| { |
| "best_global_step": 5500, |
| "best_metric": 1.8417972326278687, |
| "best_model_checkpoint": "/kaggle/working/checkpoints/checkpoint-5500", |
| "epoch": 6.0, |
| "eval_steps": 500, |
| "global_step": 5664, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0010601643254704478, |
| "grad_norm": 0.8187403082847595, |
| "learning_rate": 0.0, |
| "loss": 10.509103775024414, |
| "num_input_tokens_seen": 4096, |
| "step": 1, |
| "train_runtime": 4.0328, |
| "train_tokens_per_second": 1015.678 |
| }, |
| { |
| "epoch": 0.053008216273522396, |
| "grad_norm": 0.7271436452865601, |
| "learning_rate": 8.64705882352941e-05, |
| "loss": 9.653437400350766, |
| "num_input_tokens_seen": 204800, |
| "step": 50, |
| "train_runtime": 85.0317, |
| "train_tokens_per_second": 2408.514 |
| }, |
| { |
| "epoch": 0.10601643254704479, |
| "grad_norm": 0.6869716644287109, |
| "learning_rate": 0.00017470588235294116, |
| "loss": 9.471989135742188, |
| "num_input_tokens_seen": 409600, |
| "step": 100, |
| "train_runtime": 176.2658, |
| "train_tokens_per_second": 2323.764 |
| }, |
| { |
| "epoch": 0.1590246488205672, |
| "grad_norm": 0.6832960247993469, |
| "learning_rate": 0.0002629411764705882, |
| "loss": 9.664740600585937, |
| "num_input_tokens_seen": 614400, |
| "step": 150, |
| "train_runtime": 267.3416, |
| "train_tokens_per_second": 2298.183 |
| }, |
| { |
| "epoch": 0.21203286509408958, |
| "grad_norm": 0.6674580574035645, |
| "learning_rate": 0.00029841645431379684, |
| "loss": 9.3196337890625, |
| "num_input_tokens_seen": 819200, |
| "step": 200, |
| "train_runtime": 358.4777, |
| "train_tokens_per_second": 2285.219 |
| }, |
| { |
| "epoch": 0.26504108136761195, |
| "grad_norm": 0.6817887425422668, |
| "learning_rate": 0.00029568620313068797, |
| "loss": 9.44159423828125, |
| "num_input_tokens_seen": 1024000, |
| "step": 250, |
| "train_runtime": 449.5642, |
| "train_tokens_per_second": 2277.761 |
| }, |
| { |
| "epoch": 0.3180492976411344, |
| "grad_norm": 0.6933718919754028, |
| "learning_rate": 0.00029295595194757916, |
| "loss": 9.353170776367188, |
| "num_input_tokens_seen": 1228800, |
| "step": 300, |
| "train_runtime": 540.6374, |
| "train_tokens_per_second": 2272.873 |
| }, |
| { |
| "epoch": 0.37105751391465674, |
| "grad_norm": 0.6354528069496155, |
| "learning_rate": 0.0002902257007644703, |
| "loss": 9.46597900390625, |
| "num_input_tokens_seen": 1433600, |
| "step": 350, |
| "train_runtime": 631.7827, |
| "train_tokens_per_second": 2269.135 |
| }, |
| { |
| "epoch": 0.42406573018817917, |
| "grad_norm": 0.6813943982124329, |
| "learning_rate": 0.0002874954495813615, |
| "loss": 9.3445703125, |
| "num_input_tokens_seen": 1638400, |
| "step": 400, |
| "train_runtime": 722.6411, |
| "train_tokens_per_second": 2267.239 |
| }, |
| { |
| "epoch": 0.47707394646170154, |
| "grad_norm": 0.5886964797973633, |
| "learning_rate": 0.0002847651983982526, |
| "loss": 9.338291625976563, |
| "num_input_tokens_seen": 1843200, |
| "step": 450, |
| "train_runtime": 813.7385, |
| "train_tokens_per_second": 2265.101 |
| }, |
| { |
| "epoch": 0.5300821627352239, |
| "grad_norm": 0.6884158849716187, |
| "learning_rate": 0.0002820349472151438, |
| "loss": 9.193983764648438, |
| "num_input_tokens_seen": 2048000, |
| "step": 500, |
| "train_runtime": 904.7321, |
| "train_tokens_per_second": 2263.653 |
| }, |
| { |
| "epoch": 0.5300821627352239, |
| "eval_bleu": 4.06, |
| "eval_chrf": 19.66, |
| "eval_loss": 1.8984475135803223, |
| "eval_runtime": 87.1763, |
| "eval_samples_per_second": 3.499, |
| "eval_steps_per_second": 1.755, |
| "num_input_tokens_seen": 2048000, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5830903790087464, |
| "grad_norm": 0.6719595193862915, |
| "learning_rate": 0.0002793046960320349, |
| "loss": 9.386028442382813, |
| "num_input_tokens_seen": 2252800, |
| "step": 550, |
| "train_runtime": 1094.9482, |
| "train_tokens_per_second": 2057.449 |
| }, |
| { |
| "epoch": 0.6360985952822688, |
| "grad_norm": 0.6773673295974731, |
| "learning_rate": 0.00027657444484892606, |
| "loss": 9.26089599609375, |
| "num_input_tokens_seen": 2457600, |
| "step": 600, |
| "train_runtime": 1186.45, |
| "train_tokens_per_second": 2071.389 |
| }, |
| { |
| "epoch": 0.6891068115557911, |
| "grad_norm": 0.6533596515655518, |
| "learning_rate": 0.00027384419366581725, |
| "loss": 9.2549658203125, |
| "num_input_tokens_seen": 2662400, |
| "step": 650, |
| "train_runtime": 1277.7875, |
| "train_tokens_per_second": 2083.602 |
| }, |
| { |
| "epoch": 0.7421150278293135, |
| "grad_norm": 0.6407659649848938, |
| "learning_rate": 0.0002711139424827084, |
| "loss": 9.27232666015625, |
| "num_input_tokens_seen": 2867200, |
| "step": 700, |
| "train_runtime": 1369.0819, |
| "train_tokens_per_second": 2094.25 |
| }, |
| { |
| "epoch": 0.795123244102836, |
| "grad_norm": 0.7818734645843506, |
| "learning_rate": 0.00026838369129959956, |
| "loss": 9.223333740234375, |
| "num_input_tokens_seen": 3072000, |
| "step": 750, |
| "train_runtime": 1460.1534, |
| "train_tokens_per_second": 2103.889 |
| }, |
| { |
| "epoch": 0.8481314603763583, |
| "grad_norm": 0.6569114923477173, |
| "learning_rate": 0.0002656534401164907, |
| "loss": 9.193381958007812, |
| "num_input_tokens_seen": 3276800, |
| "step": 800, |
| "train_runtime": 1551.22, |
| "train_tokens_per_second": 2112.402 |
| }, |
| { |
| "epoch": 0.9011396766498807, |
| "grad_norm": 0.7116767168045044, |
| "learning_rate": 0.0002629231889333819, |
| "loss": 9.336693115234375, |
| "num_input_tokens_seen": 3481600, |
| "step": 850, |
| "train_runtime": 1642.3679, |
| "train_tokens_per_second": 2119.866 |
| }, |
| { |
| "epoch": 0.9541478929234031, |
| "grad_norm": 0.64178466796875, |
| "learning_rate": 0.000260192937750273, |
| "loss": 9.349353637695312, |
| "num_input_tokens_seen": 3686400, |
| "step": 900, |
| "train_runtime": 1733.6186, |
| "train_tokens_per_second": 2126.419 |
| }, |
| { |
| "epoch": 1.0063609859528226, |
| "grad_norm": 0.6793839931488037, |
| "learning_rate": 0.00025746268656716415, |
| "loss": 9.190228271484376, |
| "num_input_tokens_seen": 3887616, |
| "step": 950, |
| "train_runtime": 1823.6521, |
| "train_tokens_per_second": 2131.775 |
| }, |
| { |
| "epoch": 1.059369202226345, |
| "grad_norm": 0.6120585799217224, |
| "learning_rate": 0.00025473243538405533, |
| "loss": 8.995220336914063, |
| "num_input_tokens_seen": 4092416, |
| "step": 1000, |
| "train_runtime": 1914.9217, |
| "train_tokens_per_second": 2137.119 |
| }, |
| { |
| "epoch": 1.059369202226345, |
| "eval_bleu": 4.04, |
| "eval_chrf": 19.54, |
| "eval_loss": 1.8891278505325317, |
| "eval_runtime": 84.89, |
| "eval_samples_per_second": 3.593, |
| "eval_steps_per_second": 1.802, |
| "num_input_tokens_seen": 4092416, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.1123774184998674, |
| "grad_norm": 0.6622714996337891, |
| "learning_rate": 0.00025200218420094646, |
| "loss": 9.169678955078124, |
| "num_input_tokens_seen": 4297216, |
| "step": 1050, |
| "train_runtime": 2104.4734, |
| "train_tokens_per_second": 2041.944 |
| }, |
| { |
| "epoch": 1.16538563477339, |
| "grad_norm": 0.6801309585571289, |
| "learning_rate": 0.00024927193301783765, |
| "loss": 9.108585205078125, |
| "num_input_tokens_seen": 4502016, |
| "step": 1100, |
| "train_runtime": 2195.9257, |
| "train_tokens_per_second": 2050.168 |
| }, |
| { |
| "epoch": 1.2183938510469123, |
| "grad_norm": 0.6843289732933044, |
| "learning_rate": 0.0002465416818347288, |
| "loss": 9.04650146484375, |
| "num_input_tokens_seen": 4706816, |
| "step": 1150, |
| "train_runtime": 2287.3334, |
| "train_tokens_per_second": 2057.774 |
| }, |
| { |
| "epoch": 1.2714020673204347, |
| "grad_norm": 0.6618881821632385, |
| "learning_rate": 0.00024381143065161994, |
| "loss": 9.0776953125, |
| "num_input_tokens_seen": 4911616, |
| "step": 1200, |
| "train_runtime": 2378.5524, |
| "train_tokens_per_second": 2064.96 |
| }, |
| { |
| "epoch": 1.324410283593957, |
| "grad_norm": 0.7164718508720398, |
| "learning_rate": 0.00024108117946851107, |
| "loss": 9.070903930664063, |
| "num_input_tokens_seen": 5116416, |
| "step": 1250, |
| "train_runtime": 2470.0729, |
| "train_tokens_per_second": 2071.362 |
| }, |
| { |
| "epoch": 1.3774184998674794, |
| "grad_norm": 0.703388512134552, |
| "learning_rate": 0.00023835092828540223, |
| "loss": 9.000114135742187, |
| "num_input_tokens_seen": 5321216, |
| "step": 1300, |
| "train_runtime": 2561.2113, |
| "train_tokens_per_second": 2077.617 |
| }, |
| { |
| "epoch": 1.4304267161410018, |
| "grad_norm": 0.6863468289375305, |
| "learning_rate": 0.00023562067710229337, |
| "loss": 9.03710693359375, |
| "num_input_tokens_seen": 5526016, |
| "step": 1350, |
| "train_runtime": 2652.4931, |
| "train_tokens_per_second": 2083.329 |
| }, |
| { |
| "epoch": 1.4834349324145242, |
| "grad_norm": 0.681918740272522, |
| "learning_rate": 0.00023289042591918453, |
| "loss": 9.07092529296875, |
| "num_input_tokens_seen": 5730816, |
| "step": 1400, |
| "train_runtime": 2743.7676, |
| "train_tokens_per_second": 2088.667 |
| }, |
| { |
| "epoch": 1.5364431486880465, |
| "grad_norm": 0.6488898396492004, |
| "learning_rate": 0.00023016017473607568, |
| "loss": 9.12448974609375, |
| "num_input_tokens_seen": 5935616, |
| "step": 1450, |
| "train_runtime": 2835.0135, |
| "train_tokens_per_second": 2093.682 |
| }, |
| { |
| "epoch": 1.589451364961569, |
| "grad_norm": 0.629993736743927, |
| "learning_rate": 0.00022742992355296684, |
| "loss": 8.988504638671875, |
| "num_input_tokens_seen": 6140416, |
| "step": 1500, |
| "train_runtime": 2926.1673, |
| "train_tokens_per_second": 2098.45 |
| }, |
| { |
| "epoch": 1.589451364961569, |
| "eval_bleu": 4.15, |
| "eval_chrf": 19.86, |
| "eval_loss": 1.8749940395355225, |
| "eval_runtime": 85.3123, |
| "eval_samples_per_second": 3.575, |
| "eval_steps_per_second": 1.793, |
| "num_input_tokens_seen": 6140416, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.6424595812350913, |
| "grad_norm": 0.6674344539642334, |
| "learning_rate": 0.000224699672369858, |
| "loss": 9.002008056640625, |
| "num_input_tokens_seen": 6345216, |
| "step": 1550, |
| "train_runtime": 3115.9795, |
| "train_tokens_per_second": 2036.347 |
| }, |
| { |
| "epoch": 1.6954677975086139, |
| "grad_norm": 0.6869524717330933, |
| "learning_rate": 0.00022196942118674916, |
| "loss": 9.157560424804688, |
| "num_input_tokens_seen": 6550016, |
| "step": 1600, |
| "train_runtime": 3207.373, |
| "train_tokens_per_second": 2042.175 |
| }, |
| { |
| "epoch": 1.7484760137821362, |
| "grad_norm": 0.661709725856781, |
| "learning_rate": 0.0002192391700036403, |
| "loss": 9.031039428710937, |
| "num_input_tokens_seen": 6754816, |
| "step": 1650, |
| "train_runtime": 3298.7999, |
| "train_tokens_per_second": 2047.659 |
| }, |
| { |
| "epoch": 1.8014842300556586, |
| "grad_norm": 0.6780179738998413, |
| "learning_rate": 0.00021650891882053145, |
| "loss": 9.039593505859376, |
| "num_input_tokens_seen": 6959616, |
| "step": 1700, |
| "train_runtime": 3390.0711, |
| "train_tokens_per_second": 2052.941 |
| }, |
| { |
| "epoch": 1.854492446329181, |
| "grad_norm": 0.700842022895813, |
| "learning_rate": 0.0002137786676374226, |
| "loss": 8.978927612304688, |
| "num_input_tokens_seen": 7164416, |
| "step": 1750, |
| "train_runtime": 3481.1155, |
| "train_tokens_per_second": 2058.081 |
| }, |
| { |
| "epoch": 1.9075006626027036, |
| "grad_norm": 0.6786106824874878, |
| "learning_rate": 0.00021104841645431377, |
| "loss": 9.08431396484375, |
| "num_input_tokens_seen": 7369216, |
| "step": 1800, |
| "train_runtime": 3572.2502, |
| "train_tokens_per_second": 2062.906 |
| }, |
| { |
| "epoch": 1.960508878876226, |
| "grad_norm": 0.6586954593658447, |
| "learning_rate": 0.00020831816527120493, |
| "loss": 9.029188232421875, |
| "num_input_tokens_seen": 7574016, |
| "step": 1850, |
| "train_runtime": 3663.5445, |
| "train_tokens_per_second": 2067.401 |
| }, |
| { |
| "epoch": 2.0127219719056453, |
| "grad_norm": 0.7223443388938904, |
| "learning_rate": 0.0002055879140880961, |
| "loss": 8.716483764648437, |
| "num_input_tokens_seen": 7775232, |
| "step": 1900, |
| "train_runtime": 3753.5921, |
| "train_tokens_per_second": 2071.411 |
| }, |
| { |
| "epoch": 2.0657301881791676, |
| "grad_norm": 0.6501656770706177, |
| "learning_rate": 0.00020285766290498725, |
| "loss": 8.858229370117188, |
| "num_input_tokens_seen": 7980032, |
| "step": 1950, |
| "train_runtime": 3844.9155, |
| "train_tokens_per_second": 2075.477 |
| }, |
| { |
| "epoch": 2.11873840445269, |
| "grad_norm": 0.6163517236709595, |
| "learning_rate": 0.00020012741172187838, |
| "loss": 8.852374267578124, |
| "num_input_tokens_seen": 8184832, |
| "step": 2000, |
| "train_runtime": 3936.2252, |
| "train_tokens_per_second": 2079.361 |
| }, |
| { |
| "epoch": 2.11873840445269, |
| "eval_bleu": 4.07, |
| "eval_chrf": 19.95, |
| "eval_loss": 1.8687856197357178, |
| "eval_runtime": 85.2414, |
| "eval_samples_per_second": 3.578, |
| "eval_steps_per_second": 1.795, |
| "num_input_tokens_seen": 8184832, |
| "step": 2000 |
| }, |
| { |
| "epoch": 2.1717466207262124, |
| "grad_norm": 0.6511809825897217, |
| "learning_rate": 0.00019739716053876954, |
| "loss": 8.9257958984375, |
| "num_input_tokens_seen": 8389632, |
| "step": 2050, |
| "train_runtime": 4126.538, |
| "train_tokens_per_second": 2033.092 |
| }, |
| { |
| "epoch": 2.2247548369997348, |
| "grad_norm": 0.6239650249481201, |
| "learning_rate": 0.0001946669093556607, |
| "loss": 8.875814208984375, |
| "num_input_tokens_seen": 8594432, |
| "step": 2100, |
| "train_runtime": 4217.9271, |
| "train_tokens_per_second": 2037.596 |
| }, |
| { |
| "epoch": 2.277763053273257, |
| "grad_norm": 0.6986948847770691, |
| "learning_rate": 0.00019193665817255186, |
| "loss": 8.9848681640625, |
| "num_input_tokens_seen": 8799232, |
| "step": 2150, |
| "train_runtime": 4309.2527, |
| "train_tokens_per_second": 2041.939 |
| }, |
| { |
| "epoch": 2.33077126954678, |
| "grad_norm": 0.6011477708816528, |
| "learning_rate": 0.00018920640698944302, |
| "loss": 8.912135620117187, |
| "num_input_tokens_seen": 9004032, |
| "step": 2200, |
| "train_runtime": 4400.5936, |
| "train_tokens_per_second": 2046.095 |
| }, |
| { |
| "epoch": 2.3837794858203023, |
| "grad_norm": 0.7250493764877319, |
| "learning_rate": 0.00018647615580633418, |
| "loss": 8.793685913085938, |
| "num_input_tokens_seen": 9208832, |
| "step": 2250, |
| "train_runtime": 4492.0003, |
| "train_tokens_per_second": 2050.052 |
| }, |
| { |
| "epoch": 2.4367877020938247, |
| "grad_norm": 0.6375405788421631, |
| "learning_rate": 0.00018374590462322534, |
| "loss": 8.844403076171876, |
| "num_input_tokens_seen": 9413632, |
| "step": 2300, |
| "train_runtime": 4583.4122, |
| "train_tokens_per_second": 2053.848 |
| }, |
| { |
| "epoch": 2.489795918367347, |
| "grad_norm": 0.6628595590591431, |
| "learning_rate": 0.00018101565344011647, |
| "loss": 8.734286499023437, |
| "num_input_tokens_seen": 9618432, |
| "step": 2350, |
| "train_runtime": 4674.7698, |
| "train_tokens_per_second": 2057.52 |
| }, |
| { |
| "epoch": 2.5428041346408694, |
| "grad_norm": 0.6868587136268616, |
| "learning_rate": 0.00017828540225700763, |
| "loss": 8.970186157226562, |
| "num_input_tokens_seen": 9823232, |
| "step": 2400, |
| "train_runtime": 4765.9951, |
| "train_tokens_per_second": 2061.108 |
| }, |
| { |
| "epoch": 2.5958123509143918, |
| "grad_norm": 0.6495867371559143, |
| "learning_rate": 0.0001755551510738988, |
| "loss": 8.924597778320312, |
| "num_input_tokens_seen": 10028032, |
| "step": 2450, |
| "train_runtime": 4857.2206, |
| "train_tokens_per_second": 2064.562 |
| }, |
| { |
| "epoch": 2.648820567187914, |
| "grad_norm": 0.7620416283607483, |
| "learning_rate": 0.00017282489989078995, |
| "loss": 8.905831298828126, |
| "num_input_tokens_seen": 10232832, |
| "step": 2500, |
| "train_runtime": 4948.5715, |
| "train_tokens_per_second": 2067.836 |
| }, |
| { |
| "epoch": 2.648820567187914, |
| "eval_bleu": 4.05, |
| "eval_chrf": 19.71, |
| "eval_loss": 1.8576197624206543, |
| "eval_runtime": 85.4771, |
| "eval_samples_per_second": 3.568, |
| "eval_steps_per_second": 1.79, |
| "num_input_tokens_seen": 10232832, |
| "step": 2500 |
| }, |
| { |
| "epoch": 2.7018287834614365, |
| "grad_norm": 0.7461215853691101, |
| "learning_rate": 0.0001700946487076811, |
| "loss": 8.890812377929688, |
| "num_input_tokens_seen": 10437632, |
| "step": 2550, |
| "train_runtime": 5138.677, |
| "train_tokens_per_second": 2031.191 |
| }, |
| { |
| "epoch": 2.754836999734959, |
| "grad_norm": 0.6335775256156921, |
| "learning_rate": 0.00016736439752457227, |
| "loss": 8.868407592773437, |
| "num_input_tokens_seen": 10642432, |
| "step": 2600, |
| "train_runtime": 5230.3092, |
| "train_tokens_per_second": 2034.762 |
| }, |
| { |
| "epoch": 2.8078452160084812, |
| "grad_norm": 0.6892751455307007, |
| "learning_rate": 0.00016463414634146343, |
| "loss": 8.848526000976562, |
| "num_input_tokens_seen": 10847232, |
| "step": 2650, |
| "train_runtime": 5321.7199, |
| "train_tokens_per_second": 2038.294 |
| }, |
| { |
| "epoch": 2.8608534322820036, |
| "grad_norm": 0.7833438515663147, |
| "learning_rate": 0.00016190389515835456, |
| "loss": 8.842973022460937, |
| "num_input_tokens_seen": 11052032, |
| "step": 2700, |
| "train_runtime": 5413.0681, |
| "train_tokens_per_second": 2041.732 |
| }, |
| { |
| "epoch": 2.913861648555526, |
| "grad_norm": 0.7217938303947449, |
| "learning_rate": 0.00015917364397524572, |
| "loss": 8.9160205078125, |
| "num_input_tokens_seen": 11256832, |
| "step": 2750, |
| "train_runtime": 5504.4642, |
| "train_tokens_per_second": 2045.037 |
| }, |
| { |
| "epoch": 2.9668698648290484, |
| "grad_norm": 0.625913143157959, |
| "learning_rate": 0.00015644339279213688, |
| "loss": 8.780228271484376, |
| "num_input_tokens_seen": 11461632, |
| "step": 2800, |
| "train_runtime": 5595.8294, |
| "train_tokens_per_second": 2048.245 |
| }, |
| { |
| "epoch": 3.019082957858468, |
| "grad_norm": 0.686261773109436, |
| "learning_rate": 0.000153713141609028, |
| "loss": 8.62293212890625, |
| "num_input_tokens_seen": 11662848, |
| "step": 2850, |
| "train_runtime": 5686.1951, |
| "train_tokens_per_second": 2051.081 |
| }, |
| { |
| "epoch": 3.0720911741319905, |
| "grad_norm": 0.7150633335113525, |
| "learning_rate": 0.00015098289042591917, |
| "loss": 8.857227783203125, |
| "num_input_tokens_seen": 11867648, |
| "step": 2900, |
| "train_runtime": 5777.4806, |
| "train_tokens_per_second": 2054.122 |
| }, |
| { |
| "epoch": 3.125099390405513, |
| "grad_norm": 0.7594448328018188, |
| "learning_rate": 0.00014825263924281033, |
| "loss": 8.733943481445312, |
| "num_input_tokens_seen": 12072448, |
| "step": 2950, |
| "train_runtime": 5868.7816, |
| "train_tokens_per_second": 2057.062 |
| }, |
| { |
| "epoch": 3.1781076066790352, |
| "grad_norm": 0.6436108946800232, |
| "learning_rate": 0.00014552238805970149, |
| "loss": 8.650802612304688, |
| "num_input_tokens_seen": 12277248, |
| "step": 3000, |
| "train_runtime": 5960.2675, |
| "train_tokens_per_second": 2059.848 |
| }, |
| { |
| "epoch": 3.1781076066790352, |
| "eval_bleu": 4.2, |
| "eval_chrf": 19.81, |
| "eval_loss": 1.8580111265182495, |
| "eval_runtime": 85.4127, |
| "eval_samples_per_second": 3.571, |
| "eval_steps_per_second": 1.791, |
| "num_input_tokens_seen": 12277248, |
| "step": 3000 |
| }, |
| { |
| "epoch": 3.2311158229525576, |
| "grad_norm": 0.6634842753410339, |
| "learning_rate": 0.00014279213687659265, |
| "loss": 8.697005615234374, |
| "num_input_tokens_seen": 12482048, |
| "step": 3050, |
| "train_runtime": 6151.2346, |
| "train_tokens_per_second": 2029.194 |
| }, |
| { |
| "epoch": 3.28412403922608, |
| "grad_norm": 0.7246391773223877, |
| "learning_rate": 0.0001400618856934838, |
| "loss": 8.7168359375, |
| "num_input_tokens_seen": 12686848, |
| "step": 3100, |
| "train_runtime": 6242.7027, |
| "train_tokens_per_second": 2032.268 |
| }, |
| { |
| "epoch": 3.3371322554996024, |
| "grad_norm": 0.6836341023445129, |
| "learning_rate": 0.00013733163451037494, |
| "loss": 8.65872802734375, |
| "num_input_tokens_seen": 12891648, |
| "step": 3150, |
| "train_runtime": 6333.8948, |
| "train_tokens_per_second": 2035.343 |
| }, |
| { |
| "epoch": 3.3901404717731247, |
| "grad_norm": 0.7036194801330566, |
| "learning_rate": 0.0001346013833272661, |
| "loss": 8.667957763671875, |
| "num_input_tokens_seen": 13096448, |
| "step": 3200, |
| "train_runtime": 6425.2484, |
| "train_tokens_per_second": 2038.279 |
| }, |
| { |
| "epoch": 3.443148688046647, |
| "grad_norm": 0.7401408553123474, |
| "learning_rate": 0.00013187113214415725, |
| "loss": 8.828726806640624, |
| "num_input_tokens_seen": 13301248, |
| "step": 3250, |
| "train_runtime": 6516.5209, |
| "train_tokens_per_second": 2041.158 |
| }, |
| { |
| "epoch": 3.4961569043201695, |
| "grad_norm": 0.735714852809906, |
| "learning_rate": 0.0001291408809610484, |
| "loss": 8.744295654296875, |
| "num_input_tokens_seen": 13506048, |
| "step": 3300, |
| "train_runtime": 6607.8418, |
| "train_tokens_per_second": 2043.942 |
| }, |
| { |
| "epoch": 3.549165120593692, |
| "grad_norm": 0.7713711261749268, |
| "learning_rate": 0.00012641062977793955, |
| "loss": 8.785454711914063, |
| "num_input_tokens_seen": 13710848, |
| "step": 3350, |
| "train_runtime": 6699.3485, |
| "train_tokens_per_second": 2046.594 |
| }, |
| { |
| "epoch": 3.6021733368672146, |
| "grad_norm": 0.6046838760375977, |
| "learning_rate": 0.0001236803785948307, |
| "loss": 8.79681884765625, |
| "num_input_tokens_seen": 13915648, |
| "step": 3400, |
| "train_runtime": 6790.6621, |
| "train_tokens_per_second": 2049.233 |
| }, |
| { |
| "epoch": 3.6551815531407366, |
| "grad_norm": 0.7195361852645874, |
| "learning_rate": 0.00012095012741172186, |
| "loss": 8.713923950195312, |
| "num_input_tokens_seen": 14120448, |
| "step": 3450, |
| "train_runtime": 6881.9777, |
| "train_tokens_per_second": 2051.801 |
| }, |
| { |
| "epoch": 3.7081897694142594, |
| "grad_norm": 0.6607725620269775, |
| "learning_rate": 0.00011821987622861302, |
| "loss": 8.707265014648437, |
| "num_input_tokens_seen": 14325248, |
| "step": 3500, |
| "train_runtime": 6973.3252, |
| "train_tokens_per_second": 2054.292 |
| }, |
| { |
| "epoch": 3.7081897694142594, |
| "eval_bleu": 4.18, |
| "eval_chrf": 19.91, |
| "eval_loss": 1.8520567417144775, |
| "eval_runtime": 85.3221, |
| "eval_samples_per_second": 3.575, |
| "eval_steps_per_second": 1.793, |
| "num_input_tokens_seen": 14325248, |
| "step": 3500 |
| }, |
| { |
| "epoch": 3.7611979856877817, |
| "grad_norm": 0.7562373876571655, |
| "learning_rate": 0.00011548962504550417, |
| "loss": 8.9096435546875, |
| "num_input_tokens_seen": 14530048, |
| "step": 3550, |
| "train_runtime": 7163.8071, |
| "train_tokens_per_second": 2028.258 |
| }, |
| { |
| "epoch": 3.814206201961304, |
| "grad_norm": 0.6998844742774963, |
| "learning_rate": 0.00011275937386239533, |
| "loss": 8.541333618164062, |
| "num_input_tokens_seen": 14734848, |
| "step": 3600, |
| "train_runtime": 7255.3341, |
| "train_tokens_per_second": 2030.899 |
| }, |
| { |
| "epoch": 3.8672144182348265, |
| "grad_norm": 0.7850280404090881, |
| "learning_rate": 0.00011002912267928649, |
| "loss": 8.79550537109375, |
| "num_input_tokens_seen": 14939648, |
| "step": 3650, |
| "train_runtime": 7346.7779, |
| "train_tokens_per_second": 2033.497 |
| }, |
| { |
| "epoch": 3.920222634508349, |
| "grad_norm": 0.6771405339241028, |
| "learning_rate": 0.00010729887149617765, |
| "loss": 8.66411865234375, |
| "num_input_tokens_seen": 15144448, |
| "step": 3700, |
| "train_runtime": 7438.0137, |
| "train_tokens_per_second": 2036.088 |
| }, |
| { |
| "epoch": 3.973230850781871, |
| "grad_norm": 0.6406344771385193, |
| "learning_rate": 0.0001045686203130688, |
| "loss": 8.713903198242187, |
| "num_input_tokens_seen": 15349248, |
| "step": 3750, |
| "train_runtime": 7529.334, |
| "train_tokens_per_second": 2038.593 |
| }, |
| { |
| "epoch": 4.025443943811291, |
| "grad_norm": 0.6602168679237366, |
| "learning_rate": 0.00010183836912995995, |
| "loss": 8.64723388671875, |
| "num_input_tokens_seen": 15550464, |
| "step": 3800, |
| "train_runtime": 7619.5214, |
| "train_tokens_per_second": 2040.871 |
| }, |
| { |
| "epoch": 4.078452160084813, |
| "grad_norm": 0.6122914552688599, |
| "learning_rate": 9.910811794685111e-05, |
| "loss": 8.693761596679687, |
| "num_input_tokens_seen": 15755264, |
| "step": 3850, |
| "train_runtime": 7710.944, |
| "train_tokens_per_second": 2043.234 |
| }, |
| { |
| "epoch": 4.131460376358335, |
| "grad_norm": 0.6196545958518982, |
| "learning_rate": 9.637786676374226e-05, |
| "loss": 8.59835693359375, |
| "num_input_tokens_seen": 15960064, |
| "step": 3900, |
| "train_runtime": 7802.3986, |
| "train_tokens_per_second": 2045.533 |
| }, |
| { |
| "epoch": 4.184468592631858, |
| "grad_norm": 0.7231429815292358, |
| "learning_rate": 9.364761558063342e-05, |
| "loss": 8.621876831054687, |
| "num_input_tokens_seen": 16164864, |
| "step": 3950, |
| "train_runtime": 7893.7536, |
| "train_tokens_per_second": 2047.804 |
| }, |
| { |
| "epoch": 4.23747680890538, |
| "grad_norm": 0.6612716913223267, |
| "learning_rate": 9.091736439752458e-05, |
| "loss": 8.63590087890625, |
| "num_input_tokens_seen": 16369664, |
| "step": 4000, |
| "train_runtime": 7985.1156, |
| "train_tokens_per_second": 2050.022 |
| }, |
| { |
| "epoch": 4.23747680890538, |
| "eval_bleu": 4.17, |
| "eval_chrf": 20.0, |
| "eval_loss": 1.8474465608596802, |
| "eval_runtime": 85.4899, |
| "eval_samples_per_second": 3.568, |
| "eval_steps_per_second": 1.79, |
| "num_input_tokens_seen": 16369664, |
| "step": 4000 |
| }, |
| { |
| "epoch": 4.290485025178903, |
| "grad_norm": 0.611146092414856, |
| "learning_rate": 8.818711321441571e-05, |
| "loss": 8.60938720703125, |
| "num_input_tokens_seen": 16574464, |
| "step": 4050, |
| "train_runtime": 8174.4094, |
| "train_tokens_per_second": 2027.604 |
| }, |
| { |
| "epoch": 4.343493241452425, |
| "grad_norm": 0.6369199156761169, |
| "learning_rate": 8.545686203130687e-05, |
| "loss": 8.630716552734375, |
| "num_input_tokens_seen": 16779264, |
| "step": 4100, |
| "train_runtime": 8265.9997, |
| "train_tokens_per_second": 2029.913 |
| }, |
| { |
| "epoch": 4.396501457725948, |
| "grad_norm": 0.6628316044807434, |
| "learning_rate": 8.272661084819803e-05, |
| "loss": 8.668856811523437, |
| "num_input_tokens_seen": 16984064, |
| "step": 4150, |
| "train_runtime": 8357.3474, |
| "train_tokens_per_second": 2032.231 |
| }, |
| { |
| "epoch": 4.4495096739994695, |
| "grad_norm": 0.6495283842086792, |
| "learning_rate": 7.999635966508917e-05, |
| "loss": 8.637254638671875, |
| "num_input_tokens_seen": 17188864, |
| "step": 4200, |
| "train_runtime": 8448.7845, |
| "train_tokens_per_second": 2034.478 |
| }, |
| { |
| "epoch": 4.502517890272992, |
| "grad_norm": 0.7096723318099976, |
| "learning_rate": 7.726610848198033e-05, |
| "loss": 8.743385009765625, |
| "num_input_tokens_seen": 17393664, |
| "step": 4250, |
| "train_runtime": 8540.0931, |
| "train_tokens_per_second": 2036.707 |
| }, |
| { |
| "epoch": 4.555526106546514, |
| "grad_norm": 0.6975917816162109, |
| "learning_rate": 7.453585729887149e-05, |
| "loss": 8.73980224609375, |
| "num_input_tokens_seen": 17598464, |
| "step": 4300, |
| "train_runtime": 8631.8018, |
| "train_tokens_per_second": 2038.794 |
| }, |
| { |
| "epoch": 4.608534322820037, |
| "grad_norm": 0.6756250262260437, |
| "learning_rate": 7.180560611576264e-05, |
| "loss": 8.648682250976563, |
| "num_input_tokens_seen": 17803264, |
| "step": 4350, |
| "train_runtime": 8723.0816, |
| "train_tokens_per_second": 2040.937 |
| }, |
| { |
| "epoch": 4.66154253909356, |
| "grad_norm": 0.6954464912414551, |
| "learning_rate": 6.90753549326538e-05, |
| "loss": 8.505289916992188, |
| "num_input_tokens_seen": 18008064, |
| "step": 4400, |
| "train_runtime": 8814.4396, |
| "train_tokens_per_second": 2043.019 |
| }, |
| { |
| "epoch": 4.714550755367082, |
| "grad_norm": 0.7037732005119324, |
| "learning_rate": 6.634510374954495e-05, |
| "loss": 8.64231201171875, |
| "num_input_tokens_seen": 18212864, |
| "step": 4450, |
| "train_runtime": 8905.8242, |
| "train_tokens_per_second": 2045.051 |
| }, |
| { |
| "epoch": 4.767558971640605, |
| "grad_norm": 0.7270088195800781, |
| "learning_rate": 6.361485256643611e-05, |
| "loss": 8.542531127929687, |
| "num_input_tokens_seen": 18417664, |
| "step": 4500, |
| "train_runtime": 8997.045, |
| "train_tokens_per_second": 2047.079 |
| }, |
| { |
| "epoch": 4.767558971640605, |
| "eval_bleu": 4.19, |
| "eval_chrf": 19.94, |
| "eval_loss": 1.844648003578186, |
| "eval_runtime": 85.4507, |
| "eval_samples_per_second": 3.569, |
| "eval_steps_per_second": 1.791, |
| "num_input_tokens_seen": 18417664, |
| "step": 4500 |
| }, |
| { |
| "epoch": 4.8205671879141265, |
| "grad_norm": 0.8335739374160767, |
| "learning_rate": 6.088460138332727e-05, |
| "loss": 8.696432495117188, |
| "num_input_tokens_seen": 18622464, |
| "step": 4550, |
| "train_runtime": 9187.0284, |
| "train_tokens_per_second": 2027.039 |
| }, |
| { |
| "epoch": 4.873575404187649, |
| "grad_norm": 0.6931884288787842, |
| "learning_rate": 5.815435020021841e-05, |
| "loss": 8.64121826171875, |
| "num_input_tokens_seen": 18827264, |
| "step": 4600, |
| "train_runtime": 9278.849, |
| "train_tokens_per_second": 2029.052 |
| }, |
| { |
| "epoch": 4.926583620461171, |
| "grad_norm": 0.6581870317459106, |
| "learning_rate": 5.5424099017109565e-05, |
| "loss": 8.625559692382813, |
| "num_input_tokens_seen": 19032064, |
| "step": 4650, |
| "train_runtime": 9370.2629, |
| "train_tokens_per_second": 2031.113 |
| }, |
| { |
| "epoch": 4.979591836734694, |
| "grad_norm": 0.6969789266586304, |
| "learning_rate": 5.2693847834000724e-05, |
| "loss": 8.7086279296875, |
| "num_input_tokens_seen": 19236864, |
| "step": 4700, |
| "train_runtime": 9461.717, |
| "train_tokens_per_second": 2033.126 |
| }, |
| { |
| "epoch": 5.031804929764114, |
| "grad_norm": 0.7297267317771912, |
| "learning_rate": 4.9963596650891877e-05, |
| "loss": 8.405086059570312, |
| "num_input_tokens_seen": 19438080, |
| "step": 4750, |
| "train_runtime": 9552.1329, |
| "train_tokens_per_second": 2034.947 |
| }, |
| { |
| "epoch": 5.084813146037636, |
| "grad_norm": 0.6402953863143921, |
| "learning_rate": 4.723334546778303e-05, |
| "loss": 8.690524291992187, |
| "num_input_tokens_seen": 19642880, |
| "step": 4800, |
| "train_runtime": 9643.5906, |
| "train_tokens_per_second": 2036.884 |
| }, |
| { |
| "epoch": 5.137821362311159, |
| "grad_norm": 0.6262202262878418, |
| "learning_rate": 4.450309428467419e-05, |
| "loss": 8.471676635742188, |
| "num_input_tokens_seen": 19847680, |
| "step": 4850, |
| "train_runtime": 9735.0779, |
| "train_tokens_per_second": 2038.78 |
| }, |
| { |
| "epoch": 5.1908295785846805, |
| "grad_norm": 0.6598888039588928, |
| "learning_rate": 4.177284310156534e-05, |
| "loss": 8.572631225585937, |
| "num_input_tokens_seen": 20052480, |
| "step": 4900, |
| "train_runtime": 9826.6555, |
| "train_tokens_per_second": 2040.621 |
| }, |
| { |
| "epoch": 5.243837794858203, |
| "grad_norm": 0.7225948572158813, |
| "learning_rate": 3.90425919184565e-05, |
| "loss": 8.642888793945312, |
| "num_input_tokens_seen": 20257280, |
| "step": 4950, |
| "train_runtime": 9918.0414, |
| "train_tokens_per_second": 2042.468 |
| }, |
| { |
| "epoch": 5.296846011131725, |
| "grad_norm": 0.6872982382774353, |
| "learning_rate": 3.6312340735347646e-05, |
| "loss": 8.432411499023438, |
| "num_input_tokens_seen": 20462080, |
| "step": 5000, |
| "train_runtime": 10009.6849, |
| "train_tokens_per_second": 2044.228 |
| }, |
| { |
| "epoch": 5.296846011131725, |
| "eval_bleu": 4.09, |
| "eval_chrf": 19.82, |
| "eval_loss": 1.841970443725586, |
| "eval_runtime": 86.237, |
| "eval_samples_per_second": 3.537, |
| "eval_steps_per_second": 1.774, |
| "num_input_tokens_seen": 20462080, |
| "step": 5000 |
| }, |
| { |
| "epoch": 5.349854227405248, |
| "grad_norm": 0.6799459457397461, |
| "learning_rate": 3.3582089552238805e-05, |
| "loss": 8.610306396484376, |
| "num_input_tokens_seen": 20666880, |
| "step": 5050, |
| "train_runtime": 10200.4654, |
| "train_tokens_per_second": 2026.072 |
| }, |
| { |
| "epoch": 5.40286244367877, |
| "grad_norm": 0.6314374804496765, |
| "learning_rate": 3.085183836912996e-05, |
| "loss": 8.522672119140625, |
| "num_input_tokens_seen": 20871680, |
| "step": 5100, |
| "train_runtime": 10292.3344, |
| "train_tokens_per_second": 2027.886 |
| }, |
| { |
| "epoch": 5.455870659952293, |
| "grad_norm": 0.6840763688087463, |
| "learning_rate": 2.812158718602111e-05, |
| "loss": 8.5241162109375, |
| "num_input_tokens_seen": 21076480, |
| "step": 5150, |
| "train_runtime": 10384.0278, |
| "train_tokens_per_second": 2029.702 |
| }, |
| { |
| "epoch": 5.508878876225815, |
| "grad_norm": 0.7108559608459473, |
| "learning_rate": 2.5391336002912266e-05, |
| "loss": 8.657234497070313, |
| "num_input_tokens_seen": 21281280, |
| "step": 5200, |
| "train_runtime": 10475.6603, |
| "train_tokens_per_second": 2031.498 |
| }, |
| { |
| "epoch": 5.5618870924993375, |
| "grad_norm": 0.7527778148651123, |
| "learning_rate": 2.266108481980342e-05, |
| "loss": 8.60021240234375, |
| "num_input_tokens_seen": 21486080, |
| "step": 5250, |
| "train_runtime": 10567.164, |
| "train_tokens_per_second": 2033.287 |
| }, |
| { |
| "epoch": 5.6148953087728595, |
| "grad_norm": 0.647541880607605, |
| "learning_rate": 1.993083363669457e-05, |
| "loss": 8.624324951171875, |
| "num_input_tokens_seen": 21690880, |
| "step": 5300, |
| "train_runtime": 10658.6905, |
| "train_tokens_per_second": 2035.042 |
| }, |
| { |
| "epoch": 5.667903525046382, |
| "grad_norm": 0.5897159576416016, |
| "learning_rate": 1.7200582453585727e-05, |
| "loss": 8.644212646484375, |
| "num_input_tokens_seen": 21895680, |
| "step": 5350, |
| "train_runtime": 10750.2076, |
| "train_tokens_per_second": 2036.768 |
| }, |
| { |
| "epoch": 5.720911741319904, |
| "grad_norm": 0.6161897778511047, |
| "learning_rate": 1.4470331270476882e-05, |
| "loss": 8.561669311523438, |
| "num_input_tokens_seen": 22100480, |
| "step": 5400, |
| "train_runtime": 10841.7811, |
| "train_tokens_per_second": 2038.455 |
| }, |
| { |
| "epoch": 5.773919957593427, |
| "grad_norm": 0.6419637799263, |
| "learning_rate": 1.1740080087368037e-05, |
| "loss": 8.653330688476563, |
| "num_input_tokens_seen": 22305280, |
| "step": 5450, |
| "train_runtime": 10933.2667, |
| "train_tokens_per_second": 2040.13 |
| }, |
| { |
| "epoch": 5.82692817386695, |
| "grad_norm": 0.6721702218055725, |
| "learning_rate": 9.00982890425919e-06, |
| "loss": 8.615657348632812, |
| "num_input_tokens_seen": 22510080, |
| "step": 5500, |
| "train_runtime": 11024.7165, |
| "train_tokens_per_second": 2041.783 |
| }, |
| { |
| "epoch": 5.82692817386695, |
| "eval_bleu": 4.19, |
| "eval_chrf": 20.05, |
| "eval_loss": 1.8417972326278687, |
| "eval_runtime": 86.5279, |
| "eval_samples_per_second": 3.525, |
| "eval_steps_per_second": 1.768, |
| "num_input_tokens_seen": 22510080, |
| "step": 5500 |
| }, |
| { |
| "epoch": 5.879936390140472, |
| "grad_norm": 0.7171221375465393, |
| "learning_rate": 6.279577721150346e-06, |
| "loss": 8.624059448242187, |
| "num_input_tokens_seen": 22714880, |
| "step": 5550, |
| "train_runtime": 11215.8439, |
| "train_tokens_per_second": 2025.249 |
| }, |
| { |
| "epoch": 5.932944606413994, |
| "grad_norm": 0.6685547232627869, |
| "learning_rate": 3.549326538041499e-06, |
| "loss": 8.639024047851562, |
| "num_input_tokens_seen": 22919680, |
| "step": 5600, |
| "train_runtime": 11307.5506, |
| "train_tokens_per_second": 2026.936 |
| }, |
| { |
| "epoch": 5.9859528226875165, |
| "grad_norm": 0.6983809471130371, |
| "learning_rate": 8.190753549326538e-07, |
| "loss": 8.527723388671875, |
| "num_input_tokens_seen": 23124480, |
| "step": 5650, |
| "train_runtime": 11399.0902, |
| "train_tokens_per_second": 2028.625 |
| }, |
| { |
| "epoch": 6.0, |
| "num_input_tokens_seen": 23178240, |
| "step": 5664, |
| "total_flos": 4.229533786963968e+16, |
| "train_loss": 8.868080515309241, |
| "train_runtime": 11434.2452, |
| "train_samples_per_second": 15.837, |
| "train_steps_per_second": 0.495 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 5664, |
| "num_input_tokens_seen": 23178240, |
| "num_train_epochs": 6, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.229533786963968e+16, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
