1046 lines
28 KiB
JSON
1046 lines
28 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.39184952978056425,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74259386062622,
|
|
"epoch": 0.003918495297805642,
|
|
"grad_norm": 5.34375,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7559,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 9174.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742582988739013,
|
|
"epoch": 0.007836990595611285,
|
|
"grad_norm": 5.09375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7343,
|
|
"mean_token_accuracy": 0.00011074197245761753,
|
|
"num_tokens": 17790.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742611026763916,
|
|
"epoch": 0.011755485893416929,
|
|
"grad_norm": 5.03125,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.6946,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 25850.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742605590820313,
|
|
"epoch": 0.01567398119122257,
|
|
"grad_norm": 5.1875,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.6748,
|
|
"mean_token_accuracy": 0.001098086079582572,
|
|
"num_tokens": 35194.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742543792724609,
|
|
"epoch": 0.019592476489028215,
|
|
"grad_norm": 4.53125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.5631,
|
|
"mean_token_accuracy": 0.008560778619721533,
|
|
"num_tokens": 44218.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.74220838546753,
|
|
"epoch": 0.023510971786833857,
|
|
"grad_norm": 4.09375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.4525,
|
|
"mean_token_accuracy": 0.03252750001847744,
|
|
"num_tokens": 53397.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.740808391571045,
|
|
"epoch": 0.0274294670846395,
|
|
"grad_norm": 3.421875,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.3079,
|
|
"mean_token_accuracy": 0.0437689334154129,
|
|
"num_tokens": 62749.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.736591911315918,
|
|
"epoch": 0.03134796238244514,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1743,
|
|
"mean_token_accuracy": 0.043372325226664545,
|
|
"num_tokens": 71721.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.729655265808105,
|
|
"epoch": 0.03526645768025078,
|
|
"grad_norm": 2.4375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.0405,
|
|
"mean_token_accuracy": 0.046825164929032326,
|
|
"num_tokens": 79844.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.72356595993042,
|
|
"epoch": 0.03918495297805643,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.9917,
|
|
"mean_token_accuracy": 0.04553030654788017,
|
|
"num_tokens": 88866.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.719326305389405,
|
|
"epoch": 0.04310344827586207,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.9336,
|
|
"mean_token_accuracy": 0.044035762920975684,
|
|
"num_tokens": 97918.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.715408420562744,
|
|
"epoch": 0.047021943573667714,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.8702,
|
|
"mean_token_accuracy": 0.04544390179216862,
|
|
"num_tokens": 107043.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.710628414154053,
|
|
"epoch": 0.050940438871473356,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8037,
|
|
"mean_token_accuracy": 0.04611521698534489,
|
|
"num_tokens": 116000.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.704075908660888,
|
|
"epoch": 0.054858934169279,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.7686,
|
|
"mean_token_accuracy": 0.04174055978655815,
|
|
"num_tokens": 124559.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.700677013397216,
|
|
"epoch": 0.05877742946708464,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.6568,
|
|
"mean_token_accuracy": 0.04639651067554951,
|
|
"num_tokens": 132868.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.695802783966064,
|
|
"epoch": 0.06269592476489028,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6523,
|
|
"mean_token_accuracy": 0.04210274666547775,
|
|
"num_tokens": 141286.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.688793563842774,
|
|
"epoch": 0.06661442006269593,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.5499,
|
|
"mean_token_accuracy": 0.05031363070011139,
|
|
"num_tokens": 150406.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.679227447509765,
|
|
"epoch": 0.07053291536050156,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4859,
|
|
"mean_token_accuracy": 0.056176548823714256,
|
|
"num_tokens": 158770.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.663717269897461,
|
|
"epoch": 0.07445141065830721,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.4023,
|
|
"mean_token_accuracy": 0.06409536376595497,
|
|
"num_tokens": 167763.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.639647197723388,
|
|
"epoch": 0.07836990595611286,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.286,
|
|
"mean_token_accuracy": 0.06052683852612972,
|
|
"num_tokens": 177306.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.605971527099609,
|
|
"epoch": 0.0822884012539185,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.1915,
|
|
"mean_token_accuracy": 0.06200857125222683,
|
|
"num_tokens": 186014.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.545482730865478,
|
|
"epoch": 0.08620689655172414,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.0547,
|
|
"mean_token_accuracy": 0.06552336104214192,
|
|
"num_tokens": 194122.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.471019172668457,
|
|
"epoch": 0.09012539184952978,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9639,
|
|
"mean_token_accuracy": 0.061758537963032725,
|
|
"num_tokens": 203097.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.420811939239503,
|
|
"epoch": 0.09404388714733543,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8529,
|
|
"mean_token_accuracy": 0.06423232220113277,
|
|
"num_tokens": 211413.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.383054542541505,
|
|
"epoch": 0.09796238244514106,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7779,
|
|
"mean_token_accuracy": 0.06270350776612758,
|
|
"num_tokens": 220550.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.31447114944458,
|
|
"epoch": 0.10188087774294671,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.6079,
|
|
"mean_token_accuracy": 0.06446850374341011,
|
|
"num_tokens": 229197.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.19693603515625,
|
|
"epoch": 0.10579937304075235,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.5715,
|
|
"mean_token_accuracy": 0.05803979113698006,
|
|
"num_tokens": 238479.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.129131984710693,
|
|
"epoch": 0.109717868338558,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.4073,
|
|
"mean_token_accuracy": 0.06696602776646614,
|
|
"num_tokens": 246318.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.014921092987061,
|
|
"epoch": 0.11363636363636363,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.3497,
|
|
"mean_token_accuracy": 0.06019376777112484,
|
|
"num_tokens": 254783.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.837921333312988,
|
|
"epoch": 0.11755485893416928,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.1727,
|
|
"mean_token_accuracy": 0.06220055930316448,
|
|
"num_tokens": 263416.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.704282855987548,
|
|
"epoch": 0.12147335423197492,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.1265,
|
|
"mean_token_accuracy": 0.06537272743880748,
|
|
"num_tokens": 271930.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.534114360809326,
|
|
"epoch": 0.12539184952978055,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.9819,
|
|
"mean_token_accuracy": 0.06310658231377601,
|
|
"num_tokens": 280546.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.289366626739502,
|
|
"epoch": 0.12931034482758622,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.9162,
|
|
"mean_token_accuracy": 0.0649850957095623,
|
|
"num_tokens": 288813.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.108893775939942,
|
|
"epoch": 0.13322884012539185,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.8379,
|
|
"mean_token_accuracy": 0.06351686343550682,
|
|
"num_tokens": 297966.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.892046165466308,
|
|
"epoch": 0.1371473354231975,
|
|
"grad_norm": 0.859375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.8598,
|
|
"mean_token_accuracy": 0.06292634233832359,
|
|
"num_tokens": 307135.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.728424072265625,
|
|
"epoch": 0.14106583072100312,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.7429,
|
|
"mean_token_accuracy": 0.06494242362678052,
|
|
"num_tokens": 315546.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.619221019744874,
|
|
"epoch": 0.14498432601880878,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.7168,
|
|
"mean_token_accuracy": 0.06593996658921242,
|
|
"num_tokens": 323930.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.409439945220948,
|
|
"epoch": 0.14890282131661442,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.5689,
|
|
"mean_token_accuracy": 0.0733262188732624,
|
|
"num_tokens": 332419.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.326526832580566,
|
|
"epoch": 0.15282131661442006,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.654,
|
|
"mean_token_accuracy": 0.0696730900555849,
|
|
"num_tokens": 341362.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.3056715965271,
|
|
"epoch": 0.15673981191222572,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.5645,
|
|
"mean_token_accuracy": 0.07249549068510533,
|
|
"num_tokens": 349395.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.24736156463623,
|
|
"epoch": 0.16065830721003135,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.5705,
|
|
"mean_token_accuracy": 0.06626113764941692,
|
|
"num_tokens": 358413.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.157195377349854,
|
|
"epoch": 0.164576802507837,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.6532,
|
|
"mean_token_accuracy": 0.07263443917036057,
|
|
"num_tokens": 366489.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.1927809715271,
|
|
"epoch": 0.16849529780564262,
|
|
"grad_norm": 0.76953125,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.5231,
|
|
"mean_token_accuracy": 0.06818853206932544,
|
|
"num_tokens": 375335.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.176179027557373,
|
|
"epoch": 0.1724137931034483,
|
|
"grad_norm": 0.91015625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.5638,
|
|
"mean_token_accuracy": 0.07268193662166596,
|
|
"num_tokens": 384276.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.140285873413086,
|
|
"epoch": 0.17633228840125392,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.5991,
|
|
"mean_token_accuracy": 0.06890415549278259,
|
|
"num_tokens": 393571.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.1433687210083,
|
|
"epoch": 0.18025078369905956,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.5228,
|
|
"mean_token_accuracy": 0.07579943463206291,
|
|
"num_tokens": 401865.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.072039985656739,
|
|
"epoch": 0.1841692789968652,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.4982,
|
|
"mean_token_accuracy": 0.07117367573082448,
|
|
"num_tokens": 410518.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.076621770858765,
|
|
"epoch": 0.18808777429467086,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.5051,
|
|
"mean_token_accuracy": 0.06969159096479416,
|
|
"num_tokens": 419769.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.030782747268677,
|
|
"epoch": 0.1920062695924765,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.4391,
|
|
"mean_token_accuracy": 0.07655345052480697,
|
|
"num_tokens": 428204.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.057575464248657,
|
|
"epoch": 0.19592476489028213,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.5263,
|
|
"mean_token_accuracy": 0.07384133897721767,
|
|
"num_tokens": 436594.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.054854202270509,
|
|
"epoch": 0.19984326018808776,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3139,
|
|
"mean_token_accuracy": 0.07670355625450612,
|
|
"num_tokens": 444645.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.953717613220215,
|
|
"epoch": 0.20376175548589343,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.3561,
|
|
"mean_token_accuracy": 0.07812742255628109,
|
|
"num_tokens": 453016.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.9555809020996096,
|
|
"epoch": 0.20768025078369906,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.4192,
|
|
"mean_token_accuracy": 0.07524299174547196,
|
|
"num_tokens": 462116.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.957012748718261,
|
|
"epoch": 0.2115987460815047,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.3694,
|
|
"mean_token_accuracy": 0.07774586156010628,
|
|
"num_tokens": 470807.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.9492302417755125,
|
|
"epoch": 0.21551724137931033,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.4561,
|
|
"mean_token_accuracy": 0.07432891242206097,
|
|
"num_tokens": 479592.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.016344738006591,
|
|
"epoch": 0.219435736677116,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.3803,
|
|
"mean_token_accuracy": 0.07607875801622868,
|
|
"num_tokens": 488107.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.870184087753296,
|
|
"epoch": 0.22335423197492163,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.3792,
|
|
"mean_token_accuracy": 0.07687489986419678,
|
|
"num_tokens": 496775.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.973387956619263,
|
|
"epoch": 0.22727272727272727,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.4096,
|
|
"mean_token_accuracy": 0.07644881159067154,
|
|
"num_tokens": 505415.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.946991682052612,
|
|
"epoch": 0.23119122257053293,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.3915,
|
|
"mean_token_accuracy": 0.07934824042022229,
|
|
"num_tokens": 514447.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.891976928710937,
|
|
"epoch": 0.23510971786833856,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.381,
|
|
"mean_token_accuracy": 0.07889064475893974,
|
|
"num_tokens": 522998.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.92128734588623,
|
|
"epoch": 0.2390282131661442,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.3459,
|
|
"mean_token_accuracy": 0.08099666759371757,
|
|
"num_tokens": 531542.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.863611841201783,
|
|
"epoch": 0.24294670846394983,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2927,
|
|
"mean_token_accuracy": 0.07772218361496926,
|
|
"num_tokens": 540143.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.924414491653442,
|
|
"epoch": 0.2468652037617555,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.3679,
|
|
"mean_token_accuracy": 0.08110597431659698,
|
|
"num_tokens": 549156.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.803961706161499,
|
|
"epoch": 0.2507836990595611,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.2252,
|
|
"mean_token_accuracy": 0.08031321838498115,
|
|
"num_tokens": 557522.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.805857229232788,
|
|
"epoch": 0.2547021943573668,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.2879,
|
|
"mean_token_accuracy": 0.08254150673747063,
|
|
"num_tokens": 566650.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.8912803649902346,
|
|
"epoch": 0.25862068965517243,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.2408,
|
|
"mean_token_accuracy": 0.08324127942323685,
|
|
"num_tokens": 576219.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.849798917770386,
|
|
"epoch": 0.26253918495297807,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.3416,
|
|
"mean_token_accuracy": 0.08195492923259735,
|
|
"num_tokens": 584304.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.75479941368103,
|
|
"epoch": 0.2664576802507837,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.2574,
|
|
"mean_token_accuracy": 0.08678329288959503,
|
|
"num_tokens": 593163.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.7592004299163815,
|
|
"epoch": 0.27037617554858934,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.1773,
|
|
"mean_token_accuracy": 0.08335737958550453,
|
|
"num_tokens": 602077.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.792924070358277,
|
|
"epoch": 0.274294670846395,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.2812,
|
|
"mean_token_accuracy": 0.08297529369592667,
|
|
"num_tokens": 610009.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.816137742996216,
|
|
"epoch": 0.2782131661442006,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.3247,
|
|
"mean_token_accuracy": 0.07837995067238808,
|
|
"num_tokens": 619282.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.692404460906983,
|
|
"epoch": 0.28213166144200624,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.2457,
|
|
"mean_token_accuracy": 0.08888485580682755,
|
|
"num_tokens": 628138.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.819083976745605,
|
|
"epoch": 0.28605015673981193,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.1219,
|
|
"mean_token_accuracy": 0.09141435995697975,
|
|
"num_tokens": 637021.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.734615421295166,
|
|
"epoch": 0.28996865203761757,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.2772,
|
|
"mean_token_accuracy": 0.08298610299825668,
|
|
"num_tokens": 646703.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.772846841812134,
|
|
"epoch": 0.2938871473354232,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.1535,
|
|
"mean_token_accuracy": 0.08612042218446732,
|
|
"num_tokens": 655616.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.662965631484985,
|
|
"epoch": 0.29780564263322884,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.1548,
|
|
"mean_token_accuracy": 0.09385274946689606,
|
|
"num_tokens": 663783.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.72055869102478,
|
|
"epoch": 0.3017241379310345,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.1499,
|
|
"mean_token_accuracy": 0.09183024540543556,
|
|
"num_tokens": 671855.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.686996603012085,
|
|
"epoch": 0.3056426332288401,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.1691,
|
|
"mean_token_accuracy": 0.08594455868005753,
|
|
"num_tokens": 680981.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.710857486724853,
|
|
"epoch": 0.30956112852664575,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.1032,
|
|
"mean_token_accuracy": 0.09159721955657005,
|
|
"num_tokens": 689294.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.561884641647339,
|
|
"epoch": 0.31347962382445144,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.055,
|
|
"mean_token_accuracy": 0.09060370847582817,
|
|
"num_tokens": 697703.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.728968143463135,
|
|
"epoch": 0.31739811912225707,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.1907,
|
|
"mean_token_accuracy": 0.08998276144266129,
|
|
"num_tokens": 706792.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.705010414123535,
|
|
"epoch": 0.3213166144200627,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.0954,
|
|
"mean_token_accuracy": 0.09226684123277665,
|
|
"num_tokens": 715864.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.541502141952515,
|
|
"epoch": 0.32523510971786834,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.1394,
|
|
"mean_token_accuracy": 0.0916991800069809,
|
|
"num_tokens": 723921.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.679799509048462,
|
|
"epoch": 0.329153605015674,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.1503,
|
|
"mean_token_accuracy": 0.08857458978891372,
|
|
"num_tokens": 732797.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.634342527389526,
|
|
"epoch": 0.3330721003134796,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.0126,
|
|
"mean_token_accuracy": 0.0977866668254137,
|
|
"num_tokens": 741248.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.6327766418457035,
|
|
"epoch": 0.33699059561128525,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.0961,
|
|
"mean_token_accuracy": 0.09004503935575485,
|
|
"num_tokens": 749766.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.5497727394104,
|
|
"epoch": 0.3409090909090909,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.067,
|
|
"mean_token_accuracy": 0.09081325829029083,
|
|
"num_tokens": 758695.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.6185346126556395,
|
|
"epoch": 0.3448275862068966,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.0591,
|
|
"mean_token_accuracy": 0.08728412315249442,
|
|
"num_tokens": 767624.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.5867267608642575,
|
|
"epoch": 0.3487460815047022,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.0721,
|
|
"mean_token_accuracy": 0.09007409512996674,
|
|
"num_tokens": 776616.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.504705095291138,
|
|
"epoch": 0.35266457680250785,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.0826,
|
|
"mean_token_accuracy": 0.09070678725838661,
|
|
"num_tokens": 786147.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.628587007522583,
|
|
"epoch": 0.3565830721003135,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.1945,
|
|
"mean_token_accuracy": 0.09022908806800842,
|
|
"num_tokens": 795213.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.5562244892120365,
|
|
"epoch": 0.3605015673981191,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.0011,
|
|
"mean_token_accuracy": 0.09484408944845199,
|
|
"num_tokens": 803118.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.582311534881592,
|
|
"epoch": 0.36442006269592475,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.9727,
|
|
"mean_token_accuracy": 0.10151090547442436,
|
|
"num_tokens": 811358.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.470248889923096,
|
|
"epoch": 0.3683385579937304,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.971,
|
|
"mean_token_accuracy": 0.09536491632461548,
|
|
"num_tokens": 819653.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.467239570617676,
|
|
"epoch": 0.3722570532915361,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.9984,
|
|
"mean_token_accuracy": 0.0897679604589939,
|
|
"num_tokens": 828462.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.6044769287109375,
|
|
"epoch": 0.3761755485893417,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.0876,
|
|
"mean_token_accuracy": 0.09487015083432197,
|
|
"num_tokens": 836204.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.44920597076416,
|
|
"epoch": 0.38009404388714735,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.9686,
|
|
"mean_token_accuracy": 0.09797411188483238,
|
|
"num_tokens": 845032.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.5414040088653564,
|
|
"epoch": 0.384012539184953,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.9796,
|
|
"mean_token_accuracy": 0.09801916852593422,
|
|
"num_tokens": 853324.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.508725357055664,
|
|
"epoch": 0.3879310344827586,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.9449,
|
|
"mean_token_accuracy": 0.10975441187620164,
|
|
"num_tokens": 861640.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.553517627716064,
|
|
"epoch": 0.39184952978056425,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.116,
|
|
"mean_token_accuracy": 0.09244248941540718,
|
|
"num_tokens": 870758.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.39184952978056425,
|
|
"eval_entropy": 7.418469055678493,
|
|
"eval_loss": 7.0935492515563965,
|
|
"eval_mean_token_accuracy": 0.09673699742370798,
|
|
"eval_num_tokens": 870758.0,
|
|
"eval_runtime": 2.2092,
|
|
"eval_samples_per_second": 1865.802,
|
|
"eval_steps_per_second": 233.565,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 12750,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 10,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 171439857008640.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|