1155 lines
30 KiB
JSON
1155 lines
30 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.7822430820377432,
|
|
"eval_steps": 500,
|
|
"global_step": 4000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.004889019262735895,
|
|
"grad_norm": 1.8119502067565918,
|
|
"learning_rate": 9.411764705882352e-05,
|
|
"loss": 9.918764038085937,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.00977803852547179,
|
|
"grad_norm": 1.0673187971115112,
|
|
"learning_rate": 0.000192156862745098,
|
|
"loss": 8.179608154296876,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.014667057788207686,
|
|
"grad_norm": 0.989127516746521,
|
|
"learning_rate": 0.0002901960784313725,
|
|
"loss": 7.328287963867187,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.01955607705094358,
|
|
"grad_norm": 1.2597979307174683,
|
|
"learning_rate": 0.000388235294117647,
|
|
"loss": 6.83828125,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.024445096313679474,
|
|
"grad_norm": 0.7706186771392822,
|
|
"learning_rate": 0.00048627450980392154,
|
|
"loss": 6.4037939453125,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.02933411557641537,
|
|
"grad_norm": 0.9577609300613403,
|
|
"learning_rate": 0.0005843137254901961,
|
|
"loss": 6.09682861328125,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.034223134839151265,
|
|
"grad_norm": 0.9290236830711365,
|
|
"learning_rate": 0.0005999734732075449,
|
|
"loss": 5.854271240234375,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.03911215410188716,
|
|
"grad_norm": 0.5200176239013672,
|
|
"learning_rate": 0.0005998727266417804,
|
|
"loss": 5.61611083984375,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.04400117336462306,
|
|
"grad_norm": 0.5042134523391724,
|
|
"learning_rate": 0.0005996968231157925,
|
|
"loss": 5.415563354492187,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.04889019262735895,
|
|
"grad_norm": 0.7687386870384216,
|
|
"learning_rate": 0.000599445806716199,
|
|
"loss": 5.224786376953125,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.053779211890094845,
|
|
"grad_norm": 0.7010864019393921,
|
|
"learning_rate": 0.000599119740355122,
|
|
"loss": 5.0750430297851565,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.05866823115283074,
|
|
"grad_norm": 0.4438683092594147,
|
|
"learning_rate": 0.0005987187057544201,
|
|
"loss": 4.935214233398438,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.06355725041556663,
|
|
"grad_norm": 0.4158670902252197,
|
|
"learning_rate": 0.0005982428034252062,
|
|
"loss": 4.824454650878907,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 0.06844626967830253,
|
|
"grad_norm": 0.5508437156677246,
|
|
"learning_rate": 0.0005976921526426568,
|
|
"loss": 4.713559875488281,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.07333528894103843,
|
|
"grad_norm": 0.39377203583717346,
|
|
"learning_rate": 0.0005970668914161184,
|
|
"loss": 4.589186401367187,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 0.07822430820377432,
|
|
"grad_norm": 0.5569614171981812,
|
|
"learning_rate": 0.0005963671764545175,
|
|
"loss": 4.531298522949219,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.08311332746651022,
|
|
"grad_norm": 0.42474564909935,
|
|
"learning_rate": 0.0005955931831270854,
|
|
"loss": 4.437086181640625,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 0.08800234672924612,
|
|
"grad_norm": 0.4873526394367218,
|
|
"learning_rate": 0.0005947451054194051,
|
|
"loss": 4.356460876464844,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.09289136599198201,
|
|
"grad_norm": 0.4566892087459564,
|
|
"learning_rate": 0.0005938231558847931,
|
|
"loss": 4.277253112792969,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 0.0977803852547179,
|
|
"grad_norm": 0.5659822821617126,
|
|
"learning_rate": 0.0005928275655910269,
|
|
"loss": 4.208056640625,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.1026694045174538,
|
|
"grad_norm": 0.3656966984272003,
|
|
"learning_rate": 0.0005917585840624328,
|
|
"loss": 4.158276977539063,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 0.10755842378018969,
|
|
"grad_norm": 0.3782542943954468,
|
|
"learning_rate": 0.0005906164792173478,
|
|
"loss": 4.085022888183594,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.11244744304292559,
|
|
"grad_norm": 0.39630722999572754,
|
|
"learning_rate": 0.0005894015373009715,
|
|
"loss": 4.033250732421875,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 0.11733646230566148,
|
|
"grad_norm": 0.37524425983428955,
|
|
"learning_rate": 0.0005881140628136237,
|
|
"loss": 3.992838134765625,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.12222548156839738,
|
|
"grad_norm": 0.4007509648799896,
|
|
"learning_rate": 0.0005867543784344289,
|
|
"loss": 3.956600341796875,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 0.12711450083113326,
|
|
"grad_norm": 0.3884464204311371,
|
|
"learning_rate": 0.0005853228249404426,
|
|
"loss": 3.92346435546875,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.13200352009386918,
|
|
"grad_norm": 0.28836891055107117,
|
|
"learning_rate": 0.0005838197611212426,
|
|
"loss": 3.875860290527344,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 0.13689253935660506,
|
|
"grad_norm": 0.3205326795578003,
|
|
"learning_rate": 0.0005822455636890059,
|
|
"loss": 3.824952392578125,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.14178155861934097,
|
|
"grad_norm": 0.398606538772583,
|
|
"learning_rate": 0.0005806006271840933,
|
|
"loss": 3.82131103515625,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 0.14667057788207685,
|
|
"grad_norm": 0.7978962063789368,
|
|
"learning_rate": 0.0005788853638761664,
|
|
"loss": 3.774755859375,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.15155959714481276,
|
|
"grad_norm": 0.2822282910346985,
|
|
"learning_rate": 0.0005771002036608601,
|
|
"loss": 3.7683151245117186,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 0.15644861640754865,
|
|
"grad_norm": 0.31294363737106323,
|
|
"learning_rate": 0.000575245593952038,
|
|
"loss": 3.7285302734375,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.16133763567028453,
|
|
"grad_norm": 0.3369923532009125,
|
|
"learning_rate": 0.0005733219995696581,
|
|
"loss": 3.7152362060546875,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 0.16622665493302044,
|
|
"grad_norm": 0.334830641746521,
|
|
"learning_rate": 0.0005713299026232741,
|
|
"loss": 3.680809326171875,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.17111567419575632,
|
|
"grad_norm": 0.35272541642189026,
|
|
"learning_rate": 0.0005692698023912055,
|
|
"loss": 3.6764453125,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 0.17600469345849223,
|
|
"grad_norm": 0.36247503757476807,
|
|
"learning_rate": 0.0005671422151954029,
|
|
"loss": 3.6342178344726563,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.18089371272122812,
|
|
"grad_norm": 0.2963731586933136,
|
|
"learning_rate": 0.0005649476742720431,
|
|
"loss": 3.633841552734375,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 0.18578273198396403,
|
|
"grad_norm": 0.32037004828453064,
|
|
"learning_rate": 0.0005626867296378839,
|
|
"loss": 3.6143218994140627,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.1906717512466999,
|
|
"grad_norm": 0.3355698585510254,
|
|
"learning_rate": 0.0005603599479524141,
|
|
"loss": 3.6056793212890623,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 0.1955607705094358,
|
|
"grad_norm": 0.270993709564209,
|
|
"learning_rate": 0.0005579679123758313,
|
|
"loss": 3.5813720703125,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.2004497897721717,
|
|
"grad_norm": 0.30514687299728394,
|
|
"learning_rate": 0.0005555112224228853,
|
|
"loss": 3.560430908203125,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 0.2053388090349076,
|
|
"grad_norm": 0.2816086411476135,
|
|
"learning_rate": 0.000552990493812622,
|
|
"loss": 3.5420367431640627,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.2102278282976435,
|
|
"grad_norm": 0.3174511194229126,
|
|
"learning_rate": 0.0005504063583140653,
|
|
"loss": 3.509090576171875,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 0.21511684756037938,
|
|
"grad_norm": 0.2977951765060425,
|
|
"learning_rate": 0.0005477594635878782,
|
|
"loss": 3.516375732421875,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.2200058668231153,
|
|
"grad_norm": 0.2921285033226013,
|
|
"learning_rate": 0.000545050473024039,
|
|
"loss": 3.4910296630859374,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 0.22489488608585118,
|
|
"grad_norm": 0.28328269720077515,
|
|
"learning_rate": 0.0005422800655755768,
|
|
"loss": 3.4671282958984375,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.2297839053485871,
|
|
"grad_norm": 0.2933865487575531,
|
|
"learning_rate": 0.000539448935588405,
|
|
"loss": 3.490953369140625,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 0.23467292461132297,
|
|
"grad_norm": 0.2856893539428711,
|
|
"learning_rate": 0.0005365577926272991,
|
|
"loss": 3.46058837890625,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.23956194387405885,
|
|
"grad_norm": 0.28969162702560425,
|
|
"learning_rate": 0.0005336073612980574,
|
|
"loss": 3.45246337890625,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 0.24445096313679476,
|
|
"grad_norm": 0.28745341300964355,
|
|
"learning_rate": 0.0005305983810658942,
|
|
"loss": 3.4447247314453127,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.24933998239953065,
|
|
"grad_norm": 0.3049822747707367,
|
|
"learning_rate": 0.0005275316060701078,
|
|
"loss": 3.4294424438476563,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 0.25422900166226653,
|
|
"grad_norm": 0.3087104260921478,
|
|
"learning_rate": 0.0005244078049350704,
|
|
"loss": 3.434173889160156,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.25911802092500247,
|
|
"grad_norm": 0.27349814772605896,
|
|
"learning_rate": 0.0005212277605775895,
|
|
"loss": 3.4029391479492186,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 0.26400704018773835,
|
|
"grad_norm": 0.26429903507232666,
|
|
"learning_rate": 0.0005179922700106841,
|
|
"loss": 3.409964904785156,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.26889605945047423,
|
|
"grad_norm": 0.25591129064559937,
|
|
"learning_rate": 0.0005147021441438311,
|
|
"loss": 3.4059173583984377,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 0.2737850787132101,
|
|
"grad_norm": 0.283764123916626,
|
|
"learning_rate": 0.0005113582075797267,
|
|
"loss": 3.3744482421875,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.278674097975946,
|
|
"grad_norm": 0.24849016964435577,
|
|
"learning_rate": 0.0005079612984076167,
|
|
"loss": 3.35540771484375,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 0.28356311723868194,
|
|
"grad_norm": 0.26382192969322205,
|
|
"learning_rate": 0.0005045122679932468,
|
|
"loss": 3.372743835449219,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.2884521365014178,
|
|
"grad_norm": 0.2579769790172577,
|
|
"learning_rate": 0.0005010119807654852,
|
|
"loss": 3.359498291015625,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 0.2933411557641537,
|
|
"grad_norm": 0.27852851152420044,
|
|
"learning_rate": 0.0004974613139996708,
|
|
"loss": 3.33959228515625,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.2982301750268896,
|
|
"grad_norm": 0.25548404455184937,
|
|
"learning_rate": 0.0004938611575977429,
|
|
"loss": 3.3380303955078126,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 0.3031191942896255,
|
|
"grad_norm": 0.27390819787979126,
|
|
"learning_rate": 0.0004902124138652042,
|
|
"loss": 3.316756286621094,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.3080082135523614,
|
|
"grad_norm": 0.2832481265068054,
|
|
"learning_rate": 0.0004865159972849781,
|
|
"loss": 3.3193817138671875,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 0.3128972328150973,
|
|
"grad_norm": 0.2510474920272827,
|
|
"learning_rate": 0.0004827728342882107,
|
|
"loss": 3.3231927490234376,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.3177862520778332,
|
|
"grad_norm": 0.24408160150051117,
|
|
"learning_rate": 0.00047898386302208066,
|
|
"loss": 3.285135498046875,
|
|
"step": 1625
|
|
},
|
|
{
|
|
"epoch": 0.32267527134056906,
|
|
"grad_norm": 0.2493797391653061,
|
|
"learning_rate": 0.00047515003311467157,
|
|
"loss": 3.2993060302734376,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.327564290603305,
|
|
"grad_norm": 0.24639739096164703,
|
|
"learning_rate": 0.0004712723054369673,
|
|
"loss": 3.297099609375,
|
|
"step": 1675
|
|
},
|
|
{
|
|
"epoch": 0.3324533098660409,
|
|
"grad_norm": 0.2502795159816742,
|
|
"learning_rate": 0.00046735165186202905,
|
|
"loss": 3.295542907714844,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.33734232912877676,
|
|
"grad_norm": 0.29107826948165894,
|
|
"learning_rate": 0.00046338905502141557,
|
|
"loss": 3.2876678466796876,
|
|
"step": 1725
|
|
},
|
|
{
|
|
"epoch": 0.34223134839151265,
|
|
"grad_norm": 0.2635728418827057,
|
|
"learning_rate": 0.0004593855080589063,
|
|
"loss": 3.268152160644531,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.3471203676542486,
|
|
"grad_norm": 0.2607461214065552,
|
|
"learning_rate": 0.00045534201438159077,
|
|
"loss": 3.2758914184570314,
|
|
"step": 1775
|
|
},
|
|
{
|
|
"epoch": 0.35200938691698447,
|
|
"grad_norm": 0.2734006345272064,
|
|
"learning_rate": 0.00045125958740838454,
|
|
"loss": 3.2573211669921873,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.35689840617972035,
|
|
"grad_norm": 0.24503974616527557,
|
|
"learning_rate": 0.0004471392503160368,
|
|
"loss": 3.2612152099609375,
|
|
"step": 1825
|
|
},
|
|
{
|
|
"epoch": 0.36178742544245623,
|
|
"grad_norm": 0.2681194245815277,
|
|
"learning_rate": 0.00044298203578269237,
|
|
"loss": 3.2565509033203126,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.3666764447051921,
|
|
"grad_norm": 0.25433462858200073,
|
|
"learning_rate": 0.00043878898572907114,
|
|
"loss": 3.227747802734375,
|
|
"step": 1875
|
|
},
|
|
{
|
|
"epoch": 0.37156546396792806,
|
|
"grad_norm": 0.25087985396385193,
|
|
"learning_rate": 0.00043456115105733323,
|
|
"loss": 3.2326690673828127,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.37645448323066394,
|
|
"grad_norm": 0.2513449788093567,
|
|
"learning_rate": 0.0004302995913876907,
|
|
"loss": 3.2403411865234375,
|
|
"step": 1925
|
|
},
|
|
{
|
|
"epoch": 0.3813435024933998,
|
|
"grad_norm": 0.23991937935352325,
|
|
"learning_rate": 0.0004260053747928362,
|
|
"loss": 3.2273150634765626,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.3862325217561357,
|
|
"grad_norm": 0.25117090344429016,
|
|
"learning_rate": 0.0004216795775302522,
|
|
"loss": 3.2322503662109376,
|
|
"step": 1975
|
|
},
|
|
{
|
|
"epoch": 0.3911215410188716,
|
|
"grad_norm": 0.24637891352176666,
|
|
"learning_rate": 0.00041732328377246884,
|
|
"loss": 3.2080303955078127,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.3960105602816075,
|
|
"grad_norm": 0.2761523425579071,
|
|
"learning_rate": 0.0004129375853353386,
|
|
"loss": 3.2052261352539064,
|
|
"step": 2025
|
|
},
|
|
{
|
|
"epoch": 0.4008995795443434,
|
|
"grad_norm": 0.26527732610702515,
|
|
"learning_rate": 0.000408523581404395,
|
|
"loss": 3.219066162109375,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.4057885988070793,
|
|
"grad_norm": 0.2626402676105499,
|
|
"learning_rate": 0.00040408237825936393,
|
|
"loss": 3.1982321166992187,
|
|
"step": 2075
|
|
},
|
|
{
|
|
"epoch": 0.4106776180698152,
|
|
"grad_norm": 0.2456260770559311,
|
|
"learning_rate": 0.0003996150889968977,
|
|
"loss": 3.19399658203125,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.4155666373325511,
|
|
"grad_norm": 0.24225425720214844,
|
|
"learning_rate": 0.0003951228332515998,
|
|
"loss": 3.21568603515625,
|
|
"step": 2125
|
|
},
|
|
{
|
|
"epoch": 0.420455656595287,
|
|
"grad_norm": 0.2526324689388275,
|
|
"learning_rate": 0.00039060673691541166,
|
|
"loss": 3.20046142578125,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.4253446758580229,
|
|
"grad_norm": 0.25951486825942993,
|
|
"learning_rate": 0.0003860679318554305,
|
|
"loss": 3.18888916015625,
|
|
"step": 2175
|
|
},
|
|
{
|
|
"epoch": 0.43023369512075876,
|
|
"grad_norm": 0.2438725382089615,
|
|
"learning_rate": 0.00038150755563023085,
|
|
"loss": 3.1862054443359376,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.43512271438349465,
|
|
"grad_norm": 0.25872883200645447,
|
|
"learning_rate": 0.0003769267512047575,
|
|
"loss": 3.16793701171875,
|
|
"step": 2225
|
|
},
|
|
{
|
|
"epoch": 0.4400117336462306,
|
|
"grad_norm": 0.25001493096351624,
|
|
"learning_rate": 0.00037232666666386623,
|
|
"loss": 3.1839862060546875,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.44490075290896647,
|
|
"grad_norm": 0.25206780433654785,
|
|
"learning_rate": 0.0003677084549245782,
|
|
"loss": 3.1718374633789064,
|
|
"step": 2275
|
|
},
|
|
{
|
|
"epoch": 0.44978977217170235,
|
|
"grad_norm": 0.2607225775718689,
|
|
"learning_rate": 0.0003630732734471266,
|
|
"loss": 3.1549212646484377,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 0.45467879143443823,
|
|
"grad_norm": 0.25680533051490784,
|
|
"learning_rate": 0.0003584222839448615,
|
|
"loss": 3.1611688232421873,
|
|
"step": 2325
|
|
},
|
|
{
|
|
"epoch": 0.4595678106971742,
|
|
"grad_norm": 0.23713268339633942,
|
|
"learning_rate": 0.00035375665209309105,
|
|
"loss": 3.1543890380859376,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 0.46445682995991006,
|
|
"grad_norm": 0.23847518861293793,
|
|
"learning_rate": 0.00034907754723692856,
|
|
"loss": 3.1569989013671873,
|
|
"step": 2375
|
|
},
|
|
{
|
|
"epoch": 0.46934584922264594,
|
|
"grad_norm": 0.23781505227088928,
|
|
"learning_rate": 0.00034438614209822004,
|
|
"loss": 3.1373858642578125,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 0.4742348684853818,
|
|
"grad_norm": 0.2714940905570984,
|
|
"learning_rate": 0.00033968361248162555,
|
|
"loss": 3.1243209838867188,
|
|
"step": 2425
|
|
},
|
|
{
|
|
"epoch": 0.4791238877481177,
|
|
"grad_norm": 0.2610074281692505,
|
|
"learning_rate": 0.0003349711369799277,
|
|
"loss": 3.136983642578125,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 0.48401290701085364,
|
|
"grad_norm": 0.2547595202922821,
|
|
"learning_rate": 0.0003302498966786418,
|
|
"loss": 3.11527099609375,
|
|
"step": 2475
|
|
},
|
|
{
|
|
"epoch": 0.4889019262735895,
|
|
"grad_norm": 0.25215449929237366,
|
|
"learning_rate": 0.00032552107486000036,
|
|
"loss": 3.13005126953125,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.4937909455363254,
|
|
"grad_norm": 0.26304101943969727,
|
|
"learning_rate": 0.00032078585670638853,
|
|
"loss": 3.117969970703125,
|
|
"step": 2525
|
|
},
|
|
{
|
|
"epoch": 0.4986799647990613,
|
|
"grad_norm": 0.24783320724964142,
|
|
"learning_rate": 0.0003160454290033015,
|
|
"loss": 3.119049072265625,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 0.5035689840617972,
|
|
"grad_norm": 0.24181707203388214,
|
|
"learning_rate": 0.00031130097984190166,
|
|
"loss": 3.1037738037109377,
|
|
"step": 2575
|
|
},
|
|
{
|
|
"epoch": 0.5084580033245331,
|
|
"grad_norm": 0.2581552565097809,
|
|
"learning_rate": 0.0003065536983212473,
|
|
"loss": 3.1013323974609377,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 0.5133470225872689,
|
|
"grad_norm": 0.2646828293800354,
|
|
"learning_rate": 0.0003018047742502698,
|
|
"loss": 3.119427490234375,
|
|
"step": 2625
|
|
},
|
|
{
|
|
"epoch": 0.5182360418500049,
|
|
"grad_norm": 0.24800482392311096,
|
|
"learning_rate": 0.0002970553978495721,
|
|
"loss": 3.1079791259765623,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 0.5231250611127408,
|
|
"grad_norm": 0.25518327951431274,
|
|
"learning_rate": 0.00029230675945312424,
|
|
"loss": 3.103538818359375,
|
|
"step": 2675
|
|
},
|
|
{
|
|
"epoch": 0.5280140803754767,
|
|
"grad_norm": 0.2598397731781006,
|
|
"learning_rate": 0.0002875600492099309,
|
|
"loss": 3.0938665771484377,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 0.5329030996382126,
|
|
"grad_norm": 0.24244071543216705,
|
|
"learning_rate": 0.00028281645678574413,
|
|
"loss": 3.0996221923828124,
|
|
"step": 2725
|
|
},
|
|
{
|
|
"epoch": 0.5377921189009485,
|
|
"grad_norm": 0.2510348856449127,
|
|
"learning_rate": 0.00027807717106489885,
|
|
"loss": 3.0980859375,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 0.5426811381636844,
|
|
"grad_norm": 0.2464042454957962,
|
|
"learning_rate": 0.000273343379852343,
|
|
"loss": 3.087640380859375,
|
|
"step": 2775
|
|
},
|
|
{
|
|
"epoch": 0.5475701574264202,
|
|
"grad_norm": 0.2509900629520416,
|
|
"learning_rate": 0.00026861626957593797,
|
|
"loss": 3.086356201171875,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 0.5524591766891561,
|
|
"grad_norm": 0.25076374411582947,
|
|
"learning_rate": 0.00026389702498910586,
|
|
"loss": 3.075096435546875,
|
|
"step": 2825
|
|
},
|
|
{
|
|
"epoch": 0.557348195951892,
|
|
"grad_norm": 0.24435539543628693,
|
|
"learning_rate": 0.0002591868288738945,
|
|
"loss": 3.0860525512695314,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 0.562237215214628,
|
|
"grad_norm": 0.24108317494392395,
|
|
"learning_rate": 0.0002544868617445375,
|
|
"loss": 3.087845458984375,
|
|
"step": 2875
|
|
},
|
|
{
|
|
"epoch": 0.5671262344773639,
|
|
"grad_norm": 0.24404887855052948,
|
|
"learning_rate": 0.00024979830155158245,
|
|
"loss": 3.069549560546875,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 0.5720152537400998,
|
|
"grad_norm": 0.2537246644496918,
|
|
"learning_rate": 0.0002451223233866621,
|
|
"loss": 3.0592523193359376,
|
|
"step": 2925
|
|
},
|
|
{
|
|
"epoch": 0.5769042730028356,
|
|
"grad_norm": 0.25408676266670227,
|
|
"learning_rate": 0.00024046009918798114,
|
|
"loss": 3.054830322265625,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 0.5817932922655715,
|
|
"grad_norm": 0.24859051406383514,
|
|
"learning_rate": 0.00023581279744659458,
|
|
"loss": 3.0742523193359377,
|
|
"step": 2975
|
|
},
|
|
{
|
|
"epoch": 0.5866823115283074,
|
|
"grad_norm": 0.2513684630393982,
|
|
"learning_rate": 0.00023118158291354866,
|
|
"loss": 3.05662353515625,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.5915713307910433,
|
|
"grad_norm": 0.25194358825683594,
|
|
"learning_rate": 0.00022656761630796015,
|
|
"loss": 3.054917907714844,
|
|
"step": 3025
|
|
},
|
|
{
|
|
"epoch": 0.5964603500537792,
|
|
"grad_norm": 0.25090256333351135,
|
|
"learning_rate": 0.00022197205402610612,
|
|
"loss": 3.0644140625,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 0.6013493693165151,
|
|
"grad_norm": 0.2501232922077179,
|
|
"learning_rate": 0.00021739604785159654,
|
|
"loss": 3.0607208251953124,
|
|
"step": 3075
|
|
},
|
|
{
|
|
"epoch": 0.606238388579251,
|
|
"grad_norm": 0.25573962926864624,
|
|
"learning_rate": 0.00021284074466670336,
|
|
"loss": 3.0519308471679687,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 0.6111274078419869,
|
|
"grad_norm": 0.25536900758743286,
|
|
"learning_rate": 0.00020830728616491804,
|
|
"loss": 3.0519940185546877,
|
|
"step": 3125
|
|
},
|
|
{
|
|
"epoch": 0.6160164271047228,
|
|
"grad_norm": 0.24477079510688782,
|
|
"learning_rate": 0.00020379680856480944,
|
|
"loss": 3.0534375,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 0.6209054463674587,
|
|
"grad_norm": 0.25385400652885437,
|
|
"learning_rate": 0.00019931044232525378,
|
|
"loss": 3.0516427612304686,
|
|
"step": 3175
|
|
},
|
|
{
|
|
"epoch": 0.6257944656301946,
|
|
"grad_norm": 0.2569658160209656,
|
|
"learning_rate": 0.00019484931186210838,
|
|
"loss": 3.0533740234375,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 0.6306834848929305,
|
|
"grad_norm": 0.2463824599981308,
|
|
"learning_rate": 0.00019041453526639975,
|
|
"loss": 3.0460821533203126,
|
|
"step": 3225
|
|
},
|
|
{
|
|
"epoch": 0.6355725041556664,
|
|
"grad_norm": 0.25980037450790405,
|
|
"learning_rate": 0.0001860072240240973,
|
|
"loss": 3.0431292724609373,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 0.6404615234184022,
|
|
"grad_norm": 0.2577205002307892,
|
|
"learning_rate": 0.00018162848273754132,
|
|
"loss": 3.051102294921875,
|
|
"step": 3275
|
|
},
|
|
{
|
|
"epoch": 0.6453505426811381,
|
|
"grad_norm": 0.23692680895328522,
|
|
"learning_rate": 0.000177279408848598,
|
|
"loss": 3.01787841796875,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 0.6502395619438741,
|
|
"grad_norm": 0.2343571037054062,
|
|
"learning_rate": 0.00017296109236360753,
|
|
"loss": 3.0158380126953124,
|
|
"step": 3325
|
|
},
|
|
{
|
|
"epoch": 0.65512858120661,
|
|
"grad_norm": 0.23659783601760864,
|
|
"learning_rate": 0.00016867461558019647,
|
|
"loss": 3.0293670654296876,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 0.6600176004693459,
|
|
"grad_norm": 0.24451994895935059,
|
|
"learning_rate": 0.0001644210528160217,
|
|
"loss": 3.0010888671875,
|
|
"step": 3375
|
|
},
|
|
{
|
|
"epoch": 0.6649066197320818,
|
|
"grad_norm": 0.25063881278038025,
|
|
"learning_rate": 0.00016020147013951522,
|
|
"loss": 3.0027975463867187,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 0.6697956389948176,
|
|
"grad_norm": 0.2542019486427307,
|
|
"learning_rate": 0.00015601692510269568,
|
|
"loss": 3.0187261962890624,
|
|
"step": 3425
|
|
},
|
|
{
|
|
"epoch": 0.6746846582575535,
|
|
"grad_norm": 0.2519824504852295,
|
|
"learning_rate": 0.00015186846647611464,
|
|
"loss": 3.0060882568359375,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 0.6795736775202894,
|
|
"grad_norm": 0.24705493450164795,
|
|
"learning_rate": 0.0001477571339860033,
|
|
"loss": 3.0131549072265624,
|
|
"step": 3475
|
|
},
|
|
{
|
|
"epoch": 0.6844626967830253,
|
|
"grad_norm": 0.24704059958457947,
|
|
"learning_rate": 0.00014368395805368716,
|
|
"loss": 3.0088775634765623,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 0.6893517160457612,
|
|
"grad_norm": 0.2505241334438324,
|
|
"learning_rate": 0.00013964995953733144,
|
|
"loss": 3.012384033203125,
|
|
"step": 3525
|
|
},
|
|
{
|
|
"epoch": 0.6942407353084972,
|
|
"grad_norm": 0.26129379868507385,
|
|
"learning_rate": 0.00013565614947608418,
|
|
"loss": 2.9906854248046875,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 0.699129754571233,
|
|
"grad_norm": 0.2629392445087433,
|
|
"learning_rate": 0.00013170352883667958,
|
|
"loss": 2.987822265625,
|
|
"step": 3575
|
|
},
|
|
{
|
|
"epoch": 0.7040187738339689,
|
|
"grad_norm": 0.24119465053081512,
|
|
"learning_rate": 0.00012779308826256668,
|
|
"loss": 2.9882269287109375,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 0.7089077930967048,
|
|
"grad_norm": 0.24107669293880463,
|
|
"learning_rate": 0.00012392580782562455,
|
|
"loss": 2.9944049072265626,
|
|
"step": 3625
|
|
},
|
|
{
|
|
"epoch": 0.7137968123594407,
|
|
"grad_norm": 0.22759009897708893,
|
|
"learning_rate": 0.0001201026567805276,
|
|
"loss": 2.982850341796875,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 0.7186858316221766,
|
|
"grad_norm": 0.24127598106861115,
|
|
"learning_rate": 0.00011632459332182107,
|
|
"loss": 3.005535888671875,
|
|
"step": 3675
|
|
},
|
|
{
|
|
"epoch": 0.7235748508849125,
|
|
"grad_norm": 0.2499409168958664,
|
|
"learning_rate": 0.0001125925643437701,
|
|
"loss": 2.9897674560546874,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 0.7284638701476484,
|
|
"grad_norm": 0.24349592626094818,
|
|
"learning_rate": 0.00010890750520303963,
|
|
"loss": 2.982366943359375,
|
|
"step": 3725
|
|
},
|
|
{
|
|
"epoch": 0.7333528894103842,
|
|
"grad_norm": 0.23667509853839874,
|
|
"learning_rate": 0.00010527033948426599,
|
|
"loss": 2.9962640380859376,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 0.7382419086731201,
|
|
"grad_norm": 0.2368006855249405,
|
|
"learning_rate": 0.0001016819787685801,
|
|
"loss": 2.9754425048828126,
|
|
"step": 3775
|
|
},
|
|
{
|
|
"epoch": 0.7431309279358561,
|
|
"grad_norm": 0.2447584569454193,
|
|
"learning_rate": 9.814332240513792e-05,
|
|
"loss": 2.9829901123046874,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 0.748019947198592,
|
|
"grad_norm": 0.2414255291223526,
|
|
"learning_rate": 9.465525728571692e-05,
|
|
"loss": 2.9877850341796877,
|
|
"step": 3825
|
|
},
|
|
{
|
|
"epoch": 0.7529089664613279,
|
|
"grad_norm": 0.2383822649717331,
|
|
"learning_rate": 9.12186576224348e-05,
|
|
"loss": 2.9598373413085937,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 0.7577979857240638,
|
|
"grad_norm": 0.24468879401683807,
|
|
"learning_rate": 8.783438472864648e-05,
|
|
"loss": 2.9916671752929687,
|
|
"step": 3875
|
|
},
|
|
{
|
|
"epoch": 0.7626870049867996,
|
|
"grad_norm": 0.24762850999832153,
|
|
"learning_rate": 8.450328680307309e-05,
|
|
"loss": 2.992269287109375,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 0.7675760242495355,
|
|
"grad_norm": 0.23502996563911438,
|
|
"learning_rate": 8.122619871721876e-05,
|
|
"loss": 2.9907321166992187,
|
|
"step": 3925
|
|
},
|
|
{
|
|
"epoch": 0.7724650435122714,
|
|
"grad_norm": 0.24725691974163055,
|
|
"learning_rate": 7.800394180612657e-05,
|
|
"loss": 2.9762115478515625,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 0.7773540627750073,
|
|
"grad_norm": 0.24725252389907837,
|
|
"learning_rate": 7.483732366252843e-05,
|
|
"loss": 2.976396484375,
|
|
"step": 3975
|
|
},
|
|
{
|
|
"epoch": 0.7822430820377432,
|
|
"grad_norm": 0.24095748364925385,
|
|
"learning_rate": 7.172713793443797e-05,
|
|
"loss": 2.9821087646484377,
|
|
"step": 4000
|
|
}
|
|
],
|
|
"logging_steps": 25,
|
|
"max_steps": 5114,
|
|
"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": 5.344834255794995e+17,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|