775 lines
16 KiB
JSON
775 lines
16 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"global_step": 62988,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.998412396011939e-05,
|
|
"loss": 1.6197,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.9968247920238778e-05,
|
|
"loss": 1.5231,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 1.9952371880358167e-05,
|
|
"loss": 1.508,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 1.9936495840477554e-05,
|
|
"loss": 1.4933,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 1.992061980059694e-05,
|
|
"loss": 1.4876,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 1.990474376071633e-05,
|
|
"loss": 1.4779,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.9888867720835716e-05,
|
|
"loss": 1.4754,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1.9872991680955105e-05,
|
|
"loss": 1.4712,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.9857115641074492e-05,
|
|
"loss": 1.4677,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"learning_rate": 1.984123960119388e-05,
|
|
"loss": 1.4622,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.9825363561313268e-05,
|
|
"loss": 1.4638,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9809487521432654e-05,
|
|
"loss": 1.4594,
|
|
"step": 6000
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.9793611481552044e-05,
|
|
"loss": 1.4514,
|
|
"step": 6500
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.977773544167143e-05,
|
|
"loss": 1.4568,
|
|
"step": 7000
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.976185940179082e-05,
|
|
"loss": 1.4524,
|
|
"step": 7500
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9745983361910206e-05,
|
|
"loss": 1.4469,
|
|
"step": 8000
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9730107322029596e-05,
|
|
"loss": 1.4478,
|
|
"step": 8500
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.9714231282148982e-05,
|
|
"loss": 1.4443,
|
|
"step": 9000
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.969835524226837e-05,
|
|
"loss": 1.442,
|
|
"step": 9500
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.9682479202387758e-05,
|
|
"loss": 1.4395,
|
|
"step": 10000
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.9666603162507144e-05,
|
|
"loss": 1.4377,
|
|
"step": 10500
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.9650727122626534e-05,
|
|
"loss": 1.4381,
|
|
"step": 11000
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.963485108274592e-05,
|
|
"loss": 1.4337,
|
|
"step": 11500
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"learning_rate": 1.961897504286531e-05,
|
|
"loss": 1.4338,
|
|
"step": 12000
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.96030990029847e-05,
|
|
"loss": 1.4347,
|
|
"step": 12500
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9587222963104086e-05,
|
|
"loss": 1.436,
|
|
"step": 13000
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9571346923223472e-05,
|
|
"loss": 1.4333,
|
|
"step": 13500
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.955547088334286e-05,
|
|
"loss": 1.425,
|
|
"step": 14000
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.9539594843462248e-05,
|
|
"loss": 1.4254,
|
|
"step": 14500
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.9523718803581638e-05,
|
|
"loss": 1.4285,
|
|
"step": 15000
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.9507842763701024e-05,
|
|
"loss": 1.4226,
|
|
"step": 15500
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.9491966723820414e-05,
|
|
"loss": 1.4261,
|
|
"step": 16000
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.94760906839398e-05,
|
|
"loss": 1.4233,
|
|
"step": 16500
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.9460214644059186e-05,
|
|
"loss": 1.4206,
|
|
"step": 17000
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.9444338604178576e-05,
|
|
"loss": 1.4224,
|
|
"step": 17500
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.9428462564297962e-05,
|
|
"loss": 1.414,
|
|
"step": 18000
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.9412586524417352e-05,
|
|
"loss": 1.424,
|
|
"step": 18500
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.9396710484536738e-05,
|
|
"loss": 1.4181,
|
|
"step": 19000
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9380834444656128e-05,
|
|
"loss": 1.4168,
|
|
"step": 19500
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.9364958404775514e-05,
|
|
"loss": 1.4171,
|
|
"step": 20000
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.9349082364894904e-05,
|
|
"loss": 1.4116,
|
|
"step": 20500
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.933320632501429e-05,
|
|
"loss": 1.4055,
|
|
"step": 21000
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.9317330285133676e-05,
|
|
"loss": 1.4118,
|
|
"step": 21500
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9301454245253066e-05,
|
|
"loss": 1.4134,
|
|
"step": 22000
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.9285578205372452e-05,
|
|
"loss": 1.4205,
|
|
"step": 22500
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.9269702165491842e-05,
|
|
"loss": 1.4088,
|
|
"step": 23000
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.925382612561123e-05,
|
|
"loss": 1.4104,
|
|
"step": 23500
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.9237950085730618e-05,
|
|
"loss": 1.4109,
|
|
"step": 24000
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.9222074045850004e-05,
|
|
"loss": 1.4073,
|
|
"step": 24500
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.920619800596939e-05,
|
|
"loss": 1.4048,
|
|
"step": 25000
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.919032196608878e-05,
|
|
"loss": 1.4056,
|
|
"step": 25500
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.9174445926208167e-05,
|
|
"loss": 1.4045,
|
|
"step": 26000
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.9158569886327556e-05,
|
|
"loss": 1.4042,
|
|
"step": 26500
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.9142693846446946e-05,
|
|
"loss": 1.4001,
|
|
"step": 27000
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.9126817806566332e-05,
|
|
"loss": 1.4065,
|
|
"step": 27500
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.911094176668572e-05,
|
|
"loss": 1.4041,
|
|
"step": 28000
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.9095065726805108e-05,
|
|
"loss": 1.3994,
|
|
"step": 28500
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.9079189686924494e-05,
|
|
"loss": 1.3991,
|
|
"step": 29000
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.9063313647043884e-05,
|
|
"loss": 1.3956,
|
|
"step": 29500
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.904743760716327e-05,
|
|
"loss": 1.3995,
|
|
"step": 30000
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.903156156728266e-05,
|
|
"loss": 1.4011,
|
|
"step": 30500
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9015685527402046e-05,
|
|
"loss": 1.3964,
|
|
"step": 31000
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.8999809487521436e-05,
|
|
"loss": 1.3925,
|
|
"step": 31500
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"learning_rate": 1.8983933447640822e-05,
|
|
"loss": 1.3875,
|
|
"step": 32000
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.896805740776021e-05,
|
|
"loss": 1.3945,
|
|
"step": 32500
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.8952181367879598e-05,
|
|
"loss": 1.3919,
|
|
"step": 33000
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.8936305327998985e-05,
|
|
"loss": 1.3942,
|
|
"step": 33500
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.8920429288118374e-05,
|
|
"loss": 1.3906,
|
|
"step": 34000
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.890455324823776e-05,
|
|
"loss": 1.3916,
|
|
"step": 34500
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.888867720835715e-05,
|
|
"loss": 1.3871,
|
|
"step": 35000
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.8872801168476536e-05,
|
|
"loss": 1.3824,
|
|
"step": 35500
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.8856925128595923e-05,
|
|
"loss": 1.3919,
|
|
"step": 36000
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.8841049088715312e-05,
|
|
"loss": 1.3885,
|
|
"step": 36500
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.88251730488347e-05,
|
|
"loss": 1.3837,
|
|
"step": 37000
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.880929700895409e-05,
|
|
"loss": 1.389,
|
|
"step": 37500
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.8793420969073475e-05,
|
|
"loss": 1.3841,
|
|
"step": 38000
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8777544929192864e-05,
|
|
"loss": 1.3846,
|
|
"step": 38500
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.876166888931225e-05,
|
|
"loss": 1.3807,
|
|
"step": 39000
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.874579284943164e-05,
|
|
"loss": 1.382,
|
|
"step": 39500
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8729916809551027e-05,
|
|
"loss": 1.3831,
|
|
"step": 40000
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8714040769670413e-05,
|
|
"loss": 1.3814,
|
|
"step": 40500
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.8698164729789803e-05,
|
|
"loss": 1.3858,
|
|
"step": 41000
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"learning_rate": 1.8682288689909192e-05,
|
|
"loss": 1.3846,
|
|
"step": 41500
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.866641265002858e-05,
|
|
"loss": 1.3831,
|
|
"step": 42000
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.8650536610147968e-05,
|
|
"loss": 1.3738,
|
|
"step": 42500
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.8634660570267354e-05,
|
|
"loss": 1.3798,
|
|
"step": 43000
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.861878453038674e-05,
|
|
"loss": 1.3755,
|
|
"step": 43500
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.8602908490506127e-05,
|
|
"loss": 1.3753,
|
|
"step": 44000
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.8587032450625517e-05,
|
|
"loss": 1.3745,
|
|
"step": 44500
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.8571156410744906e-05,
|
|
"loss": 1.3814,
|
|
"step": 45000
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"learning_rate": 1.8555280370864293e-05,
|
|
"loss": 1.3759,
|
|
"step": 45500
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"learning_rate": 1.8539404330983682e-05,
|
|
"loss": 1.3722,
|
|
"step": 46000
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.852352829110307e-05,
|
|
"loss": 1.3768,
|
|
"step": 46500
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.8507652251222455e-05,
|
|
"loss": 1.3724,
|
|
"step": 47000
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.8491776211341845e-05,
|
|
"loss": 1.3745,
|
|
"step": 47500
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"learning_rate": 1.847590017146123e-05,
|
|
"loss": 1.3727,
|
|
"step": 48000
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"learning_rate": 1.846002413158062e-05,
|
|
"loss": 1.3703,
|
|
"step": 48500
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.8444148091700007e-05,
|
|
"loss": 1.3673,
|
|
"step": 49000
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.8428272051819397e-05,
|
|
"loss": 1.3686,
|
|
"step": 49500
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.8412396011938786e-05,
|
|
"loss": 1.3663,
|
|
"step": 50000
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.8396519972058173e-05,
|
|
"loss": 1.3665,
|
|
"step": 50500
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"learning_rate": 1.838064393217756e-05,
|
|
"loss": 1.3701,
|
|
"step": 51000
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.8364767892296945e-05,
|
|
"loss": 1.3678,
|
|
"step": 51500
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.8348891852416335e-05,
|
|
"loss": 1.3704,
|
|
"step": 52000
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.833301581253572e-05,
|
|
"loss": 1.3657,
|
|
"step": 52500
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"learning_rate": 1.831713977265511e-05,
|
|
"loss": 1.3683,
|
|
"step": 53000
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.83012637327745e-05,
|
|
"loss": 1.3637,
|
|
"step": 53500
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.8285387692893887e-05,
|
|
"loss": 1.3636,
|
|
"step": 54000
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.8269511653013273e-05,
|
|
"loss": 1.3626,
|
|
"step": 54500
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.825363561313266e-05,
|
|
"loss": 1.3682,
|
|
"step": 55000
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"learning_rate": 1.823775957325205e-05,
|
|
"loss": 1.3653,
|
|
"step": 55500
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.822188353337144e-05,
|
|
"loss": 1.3654,
|
|
"step": 56000
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.8206007493490825e-05,
|
|
"loss": 1.3609,
|
|
"step": 56500
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.8190131453610215e-05,
|
|
"loss": 1.3616,
|
|
"step": 57000
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"learning_rate": 1.81742554137296e-05,
|
|
"loss": 1.362,
|
|
"step": 57500
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"learning_rate": 1.815837937384899e-05,
|
|
"loss": 1.364,
|
|
"step": 58000
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.8142503333968377e-05,
|
|
"loss": 1.3573,
|
|
"step": 58500
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.8126627294087763e-05,
|
|
"loss": 1.3543,
|
|
"step": 59000
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.8110751254207153e-05,
|
|
"loss": 1.3592,
|
|
"step": 59500
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"learning_rate": 1.809487521432654e-05,
|
|
"loss": 1.3598,
|
|
"step": 60000
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.807899917444593e-05,
|
|
"loss": 1.3561,
|
|
"step": 60500
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"learning_rate": 1.8063123134565315e-05,
|
|
"loss": 1.3575,
|
|
"step": 61000
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.8047247094684705e-05,
|
|
"loss": 1.3614,
|
|
"step": 61500
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.803137105480409e-05,
|
|
"loss": 1.3575,
|
|
"step": 62000
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.8015495014923477e-05,
|
|
"loss": 1.3558,
|
|
"step": 62500
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_loss": 1.3514442443847656,
|
|
"eval_runtime": 2650.2414,
|
|
"eval_samples_per_second": 237.786,
|
|
"eval_steps_per_second": 2.973,
|
|
"step": 62988
|
|
}
|
|
],
|
|
"max_steps": 629880,
|
|
"num_train_epochs": 10,
|
|
"total_flos": 2.3398437545501e+18,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|