss": 1.3022,
|
|
"step": 61200
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"eval_loss": 1.2816342115402222,
|
|
"eval_runtime": 3.7731,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61200
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"learning_rate": 1.2244786714838057e-05,
|
|
"loss": 1.3049,
|
|
"step": 61250
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"eval_loss": 1.2803840637207031,
|
|
"eval_runtime": 3.7732,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61250
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"learning_rate": 1.2238448374215632e-05,
|
|
"loss": 1.3008,
|
|
"step": 61300
|
|
},
|
|
{
|
|
"epoch": 3.88,
|
|
"eval_loss": 1.281725287437439,
|
|
"eval_runtime": 3.7688,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 61300
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"learning_rate": 1.2232110033593205e-05,
|
|
"loss": 1.3142,
|
|
"step": 61350
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"eval_loss": 1.2848403453826904,
|
|
"eval_runtime": 3.773,
|
|
"eval_samples_per_second": 13.252,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61350
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"learning_rate": 1.2225771692970781e-05,
|
|
"loss": 1.3137,
|
|
"step": 61400
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"eval_loss": 1.2818886041641235,
|
|
"eval_runtime": 3.7741,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61400
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"learning_rate": 1.2219433352348356e-05,
|
|
"loss": 1.3005,
|
|
"step": 61450
|
|
},
|
|
{
|
|
"epoch": 3.89,
|
|
"eval_loss": 1.283537745475769,
|
|
"eval_runtime": 3.7737,
|
|
"eval_samples_per_second": 13.25,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61450
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"learning_rate": 1.221309501172593e-05,
|
|
"loss": 1.336,
|
|
"step": 61500
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"eval_loss": 1.2776293754577637,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61500
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"learning_rate": 1.2206756671103506e-05,
|
|
"loss": 1.3056,
|
|
"step": 61550
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"eval_loss": 1.283506155014038,
|
|
"eval_runtime": 3.7698,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 61550
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"learning_rate": 1.220041833048108e-05,
|
|
"loss": 1.2928,
|
|
"step": 61600
|
|
},
|
|
{
|
|
"epoch": 3.9,
|
|
"eval_loss": 1.2818597555160522,
|
|
"eval_runtime": 3.7702,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61600
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"learning_rate": 1.2194079989858655e-05,
|
|
"loss": 1.2891,
|
|
"step": 61650
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"eval_loss": 1.2851519584655762,
|
|
"eval_runtime": 3.7715,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61650
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"learning_rate": 1.2187741649236232e-05,
|
|
"loss": 1.3021,
|
|
"step": 61700
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"eval_loss": 1.2861803770065308,
|
|
"eval_runtime": 3.7703,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61700
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"learning_rate": 1.2181403308613806e-05,
|
|
"loss": 1.299,
|
|
"step": 61750
|
|
},
|
|
{
|
|
"epoch": 3.91,
|
|
"eval_loss": 1.2909579277038574,
|
|
"eval_runtime": 3.771,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61750
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"learning_rate": 1.2175064967991381e-05,
|
|
"loss": 1.3037,
|
|
"step": 61800
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"eval_loss": 1.28318190574646,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61800
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"learning_rate": 1.2168726627368956e-05,
|
|
"loss": 1.2865,
|
|
"step": 61850
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"eval_loss": 1.2850934267044067,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61850
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"learning_rate": 1.216238828674653e-05,
|
|
"loss": 1.2969,
|
|
"step": 61900
|
|
},
|
|
{
|
|
"epoch": 3.92,
|
|
"eval_loss": 1.2815029621124268,
|
|
"eval_runtime": 3.769,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 61900
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"learning_rate": 1.2156049946124107e-05,
|
|
"loss": 1.2758,
|
|
"step": 61950
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"eval_loss": 1.2869964838027954,
|
|
"eval_runtime": 3.7717,
|
|
"eval_samples_per_second": 13.256,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 61950
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"learning_rate": 1.2149711605501682e-05,
|
|
"loss": 1.3273,
|
|
"step": 62000
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"eval_loss": 1.2845485210418701,
|
|
"eval_runtime": 3.7687,
|
|
"eval_samples_per_second": 13.267,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62000
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"learning_rate": 1.2143373264879255e-05,
|
|
"loss": 1.2966,
|
|
"step": 62050
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"eval_loss": 1.2811493873596191,
|
|
"eval_runtime": 3.7727,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62050
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"learning_rate": 1.213703492425683e-05,
|
|
"loss": 1.2894,
|
|
"step": 62100
|
|
},
|
|
{
|
|
"epoch": 3.93,
|
|
"eval_loss": 1.2819486856460571,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62100
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"learning_rate": 1.2130696583634405e-05,
|
|
"loss": 1.2853,
|
|
"step": 62150
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"eval_loss": 1.285772681236267,
|
|
"eval_runtime": 3.7689,
|
|
"eval_samples_per_second": 13.266,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62150
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"learning_rate": 1.212435824301198e-05,
|
|
"loss": 1.2924,
|
|
"step": 62200
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"eval_loss": 1.2840394973754883,
|
|
"eval_runtime": 3.7672,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62200
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"learning_rate": 1.2118019902389556e-05,
|
|
"loss": 1.3079,
|
|
"step": 62250
|
|
},
|
|
{
|
|
"epoch": 3.94,
|
|
"eval_loss": 1.28691828250885,
|
|
"eval_runtime": 3.77,
|
|
"eval_samples_per_second": 13.262,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62250
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"learning_rate": 1.211168156176713e-05,
|
|
"loss": 1.3211,
|
|
"step": 62300
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"eval_loss": 1.2858519554138184,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62300
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"learning_rate": 1.2105343221144705e-05,
|
|
"loss": 1.2916,
|
|
"step": 62350
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"eval_loss": 1.2870821952819824,
|
|
"eval_runtime": 3.7693,
|
|
"eval_samples_per_second": 13.265,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62350
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"learning_rate": 1.209900488052228e-05,
|
|
"loss": 1.3148,
|
|
"step": 62400
|
|
},
|
|
{
|
|
"epoch": 3.95,
|
|
"eval_loss": 1.2813467979431152,
|
|
"eval_runtime": 3.7712,
|
|
"eval_samples_per_second": 13.259,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62400
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"learning_rate": 1.2092666539899855e-05,
|
|
"loss": 1.3038,
|
|
"step": 62450
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"eval_loss": 1.2803038358688354,
|
|
"eval_runtime": 3.7695,
|
|
"eval_samples_per_second": 13.264,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62450
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"learning_rate": 1.2086328199277431e-05,
|
|
"loss": 1.3143,
|
|
"step": 62500
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"eval_loss": 1.2869164943695068,
|
|
"eval_runtime": 3.7699,
|
|
"eval_samples_per_second": 13.263,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62500
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"learning_rate": 1.2079989858655006e-05,
|
|
"loss": 1.3129,
|
|
"step": 62550
|
|
},
|
|
{
|
|
"epoch": 3.96,
|
|
"eval_loss": 1.2853258848190308,
|
|
"eval_runtime": 3.7705,
|
|
"eval_samples_per_second": 13.261,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62550
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"learning_rate": 1.2073651518032581e-05,
|
|
"loss": 1.3094,
|
|
"step": 62600
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"eval_loss": 1.282500147819519,
|
|
"eval_runtime": 3.7734,
|
|
"eval_samples_per_second": 13.251,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62600
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"learning_rate": 1.2067313177410156e-05,
|
|
"loss": 1.2833,
|
|
"step": 62650
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"eval_loss": 1.2867021560668945,
|
|
"eval_runtime": 3.7681,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62650
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"learning_rate": 1.206097483678773e-05,
|
|
"loss": 1.2875,
|
|
"step": 62700
|
|
},
|
|
{
|
|
"epoch": 3.97,
|
|
"eval_loss": 1.2819606065750122,
|
|
"eval_runtime": 3.7673,
|
|
"eval_samples_per_second": 13.272,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62700
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"learning_rate": 1.2054636496165304e-05,
|
|
"loss": 1.295,
|
|
"step": 62750
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"eval_loss": 1.282488226890564,
|
|
"eval_runtime": 3.7728,
|
|
"eval_samples_per_second": 13.253,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62750
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"learning_rate": 1.2048298155542878e-05,
|
|
"loss": 1.2854,
|
|
"step": 62800
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"eval_loss": 1.2850754261016846,
|
|
"eval_runtime": 3.7683,
|
|
"eval_samples_per_second": 13.268,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 62800
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"learning_rate": 1.2041959814920455e-05,
|
|
"loss": 1.2985,
|
|
"step": 62850
|
|
},
|
|
{
|
|
"epoch": 3.98,
|
|
"eval_loss": 1.2898350954055786,
|
|
"eval_runtime": 3.7725,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62850
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"learning_rate": 1.203562147429803e-05,
|
|
"loss": 1.3023,
|
|
"step": 62900
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"eval_loss": 1.289136528968811,
|
|
"eval_runtime": 3.7724,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62900
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"learning_rate": 1.2029283133675604e-05,
|
|
"loss": 1.3083,
|
|
"step": 62950
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"eval_loss": 1.281229853630066,
|
|
"eval_runtime": 3.7742,
|
|
"eval_samples_per_second": 13.248,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 62950
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"learning_rate": 1.2022944793053179e-05,
|
|
"loss": 1.2947,
|
|
"step": 63000
|
|
},
|
|
{
|
|
"epoch": 3.99,
|
|
"eval_loss": 1.2780238389968872,
|
|
"eval_runtime": 3.7716,
|
|
"eval_samples_per_second": 13.257,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 63000
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"learning_rate": 1.2016606452430754e-05,
|
|
"loss": 1.2902,
|
|
"step": 63050
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_loss": 1.2823433876037598,
|
|
"eval_runtime": 3.7726,
|
|
"eval_samples_per_second": 13.254,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 63050
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"learning_rate": 1.201026811180833e-05,
|
|
"loss": 1.3094,
|
|
"step": 63100
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_loss": 1.2773927450180054,
|
|
"eval_runtime": 3.7682,
|
|
"eval_samples_per_second": 13.269,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 63100
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"learning_rate": 1.2004056537998352e-05,
|
|
"loss": 1.3274,
|
|
"step": 63150
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_loss": 1.2770581245422363,
|
|
"eval_runtime": 3.7677,
|
|
"eval_samples_per_second": 13.271,
|
|
"eval_steps_per_second": 0.531,
|
|
"step": 63150
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"learning_rate": 1.1997718197375927e-05,
|
|
"loss": 1.2868,
|
|
"step": 63200
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_loss": 1.2748980522155762,
|
|
"eval_runtime": 3.7707,
|
|
"eval_samples_per_second": 13.26,
|
|
"eval_steps_per_second": 0.53,
|
|
"step": 63200
|
|
}
|
|
],
|
|
"max_steps": 157820,
|
|
"num_train_epochs": 10,
|
|
"total_flos": 1.2041035551006628e+20,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|