1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.011172310545543924,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.69185152053833,
|
|
"epoch": 0.00011172310545543924,
|
|
"grad_norm": 18.5,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.5751,
|
|
"mean_token_accuracy": 0.00030030030757188797,
|
|
"num_tokens": 4264.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691861820220947,
|
|
"epoch": 0.00022344621091087847,
|
|
"grad_norm": 16.75,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.4625,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 8413.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.690832233428955,
|
|
"epoch": 0.0003351693163663177,
|
|
"grad_norm": 13.4375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.1956,
|
|
"mean_token_accuracy": 0.03499803415033966,
|
|
"num_tokens": 12394.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.681817626953125,
|
|
"epoch": 0.00044689242182175694,
|
|
"grad_norm": 8.4375,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 9.8891,
|
|
"mean_token_accuracy": 0.052673622593283655,
|
|
"num_tokens": 16610.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.648191356658936,
|
|
"epoch": 0.0005586155272771962,
|
|
"grad_norm": 5.84375,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.6415,
|
|
"mean_token_accuracy": 0.059712447226047516,
|
|
"num_tokens": 21243.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.600524044036865,
|
|
"epoch": 0.0006703386327326354,
|
|
"grad_norm": 4.21875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.3955,
|
|
"mean_token_accuracy": 0.06313439272344112,
|
|
"num_tokens": 25571.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.565627956390381,
|
|
"epoch": 0.0007820617381880747,
|
|
"grad_norm": 4.28125,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.2501,
|
|
"mean_token_accuracy": 0.058146678283810614,
|
|
"num_tokens": 29412.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.537028217315674,
|
|
"epoch": 0.0008937848436435139,
|
|
"grad_norm": 3.796875,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.2173,
|
|
"mean_token_accuracy": 0.061766361258924006,
|
|
"num_tokens": 33893.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.526171112060547,
|
|
"epoch": 0.0010055079490989532,
|
|
"grad_norm": 3.765625,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.1499,
|
|
"mean_token_accuracy": 0.06647200398147106,
|
|
"num_tokens": 38043.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.459559440612793,
|
|
"epoch": 0.0011172310545543925,
|
|
"grad_norm": 3.296875,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.0927,
|
|
"mean_token_accuracy": 0.06208832561969757,
|
|
"num_tokens": 42222.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.321250820159912,
|
|
"epoch": 0.0012289541600098315,
|
|
"grad_norm": 3.640625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 8.9929,
|
|
"mean_token_accuracy": 0.06888641528785229,
|
|
"num_tokens": 46321.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.221459579467773,
|
|
"epoch": 0.0013406772654652708,
|
|
"grad_norm": 3.140625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.0121,
|
|
"mean_token_accuracy": 0.07621248066425323,
|
|
"num_tokens": 50816.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.167527294158935,
|
|
"epoch": 0.0014524003709207101,
|
|
"grad_norm": 3.046875,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 8.9766,
|
|
"mean_token_accuracy": 0.06631188206374646,
|
|
"num_tokens": 55331.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.158582305908203,
|
|
"epoch": 0.0015641234763761494,
|
|
"grad_norm": 2.828125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 8.8916,
|
|
"mean_token_accuracy": 0.06474936045706273,
|
|
"num_tokens": 59379.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.132518291473389,
|
|
"epoch": 0.0016758465818315885,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 8.7691,
|
|
"mean_token_accuracy": 0.07134371958673,
|
|
"num_tokens": 63463.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.115495491027833,
|
|
"epoch": 0.0017875696872870278,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 8.8345,
|
|
"mean_token_accuracy": 0.07139550410211086,
|
|
"num_tokens": 67897.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.076316165924073,
|
|
"epoch": 0.001899292792742467,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 8.6081,
|
|
"mean_token_accuracy": 0.07515238299965858,
|
|
"num_tokens": 71995.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.003674030303955,
|
|
"epoch": 0.0020110158981979064,
|
|
"grad_norm": 2.78125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.4613,
|
|
"mean_token_accuracy": 0.07963024936616421,
|
|
"num_tokens": 75745.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 9.955099964141846,
|
|
"epoch": 0.0021227390036533456,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.4039,
|
|
"mean_token_accuracy": 0.0783398538827896,
|
|
"num_tokens": 79641.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 9.907726192474366,
|
|
"epoch": 0.002234462109108785,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.3639,
|
|
"mean_token_accuracy": 0.07564271241426468,
|
|
"num_tokens": 84368.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 9.777185916900635,
|
|
"epoch": 0.0023461852145642242,
|
|
"grad_norm": 3.046875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.3562,
|
|
"mean_token_accuracy": 0.07499068863689899,
|
|
"num_tokens": 89177.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 9.874152565002442,
|
|
"epoch": 0.002457908320019663,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.1616,
|
|
"mean_token_accuracy": 0.08200128600001336,
|
|
"num_tokens": 93266.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 9.573559665679932,
|
|
"epoch": 0.0025696314254751024,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.0628,
|
|
"mean_token_accuracy": 0.08064724728465081,
|
|
"num_tokens": 97385.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.477795314788818,
|
|
"epoch": 0.0026813545309305417,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 7.8909,
|
|
"mean_token_accuracy": 0.08364077880978585,
|
|
"num_tokens": 100911.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.406207180023193,
|
|
"epoch": 0.002793077636385981,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 7.8101,
|
|
"mean_token_accuracy": 0.08597532883286477,
|
|
"num_tokens": 104785.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.198468017578126,
|
|
"epoch": 0.0029048007418414202,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.7507,
|
|
"mean_token_accuracy": 0.08100780844688416,
|
|
"num_tokens": 109434.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.137882709503174,
|
|
"epoch": 0.0030165238472968595,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.7526,
|
|
"mean_token_accuracy": 0.08647556751966476,
|
|
"num_tokens": 113869.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 8.982115554809571,
|
|
"epoch": 0.003128246952752299,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.6855,
|
|
"mean_token_accuracy": 0.0863442875444889,
|
|
"num_tokens": 118332.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.803977680206298,
|
|
"epoch": 0.003239970058207738,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.5951,
|
|
"mean_token_accuracy": 0.07965997867286205,
|
|
"num_tokens": 122140.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.717557144165038,
|
|
"epoch": 0.003351693163663177,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.4489,
|
|
"mean_token_accuracy": 0.08853708282113075,
|
|
"num_tokens": 126515.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.480121898651124,
|
|
"epoch": 0.0034634162691186163,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.5078,
|
|
"mean_token_accuracy": 0.0842710092663765,
|
|
"num_tokens": 130397.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.55941562652588,
|
|
"epoch": 0.0035751393745740555,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.448,
|
|
"mean_token_accuracy": 0.086439348757267,
|
|
"num_tokens": 134428.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.38999900817871,
|
|
"epoch": 0.003686862480029495,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.3523,
|
|
"mean_token_accuracy": 0.09324855357408524,
|
|
"num_tokens": 138294.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.155976009368896,
|
|
"epoch": 0.003798585585484934,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.3474,
|
|
"mean_token_accuracy": 0.08743810728192329,
|
|
"num_tokens": 142750.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.256286525726319,
|
|
"epoch": 0.003910308690940373,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.3177,
|
|
"mean_token_accuracy": 0.08836911171674729,
|
|
"num_tokens": 147220.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.98900089263916,
|
|
"epoch": 0.004022031796395813,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.2059,
|
|
"mean_token_accuracy": 0.09879823476076126,
|
|
"num_tokens": 151396.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.085029649734498,
|
|
"epoch": 0.0041337549018512516,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.2565,
|
|
"mean_token_accuracy": 0.0990697756409645,
|
|
"num_tokens": 155336.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.046411085128785,
|
|
"epoch": 0.004245478007306691,
|
|
"grad_norm": 3.03125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.2963,
|
|
"mean_token_accuracy": 0.08894443586468696,
|
|
"num_tokens": 159752.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.043665552139283,
|
|
"epoch": 0.00435720111276213,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.3874,
|
|
"mean_token_accuracy": 0.08465041853487491,
|
|
"num_tokens": 163895.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.914860630035401,
|
|
"epoch": 0.00446892421821757,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.1927,
|
|
"mean_token_accuracy": 0.093280029296875,
|
|
"num_tokens": 168312.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.915866088867188,
|
|
"epoch": 0.004580647323673009,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.2334,
|
|
"mean_token_accuracy": 0.09746119827032089,
|
|
"num_tokens": 172358.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.660448884963989,
|
|
"epoch": 0.0046923704291284484,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.2206,
|
|
"mean_token_accuracy": 0.09569770470261574,
|
|
"num_tokens": 177085.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.943923139572144,
|
|
"epoch": 0.004804093534583887,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.3482,
|
|
"mean_token_accuracy": 0.08808798864483833,
|
|
"num_tokens": 181410.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.640953445434571,
|
|
"epoch": 0.004915816640039326,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.0931,
|
|
"mean_token_accuracy": 0.0959208883345127,
|
|
"num_tokens": 185649.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.714695310592651,
|
|
"epoch": 0.005027539745494766,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.1461,
|
|
"mean_token_accuracy": 0.09635853916406631,
|
|
"num_tokens": 190055.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.722128820419312,
|
|
"epoch": 0.005139262850950205,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.1545,
|
|
"mean_token_accuracy": 0.10052503794431686,
|
|
"num_tokens": 194267.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.67925763130188,
|
|
"epoch": 0.0052509859564056445,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.0976,
|
|
"mean_token_accuracy": 0.09923605695366859,
|
|
"num_tokens": 198483.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.5812575340271,
|
|
"epoch": 0.005362709061861083,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.0173,
|
|
"mean_token_accuracy": 0.10283883661031723,
|
|
"num_tokens": 202514.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.598950576782227,
|
|
"epoch": 0.005474432167316523,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.1592,
|
|
"mean_token_accuracy": 0.09925142973661423,
|
|
"num_tokens": 206748.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.573145723342895,
|
|
"epoch": 0.005586155272771962,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.0229,
|
|
"mean_token_accuracy": 0.10259764194488526,
|
|
"num_tokens": 210981.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.5371081829071045,
|
|
"epoch": 0.005697878378227402,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.1679,
|
|
"mean_token_accuracy": 0.09469496235251426,
|
|
"num_tokens": 215551.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.520426607131958,
|
|
"epoch": 0.0058096014836828405,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 6.9739,
|
|
"mean_token_accuracy": 0.11107034981250763,
|
|
"num_tokens": 219677.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.612557506561279,
|
|
"epoch": 0.005921324589138279,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.072,
|
|
"mean_token_accuracy": 0.10267340168356895,
|
|
"num_tokens": 223457.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.546310329437256,
|
|
"epoch": 0.006033047694593719,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 6.9226,
|
|
"mean_token_accuracy": 0.1154308445751667,
|
|
"num_tokens": 227135.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.572003746032715,
|
|
"epoch": 0.006144770800049158,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.1372,
|
|
"mean_token_accuracy": 0.09673138484358787,
|
|
"num_tokens": 231393.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.611022663116455,
|
|
"epoch": 0.006256493905504598,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.1052,
|
|
"mean_token_accuracy": 0.10218687430024147,
|
|
"num_tokens": 235959.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.403985118865966,
|
|
"epoch": 0.0063682170109600365,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 6.9434,
|
|
"mean_token_accuracy": 0.10860099717974663,
|
|
"num_tokens": 239875.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.420692682266235,
|
|
"epoch": 0.006479940116415476,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.0656,
|
|
"mean_token_accuracy": 0.11386927664279937,
|
|
"num_tokens": 244397.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.5308808326721195,
|
|
"epoch": 0.006591663221870915,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.0,
|
|
"mean_token_accuracy": 0.10292632281780242,
|
|
"num_tokens": 248441.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.551111650466919,
|
|
"epoch": 0.006703386327326354,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.0078,
|
|
"mean_token_accuracy": 0.10877085924148559,
|
|
"num_tokens": 252693.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.420186233520508,
|
|
"epoch": 0.006815109432781794,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.9778,
|
|
"mean_token_accuracy": 0.1077166736125946,
|
|
"num_tokens": 257135.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.455549049377441,
|
|
"epoch": 0.0069268325382372325,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.101,
|
|
"mean_token_accuracy": 0.09917000532150269,
|
|
"num_tokens": 261919.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.428725242614746,
|
|
"epoch": 0.007038555643692672,
|
|
"grad_norm": 1.71875,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0018,
|
|
"mean_token_accuracy": 0.1032940112054348,
|
|
"num_tokens": 266656.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.4056861877441404,
|
|
"epoch": 0.007150278749148111,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.9649,
|
|
"mean_token_accuracy": 0.10402853935956954,
|
|
"num_tokens": 270879.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.356624603271484,
|
|
"epoch": 0.007262001854603551,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.9156,
|
|
"mean_token_accuracy": 0.11899577528238296,
|
|
"num_tokens": 274884.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.359720230102539,
|
|
"epoch": 0.00737372496005899,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.8492,
|
|
"mean_token_accuracy": 0.10880193561315536,
|
|
"num_tokens": 279096.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.420133590698242,
|
|
"epoch": 0.007485448065514429,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.9547,
|
|
"mean_token_accuracy": 0.10481383726000786,
|
|
"num_tokens": 282923.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.375938367843628,
|
|
"epoch": 0.007597171170969868,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.8854,
|
|
"mean_token_accuracy": 0.11555279865860939,
|
|
"num_tokens": 287283.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.266809701919556,
|
|
"epoch": 0.007708894276425307,
|
|
"grad_norm": 1.671875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.8566,
|
|
"mean_token_accuracy": 0.11106687858700752,
|
|
"num_tokens": 291409.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.351815128326416,
|
|
"epoch": 0.007820617381880746,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9102,
|
|
"mean_token_accuracy": 0.10942614153027534,
|
|
"num_tokens": 295157.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.481018209457398,
|
|
"epoch": 0.007932340487336187,
|
|
"grad_norm": 1.9765625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.9203,
|
|
"mean_token_accuracy": 0.10715582817792893,
|
|
"num_tokens": 299370.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.2193928241729735,
|
|
"epoch": 0.008044063592791625,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.9256,
|
|
"mean_token_accuracy": 0.11601178944110871,
|
|
"num_tokens": 303955.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.46614179611206,
|
|
"epoch": 0.008155786698247064,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.8992,
|
|
"mean_token_accuracy": 0.11738249063491821,
|
|
"num_tokens": 308047.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.339001560211182,
|
|
"epoch": 0.008267509803702503,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.8353,
|
|
"mean_token_accuracy": 0.12007508799433708,
|
|
"num_tokens": 312050.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.393199253082275,
|
|
"epoch": 0.008379232909157944,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.789,
|
|
"mean_token_accuracy": 0.11748188808560371,
|
|
"num_tokens": 315901.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.206725311279297,
|
|
"epoch": 0.008490956014613383,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.8387,
|
|
"mean_token_accuracy": 0.11932774782180786,
|
|
"num_tokens": 320077.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.309018516540528,
|
|
"epoch": 0.008602679120068821,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.7772,
|
|
"mean_token_accuracy": 0.12389342486858368,
|
|
"num_tokens": 324365.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.1502117156982425,
|
|
"epoch": 0.00871440222552426,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.7342,
|
|
"mean_token_accuracy": 0.12538171485066413,
|
|
"num_tokens": 327971.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.207239198684692,
|
|
"epoch": 0.0088261253309797,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.7814,
|
|
"mean_token_accuracy": 0.12070702090859413,
|
|
"num_tokens": 331791.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.188077354431153,
|
|
"epoch": 0.00893784843643514,
|
|
"grad_norm": 1.7265625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.8347,
|
|
"mean_token_accuracy": 0.11772776916623115,
|
|
"num_tokens": 335616.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.238067245483398,
|
|
"epoch": 0.009049571541890579,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.7448,
|
|
"mean_token_accuracy": 0.12914832159876824,
|
|
"num_tokens": 339815.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.271863889694214,
|
|
"epoch": 0.009161294647346017,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.7314,
|
|
"mean_token_accuracy": 0.12098882421851158,
|
|
"num_tokens": 344045.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.054260921478272,
|
|
"epoch": 0.009273017752801456,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.8258,
|
|
"mean_token_accuracy": 0.12370770499110222,
|
|
"num_tokens": 348726.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.25892219543457,
|
|
"epoch": 0.009384740858256897,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.811,
|
|
"mean_token_accuracy": 0.12439991980791092,
|
|
"num_tokens": 353228.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.089362478256225,
|
|
"epoch": 0.009496463963712336,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.7731,
|
|
"mean_token_accuracy": 0.12979045361280442,
|
|
"num_tokens": 357379.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.264574861526489,
|
|
"epoch": 0.009608187069167775,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.7426,
|
|
"mean_token_accuracy": 0.11879147663712501,
|
|
"num_tokens": 361347.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.1300811767578125,
|
|
"epoch": 0.009719910174623213,
|
|
"grad_norm": 1.7265625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.7019,
|
|
"mean_token_accuracy": 0.12289151027798653,
|
|
"num_tokens": 365364.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.1097527027130125,
|
|
"epoch": 0.009831633280078652,
|
|
"grad_norm": 1.6171875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.7534,
|
|
"mean_token_accuracy": 0.12068467438220978,
|
|
"num_tokens": 369648.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.229344892501831,
|
|
"epoch": 0.009943356385534093,
|
|
"grad_norm": 1.578125,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.7309,
|
|
"mean_token_accuracy": 0.12263526245951653,
|
|
"num_tokens": 373734.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.1849524021148685,
|
|
"epoch": 0.010055079490989532,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.7276,
|
|
"mean_token_accuracy": 0.12742977142333983,
|
|
"num_tokens": 378184.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.064081621170044,
|
|
"epoch": 0.01016680259644497,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.8377,
|
|
"mean_token_accuracy": 0.12056938409805298,
|
|
"num_tokens": 382686.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.035848474502563,
|
|
"epoch": 0.01027852570190041,
|
|
"grad_norm": 1.84375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.7467,
|
|
"mean_token_accuracy": 0.13145566657185553,
|
|
"num_tokens": 387210.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.06201696395874,
|
|
"epoch": 0.01039024880735585,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.7263,
|
|
"mean_token_accuracy": 0.12420300841331482,
|
|
"num_tokens": 391247.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.2231800079345705,
|
|
"epoch": 0.010501971912811289,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.6101,
|
|
"mean_token_accuracy": 0.12484818920493126,
|
|
"num_tokens": 395188.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.963715267181397,
|
|
"epoch": 0.010613695018266728,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.6759,
|
|
"mean_token_accuracy": 0.1323227606713772,
|
|
"num_tokens": 399513.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.180826234817505,
|
|
"epoch": 0.010725418123722167,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.7397,
|
|
"mean_token_accuracy": 0.1241152174770832,
|
|
"num_tokens": 404052.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.000790357589722,
|
|
"epoch": 0.010837141229177606,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6031,
|
|
"mean_token_accuracy": 0.13010914325714112,
|
|
"num_tokens": 408299.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.951791143417358,
|
|
"epoch": 0.010948864334633046,
|
|
"grad_norm": 1.890625,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.597,
|
|
"mean_token_accuracy": 0.13655143678188325,
|
|
"num_tokens": 412376.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.940329837799072,
|
|
"epoch": 0.011060587440088485,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.6731,
|
|
"mean_token_accuracy": 0.13289386332035064,
|
|
"num_tokens": 417010.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.088097476959229,
|
|
"epoch": 0.011172310545543924,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.5194,
|
|
"mean_token_accuracy": 0.14706601724028587,
|
|
"num_tokens": 421167.0,
|
|
"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": 614281236480000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|