| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 173, | |
| "global_step": 689, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001451378809869376, | |
| "grad_norm": 0.1842290163040161, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0914, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001451378809869376, | |
| "eval_loss": 0.3193407654762268, | |
| "eval_runtime": 32.4089, | |
| "eval_samples_per_second": 8.979, | |
| "eval_steps_per_second": 4.505, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002902757619738752, | |
| "grad_norm": 0.2510640323162079, | |
| "learning_rate": 4e-05, | |
| "loss": 3.2586, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0043541364296081275, | |
| "grad_norm": 0.13390600681304932, | |
| "learning_rate": 6e-05, | |
| "loss": 0.6678, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005805515239477504, | |
| "grad_norm": 0.13102972507476807, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8674, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00725689404934688, | |
| "grad_norm": 0.3714960515499115, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7366, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008708272859216255, | |
| "grad_norm": 0.17309564352035522, | |
| "learning_rate": 0.00012, | |
| "loss": 1.0647, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010159651669085631, | |
| "grad_norm": 0.19592316448688507, | |
| "learning_rate": 0.00014, | |
| "loss": 1.7067, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.011611030478955007, | |
| "grad_norm": 0.20908816158771515, | |
| "learning_rate": 0.00016, | |
| "loss": 0.8817, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.013062409288824383, | |
| "grad_norm": 0.2101982682943344, | |
| "learning_rate": 0.00018, | |
| "loss": 1.4636, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01451378809869376, | |
| "grad_norm": 0.3246542513370514, | |
| "learning_rate": 0.0002, | |
| "loss": 1.7697, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015965166908563134, | |
| "grad_norm": 0.3263128697872162, | |
| "learning_rate": 0.00019999892964109545, | |
| "loss": 2.3342, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01741654571843251, | |
| "grad_norm": 0.240472674369812, | |
| "learning_rate": 0.00019999571858729512, | |
| "loss": 0.7013, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.018867924528301886, | |
| "grad_norm": 0.26579201221466064, | |
| "learning_rate": 0.00019999036690733865, | |
| "loss": 1.219, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.020319303338171262, | |
| "grad_norm": 0.37741780281066895, | |
| "learning_rate": 0.00019998287471579036, | |
| "loss": 1.8527, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02177068214804064, | |
| "grad_norm": 0.2223397046327591, | |
| "learning_rate": 0.00019997324217303696, | |
| "loss": 0.3127, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.023222060957910014, | |
| "grad_norm": 0.3512834906578064, | |
| "learning_rate": 0.000199961469485284, | |
| "loss": 1.4699, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02467343976777939, | |
| "grad_norm": 0.3912449777126312, | |
| "learning_rate": 0.00019994755690455152, | |
| "loss": 1.3437, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.026124818577648767, | |
| "grad_norm": 0.3254362642765045, | |
| "learning_rate": 0.00019993150472866856, | |
| "loss": 0.8423, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.027576197387518143, | |
| "grad_norm": 0.3357979357242584, | |
| "learning_rate": 0.00019991331330126698, | |
| "loss": 0.7841, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.02902757619738752, | |
| "grad_norm": 0.2792049050331116, | |
| "learning_rate": 0.00019989298301177386, | |
| "loss": 0.5499, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.030478955007256895, | |
| "grad_norm": 0.31481030583381653, | |
| "learning_rate": 0.00019987051429540332, | |
| "loss": 1.4346, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03193033381712627, | |
| "grad_norm": 0.2967741787433624, | |
| "learning_rate": 0.00019984590763314722, | |
| "loss": 0.7228, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.033381712626995644, | |
| "grad_norm": 0.33911633491516113, | |
| "learning_rate": 0.00019981916355176472, | |
| "loss": 0.871, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03483309143686502, | |
| "grad_norm": 0.32208600640296936, | |
| "learning_rate": 0.00019979028262377118, | |
| "loss": 1.1648, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.036284470246734396, | |
| "grad_norm": 0.26672494411468506, | |
| "learning_rate": 0.00019975926546742572, | |
| "loss": 0.4856, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03773584905660377, | |
| "grad_norm": 0.48599860072135925, | |
| "learning_rate": 0.00019972611274671814, | |
| "loss": 2.17, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03918722786647315, | |
| "grad_norm": 0.22017762064933777, | |
| "learning_rate": 0.00019969082517135462, | |
| "loss": 0.5839, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.040638606676342524, | |
| "grad_norm": 0.3883523941040039, | |
| "learning_rate": 0.0001996534034967426, | |
| "loss": 1.5153, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0420899854862119, | |
| "grad_norm": 0.29745909571647644, | |
| "learning_rate": 0.00019961384852397457, | |
| "loss": 0.8633, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04354136429608128, | |
| "grad_norm": 0.25378555059432983, | |
| "learning_rate": 0.00019957216109981078, | |
| "loss": 0.9834, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04499274310595065, | |
| "grad_norm": 0.3489767014980316, | |
| "learning_rate": 0.0001995283421166614, | |
| "loss": 0.5789, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.04644412191582003, | |
| "grad_norm": 0.29305726289749146, | |
| "learning_rate": 0.00019948239251256722, | |
| "loss": 0.5507, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.047895500725689405, | |
| "grad_norm": 0.5047852993011475, | |
| "learning_rate": 0.00019943431327117957, | |
| "loss": 1.4252, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.04934687953555878, | |
| "grad_norm": 0.3907155692577362, | |
| "learning_rate": 0.00019938410542173936, | |
| "loss": 1.2378, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05079825834542816, | |
| "grad_norm": 0.36083394289016724, | |
| "learning_rate": 0.00019933177003905492, | |
| "loss": 0.7357, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05224963715529753, | |
| "grad_norm": 0.35112664103507996, | |
| "learning_rate": 0.00019927730824347915, | |
| "loss": 0.9383, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05370101596516691, | |
| "grad_norm": 0.37332794070243835, | |
| "learning_rate": 0.00019922072120088537, | |
| "loss": 0.6428, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.055152394775036286, | |
| "grad_norm": 0.3495066165924072, | |
| "learning_rate": 0.00019916201012264254, | |
| "loss": 2.2821, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.05660377358490566, | |
| "grad_norm": 0.3157879412174225, | |
| "learning_rate": 0.0001991011762655891, | |
| "loss": 0.9946, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.05805515239477504, | |
| "grad_norm": 0.35262367129325867, | |
| "learning_rate": 0.00019903822093200627, | |
| "loss": 1.8988, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.059506531204644414, | |
| "grad_norm": 0.4413168728351593, | |
| "learning_rate": 0.00019897314546959014, | |
| "loss": 1.0766, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06095791001451379, | |
| "grad_norm": 0.521634578704834, | |
| "learning_rate": 0.0001989059512714227, | |
| "loss": 1.5654, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.062409288824383166, | |
| "grad_norm": 0.38300636410713196, | |
| "learning_rate": 0.00019883663977594208, | |
| "loss": 0.8673, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06386066763425254, | |
| "grad_norm": 0.3619363009929657, | |
| "learning_rate": 0.00019876521246691182, | |
| "loss": 1.2571, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06531204644412192, | |
| "grad_norm": 0.5837172865867615, | |
| "learning_rate": 0.00019869167087338907, | |
| "loss": 1.6372, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06676342525399129, | |
| "grad_norm": 0.2846507430076599, | |
| "learning_rate": 0.00019861601656969184, | |
| "loss": 0.9893, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.06821480406386067, | |
| "grad_norm": 0.3829323649406433, | |
| "learning_rate": 0.00019853825117536523, | |
| "loss": 1.5662, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.06966618287373004, | |
| "grad_norm": 0.3632449805736542, | |
| "learning_rate": 0.0001984583763551469, | |
| "loss": 1.7041, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07111756168359942, | |
| "grad_norm": 0.2692880928516388, | |
| "learning_rate": 0.00019837639381893135, | |
| "loss": 0.2875, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.07256894049346879, | |
| "grad_norm": 0.4161498248577118, | |
| "learning_rate": 0.00019829230532173337, | |
| "loss": 1.6918, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07402031930333818, | |
| "grad_norm": 0.45441752672195435, | |
| "learning_rate": 0.00019820611266365036, | |
| "loss": 0.5773, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07547169811320754, | |
| "grad_norm": 0.3953988254070282, | |
| "learning_rate": 0.0001981178176898239, | |
| "loss": 2.3749, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 0.2471064329147339, | |
| "learning_rate": 0.00019802742229040025, | |
| "loss": 0.3787, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0783744557329463, | |
| "grad_norm": 0.2859368622303009, | |
| "learning_rate": 0.0001979349284004898, | |
| "loss": 0.636, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07982583454281568, | |
| "grad_norm": 0.3482446074485779, | |
| "learning_rate": 0.00019784033800012568, | |
| "loss": 1.0302, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08127721335268505, | |
| "grad_norm": 0.7591869235038757, | |
| "learning_rate": 0.00019774365311422154, | |
| "loss": 2.006, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08272859216255443, | |
| "grad_norm": 0.33345961570739746, | |
| "learning_rate": 0.00019764487581252785, | |
| "loss": 1.4342, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0841799709724238, | |
| "grad_norm": 0.36986222863197327, | |
| "learning_rate": 0.00019754400820958798, | |
| "loss": 0.8316, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08563134978229318, | |
| "grad_norm": 0.3581068217754364, | |
| "learning_rate": 0.00019744105246469263, | |
| "loss": 0.8619, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08708272859216255, | |
| "grad_norm": 0.3945651948451996, | |
| "learning_rate": 0.00019733601078183375, | |
| "loss": 1.5655, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08853410740203194, | |
| "grad_norm": 0.4055555462837219, | |
| "learning_rate": 0.00019722888540965735, | |
| "loss": 0.843, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0899854862119013, | |
| "grad_norm": 0.3120388090610504, | |
| "learning_rate": 0.0001971196786414154, | |
| "loss": 1.7423, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.09143686502177069, | |
| "grad_norm": 0.3523203432559967, | |
| "learning_rate": 0.00019700839281491662, | |
| "loss": 1.3252, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.09288824383164006, | |
| "grad_norm": 0.49849557876586914, | |
| "learning_rate": 0.00019689503031247648, | |
| "loss": 1.0555, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.09433962264150944, | |
| "grad_norm": 0.43524911999702454, | |
| "learning_rate": 0.0001967795935608663, | |
| "loss": 1.2491, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09579100145137881, | |
| "grad_norm": 0.2813914120197296, | |
| "learning_rate": 0.00019666208503126112, | |
| "loss": 0.9261, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09724238026124818, | |
| "grad_norm": 0.3467751741409302, | |
| "learning_rate": 0.00019654250723918706, | |
| "loss": 1.1822, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.09869375907111756, | |
| "grad_norm": 0.3063952326774597, | |
| "learning_rate": 0.00019642086274446711, | |
| "loss": 1.0751, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10014513788098693, | |
| "grad_norm": 0.2954864203929901, | |
| "learning_rate": 0.0001962971541511667, | |
| "loss": 1.4695, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10159651669085631, | |
| "grad_norm": 0.2937740683555603, | |
| "learning_rate": 0.0001961713841075377, | |
| "loss": 0.5735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10304789550072568, | |
| "grad_norm": 0.36586064100265503, | |
| "learning_rate": 0.00019604355530596182, | |
| "loss": 1.2774, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.10449927431059507, | |
| "grad_norm": 0.4228026866912842, | |
| "learning_rate": 0.00019591367048289298, | |
| "loss": 1.4011, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.10595065312046444, | |
| "grad_norm": 0.26343077421188354, | |
| "learning_rate": 0.00019578173241879872, | |
| "loss": 0.9024, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.10740203193033382, | |
| "grad_norm": 0.4526161551475525, | |
| "learning_rate": 0.00019564774393810068, | |
| "loss": 1.6497, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.10885341074020319, | |
| "grad_norm": 0.4131157696247101, | |
| "learning_rate": 0.00019551170790911416, | |
| "loss": 1.351, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11030478955007257, | |
| "grad_norm": 0.251254141330719, | |
| "learning_rate": 0.0001953736272439866, | |
| "loss": 0.4524, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11175616835994194, | |
| "grad_norm": 0.43410056829452515, | |
| "learning_rate": 0.00019523350489863544, | |
| "loss": 1.2543, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.11320754716981132, | |
| "grad_norm": 0.24558404088020325, | |
| "learning_rate": 0.00019509134387268463, | |
| "loss": 0.6434, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.11465892597968069, | |
| "grad_norm": 0.40656939148902893, | |
| "learning_rate": 0.00019494714720940065, | |
| "loss": 1.1747, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.11611030478955008, | |
| "grad_norm": 0.5807479023933411, | |
| "learning_rate": 0.00019480091799562704, | |
| "loss": 1.3046, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11756168359941944, | |
| "grad_norm": 0.30079004168510437, | |
| "learning_rate": 0.00019465265936171873, | |
| "loss": 0.6232, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.11901306240928883, | |
| "grad_norm": 0.3846030533313751, | |
| "learning_rate": 0.00019450237448147463, | |
| "loss": 1.3327, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1204644412191582, | |
| "grad_norm": 0.33520063757896423, | |
| "learning_rate": 0.00019435006657206997, | |
| "loss": 1.623, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.12191582002902758, | |
| "grad_norm": 0.3635580241680145, | |
| "learning_rate": 0.0001941957388939873, | |
| "loss": 1.2121, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.12336719883889695, | |
| "grad_norm": 0.2681978642940521, | |
| "learning_rate": 0.00019403939475094663, | |
| "loss": 0.4755, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12481857764876633, | |
| "grad_norm": 0.3734763562679291, | |
| "learning_rate": 0.00019388103748983496, | |
| "loss": 0.8504, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.1262699564586357, | |
| "grad_norm": 0.34104010462760925, | |
| "learning_rate": 0.00019372067050063438, | |
| "loss": 0.838, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.12772133526850507, | |
| "grad_norm": 0.29916197061538696, | |
| "learning_rate": 0.00019355829721634952, | |
| "loss": 0.5686, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.12917271407837447, | |
| "grad_norm": 0.38376718759536743, | |
| "learning_rate": 0.00019339392111293425, | |
| "loss": 0.7435, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.13062409288824384, | |
| "grad_norm": 0.2859137952327728, | |
| "learning_rate": 0.0001932275457092171, | |
| "loss": 0.5112, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1320754716981132, | |
| "grad_norm": 0.25812411308288574, | |
| "learning_rate": 0.00019305917456682593, | |
| "loss": 0.4435, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.13352685050798258, | |
| "grad_norm": 0.5547136664390564, | |
| "learning_rate": 0.00019288881129011175, | |
| "loss": 2.2399, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.13497822931785197, | |
| "grad_norm": 0.2667323052883148, | |
| "learning_rate": 0.00019271645952607163, | |
| "loss": 0.8544, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.13642960812772134, | |
| "grad_norm": 0.3629703223705292, | |
| "learning_rate": 0.00019254212296427044, | |
| "loss": 1.0825, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1378809869375907, | |
| "grad_norm": 0.31661346554756165, | |
| "learning_rate": 0.000192365805336762, | |
| "loss": 0.9337, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13933236574746008, | |
| "grad_norm": 0.30188268423080444, | |
| "learning_rate": 0.00019218751041800918, | |
| "loss": 0.9986, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.14078374455732948, | |
| "grad_norm": 0.4080539643764496, | |
| "learning_rate": 0.00019200724202480304, | |
| "loss": 0.9081, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.14223512336719885, | |
| "grad_norm": 0.29220181703567505, | |
| "learning_rate": 0.0001918250040161812, | |
| "loss": 0.6934, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.14368650217706821, | |
| "grad_norm": 0.3137444257736206, | |
| "learning_rate": 0.00019164080029334516, | |
| "loss": 0.5737, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.14513788098693758, | |
| "grad_norm": 0.3088196814060211, | |
| "learning_rate": 0.00019145463479957678, | |
| "loss": 0.7738, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14658925979680695, | |
| "grad_norm": 0.27212125062942505, | |
| "learning_rate": 0.00019126651152015403, | |
| "loss": 0.6838, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.14804063860667635, | |
| "grad_norm": 0.2625119686126709, | |
| "learning_rate": 0.00019107643448226536, | |
| "loss": 0.4542, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.14949201741654572, | |
| "grad_norm": 0.28616583347320557, | |
| "learning_rate": 0.00019088440775492384, | |
| "loss": 0.3243, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1509433962264151, | |
| "grad_norm": 0.3205554485321045, | |
| "learning_rate": 0.00019069043544887976, | |
| "loss": 0.7802, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.15239477503628446, | |
| "grad_norm": 0.4131568968296051, | |
| "learning_rate": 0.00019049452171653289, | |
| "loss": 1.5973, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.27477991580963135, | |
| "learning_rate": 0.00019029667075184332, | |
| "loss": 1.0704, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.15529753265602322, | |
| "grad_norm": 0.23675715923309326, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.6121, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1567489114658926, | |
| "grad_norm": 0.32074788212776184, | |
| "learning_rate": 0.00018989517410853955, | |
| "loss": 0.8862, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.15820029027576196, | |
| "grad_norm": 0.26189717650413513, | |
| "learning_rate": 0.00018969153702483546, | |
| "loss": 0.8521, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.15965166908563136, | |
| "grad_norm": 0.2720199525356293, | |
| "learning_rate": 0.00018948597989842502, | |
| "loss": 0.9151, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16110304789550073, | |
| "grad_norm": 0.31710368394851685, | |
| "learning_rate": 0.00018927850712970626, | |
| "loss": 0.9438, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1625544267053701, | |
| "grad_norm": 0.3253443241119385, | |
| "learning_rate": 0.0001890691231600856, | |
| "loss": 1.2068, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.16400580551523947, | |
| "grad_norm": 0.2926581799983978, | |
| "learning_rate": 0.00018885783247188305, | |
| "loss": 1.0507, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.16545718432510886, | |
| "grad_norm": 0.4618076682090759, | |
| "learning_rate": 0.00018864463958823596, | |
| "loss": 1.5177, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.16690856313497823, | |
| "grad_norm": 0.3991358280181885, | |
| "learning_rate": 0.00018842954907300236, | |
| "loss": 0.7893, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1683599419448476, | |
| "grad_norm": 0.3550536036491394, | |
| "learning_rate": 0.00018821256553066325, | |
| "loss": 0.6983, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.16981132075471697, | |
| "grad_norm": 0.3150579035282135, | |
| "learning_rate": 0.00018799369360622393, | |
| "loss": 1.1961, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.17126269956458637, | |
| "grad_norm": 0.3520543873310089, | |
| "learning_rate": 0.00018777293798511466, | |
| "loss": 2.1461, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.17271407837445574, | |
| "grad_norm": 0.3473776876926422, | |
| "learning_rate": 0.00018755030339309036, | |
| "loss": 1.3011, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1741654571843251, | |
| "grad_norm": 0.35327035188674927, | |
| "learning_rate": 0.00018732579459612936, | |
| "loss": 0.6785, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17561683599419448, | |
| "grad_norm": 0.29838210344314575, | |
| "learning_rate": 0.0001870994164003315, | |
| "loss": 0.7142, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.17706821480406387, | |
| "grad_norm": 0.3126171827316284, | |
| "learning_rate": 0.00018687117365181512, | |
| "loss": 1.5636, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.17851959361393324, | |
| "grad_norm": 0.33095410466194153, | |
| "learning_rate": 0.00018664107123661337, | |
| "loss": 1.6195, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1799709724238026, | |
| "grad_norm": 0.2580685615539551, | |
| "learning_rate": 0.00018640911408056963, | |
| "loss": 0.5031, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.18142235123367198, | |
| "grad_norm": 0.30072328448295593, | |
| "learning_rate": 0.00018617530714923209, | |
| "loss": 0.7414, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.18287373004354138, | |
| "grad_norm": 0.3045239746570587, | |
| "learning_rate": 0.0001859396554477473, | |
| "loss": 0.7002, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.18432510885341075, | |
| "grad_norm": 0.24665483832359314, | |
| "learning_rate": 0.00018570216402075324, | |
| "loss": 0.8552, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.18577648766328012, | |
| "grad_norm": 0.31482651829719543, | |
| "learning_rate": 0.0001854628379522712, | |
| "loss": 0.7006, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.18722786647314948, | |
| "grad_norm": 0.36719033122062683, | |
| "learning_rate": 0.00018522168236559695, | |
| "loss": 1.3433, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.18867924528301888, | |
| "grad_norm": 0.33420220017433167, | |
| "learning_rate": 0.000184978702423191, | |
| "loss": 1.4389, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.19013062409288825, | |
| "grad_norm": 0.2630361020565033, | |
| "learning_rate": 0.00018473390332656838, | |
| "loss": 0.6351, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.19158200290275762, | |
| "grad_norm": 0.3021498918533325, | |
| "learning_rate": 0.00018448729031618686, | |
| "loss": 1.2948, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.193033381712627, | |
| "grad_norm": 0.2972145080566406, | |
| "learning_rate": 0.0001842388686713351, | |
| "loss": 1.4549, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.19448476052249636, | |
| "grad_norm": 0.40088534355163574, | |
| "learning_rate": 0.00018398864371001945, | |
| "loss": 1.6829, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.19593613933236576, | |
| "grad_norm": 0.46335557103157043, | |
| "learning_rate": 0.00018373662078885027, | |
| "loss": 1.4324, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.19738751814223512, | |
| "grad_norm": 0.2726857662200928, | |
| "learning_rate": 0.00018348280530292713, | |
| "loss": 0.8288, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1988388969521045, | |
| "grad_norm": 0.30235981941223145, | |
| "learning_rate": 0.00018322720268572332, | |
| "loss": 1.0587, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.20029027576197386, | |
| "grad_norm": 0.28675076365470886, | |
| "learning_rate": 0.00018296981840896953, | |
| "loss": 0.3433, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.20174165457184326, | |
| "grad_norm": 0.23998884856700897, | |
| "learning_rate": 0.00018271065798253688, | |
| "loss": 0.662, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.20319303338171263, | |
| "grad_norm": 0.39619889855384827, | |
| "learning_rate": 0.00018244972695431874, | |
| "loss": 1.6662, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.204644412191582, | |
| "grad_norm": 0.338534414768219, | |
| "learning_rate": 0.00018218703091011214, | |
| "loss": 0.6726, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.20609579100145137, | |
| "grad_norm": 0.31382206082344055, | |
| "learning_rate": 0.00018192257547349803, | |
| "loss": 0.9496, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.20754716981132076, | |
| "grad_norm": 0.3816578984260559, | |
| "learning_rate": 0.0001816563663057211, | |
| "loss": 1.8337, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.20899854862119013, | |
| "grad_norm": 0.2983052432537079, | |
| "learning_rate": 0.00018138840910556837, | |
| "loss": 0.874, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2104499274310595, | |
| "grad_norm": 0.47916677594184875, | |
| "learning_rate": 0.00018111870960924734, | |
| "loss": 1.2913, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.21190130624092887, | |
| "grad_norm": 0.28957951068878174, | |
| "learning_rate": 0.00018084727359026317, | |
| "loss": 0.836, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.21335268505079827, | |
| "grad_norm": 0.3818242847919464, | |
| "learning_rate": 0.00018057410685929506, | |
| "loss": 1.5504, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.21480406386066764, | |
| "grad_norm": 0.39021217823028564, | |
| "learning_rate": 0.00018029921526407184, | |
| "loss": 0.66, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.216255442670537, | |
| "grad_norm": 0.41055023670196533, | |
| "learning_rate": 0.00018002260468924695, | |
| "loss": 1.027, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.21770682148040638, | |
| "grad_norm": 0.3222734034061432, | |
| "learning_rate": 0.00017974428105627208, | |
| "loss": 1.223, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.21915820029027577, | |
| "grad_norm": 0.4165111184120178, | |
| "learning_rate": 0.0001794642503232709, | |
| "loss": 0.799, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.22060957910014514, | |
| "grad_norm": 0.2545410692691803, | |
| "learning_rate": 0.00017918251848491118, | |
| "loss": 0.8621, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2220609579100145, | |
| "grad_norm": 0.4528481066226959, | |
| "learning_rate": 0.00017889909157227653, | |
| "loss": 1.1478, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.22351233671988388, | |
| "grad_norm": 0.24836307764053345, | |
| "learning_rate": 0.00017861397565273733, | |
| "loss": 0.389, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.22496371552975328, | |
| "grad_norm": 0.42859530448913574, | |
| "learning_rate": 0.00017832717682982093, | |
| "loss": 1.2932, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.22641509433962265, | |
| "grad_norm": 0.3459640443325043, | |
| "learning_rate": 0.00017803870124308068, | |
| "loss": 0.7394, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.22786647314949202, | |
| "grad_norm": 0.28327664732933044, | |
| "learning_rate": 0.00017774855506796496, | |
| "loss": 0.7749, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.22931785195936139, | |
| "grad_norm": 0.2421475350856781, | |
| "learning_rate": 0.00017745674451568453, | |
| "loss": 0.4116, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.42741137742996216, | |
| "learning_rate": 0.00017716327583307989, | |
| "loss": 1.0368, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.23222060957910015, | |
| "grad_norm": 0.3373871147632599, | |
| "learning_rate": 0.0001768681553024874, | |
| "loss": 0.822, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.23367198838896952, | |
| "grad_norm": 0.4648170471191406, | |
| "learning_rate": 0.00017657138924160476, | |
| "loss": 1.9527, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2351233671988389, | |
| "grad_norm": 0.46374577283859253, | |
| "learning_rate": 0.00017627298400335598, | |
| "loss": 2.0193, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2365747460087083, | |
| "grad_norm": 0.3726603388786316, | |
| "learning_rate": 0.00017597294597575507, | |
| "loss": 1.0391, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.23802612481857766, | |
| "grad_norm": 0.35028398036956787, | |
| "learning_rate": 0.00017567128158176953, | |
| "loss": 0.5079, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.23947750362844702, | |
| "grad_norm": 0.3101751506328583, | |
| "learning_rate": 0.0001753679972791828, | |
| "loss": 0.4194, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2409288824383164, | |
| "grad_norm": 0.3468049466609955, | |
| "learning_rate": 0.0001750630995604559, | |
| "loss": 0.61, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.24238026124818576, | |
| "grad_norm": 0.38198643922805786, | |
| "learning_rate": 0.00017475659495258864, | |
| "loss": 1.2123, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.24383164005805516, | |
| "grad_norm": 0.3985172510147095, | |
| "learning_rate": 0.00017444849001697975, | |
| "loss": 1.2891, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.24528301886792453, | |
| "grad_norm": 0.5197709202766418, | |
| "learning_rate": 0.00017413879134928643, | |
| "loss": 0.9845, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.2467343976777939, | |
| "grad_norm": 0.3085668087005615, | |
| "learning_rate": 0.00017382750557928323, | |
| "loss": 1.077, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.24818577648766327, | |
| "grad_norm": 0.526063084602356, | |
| "learning_rate": 0.00017351463937072004, | |
| "loss": 1.5359, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.24963715529753266, | |
| "grad_norm": 0.3238290846347809, | |
| "learning_rate": 0.00017320019942117956, | |
| "loss": 0.8538, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.251088534107402, | |
| "grad_norm": 0.4351443946361542, | |
| "learning_rate": 0.00017288419246193373, | |
| "loss": 2.2273, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.251088534107402, | |
| "eval_loss": 0.25142431259155273, | |
| "eval_runtime": 32.4109, | |
| "eval_samples_per_second": 8.978, | |
| "eval_steps_per_second": 4.505, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2525399129172714, | |
| "grad_norm": 0.24334195256233215, | |
| "learning_rate": 0.00017256662525779983, | |
| "loss": 0.5441, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2539912917271408, | |
| "grad_norm": 0.3937956988811493, | |
| "learning_rate": 0.00017224750460699556, | |
| "loss": 1.3447, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.25544267053701014, | |
| "grad_norm": 0.32860031723976135, | |
| "learning_rate": 0.00017192683734099353, | |
| "loss": 1.1908, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.25689404934687954, | |
| "grad_norm": 0.40750887989997864, | |
| "learning_rate": 0.000171604630324375, | |
| "loss": 0.4587, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.25834542815674894, | |
| "grad_norm": 0.24471426010131836, | |
| "learning_rate": 0.00017128089045468294, | |
| "loss": 0.4731, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.2597968069666183, | |
| "grad_norm": 0.3871125876903534, | |
| "learning_rate": 0.0001709556246622744, | |
| "loss": 1.5461, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2612481857764877, | |
| "grad_norm": 0.32278481125831604, | |
| "learning_rate": 0.00017062883991017218, | |
| "loss": 1.0052, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.262699564586357, | |
| "grad_norm": 0.2906551957130432, | |
| "learning_rate": 0.00017030054319391562, | |
| "loss": 1.1581, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.2641509433962264, | |
| "grad_norm": 0.2662639319896698, | |
| "learning_rate": 0.00016997074154141096, | |
| "loss": 0.4337, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2656023222060958, | |
| "grad_norm": 0.30343329906463623, | |
| "learning_rate": 0.00016963944201278097, | |
| "loss": 0.9429, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.26705370101596515, | |
| "grad_norm": 0.48314595222473145, | |
| "learning_rate": 0.00016930665170021363, | |
| "loss": 1.146, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.26850507982583455, | |
| "grad_norm": 0.22615785896778107, | |
| "learning_rate": 0.00016897237772781044, | |
| "loss": 0.288, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.26995645863570394, | |
| "grad_norm": 0.3443828225135803, | |
| "learning_rate": 0.00016863662725143382, | |
| "loss": 1.0044, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.2714078374455733, | |
| "grad_norm": 0.3521503508090973, | |
| "learning_rate": 0.00016829940745855408, | |
| "loss": 0.917, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2728592162554427, | |
| "grad_norm": 0.3083937168121338, | |
| "learning_rate": 0.00016796072556809534, | |
| "loss": 1.1641, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.274310595065312, | |
| "grad_norm": 0.1671004444360733, | |
| "learning_rate": 0.0001676205888302812, | |
| "loss": 0.2087, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2757619738751814, | |
| "grad_norm": 0.3343052268028259, | |
| "learning_rate": 0.00016727900452647926, | |
| "loss": 0.8929, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2772133526850508, | |
| "grad_norm": 0.2790353000164032, | |
| "learning_rate": 0.00016693597996904564, | |
| "loss": 0.8818, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.27866473149492016, | |
| "grad_norm": 0.3753064274787903, | |
| "learning_rate": 0.00016659152250116812, | |
| "loss": 0.7531, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.28011611030478956, | |
| "grad_norm": 0.3089276850223541, | |
| "learning_rate": 0.00016624563949670904, | |
| "loss": 1.2895, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.28156748911465895, | |
| "grad_norm": 0.24410511553287506, | |
| "learning_rate": 0.00016589833836004752, | |
| "loss": 0.444, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2830188679245283, | |
| "grad_norm": 0.4272465705871582, | |
| "learning_rate": 0.00016554962652592077, | |
| "loss": 0.9711, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2844702467343977, | |
| "grad_norm": 0.29573267698287964, | |
| "learning_rate": 0.00016519951145926515, | |
| "loss": 0.3583, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.28592162554426703, | |
| "grad_norm": 0.42986953258514404, | |
| "learning_rate": 0.00016484800065505627, | |
| "loss": 1.6862, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.28737300435413643, | |
| "grad_norm": 0.31086039543151855, | |
| "learning_rate": 0.00016449510163814854, | |
| "loss": 0.5808, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2888243831640058, | |
| "grad_norm": 0.4590548574924469, | |
| "learning_rate": 0.000164140821963114, | |
| "loss": 1.837, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.29027576197387517, | |
| "grad_norm": 0.4661678969860077, | |
| "learning_rate": 0.00016378516921408077, | |
| "loss": 0.8179, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29172714078374457, | |
| "grad_norm": 0.5419331192970276, | |
| "learning_rate": 0.00016342815100457063, | |
| "loss": 0.8166, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2931785195936139, | |
| "grad_norm": 0.34207096695899963, | |
| "learning_rate": 0.00016306977497733592, | |
| "loss": 0.8683, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2946298984034833, | |
| "grad_norm": 0.4035762548446655, | |
| "learning_rate": 0.00016271004880419608, | |
| "loss": 0.7494, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.2960812772133527, | |
| "grad_norm": 0.40903428196907043, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.3219, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.29753265602322204, | |
| "grad_norm": 0.3800163269042969, | |
| "learning_rate": 0.000161986576851828, | |
| "loss": 1.0624, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.29898403483309144, | |
| "grad_norm": 0.20540378987789154, | |
| "learning_rate": 0.00016162284656009274, | |
| "loss": 0.2874, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.30043541364296084, | |
| "grad_norm": 0.3501240313053131, | |
| "learning_rate": 0.00016125779709710665, | |
| "loss": 0.9726, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3018867924528302, | |
| "grad_norm": 0.46438068151474, | |
| "learning_rate": 0.00016089143627754862, | |
| "loss": 0.9383, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3033381712626996, | |
| "grad_norm": 0.37778106331825256, | |
| "learning_rate": 0.00016052377194417, | |
| "loss": 1.9048, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3047895500725689, | |
| "grad_norm": 0.4382157325744629, | |
| "learning_rate": 0.0001601548119676266, | |
| "loss": 1.2582, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3062409288824383, | |
| "grad_norm": 0.442871630191803, | |
| "learning_rate": 0.00015978456424631032, | |
| "loss": 1.086, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.5801809430122375, | |
| "learning_rate": 0.00015941303670618018, | |
| "loss": 1.234, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.30914368650217705, | |
| "grad_norm": 0.2682696282863617, | |
| "learning_rate": 0.00015904023730059228, | |
| "loss": 0.6148, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.31059506531204645, | |
| "grad_norm": 0.25142902135849, | |
| "learning_rate": 0.00015866617401012994, | |
| "loss": 0.6304, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.31204644412191584, | |
| "grad_norm": 0.36662545800209045, | |
| "learning_rate": 0.00015829085484243266, | |
| "loss": 0.8005, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3134978229317852, | |
| "grad_norm": 0.2825576961040497, | |
| "learning_rate": 0.00015791428783202465, | |
| "loss": 0.7511, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3149492017416546, | |
| "grad_norm": 0.33648526668548584, | |
| "learning_rate": 0.00015753648104014297, | |
| "loss": 0.9242, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3164005805515239, | |
| "grad_norm": 0.5301381945610046, | |
| "learning_rate": 0.00015715744255456496, | |
| "loss": 0.9624, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3178519593613933, | |
| "grad_norm": 0.3330354690551758, | |
| "learning_rate": 0.00015677718048943496, | |
| "loss": 1.6149, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3193033381712627, | |
| "grad_norm": 0.2863297164440155, | |
| "learning_rate": 0.00015639570298509064, | |
| "loss": 0.5041, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.32075471698113206, | |
| "grad_norm": 0.30266642570495605, | |
| "learning_rate": 0.00015601301820788898, | |
| "loss": 0.5615, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.32220609579100146, | |
| "grad_norm": 0.3737308084964752, | |
| "learning_rate": 0.00015562913435003114, | |
| "loss": 0.5868, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.32365747460087085, | |
| "grad_norm": 0.41030675172805786, | |
| "learning_rate": 0.00015524405962938714, | |
| "loss": 0.933, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.3251088534107402, | |
| "grad_norm": 0.22345803678035736, | |
| "learning_rate": 0.0001548578022893202, | |
| "loss": 0.4072, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3265602322206096, | |
| "grad_norm": 0.4588911235332489, | |
| "learning_rate": 0.00015447037059851, | |
| "loss": 1.3702, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.32801161103047893, | |
| "grad_norm": 0.40378692746162415, | |
| "learning_rate": 0.00015408177285077565, | |
| "loss": 0.7234, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.32946298984034833, | |
| "grad_norm": 0.36572182178497314, | |
| "learning_rate": 0.0001536920173648984, | |
| "loss": 0.6561, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.3309143686502177, | |
| "grad_norm": 0.24225015938282013, | |
| "learning_rate": 0.0001533011124844433, | |
| "loss": 0.5723, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.33236574746008707, | |
| "grad_norm": 0.4880613386631012, | |
| "learning_rate": 0.00015290906657758083, | |
| "loss": 0.8659, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.33381712626995647, | |
| "grad_norm": 0.4412708878517151, | |
| "learning_rate": 0.00015251588803690753, | |
| "loss": 0.8392, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3352685050798258, | |
| "grad_norm": 0.2982364594936371, | |
| "learning_rate": 0.00015212158527926634, | |
| "loss": 0.8567, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.3367198838896952, | |
| "grad_norm": 0.3174440860748291, | |
| "learning_rate": 0.0001517261667455667, | |
| "loss": 0.6775, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3381712626995646, | |
| "grad_norm": 0.23971304297447205, | |
| "learning_rate": 0.00015132964090060357, | |
| "loss": 0.2637, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.33962264150943394, | |
| "grad_norm": 0.3869999647140503, | |
| "learning_rate": 0.00015093201623287631, | |
| "loss": 0.7993, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.34107402031930334, | |
| "grad_norm": 0.4646807014942169, | |
| "learning_rate": 0.000150533301254407, | |
| "loss": 1.0936, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.34252539912917274, | |
| "grad_norm": 0.39114901423454285, | |
| "learning_rate": 0.0001501335045005582, | |
| "loss": 1.0431, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3439767779390421, | |
| "grad_norm": 0.32341787219047546, | |
| "learning_rate": 0.00014973263452985024, | |
| "loss": 0.5171, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3454281567489115, | |
| "grad_norm": 0.3025743365287781, | |
| "learning_rate": 0.00014933069992377793, | |
| "loss": 0.59, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.3468795355587808, | |
| "grad_norm": 0.49004557728767395, | |
| "learning_rate": 0.00014892770928662697, | |
| "loss": 1.6646, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3483309143686502, | |
| "grad_norm": 0.5597532391548157, | |
| "learning_rate": 0.00014852367124528975, | |
| "loss": 1.3429, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3497822931785196, | |
| "grad_norm": 0.4048459827899933, | |
| "learning_rate": 0.00014811859444908052, | |
| "loss": 1.4101, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.35123367198838895, | |
| "grad_norm": 0.5641282796859741, | |
| "learning_rate": 0.00014771248756955042, | |
| "loss": 1.6691, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.35268505079825835, | |
| "grad_norm": 0.4494577646255493, | |
| "learning_rate": 0.00014730535930030172, | |
| "loss": 1.211, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.35413642960812775, | |
| "grad_norm": 0.3181702792644501, | |
| "learning_rate": 0.00014689721835680182, | |
| "loss": 0.4831, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3555878084179971, | |
| "grad_norm": 0.3166160583496094, | |
| "learning_rate": 0.00014648807347619663, | |
| "loss": 1.59, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3570391872278665, | |
| "grad_norm": 0.3532291054725647, | |
| "learning_rate": 0.00014607793341712334, | |
| "loss": 0.3531, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3584905660377358, | |
| "grad_norm": 0.3703453540802002, | |
| "learning_rate": 0.00014566680695952332, | |
| "loss": 0.7499, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3599419448476052, | |
| "grad_norm": 0.29574841260910034, | |
| "learning_rate": 0.00014525470290445392, | |
| "loss": 0.7956, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3613933236574746, | |
| "grad_norm": 0.41456523537635803, | |
| "learning_rate": 0.0001448416300738999, | |
| "loss": 1.2597, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.36284470246734396, | |
| "grad_norm": 0.3268081545829773, | |
| "learning_rate": 0.000144427597310585, | |
| "loss": 0.7744, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.36429608127721336, | |
| "grad_norm": 0.4419240951538086, | |
| "learning_rate": 0.00014401261347778233, | |
| "loss": 1.2126, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.36574746008708275, | |
| "grad_norm": 0.4968956410884857, | |
| "learning_rate": 0.0001435966874591247, | |
| "loss": 0.8786, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3671988388969521, | |
| "grad_norm": 0.4824519455432892, | |
| "learning_rate": 0.0001431798281584144, | |
| "loss": 3.0825, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3686502177068215, | |
| "grad_norm": 0.30820131301879883, | |
| "learning_rate": 0.0001427620444994328, | |
| "loss": 0.6051, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.37010159651669083, | |
| "grad_norm": 0.42164450883865356, | |
| "learning_rate": 0.00014234334542574906, | |
| "loss": 0.5161, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.37155297532656023, | |
| "grad_norm": 0.32636889815330505, | |
| "learning_rate": 0.00014192373990052877, | |
| "loss": 0.855, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.37300435413642963, | |
| "grad_norm": 0.332589715719223, | |
| "learning_rate": 0.0001415032369063422, | |
| "loss": 0.402, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.37445573294629897, | |
| "grad_norm": 0.37509143352508545, | |
| "learning_rate": 0.00014108184544497178, | |
| "loss": 1.1532, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.37590711175616837, | |
| "grad_norm": 0.2669030725955963, | |
| "learning_rate": 0.00014065957453721962, | |
| "loss": 0.5493, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 0.387205570936203, | |
| "learning_rate": 0.00014023643322271426, | |
| "loss": 1.3697, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3788098693759071, | |
| "grad_norm": 0.3195268213748932, | |
| "learning_rate": 0.00013981243055971712, | |
| "loss": 0.731, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.3802612481857765, | |
| "grad_norm": 0.3117748498916626, | |
| "learning_rate": 0.00013938757562492873, | |
| "loss": 0.4412, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.38171262699564584, | |
| "grad_norm": 0.43555042147636414, | |
| "learning_rate": 0.00013896187751329437, | |
| "loss": 0.9857, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.38316400580551524, | |
| "grad_norm": 0.3363146483898163, | |
| "learning_rate": 0.0001385353453378093, | |
| "loss": 0.8749, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.3038422465324402, | |
| "learning_rate": 0.00013810798822932378, | |
| "loss": 0.5955, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.386066763425254, | |
| "grad_norm": 0.47416090965270996, | |
| "learning_rate": 0.00013767981533634754, | |
| "loss": 3.7495, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.3875181422351234, | |
| "grad_norm": 0.237370103597641, | |
| "learning_rate": 0.00013725083582485397, | |
| "loss": 0.3979, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3889695210449927, | |
| "grad_norm": 0.1970636397600174, | |
| "learning_rate": 0.0001368210588780838, | |
| "loss": 0.297, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.3904208998548621, | |
| "grad_norm": 0.5174916982650757, | |
| "learning_rate": 0.00013639049369634876, | |
| "loss": 1.4907, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.3918722786647315, | |
| "grad_norm": 0.2811208665370941, | |
| "learning_rate": 0.00013595914949683432, | |
| "loss": 0.5772, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.39332365747460085, | |
| "grad_norm": 0.30495163798332214, | |
| "learning_rate": 0.00013552703551340258, | |
| "loss": 0.3624, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.39477503628447025, | |
| "grad_norm": 0.4540651738643646, | |
| "learning_rate": 0.00013509416099639457, | |
| "loss": 1.6049, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.39622641509433965, | |
| "grad_norm": 0.4534936547279358, | |
| "learning_rate": 0.00013466053521243214, | |
| "loss": 1.3386, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.397677793904209, | |
| "grad_norm": 0.23787882924079895, | |
| "learning_rate": 0.00013422616744421966, | |
| "loss": 0.5185, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3991291727140784, | |
| "grad_norm": 0.35515129566192627, | |
| "learning_rate": 0.00013379106699034537, | |
| "loss": 1.4753, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4005805515239477, | |
| "grad_norm": 0.38288307189941406, | |
| "learning_rate": 0.00013335524316508208, | |
| "loss": 1.4326, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.4020319303338171, | |
| "grad_norm": 0.27358075976371765, | |
| "learning_rate": 0.0001329187052981881, | |
| "loss": 0.683, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.4034833091436865, | |
| "grad_norm": 0.5965114831924438, | |
| "learning_rate": 0.00013248146273470726, | |
| "loss": 2.2199, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.40493468795355586, | |
| "grad_norm": 0.27985599637031555, | |
| "learning_rate": 0.00013204352483476897, | |
| "loss": 0.6224, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.40638606676342526, | |
| "grad_norm": 0.31146517395973206, | |
| "learning_rate": 0.0001316049009733879, | |
| "loss": 0.9722, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.40783744557329465, | |
| "grad_norm": 0.2670224905014038, | |
| "learning_rate": 0.0001311656005402631, | |
| "loss": 0.4706, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.409288824383164, | |
| "grad_norm": 0.30668875575065613, | |
| "learning_rate": 0.00013072563293957723, | |
| "loss": 0.8013, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.4107402031930334, | |
| "grad_norm": 1.6362825632095337, | |
| "learning_rate": 0.00013028500758979506, | |
| "loss": 1.7021, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.41219158200290273, | |
| "grad_norm": 0.3747856914997101, | |
| "learning_rate": 0.00012984373392346194, | |
| "loss": 0.7395, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.41364296081277213, | |
| "grad_norm": 0.29460155963897705, | |
| "learning_rate": 0.0001294018213870018, | |
| "loss": 0.4852, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.41509433962264153, | |
| "grad_norm": 0.3442453444004059, | |
| "learning_rate": 0.00012895927944051502, | |
| "loss": 1.0444, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.41654571843251087, | |
| "grad_norm": 0.4372914731502533, | |
| "learning_rate": 0.00012851611755757586, | |
| "loss": 0.9556, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.41799709724238027, | |
| "grad_norm": 0.36953088641166687, | |
| "learning_rate": 0.00012807234522502968, | |
| "loss": 0.6292, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.41944847605224966, | |
| "grad_norm": 0.2996106445789337, | |
| "learning_rate": 0.00012762797194278984, | |
| "loss": 0.4637, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.420899854862119, | |
| "grad_norm": 0.46969079971313477, | |
| "learning_rate": 0.0001271830072236343, | |
| "loss": 2.3145, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4223512336719884, | |
| "grad_norm": 0.377268522977829, | |
| "learning_rate": 0.00012673746059300208, | |
| "loss": 1.1514, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.42380261248185774, | |
| "grad_norm": 0.37440070509910583, | |
| "learning_rate": 0.00012629134158878918, | |
| "loss": 1.2862, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.42525399129172714, | |
| "grad_norm": 0.47031837701797485, | |
| "learning_rate": 0.0001258446597611447, | |
| "loss": 1.3154, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.42670537010159654, | |
| "grad_norm": 0.28943926095962524, | |
| "learning_rate": 0.00012539742467226601, | |
| "loss": 0.7372, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4281567489114659, | |
| "grad_norm": 0.3967966139316559, | |
| "learning_rate": 0.00012494964589619423, | |
| "loss": 1.7696, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4296081277213353, | |
| "grad_norm": 0.3762686252593994, | |
| "learning_rate": 0.00012450133301860952, | |
| "loss": 1.2166, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4310595065312046, | |
| "grad_norm": 0.3782839775085449, | |
| "learning_rate": 0.00012405249563662537, | |
| "loss": 0.9249, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.432510885341074, | |
| "grad_norm": 0.551977276802063, | |
| "learning_rate": 0.0001236031433585836, | |
| "loss": 0.7932, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4339622641509434, | |
| "grad_norm": 0.4811376631259918, | |
| "learning_rate": 0.00012315328580384842, | |
| "loss": 1.4454, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.43541364296081275, | |
| "grad_norm": 0.6780052185058594, | |
| "learning_rate": 0.00012270293260260067, | |
| "loss": 1.8719, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.43686502177068215, | |
| "grad_norm": 0.5372028350830078, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 1.13, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.43831640058055155, | |
| "grad_norm": 0.6659995317459106, | |
| "learning_rate": 0.00012180077783413601, | |
| "loss": 1.2524, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4397677793904209, | |
| "grad_norm": 0.30327922105789185, | |
| "learning_rate": 0.00012134899557950698, | |
| "loss": 0.4541, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.4412191582002903, | |
| "grad_norm": 0.28171366453170776, | |
| "learning_rate": 0.00012089675630312754, | |
| "loss": 0.5961, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4426705370101596, | |
| "grad_norm": 0.49443453550338745, | |
| "learning_rate": 0.00012044406968616432, | |
| "loss": 1.3738, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.444121915820029, | |
| "grad_norm": 0.27701106667518616, | |
| "learning_rate": 0.00011999094541936047, | |
| "loss": 0.8035, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.4455732946298984, | |
| "grad_norm": 0.44474315643310547, | |
| "learning_rate": 0.00011953739320282778, | |
| "loss": 1.9202, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.44702467343976776, | |
| "grad_norm": 0.36672115325927734, | |
| "learning_rate": 0.00011908342274583936, | |
| "loss": 1.3322, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.44847605224963716, | |
| "grad_norm": 0.47056931257247925, | |
| "learning_rate": 0.00011862904376662167, | |
| "loss": 0.6009, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.44992743105950656, | |
| "grad_norm": 0.33981776237487793, | |
| "learning_rate": 0.00011817426599214636, | |
| "loss": 0.8832, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4513788098693759, | |
| "grad_norm": 0.5165224075317383, | |
| "learning_rate": 0.0001177190991579223, | |
| "loss": 1.2614, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.4528301886792453, | |
| "grad_norm": 0.36011219024658203, | |
| "learning_rate": 0.00011726355300778693, | |
| "loss": 0.5128, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.45428156748911463, | |
| "grad_norm": 0.4018263518810272, | |
| "learning_rate": 0.00011680763729369783, | |
| "loss": 0.876, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.45573294629898403, | |
| "grad_norm": 0.36778366565704346, | |
| "learning_rate": 0.00011635136177552391, | |
| "loss": 0.6217, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.45718432510885343, | |
| "grad_norm": 0.3395918905735016, | |
| "learning_rate": 0.00011589473622083642, | |
| "loss": 0.7758, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.45863570391872277, | |
| "grad_norm": 0.30903902649879456, | |
| "learning_rate": 0.00011543777040469994, | |
| "loss": 1.1969, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.46008708272859217, | |
| "grad_norm": 0.36154720187187195, | |
| "learning_rate": 0.00011498047410946306, | |
| "loss": 0.7883, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.3478432595729828, | |
| "learning_rate": 0.00011452285712454904, | |
| "loss": 1.1028, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4629898403483309, | |
| "grad_norm": 0.8324244618415833, | |
| "learning_rate": 0.00011406492924624614, | |
| "loss": 2.98, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.4644412191582003, | |
| "grad_norm": 0.36261022090911865, | |
| "learning_rate": 0.00011360670027749807, | |
| "loss": 0.8629, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.46589259796806964, | |
| "grad_norm": 0.42329132556915283, | |
| "learning_rate": 0.00011314818002769389, | |
| "loss": 1.8163, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.46734397677793904, | |
| "grad_norm": 0.5022541284561157, | |
| "learning_rate": 0.0001126893783124583, | |
| "loss": 1.5672, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.46879535558780844, | |
| "grad_norm": 0.3574770987033844, | |
| "learning_rate": 0.00011223030495344127, | |
| "loss": 0.3025, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.4702467343976778, | |
| "grad_norm": 0.4947843551635742, | |
| "learning_rate": 0.00011177096977810803, | |
| "loss": 0.8242, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4716981132075472, | |
| "grad_norm": 0.4501658082008362, | |
| "learning_rate": 0.00011131138261952845, | |
| "loss": 0.8425, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4731494920174166, | |
| "grad_norm": 0.21305899322032928, | |
| "learning_rate": 0.00011085155331616663, | |
| "loss": 0.245, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.4746008708272859, | |
| "grad_norm": 0.38023510575294495, | |
| "learning_rate": 0.00011039149171167045, | |
| "loss": 1.0404, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.4760522496371553, | |
| "grad_norm": 0.4849722385406494, | |
| "learning_rate": 0.00010993120765466056, | |
| "loss": 1.0257, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.47750362844702465, | |
| "grad_norm": 0.4337415099143982, | |
| "learning_rate": 0.00010947071099851971, | |
| "loss": 0.787, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.47895500725689405, | |
| "grad_norm": 0.3395763635635376, | |
| "learning_rate": 0.00010901001160118189, | |
| "loss": 0.9218, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.48040638606676345, | |
| "grad_norm": 0.5438852906227112, | |
| "learning_rate": 0.00010854911932492114, | |
| "loss": 0.7867, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.4818577648766328, | |
| "grad_norm": 0.299010306596756, | |
| "learning_rate": 0.00010808804403614043, | |
| "loss": 0.7583, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.4833091436865022, | |
| "grad_norm": 0.44489172101020813, | |
| "learning_rate": 0.00010762679560516067, | |
| "loss": 1.9207, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.4847605224963715, | |
| "grad_norm": 0.3161913752555847, | |
| "learning_rate": 0.00010716538390600908, | |
| "loss": 0.7175, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.4862119013062409, | |
| "grad_norm": 0.38730329275131226, | |
| "learning_rate": 0.00010670381881620814, | |
| "loss": 1.2888, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4876632801161103, | |
| "grad_norm": 0.3712642788887024, | |
| "learning_rate": 0.00010624211021656392, | |
| "loss": 0.9692, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.48911465892597966, | |
| "grad_norm": 0.3168681859970093, | |
| "learning_rate": 0.00010578026799095464, | |
| "loss": 1.1365, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.49056603773584906, | |
| "grad_norm": 0.36398571729660034, | |
| "learning_rate": 0.00010531830202611904, | |
| "loss": 0.9235, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.49201741654571846, | |
| "grad_norm": 0.30887413024902344, | |
| "learning_rate": 0.00010485622221144484, | |
| "loss": 0.4427, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.4934687953555878, | |
| "grad_norm": 0.26171329617500305, | |
| "learning_rate": 0.0001043940384387569, | |
| "loss": 0.6965, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4949201741654572, | |
| "grad_norm": 0.44587722420692444, | |
| "learning_rate": 0.00010393176060210557, | |
| "loss": 1.8459, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.49637155297532654, | |
| "grad_norm": 0.341634064912796, | |
| "learning_rate": 0.0001034693985975548, | |
| "loss": 0.7757, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.49782293178519593, | |
| "grad_norm": 0.378426194190979, | |
| "learning_rate": 0.0001030069623229704, | |
| "loss": 1.1976, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.49927431059506533, | |
| "grad_norm": 0.4587966799736023, | |
| "learning_rate": 0.00010254446167780803, | |
| "loss": 0.7019, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5007256894049347, | |
| "grad_norm": 0.2761523127555847, | |
| "learning_rate": 0.00010208190656290137, | |
| "loss": 0.3916, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.502177068214804, | |
| "grad_norm": 0.42763715982437134, | |
| "learning_rate": 0.00010161930688025017, | |
| "loss": 0.8072, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.502177068214804, | |
| "eval_loss": 0.23982280492782593, | |
| "eval_runtime": 32.4007, | |
| "eval_samples_per_second": 8.981, | |
| "eval_steps_per_second": 4.506, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5036284470246735, | |
| "grad_norm": 0.39921316504478455, | |
| "learning_rate": 0.00010115667253280816, | |
| "loss": 0.9669, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5050798258345428, | |
| "grad_norm": 0.24880976974964142, | |
| "learning_rate": 0.00010069401342427129, | |
| "loss": 0.4452, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5065312046444121, | |
| "grad_norm": 0.27354684472084045, | |
| "learning_rate": 0.00010023133945886545, | |
| "loss": 0.2555, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5079825834542816, | |
| "grad_norm": 0.4844830334186554, | |
| "learning_rate": 9.976866054113458e-05, | |
| "loss": 0.9596, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5094339622641509, | |
| "grad_norm": 0.3840599060058594, | |
| "learning_rate": 9.930598657572873e-05, | |
| "loss": 0.7379, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5108853410740203, | |
| "grad_norm": 0.2949220836162567, | |
| "learning_rate": 9.884332746719186e-05, | |
| "loss": 0.5608, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5123367198838897, | |
| "grad_norm": 0.5040110945701599, | |
| "learning_rate": 9.838069311974986e-05, | |
| "loss": 1.3997, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5137880986937591, | |
| "grad_norm": 0.4586881697177887, | |
| "learning_rate": 9.791809343709865e-05, | |
| "loss": 0.9486, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5152394775036284, | |
| "grad_norm": 0.256457656621933, | |
| "learning_rate": 9.745553832219202e-05, | |
| "loss": 0.5397, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5166908563134979, | |
| "grad_norm": 0.4685473144054413, | |
| "learning_rate": 9.699303767702963e-05, | |
| "loss": 3.0154, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5181422351233672, | |
| "grad_norm": 0.3497532606124878, | |
| "learning_rate": 9.653060140244524e-05, | |
| "loss": 0.7937, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5195936139332366, | |
| "grad_norm": 0.31562376022338867, | |
| "learning_rate": 9.606823939789448e-05, | |
| "loss": 0.7917, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.521044992743106, | |
| "grad_norm": 0.30568790435791016, | |
| "learning_rate": 9.560596156124314e-05, | |
| "loss": 0.7585, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5224963715529753, | |
| "grad_norm": 0.3299367129802704, | |
| "learning_rate": 9.514377778855521e-05, | |
| "loss": 0.7533, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5239477503628447, | |
| "grad_norm": 0.4130513668060303, | |
| "learning_rate": 9.468169797388097e-05, | |
| "loss": 1.5789, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.525399129172714, | |
| "grad_norm": 0.21178820729255676, | |
| "learning_rate": 9.421973200904538e-05, | |
| "loss": 0.5157, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5268505079825835, | |
| "grad_norm": 0.341922789812088, | |
| "learning_rate": 9.375788978343608e-05, | |
| "loss": 0.8536, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5283018867924528, | |
| "grad_norm": 0.3225785493850708, | |
| "learning_rate": 9.329618118379187e-05, | |
| "loss": 0.8953, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5297532656023222, | |
| "grad_norm": 0.5227508544921875, | |
| "learning_rate": 9.283461609399094e-05, | |
| "loss": 1.5774, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5312046444121916, | |
| "grad_norm": 0.45989224314689636, | |
| "learning_rate": 9.237320439483937e-05, | |
| "loss": 0.6673, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.532656023222061, | |
| "grad_norm": 0.5069538950920105, | |
| "learning_rate": 9.19119559638596e-05, | |
| "loss": 3.0604, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5341074020319303, | |
| "grad_norm": 0.3407710790634155, | |
| "learning_rate": 9.14508806750789e-05, | |
| "loss": 1.0451, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5355587808417998, | |
| "grad_norm": 0.3569037616252899, | |
| "learning_rate": 9.098998839881813e-05, | |
| "loss": 0.7066, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5370101596516691, | |
| "grad_norm": 0.28697144985198975, | |
| "learning_rate": 9.052928900148031e-05, | |
| "loss": 0.9004, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.41111066937446594, | |
| "learning_rate": 9.006879234533946e-05, | |
| "loss": 1.4209, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5399129172714079, | |
| "grad_norm": 0.4514535367488861, | |
| "learning_rate": 8.960850828832957e-05, | |
| "loss": 1.0341, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5413642960812772, | |
| "grad_norm": 0.5463643670082092, | |
| "learning_rate": 8.914844668383339e-05, | |
| "loss": 1.695, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5428156748911466, | |
| "grad_norm": 0.36876559257507324, | |
| "learning_rate": 8.868861738047158e-05, | |
| "loss": 0.6762, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5442670537010159, | |
| "grad_norm": 0.3862442374229431, | |
| "learning_rate": 8.822903022189199e-05, | |
| "loss": 0.7505, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5457184325108854, | |
| "grad_norm": 0.4112395644187927, | |
| "learning_rate": 8.776969504655875e-05, | |
| "loss": 0.7731, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5471698113207547, | |
| "grad_norm": 0.3331882059574127, | |
| "learning_rate": 8.731062168754174e-05, | |
| "loss": 0.9437, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.548621190130624, | |
| "grad_norm": 0.4254298806190491, | |
| "learning_rate": 8.685181997230614e-05, | |
| "loss": 1.668, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5500725689404935, | |
| "grad_norm": 0.35162025690078735, | |
| "learning_rate": 8.639329972250198e-05, | |
| "loss": 1.1166, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5515239477503628, | |
| "grad_norm": 0.46706706285476685, | |
| "learning_rate": 8.593507075375387e-05, | |
| "loss": 1.5868, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5529753265602322, | |
| "grad_norm": 0.22348397970199585, | |
| "learning_rate": 8.5477142875451e-05, | |
| "loss": 0.2983, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5544267053701016, | |
| "grad_norm": 0.39692750573158264, | |
| "learning_rate": 8.501952589053695e-05, | |
| "loss": 0.8706, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.555878084179971, | |
| "grad_norm": 0.5857199430465698, | |
| "learning_rate": 8.456222959530008e-05, | |
| "loss": 1.8278, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5573294629898403, | |
| "grad_norm": 0.4882966876029968, | |
| "learning_rate": 8.410526377916359e-05, | |
| "loss": 1.6223, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5587808417997098, | |
| "grad_norm": 0.3702254891395569, | |
| "learning_rate": 8.36486382244761e-05, | |
| "loss": 0.7289, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5602322206095791, | |
| "grad_norm": 0.3012731075286865, | |
| "learning_rate": 8.319236270630218e-05, | |
| "loss": 0.3799, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5616835994194485, | |
| "grad_norm": 0.439186155796051, | |
| "learning_rate": 8.273644699221308e-05, | |
| "loss": 1.2072, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5631349782293179, | |
| "grad_norm": 0.4469393193721771, | |
| "learning_rate": 8.228090084207774e-05, | |
| "loss": 0.9496, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5645863570391872, | |
| "grad_norm": 0.3852115571498871, | |
| "learning_rate": 8.182573400785367e-05, | |
| "loss": 0.767, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5660377358490566, | |
| "grad_norm": 0.21346166729927063, | |
| "learning_rate": 8.137095623337836e-05, | |
| "loss": 0.171, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5674891146589259, | |
| "grad_norm": 0.3433801531791687, | |
| "learning_rate": 8.091657725416065e-05, | |
| "loss": 0.3763, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5689404934687954, | |
| "grad_norm": 0.463051438331604, | |
| "learning_rate": 8.046260679717225e-05, | |
| "loss": 0.6219, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5703918722786647, | |
| "grad_norm": 0.41982725262641907, | |
| "learning_rate": 8.000905458063957e-05, | |
| "loss": 0.7653, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5718432510885341, | |
| "grad_norm": 0.30422887206077576, | |
| "learning_rate": 7.95559303138357e-05, | |
| "loss": 0.8267, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.5732946298984035, | |
| "grad_norm": 0.4346681237220764, | |
| "learning_rate": 7.91032436968725e-05, | |
| "loss": 1.2461, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5747460087082729, | |
| "grad_norm": 0.4238103926181793, | |
| "learning_rate": 7.865100442049303e-05, | |
| "loss": 1.3248, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5761973875181422, | |
| "grad_norm": 0.30451324582099915, | |
| "learning_rate": 7.8199222165864e-05, | |
| "loss": 0.5411, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5776487663280117, | |
| "grad_norm": 0.41048821806907654, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.825, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.579100145137881, | |
| "grad_norm": 0.4246201813220978, | |
| "learning_rate": 7.729706739739939e-05, | |
| "loss": 0.879, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5805515239477503, | |
| "grad_norm": 0.436656653881073, | |
| "learning_rate": 7.684671419615163e-05, | |
| "loss": 0.9675, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5820029027576198, | |
| "grad_norm": 0.5508980751037598, | |
| "learning_rate": 7.639685664141643e-05, | |
| "loss": 1.3912, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.5834542815674891, | |
| "grad_norm": 0.4069737493991852, | |
| "learning_rate": 7.594750436337467e-05, | |
| "loss": 1.6445, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.5849056603773585, | |
| "grad_norm": 0.43757930397987366, | |
| "learning_rate": 7.549866698139048e-05, | |
| "loss": 1.0197, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.5863570391872278, | |
| "grad_norm": 0.409341961145401, | |
| "learning_rate": 7.505035410380575e-05, | |
| "loss": 0.9127, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.5878084179970973, | |
| "grad_norm": 0.42377200722694397, | |
| "learning_rate": 7.460257532773401e-05, | |
| "loss": 1.6848, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5892597968069666, | |
| "grad_norm": 0.3736453950405121, | |
| "learning_rate": 7.415534023885529e-05, | |
| "loss": 0.8693, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.590711175616836, | |
| "grad_norm": 0.3143121898174286, | |
| "learning_rate": 7.37086584112108e-05, | |
| "loss": 1.3999, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.5921625544267054, | |
| "grad_norm": 0.5278828144073486, | |
| "learning_rate": 7.326253940699794e-05, | |
| "loss": 0.5467, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5936139332365747, | |
| "grad_norm": 0.3570123612880707, | |
| "learning_rate": 7.281699277636572e-05, | |
| "loss": 0.6123, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.5950653120464441, | |
| "grad_norm": 0.383098840713501, | |
| "learning_rate": 7.237202805721019e-05, | |
| "loss": 0.7806, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5965166908563135, | |
| "grad_norm": 0.32666346430778503, | |
| "learning_rate": 7.192765477497032e-05, | |
| "loss": 0.7825, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.5979680696661829, | |
| "grad_norm": 0.38186514377593994, | |
| "learning_rate": 7.148388244242414e-05, | |
| "loss": 0.5706, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.5994194484760522, | |
| "grad_norm": 0.3630182147026062, | |
| "learning_rate": 7.1040720559485e-05, | |
| "loss": 0.7846, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6008708272859217, | |
| "grad_norm": 0.6456810832023621, | |
| "learning_rate": 7.05981786129982e-05, | |
| "loss": 0.8707, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.602322206095791, | |
| "grad_norm": 0.47593608498573303, | |
| "learning_rate": 7.015626607653806e-05, | |
| "loss": 0.3883, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6037735849056604, | |
| "grad_norm": 0.3797203600406647, | |
| "learning_rate": 6.971499241020495e-05, | |
| "loss": 0.7263, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.6052249637155298, | |
| "grad_norm": 0.3255414664745331, | |
| "learning_rate": 6.927436706042276e-05, | |
| "loss": 0.4909, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6066763425253991, | |
| "grad_norm": 0.4275079369544983, | |
| "learning_rate": 6.883439945973692e-05, | |
| "loss": 1.446, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6081277213352685, | |
| "grad_norm": 0.5153216123580933, | |
| "learning_rate": 6.839509902661216e-05, | |
| "loss": 0.7299, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6095791001451378, | |
| "grad_norm": 0.3721558451652527, | |
| "learning_rate": 6.795647516523107e-05, | |
| "loss": 1.2028, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6110304789550073, | |
| "grad_norm": 0.3619592487812042, | |
| "learning_rate": 6.75185372652928e-05, | |
| "loss": 1.1034, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.6124818577648766, | |
| "grad_norm": 0.37321385741233826, | |
| "learning_rate": 6.708129470181198e-05, | |
| "loss": 1.169, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.613933236574746, | |
| "grad_norm": 0.3966982662677765, | |
| "learning_rate": 6.664475683491796e-05, | |
| "loss": 0.6847, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.5852516293525696, | |
| "learning_rate": 6.620893300965464e-05, | |
| "loss": 2.1163, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6168359941944848, | |
| "grad_norm": 0.45382705330848694, | |
| "learning_rate": 6.577383255578033e-05, | |
| "loss": 0.9365, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6182873730043541, | |
| "grad_norm": 0.34492751955986023, | |
| "learning_rate": 6.533946478756788e-05, | |
| "loss": 0.6898, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6197387518142236, | |
| "grad_norm": 0.40735992789268494, | |
| "learning_rate": 6.490583900360543e-05, | |
| "loss": 1.0304, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6211901306240929, | |
| "grad_norm": 0.4234837293624878, | |
| "learning_rate": 6.447296448659743e-05, | |
| "loss": 1.4856, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6226415094339622, | |
| "grad_norm": 0.32544052600860596, | |
| "learning_rate": 6.404085050316569e-05, | |
| "loss": 0.4444, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.6240928882438317, | |
| "grad_norm": 0.6741891503334045, | |
| "learning_rate": 6.360950630365126e-05, | |
| "loss": 1.5752, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.625544267053701, | |
| "grad_norm": 0.3429271876811981, | |
| "learning_rate": 6.317894112191621e-05, | |
| "loss": 0.3477, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6269956458635704, | |
| "grad_norm": 0.4417281746864319, | |
| "learning_rate": 6.274916417514604e-05, | |
| "loss": 1.2727, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6284470246734397, | |
| "grad_norm": 0.33864307403564453, | |
| "learning_rate": 6.232018466365247e-05, | |
| "loss": 0.4173, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6298984034833092, | |
| "grad_norm": 0.37801864743232727, | |
| "learning_rate": 6.189201177067623e-05, | |
| "loss": 0.8458, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6313497822931785, | |
| "grad_norm": 0.3596443235874176, | |
| "learning_rate": 6.146465466219072e-05, | |
| "loss": 1.2532, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6328011611030478, | |
| "grad_norm": 0.47118690609931946, | |
| "learning_rate": 6.103812248670567e-05, | |
| "loss": 0.8174, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6342525399129173, | |
| "grad_norm": 0.4070504903793335, | |
| "learning_rate": 6.061242437507131e-05, | |
| "loss": 0.7682, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6357039187227866, | |
| "grad_norm": 0.5048978924751282, | |
| "learning_rate": 6.018756944028291e-05, | |
| "loss": 0.953, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.637155297532656, | |
| "grad_norm": 0.4599890112876892, | |
| "learning_rate": 5.9763566777285786e-05, | |
| "loss": 1.2153, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6386066763425254, | |
| "grad_norm": 0.26790425181388855, | |
| "learning_rate": 5.93404254627804e-05, | |
| "loss": 0.4613, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6400580551523948, | |
| "grad_norm": 0.3035740256309509, | |
| "learning_rate": 5.891815455502824e-05, | |
| "loss": 0.565, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6415094339622641, | |
| "grad_norm": 0.41612568497657776, | |
| "learning_rate": 5.849676309365786e-05, | |
| "loss": 0.7652, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6429608127721336, | |
| "grad_norm": 0.3889887034893036, | |
| "learning_rate": 5.807626009947127e-05, | |
| "loss": 0.9254, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6444121915820029, | |
| "grad_norm": 0.3380615711212158, | |
| "learning_rate": 5.765665457425102e-05, | |
| "loss": 0.7985, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6458635703918723, | |
| "grad_norm": 0.2893095910549164, | |
| "learning_rate": 5.7237955500567205e-05, | |
| "loss": 0.6574, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6473149492017417, | |
| "grad_norm": 0.4444977641105652, | |
| "learning_rate": 5.682017184158561e-05, | |
| "loss": 0.8707, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.648766328011611, | |
| "grad_norm": 0.4025658071041107, | |
| "learning_rate": 5.6403312540875316e-05, | |
| "loss": 1.1545, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.6502177068214804, | |
| "grad_norm": 0.3898888826370239, | |
| "learning_rate": 5.5987386522217645e-05, | |
| "loss": 2.0098, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6516690856313497, | |
| "grad_norm": 0.42759767174720764, | |
| "learning_rate": 5.557240268941497e-05, | |
| "loss": 0.8124, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6531204644412192, | |
| "grad_norm": 0.35215985774993896, | |
| "learning_rate": 5.5158369926100104e-05, | |
| "loss": 1.0221, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6545718432510885, | |
| "grad_norm": 0.4174480140209198, | |
| "learning_rate": 5.474529709554612e-05, | |
| "loss": 1.2152, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6560232220609579, | |
| "grad_norm": 0.45766574144363403, | |
| "learning_rate": 5.433319304047666e-05, | |
| "loss": 1.7925, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.6574746008708273, | |
| "grad_norm": 0.40072959661483765, | |
| "learning_rate": 5.392206658287669e-05, | |
| "loss": 0.9568, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6589259796806967, | |
| "grad_norm": 0.41294583678245544, | |
| "learning_rate": 5.3511926523803416e-05, | |
| "loss": 1.7724, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.660377358490566, | |
| "grad_norm": 0.40195024013519287, | |
| "learning_rate": 5.310278164319815e-05, | |
| "loss": 0.7026, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6618287373004355, | |
| "grad_norm": 0.21002167463302612, | |
| "learning_rate": 5.269464069969829e-05, | |
| "loss": 0.3955, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.6632801161103048, | |
| "grad_norm": 0.34857895970344543, | |
| "learning_rate": 5.2287512430449604e-05, | |
| "loss": 0.6136, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.6647314949201741, | |
| "grad_norm": 0.4372917711734772, | |
| "learning_rate": 5.1881405550919493e-05, | |
| "loss": 1.0637, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.6661828737300436, | |
| "grad_norm": 0.3765782415866852, | |
| "learning_rate": 5.147632875471029e-05, | |
| "loss": 0.7807, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.6676342525399129, | |
| "grad_norm": 0.32897019386291504, | |
| "learning_rate": 5.107229071337305e-05, | |
| "loss": 0.9325, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6690856313497823, | |
| "grad_norm": 0.48499614000320435, | |
| "learning_rate": 5.0669300076222096e-05, | |
| "loss": 1.2014, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.6705370101596516, | |
| "grad_norm": 0.29446157813072205, | |
| "learning_rate": 5.026736547014981e-05, | |
| "loss": 0.3329, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.6719883889695211, | |
| "grad_norm": 0.47891777753829956, | |
| "learning_rate": 4.9866495499441824e-05, | |
| "loss": 1.0438, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6734397677793904, | |
| "grad_norm": 0.34177759289741516, | |
| "learning_rate": 4.946669874559301e-05, | |
| "loss": 1.4163, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.6748911465892597, | |
| "grad_norm": 0.47582611441612244, | |
| "learning_rate": 4.9067983767123736e-05, | |
| "loss": 0.4841, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6763425253991292, | |
| "grad_norm": 0.3827131986618042, | |
| "learning_rate": 4.8670359099396466e-05, | |
| "loss": 0.495, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.6777939042089985, | |
| "grad_norm": 0.4850855767726898, | |
| "learning_rate": 4.827383325443331e-05, | |
| "loss": 1.6036, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.6792452830188679, | |
| "grad_norm": 0.29585230350494385, | |
| "learning_rate": 4.7878414720733664e-05, | |
| "loss": 0.6934, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.6806966618287373, | |
| "grad_norm": 0.2892296612262726, | |
| "learning_rate": 4.748411196309248e-05, | |
| "loss": 0.5579, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.6821480406386067, | |
| "grad_norm": 0.3687235713005066, | |
| "learning_rate": 4.709093342241917e-05, | |
| "loss": 1.1527, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.683599419448476, | |
| "grad_norm": 0.3160078823566437, | |
| "learning_rate": 4.6698887515556685e-05, | |
| "loss": 0.5885, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.6850507982583455, | |
| "grad_norm": 0.2962639629840851, | |
| "learning_rate": 4.630798263510162e-05, | |
| "loss": 1.2475, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.6865021770682148, | |
| "grad_norm": 0.4450148344039917, | |
| "learning_rate": 4.59182271492244e-05, | |
| "loss": 1.6886, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.6879535558780842, | |
| "grad_norm": 0.3711402118206024, | |
| "learning_rate": 4.552962940149005e-05, | |
| "loss": 1.1539, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.6894049346879536, | |
| "grad_norm": 0.38809695839881897, | |
| "learning_rate": 4.51421977106798e-05, | |
| "loss": 0.8469, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.690856313497823, | |
| "grad_norm": 0.41775164008140564, | |
| "learning_rate": 4.475594037061285e-05, | |
| "loss": 1.8853, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.4020920395851135, | |
| "learning_rate": 4.4370865649968906e-05, | |
| "loss": 0.4945, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.6937590711175616, | |
| "grad_norm": 0.3365013003349304, | |
| "learning_rate": 4.3986981792111025e-05, | |
| "loss": 0.6874, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.6952104499274311, | |
| "grad_norm": 0.26148009300231934, | |
| "learning_rate": 4.360429701490934e-05, | |
| "loss": 0.194, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.6966618287373004, | |
| "grad_norm": 0.3839820921421051, | |
| "learning_rate": 4.322281951056508e-05, | |
| "loss": 0.7245, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6981132075471698, | |
| "grad_norm": 1.308461308479309, | |
| "learning_rate": 4.284255744543505e-05, | |
| "loss": 0.7233, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.6995645863570392, | |
| "grad_norm": 0.39677149057388306, | |
| "learning_rate": 4.246351895985702e-05, | |
| "loss": 0.3733, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7010159651669086, | |
| "grad_norm": 0.4022575318813324, | |
| "learning_rate": 4.208571216797539e-05, | |
| "loss": 1.3429, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7024673439767779, | |
| "grad_norm": 0.3253962993621826, | |
| "learning_rate": 4.170914515756737e-05, | |
| "loss": 0.3215, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7039187227866474, | |
| "grad_norm": 0.1781662255525589, | |
| "learning_rate": 4.133382598987007e-05, | |
| "loss": 0.1333, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7053701015965167, | |
| "grad_norm": 0.23567990958690643, | |
| "learning_rate": 4.0959762699407766e-05, | |
| "loss": 0.441, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.706821480406386, | |
| "grad_norm": 0.29025065898895264, | |
| "learning_rate": 4.058696329381987e-05, | |
| "loss": 0.3822, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7082728592162555, | |
| "grad_norm": 0.2783532738685608, | |
| "learning_rate": 4.0215435753689676e-05, | |
| "loss": 0.3143, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7097242380261248, | |
| "grad_norm": 0.3103715181350708, | |
| "learning_rate": 3.9845188032373426e-05, | |
| "loss": 0.8413, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7111756168359942, | |
| "grad_norm": 0.5938622355461121, | |
| "learning_rate": 3.947622805583001e-05, | |
| "loss": 1.302, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7126269956458636, | |
| "grad_norm": 0.3905632495880127, | |
| "learning_rate": 3.910856372245139e-05, | |
| "loss": 0.8033, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.714078374455733, | |
| "grad_norm": 0.38237109780311584, | |
| "learning_rate": 3.874220290289336e-05, | |
| "loss": 1.2902, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7155297532656023, | |
| "grad_norm": 0.43254566192626953, | |
| "learning_rate": 3.8377153439907266e-05, | |
| "loss": 0.7916, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7169811320754716, | |
| "grad_norm": 0.45678088068962097, | |
| "learning_rate": 3.801342314817201e-05, | |
| "loss": 1.8683, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7184325108853411, | |
| "grad_norm": 0.2868676483631134, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.3818, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7198838896952104, | |
| "grad_norm": 0.45408400893211365, | |
| "learning_rate": 3.7289951195803954e-05, | |
| "loss": 1.142, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.7213352685050798, | |
| "grad_norm": 0.47678226232528687, | |
| "learning_rate": 3.693022502266413e-05, | |
| "loss": 1.5476, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7227866473149492, | |
| "grad_norm": 0.4446166753768921, | |
| "learning_rate": 3.6571848995429405e-05, | |
| "loss": 0.5653, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7242380261248186, | |
| "grad_norm": 0.31094321608543396, | |
| "learning_rate": 3.621483078591923e-05, | |
| "loss": 0.7948, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7256894049346879, | |
| "grad_norm": 0.40678948163986206, | |
| "learning_rate": 3.585917803688603e-05, | |
| "loss": 1.1627, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7271407837445574, | |
| "grad_norm": 0.5062684416770935, | |
| "learning_rate": 3.550489836185149e-05, | |
| "loss": 0.9259, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.7285921625544267, | |
| "grad_norm": 0.29026418924331665, | |
| "learning_rate": 3.5151999344943735e-05, | |
| "loss": 0.8886, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.7300435413642961, | |
| "grad_norm": 0.3942019045352936, | |
| "learning_rate": 3.4800488540734856e-05, | |
| "loss": 0.7888, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7314949201741655, | |
| "grad_norm": 0.6088467836380005, | |
| "learning_rate": 3.4450373474079276e-05, | |
| "loss": 1.9246, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7329462989840348, | |
| "grad_norm": 0.45419415831565857, | |
| "learning_rate": 3.4101661639952534e-05, | |
| "loss": 1.4949, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7343976777939042, | |
| "grad_norm": 0.4433126747608185, | |
| "learning_rate": 3.375436050329097e-05, | |
| "loss": 1.4551, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.7358490566037735, | |
| "grad_norm": 0.32893896102905273, | |
| "learning_rate": 3.340847749883191e-05, | |
| "loss": 0.8053, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.737300435413643, | |
| "grad_norm": 0.36813730001449585, | |
| "learning_rate": 3.306402003095438e-05, | |
| "loss": 0.9338, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7387518142235123, | |
| "grad_norm": 0.2666694223880768, | |
| "learning_rate": 3.2720995473520766e-05, | |
| "loss": 0.5419, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7402031930333817, | |
| "grad_norm": 0.4525779187679291, | |
| "learning_rate": 3.237941116971884e-05, | |
| "loss": 0.658, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7416545718432511, | |
| "grad_norm": 0.5365322828292847, | |
| "learning_rate": 3.203927443190464e-05, | |
| "loss": 1.3406, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.7431059506531205, | |
| "grad_norm": 0.5615587830543518, | |
| "learning_rate": 3.170059254144593e-05, | |
| "loss": 1.6261, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7445573294629898, | |
| "grad_norm": 0.42640161514282227, | |
| "learning_rate": 3.1363372748566176e-05, | |
| "loss": 1.1084, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7460087082728593, | |
| "grad_norm": 0.4219587743282318, | |
| "learning_rate": 3.102762227218957e-05, | |
| "loss": 1.109, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7474600870827286, | |
| "grad_norm": 0.3355334401130676, | |
| "learning_rate": 3.0693348299786405e-05, | |
| "loss": 1.3992, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7489114658925979, | |
| "grad_norm": 0.6047046184539795, | |
| "learning_rate": 3.0360557987219053e-05, | |
| "loss": 2.6407, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.7503628447024674, | |
| "grad_norm": 0.5143350958824158, | |
| "learning_rate": 3.0029258458589048e-05, | |
| "loss": 0.8794, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7518142235123367, | |
| "grad_norm": 0.36478573083877563, | |
| "learning_rate": 2.9699456806084426e-05, | |
| "loss": 0.5779, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7532656023222061, | |
| "grad_norm": 0.310554563999176, | |
| "learning_rate": 2.9371160089827842e-05, | |
| "loss": 0.633, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7532656023222061, | |
| "eval_loss": 0.23614992201328278, | |
| "eval_runtime": 32.4618, | |
| "eval_samples_per_second": 8.964, | |
| "eval_steps_per_second": 4.498, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 0.24123501777648926, | |
| "learning_rate": 2.9044375337725593e-05, | |
| "loss": 0.4715, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7561683599419449, | |
| "grad_norm": 0.4682220220565796, | |
| "learning_rate": 2.8719109545317103e-05, | |
| "loss": 0.8464, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.7576197387518142, | |
| "grad_norm": 0.3238670527935028, | |
| "learning_rate": 2.839536967562504e-05, | |
| "loss": 0.34, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.7590711175616836, | |
| "grad_norm": 0.46153318881988525, | |
| "learning_rate": 2.807316265900649e-05, | |
| "loss": 1.183, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.760522496371553, | |
| "grad_norm": 0.42253023386001587, | |
| "learning_rate": 2.7752495393004475e-05, | |
| "loss": 0.5708, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.7619738751814223, | |
| "grad_norm": 0.4283324182033539, | |
| "learning_rate": 2.7433374742200192e-05, | |
| "loss": 1.2854, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7634252539912917, | |
| "grad_norm": 0.33337557315826416, | |
| "learning_rate": 2.7115807538066297e-05, | |
| "loss": 0.6406, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.7648766328011611, | |
| "grad_norm": 0.33540481328964233, | |
| "learning_rate": 2.679980057882049e-05, | |
| "loss": 0.5988, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.7663280116110305, | |
| "grad_norm": 0.5348614454269409, | |
| "learning_rate": 2.6485360629279987e-05, | |
| "loss": 1.6389, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.7677793904208998, | |
| "grad_norm": 0.36704495549201965, | |
| "learning_rate": 2.6172494420716797e-05, | |
| "loss": 1.0625, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.3153030276298523, | |
| "learning_rate": 2.5861208650713588e-05, | |
| "loss": 0.3844, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7706821480406386, | |
| "grad_norm": 0.29406848549842834, | |
| "learning_rate": 2.555150998302026e-05, | |
| "loss": 0.575, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.772133526850508, | |
| "grad_norm": 0.35533541440963745, | |
| "learning_rate": 2.524340504741135e-05, | |
| "loss": 0.9053, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.7735849056603774, | |
| "grad_norm": 0.35642877221107483, | |
| "learning_rate": 2.4936900439544107e-05, | |
| "loss": 1.6089, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.7750362844702468, | |
| "grad_norm": 0.5626897811889648, | |
| "learning_rate": 2.4632002720817228e-05, | |
| "loss": 1.136, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.7764876632801161, | |
| "grad_norm": 0.3613468408584595, | |
| "learning_rate": 2.432871841823047e-05, | |
| "loss": 1.0527, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.7779390420899854, | |
| "grad_norm": 0.5131534934043884, | |
| "learning_rate": 2.4027054024244956e-05, | |
| "loss": 1.0593, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.7793904208998549, | |
| "grad_norm": 0.2713516652584076, | |
| "learning_rate": 2.372701599664404e-05, | |
| "loss": 0.3204, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.7808417997097242, | |
| "grad_norm": 0.3534812033176422, | |
| "learning_rate": 2.342861075839523e-05, | |
| "loss": 0.6906, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.7822931785195936, | |
| "grad_norm": 0.2669237554073334, | |
| "learning_rate": 2.313184469751265e-05, | |
| "loss": 0.4097, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.783744557329463, | |
| "grad_norm": 0.26740333437919617, | |
| "learning_rate": 2.283672416692013e-05, | |
| "loss": 0.2641, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7851959361393324, | |
| "grad_norm": 0.4208941161632538, | |
| "learning_rate": 2.2543255484315484e-05, | |
| "loss": 0.9359, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.7866473149492017, | |
| "grad_norm": 0.32194021344184875, | |
| "learning_rate": 2.2251444932035094e-05, | |
| "loss": 0.6239, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.7880986937590712, | |
| "grad_norm": 0.4907798767089844, | |
| "learning_rate": 2.1961298756919335e-05, | |
| "loss": 1.76, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.7895500725689405, | |
| "grad_norm": 0.45274341106414795, | |
| "learning_rate": 2.16728231701791e-05, | |
| "loss": 1.3965, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.7910014513788098, | |
| "grad_norm": 0.3342875838279724, | |
| "learning_rate": 2.138602434726269e-05, | |
| "loss": 0.2511, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.7924528301886793, | |
| "grad_norm": 0.37094488739967346, | |
| "learning_rate": 2.11009084277235e-05, | |
| "loss": 1.1932, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.7939042089985486, | |
| "grad_norm": 0.36341187357902527, | |
| "learning_rate": 2.0817481515088833e-05, | |
| "loss": 0.7164, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.795355587808418, | |
| "grad_norm": 0.4300606846809387, | |
| "learning_rate": 2.053574967672912e-05, | |
| "loss": 0.9558, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.7968069666182874, | |
| "grad_norm": 0.4252452850341797, | |
| "learning_rate": 2.025571894372794e-05, | |
| "loss": 1.1288, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.7982583454281568, | |
| "grad_norm": 0.5417127013206482, | |
| "learning_rate": 1.9977395310753087e-05, | |
| "loss": 2.1247, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7997097242380261, | |
| "grad_norm": 0.3363449275493622, | |
| "learning_rate": 1.970078473592817e-05, | |
| "loss": 0.5331, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8011611030478955, | |
| "grad_norm": 0.2750878632068634, | |
| "learning_rate": 1.9425893140704943e-05, | |
| "loss": 0.3197, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8026124818577649, | |
| "grad_norm": 0.2823106050491333, | |
| "learning_rate": 1.915272640973683e-05, | |
| "loss": 0.6639, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8040638606676342, | |
| "grad_norm": 0.41894766688346863, | |
| "learning_rate": 1.888129039075268e-05, | |
| "loss": 1.0442, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8055152394775036, | |
| "grad_norm": 0.3718733489513397, | |
| "learning_rate": 1.8611590894431653e-05, | |
| "loss": 0.7986, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.806966618287373, | |
| "grad_norm": 0.3276025950908661, | |
| "learning_rate": 1.8343633694278895e-05, | |
| "loss": 0.7076, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8084179970972424, | |
| "grad_norm": 0.3458663523197174, | |
| "learning_rate": 1.807742452650196e-05, | |
| "loss": 0.8522, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8098693759071117, | |
| "grad_norm": 0.41551870107650757, | |
| "learning_rate": 1.7812969089887864e-05, | |
| "loss": 0.7282, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8113207547169812, | |
| "grad_norm": 0.36481913924217224, | |
| "learning_rate": 1.7550273045681242e-05, | |
| "loss": 0.5842, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.8127721335268505, | |
| "grad_norm": 0.31234338879585266, | |
| "learning_rate": 1.7289342017463138e-05, | |
| "loss": 1.0055, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8142235123367199, | |
| "grad_norm": 0.29913321137428284, | |
| "learning_rate": 1.7030181591030493e-05, | |
| "loss": 0.4555, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8156748911465893, | |
| "grad_norm": 0.2602443993091583, | |
| "learning_rate": 1.6772797314276713e-05, | |
| "loss": 0.5486, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.8171262699564587, | |
| "grad_norm": 0.43661990761756897, | |
| "learning_rate": 1.65171946970729e-05, | |
| "loss": 1.7129, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.818577648766328, | |
| "grad_norm": 0.658896267414093, | |
| "learning_rate": 1.6263379211149732e-05, | |
| "loss": 2.026, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8200290275761973, | |
| "grad_norm": 0.48423880338668823, | |
| "learning_rate": 1.601135628998056e-05, | |
| "loss": 1.1105, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.8214804063860668, | |
| "grad_norm": 0.3823278844356537, | |
| "learning_rate": 1.5761131328664947e-05, | |
| "loss": 0.9354, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8229317851959361, | |
| "grad_norm": 0.38709479570388794, | |
| "learning_rate": 1.5512709683813164e-05, | |
| "loss": 1.7732, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8243831640058055, | |
| "grad_norm": 0.4736160933971405, | |
| "learning_rate": 1.526609667343163e-05, | |
| "loss": 1.0733, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.8258345428156749, | |
| "grad_norm": 0.4745509922504425, | |
| "learning_rate": 1.5021297576809001e-05, | |
| "loss": 1.4294, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.8272859216255443, | |
| "grad_norm": 0.28139880299568176, | |
| "learning_rate": 1.4778317634403083e-05, | |
| "loss": 0.2901, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8287373004354136, | |
| "grad_norm": 0.39859816431999207, | |
| "learning_rate": 1.4537162047728804e-05, | |
| "loss": 0.7626, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.8301886792452831, | |
| "grad_norm": 0.8511147499084473, | |
| "learning_rate": 1.4297835979246777e-05, | |
| "loss": 1.482, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8316400580551524, | |
| "grad_norm": 0.4506050646305084, | |
| "learning_rate": 1.4060344552252702e-05, | |
| "loss": 1.2382, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.8330914368650217, | |
| "grad_norm": 0.22067701816558838, | |
| "learning_rate": 1.3824692850767929e-05, | |
| "loss": 0.571, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.8345428156748912, | |
| "grad_norm": 0.3021042048931122, | |
| "learning_rate": 1.3590885919430384e-05, | |
| "loss": 0.522, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8359941944847605, | |
| "grad_norm": 0.49198654294013977, | |
| "learning_rate": 1.3358928763386646e-05, | |
| "loss": 0.9773, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8374455732946299, | |
| "grad_norm": 0.35230767726898193, | |
| "learning_rate": 1.3128826348184887e-05, | |
| "loss": 0.6312, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.8388969521044993, | |
| "grad_norm": 0.3485874831676483, | |
| "learning_rate": 1.2900583599668515e-05, | |
| "loss": 0.6788, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.8403483309143687, | |
| "grad_norm": 0.3603363335132599, | |
| "learning_rate": 1.2674205403870642e-05, | |
| "loss": 0.7338, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.841799709724238, | |
| "grad_norm": 0.25826606154441833, | |
| "learning_rate": 1.2449696606909655e-05, | |
| "loss": 0.4404, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8432510885341074, | |
| "grad_norm": 0.3804086744785309, | |
| "learning_rate": 1.222706201488536e-05, | |
| "loss": 0.4955, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8447024673439768, | |
| "grad_norm": 0.47231999039649963, | |
| "learning_rate": 1.200630639377609e-05, | |
| "loss": 1.6225, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.44939887523651123, | |
| "learning_rate": 1.1787434469336766e-05, | |
| "loss": 1.252, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.8476052249637155, | |
| "grad_norm": 0.3787010610103607, | |
| "learning_rate": 1.1570450926997655e-05, | |
| "loss": 0.544, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.8490566037735849, | |
| "grad_norm": 0.4327794909477234, | |
| "learning_rate": 1.135536041176406e-05, | |
| "loss": 0.3797, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8505079825834543, | |
| "grad_norm": 0.35290834307670593, | |
| "learning_rate": 1.1142167528116965e-05, | |
| "loss": 0.8393, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.8519593613933236, | |
| "grad_norm": 0.24693018198013306, | |
| "learning_rate": 1.0930876839914416e-05, | |
| "loss": 0.1381, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.8534107402031931, | |
| "grad_norm": 0.3256937265396118, | |
| "learning_rate": 1.072149287029377e-05, | |
| "loss": 0.5073, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.8548621190130624, | |
| "grad_norm": 0.32373976707458496, | |
| "learning_rate": 1.0514020101574972e-05, | |
| "loss": 0.4282, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.8563134978229318, | |
| "grad_norm": 0.3258521854877472, | |
| "learning_rate": 1.0308462975164545e-05, | |
| "loss": 0.348, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8577648766328012, | |
| "grad_norm": 0.3401307165622711, | |
| "learning_rate": 1.010482589146048e-05, | |
| "loss": 1.1884, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.8592162554426706, | |
| "grad_norm": 0.39415818452835083, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 1.2618, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.8606676342525399, | |
| "grad_norm": 0.3865334689617157, | |
| "learning_rate": 9.703329248156711e-06, | |
| "loss": 0.9641, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.8621190130624092, | |
| "grad_norm": 0.5107492208480835, | |
| "learning_rate": 9.505478283467128e-06, | |
| "loss": 0.6755, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.8635703918722787, | |
| "grad_norm": 0.3883417248725891, | |
| "learning_rate": 9.309564551120254e-06, | |
| "loss": 1.4845, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.865021770682148, | |
| "grad_norm": 0.5274085998535156, | |
| "learning_rate": 9.115592245076177e-06, | |
| "loss": 1.4281, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.8664731494920174, | |
| "grad_norm": 0.4148600399494171, | |
| "learning_rate": 8.923565517734633e-06, | |
| "loss": 1.4102, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.8679245283018868, | |
| "grad_norm": 0.3156782388687134, | |
| "learning_rate": 8.733488479845997e-06, | |
| "loss": 0.7565, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.8693759071117562, | |
| "grad_norm": 0.41858044266700745, | |
| "learning_rate": 8.545365200423217e-06, | |
| "loss": 1.3413, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.8708272859216255, | |
| "grad_norm": 0.3816291391849518, | |
| "learning_rate": 8.359199706654853e-06, | |
| "loss": 0.9125, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.872278664731495, | |
| "grad_norm": 0.3754681944847107, | |
| "learning_rate": 8.1749959838188e-06, | |
| "loss": 0.5213, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.8737300435413643, | |
| "grad_norm": 0.32858067750930786, | |
| "learning_rate": 7.992757975196974e-06, | |
| "loss": 0.9199, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.8751814223512336, | |
| "grad_norm": 0.4272098243236542, | |
| "learning_rate": 7.812489581990844e-06, | |
| "loss": 1.4278, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.8766328011611031, | |
| "grad_norm": 0.35721269249916077, | |
| "learning_rate": 7.634194663238015e-06, | |
| "loss": 0.4783, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.8780841799709724, | |
| "grad_norm": 0.38133150339126587, | |
| "learning_rate": 7.457877035729588e-06, | |
| "loss": 0.6231, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.8795355587808418, | |
| "grad_norm": 0.3652817904949188, | |
| "learning_rate": 7.283540473928385e-06, | |
| "loss": 0.7934, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.8809869375907112, | |
| "grad_norm": 0.3659121096134186, | |
| "learning_rate": 7.11118870988825e-06, | |
| "loss": 0.9149, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.8824383164005806, | |
| "grad_norm": 0.4516978859901428, | |
| "learning_rate": 6.940825433174103e-06, | |
| "loss": 1.2015, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.8838896952104499, | |
| "grad_norm": 0.3403629660606384, | |
| "learning_rate": 6.772454290782926e-06, | |
| "loss": 0.552, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.8853410740203193, | |
| "grad_norm": 0.24706141650676727, | |
| "learning_rate": 6.606078887065748e-06, | |
| "loss": 0.2273, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8867924528301887, | |
| "grad_norm": 0.3536119759082794, | |
| "learning_rate": 6.441702783650494e-06, | |
| "loss": 0.7869, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.888243831640058, | |
| "grad_norm": 0.43990838527679443, | |
| "learning_rate": 6.2793294993656494e-06, | |
| "loss": 0.745, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.8896952104499274, | |
| "grad_norm": 0.560270369052887, | |
| "learning_rate": 6.118962510165038e-06, | |
| "loss": 1.311, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.8911465892597968, | |
| "grad_norm": 0.3678494691848755, | |
| "learning_rate": 5.960605249053386e-06, | |
| "loss": 1.182, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.8925979680696662, | |
| "grad_norm": 0.36771297454833984, | |
| "learning_rate": 5.8042611060127315e-06, | |
| "loss": 0.7735, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.8940493468795355, | |
| "grad_norm": 0.43662601709365845, | |
| "learning_rate": 5.649933427930043e-06, | |
| "loss": 0.9749, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.895500725689405, | |
| "grad_norm": 0.4699442386627197, | |
| "learning_rate": 5.497625518525373e-06, | |
| "loss": 0.9527, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.8969521044992743, | |
| "grad_norm": 0.3796102702617645, | |
| "learning_rate": 5.347340638281273e-06, | |
| "loss": 0.7223, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.8984034833091437, | |
| "grad_norm": 0.3132692277431488, | |
| "learning_rate": 5.199082004372957e-06, | |
| "loss": 0.3637, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.8998548621190131, | |
| "grad_norm": 0.46878132224082947, | |
| "learning_rate": 5.052852790599383e-06, | |
| "loss": 1.131, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9013062409288825, | |
| "grad_norm": 0.359684556722641, | |
| "learning_rate": 4.908656127315359e-06, | |
| "loss": 0.6985, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9027576197387518, | |
| "grad_norm": 0.32585737109184265, | |
| "learning_rate": 4.7664951013645875e-06, | |
| "loss": 0.6099, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9042089985486212, | |
| "grad_norm": 0.4055173397064209, | |
| "learning_rate": 4.626372756013409e-06, | |
| "loss": 0.6696, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.9056603773584906, | |
| "grad_norm": 0.3793414235115051, | |
| "learning_rate": 4.48829209088586e-06, | |
| "loss": 1.2006, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9071117561683599, | |
| "grad_norm": 0.41631418466567993, | |
| "learning_rate": 4.352256061899329e-06, | |
| "loss": 0.6615, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9085631349782293, | |
| "grad_norm": 0.28770431876182556, | |
| "learning_rate": 4.2182675812012965e-06, | |
| "loss": 0.4437, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.9100145137880987, | |
| "grad_norm": 0.4755733609199524, | |
| "learning_rate": 4.086329517107046e-06, | |
| "loss": 0.9437, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9114658925979681, | |
| "grad_norm": 0.4184425473213196, | |
| "learning_rate": 3.9564446940382084e-06, | |
| "loss": 1.8477, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9129172714078374, | |
| "grad_norm": 0.515277087688446, | |
| "learning_rate": 3.828615892462328e-06, | |
| "loss": 1.0525, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.9143686502177069, | |
| "grad_norm": 0.35891175270080566, | |
| "learning_rate": 3.7028458488333138e-06, | |
| "loss": 1.4299, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9158200290275762, | |
| "grad_norm": 0.34797385334968567, | |
| "learning_rate": 3.5791372555328937e-06, | |
| "loss": 0.3932, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9172714078374455, | |
| "grad_norm": 0.4482119679450989, | |
| "learning_rate": 3.457492760812975e-06, | |
| "loss": 0.9157, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.918722786647315, | |
| "grad_norm": 0.43506234884262085, | |
| "learning_rate": 3.3379149687388867e-06, | |
| "loss": 0.6012, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9201741654571843, | |
| "grad_norm": 0.539933443069458, | |
| "learning_rate": 3.2204064391337273e-06, | |
| "loss": 0.8629, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.9216255442670537, | |
| "grad_norm": 0.32083696126937866, | |
| "learning_rate": 3.1049696875235512e-06, | |
| "loss": 0.4713, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.5421581864356995, | |
| "learning_rate": 2.9916071850834005e-06, | |
| "loss": 1.2597, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9245283018867925, | |
| "grad_norm": 0.2753216028213501, | |
| "learning_rate": 2.8803213585846035e-06, | |
| "loss": 0.4526, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9259796806966618, | |
| "grad_norm": 0.5353147983551025, | |
| "learning_rate": 2.7711145903426496e-06, | |
| "loss": 1.0714, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9274310595065312, | |
| "grad_norm": 0.26769256591796875, | |
| "learning_rate": 2.663989218166274e-06, | |
| "loss": 1.1174, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.9288824383164006, | |
| "grad_norm": 0.2907368838787079, | |
| "learning_rate": 2.5589475353073988e-06, | |
| "loss": 1.2706, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.93033381712627, | |
| "grad_norm": 0.49619778990745544, | |
| "learning_rate": 2.455991790412027e-06, | |
| "loss": 1.0534, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9317851959361393, | |
| "grad_norm": 0.32665136456489563, | |
| "learning_rate": 2.355124187472135e-06, | |
| "loss": 0.9813, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9332365747460087, | |
| "grad_norm": 0.24648268520832062, | |
| "learning_rate": 2.2563468857784818e-06, | |
| "loss": 0.3728, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9346879535558781, | |
| "grad_norm": 0.45106983184814453, | |
| "learning_rate": 2.159661999874307e-06, | |
| "loss": 0.6732, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.9361393323657474, | |
| "grad_norm": 0.4321921467781067, | |
| "learning_rate": 2.0650715995102267e-06, | |
| "loss": 1.9884, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9375907111756169, | |
| "grad_norm": 0.33670032024383545, | |
| "learning_rate": 1.9725777095997657e-06, | |
| "loss": 0.6079, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.9390420899854862, | |
| "grad_norm": 0.324014276266098, | |
| "learning_rate": 1.882182310176095e-06, | |
| "loss": 0.2433, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.9404934687953556, | |
| "grad_norm": 0.41200318932533264, | |
| "learning_rate": 1.7938873363496422e-06, | |
| "loss": 1.2887, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.941944847605225, | |
| "grad_norm": 0.4197370707988739, | |
| "learning_rate": 1.707694678266636e-06, | |
| "loss": 1.1368, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 0.5854864120483398, | |
| "learning_rate": 1.6236061810686487e-06, | |
| "loss": 0.5592, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9448476052249637, | |
| "grad_norm": 0.426692932844162, | |
| "learning_rate": 1.5416236448531074e-06, | |
| "loss": 1.1019, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.9462989840348331, | |
| "grad_norm": 0.4369201362133026, | |
| "learning_rate": 1.461748824634801e-06, | |
| "loss": 1.0856, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.9477503628447025, | |
| "grad_norm": 0.3605095148086548, | |
| "learning_rate": 1.383983430308189e-06, | |
| "loss": 1.3698, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.9492017416545718, | |
| "grad_norm": 0.28937241435050964, | |
| "learning_rate": 1.30832912661093e-06, | |
| "loss": 0.4791, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.9506531204644412, | |
| "grad_norm": 0.38452112674713135, | |
| "learning_rate": 1.2347875330881886e-06, | |
| "loss": 1.3005, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9521044992743106, | |
| "grad_norm": 0.44588330388069153, | |
| "learning_rate": 1.1633602240579522e-06, | |
| "loss": 1.1278, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.95355587808418, | |
| "grad_norm": 0.34775638580322266, | |
| "learning_rate": 1.0940487285773459e-06, | |
| "loss": 0.6701, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.9550072568940493, | |
| "grad_norm": 0.34269067645072937, | |
| "learning_rate": 1.026854530409882e-06, | |
| "loss": 1.3742, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.9564586357039188, | |
| "grad_norm": 0.3833567500114441, | |
| "learning_rate": 9.61779067993729e-07, | |
| "loss": 1.387, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.9579100145137881, | |
| "grad_norm": 0.37239694595336914, | |
| "learning_rate": 8.988237344109251e-07, | |
| "loss": 0.8695, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9593613933236574, | |
| "grad_norm": 0.30777066946029663, | |
| "learning_rate": 8.379898773574924e-07, | |
| "loss": 0.7924, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.9608127721335269, | |
| "grad_norm": 0.38899528980255127, | |
| "learning_rate": 7.792787991146355e-07, | |
| "loss": 1.4396, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.9622641509433962, | |
| "grad_norm": 0.4895128309726715, | |
| "learning_rate": 7.226917565208657e-07, | |
| "loss": 1.3093, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.9637155297532656, | |
| "grad_norm": 0.5337472558021545, | |
| "learning_rate": 6.682299609450993e-07, | |
| "loss": 0.8694, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.965166908563135, | |
| "grad_norm": 0.32183799147605896, | |
| "learning_rate": 6.158945782606673e-07, | |
| "loss": 0.5548, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9666182873730044, | |
| "grad_norm": 0.3950204849243164, | |
| "learning_rate": 5.65686728820436e-07, | |
| "loss": 0.4285, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.9680696661828737, | |
| "grad_norm": 0.4570080041885376, | |
| "learning_rate": 5.176074874327919e-07, | |
| "loss": 1.5729, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.969521044992743, | |
| "grad_norm": 0.42208755016326904, | |
| "learning_rate": 4.7165788333860536e-07, | |
| "loss": 0.8863, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.9709724238026125, | |
| "grad_norm": 0.29929739236831665, | |
| "learning_rate": 4.278389001892369e-07, | |
| "loss": 0.7099, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.9724238026124818, | |
| "grad_norm": 0.46631020307540894, | |
| "learning_rate": 3.8615147602546473e-07, | |
| "loss": 0.4506, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9738751814223512, | |
| "grad_norm": 0.4783049523830414, | |
| "learning_rate": 3.4659650325740147e-07, | |
| "loss": 1.1751, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.9753265602322206, | |
| "grad_norm": 0.3503381907939911, | |
| "learning_rate": 3.091748286453866e-07, | |
| "loss": 0.9895, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.97677793904209, | |
| "grad_norm": 0.28899291157722473, | |
| "learning_rate": 2.7388725328189036e-07, | |
| "loss": 0.5174, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.9782293178519593, | |
| "grad_norm": 0.33355581760406494, | |
| "learning_rate": 2.407345325743049e-07, | |
| "loss": 0.4006, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.9796806966618288, | |
| "grad_norm": 0.40867161750793457, | |
| "learning_rate": 2.0971737622883515e-07, | |
| "loss": 1.0604, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.9811320754716981, | |
| "grad_norm": 0.40149986743927, | |
| "learning_rate": 1.8083644823526647e-07, | |
| "loss": 0.7342, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.9825834542815675, | |
| "grad_norm": 0.4048071503639221, | |
| "learning_rate": 1.5409236685277605e-07, | |
| "loss": 0.966, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.9840348330914369, | |
| "grad_norm": 0.4240354895591736, | |
| "learning_rate": 1.2948570459667686e-07, | |
| "loss": 1.29, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.9854862119013063, | |
| "grad_norm": 0.429208904504776, | |
| "learning_rate": 1.0701698822614959e-07, | |
| "loss": 0.985, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.9869375907111756, | |
| "grad_norm": 0.29551103711128235, | |
| "learning_rate": 8.668669873304058e-08, | |
| "loss": 0.6212, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.988388969521045, | |
| "grad_norm": 0.4513908922672272, | |
| "learning_rate": 6.849527133144796e-08, | |
| "loss": 1.3823, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.9898403483309144, | |
| "grad_norm": 0.38480904698371887, | |
| "learning_rate": 5.2443095448506674e-08, | |
| "loss": 0.8696, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.9912917271407837, | |
| "grad_norm": 0.3505820035934448, | |
| "learning_rate": 3.8530514716017544e-08, | |
| "loss": 0.3473, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.9927431059506531, | |
| "grad_norm": 0.37889888882637024, | |
| "learning_rate": 2.6757826963053066e-08, | |
| "loss": 0.3755, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.9941944847605225, | |
| "grad_norm": 0.3586353063583374, | |
| "learning_rate": 1.712528420966253e-08, | |
| "loss": 1.2269, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.9956458635703919, | |
| "grad_norm": 0.2834562361240387, | |
| "learning_rate": 9.633092661376352e-09, | |
| "loss": 0.1832, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.9970972423802612, | |
| "grad_norm": 0.3964359760284424, | |
| "learning_rate": 4.281412704887356e-09, | |
| "loss": 1.9892, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.9985486211901307, | |
| "grad_norm": 0.543126106262207, | |
| "learning_rate": 1.070358904564639e-09, | |
| "loss": 0.7731, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.22489748895168304, | |
| "learning_rate": 0.0, | |
| "loss": 0.4556, | |
| "step": 689 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 689, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 173, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.357516200247296e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |