| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.944206008583691, | |
| "eval_steps": 500, | |
| "global_step": 232, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017167381974248927, | |
| "grad_norm": 24.894664359139377, | |
| "learning_rate": 0.0, | |
| "loss": 2.0734, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.034334763948497854, | |
| "grad_norm": 24.00031672844471, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.0836, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05150214592274678, | |
| "grad_norm": 29.716249933336616, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.1693, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06866952789699571, | |
| "grad_norm": 24.37331054435917, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.0201, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08583690987124463, | |
| "grad_norm": 19.789582138660975, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.9014, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10300429184549356, | |
| "grad_norm": 18.92438805672366, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.0578, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12017167381974249, | |
| "grad_norm": 13.872256454769323, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.9303, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13733905579399142, | |
| "grad_norm": 17.41216169856372, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 1.9427, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15450643776824036, | |
| "grad_norm": 13.884818846193818, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 1.7924, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17167381974248927, | |
| "grad_norm": 10.563149359389286, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 1.8341, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1888412017167382, | |
| "grad_norm": 17.065510513137394, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.8003, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20600858369098712, | |
| "grad_norm": 13.365031970331449, | |
| "learning_rate": 2.2e-06, | |
| "loss": 1.7405, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22317596566523606, | |
| "grad_norm": 10.184190757772281, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.7369, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24034334763948498, | |
| "grad_norm": 8.522394959606002, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.7161, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2575107296137339, | |
| "grad_norm": 7.4639532795389885, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.7516, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.27467811158798283, | |
| "grad_norm": 6.885531173304724, | |
| "learning_rate": 3e-06, | |
| "loss": 1.7291, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2918454935622318, | |
| "grad_norm": 8.124663030531377, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.8113, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3090128755364807, | |
| "grad_norm": 7.505534742466626, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.7011, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3261802575107296, | |
| "grad_norm": 6.546353703058948, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.679, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.34334763948497854, | |
| "grad_norm": 8.084166970057376, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 1.7672, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3605150214592275, | |
| "grad_norm": 7.39493724533185, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.7273, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3776824034334764, | |
| "grad_norm": 7.897097334092764, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 1.723, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3948497854077253, | |
| "grad_norm": 6.969310652434427, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.6652, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.41201716738197425, | |
| "grad_norm": 6.593881510049774, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.6471, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4291845493562232, | |
| "grad_norm": 7.447075197928159, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.7447, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.44635193133047213, | |
| "grad_norm": 7.593703083280446, | |
| "learning_rate": 5e-06, | |
| "loss": 1.7199, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.463519313304721, | |
| "grad_norm": 7.670358465509374, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.7528, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.48068669527896996, | |
| "grad_norm": 6.816789703879088, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.69, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4978540772532189, | |
| "grad_norm": 6.703285346233008, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.6589, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5150214592274678, | |
| "grad_norm": 6.319351833531301, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.6783, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5321888412017167, | |
| "grad_norm": 7.668634230120905, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7307, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5493562231759657, | |
| "grad_norm": 7.216605386564187, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 1.6241, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5665236051502146, | |
| "grad_norm": 7.95986167238638, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.7027, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5836909871244635, | |
| "grad_norm": 7.567869929265614, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 1.6541, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6008583690987125, | |
| "grad_norm": 7.583204324051831, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.6605, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6180257510729614, | |
| "grad_norm": 7.927092569421214, | |
| "learning_rate": 7e-06, | |
| "loss": 1.6376, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6351931330472103, | |
| "grad_norm": 6.777169937117983, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.6076, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6523605150214592, | |
| "grad_norm": 7.376746584211704, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.6631, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6695278969957081, | |
| "grad_norm": 5.931441717917702, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 1.6105, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6866952789699571, | |
| "grad_norm": 7.788470294493901, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 1.6777, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.703862660944206, | |
| "grad_norm": 7.201871110419962, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6622, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.721030042918455, | |
| "grad_norm": 7.351004222204919, | |
| "learning_rate": 8.2e-06, | |
| "loss": 1.6343, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7381974248927039, | |
| "grad_norm": 7.116370756646885, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.6713, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7553648068669528, | |
| "grad_norm": 6.804638326754223, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.5942, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7725321888412017, | |
| "grad_norm": 7.1784709545283025, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.6453, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7896995708154506, | |
| "grad_norm": 6.578248052500373, | |
| "learning_rate": 9e-06, | |
| "loss": 1.6493, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8068669527896996, | |
| "grad_norm": 5.835458640344457, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.5686, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8240343347639485, | |
| "grad_norm": 6.024182879280123, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.5686, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8412017167381974, | |
| "grad_norm": 5.931203630890797, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.5948, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8583690987124464, | |
| "grad_norm": 5.741506735885901, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.6167, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8755364806866953, | |
| "grad_norm": 6.248428015586314, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6021, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8927038626609443, | |
| "grad_norm": 5.523348714272916, | |
| "learning_rate": 9.999255120204248e-06, | |
| "loss": 1.6021, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9098712446351931, | |
| "grad_norm": 5.486958421466518, | |
| "learning_rate": 9.997020702755353e-06, | |
| "loss": 1.5414, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.927038626609442, | |
| "grad_norm": 4.906442579745384, | |
| "learning_rate": 9.993297413402282e-06, | |
| "loss": 1.6355, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.944206008583691, | |
| "grad_norm": 3.4477648478182217, | |
| "learning_rate": 9.98808636150624e-06, | |
| "loss": 1.5204, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9613733905579399, | |
| "grad_norm": 5.283485667760506, | |
| "learning_rate": 9.981389099710132e-06, | |
| "loss": 1.557, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9785407725321889, | |
| "grad_norm": 4.864774414469769, | |
| "learning_rate": 9.973207623475964e-06, | |
| "loss": 1.5847, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9957081545064378, | |
| "grad_norm": 4.99512035052619, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 1.4972, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.99512035052619, | |
| "learning_rate": 9.952402219937817e-06, | |
| "loss": 1.6505, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.425569772720337, | |
| "eval_runtime": 55.3506, | |
| "eval_samples_per_second": 6.504, | |
| "eval_steps_per_second": 0.217, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0171673819742488, | |
| "grad_norm": 3.9912629226494145, | |
| "learning_rate": 9.939784491643734e-06, | |
| "loss": 1.5689, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0343347639484979, | |
| "grad_norm": 3.5037206814091273, | |
| "learning_rate": 9.925694945084369e-06, | |
| "loss": 1.5846, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0515021459227467, | |
| "grad_norm": 3.096134013860231, | |
| "learning_rate": 9.910137778267153e-06, | |
| "loss": 1.561, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0686695278969958, | |
| "grad_norm": 6.745525969869406, | |
| "learning_rate": 9.893117626479778e-06, | |
| "loss": 1.5086, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0858369098712446, | |
| "grad_norm": 3.771654699845576, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 1.4463, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1030042918454936, | |
| "grad_norm": 7.348146928107017, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 1.5732, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1201716738197425, | |
| "grad_norm": 4.285886261726828, | |
| "learning_rate": 9.833332143466099e-06, | |
| "loss": 1.5634, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1373390557939915, | |
| "grad_norm": 9.50651645737036, | |
| "learning_rate": 9.810515099218004e-06, | |
| "loss": 1.5832, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1545064377682404, | |
| "grad_norm": 9.85903247254424, | |
| "learning_rate": 9.78626475276808e-06, | |
| "loss": 1.5158, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.1716738197424892, | |
| "grad_norm": 4.500325598034437, | |
| "learning_rate": 9.76058832955357e-06, | |
| "loss": 1.5577, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1888412017167382, | |
| "grad_norm": 6.726627673373393, | |
| "learning_rate": 9.733493479914031e-06, | |
| "loss": 1.5402, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.206008583690987, | |
| "grad_norm": 5.992547800498691, | |
| "learning_rate": 9.704988276811883e-06, | |
| "loss": 1.4771, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2231759656652361, | |
| "grad_norm": 3.966804454715501, | |
| "learning_rate": 9.675081213427076e-06, | |
| "loss": 1.4715, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.240343347639485, | |
| "grad_norm": 5.018712503166457, | |
| "learning_rate": 9.643781200626512e-06, | |
| "loss": 1.483, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2575107296137338, | |
| "grad_norm": 5.147155410650178, | |
| "learning_rate": 9.611097564309054e-06, | |
| "loss": 1.529, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2746781115879828, | |
| "grad_norm": 4.430249361377804, | |
| "learning_rate": 9.577040042626832e-06, | |
| "loss": 1.5055, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2918454935622319, | |
| "grad_norm": 3.3331885191043824, | |
| "learning_rate": 9.54161878308377e-06, | |
| "loss": 1.5686, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3090128755364807, | |
| "grad_norm": 3.3765624318373684, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 1.4806, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3261802575107295, | |
| "grad_norm": 3.0122848173838146, | |
| "learning_rate": 9.466727668927817e-06, | |
| "loss": 1.483, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3433476394849786, | |
| "grad_norm": 2.8374889256024325, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 1.5371, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.3605150214592274, | |
| "grad_norm": 3.8591143057415067, | |
| "learning_rate": 9.38651347099721e-06, | |
| "loss": 1.515, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3776824034334765, | |
| "grad_norm": 2.702387816445196, | |
| "learning_rate": 9.344439843625034e-06, | |
| "loss": 1.5065, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3948497854077253, | |
| "grad_norm": 3.231309074741847, | |
| "learning_rate": 9.301071782067504e-06, | |
| "loss": 1.4691, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4120171673819741, | |
| "grad_norm": 2.569345186210359, | |
| "learning_rate": 9.256422207921757e-06, | |
| "loss": 1.4551, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4291845493562232, | |
| "grad_norm": 2.6936841928298763, | |
| "learning_rate": 9.21050442461406e-06, | |
| "loss": 1.5342, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4463519313304722, | |
| "grad_norm": 3.1643389084628923, | |
| "learning_rate": 9.163332113436031e-06, | |
| "loss": 1.5134, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.463519313304721, | |
| "grad_norm": 2.0547003373714317, | |
| "learning_rate": 9.114919329468283e-06, | |
| "loss": 1.543, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.48068669527897, | |
| "grad_norm": 3.4765227820159628, | |
| "learning_rate": 9.065280497392663e-06, | |
| "loss": 1.4952, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.497854077253219, | |
| "grad_norm": 2.3605853021000245, | |
| "learning_rate": 9.014430407194413e-06, | |
| "loss": 1.4706, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5150214592274678, | |
| "grad_norm": 2.79040585245284, | |
| "learning_rate": 8.962384209755453e-06, | |
| "loss": 1.5038, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5321888412017168, | |
| "grad_norm": 3.130637987619235, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 1.5331, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5493562231759657, | |
| "grad_norm": 2.1702749562184036, | |
| "learning_rate": 8.854765873974898e-06, | |
| "loss": 1.4362, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.5665236051502145, | |
| "grad_norm": 3.529361540533676, | |
| "learning_rate": 8.799225800722895e-06, | |
| "loss": 1.5103, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.5836909871244635, | |
| "grad_norm": 2.1288633137671362, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 1.4735, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6008583690987126, | |
| "grad_norm": 4.653595880593553, | |
| "learning_rate": 8.684766579921684e-06, | |
| "loss": 1.4837, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6180257510729614, | |
| "grad_norm": 4.183036771208689, | |
| "learning_rate": 8.625881535716883e-06, | |
| "loss": 1.4616, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6351931330472103, | |
| "grad_norm": 3.583624228870984, | |
| "learning_rate": 8.565916153152982e-06, | |
| "loss": 1.452, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.652360515021459, | |
| "grad_norm": 2.9353226938465538, | |
| "learning_rate": 8.504888299030748e-06, | |
| "loss": 1.5003, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.6695278969957081, | |
| "grad_norm": 3.9811908362996387, | |
| "learning_rate": 8.442816156716386e-06, | |
| "loss": 1.4712, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.6866952789699572, | |
| "grad_norm": 4.507840289564422, | |
| "learning_rate": 8.379718220723772e-06, | |
| "loss": 1.5154, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.703862660944206, | |
| "grad_norm": 1.7351061691281462, | |
| "learning_rate": 8.315613291203977e-06, | |
| "loss": 1.5071, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7210300429184548, | |
| "grad_norm": 2.1548506371633476, | |
| "learning_rate": 8.250520468343722e-06, | |
| "loss": 1.4847, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.738197424892704, | |
| "grad_norm": 2.915800387932283, | |
| "learning_rate": 8.184459146674447e-06, | |
| "loss": 1.5255, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.755364806866953, | |
| "grad_norm": 1.8127751226762168, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 1.4585, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.7725321888412018, | |
| "grad_norm": 3.2104447847604884, | |
| "learning_rate": 8.049510022000365e-06, | |
| "loss": 1.5145, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.7896995708154506, | |
| "grad_norm": 1.7409586719583194, | |
| "learning_rate": 7.980662427346127e-06, | |
| "loss": 1.5268, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8068669527896994, | |
| "grad_norm": 4.142279628383546, | |
| "learning_rate": 7.910926738603855e-06, | |
| "loss": 1.4615, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8240343347639485, | |
| "grad_norm": 3.877668467895349, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 1.465, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8412017167381975, | |
| "grad_norm": 2.656628633387025, | |
| "learning_rate": 7.768874448802665e-06, | |
| "loss": 1.4922, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.8583690987124464, | |
| "grad_norm": 2.584883160911178, | |
| "learning_rate": 7.696600172495997e-06, | |
| "loss": 1.5179, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.8755364806866952, | |
| "grad_norm": 2.587010532927385, | |
| "learning_rate": 7.62352243899504e-06, | |
| "loss": 1.5014, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8927038626609443, | |
| "grad_norm": 2.5107207102352915, | |
| "learning_rate": 7.5496630219506805e-06, | |
| "loss": 1.5064, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.909871244635193, | |
| "grad_norm": 2.0882468780951915, | |
| "learning_rate": 7.475043927917908e-06, | |
| "loss": 1.4503, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9270386266094421, | |
| "grad_norm": 2.1486369526322924, | |
| "learning_rate": 7.399687389798933e-06, | |
| "loss": 1.5461, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.944206008583691, | |
| "grad_norm": 1.8242630296650162, | |
| "learning_rate": 7.323615860218844e-06, | |
| "loss": 1.4363, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.9613733905579398, | |
| "grad_norm": 2.145842755842824, | |
| "learning_rate": 7.246852004835807e-06, | |
| "loss": 1.4766, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.9785407725321889, | |
| "grad_norm": 1.6691578334566337, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 1.499, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.995708154506438, | |
| "grad_norm": 2.2108584637006863, | |
| "learning_rate": 7.091339003877826e-06, | |
| "loss": 1.4225, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.2108584637006863, | |
| "learning_rate": 7.012636193699838e-06, | |
| "loss": 1.5231, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.3721916675567627, | |
| "eval_runtime": 55.5761, | |
| "eval_samples_per_second": 6.478, | |
| "eval_steps_per_second": 0.216, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.017167381974249, | |
| "grad_norm": 3.7388997063310248, | |
| "learning_rate": 6.933333714707094e-06, | |
| "loss": 1.4963, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0343347639484977, | |
| "grad_norm": 1.9374219671850537, | |
| "learning_rate": 6.8534551952253395e-06, | |
| "loss": 1.511, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.051502145922747, | |
| "grad_norm": 2.137490994962777, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 1.4785, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.0686695278969958, | |
| "grad_norm": 1.5840707110043721, | |
| "learning_rate": 6.692065399168352e-06, | |
| "loss": 1.4386, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0858369098712446, | |
| "grad_norm": 2.0353319612062175, | |
| "learning_rate": 6.6106022089924535e-06, | |
| "loss": 1.3749, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.1030042918454934, | |
| "grad_norm": 1.8137709307149847, | |
| "learning_rate": 6.5286591367987655e-06, | |
| "loss": 1.4849, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.1201716738197427, | |
| "grad_norm": 2.7161222343926212, | |
| "learning_rate": 6.4462605976828395e-06, | |
| "loss": 1.4845, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.1373390557939915, | |
| "grad_norm": 1.3620483576209157, | |
| "learning_rate": 6.363431142447469e-06, | |
| "loss": 1.4977, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.1545064377682404, | |
| "grad_norm": 1.8781921403392814, | |
| "learning_rate": 6.280195450287736e-06, | |
| "loss": 1.4339, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.171673819742489, | |
| "grad_norm": 1.8249006642987649, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 1.4837, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.188841201716738, | |
| "grad_norm": 1.4576334521913883, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 1.4642, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.2060085836909873, | |
| "grad_norm": 1.8699628551915364, | |
| "learning_rate": 6.028299515429683e-06, | |
| "loss": 1.3993, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.223175965665236, | |
| "grad_norm": 1.1893820742085242, | |
| "learning_rate": 5.943687977264584e-06, | |
| "loss": 1.3989, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.240343347639485, | |
| "grad_norm": 1.7116247132028521, | |
| "learning_rate": 5.858795265456382e-06, | |
| "loss": 1.4183, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.257510729613734, | |
| "grad_norm": 1.5102064250019567, | |
| "learning_rate": 5.773646673951406e-06, | |
| "loss": 1.457, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.274678111587983, | |
| "grad_norm": 1.2667706935607246, | |
| "learning_rate": 5.688267572935843e-06, | |
| "loss": 1.436, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.291845493562232, | |
| "grad_norm": 1.6963247346173285, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 1.4965, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.3090128755364807, | |
| "grad_norm": 1.2569488026568798, | |
| "learning_rate": 5.51691965894185e-06, | |
| "loss": 1.4098, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.3261802575107295, | |
| "grad_norm": 1.1272036077401364, | |
| "learning_rate": 5.4310018994030974e-06, | |
| "loss": 1.4119, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.3433476394849784, | |
| "grad_norm": 1.673271865613411, | |
| "learning_rate": 5.3449557220216245e-06, | |
| "loss": 1.4661, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.3605150214592276, | |
| "grad_norm": 1.2078563404488671, | |
| "learning_rate": 5.258806764421048e-06, | |
| "loss": 1.4497, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.3776824034334765, | |
| "grad_norm": 1.308716503746094, | |
| "learning_rate": 5.172580694848541e-06, | |
| "loss": 1.4351, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3948497854077253, | |
| "grad_norm": 1.6067743882863361, | |
| "learning_rate": 5.0863032045269435e-06, | |
| "loss": 1.4017, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.412017167381974, | |
| "grad_norm": 1.15329126746228, | |
| "learning_rate": 5e-06, | |
| "loss": 1.3885, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.429184549356223, | |
| "grad_norm": 1.3029425990628287, | |
| "learning_rate": 4.913696795473058e-06, | |
| "loss": 1.4686, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.4463519313304722, | |
| "grad_norm": 1.3782312780671255, | |
| "learning_rate": 4.827419305151461e-06, | |
| "loss": 1.4472, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.463519313304721, | |
| "grad_norm": 1.1387922784921702, | |
| "learning_rate": 4.741193235578953e-06, | |
| "loss": 1.4717, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.48068669527897, | |
| "grad_norm": 1.4489271450805266, | |
| "learning_rate": 4.6550442779783755e-06, | |
| "loss": 1.4325, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.4978540772532187, | |
| "grad_norm": 1.3782527667091915, | |
| "learning_rate": 4.568998100596903e-06, | |
| "loss": 1.405, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.5150214592274676, | |
| "grad_norm": 1.2668976356990511, | |
| "learning_rate": 4.4830803410581506e-06, | |
| "loss": 1.4427, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.532188841201717, | |
| "grad_norm": 1.4950710057650962, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 1.4714, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.5493562231759657, | |
| "grad_norm": 1.0576970824111924, | |
| "learning_rate": 4.31173242706416e-06, | |
| "loss": 1.3698, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.5665236051502145, | |
| "grad_norm": 1.6518504203265545, | |
| "learning_rate": 4.226353326048594e-06, | |
| "loss": 1.4462, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.5836909871244638, | |
| "grad_norm": 1.0690977885341415, | |
| "learning_rate": 4.14120473454362e-06, | |
| "loss": 1.4094, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.6008583690987126, | |
| "grad_norm": 1.2319353745161556, | |
| "learning_rate": 4.056312022735417e-06, | |
| "loss": 1.4279, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.6180257510729614, | |
| "grad_norm": 1.4981453391586554, | |
| "learning_rate": 3.9717004845703175e-06, | |
| "loss": 1.404, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.6351931330472103, | |
| "grad_norm": 1.1718194614512722, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 1.3936, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.652360515021459, | |
| "grad_norm": 1.382020649765366, | |
| "learning_rate": 3.803421678562213e-06, | |
| "loss": 1.4432, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.6695278969957084, | |
| "grad_norm": 1.0295860006980109, | |
| "learning_rate": 3.7198045497122647e-06, | |
| "loss": 1.4189, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.686695278969957, | |
| "grad_norm": 1.6145284029231795, | |
| "learning_rate": 3.6365688575525315e-06, | |
| "loss": 1.4579, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.703862660944206, | |
| "grad_norm": 1.1519727791811123, | |
| "learning_rate": 3.553739402317162e-06, | |
| "loss": 1.4404, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.721030042918455, | |
| "grad_norm": 1.6702535379739811, | |
| "learning_rate": 3.471340863201237e-06, | |
| "loss": 1.4247, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.7381974248927037, | |
| "grad_norm": 1.4413775372344653, | |
| "learning_rate": 3.389397791007548e-06, | |
| "loss": 1.4669, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.755364806866953, | |
| "grad_norm": 1.6769364420250672, | |
| "learning_rate": 3.307934600831648e-06, | |
| "loss": 1.3987, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.772532188841202, | |
| "grad_norm": 1.4127696099779041, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 1.4577, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.7896995708154506, | |
| "grad_norm": 1.3981396160008468, | |
| "learning_rate": 3.1465448047746626e-06, | |
| "loss": 1.4643, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.8068669527896994, | |
| "grad_norm": 1.1640956539171732, | |
| "learning_rate": 3.0666662852929063e-06, | |
| "loss": 1.4127, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.8240343347639483, | |
| "grad_norm": 1.524450627139972, | |
| "learning_rate": 2.9873638063001633e-06, | |
| "loss": 1.4114, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.8412017167381975, | |
| "grad_norm": 1.3456136782749846, | |
| "learning_rate": 2.9086609961221758e-06, | |
| "loss": 1.435, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.8583690987124464, | |
| "grad_norm": 1.2127544846215854, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 1.464, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.875536480686695, | |
| "grad_norm": 1.1389112761416087, | |
| "learning_rate": 2.7531479951641928e-06, | |
| "loss": 1.4468, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.8927038626609445, | |
| "grad_norm": 1.2715694866770404, | |
| "learning_rate": 2.6763841397811576e-06, | |
| "loss": 1.451, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.909871244635193, | |
| "grad_norm": 1.2899452182832583, | |
| "learning_rate": 2.6003126102010696e-06, | |
| "loss": 1.3969, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.927038626609442, | |
| "grad_norm": 1.1876736667702164, | |
| "learning_rate": 2.524956072082093e-06, | |
| "loss": 1.4894, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.944206008583691, | |
| "grad_norm": 1.0044043551342703, | |
| "learning_rate": 2.450336978049322e-06, | |
| "loss": 1.3833, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.96137339055794, | |
| "grad_norm": 1.1006453119175361, | |
| "learning_rate": 2.37647756100496e-06, | |
| "loss": 1.422, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.978540772532189, | |
| "grad_norm": 1.1339844183954066, | |
| "learning_rate": 2.3033998275040047e-06, | |
| "loss": 1.4416, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.995708154506438, | |
| "grad_norm": 1.0443241596589774, | |
| "learning_rate": 2.2311255511973347e-06, | |
| "loss": 1.3739, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0443241596589774, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 1.4492, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.3466758728027344, | |
| "eval_runtime": 55.4567, | |
| "eval_samples_per_second": 6.492, | |
| "eval_steps_per_second": 0.216, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.017167381974249, | |
| "grad_norm": 2.208608681250419, | |
| "learning_rate": 2.089073261396148e-06, | |
| "loss": 1.4496, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.0343347639484977, | |
| "grad_norm": 2.068961082241978, | |
| "learning_rate": 2.0193375726538737e-06, | |
| "loss": 1.4599, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.051502145922747, | |
| "grad_norm": 1.2686715786339267, | |
| "learning_rate": 1.9504899779996354e-06, | |
| "loss": 1.4241, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.0686695278969958, | |
| "grad_norm": 0.9766943956912364, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 1.3856, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.0858369098712446, | |
| "grad_norm": 0.9570891019240637, | |
| "learning_rate": 1.8155408533255553e-06, | |
| "loss": 1.3267, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.1030042918454934, | |
| "grad_norm": 0.9858819209987062, | |
| "learning_rate": 1.7494795316562791e-06, | |
| "loss": 1.4371, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.1201716738197427, | |
| "grad_norm": 0.9400585321740289, | |
| "learning_rate": 1.6843867087960252e-06, | |
| "loss": 1.4347, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.1373390557939915, | |
| "grad_norm": 1.0563349262899193, | |
| "learning_rate": 1.6202817792762283e-06, | |
| "loss": 1.4431, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.1545064377682404, | |
| "grad_norm": 0.9217174697986299, | |
| "learning_rate": 1.557183843283614e-06, | |
| "loss": 1.3872, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.171673819742489, | |
| "grad_norm": 0.9537020658805889, | |
| "learning_rate": 1.4951117009692528e-06, | |
| "loss": 1.4342, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.188841201716738, | |
| "grad_norm": 0.931415129363885, | |
| "learning_rate": 1.4340838468470198e-06, | |
| "loss": 1.4167, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.2060085836909873, | |
| "grad_norm": 0.8848653838809811, | |
| "learning_rate": 1.374118464283119e-06, | |
| "loss": 1.3483, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.223175965665236, | |
| "grad_norm": 0.8999049162268763, | |
| "learning_rate": 1.3152334200783167e-06, | |
| "loss": 1.3486, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.240343347639485, | |
| "grad_norm": 0.913466824589116, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 1.372, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.257510729613734, | |
| "grad_norm": 0.9910112277396748, | |
| "learning_rate": 1.2007741992771065e-06, | |
| "loss": 1.4064, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.274678111587983, | |
| "grad_norm": 0.883357238281907, | |
| "learning_rate": 1.145234126025102e-06, | |
| "loss": 1.3869, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.291845493562232, | |
| "grad_norm": 0.9234833713444595, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 1.4469, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.3090128755364807, | |
| "grad_norm": 0.859382750295263, | |
| "learning_rate": 1.037615790244549e-06, | |
| "loss": 1.3571, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.3261802575107295, | |
| "grad_norm": 0.815275008698639, | |
| "learning_rate": 9.85569592805588e-07, | |
| "loss": 1.3635, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.3433476394849784, | |
| "grad_norm": 0.8740357211724161, | |
| "learning_rate": 9.347195026073369e-07, | |
| "loss": 1.4195, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.3605150214592276, | |
| "grad_norm": 0.8549841652198367, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 1.4011, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.3776824034334765, | |
| "grad_norm": 0.8255973582968239, | |
| "learning_rate": 8.366678865639688e-07, | |
| "loss": 1.3907, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.3948497854077253, | |
| "grad_norm": 0.8469777726661422, | |
| "learning_rate": 7.894955753859412e-07, | |
| "loss": 1.3586, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.412017167381974, | |
| "grad_norm": 0.84555567071041, | |
| "learning_rate": 7.435777920782444e-07, | |
| "loss": 1.344, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.429184549356223, | |
| "grad_norm": 0.8100696398102393, | |
| "learning_rate": 6.989282179324963e-07, | |
| "loss": 1.4253, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.4463519313304722, | |
| "grad_norm": 0.853525501890134, | |
| "learning_rate": 6.555601563749675e-07, | |
| "loss": 1.4028, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.463519313304721, | |
| "grad_norm": 2.5389247756358455, | |
| "learning_rate": 6.134865290027903e-07, | |
| "loss": 1.4282, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.48068669527897, | |
| "grad_norm": 1.2453666685589984, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 1.3921, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.4978540772532187, | |
| "grad_norm": 0.7642367373812298, | |
| "learning_rate": 5.332723310721855e-07, | |
| "loss": 1.365, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 3.5150214592274676, | |
| "grad_norm": 0.9270502698949785, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 1.4041, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 3.532188841201717, | |
| "grad_norm": 0.8098170144667726, | |
| "learning_rate": 4.5838121691622995e-07, | |
| "loss": 1.4315, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 3.5493562231759657, | |
| "grad_norm": 0.7857989969064568, | |
| "learning_rate": 4.2295995737316854e-07, | |
| "loss": 1.332, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 3.5665236051502145, | |
| "grad_norm": 0.8648677468818124, | |
| "learning_rate": 3.8890243569094874e-07, | |
| "loss": 1.4088, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.5836909871244638, | |
| "grad_norm": 0.8148105107754193, | |
| "learning_rate": 3.5621879937348836e-07, | |
| "loss": 1.3708, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 3.6008583690987126, | |
| "grad_norm": 0.8784091247812043, | |
| "learning_rate": 3.2491878657292643e-07, | |
| "loss": 1.3931, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.6180257510729614, | |
| "grad_norm": 0.8314788876032824, | |
| "learning_rate": 2.9501172318811834e-07, | |
| "loss": 1.3706, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 3.6351931330472103, | |
| "grad_norm": 0.8122583055213561, | |
| "learning_rate": 2.6650652008597067e-07, | |
| "loss": 1.3595, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 3.652360515021459, | |
| "grad_norm": 0.8988415508956524, | |
| "learning_rate": 2.394116704464294e-07, | |
| "loss": 1.4083, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.6695278969957084, | |
| "grad_norm": 0.7489347389916948, | |
| "learning_rate": 2.137352472319215e-07, | |
| "loss": 1.3862, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 3.686695278969957, | |
| "grad_norm": 0.8026525023223839, | |
| "learning_rate": 1.8948490078199767e-07, | |
| "loss": 1.4263, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 3.703862660944206, | |
| "grad_norm": 0.8691178194849434, | |
| "learning_rate": 1.666678565339025e-07, | |
| "loss": 1.4107, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 3.721030042918455, | |
| "grad_norm": 0.8295342705741752, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 1.3977, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 3.7381974248927037, | |
| "grad_norm": 0.9952946458735904, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 1.44, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.755364806866953, | |
| "grad_norm": 0.8929004736914973, | |
| "learning_rate": 1.0688237352022346e-07, | |
| "loss": 1.3744, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 3.772532188841202, | |
| "grad_norm": 0.7817159407340549, | |
| "learning_rate": 8.986222173284876e-08, | |
| "loss": 1.4323, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 3.7896995708154506, | |
| "grad_norm": 0.7269772041119471, | |
| "learning_rate": 7.430505491563101e-08, | |
| "loss": 1.4378, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 3.8068669527896994, | |
| "grad_norm": 0.846076015787421, | |
| "learning_rate": 6.021550835626777e-08, | |
| "loss": 1.3876, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 3.8240343347639483, | |
| "grad_norm": 0.7622844494333099, | |
| "learning_rate": 4.759778006218407e-08, | |
| "loss": 1.3879, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.8412017167381975, | |
| "grad_norm": 0.7517386279788537, | |
| "learning_rate": 3.645562950973014e-08, | |
| "loss": 1.4094, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 3.8583690987124464, | |
| "grad_norm": 0.7555615969516848, | |
| "learning_rate": 2.6792376524036878e-08, | |
| "loss": 1.4385, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 3.875536480686695, | |
| "grad_norm": 0.759032041862736, | |
| "learning_rate": 1.8610900289867673e-08, | |
| "loss": 1.4219, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 3.8927038626609445, | |
| "grad_norm": 0.8896057625292756, | |
| "learning_rate": 1.1913638493762369e-08, | |
| "loss": 1.4286, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 3.909871244635193, | |
| "grad_norm": 0.8049879121289703, | |
| "learning_rate": 6.702586597719385e-09, | |
| "loss": 1.3756, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.927038626609442, | |
| "grad_norm": 0.8081459120471564, | |
| "learning_rate": 2.9792972446479605e-09, | |
| "loss": 1.4649, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 3.944206008583691, | |
| "grad_norm": 0.7889695020980092, | |
| "learning_rate": 7.448797957526621e-10, | |
| "loss": 1.3612, | |
| "step": 232 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 232, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 29, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9152650971882455e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |