| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 1431, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0006988120195667365, |
| "grad_norm": 15.825864096286791, |
| "learning_rate": 0.0, |
| "loss": 26.435, |
| "num_tokens": 7466.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.001397624039133473, |
| "grad_norm": 12.251496631056742, |
| "learning_rate": 4.651162790697674e-06, |
| "loss": 26.4932, |
| "num_tokens": 14867.0, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0020964360587002098, |
| "grad_norm": 13.519476973182835, |
| "learning_rate": 9.302325581395349e-06, |
| "loss": 26.4287, |
| "num_tokens": 22758.0, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.002795248078266946, |
| "grad_norm": 15.38208620339498, |
| "learning_rate": 1.3953488372093024e-05, |
| "loss": 26.07, |
| "num_tokens": 30251.0, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0034940600978336828, |
| "grad_norm": 14.123335556294114, |
| "learning_rate": 1.8604651162790697e-05, |
| "loss": 25.6902, |
| "num_tokens": 37442.0, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0041928721174004195, |
| "grad_norm": 38.41325227449582, |
| "learning_rate": 2.3255813953488374e-05, |
| "loss": 25.6086, |
| "num_tokens": 44450.0, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.004891684136967156, |
| "grad_norm": 49.20044694933379, |
| "learning_rate": 2.7906976744186048e-05, |
| "loss": 25.6802, |
| "num_tokens": 51596.0, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.005590496156533892, |
| "grad_norm": 24.837518059934986, |
| "learning_rate": 3.2558139534883724e-05, |
| "loss": 25.0283, |
| "num_tokens": 58501.0, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.006289308176100629, |
| "grad_norm": 35.02610727478617, |
| "learning_rate": 3.7209302325581394e-05, |
| "loss": 24.5402, |
| "num_tokens": 65085.0, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0069881201956673656, |
| "grad_norm": 18.49743608785304, |
| "learning_rate": 4.186046511627907e-05, |
| "loss": 24.1431, |
| "num_tokens": 72520.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.007686932215234102, |
| "grad_norm": 16.142798556485957, |
| "learning_rate": 4.651162790697675e-05, |
| "loss": 23.9589, |
| "num_tokens": 79490.0, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.008385744234800839, |
| "grad_norm": 15.957529997417875, |
| "learning_rate": 5.1162790697674425e-05, |
| "loss": 23.7223, |
| "num_tokens": 86902.0, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.009084556254367574, |
| "grad_norm": 10.570962604630072, |
| "learning_rate": 5.5813953488372095e-05, |
| "loss": 22.9895, |
| "num_tokens": 92738.0, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.009783368273934312, |
| "grad_norm": 18.77079227590805, |
| "learning_rate": 6.0465116279069765e-05, |
| "loss": 23.1725, |
| "num_tokens": 100060.0, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.010482180293501049, |
| "grad_norm": 7.3705046557873635, |
| "learning_rate": 6.511627906976745e-05, |
| "loss": 22.0704, |
| "num_tokens": 107649.0, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.011180992313067784, |
| "grad_norm": 9.756769805700404, |
| "learning_rate": 6.976744186046513e-05, |
| "loss": 21.7304, |
| "num_tokens": 114327.0, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.011879804332634521, |
| "grad_norm": 7.1793383883530995, |
| "learning_rate": 7.441860465116279e-05, |
| "loss": 21.4763, |
| "num_tokens": 121566.0, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.012578616352201259, |
| "grad_norm": 7.137243029031286, |
| "learning_rate": 7.906976744186047e-05, |
| "loss": 21.3598, |
| "num_tokens": 128982.0, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.013277428371767994, |
| "grad_norm": 13.638520130002313, |
| "learning_rate": 8.372093023255814e-05, |
| "loss": 21.1434, |
| "num_tokens": 136536.0, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.013976240391334731, |
| "grad_norm": 11.69983873670843, |
| "learning_rate": 8.837209302325582e-05, |
| "loss": 21.0494, |
| "num_tokens": 144292.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.014675052410901468, |
| "grad_norm": 5.355565758628784, |
| "learning_rate": 9.30232558139535e-05, |
| "loss": 20.9966, |
| "num_tokens": 151133.0, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.015373864430468204, |
| "grad_norm": 4.717512683910897, |
| "learning_rate": 9.767441860465116e-05, |
| "loss": 20.7461, |
| "num_tokens": 158252.0, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01607267645003494, |
| "grad_norm": 7.555779518574644, |
| "learning_rate": 0.00010232558139534885, |
| "loss": 20.3074, |
| "num_tokens": 165457.0, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.016771488469601678, |
| "grad_norm": 6.72771421928895, |
| "learning_rate": 0.00010697674418604651, |
| "loss": 20.2158, |
| "num_tokens": 172785.0, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.017470300489168415, |
| "grad_norm": 9.911658022140173, |
| "learning_rate": 0.00011162790697674419, |
| "loss": 20.2821, |
| "num_tokens": 179494.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01816911250873515, |
| "grad_norm": 4.297543791874311, |
| "learning_rate": 0.00011627906976744187, |
| "loss": 20.2236, |
| "num_tokens": 187501.0, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.018867924528301886, |
| "grad_norm": 3.398148288869523, |
| "learning_rate": 0.00012093023255813953, |
| "loss": 20.1017, |
| "num_tokens": 194486.0, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.019566736547868623, |
| "grad_norm": 8.051576045312112, |
| "learning_rate": 0.0001255813953488372, |
| "loss": 19.7799, |
| "num_tokens": 201489.0, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.02026554856743536, |
| "grad_norm": 15.488905000197105, |
| "learning_rate": 0.0001302325581395349, |
| "loss": 19.9241, |
| "num_tokens": 209354.0, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.020964360587002098, |
| "grad_norm": 3.283628115165652, |
| "learning_rate": 0.00013488372093023256, |
| "loss": 19.404, |
| "num_tokens": 216345.0, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.02166317260656883, |
| "grad_norm": 3.9857224790977117, |
| "learning_rate": 0.00013953488372093025, |
| "loss": 19.6057, |
| "num_tokens": 223037.0, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.02236198462613557, |
| "grad_norm": 3.22352219751297, |
| "learning_rate": 0.00014418604651162791, |
| "loss": 19.5688, |
| "num_tokens": 230416.0, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.023060796645702306, |
| "grad_norm": 3.8491447527833986, |
| "learning_rate": 0.00014883720930232558, |
| "loss": 19.2043, |
| "num_tokens": 237115.0, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.023759608665269043, |
| "grad_norm": 3.7874852298026425, |
| "learning_rate": 0.00015348837209302327, |
| "loss": 19.084, |
| "num_tokens": 244240.0, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.02445842068483578, |
| "grad_norm": 3.8289674897719554, |
| "learning_rate": 0.00015813953488372093, |
| "loss": 19.075, |
| "num_tokens": 251723.0, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.025157232704402517, |
| "grad_norm": 3.9440239720411228, |
| "learning_rate": 0.00016279069767441862, |
| "loss": 18.5376, |
| "num_tokens": 258362.0, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.02585604472396925, |
| "grad_norm": 3.0319951387170065, |
| "learning_rate": 0.00016744186046511629, |
| "loss": 18.7768, |
| "num_tokens": 265593.0, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.026554856743535988, |
| "grad_norm": 3.3898521484712276, |
| "learning_rate": 0.00017209302325581395, |
| "loss": 18.1432, |
| "num_tokens": 272624.0, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.027253668763102725, |
| "grad_norm": 2.9301870676453463, |
| "learning_rate": 0.00017674418604651164, |
| "loss": 18.588, |
| "num_tokens": 279181.0, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.027952480782669462, |
| "grad_norm": 2.654404660666782, |
| "learning_rate": 0.0001813953488372093, |
| "loss": 18.2087, |
| "num_tokens": 285979.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.0286512928022362, |
| "grad_norm": 3.7416042387150528, |
| "learning_rate": 0.000186046511627907, |
| "loss": 18.2133, |
| "num_tokens": 293070.0, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.029350104821802937, |
| "grad_norm": 3.105130611534875, |
| "learning_rate": 0.00019069767441860466, |
| "loss": 17.687, |
| "num_tokens": 299558.0, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.03004891684136967, |
| "grad_norm": 3.3080203495958673, |
| "learning_rate": 0.00019534883720930232, |
| "loss": 17.8142, |
| "num_tokens": 306733.0, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.030747728860936407, |
| "grad_norm": 4.550644498461186, |
| "learning_rate": 0.0002, |
| "loss": 17.5783, |
| "num_tokens": 313625.0, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.031446540880503145, |
| "grad_norm": 2.89083091176252, |
| "learning_rate": 0.00019999974385219888, |
| "loss": 17.6028, |
| "num_tokens": 320835.0, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.03214535290006988, |
| "grad_norm": 2.688470184996243, |
| "learning_rate": 0.00019999897541010772, |
| "loss": 17.0988, |
| "num_tokens": 328348.0, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.03284416491963662, |
| "grad_norm": 4.12814028592051, |
| "learning_rate": 0.00019999769467766323, |
| "loss": 16.9933, |
| "num_tokens": 335913.0, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.033542976939203356, |
| "grad_norm": 2.6768197525570696, |
| "learning_rate": 0.00019999590166142655, |
| "loss": 16.9533, |
| "num_tokens": 343205.0, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.03424178895877009, |
| "grad_norm": 4.933854683750753, |
| "learning_rate": 0.0001999935963705832, |
| "loss": 16.78, |
| "num_tokens": 350261.0, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.03494060097833683, |
| "grad_norm": 2.4948269682154813, |
| "learning_rate": 0.0001999907788169431, |
| "loss": 16.986, |
| "num_tokens": 357367.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.03563941299790356, |
| "grad_norm": 2.5971622220228503, |
| "learning_rate": 0.00019998744901494049, |
| "loss": 16.3867, |
| "num_tokens": 364705.0, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.0363382250174703, |
| "grad_norm": 3.5285288483689126, |
| "learning_rate": 0.00019998360698163375, |
| "loss": 16.6009, |
| "num_tokens": 371422.0, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.037037037037037035, |
| "grad_norm": 2.3072905364363407, |
| "learning_rate": 0.00019997925273670543, |
| "loss": 16.3405, |
| "num_tokens": 378970.0, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.03773584905660377, |
| "grad_norm": 2.111981005938095, |
| "learning_rate": 0.0001999743863024622, |
| "loss": 16.1949, |
| "num_tokens": 386189.0, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.03843466107617051, |
| "grad_norm": 2.296216898523924, |
| "learning_rate": 0.00019996900770383454, |
| "loss": 16.3281, |
| "num_tokens": 393815.0, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.039133473095737246, |
| "grad_norm": 2.666184126601885, |
| "learning_rate": 0.0001999631169683768, |
| "loss": 16.1931, |
| "num_tokens": 399923.0, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.039832285115303984, |
| "grad_norm": 3.7207866770964886, |
| "learning_rate": 0.0001999567141262669, |
| "loss": 16.1543, |
| "num_tokens": 406458.0, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.04053109713487072, |
| "grad_norm": 1.8938232560374766, |
| "learning_rate": 0.0001999497992103064, |
| "loss": 16.0368, |
| "num_tokens": 414710.0, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.04122990915443746, |
| "grad_norm": 2.412458505414354, |
| "learning_rate": 0.00019994237225592012, |
| "loss": 15.8979, |
| "num_tokens": 422447.0, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.041928721174004195, |
| "grad_norm": 2.5317332125646894, |
| "learning_rate": 0.00019993443330115592, |
| "loss": 15.5823, |
| "num_tokens": 428907.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.04262753319357093, |
| "grad_norm": 2.8439927802342164, |
| "learning_rate": 0.0001999259823866848, |
| "loss": 15.7681, |
| "num_tokens": 436247.0, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.04332634521313766, |
| "grad_norm": 2.3124546860656836, |
| "learning_rate": 0.0001999170195558004, |
| "loss": 15.496, |
| "num_tokens": 443409.0, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.0440251572327044, |
| "grad_norm": 2.1789745411385413, |
| "learning_rate": 0.0001999075448544189, |
| "loss": 15.4492, |
| "num_tokens": 450956.0, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.04472396925227114, |
| "grad_norm": 2.62233287160653, |
| "learning_rate": 0.00019989755833107876, |
| "loss": 15.2355, |
| "num_tokens": 458417.0, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.045422781271837874, |
| "grad_norm": 1.8545665130668028, |
| "learning_rate": 0.00019988706003694055, |
| "loss": 15.3334, |
| "num_tokens": 465196.0, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.04612159329140461, |
| "grad_norm": 3.6525347342712147, |
| "learning_rate": 0.00019987605002578653, |
| "loss": 15.1933, |
| "num_tokens": 472489.0, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.04682040531097135, |
| "grad_norm": 2.7277334237717783, |
| "learning_rate": 0.0001998645283540205, |
| "loss": 15.1653, |
| "num_tokens": 479315.0, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.047519217330538085, |
| "grad_norm": 1.97496245060142, |
| "learning_rate": 0.00019985249508066755, |
| "loss": 15.3583, |
| "num_tokens": 486906.0, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.04821802935010482, |
| "grad_norm": 2.365948985865034, |
| "learning_rate": 0.0001998399502673735, |
| "loss": 14.9784, |
| "num_tokens": 493938.0, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.04891684136967156, |
| "grad_norm": 2.0207852222510008, |
| "learning_rate": 0.00019982689397840496, |
| "loss": 15.0744, |
| "num_tokens": 500590.0, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.0496156533892383, |
| "grad_norm": 2.137629225341362, |
| "learning_rate": 0.00019981332628064865, |
| "loss": 15.0679, |
| "num_tokens": 508345.0, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.050314465408805034, |
| "grad_norm": 2.709600688855553, |
| "learning_rate": 0.0001997992472436114, |
| "loss": 14.9663, |
| "num_tokens": 515272.0, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.05101327742837177, |
| "grad_norm": 2.769437204379797, |
| "learning_rate": 0.0001997846569394194, |
| "loss": 15.0369, |
| "num_tokens": 522104.0, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.0517120894479385, |
| "grad_norm": 2.1465516152240407, |
| "learning_rate": 0.00019976955544281815, |
| "loss": 14.9015, |
| "num_tokens": 529458.0, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.05241090146750524, |
| "grad_norm": 2.554970599066564, |
| "learning_rate": 0.000199753942831172, |
| "loss": 14.563, |
| "num_tokens": 537278.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.053109713487071976, |
| "grad_norm": 2.388126495406421, |
| "learning_rate": 0.0001997378191844636, |
| "loss": 14.813, |
| "num_tokens": 544549.0, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.05380852550663871, |
| "grad_norm": 2.1919533051739863, |
| "learning_rate": 0.00019972118458529375, |
| "loss": 14.4483, |
| "num_tokens": 551068.0, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.05450733752620545, |
| "grad_norm": 1.6927300542355745, |
| "learning_rate": 0.00019970403911888078, |
| "loss": 14.7465, |
| "num_tokens": 558188.0, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.05520614954577219, |
| "grad_norm": 2.2064497543576587, |
| "learning_rate": 0.0001996863828730601, |
| "loss": 14.6674, |
| "num_tokens": 565805.0, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.055904961565338925, |
| "grad_norm": 2.4114054632830295, |
| "learning_rate": 0.00019966821593828392, |
| "loss": 14.5352, |
| "num_tokens": 573381.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.05660377358490566, |
| "grad_norm": 2.4134183889931102, |
| "learning_rate": 0.0001996495384076206, |
| "loss": 14.3595, |
| "num_tokens": 581350.0, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.0573025856044724, |
| "grad_norm": 1.8864006743237123, |
| "learning_rate": 0.0001996303503767544, |
| "loss": 14.3631, |
| "num_tokens": 588473.0, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.058001397624039136, |
| "grad_norm": 3.13170686212352, |
| "learning_rate": 0.00019961065194398466, |
| "loss": 14.2639, |
| "num_tokens": 595623.0, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.05870020964360587, |
| "grad_norm": 1.8895852504927524, |
| "learning_rate": 0.00019959044321022563, |
| "loss": 14.5274, |
| "num_tokens": 602999.0, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.0593990216631726, |
| "grad_norm": 2.0439231597559364, |
| "learning_rate": 0.00019956972427900578, |
| "loss": 14.1757, |
| "num_tokens": 610543.0, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.06009783368273934, |
| "grad_norm": 2.2725286826693627, |
| "learning_rate": 0.00019954849525646726, |
| "loss": 14.0905, |
| "num_tokens": 617203.0, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.06079664570230608, |
| "grad_norm": 1.7055612047248596, |
| "learning_rate": 0.0001995267562513654, |
| "loss": 14.5731, |
| "num_tokens": 623602.0, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.061495457721872815, |
| "grad_norm": 2.0031518563876785, |
| "learning_rate": 0.00019950450737506824, |
| "loss": 14.2865, |
| "num_tokens": 631160.0, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.06219426974143955, |
| "grad_norm": 1.815051060076169, |
| "learning_rate": 0.00019948174874155573, |
| "loss": 13.8439, |
| "num_tokens": 638355.0, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.06289308176100629, |
| "grad_norm": 1.8441847733219316, |
| "learning_rate": 0.00019945848046741934, |
| "loss": 14.061, |
| "num_tokens": 645362.0, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.06359189378057302, |
| "grad_norm": 2.155812571108576, |
| "learning_rate": 0.00019943470267186144, |
| "loss": 13.9919, |
| "num_tokens": 652539.0, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.06429070580013976, |
| "grad_norm": 1.817937039545139, |
| "learning_rate": 0.00019941041547669465, |
| "loss": 14.1473, |
| "num_tokens": 659738.0, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.0649895178197065, |
| "grad_norm": 1.6248726317728508, |
| "learning_rate": 0.0001993856190063412, |
| "loss": 13.9793, |
| "num_tokens": 667862.0, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.06568832983927324, |
| "grad_norm": 2.094683855819787, |
| "learning_rate": 0.00019936031338783225, |
| "loss": 14.0755, |
| "num_tokens": 675183.0, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.06638714185883997, |
| "grad_norm": 2.423753515861936, |
| "learning_rate": 0.00019933449875080746, |
| "loss": 13.7537, |
| "num_tokens": 682199.0, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.06708595387840671, |
| "grad_norm": 1.8615093715014854, |
| "learning_rate": 0.00019930817522751401, |
| "loss": 13.7573, |
| "num_tokens": 689412.0, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.06778476589797344, |
| "grad_norm": 1.922603628179396, |
| "learning_rate": 0.0001992813429528062, |
| "loss": 13.7809, |
| "num_tokens": 696608.0, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.06848357791754019, |
| "grad_norm": 1.885694054680989, |
| "learning_rate": 0.0001992540020641446, |
| "loss": 13.6316, |
| "num_tokens": 703838.0, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.06918238993710692, |
| "grad_norm": 1.830733189483595, |
| "learning_rate": 0.0001992261527015953, |
| "loss": 13.8526, |
| "num_tokens": 711432.0, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.06988120195667366, |
| "grad_norm": 2.0529905069418053, |
| "learning_rate": 0.00019919779500782948, |
| "loss": 13.8049, |
| "num_tokens": 717755.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.07058001397624039, |
| "grad_norm": 1.9178698722664131, |
| "learning_rate": 0.0001991689291281223, |
| "loss": 13.9275, |
| "num_tokens": 725381.0, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.07127882599580712, |
| "grad_norm": 1.7826032618004608, |
| "learning_rate": 0.00019913955521035234, |
| "loss": 13.4739, |
| "num_tokens": 732317.0, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.07197763801537387, |
| "grad_norm": 1.7430765768053889, |
| "learning_rate": 0.00019910967340500094, |
| "loss": 13.7472, |
| "num_tokens": 739043.0, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.0726764500349406, |
| "grad_norm": 1.9327097576731953, |
| "learning_rate": 0.00019907928386515126, |
| "loss": 13.7117, |
| "num_tokens": 745729.0, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.07337526205450734, |
| "grad_norm": 1.6523098222253636, |
| "learning_rate": 0.00019904838674648763, |
| "loss": 13.5757, |
| "num_tokens": 753195.0, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.07407407407407407, |
| "grad_norm": 1.6969490994020622, |
| "learning_rate": 0.00019901698220729458, |
| "loss": 13.748, |
| "num_tokens": 759938.0, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.07477288609364081, |
| "grad_norm": 1.7677649372168398, |
| "learning_rate": 0.00019898507040845616, |
| "loss": 13.5139, |
| "num_tokens": 767625.0, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.07547169811320754, |
| "grad_norm": 1.9518708854729268, |
| "learning_rate": 0.00019895265151345518, |
| "loss": 13.4042, |
| "num_tokens": 775131.0, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.07617051013277429, |
| "grad_norm": 1.5910502551681978, |
| "learning_rate": 0.00019891972568837214, |
| "loss": 13.5223, |
| "num_tokens": 782395.0, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.07686932215234102, |
| "grad_norm": 1.7329435914061764, |
| "learning_rate": 0.00019888629310188465, |
| "loss": 13.513, |
| "num_tokens": 789064.0, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.07756813417190776, |
| "grad_norm": 1.483874271007651, |
| "learning_rate": 0.00019885235392526636, |
| "loss": 13.4427, |
| "num_tokens": 796698.0, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.07826694619147449, |
| "grad_norm": 1.6804353651923707, |
| "learning_rate": 0.00019881790833238617, |
| "loss": 13.4666, |
| "num_tokens": 803919.0, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.07896575821104122, |
| "grad_norm": 1.9157840259811565, |
| "learning_rate": 0.00019878295649970734, |
| "loss": 13.3714, |
| "num_tokens": 810971.0, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.07966457023060797, |
| "grad_norm": 1.7551309996052762, |
| "learning_rate": 0.0001987474986062866, |
| "loss": 13.2645, |
| "num_tokens": 817887.0, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.0803633822501747, |
| "grad_norm": 1.5349632976896173, |
| "learning_rate": 0.00019871153483377315, |
| "loss": 13.3252, |
| "num_tokens": 824738.0, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.08106219426974144, |
| "grad_norm": 1.924756300417848, |
| "learning_rate": 0.0001986750653664078, |
| "loss": 13.4228, |
| "num_tokens": 832079.0, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.08176100628930817, |
| "grad_norm": 1.5896909041784393, |
| "learning_rate": 0.0001986380903910221, |
| "loss": 13.27, |
| "num_tokens": 838908.0, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.08245981830887492, |
| "grad_norm": 1.575684845736584, |
| "learning_rate": 0.00019860061009703713, |
| "loss": 13.4192, |
| "num_tokens": 845348.0, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.08315863032844165, |
| "grad_norm": 1.8295088795123042, |
| "learning_rate": 0.00019856262467646282, |
| "loss": 13.4711, |
| "num_tokens": 852162.0, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.08385744234800839, |
| "grad_norm": 1.7581038567698788, |
| "learning_rate": 0.00019852413432389684, |
| "loss": 13.3141, |
| "num_tokens": 860170.0, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.08455625436757512, |
| "grad_norm": 1.526290883383786, |
| "learning_rate": 0.00019848513923652358, |
| "loss": 13.2942, |
| "num_tokens": 867476.0, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.08525506638714186, |
| "grad_norm": 1.878529626375439, |
| "learning_rate": 0.00019844563961411309, |
| "loss": 13.2712, |
| "num_tokens": 874866.0, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.0859538784067086, |
| "grad_norm": 1.489846227446125, |
| "learning_rate": 0.00019840563565902026, |
| "loss": 13.2977, |
| "num_tokens": 881774.0, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.08665269042627533, |
| "grad_norm": 1.6815017809617654, |
| "learning_rate": 0.00019836512757618355, |
| "loss": 13.2761, |
| "num_tokens": 888149.0, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.08735150244584207, |
| "grad_norm": 1.6091641966868473, |
| "learning_rate": 0.00019832411557312414, |
| "loss": 13.4763, |
| "num_tokens": 894693.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.0880503144654088, |
| "grad_norm": 1.5095813931005195, |
| "learning_rate": 0.00019828259985994463, |
| "loss": 13.5071, |
| "num_tokens": 901024.0, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.08874912648497554, |
| "grad_norm": 1.6693788073219258, |
| "learning_rate": 0.00019824058064932831, |
| "loss": 13.2807, |
| "num_tokens": 908206.0, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.08944793850454227, |
| "grad_norm": 1.9687315987537344, |
| "learning_rate": 0.00019819805815653768, |
| "loss": 13.363, |
| "num_tokens": 914376.0, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.09014675052410902, |
| "grad_norm": 1.7385897372739076, |
| "learning_rate": 0.00019815503259941358, |
| "loss": 13.2634, |
| "num_tokens": 921721.0, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.09084556254367575, |
| "grad_norm": 1.428721148603436, |
| "learning_rate": 0.0001981115041983741, |
| "loss": 13.2458, |
| "num_tokens": 928937.0, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.09154437456324249, |
| "grad_norm": 1.662634637805847, |
| "learning_rate": 0.0001980674731764133, |
| "loss": 13.2113, |
| "num_tokens": 936881.0, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.09224318658280922, |
| "grad_norm": 1.6355733105429533, |
| "learning_rate": 0.00019802293975910016, |
| "loss": 13.3508, |
| "num_tokens": 943685.0, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.09294199860237597, |
| "grad_norm": 1.6463942231883306, |
| "learning_rate": 0.00019797790417457742, |
| "loss": 13.0835, |
| "num_tokens": 950976.0, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.0936408106219427, |
| "grad_norm": 1.4062728688315147, |
| "learning_rate": 0.0001979323666535604, |
| "loss": 13.2962, |
| "num_tokens": 957928.0, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.09433962264150944, |
| "grad_norm": 1.5672116073719475, |
| "learning_rate": 0.00019788632742933585, |
| "loss": 13.105, |
| "num_tokens": 964412.0, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.09503843466107617, |
| "grad_norm": 1.8304590876950395, |
| "learning_rate": 0.00019783978673776063, |
| "loss": 13.1619, |
| "num_tokens": 971468.0, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.0957372466806429, |
| "grad_norm": 1.5169514478830495, |
| "learning_rate": 0.00019779274481726073, |
| "loss": 13.223, |
| "num_tokens": 978459.0, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.09643605870020965, |
| "grad_norm": 1.320041796641737, |
| "learning_rate": 0.00019774520190882978, |
| "loss": 13.1441, |
| "num_tokens": 985920.0, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.09713487071977638, |
| "grad_norm": 1.579778797268716, |
| "learning_rate": 0.00019769715825602803, |
| "loss": 13.0779, |
| "num_tokens": 992764.0, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.09783368273934312, |
| "grad_norm": 1.6097268033651382, |
| "learning_rate": 0.00019764861410498098, |
| "loss": 13.2544, |
| "num_tokens": 999854.0, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.09853249475890985, |
| "grad_norm": 1.4573003230376171, |
| "learning_rate": 0.00019759956970437825, |
| "loss": 13.0092, |
| "num_tokens": 1006667.0, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.0992313067784766, |
| "grad_norm": 1.5921140115600168, |
| "learning_rate": 0.00019755002530547208, |
| "loss": 13.2447, |
| "num_tokens": 1014089.0, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.09993011879804332, |
| "grad_norm": 1.3438414888588825, |
| "learning_rate": 0.00019749998116207621, |
| "loss": 13.2542, |
| "num_tokens": 1020912.0, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.10062893081761007, |
| "grad_norm": 1.7285759166278971, |
| "learning_rate": 0.00019744943753056472, |
| "loss": 12.9093, |
| "num_tokens": 1028495.0, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.1013277428371768, |
| "grad_norm": 1.3903171195019095, |
| "learning_rate": 0.0001973983946698703, |
| "loss": 13.0009, |
| "num_tokens": 1036332.0, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.10202655485674354, |
| "grad_norm": 1.7455056295429205, |
| "learning_rate": 0.0001973468528414833, |
| "loss": 13.1304, |
| "num_tokens": 1043895.0, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.10272536687631027, |
| "grad_norm": 1.4497939444434873, |
| "learning_rate": 0.0001972948123094503, |
| "loss": 13.0796, |
| "num_tokens": 1051123.0, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.103424178895877, |
| "grad_norm": 1.383580156751911, |
| "learning_rate": 0.00019724227334037256, |
| "loss": 13.1124, |
| "num_tokens": 1058636.0, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.10412299091544375, |
| "grad_norm": 1.4838421489362628, |
| "learning_rate": 0.00019718923620340496, |
| "loss": 13.142, |
| "num_tokens": 1065532.0, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.10482180293501048, |
| "grad_norm": 1.4242535011413662, |
| "learning_rate": 0.00019713570117025443, |
| "loss": 12.8958, |
| "num_tokens": 1073092.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.10552061495457722, |
| "grad_norm": 1.40181080040232, |
| "learning_rate": 0.0001970816685151786, |
| "loss": 12.9283, |
| "num_tokens": 1081083.0, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.10621942697414395, |
| "grad_norm": 1.5313341039141877, |
| "learning_rate": 0.00019702713851498435, |
| "loss": 12.9873, |
| "num_tokens": 1088405.0, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.1069182389937107, |
| "grad_norm": 1.3601093201693377, |
| "learning_rate": 0.00019697211144902648, |
| "loss": 12.9683, |
| "num_tokens": 1095299.0, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.10761705101327743, |
| "grad_norm": 1.6794919624091056, |
| "learning_rate": 0.00019691658759920624, |
| "loss": 12.9984, |
| "num_tokens": 1102272.0, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.10831586303284417, |
| "grad_norm": 1.2881559193849654, |
| "learning_rate": 0.00019686056724996988, |
| "loss": 13.0438, |
| "num_tokens": 1108878.0, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.1090146750524109, |
| "grad_norm": 1.5180984823138908, |
| "learning_rate": 0.00019680405068830717, |
| "loss": 12.9799, |
| "num_tokens": 1116345.0, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.10971348707197764, |
| "grad_norm": 1.4196966779586535, |
| "learning_rate": 0.00019674703820374994, |
| "loss": 13.065, |
| "num_tokens": 1123338.0, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.11041229909154437, |
| "grad_norm": 1.3929707667138889, |
| "learning_rate": 0.0001966895300883707, |
| "loss": 12.8968, |
| "num_tokens": 1130116.0, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.1111111111111111, |
| "grad_norm": 1.4996662049248695, |
| "learning_rate": 0.00019663152663678099, |
| "loss": 12.7327, |
| "num_tokens": 1137314.0, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.11180992313067785, |
| "grad_norm": 1.369079829503796, |
| "learning_rate": 0.0001965730281461299, |
| "loss": 13.0669, |
| "num_tokens": 1143586.0, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.11250873515024458, |
| "grad_norm": 1.4354897097544304, |
| "learning_rate": 0.00019651403491610268, |
| "loss": 12.9036, |
| "num_tokens": 1150678.0, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.11320754716981132, |
| "grad_norm": 1.352851751508517, |
| "learning_rate": 0.000196454547248919, |
| "loss": 12.9177, |
| "num_tokens": 1158316.0, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.11390635918937805, |
| "grad_norm": 1.3710181240525572, |
| "learning_rate": 0.00019639456544933155, |
| "loss": 13.0527, |
| "num_tokens": 1165236.0, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.1146051712089448, |
| "grad_norm": 1.3068223937565793, |
| "learning_rate": 0.0001963340898246245, |
| "loss": 13.0438, |
| "num_tokens": 1172589.0, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.11530398322851153, |
| "grad_norm": 1.346519471881747, |
| "learning_rate": 0.00019627312068461184, |
| "loss": 13.0086, |
| "num_tokens": 1179343.0, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.11600279524807827, |
| "grad_norm": 1.446789885424941, |
| "learning_rate": 0.00019621165834163572, |
| "loss": 12.9444, |
| "num_tokens": 1185779.0, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.116701607267645, |
| "grad_norm": 1.444884622592261, |
| "learning_rate": 0.00019614970311056503, |
| "loss": 13.2269, |
| "num_tokens": 1192364.0, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.11740041928721175, |
| "grad_norm": 1.2891615557860914, |
| "learning_rate": 0.00019608725530879375, |
| "loss": 12.9378, |
| "num_tokens": 1199385.0, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.11809923130677848, |
| "grad_norm": 1.3524191128592733, |
| "learning_rate": 0.00019602431525623918, |
| "loss": 13.1315, |
| "num_tokens": 1206524.0, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.1187980433263452, |
| "grad_norm": 1.466681450368447, |
| "learning_rate": 0.00019596088327534047, |
| "loss": 12.8783, |
| "num_tokens": 1213487.0, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.11949685534591195, |
| "grad_norm": 1.4000232584973735, |
| "learning_rate": 0.0001958969596910568, |
| "loss": 12.887, |
| "num_tokens": 1220301.0, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.12019566736547868, |
| "grad_norm": 1.250384646500827, |
| "learning_rate": 0.000195832544830866, |
| "loss": 12.8478, |
| "num_tokens": 1227778.0, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.12089447938504543, |
| "grad_norm": 1.5137312829748157, |
| "learning_rate": 0.00019576763902476242, |
| "loss": 12.9285, |
| "num_tokens": 1234261.0, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.12159329140461216, |
| "grad_norm": 1.2791777538329325, |
| "learning_rate": 0.0001957022426052558, |
| "loss": 12.8265, |
| "num_tokens": 1241757.0, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.1222921034241789, |
| "grad_norm": 1.5901299698985887, |
| "learning_rate": 0.00019563635590736901, |
| "loss": 12.8283, |
| "num_tokens": 1248424.0, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.12299091544374563, |
| "grad_norm": 1.4082846461212457, |
| "learning_rate": 0.00019556997926863673, |
| "loss": 13.054, |
| "num_tokens": 1255116.0, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.12368972746331237, |
| "grad_norm": 1.336316502475352, |
| "learning_rate": 0.0001955031130291036, |
| "loss": 12.8992, |
| "num_tokens": 1262373.0, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.1243885394828791, |
| "grad_norm": 1.2337463600865273, |
| "learning_rate": 0.0001954357575313224, |
| "loss": 13.0176, |
| "num_tokens": 1268591.0, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.12508735150244585, |
| "grad_norm": 1.3019629873776766, |
| "learning_rate": 0.0001953679131203524, |
| "loss": 12.8964, |
| "num_tokens": 1276309.0, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.12578616352201258, |
| "grad_norm": 1.359626952499068, |
| "learning_rate": 0.00019529958014375746, |
| "loss": 12.8146, |
| "num_tokens": 1283604.0, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.1264849755415793, |
| "grad_norm": 1.1515862878249248, |
| "learning_rate": 0.0001952307589516045, |
| "loss": 12.9346, |
| "num_tokens": 1290423.0, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.12718378756114604, |
| "grad_norm": 1.3387432887733188, |
| "learning_rate": 0.00019516144989646143, |
| "loss": 13.0495, |
| "num_tokens": 1297162.0, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.1278825995807128, |
| "grad_norm": 1.2279239336305443, |
| "learning_rate": 0.00019509165333339551, |
| "loss": 12.798, |
| "num_tokens": 1304042.0, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.12858141160027953, |
| "grad_norm": 1.43782424281205, |
| "learning_rate": 0.0001950213696199714, |
| "loss": 12.7759, |
| "num_tokens": 1311266.0, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.12928022361984626, |
| "grad_norm": 1.3241777082441377, |
| "learning_rate": 0.00019495059911624958, |
| "loss": 13.0, |
| "num_tokens": 1317490.0, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.129979035639413, |
| "grad_norm": 1.4217170996471962, |
| "learning_rate": 0.00019487934218478413, |
| "loss": 12.9661, |
| "num_tokens": 1324708.0, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.13067784765897975, |
| "grad_norm": 1.34114667099107, |
| "learning_rate": 0.0001948075991906212, |
| "loss": 12.7199, |
| "num_tokens": 1331506.0, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.13137665967854648, |
| "grad_norm": 1.2485002323335002, |
| "learning_rate": 0.00019473537050129704, |
| "loss": 12.8164, |
| "num_tokens": 1338737.0, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.1320754716981132, |
| "grad_norm": 1.2386002612681404, |
| "learning_rate": 0.00019466265648683602, |
| "loss": 12.8269, |
| "num_tokens": 1346238.0, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.13277428371767994, |
| "grad_norm": 1.193140463955583, |
| "learning_rate": 0.0001945894575197488, |
| "loss": 12.8081, |
| "num_tokens": 1353786.0, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.1334730957372467, |
| "grad_norm": 1.5872251747340467, |
| "learning_rate": 0.00019451577397503053, |
| "loss": 12.9692, |
| "num_tokens": 1360969.0, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.13417190775681342, |
| "grad_norm": 1.2664833823468389, |
| "learning_rate": 0.00019444160623015874, |
| "loss": 13.0089, |
| "num_tokens": 1368167.0, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.13487071977638015, |
| "grad_norm": 1.3919751858860443, |
| "learning_rate": 0.00019436695466509152, |
| "loss": 12.7512, |
| "num_tokens": 1375092.0, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.13556953179594688, |
| "grad_norm": 1.7287267721473023, |
| "learning_rate": 0.00019429181966226558, |
| "loss": 12.583, |
| "num_tokens": 1383015.0, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.13626834381551362, |
| "grad_norm": 1.3019053109225178, |
| "learning_rate": 0.00019421620160659417, |
| "loss": 12.6838, |
| "num_tokens": 1389785.0, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.13696715583508037, |
| "grad_norm": 1.427516186724372, |
| "learning_rate": 0.00019414010088546535, |
| "loss": 12.6612, |
| "num_tokens": 1397770.0, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.1376659678546471, |
| "grad_norm": 1.5627247304701708, |
| "learning_rate": 0.00019406351788873972, |
| "loss": 12.6157, |
| "num_tokens": 1404674.0, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.13836477987421383, |
| "grad_norm": 1.3129375645013237, |
| "learning_rate": 0.00019398645300874865, |
| "loss": 12.9586, |
| "num_tokens": 1411618.0, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.13906359189378056, |
| "grad_norm": 1.2811801120707738, |
| "learning_rate": 0.00019390890664029204, |
| "loss": 12.6556, |
| "num_tokens": 1418834.0, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.13976240391334732, |
| "grad_norm": 1.2764420734679425, |
| "learning_rate": 0.0001938308791806366, |
| "loss": 12.8285, |
| "num_tokens": 1425877.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.14046121593291405, |
| "grad_norm": 1.4621638790890457, |
| "learning_rate": 0.0001937523710295136, |
| "loss": 12.7881, |
| "num_tokens": 1432515.0, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.14116002795248078, |
| "grad_norm": 1.2904825789726404, |
| "learning_rate": 0.00019367338258911675, |
| "loss": 12.7984, |
| "num_tokens": 1439548.0, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.1418588399720475, |
| "grad_norm": 1.186391641270174, |
| "learning_rate": 0.0001935939142641004, |
| "loss": 12.634, |
| "num_tokens": 1446322.0, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.14255765199161424, |
| "grad_norm": 1.1674776839727241, |
| "learning_rate": 0.0001935139664615773, |
| "loss": 12.8383, |
| "num_tokens": 1453298.0, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.143256464011181, |
| "grad_norm": 1.2683482772147976, |
| "learning_rate": 0.00019343353959111652, |
| "loss": 12.6733, |
| "num_tokens": 1460188.0, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.14395527603074773, |
| "grad_norm": 1.216633183257395, |
| "learning_rate": 0.00019335263406474137, |
| "loss": 12.7315, |
| "num_tokens": 1467199.0, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.14465408805031446, |
| "grad_norm": 1.2669421899729538, |
| "learning_rate": 0.00019327125029692735, |
| "loss": 12.6752, |
| "num_tokens": 1474116.0, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.1453529000698812, |
| "grad_norm": 1.1816422309516645, |
| "learning_rate": 0.00019318938870459984, |
| "loss": 12.7585, |
| "num_tokens": 1480988.0, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.14605171208944795, |
| "grad_norm": 1.593879641391394, |
| "learning_rate": 0.00019310704970713224, |
| "loss": 12.5382, |
| "num_tokens": 1487900.0, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.14675052410901468, |
| "grad_norm": 1.4309781033240252, |
| "learning_rate": 0.0001930242337263436, |
| "loss": 12.672, |
| "num_tokens": 1495543.0, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.1474493361285814, |
| "grad_norm": 1.3370449709559875, |
| "learning_rate": 0.00019294094118649653, |
| "loss": 12.542, |
| "num_tokens": 1502498.0, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.14814814814814814, |
| "grad_norm": 1.7429475926881917, |
| "learning_rate": 0.00019285717251429506, |
| "loss": 12.5638, |
| "num_tokens": 1509580.0, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.1488469601677149, |
| "grad_norm": 1.3587122905893039, |
| "learning_rate": 0.00019277292813888244, |
| "loss": 12.689, |
| "num_tokens": 1516376.0, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.14954577218728163, |
| "grad_norm": 1.4852579200171538, |
| "learning_rate": 0.00019268820849183883, |
| "loss": 12.7086, |
| "num_tokens": 1523015.0, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.15024458420684836, |
| "grad_norm": 1.3259646669042442, |
| "learning_rate": 0.00019260301400717938, |
| "loss": 12.8351, |
| "num_tokens": 1530696.0, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.1509433962264151, |
| "grad_norm": 1.432491095454994, |
| "learning_rate": 0.00019251734512135157, |
| "loss": 13.0261, |
| "num_tokens": 1537893.0, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.15164220824598182, |
| "grad_norm": 1.5012452677994659, |
| "learning_rate": 0.00019243120227323333, |
| "loss": 12.5507, |
| "num_tokens": 1545460.0, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.15234102026554858, |
| "grad_norm": 1.449104510035988, |
| "learning_rate": 0.00019234458590413077, |
| "loss": 12.6152, |
| "num_tokens": 1552764.0, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.1530398322851153, |
| "grad_norm": 1.2430393467361687, |
| "learning_rate": 0.0001922574964577757, |
| "loss": 12.6997, |
| "num_tokens": 1559826.0, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.15373864430468204, |
| "grad_norm": 1.1751985575150206, |
| "learning_rate": 0.0001921699343803235, |
| "loss": 12.7505, |
| "num_tokens": 1567575.0, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.15443745632424877, |
| "grad_norm": 1.224128766914656, |
| "learning_rate": 0.00019208190012035087, |
| "loss": 12.6943, |
| "num_tokens": 1574362.0, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.15513626834381553, |
| "grad_norm": 1.4730426557595246, |
| "learning_rate": 0.00019199339412885347, |
| "loss": 12.4975, |
| "num_tokens": 1581335.0, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.15583508036338226, |
| "grad_norm": 1.245219689341052, |
| "learning_rate": 0.00019190441685924353, |
| "loss": 12.52, |
| "num_tokens": 1588536.0, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.15653389238294899, |
| "grad_norm": 1.3566970317678715, |
| "learning_rate": 0.00019181496876734776, |
| "loss": 12.67, |
| "num_tokens": 1595480.0, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.15723270440251572, |
| "grad_norm": 1.2125862352464616, |
| "learning_rate": 0.0001917250503114048, |
| "loss": 12.6823, |
| "num_tokens": 1602858.0, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.15793151642208245, |
| "grad_norm": 1.2658499521661504, |
| "learning_rate": 0.0001916346619520629, |
| "loss": 12.4635, |
| "num_tokens": 1610558.0, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.1586303284416492, |
| "grad_norm": 1.1986536717425549, |
| "learning_rate": 0.00019154380415237768, |
| "loss": 12.7901, |
| "num_tokens": 1617490.0, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.15932914046121593, |
| "grad_norm": 1.2965597058989395, |
| "learning_rate": 0.00019145247737780961, |
| "loss": 12.6139, |
| "num_tokens": 1624406.0, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.16002795248078266, |
| "grad_norm": 1.0789453042961938, |
| "learning_rate": 0.00019136068209622183, |
| "loss": 12.6052, |
| "num_tokens": 1631539.0, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.1607267645003494, |
| "grad_norm": 1.2665628322889513, |
| "learning_rate": 0.00019126841877787745, |
| "loss": 12.5022, |
| "num_tokens": 1638417.0, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.16142557651991615, |
| "grad_norm": 1.1381170673306098, |
| "learning_rate": 0.00019117568789543742, |
| "loss": 12.5689, |
| "num_tokens": 1645769.0, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.16212438853948288, |
| "grad_norm": 1.2090318652793792, |
| "learning_rate": 0.00019108248992395795, |
| "loss": 12.6484, |
| "num_tokens": 1653953.0, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.1628232005590496, |
| "grad_norm": 1.0651593003631354, |
| "learning_rate": 0.0001909888253408882, |
| "loss": 12.3967, |
| "num_tokens": 1661632.0, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.16352201257861634, |
| "grad_norm": 1.289629674228255, |
| "learning_rate": 0.00019089469462606765, |
| "loss": 12.463, |
| "num_tokens": 1668594.0, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.1642208245981831, |
| "grad_norm": 1.2696209740727902, |
| "learning_rate": 0.00019080009826172387, |
| "loss": 12.4893, |
| "num_tokens": 1675137.0, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.16491963661774983, |
| "grad_norm": 1.295734136130956, |
| "learning_rate": 0.00019070503673246982, |
| "loss": 12.4947, |
| "num_tokens": 1682132.0, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.16561844863731656, |
| "grad_norm": 1.186085096495133, |
| "learning_rate": 0.0001906095105253016, |
| "loss": 12.803, |
| "num_tokens": 1688115.0, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.1663172606568833, |
| "grad_norm": 1.3469233034400552, |
| "learning_rate": 0.00019051352012959568, |
| "loss": 12.5784, |
| "num_tokens": 1694968.0, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.16701607267645002, |
| "grad_norm": 1.1030332713184419, |
| "learning_rate": 0.0001904170660371067, |
| "loss": 12.7819, |
| "num_tokens": 1702568.0, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.16771488469601678, |
| "grad_norm": 1.5256563801599314, |
| "learning_rate": 0.00019032014874196474, |
| "loss": 12.7243, |
| "num_tokens": 1709560.0, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.1684136967155835, |
| "grad_norm": 1.0954893842833933, |
| "learning_rate": 0.0001902227687406728, |
| "loss": 12.5503, |
| "num_tokens": 1717750.0, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.16911250873515024, |
| "grad_norm": 1.2813575980659235, |
| "learning_rate": 0.0001901249265321044, |
| "loss": 12.5196, |
| "num_tokens": 1724736.0, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.16981132075471697, |
| "grad_norm": 1.2440300161189415, |
| "learning_rate": 0.00019002662261750078, |
| "loss": 12.5384, |
| "num_tokens": 1732240.0, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.17051013277428373, |
| "grad_norm": 1.2129795823903677, |
| "learning_rate": 0.00018992785750046863, |
| "loss": 12.3593, |
| "num_tokens": 1739543.0, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.17120894479385046, |
| "grad_norm": 1.1970012785599058, |
| "learning_rate": 0.00018982863168697734, |
| "loss": 12.587, |
| "num_tokens": 1746459.0, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.1719077568134172, |
| "grad_norm": 1.3170795769984949, |
| "learning_rate": 0.00018972894568535634, |
| "loss": 12.705, |
| "num_tokens": 1753478.0, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.17260656883298392, |
| "grad_norm": 1.1706949815019843, |
| "learning_rate": 0.00018962880000629258, |
| "loss": 12.3755, |
| "num_tokens": 1760374.0, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.17330538085255065, |
| "grad_norm": 1.3094831534974631, |
| "learning_rate": 0.0001895281951628281, |
| "loss": 12.7245, |
| "num_tokens": 1767809.0, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.1740041928721174, |
| "grad_norm": 1.044605208635119, |
| "learning_rate": 0.000189427131670357, |
| "loss": 12.6817, |
| "num_tokens": 1775284.0, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.17470300489168414, |
| "grad_norm": 1.1554679184494987, |
| "learning_rate": 0.00018932561004662312, |
| "loss": 12.3913, |
| "num_tokens": 1782896.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.17540181691125087, |
| "grad_norm": 1.3348111022458111, |
| "learning_rate": 0.00018922363081171723, |
| "loss": 12.5296, |
| "num_tokens": 1790348.0, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.1761006289308176, |
| "grad_norm": 1.1120352895593362, |
| "learning_rate": 0.0001891211944880746, |
| "loss": 12.8437, |
| "num_tokens": 1796663.0, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.17679944095038436, |
| "grad_norm": 1.5043333334071107, |
| "learning_rate": 0.00018901830160047184, |
| "loss": 12.4049, |
| "num_tokens": 1804575.0, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.1774982529699511, |
| "grad_norm": 1.1726144920737067, |
| "learning_rate": 0.0001889149526760248, |
| "loss": 12.5944, |
| "num_tokens": 1810818.0, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.17819706498951782, |
| "grad_norm": 1.2408621835003835, |
| "learning_rate": 0.0001888111482441855, |
| "loss": 12.4801, |
| "num_tokens": 1817813.0, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.17889587700908455, |
| "grad_norm": 1.379904171536855, |
| "learning_rate": 0.00018870688883673936, |
| "loss": 12.5618, |
| "num_tokens": 1824365.0, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.1795946890286513, |
| "grad_norm": 1.1840265644520456, |
| "learning_rate": 0.00018860217498780285, |
| "loss": 12.5814, |
| "num_tokens": 1831336.0, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.18029350104821804, |
| "grad_norm": 1.3292220144802318, |
| "learning_rate": 0.00018849700723382035, |
| "loss": 12.385, |
| "num_tokens": 1838657.0, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.18099231306778477, |
| "grad_norm": 1.211001471258756, |
| "learning_rate": 0.0001883913861135617, |
| "loss": 12.4776, |
| "num_tokens": 1845965.0, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.1816911250873515, |
| "grad_norm": 1.2646806883821105, |
| "learning_rate": 0.00018828531216811913, |
| "loss": 12.7187, |
| "num_tokens": 1852638.0, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.18238993710691823, |
| "grad_norm": 1.3716555848937888, |
| "learning_rate": 0.00018817878594090494, |
| "loss": 12.598, |
| "num_tokens": 1859877.0, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.18308874912648498, |
| "grad_norm": 1.1204826654463, |
| "learning_rate": 0.00018807180797764822, |
| "loss": 12.5591, |
| "num_tokens": 1866923.0, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.18378756114605171, |
| "grad_norm": 1.3449240572237333, |
| "learning_rate": 0.00018796437882639242, |
| "loss": 12.3795, |
| "num_tokens": 1873292.0, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.18448637316561844, |
| "grad_norm": 1.1392064399539101, |
| "learning_rate": 0.00018785649903749234, |
| "loss": 12.43, |
| "num_tokens": 1879744.0, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.18518518518518517, |
| "grad_norm": 1.390099384211355, |
| "learning_rate": 0.00018774816916361137, |
| "loss": 12.5114, |
| "num_tokens": 1886064.0, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.18588399720475193, |
| "grad_norm": 1.1814107601446537, |
| "learning_rate": 0.00018763938975971872, |
| "loss": 12.3272, |
| "num_tokens": 1893813.0, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.18658280922431866, |
| "grad_norm": 1.2818010177186572, |
| "learning_rate": 0.0001875301613830865, |
| "loss": 12.8937, |
| "num_tokens": 1901211.0, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.1872816212438854, |
| "grad_norm": 1.400621329298869, |
| "learning_rate": 0.00018742048459328682, |
| "loss": 12.51, |
| "num_tokens": 1907987.0, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.18798043326345212, |
| "grad_norm": 1.2330141005937425, |
| "learning_rate": 0.00018731035995218914, |
| "loss": 12.7508, |
| "num_tokens": 1915853.0, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.18867924528301888, |
| "grad_norm": 1.37904712394343, |
| "learning_rate": 0.00018719978802395705, |
| "loss": 12.2983, |
| "num_tokens": 1923310.0, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.1893780573025856, |
| "grad_norm": 1.2467103995837752, |
| "learning_rate": 0.0001870887693750458, |
| "loss": 12.3118, |
| "num_tokens": 1930169.0, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.19007686932215234, |
| "grad_norm": 1.1402621781657731, |
| "learning_rate": 0.00018697730457419893, |
| "loss": 12.452, |
| "num_tokens": 1937617.0, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.19077568134171907, |
| "grad_norm": 1.215321051813323, |
| "learning_rate": 0.00018686539419244578, |
| "loss": 12.4004, |
| "num_tokens": 1944358.0, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.1914744933612858, |
| "grad_norm": 1.3080313421863534, |
| "learning_rate": 0.0001867530388030983, |
| "loss": 12.4736, |
| "num_tokens": 1951629.0, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.19217330538085256, |
| "grad_norm": 1.1294773048320363, |
| "learning_rate": 0.00018664023898174817, |
| "loss": 12.4207, |
| "num_tokens": 1958779.0, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.1928721174004193, |
| "grad_norm": 1.2033795977734214, |
| "learning_rate": 0.00018652699530626398, |
| "loss": 12.3521, |
| "num_tokens": 1966253.0, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.19357092941998602, |
| "grad_norm": 1.152683402138601, |
| "learning_rate": 0.00018641330835678804, |
| "loss": 12.5124, |
| "num_tokens": 1973038.0, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.19426974143955275, |
| "grad_norm": 1.008307851183863, |
| "learning_rate": 0.00018629917871573366, |
| "loss": 12.621, |
| "num_tokens": 1980735.0, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.1949685534591195, |
| "grad_norm": 1.2051981110164842, |
| "learning_rate": 0.0001861846069677819, |
| "loss": 12.4481, |
| "num_tokens": 1988250.0, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.19566736547868624, |
| "grad_norm": 1.0838581398000466, |
| "learning_rate": 0.00018606959369987883, |
| "loss": 12.4857, |
| "num_tokens": 1995184.0, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.19636617749825297, |
| "grad_norm": 1.0917822922153184, |
| "learning_rate": 0.00018595413950123235, |
| "loss": 12.3827, |
| "num_tokens": 2001901.0, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.1970649895178197, |
| "grad_norm": 1.1018335599813311, |
| "learning_rate": 0.00018583824496330923, |
| "loss": 12.685, |
| "num_tokens": 2009242.0, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.19776380153738643, |
| "grad_norm": 1.2127896142017403, |
| "learning_rate": 0.00018572191067983216, |
| "loss": 12.5809, |
| "num_tokens": 2016167.0, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.1984626135569532, |
| "grad_norm": 1.348517937802105, |
| "learning_rate": 0.00018560513724677643, |
| "loss": 12.4575, |
| "num_tokens": 2023059.0, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.19916142557651992, |
| "grad_norm": 1.1090341293586254, |
| "learning_rate": 0.00018548792526236732, |
| "loss": 12.3595, |
| "num_tokens": 2030297.0, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.19986023759608665, |
| "grad_norm": 1.342568850797832, |
| "learning_rate": 0.00018537027532707662, |
| "loss": 12.5375, |
| "num_tokens": 2036674.0, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.20055904961565338, |
| "grad_norm": 1.1060237909798654, |
| "learning_rate": 0.00018525218804361977, |
| "loss": 12.2509, |
| "num_tokens": 2043766.0, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.20125786163522014, |
| "grad_norm": 1.2327170579459596, |
| "learning_rate": 0.00018513366401695276, |
| "loss": 12.5329, |
| "num_tokens": 2051302.0, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.20195667365478687, |
| "grad_norm": 1.1508656944870255, |
| "learning_rate": 0.00018501470385426892, |
| "loss": 12.3698, |
| "num_tokens": 2058562.0, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.2026554856743536, |
| "grad_norm": 1.1595543958178998, |
| "learning_rate": 0.00018489530816499596, |
| "loss": 12.4003, |
| "num_tokens": 2065605.0, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.20335429769392033, |
| "grad_norm": 1.3192276014407542, |
| "learning_rate": 0.00018477547756079276, |
| "loss": 12.4305, |
| "num_tokens": 2072590.0, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.20405310971348709, |
| "grad_norm": 1.1135384173094371, |
| "learning_rate": 0.0001846552126555462, |
| "loss": 12.5111, |
| "num_tokens": 2080039.0, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.20475192173305382, |
| "grad_norm": 1.3261734695119327, |
| "learning_rate": 0.00018453451406536816, |
| "loss": 12.6033, |
| "num_tokens": 2086481.0, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.20545073375262055, |
| "grad_norm": 1.1201785620737152, |
| "learning_rate": 0.00018441338240859215, |
| "loss": 12.5272, |
| "num_tokens": 2093192.0, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.20614954577218728, |
| "grad_norm": 1.1183523903788608, |
| "learning_rate": 0.00018429181830577034, |
| "loss": 12.3886, |
| "num_tokens": 2100572.0, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.206848357791754, |
| "grad_norm": 1.2121568257432795, |
| "learning_rate": 0.00018416982237967028, |
| "loss": 12.5291, |
| "num_tokens": 2107911.0, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.20754716981132076, |
| "grad_norm": 1.162263554260533, |
| "learning_rate": 0.00018404739525527174, |
| "loss": 12.334, |
| "num_tokens": 2115264.0, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.2082459818308875, |
| "grad_norm": 1.1525516113964571, |
| "learning_rate": 0.0001839245375597635, |
| "loss": 12.2785, |
| "num_tokens": 2122114.0, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.20894479385045422, |
| "grad_norm": 1.186494728566964, |
| "learning_rate": 0.0001838012499225401, |
| "loss": 12.3004, |
| "num_tokens": 2129186.0, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.20964360587002095, |
| "grad_norm": 1.2912767009253392, |
| "learning_rate": 0.00018367753297519873, |
| "loss": 12.5505, |
| "num_tokens": 2136056.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.2103424178895877, |
| "grad_norm": 1.1093100716237951, |
| "learning_rate": 0.00018355338735153587, |
| "loss": 12.3422, |
| "num_tokens": 2143135.0, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.21104122990915444, |
| "grad_norm": 1.1832901006966179, |
| "learning_rate": 0.00018342881368754404, |
| "loss": 12.4585, |
| "num_tokens": 2149855.0, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.21174004192872117, |
| "grad_norm": 1.304549193251159, |
| "learning_rate": 0.00018330381262140864, |
| "loss": 12.6785, |
| "num_tokens": 2156629.0, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.2124388539482879, |
| "grad_norm": 1.02534115696337, |
| "learning_rate": 0.00018317838479350472, |
| "loss": 12.5892, |
| "num_tokens": 2163993.0, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.21313766596785463, |
| "grad_norm": 1.2301685746042796, |
| "learning_rate": 0.0001830525308463934, |
| "loss": 12.4886, |
| "num_tokens": 2170517.0, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.2138364779874214, |
| "grad_norm": 1.075671835098473, |
| "learning_rate": 0.00018292625142481906, |
| "loss": 12.4849, |
| "num_tokens": 2177805.0, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.21453529000698812, |
| "grad_norm": 1.1292457741896016, |
| "learning_rate": 0.00018279954717570553, |
| "loss": 12.467, |
| "num_tokens": 2184824.0, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.21523410202655485, |
| "grad_norm": 1.052027926129132, |
| "learning_rate": 0.00018267241874815314, |
| "loss": 12.1912, |
| "num_tokens": 2192640.0, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.21593291404612158, |
| "grad_norm": 1.1526308912958079, |
| "learning_rate": 0.00018254486679343516, |
| "loss": 12.4349, |
| "num_tokens": 2199963.0, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.21663172606568834, |
| "grad_norm": 1.0944974834134447, |
| "learning_rate": 0.00018241689196499475, |
| "loss": 12.425, |
| "num_tokens": 2207388.0, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.21733053808525507, |
| "grad_norm": 1.2455487452556113, |
| "learning_rate": 0.00018228849491844129, |
| "loss": 12.2816, |
| "num_tokens": 2214115.0, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.2180293501048218, |
| "grad_norm": 1.1660092991408635, |
| "learning_rate": 0.00018215967631154717, |
| "loss": 12.3521, |
| "num_tokens": 2221801.0, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.21872816212438853, |
| "grad_norm": 1.2793656576637555, |
| "learning_rate": 0.00018203043680424448, |
| "loss": 12.4023, |
| "num_tokens": 2229449.0, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.2194269741439553, |
| "grad_norm": 1.2171298749865551, |
| "learning_rate": 0.00018190077705862155, |
| "loss": 12.563, |
| "num_tokens": 2236249.0, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.22012578616352202, |
| "grad_norm": 1.3096258411240842, |
| "learning_rate": 0.00018177069773891953, |
| "loss": 12.3427, |
| "num_tokens": 2243354.0, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.22082459818308875, |
| "grad_norm": 1.2231138275461113, |
| "learning_rate": 0.00018164019951152902, |
| "loss": 12.4167, |
| "num_tokens": 2249837.0, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.22152341020265548, |
| "grad_norm": 1.6783299447406501, |
| "learning_rate": 0.00018150928304498675, |
| "loss": 12.3678, |
| "num_tokens": 2256520.0, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.2222222222222222, |
| "grad_norm": 1.2789357635089038, |
| "learning_rate": 0.00018137794900997201, |
| "loss": 12.5104, |
| "num_tokens": 2263145.0, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.22292103424178897, |
| "grad_norm": 1.3687005934905156, |
| "learning_rate": 0.0001812461980793033, |
| "loss": 12.4424, |
| "num_tokens": 2269862.0, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.2236198462613557, |
| "grad_norm": 1.0318168591263561, |
| "learning_rate": 0.0001811140309279348, |
| "loss": 12.5401, |
| "num_tokens": 2276820.0, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.22431865828092243, |
| "grad_norm": 1.3080904924929653, |
| "learning_rate": 0.00018098144823295304, |
| "loss": 12.2369, |
| "num_tokens": 2284748.0, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.22501747030048916, |
| "grad_norm": 1.1758761094017982, |
| "learning_rate": 0.00018084845067357336, |
| "loss": 12.3439, |
| "num_tokens": 2292140.0, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.22571628232005592, |
| "grad_norm": 1.0519584181435369, |
| "learning_rate": 0.00018071503893113638, |
| "loss": 12.5563, |
| "num_tokens": 2300108.0, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.22641509433962265, |
| "grad_norm": 1.3244204464017082, |
| "learning_rate": 0.00018058121368910458, |
| "loss": 12.1751, |
| "num_tokens": 2307811.0, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.22711390635918938, |
| "grad_norm": 1.3441901462036563, |
| "learning_rate": 0.00018044697563305876, |
| "loss": 12.2075, |
| "num_tokens": 2315254.0, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.2278127183787561, |
| "grad_norm": 1.1698154655629684, |
| "learning_rate": 0.00018031232545069468, |
| "loss": 12.3967, |
| "num_tokens": 2323116.0, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.22851153039832284, |
| "grad_norm": 1.2327000749220458, |
| "learning_rate": 0.00018017726383181925, |
| "loss": 12.4241, |
| "num_tokens": 2330812.0, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.2292103424178896, |
| "grad_norm": 1.1494561411366055, |
| "learning_rate": 0.0001800417914683471, |
| "loss": 12.3747, |
| "num_tokens": 2338451.0, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.22990915443745633, |
| "grad_norm": 1.035155095769357, |
| "learning_rate": 0.0001799059090542974, |
| "loss": 12.4331, |
| "num_tokens": 2346026.0, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.23060796645702306, |
| "grad_norm": 1.036281954431926, |
| "learning_rate": 0.00017976961728578963, |
| "loss": 12.2998, |
| "num_tokens": 2353605.0, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.23130677847658979, |
| "grad_norm": 0.9491052072879148, |
| "learning_rate": 0.00017963291686104053, |
| "loss": 12.2493, |
| "num_tokens": 2360509.0, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.23200559049615654, |
| "grad_norm": 1.0796880075257131, |
| "learning_rate": 0.00017949580848036046, |
| "loss": 12.4595, |
| "num_tokens": 2367035.0, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.23270440251572327, |
| "grad_norm": 1.084018095842082, |
| "learning_rate": 0.00017935829284614952, |
| "loss": 12.3212, |
| "num_tokens": 2373702.0, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.23340321453529, |
| "grad_norm": 1.1049395371201602, |
| "learning_rate": 0.00017922037066289432, |
| "loss": 12.4234, |
| "num_tokens": 2380174.0, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.23410202655485673, |
| "grad_norm": 1.0508677886350823, |
| "learning_rate": 0.0001790820426371641, |
| "loss": 12.3334, |
| "num_tokens": 2387052.0, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.2348008385744235, |
| "grad_norm": 1.0785045718409343, |
| "learning_rate": 0.00017894330947760726, |
| "loss": 12.1805, |
| "num_tokens": 2393866.0, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.23549965059399022, |
| "grad_norm": 0.9921298050576933, |
| "learning_rate": 0.0001788041718949477, |
| "loss": 12.2666, |
| "num_tokens": 2401040.0, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.23619846261355695, |
| "grad_norm": 1.0660998539990059, |
| "learning_rate": 0.00017866463060198115, |
| "loss": 12.4653, |
| "num_tokens": 2408627.0, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.23689727463312368, |
| "grad_norm": 0.995380832073994, |
| "learning_rate": 0.00017852468631357146, |
| "loss": 12.3646, |
| "num_tokens": 2415390.0, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.2375960866526904, |
| "grad_norm": 1.075956806450121, |
| "learning_rate": 0.00017838433974664712, |
| "loss": 12.221, |
| "num_tokens": 2422275.0, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.23829489867225717, |
| "grad_norm": 1.0139961269153555, |
| "learning_rate": 0.00017824359162019738, |
| "loss": 12.2838, |
| "num_tokens": 2429408.0, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.2389937106918239, |
| "grad_norm": 1.1457365908463537, |
| "learning_rate": 0.00017810244265526875, |
| "loss": 12.3507, |
| "num_tokens": 2436362.0, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.23969252271139063, |
| "grad_norm": 1.1147345258046557, |
| "learning_rate": 0.00017796089357496108, |
| "loss": 12.4508, |
| "num_tokens": 2442668.0, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.24039133473095736, |
| "grad_norm": 1.0930232074967394, |
| "learning_rate": 0.0001778189451044242, |
| "loss": 12.4069, |
| "num_tokens": 2449013.0, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.24109014675052412, |
| "grad_norm": 1.0438980257850956, |
| "learning_rate": 0.00017767659797085375, |
| "loss": 12.3327, |
| "num_tokens": 2455131.0, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.24178895877009085, |
| "grad_norm": 1.073359782419257, |
| "learning_rate": 0.000177533852903488, |
| "loss": 12.4483, |
| "num_tokens": 2462060.0, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.24248777078965758, |
| "grad_norm": 0.9628449472719576, |
| "learning_rate": 0.0001773907106336035, |
| "loss": 12.4128, |
| "num_tokens": 2469590.0, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.2431865828092243, |
| "grad_norm": 1.0691074828768792, |
| "learning_rate": 0.0001772471718945119, |
| "loss": 12.1702, |
| "num_tokens": 2476488.0, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.24388539482879107, |
| "grad_norm": 0.9952237415704821, |
| "learning_rate": 0.0001771032374215558, |
| "loss": 12.2478, |
| "num_tokens": 2484657.0, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.2445842068483578, |
| "grad_norm": 1.0036029456137867, |
| "learning_rate": 0.00017695890795210517, |
| "loss": 12.494, |
| "num_tokens": 2491489.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.24528301886792453, |
| "grad_norm": 1.1069581308411622, |
| "learning_rate": 0.00017681418422555356, |
| "loss": 12.4616, |
| "num_tokens": 2498631.0, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.24598183088749126, |
| "grad_norm": 0.9149619351800009, |
| "learning_rate": 0.00017666906698331428, |
| "loss": 12.0364, |
| "num_tokens": 2506017.0, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.246680642907058, |
| "grad_norm": 1.108978780022514, |
| "learning_rate": 0.00017652355696881652, |
| "loss": 12.3213, |
| "num_tokens": 2513168.0, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.24737945492662475, |
| "grad_norm": 1.0744303402500144, |
| "learning_rate": 0.0001763776549275017, |
| "loss": 12.2812, |
| "num_tokens": 2520301.0, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.24807826694619148, |
| "grad_norm": 1.1831634927540147, |
| "learning_rate": 0.00017623136160681963, |
| "loss": 12.369, |
| "num_tokens": 2527172.0, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.2487770789657582, |
| "grad_norm": 1.052250021391116, |
| "learning_rate": 0.00017608467775622445, |
| "loss": 12.425, |
| "num_tokens": 2534008.0, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.24947589098532494, |
| "grad_norm": 1.0564729282720158, |
| "learning_rate": 0.00017593760412717117, |
| "loss": 12.3877, |
| "num_tokens": 2541533.0, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.2501747030048917, |
| "grad_norm": 1.1910914093601987, |
| "learning_rate": 0.0001757901414731115, |
| "loss": 12.4473, |
| "num_tokens": 2548063.0, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.2508735150244584, |
| "grad_norm": 1.0694263327013265, |
| "learning_rate": 0.00017564229054949006, |
| "loss": 12.2255, |
| "num_tokens": 2555500.0, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.25157232704402516, |
| "grad_norm": 1.2542503960072244, |
| "learning_rate": 0.0001754940521137407, |
| "loss": 12.2771, |
| "num_tokens": 2562431.0, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.2522711390635919, |
| "grad_norm": 1.1158939880015024, |
| "learning_rate": 0.0001753454269252824, |
| "loss": 12.4299, |
| "num_tokens": 2569063.0, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.2529699510831586, |
| "grad_norm": 1.0409199378943457, |
| "learning_rate": 0.00017519641574551546, |
| "loss": 12.4908, |
| "num_tokens": 2576257.0, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.25366876310272535, |
| "grad_norm": 1.2189765723091746, |
| "learning_rate": 0.0001750470193378176, |
| "loss": 12.4082, |
| "num_tokens": 2582076.0, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.2543675751222921, |
| "grad_norm": 1.0384789257844729, |
| "learning_rate": 0.00017489723846754002, |
| "loss": 12.5168, |
| "num_tokens": 2588634.0, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.25506638714185886, |
| "grad_norm": 1.0719298770534975, |
| "learning_rate": 0.0001747470739020036, |
| "loss": 12.1278, |
| "num_tokens": 2595988.0, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.2557651991614256, |
| "grad_norm": 1.1294326548073568, |
| "learning_rate": 0.00017459652641049474, |
| "loss": 12.3269, |
| "num_tokens": 2602882.0, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.2564640111809923, |
| "grad_norm": 1.173317995313158, |
| "learning_rate": 0.0001744455967642616, |
| "loss": 12.2611, |
| "num_tokens": 2610428.0, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.25716282320055905, |
| "grad_norm": 1.0410607823903442, |
| "learning_rate": 0.00017429428573651024, |
| "loss": 12.0204, |
| "num_tokens": 2617338.0, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.2578616352201258, |
| "grad_norm": 1.1884465012383119, |
| "learning_rate": 0.00017414259410240026, |
| "loss": 12.226, |
| "num_tokens": 2624172.0, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.2585604472396925, |
| "grad_norm": 1.0099285124263113, |
| "learning_rate": 0.0001739905226390413, |
| "loss": 12.2136, |
| "num_tokens": 2631185.0, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.25925925925925924, |
| "grad_norm": 1.121309544418593, |
| "learning_rate": 0.0001738380721254888, |
| "loss": 12.349, |
| "num_tokens": 2638925.0, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.259958071278826, |
| "grad_norm": 1.0686531979747818, |
| "learning_rate": 0.00017368524334273998, |
| "loss": 12.2824, |
| "num_tokens": 2645812.0, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.2606568832983927, |
| "grad_norm": 1.0070897139519819, |
| "learning_rate": 0.00017353203707373, |
| "loss": 12.4482, |
| "num_tokens": 2652935.0, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.2613556953179595, |
| "grad_norm": 1.1557578122673882, |
| "learning_rate": 0.00017337845410332782, |
| "loss": 12.289, |
| "num_tokens": 2659882.0, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.2620545073375262, |
| "grad_norm": 1.0827441091133332, |
| "learning_rate": 0.0001732244952183323, |
| "loss": 12.2165, |
| "num_tokens": 2667013.0, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.26275331935709295, |
| "grad_norm": 0.9155364485477532, |
| "learning_rate": 0.000173070161207468, |
| "loss": 12.2606, |
| "num_tokens": 2675131.0, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.2634521313766597, |
| "grad_norm": 1.0579891464839146, |
| "learning_rate": 0.00017291545286138126, |
| "loss": 12.3224, |
| "num_tokens": 2681743.0, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.2641509433962264, |
| "grad_norm": 0.9495150325149874, |
| "learning_rate": 0.00017276037097263612, |
| "loss": 12.2971, |
| "num_tokens": 2688355.0, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.26484975541579314, |
| "grad_norm": 1.0942465116250388, |
| "learning_rate": 0.00017260491633571033, |
| "loss": 12.2816, |
| "num_tokens": 2695315.0, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.2655485674353599, |
| "grad_norm": 0.9236055818886025, |
| "learning_rate": 0.0001724490897469911, |
| "loss": 12.154, |
| "num_tokens": 2703258.0, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.2662473794549266, |
| "grad_norm": 1.0922583573424403, |
| "learning_rate": 0.00017229289200477123, |
| "loss": 12.1475, |
| "num_tokens": 2710326.0, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.2669461914744934, |
| "grad_norm": 0.9557123291968079, |
| "learning_rate": 0.00017213632390924486, |
| "loss": 12.2181, |
| "num_tokens": 2716825.0, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.2676450034940601, |
| "grad_norm": 1.0051833679431164, |
| "learning_rate": 0.00017197938626250348, |
| "loss": 12.265, |
| "num_tokens": 2723868.0, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.26834381551362685, |
| "grad_norm": 0.941677244029484, |
| "learning_rate": 0.00017182207986853176, |
| "loss": 12.3948, |
| "num_tokens": 2730711.0, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.2690426275331936, |
| "grad_norm": 1.0600159793133541, |
| "learning_rate": 0.00017166440553320337, |
| "loss": 12.1216, |
| "num_tokens": 2737540.0, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.2697414395527603, |
| "grad_norm": 1.0303448078848036, |
| "learning_rate": 0.0001715063640642771, |
| "loss": 12.2437, |
| "num_tokens": 2744595.0, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.27044025157232704, |
| "grad_norm": 1.0844807651487938, |
| "learning_rate": 0.00017134795627139236, |
| "loss": 12.324, |
| "num_tokens": 2751663.0, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.27113906359189377, |
| "grad_norm": 0.9321137382225717, |
| "learning_rate": 0.00017118918296606537, |
| "loss": 12.2311, |
| "num_tokens": 2759081.0, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.2718378756114605, |
| "grad_norm": 0.99797444758753, |
| "learning_rate": 0.00017103004496168473, |
| "loss": 12.2611, |
| "num_tokens": 2766249.0, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.27253668763102723, |
| "grad_norm": 1.0201352413756293, |
| "learning_rate": 0.0001708705430735075, |
| "loss": 12.3475, |
| "num_tokens": 2773320.0, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.273235499650594, |
| "grad_norm": 0.9230415417655204, |
| "learning_rate": 0.00017071067811865476, |
| "loss": 12.2765, |
| "num_tokens": 2780968.0, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.27393431167016075, |
| "grad_norm": 1.058464339332179, |
| "learning_rate": 0.0001705504509161077, |
| "loss": 12.2317, |
| "num_tokens": 2788319.0, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.2746331236897275, |
| "grad_norm": 0.935810740189098, |
| "learning_rate": 0.00017038986228670323, |
| "loss": 12.3109, |
| "num_tokens": 2796120.0, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.2753319357092942, |
| "grad_norm": 1.0048031215221804, |
| "learning_rate": 0.00017022891305312987, |
| "loss": 12.1342, |
| "num_tokens": 2803198.0, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.27603074772886094, |
| "grad_norm": 1.0859371888536073, |
| "learning_rate": 0.00017006760403992337, |
| "loss": 12.1707, |
| "num_tokens": 2809978.0, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.27672955974842767, |
| "grad_norm": 1.057763218829223, |
| "learning_rate": 0.00016990593607346276, |
| "loss": 12.1999, |
| "num_tokens": 2816932.0, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.2774283717679944, |
| "grad_norm": 0.9696692663519124, |
| "learning_rate": 0.00016974390998196595, |
| "loss": 12.1105, |
| "num_tokens": 2824857.0, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.2781271837875611, |
| "grad_norm": 1.0912831454596232, |
| "learning_rate": 0.00016958152659548548, |
| "loss": 12.161, |
| "num_tokens": 2832029.0, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.27882599580712786, |
| "grad_norm": 1.0272883066524952, |
| "learning_rate": 0.00016941878674590425, |
| "loss": 12.1335, |
| "num_tokens": 2839252.0, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.27952480782669464, |
| "grad_norm": 0.9719847308428423, |
| "learning_rate": 0.00016925569126693136, |
| "loss": 12.1187, |
| "num_tokens": 2847177.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.2802236198462614, |
| "grad_norm": 0.9720490101732504, |
| "learning_rate": 0.0001690922409940978, |
| "loss": 12.2379, |
| "num_tokens": 2854561.0, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.2809224318658281, |
| "grad_norm": 1.0410666937197113, |
| "learning_rate": 0.00016892843676475212, |
| "loss": 12.0956, |
| "num_tokens": 2861654.0, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.28162124388539483, |
| "grad_norm": 1.013809534538909, |
| "learning_rate": 0.00016876427941805622, |
| "loss": 12.1831, |
| "num_tokens": 2868359.0, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.28232005590496156, |
| "grad_norm": 0.8592336007443362, |
| "learning_rate": 0.00016859976979498092, |
| "loss": 12.1739, |
| "num_tokens": 2875713.0, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.2830188679245283, |
| "grad_norm": 0.9975671907769668, |
| "learning_rate": 0.00016843490873830178, |
| "loss": 12.1526, |
| "num_tokens": 2882164.0, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.283717679944095, |
| "grad_norm": 1.046642053431646, |
| "learning_rate": 0.00016826969709259477, |
| "loss": 12.4811, |
| "num_tokens": 2888466.0, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.28441649196366176, |
| "grad_norm": 0.9951638046201554, |
| "learning_rate": 0.0001681041357042319, |
| "loss": 12.1391, |
| "num_tokens": 2896054.0, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.2851153039832285, |
| "grad_norm": 1.0120154348914012, |
| "learning_rate": 0.0001679382254213768, |
| "loss": 12.2303, |
| "num_tokens": 2903381.0, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.28581411600279527, |
| "grad_norm": 0.971103244517684, |
| "learning_rate": 0.00016777196709398065, |
| "loss": 12.3491, |
| "num_tokens": 2910764.0, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.286512928022362, |
| "grad_norm": 1.008656433224794, |
| "learning_rate": 0.00016760536157377754, |
| "loss": 12.1785, |
| "num_tokens": 2917852.0, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.28721174004192873, |
| "grad_norm": 1.0230579216157252, |
| "learning_rate": 0.00016743840971428017, |
| "loss": 12.0658, |
| "num_tokens": 2925426.0, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.28791055206149546, |
| "grad_norm": 1.020310009862771, |
| "learning_rate": 0.00016727111237077559, |
| "loss": 12.223, |
| "num_tokens": 2932534.0, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.2886093640810622, |
| "grad_norm": 0.9125469963918674, |
| "learning_rate": 0.00016710347040032076, |
| "loss": 12.2343, |
| "num_tokens": 2939544.0, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.2893081761006289, |
| "grad_norm": 1.152646459857617, |
| "learning_rate": 0.0001669354846617381, |
| "loss": 12.1026, |
| "num_tokens": 2946773.0, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.29000698812019565, |
| "grad_norm": 0.9487006632604599, |
| "learning_rate": 0.00016676715601561117, |
| "loss": 12.2828, |
| "num_tokens": 2954251.0, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.2907058001397624, |
| "grad_norm": 1.0339989063419985, |
| "learning_rate": 0.00016659848532428023, |
| "loss": 12.4354, |
| "num_tokens": 2961656.0, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.2914046121593291, |
| "grad_norm": 0.9810692884344294, |
| "learning_rate": 0.00016642947345183774, |
| "loss": 12.2144, |
| "num_tokens": 2969059.0, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.2921034241788959, |
| "grad_norm": 1.041382797288484, |
| "learning_rate": 0.0001662601212641242, |
| "loss": 12.1039, |
| "num_tokens": 2977234.0, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.29280223619846263, |
| "grad_norm": 0.9394420025529445, |
| "learning_rate": 0.00016609042962872333, |
| "loss": 12.1398, |
| "num_tokens": 2984854.0, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.29350104821802936, |
| "grad_norm": 1.0586381513052323, |
| "learning_rate": 0.00016592039941495804, |
| "loss": 11.9381, |
| "num_tokens": 2991537.0, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.2941998602375961, |
| "grad_norm": 0.9975034902351064, |
| "learning_rate": 0.00016575003149388548, |
| "loss": 12.1831, |
| "num_tokens": 2998088.0, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.2948986722571628, |
| "grad_norm": 1.0160628218220324, |
| "learning_rate": 0.00016557932673829311, |
| "loss": 12.2999, |
| "num_tokens": 3004876.0, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.29559748427672955, |
| "grad_norm": 1.093919382310838, |
| "learning_rate": 0.0001654082860226939, |
| "loss": 12.1683, |
| "num_tokens": 3011520.0, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.2962962962962963, |
| "grad_norm": 0.9568683146100065, |
| "learning_rate": 0.00016523691022332185, |
| "loss": 12.2034, |
| "num_tokens": 3018294.0, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.296995108315863, |
| "grad_norm": 1.133923086552664, |
| "learning_rate": 0.00016506520021812766, |
| "loss": 11.8777, |
| "num_tokens": 3026301.0, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.2976939203354298, |
| "grad_norm": 0.9273115968347424, |
| "learning_rate": 0.00016489315688677416, |
| "loss": 12.1574, |
| "num_tokens": 3034166.0, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.2983927323549965, |
| "grad_norm": 0.8779205899305784, |
| "learning_rate": 0.00016472078111063175, |
| "loss": 12.0526, |
| "num_tokens": 3041930.0, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.29909154437456326, |
| "grad_norm": 1.0184105137011932, |
| "learning_rate": 0.00016454807377277398, |
| "loss": 12.2997, |
| "num_tokens": 3048767.0, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.29979035639413, |
| "grad_norm": 0.961478703601268, |
| "learning_rate": 0.00016437503575797297, |
| "loss": 12.0761, |
| "num_tokens": 3055753.0, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.3004891684136967, |
| "grad_norm": 0.9946551332938962, |
| "learning_rate": 0.00016420166795269475, |
| "loss": 12.3698, |
| "num_tokens": 3063120.0, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.30118798043326345, |
| "grad_norm": 0.9437621585892487, |
| "learning_rate": 0.00016402797124509508, |
| "loss": 12.235, |
| "num_tokens": 3070742.0, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.3018867924528302, |
| "grad_norm": 1.0369518319538322, |
| "learning_rate": 0.00016385394652501445, |
| "loss": 12.1428, |
| "num_tokens": 3078056.0, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.3025856044723969, |
| "grad_norm": 0.9296598663042578, |
| "learning_rate": 0.00016367959468397393, |
| "loss": 12.1666, |
| "num_tokens": 3085214.0, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.30328441649196364, |
| "grad_norm": 0.9998913789930738, |
| "learning_rate": 0.00016350491661517032, |
| "loss": 12.199, |
| "num_tokens": 3092548.0, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.3039832285115304, |
| "grad_norm": 0.9317649079361281, |
| "learning_rate": 0.00016332991321347167, |
| "loss": 12.0497, |
| "num_tokens": 3100393.0, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.30468204053109715, |
| "grad_norm": 1.0873578371369284, |
| "learning_rate": 0.0001631545853754127, |
| "loss": 12.396, |
| "num_tokens": 3106325.0, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.3053808525506639, |
| "grad_norm": 1.111865839622246, |
| "learning_rate": 0.0001629789339991902, |
| "loss": 12.2868, |
| "num_tokens": 3113193.0, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.3060796645702306, |
| "grad_norm": 1.0625922180498937, |
| "learning_rate": 0.0001628029599846585, |
| "loss": 12.2205, |
| "num_tokens": 3120358.0, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.30677847658979734, |
| "grad_norm": 1.0810749864557256, |
| "learning_rate": 0.00016262666423332473, |
| "loss": 12.118, |
| "num_tokens": 3127109.0, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.3074772886093641, |
| "grad_norm": 1.0039089118126248, |
| "learning_rate": 0.00016245004764834422, |
| "loss": 11.9709, |
| "num_tokens": 3134426.0, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.3081761006289308, |
| "grad_norm": 0.9760746122171329, |
| "learning_rate": 0.000162273111134516, |
| "loss": 12.2341, |
| "num_tokens": 3141188.0, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.30887491264849753, |
| "grad_norm": 1.0221181358646754, |
| "learning_rate": 0.00016209585559827806, |
| "loss": 12.1916, |
| "num_tokens": 3149034.0, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.30957372466806427, |
| "grad_norm": 0.9632877916386517, |
| "learning_rate": 0.0001619182819477027, |
| "loss": 12.3057, |
| "num_tokens": 3156062.0, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.31027253668763105, |
| "grad_norm": 1.0519770843397107, |
| "learning_rate": 0.0001617403910924919, |
| "loss": 12.0121, |
| "num_tokens": 3164141.0, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.3109713487071978, |
| "grad_norm": 0.9570699017638297, |
| "learning_rate": 0.00016156218394397273, |
| "loss": 12.1094, |
| "num_tokens": 3172103.0, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.3116701607267645, |
| "grad_norm": 1.078199682948591, |
| "learning_rate": 0.0001613836614150926, |
| "loss": 12.2273, |
| "num_tokens": 3178950.0, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.31236897274633124, |
| "grad_norm": 0.9297993503765057, |
| "learning_rate": 0.00016120482442041447, |
| "loss": 12.0441, |
| "num_tokens": 3186094.0, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.31306778476589797, |
| "grad_norm": 1.052582853013476, |
| "learning_rate": 0.0001610256738761125, |
| "loss": 12.276, |
| "num_tokens": 3193281.0, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.3137665967854647, |
| "grad_norm": 0.9525142248097075, |
| "learning_rate": 0.000160846210699967, |
| "loss": 12.089, |
| "num_tokens": 3200265.0, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.31446540880503143, |
| "grad_norm": 0.9973983515962607, |
| "learning_rate": 0.0001606664358113599, |
| "loss": 12.0084, |
| "num_tokens": 3207548.0, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.31516422082459816, |
| "grad_norm": 0.9695049891418782, |
| "learning_rate": 0.00016048635013127016, |
| "loss": 12.2589, |
| "num_tokens": 3214372.0, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.3158630328441649, |
| "grad_norm": 1.1601028592283664, |
| "learning_rate": 0.00016030595458226872, |
| "loss": 11.9678, |
| "num_tokens": 3221203.0, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.3165618448637317, |
| "grad_norm": 0.9617505431282883, |
| "learning_rate": 0.00016012525008851403, |
| "loss": 12.0657, |
| "num_tokens": 3229249.0, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.3172606568832984, |
| "grad_norm": 1.1766775349257634, |
| "learning_rate": 0.0001599442375757473, |
| "loss": 12.1183, |
| "num_tokens": 3236472.0, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.31795946890286514, |
| "grad_norm": 0.9825393901516871, |
| "learning_rate": 0.00015976291797128767, |
| "loss": 12.1128, |
| "num_tokens": 3243897.0, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.31865828092243187, |
| "grad_norm": 1.115098077594373, |
| "learning_rate": 0.00015958129220402744, |
| "loss": 11.989, |
| "num_tokens": 3250533.0, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.3193570929419986, |
| "grad_norm": 1.112270034985869, |
| "learning_rate": 0.00015939936120442752, |
| "loss": 12.2418, |
| "num_tokens": 3257466.0, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.32005590496156533, |
| "grad_norm": 1.0490822017700396, |
| "learning_rate": 0.00015921712590451236, |
| "loss": 12.0436, |
| "num_tokens": 3263944.0, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.32075471698113206, |
| "grad_norm": 1.19266057918563, |
| "learning_rate": 0.00015903458723786544, |
| "loss": 12.3036, |
| "num_tokens": 3270543.0, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.3214535290006988, |
| "grad_norm": 0.9754576421413539, |
| "learning_rate": 0.00015885174613962426, |
| "loss": 12.2576, |
| "num_tokens": 3277894.0, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.3221523410202656, |
| "grad_norm": 1.2240651695002909, |
| "learning_rate": 0.00015866860354647576, |
| "loss": 12.1746, |
| "num_tokens": 3284793.0, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.3228511530398323, |
| "grad_norm": 0.9220005514021452, |
| "learning_rate": 0.00015848516039665138, |
| "loss": 12.1768, |
| "num_tokens": 3292856.0, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.32354996505939904, |
| "grad_norm": 1.1405752061575267, |
| "learning_rate": 0.0001583014176299223, |
| "loss": 11.9588, |
| "num_tokens": 3300134.0, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.32424877707896577, |
| "grad_norm": 0.9785862767305678, |
| "learning_rate": 0.00015811737618759468, |
| "loss": 12.0556, |
| "num_tokens": 3307091.0, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.3249475890985325, |
| "grad_norm": 1.1239069485026847, |
| "learning_rate": 0.00015793303701250468, |
| "loss": 11.9657, |
| "num_tokens": 3314380.0, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.3256464011180992, |
| "grad_norm": 0.9230619844697087, |
| "learning_rate": 0.00015774840104901378, |
| "loss": 12.4159, |
| "num_tokens": 3322081.0, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.32634521313766596, |
| "grad_norm": 1.045958434033744, |
| "learning_rate": 0.000157563469243004, |
| "loss": 12.2218, |
| "num_tokens": 3329105.0, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.3270440251572327, |
| "grad_norm": 1.0386841546884902, |
| "learning_rate": 0.00015737824254187275, |
| "loss": 12.1626, |
| "num_tokens": 3336405.0, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.3277428371767994, |
| "grad_norm": 1.0497667874202588, |
| "learning_rate": 0.00015719272189452824, |
| "loss": 12.0446, |
| "num_tokens": 3343087.0, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.3284416491963662, |
| "grad_norm": 0.9471293053870683, |
| "learning_rate": 0.00015700690825138473, |
| "loss": 12.2638, |
| "num_tokens": 3350235.0, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.32914046121593293, |
| "grad_norm": 0.9131515390790194, |
| "learning_rate": 0.00015682080256435724, |
| "loss": 12.1216, |
| "num_tokens": 3357485.0, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.32983927323549966, |
| "grad_norm": 0.9820518461252037, |
| "learning_rate": 0.00015663440578685703, |
| "loss": 12.2044, |
| "num_tokens": 3364351.0, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.3305380852550664, |
| "grad_norm": 0.8992691022961555, |
| "learning_rate": 0.00015644771887378663, |
| "loss": 12.1364, |
| "num_tokens": 3372249.0, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.3312368972746331, |
| "grad_norm": 0.9980561034084768, |
| "learning_rate": 0.00015626074278153485, |
| "loss": 12.1517, |
| "num_tokens": 3379136.0, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.33193570929419985, |
| "grad_norm": 0.9243681679240103, |
| "learning_rate": 0.000156073478467972, |
| "loss": 12.0981, |
| "num_tokens": 3386556.0, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.3326345213137666, |
| "grad_norm": 0.9256289781367883, |
| "learning_rate": 0.0001558859268924449, |
| "loss": 11.8951, |
| "num_tokens": 3394436.0, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.3333333333333333, |
| "grad_norm": 0.9900816750625112, |
| "learning_rate": 0.0001556980890157721, |
| "loss": 12.029, |
| "num_tokens": 3401842.0, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.33403214535290005, |
| "grad_norm": 0.9703349089213797, |
| "learning_rate": 0.00015550996580023868, |
| "loss": 11.9128, |
| "num_tokens": 3409180.0, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.33473095737246683, |
| "grad_norm": 0.9630292571815687, |
| "learning_rate": 0.00015532155820959165, |
| "loss": 12.2213, |
| "num_tokens": 3416900.0, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.33542976939203356, |
| "grad_norm": 0.9051616995348135, |
| "learning_rate": 0.00015513286720903485, |
| "loss": 12.2121, |
| "num_tokens": 3424074.0, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.3361285814116003, |
| "grad_norm": 0.9690008678265436, |
| "learning_rate": 0.00015494389376522388, |
| "loss": 12.2515, |
| "num_tokens": 3431040.0, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.336827393431167, |
| "grad_norm": 1.0166496850552176, |
| "learning_rate": 0.0001547546388462615, |
| "loss": 12.1705, |
| "num_tokens": 3437949.0, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.33752620545073375, |
| "grad_norm": 0.9396624167132782, |
| "learning_rate": 0.00015456510342169225, |
| "loss": 11.8757, |
| "num_tokens": 3444767.0, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.3382250174703005, |
| "grad_norm": 0.9809819891433745, |
| "learning_rate": 0.00015437528846249784, |
| "loss": 12.1112, |
| "num_tokens": 3452455.0, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.3389238294898672, |
| "grad_norm": 0.9286838895806877, |
| "learning_rate": 0.00015418519494109185, |
| "loss": 12.0485, |
| "num_tokens": 3459075.0, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.33962264150943394, |
| "grad_norm": 0.9139552415362666, |
| "learning_rate": 0.00015399482383131517, |
| "loss": 12.0818, |
| "num_tokens": 3466194.0, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.3403214535290007, |
| "grad_norm": 0.9304856467336878, |
| "learning_rate": 0.0001538041761084305, |
| "loss": 12.1795, |
| "num_tokens": 3472861.0, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.34102026554856746, |
| "grad_norm": 0.9208992780595067, |
| "learning_rate": 0.00015361325274911779, |
| "loss": 12.0489, |
| "num_tokens": 3480127.0, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.3417190775681342, |
| "grad_norm": 0.8997956372118784, |
| "learning_rate": 0.00015342205473146904, |
| "loss": 12.1201, |
| "num_tokens": 3487155.0, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.3424178895877009, |
| "grad_norm": 0.9777088108576337, |
| "learning_rate": 0.00015323058303498324, |
| "loss": 12.0205, |
| "num_tokens": 3494773.0, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.34311670160726765, |
| "grad_norm": 0.879745446242411, |
| "learning_rate": 0.00015303883864056154, |
| "loss": 11.9763, |
| "num_tokens": 3502169.0, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.3438155136268344, |
| "grad_norm": 0.9329909041798102, |
| "learning_rate": 0.00015284682253050198, |
| "loss": 12.3791, |
| "num_tokens": 3509575.0, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.3445143256464011, |
| "grad_norm": 0.9682786444676359, |
| "learning_rate": 0.00015265453568849463, |
| "loss": 12.4002, |
| "num_tokens": 3516257.0, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.34521313766596784, |
| "grad_norm": 0.8882073075846929, |
| "learning_rate": 0.0001524619790996166, |
| "loss": 12.2699, |
| "num_tokens": 3523550.0, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.34591194968553457, |
| "grad_norm": 0.9501809761434933, |
| "learning_rate": 0.00015226915375032675, |
| "loss": 12.0807, |
| "num_tokens": 3530386.0, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.3466107617051013, |
| "grad_norm": 0.8935380130451016, |
| "learning_rate": 0.00015207606062846092, |
| "loss": 12.0567, |
| "num_tokens": 3537432.0, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.3473095737246681, |
| "grad_norm": 0.9513080986369654, |
| "learning_rate": 0.00015188270072322664, |
| "loss": 12.1846, |
| "num_tokens": 3544343.0, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.3480083857442348, |
| "grad_norm": 0.9311057870656899, |
| "learning_rate": 0.00015168907502519823, |
| "loss": 12.0444, |
| "num_tokens": 3551437.0, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.34870719776380155, |
| "grad_norm": 1.0017415906580418, |
| "learning_rate": 0.00015149518452631163, |
| "loss": 12.1657, |
| "num_tokens": 3558381.0, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.3494060097833683, |
| "grad_norm": 1.0960810177794507, |
| "learning_rate": 0.00015130103021985928, |
| "loss": 12.0332, |
| "num_tokens": 3565537.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.350104821802935, |
| "grad_norm": 0.9446673820098566, |
| "learning_rate": 0.00015110661310048523, |
| "loss": 12.3164, |
| "num_tokens": 3572664.0, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.35080363382250174, |
| "grad_norm": 1.00955257241026, |
| "learning_rate": 0.00015091193416417981, |
| "loss": 11.9365, |
| "num_tokens": 3579863.0, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.35150244584206847, |
| "grad_norm": 0.9056375701048143, |
| "learning_rate": 0.00015071699440827462, |
| "loss": 12.1319, |
| "num_tokens": 3587973.0, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.3522012578616352, |
| "grad_norm": 0.9234279886277251, |
| "learning_rate": 0.00015052179483143752, |
| "loss": 12.0822, |
| "num_tokens": 3594430.0, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.352900069881202, |
| "grad_norm": 0.9306165895829293, |
| "learning_rate": 0.00015032633643366727, |
| "loss": 11.9546, |
| "num_tokens": 3600877.0, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.3535988819007687, |
| "grad_norm": 0.916209027167804, |
| "learning_rate": 0.0001501306202162887, |
| "loss": 12.0369, |
| "num_tokens": 3607903.0, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.35429769392033544, |
| "grad_norm": 0.8278388004090699, |
| "learning_rate": 0.0001499346471819474, |
| "loss": 12.0787, |
| "num_tokens": 3614967.0, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.3549965059399022, |
| "grad_norm": 0.8626281789322203, |
| "learning_rate": 0.00014973841833460457, |
| "loss": 12.158, |
| "num_tokens": 3622644.0, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.3556953179594689, |
| "grad_norm": 1.0169394719056057, |
| "learning_rate": 0.00014954193467953196, |
| "loss": 12.0281, |
| "num_tokens": 3629496.0, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.35639412997903563, |
| "grad_norm": 1.0391735155244062, |
| "learning_rate": 0.0001493451972233067, |
| "loss": 12.1707, |
| "num_tokens": 3636251.0, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.35709294199860236, |
| "grad_norm": 0.8580636790874151, |
| "learning_rate": 0.0001491482069738062, |
| "loss": 12.0772, |
| "num_tokens": 3643640.0, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.3577917540181691, |
| "grad_norm": 0.8826702226607871, |
| "learning_rate": 0.00014895096494020274, |
| "loss": 12.0327, |
| "num_tokens": 3651194.0, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.3584905660377358, |
| "grad_norm": 0.8613455261520183, |
| "learning_rate": 0.00014875347213295863, |
| "loss": 11.9211, |
| "num_tokens": 3658441.0, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.3591893780573026, |
| "grad_norm": 0.9064281924248683, |
| "learning_rate": 0.00014855572956382082, |
| "loss": 12.0293, |
| "num_tokens": 3665170.0, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.35988819007686934, |
| "grad_norm": 0.8367714960887371, |
| "learning_rate": 0.0001483577382458158, |
| "loss": 12.0457, |
| "num_tokens": 3672474.0, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.36058700209643607, |
| "grad_norm": 0.9241907055749421, |
| "learning_rate": 0.00014815949919324444, |
| "loss": 11.9374, |
| "num_tokens": 3679386.0, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.3612858141160028, |
| "grad_norm": 0.824152296178085, |
| "learning_rate": 0.00014796101342167664, |
| "loss": 12.1566, |
| "num_tokens": 3687308.0, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.36198462613556953, |
| "grad_norm": 0.8944396761994626, |
| "learning_rate": 0.00014776228194794623, |
| "loss": 11.9896, |
| "num_tokens": 3694073.0, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.36268343815513626, |
| "grad_norm": 0.858271930480037, |
| "learning_rate": 0.00014756330579014591, |
| "loss": 12.1352, |
| "num_tokens": 3701387.0, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.363382250174703, |
| "grad_norm": 1.154248735078944, |
| "learning_rate": 0.0001473640859676217, |
| "loss": 11.858, |
| "num_tokens": 3708600.0, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.3640810621942697, |
| "grad_norm": 0.9591101732020725, |
| "learning_rate": 0.00014716462350096803, |
| "loss": 12.0315, |
| "num_tokens": 3716979.0, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.36477987421383645, |
| "grad_norm": 1.17298705448662, |
| "learning_rate": 0.0001469649194120224, |
| "loss": 12.0713, |
| "num_tokens": 3723012.0, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.36547868623340324, |
| "grad_norm": 0.9650929504820438, |
| "learning_rate": 0.00014676497472385994, |
| "loss": 11.8546, |
| "num_tokens": 3730363.0, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.36617749825296997, |
| "grad_norm": 1.0598732657131429, |
| "learning_rate": 0.0001465647904607886, |
| "loss": 12.0821, |
| "num_tokens": 3737054.0, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.3668763102725367, |
| "grad_norm": 1.055588049526762, |
| "learning_rate": 0.00014636436764834353, |
| "loss": 11.9764, |
| "num_tokens": 3743973.0, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.36757512229210343, |
| "grad_norm": 0.9634047864594488, |
| "learning_rate": 0.000146163707313282, |
| "loss": 12.2426, |
| "num_tokens": 3750646.0, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.36827393431167016, |
| "grad_norm": 0.9379570298735684, |
| "learning_rate": 0.00014596281048357806, |
| "loss": 12.0747, |
| "num_tokens": 3758294.0, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.3689727463312369, |
| "grad_norm": 0.8475926897769003, |
| "learning_rate": 0.0001457616781884173, |
| "loss": 11.9549, |
| "num_tokens": 3765428.0, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.3696715583508036, |
| "grad_norm": 0.9785101548217052, |
| "learning_rate": 0.00014556031145819168, |
| "loss": 12.0683, |
| "num_tokens": 3772483.0, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.37037037037037035, |
| "grad_norm": 0.8657231796104006, |
| "learning_rate": 0.0001453587113244941, |
| "loss": 12.0748, |
| "num_tokens": 3779599.0, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.3710691823899371, |
| "grad_norm": 0.9948561473315793, |
| "learning_rate": 0.00014515687882011313, |
| "loss": 12.0452, |
| "num_tokens": 3786663.0, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.37176799440950387, |
| "grad_norm": 0.8961368986877242, |
| "learning_rate": 0.00014495481497902788, |
| "loss": 12.0502, |
| "num_tokens": 3793917.0, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.3724668064290706, |
| "grad_norm": 1.0039214702487007, |
| "learning_rate": 0.00014475252083640246, |
| "loss": 12.1103, |
| "num_tokens": 3800383.0, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.3731656184486373, |
| "grad_norm": 1.104043842808782, |
| "learning_rate": 0.00014454999742858092, |
| "loss": 12.0706, |
| "num_tokens": 3807421.0, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.37386443046820406, |
| "grad_norm": 0.9706944725637485, |
| "learning_rate": 0.0001443472457930817, |
| "loss": 12.1253, |
| "num_tokens": 3814470.0, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.3745632424877708, |
| "grad_norm": 1.0708871567151401, |
| "learning_rate": 0.0001441442669685926, |
| "loss": 12.175, |
| "num_tokens": 3821117.0, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.3752620545073375, |
| "grad_norm": 0.9202934515725604, |
| "learning_rate": 0.00014394106199496517, |
| "loss": 12.2001, |
| "num_tokens": 3828050.0, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.37596086652690425, |
| "grad_norm": 0.887859566885173, |
| "learning_rate": 0.00014373763191320954, |
| "loss": 12.1075, |
| "num_tokens": 3835858.0, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.376659678546471, |
| "grad_norm": 0.9187757022639819, |
| "learning_rate": 0.00014353397776548912, |
| "loss": 12.0908, |
| "num_tokens": 3843141.0, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.37735849056603776, |
| "grad_norm": 0.9155332566153289, |
| "learning_rate": 0.00014333010059511505, |
| "loss": 12.1828, |
| "num_tokens": 3850497.0, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.3780573025856045, |
| "grad_norm": 0.8284408148826566, |
| "learning_rate": 0.0001431260014465412, |
| "loss": 11.6563, |
| "num_tokens": 3857621.0, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.3787561146051712, |
| "grad_norm": 0.9716124370861428, |
| "learning_rate": 0.00014292168136535854, |
| "loss": 12.0896, |
| "num_tokens": 3864403.0, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.37945492662473795, |
| "grad_norm": 0.8208484050542351, |
| "learning_rate": 0.00014271714139828983, |
| "loss": 11.8759, |
| "num_tokens": 3871744.0, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.3801537386443047, |
| "grad_norm": 0.9605228173140264, |
| "learning_rate": 0.0001425123825931843, |
| "loss": 11.994, |
| "num_tokens": 3879289.0, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.3808525506638714, |
| "grad_norm": 0.9031441753054636, |
| "learning_rate": 0.00014230740599901231, |
| "loss": 11.8901, |
| "num_tokens": 3886810.0, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.38155136268343814, |
| "grad_norm": 0.907168225729437, |
| "learning_rate": 0.00014210221266585998, |
| "loss": 12.1076, |
| "num_tokens": 3894520.0, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.3822501747030049, |
| "grad_norm": 0.9280949877436617, |
| "learning_rate": 0.0001418968036449237, |
| "loss": 12.0276, |
| "num_tokens": 3901356.0, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.3829489867225716, |
| "grad_norm": 0.8801303118077374, |
| "learning_rate": 0.0001416911799885049, |
| "loss": 12.0104, |
| "num_tokens": 3908231.0, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.3836477987421384, |
| "grad_norm": 0.8825862875085414, |
| "learning_rate": 0.00014148534275000444, |
| "loss": 12.2931, |
| "num_tokens": 3915548.0, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.3843466107617051, |
| "grad_norm": 0.9636172523939012, |
| "learning_rate": 0.0001412792929839175, |
| "loss": 11.9835, |
| "num_tokens": 3922294.0, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.38504542278127185, |
| "grad_norm": 0.8577046886760447, |
| "learning_rate": 0.00014107303174582794, |
| "loss": 11.95, |
| "num_tokens": 3929145.0, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.3857442348008386, |
| "grad_norm": 0.9191735227445484, |
| "learning_rate": 0.00014086656009240306, |
| "loss": 12.0281, |
| "num_tokens": 3936061.0, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.3864430468204053, |
| "grad_norm": 0.9127134286950958, |
| "learning_rate": 0.00014065987908138804, |
| "loss": 12.1853, |
| "num_tokens": 3942508.0, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.38714185883997204, |
| "grad_norm": 0.9287174901506331, |
| "learning_rate": 0.00014045298977160057, |
| "loss": 12.189, |
| "num_tokens": 3949004.0, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.38784067085953877, |
| "grad_norm": 0.8083833012878675, |
| "learning_rate": 0.00014024589322292555, |
| "loss": 11.9526, |
| "num_tokens": 3956901.0, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.3885394828791055, |
| "grad_norm": 0.937342232458656, |
| "learning_rate": 0.00014003859049630942, |
| "loss": 12.1259, |
| "num_tokens": 3964133.0, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.38923829489867223, |
| "grad_norm": 0.8744910335409934, |
| "learning_rate": 0.000139831082653755, |
| "loss": 12.2569, |
| "num_tokens": 3971294.0, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.389937106918239, |
| "grad_norm": 0.9325435804798502, |
| "learning_rate": 0.00013962337075831583, |
| "loss": 12.0983, |
| "num_tokens": 3977426.0, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.39063591893780575, |
| "grad_norm": 0.922661936419083, |
| "learning_rate": 0.00013941545587409075, |
| "loss": 11.9229, |
| "num_tokens": 3984492.0, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.3913347309573725, |
| "grad_norm": 0.9600683664822316, |
| "learning_rate": 0.00013920733906621862, |
| "loss": 12.1813, |
| "num_tokens": 3991120.0, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.3920335429769392, |
| "grad_norm": 0.9723044506449994, |
| "learning_rate": 0.00013899902140087272, |
| "loss": 12.1032, |
| "num_tokens": 3997971.0, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.39273235499650594, |
| "grad_norm": 0.8414594620114867, |
| "learning_rate": 0.00013879050394525523, |
| "loss": 11.9443, |
| "num_tokens": 4005376.0, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.39343116701607267, |
| "grad_norm": 1.0202475102587107, |
| "learning_rate": 0.00013858178776759197, |
| "loss": 12.1481, |
| "num_tokens": 4012471.0, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.3941299790356394, |
| "grad_norm": 0.8839072059833434, |
| "learning_rate": 0.00013837287393712666, |
| "loss": 12.0321, |
| "num_tokens": 4019054.0, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.39482879105520613, |
| "grad_norm": 0.9555850931145216, |
| "learning_rate": 0.00013816376352411574, |
| "loss": 12.1077, |
| "num_tokens": 4025454.0, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.39552760307477286, |
| "grad_norm": 0.8679296924666821, |
| "learning_rate": 0.00013795445759982262, |
| "loss": 11.7888, |
| "num_tokens": 4032992.0, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.39622641509433965, |
| "grad_norm": 0.8722743434457823, |
| "learning_rate": 0.00013774495723651236, |
| "loss": 12.024, |
| "num_tokens": 4040064.0, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.3969252271139064, |
| "grad_norm": 0.9600855639345066, |
| "learning_rate": 0.0001375352635074461, |
| "loss": 11.8821, |
| "num_tokens": 4046967.0, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.3976240391334731, |
| "grad_norm": 0.7877602768482204, |
| "learning_rate": 0.0001373253774868756, |
| "loss": 12.0148, |
| "num_tokens": 4055301.0, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.39832285115303984, |
| "grad_norm": 0.9074208297899601, |
| "learning_rate": 0.00013711530025003766, |
| "loss": 12.0443, |
| "num_tokens": 4062431.0, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.39902166317260657, |
| "grad_norm": 0.8287729757656834, |
| "learning_rate": 0.00013690503287314883, |
| "loss": 11.8087, |
| "num_tokens": 4070230.0, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.3997204751921733, |
| "grad_norm": 0.9316554946864943, |
| "learning_rate": 0.00013669457643339955, |
| "loss": 11.8952, |
| "num_tokens": 4077148.0, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.40041928721174, |
| "grad_norm": 0.8817289397600743, |
| "learning_rate": 0.00013648393200894893, |
| "loss": 11.8121, |
| "num_tokens": 4084168.0, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.40111809923130676, |
| "grad_norm": 0.8430911631310684, |
| "learning_rate": 0.00013627310067891913, |
| "loss": 12.0071, |
| "num_tokens": 4092313.0, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.4018169112508735, |
| "grad_norm": 0.8381632302819609, |
| "learning_rate": 0.00013606208352338973, |
| "loss": 11.8803, |
| "num_tokens": 4099578.0, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.4025157232704403, |
| "grad_norm": 0.8404927564721982, |
| "learning_rate": 0.00013585088162339231, |
| "loss": 12.0778, |
| "num_tokens": 4106534.0, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.403214535290007, |
| "grad_norm": 0.9183215666016679, |
| "learning_rate": 0.00013563949606090503, |
| "loss": 11.9381, |
| "num_tokens": 4113134.0, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.40391334730957373, |
| "grad_norm": 0.8298708376682066, |
| "learning_rate": 0.00013542792791884674, |
| "loss": 11.9096, |
| "num_tokens": 4120863.0, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.40461215932914046, |
| "grad_norm": 0.8725360483372107, |
| "learning_rate": 0.00013521617828107175, |
| "loss": 11.8436, |
| "num_tokens": 4127780.0, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.4053109713487072, |
| "grad_norm": 0.8961375996287267, |
| "learning_rate": 0.00013500424823236412, |
| "loss": 11.95, |
| "num_tokens": 4134829.0, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.4060097833682739, |
| "grad_norm": 0.7861779536888713, |
| "learning_rate": 0.0001347921388584322, |
| "loss": 11.7022, |
| "num_tokens": 4142834.0, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.40670859538784065, |
| "grad_norm": 0.8496401776165243, |
| "learning_rate": 0.000134579851245903, |
| "loss": 12.1181, |
| "num_tokens": 4149774.0, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.4074074074074074, |
| "grad_norm": 0.8468938520035636, |
| "learning_rate": 0.00013436738648231656, |
| "loss": 12.0059, |
| "num_tokens": 4156412.0, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.40810621942697417, |
| "grad_norm": 0.8790337397038394, |
| "learning_rate": 0.00013415474565612058, |
| "loss": 12.1166, |
| "num_tokens": 4163132.0, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.4088050314465409, |
| "grad_norm": 0.8589311219069107, |
| "learning_rate": 0.00013394192985666465, |
| "loss": 12.0116, |
| "num_tokens": 4170145.0, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.40950384346610763, |
| "grad_norm": 0.8696135886818056, |
| "learning_rate": 0.0001337289401741947, |
| "loss": 11.9189, |
| "num_tokens": 4177826.0, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.41020265548567436, |
| "grad_norm": 1.2530218930273918, |
| "learning_rate": 0.0001335157776998476, |
| "loss": 11.873, |
| "num_tokens": 4184556.0, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.4109014675052411, |
| "grad_norm": 1.017543600554075, |
| "learning_rate": 0.00013330244352564527, |
| "loss": 11.9424, |
| "num_tokens": 4191042.0, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.4116002795248078, |
| "grad_norm": 0.9275163950537736, |
| "learning_rate": 0.0001330889387444893, |
| "loss": 12.1217, |
| "num_tokens": 4197862.0, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.41229909154437455, |
| "grad_norm": 0.8853093500307286, |
| "learning_rate": 0.00013287526445015531, |
| "loss": 12.0612, |
| "num_tokens": 4205404.0, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.4129979035639413, |
| "grad_norm": 0.8464143780556126, |
| "learning_rate": 0.0001326614217372873, |
| "loss": 11.897, |
| "num_tokens": 4212599.0, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.413696715583508, |
| "grad_norm": 0.9073287957293683, |
| "learning_rate": 0.0001324474117013921, |
| "loss": 12.1639, |
| "num_tokens": 4219704.0, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.4143955276030748, |
| "grad_norm": 0.8985675842119918, |
| "learning_rate": 0.00013223323543883373, |
| "loss": 12.0619, |
| "num_tokens": 4226784.0, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.41509433962264153, |
| "grad_norm": 0.9107156893964011, |
| "learning_rate": 0.0001320188940468277, |
| "loss": 11.903, |
| "num_tokens": 4233717.0, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.41579315164220826, |
| "grad_norm": 0.8989603756831062, |
| "learning_rate": 0.0001318043886234356, |
| "loss": 12.0936, |
| "num_tokens": 4240647.0, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.416491963661775, |
| "grad_norm": 0.9806315936138839, |
| "learning_rate": 0.00013158972026755926, |
| "loss": 11.9643, |
| "num_tokens": 4247672.0, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.4171907756813417, |
| "grad_norm": 0.8241384074500768, |
| "learning_rate": 0.0001313748900789352, |
| "loss": 11.8706, |
| "num_tokens": 4255152.0, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.41788958770090845, |
| "grad_norm": 0.9271536643066998, |
| "learning_rate": 0.0001311598991581291, |
| "loss": 12.1374, |
| "num_tokens": 4261453.0, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.4185883997204752, |
| "grad_norm": 0.9002546899881367, |
| "learning_rate": 0.00013094474860652987, |
| "loss": 11.9134, |
| "num_tokens": 4268508.0, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.4192872117400419, |
| "grad_norm": 0.8757592331202514, |
| "learning_rate": 0.00013072943952634447, |
| "loss": 11.9621, |
| "num_tokens": 4275543.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.41998602375960864, |
| "grad_norm": 0.9204020510434368, |
| "learning_rate": 0.00013051397302059171, |
| "loss": 12.1479, |
| "num_tokens": 4282191.0, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.4206848357791754, |
| "grad_norm": 0.8328460955320042, |
| "learning_rate": 0.00013029835019309714, |
| "loss": 12.0701, |
| "num_tokens": 4289593.0, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.42138364779874216, |
| "grad_norm": 0.7838151120132228, |
| "learning_rate": 0.000130082572148487, |
| "loss": 11.8089, |
| "num_tokens": 4297297.0, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.4220824598183089, |
| "grad_norm": 0.8090146795838783, |
| "learning_rate": 0.00012986663999218261, |
| "loss": 12.0356, |
| "num_tokens": 4304161.0, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.4227812718378756, |
| "grad_norm": 0.8425172850446164, |
| "learning_rate": 0.00012965055483039507, |
| "loss": 11.7753, |
| "num_tokens": 4311640.0, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.42348008385744235, |
| "grad_norm": 0.8445465278085014, |
| "learning_rate": 0.00012943431777011902, |
| "loss": 12.0445, |
| "num_tokens": 4318619.0, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.4241788958770091, |
| "grad_norm": 0.8513080972741544, |
| "learning_rate": 0.00012921792991912753, |
| "loss": 12.0995, |
| "num_tokens": 4325488.0, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.4248777078965758, |
| "grad_norm": 0.7722421017802668, |
| "learning_rate": 0.00012900139238596598, |
| "loss": 11.7651, |
| "num_tokens": 4332936.0, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.42557651991614254, |
| "grad_norm": 0.8918135734711797, |
| "learning_rate": 0.00012878470627994664, |
| "loss": 11.8054, |
| "num_tokens": 4339915.0, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.42627533193570927, |
| "grad_norm": 0.8189265441430075, |
| "learning_rate": 0.0001285678727111429, |
| "loss": 11.7899, |
| "num_tokens": 4347052.0, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.42697414395527605, |
| "grad_norm": 0.8964986737425554, |
| "learning_rate": 0.00012835089279038362, |
| "loss": 11.9812, |
| "num_tokens": 4353752.0, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.4276729559748428, |
| "grad_norm": 0.8413717581563538, |
| "learning_rate": 0.00012813376762924733, |
| "loss": 12.1173, |
| "num_tokens": 4361038.0, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.4283717679944095, |
| "grad_norm": 0.850087071440023, |
| "learning_rate": 0.0001279164983400568, |
| "loss": 11.9678, |
| "num_tokens": 4368560.0, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.42907058001397624, |
| "grad_norm": 0.8010995265019617, |
| "learning_rate": 0.00012769908603587292, |
| "loss": 11.7605, |
| "num_tokens": 4376222.0, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.429769392033543, |
| "grad_norm": 0.8025102750844321, |
| "learning_rate": 0.0001274815318304894, |
| "loss": 11.8735, |
| "num_tokens": 4383884.0, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.4304682040531097, |
| "grad_norm": 0.8707077238242065, |
| "learning_rate": 0.0001272638368384269, |
| "loss": 11.9285, |
| "num_tokens": 4390978.0, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.43116701607267643, |
| "grad_norm": 0.8116822458689641, |
| "learning_rate": 0.00012704600217492725, |
| "loss": 12.1714, |
| "num_tokens": 4398693.0, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.43186582809224316, |
| "grad_norm": 0.826675965931149, |
| "learning_rate": 0.0001268280289559479, |
| "loss": 11.9573, |
| "num_tokens": 4405911.0, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.43256464011180995, |
| "grad_norm": 0.8534971115758093, |
| "learning_rate": 0.00012660991829815602, |
| "loss": 11.9621, |
| "num_tokens": 4413468.0, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.4332634521313767, |
| "grad_norm": 0.8760979261507388, |
| "learning_rate": 0.00012639167131892293, |
| "loss": 12.0476, |
| "num_tokens": 4420628.0, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.4339622641509434, |
| "grad_norm": 0.8714727411601618, |
| "learning_rate": 0.0001261732891363183, |
| "loss": 11.9621, |
| "num_tokens": 4427900.0, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.43466107617051014, |
| "grad_norm": 0.8922177474156737, |
| "learning_rate": 0.0001259547728691045, |
| "loss": 12.0367, |
| "num_tokens": 4434789.0, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.43535988819007687, |
| "grad_norm": 0.841294675830371, |
| "learning_rate": 0.00012573612363673067, |
| "loss": 12.1268, |
| "num_tokens": 4442355.0, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.4360587002096436, |
| "grad_norm": 0.8662418572358458, |
| "learning_rate": 0.00012551734255932727, |
| "loss": 11.8095, |
| "num_tokens": 4449354.0, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.43675751222921033, |
| "grad_norm": 0.8966435415783298, |
| "learning_rate": 0.0001252984307577001, |
| "loss": 11.7621, |
| "num_tokens": 4456996.0, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.43745632424877706, |
| "grad_norm": 0.9140099642044371, |
| "learning_rate": 0.00012507938935332478, |
| "loss": 11.8998, |
| "num_tokens": 4464108.0, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.4381551362683438, |
| "grad_norm": 0.9252375343477144, |
| "learning_rate": 0.00012486021946834068, |
| "loss": 11.7647, |
| "num_tokens": 4471782.0, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.4388539482879106, |
| "grad_norm": 0.9539679302319235, |
| "learning_rate": 0.00012464092222554552, |
| "loss": 11.768, |
| "num_tokens": 4478870.0, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.4395527603074773, |
| "grad_norm": 0.8265019024489157, |
| "learning_rate": 0.00012442149874838948, |
| "loss": 11.7075, |
| "num_tokens": 4486006.0, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.44025157232704404, |
| "grad_norm": 0.8259779874857927, |
| "learning_rate": 0.00012420195016096933, |
| "loss": 12.1217, |
| "num_tokens": 4493433.0, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.44095038434661077, |
| "grad_norm": 0.8504634346376905, |
| "learning_rate": 0.00012398227758802285, |
| "loss": 11.6897, |
| "num_tokens": 4501615.0, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.4416491963661775, |
| "grad_norm": 0.8422557275254934, |
| "learning_rate": 0.00012376248215492297, |
| "loss": 12.0223, |
| "num_tokens": 4508453.0, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.44234800838574423, |
| "grad_norm": 0.8739580752551247, |
| "learning_rate": 0.000123542564987672, |
| "loss": 11.9015, |
| "num_tokens": 4515645.0, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.44304682040531096, |
| "grad_norm": 0.7878006817363354, |
| "learning_rate": 0.00012332252721289594, |
| "loss": 11.6232, |
| "num_tokens": 4522914.0, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.4437456324248777, |
| "grad_norm": 0.8701782429674091, |
| "learning_rate": 0.00012310236995783866, |
| "loss": 12.065, |
| "num_tokens": 4530012.0, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.4444444444444444, |
| "grad_norm": 0.8206317742230089, |
| "learning_rate": 0.00012288209435035605, |
| "loss": 11.8848, |
| "num_tokens": 4536666.0, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.4451432564640112, |
| "grad_norm": 0.8478667503316676, |
| "learning_rate": 0.00012266170151891036, |
| "loss": 11.8555, |
| "num_tokens": 4543673.0, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.44584206848357794, |
| "grad_norm": 0.8711894725771603, |
| "learning_rate": 0.00012244119259256442, |
| "loss": 11.9741, |
| "num_tokens": 4550285.0, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.44654088050314467, |
| "grad_norm": 0.8073562820853625, |
| "learning_rate": 0.00012222056870097572, |
| "loss": 12.1476, |
| "num_tokens": 4557630.0, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.4472396925227114, |
| "grad_norm": 0.8550949203116837, |
| "learning_rate": 0.00012199983097439079, |
| "loss": 12.0266, |
| "num_tokens": 4564552.0, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.4479385045422781, |
| "grad_norm": 0.8361717391483878, |
| "learning_rate": 0.00012177898054363923, |
| "loss": 11.8465, |
| "num_tokens": 4572109.0, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.44863731656184486, |
| "grad_norm": 0.8342965373541649, |
| "learning_rate": 0.00012155801854012816, |
| "loss": 11.8991, |
| "num_tokens": 4579365.0, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.4493361285814116, |
| "grad_norm": 0.9064601938720505, |
| "learning_rate": 0.00012133694609583615, |
| "loss": 11.8593, |
| "num_tokens": 4585759.0, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.4500349406009783, |
| "grad_norm": 0.9010741359394927, |
| "learning_rate": 0.00012111576434330766, |
| "loss": 11.9357, |
| "num_tokens": 4592247.0, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.45073375262054505, |
| "grad_norm": 0.8334005590912353, |
| "learning_rate": 0.00012089447441564705, |
| "loss": 11.8785, |
| "num_tokens": 4599348.0, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.45143256464011183, |
| "grad_norm": 0.9178185738773574, |
| "learning_rate": 0.00012067307744651288, |
| "loss": 11.7183, |
| "num_tokens": 4606162.0, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.45213137665967856, |
| "grad_norm": 0.875013015269982, |
| "learning_rate": 0.00012045157457011211, |
| "loss": 11.6998, |
| "num_tokens": 4612972.0, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.4528301886792453, |
| "grad_norm": 0.9094140975159261, |
| "learning_rate": 0.00012022996692119424, |
| "loss": 11.9093, |
| "num_tokens": 4620976.0, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.453529000698812, |
| "grad_norm": 0.8153851869591483, |
| "learning_rate": 0.00012000825563504547, |
| "loss": 12.0035, |
| "num_tokens": 4629303.0, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.45422781271837875, |
| "grad_norm": 0.8950570859113985, |
| "learning_rate": 0.000119786441847483, |
| "loss": 11.8635, |
| "num_tokens": 4636638.0, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.4549266247379455, |
| "grad_norm": 0.8741693623971509, |
| "learning_rate": 0.00011956452669484908, |
| "loss": 11.824, |
| "num_tokens": 4643800.0, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.4556254367575122, |
| "grad_norm": 0.8940551480830519, |
| "learning_rate": 0.0001193425113140053, |
| "loss": 11.938, |
| "num_tokens": 4650901.0, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.45632424877707894, |
| "grad_norm": 0.9707279366744572, |
| "learning_rate": 0.00011912039684232674, |
| "loss": 12.0174, |
| "num_tokens": 4657604.0, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.4570230607966457, |
| "grad_norm": 0.9648273198741646, |
| "learning_rate": 0.000118898184417696, |
| "loss": 11.8252, |
| "num_tokens": 4664504.0, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.45772187281621246, |
| "grad_norm": 0.9945189728309309, |
| "learning_rate": 0.00011867587517849757, |
| "loss": 11.9346, |
| "num_tokens": 4670477.0, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.4584206848357792, |
| "grad_norm": 0.7900857843512296, |
| "learning_rate": 0.0001184534702636119, |
| "loss": 11.8475, |
| "num_tokens": 4677858.0, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.4591194968553459, |
| "grad_norm": 1.0503756020070543, |
| "learning_rate": 0.00011823097081240964, |
| "loss": 11.862, |
| "num_tokens": 4684626.0, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.45981830887491265, |
| "grad_norm": 0.8546502745711176, |
| "learning_rate": 0.00011800837796474561, |
| "loss": 12.1335, |
| "num_tokens": 4691582.0, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.4605171208944794, |
| "grad_norm": 0.8622254441465562, |
| "learning_rate": 0.00011778569286095329, |
| "loss": 11.9545, |
| "num_tokens": 4699398.0, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.4612159329140461, |
| "grad_norm": 0.7788567788431954, |
| "learning_rate": 0.00011756291664183859, |
| "loss": 11.9524, |
| "num_tokens": 4707448.0, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.46191474493361284, |
| "grad_norm": 0.908871171310194, |
| "learning_rate": 0.00011734005044867426, |
| "loss": 11.8801, |
| "num_tokens": 4714120.0, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.46261355695317957, |
| "grad_norm": 0.9693707218421472, |
| "learning_rate": 0.00011711709542319411, |
| "loss": 11.9347, |
| "num_tokens": 4721137.0, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.46331236897274636, |
| "grad_norm": 0.9709352688677537, |
| "learning_rate": 0.00011689405270758684, |
| "loss": 12.1173, |
| "num_tokens": 4727881.0, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.4640111809923131, |
| "grad_norm": 0.8758468731721972, |
| "learning_rate": 0.00011667092344449053, |
| "loss": 11.944, |
| "num_tokens": 4735809.0, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.4647099930118798, |
| "grad_norm": 0.9469256976252346, |
| "learning_rate": 0.00011644770877698654, |
| "loss": 12.1369, |
| "num_tokens": 4742094.0, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.46540880503144655, |
| "grad_norm": 0.8237527876349462, |
| "learning_rate": 0.00011622440984859384, |
| "loss": 11.8306, |
| "num_tokens": 4749810.0, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.4661076170510133, |
| "grad_norm": 1.0185008554050337, |
| "learning_rate": 0.00011600102780326296, |
| "loss": 11.9714, |
| "num_tokens": 4756916.0, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.46680642907058, |
| "grad_norm": 0.9390523791728513, |
| "learning_rate": 0.00011577756378537033, |
| "loss": 11.9079, |
| "num_tokens": 4764365.0, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.46750524109014674, |
| "grad_norm": 0.8903990493176246, |
| "learning_rate": 0.00011555401893971229, |
| "loss": 11.77, |
| "num_tokens": 4771335.0, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.46820405310971347, |
| "grad_norm": 0.8495503404738405, |
| "learning_rate": 0.00011533039441149926, |
| "loss": 11.8505, |
| "num_tokens": 4778789.0, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.4689028651292802, |
| "grad_norm": 0.8080825519779163, |
| "learning_rate": 0.00011510669134634984, |
| "loss": 11.858, |
| "num_tokens": 4786533.0, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.469601677148847, |
| "grad_norm": 0.8940185008450258, |
| "learning_rate": 0.000114882910890285, |
| "loss": 12.0071, |
| "num_tokens": 4793395.0, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.4703004891684137, |
| "grad_norm": 0.8102336082772988, |
| "learning_rate": 0.00011465905418972216, |
| "loss": 11.85, |
| "num_tokens": 4800501.0, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.47099930118798045, |
| "grad_norm": 0.8684446349073909, |
| "learning_rate": 0.00011443512239146941, |
| "loss": 11.7498, |
| "num_tokens": 4807730.0, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.4716981132075472, |
| "grad_norm": 0.8972831808941225, |
| "learning_rate": 0.00011421111664271946, |
| "loss": 12.1878, |
| "num_tokens": 4814631.0, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.4723969252271139, |
| "grad_norm": 0.7811357044524173, |
| "learning_rate": 0.00011398703809104391, |
| "loss": 11.977, |
| "num_tokens": 4821859.0, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.47309573724668064, |
| "grad_norm": 0.8878196813815282, |
| "learning_rate": 0.00011376288788438734, |
| "loss": 12.0799, |
| "num_tokens": 4829001.0, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.47379454926624737, |
| "grad_norm": 0.7728940561788115, |
| "learning_rate": 0.00011353866717106137, |
| "loss": 12.0937, |
| "num_tokens": 4836233.0, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.4744933612858141, |
| "grad_norm": 0.7652381421322222, |
| "learning_rate": 0.0001133143770997389, |
| "loss": 11.8881, |
| "num_tokens": 4844216.0, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.4751921733053808, |
| "grad_norm": 0.805128881427533, |
| "learning_rate": 0.00011309001881944809, |
| "loss": 11.9522, |
| "num_tokens": 4851705.0, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.4758909853249476, |
| "grad_norm": 0.8597585731076033, |
| "learning_rate": 0.00011286559347956651, |
| "loss": 12.1198, |
| "num_tokens": 4858279.0, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.47658979734451434, |
| "grad_norm": 0.8453118490353572, |
| "learning_rate": 0.00011264110222981535, |
| "loss": 11.9634, |
| "num_tokens": 4866344.0, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.4772886093640811, |
| "grad_norm": 0.8093626293120404, |
| "learning_rate": 0.00011241654622025334, |
| "loss": 11.8598, |
| "num_tokens": 4873494.0, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.4779874213836478, |
| "grad_norm": 0.8105493286965195, |
| "learning_rate": 0.00011219192660127116, |
| "loss": 11.7991, |
| "num_tokens": 4880904.0, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.47868623340321453, |
| "grad_norm": 0.7479575202425699, |
| "learning_rate": 0.00011196724452358516, |
| "loss": 11.7939, |
| "num_tokens": 4888663.0, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.47938504542278126, |
| "grad_norm": 0.9078873016468286, |
| "learning_rate": 0.00011174250113823173, |
| "loss": 11.8651, |
| "num_tokens": 4895524.0, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.480083857442348, |
| "grad_norm": 0.82828599173561, |
| "learning_rate": 0.00011151769759656136, |
| "loss": 11.8728, |
| "num_tokens": 4902875.0, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.4807826694619147, |
| "grad_norm": 0.8498893366628105, |
| "learning_rate": 0.00011129283505023274, |
| "loss": 11.988, |
| "num_tokens": 4910339.0, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.48148148148148145, |
| "grad_norm": 0.8691199007501399, |
| "learning_rate": 0.00011106791465120678, |
| "loss": 11.8249, |
| "num_tokens": 4917859.0, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.48218029350104824, |
| "grad_norm": 0.8363420468675724, |
| "learning_rate": 0.00011084293755174083, |
| "loss": 11.7725, |
| "num_tokens": 4924845.0, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.48287910552061497, |
| "grad_norm": 0.8342914301300645, |
| "learning_rate": 0.0001106179049043826, |
| "loss": 11.9105, |
| "num_tokens": 4932581.0, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.4835779175401817, |
| "grad_norm": 0.9041859543752735, |
| "learning_rate": 0.00011039281786196454, |
| "loss": 11.8527, |
| "num_tokens": 4938840.0, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.48427672955974843, |
| "grad_norm": 0.8332056105830793, |
| "learning_rate": 0.00011016767757759758, |
| "loss": 11.8375, |
| "num_tokens": 4946122.0, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.48497554157931516, |
| "grad_norm": 0.7999356426489574, |
| "learning_rate": 0.00010994248520466555, |
| "loss": 11.8864, |
| "num_tokens": 4953325.0, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.4856743535988819, |
| "grad_norm": 0.8840646203709945, |
| "learning_rate": 0.00010971724189681907, |
| "loss": 11.7926, |
| "num_tokens": 4960210.0, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.4863731656184486, |
| "grad_norm": 0.7161655467219089, |
| "learning_rate": 0.00010949194880796966, |
| "loss": 11.6796, |
| "num_tokens": 4968396.0, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.48707197763801535, |
| "grad_norm": 0.8391872409460347, |
| "learning_rate": 0.000109266607092284, |
| "loss": 11.8614, |
| "num_tokens": 4975812.0, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.48777078965758214, |
| "grad_norm": 0.8136075034785465, |
| "learning_rate": 0.00010904121790417767, |
| "loss": 11.7733, |
| "num_tokens": 4983615.0, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.48846960167714887, |
| "grad_norm": 0.8639838769972311, |
| "learning_rate": 0.00010881578239830965, |
| "loss": 11.7194, |
| "num_tokens": 4990664.0, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.4891684136967156, |
| "grad_norm": 0.771704690891524, |
| "learning_rate": 0.0001085903017295761, |
| "loss": 11.8352, |
| "num_tokens": 4997991.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.48986722571628233, |
| "grad_norm": 0.8082632295628299, |
| "learning_rate": 0.00010836477705310457, |
| "loss": 11.7211, |
| "num_tokens": 5005535.0, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.49056603773584906, |
| "grad_norm": 0.8221099777163831, |
| "learning_rate": 0.00010813920952424805, |
| "loss": 11.7424, |
| "num_tokens": 5012725.0, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.4912648497554158, |
| "grad_norm": 0.8526099066187208, |
| "learning_rate": 0.00010791360029857908, |
| "loss": 11.7562, |
| "num_tokens": 5020226.0, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.4919636617749825, |
| "grad_norm": 0.8830746317686873, |
| "learning_rate": 0.00010768795053188378, |
| "loss": 11.8946, |
| "num_tokens": 5027841.0, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.49266247379454925, |
| "grad_norm": 0.8467987158546797, |
| "learning_rate": 0.00010746226138015605, |
| "loss": 11.8602, |
| "num_tokens": 5034672.0, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.493361285814116, |
| "grad_norm": 0.9110055267989619, |
| "learning_rate": 0.00010723653399959141, |
| "loss": 11.8089, |
| "num_tokens": 5041763.0, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.49406009783368277, |
| "grad_norm": 0.8590754661724935, |
| "learning_rate": 0.00010701076954658133, |
| "loss": 12.0917, |
| "num_tokens": 5048922.0, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.4947589098532495, |
| "grad_norm": 0.8276297434047943, |
| "learning_rate": 0.00010678496917770719, |
| "loss": 11.9569, |
| "num_tokens": 5056008.0, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.4954577218728162, |
| "grad_norm": 0.8873635245443052, |
| "learning_rate": 0.00010655913404973432, |
| "loss": 11.9105, |
| "num_tokens": 5062729.0, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.49615653389238296, |
| "grad_norm": 0.7860622687601649, |
| "learning_rate": 0.0001063332653196062, |
| "loss": 11.9105, |
| "num_tokens": 5069602.0, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.4968553459119497, |
| "grad_norm": 0.8585551672598832, |
| "learning_rate": 0.00010610736414443836, |
| "loss": 11.8514, |
| "num_tokens": 5076146.0, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.4975541579315164, |
| "grad_norm": 0.813654694030058, |
| "learning_rate": 0.00010588143168151257, |
| "loss": 11.7629, |
| "num_tokens": 5083612.0, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.49825296995108315, |
| "grad_norm": 0.88870371822207, |
| "learning_rate": 0.00010565546908827093, |
| "loss": 11.7497, |
| "num_tokens": 5090353.0, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.4989517819706499, |
| "grad_norm": 0.8315503003113751, |
| "learning_rate": 0.00010542947752230987, |
| "loss": 11.7071, |
| "num_tokens": 5098640.0, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.4996505939902166, |
| "grad_norm": 0.7808060509465453, |
| "learning_rate": 0.00010520345814137422, |
| "loss": 11.7736, |
| "num_tokens": 5106336.0, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.5003494060097834, |
| "grad_norm": 0.9832986052690458, |
| "learning_rate": 0.0001049774121033514, |
| "loss": 11.8544, |
| "num_tokens": 5113949.0, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.5010482180293501, |
| "grad_norm": 0.8593993063532692, |
| "learning_rate": 0.00010475134056626521, |
| "loss": 11.5894, |
| "num_tokens": 5120936.0, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.5017470300489169, |
| "grad_norm": 0.8555509359111242, |
| "learning_rate": 0.00010452524468827028, |
| "loss": 11.8802, |
| "num_tokens": 5127936.0, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.5024458420684835, |
| "grad_norm": 0.8768545835966081, |
| "learning_rate": 0.00010429912562764582, |
| "loss": 11.7651, |
| "num_tokens": 5135181.0, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.5031446540880503, |
| "grad_norm": 0.8954332003506361, |
| "learning_rate": 0.00010407298454278983, |
| "loss": 11.9635, |
| "num_tokens": 5141581.0, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.5038434661076171, |
| "grad_norm": 0.7787241308163475, |
| "learning_rate": 0.00010384682259221314, |
| "loss": 11.8157, |
| "num_tokens": 5149210.0, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.5045422781271838, |
| "grad_norm": 0.8043762411091894, |
| "learning_rate": 0.00010362064093453347, |
| "loss": 11.9794, |
| "num_tokens": 5156153.0, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.5052410901467506, |
| "grad_norm": 0.8494247945373716, |
| "learning_rate": 0.00010339444072846955, |
| "loss": 11.8751, |
| "num_tokens": 5163697.0, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.5059399021663172, |
| "grad_norm": 0.8004411120888781, |
| "learning_rate": 0.00010316822313283503, |
| "loss": 11.8148, |
| "num_tokens": 5170939.0, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.506638714185884, |
| "grad_norm": 0.8098718525402316, |
| "learning_rate": 0.00010294198930653273, |
| "loss": 11.7984, |
| "num_tokens": 5178058.0, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.5073375262054507, |
| "grad_norm": 0.8570224646790227, |
| "learning_rate": 0.00010271574040854863, |
| "loss": 11.7122, |
| "num_tokens": 5185093.0, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.5080363382250175, |
| "grad_norm": 0.9625030998434088, |
| "learning_rate": 0.00010248947759794583, |
| "loss": 11.9835, |
| "num_tokens": 5191726.0, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.5087351502445842, |
| "grad_norm": 0.8635080331426315, |
| "learning_rate": 0.00010226320203385878, |
| "loss": 11.7241, |
| "num_tokens": 5198609.0, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.5094339622641509, |
| "grad_norm": 0.8677572413713257, |
| "learning_rate": 0.00010203691487548721, |
| "loss": 11.8883, |
| "num_tokens": 5206060.0, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.5101327742837177, |
| "grad_norm": 0.8979446096984454, |
| "learning_rate": 0.00010181061728209034, |
| "loss": 11.9268, |
| "num_tokens": 5213009.0, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.5108315863032844, |
| "grad_norm": 0.8252839906702719, |
| "learning_rate": 0.00010158431041298076, |
| "loss": 12.0079, |
| "num_tokens": 5220145.0, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.5115303983228512, |
| "grad_norm": 0.8819812351285876, |
| "learning_rate": 0.00010135799542751861, |
| "loss": 11.8362, |
| "num_tokens": 5226851.0, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.5122292103424179, |
| "grad_norm": 1.1383179013782923, |
| "learning_rate": 0.0001011316734851056, |
| "loss": 11.7761, |
| "num_tokens": 5234276.0, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.5129280223619846, |
| "grad_norm": 0.8044321074171096, |
| "learning_rate": 0.00010090534574517907, |
| "loss": 11.7191, |
| "num_tokens": 5241284.0, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.5136268343815513, |
| "grad_norm": 1.1007084709280315, |
| "learning_rate": 0.00010067901336720611, |
| "loss": 11.6657, |
| "num_tokens": 5248568.0, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.5143256464011181, |
| "grad_norm": 0.8870149031390947, |
| "learning_rate": 0.00010045267751067757, |
| "loss": 12.0822, |
| "num_tokens": 5255337.0, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.5150244584206848, |
| "grad_norm": 0.9108240607842385, |
| "learning_rate": 0.00010022633933510201, |
| "loss": 11.6611, |
| "num_tokens": 5262391.0, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.5157232704402516, |
| "grad_norm": 0.8605558610574175, |
| "learning_rate": 0.0001, |
| "loss": 11.8638, |
| "num_tokens": 5270611.0, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.5164220824598184, |
| "grad_norm": 0.8728553331392952, |
| "learning_rate": 9.977366066489801e-05, |
| "loss": 11.74, |
| "num_tokens": 5278249.0, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.517120894479385, |
| "grad_norm": 0.9400700814466663, |
| "learning_rate": 9.954732248932244e-05, |
| "loss": 11.8641, |
| "num_tokens": 5285271.0, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.5178197064989518, |
| "grad_norm": 0.8495250755418514, |
| "learning_rate": 9.932098663279392e-05, |
| "loss": 11.7304, |
| "num_tokens": 5292168.0, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.5185185185185185, |
| "grad_norm": 0.9155703182347676, |
| "learning_rate": 9.909465425482093e-05, |
| "loss": 11.9257, |
| "num_tokens": 5298892.0, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.5192173305380853, |
| "grad_norm": 0.9016515273132614, |
| "learning_rate": 9.886832651489444e-05, |
| "loss": 11.8944, |
| "num_tokens": 5305743.0, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.519916142557652, |
| "grad_norm": 0.8118283003676506, |
| "learning_rate": 9.864200457248144e-05, |
| "loss": 11.6752, |
| "num_tokens": 5313196.0, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.5206149545772187, |
| "grad_norm": 0.8119495179861029, |
| "learning_rate": 9.841568958701924e-05, |
| "loss": 11.713, |
| "num_tokens": 5320688.0, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.5213137665967854, |
| "grad_norm": 0.8445755422496691, |
| "learning_rate": 9.81893827179097e-05, |
| "loss": 11.712, |
| "num_tokens": 5328457.0, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.5220125786163522, |
| "grad_norm": 0.8525238796762149, |
| "learning_rate": 9.796308512451284e-05, |
| "loss": 11.8845, |
| "num_tokens": 5335909.0, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.522711390635919, |
| "grad_norm": 0.9259968525833697, |
| "learning_rate": 9.773679796614124e-05, |
| "loss": 11.6382, |
| "num_tokens": 5343377.0, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.5234102026554857, |
| "grad_norm": 0.7746588754673115, |
| "learning_rate": 9.751052240205421e-05, |
| "loss": 11.8223, |
| "num_tokens": 5350725.0, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.5241090146750524, |
| "grad_norm": 0.9090044410487871, |
| "learning_rate": 9.728425959145139e-05, |
| "loss": 11.736, |
| "num_tokens": 5358235.0, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.5248078266946191, |
| "grad_norm": 0.8119032545667709, |
| "learning_rate": 9.705801069346729e-05, |
| "loss": 11.8128, |
| "num_tokens": 5365590.0, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.5255066387141859, |
| "grad_norm": 0.8593779120663517, |
| "learning_rate": 9.683177686716501e-05, |
| "loss": 11.7261, |
| "num_tokens": 5371915.0, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.5262054507337526, |
| "grad_norm": 0.9901347880257034, |
| "learning_rate": 9.660555927153047e-05, |
| "loss": 11.9423, |
| "num_tokens": 5378290.0, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.5269042627533194, |
| "grad_norm": 0.7442310095588829, |
| "learning_rate": 9.637935906546655e-05, |
| "loss": 11.7566, |
| "num_tokens": 5385294.0, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.527603074772886, |
| "grad_norm": 0.8575812640939505, |
| "learning_rate": 9.615317740778689e-05, |
| "loss": 11.9029, |
| "num_tokens": 5392707.0, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.5283018867924528, |
| "grad_norm": 0.8047709007984075, |
| "learning_rate": 9.592701545721021e-05, |
| "loss": 11.8251, |
| "num_tokens": 5400588.0, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.5290006988120196, |
| "grad_norm": 0.7860377665780968, |
| "learning_rate": 9.570087437235423e-05, |
| "loss": 11.905, |
| "num_tokens": 5407337.0, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.5296995108315863, |
| "grad_norm": 0.8194379087299098, |
| "learning_rate": 9.547475531172973e-05, |
| "loss": 11.8038, |
| "num_tokens": 5414577.0, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.5303983228511531, |
| "grad_norm": 0.8070308802592453, |
| "learning_rate": 9.524865943373481e-05, |
| "loss": 11.8402, |
| "num_tokens": 5421211.0, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.5310971348707197, |
| "grad_norm": 0.8803852392568843, |
| "learning_rate": 9.502258789664865e-05, |
| "loss": 11.944, |
| "num_tokens": 5427910.0, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.5317959468902865, |
| "grad_norm": 0.8457383467882692, |
| "learning_rate": 9.479654185862579e-05, |
| "loss": 11.9176, |
| "num_tokens": 5434742.0, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.5324947589098532, |
| "grad_norm": 0.7743210614507713, |
| "learning_rate": 9.457052247769017e-05, |
| "loss": 11.847, |
| "num_tokens": 5441794.0, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.53319357092942, |
| "grad_norm": 0.7600921058483522, |
| "learning_rate": 9.434453091172908e-05, |
| "loss": 11.7134, |
| "num_tokens": 5449915.0, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.5338923829489868, |
| "grad_norm": 0.8410587669442078, |
| "learning_rate": 9.411856831848745e-05, |
| "loss": 11.6731, |
| "num_tokens": 5456843.0, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.5345911949685535, |
| "grad_norm": 0.8081870610413466, |
| "learning_rate": 9.38926358555617e-05, |
| "loss": 11.8152, |
| "num_tokens": 5464023.0, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.5352900069881202, |
| "grad_norm": 0.8361358425721516, |
| "learning_rate": 9.366673468039383e-05, |
| "loss": 11.9436, |
| "num_tokens": 5471406.0, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.5359888190076869, |
| "grad_norm": 0.8862719304918909, |
| "learning_rate": 9.34408659502657e-05, |
| "loss": 11.8146, |
| "num_tokens": 5478339.0, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.5366876310272537, |
| "grad_norm": 0.8320140788538509, |
| "learning_rate": 9.321503082229282e-05, |
| "loss": 11.79, |
| "num_tokens": 5485853.0, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.5373864430468204, |
| "grad_norm": 0.8427164891185774, |
| "learning_rate": 9.298923045341869e-05, |
| "loss": 11.6928, |
| "num_tokens": 5493547.0, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.5380852550663872, |
| "grad_norm": 0.9065833012454514, |
| "learning_rate": 9.276346600040862e-05, |
| "loss": 11.4978, |
| "num_tokens": 5500515.0, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.5387840670859538, |
| "grad_norm": 0.8367627544247601, |
| "learning_rate": 9.253773861984397e-05, |
| "loss": 11.6405, |
| "num_tokens": 5507599.0, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.5394828791055206, |
| "grad_norm": 0.8541655457970996, |
| "learning_rate": 9.231204946811624e-05, |
| "loss": 11.7868, |
| "num_tokens": 5514593.0, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.5401816911250874, |
| "grad_norm": 0.8505734247130968, |
| "learning_rate": 9.208639970142093e-05, |
| "loss": 11.8589, |
| "num_tokens": 5521763.0, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.5408805031446541, |
| "grad_norm": 0.7014079921724546, |
| "learning_rate": 9.186079047575197e-05, |
| "loss": 11.7853, |
| "num_tokens": 5529580.0, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.5415793151642209, |
| "grad_norm": 0.775435021642862, |
| "learning_rate": 9.163522294689546e-05, |
| "loss": 11.7545, |
| "num_tokens": 5536873.0, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.5422781271837875, |
| "grad_norm": 0.8986198411125287, |
| "learning_rate": 9.140969827042391e-05, |
| "loss": 11.8047, |
| "num_tokens": 5543550.0, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.5429769392033543, |
| "grad_norm": 0.7809603754618126, |
| "learning_rate": 9.118421760169038e-05, |
| "loss": 11.9793, |
| "num_tokens": 5550884.0, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.543675751222921, |
| "grad_norm": 0.8057085091276677, |
| "learning_rate": 9.095878209582237e-05, |
| "loss": 11.5977, |
| "num_tokens": 5557807.0, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.5443745632424878, |
| "grad_norm": 0.891977933987541, |
| "learning_rate": 9.073339290771603e-05, |
| "loss": 11.8543, |
| "num_tokens": 5564576.0, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.5450733752620545, |
| "grad_norm": 0.7388141711422563, |
| "learning_rate": 9.050805119203035e-05, |
| "loss": 11.7053, |
| "num_tokens": 5572504.0, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.5457721872816212, |
| "grad_norm": 0.7625862495200397, |
| "learning_rate": 9.028275810318095e-05, |
| "loss": 11.6129, |
| "num_tokens": 5579720.0, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.546470999301188, |
| "grad_norm": 0.7581306071682549, |
| "learning_rate": 9.005751479533449e-05, |
| "loss": 11.8261, |
| "num_tokens": 5586866.0, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.5471698113207547, |
| "grad_norm": 0.8062850504793252, |
| "learning_rate": 8.983232242240247e-05, |
| "loss": 11.7799, |
| "num_tokens": 5593273.0, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.5478686233403215, |
| "grad_norm": 0.7696652830589925, |
| "learning_rate": 8.96071821380355e-05, |
| "loss": 11.9284, |
| "num_tokens": 5600825.0, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.5485674353598882, |
| "grad_norm": 0.7465186876499212, |
| "learning_rate": 8.938209509561741e-05, |
| "loss": 11.8043, |
| "num_tokens": 5608068.0, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.549266247379455, |
| "grad_norm": 0.7475919153703743, |
| "learning_rate": 8.91570624482592e-05, |
| "loss": 11.7928, |
| "num_tokens": 5615637.0, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.5499650593990216, |
| "grad_norm": 0.8374151644755823, |
| "learning_rate": 8.893208534879324e-05, |
| "loss": 11.6898, |
| "num_tokens": 5622822.0, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.5506638714185884, |
| "grad_norm": 0.747989105129382, |
| "learning_rate": 8.87071649497673e-05, |
| "loss": 11.7666, |
| "num_tokens": 5629785.0, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.5513626834381551, |
| "grad_norm": 0.7714546710718322, |
| "learning_rate": 8.848230240343865e-05, |
| "loss": 11.7869, |
| "num_tokens": 5636551.0, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.5520614954577219, |
| "grad_norm": 0.8093924548009458, |
| "learning_rate": 8.82574988617683e-05, |
| "loss": 11.7715, |
| "num_tokens": 5643210.0, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.5527603074772887, |
| "grad_norm": 0.7791258807969518, |
| "learning_rate": 8.803275547641488e-05, |
| "loss": 11.6265, |
| "num_tokens": 5650927.0, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.5534591194968553, |
| "grad_norm": 0.7501244610969549, |
| "learning_rate": 8.780807339872886e-05, |
| "loss": 11.8536, |
| "num_tokens": 5658354.0, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.5541579315164221, |
| "grad_norm": 0.9213911806759102, |
| "learning_rate": 8.758345377974667e-05, |
| "loss": 11.9677, |
| "num_tokens": 5664499.0, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.5548567435359888, |
| "grad_norm": 0.8448158401403486, |
| "learning_rate": 8.735889777018465e-05, |
| "loss": 11.8338, |
| "num_tokens": 5670823.0, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 0.7961882205740662, |
| "learning_rate": 8.71344065204335e-05, |
| "loss": 11.6791, |
| "num_tokens": 5678464.0, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.5562543675751223, |
| "grad_norm": 0.9026550542048194, |
| "learning_rate": 8.690998118055193e-05, |
| "loss": 11.701, |
| "num_tokens": 5685313.0, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.556953179594689, |
| "grad_norm": 0.763089852499311, |
| "learning_rate": 8.66856229002611e-05, |
| "loss": 11.819, |
| "num_tokens": 5692972.0, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.5576519916142557, |
| "grad_norm": 0.8827640381152261, |
| "learning_rate": 8.646133282893864e-05, |
| "loss": 11.6831, |
| "num_tokens": 5699327.0, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.5583508036338225, |
| "grad_norm": 0.7797458428691016, |
| "learning_rate": 8.623711211561267e-05, |
| "loss": 11.6918, |
| "num_tokens": 5706702.0, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.5590496156533893, |
| "grad_norm": 0.7838131784869478, |
| "learning_rate": 8.601296190895611e-05, |
| "loss": 11.8356, |
| "num_tokens": 5713918.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.559748427672956, |
| "grad_norm": 0.8784533058303018, |
| "learning_rate": 8.578888335728057e-05, |
| "loss": 11.6701, |
| "num_tokens": 5721220.0, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.5604472396925227, |
| "grad_norm": 0.8123001209869102, |
| "learning_rate": 8.55648776085306e-05, |
| "loss": 11.7312, |
| "num_tokens": 5728492.0, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.5611460517120894, |
| "grad_norm": 0.7858814515665766, |
| "learning_rate": 8.534094581027785e-05, |
| "loss": 11.876, |
| "num_tokens": 5735983.0, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.5618448637316562, |
| "grad_norm": 0.8844901910925126, |
| "learning_rate": 8.511708910971505e-05, |
| "loss": 11.7999, |
| "num_tokens": 5742678.0, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.5625436757512229, |
| "grad_norm": 0.8078853147041326, |
| "learning_rate": 8.489330865365018e-05, |
| "loss": 11.7124, |
| "num_tokens": 5750553.0, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.5632424877707897, |
| "grad_norm": 0.7784603290334375, |
| "learning_rate": 8.466960558850077e-05, |
| "loss": 11.4818, |
| "num_tokens": 5758022.0, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.5639412997903563, |
| "grad_norm": 0.7639221830150267, |
| "learning_rate": 8.444598106028773e-05, |
| "loss": 11.6915, |
| "num_tokens": 5765255.0, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.5646401118099231, |
| "grad_norm": 0.802882077997663, |
| "learning_rate": 8.422243621462969e-05, |
| "loss": 11.7076, |
| "num_tokens": 5772746.0, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.5653389238294899, |
| "grad_norm": 0.7669134365996498, |
| "learning_rate": 8.399897219673709e-05, |
| "loss": 11.6297, |
| "num_tokens": 5779819.0, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.5660377358490566, |
| "grad_norm": 0.7376107274278996, |
| "learning_rate": 8.37755901514062e-05, |
| "loss": 11.7143, |
| "num_tokens": 5787293.0, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.5667365478686234, |
| "grad_norm": 0.7229725613909326, |
| "learning_rate": 8.355229122301348e-05, |
| "loss": 11.6863, |
| "num_tokens": 5795358.0, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.56743535988819, |
| "grad_norm": 0.8065237716277036, |
| "learning_rate": 8.332907655550948e-05, |
| "loss": 11.8595, |
| "num_tokens": 5802192.0, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.5681341719077568, |
| "grad_norm": 0.8397707150038022, |
| "learning_rate": 8.310594729241317e-05, |
| "loss": 11.8714, |
| "num_tokens": 5809107.0, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.5688329839273235, |
| "grad_norm": 0.7728890632051899, |
| "learning_rate": 8.288290457680591e-05, |
| "loss": 11.6326, |
| "num_tokens": 5815896.0, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.5695317959468903, |
| "grad_norm": 0.7647611926282754, |
| "learning_rate": 8.265994955132572e-05, |
| "loss": 11.8586, |
| "num_tokens": 5823124.0, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.570230607966457, |
| "grad_norm": 0.835572575499216, |
| "learning_rate": 8.243708335816145e-05, |
| "loss": 11.9913, |
| "num_tokens": 5829951.0, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.5709294199860238, |
| "grad_norm": 0.7771857677434221, |
| "learning_rate": 8.221430713904672e-05, |
| "loss": 11.5943, |
| "num_tokens": 5836842.0, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.5716282320055905, |
| "grad_norm": 0.7915716967198777, |
| "learning_rate": 8.19916220352544e-05, |
| "loss": 11.7728, |
| "num_tokens": 5844342.0, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.5723270440251572, |
| "grad_norm": 0.7337383433889063, |
| "learning_rate": 8.176902918759041e-05, |
| "loss": 11.7272, |
| "num_tokens": 5851608.0, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.573025856044724, |
| "grad_norm": 0.7784845393576943, |
| "learning_rate": 8.15465297363881e-05, |
| "loss": 11.5759, |
| "num_tokens": 5858938.0, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.5737246680642907, |
| "grad_norm": 0.7834054172956932, |
| "learning_rate": 8.132412482150245e-05, |
| "loss": 11.7405, |
| "num_tokens": 5865983.0, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.5744234800838575, |
| "grad_norm": 0.8342582004871633, |
| "learning_rate": 8.110181558230404e-05, |
| "loss": 11.8399, |
| "num_tokens": 5872600.0, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.5751222921034241, |
| "grad_norm": 0.8664104035122296, |
| "learning_rate": 8.087960315767328e-05, |
| "loss": 11.7264, |
| "num_tokens": 5878398.0, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.5758211041229909, |
| "grad_norm": 0.728258050995661, |
| "learning_rate": 8.06574886859947e-05, |
| "loss": 11.8376, |
| "num_tokens": 5886022.0, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.5765199161425576, |
| "grad_norm": 0.7762790042757487, |
| "learning_rate": 8.043547330515092e-05, |
| "loss": 11.7216, |
| "num_tokens": 5893141.0, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.5772187281621244, |
| "grad_norm": 0.76741040938386, |
| "learning_rate": 8.021355815251703e-05, |
| "loss": 11.7179, |
| "num_tokens": 5900162.0, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.5779175401816912, |
| "grad_norm": 0.8431161282769128, |
| "learning_rate": 7.999174436495456e-05, |
| "loss": 11.5942, |
| "num_tokens": 5907169.0, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.5786163522012578, |
| "grad_norm": 0.761803797183389, |
| "learning_rate": 7.97700330788058e-05, |
| "loss": 11.866, |
| "num_tokens": 5914949.0, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.5793151642208246, |
| "grad_norm": 0.8127619905274301, |
| "learning_rate": 7.954842542988792e-05, |
| "loss": 11.8023, |
| "num_tokens": 5921813.0, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.5800139762403913, |
| "grad_norm": 0.7588539365182271, |
| "learning_rate": 7.932692255348711e-05, |
| "loss": 11.9015, |
| "num_tokens": 5929467.0, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.5807127882599581, |
| "grad_norm": 0.722417526935907, |
| "learning_rate": 7.910552558435297e-05, |
| "loss": 11.5916, |
| "num_tokens": 5936756.0, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.5814116002795248, |
| "grad_norm": 0.734304834951851, |
| "learning_rate": 7.888423565669236e-05, |
| "loss": 11.7022, |
| "num_tokens": 5944825.0, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.5821104122990916, |
| "grad_norm": 0.7651356943028277, |
| "learning_rate": 7.866305390416385e-05, |
| "loss": 11.7116, |
| "num_tokens": 5951987.0, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.5828092243186582, |
| "grad_norm": 0.7263482640928406, |
| "learning_rate": 7.844198145987187e-05, |
| "loss": 11.6426, |
| "num_tokens": 5959236.0, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.583508036338225, |
| "grad_norm": 0.8136003869167677, |
| "learning_rate": 7.82210194563608e-05, |
| "loss": 12.0265, |
| "num_tokens": 5966125.0, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.5842068483577918, |
| "grad_norm": 0.7931185544520268, |
| "learning_rate": 7.800016902560924e-05, |
| "loss": 11.7356, |
| "num_tokens": 5974067.0, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.5849056603773585, |
| "grad_norm": 0.77986440427414, |
| "learning_rate": 7.77794312990243e-05, |
| "loss": 11.7585, |
| "num_tokens": 5980876.0, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.5856044723969253, |
| "grad_norm": 0.7930623347892715, |
| "learning_rate": 7.755880740743559e-05, |
| "loss": 11.9373, |
| "num_tokens": 5988494.0, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.5863032844164919, |
| "grad_norm": 0.817088280185177, |
| "learning_rate": 7.733829848108965e-05, |
| "loss": 11.6533, |
| "num_tokens": 5995555.0, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.5870020964360587, |
| "grad_norm": 0.816574051747384, |
| "learning_rate": 7.7117905649644e-05, |
| "loss": 11.5985, |
| "num_tokens": 6002655.0, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.5877009084556254, |
| "grad_norm": 0.7920959598315253, |
| "learning_rate": 7.689763004216135e-05, |
| "loss": 11.8124, |
| "num_tokens": 6009185.0, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.5883997204751922, |
| "grad_norm": 0.8852589257704687, |
| "learning_rate": 7.667747278710406e-05, |
| "loss": 11.54, |
| "num_tokens": 6016823.0, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.589098532494759, |
| "grad_norm": 0.7548742035632593, |
| "learning_rate": 7.6457435012328e-05, |
| "loss": 11.7246, |
| "num_tokens": 6024771.0, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.5897973445143256, |
| "grad_norm": 0.854098594965677, |
| "learning_rate": 7.623751784507706e-05, |
| "loss": 11.664, |
| "num_tokens": 6031975.0, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.5904961565338924, |
| "grad_norm": 0.7853779219200732, |
| "learning_rate": 7.601772241197719e-05, |
| "loss": 11.9128, |
| "num_tokens": 6039124.0, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.5911949685534591, |
| "grad_norm": 0.8383469114270452, |
| "learning_rate": 7.579804983903067e-05, |
| "loss": 11.8526, |
| "num_tokens": 6045752.0, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.5918937805730259, |
| "grad_norm": 0.8878587452859725, |
| "learning_rate": 7.557850125161053e-05, |
| "loss": 11.856, |
| "num_tokens": 6052231.0, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.5925925925925926, |
| "grad_norm": 0.8333451117092037, |
| "learning_rate": 7.535907777445449e-05, |
| "loss": 11.6259, |
| "num_tokens": 6059886.0, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.5932914046121593, |
| "grad_norm": 0.8078970484894613, |
| "learning_rate": 7.513978053165934e-05, |
| "loss": 11.8283, |
| "num_tokens": 6066548.0, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.593990216631726, |
| "grad_norm": 0.7637621169363601, |
| "learning_rate": 7.492061064667526e-05, |
| "loss": 11.7571, |
| "num_tokens": 6074097.0, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.5946890286512928, |
| "grad_norm": 0.8517447452369387, |
| "learning_rate": 7.470156924229988e-05, |
| "loss": 11.7946, |
| "num_tokens": 6080717.0, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.5953878406708596, |
| "grad_norm": 0.7944555032647729, |
| "learning_rate": 7.448265744067275e-05, |
| "loss": 11.6181, |
| "num_tokens": 6088678.0, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.5960866526904263, |
| "grad_norm": 0.8290060549583308, |
| "learning_rate": 7.426387636326936e-05, |
| "loss": 11.7951, |
| "num_tokens": 6095534.0, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.596785464709993, |
| "grad_norm": 0.7769775625071952, |
| "learning_rate": 7.404522713089554e-05, |
| "loss": 11.8202, |
| "num_tokens": 6102965.0, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.5974842767295597, |
| "grad_norm": 0.788894868014834, |
| "learning_rate": 7.382671086368172e-05, |
| "loss": 11.5052, |
| "num_tokens": 6109857.0, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.5981830887491265, |
| "grad_norm": 0.9269678831820426, |
| "learning_rate": 7.360832868107708e-05, |
| "loss": 12.0685, |
| "num_tokens": 6116670.0, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.5988819007686932, |
| "grad_norm": 0.7500023253076064, |
| "learning_rate": 7.3390081701844e-05, |
| "loss": 11.6106, |
| "num_tokens": 6123896.0, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.59958071278826, |
| "grad_norm": 0.8334471335860142, |
| "learning_rate": 7.317197104405213e-05, |
| "loss": 11.8254, |
| "num_tokens": 6130750.0, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.6002795248078266, |
| "grad_norm": 0.9036108784233868, |
| "learning_rate": 7.295399782507275e-05, |
| "loss": 11.6364, |
| "num_tokens": 6137850.0, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.6009783368273934, |
| "grad_norm": 0.773889565114804, |
| "learning_rate": 7.273616316157312e-05, |
| "loss": 11.6782, |
| "num_tokens": 6144967.0, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.6016771488469602, |
| "grad_norm": 0.991907637656795, |
| "learning_rate": 7.251846816951063e-05, |
| "loss": 11.6562, |
| "num_tokens": 6151858.0, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.6023759608665269, |
| "grad_norm": 0.8650022175317262, |
| "learning_rate": 7.23009139641271e-05, |
| "loss": 11.7501, |
| "num_tokens": 6158407.0, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.6030747728860937, |
| "grad_norm": 0.7830337647367694, |
| "learning_rate": 7.208350165994325e-05, |
| "loss": 11.9671, |
| "num_tokens": 6165454.0, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.6037735849056604, |
| "grad_norm": 0.9634536848543019, |
| "learning_rate": 7.186623237075265e-05, |
| "loss": 11.718, |
| "num_tokens": 6172278.0, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.6044723969252271, |
| "grad_norm": 0.8009388532348747, |
| "learning_rate": 7.16491072096164e-05, |
| "loss": 11.797, |
| "num_tokens": 6179334.0, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.6051712089447938, |
| "grad_norm": 0.7794753187578644, |
| "learning_rate": 7.143212728885714e-05, |
| "loss": 11.7032, |
| "num_tokens": 6186949.0, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.6058700209643606, |
| "grad_norm": 0.8209269866004077, |
| "learning_rate": 7.121529372005335e-05, |
| "loss": 11.7299, |
| "num_tokens": 6194084.0, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.6065688329839273, |
| "grad_norm": 0.788054239481001, |
| "learning_rate": 7.099860761403403e-05, |
| "loss": 11.8418, |
| "num_tokens": 6201312.0, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.6072676450034941, |
| "grad_norm": 0.7412690116578288, |
| "learning_rate": 7.078207008087248e-05, |
| "loss": 11.7934, |
| "num_tokens": 6208547.0, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.6079664570230608, |
| "grad_norm": 0.7451063882822027, |
| "learning_rate": 7.056568222988099e-05, |
| "loss": 11.8384, |
| "num_tokens": 6215891.0, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.6086652690426275, |
| "grad_norm": 0.7488238631148831, |
| "learning_rate": 7.034944516960498e-05, |
| "loss": 11.7576, |
| "num_tokens": 6223503.0, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.6093640810621943, |
| "grad_norm": 0.7457484829284764, |
| "learning_rate": 7.013336000781738e-05, |
| "loss": 11.8307, |
| "num_tokens": 6230239.0, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.610062893081761, |
| "grad_norm": 0.7550948672495834, |
| "learning_rate": 6.991742785151305e-05, |
| "loss": 11.6902, |
| "num_tokens": 6237658.0, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.6107617051013278, |
| "grad_norm": 0.7831287226929119, |
| "learning_rate": 6.970164980690285e-05, |
| "loss": 11.8591, |
| "num_tokens": 6244558.0, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.6114605171208944, |
| "grad_norm": 0.7615226027016072, |
| "learning_rate": 6.94860269794083e-05, |
| "loss": 11.8389, |
| "num_tokens": 6251785.0, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.6121593291404612, |
| "grad_norm": 0.7467213227226704, |
| "learning_rate": 6.927056047365557e-05, |
| "loss": 11.6384, |
| "num_tokens": 6258748.0, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.6128581411600279, |
| "grad_norm": 0.7521897728411195, |
| "learning_rate": 6.905525139347011e-05, |
| "loss": 11.7597, |
| "num_tokens": 6266062.0, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.6135569531795947, |
| "grad_norm": 0.7647668153792994, |
| "learning_rate": 6.884010084187093e-05, |
| "loss": 11.7313, |
| "num_tokens": 6272639.0, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.6142557651991615, |
| "grad_norm": 0.7902525252819772, |
| "learning_rate": 6.86251099210648e-05, |
| "loss": 11.8622, |
| "num_tokens": 6279204.0, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.6149545772187281, |
| "grad_norm": 0.7303818732013376, |
| "learning_rate": 6.841027973244076e-05, |
| "loss": 11.7259, |
| "num_tokens": 6286372.0, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.6156533892382949, |
| "grad_norm": 0.794393939694085, |
| "learning_rate": 6.819561137656443e-05, |
| "loss": 11.6617, |
| "num_tokens": 6292809.0, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.6163522012578616, |
| "grad_norm": 0.7485246654196547, |
| "learning_rate": 6.798110595317229e-05, |
| "loss": 11.6752, |
| "num_tokens": 6300059.0, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.6170510132774284, |
| "grad_norm": 0.7892650884136205, |
| "learning_rate": 6.776676456116629e-05, |
| "loss": 11.8064, |
| "num_tokens": 6306992.0, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.6177498252969951, |
| "grad_norm": 0.7816686405399219, |
| "learning_rate": 6.755258829860791e-05, |
| "loss": 11.6161, |
| "num_tokens": 6313854.0, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.6184486373165619, |
| "grad_norm": 0.7673507811300475, |
| "learning_rate": 6.733857826271271e-05, |
| "loss": 11.6007, |
| "num_tokens": 6320866.0, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.6191474493361285, |
| "grad_norm": 0.7482768854684733, |
| "learning_rate": 6.712473554984472e-05, |
| "loss": 11.6434, |
| "num_tokens": 6328614.0, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.6198462613556953, |
| "grad_norm": 0.7084508806822576, |
| "learning_rate": 6.69110612555107e-05, |
| "loss": 11.7712, |
| "num_tokens": 6336010.0, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.6205450733752621, |
| "grad_norm": 0.8161348307383346, |
| "learning_rate": 6.669755647435474e-05, |
| "loss": 11.6412, |
| "num_tokens": 6342374.0, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.6212438853948288, |
| "grad_norm": 0.7548031047135115, |
| "learning_rate": 6.648422230015242e-05, |
| "loss": 11.5709, |
| "num_tokens": 6349581.0, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.6219426974143956, |
| "grad_norm": 0.7717987232600103, |
| "learning_rate": 6.627105982580528e-05, |
| "loss": 11.784, |
| "num_tokens": 6356441.0, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.6226415094339622, |
| "grad_norm": 0.7798435061265528, |
| "learning_rate": 6.605807014333538e-05, |
| "loss": 11.646, |
| "num_tokens": 6363148.0, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.623340321453529, |
| "grad_norm": 0.7800404194831362, |
| "learning_rate": 6.584525434387944e-05, |
| "loss": 11.6328, |
| "num_tokens": 6370176.0, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.6240391334730957, |
| "grad_norm": 0.7682712236013818, |
| "learning_rate": 6.563261351768345e-05, |
| "loss": 11.8973, |
| "num_tokens": 6377384.0, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.6247379454926625, |
| "grad_norm": 0.7258219997442539, |
| "learning_rate": 6.542014875409703e-05, |
| "loss": 11.6427, |
| "num_tokens": 6384876.0, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.6254367575122292, |
| "grad_norm": 0.7952066180536073, |
| "learning_rate": 6.52078611415678e-05, |
| "loss": 11.6802, |
| "num_tokens": 6391664.0, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.6261355695317959, |
| "grad_norm": 0.8232647183093804, |
| "learning_rate": 6.49957517676359e-05, |
| "loss": 11.6117, |
| "num_tokens": 6398231.0, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.6268343815513627, |
| "grad_norm": 0.7882657323119439, |
| "learning_rate": 6.47838217189283e-05, |
| "loss": 11.6319, |
| "num_tokens": 6405010.0, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.6275331935709294, |
| "grad_norm": 0.7265020915206697, |
| "learning_rate": 6.457207208115328e-05, |
| "loss": 11.7671, |
| "num_tokens": 6412293.0, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.6282320055904962, |
| "grad_norm": 0.7966290295146391, |
| "learning_rate": 6.436050393909499e-05, |
| "loss": 11.8313, |
| "num_tokens": 6419190.0, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "grad_norm": 0.7395527198077914, |
| "learning_rate": 6.414911837660768e-05, |
| "loss": 11.4508, |
| "num_tokens": 6426719.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.6296296296296297, |
| "grad_norm": 0.7299541853834819, |
| "learning_rate": 6.393791647661032e-05, |
| "loss": 11.8029, |
| "num_tokens": 6434060.0, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.6303284416491963, |
| "grad_norm": 0.7463396559075024, |
| "learning_rate": 6.372689932108091e-05, |
| "loss": 11.7718, |
| "num_tokens": 6441144.0, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.6310272536687631, |
| "grad_norm": 0.7959819255730147, |
| "learning_rate": 6.351606799105107e-05, |
| "loss": 11.545, |
| "num_tokens": 6448000.0, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.6317260656883298, |
| "grad_norm": 0.6790515513892001, |
| "learning_rate": 6.330542356660046e-05, |
| "loss": 11.7301, |
| "num_tokens": 6455854.0, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.6324248777078966, |
| "grad_norm": 0.7426253486979305, |
| "learning_rate": 6.309496712685122e-05, |
| "loss": 11.6712, |
| "num_tokens": 6462687.0, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.6331236897274634, |
| "grad_norm": 0.7600216560906856, |
| "learning_rate": 6.288469974996234e-05, |
| "loss": 11.7371, |
| "num_tokens": 6469302.0, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.63382250174703, |
| "grad_norm": 0.8364647552182698, |
| "learning_rate": 6.267462251312445e-05, |
| "loss": 11.7523, |
| "num_tokens": 6475494.0, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.6345213137665968, |
| "grad_norm": 0.7196566382149244, |
| "learning_rate": 6.24647364925539e-05, |
| "loss": 11.6858, |
| "num_tokens": 6483132.0, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.6352201257861635, |
| "grad_norm": 0.7234513446495888, |
| "learning_rate": 6.225504276348766e-05, |
| "loss": 11.5524, |
| "num_tokens": 6490796.0, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.6359189378057303, |
| "grad_norm": 0.7221735536071997, |
| "learning_rate": 6.204554240017742e-05, |
| "loss": 11.6109, |
| "num_tokens": 6498706.0, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.636617749825297, |
| "grad_norm": 0.7589093046517474, |
| "learning_rate": 6.183623647588427e-05, |
| "loss": 11.8576, |
| "num_tokens": 6505674.0, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.6373165618448637, |
| "grad_norm": 0.8427489792553641, |
| "learning_rate": 6.162712606287335e-05, |
| "loss": 11.6848, |
| "num_tokens": 6512122.0, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.6380153738644304, |
| "grad_norm": 0.7509384033664809, |
| "learning_rate": 6.141821223240804e-05, |
| "loss": 11.6497, |
| "num_tokens": 6519491.0, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.6387141858839972, |
| "grad_norm": 0.7727801776974009, |
| "learning_rate": 6.120949605474478e-05, |
| "loss": 11.7904, |
| "num_tokens": 6526251.0, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.639412997903564, |
| "grad_norm": 0.7698221947163191, |
| "learning_rate": 6.100097859912732e-05, |
| "loss": 11.7356, |
| "num_tokens": 6532851.0, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.6401118099231307, |
| "grad_norm": 0.8484210212133271, |
| "learning_rate": 6.0792660933781375e-05, |
| "loss": 11.6482, |
| "num_tokens": 6539983.0, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.6408106219426974, |
| "grad_norm": 0.8146973477377301, |
| "learning_rate": 6.058454412590928e-05, |
| "loss": 11.682, |
| "num_tokens": 6547267.0, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.6415094339622641, |
| "grad_norm": 0.8066477438364413, |
| "learning_rate": 6.037662924168419e-05, |
| "loss": 11.5587, |
| "num_tokens": 6554111.0, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.6422082459818309, |
| "grad_norm": 0.7961568117356921, |
| "learning_rate": 6.016891734624501e-05, |
| "loss": 11.7357, |
| "num_tokens": 6561069.0, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.6429070580013976, |
| "grad_norm": 0.8791796417982806, |
| "learning_rate": 5.9961409503690605e-05, |
| "loss": 11.884, |
| "num_tokens": 6568238.0, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.6436058700209644, |
| "grad_norm": 0.712988301703953, |
| "learning_rate": 5.975410677707447e-05, |
| "loss": 11.7497, |
| "num_tokens": 6575880.0, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.6443046820405312, |
| "grad_norm": 0.7647107034158884, |
| "learning_rate": 5.954701022839944e-05, |
| "loss": 11.6668, |
| "num_tokens": 6582676.0, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.6450034940600978, |
| "grad_norm": 0.7279408863948003, |
| "learning_rate": 5.9340120918611994e-05, |
| "loss": 11.8926, |
| "num_tokens": 6590130.0, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.6457023060796646, |
| "grad_norm": 0.7858715219906657, |
| "learning_rate": 5.913343990759695e-05, |
| "loss": 11.8346, |
| "num_tokens": 6596815.0, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.6464011180992313, |
| "grad_norm": 0.7662237988937426, |
| "learning_rate": 5.8926968254172076e-05, |
| "loss": 11.3992, |
| "num_tokens": 6604041.0, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.6470999301187981, |
| "grad_norm": 0.7807542174388001, |
| "learning_rate": 5.872070701608251e-05, |
| "loss": 11.7433, |
| "num_tokens": 6611449.0, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.6477987421383647, |
| "grad_norm": 0.7770790305426569, |
| "learning_rate": 5.851465724999559e-05, |
| "loss": 11.8598, |
| "num_tokens": 6618417.0, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.6484975541579315, |
| "grad_norm": 0.7276280775957402, |
| "learning_rate": 5.830882001149517e-05, |
| "loss": 11.6542, |
| "num_tokens": 6626409.0, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.6491963661774982, |
| "grad_norm": 0.7659912850579355, |
| "learning_rate": 5.8103196355076305e-05, |
| "loss": 11.7479, |
| "num_tokens": 6633421.0, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.649895178197065, |
| "grad_norm": 0.7488754867958327, |
| "learning_rate": 5.789778733414004e-05, |
| "loss": 11.681, |
| "num_tokens": 6641387.0, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.6505939902166318, |
| "grad_norm": 0.7426656034656497, |
| "learning_rate": 5.769259400098769e-05, |
| "loss": 11.5628, |
| "num_tokens": 6648880.0, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.6512928022361985, |
| "grad_norm": 0.7485244300975621, |
| "learning_rate": 5.748761740681573e-05, |
| "loss": 11.7034, |
| "num_tokens": 6656518.0, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.6519916142557652, |
| "grad_norm": 0.7280973172346309, |
| "learning_rate": 5.728285860171021e-05, |
| "loss": 11.5834, |
| "num_tokens": 6663873.0, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.6526904262753319, |
| "grad_norm": 0.7690745999643336, |
| "learning_rate": 5.7078318634641456e-05, |
| "loss": 11.7479, |
| "num_tokens": 6671021.0, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.6533892382948987, |
| "grad_norm": 0.8091833500142877, |
| "learning_rate": 5.687399855345879e-05, |
| "loss": 11.7293, |
| "num_tokens": 6677620.0, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.6540880503144654, |
| "grad_norm": 0.7583842125434499, |
| "learning_rate": 5.666989940488496e-05, |
| "loss": 11.6877, |
| "num_tokens": 6685101.0, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.6547868623340322, |
| "grad_norm": 0.7756271660445293, |
| "learning_rate": 5.646602223451094e-05, |
| "loss": 11.5556, |
| "num_tokens": 6692207.0, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.6554856743535988, |
| "grad_norm": 0.732585726312475, |
| "learning_rate": 5.6262368086790504e-05, |
| "loss": 11.6001, |
| "num_tokens": 6699759.0, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.6561844863731656, |
| "grad_norm": 0.7822132210828396, |
| "learning_rate": 5.605893800503484e-05, |
| "loss": 11.5981, |
| "num_tokens": 6706906.0, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.6568832983927324, |
| "grad_norm": 0.80455226020786, |
| "learning_rate": 5.585573303140741e-05, |
| "loss": 11.7701, |
| "num_tokens": 6713394.0, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.6575821104122991, |
| "grad_norm": 0.7515507860338341, |
| "learning_rate": 5.565275420691831e-05, |
| "loss": 11.6546, |
| "num_tokens": 6720211.0, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.6582809224318659, |
| "grad_norm": 0.7279100113231547, |
| "learning_rate": 5.5450002571419104e-05, |
| "loss": 11.8563, |
| "num_tokens": 6727718.0, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.6589797344514325, |
| "grad_norm": 0.740532850900047, |
| "learning_rate": 5.524747916359756e-05, |
| "loss": 11.5629, |
| "num_tokens": 6735027.0, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.6596785464709993, |
| "grad_norm": 0.7190050209844847, |
| "learning_rate": 5.504518502097212e-05, |
| "loss": 11.7309, |
| "num_tokens": 6742667.0, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.660377358490566, |
| "grad_norm": 0.7281631801675564, |
| "learning_rate": 5.484312117988687e-05, |
| "loss": 11.6905, |
| "num_tokens": 6750129.0, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.6610761705101328, |
| "grad_norm": 0.7642082512718011, |
| "learning_rate": 5.464128867550593e-05, |
| "loss": 11.5831, |
| "num_tokens": 6756898.0, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.6617749825296995, |
| "grad_norm": 0.7353524837824559, |
| "learning_rate": 5.4439688541808345e-05, |
| "loss": 11.8927, |
| "num_tokens": 6763921.0, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.6624737945492662, |
| "grad_norm": 0.7252766686868163, |
| "learning_rate": 5.423832181158274e-05, |
| "loss": 11.665, |
| "num_tokens": 6771091.0, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.663172606568833, |
| "grad_norm": 0.7113069997901473, |
| "learning_rate": 5.4037189516422e-05, |
| "loss": 11.5554, |
| "num_tokens": 6778488.0, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.6638714185883997, |
| "grad_norm": 0.7014962329353019, |
| "learning_rate": 5.383629268671804e-05, |
| "loss": 11.5834, |
| "num_tokens": 6785464.0, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.6645702306079665, |
| "grad_norm": 0.7875873029591225, |
| "learning_rate": 5.3635632351656495e-05, |
| "loss": 11.8594, |
| "num_tokens": 6792243.0, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.6652690426275332, |
| "grad_norm": 0.7213066317257657, |
| "learning_rate": 5.3435209539211394e-05, |
| "loss": 11.6927, |
| "num_tokens": 6799493.0, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.6659678546471, |
| "grad_norm": 0.7035828384585064, |
| "learning_rate": 5.323502527614007e-05, |
| "loss": 11.6259, |
| "num_tokens": 6806467.0, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 0.7809434468803235, |
| "learning_rate": 5.303508058797766e-05, |
| "loss": 11.6873, |
| "num_tokens": 6813306.0, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.6673654786862334, |
| "grad_norm": 0.7283465666823379, |
| "learning_rate": 5.2835376499031955e-05, |
| "loss": 11.4497, |
| "num_tokens": 6820518.0, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.6680642907058001, |
| "grad_norm": 0.7617743650609611, |
| "learning_rate": 5.263591403237831e-05, |
| "loss": 11.668, |
| "num_tokens": 6827100.0, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.6687631027253669, |
| "grad_norm": 0.8177076699367816, |
| "learning_rate": 5.243669420985413e-05, |
| "loss": 11.6849, |
| "num_tokens": 6833940.0, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.6694619147449337, |
| "grad_norm": 0.7408475483142893, |
| "learning_rate": 5.22377180520538e-05, |
| "loss": 11.8671, |
| "num_tokens": 6841696.0, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.6701607267645003, |
| "grad_norm": 0.7169014961837786, |
| "learning_rate": 5.2038986578323437e-05, |
| "loss": 11.8537, |
| "num_tokens": 6849308.0, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.6708595387840671, |
| "grad_norm": 0.7580468191959683, |
| "learning_rate": 5.1840500806755575e-05, |
| "loss": 11.5533, |
| "num_tokens": 6856816.0, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.6715583508036338, |
| "grad_norm": 0.748707497098157, |
| "learning_rate": 5.164226175418421e-05, |
| "loss": 11.7469, |
| "num_tokens": 6863209.0, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.6722571628232006, |
| "grad_norm": 0.7525456270124457, |
| "learning_rate": 5.1444270436179185e-05, |
| "loss": 11.4797, |
| "num_tokens": 6869856.0, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.6729559748427673, |
| "grad_norm": 0.7211635471907168, |
| "learning_rate": 5.12465278670414e-05, |
| "loss": 11.6617, |
| "num_tokens": 6877774.0, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.673654786862334, |
| "grad_norm": 0.7553976822609867, |
| "learning_rate": 5.10490350597973e-05, |
| "loss": 11.5848, |
| "num_tokens": 6884547.0, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.6743535988819007, |
| "grad_norm": 0.7022807434186031, |
| "learning_rate": 5.085179302619383e-05, |
| "loss": 11.7242, |
| "num_tokens": 6892114.0, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.6750524109014675, |
| "grad_norm": 0.780693519873508, |
| "learning_rate": 5.06548027766933e-05, |
| "loss": 11.8297, |
| "num_tokens": 6899318.0, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.6757512229210343, |
| "grad_norm": 0.7267581172978977, |
| "learning_rate": 5.045806532046806e-05, |
| "loss": 11.6526, |
| "num_tokens": 6906566.0, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.676450034940601, |
| "grad_norm": 0.7846570944458914, |
| "learning_rate": 5.0261581665395475e-05, |
| "loss": 11.6798, |
| "num_tokens": 6913114.0, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.6771488469601677, |
| "grad_norm": 0.7112948402372895, |
| "learning_rate": 5.006535281805265e-05, |
| "loss": 11.5148, |
| "num_tokens": 6920642.0, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.6778476589797344, |
| "grad_norm": 0.7583944622835155, |
| "learning_rate": 4.9869379783711315e-05, |
| "loss": 11.5235, |
| "num_tokens": 6927787.0, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.6785464709993012, |
| "grad_norm": 0.7573666973845233, |
| "learning_rate": 4.967366356633275e-05, |
| "loss": 11.5615, |
| "num_tokens": 6935406.0, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.6792452830188679, |
| "grad_norm": 0.7227775502190015, |
| "learning_rate": 4.947820516856253e-05, |
| "loss": 11.6955, |
| "num_tokens": 6942968.0, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.6799440950384347, |
| "grad_norm": 0.7814966089422306, |
| "learning_rate": 4.9283005591725375e-05, |
| "loss": 11.752, |
| "num_tokens": 6949948.0, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.6806429070580013, |
| "grad_norm": 0.7175546627631187, |
| "learning_rate": 4.908806583582021e-05, |
| "loss": 11.7433, |
| "num_tokens": 6957454.0, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.6813417190775681, |
| "grad_norm": 0.7517273070787456, |
| "learning_rate": 4.8893386899514746e-05, |
| "loss": 11.6715, |
| "num_tokens": 6964979.0, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.6820405310971349, |
| "grad_norm": 0.669368555621721, |
| "learning_rate": 4.869896978014071e-05, |
| "loss": 11.6769, |
| "num_tokens": 6972329.0, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.6827393431167016, |
| "grad_norm": 0.7504037459255867, |
| "learning_rate": 4.85048154736884e-05, |
| "loss": 11.7083, |
| "num_tokens": 6978773.0, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.6834381551362684, |
| "grad_norm": 0.7077905666186198, |
| "learning_rate": 4.831092497480179e-05, |
| "loss": 11.6217, |
| "num_tokens": 6986336.0, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.684136967155835, |
| "grad_norm": 0.8637933521061518, |
| "learning_rate": 4.81172992767734e-05, |
| "loss": 11.6379, |
| "num_tokens": 6992944.0, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.6848357791754018, |
| "grad_norm": 0.7721626875146355, |
| "learning_rate": 4.792393937153914e-05, |
| "loss": 11.7463, |
| "num_tokens": 6999683.0, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.6855345911949685, |
| "grad_norm": 0.7117222748710471, |
| "learning_rate": 4.773084624967327e-05, |
| "loss": 11.5521, |
| "num_tokens": 7006810.0, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.6862334032145353, |
| "grad_norm": 0.7539774224870581, |
| "learning_rate": 4.753802090038344e-05, |
| "loss": 11.8959, |
| "num_tokens": 7013925.0, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.686932215234102, |
| "grad_norm": 0.8327203871710336, |
| "learning_rate": 4.734546431150536e-05, |
| "loss": 11.8562, |
| "num_tokens": 7020571.0, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.6876310272536688, |
| "grad_norm": 0.7708657591442682, |
| "learning_rate": 4.715317746949804e-05, |
| "loss": 11.8254, |
| "num_tokens": 7027401.0, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.6883298392732355, |
| "grad_norm": 0.7258834917462351, |
| "learning_rate": 4.6961161359438486e-05, |
| "loss": 11.4523, |
| "num_tokens": 7034682.0, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.6890286512928022, |
| "grad_norm": 0.7330377602083751, |
| "learning_rate": 4.676941696501673e-05, |
| "loss": 11.5904, |
| "num_tokens": 7042175.0, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.689727463312369, |
| "grad_norm": 0.8028105455984438, |
| "learning_rate": 4.657794526853096e-05, |
| "loss": 11.6372, |
| "num_tokens": 7048976.0, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.6904262753319357, |
| "grad_norm": 0.7215274112971415, |
| "learning_rate": 4.6386747250882224e-05, |
| "loss": 11.6798, |
| "num_tokens": 7056210.0, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.6911250873515025, |
| "grad_norm": 0.7769719500087231, |
| "learning_rate": 4.6195823891569545e-05, |
| "loss": 11.8147, |
| "num_tokens": 7063291.0, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.6918238993710691, |
| "grad_norm": 0.7633412054706588, |
| "learning_rate": 4.60051761686849e-05, |
| "loss": 11.7149, |
| "num_tokens": 7070074.0, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.6925227113906359, |
| "grad_norm": 0.7299361913347738, |
| "learning_rate": 4.581480505890816e-05, |
| "loss": 11.5324, |
| "num_tokens": 7077328.0, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.6932215234102026, |
| "grad_norm": 0.7373024767213333, |
| "learning_rate": 4.5624711537502206e-05, |
| "loss": 11.7304, |
| "num_tokens": 7084251.0, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.6939203354297694, |
| "grad_norm": 0.7830459462760436, |
| "learning_rate": 4.543489657830777e-05, |
| "loss": 11.6188, |
| "num_tokens": 7091370.0, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.6946191474493362, |
| "grad_norm": 0.8182215566071669, |
| "learning_rate": 4.52453611537385e-05, |
| "loss": 11.6924, |
| "num_tokens": 7097992.0, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.6953179594689028, |
| "grad_norm": 0.7946888243423976, |
| "learning_rate": 4.505610623477611e-05, |
| "loss": 11.6498, |
| "num_tokens": 7104219.0, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.6960167714884696, |
| "grad_norm": 0.7549468174659378, |
| "learning_rate": 4.486713279096515e-05, |
| "loss": 11.5893, |
| "num_tokens": 7110986.0, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.6967155835080363, |
| "grad_norm": 0.6921174078679769, |
| "learning_rate": 4.4678441790408335e-05, |
| "loss": 11.7298, |
| "num_tokens": 7118553.0, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.6974143955276031, |
| "grad_norm": 0.7613107064310906, |
| "learning_rate": 4.449003419976133e-05, |
| "loss": 11.4912, |
| "num_tokens": 7125671.0, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.6981132075471698, |
| "grad_norm": 0.7419182676475446, |
| "learning_rate": 4.430191098422795e-05, |
| "loss": 11.4444, |
| "num_tokens": 7132247.0, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.6988120195667366, |
| "grad_norm": 0.8116675732612285, |
| "learning_rate": 4.411407310755513e-05, |
| "loss": 11.7335, |
| "num_tokens": 7138298.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.6995108315863033, |
| "grad_norm": 0.7569228200974064, |
| "learning_rate": 4.392652153202802e-05, |
| "loss": 11.7441, |
| "num_tokens": 7144756.0, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.70020964360587, |
| "grad_norm": 0.7200908962994922, |
| "learning_rate": 4.373925721846519e-05, |
| "loss": 11.6201, |
| "num_tokens": 7152146.0, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.7009084556254368, |
| "grad_norm": 0.6993029376311879, |
| "learning_rate": 4.355228112621341e-05, |
| "loss": 11.6431, |
| "num_tokens": 7159343.0, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.7016072676450035, |
| "grad_norm": 0.7461449573644776, |
| "learning_rate": 4.336559421314298e-05, |
| "loss": 11.6424, |
| "num_tokens": 7165846.0, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.7023060796645703, |
| "grad_norm": 0.7485912923255708, |
| "learning_rate": 4.317919743564278e-05, |
| "loss": 11.7013, |
| "num_tokens": 7173032.0, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.7030048916841369, |
| "grad_norm": 0.7444500724716369, |
| "learning_rate": 4.29930917486153e-05, |
| "loss": 11.8321, |
| "num_tokens": 7179676.0, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.7037037037037037, |
| "grad_norm": 0.7172721105055976, |
| "learning_rate": 4.2807278105471735e-05, |
| "loss": 11.551, |
| "num_tokens": 7187102.0, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.7044025157232704, |
| "grad_norm": 0.7075647979765408, |
| "learning_rate": 4.2621757458127285e-05, |
| "loss": 11.6526, |
| "num_tokens": 7194233.0, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.7051013277428372, |
| "grad_norm": 0.7356820423823094, |
| "learning_rate": 4.243653075699604e-05, |
| "loss": 11.6294, |
| "num_tokens": 7200796.0, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.705800139762404, |
| "grad_norm": 0.7823468359047476, |
| "learning_rate": 4.2251598950986226e-05, |
| "loss": 11.678, |
| "num_tokens": 7207142.0, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.7064989517819706, |
| "grad_norm": 0.764640535949318, |
| "learning_rate": 4.2066962987495376e-05, |
| "loss": 11.7764, |
| "num_tokens": 7213920.0, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.7071977638015374, |
| "grad_norm": 0.7037309726706639, |
| "learning_rate": 4.188262381240534e-05, |
| "loss": 11.8543, |
| "num_tokens": 7221271.0, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.7078965758211041, |
| "grad_norm": 0.6676818416659133, |
| "learning_rate": 4.169858237007772e-05, |
| "loss": 11.6132, |
| "num_tokens": 7229149.0, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.7085953878406709, |
| "grad_norm": 0.7253896868171327, |
| "learning_rate": 4.151483960334862e-05, |
| "loss": 11.5597, |
| "num_tokens": 7236889.0, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.7092941998602376, |
| "grad_norm": 0.7392536296757929, |
| "learning_rate": 4.133139645352425e-05, |
| "loss": 11.7499, |
| "num_tokens": 7243535.0, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.7099930118798043, |
| "grad_norm": 0.7859312545956098, |
| "learning_rate": 4.114825386037576e-05, |
| "loss": 11.8108, |
| "num_tokens": 7249939.0, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.710691823899371, |
| "grad_norm": 0.6840777175495678, |
| "learning_rate": 4.0965412762134556e-05, |
| "loss": 11.6202, |
| "num_tokens": 7257993.0, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.7113906359189378, |
| "grad_norm": 0.7796568358822769, |
| "learning_rate": 4.078287409548763e-05, |
| "loss": 11.7883, |
| "num_tokens": 7264252.0, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.7120894479385046, |
| "grad_norm": 0.7228198060469493, |
| "learning_rate": 4.060063879557249e-05, |
| "loss": 11.4887, |
| "num_tokens": 7271420.0, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.7127882599580713, |
| "grad_norm": 0.730805955091614, |
| "learning_rate": 4.0418707795972574e-05, |
| "loss": 11.8562, |
| "num_tokens": 7278894.0, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.713487071977638, |
| "grad_norm": 0.7206633340955824, |
| "learning_rate": 4.023708202871239e-05, |
| "loss": 11.494, |
| "num_tokens": 7286940.0, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.7141858839972047, |
| "grad_norm": 0.7074645446742183, |
| "learning_rate": 4.005576242425272e-05, |
| "loss": 11.2935, |
| "num_tokens": 7294745.0, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.7148846960167715, |
| "grad_norm": 0.7320420207557593, |
| "learning_rate": 3.9874749911485995e-05, |
| "loss": 11.6363, |
| "num_tokens": 7301959.0, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.7155835080363382, |
| "grad_norm": 0.771063008702391, |
| "learning_rate": 3.969404541773132e-05, |
| "loss": 11.5651, |
| "num_tokens": 7308592.0, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.716282320055905, |
| "grad_norm": 0.7590205282163587, |
| "learning_rate": 3.951364986872984e-05, |
| "loss": 11.589, |
| "num_tokens": 7315715.0, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.7169811320754716, |
| "grad_norm": 0.7763995366489977, |
| "learning_rate": 3.933356418864008e-05, |
| "loss": 11.6557, |
| "num_tokens": 7322229.0, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.7176799440950384, |
| "grad_norm": 0.7390841816730954, |
| "learning_rate": 3.9153789300033e-05, |
| "loss": 11.7646, |
| "num_tokens": 7329104.0, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.7183787561146052, |
| "grad_norm": 0.6928548796948936, |
| "learning_rate": 3.8974326123887515e-05, |
| "loss": 11.4572, |
| "num_tokens": 7336700.0, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.7190775681341719, |
| "grad_norm": 0.7395323821851384, |
| "learning_rate": 3.879517557958554e-05, |
| "loss": 11.8653, |
| "num_tokens": 7343801.0, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.7197763801537387, |
| "grad_norm": 0.7538585339891928, |
| "learning_rate": 3.861633858490745e-05, |
| "loss": 11.7817, |
| "num_tokens": 7350746.0, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.7204751921733054, |
| "grad_norm": 0.6744251530070715, |
| "learning_rate": 3.8437816056027296e-05, |
| "loss": 11.4897, |
| "num_tokens": 7358560.0, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.7211740041928721, |
| "grad_norm": 0.7512447279022534, |
| "learning_rate": 3.82596089075081e-05, |
| "loss": 11.9019, |
| "num_tokens": 7365191.0, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.7218728162124388, |
| "grad_norm": 0.7267257258570513, |
| "learning_rate": 3.808171805229733e-05, |
| "loss": 11.687, |
| "num_tokens": 7372368.0, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.7225716282320056, |
| "grad_norm": 0.7326891371575754, |
| "learning_rate": 3.790414440172197e-05, |
| "loss": 11.6076, |
| "num_tokens": 7379881.0, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.7232704402515723, |
| "grad_norm": 0.7370905158754345, |
| "learning_rate": 3.7726888865484e-05, |
| "loss": 11.4954, |
| "num_tokens": 7386744.0, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.7239692522711391, |
| "grad_norm": 0.6431024730369418, |
| "learning_rate": 3.754995235165579e-05, |
| "loss": 11.4018, |
| "num_tokens": 7394571.0, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.7246680642907058, |
| "grad_norm": 0.6783758768249005, |
| "learning_rate": 3.73733357666753e-05, |
| "loss": 11.5327, |
| "num_tokens": 7402101.0, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.7253668763102725, |
| "grad_norm": 0.7079863175153168, |
| "learning_rate": 3.719704001534149e-05, |
| "loss": 11.6665, |
| "num_tokens": 7409120.0, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.7260656883298393, |
| "grad_norm": 0.7085916478726362, |
| "learning_rate": 3.702106600080979e-05, |
| "loss": 11.7719, |
| "num_tokens": 7416318.0, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.726764500349406, |
| "grad_norm": 0.7136983516629083, |
| "learning_rate": 3.6845414624587326e-05, |
| "loss": 11.6003, |
| "num_tokens": 7423580.0, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.7274633123689728, |
| "grad_norm": 0.6870755578878269, |
| "learning_rate": 3.667008678652837e-05, |
| "loss": 11.5949, |
| "num_tokens": 7431059.0, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.7281621243885394, |
| "grad_norm": 0.6859117308717101, |
| "learning_rate": 3.6495083384829723e-05, |
| "loss": 11.5083, |
| "num_tokens": 7438722.0, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.7288609364081062, |
| "grad_norm": 0.7632867590575223, |
| "learning_rate": 3.6320405316026074e-05, |
| "loss": 11.9206, |
| "num_tokens": 7445317.0, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.7295597484276729, |
| "grad_norm": 0.750633699481274, |
| "learning_rate": 3.6146053474985564e-05, |
| "loss": 11.8181, |
| "num_tokens": 7452194.0, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.7302585604472397, |
| "grad_norm": 0.6731058653737066, |
| "learning_rate": 3.597202875490494e-05, |
| "loss": 11.5177, |
| "num_tokens": 7460052.0, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.7309573724668065, |
| "grad_norm": 0.7296995993926372, |
| "learning_rate": 3.579833204730525e-05, |
| "loss": 11.66, |
| "num_tokens": 7467357.0, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.7316561844863732, |
| "grad_norm": 0.7563581339981545, |
| "learning_rate": 3.562496424202707e-05, |
| "loss": 11.6336, |
| "num_tokens": 7474027.0, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.7323549965059399, |
| "grad_norm": 0.7152448422103289, |
| "learning_rate": 3.5451926227225997e-05, |
| "loss": 11.5833, |
| "num_tokens": 7481232.0, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.7330538085255066, |
| "grad_norm": 0.7233352244440772, |
| "learning_rate": 3.5279218889368225e-05, |
| "loss": 11.6259, |
| "num_tokens": 7487946.0, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.7337526205450734, |
| "grad_norm": 0.7201818429672866, |
| "learning_rate": 3.5106843113225854e-05, |
| "loss": 11.6353, |
| "num_tokens": 7494713.0, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.7344514325646401, |
| "grad_norm": 0.6925680193359807, |
| "learning_rate": 3.493479978187236e-05, |
| "loss": 11.6771, |
| "num_tokens": 7502002.0, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.7351502445842069, |
| "grad_norm": 0.7114375668835355, |
| "learning_rate": 3.4763089776678203e-05, |
| "loss": 11.6443, |
| "num_tokens": 7509025.0, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.7358490566037735, |
| "grad_norm": 0.7602494623410806, |
| "learning_rate": 3.459171397730614e-05, |
| "loss": 11.6227, |
| "num_tokens": 7515456.0, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.7365478686233403, |
| "grad_norm": 0.7069387502451085, |
| "learning_rate": 3.44206732617069e-05, |
| "loss": 11.6711, |
| "num_tokens": 7522493.0, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.7372466806429071, |
| "grad_norm": 0.7502241449288016, |
| "learning_rate": 3.424996850611455e-05, |
| "loss": 11.4776, |
| "num_tokens": 7529387.0, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.7379454926624738, |
| "grad_norm": 0.6835406817242182, |
| "learning_rate": 3.4079600585041996e-05, |
| "loss": 11.1981, |
| "num_tokens": 7537726.0, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.7386443046820406, |
| "grad_norm": 0.6792540808215116, |
| "learning_rate": 3.3909570371276654e-05, |
| "loss": 11.7134, |
| "num_tokens": 7545113.0, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.7393431167016072, |
| "grad_norm": 0.773711812243068, |
| "learning_rate": 3.3739878735875796e-05, |
| "loss": 11.7631, |
| "num_tokens": 7552056.0, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.740041928721174, |
| "grad_norm": 0.6829312703518382, |
| "learning_rate": 3.357052654816225e-05, |
| "loss": 11.5149, |
| "num_tokens": 7559863.0, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.7407407407407407, |
| "grad_norm": 0.703661171294156, |
| "learning_rate": 3.3401514675719816e-05, |
| "loss": 11.7388, |
| "num_tokens": 7567323.0, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.7414395527603075, |
| "grad_norm": 0.7508429715088617, |
| "learning_rate": 3.323284398438886e-05, |
| "loss": 11.5236, |
| "num_tokens": 7573959.0, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.7421383647798742, |
| "grad_norm": 0.7168308854044466, |
| "learning_rate": 3.306451533826194e-05, |
| "loss": 11.8019, |
| "num_tokens": 7580996.0, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.742837176799441, |
| "grad_norm": 0.734464873386508, |
| "learning_rate": 3.289652959967925e-05, |
| "loss": 11.8128, |
| "num_tokens": 7587525.0, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.7435359888190077, |
| "grad_norm": 0.6971831147564971, |
| "learning_rate": 3.272888762922442e-05, |
| "loss": 11.7126, |
| "num_tokens": 7595094.0, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.7442348008385744, |
| "grad_norm": 0.7113897493331813, |
| "learning_rate": 3.2561590285719856e-05, |
| "loss": 11.5176, |
| "num_tokens": 7602618.0, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.7449336128581412, |
| "grad_norm": 0.6655971222584358, |
| "learning_rate": 3.2394638426222467e-05, |
| "loss": 11.4715, |
| "num_tokens": 7610143.0, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.7456324248777079, |
| "grad_norm": 0.6764218802393406, |
| "learning_rate": 3.222803290601934e-05, |
| "loss": 11.5741, |
| "num_tokens": 7617786.0, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.7463312368972747, |
| "grad_norm": 0.7070245163442258, |
| "learning_rate": 3.20617745786232e-05, |
| "loss": 11.4435, |
| "num_tokens": 7624874.0, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.7470300489168413, |
| "grad_norm": 0.7230608548627172, |
| "learning_rate": 3.189586429576812e-05, |
| "loss": 11.622, |
| "num_tokens": 7631644.0, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.7477288609364081, |
| "grad_norm": 0.833660300984025, |
| "learning_rate": 3.173030290740524e-05, |
| "loss": 11.7174, |
| "num_tokens": 7637610.0, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.7484276729559748, |
| "grad_norm": 0.6958777713086013, |
| "learning_rate": 3.1565091261698245e-05, |
| "loss": 11.4735, |
| "num_tokens": 7644746.0, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.7491264849755416, |
| "grad_norm": 0.7304731992070302, |
| "learning_rate": 3.140023020501912e-05, |
| "loss": 11.4508, |
| "num_tokens": 7651980.0, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.7498252969951084, |
| "grad_norm": 0.7441043526708394, |
| "learning_rate": 3.1235720581943827e-05, |
| "loss": 11.6389, |
| "num_tokens": 7658633.0, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.750524109014675, |
| "grad_norm": 0.6853622284268183, |
| "learning_rate": 3.107156323524788e-05, |
| "loss": 11.3994, |
| "num_tokens": 7666388.0, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.7512229210342418, |
| "grad_norm": 0.773787030607795, |
| "learning_rate": 3.0907759005902224e-05, |
| "loss": 11.6966, |
| "num_tokens": 7673401.0, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.7519217330538085, |
| "grad_norm": 0.7387364189450014, |
| "learning_rate": 3.074430873306865e-05, |
| "loss": 11.6986, |
| "num_tokens": 7680465.0, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.7526205450733753, |
| "grad_norm": 0.7153460725520698, |
| "learning_rate": 3.058121325409579e-05, |
| "loss": 11.6219, |
| "num_tokens": 7687681.0, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.753319357092942, |
| "grad_norm": 0.7069130545875086, |
| "learning_rate": 3.041847340451456e-05, |
| "loss": 11.6594, |
| "num_tokens": 7694753.0, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.7540181691125087, |
| "grad_norm": 0.7643919239299855, |
| "learning_rate": 3.0256090018034046e-05, |
| "loss": 11.6039, |
| "num_tokens": 7700964.0, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "grad_norm": 0.7487623121730519, |
| "learning_rate": 3.009406392653723e-05, |
| "loss": 11.5454, |
| "num_tokens": 7707592.0, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.7554157931516422, |
| "grad_norm": 0.6759735347117747, |
| "learning_rate": 2.993239596007669e-05, |
| "loss": 11.486, |
| "num_tokens": 7715404.0, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.756114605171209, |
| "grad_norm": 0.688840054711059, |
| "learning_rate": 2.9771086946870175e-05, |
| "loss": 11.4369, |
| "num_tokens": 7722495.0, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.7568134171907757, |
| "grad_norm": 0.8140590429556787, |
| "learning_rate": 2.9610137713296783e-05, |
| "loss": 11.8346, |
| "num_tokens": 7728379.0, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.7575122292103424, |
| "grad_norm": 0.7365687944687868, |
| "learning_rate": 2.9449549083892292e-05, |
| "loss": 11.751, |
| "num_tokens": 7735201.0, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.7582110412299091, |
| "grad_norm": 0.8133640666754306, |
| "learning_rate": 2.9289321881345254e-05, |
| "loss": 11.928, |
| "num_tokens": 7741618.0, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.7589098532494759, |
| "grad_norm": 0.7657442876889771, |
| "learning_rate": 2.9129456926492548e-05, |
| "loss": 11.4964, |
| "num_tokens": 7748854.0, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.7596086652690426, |
| "grad_norm": 0.6907962618994248, |
| "learning_rate": 2.8969955038315277e-05, |
| "loss": 11.4488, |
| "num_tokens": 7756291.0, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.7603074772886094, |
| "grad_norm": 0.7152174671204918, |
| "learning_rate": 2.8810817033934656e-05, |
| "loss": 11.4593, |
| "num_tokens": 7763508.0, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.7610062893081762, |
| "grad_norm": 0.6975589242543092, |
| "learning_rate": 2.8652043728607625e-05, |
| "loss": 11.6335, |
| "num_tokens": 7771093.0, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.7617051013277428, |
| "grad_norm": 0.7668716962872764, |
| "learning_rate": 2.8493635935722928e-05, |
| "loss": 11.5716, |
| "num_tokens": 7777723.0, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.7624039133473096, |
| "grad_norm": 0.7100709643860577, |
| "learning_rate": 2.8335594466796656e-05, |
| "loss": 11.5777, |
| "num_tokens": 7784753.0, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.7631027253668763, |
| "grad_norm": 0.7708747787804404, |
| "learning_rate": 2.8177920131468273e-05, |
| "loss": 11.8262, |
| "num_tokens": 7791905.0, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.7638015373864431, |
| "grad_norm": 0.7096698780955707, |
| "learning_rate": 2.8020613737496547e-05, |
| "loss": 11.6567, |
| "num_tokens": 7799579.0, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.7645003494060097, |
| "grad_norm": 0.7255469439275377, |
| "learning_rate": 2.7863676090755176e-05, |
| "loss": 11.6891, |
| "num_tokens": 7806386.0, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.7651991614255765, |
| "grad_norm": 0.6472085969452891, |
| "learning_rate": 2.770710799522879e-05, |
| "loss": 11.4972, |
| "num_tokens": 7814116.0, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.7658979734451432, |
| "grad_norm": 0.7602480524843874, |
| "learning_rate": 2.7550910253008933e-05, |
| "loss": 11.5528, |
| "num_tokens": 7820763.0, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.76659678546471, |
| "grad_norm": 0.7067899715900233, |
| "learning_rate": 2.739508366428969e-05, |
| "loss": 11.5087, |
| "num_tokens": 7828321.0, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.7672955974842768, |
| "grad_norm": 0.7107532648880457, |
| "learning_rate": 2.723962902736389e-05, |
| "loss": 11.5416, |
| "num_tokens": 7835798.0, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.7679944095038435, |
| "grad_norm": 0.7031234959761055, |
| "learning_rate": 2.7084547138618778e-05, |
| "loss": 11.614, |
| "num_tokens": 7842754.0, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.7686932215234102, |
| "grad_norm": 0.7256785231783027, |
| "learning_rate": 2.6929838792532037e-05, |
| "loss": 11.4756, |
| "num_tokens": 7849613.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.7693920335429769, |
| "grad_norm": 0.711318035418091, |
| "learning_rate": 2.6775504781667725e-05, |
| "loss": 11.6399, |
| "num_tokens": 7856848.0, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.7700908455625437, |
| "grad_norm": 0.7037219078316764, |
| "learning_rate": 2.6621545896672174e-05, |
| "loss": 11.6242, |
| "num_tokens": 7863992.0, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.7707896575821104, |
| "grad_norm": 0.6745280912886404, |
| "learning_rate": 2.6467962926270017e-05, |
| "loss": 11.5943, |
| "num_tokens": 7872390.0, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.7714884696016772, |
| "grad_norm": 0.7213064153858593, |
| "learning_rate": 2.6314756657260054e-05, |
| "loss": 11.5445, |
| "num_tokens": 7879165.0, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.7721872816212438, |
| "grad_norm": 0.7348100469235388, |
| "learning_rate": 2.6161927874511216e-05, |
| "loss": 11.4994, |
| "num_tokens": 7886459.0, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.7728860936408106, |
| "grad_norm": 0.6935894440933058, |
| "learning_rate": 2.6009477360958712e-05, |
| "loss": 11.6117, |
| "num_tokens": 7893611.0, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.7735849056603774, |
| "grad_norm": 0.6940669280906595, |
| "learning_rate": 2.585740589759976e-05, |
| "loss": 11.4804, |
| "num_tokens": 7900850.0, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.7742837176799441, |
| "grad_norm": 0.751295238412207, |
| "learning_rate": 2.5705714263489776e-05, |
| "loss": 11.5352, |
| "num_tokens": 7907796.0, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.7749825296995109, |
| "grad_norm": 0.7391047262828542, |
| "learning_rate": 2.555440323573839e-05, |
| "loss": 11.708, |
| "num_tokens": 7915024.0, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.7756813417190775, |
| "grad_norm": 0.8103062577331461, |
| "learning_rate": 2.540347358950529e-05, |
| "loss": 11.4636, |
| "num_tokens": 7921430.0, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.7763801537386443, |
| "grad_norm": 0.7421607619269197, |
| "learning_rate": 2.5252926097996445e-05, |
| "loss": 11.7304, |
| "num_tokens": 7928272.0, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.777078965758211, |
| "grad_norm": 0.7719958048167995, |
| "learning_rate": 2.5102761532460008e-05, |
| "loss": 11.6926, |
| "num_tokens": 7934944.0, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.7777777777777778, |
| "grad_norm": 0.7401622976674673, |
| "learning_rate": 2.4952980662182425e-05, |
| "loss": 11.4233, |
| "num_tokens": 7941339.0, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.7784765897973445, |
| "grad_norm": 0.7248195083161828, |
| "learning_rate": 2.4803584254484568e-05, |
| "loss": 11.6312, |
| "num_tokens": 7948332.0, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.7791754018169113, |
| "grad_norm": 0.732259203116434, |
| "learning_rate": 2.4654573074717602e-05, |
| "loss": 11.412, |
| "num_tokens": 7955689.0, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.779874213836478, |
| "grad_norm": 0.7059818698633936, |
| "learning_rate": 2.4505947886259318e-05, |
| "loss": 11.742, |
| "num_tokens": 7962958.0, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.7805730258560447, |
| "grad_norm": 0.6751946136877488, |
| "learning_rate": 2.435770945050997e-05, |
| "loss": 11.3857, |
| "num_tokens": 7970832.0, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.7812718378756115, |
| "grad_norm": 0.7509321728923618, |
| "learning_rate": 2.420985852688854e-05, |
| "loss": 11.7448, |
| "num_tokens": 7977881.0, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.7819706498951782, |
| "grad_norm": 0.6561745904495563, |
| "learning_rate": 2.4062395872828846e-05, |
| "loss": 11.531, |
| "num_tokens": 7985659.0, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.782669461914745, |
| "grad_norm": 0.6810374482460928, |
| "learning_rate": 2.3915322243775562e-05, |
| "loss": 11.511, |
| "num_tokens": 7993219.0, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.7833682739343116, |
| "grad_norm": 0.6896502817674913, |
| "learning_rate": 2.3768638393180407e-05, |
| "loss": 11.5836, |
| "num_tokens": 8000242.0, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.7840670859538784, |
| "grad_norm": 0.7145964062311239, |
| "learning_rate": 2.362234507249832e-05, |
| "loss": 11.6156, |
| "num_tokens": 8007589.0, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.7847658979734451, |
| "grad_norm": 0.6651730755964388, |
| "learning_rate": 2.3476443031183503e-05, |
| "loss": 11.6697, |
| "num_tokens": 8015113.0, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.7854647099930119, |
| "grad_norm": 0.730162607136164, |
| "learning_rate": 2.3330933016685754e-05, |
| "loss": 11.6406, |
| "num_tokens": 8021703.0, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.7861635220125787, |
| "grad_norm": 0.706585916019649, |
| "learning_rate": 2.318581577444646e-05, |
| "loss": 11.5674, |
| "num_tokens": 8028563.0, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.7868623340321453, |
| "grad_norm": 0.7261019540757417, |
| "learning_rate": 2.304109204789484e-05, |
| "loss": 11.6464, |
| "num_tokens": 8035481.0, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.7875611460517121, |
| "grad_norm": 0.7230591047259078, |
| "learning_rate": 2.289676257844423e-05, |
| "loss": 11.7388, |
| "num_tokens": 8042374.0, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.7882599580712788, |
| "grad_norm": 0.6710460854679413, |
| "learning_rate": 2.275282810548811e-05, |
| "loss": 11.5265, |
| "num_tokens": 8050149.0, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.7889587700908456, |
| "grad_norm": 0.7303492099047564, |
| "learning_rate": 2.2609289366396502e-05, |
| "loss": 11.7182, |
| "num_tokens": 8056810.0, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.7896575821104123, |
| "grad_norm": 0.7108633933438501, |
| "learning_rate": 2.2466147096512035e-05, |
| "loss": 11.8337, |
| "num_tokens": 8063712.0, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.790356394129979, |
| "grad_norm": 0.7013257263082335, |
| "learning_rate": 2.2323402029146244e-05, |
| "loss": 11.5153, |
| "num_tokens": 8070844.0, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.7910552061495457, |
| "grad_norm": 0.6692578889529726, |
| "learning_rate": 2.2181054895575847e-05, |
| "loss": 11.4817, |
| "num_tokens": 8078320.0, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.7917540181691125, |
| "grad_norm": 0.6722679420030302, |
| "learning_rate": 2.2039106425038924e-05, |
| "loss": 11.7263, |
| "num_tokens": 8085826.0, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.7924528301886793, |
| "grad_norm": 0.7022394162218761, |
| "learning_rate": 2.189755734473129e-05, |
| "loss": 11.6687, |
| "num_tokens": 8092777.0, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.793151642208246, |
| "grad_norm": 0.8000656873886911, |
| "learning_rate": 2.175640837980265e-05, |
| "loss": 11.8119, |
| "num_tokens": 8099130.0, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.7938504542278128, |
| "grad_norm": 0.781499895574501, |
| "learning_rate": 2.161566025335289e-05, |
| "loss": 11.6872, |
| "num_tokens": 8105524.0, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.7945492662473794, |
| "grad_norm": 0.7074827094878331, |
| "learning_rate": 2.1475313686428544e-05, |
| "loss": 11.4998, |
| "num_tokens": 8112500.0, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.7952480782669462, |
| "grad_norm": 0.7135705955177107, |
| "learning_rate": 2.133536939801888e-05, |
| "loss": 11.6275, |
| "num_tokens": 8119456.0, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.7959468902865129, |
| "grad_norm": 0.7271071660736487, |
| "learning_rate": 2.1195828105052283e-05, |
| "loss": 11.5599, |
| "num_tokens": 8126211.0, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.7966457023060797, |
| "grad_norm": 0.757471440304504, |
| "learning_rate": 2.105669052239274e-05, |
| "loss": 11.605, |
| "num_tokens": 8132659.0, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.7973445143256463, |
| "grad_norm": 0.731869331936842, |
| "learning_rate": 2.091795736283593e-05, |
| "loss": 11.7303, |
| "num_tokens": 8139414.0, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.7980433263452131, |
| "grad_norm": 0.7140975321778861, |
| "learning_rate": 2.0779629337105722e-05, |
| "loss": 11.413, |
| "num_tokens": 8146787.0, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.7987421383647799, |
| "grad_norm": 0.7644539350392501, |
| "learning_rate": 2.064170715385052e-05, |
| "loss": 11.5113, |
| "num_tokens": 8153506.0, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.7994409503843466, |
| "grad_norm": 0.65418077450009, |
| "learning_rate": 2.050419151963957e-05, |
| "loss": 11.7101, |
| "num_tokens": 8161093.0, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.8001397624039134, |
| "grad_norm": 0.7413639644128542, |
| "learning_rate": 2.0367083138959476e-05, |
| "loss": 11.6071, |
| "num_tokens": 8167975.0, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.80083857442348, |
| "grad_norm": 0.7635154540705584, |
| "learning_rate": 2.0230382714210384e-05, |
| "loss": 11.6118, |
| "num_tokens": 8174799.0, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.8015373864430468, |
| "grad_norm": 0.716235149002391, |
| "learning_rate": 2.0094090945702616e-05, |
| "loss": 11.6827, |
| "num_tokens": 8181864.0, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.8022361984626135, |
| "grad_norm": 0.6800488359307103, |
| "learning_rate": 1.9958208531652877e-05, |
| "loss": 11.6604, |
| "num_tokens": 8188903.0, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.8029350104821803, |
| "grad_norm": 0.6393895545972774, |
| "learning_rate": 1.9822736168180778e-05, |
| "loss": 11.641, |
| "num_tokens": 8196709.0, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.803633822501747, |
| "grad_norm": 0.6983448761591593, |
| "learning_rate": 1.9687674549305335e-05, |
| "loss": 11.6795, |
| "num_tokens": 8204133.0, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.8043326345213138, |
| "grad_norm": 0.659945277737391, |
| "learning_rate": 1.9553024366941242e-05, |
| "loss": 11.5617, |
| "num_tokens": 8211596.0, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.8050314465408805, |
| "grad_norm": 0.7114940504611794, |
| "learning_rate": 1.9418786310895464e-05, |
| "loss": 11.677, |
| "num_tokens": 8218648.0, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.8057302585604472, |
| "grad_norm": 0.6498997833303469, |
| "learning_rate": 1.9284961068863673e-05, |
| "loss": 11.5807, |
| "num_tokens": 8226878.0, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.806429070580014, |
| "grad_norm": 0.6851026198320295, |
| "learning_rate": 1.9151549326426656e-05, |
| "loss": 11.6532, |
| "num_tokens": 8234332.0, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.8071278825995807, |
| "grad_norm": 0.7304658584255903, |
| "learning_rate": 1.9018551767046966e-05, |
| "loss": 11.4973, |
| "num_tokens": 8241077.0, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.8078266946191475, |
| "grad_norm": 0.7013604887509809, |
| "learning_rate": 1.8885969072065225e-05, |
| "loss": 11.421, |
| "num_tokens": 8248516.0, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.8085255066387141, |
| "grad_norm": 0.7053816647394219, |
| "learning_rate": 1.8753801920696712e-05, |
| "loss": 11.6213, |
| "num_tokens": 8255076.0, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.8092243186582809, |
| "grad_norm": 0.6535806078494023, |
| "learning_rate": 1.8622050990027995e-05, |
| "loss": 11.5572, |
| "num_tokens": 8262629.0, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.8099231306778477, |
| "grad_norm": 0.7139779119311859, |
| "learning_rate": 1.8490716955013232e-05, |
| "loss": 11.5717, |
| "num_tokens": 8269455.0, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.8106219426974144, |
| "grad_norm": 0.7532169883528747, |
| "learning_rate": 1.8359800488470978e-05, |
| "loss": 11.5791, |
| "num_tokens": 8276214.0, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.8113207547169812, |
| "grad_norm": 0.6616779523982886, |
| "learning_rate": 1.8229302261080495e-05, |
| "loss": 11.6381, |
| "num_tokens": 8283578.0, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.8120195667365478, |
| "grad_norm": 0.7377501630255389, |
| "learning_rate": 1.809922294137847e-05, |
| "loss": 11.6844, |
| "num_tokens": 8290424.0, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.8127183787561146, |
| "grad_norm": 0.6456460863732775, |
| "learning_rate": 1.7969563195755535e-05, |
| "loss": 11.4604, |
| "num_tokens": 8298504.0, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.8134171907756813, |
| "grad_norm": 0.7061897389733708, |
| "learning_rate": 1.784032368845283e-05, |
| "loss": 11.7079, |
| "num_tokens": 8305815.0, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.8141160027952481, |
| "grad_norm": 0.7107549537689926, |
| "learning_rate": 1.7711505081558734e-05, |
| "loss": 11.5782, |
| "num_tokens": 8312752.0, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.8148148148148148, |
| "grad_norm": 0.6668197053834488, |
| "learning_rate": 1.758310803500527e-05, |
| "loss": 11.6164, |
| "num_tokens": 8320180.0, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.8155136268343816, |
| "grad_norm": 0.7170983306456978, |
| "learning_rate": 1.7455133206564832e-05, |
| "loss": 11.4534, |
| "num_tokens": 8327115.0, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.8162124388539483, |
| "grad_norm": 0.6641025870433105, |
| "learning_rate": 1.73275812518469e-05, |
| "loss": 11.4965, |
| "num_tokens": 8334667.0, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.816911250873515, |
| "grad_norm": 0.7537451851693961, |
| "learning_rate": 1.7200452824294498e-05, |
| "loss": 11.6092, |
| "num_tokens": 8340918.0, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.8176100628930818, |
| "grad_norm": 0.721848127760393, |
| "learning_rate": 1.707374857518094e-05, |
| "loss": 11.6172, |
| "num_tokens": 8347718.0, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.8183088749126485, |
| "grad_norm": 0.692152492264296, |
| "learning_rate": 1.6947469153606577e-05, |
| "loss": 11.7099, |
| "num_tokens": 8355388.0, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.8190076869322153, |
| "grad_norm": 0.7464531132554777, |
| "learning_rate": 1.6821615206495312e-05, |
| "loss": 11.7093, |
| "num_tokens": 8362353.0, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.8197064989517819, |
| "grad_norm": 0.7371225093942753, |
| "learning_rate": 1.6696187378591376e-05, |
| "loss": 11.5761, |
| "num_tokens": 8369668.0, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.8204053109713487, |
| "grad_norm": 0.7366402368479578, |
| "learning_rate": 1.657118631245601e-05, |
| "loss": 11.7975, |
| "num_tokens": 8376453.0, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.8211041229909154, |
| "grad_norm": 0.6621056492574404, |
| "learning_rate": 1.6446612648464164e-05, |
| "loss": 11.537, |
| "num_tokens": 8384073.0, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.8218029350104822, |
| "grad_norm": 0.6531556068495009, |
| "learning_rate": 1.632246702480128e-05, |
| "loss": 11.3117, |
| "num_tokens": 8391615.0, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.822501747030049, |
| "grad_norm": 0.6820117221205548, |
| "learning_rate": 1.619875007745989e-05, |
| "loss": 11.5475, |
| "num_tokens": 8399117.0, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.8232005590496156, |
| "grad_norm": 0.7645685404468363, |
| "learning_rate": 1.607546244023651e-05, |
| "loss": 11.6094, |
| "num_tokens": 8405508.0, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.8238993710691824, |
| "grad_norm": 0.7145369409191769, |
| "learning_rate": 1.5952604744728272e-05, |
| "loss": 11.7182, |
| "num_tokens": 8412474.0, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.8245981830887491, |
| "grad_norm": 0.7674901271848729, |
| "learning_rate": 1.5830177620329712e-05, |
| "loss": 11.7028, |
| "num_tokens": 8419536.0, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.8252969951083159, |
| "grad_norm": 0.708484709151861, |
| "learning_rate": 1.570818169422966e-05, |
| "loss": 11.6668, |
| "num_tokens": 8426032.0, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.8259958071278826, |
| "grad_norm": 0.7730341374909616, |
| "learning_rate": 1.558661759140786e-05, |
| "loss": 11.6669, |
| "num_tokens": 8432351.0, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.8266946191474493, |
| "grad_norm": 0.6863726385167277, |
| "learning_rate": 1.5465485934631853e-05, |
| "loss": 11.4615, |
| "num_tokens": 8439781.0, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.827393431167016, |
| "grad_norm": 0.6979815579089426, |
| "learning_rate": 1.5344787344453805e-05, |
| "loss": 11.6105, |
| "num_tokens": 8446877.0, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.8280922431865828, |
| "grad_norm": 0.7178919451347634, |
| "learning_rate": 1.5224522439207246e-05, |
| "loss": 11.4421, |
| "num_tokens": 8453354.0, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.8287910552061496, |
| "grad_norm": 0.7224288461147278, |
| "learning_rate": 1.5104691835004048e-05, |
| "loss": 11.6013, |
| "num_tokens": 8460350.0, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.8294898672257163, |
| "grad_norm": 0.6610925527680873, |
| "learning_rate": 1.498529614573111e-05, |
| "loss": 11.4952, |
| "num_tokens": 8467882.0, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.8301886792452831, |
| "grad_norm": 0.7644397218225072, |
| "learning_rate": 1.4866335983047264e-05, |
| "loss": 11.5023, |
| "num_tokens": 8474603.0, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.8308874912648497, |
| "grad_norm": 0.710571517689817, |
| "learning_rate": 1.4747811956380242e-05, |
| "loss": 11.6805, |
| "num_tokens": 8481543.0, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.8315863032844165, |
| "grad_norm": 0.709592960585445, |
| "learning_rate": 1.4629724672923384e-05, |
| "loss": 11.5167, |
| "num_tokens": 8488957.0, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.8322851153039832, |
| "grad_norm": 0.66919107865293, |
| "learning_rate": 1.4512074737632686e-05, |
| "loss": 11.4268, |
| "num_tokens": 8496305.0, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.83298392732355, |
| "grad_norm": 0.6974868700924324, |
| "learning_rate": 1.439486275322357e-05, |
| "loss": 11.5845, |
| "num_tokens": 8503384.0, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.8336827393431167, |
| "grad_norm": 0.7031493852558214, |
| "learning_rate": 1.4278089320167876e-05, |
| "loss": 11.4762, |
| "num_tokens": 8510477.0, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.8343815513626834, |
| "grad_norm": 0.7221258353637122, |
| "learning_rate": 1.4161755036690771e-05, |
| "loss": 11.4896, |
| "num_tokens": 8517882.0, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.8350803633822502, |
| "grad_norm": 0.7166924081429917, |
| "learning_rate": 1.4045860498767671e-05, |
| "loss": 11.7326, |
| "num_tokens": 8524376.0, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.8357791754018169, |
| "grad_norm": 0.7081783346953887, |
| "learning_rate": 1.3930406300121179e-05, |
| "loss": 11.6734, |
| "num_tokens": 8531630.0, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.8364779874213837, |
| "grad_norm": 0.7202219011638747, |
| "learning_rate": 1.3815393032218115e-05, |
| "loss": 11.4651, |
| "num_tokens": 8538262.0, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.8371767994409504, |
| "grad_norm": 0.7072816743004284, |
| "learning_rate": 1.3700821284266351e-05, |
| "loss": 11.4478, |
| "num_tokens": 8545378.0, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.8378756114605171, |
| "grad_norm": 0.6675056875060252, |
| "learning_rate": 1.3586691643211957e-05, |
| "loss": 11.6135, |
| "num_tokens": 8553221.0, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.8385744234800838, |
| "grad_norm": 0.7859448407763275, |
| "learning_rate": 1.3473004693736036e-05, |
| "loss": 11.647, |
| "num_tokens": 8559107.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.8392732354996506, |
| "grad_norm": 0.7094999294935215, |
| "learning_rate": 1.3359761018251826e-05, |
| "loss": 11.4291, |
| "num_tokens": 8565926.0, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.8399720475192173, |
| "grad_norm": 0.692951631444227, |
| "learning_rate": 1.324696119690173e-05, |
| "loss": 11.7024, |
| "num_tokens": 8573174.0, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.8406708595387841, |
| "grad_norm": 0.7146683634720493, |
| "learning_rate": 1.3134605807554246e-05, |
| "loss": 11.161, |
| "num_tokens": 8579902.0, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.8413696715583509, |
| "grad_norm": 0.7185790382607279, |
| "learning_rate": 1.302269542580109e-05, |
| "loss": 11.7996, |
| "num_tokens": 8586798.0, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.8420684835779175, |
| "grad_norm": 0.667557438188177, |
| "learning_rate": 1.291123062495424e-05, |
| "loss": 11.3108, |
| "num_tokens": 8593890.0, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.8427672955974843, |
| "grad_norm": 0.7789703868272325, |
| "learning_rate": 1.2800211976042941e-05, |
| "loss": 11.7473, |
| "num_tokens": 8600460.0, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.843466107617051, |
| "grad_norm": 0.6996322937515739, |
| "learning_rate": 1.268964004781089e-05, |
| "loss": 11.4618, |
| "num_tokens": 8607572.0, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.8441649196366178, |
| "grad_norm": 0.6947717087779574, |
| "learning_rate": 1.2579515406713193e-05, |
| "loss": 11.6363, |
| "num_tokens": 8614609.0, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.8448637316561844, |
| "grad_norm": 0.690200182048955, |
| "learning_rate": 1.246983861691352e-05, |
| "loss": 11.5823, |
| "num_tokens": 8621756.0, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.8455625436757512, |
| "grad_norm": 0.7267694145321842, |
| "learning_rate": 1.236061024028129e-05, |
| "loss": 11.4435, |
| "num_tokens": 8628716.0, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.8462613556953179, |
| "grad_norm": 0.6483287697138146, |
| "learning_rate": 1.2251830836388622e-05, |
| "loss": 11.5128, |
| "num_tokens": 8636472.0, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.8469601677148847, |
| "grad_norm": 0.6866343015663074, |
| "learning_rate": 1.214350096250767e-05, |
| "loss": 11.7093, |
| "num_tokens": 8643444.0, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.8476589797344515, |
| "grad_norm": 0.6681580155044914, |
| "learning_rate": 1.2035621173607581e-05, |
| "loss": 11.5512, |
| "num_tokens": 8650747.0, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.8483577917540182, |
| "grad_norm": 0.7119552317792336, |
| "learning_rate": 1.192819202235178e-05, |
| "loss": 11.503, |
| "num_tokens": 8657676.0, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.8490566037735849, |
| "grad_norm": 0.6475705814393042, |
| "learning_rate": 1.1821214059095088e-05, |
| "loss": 11.4912, |
| "num_tokens": 8665448.0, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.8497554157931516, |
| "grad_norm": 0.6749727921972105, |
| "learning_rate": 1.1714687831880865e-05, |
| "loss": 11.6501, |
| "num_tokens": 8672624.0, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.8504542278127184, |
| "grad_norm": 0.6984478602069423, |
| "learning_rate": 1.1608613886438346e-05, |
| "loss": 11.437, |
| "num_tokens": 8679416.0, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.8511530398322851, |
| "grad_norm": 0.689356359799714, |
| "learning_rate": 1.1502992766179666e-05, |
| "loss": 11.6217, |
| "num_tokens": 8686327.0, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.8518518518518519, |
| "grad_norm": 0.7150930925806451, |
| "learning_rate": 1.139782501219715e-05, |
| "loss": 11.5312, |
| "num_tokens": 8693164.0, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.8525506638714185, |
| "grad_norm": 0.7085400286578597, |
| "learning_rate": 1.1293111163260639e-05, |
| "loss": 11.4836, |
| "num_tokens": 8700212.0, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.8532494758909853, |
| "grad_norm": 0.6496563699098009, |
| "learning_rate": 1.118885175581451e-05, |
| "loss": 11.5562, |
| "num_tokens": 8707503.0, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.8539482879105521, |
| "grad_norm": 0.6396886618398422, |
| "learning_rate": 1.1085047323975173e-05, |
| "loss": 11.4745, |
| "num_tokens": 8715062.0, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.8546470999301188, |
| "grad_norm": 0.6921296350375813, |
| "learning_rate": 1.0981698399528151e-05, |
| "loss": 11.6231, |
| "num_tokens": 8722494.0, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.8553459119496856, |
| "grad_norm": 0.6599166302398856, |
| "learning_rate": 1.0878805511925438e-05, |
| "loss": 11.537, |
| "num_tokens": 8730292.0, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.8560447239692522, |
| "grad_norm": 0.6992578050598631, |
| "learning_rate": 1.0776369188282775e-05, |
| "loss": 11.6358, |
| "num_tokens": 8737792.0, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.856743535988819, |
| "grad_norm": 0.662340520996625, |
| "learning_rate": 1.0674389953376928e-05, |
| "loss": 11.4804, |
| "num_tokens": 8745484.0, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.8574423480083857, |
| "grad_norm": 0.6668124181859917, |
| "learning_rate": 1.0572868329643027e-05, |
| "loss": 11.4285, |
| "num_tokens": 8752746.0, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.8581411600279525, |
| "grad_norm": 0.747299965413557, |
| "learning_rate": 1.0471804837171916e-05, |
| "loss": 11.7335, |
| "num_tokens": 8759307.0, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.8588399720475192, |
| "grad_norm": 0.664637762098333, |
| "learning_rate": 1.0371199993707392e-05, |
| "loss": 11.3528, |
| "num_tokens": 8766894.0, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.859538784067086, |
| "grad_norm": 0.7080136942960862, |
| "learning_rate": 1.027105431464368e-05, |
| "loss": 11.5728, |
| "num_tokens": 8773818.0, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.8602375960866527, |
| "grad_norm": 0.6584697619389641, |
| "learning_rate": 1.0171368313022677e-05, |
| "loss": 11.4522, |
| "num_tokens": 8781325.0, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.8609364081062194, |
| "grad_norm": 0.7066646119627334, |
| "learning_rate": 1.0072142499531344e-05, |
| "loss": 11.6604, |
| "num_tokens": 8788576.0, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.8616352201257862, |
| "grad_norm": 0.70483414399539, |
| "learning_rate": 9.973377382499227e-06, |
| "loss": 11.7258, |
| "num_tokens": 8794997.0, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.8623340321453529, |
| "grad_norm": 0.6487083119100663, |
| "learning_rate": 9.875073467895634e-06, |
| "loss": 11.4893, |
| "num_tokens": 8802443.0, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.8630328441649197, |
| "grad_norm": 0.6216707927603725, |
| "learning_rate": 9.777231259327212e-06, |
| "loss": 11.4552, |
| "num_tokens": 8810645.0, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.8637316561844863, |
| "grad_norm": 0.6838310230373638, |
| "learning_rate": 9.679851258035277e-06, |
| "loss": 11.4112, |
| "num_tokens": 8818550.0, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.8644304682040531, |
| "grad_norm": 0.6665502799252908, |
| "learning_rate": 9.582933962893293e-06, |
| "loss": 11.4809, |
| "num_tokens": 8825989.0, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.8651292802236199, |
| "grad_norm": 0.6544525789430773, |
| "learning_rate": 9.48647987040433e-06, |
| "loss": 11.4967, |
| "num_tokens": 8833884.0, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.8658280922431866, |
| "grad_norm": 0.7756422804378578, |
| "learning_rate": 9.390489474698439e-06, |
| "loss": 11.5953, |
| "num_tokens": 8840327.0, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.8665269042627534, |
| "grad_norm": 0.6885918162181854, |
| "learning_rate": 9.294963267530176e-06, |
| "loss": 11.6883, |
| "num_tokens": 8847713.0, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.86722571628232, |
| "grad_norm": 0.6614558583260526, |
| "learning_rate": 9.19990173827615e-06, |
| "loss": 11.3745, |
| "num_tokens": 8855278.0, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.8679245283018868, |
| "grad_norm": 0.7179694406243424, |
| "learning_rate": 9.105305373932338e-06, |
| "loss": 11.4962, |
| "num_tokens": 8862156.0, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.8686233403214535, |
| "grad_norm": 0.6628392505284637, |
| "learning_rate": 9.01117465911181e-06, |
| "loss": 11.53, |
| "num_tokens": 8869458.0, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.8693221523410203, |
| "grad_norm": 0.7319153659663948, |
| "learning_rate": 8.917510076042057e-06, |
| "loss": 11.659, |
| "num_tokens": 8875858.0, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.870020964360587, |
| "grad_norm": 0.7259443353311615, |
| "learning_rate": 8.824312104562615e-06, |
| "loss": 11.4954, |
| "num_tokens": 8882524.0, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.8707197763801537, |
| "grad_norm": 0.6508593743959824, |
| "learning_rate": 8.731581222122587e-06, |
| "loss": 11.6273, |
| "num_tokens": 8890194.0, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.8714185883997205, |
| "grad_norm": 0.6796451861284053, |
| "learning_rate": 8.639317903778189e-06, |
| "loss": 11.5158, |
| "num_tokens": 8897718.0, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.8721174004192872, |
| "grad_norm": 0.6850555867708901, |
| "learning_rate": 8.547522622190385e-06, |
| "loss": 11.3543, |
| "num_tokens": 8904932.0, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.872816212438854, |
| "grad_norm": 0.663900610727234, |
| "learning_rate": 8.45619584762235e-06, |
| "loss": 11.6675, |
| "num_tokens": 8912001.0, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.8735150244584207, |
| "grad_norm": 0.6824952234223884, |
| "learning_rate": 8.365338047937121e-06, |
| "loss": 11.3852, |
| "num_tokens": 8919388.0, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.8742138364779874, |
| "grad_norm": 0.7182084938264671, |
| "learning_rate": 8.274949688595224e-06, |
| "loss": 11.7609, |
| "num_tokens": 8926115.0, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.8749126484975541, |
| "grad_norm": 0.7430847007855079, |
| "learning_rate": 8.185031232652251e-06, |
| "loss": 11.633, |
| "num_tokens": 8932498.0, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.8756114605171209, |
| "grad_norm": 0.7147212955436707, |
| "learning_rate": 8.095583140756468e-06, |
| "loss": 11.6627, |
| "num_tokens": 8939059.0, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.8763102725366876, |
| "grad_norm": 0.684812308930803, |
| "learning_rate": 8.006605871146577e-06, |
| "loss": 11.5533, |
| "num_tokens": 8946032.0, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.8770090845562544, |
| "grad_norm": 0.6692098195446166, |
| "learning_rate": 7.918099879649144e-06, |
| "loss": 11.6309, |
| "num_tokens": 8953182.0, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.8777078965758212, |
| "grad_norm": 0.7329944739560608, |
| "learning_rate": 7.830065619676518e-06, |
| "loss": 11.4119, |
| "num_tokens": 8959948.0, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.8784067085953878, |
| "grad_norm": 0.6706051183750147, |
| "learning_rate": 7.742503542224334e-06, |
| "loss": 11.5542, |
| "num_tokens": 8967644.0, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.8791055206149546, |
| "grad_norm": 0.6496787024300155, |
| "learning_rate": 7.65541409586924e-06, |
| "loss": 11.5373, |
| "num_tokens": 8975384.0, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.8798043326345213, |
| "grad_norm": 0.8124874424417432, |
| "learning_rate": 7.568797726766686e-06, |
| "loss": 11.4112, |
| "num_tokens": 8981552.0, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.8805031446540881, |
| "grad_norm": 0.6599550037061661, |
| "learning_rate": 7.482654878648465e-06, |
| "loss": 11.6658, |
| "num_tokens": 8989062.0, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.8812019566736548, |
| "grad_norm": 0.6657122628630363, |
| "learning_rate": 7.396985992820648e-06, |
| "loss": 11.5582, |
| "num_tokens": 8996294.0, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.8819007686932215, |
| "grad_norm": 0.6858647720677542, |
| "learning_rate": 7.311791508161159e-06, |
| "loss": 11.646, |
| "num_tokens": 9003435.0, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.8825995807127882, |
| "grad_norm": 0.6364378593415492, |
| "learning_rate": 7.227071861117562e-06, |
| "loss": 11.7278, |
| "num_tokens": 9011341.0, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.883298392732355, |
| "grad_norm": 0.710629114263608, |
| "learning_rate": 7.14282748570495e-06, |
| "loss": 11.5413, |
| "num_tokens": 9017863.0, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.8839972047519218, |
| "grad_norm": 0.7150617897334426, |
| "learning_rate": 7.059058813503483e-06, |
| "loss": 11.6807, |
| "num_tokens": 9024474.0, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.8846960167714885, |
| "grad_norm": 0.6571356924743433, |
| "learning_rate": 6.975766273656425e-06, |
| "loss": 11.5295, |
| "num_tokens": 9032345.0, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.8853948287910552, |
| "grad_norm": 0.7275281273595156, |
| "learning_rate": 6.892950292867784e-06, |
| "loss": 11.6873, |
| "num_tokens": 9038850.0, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.8860936408106219, |
| "grad_norm": 0.6878221992679487, |
| "learning_rate": 6.810611295400171e-06, |
| "loss": 11.6192, |
| "num_tokens": 9045840.0, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.8867924528301887, |
| "grad_norm": 0.6586995132708238, |
| "learning_rate": 6.728749703072679e-06, |
| "loss": 11.5176, |
| "num_tokens": 9053258.0, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.8874912648497554, |
| "grad_norm": 0.7177747106842158, |
| "learning_rate": 6.647365935258642e-06, |
| "loss": 11.4239, |
| "num_tokens": 9060449.0, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.8881900768693222, |
| "grad_norm": 0.7151360360117833, |
| "learning_rate": 6.56646040888349e-06, |
| "loss": 11.6079, |
| "num_tokens": 9067366.0, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.8888888888888888, |
| "grad_norm": 0.6699876527810528, |
| "learning_rate": 6.48603353842272e-06, |
| "loss": 11.5647, |
| "num_tokens": 9074722.0, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.8895877009084556, |
| "grad_norm": 0.6934463748378176, |
| "learning_rate": 6.406085735899625e-06, |
| "loss": 11.5854, |
| "num_tokens": 9082068.0, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.8902865129280224, |
| "grad_norm": 0.6559017435267815, |
| "learning_rate": 6.326617410883295e-06, |
| "loss": 11.7271, |
| "num_tokens": 9089269.0, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.8909853249475891, |
| "grad_norm": 0.7167601840979251, |
| "learning_rate": 6.247628970486463e-06, |
| "loss": 11.3844, |
| "num_tokens": 9096281.0, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.8916841369671559, |
| "grad_norm": 0.6498437618428198, |
| "learning_rate": 6.169120819363405e-06, |
| "loss": 11.6058, |
| "num_tokens": 9103954.0, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.8923829489867225, |
| "grad_norm": 0.7045432270958085, |
| "learning_rate": 6.091093359707977e-06, |
| "loss": 11.5676, |
| "num_tokens": 9110403.0, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.8930817610062893, |
| "grad_norm": 0.691785790482857, |
| "learning_rate": 6.013546991251373e-06, |
| "loss": 11.3183, |
| "num_tokens": 9117361.0, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.893780573025856, |
| "grad_norm": 0.731188326419675, |
| "learning_rate": 5.936482111260278e-06, |
| "loss": 11.6684, |
| "num_tokens": 9123889.0, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.8944793850454228, |
| "grad_norm": 0.7114338834078112, |
| "learning_rate": 5.859899114534661e-06, |
| "loss": 11.354, |
| "num_tokens": 9130750.0, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.8951781970649895, |
| "grad_norm": 0.6830212183885706, |
| "learning_rate": 5.783798393405826e-06, |
| "loss": 11.732, |
| "num_tokens": 9137972.0, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.8958770090845563, |
| "grad_norm": 0.687450384519889, |
| "learning_rate": 5.708180337734448e-06, |
| "loss": 11.6706, |
| "num_tokens": 9144975.0, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.896575821104123, |
| "grad_norm": 0.6852418292497463, |
| "learning_rate": 5.633045334908493e-06, |
| "loss": 11.7086, |
| "num_tokens": 9152166.0, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.8972746331236897, |
| "grad_norm": 0.7322479114218022, |
| "learning_rate": 5.5583937698412856e-06, |
| "loss": 11.5725, |
| "num_tokens": 9158905.0, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.8979734451432565, |
| "grad_norm": 0.7282825482109372, |
| "learning_rate": 5.4842260249694964e-06, |
| "loss": 11.8095, |
| "num_tokens": 9165569.0, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.8986722571628232, |
| "grad_norm": 0.7137772828071814, |
| "learning_rate": 5.410542480251202e-06, |
| "loss": 11.6792, |
| "num_tokens": 9172201.0, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.89937106918239, |
| "grad_norm": 0.7233254306106957, |
| "learning_rate": 5.337343513164006e-06, |
| "loss": 11.4407, |
| "num_tokens": 9179084.0, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.9000698812019566, |
| "grad_norm": 0.674280395459097, |
| "learning_rate": 5.264629498702967e-06, |
| "loss": 11.5151, |
| "num_tokens": 9186520.0, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.9007686932215234, |
| "grad_norm": 0.6428729546255919, |
| "learning_rate": 5.192400809378783e-06, |
| "loss": 11.5095, |
| "num_tokens": 9193927.0, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.9014675052410901, |
| "grad_norm": 0.6726081450868195, |
| "learning_rate": 5.120657815215879e-06, |
| "loss": 11.5003, |
| "num_tokens": 9201756.0, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.9021663172606569, |
| "grad_norm": 0.6433988501725089, |
| "learning_rate": 5.0494008837504214e-06, |
| "loss": 11.4588, |
| "num_tokens": 9209398.0, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.9028651292802237, |
| "grad_norm": 0.7420404359119663, |
| "learning_rate": 4.978630380028582e-06, |
| "loss": 11.516, |
| "num_tokens": 9215890.0, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.9035639412997903, |
| "grad_norm": 0.7831591074910151, |
| "learning_rate": 4.908346666604502e-06, |
| "loss": 11.7173, |
| "num_tokens": 9222461.0, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.9042627533193571, |
| "grad_norm": 0.7071237983835428, |
| "learning_rate": 4.8385501035385746e-06, |
| "loss": 11.7076, |
| "num_tokens": 9229085.0, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.9049615653389238, |
| "grad_norm": 0.6427994454056646, |
| "learning_rate": 4.769241048395512e-06, |
| "loss": 11.416, |
| "num_tokens": 9236669.0, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.9056603773584906, |
| "grad_norm": 0.6884218952694207, |
| "learning_rate": 4.700419856242555e-06, |
| "loss": 11.4504, |
| "num_tokens": 9243791.0, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.9063591893780573, |
| "grad_norm": 0.7146280973640139, |
| "learning_rate": 4.632086879647635e-06, |
| "loss": 11.6455, |
| "num_tokens": 9250468.0, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.907058001397624, |
| "grad_norm": 0.6551040722182023, |
| "learning_rate": 4.564242468677615e-06, |
| "loss": 11.4883, |
| "num_tokens": 9257850.0, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.9077568134171907, |
| "grad_norm": 0.6502907151046965, |
| "learning_rate": 4.496886970896396e-06, |
| "loss": 11.5783, |
| "num_tokens": 9265236.0, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.9084556254367575, |
| "grad_norm": 0.7049840144986045, |
| "learning_rate": 4.430020731363271e-06, |
| "loss": 11.5172, |
| "num_tokens": 9272210.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.9091544374563243, |
| "grad_norm": 0.6780643846526679, |
| "learning_rate": 4.3636440926310144e-06, |
| "loss": 11.8368, |
| "num_tokens": 9279337.0, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.909853249475891, |
| "grad_norm": 0.7179333140546513, |
| "learning_rate": 4.2977573947442175e-06, |
| "loss": 11.5682, |
| "num_tokens": 9286092.0, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.9105520614954578, |
| "grad_norm": 0.7014760945022329, |
| "learning_rate": 4.232360975237571e-06, |
| "loss": 11.5352, |
| "num_tokens": 9292888.0, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.9112508735150244, |
| "grad_norm": 0.6757947711055604, |
| "learning_rate": 4.167455169134027e-06, |
| "loss": 11.6058, |
| "num_tokens": 9299710.0, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.9119496855345912, |
| "grad_norm": 0.7019157456554327, |
| "learning_rate": 4.103040308943195e-06, |
| "loss": 11.6906, |
| "num_tokens": 9306349.0, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.9126484975541579, |
| "grad_norm": 0.6500720977440768, |
| "learning_rate": 4.039116724659564e-06, |
| "loss": 11.6691, |
| "num_tokens": 9313892.0, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.9133473095737247, |
| "grad_norm": 0.6984094359716958, |
| "learning_rate": 3.975684743760832e-06, |
| "loss": 11.6063, |
| "num_tokens": 9320784.0, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.9140461215932913, |
| "grad_norm": 0.626177459402382, |
| "learning_rate": 3.91274469120626e-06, |
| "loss": 11.3268, |
| "num_tokens": 9328759.0, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.9147449336128581, |
| "grad_norm": 0.6866335858432335, |
| "learning_rate": 3.850296889434968e-06, |
| "loss": 11.9733, |
| "num_tokens": 9335584.0, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.9154437456324249, |
| "grad_norm": 0.61356922536621, |
| "learning_rate": 3.788341658364314e-06, |
| "loss": 11.5132, |
| "num_tokens": 9343277.0, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.9161425576519916, |
| "grad_norm": 0.6604269460426199, |
| "learning_rate": 3.726879315388199e-06, |
| "loss": 11.714, |
| "num_tokens": 9350951.0, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.9168413696715584, |
| "grad_norm": 0.7222165402296626, |
| "learning_rate": 3.665910175375498e-06, |
| "loss": 11.5123, |
| "num_tokens": 9357659.0, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.9175401816911251, |
| "grad_norm": 0.6523514767911169, |
| "learning_rate": 3.6054345506684627e-06, |
| "loss": 11.5572, |
| "num_tokens": 9365388.0, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.9182389937106918, |
| "grad_norm": 0.6545144502149857, |
| "learning_rate": 3.5454527510810352e-06, |
| "loss": 11.4776, |
| "num_tokens": 9372586.0, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.9189378057302585, |
| "grad_norm": 0.6412598756459515, |
| "learning_rate": 3.485965083897347e-06, |
| "loss": 11.4683, |
| "num_tokens": 9380241.0, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.9196366177498253, |
| "grad_norm": 0.7111845448756704, |
| "learning_rate": 3.426971853870109e-06, |
| "loss": 11.7242, |
| "num_tokens": 9387292.0, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.9203354297693921, |
| "grad_norm": 0.7099462928666012, |
| "learning_rate": 3.3684733632190157e-06, |
| "loss": 11.4538, |
| "num_tokens": 9394171.0, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.9210342417889588, |
| "grad_norm": 0.6526692218044201, |
| "learning_rate": 3.310469911629288e-06, |
| "loss": 11.5081, |
| "num_tokens": 9401400.0, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.9217330538085255, |
| "grad_norm": 0.7069103876866893, |
| "learning_rate": 3.252961796250054e-06, |
| "loss": 11.4752, |
| "num_tokens": 9408486.0, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.9224318658280922, |
| "grad_norm": 0.6889792496785059, |
| "learning_rate": 3.1959493116928476e-06, |
| "loss": 11.7564, |
| "num_tokens": 9415285.0, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.923130677847659, |
| "grad_norm": 0.7002775125728018, |
| "learning_rate": 3.1394327500301357e-06, |
| "loss": 11.6422, |
| "num_tokens": 9422270.0, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.9238294898672257, |
| "grad_norm": 0.6264289816602666, |
| "learning_rate": 3.0834124007937614e-06, |
| "loss": 11.6305, |
| "num_tokens": 9430223.0, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.9245283018867925, |
| "grad_norm": 0.6590788666191844, |
| "learning_rate": 3.0278885509735234e-06, |
| "loss": 11.5625, |
| "num_tokens": 9437454.0, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.9252271139063591, |
| "grad_norm": 0.6614466147979401, |
| "learning_rate": 2.9728614850156653e-06, |
| "loss": 11.3846, |
| "num_tokens": 9444734.0, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.9259259259259259, |
| "grad_norm": 0.7243775214182312, |
| "learning_rate": 2.9183314848214127e-06, |
| "loss": 11.6559, |
| "num_tokens": 9451233.0, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.9266247379454927, |
| "grad_norm": 0.6762857814271098, |
| "learning_rate": 2.864298829745571e-06, |
| "loss": 11.3847, |
| "num_tokens": 9458292.0, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.9273235499650594, |
| "grad_norm": 0.6402745881914114, |
| "learning_rate": 2.8107637965950506e-06, |
| "loss": 11.4931, |
| "num_tokens": 9465760.0, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.9280223619846262, |
| "grad_norm": 0.7081679785796786, |
| "learning_rate": 2.7577266596274576e-06, |
| "loss": 11.5325, |
| "num_tokens": 9473090.0, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.9287211740041929, |
| "grad_norm": 0.6762077375589634, |
| "learning_rate": 2.7051876905497375e-06, |
| "loss": 11.6349, |
| "num_tokens": 9480091.0, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.9294199860237596, |
| "grad_norm": 0.6469009183554926, |
| "learning_rate": 2.6531471585167e-06, |
| "loss": 11.6433, |
| "num_tokens": 9487848.0, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.9301187980433263, |
| "grad_norm": 0.657883502355354, |
| "learning_rate": 2.6016053301297196e-06, |
| "loss": 11.4191, |
| "num_tokens": 9495032.0, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.9308176100628931, |
| "grad_norm": 0.6166537222912691, |
| "learning_rate": 2.5505624694353024e-06, |
| "loss": 11.4717, |
| "num_tokens": 9502796.0, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.9315164220824598, |
| "grad_norm": 0.7211875437218437, |
| "learning_rate": 2.5000188379237786e-06, |
| "loss": 11.5148, |
| "num_tokens": 9509344.0, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.9322152341020266, |
| "grad_norm": 0.6677353668427028, |
| "learning_rate": 2.4499746945279566e-06, |
| "loss": 11.5097, |
| "num_tokens": 9516077.0, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.9329140461215933, |
| "grad_norm": 0.6567523061369096, |
| "learning_rate": 2.4004302956217804e-06, |
| "loss": 11.5162, |
| "num_tokens": 9523359.0, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.93361285814116, |
| "grad_norm": 0.708453970623699, |
| "learning_rate": 2.3513858950190204e-06, |
| "loss": 11.5558, |
| "num_tokens": 9530186.0, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.9343116701607268, |
| "grad_norm": 0.6988749660792845, |
| "learning_rate": 2.302841743971995e-06, |
| "loss": 11.3291, |
| "num_tokens": 9537136.0, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.9350104821802935, |
| "grad_norm": 0.6776722128963621, |
| "learning_rate": 2.2547980911702404e-06, |
| "loss": 11.6395, |
| "num_tokens": 9544223.0, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.9357092941998603, |
| "grad_norm": 0.7115389242403719, |
| "learning_rate": 2.2072551827392983e-06, |
| "loss": 11.6698, |
| "num_tokens": 9550920.0, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.9364081062194269, |
| "grad_norm": 0.6466374296903232, |
| "learning_rate": 2.1602132622393746e-06, |
| "loss": 11.532, |
| "num_tokens": 9558586.0, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.9371069182389937, |
| "grad_norm": 0.6524509216771044, |
| "learning_rate": 2.1136725706641712e-06, |
| "loss": 11.7108, |
| "num_tokens": 9566214.0, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.9378057302585604, |
| "grad_norm": 0.6840710486627725, |
| "learning_rate": 2.0676333464396126e-06, |
| "loss": 11.5571, |
| "num_tokens": 9573631.0, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.9385045422781272, |
| "grad_norm": 0.6908673649722832, |
| "learning_rate": 2.0220958254225984e-06, |
| "loss": 11.6546, |
| "num_tokens": 9580814.0, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.939203354297694, |
| "grad_norm": 0.7288555180137898, |
| "learning_rate": 1.977060240899864e-06, |
| "loss": 11.5905, |
| "num_tokens": 9587261.0, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.9399021663172606, |
| "grad_norm": 0.7220132561388931, |
| "learning_rate": 1.932526823586722e-06, |
| "loss": 11.7036, |
| "num_tokens": 9593973.0, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.9406009783368274, |
| "grad_norm": 0.6198864368807762, |
| "learning_rate": 1.8884958016259113e-06, |
| "loss": 11.5988, |
| "num_tokens": 9601821.0, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.9412997903563941, |
| "grad_norm": 0.6436118875785196, |
| "learning_rate": 1.844967400586428e-06, |
| "loss": 11.5774, |
| "num_tokens": 9609343.0, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.9419986023759609, |
| "grad_norm": 0.6336553658975224, |
| "learning_rate": 1.8019418434623404e-06, |
| "loss": 11.5016, |
| "num_tokens": 9616602.0, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.9426974143955276, |
| "grad_norm": 0.7075508611037244, |
| "learning_rate": 1.7594193506716983e-06, |
| "loss": 11.5928, |
| "num_tokens": 9623622.0, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.9433962264150944, |
| "grad_norm": 0.6735353557128455, |
| "learning_rate": 1.7174001400553586e-06, |
| "loss": 11.6306, |
| "num_tokens": 9631105.0, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.944095038434661, |
| "grad_norm": 0.6532951400716054, |
| "learning_rate": 1.6758844268758843e-06, |
| "loss": 11.7102, |
| "num_tokens": 9638663.0, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.9447938504542278, |
| "grad_norm": 0.6033206298760382, |
| "learning_rate": 1.634872423816458e-06, |
| "loss": 11.3447, |
| "num_tokens": 9646957.0, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.9454926624737946, |
| "grad_norm": 0.6789421018738463, |
| "learning_rate": 1.5943643409797594e-06, |
| "loss": 11.4869, |
| "num_tokens": 9653670.0, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.9461914744933613, |
| "grad_norm": 0.6488503290113109, |
| "learning_rate": 1.5543603858869215e-06, |
| "loss": 11.5039, |
| "num_tokens": 9660888.0, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.9468902865129281, |
| "grad_norm": 0.6679560200765636, |
| "learning_rate": 1.5148607634764446e-06, |
| "loss": 11.5494, |
| "num_tokens": 9668331.0, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.9475890985324947, |
| "grad_norm": 0.6993067195845916, |
| "learning_rate": 1.475865676103161e-06, |
| "loss": 11.5813, |
| "num_tokens": 9675259.0, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.9482879105520615, |
| "grad_norm": 0.6595111575445176, |
| "learning_rate": 1.4373753235371823e-06, |
| "loss": 11.2185, |
| "num_tokens": 9682674.0, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.9489867225716282, |
| "grad_norm": 0.7788936697274639, |
| "learning_rate": 1.3993899029628997e-06, |
| "loss": 11.6757, |
| "num_tokens": 9689182.0, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.949685534591195, |
| "grad_norm": 0.6643129196285505, |
| "learning_rate": 1.3619096089779293e-06, |
| "loss": 11.7565, |
| "num_tokens": 9696437.0, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.9503843466107617, |
| "grad_norm": 0.7004411450914478, |
| "learning_rate": 1.3249346335922007e-06, |
| "loss": 11.7835, |
| "num_tokens": 9703396.0, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.9510831586303284, |
| "grad_norm": 0.670408766200783, |
| "learning_rate": 1.2884651662268709e-06, |
| "loss": 11.6038, |
| "num_tokens": 9711059.0, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.9517819706498952, |
| "grad_norm": 0.6844685491462128, |
| "learning_rate": 1.2525013937134122e-06, |
| "loss": 11.5823, |
| "num_tokens": 9718400.0, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.9524807826694619, |
| "grad_norm": 0.6595213178977267, |
| "learning_rate": 1.2170435002926694e-06, |
| "loss": 11.4552, |
| "num_tokens": 9725645.0, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.9531795946890287, |
| "grad_norm": 0.6789665079944774, |
| "learning_rate": 1.1820916676138382e-06, |
| "loss": 11.4484, |
| "num_tokens": 9733005.0, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.9538784067085954, |
| "grad_norm": 0.6607462877401795, |
| "learning_rate": 1.147646074733655e-06, |
| "loss": 11.5541, |
| "num_tokens": 9740218.0, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.9545772187281621, |
| "grad_norm": 0.7142120338643752, |
| "learning_rate": 1.1137068981153632e-06, |
| "loss": 11.8216, |
| "num_tokens": 9746821.0, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.9552760307477288, |
| "grad_norm": 0.7459534452479601, |
| "learning_rate": 1.0802743116278714e-06, |
| "loss": 11.5283, |
| "num_tokens": 9753374.0, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.9559748427672956, |
| "grad_norm": 0.7015457824024218, |
| "learning_rate": 1.0473484865448525e-06, |
| "loss": 11.6701, |
| "num_tokens": 9760505.0, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.9566736547868623, |
| "grad_norm": 0.6560790041556968, |
| "learning_rate": 1.014929591543845e-06, |
| "loss": 11.6314, |
| "num_tokens": 9767972.0, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.9573724668064291, |
| "grad_norm": 0.7033378425039997, |
| "learning_rate": 9.830177927054428e-07, |
| "loss": 11.5456, |
| "num_tokens": 9774744.0, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.9580712788259959, |
| "grad_norm": 0.6798168243800103, |
| "learning_rate": 9.516132535123846e-07, |
| "loss": 11.2908, |
| "num_tokens": 9782050.0, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.9587700908455625, |
| "grad_norm": 0.6486824565620933, |
| "learning_rate": 9.207161348487315e-07, |
| "loss": 11.5991, |
| "num_tokens": 9789435.0, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.9594689028651293, |
| "grad_norm": 0.6277062623866081, |
| "learning_rate": 8.903265949990691e-07, |
| "loss": 11.4873, |
| "num_tokens": 9797498.0, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.960167714884696, |
| "grad_norm": 0.6779506091362029, |
| "learning_rate": 8.604447896476852e-07, |
| "loss": 11.4729, |
| "num_tokens": 9805014.0, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.9608665269042628, |
| "grad_norm": 0.6345571679344191, |
| "learning_rate": 8.310708718777371e-07, |
| "loss": 11.314, |
| "num_tokens": 9812890.0, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.9615653389238294, |
| "grad_norm": 0.6829253388826183, |
| "learning_rate": 8.022049921705299e-07, |
| "loss": 11.4405, |
| "num_tokens": 9819758.0, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.9622641509433962, |
| "grad_norm": 0.659631396208915, |
| "learning_rate": 7.73847298404684e-07, |
| "loss": 11.4128, |
| "num_tokens": 9826786.0, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.9629629629629629, |
| "grad_norm": 0.6407730482855021, |
| "learning_rate": 7.459979358554248e-07, |
| "loss": 11.565, |
| "num_tokens": 9834179.0, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.9636617749825297, |
| "grad_norm": 0.6521536841219591, |
| "learning_rate": 7.186570471937937e-07, |
| "loss": 11.7037, |
| "num_tokens": 9841706.0, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.9643605870020965, |
| "grad_norm": 0.6407279577165172, |
| "learning_rate": 6.918247724859939e-07, |
| "loss": 11.3862, |
| "num_tokens": 9849180.0, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.9650593990216632, |
| "grad_norm": 0.6564743299030316, |
| "learning_rate": 6.655012491925683e-07, |
| "loss": 11.6232, |
| "num_tokens": 9856546.0, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.9657582110412299, |
| "grad_norm": 0.8390466688972114, |
| "learning_rate": 6.396866121677559e-07, |
| "loss": 11.7295, |
| "num_tokens": 9863096.0, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.9664570230607966, |
| "grad_norm": 0.6392241129240702, |
| "learning_rate": 6.143809936588363e-07, |
| "loss": 11.4262, |
| "num_tokens": 9870255.0, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.9671558350803634, |
| "grad_norm": 0.609684179142815, |
| "learning_rate": 5.895845233053643e-07, |
| "loss": 11.5644, |
| "num_tokens": 9878236.0, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.9678546470999301, |
| "grad_norm": 0.6981982820151982, |
| "learning_rate": 5.652973281385588e-07, |
| "loss": 11.6708, |
| "num_tokens": 9885146.0, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.9685534591194969, |
| "grad_norm": 0.6785618783351172, |
| "learning_rate": 5.415195325806699e-07, |
| "loss": 11.5522, |
| "num_tokens": 9892008.0, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.9692522711390635, |
| "grad_norm": 0.6523225408549266, |
| "learning_rate": 5.182512584443022e-07, |
| "loss": 11.6692, |
| "num_tokens": 9899023.0, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.9699510831586303, |
| "grad_norm": 0.6459246784791243, |
| "learning_rate": 4.954926249317815e-07, |
| "loss": 11.363, |
| "num_tokens": 9906287.0, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.9706498951781971, |
| "grad_norm": 0.6632139942263436, |
| "learning_rate": 4.732437486345886e-07, |
| "loss": 11.2905, |
| "num_tokens": 9913471.0, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.9713487071977638, |
| "grad_norm": 0.6526848721218659, |
| "learning_rate": 4.515047435327491e-07, |
| "loss": 11.6156, |
| "num_tokens": 9920776.0, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.9720475192173306, |
| "grad_norm": 0.6772714564607265, |
| "learning_rate": 4.3027572099422207e-07, |
| "loss": 11.5893, |
| "num_tokens": 9927914.0, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.9727463312368972, |
| "grad_norm": 0.7303242171313036, |
| "learning_rate": 4.0955678977436797e-07, |
| "loss": 11.6023, |
| "num_tokens": 9934401.0, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.973445143256464, |
| "grad_norm": 0.6934874178347061, |
| "learning_rate": 3.893480560153484e-07, |
| "loss": 11.4115, |
| "num_tokens": 9941705.0, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.9741439552760307, |
| "grad_norm": 0.6461084160572809, |
| "learning_rate": 3.6964962324561593e-07, |
| "loss": 11.5293, |
| "num_tokens": 9949413.0, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.9748427672955975, |
| "grad_norm": 0.6736414342784565, |
| "learning_rate": 3.504615923793919e-07, |
| "loss": 11.3817, |
| "num_tokens": 9956435.0, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.9755415793151643, |
| "grad_norm": 0.701011829428076, |
| "learning_rate": 3.317840617160894e-07, |
| "loss": 11.598, |
| "num_tokens": 9963267.0, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.976240391334731, |
| "grad_norm": 0.681775890484585, |
| "learning_rate": 3.136171269399024e-07, |
| "loss": 11.2813, |
| "num_tokens": 9970795.0, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.9769392033542977, |
| "grad_norm": 0.7029694063122361, |
| "learning_rate": 2.959608811192283e-07, |
| "loss": 11.5288, |
| "num_tokens": 9977599.0, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.9776380153738644, |
| "grad_norm": 0.6397069793309575, |
| "learning_rate": 2.7881541470623494e-07, |
| "loss": 11.5876, |
| "num_tokens": 9985126.0, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.9783368273934312, |
| "grad_norm": 0.6884389472291744, |
| "learning_rate": 2.6218081553638364e-07, |
| "loss": 11.3874, |
| "num_tokens": 9991636.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.9790356394129979, |
| "grad_norm": 0.652796733566214, |
| "learning_rate": 2.4605716882801776e-07, |
| "loss": 11.5593, |
| "num_tokens": 9999333.0, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.9797344514325647, |
| "grad_norm": 0.7054950985878522, |
| "learning_rate": 2.3044455718185253e-07, |
| "loss": 11.577, |
| "num_tokens": 10006127.0, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.9804332634521313, |
| "grad_norm": 0.7053254709063593, |
| "learning_rate": 2.153430605806195e-07, |
| "loss": 11.5192, |
| "num_tokens": 10012832.0, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.9811320754716981, |
| "grad_norm": 0.6350916351726831, |
| "learning_rate": 2.0075275638862246e-07, |
| "loss": 11.2378, |
| "num_tokens": 10020535.0, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.9818308874912649, |
| "grad_norm": 0.6690972398564283, |
| "learning_rate": 1.8667371935133792e-07, |
| "loss": 11.7258, |
| "num_tokens": 10027624.0, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.9825296995108316, |
| "grad_norm": 0.6592830981787241, |
| "learning_rate": 1.7310602159505974e-07, |
| "loss": 11.553, |
| "num_tokens": 10034683.0, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.9832285115303984, |
| "grad_norm": 0.6688640725405685, |
| "learning_rate": 1.6004973262651047e-07, |
| "loss": 11.6536, |
| "num_tokens": 10042053.0, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.983927323549965, |
| "grad_norm": 0.6527104327096959, |
| "learning_rate": 1.4750491933247512e-07, |
| "loss": 11.7204, |
| "num_tokens": 10049229.0, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.9846261355695318, |
| "grad_norm": 0.7319590410040872, |
| "learning_rate": 1.3547164597949026e-07, |
| "loss": 11.8563, |
| "num_tokens": 10055821.0, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.9853249475890985, |
| "grad_norm": 0.5893377672732698, |
| "learning_rate": 1.2394997421347753e-07, |
| "loss": 11.5842, |
| "num_tokens": 10064206.0, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.9860237596086653, |
| "grad_norm": 0.7685477405141135, |
| "learning_rate": 1.1293996305946631e-07, |
| "loss": 11.5083, |
| "num_tokens": 10070918.0, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.986722571628232, |
| "grad_norm": 0.6450315989039693, |
| "learning_rate": 1.0244166892124928e-07, |
| "loss": 11.5581, |
| "num_tokens": 10078257.0, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.9874213836477987, |
| "grad_norm": 0.6761611605290332, |
| "learning_rate": 9.245514558112733e-08, |
| "loss": 11.5999, |
| "num_tokens": 10085142.0, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.9881201956673655, |
| "grad_norm": 0.7301424243829707, |
| "learning_rate": 8.298044419962069e-08, |
| "loss": 11.5774, |
| "num_tokens": 10091532.0, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.9888190076869322, |
| "grad_norm": 0.6810160455763407, |
| "learning_rate": 7.401761331521372e-08, |
| "loss": 11.5121, |
| "num_tokens": 10098630.0, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.989517819706499, |
| "grad_norm": 0.6946045633198202, |
| "learning_rate": 6.556669884408839e-08, |
| "loss": 11.5327, |
| "num_tokens": 10105381.0, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.9902166317260657, |
| "grad_norm": 0.6767554553393855, |
| "learning_rate": 5.7627744079902235e-08, |
| "loss": 11.5641, |
| "num_tokens": 10112328.0, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.9909154437456325, |
| "grad_norm": 0.7223008729413601, |
| "learning_rate": 5.0200789693588544e-08, |
| "loss": 11.6189, |
| "num_tokens": 10118964.0, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.9916142557651991, |
| "grad_norm": 0.6661158640448771, |
| "learning_rate": 4.32858737330899e-08, |
| "loss": 11.39, |
| "num_tokens": 10125874.0, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.9923130677847659, |
| "grad_norm": 0.7615114163388279, |
| "learning_rate": 3.6883031623224926e-08, |
| "loss": 11.8602, |
| "num_tokens": 10132274.0, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.9930118798043326, |
| "grad_norm": 0.7254778236725256, |
| "learning_rate": 3.099229616547739e-08, |
| "loss": 11.6168, |
| "num_tokens": 10138845.0, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.9937106918238994, |
| "grad_norm": 0.6847796254122286, |
| "learning_rate": 2.5613697537818504e-08, |
| "loss": 11.6849, |
| "num_tokens": 10145801.0, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.9944095038434662, |
| "grad_norm": 0.6968642739944435, |
| "learning_rate": 2.074726329457377e-08, |
| "loss": 11.5605, |
| "num_tokens": 10152749.0, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.9951083158630328, |
| "grad_norm": 0.6498028004501105, |
| "learning_rate": 1.6393018366278602e-08, |
| "loss": 11.5241, |
| "num_tokens": 10160312.0, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.9958071278825996, |
| "grad_norm": 0.691794824675225, |
| "learning_rate": 1.2550985059522902e-08, |
| "loss": 11.5186, |
| "num_tokens": 10167099.0, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.9965059399021663, |
| "grad_norm": 0.6808541350424336, |
| "learning_rate": 9.221183056895566e-09, |
| "loss": 11.468, |
| "num_tokens": 10174489.0, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.9972047519217331, |
| "grad_norm": 0.6518646820012849, |
| "learning_rate": 6.4036294168068335e-09, |
| "loss": 11.6372, |
| "num_tokens": 10182102.0, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.9979035639412998, |
| "grad_norm": 0.6820092282899476, |
| "learning_rate": 4.0983385734660875e-09, |
| "loss": 11.4125, |
| "num_tokens": 10188883.0, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.9986023759608665, |
| "grad_norm": 0.6891039074476841, |
| "learning_rate": 2.305322336781934e-09, |
| "loss": 11.593, |
| "num_tokens": 10195672.0, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.9993011879804332, |
| "grad_norm": 0.6506047130474067, |
| "learning_rate": 1.0245898922844888e-09, |
| "loss": 11.7408, |
| "num_tokens": 10203012.0, |
| "step": 1430 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.6706512392208532, |
| "learning_rate": 2.561478011253726e-10, |
| "loss": 11.5274, |
| "num_tokens": 10210317.0, |
| "step": 1431 |
| }, |
| { |
| "epoch": 1.0, |
| "step": 1431, |
| "total_flos": 638292458766336.0, |
| "train_loss": 12.375564571863118, |
| "train_runtime": 11888.1182, |
| "train_samples_per_second": 7.706, |
| "train_steps_per_second": 0.12 |
| } |
| ], |
| "logging_steps": 1.0, |
| "max_steps": 1431, |
| "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": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 638292458766336.0, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
