| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984399375975039, | |
| "eval_steps": 500, | |
| "global_step": 160, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0062402496099844, | |
| "grad_norm": 10.74402992025825, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.0885, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0124804992199688, | |
| "grad_norm": 10.414041822411015, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 2.0268, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0187207488299532, | |
| "grad_norm": 9.463277664294495, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 2.0224, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0249609984399376, | |
| "grad_norm": 3.816015682685396, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.0739, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.031201248049921998, | |
| "grad_norm": 2.45701507391519, | |
| "learning_rate": 2e-06, | |
| "loss": 2.0331, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0374414976599064, | |
| "grad_norm": 2.2119224271122664, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 2.0191, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0436817472698908, | |
| "grad_norm": 2.282301056375637, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.9805, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0499219968798752, | |
| "grad_norm": 2.8111235551059672, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 2.0581, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.056162246489859596, | |
| "grad_norm": 2.290032450883113, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 2.0564, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.062402496099843996, | |
| "grad_norm": 2.7448772203230587, | |
| "learning_rate": 4e-06, | |
| "loss": 2.0477, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0686427457098284, | |
| "grad_norm": 2.426609984220206, | |
| "learning_rate": 4.4e-06, | |
| "loss": 2.021, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0748829953198128, | |
| "grad_norm": 2.8668649784307085, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 2.0695, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0811232449297972, | |
| "grad_norm": 3.3242641740190484, | |
| "learning_rate": 5.200000000000001e-06, | |
| "loss": 2.0124, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0873634945397816, | |
| "grad_norm": 2.61279395903613, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 2.1133, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.093603744149766, | |
| "grad_norm": 5.14184681992228, | |
| "learning_rate": 6e-06, | |
| "loss": 2.043, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0998439937597504, | |
| "grad_norm": 2.8607405485409276, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 2.0186, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1060842433697348, | |
| "grad_norm": 3.907631913383047, | |
| "learning_rate": 6.8e-06, | |
| "loss": 2.0141, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11232449297971919, | |
| "grad_norm": 2.6838896552707014, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 2.0818, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.11856474258970359, | |
| "grad_norm": 3.6972529763342337, | |
| "learning_rate": 7.6e-06, | |
| "loss": 2.0148, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12480499219968799, | |
| "grad_norm": 2.887987766429437, | |
| "learning_rate": 8e-06, | |
| "loss": 1.9952, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1310452418096724, | |
| "grad_norm": 3.507471022279554, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 2.0801, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1372854914196568, | |
| "grad_norm": 1.8196990054764854, | |
| "learning_rate": 8.8e-06, | |
| "loss": 2.0431, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1435257410296412, | |
| "grad_norm": 4.634858392717085, | |
| "learning_rate": 9.2e-06, | |
| "loss": 2.0264, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1497659906396256, | |
| "grad_norm": 3.62013301472126, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 2.0417, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.15600624024961, | |
| "grad_norm": 3.8817500707011305, | |
| "learning_rate": 1e-05, | |
| "loss": 2.0338, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1622464898595944, | |
| "grad_norm": 4.087718448832571, | |
| "learning_rate": 1.0400000000000002e-05, | |
| "loss": 1.9754, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1684867394695788, | |
| "grad_norm": 3.0490578592864828, | |
| "learning_rate": 1.08e-05, | |
| "loss": 2.0386, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1747269890795632, | |
| "grad_norm": 2.9939386802357517, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 2.0044, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1809672386895476, | |
| "grad_norm": 3.935708629623511, | |
| "learning_rate": 1.16e-05, | |
| "loss": 2.0386, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.187207488299532, | |
| "grad_norm": 3.3632255605488686, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.0736, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1934477379095164, | |
| "grad_norm": 4.075768572911316, | |
| "learning_rate": 1.24e-05, | |
| "loss": 2.0516, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1996879875195008, | |
| "grad_norm": 1.9953301732913695, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 2.0221, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2059282371294852, | |
| "grad_norm": 5.026680509494266, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 1.9922, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2121684867394696, | |
| "grad_norm": 3.335169761696431, | |
| "learning_rate": 1.36e-05, | |
| "loss": 2.0422, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.21840873634945399, | |
| "grad_norm": 4.139988627533733, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 2.0286, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22464898595943839, | |
| "grad_norm": 3.7359622657764002, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 2.0034, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.23088923556942278, | |
| "grad_norm": 3.2432139656379233, | |
| "learning_rate": 1.4800000000000002e-05, | |
| "loss": 2.0811, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.23712948517940718, | |
| "grad_norm": 2.6586955034097532, | |
| "learning_rate": 1.52e-05, | |
| "loss": 2.02, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.24336973478939158, | |
| "grad_norm": 4.013453056484069, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 1.9795, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.24960998439937598, | |
| "grad_norm": 2.521776704363988, | |
| "learning_rate": 1.6e-05, | |
| "loss": 1.9903, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25585023400936036, | |
| "grad_norm": 4.065575685396291, | |
| "learning_rate": 1.6400000000000002e-05, | |
| "loss": 2.0221, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2620904836193448, | |
| "grad_norm": 3.5403076970966643, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 2.0709, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.26833073322932915, | |
| "grad_norm": 3.702151042696706, | |
| "learning_rate": 1.72e-05, | |
| "loss": 2.0208, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2745709828393136, | |
| "grad_norm": 2.591541370214587, | |
| "learning_rate": 1.76e-05, | |
| "loss": 2.0225, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.28081123244929795, | |
| "grad_norm": 3.486718721357557, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.0365, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2870514820592824, | |
| "grad_norm": 4.953839109375147, | |
| "learning_rate": 1.84e-05, | |
| "loss": 2.0201, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.29329173166926675, | |
| "grad_norm": 3.293398181575717, | |
| "learning_rate": 1.8800000000000003e-05, | |
| "loss": 2.0134, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2995319812792512, | |
| "grad_norm": 5.579089802583317, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 1.9558, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.30577223088923555, | |
| "grad_norm": 4.544122172306774, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.9454, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.31201248049922, | |
| "grad_norm": 5.4447638064914985, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0016, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.31825273010920435, | |
| "grad_norm": 9.597372129578162, | |
| "learning_rate": 1.9981668194317142e-05, | |
| "loss": 1.9972, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3244929797191888, | |
| "grad_norm": 27.057608263740534, | |
| "learning_rate": 1.9963031423290204e-05, | |
| "loss": 2.0419, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.33073322932917315, | |
| "grad_norm": 7.246050894562496, | |
| "learning_rate": 1.994408201304753e-05, | |
| "loss": 2.03, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3369734789391576, | |
| "grad_norm": 11.339707951526051, | |
| "learning_rate": 1.9924812030075188e-05, | |
| "loss": 2.0708, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.34321372854914195, | |
| "grad_norm": 54.33866443144764, | |
| "learning_rate": 1.9905213270142184e-05, | |
| "loss": 2.1948, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3494539781591264, | |
| "grad_norm": 10.82899390141888, | |
| "learning_rate": 1.988527724665392e-05, | |
| "loss": 2.1432, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.35569422776911075, | |
| "grad_norm": 4.067023362710712, | |
| "learning_rate": 1.986499517839923e-05, | |
| "loss": 2.0887, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3619344773790952, | |
| "grad_norm": 5.822340342195218, | |
| "learning_rate": 1.9844357976653696e-05, | |
| "loss": 2.0353, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.36817472698907955, | |
| "grad_norm": 3.5192080522606117, | |
| "learning_rate": 1.9823356231599608e-05, | |
| "loss": 2.0402, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.374414976599064, | |
| "grad_norm": 4.0093539975984065, | |
| "learning_rate": 1.9801980198019806e-05, | |
| "loss": 2.0968, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.38065522620904835, | |
| "grad_norm": 3.248210484978779, | |
| "learning_rate": 1.9780219780219784e-05, | |
| "loss": 2.0698, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3868954758190328, | |
| "grad_norm": 9.86816642625139, | |
| "learning_rate": 1.9758064516129035e-05, | |
| "loss": 2.1396, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.39313572542901715, | |
| "grad_norm": 6.803518728015235, | |
| "learning_rate": 1.9735503560528995e-05, | |
| "loss": 2.0915, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3993759750390016, | |
| "grad_norm": 9.116753527737151, | |
| "learning_rate": 1.9712525667351133e-05, | |
| "loss": 2.1454, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.40561622464898595, | |
| "grad_norm": 5.696843981671945, | |
| "learning_rate": 1.9689119170984456e-05, | |
| "loss": 2.1245, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4118564742589704, | |
| "grad_norm": 14.915888442988669, | |
| "learning_rate": 1.96652719665272e-05, | |
| "loss": 2.121, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.41809672386895474, | |
| "grad_norm": 96.78241254851471, | |
| "learning_rate": 1.9640971488912355e-05, | |
| "loss": 2.3768, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4243369734789392, | |
| "grad_norm": 11.648041872276787, | |
| "learning_rate": 1.961620469083156e-05, | |
| "loss": 2.2285, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.43057722308892354, | |
| "grad_norm": 7.174165495387168, | |
| "learning_rate": 1.9590958019375676e-05, | |
| "loss": 2.1671, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.43681747269890797, | |
| "grad_norm": 6.48742758649308, | |
| "learning_rate": 1.956521739130435e-05, | |
| "loss": 2.1101, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.44305772230889234, | |
| "grad_norm": 3.8309613955567263, | |
| "learning_rate": 1.953896816684962e-05, | |
| "loss": 2.1285, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.44929797191887677, | |
| "grad_norm": 4.383193420557969, | |
| "learning_rate": 1.9512195121951222e-05, | |
| "loss": 2.1133, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.45553822152886114, | |
| "grad_norm": 3.9756337036474765, | |
| "learning_rate": 1.9484882418812992e-05, | |
| "loss": 2.05, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.46177847113884557, | |
| "grad_norm": 2.628853316257943, | |
| "learning_rate": 1.9457013574660635e-05, | |
| "loss": 2.0764, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.46801872074882994, | |
| "grad_norm": 3.9077882447399372, | |
| "learning_rate": 1.942857142857143e-05, | |
| "loss": 2.0569, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.47425897035881437, | |
| "grad_norm": 2.153564772831578, | |
| "learning_rate": 1.9399538106235567e-05, | |
| "loss": 1.9753, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.48049921996879874, | |
| "grad_norm": 3.2742987852933907, | |
| "learning_rate": 1.9369894982497083e-05, | |
| "loss": 2.0728, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.48673946957878317, | |
| "grad_norm": 3.0871656421066236, | |
| "learning_rate": 1.9339622641509436e-05, | |
| "loss": 2.0725, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.49297971918876754, | |
| "grad_norm": 4.974500703202643, | |
| "learning_rate": 1.9308700834326583e-05, | |
| "loss": 2.0574, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.49921996879875197, | |
| "grad_norm": 1.9848768489870379, | |
| "learning_rate": 1.9277108433734944e-05, | |
| "loss": 2.039, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5054602184087363, | |
| "grad_norm": 3.675535765562546, | |
| "learning_rate": 1.9244823386114496e-05, | |
| "loss": 2.0498, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5117004680187207, | |
| "grad_norm": 3.646020776409229, | |
| "learning_rate": 1.9211822660098524e-05, | |
| "loss": 2.09, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5179407176287052, | |
| "grad_norm": 3.2885099554782, | |
| "learning_rate": 1.9178082191780822e-05, | |
| "loss": 2.0594, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5241809672386896, | |
| "grad_norm": 3.86257607313755, | |
| "learning_rate": 1.9143576826196474e-05, | |
| "loss": 2.0982, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5304212168486739, | |
| "grad_norm": 2.9398571293411337, | |
| "learning_rate": 1.9108280254777068e-05, | |
| "loss": 2.0192, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5366614664586583, | |
| "grad_norm": 5.691015626741047, | |
| "learning_rate": 1.907216494845361e-05, | |
| "loss": 2.0719, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5429017160686428, | |
| "grad_norm": 3.2393091526236204, | |
| "learning_rate": 1.9035202086049548e-05, | |
| "loss": 2.0201, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5491419656786272, | |
| "grad_norm": 7.878611924949738, | |
| "learning_rate": 1.899736147757256e-05, | |
| "loss": 2.0513, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5553822152886115, | |
| "grad_norm": 5.586452362544675, | |
| "learning_rate": 1.895861148197597e-05, | |
| "loss": 2.008, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5616224648985959, | |
| "grad_norm": 6.366113686688125, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 2.0222, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5678627145085804, | |
| "grad_norm": 5.5469754048863305, | |
| "learning_rate": 1.887824897400821e-05, | |
| "loss": 2.0377, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5741029641185648, | |
| "grad_norm": 4.927264399708736, | |
| "learning_rate": 1.883656509695291e-05, | |
| "loss": 2.0055, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5803432137285491, | |
| "grad_norm": 4.400828802773919, | |
| "learning_rate": 1.8793828892005612e-05, | |
| "loss": 2.0385, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5865834633385335, | |
| "grad_norm": 5.1654734914491, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 2.033, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.592823712948518, | |
| "grad_norm": 2.8256214952031224, | |
| "learning_rate": 1.8705035971223024e-05, | |
| "loss": 2.0456, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5990639625585024, | |
| "grad_norm": 5.945292262432217, | |
| "learning_rate": 1.8658892128279884e-05, | |
| "loss": 2.0766, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6053042121684867, | |
| "grad_norm": 3.1462260207166923, | |
| "learning_rate": 1.861152141802068e-05, | |
| "loss": 2.0502, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6115444617784711, | |
| "grad_norm": 6.002424537736364, | |
| "learning_rate": 1.8562874251497008e-05, | |
| "loss": 2.0255, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6177847113884556, | |
| "grad_norm": 4.92987161539047, | |
| "learning_rate": 1.851289833080425e-05, | |
| "loss": 2.0076, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.62402496099844, | |
| "grad_norm": 4.302040549632859, | |
| "learning_rate": 1.846153846153846e-05, | |
| "loss": 2.0523, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6302652106084243, | |
| "grad_norm": 2.838409708113077, | |
| "learning_rate": 1.840873634945398e-05, | |
| "loss": 1.9646, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6365054602184087, | |
| "grad_norm": 3.295108362363732, | |
| "learning_rate": 1.8354430379746836e-05, | |
| "loss": 1.9575, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6427457098283932, | |
| "grad_norm": 2.2894649638697557, | |
| "learning_rate": 1.8298555377207063e-05, | |
| "loss": 1.9964, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6489859594383776, | |
| "grad_norm": 3.7807867883237343, | |
| "learning_rate": 1.8241042345276876e-05, | |
| "loss": 1.9929, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6552262090483619, | |
| "grad_norm": 3.5213953795099817, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 2.0305, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6614664586583463, | |
| "grad_norm": 3.180581118333888, | |
| "learning_rate": 1.8120805369127517e-05, | |
| "loss": 2.0, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6677067082683308, | |
| "grad_norm": 3.7407765148739154, | |
| "learning_rate": 1.8057921635434413e-05, | |
| "loss": 1.9947, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6739469578783152, | |
| "grad_norm": 2.4800887886136302, | |
| "learning_rate": 1.7993079584775085e-05, | |
| "loss": 1.945, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6801872074882995, | |
| "grad_norm": 2.280159270411404, | |
| "learning_rate": 1.792618629173989e-05, | |
| "loss": 2.0461, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6864274570982839, | |
| "grad_norm": 1.8103075862546911, | |
| "learning_rate": 1.785714285714286e-05, | |
| "loss": 2.0384, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6926677067082684, | |
| "grad_norm": 2.7530066817239116, | |
| "learning_rate": 1.7785843920145195e-05, | |
| "loss": 2.0422, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6989079563182528, | |
| "grad_norm": 2.385559179108808, | |
| "learning_rate": 1.771217712177122e-05, | |
| "loss": 2.0201, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7051482059282371, | |
| "grad_norm": 3.4132861718816656, | |
| "learning_rate": 1.7636022514071294e-05, | |
| "loss": 2.0059, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7113884555382215, | |
| "grad_norm": 1.9687493474305817, | |
| "learning_rate": 1.755725190839695e-05, | |
| "loss": 1.9424, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.717628705148206, | |
| "grad_norm": 3.3664419654172812, | |
| "learning_rate": 1.7475728155339808e-05, | |
| "loss": 1.9917, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7238689547581904, | |
| "grad_norm": 1.7453289663664593, | |
| "learning_rate": 1.7391304347826085e-05, | |
| "loss": 1.9943, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7301092043681747, | |
| "grad_norm": 3.513395200138233, | |
| "learning_rate": 1.7303822937625752e-05, | |
| "loss": 2.0497, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7363494539781591, | |
| "grad_norm": 2.680135793492889, | |
| "learning_rate": 1.721311475409836e-05, | |
| "loss": 2.0534, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7425897035881436, | |
| "grad_norm": 2.921520804571559, | |
| "learning_rate": 1.711899791231733e-05, | |
| "loss": 1.9642, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.748829953198128, | |
| "grad_norm": 2.3783552559811594, | |
| "learning_rate": 1.7021276595744686e-05, | |
| "loss": 2.0363, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7550702028081123, | |
| "grad_norm": 3.4009796330846567, | |
| "learning_rate": 1.691973969631237e-05, | |
| "loss": 2.0112, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7613104524180967, | |
| "grad_norm": 3.173059834064885, | |
| "learning_rate": 1.6814159292035402e-05, | |
| "loss": 2.036, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7675507020280812, | |
| "grad_norm": 1.9816975235472727, | |
| "learning_rate": 1.670428893905192e-05, | |
| "loss": 2.0391, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7737909516380655, | |
| "grad_norm": 2.3017687165242755, | |
| "learning_rate": 1.6589861751152075e-05, | |
| "loss": 2.0232, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7800312012480499, | |
| "grad_norm": 2.094432126031965, | |
| "learning_rate": 1.647058823529412e-05, | |
| "loss": 2.0496, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7862714508580343, | |
| "grad_norm": 4.369204228615871, | |
| "learning_rate": 1.6346153846153847e-05, | |
| "loss": 1.995, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7925117004680188, | |
| "grad_norm": 2.707513968938277, | |
| "learning_rate": 1.6216216216216215e-05, | |
| "loss": 2.0909, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7987519500780031, | |
| "grad_norm": 3.3723262775787437, | |
| "learning_rate": 1.608040201005025e-05, | |
| "loss": 2.0436, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8049921996879875, | |
| "grad_norm": 4.542572647672642, | |
| "learning_rate": 1.5938303341902316e-05, | |
| "loss": 2.0467, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8112324492979719, | |
| "grad_norm": 2.752970144346649, | |
| "learning_rate": 1.5789473684210526e-05, | |
| "loss": 2.026, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8174726989079563, | |
| "grad_norm": 5.8547013314456695, | |
| "learning_rate": 1.5633423180592992e-05, | |
| "loss": 2.0379, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8237129485179407, | |
| "grad_norm": 3.6188017739630824, | |
| "learning_rate": 1.5469613259668508e-05, | |
| "loss": 1.9989, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8299531981279251, | |
| "grad_norm": 7.450963398151789, | |
| "learning_rate": 1.529745042492918e-05, | |
| "loss": 1.9937, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8361934477379095, | |
| "grad_norm": 6.338885342053262, | |
| "learning_rate": 1.5116279069767441e-05, | |
| "loss": 2.0268, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8424336973478939, | |
| "grad_norm": 4.068350711677805, | |
| "learning_rate": 1.4925373134328359e-05, | |
| "loss": 2.0657, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8486739469578783, | |
| "grad_norm": 5.595793283481958, | |
| "learning_rate": 1.4723926380368096e-05, | |
| "loss": 2.0018, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8549141965678627, | |
| "grad_norm": 2.188112379086696, | |
| "learning_rate": 1.4511041009463722e-05, | |
| "loss": 1.9973, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8611544461778471, | |
| "grad_norm": 2.712340447399982, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 1.9837, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8673946957878315, | |
| "grad_norm": 4.524782432904714, | |
| "learning_rate": 1.4046822742474917e-05, | |
| "loss": 1.9805, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.8736349453978159, | |
| "grad_norm": 2.5304287757872026, | |
| "learning_rate": 1.3793103448275862e-05, | |
| "loss": 2.0088, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8798751950078003, | |
| "grad_norm": 5.924048202005877, | |
| "learning_rate": 1.3523131672597865e-05, | |
| "loss": 1.9997, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8861154446177847, | |
| "grad_norm": 5.465435613727891, | |
| "learning_rate": 1.323529411764706e-05, | |
| "loss": 1.9556, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8923556942277691, | |
| "grad_norm": 3.404337115153408, | |
| "learning_rate": 1.2927756653992395e-05, | |
| "loss": 2.0268, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8985959438377535, | |
| "grad_norm": 4.2172697858979165, | |
| "learning_rate": 1.2598425196850392e-05, | |
| "loss": 1.9997, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9048361934477379, | |
| "grad_norm": 2.4532813229494255, | |
| "learning_rate": 1.2244897959183674e-05, | |
| "loss": 2.0326, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9110764430577223, | |
| "grad_norm": 1.9619593022855364, | |
| "learning_rate": 1.1864406779661018e-05, | |
| "loss": 1.9498, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9173166926677067, | |
| "grad_norm": 5.121507704839843, | |
| "learning_rate": 1.1453744493392071e-05, | |
| "loss": 1.9838, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9235569422776911, | |
| "grad_norm": 4.5855609090084855, | |
| "learning_rate": 1.1009174311926605e-05, | |
| "loss": 1.9952, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9297971918876755, | |
| "grad_norm": 2.8048531358016215, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 1.9376, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9360374414976599, | |
| "grad_norm": 3.1695682113009767, | |
| "learning_rate": 1e-05, | |
| "loss": 1.9892, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9422776911076443, | |
| "grad_norm": 2.562068085601565, | |
| "learning_rate": 9.424083769633508e-06, | |
| "loss": 1.989, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9485179407176287, | |
| "grad_norm": 2.5588451099258958, | |
| "learning_rate": 8.79120879120879e-06, | |
| "loss": 2.0055, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9547581903276131, | |
| "grad_norm": 2.346708882799627, | |
| "learning_rate": 8.092485549132947e-06, | |
| "loss": 1.9344, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9609984399375975, | |
| "grad_norm": 1.9275298796427967, | |
| "learning_rate": 7.3170731707317065e-06, | |
| "loss": 1.9777, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9672386895475819, | |
| "grad_norm": 2.3144404690009934, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 2.0236, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9734789391575663, | |
| "grad_norm": 2.0378918646209385, | |
| "learning_rate": 5.479452054794519e-06, | |
| "loss": 1.9846, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.9797191887675507, | |
| "grad_norm": 1.653128950918772, | |
| "learning_rate": 4.379562043795621e-06, | |
| "loss": 1.933, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9859594383775351, | |
| "grad_norm": 1.5724161203243192, | |
| "learning_rate": 3.125e-06, | |
| "loss": 2.0316, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9921996879875195, | |
| "grad_norm": 0.8954010032529827, | |
| "learning_rate": 1.680672268907563e-06, | |
| "loss": 1.9258, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9984399375975039, | |
| "grad_norm": 1.0745203334736726, | |
| "learning_rate": 0, | |
| "loss": 2.0094, | |
| "step": 160 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 160, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 80, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0297691278239334e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |