2035 lines
54 KiB
JSON
2035 lines
54 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.06052719184093454,
|
|
"eval_steps": 500,
|
|
"global_step": 1000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74259672164917,
|
|
"epoch": 0.0003026359592046727,
|
|
"grad_norm": 5.3125,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7761,
|
|
"mean_token_accuracy": 0.00015748031437397004,
|
|
"num_tokens": 10868.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.74251298904419,
|
|
"epoch": 0.0006052719184093454,
|
|
"grad_norm": 4.65625,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7789,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 22589.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.74259262084961,
|
|
"epoch": 0.0009079078776140181,
|
|
"grad_norm": 5.0625,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.7312,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 32471.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742576599121094,
|
|
"epoch": 0.0012105438368186909,
|
|
"grad_norm": 4.90625,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.6762,
|
|
"mean_token_accuracy": 0.0004521072842180729,
|
|
"num_tokens": 42231.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742555141448975,
|
|
"epoch": 0.0015131797960233635,
|
|
"grad_norm": 4.71875,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.5709,
|
|
"mean_token_accuracy": 0.0037700447253882887,
|
|
"num_tokens": 52966.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.742278385162354,
|
|
"epoch": 0.0018158157552280363,
|
|
"grad_norm": 4.1875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.4476,
|
|
"mean_token_accuracy": 0.03040276560932398,
|
|
"num_tokens": 63711.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.741186809539794,
|
|
"epoch": 0.002118451714432709,
|
|
"grad_norm": 3.09375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.3036,
|
|
"mean_token_accuracy": 0.040517190098762514,
|
|
"num_tokens": 74046.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.73850612640381,
|
|
"epoch": 0.0024210876736373817,
|
|
"grad_norm": 2.75,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 10.1339,
|
|
"mean_token_accuracy": 0.043018221482634546,
|
|
"num_tokens": 84275.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.73492136001587,
|
|
"epoch": 0.0027237236328420543,
|
|
"grad_norm": 2.234375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 10.0376,
|
|
"mean_token_accuracy": 0.04388587549328804,
|
|
"num_tokens": 93302.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.731941604614258,
|
|
"epoch": 0.003026359592046727,
|
|
"grad_norm": 2.21875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.9644,
|
|
"mean_token_accuracy": 0.041650505177676676,
|
|
"num_tokens": 103414.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.730784225463868,
|
|
"epoch": 0.0033289955512513995,
|
|
"grad_norm": 1.9921875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.8956,
|
|
"mean_token_accuracy": 0.04557133205235005,
|
|
"num_tokens": 113749.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.728888988494873,
|
|
"epoch": 0.0036316315104560726,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.8525,
|
|
"mean_token_accuracy": 0.04352857358753681,
|
|
"num_tokens": 123922.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.72574872970581,
|
|
"epoch": 0.003934267469660745,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.8144,
|
|
"mean_token_accuracy": 0.04175829254090786,
|
|
"num_tokens": 134720.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.723058605194092,
|
|
"epoch": 0.004236903428865418,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.7116,
|
|
"mean_token_accuracy": 0.04281677231192589,
|
|
"num_tokens": 145025.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.72093152999878,
|
|
"epoch": 0.00453953938807009,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.6537,
|
|
"mean_token_accuracy": 0.04347668327391148,
|
|
"num_tokens": 155182.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.716725921630859,
|
|
"epoch": 0.004842175347274763,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.6032,
|
|
"mean_token_accuracy": 0.04583056308329105,
|
|
"num_tokens": 166231.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.70965871810913,
|
|
"epoch": 0.005144811306479436,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.4954,
|
|
"mean_token_accuracy": 0.04315895065665245,
|
|
"num_tokens": 175799.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.702201461791992,
|
|
"epoch": 0.005447447265684109,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.4271,
|
|
"mean_token_accuracy": 0.04461735654622316,
|
|
"num_tokens": 185455.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.692846012115478,
|
|
"epoch": 0.005750083224888782,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.4192,
|
|
"mean_token_accuracy": 0.04293463826179504,
|
|
"num_tokens": 196210.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.680758380889893,
|
|
"epoch": 0.006052719184093454,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.3191,
|
|
"mean_token_accuracy": 0.047341904044151305,
|
|
"num_tokens": 206618.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.663753604888916,
|
|
"epoch": 0.006355355143298127,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.1872,
|
|
"mean_token_accuracy": 0.05284549556672573,
|
|
"num_tokens": 217020.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.640412425994873,
|
|
"epoch": 0.006657991102502799,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.06,
|
|
"mean_token_accuracy": 0.057008244842290876,
|
|
"num_tokens": 227195.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.60520315170288,
|
|
"epoch": 0.006960627061707472,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9862,
|
|
"mean_token_accuracy": 0.057433782145380974,
|
|
"num_tokens": 238473.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.55847454071045,
|
|
"epoch": 0.007263263020912145,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.8089,
|
|
"mean_token_accuracy": 0.06153528727591038,
|
|
"num_tokens": 247778.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.495486640930176,
|
|
"epoch": 0.007565898980116817,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7828,
|
|
"mean_token_accuracy": 0.06823135390877724,
|
|
"num_tokens": 258424.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.432032680511474,
|
|
"epoch": 0.00786853493932149,
|
|
"grad_norm": 1.7578125,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.6115,
|
|
"mean_token_accuracy": 0.06918121092021465,
|
|
"num_tokens": 267715.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.359144115447998,
|
|
"epoch": 0.008171170898526163,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.4638,
|
|
"mean_token_accuracy": 0.07366116344928741,
|
|
"num_tokens": 277127.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.267242240905762,
|
|
"epoch": 0.008473806857730836,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.3343,
|
|
"mean_token_accuracy": 0.0729695912450552,
|
|
"num_tokens": 287275.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 10.183104038238525,
|
|
"epoch": 0.008776442816935508,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.2075,
|
|
"mean_token_accuracy": 0.07483375370502472,
|
|
"num_tokens": 297131.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 10.032635116577149,
|
|
"epoch": 0.00907907877614018,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.0837,
|
|
"mean_token_accuracy": 0.0766814447939396,
|
|
"num_tokens": 307649.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.881086444854736,
|
|
"epoch": 0.009381714735344854,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.9254,
|
|
"mean_token_accuracy": 0.0791646283119917,
|
|
"num_tokens": 317655.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.67042417526245,
|
|
"epoch": 0.009684350694549527,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.9643,
|
|
"mean_token_accuracy": 0.07545723952353,
|
|
"num_tokens": 329286.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.513041305541993,
|
|
"epoch": 0.0099869866537542,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.7586,
|
|
"mean_token_accuracy": 0.0771209493279457,
|
|
"num_tokens": 339212.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.245701503753661,
|
|
"epoch": 0.010289622612958871,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.7092,
|
|
"mean_token_accuracy": 0.07468102127313614,
|
|
"num_tokens": 349205.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 9.026909351348877,
|
|
"epoch": 0.010592258572163544,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.6824,
|
|
"mean_token_accuracy": 0.08052874132990837,
|
|
"num_tokens": 360548.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.78289623260498,
|
|
"epoch": 0.010894894531368217,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.5607,
|
|
"mean_token_accuracy": 0.07696975283324718,
|
|
"num_tokens": 371049.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.596883869171142,
|
|
"epoch": 0.01119753049057289,
|
|
"grad_norm": 0.7890625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.6138,
|
|
"mean_token_accuracy": 0.07861833088099957,
|
|
"num_tokens": 382334.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.436706256866454,
|
|
"epoch": 0.011500166449777563,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.5283,
|
|
"mean_token_accuracy": 0.07383836694061756,
|
|
"num_tokens": 393169.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.288833332061767,
|
|
"epoch": 0.011802802408982235,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4185,
|
|
"mean_token_accuracy": 0.0791880264878273,
|
|
"num_tokens": 403826.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.156514549255371,
|
|
"epoch": 0.012105438368186908,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.3629,
|
|
"mean_token_accuracy": 0.08756700083613396,
|
|
"num_tokens": 414231.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.109231567382812,
|
|
"epoch": 0.01240807432739158,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.3613,
|
|
"mean_token_accuracy": 0.07943096160888671,
|
|
"num_tokens": 424532.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.120400428771973,
|
|
"epoch": 0.012710710286596254,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.4376,
|
|
"mean_token_accuracy": 0.0809888195246458,
|
|
"num_tokens": 435002.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.078549480438232,
|
|
"epoch": 0.013013346245800927,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.3373,
|
|
"mean_token_accuracy": 0.088172397762537,
|
|
"num_tokens": 444954.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.021297836303711,
|
|
"epoch": 0.013315982205005598,
|
|
"grad_norm": 0.87890625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.3151,
|
|
"mean_token_accuracy": 0.08277993872761727,
|
|
"num_tokens": 455078.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.960602712631226,
|
|
"epoch": 0.013618618164210271,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.3156,
|
|
"mean_token_accuracy": 0.08474647626280785,
|
|
"num_tokens": 464688.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.901602554321289,
|
|
"epoch": 0.013921254123414944,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2243,
|
|
"mean_token_accuracy": 0.08522045239806175,
|
|
"num_tokens": 475365.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.995245122909546,
|
|
"epoch": 0.014223890082619617,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.3312,
|
|
"mean_token_accuracy": 0.08323194459080696,
|
|
"num_tokens": 487238.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.879671859741211,
|
|
"epoch": 0.01452652604182429,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.2215,
|
|
"mean_token_accuracy": 0.08390360698103905,
|
|
"num_tokens": 497485.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.803989362716675,
|
|
"epoch": 0.014829162001028962,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2663,
|
|
"mean_token_accuracy": 0.09161997810006142,
|
|
"num_tokens": 508228.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.834836626052857,
|
|
"epoch": 0.015131797960233635,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.1851,
|
|
"mean_token_accuracy": 0.08978703171014786,
|
|
"num_tokens": 518645.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.774400329589843,
|
|
"epoch": 0.015434433919438308,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.1447,
|
|
"mean_token_accuracy": 0.09164771437644958,
|
|
"num_tokens": 528151.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.811225557327271,
|
|
"epoch": 0.01573706987864298,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.2129,
|
|
"mean_token_accuracy": 0.08425109386444092,
|
|
"num_tokens": 539160.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.784347248077393,
|
|
"epoch": 0.016039705837847654,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.2983,
|
|
"mean_token_accuracy": 0.08336320072412491,
|
|
"num_tokens": 550053.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.730723190307617,
|
|
"epoch": 0.016342341797052327,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.0826,
|
|
"mean_token_accuracy": 0.09093882888555527,
|
|
"num_tokens": 560229.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.7058757781982425,
|
|
"epoch": 0.016644977756257,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1653,
|
|
"mean_token_accuracy": 0.08866406977176666,
|
|
"num_tokens": 570630.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.813187074661255,
|
|
"epoch": 0.016947613715461673,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2263,
|
|
"mean_token_accuracy": 0.08829710632562637,
|
|
"num_tokens": 581208.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.697104406356812,
|
|
"epoch": 0.017250249674666342,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.0955,
|
|
"mean_token_accuracy": 0.09898408353328705,
|
|
"num_tokens": 591352.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.6714756965637205,
|
|
"epoch": 0.017552885633871015,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.1736,
|
|
"mean_token_accuracy": 0.09172456562519074,
|
|
"num_tokens": 600670.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.675610876083374,
|
|
"epoch": 0.01785552159307569,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.1785,
|
|
"mean_token_accuracy": 0.09298023506999016,
|
|
"num_tokens": 610625.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.745320653915405,
|
|
"epoch": 0.01815815755228036,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.1448,
|
|
"mean_token_accuracy": 0.09545003101229668,
|
|
"num_tokens": 621096.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.587613916397094,
|
|
"epoch": 0.018460793511485035,
|
|
"grad_norm": 0.88671875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.1062,
|
|
"mean_token_accuracy": 0.09240609779953957,
|
|
"num_tokens": 632168.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.6709612846374515,
|
|
"epoch": 0.018763429470689708,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2109,
|
|
"mean_token_accuracy": 0.09153169840574264,
|
|
"num_tokens": 642277.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.587620401382447,
|
|
"epoch": 0.01906606542989438,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0974,
|
|
"mean_token_accuracy": 0.09367592856287957,
|
|
"num_tokens": 652429.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.541977453231811,
|
|
"epoch": 0.019368701389099054,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.989,
|
|
"mean_token_accuracy": 0.09826880320906639,
|
|
"num_tokens": 662805.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.580408334732056,
|
|
"epoch": 0.019671337348303727,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.9876,
|
|
"mean_token_accuracy": 0.10082038789987564,
|
|
"num_tokens": 673015.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.452456760406494,
|
|
"epoch": 0.0199739733075084,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.9221,
|
|
"mean_token_accuracy": 0.09537633135914803,
|
|
"num_tokens": 683357.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.544377756118775,
|
|
"epoch": 0.02027660926671307,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.0466,
|
|
"mean_token_accuracy": 0.09294784143567085,
|
|
"num_tokens": 693310.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.5124753475189205,
|
|
"epoch": 0.020579245225917742,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.0218,
|
|
"mean_token_accuracy": 0.09687028974294662,
|
|
"num_tokens": 703189.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.457939147949219,
|
|
"epoch": 0.020881881185122415,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.9878,
|
|
"mean_token_accuracy": 0.09731099009513855,
|
|
"num_tokens": 714359.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.501646852493286,
|
|
"epoch": 0.02118451714432709,
|
|
"grad_norm": 0.9296875,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.0421,
|
|
"mean_token_accuracy": 0.0931004598736763,
|
|
"num_tokens": 724991.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.500136518478394,
|
|
"epoch": 0.02148715310353176,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.8986,
|
|
"mean_token_accuracy": 0.09156099259853363,
|
|
"num_tokens": 735879.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.451840782165528,
|
|
"epoch": 0.021789789062736434,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9577,
|
|
"mean_token_accuracy": 0.10013179183006286,
|
|
"num_tokens": 745887.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.5548046112060545,
|
|
"epoch": 0.022092425021941108,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.9763,
|
|
"mean_token_accuracy": 0.1025528259575367,
|
|
"num_tokens": 755822.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.366734838485717,
|
|
"epoch": 0.02239506098114578,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.9315,
|
|
"mean_token_accuracy": 0.10159048363566399,
|
|
"num_tokens": 765330.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.413720750808716,
|
|
"epoch": 0.022697696940350454,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.8919,
|
|
"mean_token_accuracy": 0.10146145895123482,
|
|
"num_tokens": 775577.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.391197729110718,
|
|
"epoch": 0.023000332899555127,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.9459,
|
|
"mean_token_accuracy": 0.0984920509159565,
|
|
"num_tokens": 784999.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.430153036117554,
|
|
"epoch": 0.023302968858759796,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.8889,
|
|
"mean_token_accuracy": 0.10047663822770118,
|
|
"num_tokens": 794917.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.323439264297486,
|
|
"epoch": 0.02360560481796447,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8756,
|
|
"mean_token_accuracy": 0.09725546389818192,
|
|
"num_tokens": 805772.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.467493295669556,
|
|
"epoch": 0.023908240777169142,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.0954,
|
|
"mean_token_accuracy": 0.0907193399965763,
|
|
"num_tokens": 816326.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.409446811676025,
|
|
"epoch": 0.024210876736373815,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.8854,
|
|
"mean_token_accuracy": 0.0965262807905674,
|
|
"num_tokens": 825723.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.347925472259521,
|
|
"epoch": 0.02451351269557849,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8529,
|
|
"mean_token_accuracy": 0.10136102885007858,
|
|
"num_tokens": 834939.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.287171125411987,
|
|
"epoch": 0.02481614865478316,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7984,
|
|
"mean_token_accuracy": 0.10428313314914703,
|
|
"num_tokens": 844093.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.37103533744812,
|
|
"epoch": 0.025118784613987834,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.8791,
|
|
"mean_token_accuracy": 0.09733516573905945,
|
|
"num_tokens": 853879.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.328188419342041,
|
|
"epoch": 0.025421420573192507,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.8858,
|
|
"mean_token_accuracy": 0.09786264151334763,
|
|
"num_tokens": 864013.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.331731557846069,
|
|
"epoch": 0.02572405653239718,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.8476,
|
|
"mean_token_accuracy": 0.10083968937397003,
|
|
"num_tokens": 874295.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.2822911739349365,
|
|
"epoch": 0.026026692491601854,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.8964,
|
|
"mean_token_accuracy": 0.09731603786349297,
|
|
"num_tokens": 884748.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.363057661056518,
|
|
"epoch": 0.026329328450806523,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.9178,
|
|
"mean_token_accuracy": 0.09977400824427604,
|
|
"num_tokens": 895809.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.325506687164307,
|
|
"epoch": 0.026631964410011196,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.8938,
|
|
"mean_token_accuracy": 0.10050061941146851,
|
|
"num_tokens": 905252.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.357236623764038,
|
|
"epoch": 0.02693460036921587,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.8265,
|
|
"mean_token_accuracy": 0.09752592518925667,
|
|
"num_tokens": 913876.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.245391464233398,
|
|
"epoch": 0.027237236328420542,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.8026,
|
|
"mean_token_accuracy": 0.09813540726900101,
|
|
"num_tokens": 924255.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.230363607406616,
|
|
"epoch": 0.027539872287625215,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.7995,
|
|
"mean_token_accuracy": 0.10496273562312126,
|
|
"num_tokens": 934799.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.288649463653565,
|
|
"epoch": 0.02784250824682989,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.8847,
|
|
"mean_token_accuracy": 0.10213120430707931,
|
|
"num_tokens": 945757.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.276965951919555,
|
|
"epoch": 0.02814514420603456,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.8831,
|
|
"mean_token_accuracy": 0.0962677277624607,
|
|
"num_tokens": 956143.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.307726716995239,
|
|
"epoch": 0.028447780165239234,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.7977,
|
|
"mean_token_accuracy": 0.10646740570664406,
|
|
"num_tokens": 966437.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.251181745529175,
|
|
"epoch": 0.028750416124443907,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.8266,
|
|
"mean_token_accuracy": 0.09941198974847794,
|
|
"num_tokens": 977350.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.3098362445831295,
|
|
"epoch": 0.02905305208364858,
|
|
"grad_norm": 0.890625,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.9111,
|
|
"mean_token_accuracy": 0.09969582483172416,
|
|
"num_tokens": 988340.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.254582738876342,
|
|
"epoch": 0.029355688042853254,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.7656,
|
|
"mean_token_accuracy": 0.10246146842837334,
|
|
"num_tokens": 998323.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.18165078163147,
|
|
"epoch": 0.029658324002057923,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.6808,
|
|
"mean_token_accuracy": 0.10682199671864509,
|
|
"num_tokens": 1007731.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.063388872146606,
|
|
"epoch": 0.029960959961262596,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.6377,
|
|
"mean_token_accuracy": 0.10866327434778214,
|
|
"num_tokens": 1017543.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.027464437484741,
|
|
"epoch": 0.03026359592046727,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.7361,
|
|
"mean_token_accuracy": 0.10657615810632706,
|
|
"num_tokens": 1028558.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 7.175600576400757,
|
|
"epoch": 0.030566231879671942,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000252,
|
|
"loss": 6.7472,
|
|
"mean_token_accuracy": 0.10892508924007416,
|
|
"num_tokens": 1038904.0,
|
|
"step": 505
|
|
},
|
|
{
|
|
"entropy": 7.212454414367675,
|
|
"epoch": 0.030868867838876615,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0002545,
|
|
"loss": 6.7726,
|
|
"mean_token_accuracy": 0.10466748997569084,
|
|
"num_tokens": 1049143.0,
|
|
"step": 510
|
|
},
|
|
{
|
|
"entropy": 7.130760240554809,
|
|
"epoch": 0.03117150379808129,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000257,
|
|
"loss": 6.7614,
|
|
"mean_token_accuracy": 0.10434122681617737,
|
|
"num_tokens": 1059914.0,
|
|
"step": 515
|
|
},
|
|
{
|
|
"entropy": 7.148777818679809,
|
|
"epoch": 0.03147413975728596,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002595,
|
|
"loss": 6.8028,
|
|
"mean_token_accuracy": 0.10173705667257309,
|
|
"num_tokens": 1071007.0,
|
|
"step": 520
|
|
},
|
|
{
|
|
"entropy": 7.233448696136475,
|
|
"epoch": 0.03177677571649063,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000262,
|
|
"loss": 6.7909,
|
|
"mean_token_accuracy": 0.10418517664074897,
|
|
"num_tokens": 1081983.0,
|
|
"step": 525
|
|
},
|
|
{
|
|
"entropy": 7.17897367477417,
|
|
"epoch": 0.03207941167569531,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00026450000000000003,
|
|
"loss": 6.8365,
|
|
"mean_token_accuracy": 0.0993827037513256,
|
|
"num_tokens": 1093200.0,
|
|
"step": 530
|
|
},
|
|
{
|
|
"entropy": 7.124456787109375,
|
|
"epoch": 0.03238204763489998,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00026700000000000004,
|
|
"loss": 6.7039,
|
|
"mean_token_accuracy": 0.10733622834086418,
|
|
"num_tokens": 1103831.0,
|
|
"step": 535
|
|
},
|
|
{
|
|
"entropy": 7.158724975585938,
|
|
"epoch": 0.03268468359410465,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.00026950000000000005,
|
|
"loss": 6.7492,
|
|
"mean_token_accuracy": 0.10417231321334838,
|
|
"num_tokens": 1114949.0,
|
|
"step": 540
|
|
},
|
|
{
|
|
"entropy": 7.080250072479248,
|
|
"epoch": 0.03298731955330932,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00027200000000000005,
|
|
"loss": 6.6037,
|
|
"mean_token_accuracy": 0.10863222777843476,
|
|
"num_tokens": 1125079.0,
|
|
"step": 545
|
|
},
|
|
{
|
|
"entropy": 7.089680099487305,
|
|
"epoch": 0.033289955512514,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002745,
|
|
"loss": 6.7476,
|
|
"mean_token_accuracy": 0.10140802189707757,
|
|
"num_tokens": 1136097.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 7.035984706878662,
|
|
"epoch": 0.03359259147171867,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000277,
|
|
"loss": 6.6451,
|
|
"mean_token_accuracy": 0.10919994190335273,
|
|
"num_tokens": 1146960.0,
|
|
"step": 555
|
|
},
|
|
{
|
|
"entropy": 7.023309707641602,
|
|
"epoch": 0.033895227430923346,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0002795,
|
|
"loss": 6.6812,
|
|
"mean_token_accuracy": 0.10784762799739837,
|
|
"num_tokens": 1156752.0,
|
|
"step": 560
|
|
},
|
|
{
|
|
"entropy": 7.035311365127564,
|
|
"epoch": 0.034197863390128015,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00028199999999999997,
|
|
"loss": 6.6513,
|
|
"mean_token_accuracy": 0.10291712656617165,
|
|
"num_tokens": 1166831.0,
|
|
"step": 565
|
|
},
|
|
{
|
|
"entropy": 7.112051916122437,
|
|
"epoch": 0.034500499349332685,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0002845,
|
|
"loss": 6.7331,
|
|
"mean_token_accuracy": 0.10639557167887688,
|
|
"num_tokens": 1178318.0,
|
|
"step": 570
|
|
},
|
|
{
|
|
"entropy": 7.135074758529663,
|
|
"epoch": 0.03480313530853736,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000287,
|
|
"loss": 6.7002,
|
|
"mean_token_accuracy": 0.104981629550457,
|
|
"num_tokens": 1187822.0,
|
|
"step": 575
|
|
},
|
|
{
|
|
"entropy": 7.064349842071533,
|
|
"epoch": 0.03510577126774203,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002895,
|
|
"loss": 6.713,
|
|
"mean_token_accuracy": 0.10567119121551513,
|
|
"num_tokens": 1198243.0,
|
|
"step": 580
|
|
},
|
|
{
|
|
"entropy": 7.098157596588135,
|
|
"epoch": 0.03540840722694671,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000292,
|
|
"loss": 6.6459,
|
|
"mean_token_accuracy": 0.10842026993632317,
|
|
"num_tokens": 1208688.0,
|
|
"step": 585
|
|
},
|
|
{
|
|
"entropy": 6.983540201187134,
|
|
"epoch": 0.03571104318615138,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.0002945,
|
|
"loss": 6.7001,
|
|
"mean_token_accuracy": 0.10743521004915238,
|
|
"num_tokens": 1218928.0,
|
|
"step": 590
|
|
},
|
|
{
|
|
"entropy": 7.1224036693573,
|
|
"epoch": 0.03601367914535605,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.000297,
|
|
"loss": 6.7705,
|
|
"mean_token_accuracy": 0.10451584979891777,
|
|
"num_tokens": 1229582.0,
|
|
"step": 595
|
|
},
|
|
{
|
|
"entropy": 6.972618961334229,
|
|
"epoch": 0.03631631510456072,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0002995,
|
|
"loss": 6.5813,
|
|
"mean_token_accuracy": 0.10848907530307769,
|
|
"num_tokens": 1239887.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 7.055910205841064,
|
|
"epoch": 0.0366189510637654,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000302,
|
|
"loss": 6.6538,
|
|
"mean_token_accuracy": 0.1105770394206047,
|
|
"num_tokens": 1249278.0,
|
|
"step": 605
|
|
},
|
|
{
|
|
"entropy": 6.987835693359375,
|
|
"epoch": 0.03692158702297007,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.0003045,
|
|
"loss": 6.5816,
|
|
"mean_token_accuracy": 0.11098704561591148,
|
|
"num_tokens": 1259931.0,
|
|
"step": 610
|
|
},
|
|
{
|
|
"entropy": 7.000305461883545,
|
|
"epoch": 0.03722422298217474,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000307,
|
|
"loss": 6.6462,
|
|
"mean_token_accuracy": 0.10493655204772949,
|
|
"num_tokens": 1269299.0,
|
|
"step": 615
|
|
},
|
|
{
|
|
"entropy": 7.056825733184814,
|
|
"epoch": 0.037526858941379415,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.0003095,
|
|
"loss": 6.6602,
|
|
"mean_token_accuracy": 0.11132641285657882,
|
|
"num_tokens": 1279662.0,
|
|
"step": 620
|
|
},
|
|
{
|
|
"entropy": 6.947488164901733,
|
|
"epoch": 0.037829494900584085,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000312,
|
|
"loss": 6.6816,
|
|
"mean_token_accuracy": 0.1092316597700119,
|
|
"num_tokens": 1289513.0,
|
|
"step": 625
|
|
},
|
|
{
|
|
"entropy": 6.978721809387207,
|
|
"epoch": 0.03813213085978876,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.0003145,
|
|
"loss": 6.5508,
|
|
"mean_token_accuracy": 0.11456230357289314,
|
|
"num_tokens": 1299361.0,
|
|
"step": 630
|
|
},
|
|
{
|
|
"entropy": 6.961810827255249,
|
|
"epoch": 0.03843476681899343,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000317,
|
|
"loss": 6.5855,
|
|
"mean_token_accuracy": 0.11074020341038704,
|
|
"num_tokens": 1310263.0,
|
|
"step": 635
|
|
},
|
|
{
|
|
"entropy": 6.986895895004272,
|
|
"epoch": 0.03873740277819811,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0003195,
|
|
"loss": 6.6711,
|
|
"mean_token_accuracy": 0.11129543110728264,
|
|
"num_tokens": 1320120.0,
|
|
"step": 640
|
|
},
|
|
{
|
|
"entropy": 7.036265087127686,
|
|
"epoch": 0.03904003873740278,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000322,
|
|
"loss": 6.7857,
|
|
"mean_token_accuracy": 0.10901256203651429,
|
|
"num_tokens": 1330575.0,
|
|
"step": 645
|
|
},
|
|
{
|
|
"entropy": 7.012125444412232,
|
|
"epoch": 0.03934267469660745,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00032450000000000003,
|
|
"loss": 6.6537,
|
|
"mean_token_accuracy": 0.10976160243153572,
|
|
"num_tokens": 1340416.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 6.944071865081787,
|
|
"epoch": 0.03964531065581212,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.00032700000000000003,
|
|
"loss": 6.5542,
|
|
"mean_token_accuracy": 0.11052674353122711,
|
|
"num_tokens": 1350263.0,
|
|
"step": 655
|
|
},
|
|
{
|
|
"entropy": 6.9588096141815186,
|
|
"epoch": 0.0399479466150168,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.00032950000000000004,
|
|
"loss": 6.6478,
|
|
"mean_token_accuracy": 0.10758454948663712,
|
|
"num_tokens": 1360455.0,
|
|
"step": 660
|
|
},
|
|
{
|
|
"entropy": 6.912649297714234,
|
|
"epoch": 0.04025058257422147,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.00033200000000000005,
|
|
"loss": 6.6268,
|
|
"mean_token_accuracy": 0.10876741632819176,
|
|
"num_tokens": 1370250.0,
|
|
"step": 665
|
|
},
|
|
{
|
|
"entropy": 6.955088329315186,
|
|
"epoch": 0.04055321853342614,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00033450000000000005,
|
|
"loss": 6.589,
|
|
"mean_token_accuracy": 0.1123467318713665,
|
|
"num_tokens": 1380625.0,
|
|
"step": 670
|
|
},
|
|
{
|
|
"entropy": 6.866590213775635,
|
|
"epoch": 0.040855854492630815,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.000337,
|
|
"loss": 6.5833,
|
|
"mean_token_accuracy": 0.11681398823857307,
|
|
"num_tokens": 1391799.0,
|
|
"step": 675
|
|
},
|
|
{
|
|
"entropy": 6.9002148628234865,
|
|
"epoch": 0.041158490451835485,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0003395,
|
|
"loss": 6.5222,
|
|
"mean_token_accuracy": 0.11345557197928428,
|
|
"num_tokens": 1402386.0,
|
|
"step": 680
|
|
},
|
|
{
|
|
"entropy": 6.880913877487183,
|
|
"epoch": 0.04146112641104016,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000342,
|
|
"loss": 6.5282,
|
|
"mean_token_accuracy": 0.1147773802280426,
|
|
"num_tokens": 1413076.0,
|
|
"step": 685
|
|
},
|
|
{
|
|
"entropy": 6.8401960849761965,
|
|
"epoch": 0.04176376237024483,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00034449999999999997,
|
|
"loss": 6.4883,
|
|
"mean_token_accuracy": 0.1113433837890625,
|
|
"num_tokens": 1423406.0,
|
|
"step": 690
|
|
},
|
|
{
|
|
"entropy": 6.829567813873291,
|
|
"epoch": 0.04206639832944951,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000347,
|
|
"loss": 6.5338,
|
|
"mean_token_accuracy": 0.11748597323894501,
|
|
"num_tokens": 1434166.0,
|
|
"step": 695
|
|
},
|
|
{
|
|
"entropy": 6.7734236240386965,
|
|
"epoch": 0.04236903428865418,
|
|
"grad_norm": 0.90234375,
|
|
"learning_rate": 0.0003495,
|
|
"loss": 6.4686,
|
|
"mean_token_accuracy": 0.1190357156097889,
|
|
"num_tokens": 1444492.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 6.852633428573609,
|
|
"epoch": 0.04267167024785885,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000352,
|
|
"loss": 6.5249,
|
|
"mean_token_accuracy": 0.11431135833263398,
|
|
"num_tokens": 1454499.0,
|
|
"step": 705
|
|
},
|
|
{
|
|
"entropy": 6.920568561553955,
|
|
"epoch": 0.04297430620706352,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0003545,
|
|
"loss": 6.5507,
|
|
"mean_token_accuracy": 0.11268127039074897,
|
|
"num_tokens": 1464854.0,
|
|
"step": 710
|
|
},
|
|
{
|
|
"entropy": 6.81124439239502,
|
|
"epoch": 0.0432769421662682,
|
|
"grad_norm": 0.93359375,
|
|
"learning_rate": 0.000357,
|
|
"loss": 6.5423,
|
|
"mean_token_accuracy": 0.1138940081000328,
|
|
"num_tokens": 1475902.0,
|
|
"step": 715
|
|
},
|
|
{
|
|
"entropy": 6.867471027374267,
|
|
"epoch": 0.04357957812547287,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0003595,
|
|
"loss": 6.5599,
|
|
"mean_token_accuracy": 0.11219575852155686,
|
|
"num_tokens": 1486215.0,
|
|
"step": 720
|
|
},
|
|
{
|
|
"entropy": 6.818832921981811,
|
|
"epoch": 0.04388221408467754,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000362,
|
|
"loss": 6.4046,
|
|
"mean_token_accuracy": 0.12129560708999634,
|
|
"num_tokens": 1495981.0,
|
|
"step": 725
|
|
},
|
|
{
|
|
"entropy": 6.84439525604248,
|
|
"epoch": 0.044184850043882215,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0003645,
|
|
"loss": 6.5226,
|
|
"mean_token_accuracy": 0.11364036872982979,
|
|
"num_tokens": 1506298.0,
|
|
"step": 730
|
|
},
|
|
{
|
|
"entropy": 6.711646842956543,
|
|
"epoch": 0.044487486003086885,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000367,
|
|
"loss": 6.5304,
|
|
"mean_token_accuracy": 0.11716336011886597,
|
|
"num_tokens": 1516032.0,
|
|
"step": 735
|
|
},
|
|
{
|
|
"entropy": 6.956787157058716,
|
|
"epoch": 0.04479012196229156,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0003695,
|
|
"loss": 6.6011,
|
|
"mean_token_accuracy": 0.10908963233232498,
|
|
"num_tokens": 1527335.0,
|
|
"step": 740
|
|
},
|
|
{
|
|
"entropy": 6.775584411621094,
|
|
"epoch": 0.04509275792149623,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000372,
|
|
"loss": 6.4768,
|
|
"mean_token_accuracy": 0.11435298025608062,
|
|
"num_tokens": 1538186.0,
|
|
"step": 745
|
|
},
|
|
{
|
|
"entropy": 6.9018229961395265,
|
|
"epoch": 0.04539539388070091,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0003745,
|
|
"loss": 6.6299,
|
|
"mean_token_accuracy": 0.11318550854921341,
|
|
"num_tokens": 1548197.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 6.855479621887207,
|
|
"epoch": 0.04569802983990558,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000377,
|
|
"loss": 6.6064,
|
|
"mean_token_accuracy": 0.11473771557211876,
|
|
"num_tokens": 1557623.0,
|
|
"step": 755
|
|
},
|
|
{
|
|
"entropy": 6.816193437576294,
|
|
"epoch": 0.04600066579911025,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0003795,
|
|
"loss": 6.509,
|
|
"mean_token_accuracy": 0.1118078961968422,
|
|
"num_tokens": 1568205.0,
|
|
"step": 760
|
|
},
|
|
{
|
|
"entropy": 6.835904216766357,
|
|
"epoch": 0.04630330175831492,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.000382,
|
|
"loss": 6.5306,
|
|
"mean_token_accuracy": 0.1150369681417942,
|
|
"num_tokens": 1578558.0,
|
|
"step": 765
|
|
},
|
|
{
|
|
"entropy": 6.773059940338134,
|
|
"epoch": 0.04660593771751959,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0003845,
|
|
"loss": 6.4628,
|
|
"mean_token_accuracy": 0.1139510601758957,
|
|
"num_tokens": 1589240.0,
|
|
"step": 770
|
|
},
|
|
{
|
|
"entropy": 6.803834056854248,
|
|
"epoch": 0.04690857367672427,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00038700000000000003,
|
|
"loss": 6.5198,
|
|
"mean_token_accuracy": 0.11591631695628166,
|
|
"num_tokens": 1600318.0,
|
|
"step": 775
|
|
},
|
|
{
|
|
"entropy": 6.803225708007813,
|
|
"epoch": 0.04721120963592894,
|
|
"grad_norm": 0.95703125,
|
|
"learning_rate": 0.00038950000000000003,
|
|
"loss": 6.5481,
|
|
"mean_token_accuracy": 0.11747798472642898,
|
|
"num_tokens": 1610606.0,
|
|
"step": 780
|
|
},
|
|
{
|
|
"entropy": 6.932116937637329,
|
|
"epoch": 0.047513845595133615,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.00039200000000000004,
|
|
"loss": 6.6079,
|
|
"mean_token_accuracy": 0.11167082041502,
|
|
"num_tokens": 1621517.0,
|
|
"step": 785
|
|
},
|
|
{
|
|
"entropy": 6.681528043746948,
|
|
"epoch": 0.047816481554338285,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.00039450000000000005,
|
|
"loss": 6.5522,
|
|
"mean_token_accuracy": 0.1144383728504181,
|
|
"num_tokens": 1631929.0,
|
|
"step": 790
|
|
},
|
|
{
|
|
"entropy": 6.830399990081787,
|
|
"epoch": 0.04811911751354296,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.00039700000000000005,
|
|
"loss": 6.4941,
|
|
"mean_token_accuracy": 0.11434257254004479,
|
|
"num_tokens": 1642393.0,
|
|
"step": 795
|
|
},
|
|
{
|
|
"entropy": 6.8345434188842775,
|
|
"epoch": 0.04842175347274763,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0003995,
|
|
"loss": 6.6014,
|
|
"mean_token_accuracy": 0.10618331655859947,
|
|
"num_tokens": 1653210.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 6.824073791503906,
|
|
"epoch": 0.04872438943195231,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000402,
|
|
"loss": 6.4797,
|
|
"mean_token_accuracy": 0.1194666214287281,
|
|
"num_tokens": 1662984.0,
|
|
"step": 805
|
|
},
|
|
{
|
|
"entropy": 6.756822013854981,
|
|
"epoch": 0.04902702539115698,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0004045,
|
|
"loss": 6.4334,
|
|
"mean_token_accuracy": 0.1138382039964199,
|
|
"num_tokens": 1672976.0,
|
|
"step": 810
|
|
},
|
|
{
|
|
"entropy": 6.846238517761231,
|
|
"epoch": 0.04932966135036165,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00040699999999999997,
|
|
"loss": 6.5513,
|
|
"mean_token_accuracy": 0.11465490832924843,
|
|
"num_tokens": 1682799.0,
|
|
"step": 815
|
|
},
|
|
{
|
|
"entropy": 6.752606964111328,
|
|
"epoch": 0.04963229730956632,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0004095,
|
|
"loss": 6.4188,
|
|
"mean_token_accuracy": 0.1187475048005581,
|
|
"num_tokens": 1692288.0,
|
|
"step": 820
|
|
},
|
|
{
|
|
"entropy": 6.809905910491944,
|
|
"epoch": 0.04993493326877099,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.000412,
|
|
"loss": 6.4824,
|
|
"mean_token_accuracy": 0.11313270628452302,
|
|
"num_tokens": 1703602.0,
|
|
"step": 825
|
|
},
|
|
{
|
|
"entropy": 6.7061944007873535,
|
|
"epoch": 0.05023756922797567,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.0004145,
|
|
"loss": 6.4926,
|
|
"mean_token_accuracy": 0.11884824633598327,
|
|
"num_tokens": 1713413.0,
|
|
"step": 830
|
|
},
|
|
{
|
|
"entropy": 6.731452989578247,
|
|
"epoch": 0.05054020518718034,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000417,
|
|
"loss": 6.5301,
|
|
"mean_token_accuracy": 0.11357911452651023,
|
|
"num_tokens": 1724142.0,
|
|
"step": 835
|
|
},
|
|
{
|
|
"entropy": 6.763986968994141,
|
|
"epoch": 0.050842841146385015,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0004195,
|
|
"loss": 6.5569,
|
|
"mean_token_accuracy": 0.11586036384105683,
|
|
"num_tokens": 1734202.0,
|
|
"step": 840
|
|
},
|
|
{
|
|
"entropy": 6.817945718765259,
|
|
"epoch": 0.051145477105589685,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.000422,
|
|
"loss": 6.5204,
|
|
"mean_token_accuracy": 0.11986512243747711,
|
|
"num_tokens": 1744736.0,
|
|
"step": 845
|
|
},
|
|
{
|
|
"entropy": 6.76025652885437,
|
|
"epoch": 0.05144811306479436,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.0004245,
|
|
"loss": 6.398,
|
|
"mean_token_accuracy": 0.11987611725926399,
|
|
"num_tokens": 1754142.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 6.619945669174195,
|
|
"epoch": 0.05175074902399903,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000427,
|
|
"loss": 6.4384,
|
|
"mean_token_accuracy": 0.12378875017166138,
|
|
"num_tokens": 1764045.0,
|
|
"step": 855
|
|
},
|
|
{
|
|
"entropy": 6.711723947525025,
|
|
"epoch": 0.05205338498320371,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0004295,
|
|
"loss": 6.4446,
|
|
"mean_token_accuracy": 0.113596211373806,
|
|
"num_tokens": 1774580.0,
|
|
"step": 860
|
|
},
|
|
{
|
|
"entropy": 6.655456399917602,
|
|
"epoch": 0.05235602094240838,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.000432,
|
|
"loss": 6.4295,
|
|
"mean_token_accuracy": 0.11723949536681175,
|
|
"num_tokens": 1785741.0,
|
|
"step": 865
|
|
},
|
|
{
|
|
"entropy": 6.785784339904785,
|
|
"epoch": 0.052658656901613046,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004345,
|
|
"loss": 6.4901,
|
|
"mean_token_accuracy": 0.11614948362112046,
|
|
"num_tokens": 1796131.0,
|
|
"step": 870
|
|
},
|
|
{
|
|
"entropy": 6.729324054718018,
|
|
"epoch": 0.05296129286081772,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000437,
|
|
"loss": 6.4712,
|
|
"mean_token_accuracy": 0.11509689018130302,
|
|
"num_tokens": 1807404.0,
|
|
"step": 875
|
|
},
|
|
{
|
|
"entropy": 6.784345436096191,
|
|
"epoch": 0.05326392882002239,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0004395,
|
|
"loss": 6.4961,
|
|
"mean_token_accuracy": 0.1187933437526226,
|
|
"num_tokens": 1817579.0,
|
|
"step": 880
|
|
},
|
|
{
|
|
"entropy": 6.643318367004395,
|
|
"epoch": 0.05356656477922707,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000442,
|
|
"loss": 6.396,
|
|
"mean_token_accuracy": 0.12219256609678268,
|
|
"num_tokens": 1827075.0,
|
|
"step": 885
|
|
},
|
|
{
|
|
"entropy": 6.709542989730835,
|
|
"epoch": 0.05386920073843174,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0004445,
|
|
"loss": 6.5232,
|
|
"mean_token_accuracy": 0.1142655372619629,
|
|
"num_tokens": 1837556.0,
|
|
"step": 890
|
|
},
|
|
{
|
|
"entropy": 6.814684057235718,
|
|
"epoch": 0.054171836697636415,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000447,
|
|
"loss": 6.4722,
|
|
"mean_token_accuracy": 0.12146498113870621,
|
|
"num_tokens": 1847939.0,
|
|
"step": 895
|
|
},
|
|
{
|
|
"entropy": 6.674838018417359,
|
|
"epoch": 0.054474472656841084,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.00044950000000000003,
|
|
"loss": 6.4691,
|
|
"mean_token_accuracy": 0.1165416806936264,
|
|
"num_tokens": 1857837.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 6.669571352005005,
|
|
"epoch": 0.05477710861604576,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00045200000000000004,
|
|
"loss": 6.3874,
|
|
"mean_token_accuracy": 0.12307673171162606,
|
|
"num_tokens": 1868341.0,
|
|
"step": 905
|
|
},
|
|
{
|
|
"entropy": 6.622711086273194,
|
|
"epoch": 0.05507974457525043,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.00045450000000000004,
|
|
"loss": 6.468,
|
|
"mean_token_accuracy": 0.11901665478944778,
|
|
"num_tokens": 1879798.0,
|
|
"step": 910
|
|
},
|
|
{
|
|
"entropy": 6.7006079196929935,
|
|
"epoch": 0.05538238053445511,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00045700000000000005,
|
|
"loss": 6.449,
|
|
"mean_token_accuracy": 0.12047585397958756,
|
|
"num_tokens": 1890685.0,
|
|
"step": 915
|
|
},
|
|
{
|
|
"entropy": 6.603058910369873,
|
|
"epoch": 0.05568501649365978,
|
|
"grad_norm": 0.9765625,
|
|
"learning_rate": 0.00045950000000000006,
|
|
"loss": 6.4091,
|
|
"mean_token_accuracy": 0.1215558484196663,
|
|
"num_tokens": 1900870.0,
|
|
"step": 920
|
|
},
|
|
{
|
|
"entropy": 6.610906982421875,
|
|
"epoch": 0.055987652452864446,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000462,
|
|
"loss": 6.44,
|
|
"mean_token_accuracy": 0.11300771608948708,
|
|
"num_tokens": 1911562.0,
|
|
"step": 925
|
|
},
|
|
{
|
|
"entropy": 6.643617773056031,
|
|
"epoch": 0.05629028841206912,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0004645,
|
|
"loss": 6.4238,
|
|
"mean_token_accuracy": 0.11745399534702301,
|
|
"num_tokens": 1922302.0,
|
|
"step": 930
|
|
},
|
|
{
|
|
"entropy": 6.712780857086182,
|
|
"epoch": 0.05659292437127379,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000467,
|
|
"loss": 6.4711,
|
|
"mean_token_accuracy": 0.11841732487082482,
|
|
"num_tokens": 1933022.0,
|
|
"step": 935
|
|
},
|
|
{
|
|
"entropy": 6.693202972412109,
|
|
"epoch": 0.05689556033047847,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0004695,
|
|
"loss": 6.5237,
|
|
"mean_token_accuracy": 0.12020932510495186,
|
|
"num_tokens": 1943630.0,
|
|
"step": 940
|
|
},
|
|
{
|
|
"entropy": 6.706296968460083,
|
|
"epoch": 0.05719819628968314,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000472,
|
|
"loss": 6.4836,
|
|
"mean_token_accuracy": 0.11581766158342362,
|
|
"num_tokens": 1954665.0,
|
|
"step": 945
|
|
},
|
|
{
|
|
"entropy": 6.666206693649292,
|
|
"epoch": 0.057500832248887815,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0004745,
|
|
"loss": 6.516,
|
|
"mean_token_accuracy": 0.1156513161957264,
|
|
"num_tokens": 1965383.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 6.690737009048462,
|
|
"epoch": 0.057803468208092484,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000477,
|
|
"loss": 6.4326,
|
|
"mean_token_accuracy": 0.1202012687921524,
|
|
"num_tokens": 1975087.0,
|
|
"step": 955
|
|
},
|
|
{
|
|
"entropy": 6.634597635269165,
|
|
"epoch": 0.05810610416729716,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0004795,
|
|
"loss": 6.4259,
|
|
"mean_token_accuracy": 0.12529866620898247,
|
|
"num_tokens": 1985507.0,
|
|
"step": 960
|
|
},
|
|
{
|
|
"entropy": 6.632993459701538,
|
|
"epoch": 0.05840874012650183,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000482,
|
|
"loss": 6.376,
|
|
"mean_token_accuracy": 0.11986934766173363,
|
|
"num_tokens": 1997943.0,
|
|
"step": 965
|
|
},
|
|
{
|
|
"entropy": 6.570653009414673,
|
|
"epoch": 0.05871137608570651,
|
|
"grad_norm": 0.921875,
|
|
"learning_rate": 0.0004845,
|
|
"loss": 6.3299,
|
|
"mean_token_accuracy": 0.12246516719460487,
|
|
"num_tokens": 2008472.0,
|
|
"step": 970
|
|
},
|
|
{
|
|
"entropy": 6.606798601150513,
|
|
"epoch": 0.05901401204491118,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000487,
|
|
"loss": 6.4283,
|
|
"mean_token_accuracy": 0.11940227895975113,
|
|
"num_tokens": 2018448.0,
|
|
"step": 975
|
|
},
|
|
{
|
|
"entropy": 6.714173650741577,
|
|
"epoch": 0.059316648004115846,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.0004895,
|
|
"loss": 6.4469,
|
|
"mean_token_accuracy": 0.1235787332057953,
|
|
"num_tokens": 2029030.0,
|
|
"step": 980
|
|
},
|
|
{
|
|
"entropy": 6.552170896530152,
|
|
"epoch": 0.05961928396332052,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.000492,
|
|
"loss": 6.3973,
|
|
"mean_token_accuracy": 0.12218492925167083,
|
|
"num_tokens": 2040111.0,
|
|
"step": 985
|
|
},
|
|
{
|
|
"entropy": 6.676973581314087,
|
|
"epoch": 0.05992191992252519,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004945,
|
|
"loss": 6.4198,
|
|
"mean_token_accuracy": 0.11496246978640556,
|
|
"num_tokens": 2050392.0,
|
|
"step": 990
|
|
},
|
|
{
|
|
"entropy": 6.612374591827392,
|
|
"epoch": 0.06022455588172987,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000497,
|
|
"loss": 6.4309,
|
|
"mean_token_accuracy": 0.12264466732740402,
|
|
"num_tokens": 2060915.0,
|
|
"step": 995
|
|
},
|
|
{
|
|
"entropy": 6.635892105102539,
|
|
"epoch": 0.06052719184093454,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004995,
|
|
"loss": 6.472,
|
|
"mean_token_accuracy": 0.1088255800306797,
|
|
"num_tokens": 2072710.0,
|
|
"step": 1000
|
|
}
|
|
],
|
|
"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": 447677704765440.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|