| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 4.989190088142358, |
| "eval_steps": 10000, |
| "global_step": 60000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.04157658406785299, |
| "grad_norm": 0.8059402108192444, |
| "learning_rate": 0.00025, |
| "loss": 7.9387, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.08315316813570597, |
| "grad_norm": 0.18919387459754944, |
| "learning_rate": 0.0005, |
| "loss": 7.951, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.12472975220355896, |
| "grad_norm": 0.4990023076534271, |
| "learning_rate": 0.000495772027735498, |
| "loss": 7.9528, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.16630633627141195, |
| "grad_norm": 0.5437057018280029, |
| "learning_rate": 0.0004915440554709961, |
| "loss": 7.9422, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.20788292033926492, |
| "grad_norm": 0.35262131690979004, |
| "learning_rate": 0.00048731608320649417, |
| "loss": 7.9496, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.24945950440711792, |
| "grad_norm": 1.5567623376846313, |
| "learning_rate": 0.00048308811094199225, |
| "loss": 7.9473, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.2910360884749709, |
| "grad_norm": 1.070528507232666, |
| "learning_rate": 0.0004788601386774903, |
| "loss": 7.9368, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3326126725428239, |
| "grad_norm": 0.2238803505897522, |
| "learning_rate": 0.0004746321664129883, |
| "loss": 7.9446, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.37418925661067687, |
| "grad_norm": 0.6900185346603394, |
| "learning_rate": 0.0004704041941484864, |
| "loss": 7.9435, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.41576584067852984, |
| "grad_norm": 0.4404136538505554, |
| "learning_rate": 0.00046617622188398444, |
| "loss": 7.9488, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.4573424247463828, |
| "grad_norm": 0.3816683888435364, |
| "learning_rate": 0.00046195670556401153, |
| "loss": 7.9463, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.49891900881423584, |
| "grad_norm": 0.41241884231567383, |
| "learning_rate": 0.00045772873329950957, |
| "loss": 7.9452, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.5404955928820888, |
| "grad_norm": 0.8631793260574341, |
| "learning_rate": 0.00045350076103500765, |
| "loss": 7.94, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.5820721769499418, |
| "grad_norm": 0.44780728220939636, |
| "learning_rate": 0.00044927278877050563, |
| "loss": 7.9472, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.6236487610177948, |
| "grad_norm": 0.2136773020029068, |
| "learning_rate": 0.0004450617283950618, |
| "loss": 7.9409, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.6652253450856478, |
| "grad_norm": 0.2820763885974884, |
| "learning_rate": 0.0004408337561305598, |
| "loss": 7.9483, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.7068019291535007, |
| "grad_norm": 0.694266140460968, |
| "learning_rate": 0.00043660578386605785, |
| "loss": 7.9427, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.7483785132213537, |
| "grad_norm": 1.1062744855880737, |
| "learning_rate": 0.00043237781160155594, |
| "loss": 7.9448, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.7899550972892068, |
| "grad_norm": 0.6461020708084106, |
| "learning_rate": 0.00042814983933705397, |
| "loss": 7.9392, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.8315316813570597, |
| "grad_norm": 0.7867754101753235, |
| "learning_rate": 0.00042392186707255206, |
| "loss": 7.9435, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.8315316813570597, |
| "eval_loss": 9.379227638244629, |
| "eval_runtime": 2903.9955, |
| "eval_samples_per_second": 132.515, |
| "eval_steps_per_second": 4.141, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.8731082654249127, |
| "grad_norm": 0.5909550786018372, |
| "learning_rate": 0.00041970235075257904, |
| "loss": 7.948, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.9146848494927656, |
| "grad_norm": 0.24700529873371124, |
| "learning_rate": 0.00041547437848807713, |
| "loss": 7.9407, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.9562614335606187, |
| "grad_norm": 0.3503686487674713, |
| "learning_rate": 0.00041124640622357516, |
| "loss": 7.9396, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.9978380176284717, |
| "grad_norm": 0.4738515019416809, |
| "learning_rate": 0.00040701843395907325, |
| "loss": 7.9312, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.0394146016963246, |
| "grad_norm": 0.32388758659362793, |
| "learning_rate": 0.0004027904616945713, |
| "loss": 7.9352, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.0809911857641776, |
| "grad_norm": 1.2330721616744995, |
| "learning_rate": 0.0003985624894300693, |
| "loss": 7.9493, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.1225677698320307, |
| "grad_norm": 0.21379995346069336, |
| "learning_rate": 0.0003943345171655674, |
| "loss": 7.9334, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.1641443538998835, |
| "grad_norm": 0.3356311023235321, |
| "learning_rate": 0.00039010654490106544, |
| "loss": 7.9524, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.2057209379677365, |
| "grad_norm": 0.69386225938797, |
| "learning_rate": 0.00038587857263656353, |
| "loss": 7.9395, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.2472975220355895, |
| "grad_norm": 0.3345106542110443, |
| "learning_rate": 0.00038165060037206156, |
| "loss": 7.9449, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.2888741061034426, |
| "grad_norm": 0.8065163493156433, |
| "learning_rate": 0.00037743108405208865, |
| "loss": 7.945, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.3304506901712956, |
| "grad_norm": 0.5001415014266968, |
| "learning_rate": 0.0003732115677321157, |
| "loss": 7.9435, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.3720272742391484, |
| "grad_norm": 1.0568392276763916, |
| "learning_rate": 0.0003689835954676137, |
| "loss": 7.9435, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.4136038583070014, |
| "grad_norm": 0.36474671959877014, |
| "learning_rate": 0.0003647556232031118, |
| "loss": 7.9396, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.4551804423748544, |
| "grad_norm": 0.3686589300632477, |
| "learning_rate": 0.00036053610688313885, |
| "loss": 7.9323, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.4967570264427075, |
| "grad_norm": 0.8043748736381531, |
| "learning_rate": 0.0003563165905631659, |
| "loss": 7.9493, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.5383336105105605, |
| "grad_norm": 0.17725110054016113, |
| "learning_rate": 0.000352097074243193, |
| "loss": 7.9321, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.5799101945784133, |
| "grad_norm": 0.6144613027572632, |
| "learning_rate": 0.00034786910197869106, |
| "loss": 7.9382, |
| "step": 19000 |
| }, |
| { |
| "epoch": 1.6214867786462666, |
| "grad_norm": 0.3076309859752655, |
| "learning_rate": 0.0003436411297141891, |
| "loss": 7.9424, |
| "step": 19500 |
| }, |
| { |
| "epoch": 1.6630633627141194, |
| "grad_norm": 0.45199695229530334, |
| "learning_rate": 0.00033941315744968713, |
| "loss": 7.9344, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.6630633627141194, |
| "eval_loss": 9.162654876708984, |
| "eval_runtime": 2896.236, |
| "eval_samples_per_second": 132.87, |
| "eval_steps_per_second": 4.152, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.7046399467819724, |
| "grad_norm": 0.6658689975738525, |
| "learning_rate": 0.0003351851851851852, |
| "loss": 7.9352, |
| "step": 20500 |
| }, |
| { |
| "epoch": 1.7462165308498254, |
| "grad_norm": 0.5613082051277161, |
| "learning_rate": 0.00033095721292068325, |
| "loss": 7.9292, |
| "step": 21000 |
| }, |
| { |
| "epoch": 1.7877931149176782, |
| "grad_norm": 0.38716328144073486, |
| "learning_rate": 0.00032672924065618134, |
| "loss": 7.9327, |
| "step": 21500 |
| }, |
| { |
| "epoch": 1.8293696989855315, |
| "grad_norm": 0.7330873608589172, |
| "learning_rate": 0.00032250126839167937, |
| "loss": 7.928, |
| "step": 22000 |
| }, |
| { |
| "epoch": 1.8709462830533843, |
| "grad_norm": 0.28898411989212036, |
| "learning_rate": 0.0003182986639607644, |
| "loss": 7.9521, |
| "step": 22500 |
| }, |
| { |
| "epoch": 1.9125228671212373, |
| "grad_norm": 0.7989892959594727, |
| "learning_rate": 0.00031407069169626245, |
| "loss": 7.9357, |
| "step": 23000 |
| }, |
| { |
| "epoch": 1.9540994511890903, |
| "grad_norm": 2.788367509841919, |
| "learning_rate": 0.00030985117537628954, |
| "loss": 7.9401, |
| "step": 23500 |
| }, |
| { |
| "epoch": 1.9956760352569431, |
| "grad_norm": 0.5419213175773621, |
| "learning_rate": 0.00030562320311178757, |
| "loss": 7.9396, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.0372526193247964, |
| "grad_norm": 0.6582406759262085, |
| "learning_rate": 0.00030139523084728566, |
| "loss": 7.9346, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.078829203392649, |
| "grad_norm": 0.5304723381996155, |
| "learning_rate": 0.0002971672585827837, |
| "loss": 7.9344, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.1204057874605025, |
| "grad_norm": 0.5516705513000488, |
| "learning_rate": 0.0002929392863182818, |
| "loss": 7.9274, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.1619823715283553, |
| "grad_norm": 1.1628555059432983, |
| "learning_rate": 0.00028871976999830876, |
| "loss": 7.9465, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.203558955596208, |
| "grad_norm": 0.3907375931739807, |
| "learning_rate": 0.00028449179773380685, |
| "loss": 7.9391, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.2451355396640613, |
| "grad_norm": 1.1125935316085815, |
| "learning_rate": 0.0002802638254693049, |
| "loss": 7.9292, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.286712123731914, |
| "grad_norm": 0.8965011835098267, |
| "learning_rate": 0.000276035853204803, |
| "loss": 7.9489, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.328288707799767, |
| "grad_norm": 0.9403777122497559, |
| "learning_rate": 0.000271807880940301, |
| "loss": 7.9436, |
| "step": 28000 |
| }, |
| { |
| "epoch": 2.36986529186762, |
| "grad_norm": 0.8826812505722046, |
| "learning_rate": 0.00026757990867579904, |
| "loss": 7.9383, |
| "step": 28500 |
| }, |
| { |
| "epoch": 2.411441875935473, |
| "grad_norm": 1.86482572555542, |
| "learning_rate": 0.00026335193641129713, |
| "loss": 7.9397, |
| "step": 29000 |
| }, |
| { |
| "epoch": 2.4530184600033262, |
| "grad_norm": 0.2199900895357132, |
| "learning_rate": 0.00025912396414679516, |
| "loss": 7.9301, |
| "step": 29500 |
| }, |
| { |
| "epoch": 2.494595044071179, |
| "grad_norm": 4.152510166168213, |
| "learning_rate": 0.00025489599188229325, |
| "loss": 7.9276, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.494595044071179, |
| "eval_loss": 9.38607406616211, |
| "eval_runtime": 19770.2885, |
| "eval_samples_per_second": 19.465, |
| "eval_steps_per_second": 0.608, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.536171628139032, |
| "grad_norm": 1.1997425556182861, |
| "learning_rate": 0.0002506680196177913, |
| "loss": 7.9276, |
| "step": 30500 |
| }, |
| { |
| "epoch": 2.577748212206885, |
| "grad_norm": 1.2270859479904175, |
| "learning_rate": 0.00024644004735328937, |
| "loss": 7.924, |
| "step": 31000 |
| }, |
| { |
| "epoch": 2.619324796274738, |
| "grad_norm": 0.8749690651893616, |
| "learning_rate": 0.00024221207508878743, |
| "loss": 7.9268, |
| "step": 31500 |
| }, |
| { |
| "epoch": 2.660901380342591, |
| "grad_norm": 1.2462141513824463, |
| "learning_rate": 0.00023799255876881447, |
| "loss": 7.9462, |
| "step": 32000 |
| }, |
| { |
| "epoch": 2.702477964410444, |
| "grad_norm": 0.6699239015579224, |
| "learning_rate": 0.00023376458650431253, |
| "loss": 7.9261, |
| "step": 32500 |
| }, |
| { |
| "epoch": 2.7440545484782968, |
| "grad_norm": 0.8309171795845032, |
| "learning_rate": 0.0002295450701843396, |
| "loss": 7.9316, |
| "step": 33000 |
| }, |
| { |
| "epoch": 2.78563113254615, |
| "grad_norm": 1.003183364868164, |
| "learning_rate": 0.00022531709791983765, |
| "loss": 7.9262, |
| "step": 33500 |
| }, |
| { |
| "epoch": 2.827207716614003, |
| "grad_norm": 0.7003466486930847, |
| "learning_rate": 0.00022108912565533572, |
| "loss": 7.9384, |
| "step": 34000 |
| }, |
| { |
| "epoch": 2.868784300681856, |
| "grad_norm": 0.6416522264480591, |
| "learning_rate": 0.00021686115339083375, |
| "loss": 7.9367, |
| "step": 34500 |
| }, |
| { |
| "epoch": 2.910360884749709, |
| "grad_norm": 0.7745890021324158, |
| "learning_rate": 0.0002126331811263318, |
| "loss": 7.9493, |
| "step": 35000 |
| }, |
| { |
| "epoch": 2.9519374688175617, |
| "grad_norm": 2.5929224491119385, |
| "learning_rate": 0.00020840520886182987, |
| "loss": 7.939, |
| "step": 35500 |
| }, |
| { |
| "epoch": 2.993514052885415, |
| "grad_norm": 0.6639001965522766, |
| "learning_rate": 0.00020417723659732793, |
| "loss": 7.932, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.0350906369532678, |
| "grad_norm": 4.156096935272217, |
| "learning_rate": 0.000199949264332826, |
| "loss": 7.9312, |
| "step": 36500 |
| }, |
| { |
| "epoch": 3.076667221021121, |
| "grad_norm": 0.6016917824745178, |
| "learning_rate": 0.00019572129206832403, |
| "loss": 7.9321, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.118243805088974, |
| "grad_norm": 0.6136893630027771, |
| "learning_rate": 0.0001914933198038221, |
| "loss": 7.934, |
| "step": 37500 |
| }, |
| { |
| "epoch": 3.159820389156827, |
| "grad_norm": 1.5245190858840942, |
| "learning_rate": 0.00018726534753932015, |
| "loss": 7.9371, |
| "step": 38000 |
| }, |
| { |
| "epoch": 3.20139697322468, |
| "grad_norm": 1.779859185218811, |
| "learning_rate": 0.0001830373752748182, |
| "loss": 7.9355, |
| "step": 38500 |
| }, |
| { |
| "epoch": 3.2429735572925327, |
| "grad_norm": 0.7921754717826843, |
| "learning_rate": 0.00017880940301031627, |
| "loss": 7.9297, |
| "step": 39000 |
| }, |
| { |
| "epoch": 3.284550141360386, |
| "grad_norm": 0.40482184290885925, |
| "learning_rate": 0.0001745814307458143, |
| "loss": 7.943, |
| "step": 39500 |
| }, |
| { |
| "epoch": 3.3261267254282387, |
| "grad_norm": 0.950188398361206, |
| "learning_rate": 0.00017036191442584137, |
| "loss": 7.9307, |
| "step": 40000 |
| }, |
| { |
| "epoch": 3.3261267254282387, |
| "eval_loss": 8.992598533630371, |
| "eval_runtime": 19781.8489, |
| "eval_samples_per_second": 19.453, |
| "eval_steps_per_second": 0.608, |
| "step": 40000 |
| }, |
| { |
| "epoch": 3.367703309496092, |
| "grad_norm": 3.409175157546997, |
| "learning_rate": 0.00016614239810586843, |
| "loss": 7.9407, |
| "step": 40500 |
| }, |
| { |
| "epoch": 3.409279893563945, |
| "grad_norm": 0.41312387585639954, |
| "learning_rate": 0.0001619228817858955, |
| "loss": 7.9386, |
| "step": 41000 |
| }, |
| { |
| "epoch": 3.4508564776317976, |
| "grad_norm": 0.1976231038570404, |
| "learning_rate": 0.00015769490952139353, |
| "loss": 7.9346, |
| "step": 41500 |
| }, |
| { |
| "epoch": 3.492433061699651, |
| "grad_norm": 0.5483559966087341, |
| "learning_rate": 0.0001534669372568916, |
| "loss": 7.9234, |
| "step": 42000 |
| }, |
| { |
| "epoch": 3.5340096457675036, |
| "grad_norm": 0.7013275027275085, |
| "learning_rate": 0.00014923896499238965, |
| "loss": 7.9295, |
| "step": 42500 |
| }, |
| { |
| "epoch": 3.5755862298353565, |
| "grad_norm": 1.0993248224258423, |
| "learning_rate": 0.0001450109927278877, |
| "loss": 7.9313, |
| "step": 43000 |
| }, |
| { |
| "epoch": 3.6171628139032097, |
| "grad_norm": 0.9102666974067688, |
| "learning_rate": 0.00014079147640791475, |
| "loss": 7.9169, |
| "step": 43500 |
| }, |
| { |
| "epoch": 3.658739397971063, |
| "grad_norm": 0.3310043215751648, |
| "learning_rate": 0.0001365635041434128, |
| "loss": 7.9331, |
| "step": 44000 |
| }, |
| { |
| "epoch": 3.7003159820389158, |
| "grad_norm": 1.6578195095062256, |
| "learning_rate": 0.00013233553187891087, |
| "loss": 7.9408, |
| "step": 44500 |
| }, |
| { |
| "epoch": 3.7418925661067686, |
| "grad_norm": 1.0133479833602905, |
| "learning_rate": 0.00012810755961440893, |
| "loss": 7.9278, |
| "step": 45000 |
| }, |
| { |
| "epoch": 3.7834691501746214, |
| "grad_norm": 1.0534777641296387, |
| "learning_rate": 0.000123879587349907, |
| "loss": 7.9222, |
| "step": 45500 |
| }, |
| { |
| "epoch": 3.8250457342424746, |
| "grad_norm": 1.0995186567306519, |
| "learning_rate": 0.00011965161508540504, |
| "loss": 7.9222, |
| "step": 46000 |
| }, |
| { |
| "epoch": 3.866622318310328, |
| "grad_norm": 1.2467023134231567, |
| "learning_rate": 0.00011542364282090308, |
| "loss": 7.9374, |
| "step": 46500 |
| }, |
| { |
| "epoch": 3.9081989023781807, |
| "grad_norm": 0.6253928542137146, |
| "learning_rate": 0.00011119567055640114, |
| "loss": 7.9241, |
| "step": 47000 |
| }, |
| { |
| "epoch": 3.9497754864460335, |
| "grad_norm": 0.3281259536743164, |
| "learning_rate": 0.0001069676982918992, |
| "loss": 7.9298, |
| "step": 47500 |
| }, |
| { |
| "epoch": 3.9913520705138867, |
| "grad_norm": 0.6171555519104004, |
| "learning_rate": 0.00010273972602739725, |
| "loss": 7.9352, |
| "step": 48000 |
| }, |
| { |
| "epoch": 4.03292865458174, |
| "grad_norm": 0.41333335638046265, |
| "learning_rate": 9.851175376289531e-05, |
| "loss": 7.9356, |
| "step": 48500 |
| }, |
| { |
| "epoch": 4.074505238649593, |
| "grad_norm": 0.2815527021884918, |
| "learning_rate": 9.428378149839336e-05, |
| "loss": 7.9241, |
| "step": 49000 |
| }, |
| { |
| "epoch": 4.116081822717446, |
| "grad_norm": 0.8225882649421692, |
| "learning_rate": 9.005580923389142e-05, |
| "loss": 7.9214, |
| "step": 49500 |
| }, |
| { |
| "epoch": 4.157658406785298, |
| "grad_norm": 0.44963425397872925, |
| "learning_rate": 8.582783696938948e-05, |
| "loss": 7.9237, |
| "step": 50000 |
| }, |
| { |
| "epoch": 4.157658406785298, |
| "eval_loss": 8.952984809875488, |
| "eval_runtime": 2779.282, |
| "eval_samples_per_second": 138.462, |
| "eval_steps_per_second": 4.327, |
| "step": 50000 |
| }, |
| { |
| "epoch": 4.199234990853151, |
| "grad_norm": 0.5252857208251953, |
| "learning_rate": 8.159986470488753e-05, |
| "loss": 7.9331, |
| "step": 50500 |
| }, |
| { |
| "epoch": 4.240811574921005, |
| "grad_norm": 0.4898480176925659, |
| "learning_rate": 7.737189244038559e-05, |
| "loss": 7.9296, |
| "step": 51000 |
| }, |
| { |
| "epoch": 4.282388158988858, |
| "grad_norm": 0.597152054309845, |
| "learning_rate": 7.314392017588364e-05, |
| "loss": 7.9371, |
| "step": 51500 |
| }, |
| { |
| "epoch": 4.3239647430567105, |
| "grad_norm": 0.3579633831977844, |
| "learning_rate": 6.89159479113817e-05, |
| "loss": 7.92, |
| "step": 52000 |
| }, |
| { |
| "epoch": 4.365541327124563, |
| "grad_norm": 0.8714535236358643, |
| "learning_rate": 6.468797564687976e-05, |
| "loss": 7.9259, |
| "step": 52500 |
| }, |
| { |
| "epoch": 4.407117911192416, |
| "grad_norm": 0.4833739101886749, |
| "learning_rate": 6.046000338237781e-05, |
| "loss": 7.93, |
| "step": 53000 |
| }, |
| { |
| "epoch": 4.44869449526027, |
| "grad_norm": 0.7422254681587219, |
| "learning_rate": 5.623203111787587e-05, |
| "loss": 7.9443, |
| "step": 53500 |
| }, |
| { |
| "epoch": 4.490271079328123, |
| "grad_norm": 0.3269445598125458, |
| "learning_rate": 5.201251479790293e-05, |
| "loss": 7.9211, |
| "step": 54000 |
| }, |
| { |
| "epoch": 4.5318476633959754, |
| "grad_norm": 0.3518795073032379, |
| "learning_rate": 4.779299847792998e-05, |
| "loss": 7.9327, |
| "step": 54500 |
| }, |
| { |
| "epoch": 4.573424247463828, |
| "grad_norm": 0.29840198159217834, |
| "learning_rate": 4.356502621342804e-05, |
| "loss": 7.9341, |
| "step": 55000 |
| }, |
| { |
| "epoch": 4.615000831531681, |
| "grad_norm": 0.6424156427383423, |
| "learning_rate": 3.9337053948926096e-05, |
| "loss": 7.943, |
| "step": 55500 |
| }, |
| { |
| "epoch": 4.656577415599534, |
| "grad_norm": 0.2546742260456085, |
| "learning_rate": 3.511753762895315e-05, |
| "loss": 7.9313, |
| "step": 56000 |
| }, |
| { |
| "epoch": 4.698153999667388, |
| "grad_norm": 0.5346771478652954, |
| "learning_rate": 3.0889565364451214e-05, |
| "loss": 7.9275, |
| "step": 56500 |
| }, |
| { |
| "epoch": 4.73973058373524, |
| "grad_norm": 0.7033748030662537, |
| "learning_rate": 2.6661593099949264e-05, |
| "loss": 7.9358, |
| "step": 57000 |
| }, |
| { |
| "epoch": 4.781307167803093, |
| "grad_norm": 0.28617945313453674, |
| "learning_rate": 2.243362083544732e-05, |
| "loss": 7.9299, |
| "step": 57500 |
| }, |
| { |
| "epoch": 4.822883751870946, |
| "grad_norm": 0.5841323733329773, |
| "learning_rate": 1.8205648570945376e-05, |
| "loss": 7.9234, |
| "step": 58000 |
| }, |
| { |
| "epoch": 4.8644603359388, |
| "grad_norm": 0.4925735890865326, |
| "learning_rate": 1.397767630644343e-05, |
| "loss": 7.9181, |
| "step": 58500 |
| }, |
| { |
| "epoch": 4.9060369200066525, |
| "grad_norm": 0.7338876128196716, |
| "learning_rate": 9.749704041941485e-06, |
| "loss": 7.9225, |
| "step": 59000 |
| }, |
| { |
| "epoch": 4.947613504074505, |
| "grad_norm": 0.4279155731201172, |
| "learning_rate": 5.52173177743954e-06, |
| "loss": 7.9266, |
| "step": 59500 |
| }, |
| { |
| "epoch": 4.989190088142358, |
| "grad_norm": 0.9879505038261414, |
| "learning_rate": 1.2937595129375953e-06, |
| "loss": 7.9321, |
| "step": 60000 |
| }, |
| { |
| "epoch": 4.989190088142358, |
| "eval_loss": 8.944025993347168, |
| "eval_runtime": 2914.4582, |
| "eval_samples_per_second": 132.04, |
| "eval_steps_per_second": 4.126, |
| "step": 60000 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 60130, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 6.726369615519345e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
