rope_instruct / trainer_state.json
IvenWu123's picture
Upload folder using huggingface_hub
b31818d verified
Raw
History Blame Contribute Delete
34.6 kB
{
"best_global_step": 7607,
"best_metric": 3.00644255,
"best_model_checkpoint": "/scratch/prj0000000267/yuefan/UnifyTrajLLM/output_rope_instruct_llmlow5_5e-4/v1-20251106-152726/checkpoint-7607",
"epoch": 1.0,
"eval_steps": 2000,
"global_step": 7607,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00013145786775338504,
"grad_norm": 304.0,
"learning_rate": 2.624671916010499e-07,
"loss": 30.29734230041504,
"step": 1,
"token_acc": 0.0
},
{
"epoch": 0.006572893387669252,
"grad_norm": 73.0,
"learning_rate": 1.3123359580052494e-05,
"loss": 21.377688038105866,
"step": 50,
"token_acc": 0.008780085457147575
},
{
"epoch": 0.013145786775338505,
"grad_norm": 15.3125,
"learning_rate": 2.6246719160104988e-05,
"loss": 7.8677734375,
"step": 100,
"token_acc": 0.4647903181356832
},
{
"epoch": 0.019718680163007755,
"grad_norm": 17.125,
"learning_rate": 3.9370078740157485e-05,
"loss": 6.155509033203125,
"step": 150,
"token_acc": 0.5138413431527499
},
{
"epoch": 0.02629157355067701,
"grad_norm": 35.75,
"learning_rate": 5.2493438320209976e-05,
"loss": 5.588656005859375,
"step": 200,
"token_acc": 0.5934439028612903
},
{
"epoch": 0.03286446693834626,
"grad_norm": 15.125,
"learning_rate": 6.561679790026247e-05,
"loss": 5.309133911132813,
"step": 250,
"token_acc": 0.6521334320201939
},
{
"epoch": 0.03943736032601551,
"grad_norm": 18.5,
"learning_rate": 7.874015748031497e-05,
"loss": 5.129273681640625,
"step": 300,
"token_acc": 0.6742803871946512
},
{
"epoch": 0.046010253713684765,
"grad_norm": 4.4375,
"learning_rate": 9.186351706036746e-05,
"loss": 5.0163330078125,
"step": 350,
"token_acc": 0.6809492214844467
},
{
"epoch": 0.05258314710135402,
"grad_norm": 8.875,
"learning_rate": 9.99982941196289e-05,
"loss": 4.9205517578125,
"step": 400,
"token_acc": 0.683060363678607
},
{
"epoch": 0.059156040489023266,
"grad_norm": 5.03125,
"learning_rate": 9.997750378511423e-05,
"loss": 4.752092895507812,
"step": 450,
"token_acc": 0.6940432274724169
},
{
"epoch": 0.06572893387669251,
"grad_norm": 5.3125,
"learning_rate": 9.993309772476504e-05,
"loss": 4.64461181640625,
"step": 500,
"token_acc": 0.6959406283972602
},
{
"epoch": 0.07230182726436177,
"grad_norm": 5.65625,
"learning_rate": 9.986509692164903e-05,
"loss": 4.5522998046875,
"step": 550,
"token_acc": 0.6924621588959144
},
{
"epoch": 0.07887472065203102,
"grad_norm": 5.4375,
"learning_rate": 9.977353350798975e-05,
"loss": 4.359736328125,
"step": 600,
"token_acc": 0.7033329074945415
},
{
"epoch": 0.08544761403970028,
"grad_norm": 11.4375,
"learning_rate": 9.965845074998316e-05,
"loss": 4.249523010253906,
"step": 650,
"token_acc": 0.7017039526383457
},
{
"epoch": 0.09202050742736953,
"grad_norm": 2.859375,
"learning_rate": 9.951990302735326e-05,
"loss": 4.373087158203125,
"step": 700,
"token_acc": 0.6699921322318458
},
{
"epoch": 0.09859340081503878,
"grad_norm": 4.34375,
"learning_rate": 9.935795580765613e-05,
"loss": 4.105689697265625,
"step": 750,
"token_acc": 0.6975005467269281
},
{
"epoch": 0.10516629420270804,
"grad_norm": 3.46875,
"learning_rate": 9.91726856153448e-05,
"loss": 4.0196990966796875,
"step": 800,
"token_acc": 0.6986118861143772
},
{
"epoch": 0.11173918759037728,
"grad_norm": 5.40625,
"learning_rate": 9.896417999560932e-05,
"loss": 3.95506591796875,
"step": 850,
"token_acc": 0.703784619858412
},
{
"epoch": 0.11831208097804653,
"grad_norm": 2.21875,
"learning_rate": 9.873253747300928e-05,
"loss": 3.8350949096679687,
"step": 900,
"token_acc": 0.7058728101329687
},
{
"epoch": 0.12488497436571579,
"grad_norm": 6.46875,
"learning_rate": 9.847786750491832e-05,
"loss": 3.813152160644531,
"step": 950,
"token_acc": 0.7039227968986385
},
{
"epoch": 0.13145786775338503,
"grad_norm": 2.046875,
"learning_rate": 9.820029042980254e-05,
"loss": 3.7669647216796873,
"step": 1000,
"token_acc": 0.705981499295243
},
{
"epoch": 0.13803076114105428,
"grad_norm": 3.484375,
"learning_rate": 9.78999374103573e-05,
"loss": 3.685584716796875,
"step": 1050,
"token_acc": 0.7128186768590038
},
{
"epoch": 0.14460365452872354,
"grad_norm": 2.75,
"learning_rate": 9.757695037152929e-05,
"loss": 3.613662109375,
"step": 1100,
"token_acc": 0.7169065987963625
},
{
"epoch": 0.1511765479163928,
"grad_norm": 4.46875,
"learning_rate": 9.723148193345319e-05,
"loss": 3.614210510253906,
"step": 1150,
"token_acc": 0.7133252376738136
},
{
"epoch": 0.15774944130406204,
"grad_norm": 2.953125,
"learning_rate": 9.686369533933442e-05,
"loss": 3.5963873291015624,
"step": 1200,
"token_acc": 0.7142219755262533
},
{
"epoch": 0.1643223346917313,
"grad_norm": 1.5390625,
"learning_rate": 9.647376437831236e-05,
"loss": 3.588519287109375,
"step": 1250,
"token_acc": 0.7139076732444867
},
{
"epoch": 0.17089522807940055,
"grad_norm": 2.625,
"learning_rate": 9.606187330334036e-05,
"loss": 3.5846401977539064,
"step": 1300,
"token_acc": 0.7123157676600218
},
{
"epoch": 0.1774681214670698,
"grad_norm": 2.71875,
"learning_rate": 9.562821674412097e-05,
"loss": 3.481658935546875,
"step": 1350,
"token_acc": 0.7204118049994375
},
{
"epoch": 0.18404101485473906,
"grad_norm": 2.75,
"learning_rate": 9.517299961513844e-05,
"loss": 3.453585205078125,
"step": 1400,
"token_acc": 0.7226983300292329
},
{
"epoch": 0.19061390824240831,
"grad_norm": 2.453125,
"learning_rate": 9.46964370188308e-05,
"loss": 3.4323849487304687,
"step": 1450,
"token_acc": 0.724599297306341
},
{
"epoch": 0.19718680163007757,
"grad_norm": 2.109375,
"learning_rate": 9.419875414394832e-05,
"loss": 3.461167297363281,
"step": 1500,
"token_acc": 0.7180978070116231
},
{
"epoch": 0.20375969501774682,
"grad_norm": 3.140625,
"learning_rate": 9.36801861591456e-05,
"loss": 3.440343322753906,
"step": 1550,
"token_acc": 0.7220378064338172
},
{
"epoch": 0.21033258840541608,
"grad_norm": 2.640625,
"learning_rate": 9.314097810185806e-05,
"loss": 3.382811279296875,
"step": 1600,
"token_acc": 0.7226561188235847
},
{
"epoch": 0.2169054817930853,
"grad_norm": 2.140625,
"learning_rate": 9.258138476251506e-05,
"loss": 3.398974609375,
"step": 1650,
"token_acc": 0.7228855184009217
},
{
"epoch": 0.22347837518075456,
"grad_norm": 3.375,
"learning_rate": 9.200167056414455e-05,
"loss": 3.39942626953125,
"step": 1700,
"token_acc": 0.7236204717775906
},
{
"epoch": 0.2300512685684238,
"grad_norm": 2.328125,
"learning_rate": 9.140210943742604e-05,
"loss": 3.3529150390625,
"step": 1750,
"token_acc": 0.72407514517165
},
{
"epoch": 0.23662416195609307,
"grad_norm": 3.859375,
"learning_rate": 9.078298469125089e-05,
"loss": 3.3407843017578127,
"step": 1800,
"token_acc": 0.7253981347681612
},
{
"epoch": 0.24319705534376232,
"grad_norm": 2.9375,
"learning_rate": 9.014458887885113e-05,
"loss": 3.3440185546875,
"step": 1850,
"token_acc": 0.7241830660632759
},
{
"epoch": 0.24976994873143157,
"grad_norm": 2.40625,
"learning_rate": 8.948722365956027e-05,
"loss": 3.3022760009765624,
"step": 1900,
"token_acc": 0.7303254561921374
},
{
"epoch": 0.25634284211910086,
"grad_norm": 1.8359375,
"learning_rate": 8.881119965627087e-05,
"loss": 3.3453948974609373,
"step": 1950,
"token_acc": 0.7209394423993724
},
{
"epoch": 0.26291573550677005,
"grad_norm": 1.7265625,
"learning_rate": 8.811683630865702e-05,
"loss": 3.3039337158203126,
"step": 2000,
"token_acc": 0.7238227911646586
},
{
"epoch": 0.26291573550677005,
"eval_loss": 3.321227788925171,
"eval_runtime": 233.9893,
"eval_samples_per_second": 188.355,
"eval_steps_per_second": 2.945,
"eval_token_acc": 0.727985502110483,
"step": 2000
},
{
"epoch": 0.2694886288944393,
"grad_norm": 2.3125,
"learning_rate": 8.74044617222303e-05,
"loss": 3.322220764160156,
"step": 2050,
"token_acc": 0.7210059136942014
},
{
"epoch": 0.27606152228210856,
"grad_norm": 1.5078125,
"learning_rate": 8.667441251330106e-05,
"loss": 3.2644619750976562,
"step": 2100,
"token_acc": 0.7271068777584593
},
{
"epoch": 0.2826344156697778,
"grad_norm": 2.59375,
"learning_rate": 8.592703364991808e-05,
"loss": 3.2546685791015624,
"step": 2150,
"token_acc": 0.72905454647382
},
{
"epoch": 0.28920730905744707,
"grad_norm": 1.03125,
"learning_rate": 8.51626782888618e-05,
"loss": 3.273728332519531,
"step": 2200,
"token_acc": 0.7228298860800985
},
{
"epoch": 0.2957802024451163,
"grad_norm": 1.28125,
"learning_rate": 8.438170760876823e-05,
"loss": 3.2248263549804688,
"step": 2250,
"token_acc": 0.7292383849594279
},
{
"epoch": 0.3023530958327856,
"grad_norm": 1.921875,
"learning_rate": 8.358449063946216e-05,
"loss": 3.2212490844726562,
"step": 2300,
"token_acc": 0.7281072030159621
},
{
"epoch": 0.30892598922045483,
"grad_norm": 2.4375,
"learning_rate": 8.27714040875806e-05,
"loss": 3.24546875,
"step": 2350,
"token_acc": 0.7240473211334678
},
{
"epoch": 0.3154988826081241,
"grad_norm": 1.09375,
"learning_rate": 8.194283215856873e-05,
"loss": 3.24426513671875,
"step": 2400,
"token_acc": 0.7221157836260441
},
{
"epoch": 0.32207177599579334,
"grad_norm": 2.03125,
"learning_rate": 8.109916637513229e-05,
"loss": 3.2446798706054687,
"step": 2450,
"token_acc": 0.7259417210451409
},
{
"epoch": 0.3286446693834626,
"grad_norm": 2.265625,
"learning_rate": 8.024080539223259e-05,
"loss": 3.21652587890625,
"step": 2500,
"token_acc": 0.7262627446088282
},
{
"epoch": 0.33521756277113185,
"grad_norm": 1.4453125,
"learning_rate": 7.936815480871121e-05,
"loss": 3.2104052734375,
"step": 2550,
"token_acc": 0.7240804261180275
},
{
"epoch": 0.3417904561588011,
"grad_norm": 1.8515625,
"learning_rate": 7.84816269756335e-05,
"loss": 3.20545166015625,
"step": 2600,
"token_acc": 0.7257663748400122
},
{
"epoch": 0.34836334954647036,
"grad_norm": 1.8203125,
"learning_rate": 7.758164080144156e-05,
"loss": 3.1838531494140625,
"step": 2650,
"token_acc": 0.7265359877539365
},
{
"epoch": 0.3549362429341396,
"grad_norm": 2.296875,
"learning_rate": 7.66686215540086e-05,
"loss": 3.1726068115234374,
"step": 2700,
"token_acc": 0.7272715236172255
},
{
"epoch": 0.36150913632180887,
"grad_norm": 3.703125,
"learning_rate": 7.574300065968846e-05,
"loss": 3.20809326171875,
"step": 2750,
"token_acc": 0.7240619199151419
},
{
"epoch": 0.3680820297094781,
"grad_norm": 1.2109375,
"learning_rate": 7.480521549945475e-05,
"loss": 3.137190856933594,
"step": 2800,
"token_acc": 0.7298223239809069
},
{
"epoch": 0.3746549230971474,
"grad_norm": 1.296875,
"learning_rate": 7.385570920222668e-05,
"loss": 3.177913818359375,
"step": 2850,
"token_acc": 0.7272812579594353
},
{
"epoch": 0.38122781648481663,
"grad_norm": 3.671875,
"learning_rate": 7.289493043547852e-05,
"loss": 3.1652337646484376,
"step": 2900,
"token_acc": 0.7268903808412633
},
{
"epoch": 0.3878007098724859,
"grad_norm": 0.92578125,
"learning_rate": 7.192333319323207e-05,
"loss": 3.137171325683594,
"step": 2950,
"token_acc": 0.7261608651121959
},
{
"epoch": 0.39437360326015514,
"grad_norm": 1.6796875,
"learning_rate": 7.094137658153218e-05,
"loss": 3.15204345703125,
"step": 3000,
"token_acc": 0.7257778858755249
},
{
"epoch": 0.4009464966478244,
"grad_norm": 2.609375,
"learning_rate": 6.99495246015067e-05,
"loss": 3.1364361572265627,
"step": 3050,
"token_acc": 0.7274269135603249
},
{
"epoch": 0.40751939003549364,
"grad_norm": 2.5625,
"learning_rate": 6.894824593011337e-05,
"loss": 3.137897644042969,
"step": 3100,
"token_acc": 0.7253378311126446
},
{
"epoch": 0.4140922834231629,
"grad_norm": 1.1796875,
"learning_rate": 6.793801369867719e-05,
"loss": 3.13400390625,
"step": 3150,
"token_acc": 0.7259136973792957
},
{
"epoch": 0.42066517681083215,
"grad_norm": 1.8359375,
"learning_rate": 6.691930526932315e-05,
"loss": 3.1219122314453127,
"step": 3200,
"token_acc": 0.7273963236229487
},
{
"epoch": 0.4272380701985014,
"grad_norm": 1.4296875,
"learning_rate": 6.589260200940954e-05,
"loss": 3.1406710815429686,
"step": 3250,
"token_acc": 0.723792698293771
},
{
"epoch": 0.4338109635861706,
"grad_norm": 0.85546875,
"learning_rate": 6.485838906406897e-05,
"loss": 3.1084765625,
"step": 3300,
"token_acc": 0.7265385793018984
},
{
"epoch": 0.44038385697383986,
"grad_norm": 1.9921875,
"learning_rate": 6.381715512696402e-05,
"loss": 3.124567565917969,
"step": 3350,
"token_acc": 0.7251010373164489
},
{
"epoch": 0.4469567503615091,
"grad_norm": 0.92578125,
"learning_rate": 6.276939220936628e-05,
"loss": 3.1119622802734375,
"step": 3400,
"token_acc": 0.725668361335667
},
{
"epoch": 0.45352964374917837,
"grad_norm": 0.78125,
"learning_rate": 6.171559540766767e-05,
"loss": 3.096336669921875,
"step": 3450,
"token_acc": 0.7288083839011799
},
{
"epoch": 0.4601025371368476,
"grad_norm": 1.1484375,
"learning_rate": 6.0656262669433984e-05,
"loss": 3.082548828125,
"step": 3500,
"token_acc": 0.7270383544852433
},
{
"epoch": 0.4666754305245169,
"grad_norm": 1.828125,
"learning_rate": 5.959189455811114e-05,
"loss": 3.5198748779296873,
"step": 3550,
"token_acc": 0.6858720558007885
},
{
"epoch": 0.47324832391218613,
"grad_norm": 0.875,
"learning_rate": 5.852299401649544e-05,
"loss": 3.080078430175781,
"step": 3600,
"token_acc": 0.7262006761379108
},
{
"epoch": 0.4798212172998554,
"grad_norm": 1.0546875,
"learning_rate": 5.7450066129079325e-05,
"loss": 3.0630398559570313,
"step": 3650,
"token_acc": 0.7282769091419556
},
{
"epoch": 0.48639411068752464,
"grad_norm": 2.265625,
"learning_rate": 5.637361788338543e-05,
"loss": 3.056205139160156,
"step": 3700,
"token_acc": 0.7262423587864033
},
{
"epoch": 0.4929670040751939,
"grad_norm": 1.84375,
"learning_rate": 5.529415793040109e-05,
"loss": 3.0689117431640627,
"step": 3750,
"token_acc": 0.7272350853615711
},
{
"epoch": 0.49953989746286315,
"grad_norm": 0.828125,
"learning_rate": 5.4212196344227005e-05,
"loss": 3.089296569824219,
"step": 3800,
"token_acc": 0.7252981429762677
},
{
"epoch": 0.5061127908505324,
"grad_norm": 2.203125,
"learning_rate": 5.31282443810534e-05,
"loss": 3.053915710449219,
"step": 3850,
"token_acc": 0.7280129292643683
},
{
"epoch": 0.5126856842382017,
"grad_norm": 0.91015625,
"learning_rate": 5.2042814237577665e-05,
"loss": 3.083997802734375,
"step": 3900,
"token_acc": 0.7262975281961235
},
{
"epoch": 0.5192585776258709,
"grad_norm": 2.734375,
"learning_rate": 5.095641880897737e-05,
"loss": 3.039005126953125,
"step": 3950,
"token_acc": 0.7293689888181156
},
{
"epoch": 0.5258314710135401,
"grad_norm": 1.375,
"learning_rate": 4.9869571446553556e-05,
"loss": 3.049577331542969,
"step": 4000,
"token_acc": 0.7270735903036475
},
{
"epoch": 0.5258314710135401,
"eval_loss": 3.0929133892059326,
"eval_runtime": 233.729,
"eval_samples_per_second": 188.565,
"eval_steps_per_second": 2.948,
"eval_token_acc": 0.7304881660036263,
"step": 4000
},
{
"epoch": 0.5324043644012094,
"grad_norm": 0.7109375,
"learning_rate": 4.8782785715158214e-05,
"loss": 3.0205902099609374,
"step": 4050,
"token_acc": 0.7281370087445305
},
{
"epoch": 0.5389772577888786,
"grad_norm": 2.203125,
"learning_rate": 4.769657515052103e-05,
"loss": 3.0527685546875,
"step": 4100,
"token_acc": 0.7264239882694029
},
{
"epoch": 0.5455501511765479,
"grad_norm": 0.9140625,
"learning_rate": 4.66114530165898e-05,
"loss": 3.034259033203125,
"step": 4150,
"token_acc": 0.7264680410557581
},
{
"epoch": 0.5521230445642171,
"grad_norm": 2.0,
"learning_rate": 4.552793206299924e-05,
"loss": 3.04488525390625,
"step": 4200,
"token_acc": 0.7272732236581252
},
{
"epoch": 0.5586959379518864,
"grad_norm": 1.2890625,
"learning_rate": 4.444652428278301e-05,
"loss": 3.0357012939453125,
"step": 4250,
"token_acc": 0.7265666225811408
},
{
"epoch": 0.5652688313395556,
"grad_norm": 0.76171875,
"learning_rate": 4.3367740670443e-05,
"loss": 3.03507080078125,
"step": 4300,
"token_acc": 0.726759526626428
},
{
"epoch": 0.5718417247272249,
"grad_norm": 1.9296875,
"learning_rate": 4.229209098049062e-05,
"loss": 3.00672119140625,
"step": 4350,
"token_acc": 0.7276024203600877
},
{
"epoch": 0.5784146181148941,
"grad_norm": 0.78125,
"learning_rate": 4.122008348657398e-05,
"loss": 3.0070254516601564,
"step": 4400,
"token_acc": 0.7283590686769877
},
{
"epoch": 0.5849875115025635,
"grad_norm": 2.8125,
"learning_rate": 4.015222474130479e-05,
"loss": 3.0119451904296874,
"step": 4450,
"token_acc": 0.7240358952057842
},
{
"epoch": 0.5915604048902326,
"grad_norm": 2.140625,
"learning_rate": 3.9089019336898495e-05,
"loss": 2.9994842529296877,
"step": 4500,
"token_acc": 0.7290181024891761
},
{
"epoch": 0.598133298277902,
"grad_norm": 1.25,
"learning_rate": 3.8030969666740843e-05,
"loss": 2.986125793457031,
"step": 4550,
"token_acc": 0.7274355810753096
},
{
"epoch": 0.6047061916655712,
"grad_norm": 0.69140625,
"learning_rate": 3.697857568799328e-05,
"loss": 2.993693542480469,
"step": 4600,
"token_acc": 0.728050380036322
},
{
"epoch": 0.6112790850532405,
"grad_norm": 0.6796875,
"learning_rate": 3.5932334685349714e-05,
"loss": 2.9861276245117185,
"step": 4650,
"token_acc": 0.7287345375741995
},
{
"epoch": 0.6178519784409097,
"grad_norm": 1.5234375,
"learning_rate": 3.4892741036055995e-05,
"loss": 2.9960702514648436,
"step": 4700,
"token_acc": 0.7270325579049989
},
{
"epoch": 0.624424871828579,
"grad_norm": 0.80078125,
"learning_rate": 3.386028597630324e-05,
"loss": 2.9759326171875,
"step": 4750,
"token_acc": 0.730886205943149
},
{
"epoch": 0.6309977652162482,
"grad_norm": 2.234375,
"learning_rate": 3.2835457369105405e-05,
"loss": 2.9792401123046877,
"step": 4800,
"token_acc": 0.7275496532418397
},
{
"epoch": 0.6375706586039175,
"grad_norm": 0.98046875,
"learning_rate": 3.181873947377085e-05,
"loss": 2.989874267578125,
"step": 4850,
"token_acc": 0.7254225618317313
},
{
"epoch": 0.6441435519915867,
"grad_norm": 1.6875,
"learning_rate": 3.0810612717076605e-05,
"loss": 2.9674273681640626,
"step": 4900,
"token_acc": 0.7308766261910954
},
{
"epoch": 0.650716445379256,
"grad_norm": 1.578125,
"learning_rate": 2.9811553466253662e-05,
"loss": 2.989566650390625,
"step": 4950,
"token_acc": 0.7270966777763678
},
{
"epoch": 0.6572893387669252,
"grad_norm": 0.7734375,
"learning_rate": 2.8822033803890703e-05,
"loss": 2.990399169921875,
"step": 5000,
"token_acc": 0.7270503190482351
},
{
"epoch": 0.6638622321545945,
"grad_norm": 1.2265625,
"learning_rate": 2.784252130486208e-05,
"loss": 2.989596252441406,
"step": 5050,
"token_acc": 0.7223299967469977
},
{
"epoch": 0.6704351255422637,
"grad_norm": 0.76171875,
"learning_rate": 2.6873478815386132e-05,
"loss": 2.9679669189453124,
"step": 5100,
"token_acc": 0.7294171434167955
},
{
"epoch": 0.677008018929933,
"grad_norm": 0.78515625,
"learning_rate": 2.5915364234317747e-05,
"loss": 2.96373046875,
"step": 5150,
"token_acc": 0.7294911716853812
},
{
"epoch": 0.6835809123176022,
"grad_norm": 1.2890625,
"learning_rate": 2.4968630296778744e-05,
"loss": 2.9356533813476564,
"step": 5200,
"token_acc": 0.7306216439695229
},
{
"epoch": 0.6901538057052714,
"grad_norm": 0.8984375,
"learning_rate": 2.4033724360228265e-05,
"loss": 2.9301898193359377,
"step": 5250,
"token_acc": 0.7289044204974737
},
{
"epoch": 0.6967266990929407,
"grad_norm": 0.6328125,
"learning_rate": 2.3111088193074198e-05,
"loss": 2.92695556640625,
"step": 5300,
"token_acc": 0.7309000371044998
},
{
"epoch": 0.7032995924806099,
"grad_norm": 1.8203125,
"learning_rate": 2.220115776592573e-05,
"loss": 2.957354736328125,
"step": 5350,
"token_acc": 0.7266267729312463
},
{
"epoch": 0.7098724858682792,
"grad_norm": 0.80859375,
"learning_rate": 2.1304363045585285e-05,
"loss": 2.942122497558594,
"step": 5400,
"token_acc": 0.7282598715978182
},
{
"epoch": 0.7164453792559484,
"grad_norm": 1.7734375,
"learning_rate": 2.042112779187756e-05,
"loss": 2.9434585571289062,
"step": 5450,
"token_acc": 0.7287332467339165
},
{
"epoch": 0.7230182726436177,
"grad_norm": 0.69921875,
"learning_rate": 1.9551869357411675e-05,
"loss": 2.9658978271484373,
"step": 5500,
"token_acc": 0.7261902596027892
},
{
"epoch": 0.7295911660312869,
"grad_norm": 0.68359375,
"learning_rate": 1.8696998490370536e-05,
"loss": 2.919261474609375,
"step": 5550,
"token_acc": 0.7296815852994373
},
{
"epoch": 0.7361640594189562,
"grad_norm": 1.71875,
"learning_rate": 1.7856919140421336e-05,
"loss": 2.92064208984375,
"step": 5600,
"token_acc": 0.7321287132122515
},
{
"epoch": 0.7427369528066254,
"grad_norm": 0.6796875,
"learning_rate": 1.7032028267838285e-05,
"loss": 2.9069830322265626,
"step": 5650,
"token_acc": 0.7334535738253919
},
{
"epoch": 0.7493098461942947,
"grad_norm": 1.21875,
"learning_rate": 1.6222715655928173e-05,
"loss": 2.9308148193359376,
"step": 5700,
"token_acc": 0.7275641449325173
},
{
"epoch": 0.755882739581964,
"grad_norm": 1.3359375,
"learning_rate": 1.542936372684715e-05,
"loss": 2.944654846191406,
"step": 5750,
"token_acc": 0.7283420737645054
},
{
"epoch": 0.7624556329696333,
"grad_norm": 1.875,
"learning_rate": 1.4652347360895974e-05,
"loss": 2.939736328125,
"step": 5800,
"token_acc": 0.7268706393279225
},
{
"epoch": 0.7690285263573025,
"grad_norm": 0.76953125,
"learning_rate": 1.3892033719378733e-05,
"loss": 2.93244384765625,
"step": 5850,
"token_acc": 0.7285302853931682
},
{
"epoch": 0.7756014197449718,
"grad_norm": 0.94140625,
"learning_rate": 1.3148782071109316e-05,
"loss": 2.9408578491210937,
"step": 5900,
"token_acc": 0.7278175440358985
},
{
"epoch": 0.782174313132641,
"grad_norm": 1.484375,
"learning_rate": 1.242294362264716e-05,
"loss": 2.9430950927734374,
"step": 5950,
"token_acc": 0.7295751633986928
},
{
"epoch": 0.7887472065203103,
"grad_norm": 1.4296875,
"learning_rate": 1.1714861352342516e-05,
"loss": 2.9315008544921874,
"step": 6000,
"token_acc": 0.7277595148644234
},
{
"epoch": 0.7887472065203103,
"eval_loss": 3.0107288360595703,
"eval_runtime": 233.9306,
"eval_samples_per_second": 188.402,
"eval_steps_per_second": 2.945,
"eval_token_acc": 0.731372226577765,
"step": 6000
},
{
"epoch": 0.7953200999079795,
"grad_norm": 0.82421875,
"learning_rate": 1.102486984826997e-05,
"loss": 2.943741760253906,
"step": 6050,
"token_acc": 0.7261186754076315
},
{
"epoch": 0.8018929932956488,
"grad_norm": 2.890625,
"learning_rate": 1.0353295150126464e-05,
"loss": 2.950082092285156,
"step": 6100,
"token_acc": 0.7274443740200821
},
{
"epoch": 0.808465886683318,
"grad_norm": 1.3046875,
"learning_rate": 9.700454595168722e-06,
"loss": 2.9356060791015626,
"step": 6150,
"token_acc": 0.7293063981153356
},
{
"epoch": 0.8150387800709873,
"grad_norm": 1.65625,
"learning_rate": 9.066656668262768e-06,
"loss": 2.9104058837890623,
"step": 6200,
"token_acc": 0.7318215067017955
},
{
"epoch": 0.8216116734586565,
"grad_norm": 0.77734375,
"learning_rate": 8.452200856116476e-06,
"loss": 2.9222601318359374,
"step": 6250,
"token_acc": 0.7299152664201609
},
{
"epoch": 0.8281845668463258,
"grad_norm": 0.96875,
"learning_rate": 7.857377505763975e-06,
"loss": 2.9406292724609373,
"step": 6300,
"token_acc": 0.726691403760897
},
{
"epoch": 0.834757460233995,
"grad_norm": 1.3984375,
"learning_rate": 7.282467687368782e-06,
"loss": 2.9068536376953125,
"step": 6350,
"token_acc": 0.7295383005455669
},
{
"epoch": 0.8413303536216643,
"grad_norm": 1.328125,
"learning_rate": 6.727743061410596e-06,
"loss": 2.888341369628906,
"step": 6400,
"token_acc": 0.7322604328702462
},
{
"epoch": 0.8479032470093335,
"grad_norm": 1.046875,
"learning_rate": 6.193465750318261e-06,
"loss": 2.912044677734375,
"step": 6450,
"token_acc": 0.7291989862335507
},
{
"epoch": 0.8544761403970028,
"grad_norm": 1.609375,
"learning_rate": 5.679888214609841e-06,
"loss": 2.9116754150390625,
"step": 6500,
"token_acc": 0.7306753128197583
},
{
"epoch": 0.861049033784672,
"grad_norm": 0.66796875,
"learning_rate": 5.187253133598246e-06,
"loss": 2.9217819213867187,
"step": 6550,
"token_acc": 0.7290718677559322
},
{
"epoch": 0.8676219271723412,
"grad_norm": 0.6640625,
"learning_rate": 4.7157932907185845e-06,
"loss": 2.9495709228515623,
"step": 6600,
"token_acc": 0.7272880433697411
},
{
"epoch": 0.8741948205600105,
"grad_norm": 0.87890625,
"learning_rate": 4.265731463531719e-06,
"loss": 2.9202850341796873,
"step": 6650,
"token_acc": 0.7294969244142452
},
{
"epoch": 0.8807677139476797,
"grad_norm": 1.5234375,
"learning_rate": 3.837280318455827e-06,
"loss": 2.9236526489257812,
"step": 6700,
"token_acc": 0.7294696189323149
},
{
"epoch": 0.887340607335349,
"grad_norm": 0.94921875,
"learning_rate": 3.43064231027575e-06,
"loss": 2.9318450927734374,
"step": 6750,
"token_acc": 0.729651563898742
},
{
"epoch": 0.8939135007230182,
"grad_norm": 1.0,
"learning_rate": 3.046009586477605e-06,
"loss": 2.9082293701171875,
"step": 6800,
"token_acc": 0.7320524294069299
},
{
"epoch": 0.9004863941106875,
"grad_norm": 1.484375,
"learning_rate": 2.683563896453917e-06,
"loss": 2.9276959228515627,
"step": 6850,
"token_acc": 0.7282627260549229
},
{
"epoch": 0.9070592874983567,
"grad_norm": 1.7890625,
"learning_rate": 2.343476505622072e-06,
"loss": 2.974371337890625,
"step": 6900,
"token_acc": 0.7233268070551668
},
{
"epoch": 0.913632180886026,
"grad_norm": 0.7421875,
"learning_rate": 2.025908114496744e-06,
"loss": 2.9127767944335936,
"step": 6950,
"token_acc": 0.7283729460872796
},
{
"epoch": 0.9202050742736952,
"grad_norm": 2.296875,
"learning_rate": 1.7310087827546173e-06,
"loss": 2.9310015869140624,
"step": 7000,
"token_acc": 0.7272420944365269
},
{
"epoch": 0.9267779676613646,
"grad_norm": 1.4375,
"learning_rate": 1.458917858327058e-06,
"loss": 2.91599853515625,
"step": 7050,
"token_acc": 0.7301543818025843
},
{
"epoch": 0.9333508610490338,
"grad_norm": 0.91015625,
"learning_rate": 1.2097639115544456e-06,
"loss": 2.923740234375,
"step": 7100,
"token_acc": 0.7292321100855983
},
{
"epoch": 0.9399237544367031,
"grad_norm": 0.87109375,
"learning_rate": 9.836646744332156e-07,
"loss": 2.9268426513671875,
"step": 7150,
"token_acc": 0.7261143873475261
},
{
"epoch": 0.9464966478243723,
"grad_norm": 1.25,
"learning_rate": 7.807269849843035e-07,
"loss": 2.9185272216796876,
"step": 7200,
"token_acc": 0.730116966408788
},
{
"epoch": 0.9530695412120416,
"grad_norm": 2.125,
"learning_rate": 6.010467367692562e-07,
"loss": 2.92894775390625,
"step": 7250,
"token_acc": 0.7276647733776859
},
{
"epoch": 0.9596424345997108,
"grad_norm": 0.6640625,
"learning_rate": 4.447088335779559e-07,
"loss": 2.927828369140625,
"step": 7300,
"token_acc": 0.7292929216155228
},
{
"epoch": 0.9662153279873801,
"grad_norm": 1.109375,
"learning_rate": 3.1178714930923394e-07,
"loss": 2.9167010498046877,
"step": 7350,
"token_acc": 0.7269486212165064
},
{
"epoch": 0.9727882213750493,
"grad_norm": 0.62890625,
"learning_rate": 2.0234449306350544e-07,
"loss": 2.953262939453125,
"step": 7400,
"token_acc": 0.7281943615347045
},
{
"epoch": 0.9793611147627186,
"grad_norm": 1.859375,
"learning_rate": 1.1643257946372665e-07,
"loss": 2.9185284423828124,
"step": 7450,
"token_acc": 0.7292828366372295
},
{
"epoch": 0.9859340081503878,
"grad_norm": 0.91015625,
"learning_rate": 5.409200421883642e-08,
"loss": 2.90837890625,
"step": 7500,
"token_acc": 0.730748791161207
},
{
"epoch": 0.9925069015380571,
"grad_norm": 0.84765625,
"learning_rate": 1.535222494120636e-08,
"loss": 2.93629150390625,
"step": 7550,
"token_acc": 0.7281583011324892
},
{
"epoch": 0.9990797949257263,
"grad_norm": 1.6015625,
"learning_rate": 2.3154722708640387e-10,
"loss": 2.9230780029296874,
"step": 7600,
"token_acc": 0.7279480983015442
},
{
"epoch": 1.0,
"eval_loss": 3.0064425468444824,
"eval_runtime": 233.1422,
"eval_samples_per_second": 189.039,
"eval_steps_per_second": 2.955,
"eval_token_acc": 0.7314566337614171,
"step": 7607
}
],
"logging_steps": 50,
"max_steps": 7607,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.331969097391669e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}