| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3412726626377533, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011375755421258443, | |
| "grad_norm": 7.252706903111892, | |
| "learning_rate": 7.407407407407407e-07, | |
| "loss": 1.412, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0022751510842516885, | |
| "grad_norm": 8.622248143674112, | |
| "learning_rate": 1.4814814814814815e-06, | |
| "loss": 1.486, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0034127266263775328, | |
| "grad_norm": 7.023012815723608, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.4828, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.004550302168503377, | |
| "grad_norm": 6.665729995787034, | |
| "learning_rate": 2.962962962962963e-06, | |
| "loss": 1.4857, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005687877710629221, | |
| "grad_norm": 6.544551660595376, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 1.2733, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0068254532527550656, | |
| "grad_norm": 5.76674887523466, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.3368, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00796302879488091, | |
| "grad_norm": 3.496668631685638, | |
| "learning_rate": 5.185185185185185e-06, | |
| "loss": 1.2026, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.009100604337006754, | |
| "grad_norm": 2.871434183971456, | |
| "learning_rate": 5.925925925925926e-06, | |
| "loss": 1.1447, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.010238179879132598, | |
| "grad_norm": 4.375669967214979, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.1398, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.011375755421258443, | |
| "grad_norm": 2.6750161678621494, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 1.1125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.012513330963384287, | |
| "grad_norm": 2.5960245365420795, | |
| "learning_rate": 8.148148148148148e-06, | |
| "loss": 1.0485, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.013650906505510131, | |
| "grad_norm": 2.2437210938587584, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.9705, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.014788482047635975, | |
| "grad_norm": 1.7076930327589486, | |
| "learning_rate": 9.62962962962963e-06, | |
| "loss": 0.9693, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01592605758976182, | |
| "grad_norm": 1.5720016563835741, | |
| "learning_rate": 1.037037037037037e-05, | |
| "loss": 0.9268, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.017063633131887666, | |
| "grad_norm": 1.5049616059812596, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.0427, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.018201208674013508, | |
| "grad_norm": 1.670625778019347, | |
| "learning_rate": 1.1851851851851852e-05, | |
| "loss": 0.7984, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.019338784216139354, | |
| "grad_norm": 1.5156920640819538, | |
| "learning_rate": 1.2592592592592593e-05, | |
| "loss": 0.8934, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.020476359758265197, | |
| "grad_norm": 1.3859154498920394, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.9602, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.021613935300391043, | |
| "grad_norm": 1.3086540529666517, | |
| "learning_rate": 1.4074074074074075e-05, | |
| "loss": 0.8253, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.022751510842516885, | |
| "grad_norm": 1.3062556535833154, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.8987, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02388908638464273, | |
| "grad_norm": 1.3475867604944103, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 0.8483, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.025026661926768574, | |
| "grad_norm": 1.4066019321159149, | |
| "learning_rate": 1.6296296296296297e-05, | |
| "loss": 0.8765, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.02616423746889442, | |
| "grad_norm": 1.356056753489974, | |
| "learning_rate": 1.7037037037037038e-05, | |
| "loss": 0.7376, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.027301813011020262, | |
| "grad_norm": 1.2273100266100156, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 0.7735, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.028439388553146108, | |
| "grad_norm": 1.3325688789427055, | |
| "learning_rate": 1.851851851851852e-05, | |
| "loss": 0.9376, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02957696409527195, | |
| "grad_norm": 1.2459636758267958, | |
| "learning_rate": 1.925925925925926e-05, | |
| "loss": 0.7574, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.030714539637397797, | |
| "grad_norm": 1.2914174266728426, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8459, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03185211517952364, | |
| "grad_norm": 1.1150255281605015, | |
| "learning_rate": 1.999993201860564e-05, | |
| "loss": 0.9413, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.032989690721649485, | |
| "grad_norm": 1.1925573330854713, | |
| "learning_rate": 1.9999728075346848e-05, | |
| "loss": 0.7652, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.03412726626377533, | |
| "grad_norm": 1.1659470654782442, | |
| "learning_rate": 1.9999388172996495e-05, | |
| "loss": 0.9115, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03526484180590117, | |
| "grad_norm": 1.1687557894773881, | |
| "learning_rate": 1.999891231617599e-05, | |
| "loss": 0.8744, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.036402417348027016, | |
| "grad_norm": 1.1588348230291718, | |
| "learning_rate": 1.999830051135521e-05, | |
| "loss": 0.8677, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03753999289015286, | |
| "grad_norm": 1.132149944460972, | |
| "learning_rate": 1.9997552766852434e-05, | |
| "loss": 0.9, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03867756843227871, | |
| "grad_norm": 1.1863064139153197, | |
| "learning_rate": 1.9996669092834194e-05, | |
| "loss": 0.8074, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03981514397440455, | |
| "grad_norm": 1.220841646039248, | |
| "learning_rate": 1.9995649501315172e-05, | |
| "loss": 0.8674, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04095271951653039, | |
| "grad_norm": 1.23854307526539, | |
| "learning_rate": 1.9994494006158018e-05, | |
| "loss": 0.9159, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04209029505865624, | |
| "grad_norm": 1.090587831686784, | |
| "learning_rate": 1.9993202623073173e-05, | |
| "loss": 0.8779, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.043227870600782085, | |
| "grad_norm": 1.0923548595966273, | |
| "learning_rate": 1.999177536961863e-05, | |
| "loss": 0.8266, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.044365446142907924, | |
| "grad_norm": 1.1486258645801408, | |
| "learning_rate": 1.9990212265199738e-05, | |
| "loss": 0.8896, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04550302168503377, | |
| "grad_norm": 1.1394380414139478, | |
| "learning_rate": 1.998851333106889e-05, | |
| "loss": 0.8282, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.046640597227159616, | |
| "grad_norm": 1.0624229488933081, | |
| "learning_rate": 1.9986678590325273e-05, | |
| "loss": 0.731, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04777817276928546, | |
| "grad_norm": 1.0270694900173014, | |
| "learning_rate": 1.9984708067914533e-05, | |
| "loss": 0.8598, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0489157483114113, | |
| "grad_norm": 1.053755980685397, | |
| "learning_rate": 1.998260179062844e-05, | |
| "loss": 0.7883, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.05005332385353715, | |
| "grad_norm": 1.1027980315848285, | |
| "learning_rate": 1.9980359787104533e-05, | |
| "loss": 0.8248, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.05119089939566299, | |
| "grad_norm": 1.054761323115589, | |
| "learning_rate": 1.9977982087825714e-05, | |
| "loss": 0.8647, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05232847493778884, | |
| "grad_norm": 1.0147641458879657, | |
| "learning_rate": 1.9975468725119842e-05, | |
| "loss": 0.8707, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05346605047991468, | |
| "grad_norm": 1.0320732494589604, | |
| "learning_rate": 1.99728197331593e-05, | |
| "loss": 0.7407, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.054603626022040525, | |
| "grad_norm": 1.1478872729853458, | |
| "learning_rate": 1.9970035147960524e-05, | |
| "loss": 0.7818, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.05574120156416637, | |
| "grad_norm": 1.1399868087609295, | |
| "learning_rate": 1.996711500738351e-05, | |
| "loss": 0.8373, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.056878777106292217, | |
| "grad_norm": 1.0526771432986117, | |
| "learning_rate": 1.99640593511313e-05, | |
| "loss": 0.8203, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.058016352648418056, | |
| "grad_norm": 1.0653301252742073, | |
| "learning_rate": 1.996086822074945e-05, | |
| "loss": 0.8281, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0591539281905439, | |
| "grad_norm": 1.0561063282716783, | |
| "learning_rate": 1.995754165962546e-05, | |
| "loss": 0.72, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.06029150373266975, | |
| "grad_norm": 1.0839260823800712, | |
| "learning_rate": 1.9954079712988183e-05, | |
| "loss": 0.791, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.061429079274795594, | |
| "grad_norm": 1.0146390764424258, | |
| "learning_rate": 1.995048242790721e-05, | |
| "loss": 0.7455, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06256665481692143, | |
| "grad_norm": 1.0962320578557068, | |
| "learning_rate": 1.9946749853292233e-05, | |
| "loss": 0.8347, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06370423035904728, | |
| "grad_norm": 1.1182191442676654, | |
| "learning_rate": 1.9942882039892378e-05, | |
| "loss": 0.8298, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06484180590117312, | |
| "grad_norm": 0.985261942271752, | |
| "learning_rate": 1.9938879040295508e-05, | |
| "loss": 0.7321, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06597938144329897, | |
| "grad_norm": 1.0605922824848177, | |
| "learning_rate": 1.993474090892753e-05, | |
| "loss": 0.864, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06711695698542482, | |
| "grad_norm": 1.1264056870803152, | |
| "learning_rate": 1.9930467702051632e-05, | |
| "loss": 0.8403, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06825453252755066, | |
| "grad_norm": 1.0395490010303334, | |
| "learning_rate": 1.992605947776752e-05, | |
| "loss": 0.8243, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06939210806967651, | |
| "grad_norm": 1.0708140874889909, | |
| "learning_rate": 1.9921516296010645e-05, | |
| "loss": 0.7053, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.07052968361180234, | |
| "grad_norm": 1.1040066296881672, | |
| "learning_rate": 1.991683821855137e-05, | |
| "loss": 0.8089, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07166725915392819, | |
| "grad_norm": 1.0964641303072902, | |
| "learning_rate": 1.9912025308994146e-05, | |
| "loss": 0.9017, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07280483469605403, | |
| "grad_norm": 1.0643750074791034, | |
| "learning_rate": 1.9907077632776632e-05, | |
| "loss": 0.8678, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07394241023817988, | |
| "grad_norm": 1.0767194772302087, | |
| "learning_rate": 1.9901995257168808e-05, | |
| "loss": 0.8098, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07507998578030572, | |
| "grad_norm": 0.9814291648591228, | |
| "learning_rate": 1.989677825127208e-05, | |
| "loss": 0.9009, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.07621756132243157, | |
| "grad_norm": 1.0424860021982505, | |
| "learning_rate": 1.9891426686018308e-05, | |
| "loss": 0.7733, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.07735513686455742, | |
| "grad_norm": 1.0702692819494843, | |
| "learning_rate": 1.9885940634168863e-05, | |
| "loss": 0.8821, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.07849271240668326, | |
| "grad_norm": 1.1408260155810739, | |
| "learning_rate": 1.988032017031364e-05, | |
| "loss": 0.8307, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0796302879488091, | |
| "grad_norm": 1.0327202463855523, | |
| "learning_rate": 1.9874565370870037e-05, | |
| "loss": 0.8242, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08076786349093494, | |
| "grad_norm": 1.0467929360052353, | |
| "learning_rate": 1.9868676314081907e-05, | |
| "loss": 0.8197, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.08190543903306079, | |
| "grad_norm": 1.0401393378664532, | |
| "learning_rate": 1.9862653080018508e-05, | |
| "loss": 0.7963, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.08304301457518663, | |
| "grad_norm": 1.0193314873859287, | |
| "learning_rate": 1.985649575057341e-05, | |
| "loss": 0.7183, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.08418059011731248, | |
| "grad_norm": 1.1038745628343372, | |
| "learning_rate": 1.9850204409463387e-05, | |
| "loss": 0.8222, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08531816565943832, | |
| "grad_norm": 1.2332738343888963, | |
| "learning_rate": 1.9843779142227258e-05, | |
| "loss": 0.9721, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08645574120156417, | |
| "grad_norm": 1.091825396970584, | |
| "learning_rate": 1.9837220036224755e-05, | |
| "loss": 0.8735, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08759331674369002, | |
| "grad_norm": 1.0836987966746667, | |
| "learning_rate": 1.983052718063531e-05, | |
| "loss": 0.8209, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.08873089228581585, | |
| "grad_norm": 1.1355629551230897, | |
| "learning_rate": 1.9823700666456854e-05, | |
| "loss": 0.8062, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0898684678279417, | |
| "grad_norm": 1.0972904393016396, | |
| "learning_rate": 1.981674058650458e-05, | |
| "loss": 0.79, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.09100604337006754, | |
| "grad_norm": 1.0822786867657195, | |
| "learning_rate": 1.9809647035409673e-05, | |
| "loss": 0.8685, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09214361891219339, | |
| "grad_norm": 1.0499861096027099, | |
| "learning_rate": 1.980242010961803e-05, | |
| "loss": 0.8869, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.09328119445431923, | |
| "grad_norm": 1.1565333695689672, | |
| "learning_rate": 1.9795059907388955e-05, | |
| "loss": 0.7683, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.09441876999644508, | |
| "grad_norm": 1.1404697343493142, | |
| "learning_rate": 1.9787566528793806e-05, | |
| "loss": 0.8392, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.09555634553857092, | |
| "grad_norm": 1.0072701040551242, | |
| "learning_rate": 1.977994007571465e-05, | |
| "loss": 0.7045, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.09669392108069677, | |
| "grad_norm": 1.019073621193679, | |
| "learning_rate": 1.977218065184287e-05, | |
| "loss": 0.7104, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0978314966228226, | |
| "grad_norm": 1.0557815862870046, | |
| "learning_rate": 1.9764288362677757e-05, | |
| "loss": 0.7672, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09896907216494845, | |
| "grad_norm": 1.1380481668336246, | |
| "learning_rate": 1.975626331552507e-05, | |
| "loss": 0.9107, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1001066477070743, | |
| "grad_norm": 1.093169407323706, | |
| "learning_rate": 1.9748105619495593e-05, | |
| "loss": 0.7914, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.10124422324920014, | |
| "grad_norm": 1.087351595099475, | |
| "learning_rate": 1.973981538550364e-05, | |
| "loss": 0.7814, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.10238179879132599, | |
| "grad_norm": 1.0728410820828704, | |
| "learning_rate": 1.9731392726265538e-05, | |
| "loss": 0.7555, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10351937433345183, | |
| "grad_norm": 1.2233426634088327, | |
| "learning_rate": 1.9722837756298112e-05, | |
| "loss": 0.818, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.10465694987557768, | |
| "grad_norm": 1.0575926671469376, | |
| "learning_rate": 1.971415059191712e-05, | |
| "loss": 0.7764, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.10579452541770352, | |
| "grad_norm": 1.088708118649049, | |
| "learning_rate": 1.9705331351235673e-05, | |
| "loss": 0.7902, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.10693210095982936, | |
| "grad_norm": 1.128811241106672, | |
| "learning_rate": 1.9696380154162628e-05, | |
| "loss": 0.7237, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1080696765019552, | |
| "grad_norm": 1.193909132920539, | |
| "learning_rate": 1.9687297122400952e-05, | |
| "loss": 0.8515, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10920725204408105, | |
| "grad_norm": 1.0847877557011065, | |
| "learning_rate": 1.967808237944608e-05, | |
| "loss": 0.7833, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "grad_norm": 1.0860109471141504, | |
| "learning_rate": 1.9668736050584224e-05, | |
| "loss": 0.87, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.11148240312833274, | |
| "grad_norm": 1.0232679660737842, | |
| "learning_rate": 1.9659258262890683e-05, | |
| "loss": 0.864, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.11261997867045859, | |
| "grad_norm": 1.0289877842833908, | |
| "learning_rate": 1.96496491452281e-05, | |
| "loss": 0.7928, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.11375755421258443, | |
| "grad_norm": 1.1247393636130563, | |
| "learning_rate": 1.963990882824472e-05, | |
| "loss": 0.7534, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11489512975471028, | |
| "grad_norm": 1.0570565178567373, | |
| "learning_rate": 1.96300374443726e-05, | |
| "loss": 0.8005, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.11603270529683611, | |
| "grad_norm": 0.9867357643569166, | |
| "learning_rate": 1.962003512782584e-05, | |
| "loss": 0.7223, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.11717028083896196, | |
| "grad_norm": 1.0928156643350972, | |
| "learning_rate": 1.960990201459872e-05, | |
| "loss": 0.7677, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1183078563810878, | |
| "grad_norm": 1.021316668974268, | |
| "learning_rate": 1.959963824246387e-05, | |
| "loss": 0.7598, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.11944543192321365, | |
| "grad_norm": 1.003907848763008, | |
| "learning_rate": 1.95892439509704e-05, | |
| "loss": 0.7464, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1205830074653395, | |
| "grad_norm": 1.0107143173657256, | |
| "learning_rate": 1.9578719281442002e-05, | |
| "loss": 0.8703, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.12172058300746534, | |
| "grad_norm": 1.0427993123321293, | |
| "learning_rate": 1.9568064376975013e-05, | |
| "loss": 0.8458, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.12285815854959119, | |
| "grad_norm": 0.9471895655219162, | |
| "learning_rate": 1.9557279382436483e-05, | |
| "loss": 0.8744, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.12399573409171703, | |
| "grad_norm": 1.0474061833411126, | |
| "learning_rate": 1.9546364444462207e-05, | |
| "loss": 0.7278, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.12513330963384287, | |
| "grad_norm": 0.9851209978746495, | |
| "learning_rate": 1.953531971145473e-05, | |
| "loss": 0.8167, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1262708851759687, | |
| "grad_norm": 1.0280704738378832, | |
| "learning_rate": 1.9524145333581315e-05, | |
| "loss": 0.7167, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.12740846071809456, | |
| "grad_norm": 0.9443890480132111, | |
| "learning_rate": 1.9512841462771927e-05, | |
| "loss": 0.8822, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1285460362602204, | |
| "grad_norm": 0.9717455193006768, | |
| "learning_rate": 1.9501408252717136e-05, | |
| "loss": 0.8707, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.12968361180234625, | |
| "grad_norm": 1.092473392232703, | |
| "learning_rate": 1.9489845858866066e-05, | |
| "loss": 0.7959, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.1308211873444721, | |
| "grad_norm": 1.1056495100925654, | |
| "learning_rate": 1.947815443842424e-05, | |
| "loss": 0.7132, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.13195876288659794, | |
| "grad_norm": 1.0058537198089237, | |
| "learning_rate": 1.9466334150351475e-05, | |
| "loss": 0.8592, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1330963384287238, | |
| "grad_norm": 1.0234629362307024, | |
| "learning_rate": 1.9454385155359704e-05, | |
| "loss": 0.7982, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.13423391397084963, | |
| "grad_norm": 0.9907144940751352, | |
| "learning_rate": 1.9442307615910793e-05, | |
| "loss": 0.7626, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.13537148951297548, | |
| "grad_norm": 1.0610884551525839, | |
| "learning_rate": 1.9430101696214335e-05, | |
| "loss": 0.7694, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.13650906505510133, | |
| "grad_norm": 1.136255401471595, | |
| "learning_rate": 1.9417767562225422e-05, | |
| "loss": 0.8427, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13764664059722717, | |
| "grad_norm": 0.983990428831828, | |
| "learning_rate": 1.9405305381642376e-05, | |
| "loss": 0.8446, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.13878421613935302, | |
| "grad_norm": 1.1301380622416057, | |
| "learning_rate": 1.9392715323904483e-05, | |
| "loss": 0.7658, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.13992179168147884, | |
| "grad_norm": 1.0031223316843825, | |
| "learning_rate": 1.9379997560189677e-05, | |
| "loss": 0.7684, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.14105936722360468, | |
| "grad_norm": 1.0654113241033707, | |
| "learning_rate": 1.936715226341222e-05, | |
| "loss": 0.781, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.14219694276573053, | |
| "grad_norm": 0.9271001409058074, | |
| "learning_rate": 1.9354179608220347e-05, | |
| "loss": 0.7363, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.14333451830785637, | |
| "grad_norm": 1.0866976883120223, | |
| "learning_rate": 1.93410797709939e-05, | |
| "loss": 0.8719, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.14447209384998222, | |
| "grad_norm": 1.0143165560250647, | |
| "learning_rate": 1.9327852929841918e-05, | |
| "loss": 0.7418, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.14560966939210807, | |
| "grad_norm": 1.0794908238570715, | |
| "learning_rate": 1.9314499264600218e-05, | |
| "loss": 0.9511, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1467472449342339, | |
| "grad_norm": 1.0543849611899816, | |
| "learning_rate": 1.9301018956828966e-05, | |
| "loss": 0.7261, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.14788482047635976, | |
| "grad_norm": 0.9853708931945785, | |
| "learning_rate": 1.9287412189810174e-05, | |
| "loss": 0.7949, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1490223960184856, | |
| "grad_norm": 0.9782889131523694, | |
| "learning_rate": 1.9273679148545246e-05, | |
| "loss": 0.7894, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.15015997156061145, | |
| "grad_norm": 1.0694826655849514, | |
| "learning_rate": 1.9259820019752445e-05, | |
| "loss": 0.8288, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.1512975471027373, | |
| "grad_norm": 0.9707247690290827, | |
| "learning_rate": 1.9245834991864344e-05, | |
| "loss": 0.7973, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.15243512264486314, | |
| "grad_norm": 0.97334767003227, | |
| "learning_rate": 1.9231724255025286e-05, | |
| "loss": 0.8146, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.153572698186989, | |
| "grad_norm": 0.9354396394568825, | |
| "learning_rate": 1.9217488001088784e-05, | |
| "loss": 0.7446, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15471027372911483, | |
| "grad_norm": 0.9395461051827337, | |
| "learning_rate": 1.9203126423614916e-05, | |
| "loss": 0.8077, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.15584784927124068, | |
| "grad_norm": 0.9750295198016627, | |
| "learning_rate": 1.9188639717867695e-05, | |
| "loss": 0.7134, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.15698542481336653, | |
| "grad_norm": 1.004306726572393, | |
| "learning_rate": 1.9174028080812415e-05, | |
| "loss": 0.7295, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.15812300035549234, | |
| "grad_norm": 1.0670488030887912, | |
| "learning_rate": 1.9159291711112962e-05, | |
| "loss": 0.9652, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1592605758976182, | |
| "grad_norm": 0.959021129913344, | |
| "learning_rate": 1.914443080912913e-05, | |
| "loss": 0.7656, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16039815143974404, | |
| "grad_norm": 1.0733667782620868, | |
| "learning_rate": 1.9129445576913886e-05, | |
| "loss": 0.7651, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.16153572698186988, | |
| "grad_norm": 0.9954726556545452, | |
| "learning_rate": 1.9114336218210635e-05, | |
| "loss": 0.715, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.16267330252399573, | |
| "grad_norm": 0.9972323730973797, | |
| "learning_rate": 1.909910293845042e-05, | |
| "loss": 0.6975, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.16381087806612157, | |
| "grad_norm": 0.9913964052548584, | |
| "learning_rate": 1.9083745944749163e-05, | |
| "loss": 0.7253, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 0.9819193678083898, | |
| "learning_rate": 1.9068265445904838e-05, | |
| "loss": 0.7467, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.16608602915037327, | |
| "grad_norm": 1.0151841007728057, | |
| "learning_rate": 1.905266165239462e-05, | |
| "loss": 0.8109, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.1672236046924991, | |
| "grad_norm": 0.9838304949608468, | |
| "learning_rate": 1.903693477637204e-05, | |
| "loss": 0.7765, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.16836118023462496, | |
| "grad_norm": 0.942271667758485, | |
| "learning_rate": 1.902108503166409e-05, | |
| "loss": 0.7697, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1694987557767508, | |
| "grad_norm": 0.9992910291972015, | |
| "learning_rate": 1.9005112633768315e-05, | |
| "loss": 0.758, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.17063633131887665, | |
| "grad_norm": 1.034597439114661, | |
| "learning_rate": 1.8989017799849896e-05, | |
| "loss": 0.7424, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1717739068610025, | |
| "grad_norm": 0.9794734575939594, | |
| "learning_rate": 1.897280074873868e-05, | |
| "loss": 0.6354, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.17291148240312834, | |
| "grad_norm": 0.9930145480636745, | |
| "learning_rate": 1.8956461700926215e-05, | |
| "loss": 0.8608, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1740490579452542, | |
| "grad_norm": 1.028876703411256, | |
| "learning_rate": 1.8940000878562758e-05, | |
| "loss": 0.717, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.17518663348738003, | |
| "grad_norm": 0.9475138167941232, | |
| "learning_rate": 1.8923418505454238e-05, | |
| "loss": 0.7061, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.17632420902950588, | |
| "grad_norm": 1.0267930452512193, | |
| "learning_rate": 1.8906714807059218e-05, | |
| "loss": 0.8186, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.1774617845716317, | |
| "grad_norm": 1.0643772128505595, | |
| "learning_rate": 1.8889890010485847e-05, | |
| "loss": 0.6661, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.17859936011375754, | |
| "grad_norm": 1.066349966150257, | |
| "learning_rate": 1.8872944344488747e-05, | |
| "loss": 0.7617, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1797369356558834, | |
| "grad_norm": 0.9619934443330037, | |
| "learning_rate": 1.885587803946592e-05, | |
| "loss": 0.7127, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.18087451119800924, | |
| "grad_norm": 1.004371603307917, | |
| "learning_rate": 1.883869132745561e-05, | |
| "loss": 0.7539, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.18201208674013508, | |
| "grad_norm": 1.0106558200612026, | |
| "learning_rate": 1.8821384442133145e-05, | |
| "loss": 0.8286, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18314966228226093, | |
| "grad_norm": 1.0380867326359198, | |
| "learning_rate": 1.8803957618807764e-05, | |
| "loss": 0.826, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.18428723782438677, | |
| "grad_norm": 1.0155843819783557, | |
| "learning_rate": 1.878641109441942e-05, | |
| "loss": 0.739, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.18542481336651262, | |
| "grad_norm": 0.9916382609123128, | |
| "learning_rate": 1.876874510753554e-05, | |
| "loss": 0.9736, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.18656238890863847, | |
| "grad_norm": 0.9382307275493679, | |
| "learning_rate": 1.8750959898347828e-05, | |
| "loss": 0.8839, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1876999644507643, | |
| "grad_norm": 1.038587474509486, | |
| "learning_rate": 1.8733055708668928e-05, | |
| "loss": 0.8258, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18883753999289016, | |
| "grad_norm": 1.133352025192882, | |
| "learning_rate": 1.871503278192921e-05, | |
| "loss": 0.7396, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.189975115535016, | |
| "grad_norm": 1.0035615462799168, | |
| "learning_rate": 1.8696891363173405e-05, | |
| "loss": 0.736, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.19111269107714185, | |
| "grad_norm": 0.9854553689824346, | |
| "learning_rate": 1.86786316990573e-05, | |
| "loss": 0.7483, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.1922502666192677, | |
| "grad_norm": 1.0713592511805936, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.7571, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.19338784216139354, | |
| "grad_norm": 1.0505443529749097, | |
| "learning_rate": 1.8641758629402468e-05, | |
| "loss": 0.8198, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1945254177035194, | |
| "grad_norm": 0.999013320748331, | |
| "learning_rate": 1.862314572520028e-05, | |
| "loss": 0.8081, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.1956629932456452, | |
| "grad_norm": 1.0043036302243753, | |
| "learning_rate": 1.8604415578304052e-05, | |
| "loss": 0.8008, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.19680056878777105, | |
| "grad_norm": 0.9628077082215138, | |
| "learning_rate": 1.8585568443374087e-05, | |
| "loss": 0.7795, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.1979381443298969, | |
| "grad_norm": 0.9838858995474212, | |
| "learning_rate": 1.8566604576661288e-05, | |
| "loss": 0.7328, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.19907571987202274, | |
| "grad_norm": 0.9813195544795628, | |
| "learning_rate": 1.8547524236003675e-05, | |
| "loss": 0.8816, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2002132954141486, | |
| "grad_norm": 0.9499983935703465, | |
| "learning_rate": 1.852832768082288e-05, | |
| "loss": 0.8321, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.20135087095627444, | |
| "grad_norm": 1.0242912390679038, | |
| "learning_rate": 1.850901517212062e-05, | |
| "loss": 0.715, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.20248844649840028, | |
| "grad_norm": 1.0605111053806962, | |
| "learning_rate": 1.8489586972475154e-05, | |
| "loss": 0.7838, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.20362602204052613, | |
| "grad_norm": 0.9853924717398025, | |
| "learning_rate": 1.8470043346037698e-05, | |
| "loss": 0.8206, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.20476359758265197, | |
| "grad_norm": 0.9475510993055148, | |
| "learning_rate": 1.8450384558528848e-05, | |
| "loss": 0.8305, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.20590117312477782, | |
| "grad_norm": 0.9554542547364111, | |
| "learning_rate": 1.843061087723496e-05, | |
| "loss": 0.7735, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.20703874866690367, | |
| "grad_norm": 0.9392624041198877, | |
| "learning_rate": 1.8410722571004524e-05, | |
| "loss": 0.7179, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2081763242090295, | |
| "grad_norm": 0.9205638857801508, | |
| "learning_rate": 1.8390719910244487e-05, | |
| "loss": 0.7482, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.20931389975115536, | |
| "grad_norm": 1.0155955750650456, | |
| "learning_rate": 1.8370603166916616e-05, | |
| "loss": 0.8632, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.2104514752932812, | |
| "grad_norm": 0.9256662805897953, | |
| "learning_rate": 1.8350372614533753e-05, | |
| "loss": 0.7618, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.21158905083540705, | |
| "grad_norm": 0.9614934193864958, | |
| "learning_rate": 1.8330028528156138e-05, | |
| "loss": 0.7873, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.2127266263775329, | |
| "grad_norm": 0.9409854803070374, | |
| "learning_rate": 1.830957118438764e-05, | |
| "loss": 0.7676, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.21386420191965871, | |
| "grad_norm": 1.0117832834333376, | |
| "learning_rate": 1.8289000861372006e-05, | |
| "loss": 0.6853, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.21500177746178456, | |
| "grad_norm": 0.9436331107279641, | |
| "learning_rate": 1.826831783878909e-05, | |
| "loss": 0.7664, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2161393530039104, | |
| "grad_norm": 0.9518709375119045, | |
| "learning_rate": 1.8247522397851028e-05, | |
| "loss": 0.7979, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.21727692854603625, | |
| "grad_norm": 0.950510190670522, | |
| "learning_rate": 1.8226614821298444e-05, | |
| "loss": 0.9146, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.2184145040881621, | |
| "grad_norm": 1.0001640495640078, | |
| "learning_rate": 1.820559539339657e-05, | |
| "loss": 0.8066, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.21955207963028794, | |
| "grad_norm": 1.0515205356048116, | |
| "learning_rate": 1.8184464399931414e-05, | |
| "loss": 0.7605, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "grad_norm": 0.9518221695117551, | |
| "learning_rate": 1.8163222128205853e-05, | |
| "loss": 0.77, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.22182723071453964, | |
| "grad_norm": 0.9658190130336755, | |
| "learning_rate": 1.8141868867035745e-05, | |
| "loss": 0.7553, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.22296480625666548, | |
| "grad_norm": 0.944099728059088, | |
| "learning_rate": 1.8120404906745973e-05, | |
| "loss": 0.7851, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.22410238179879133, | |
| "grad_norm": 0.9818110526183947, | |
| "learning_rate": 1.8098830539166536e-05, | |
| "loss": 0.7935, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.22523995734091717, | |
| "grad_norm": 1.0679918663170498, | |
| "learning_rate": 1.8077146057628547e-05, | |
| "loss": 0.7782, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.22637753288304302, | |
| "grad_norm": 0.9636654090223082, | |
| "learning_rate": 1.8055351756960262e-05, | |
| "loss": 0.7111, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.22751510842516887, | |
| "grad_norm": 1.0273549839107226, | |
| "learning_rate": 1.8033447933483076e-05, | |
| "loss": 0.7937, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2286526839672947, | |
| "grad_norm": 0.9768073027947863, | |
| "learning_rate": 1.8011434885007482e-05, | |
| "loss": 0.6585, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.22979025950942056, | |
| "grad_norm": 1.0757919457820178, | |
| "learning_rate": 1.7989312910829023e-05, | |
| "loss": 0.8112, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2309278350515464, | |
| "grad_norm": 1.07129922697845, | |
| "learning_rate": 1.796708231172423e-05, | |
| "loss": 0.7197, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.23206541059367222, | |
| "grad_norm": 0.9064139244148179, | |
| "learning_rate": 1.7944743389946524e-05, | |
| "loss": 0.8407, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.23320298613579807, | |
| "grad_norm": 1.0566497130819101, | |
| "learning_rate": 1.792229644922212e-05, | |
| "loss": 0.6641, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.23434056167792391, | |
| "grad_norm": 1.1821144067979144, | |
| "learning_rate": 1.789974179474588e-05, | |
| "loss": 0.8014, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.23547813722004976, | |
| "grad_norm": 1.004667649949057, | |
| "learning_rate": 1.7877079733177185e-05, | |
| "loss": 0.7442, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2366157127621756, | |
| "grad_norm": 1.0680994058446738, | |
| "learning_rate": 1.7854310572635733e-05, | |
| "loss": 0.8034, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.23775328830430145, | |
| "grad_norm": 1.0075812686052639, | |
| "learning_rate": 1.7831434622697386e-05, | |
| "loss": 0.6922, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2388908638464273, | |
| "grad_norm": 1.0078124911781743, | |
| "learning_rate": 1.780845219438994e-05, | |
| "loss": 0.8086, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.24002843938855314, | |
| "grad_norm": 1.0012945867152467, | |
| "learning_rate": 1.7785363600188894e-05, | |
| "loss": 0.7705, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.241166014930679, | |
| "grad_norm": 1.1225575837862554, | |
| "learning_rate": 1.776216915401322e-05, | |
| "loss": 0.7568, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.24230359047280484, | |
| "grad_norm": 0.9417229426258661, | |
| "learning_rate": 1.773886917122107e-05, | |
| "loss": 0.7634, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.24344116601493068, | |
| "grad_norm": 1.0736164163302175, | |
| "learning_rate": 1.771546396860551e-05, | |
| "loss": 0.7334, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.24457874155705653, | |
| "grad_norm": 1.0980435254209089, | |
| "learning_rate": 1.7691953864390208e-05, | |
| "loss": 0.8121, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.24571631709918237, | |
| "grad_norm": 0.9581368725405702, | |
| "learning_rate": 1.766833917822509e-05, | |
| "loss": 0.7622, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.24685389264130822, | |
| "grad_norm": 0.94688646103546, | |
| "learning_rate": 1.7644620231182015e-05, | |
| "loss": 0.8606, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.24799146818343407, | |
| "grad_norm": 1.000027690442481, | |
| "learning_rate": 1.7620797345750403e-05, | |
| "loss": 0.6355, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2491290437255599, | |
| "grad_norm": 1.0030472762272575, | |
| "learning_rate": 1.759687084583285e-05, | |
| "loss": 0.7331, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.25026661926768573, | |
| "grad_norm": 0.933093406691675, | |
| "learning_rate": 1.7572841056740722e-05, | |
| "loss": 0.8011, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2514041948098116, | |
| "grad_norm": 0.9421445936097415, | |
| "learning_rate": 1.7548708305189724e-05, | |
| "loss": 0.7891, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.2525417703519374, | |
| "grad_norm": 0.9681375218542562, | |
| "learning_rate": 1.7524472919295488e-05, | |
| "loss": 0.8035, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.25367934589406327, | |
| "grad_norm": 0.9614473856459215, | |
| "learning_rate": 1.7500135228569067e-05, | |
| "loss": 0.7541, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.2548169214361891, | |
| "grad_norm": 1.0407809945974127, | |
| "learning_rate": 1.7475695563912506e-05, | |
| "loss": 0.7616, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.25595449697831496, | |
| "grad_norm": 0.9557104322965116, | |
| "learning_rate": 1.7451154257614287e-05, | |
| "loss": 0.7465, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2570920725204408, | |
| "grad_norm": 1.0178852691492593, | |
| "learning_rate": 1.742651164334486e-05, | |
| "loss": 0.8887, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.25822964806256665, | |
| "grad_norm": 1.0183556677477257, | |
| "learning_rate": 1.7401768056152083e-05, | |
| "loss": 0.7316, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.2593672236046925, | |
| "grad_norm": 0.9967963519853643, | |
| "learning_rate": 1.7376923832456665e-05, | |
| "loss": 0.7435, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.26050479914681834, | |
| "grad_norm": 0.9398084773931321, | |
| "learning_rate": 1.7351979310047603e-05, | |
| "loss": 0.8236, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.2616423746889442, | |
| "grad_norm": 0.9702704929040705, | |
| "learning_rate": 1.7326934828077574e-05, | |
| "loss": 0.7236, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.26277995023107004, | |
| "grad_norm": 0.9607473172752993, | |
| "learning_rate": 1.7301790727058344e-05, | |
| "loss": 0.8514, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2639175257731959, | |
| "grad_norm": 0.99121388037334, | |
| "learning_rate": 1.727654734885612e-05, | |
| "loss": 0.7258, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.26505510131532173, | |
| "grad_norm": 0.9178575823375061, | |
| "learning_rate": 1.7251205036686913e-05, | |
| "loss": 0.7234, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2661926768574476, | |
| "grad_norm": 0.9301556936555667, | |
| "learning_rate": 1.7225764135111867e-05, | |
| "loss": 0.7478, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2673302523995734, | |
| "grad_norm": 0.951109081614099, | |
| "learning_rate": 1.7200224990032577e-05, | |
| "loss": 0.7419, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.26846782794169927, | |
| "grad_norm": 0.9160961470408167, | |
| "learning_rate": 1.7174587948686375e-05, | |
| "loss": 0.826, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2696054034838251, | |
| "grad_norm": 1.057499137690602, | |
| "learning_rate": 1.7148853359641627e-05, | |
| "loss": 0.7286, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.27074297902595096, | |
| "grad_norm": 0.9295330730635258, | |
| "learning_rate": 1.7123021572792982e-05, | |
| "loss": 0.7782, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.2718805545680768, | |
| "grad_norm": 0.9975452115416533, | |
| "learning_rate": 1.7097092939356622e-05, | |
| "loss": 0.7519, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.27301813011020265, | |
| "grad_norm": 0.9230592215185579, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.7086, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2741557056523285, | |
| "grad_norm": 0.9191258196691511, | |
| "learning_rate": 1.7044946544164434e-05, | |
| "loss": 0.7791, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.27529328119445434, | |
| "grad_norm": 0.9937503180909131, | |
| "learning_rate": 1.7018729491405537e-05, | |
| "loss": 0.7141, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2764308567365802, | |
| "grad_norm": 0.9546990458403812, | |
| "learning_rate": 1.6992417010043144e-05, | |
| "loss": 0.9387, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.27756843227870603, | |
| "grad_norm": 0.9694759038261794, | |
| "learning_rate": 1.6966009457829088e-05, | |
| "loss": 0.7102, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2787060078208318, | |
| "grad_norm": 1.010514886312666, | |
| "learning_rate": 1.6939507193807818e-05, | |
| "loss": 0.7188, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.27984358336295767, | |
| "grad_norm": 0.9552012342525721, | |
| "learning_rate": 1.6912910578311503e-05, | |
| "loss": 0.7223, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2809811589050835, | |
| "grad_norm": 0.9577918842168219, | |
| "learning_rate": 1.688621997295515e-05, | |
| "loss": 0.8556, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.28211873444720936, | |
| "grad_norm": 0.9347260963706611, | |
| "learning_rate": 1.685943574063166e-05, | |
| "loss": 0.6894, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2832563099893352, | |
| "grad_norm": 0.9796117501965532, | |
| "learning_rate": 1.6832558245506937e-05, | |
| "loss": 0.8972, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.28439388553146105, | |
| "grad_norm": 0.9620774960278952, | |
| "learning_rate": 1.6805587853014895e-05, | |
| "loss": 0.8795, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2855314610735869, | |
| "grad_norm": 0.9779292355070162, | |
| "learning_rate": 1.6778524929852513e-05, | |
| "loss": 0.7849, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.28666903661571275, | |
| "grad_norm": 1.1391374042408926, | |
| "learning_rate": 1.6751369843974842e-05, | |
| "loss": 0.8935, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2878066121578386, | |
| "grad_norm": 0.9763354939311976, | |
| "learning_rate": 1.6724122964590002e-05, | |
| "loss": 0.6899, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.28894418769996444, | |
| "grad_norm": 1.0505980346872885, | |
| "learning_rate": 1.6696784662154165e-05, | |
| "loss": 0.7091, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2900817632420903, | |
| "grad_norm": 0.961207072486454, | |
| "learning_rate": 1.666935530836651e-05, | |
| "loss": 0.77, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.29121933878421613, | |
| "grad_norm": 1.022189843385111, | |
| "learning_rate": 1.6641835276164182e-05, | |
| "loss": 0.8021, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.292356914326342, | |
| "grad_norm": 0.9676687321518108, | |
| "learning_rate": 1.6614224939717217e-05, | |
| "loss": 0.7338, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.2934944898684678, | |
| "grad_norm": 1.031020389826052, | |
| "learning_rate": 1.658652467442345e-05, | |
| "loss": 0.8083, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.29463206541059367, | |
| "grad_norm": 0.9420183191034406, | |
| "learning_rate": 1.6558734856903406e-05, | |
| "loss": 0.651, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.2957696409527195, | |
| "grad_norm": 0.9746154980433869, | |
| "learning_rate": 1.6530855864995194e-05, | |
| "loss": 0.8235, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.29690721649484536, | |
| "grad_norm": 0.9637613954906972, | |
| "learning_rate": 1.650288807774937e-05, | |
| "loss": 0.7806, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.2980447920369712, | |
| "grad_norm": 0.9521280823765289, | |
| "learning_rate": 1.6474831875423765e-05, | |
| "loss": 0.7848, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.29918236757909705, | |
| "grad_norm": 0.9106705929775119, | |
| "learning_rate": 1.644668763947833e-05, | |
| "loss": 0.7158, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.3003199431212229, | |
| "grad_norm": 1.1111068932353543, | |
| "learning_rate": 1.6418455752569945e-05, | |
| "loss": 0.7949, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.30145751866334874, | |
| "grad_norm": 0.993566678238848, | |
| "learning_rate": 1.639013659854722e-05, | |
| "loss": 0.7514, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.3025950942054746, | |
| "grad_norm": 0.9530703838625129, | |
| "learning_rate": 1.6361730562445264e-05, | |
| "loss": 0.7807, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.30373266974760044, | |
| "grad_norm": 0.8949330809420218, | |
| "learning_rate": 1.6333238030480473e-05, | |
| "loss": 0.6888, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3048702452897263, | |
| "grad_norm": 0.8988538789326558, | |
| "learning_rate": 1.6304659390045253e-05, | |
| "loss": 0.7264, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.30600782083185213, | |
| "grad_norm": 1.0074831146025507, | |
| "learning_rate": 1.627599502970277e-05, | |
| "loss": 0.7832, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.307145396373978, | |
| "grad_norm": 0.9898618247508415, | |
| "learning_rate": 1.624724533918166e-05, | |
| "loss": 0.7412, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3082829719161038, | |
| "grad_norm": 0.9571059882247707, | |
| "learning_rate": 1.6218410709370735e-05, | |
| "loss": 0.7214, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.30942054745822967, | |
| "grad_norm": 1.0048100811814527, | |
| "learning_rate": 1.6189491532313665e-05, | |
| "loss": 0.8121, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3105581230003555, | |
| "grad_norm": 0.887082066422311, | |
| "learning_rate": 1.6160488201203643e-05, | |
| "loss": 0.8081, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.31169569854248136, | |
| "grad_norm": 1.003918167712565, | |
| "learning_rate": 1.6131401110378045e-05, | |
| "loss": 0.7442, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3128332740846072, | |
| "grad_norm": 0.9479739767796763, | |
| "learning_rate": 1.6102230655313076e-05, | |
| "loss": 0.7854, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.31397084962673305, | |
| "grad_norm": 0.9644306636148626, | |
| "learning_rate": 1.607297723261837e-05, | |
| "loss": 0.852, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3151084251688589, | |
| "grad_norm": 0.9590037676007116, | |
| "learning_rate": 1.6043641240031623e-05, | |
| "loss": 0.7551, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.3162460007109847, | |
| "grad_norm": 0.8703921927089772, | |
| "learning_rate": 1.6014223076413173e-05, | |
| "loss": 0.7418, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.31738357625311053, | |
| "grad_norm": 0.9581454527420551, | |
| "learning_rate": 1.5984723141740578e-05, | |
| "loss": 0.7117, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.3185211517952364, | |
| "grad_norm": 0.9806012064667501, | |
| "learning_rate": 1.595514183710317e-05, | |
| "loss": 0.8572, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3196587273373622, | |
| "grad_norm": 0.8953483309121879, | |
| "learning_rate": 1.592547956469662e-05, | |
| "loss": 0.8641, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.32079630287948807, | |
| "grad_norm": 0.8922059600985867, | |
| "learning_rate": 1.5895736727817457e-05, | |
| "loss": 0.7629, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.3219338784216139, | |
| "grad_norm": 0.924903003131493, | |
| "learning_rate": 1.5865913730857583e-05, | |
| "loss": 0.6772, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.32307145396373976, | |
| "grad_norm": 0.9253210263661452, | |
| "learning_rate": 1.5836010979298785e-05, | |
| "loss": 0.7024, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3242090295058656, | |
| "grad_norm": 0.9598832346761834, | |
| "learning_rate": 1.580602887970721e-05, | |
| "loss": 0.6624, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.32534660504799146, | |
| "grad_norm": 0.9502477784841414, | |
| "learning_rate": 1.5775967839727843e-05, | |
| "loss": 0.7923, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3264841805901173, | |
| "grad_norm": 0.9306848601779047, | |
| "learning_rate": 1.574582826807897e-05, | |
| "loss": 0.6541, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.32762175613224315, | |
| "grad_norm": 0.9355305194278158, | |
| "learning_rate": 1.5715610574546612e-05, | |
| "loss": 0.6918, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.328759331674369, | |
| "grad_norm": 1.0272254475487725, | |
| "learning_rate": 1.5685315169978955e-05, | |
| "loss": 0.8111, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 1.0334588348504568, | |
| "learning_rate": 1.565494246628077e-05, | |
| "loss": 0.7289, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "grad_norm": 0.9170599752827013, | |
| "learning_rate": 1.562449287640781e-05, | |
| "loss": 0.7852, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.33217205830074653, | |
| "grad_norm": 0.9222463038652008, | |
| "learning_rate": 1.5593966814361183e-05, | |
| "loss": 0.7118, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3333096338428724, | |
| "grad_norm": 1.0228401075311697, | |
| "learning_rate": 1.556336469518174e-05, | |
| "loss": 0.7166, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3344472093849982, | |
| "grad_norm": 0.9867811982354191, | |
| "learning_rate": 1.553268693494444e-05, | |
| "loss": 0.7237, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.33558478492712407, | |
| "grad_norm": 0.929004653861396, | |
| "learning_rate": 1.5501933950752655e-05, | |
| "loss": 0.7315, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3367223604692499, | |
| "grad_norm": 0.9711989755609595, | |
| "learning_rate": 1.5471106160732543e-05, | |
| "loss": 0.6733, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.33785993601137576, | |
| "grad_norm": 1.0543195742090818, | |
| "learning_rate": 1.5440203984027323e-05, | |
| "loss": 0.7116, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3389975115535016, | |
| "grad_norm": 1.0061655744197207, | |
| "learning_rate": 1.5409227840791617e-05, | |
| "loss": 0.7517, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.34013508709562745, | |
| "grad_norm": 0.9690488111835885, | |
| "learning_rate": 1.5378178152185703e-05, | |
| "loss": 0.8389, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.3412726626377533, | |
| "grad_norm": 0.9582246999248494, | |
| "learning_rate": 1.5347055340369806e-05, | |
| "loss": 0.798, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 879, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 435534387740672.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |