| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.514661547456439, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0549, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.533257734742273, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0511, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.5947055173496105, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.0383, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.897727392552024, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9997, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.848965721562149, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.9823, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.4837358326339793, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9848, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.917146851772698, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9749, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.273073471577452, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9146, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.334372987494823, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9845, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.9629922479937116, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9391, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.382492770804012, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.8903, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.301626836469744, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.853, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.6179229339847265, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8749, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.9312036795892737, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8362, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.7051780756675279, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.7995, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.5030467576206636, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.8066, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.5439271706140738, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.8256, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.1453811996350687, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7774, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.1888966844740092, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7615, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1643146198907204, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7339, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.0783369819320254, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.6831, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.109436748040121, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.7157, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.9560383867296256, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.7293, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.117357157247665, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.6747, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.0753463950874016, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6971, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.9143616578209675, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.6837, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.9629882797454461, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6962, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.9546749884256414, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6589, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.8394119932746013, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6753, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8658323051506474, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6649, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.9365436247619137, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.6274, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7958921177535546, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.6116, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.7041434098820389, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.6306, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.7542640203435573, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.5907, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.8050782710234841, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.5944, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.8395441860638456, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.6054, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.6533520807886277, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.5681, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.8712063195963011, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.594, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.7058104742293924, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5539, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7507552480835233, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.5393, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.8282932005815182, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.5023, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.8273806301363852, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.4942, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8902697044853984, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.4967, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.7364831159921146, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.514, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.8288131573506206, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.5063, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.9376885983764791, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.5441, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.6835316034344262, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.4926, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.8413196098724388, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.4946, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.87099746359794, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.4776, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7021081208150572, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.4644, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.7470274598476523, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.4843, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.854534847550328, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4261, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.6949880463483018, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.4571, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.7860591810795131, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.4548, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.7197876269702297, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.4272, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.6705979898673782, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.3888, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.6866795582640342, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.3923, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6428479429474946, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.419, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.6775894435925355, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.3869, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6330137073638697, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.4152, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.652913667892219, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3995, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.672730542064185, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.3399, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7690027705801818, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.3711, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.6776531072398, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.3412, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.7059136278652931, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3569, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.745808315425912, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.3249, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.6426994235256039, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.3443, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.749325937029054, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.3375, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.6816631607927595, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.364, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.6061610214861256, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.3382, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.7595170940871228, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.2923, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.6020874715367359, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.2953, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.7912139007597716, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.2742, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.7338816222416935, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.3359, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6154369816576161, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.2916, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.5722970079784565, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.2752, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.6167620552547031, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.276, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.6544524216832458, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2568, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6365023079127102, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.2625, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6346319771918996, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.2543, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.7935687175284898, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.2188, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.563137680740866, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.2594, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.7287091402892747, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.2511, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6979911227285388, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.2189, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.6550970784683662, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.2019, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.6593022512925845, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.197, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6320539903327664, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.2396, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.695082578305977, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.2237, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.6032572012823637, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.2363, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5868210308406878, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.2338, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.8246021334217514, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1778, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.6326337124189978, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.2031, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.7735059024636389, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.2091, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.9856385466355869, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.2129, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.6339059124868586, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.2152, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.6633867202226957, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.1992, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.712977947778065, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1956, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.7242428987534822, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.1857, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.5817520512410859, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.1572, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.6485964844533239, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1681, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.5917058155229449, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.1885, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.5816880146125386, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.2073, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.556220391629839, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1688, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.5717430073099484, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.1603, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.6232664790444935, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.1668, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.5620523647118822, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.1304, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.5648688866765015, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.1881, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.6164463929550619, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1598, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.5255323924827229, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1321, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.5354541207110695, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1632, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.572740834084801, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1716, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.527643488384967, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1612, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.5071051634331077, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.1264, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.5120761265774432, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.1906, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.47424525910135223, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.1318, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.5124800689018203, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.1424, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.4835960985123145, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1546, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.519568896032822, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.1629, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.47404692918153074, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1459, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5218029774258995, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.1524, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.48716199602214394, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1433, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.5484534868846779, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.1647, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.49480611722982415, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1422, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.43267619600382706, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.1295, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.47787014903068536, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1799, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.4620611815233099, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1329, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.4658866845782663, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1432, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.4684502250321844, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1537, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.46150264180009387, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.1825, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.4392945731575745, | |
| "learning_rate": 0.0, | |
| "loss": 0.1127, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 7.177620479279104e+16, | |
| "train_loss": 0.4257054453858962, | |
| "train_runtime": 1962.7651, | |
| "train_samples_per_second": 2.093, | |
| "train_steps_per_second": 0.066 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "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": 7.177620479279104e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |