Yaxin's picture
update from Yaxin
a671727
raw
history blame
13.9 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 54528,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 4.954151995305165e-05,
"loss": 4.1243,
"step": 500
},
{
"epoch": 0.06,
"learning_rate": 4.908303990610329e-05,
"loss": 3.0378,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 4.8624559859154934e-05,
"loss": 2.767,
"step": 1500
},
{
"epoch": 0.11,
"learning_rate": 4.816607981220658e-05,
"loss": 2.5961,
"step": 2000
},
{
"epoch": 0.14,
"learning_rate": 4.770759976525822e-05,
"loss": 2.4848,
"step": 2500
},
{
"epoch": 0.17,
"learning_rate": 4.724911971830986e-05,
"loss": 2.4087,
"step": 3000
},
{
"epoch": 0.19,
"learning_rate": 4.6790639671361505e-05,
"loss": 2.34,
"step": 3500
},
{
"epoch": 0.22,
"learning_rate": 4.6332159624413144e-05,
"loss": 2.3038,
"step": 4000
},
{
"epoch": 0.25,
"learning_rate": 4.587367957746479e-05,
"loss": 2.238,
"step": 4500
},
{
"epoch": 0.28,
"learning_rate": 4.5415199530516436e-05,
"loss": 2.224,
"step": 5000
},
{
"epoch": 0.3,
"learning_rate": 4.4956719483568075e-05,
"loss": 2.1801,
"step": 5500
},
{
"epoch": 0.33,
"learning_rate": 4.449823943661972e-05,
"loss": 2.1474,
"step": 6000
},
{
"epoch": 0.36,
"learning_rate": 4.403975938967137e-05,
"loss": 2.1175,
"step": 6500
},
{
"epoch": 0.39,
"learning_rate": 4.3581279342723007e-05,
"loss": 2.0952,
"step": 7000
},
{
"epoch": 0.41,
"learning_rate": 4.312279929577465e-05,
"loss": 2.0845,
"step": 7500
},
{
"epoch": 0.44,
"learning_rate": 4.26643192488263e-05,
"loss": 2.056,
"step": 8000
},
{
"epoch": 0.47,
"learning_rate": 4.220583920187793e-05,
"loss": 2.0451,
"step": 8500
},
{
"epoch": 0.5,
"learning_rate": 4.174735915492958e-05,
"loss": 2.0272,
"step": 9000
},
{
"epoch": 0.52,
"learning_rate": 4.128887910798122e-05,
"loss": 2.0085,
"step": 9500
},
{
"epoch": 0.55,
"learning_rate": 4.083039906103286e-05,
"loss": 1.989,
"step": 10000
},
{
"epoch": 0.58,
"learning_rate": 4.037191901408451e-05,
"loss": 1.9824,
"step": 10500
},
{
"epoch": 0.61,
"learning_rate": 3.9913438967136155e-05,
"loss": 1.966,
"step": 11000
},
{
"epoch": 0.63,
"learning_rate": 3.9454958920187794e-05,
"loss": 1.9524,
"step": 11500
},
{
"epoch": 0.66,
"learning_rate": 3.899647887323944e-05,
"loss": 1.9426,
"step": 12000
},
{
"epoch": 0.69,
"learning_rate": 3.8537998826291086e-05,
"loss": 1.9391,
"step": 12500
},
{
"epoch": 0.72,
"learning_rate": 3.8079518779342725e-05,
"loss": 1.9316,
"step": 13000
},
{
"epoch": 0.74,
"learning_rate": 3.762103873239437e-05,
"loss": 1.9163,
"step": 13500
},
{
"epoch": 0.77,
"learning_rate": 3.716255868544601e-05,
"loss": 1.9031,
"step": 14000
},
{
"epoch": 0.8,
"learning_rate": 3.670407863849765e-05,
"loss": 1.8949,
"step": 14500
},
{
"epoch": 0.83,
"learning_rate": 3.6245598591549296e-05,
"loss": 1.8862,
"step": 15000
},
{
"epoch": 0.85,
"learning_rate": 3.578711854460094e-05,
"loss": 1.8856,
"step": 15500
},
{
"epoch": 0.88,
"learning_rate": 3.532863849765258e-05,
"loss": 1.8817,
"step": 16000
},
{
"epoch": 0.91,
"learning_rate": 3.487015845070423e-05,
"loss": 1.8613,
"step": 16500
},
{
"epoch": 0.94,
"learning_rate": 3.4411678403755874e-05,
"loss": 1.8533,
"step": 17000
},
{
"epoch": 0.96,
"learning_rate": 3.395319835680751e-05,
"loss": 1.8591,
"step": 17500
},
{
"epoch": 0.99,
"learning_rate": 3.349471830985916e-05,
"loss": 1.8461,
"step": 18000
},
{
"epoch": 1.02,
"learning_rate": 3.30362382629108e-05,
"loss": 1.8417,
"step": 18500
},
{
"epoch": 1.05,
"learning_rate": 3.2577758215962444e-05,
"loss": 1.825,
"step": 19000
},
{
"epoch": 1.07,
"learning_rate": 3.211927816901409e-05,
"loss": 1.8193,
"step": 19500
},
{
"epoch": 1.1,
"learning_rate": 3.166079812206573e-05,
"loss": 1.817,
"step": 20000
},
{
"epoch": 1.13,
"learning_rate": 3.120231807511737e-05,
"loss": 1.8042,
"step": 20500
},
{
"epoch": 1.16,
"learning_rate": 3.0743838028169015e-05,
"loss": 1.7981,
"step": 21000
},
{
"epoch": 1.18,
"learning_rate": 3.0285357981220658e-05,
"loss": 1.8042,
"step": 21500
},
{
"epoch": 1.21,
"learning_rate": 2.98268779342723e-05,
"loss": 1.7944,
"step": 22000
},
{
"epoch": 1.24,
"learning_rate": 2.9368397887323946e-05,
"loss": 1.7904,
"step": 22500
},
{
"epoch": 1.27,
"learning_rate": 2.890991784037559e-05,
"loss": 1.7852,
"step": 23000
},
{
"epoch": 1.29,
"learning_rate": 2.8451437793427232e-05,
"loss": 1.7744,
"step": 23500
},
{
"epoch": 1.32,
"learning_rate": 2.7992957746478874e-05,
"loss": 1.7684,
"step": 24000
},
{
"epoch": 1.35,
"learning_rate": 2.753447769953052e-05,
"loss": 1.7698,
"step": 24500
},
{
"epoch": 1.38,
"learning_rate": 2.7075997652582163e-05,
"loss": 1.7746,
"step": 25000
},
{
"epoch": 1.4,
"learning_rate": 2.6617517605633802e-05,
"loss": 1.7667,
"step": 25500
},
{
"epoch": 1.43,
"learning_rate": 2.6159037558685445e-05,
"loss": 1.7642,
"step": 26000
},
{
"epoch": 1.46,
"learning_rate": 2.5700557511737088e-05,
"loss": 1.7575,
"step": 26500
},
{
"epoch": 1.49,
"learning_rate": 2.5242077464788734e-05,
"loss": 1.7573,
"step": 27000
},
{
"epoch": 1.51,
"learning_rate": 2.4783597417840377e-05,
"loss": 1.7442,
"step": 27500
},
{
"epoch": 1.54,
"learning_rate": 2.432511737089202e-05,
"loss": 1.7433,
"step": 28000
},
{
"epoch": 1.57,
"learning_rate": 2.3866637323943662e-05,
"loss": 1.7487,
"step": 28500
},
{
"epoch": 1.6,
"learning_rate": 2.3408157276995308e-05,
"loss": 1.7338,
"step": 29000
},
{
"epoch": 1.62,
"learning_rate": 2.2949677230046947e-05,
"loss": 1.7282,
"step": 29500
},
{
"epoch": 1.65,
"learning_rate": 2.2491197183098593e-05,
"loss": 1.7302,
"step": 30000
},
{
"epoch": 1.68,
"learning_rate": 2.2032717136150236e-05,
"loss": 1.725,
"step": 30500
},
{
"epoch": 1.71,
"learning_rate": 2.157423708920188e-05,
"loss": 1.7152,
"step": 31000
},
{
"epoch": 1.73,
"learning_rate": 2.111575704225352e-05,
"loss": 1.723,
"step": 31500
},
{
"epoch": 1.76,
"learning_rate": 2.0657276995305167e-05,
"loss": 1.7078,
"step": 32000
},
{
"epoch": 1.79,
"learning_rate": 2.0198796948356807e-05,
"loss": 1.7119,
"step": 32500
},
{
"epoch": 1.82,
"learning_rate": 1.974031690140845e-05,
"loss": 1.7088,
"step": 33000
},
{
"epoch": 1.84,
"learning_rate": 1.9281836854460095e-05,
"loss": 1.7054,
"step": 33500
},
{
"epoch": 1.87,
"learning_rate": 1.8823356807511738e-05,
"loss": 1.7057,
"step": 34000
},
{
"epoch": 1.9,
"learning_rate": 1.836487676056338e-05,
"loss": 1.7088,
"step": 34500
},
{
"epoch": 1.93,
"learning_rate": 1.7906396713615027e-05,
"loss": 1.689,
"step": 35000
},
{
"epoch": 1.95,
"learning_rate": 1.7447916666666666e-05,
"loss": 1.6948,
"step": 35500
},
{
"epoch": 1.98,
"learning_rate": 1.698943661971831e-05,
"loss": 1.6877,
"step": 36000
},
{
"epoch": 2.01,
"learning_rate": 1.6530956572769955e-05,
"loss": 1.6983,
"step": 36500
},
{
"epoch": 2.04,
"learning_rate": 1.6072476525821598e-05,
"loss": 1.6694,
"step": 37000
},
{
"epoch": 2.06,
"learning_rate": 1.561399647887324e-05,
"loss": 1.6874,
"step": 37500
},
{
"epoch": 2.09,
"learning_rate": 1.5155516431924883e-05,
"loss": 1.6714,
"step": 38000
},
{
"epoch": 2.12,
"learning_rate": 1.4697036384976526e-05,
"loss": 1.6783,
"step": 38500
},
{
"epoch": 2.15,
"learning_rate": 1.423855633802817e-05,
"loss": 1.6766,
"step": 39000
},
{
"epoch": 2.17,
"learning_rate": 1.3780076291079813e-05,
"loss": 1.6707,
"step": 39500
},
{
"epoch": 2.2,
"learning_rate": 1.3321596244131457e-05,
"loss": 1.6723,
"step": 40000
},
{
"epoch": 2.23,
"learning_rate": 1.28631161971831e-05,
"loss": 1.6708,
"step": 40500
},
{
"epoch": 2.26,
"learning_rate": 1.2404636150234742e-05,
"loss": 1.6667,
"step": 41000
},
{
"epoch": 2.28,
"learning_rate": 1.1946156103286385e-05,
"loss": 1.6594,
"step": 41500
},
{
"epoch": 2.31,
"learning_rate": 1.148767605633803e-05,
"loss": 1.659,
"step": 42000
},
{
"epoch": 2.34,
"learning_rate": 1.1029196009389672e-05,
"loss": 1.6619,
"step": 42500
},
{
"epoch": 2.37,
"learning_rate": 1.0570715962441315e-05,
"loss": 1.6518,
"step": 43000
},
{
"epoch": 2.39,
"learning_rate": 1.0112235915492959e-05,
"loss": 1.6567,
"step": 43500
},
{
"epoch": 2.42,
"learning_rate": 9.653755868544602e-06,
"loss": 1.6555,
"step": 44000
},
{
"epoch": 2.45,
"learning_rate": 9.195275821596244e-06,
"loss": 1.6434,
"step": 44500
},
{
"epoch": 2.48,
"learning_rate": 8.736795774647887e-06,
"loss": 1.6505,
"step": 45000
},
{
"epoch": 2.5,
"learning_rate": 8.278315727699531e-06,
"loss": 1.6449,
"step": 45500
},
{
"epoch": 2.53,
"learning_rate": 7.819835680751174e-06,
"loss": 1.6454,
"step": 46000
},
{
"epoch": 2.56,
"learning_rate": 7.361355633802818e-06,
"loss": 1.631,
"step": 46500
},
{
"epoch": 2.59,
"learning_rate": 6.9028755868544595e-06,
"loss": 1.6445,
"step": 47000
},
{
"epoch": 2.61,
"learning_rate": 6.444395539906104e-06,
"loss": 1.6401,
"step": 47500
},
{
"epoch": 2.64,
"learning_rate": 5.9859154929577465e-06,
"loss": 1.6463,
"step": 48000
},
{
"epoch": 2.67,
"learning_rate": 5.52743544600939e-06,
"loss": 1.6307,
"step": 48500
},
{
"epoch": 2.7,
"learning_rate": 5.068955399061033e-06,
"loss": 1.6453,
"step": 49000
},
{
"epoch": 2.72,
"learning_rate": 4.610475352112676e-06,
"loss": 1.6415,
"step": 49500
},
{
"epoch": 2.75,
"learning_rate": 4.15199530516432e-06,
"loss": 1.628,
"step": 50000
},
{
"epoch": 2.78,
"learning_rate": 3.693515258215963e-06,
"loss": 1.6362,
"step": 50500
},
{
"epoch": 2.81,
"learning_rate": 3.2350352112676055e-06,
"loss": 1.6241,
"step": 51000
},
{
"epoch": 2.83,
"learning_rate": 2.776555164319249e-06,
"loss": 1.6352,
"step": 51500
},
{
"epoch": 2.86,
"learning_rate": 2.318075117370892e-06,
"loss": 1.6413,
"step": 52000
},
{
"epoch": 2.89,
"learning_rate": 1.8595950704225353e-06,
"loss": 1.6281,
"step": 52500
},
{
"epoch": 2.92,
"learning_rate": 1.4011150234741783e-06,
"loss": 1.6331,
"step": 53000
},
{
"epoch": 2.94,
"learning_rate": 9.426349765258218e-07,
"loss": 1.6384,
"step": 53500
},
{
"epoch": 2.97,
"learning_rate": 4.841549295774648e-07,
"loss": 1.6336,
"step": 54000
},
{
"epoch": 3.0,
"learning_rate": 2.5674882629107985e-08,
"loss": 1.6306,
"step": 54500
},
{
"epoch": 3.0,
"step": 54528,
"total_flos": 1.722156986294784e+17,
"train_loss": 1.8530674359188393,
"train_runtime": 31139.3919,
"train_samples_per_second": 21.012,
"train_steps_per_second": 1.751
}
],
"max_steps": 54528,
"num_train_epochs": 3,
"total_flos": 1.722156986294784e+17,
"trial_name": null,
"trial_params": null
}