| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0, |
| "eval_steps": 500, |
| "global_step": 176, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.011363636363636364, |
| "grad_norm": 0.05071757916612924, |
| "learning_rate": 1.6666666666666667e-06, |
| "loss": 0.9533, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.022727272727272728, |
| "grad_norm": 0.04259673661044319, |
| "learning_rate": 3.3333333333333333e-06, |
| "loss": 0.8187, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.03409090909090909, |
| "grad_norm": 0.038167269228831166, |
| "learning_rate": 4.9999999999999996e-06, |
| "loss": 0.7078, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.045454545454545456, |
| "grad_norm": 0.05050074141516605, |
| "learning_rate": 6.666666666666667e-06, |
| "loss": 0.9871, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.056818181818181816, |
| "grad_norm": 0.0519749083650117, |
| "learning_rate": 8.333333333333334e-06, |
| "loss": 0.9578, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.06818181818181818, |
| "grad_norm": 0.04574809279613752, |
| "learning_rate": 9.999999999999999e-06, |
| "loss": 0.9209, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.07954545454545454, |
| "grad_norm": 0.044715017183707254, |
| "learning_rate": 1.1666666666666668e-05, |
| "loss": 0.7979, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.09090909090909091, |
| "grad_norm": 0.05226409537083054, |
| "learning_rate": 1.3333333333333333e-05, |
| "loss": 0.877, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.10227272727272728, |
| "grad_norm": 0.04668468864592418, |
| "learning_rate": 1.5e-05, |
| "loss": 0.7896, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.11363636363636363, |
| "grad_norm": 0.047115395620200826, |
| "learning_rate": 1.6666666666666667e-05, |
| "loss": 0.7995, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.125, |
| "grad_norm": 0.05014552073730939, |
| "learning_rate": 1.8333333333333336e-05, |
| "loss": 0.9263, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.13636363636363635, |
| "grad_norm": 0.05372632215952556, |
| "learning_rate": 1.9999999999999998e-05, |
| "loss": 0.9383, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.14772727272727273, |
| "grad_norm": 0.05707657034738693, |
| "learning_rate": 2.1666666666666667e-05, |
| "loss": 0.8975, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.1590909090909091, |
| "grad_norm": 0.051859271383260366, |
| "learning_rate": 2.3333333333333336e-05, |
| "loss": 0.8436, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.17045454545454544, |
| "grad_norm": 0.043928447365915155, |
| "learning_rate": 2.5e-05, |
| "loss": 0.8173, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.18181818181818182, |
| "grad_norm": 0.041643505841243426, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.8475, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.19318181818181818, |
| "grad_norm": 0.035898193549085154, |
| "learning_rate": 2.8333333333333332e-05, |
| "loss": 0.7011, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.20454545454545456, |
| "grad_norm": 0.03546003498547787, |
| "learning_rate": 3e-05, |
| "loss": 0.7448, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.2159090909090909, |
| "grad_norm": 0.03349420007958397, |
| "learning_rate": 2.9997034946550984e-05, |
| "loss": 0.787, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.22727272727272727, |
| "grad_norm": 0.034648139602108434, |
| "learning_rate": 2.9988140958409528e-05, |
| "loss": 0.8557, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.23863636363636365, |
| "grad_norm": 0.03180472516603228, |
| "learning_rate": 2.9973321551728995e-05, |
| "loss": 0.8827, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 0.03533912995703418, |
| "learning_rate": 2.995258258522044e-05, |
| "loss": 0.8307, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.26136363636363635, |
| "grad_norm": 0.02773960323166012, |
| "learning_rate": 2.992593225783641e-05, |
| "loss": 0.7875, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.2727272727272727, |
| "grad_norm": 0.03211541529448073, |
| "learning_rate": 2.98933811055296e-05, |
| "loss": 0.7867, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.2840909090909091, |
| "grad_norm": 0.03484278990529821, |
| "learning_rate": 2.985494199708753e-05, |
| "loss": 0.7932, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.29545454545454547, |
| "grad_norm": 0.03540049270625024, |
| "learning_rate": 2.9810630129045003e-05, |
| "loss": 0.9371, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.3068181818181818, |
| "grad_norm": 0.03612849262444276, |
| "learning_rate": 2.976046301967631e-05, |
| "loss": 0.7899, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.3181818181818182, |
| "grad_norm": 0.03728774007148355, |
| "learning_rate": 2.9704460502069544e-05, |
| "loss": 0.8348, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.32954545454545453, |
| "grad_norm": 0.043320831531330754, |
| "learning_rate": 2.9642644716285765e-05, |
| "loss": 0.7098, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.3409090909090909, |
| "grad_norm": 0.046393197374255936, |
| "learning_rate": 2.957504010060615e-05, |
| "loss": 0.7655, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.3522727272727273, |
| "grad_norm": 0.03304316691451998, |
| "learning_rate": 2.950167338187056e-05, |
| "loss": 0.7992, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.36363636363636365, |
| "grad_norm": 0.038820818042923616, |
| "learning_rate": 2.9422573564911305e-05, |
| "loss": 0.7544, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.375, |
| "grad_norm": 0.0404422735598537, |
| "learning_rate": 2.933777192108641e-05, |
| "loss": 0.7165, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.38636363636363635, |
| "grad_norm": 0.03617558322088643, |
| "learning_rate": 2.924730197591674e-05, |
| "loss": 0.7778, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.3977272727272727, |
| "grad_norm": 0.03530529645983289, |
| "learning_rate": 2.9151199495832017e-05, |
| "loss": 0.7011, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.4090909090909091, |
| "grad_norm": 0.0314407433278301, |
| "learning_rate": 2.904950247403092e-05, |
| "loss": 0.7714, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.42045454545454547, |
| "grad_norm": 0.02890718034687669, |
| "learning_rate": 2.8942251115460808e-05, |
| "loss": 0.8296, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.4318181818181818, |
| "grad_norm": 0.026351375691308618, |
| "learning_rate": 2.88294878209231e-05, |
| "loss": 0.7224, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.4431818181818182, |
| "grad_norm": 0.026057911050370135, |
| "learning_rate": 2.871125717031052e-05, |
| "loss": 0.7434, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.45454545454545453, |
| "grad_norm": 0.026073913399806834, |
| "learning_rate": 2.858760590498285e-05, |
| "loss": 0.8207, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.4659090909090909, |
| "grad_norm": 0.020463452540514345, |
| "learning_rate": 2.8458582909288185e-05, |
| "loss": 0.7069, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.4772727272727273, |
| "grad_norm": 0.02324047128233221, |
| "learning_rate": 2.832423919123698e-05, |
| "loss": 0.7914, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.48863636363636365, |
| "grad_norm": 0.019473403032255244, |
| "learning_rate": 2.8184627862336507e-05, |
| "loss": 0.6676, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 0.018395880480375864, |
| "learning_rate": 2.8039804116593743e-05, |
| "loss": 0.7095, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.5113636363636364, |
| "grad_norm": 0.02161835607931781, |
| "learning_rate": 2.7889825208694915e-05, |
| "loss": 0.8338, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.5227272727272727, |
| "grad_norm": 0.016724773851591157, |
| "learning_rate": 2.7734750431370462e-05, |
| "loss": 0.7122, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.5340909090909091, |
| "grad_norm": 0.017846968866685832, |
| "learning_rate": 2.7574641091954148e-05, |
| "loss": 0.6685, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.5454545454545454, |
| "grad_norm": 0.016985703113680868, |
| "learning_rate": 2.7409560488145863e-05, |
| "loss": 0.7548, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.5568181818181818, |
| "grad_norm": 0.01759149371654561, |
| "learning_rate": 2.7239573882987418e-05, |
| "loss": 0.7099, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.5681818181818182, |
| "grad_norm": 0.015018271516298119, |
| "learning_rate": 2.7064748479061476e-05, |
| "loss": 0.7512, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.5795454545454546, |
| "grad_norm": 0.017114886858124977, |
| "learning_rate": 2.6885153391923615e-05, |
| "loss": 0.6785, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.5909090909090909, |
| "grad_norm": 0.019418726094752342, |
| "learning_rate": 2.6700859622778184e-05, |
| "loss": 0.6766, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.6022727272727273, |
| "grad_norm": 0.016895532120200757, |
| "learning_rate": 2.651194003040862e-05, |
| "loss": 0.7728, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.6136363636363636, |
| "grad_norm": 0.018451999655787327, |
| "learning_rate": 2.6318469302373453e-05, |
| "loss": 0.6993, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.625, |
| "grad_norm": 0.018521256672385557, |
| "learning_rate": 2.6120523925479275e-05, |
| "loss": 0.6963, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.6363636363636364, |
| "grad_norm": 0.01628280369064653, |
| "learning_rate": 2.5918182155542415e-05, |
| "loss": 0.6873, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.6477272727272727, |
| "grad_norm": 0.016688951971934624, |
| "learning_rate": 2.571152398645125e-05, |
| "loss": 0.754, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.6590909090909091, |
| "grad_norm": 0.015191399660727446, |
| "learning_rate": 2.550063111854138e-05, |
| "loss": 0.5572, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.6704545454545454, |
| "grad_norm": 0.014810629906998002, |
| "learning_rate": 2.52855869262962e-05, |
| "loss": 0.7199, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.6818181818181818, |
| "grad_norm": 0.01711427193939343, |
| "learning_rate": 2.5066476425385552e-05, |
| "loss": 0.7315, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.6931818181818182, |
| "grad_norm": 0.015569879054517195, |
| "learning_rate": 2.4843386239055634e-05, |
| "loss": 0.6592, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.7045454545454546, |
| "grad_norm": 0.013751124893475849, |
| "learning_rate": 2.4616404563883302e-05, |
| "loss": 0.7748, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.7159090909090909, |
| "grad_norm": 0.015005014515577788, |
| "learning_rate": 2.43856211349084e-05, |
| "loss": 0.7134, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.7272727272727273, |
| "grad_norm": 0.013911288664586612, |
| "learning_rate": 2.4151127190157864e-05, |
| "loss": 0.7309, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.7386363636363636, |
| "grad_norm": 0.01508230536135687, |
| "learning_rate": 2.3913015434575658e-05, |
| "loss": 0.7162, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 0.01371246304799072, |
| "learning_rate": 2.3671380003372726e-05, |
| "loss": 0.7309, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.7613636363636364, |
| "grad_norm": 0.01587857420654162, |
| "learning_rate": 2.3426316424811563e-05, |
| "loss": 0.6933, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.7727272727272727, |
| "grad_norm": 0.02112505189200146, |
| "learning_rate": 2.3177921582440015e-05, |
| "loss": 0.8161, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.7840909090909091, |
| "grad_norm": 0.015148764288189093, |
| "learning_rate": 2.2926293676789295e-05, |
| "loss": 0.6977, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.7954545454545454, |
| "grad_norm": 0.0151890594638358, |
| "learning_rate": 2.2671532186551335e-05, |
| "loss": 0.6527, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.8068181818181818, |
| "grad_norm": 0.014515147138003984, |
| "learning_rate": 2.2413737829250842e-05, |
| "loss": 0.7171, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.8181818181818182, |
| "grad_norm": 0.016798521672122816, |
| "learning_rate": 2.2153012521427593e-05, |
| "loss": 0.725, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.8295454545454546, |
| "grad_norm": 0.016175606452070515, |
| "learning_rate": 2.1889459338344667e-05, |
| "loss": 0.7314, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.8409090909090909, |
| "grad_norm": 0.01731472464207072, |
| "learning_rate": 2.162318247323868e-05, |
| "loss": 0.6708, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.8522727272727273, |
| "grad_norm": 0.015934196182729173, |
| "learning_rate": 2.1354287196127925e-05, |
| "loss": 0.6836, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.8636363636363636, |
| "grad_norm": 0.013335486137157236, |
| "learning_rate": 2.108287981219491e-05, |
| "loss": 0.7366, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.875, |
| "grad_norm": 0.016178491108168734, |
| "learning_rate": 2.0809067619759618e-05, |
| "loss": 0.7951, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.8863636363636364, |
| "grad_norm": 0.0176795976865852, |
| "learning_rate": 2.0532958867860115e-05, |
| "loss": 0.7131, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.8977272727272727, |
| "grad_norm": 0.015333567208780848, |
| "learning_rate": 2.0254662713457366e-05, |
| "loss": 0.7524, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.9090909090909091, |
| "grad_norm": 0.013938237896385245, |
| "learning_rate": 1.997428917828102e-05, |
| "loss": 0.6899, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.9204545454545454, |
| "grad_norm": 0.014571786919497603, |
| "learning_rate": 1.969194910533341e-05, |
| "loss": 0.6546, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.9318181818181818, |
| "grad_norm": 0.015175277215365669, |
| "learning_rate": 1.9407754115068814e-05, |
| "loss": 0.6288, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.9431818181818182, |
| "grad_norm": 0.016148886357585004, |
| "learning_rate": 1.9121816561265377e-05, |
| "loss": 0.6456, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.9545454545454546, |
| "grad_norm": 0.017478015375242625, |
| "learning_rate": 1.883424948660712e-05, |
| "loss": 0.6011, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.9659090909090909, |
| "grad_norm": 0.015501462455188693, |
| "learning_rate": 1.8545166577993576e-05, |
| "loss": 0.6796, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.9772727272727273, |
| "grad_norm": 0.016853145780441673, |
| "learning_rate": 1.825468212159477e-05, |
| "loss": 0.6703, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.9886363636363636, |
| "grad_norm": 0.013534100626216232, |
| "learning_rate": 1.7962910957669292e-05, |
| "loss": 0.6525, |
| "step": 87 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.01437747648796264, |
| "learning_rate": 1.766996843516326e-05, |
| "loss": 0.703, |
| "step": 88 |
| }, |
| { |
| "epoch": 1.0113636363636365, |
| "grad_norm": 0.014962989333621064, |
| "learning_rate": 1.7375970366108225e-05, |
| "loss": 0.6708, |
| "step": 89 |
| }, |
| { |
| "epoch": 1.0227272727272727, |
| "grad_norm": 0.013628354300466874, |
| "learning_rate": 1.7081032979836027e-05, |
| "loss": 0.6684, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.0340909090909092, |
| "grad_norm": 0.014757516119950007, |
| "learning_rate": 1.6785272877028573e-05, |
| "loss": 0.8039, |
| "step": 91 |
| }, |
| { |
| "epoch": 1.0454545454545454, |
| "grad_norm": 0.016583346086451647, |
| "learning_rate": 1.6488806983620927e-05, |
| "loss": 0.6698, |
| "step": 92 |
| }, |
| { |
| "epoch": 1.0568181818181819, |
| "grad_norm": 0.01557946748011454, |
| "learning_rate": 1.619175250457572e-05, |
| "loss": 0.7158, |
| "step": 93 |
| }, |
| { |
| "epoch": 1.0681818181818181, |
| "grad_norm": 0.015238083013160661, |
| "learning_rate": 1.5894226877547298e-05, |
| "loss": 0.6552, |
| "step": 94 |
| }, |
| { |
| "epoch": 1.0795454545454546, |
| "grad_norm": 0.01786276315466234, |
| "learning_rate": 1.5596347726453887e-05, |
| "loss": 0.751, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.0909090909090908, |
| "grad_norm": 0.016043977197405003, |
| "learning_rate": 1.5298232814976053e-05, |
| "loss": 0.7216, |
| "step": 96 |
| }, |
| { |
| "epoch": 1.1022727272727273, |
| "grad_norm": 0.014581212655384385, |
| "learning_rate": 1.5e-05, |
| "loss": 0.7099, |
| "step": 97 |
| }, |
| { |
| "epoch": 1.1136363636363635, |
| "grad_norm": 0.014614656214189087, |
| "learning_rate": 1.4701767185023948e-05, |
| "loss": 0.6997, |
| "step": 98 |
| }, |
| { |
| "epoch": 1.125, |
| "grad_norm": 0.014304947288667047, |
| "learning_rate": 1.4403652273546118e-05, |
| "loss": 0.6585, |
| "step": 99 |
| }, |
| { |
| "epoch": 1.1363636363636362, |
| "grad_norm": 0.013971774404435894, |
| "learning_rate": 1.4105773122452703e-05, |
| "loss": 0.6652, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.1477272727272727, |
| "grad_norm": 0.014471095814527176, |
| "learning_rate": 1.3808247495424285e-05, |
| "loss": 0.6853, |
| "step": 101 |
| }, |
| { |
| "epoch": 1.1590909090909092, |
| "grad_norm": 0.015391316610067228, |
| "learning_rate": 1.3511193016379079e-05, |
| "loss": 0.6815, |
| "step": 102 |
| }, |
| { |
| "epoch": 1.1704545454545454, |
| "grad_norm": 0.017920345666710375, |
| "learning_rate": 1.3214727122971431e-05, |
| "loss": 0.6455, |
| "step": 103 |
| }, |
| { |
| "epoch": 1.1818181818181819, |
| "grad_norm": 0.016706005732186276, |
| "learning_rate": 1.2918967020163978e-05, |
| "loss": 0.7431, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.1931818181818181, |
| "grad_norm": 0.014877322443809265, |
| "learning_rate": 1.2624029633891776e-05, |
| "loss": 0.7597, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.2045454545454546, |
| "grad_norm": 0.015368071315037188, |
| "learning_rate": 1.2330031564836749e-05, |
| "loss": 0.7589, |
| "step": 106 |
| }, |
| { |
| "epoch": 1.2159090909090908, |
| "grad_norm": 0.016382740958481367, |
| "learning_rate": 1.203708904233071e-05, |
| "loss": 0.5941, |
| "step": 107 |
| }, |
| { |
| "epoch": 1.2272727272727273, |
| "grad_norm": 0.015595461127570374, |
| "learning_rate": 1.1745317878405229e-05, |
| "loss": 0.8026, |
| "step": 108 |
| }, |
| { |
| "epoch": 1.2386363636363638, |
| "grad_norm": 0.01824573515682411, |
| "learning_rate": 1.1454833422006428e-05, |
| "loss": 0.7106, |
| "step": 109 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 0.014246226061057972, |
| "learning_rate": 1.116575051339288e-05, |
| "loss": 0.6226, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.2613636363636362, |
| "grad_norm": 0.015194199552674691, |
| "learning_rate": 1.087818343873462e-05, |
| "loss": 0.6878, |
| "step": 111 |
| }, |
| { |
| "epoch": 1.2727272727272727, |
| "grad_norm": 0.018903921584870875, |
| "learning_rate": 1.0592245884931188e-05, |
| "loss": 0.6735, |
| "step": 112 |
| }, |
| { |
| "epoch": 1.2840909090909092, |
| "grad_norm": 0.014814829445622978, |
| "learning_rate": 1.0308050894666596e-05, |
| "loss": 0.7126, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.2954545454545454, |
| "grad_norm": 0.0175575648464779, |
| "learning_rate": 1.0025710821718983e-05, |
| "loss": 0.6755, |
| "step": 114 |
| }, |
| { |
| "epoch": 1.3068181818181819, |
| "grad_norm": 0.0193390892171132, |
| "learning_rate": 9.745337286542634e-06, |
| "loss": 0.6402, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.3181818181818181, |
| "grad_norm": 0.017074854773535725, |
| "learning_rate": 9.467041132139884e-06, |
| "loss": 0.7193, |
| "step": 116 |
| }, |
| { |
| "epoch": 1.3295454545454546, |
| "grad_norm": 0.016645927361554368, |
| "learning_rate": 9.190932380240386e-06, |
| "loss": 0.6427, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.3409090909090908, |
| "grad_norm": 0.01829438342338641, |
| "learning_rate": 8.917120187805091e-06, |
| "loss": 0.7091, |
| "step": 118 |
| }, |
| { |
| "epoch": 1.3522727272727273, |
| "grad_norm": 0.018172057943652505, |
| "learning_rate": 8.645712803872084e-06, |
| "loss": 0.7768, |
| "step": 119 |
| }, |
| { |
| "epoch": 1.3636363636363638, |
| "grad_norm": 0.019067265925995506, |
| "learning_rate": 8.376817526761327e-06, |
| "loss": 0.8617, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.375, |
| "grad_norm": 0.016103807777135562, |
| "learning_rate": 8.110540661655329e-06, |
| "loss": 0.6217, |
| "step": 121 |
| }, |
| { |
| "epoch": 1.3863636363636362, |
| "grad_norm": 0.01791717347539582, |
| "learning_rate": 7.846987478572411e-06, |
| "loss": 0.7356, |
| "step": 122 |
| }, |
| { |
| "epoch": 1.3977272727272727, |
| "grad_norm": 0.018325232410528938, |
| "learning_rate": 7.586262170749158e-06, |
| "loss": 0.6521, |
| "step": 123 |
| }, |
| { |
| "epoch": 1.4090909090909092, |
| "grad_norm": 0.018905797778818914, |
| "learning_rate": 7.3284678134486685e-06, |
| "loss": 0.7509, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.4204545454545454, |
| "grad_norm": 0.018602260408145506, |
| "learning_rate": 7.073706323210714e-06, |
| "loss": 0.7016, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.4318181818181819, |
| "grad_norm": 0.016514651244433562, |
| "learning_rate": 6.822078417559991e-06, |
| "loss": 0.8002, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.4431818181818181, |
| "grad_norm": 0.018833593987548795, |
| "learning_rate": 6.573683575188436e-06, |
| "loss": 0.803, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.4545454545454546, |
| "grad_norm": 0.017644785970119186, |
| "learning_rate": 6.328619996627272e-06, |
| "loss": 0.677, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.4659090909090908, |
| "grad_norm": 0.019347828980988285, |
| "learning_rate": 6.086984565424345e-06, |
| "loss": 0.7148, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.4772727272727273, |
| "grad_norm": 0.019300543348893917, |
| "learning_rate": 5.8488728098421345e-06, |
| "loss": 0.7655, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.4886363636363638, |
| "grad_norm": 0.016574344974984206, |
| "learning_rate": 5.614378865091602e-06, |
| "loss": 0.6653, |
| "step": 131 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 0.016118492967204696, |
| "learning_rate": 5.383595436116703e-06, |
| "loss": 0.6845, |
| "step": 132 |
| }, |
| { |
| "epoch": 1.5113636363636362, |
| "grad_norm": 0.016913099517869973, |
| "learning_rate": 5.1566137609443696e-06, |
| "loss": 0.5999, |
| "step": 133 |
| }, |
| { |
| "epoch": 1.5227272727272727, |
| "grad_norm": 0.017324770814363996, |
| "learning_rate": 4.933523574614447e-06, |
| "loss": 0.6307, |
| "step": 134 |
| }, |
| { |
| "epoch": 1.5340909090909092, |
| "grad_norm": 0.018088958011481436, |
| "learning_rate": 4.714413073703804e-06, |
| "loss": 0.8114, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.5454545454545454, |
| "grad_norm": 0.018107305774406484, |
| "learning_rate": 4.4993688814586215e-06, |
| "loss": 0.7423, |
| "step": 136 |
| }, |
| { |
| "epoch": 1.5568181818181817, |
| "grad_norm": 0.017510667127987316, |
| "learning_rate": 4.2884760135487535e-06, |
| "loss": 0.759, |
| "step": 137 |
| }, |
| { |
| "epoch": 1.5681818181818183, |
| "grad_norm": 0.016966967177959615, |
| "learning_rate": 4.081817844457589e-06, |
| "loss": 0.7095, |
| "step": 138 |
| }, |
| { |
| "epoch": 1.5795454545454546, |
| "grad_norm": 0.01600098768790145, |
| "learning_rate": 3.879476074520731e-06, |
| "loss": 0.6216, |
| "step": 139 |
| }, |
| { |
| "epoch": 1.5909090909090908, |
| "grad_norm": 0.0172344148854807, |
| "learning_rate": 3.6815306976265466e-06, |
| "loss": 0.6527, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.6022727272727273, |
| "grad_norm": 0.017950252109441598, |
| "learning_rate": 3.4880599695913786e-06, |
| "loss": 0.6526, |
| "step": 141 |
| }, |
| { |
| "epoch": 1.6136363636363638, |
| "grad_norm": 0.02014187993227391, |
| "learning_rate": 3.29914037722182e-06, |
| "loss": 0.6438, |
| "step": 142 |
| }, |
| { |
| "epoch": 1.625, |
| "grad_norm": 0.01738551549819398, |
| "learning_rate": 3.114846608076387e-06, |
| "loss": 0.6948, |
| "step": 143 |
| }, |
| { |
| "epoch": 1.6363636363636362, |
| "grad_norm": 0.020356218803820863, |
| "learning_rate": 2.935251520938528e-06, |
| "loss": 0.7257, |
| "step": 144 |
| }, |
| { |
| "epoch": 1.6477272727272727, |
| "grad_norm": 0.017056708535399843, |
| "learning_rate": 2.7604261170125895e-06, |
| "loss": 0.7378, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.6590909090909092, |
| "grad_norm": 0.018468485707882423, |
| "learning_rate": 2.590439511854144e-06, |
| "loss": 0.6424, |
| "step": 146 |
| }, |
| { |
| "epoch": 1.6704545454545454, |
| "grad_norm": 0.019802120923836132, |
| "learning_rate": 2.425358908045851e-06, |
| "loss": 0.6856, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.6818181818181817, |
| "grad_norm": 0.01785608788569529, |
| "learning_rate": 2.265249568629539e-06, |
| "loss": 0.7243, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.6931818181818183, |
| "grad_norm": 0.017835761915165615, |
| "learning_rate": 2.1101747913050855e-06, |
| "loss": 0.6239, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.7045454545454546, |
| "grad_norm": 0.02106800708452751, |
| "learning_rate": 1.9601958834062627e-06, |
| "loss": 0.7743, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.7159090909090908, |
| "grad_norm": 0.02012027454636489, |
| "learning_rate": 1.8153721376634947e-06, |
| "loss": 0.7849, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.7272727272727273, |
| "grad_norm": 0.017124639920244208, |
| "learning_rate": 1.6757608087630249e-06, |
| "loss": 0.7285, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.7386363636363638, |
| "grad_norm": 0.016243073549560335, |
| "learning_rate": 1.5414170907118153e-06, |
| "loss": 0.7609, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 0.01700333660952848, |
| "learning_rate": 1.412394095017151e-06, |
| "loss": 0.6887, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.7613636363636362, |
| "grad_norm": 0.01661324924856346, |
| "learning_rate": 1.2887428296894809e-06, |
| "loss": 0.7988, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.7727272727272727, |
| "grad_norm": 0.016555144729019467, |
| "learning_rate": 1.1705121790769014e-06, |
| "loss": 0.7122, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.7840909090909092, |
| "grad_norm": 0.01645576417466405, |
| "learning_rate": 1.0577488845391946e-06, |
| "loss": 0.7107, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.7954545454545454, |
| "grad_norm": 0.01714114500980707, |
| "learning_rate": 9.504975259690835e-07, |
| "loss": 0.7371, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.8068181818181817, |
| "grad_norm": 0.01725824524428257, |
| "learning_rate": 8.488005041679842e-07, |
| "loss": 0.6346, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.8181818181818183, |
| "grad_norm": 0.018332594317410387, |
| "learning_rate": 7.526980240832598e-07, |
| "loss": 0.7321, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.8295454545454546, |
| "grad_norm": 0.017728012181319464, |
| "learning_rate": 6.622280789135876e-07, |
| "loss": 0.6729, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.8409090909090908, |
| "grad_norm": 0.018834032403160974, |
| "learning_rate": 5.774264350886937e-07, |
| "loss": 0.6841, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.8522727272727273, |
| "grad_norm": 0.016746960537530962, |
| "learning_rate": 4.983266181294427e-07, |
| "loss": 0.6954, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.8636363636363638, |
| "grad_norm": 0.01747119672545748, |
| "learning_rate": 4.2495989939384916e-07, |
| "loss": 0.7679, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.875, |
| "grad_norm": 0.016865810971941564, |
| "learning_rate": 3.573552837142391e-07, |
| "loss": 0.6358, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.8863636363636362, |
| "grad_norm": 0.022834024536834356, |
| "learning_rate": 2.9553949793045874e-07, |
| "loss": 0.7111, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.8977272727272727, |
| "grad_norm": 0.016510566074369487, |
| "learning_rate": 2.395369803236902e-07, |
| "loss": 0.603, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.9090909090909092, |
| "grad_norm": 0.01701585871989393, |
| "learning_rate": 1.893698709549996e-07, |
| "loss": 0.6871, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.9204545454545454, |
| "grad_norm": 0.019587545442803056, |
| "learning_rate": 1.4505800291247207e-07, |
| "loss": 0.795, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.9318181818181817, |
| "grad_norm": 0.017848972998695284, |
| "learning_rate": 1.0661889447039886e-07, |
| "loss": 0.7491, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.9431818181818183, |
| "grad_norm": 0.023343303419197096, |
| "learning_rate": 7.40677421635888e-08, |
| "loss": 0.7785, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.9545454545454546, |
| "grad_norm": 0.017228097663166443, |
| "learning_rate": 4.741741477956252e-08, |
| "loss": 0.7956, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.9659090909090908, |
| "grad_norm": 0.01701019181440642, |
| "learning_rate": 2.6678448271005296e-08, |
| "loss": 0.7096, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.9772727272727273, |
| "grad_norm": 0.018268742287283965, |
| "learning_rate": 1.1859041590472352e-08, |
| "loss": 0.7675, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.9886363636363638, |
| "grad_norm": 0.01786805660364942, |
| "learning_rate": 2.9650534490166527e-09, |
| "loss": 0.6691, |
| "step": 175 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.01895454835226293, |
| "learning_rate": 0.0, |
| "loss": 0.7506, |
| "step": 176 |
| }, |
| { |
| "epoch": 2.0, |
| "step": 176, |
| "total_flos": 3054747482324992.0, |
| "train_loss": 0.7323629280382936, |
| "train_runtime": 3676.475, |
| "train_samples_per_second": 1.526, |
| "train_steps_per_second": 0.048 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 176, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "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": 3054747482324992.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
