1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.029265437518290898,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.742143154144287,
|
|
"epoch": 0.000292654375182909,
|
|
"grad_norm": 5.34375,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7851,
|
|
"mean_token_accuracy": 7.092198356986045e-05,
|
|
"num_tokens": 10253.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.742099285125732,
|
|
"epoch": 0.000585308750365818,
|
|
"grad_norm": 5.09375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7585,
|
|
"mean_token_accuracy": 7.072135922499001e-05,
|
|
"num_tokens": 19586.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742091464996339,
|
|
"epoch": 0.000877963125548727,
|
|
"grad_norm": 5.71875,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.7338,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 28743.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742110919952392,
|
|
"epoch": 0.001170617500731636,
|
|
"grad_norm": 4.75,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.702,
|
|
"mean_token_accuracy": 7.518797065131366e-05,
|
|
"num_tokens": 39263.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.7420015335083,
|
|
"epoch": 0.0014632718759145448,
|
|
"grad_norm": 5.0625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.6176,
|
|
"mean_token_accuracy": 0.00650847667711787,
|
|
"num_tokens": 50195.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.741524696350098,
|
|
"epoch": 0.001755926251097454,
|
|
"grad_norm": 4.5,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.4398,
|
|
"mean_token_accuracy": 0.04643158838152885,
|
|
"num_tokens": 58982.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.739701747894287,
|
|
"epoch": 0.002048580626280363,
|
|
"grad_norm": 3.375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.3385,
|
|
"mean_token_accuracy": 0.04858696423470974,
|
|
"num_tokens": 68619.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.733176803588867,
|
|
"epoch": 0.002341235001463272,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1755,
|
|
"mean_token_accuracy": 0.054676258191466334,
|
|
"num_tokens": 78276.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.719074249267578,
|
|
"epoch": 0.0026338893766461808,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.085,
|
|
"mean_token_accuracy": 0.05106211043894291,
|
|
"num_tokens": 87977.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.702382278442382,
|
|
"epoch": 0.0029265437518290896,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 10.007,
|
|
"mean_token_accuracy": 0.05418388731777668,
|
|
"num_tokens": 97168.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.688697719573975,
|
|
"epoch": 0.003219198127011999,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.978,
|
|
"mean_token_accuracy": 0.0560054711997509,
|
|
"num_tokens": 108162.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.677631187438966,
|
|
"epoch": 0.003511852502194908,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.8907,
|
|
"mean_token_accuracy": 0.05185089595615864,
|
|
"num_tokens": 117963.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.672104358673096,
|
|
"epoch": 0.0038045068773778167,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8844,
|
|
"mean_token_accuracy": 0.05049301944673061,
|
|
"num_tokens": 128067.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.665418434143067,
|
|
"epoch": 0.004097161252560726,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.8046,
|
|
"mean_token_accuracy": 0.05261277854442596,
|
|
"num_tokens": 138347.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.655702877044678,
|
|
"epoch": 0.004389815627743635,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.7713,
|
|
"mean_token_accuracy": 0.05200894549489021,
|
|
"num_tokens": 148883.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.648012065887452,
|
|
"epoch": 0.004682470002926544,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.7304,
|
|
"mean_token_accuracy": 0.053603590652346614,
|
|
"num_tokens": 158416.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.638986778259277,
|
|
"epoch": 0.004975124378109453,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.624,
|
|
"mean_token_accuracy": 0.056516367569565774,
|
|
"num_tokens": 168380.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.622650527954102,
|
|
"epoch": 0.0052677787532923615,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.5594,
|
|
"mean_token_accuracy": 0.058524899929761884,
|
|
"num_tokens": 178623.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.598152160644531,
|
|
"epoch": 0.00556043312847527,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.5209,
|
|
"mean_token_accuracy": 0.05661297850310802,
|
|
"num_tokens": 189058.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.577679920196534,
|
|
"epoch": 0.005853087503658179,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.4548,
|
|
"mean_token_accuracy": 0.056520416587591174,
|
|
"num_tokens": 199465.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.56387767791748,
|
|
"epoch": 0.006145741878841089,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.4,
|
|
"mean_token_accuracy": 0.054618718847632405,
|
|
"num_tokens": 208712.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.543546962738038,
|
|
"epoch": 0.006438396254023998,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.3224,
|
|
"mean_token_accuracy": 0.05899658054113388,
|
|
"num_tokens": 218557.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.48940305709839,
|
|
"epoch": 0.006731050629206907,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 9.2068,
|
|
"mean_token_accuracy": 0.06870373338460922,
|
|
"num_tokens": 228961.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.402587509155273,
|
|
"epoch": 0.007023705004389816,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 9.1227,
|
|
"mean_token_accuracy": 0.06810536533594132,
|
|
"num_tokens": 238987.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.348681640625,
|
|
"epoch": 0.0073163593795727245,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 9.0633,
|
|
"mean_token_accuracy": 0.0605758685618639,
|
|
"num_tokens": 249600.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.30824089050293,
|
|
"epoch": 0.007609013754755633,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.8758,
|
|
"mean_token_accuracy": 0.07409285828471183,
|
|
"num_tokens": 259215.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.212008476257324,
|
|
"epoch": 0.007901668129938543,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.8303,
|
|
"mean_token_accuracy": 0.06557713933289051,
|
|
"num_tokens": 268888.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.150605010986329,
|
|
"epoch": 0.008194322505121452,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.74,
|
|
"mean_token_accuracy": 0.06668962053954601,
|
|
"num_tokens": 278451.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.055028533935547,
|
|
"epoch": 0.008486976880304361,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.7012,
|
|
"mean_token_accuracy": 0.06709126047790051,
|
|
"num_tokens": 289577.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.99300413131714,
|
|
"epoch": 0.00877963125548727,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.5708,
|
|
"mean_token_accuracy": 0.06588217578828334,
|
|
"num_tokens": 299057.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.817628383636475,
|
|
"epoch": 0.009072285630670179,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.4885,
|
|
"mean_token_accuracy": 0.06879487000405789,
|
|
"num_tokens": 308802.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.719903469085693,
|
|
"epoch": 0.009364940005853088,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.3738,
|
|
"mean_token_accuracy": 0.07012738101184368,
|
|
"num_tokens": 317365.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.5343448638916,
|
|
"epoch": 0.009657594381035996,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.3035,
|
|
"mean_token_accuracy": 0.07032971270382404,
|
|
"num_tokens": 326927.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.404919242858886,
|
|
"epoch": 0.009950248756218905,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.317,
|
|
"mean_token_accuracy": 0.06643726415932179,
|
|
"num_tokens": 337408.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.28181962966919,
|
|
"epoch": 0.010242903131401814,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.2328,
|
|
"mean_token_accuracy": 0.06949336491525174,
|
|
"num_tokens": 348282.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 9.099757957458497,
|
|
"epoch": 0.010535557506584723,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.1847,
|
|
"mean_token_accuracy": 0.06538042239844799,
|
|
"num_tokens": 357767.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.947648048400879,
|
|
"epoch": 0.010828211881767632,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.1401,
|
|
"mean_token_accuracy": 0.07173748835921287,
|
|
"num_tokens": 368665.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.832690525054932,
|
|
"epoch": 0.01112086625695054,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.125,
|
|
"mean_token_accuracy": 0.0688393272459507,
|
|
"num_tokens": 378741.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.722904491424561,
|
|
"epoch": 0.01141352063213345,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.1228,
|
|
"mean_token_accuracy": 0.07389259040355682,
|
|
"num_tokens": 388324.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.738927555084228,
|
|
"epoch": 0.011706175007316359,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.1796,
|
|
"mean_token_accuracy": 0.06952995508909225,
|
|
"num_tokens": 398450.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.615283966064453,
|
|
"epoch": 0.01199882938249927,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.0899,
|
|
"mean_token_accuracy": 0.07332116328179836,
|
|
"num_tokens": 408961.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.581227874755859,
|
|
"epoch": 0.012291483757682178,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.1226,
|
|
"mean_token_accuracy": 0.0683289546519518,
|
|
"num_tokens": 418517.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.532517910003662,
|
|
"epoch": 0.012584138132865087,
|
|
"grad_norm": 0.74609375,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.035,
|
|
"mean_token_accuracy": 0.07064232043921947,
|
|
"num_tokens": 428059.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.60868797302246,
|
|
"epoch": 0.012876792508047996,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.0513,
|
|
"mean_token_accuracy": 0.06911276690661908,
|
|
"num_tokens": 438499.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.50686674118042,
|
|
"epoch": 0.013169446883230905,
|
|
"grad_norm": 0.73046875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.0542,
|
|
"mean_token_accuracy": 0.06983353272080421,
|
|
"num_tokens": 448761.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.424305152893066,
|
|
"epoch": 0.013462101258413814,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.0471,
|
|
"mean_token_accuracy": 0.07090565152466297,
|
|
"num_tokens": 458386.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.447117614746094,
|
|
"epoch": 0.013754755633596722,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.991,
|
|
"mean_token_accuracy": 0.06929317899048329,
|
|
"num_tokens": 468198.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.46488447189331,
|
|
"epoch": 0.014047410008779631,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.0203,
|
|
"mean_token_accuracy": 0.07129090093076229,
|
|
"num_tokens": 480142.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.370715618133545,
|
|
"epoch": 0.01434006438396254,
|
|
"grad_norm": 0.79296875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.8998,
|
|
"mean_token_accuracy": 0.08097687140107154,
|
|
"num_tokens": 490686.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.408738231658935,
|
|
"epoch": 0.014632718759145449,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.9758,
|
|
"mean_token_accuracy": 0.07468959763646126,
|
|
"num_tokens": 500099.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.416247749328614,
|
|
"epoch": 0.014925373134328358,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.9005,
|
|
"mean_token_accuracy": 0.07769993916153908,
|
|
"num_tokens": 509525.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.347339820861816,
|
|
"epoch": 0.015218027509511267,
|
|
"grad_norm": 0.7421875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.8978,
|
|
"mean_token_accuracy": 0.07274740003049374,
|
|
"num_tokens": 520141.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.357079029083252,
|
|
"epoch": 0.015510681884694176,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.0384,
|
|
"mean_token_accuracy": 0.06597915887832642,
|
|
"num_tokens": 531645.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.3228253364563,
|
|
"epoch": 0.015803336259877086,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.9109,
|
|
"mean_token_accuracy": 0.07127318009734154,
|
|
"num_tokens": 541779.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.300227451324464,
|
|
"epoch": 0.016095990635059995,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.8871,
|
|
"mean_token_accuracy": 0.07985152080655097,
|
|
"num_tokens": 551770.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.360063171386718,
|
|
"epoch": 0.016388645010242904,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.8829,
|
|
"mean_token_accuracy": 0.0752606987953186,
|
|
"num_tokens": 561814.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.235567474365235,
|
|
"epoch": 0.016681299385425813,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.9551,
|
|
"mean_token_accuracy": 0.06852101050317287,
|
|
"num_tokens": 571988.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.42728681564331,
|
|
"epoch": 0.016973953760608722,
|
|
"grad_norm": 0.7109375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.9876,
|
|
"mean_token_accuracy": 0.07176770456135273,
|
|
"num_tokens": 582814.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.26517457962036,
|
|
"epoch": 0.01726660813579163,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.9178,
|
|
"mean_token_accuracy": 0.07443804480135441,
|
|
"num_tokens": 592301.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.199700450897216,
|
|
"epoch": 0.01755926251097454,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.8119,
|
|
"mean_token_accuracy": 0.077539586648345,
|
|
"num_tokens": 602898.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.289494514465332,
|
|
"epoch": 0.01785191688615745,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.8914,
|
|
"mean_token_accuracy": 0.07545375041663646,
|
|
"num_tokens": 612731.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.186264896392823,
|
|
"epoch": 0.018144571261340357,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.843,
|
|
"mean_token_accuracy": 0.07743252031505107,
|
|
"num_tokens": 623067.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.256502819061279,
|
|
"epoch": 0.018437225636523266,
|
|
"grad_norm": 0.76171875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.8317,
|
|
"mean_token_accuracy": 0.08097588121891022,
|
|
"num_tokens": 633538.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.319536399841308,
|
|
"epoch": 0.018729880011706175,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.8898,
|
|
"mean_token_accuracy": 0.07923161759972572,
|
|
"num_tokens": 643704.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.158168601989747,
|
|
"epoch": 0.019022534386889084,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.9204,
|
|
"mean_token_accuracy": 0.07442782260477543,
|
|
"num_tokens": 653766.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.432480907440185,
|
|
"epoch": 0.019315188762071993,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.9175,
|
|
"mean_token_accuracy": 0.07209107242524623,
|
|
"num_tokens": 663289.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.145325660705566,
|
|
"epoch": 0.0196078431372549,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.8214,
|
|
"mean_token_accuracy": 0.07566024884581565,
|
|
"num_tokens": 673196.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.328914070129395,
|
|
"epoch": 0.01990049751243781,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.8481,
|
|
"mean_token_accuracy": 0.0807725053280592,
|
|
"num_tokens": 682694.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.15658712387085,
|
|
"epoch": 0.02019315188762072,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.8414,
|
|
"mean_token_accuracy": 0.07744252718985081,
|
|
"num_tokens": 692880.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.265804481506347,
|
|
"epoch": 0.02048580626280363,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.9339,
|
|
"mean_token_accuracy": 0.07408964931964875,
|
|
"num_tokens": 703972.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.137424850463868,
|
|
"epoch": 0.020778460637986537,
|
|
"grad_norm": 0.85546875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.7467,
|
|
"mean_token_accuracy": 0.07930146306753158,
|
|
"num_tokens": 713571.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.209496688842773,
|
|
"epoch": 0.021071115013169446,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.8355,
|
|
"mean_token_accuracy": 0.08222270905971527,
|
|
"num_tokens": 723642.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.136897945404053,
|
|
"epoch": 0.021363769388352355,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.8793,
|
|
"mean_token_accuracy": 0.07310881391167641,
|
|
"num_tokens": 733848.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.176177358627319,
|
|
"epoch": 0.021656423763535264,
|
|
"grad_norm": 0.7265625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.6089,
|
|
"mean_token_accuracy": 0.10752813518047333,
|
|
"num_tokens": 743752.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.144378280639648,
|
|
"epoch": 0.021949078138718173,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.8031,
|
|
"mean_token_accuracy": 0.08254543766379356,
|
|
"num_tokens": 754145.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.118795394897461,
|
|
"epoch": 0.02224173251390108,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.766,
|
|
"mean_token_accuracy": 0.07592462375760078,
|
|
"num_tokens": 763930.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.178821086883545,
|
|
"epoch": 0.02253438688908399,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.7934,
|
|
"mean_token_accuracy": 0.08122270852327347,
|
|
"num_tokens": 774851.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.122542095184325,
|
|
"epoch": 0.0228270412642669,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.7915,
|
|
"mean_token_accuracy": 0.07377454079687595,
|
|
"num_tokens": 784251.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.100419902801514,
|
|
"epoch": 0.023119695639449808,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.8395,
|
|
"mean_token_accuracy": 0.0817381426692009,
|
|
"num_tokens": 794089.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.19840030670166,
|
|
"epoch": 0.023412350014632717,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.8242,
|
|
"mean_token_accuracy": 0.07439705729484558,
|
|
"num_tokens": 805274.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.17535228729248,
|
|
"epoch": 0.02370500438981563,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.7672,
|
|
"mean_token_accuracy": 0.07779609337449074,
|
|
"num_tokens": 816036.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.004063892364503,
|
|
"epoch": 0.02399765876499854,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.7613,
|
|
"mean_token_accuracy": 0.07725987881422043,
|
|
"num_tokens": 826393.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.159784364700318,
|
|
"epoch": 0.024290313140181447,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.7542,
|
|
"mean_token_accuracy": 0.0782277960330248,
|
|
"num_tokens": 836456.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.072831869125366,
|
|
"epoch": 0.024582967515364356,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.759,
|
|
"mean_token_accuracy": 0.07723330594599247,
|
|
"num_tokens": 846493.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.16247329711914,
|
|
"epoch": 0.024875621890547265,
|
|
"grad_norm": 0.8125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.7584,
|
|
"mean_token_accuracy": 0.07505144067108631,
|
|
"num_tokens": 857041.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.9756200313568115,
|
|
"epoch": 0.025168276265730174,
|
|
"grad_norm": 0.92578125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.6788,
|
|
"mean_token_accuracy": 0.08394450098276138,
|
|
"num_tokens": 867053.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.086863994598389,
|
|
"epoch": 0.025460930640913083,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.7398,
|
|
"mean_token_accuracy": 0.0781655989587307,
|
|
"num_tokens": 876944.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.080809020996094,
|
|
"epoch": 0.02575358501609599,
|
|
"grad_norm": 0.828125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.6902,
|
|
"mean_token_accuracy": 0.08105823285877704,
|
|
"num_tokens": 887257.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.091656923294067,
|
|
"epoch": 0.0260462393912789,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.7086,
|
|
"mean_token_accuracy": 0.0815083347260952,
|
|
"num_tokens": 896877.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.037679195404053,
|
|
"epoch": 0.02633889376646181,
|
|
"grad_norm": 0.83203125,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.6096,
|
|
"mean_token_accuracy": 0.08536934405565262,
|
|
"num_tokens": 906930.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.04052267074585,
|
|
"epoch": 0.026631548141644718,
|
|
"grad_norm": 0.8515625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.7471,
|
|
"mean_token_accuracy": 0.07020874097943305,
|
|
"num_tokens": 916841.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.061063146591186,
|
|
"epoch": 0.026924202516827627,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.6914,
|
|
"mean_token_accuracy": 0.07996720150113105,
|
|
"num_tokens": 927024.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.085123777389526,
|
|
"epoch": 0.027216856892010536,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.7227,
|
|
"mean_token_accuracy": 0.07815472632646561,
|
|
"num_tokens": 935786.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.955375385284424,
|
|
"epoch": 0.027509511267193445,
|
|
"grad_norm": 0.8359375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.6349,
|
|
"mean_token_accuracy": 0.08305523693561553,
|
|
"num_tokens": 946819.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.01088285446167,
|
|
"epoch": 0.027802165642376354,
|
|
"grad_norm": 0.84375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.6793,
|
|
"mean_token_accuracy": 0.08371292352676392,
|
|
"num_tokens": 957876.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.070115327835083,
|
|
"epoch": 0.028094820017559263,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.7627,
|
|
"mean_token_accuracy": 0.07754282280802727,
|
|
"num_tokens": 969211.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.086713933944703,
|
|
"epoch": 0.02838747439274217,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.7091,
|
|
"mean_token_accuracy": 0.07998017743229865,
|
|
"num_tokens": 979090.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.120463180541993,
|
|
"epoch": 0.02868012876792508,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.7649,
|
|
"mean_token_accuracy": 0.07725568227469921,
|
|
"num_tokens": 990510.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.065446138381958,
|
|
"epoch": 0.02897278314310799,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.6837,
|
|
"mean_token_accuracy": 0.07927028760313988,
|
|
"num_tokens": 1000890.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.89040241241455,
|
|
"epoch": 0.029265437518290898,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.633,
|
|
"mean_token_accuracy": 0.08539719134569168,
|
|
"num_tokens": 1011019.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"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": 217925004165120.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|