| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.8393662784597629, |
| "eval_steps": 500, |
| "global_step": 1500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.6571732765436173, |
| "epoch": 0.013989437974329381, |
| "grad_norm": 0.0257568359375, |
| "learning_rate": 8.888888888888889e-05, |
| "loss": 1.9518862915039064, |
| "mean_token_accuracy": 0.5733682353049516, |
| "num_tokens": 294540.0, |
| "step": 25 |
| }, |
| { |
| "entropy": 1.7126372032612562, |
| "epoch": 0.027978875948658763, |
| "grad_norm": 0.01416015625, |
| "learning_rate": 0.0001814814814814815, |
| "loss": 1.6628182983398438, |
| "mean_token_accuracy": 0.6153687073290348, |
| "num_tokens": 593048.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 1.6262741280719637, |
| "epoch": 0.041968313922988144, |
| "grad_norm": 0.01806640625, |
| "learning_rate": 0.00019993435766282977, |
| "loss": 1.584449005126953, |
| "mean_token_accuracy": 0.6257701878249645, |
| "num_tokens": 891346.0, |
| "step": 75 |
| }, |
| { |
| "entropy": 1.58288999453187, |
| "epoch": 0.055957751897317526, |
| "grad_norm": 0.016357421875, |
| "learning_rate": 0.00019966783335941717, |
| "loss": 1.5507936096191406, |
| "mean_token_accuracy": 0.6367124810814857, |
| "num_tokens": 1184611.0, |
| "step": 100 |
| }, |
| { |
| "entropy": 1.5433334041386844, |
| "epoch": 0.0699471898716469, |
| "grad_norm": 0.0189208984375, |
| "learning_rate": 0.00019919687071661834, |
| "loss": 1.512821044921875, |
| "mean_token_accuracy": 0.6423260240256786, |
| "num_tokens": 1480125.0, |
| "step": 125 |
| }, |
| { |
| "entropy": 1.560064207613468, |
| "epoch": 0.08393662784597629, |
| "grad_norm": 0.017822265625, |
| "learning_rate": 0.00019852243577149215, |
| "loss": 1.520980224609375, |
| "mean_token_accuracy": 0.6398584572970867, |
| "num_tokens": 1770905.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 1.4850163259357214, |
| "epoch": 0.09792606582030566, |
| "grad_norm": 0.0235595703125, |
| "learning_rate": 0.00019764591192282993, |
| "loss": 1.460587158203125, |
| "mean_token_accuracy": 0.6578983479738235, |
| "num_tokens": 2061126.0, |
| "step": 175 |
| }, |
| { |
| "entropy": 1.4848217015340923, |
| "epoch": 0.11191550379463505, |
| "grad_norm": 0.03662109375, |
| "learning_rate": 0.00019656909709353236, |
| "loss": 1.4420854187011718, |
| "mean_token_accuracy": 0.6592926564067603, |
| "num_tokens": 2352476.0, |
| "step": 200 |
| }, |
| { |
| "entropy": 1.4575298309139908, |
| "epoch": 0.12590494176896444, |
| "grad_norm": 0.0196533203125, |
| "learning_rate": 0.00019529420004271567, |
| "loss": 1.4135537719726563, |
| "mean_token_accuracy": 0.6653153513371944, |
| "num_tokens": 2647293.0, |
| "step": 225 |
| }, |
| { |
| "entropy": 1.440113589912653, |
| "epoch": 0.1398943797432938, |
| "grad_norm": 0.0238037109375, |
| "learning_rate": 0.00019382383583511206, |
| "loss": 1.416180877685547, |
| "mean_token_accuracy": 0.6659375876933337, |
| "num_tokens": 2941859.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 1.4387904461845755, |
| "epoch": 0.1538838177176232, |
| "grad_norm": 0.0263671875, |
| "learning_rate": 0.00019216102047705746, |
| "loss": 1.38541015625, |
| "mean_token_accuracy": 0.6683934908360243, |
| "num_tokens": 3237763.0, |
| "step": 275 |
| }, |
| { |
| "entropy": 1.3664818671159447, |
| "epoch": 0.16787325569195258, |
| "grad_norm": 0.02001953125, |
| "learning_rate": 0.00019030916473006888, |
| "loss": 1.3284611511230469, |
| "mean_token_accuracy": 0.6843961907178163, |
| "num_tokens": 3541854.0, |
| "step": 300 |
| }, |
| { |
| "entropy": 1.3906436143629253, |
| "epoch": 0.18186269366628197, |
| "grad_norm": 0.0296630859375, |
| "learning_rate": 0.00018827206711470137, |
| "loss": 1.3411039733886718, |
| "mean_token_accuracy": 0.6773389867693186, |
| "num_tokens": 3832012.0, |
| "step": 325 |
| }, |
| { |
| "entropy": 1.3563544462341814, |
| "epoch": 0.19585213164061133, |
| "grad_norm": 0.0233154296875, |
| "learning_rate": 0.000186053906119035, |
| "loss": 1.3131784057617188, |
| "mean_token_accuracy": 0.686054406017065, |
| "num_tokens": 4129238.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 1.3752087260596455, |
| "epoch": 0.20984156961494072, |
| "grad_norm": 0.0272216796875, |
| "learning_rate": 0.00018365923162777336, |
| "loss": 1.338323974609375, |
| "mean_token_accuracy": 0.6795472690463066, |
| "num_tokens": 4414939.0, |
| "step": 375 |
| }, |
| { |
| "entropy": 1.3106766785029322, |
| "epoch": 0.2238310075892701, |
| "grad_norm": 0.027099609375, |
| "learning_rate": 0.0001810929555895348, |
| "loss": 1.2676075744628905, |
| "mean_token_accuracy": 0.6990727451443672, |
| "num_tokens": 4711065.0, |
| "step": 400 |
| }, |
| { |
| "entropy": 1.2741564935538918, |
| "epoch": 0.2378204455635995, |
| "grad_norm": 0.0341796875, |
| "learning_rate": 0.00017836034194147922, |
| "loss": 1.231330337524414, |
| "mean_token_accuracy": 0.7052560625225306, |
| "num_tokens": 5006515.0, |
| "step": 425 |
| }, |
| { |
| "entropy": 1.255356255499646, |
| "epoch": 0.2518098835379289, |
| "grad_norm": 0.025390625, |
| "learning_rate": 0.000175466995811937, |
| "loss": 1.2256485748291015, |
| "mean_token_accuracy": 0.709921350106597, |
| "num_tokens": 5298743.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 1.2404796612542122, |
| "epoch": 0.26579932151225827, |
| "grad_norm": 0.029296875, |
| "learning_rate": 0.00017241885202318787, |
| "loss": 1.199592056274414, |
| "mean_token_accuracy": 0.7124969442933797, |
| "num_tokens": 5587292.0, |
| "step": 475 |
| }, |
| { |
| "entropy": 1.2135492172930389, |
| "epoch": 0.2797887594865876, |
| "grad_norm": 0.0341796875, |
| "learning_rate": 0.00016922216291797248, |
| "loss": 1.1718869781494141, |
| "mean_token_accuracy": 0.7194690197706223, |
| "num_tokens": 5878179.0, |
| "step": 500 |
| }, |
| { |
| "entropy": 1.1519292463734745, |
| "epoch": 0.293778197460917, |
| "grad_norm": 0.044921875, |
| "learning_rate": 0.0001658834855347071, |
| "loss": 1.1038724517822265, |
| "mean_token_accuracy": 0.7360028678923846, |
| "num_tokens": 6174073.0, |
| "step": 525 |
| }, |
| { |
| "entropy": 1.2023496214020997, |
| "epoch": 0.3077676354352464, |
| "grad_norm": 0.035400390625, |
| "learning_rate": 0.00016240966815770754, |
| "loss": 1.1433139801025392, |
| "mean_token_accuracy": 0.7237949234992266, |
| "num_tokens": 6471377.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 1.2287814150052145, |
| "epoch": 0.32175707340957577, |
| "grad_norm": 0.0262451171875, |
| "learning_rate": 0.00015880783627001026, |
| "loss": 1.196635513305664, |
| "mean_token_accuracy": 0.7167265053838492, |
| "num_tokens": 6762946.0, |
| "step": 575 |
| }, |
| { |
| "entropy": 1.1513058760762214, |
| "epoch": 0.33574651138390516, |
| "grad_norm": 0.03759765625, |
| "learning_rate": 0.0001550853779376045, |
| "loss": 1.104804229736328, |
| "mean_token_accuracy": 0.7350956991314888, |
| "num_tokens": 7058166.0, |
| "step": 600 |
| }, |
| { |
| "entropy": 1.1233506935788318, |
| "epoch": 0.34973594935823454, |
| "grad_norm": 0.038818359375, |
| "learning_rate": 0.00015124992865505523, |
| "loss": 1.0836825561523438, |
| "mean_token_accuracy": 0.7427010469138622, |
| "num_tokens": 7351236.0, |
| "step": 625 |
| }, |
| { |
| "entropy": 1.1373229866661132, |
| "epoch": 0.36372538733256393, |
| "grad_norm": 0.03466796875, |
| "learning_rate": 0.00014730935568360102, |
| "loss": 1.0919375610351563, |
| "mean_token_accuracy": 0.7399685283005237, |
| "num_tokens": 7642275.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 1.1448556556599214, |
| "epoch": 0.3777148253068933, |
| "grad_norm": 0.033447265625, |
| "learning_rate": 0.0001432717419138532, |
| "loss": 1.0814331817626952, |
| "mean_token_accuracy": 0.7378807676583529, |
| "num_tokens": 7935222.0, |
| "step": 675 |
| }, |
| { |
| "entropy": 1.040786258596927, |
| "epoch": 0.39170426328122265, |
| "grad_norm": 0.0311279296875, |
| "learning_rate": 0.0001391453692861967, |
| "loss": 0.9921630096435546, |
| "mean_token_accuracy": 0.7638325883448124, |
| "num_tokens": 8234237.0, |
| "step": 700 |
| }, |
| { |
| "entropy": 1.1662991707585753, |
| "epoch": 0.40569370125555204, |
| "grad_norm": 0.0302734375, |
| "learning_rate": 0.0001349387018029003, |
| "loss": 1.118931884765625, |
| "mean_token_accuracy": 0.732994829416275, |
| "num_tokens": 8529663.0, |
| "step": 725 |
| }, |
| { |
| "entropy": 1.049061678093858, |
| "epoch": 0.41968313922988143, |
| "grad_norm": 0.041015625, |
| "learning_rate": 0.00013066036816678273, |
| "loss": 0.9903451538085938, |
| "mean_token_accuracy": 0.7613946237415076, |
| "num_tokens": 8817967.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 1.1110075389547274, |
| "epoch": 0.4336725772042108, |
| "grad_norm": 0.0361328125, |
| "learning_rate": 0.0001263191440820448, |
| "loss": 1.0462547302246095, |
| "mean_token_accuracy": 0.7477244459837675, |
| "num_tokens": 9107363.0, |
| "step": 775 |
| }, |
| { |
| "entropy": 1.024681367701851, |
| "epoch": 0.4476620151785402, |
| "grad_norm": 0.039306640625, |
| "learning_rate": 0.00012192393425357353, |
| "loss": 0.9708639526367188, |
| "mean_token_accuracy": 0.7660154252499342, |
| "num_tokens": 9398385.0, |
| "step": 800 |
| }, |
| { |
| "entropy": 1.0914598937472328, |
| "epoch": 0.4616514531528696, |
| "grad_norm": 0.03662109375, |
| "learning_rate": 0.00011748375412163984, |
| "loss": 1.0285856628417969, |
| "mean_token_accuracy": 0.7533510192483663, |
| "num_tokens": 9692995.0, |
| "step": 825 |
| }, |
| { |
| "entropy": 1.070437773577869, |
| "epoch": 0.475640891127199, |
| "grad_norm": 0.0308837890625, |
| "learning_rate": 0.00011300771136945658, |
| "loss": 1.002013931274414, |
| "mean_token_accuracy": 0.7574464529007673, |
| "num_tokens": 9985540.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 1.0500496013974772, |
| "epoch": 0.48963032910152837, |
| "grad_norm": 0.033935546875, |
| "learning_rate": 0.00010850498724152798, |
| "loss": 1.0061278533935547, |
| "mean_token_accuracy": 0.7601391483098269, |
| "num_tokens": 10280858.0, |
| "step": 875 |
| }, |
| { |
| "entropy": 0.9829169751703739, |
| "epoch": 0.5036197670758578, |
| "grad_norm": 0.035888671875, |
| "learning_rate": 0.00010398481771111037, |
| "loss": 0.9166652679443359, |
| "mean_token_accuracy": 0.7795024861395359, |
| "num_tokens": 10571194.0, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.9858249768940732, |
| "epoch": 0.5176092050501871, |
| "grad_norm": 0.038818359375, |
| "learning_rate": 9.94564745354137e-05, |
| "loss": 0.9250814819335937, |
| "mean_token_accuracy": 0.7783324559032917, |
| "num_tokens": 10861535.0, |
| "step": 925 |
| }, |
| { |
| "entropy": 0.9617049301974475, |
| "epoch": 0.5315986430245165, |
| "grad_norm": 0.0419921875, |
| "learning_rate": 9.492924623740271e-05, |
| "loss": 0.9093881988525391, |
| "mean_token_accuracy": 0.7818463468551635, |
| "num_tokens": 11157232.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.9786588902026415, |
| "epoch": 0.5455880809988459, |
| "grad_norm": 0.031494140625, |
| "learning_rate": 9.04124190532087e-05, |
| "loss": 0.9290876007080078, |
| "mean_token_accuracy": 0.7784507688879967, |
| "num_tokens": 11451817.0, |
| "step": 975 |
| }, |
| { |
| "entropy": 0.9941231621522456, |
| "epoch": 0.5595775189731752, |
| "grad_norm": 0.03662109375, |
| "learning_rate": 8.591525788423168e-05, |
| "loss": 0.9281739807128906, |
| "mean_token_accuracy": 0.7773911864310503, |
| "num_tokens": 11746337.0, |
| "step": 1000 |
| }, |
| { |
| "entropy": 0.9537221440812573, |
| "epoch": 0.5735669569475046, |
| "grad_norm": 0.0303955078125, |
| "learning_rate": 8.144698729300455e-05, |
| "loss": 0.8921097564697266, |
| "mean_token_accuracy": 0.7855567722022534, |
| "num_tokens": 12037412.0, |
| "step": 1025 |
| }, |
| { |
| "entropy": 0.9557448910363019, |
| "epoch": 0.587556394921834, |
| "grad_norm": 0.029052734375, |
| "learning_rate": 7.701677258179996e-05, |
| "loss": 0.9018702697753906, |
| "mean_token_accuracy": 0.78552112005651, |
| "num_tokens": 12330073.0, |
| "step": 1050 |
| }, |
| { |
| "entropy": 0.9831135825067758, |
| "epoch": 0.6015458328961634, |
| "grad_norm": 0.031005859375, |
| "learning_rate": 7.263370099279172e-05, |
| "loss": 0.9220442199707031, |
| "mean_token_accuracy": 0.7789029122143983, |
| "num_tokens": 12623027.0, |
| "step": 1075 |
| }, |
| { |
| "entropy": 0.8838035979354754, |
| "epoch": 0.6155352708704928, |
| "grad_norm": 0.040283203125, |
| "learning_rate": 6.83067630683331e-05, |
| "loss": 0.8227420043945313, |
| "mean_token_accuracy": 0.8038330339640379, |
| "num_tokens": 12916473.0, |
| "step": 1100 |
| }, |
| { |
| "entropy": 0.883046350381337, |
| "epoch": 0.6295247088448221, |
| "grad_norm": 0.037353515625, |
| "learning_rate": 6.404483420958494e-05, |
| "loss": 0.8318240356445312, |
| "mean_token_accuracy": 0.8026394218951464, |
| "num_tokens": 13210350.0, |
| "step": 1125 |
| }, |
| { |
| "entropy": 0.9277670627878979, |
| "epoch": 0.6435141468191515, |
| "grad_norm": 0.02783203125, |
| "learning_rate": 5.9856656471321636e-05, |
| "loss": 0.8780873870849609, |
| "mean_token_accuracy": 0.7957132039964199, |
| "num_tokens": 13501483.0, |
| "step": 1150 |
| }, |
| { |
| "entropy": 0.9232714768499136, |
| "epoch": 0.6575035847934809, |
| "grad_norm": 0.0255126953125, |
| "learning_rate": 5.575082063025619e-05, |
| "loss": 0.8550299835205079, |
| "mean_token_accuracy": 0.7936624947935342, |
| "num_tokens": 13797592.0, |
| "step": 1175 |
| }, |
| { |
| "entropy": 0.899801154313609, |
| "epoch": 0.6714930227678103, |
| "grad_norm": 0.03369140625, |
| "learning_rate": 5.173574856366683e-05, |
| "loss": 0.8404230499267578, |
| "mean_token_accuracy": 0.7989044986665249, |
| "num_tokens": 14085126.0, |
| "step": 1200 |
| }, |
| { |
| "entropy": 0.8731577204307541, |
| "epoch": 0.6854824607421397, |
| "grad_norm": 0.033203125, |
| "learning_rate": 4.7819675974469356e-05, |
| "loss": 0.8161344146728515, |
| "mean_token_accuracy": 0.8059349462389946, |
| "num_tokens": 14383100.0, |
| "step": 1225 |
| }, |
| { |
| "entropy": 0.9630948188621551, |
| "epoch": 0.6994718987164691, |
| "grad_norm": 0.033935546875, |
| "learning_rate": 4.401063549816984e-05, |
| "loss": 0.91357177734375, |
| "mean_token_accuracy": 0.7841827914863825, |
| "num_tokens": 14682819.0, |
| "step": 1250 |
| }, |
| { |
| "entropy": 0.9606740972073748, |
| "epoch": 0.7134613366907985, |
| "grad_norm": 0.02783203125, |
| "learning_rate": 4.031644022634831e-05, |
| "loss": 0.896214599609375, |
| "mean_token_accuracy": 0.7852722837030888, |
| "num_tokens": 14972956.0, |
| "step": 1275 |
| }, |
| { |
| "entropy": 0.9392211610358209, |
| "epoch": 0.7274507746651279, |
| "grad_norm": 0.033447265625, |
| "learning_rate": 3.674466768047078e-05, |
| "loss": 0.8826549530029297, |
| "mean_token_accuracy": 0.7912480696290731, |
| "num_tokens": 15271172.0, |
| "step": 1300 |
| }, |
| { |
| "entropy": 0.9369741760706529, |
| "epoch": 0.7414402126394573, |
| "grad_norm": 0.0228271484375, |
| "learning_rate": 3.330264426890114e-05, |
| "loss": 0.8716287994384766, |
| "mean_token_accuracy": 0.7906265539675951, |
| "num_tokens": 15567414.0, |
| "step": 1325 |
| }, |
| { |
| "entropy": 0.8475008888402954, |
| "epoch": 0.7554296506137866, |
| "grad_norm": 0.02734375, |
| "learning_rate": 2.9997430258996208e-05, |
| "loss": 0.7837786865234375, |
| "mean_token_accuracy": 0.8113828019797802, |
| "num_tokens": 15862493.0, |
| "step": 1350 |
| }, |
| { |
| "entropy": 0.8544112277310342, |
| "epoch": 0.7694190885881159, |
| "grad_norm": 0.0400390625, |
| "learning_rate": 2.6835805295107897e-05, |
| "loss": 0.7910999298095703, |
| "mean_token_accuracy": 0.8121292810887099, |
| "num_tokens": 16154298.0, |
| "step": 1375 |
| }, |
| { |
| "entropy": 0.8836554439552128, |
| "epoch": 0.7834085265624453, |
| "grad_norm": 0.03271484375, |
| "learning_rate": 2.382425449219854e-05, |
| "loss": 0.8322444152832031, |
| "mean_token_accuracy": 0.8058089179545641, |
| "num_tokens": 16450244.0, |
| "step": 1400 |
| }, |
| { |
| "entropy": 0.9750821837875993, |
| "epoch": 0.7973979645367747, |
| "grad_norm": 0.0279541015625, |
| "learning_rate": 2.0968955133593805e-05, |
| "loss": 0.9110565948486328, |
| "mean_token_accuracy": 0.7827954424917698, |
| "num_tokens": 16742734.0, |
| "step": 1425 |
| }, |
| { |
| "entropy": 0.8045650019776076, |
| "epoch": 0.8113874025111041, |
| "grad_norm": 0.037841796875, |
| "learning_rate": 1.8275764000159222e-05, |
| "loss": 0.7386186218261719, |
| "mean_token_accuracy": 0.8229204141348601, |
| "num_tokens": 17040960.0, |
| "step": 1450 |
| }, |
| { |
| "entropy": 0.9300361970160157, |
| "epoch": 0.8253768404854335, |
| "grad_norm": 0.0303955078125, |
| "learning_rate": 1.5750205356889937e-05, |
| "loss": 0.8793846130371094, |
| "mean_token_accuracy": 0.7934108532965183, |
| "num_tokens": 17340398.0, |
| "step": 1475 |
| }, |
| { |
| "entropy": 0.9426807796442881, |
| "epoch": 0.8393662784597629, |
| "grad_norm": 0.032958984375, |
| "learning_rate": 1.339745962155613e-05, |
| "loss": 0.8905538940429687, |
| "mean_token_accuracy": 0.7892987384647131, |
| "num_tokens": 17629025.0, |
| "step": 1500 |
| } |
| ], |
| "logging_steps": 25, |
| "max_steps": 1788, |
| "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": 7.521588501358848e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
