| { | |
| "best_metric": 0.8302884697914124, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-450", | |
| "epoch": 0.08468197214904027, | |
| "eval_steps": 150, | |
| "global_step": 450, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0001881821603312006, | |
| "grad_norm": 8.021297454833984, | |
| "learning_rate": 5e-06, | |
| "loss": 4.7627, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001881821603312006, | |
| "eval_loss": 1.206099271774292, | |
| "eval_runtime": 977.2207, | |
| "eval_samples_per_second": 9.159, | |
| "eval_steps_per_second": 2.29, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0003763643206624012, | |
| "grad_norm": 8.330506324768066, | |
| "learning_rate": 1e-05, | |
| "loss": 4.7186, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0005645464809936018, | |
| "grad_norm": 7.244218826293945, | |
| "learning_rate": 1.5e-05, | |
| "loss": 4.7416, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0007527286413248024, | |
| "grad_norm": 7.7417449951171875, | |
| "learning_rate": 2e-05, | |
| "loss": 4.8388, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.000940910801656003, | |
| "grad_norm": 7.4192376136779785, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.4709, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0011290929619872036, | |
| "grad_norm": 4.583150386810303, | |
| "learning_rate": 3e-05, | |
| "loss": 4.305, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0013172751223184042, | |
| "grad_norm": 3.533625602722168, | |
| "learning_rate": 3.5e-05, | |
| "loss": 4.0565, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0015054572826496049, | |
| "grad_norm": 4.560721397399902, | |
| "learning_rate": 4e-05, | |
| "loss": 4.0204, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0016936394429808055, | |
| "grad_norm": 13.375167846679688, | |
| "learning_rate": 4.5e-05, | |
| "loss": 4.3096, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.001881821603312006, | |
| "grad_norm": 6.8327813148498535, | |
| "learning_rate": 5e-05, | |
| "loss": 4.279, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0020700037636432068, | |
| "grad_norm": 5.540746688842773, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 4.2908, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.002258185923974407, | |
| "grad_norm": 4.140532970428467, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2689, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0024463680843056076, | |
| "grad_norm": 1.9986830949783325, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.0747, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0026345502446368085, | |
| "grad_norm": 1.936602234840393, | |
| "learning_rate": 7e-05, | |
| "loss": 3.956, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.002822732404968009, | |
| "grad_norm": 1.9803358316421509, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.0274, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0030109145652992097, | |
| "grad_norm": 2.102954864501953, | |
| "learning_rate": 8e-05, | |
| "loss": 3.9315, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.00319909672563041, | |
| "grad_norm": 1.8604391813278198, | |
| "learning_rate": 8.5e-05, | |
| "loss": 3.9579, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.003387278885961611, | |
| "grad_norm": 5.816336631774902, | |
| "learning_rate": 9e-05, | |
| "loss": 3.8109, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0035754610462928114, | |
| "grad_norm": 1.8846168518066406, | |
| "learning_rate": 9.5e-05, | |
| "loss": 3.8936, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003763643206624012, | |
| "grad_norm": 2.1951682567596436, | |
| "learning_rate": 0.0001, | |
| "loss": 3.8866, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003951825366955213, | |
| "grad_norm": 1.8232979774475098, | |
| "learning_rate": 9.999866555428618e-05, | |
| "loss": 3.7866, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0041400075272864136, | |
| "grad_norm": 1.7551851272583008, | |
| "learning_rate": 9.999466228837451e-05, | |
| "loss": 3.848, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0043281896876176135, | |
| "grad_norm": 3.396578788757324, | |
| "learning_rate": 9.998799041595064e-05, | |
| "loss": 3.8432, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.004516371847948814, | |
| "grad_norm": 1.9480079412460327, | |
| "learning_rate": 9.997865029314463e-05, | |
| "loss": 3.9188, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.004704554008280015, | |
| "grad_norm": 1.8144193887710571, | |
| "learning_rate": 9.996664241851197e-05, | |
| "loss": 3.9241, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004892736168611215, | |
| "grad_norm": 1.6157795190811157, | |
| "learning_rate": 9.995196743300692e-05, | |
| "loss": 3.7515, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.005080918328942416, | |
| "grad_norm": 1.8001606464385986, | |
| "learning_rate": 9.993462611994832e-05, | |
| "loss": 3.8946, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.005269100489273617, | |
| "grad_norm": 2.1896607875823975, | |
| "learning_rate": 9.991461940497786e-05, | |
| "loss": 3.6935, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.005457282649604818, | |
| "grad_norm": 1.9953135251998901, | |
| "learning_rate": 9.989194835601048e-05, | |
| "loss": 3.6832, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.005645464809936018, | |
| "grad_norm": 1.9537560939788818, | |
| "learning_rate": 9.986661418317759e-05, | |
| "loss": 3.6689, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005833646970267219, | |
| "grad_norm": 1.7720470428466797, | |
| "learning_rate": 9.983861823876231e-05, | |
| "loss": 3.8731, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0060218291305984195, | |
| "grad_norm": 1.8037577867507935, | |
| "learning_rate": 9.980796201712734e-05, | |
| "loss": 3.7388, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0062100112909296195, | |
| "grad_norm": 1.558807611465454, | |
| "learning_rate": 9.977464715463524e-05, | |
| "loss": 3.5732, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00639819345126082, | |
| "grad_norm": 1.7031415700912476, | |
| "learning_rate": 9.973867542956104e-05, | |
| "loss": 3.7046, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.006586375611592021, | |
| "grad_norm": 1.8096400499343872, | |
| "learning_rate": 9.97000487619973e-05, | |
| "loss": 3.7951, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.006774557771923222, | |
| "grad_norm": 1.6755597591400146, | |
| "learning_rate": 9.965876921375165e-05, | |
| "loss": 3.7345, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.006962739932254422, | |
| "grad_norm": 1.6606807708740234, | |
| "learning_rate": 9.961483898823678e-05, | |
| "loss": 3.6923, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.007150922092585623, | |
| "grad_norm": 1.602203130722046, | |
| "learning_rate": 9.956826043035268e-05, | |
| "loss": 3.6913, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.007339104252916824, | |
| "grad_norm": 1.6571152210235596, | |
| "learning_rate": 9.951903602636166e-05, | |
| "loss": 3.6178, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.007527286413248024, | |
| "grad_norm": 1.737025499343872, | |
| "learning_rate": 9.946716840375551e-05, | |
| "loss": 3.6084, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0077154685735792245, | |
| "grad_norm": 1.6381107568740845, | |
| "learning_rate": 9.94126603311153e-05, | |
| "loss": 3.4131, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.007903650733910425, | |
| "grad_norm": 1.668062686920166, | |
| "learning_rate": 9.935551471796358e-05, | |
| "loss": 3.4251, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.008091832894241625, | |
| "grad_norm": 1.6551685333251953, | |
| "learning_rate": 9.92957346146091e-05, | |
| "loss": 3.4862, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.008280015054572827, | |
| "grad_norm": 1.5389114618301392, | |
| "learning_rate": 9.923332321198395e-05, | |
| "loss": 3.3558, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.008468197214904027, | |
| "grad_norm": 1.6278464794158936, | |
| "learning_rate": 9.916828384147331e-05, | |
| "loss": 3.6091, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.008656379375235227, | |
| "grad_norm": 1.642901062965393, | |
| "learning_rate": 9.910061997473752e-05, | |
| "loss": 3.3503, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.008844561535566429, | |
| "grad_norm": 2.450819730758667, | |
| "learning_rate": 9.903033522352687e-05, | |
| "loss": 3.7451, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.009032743695897629, | |
| "grad_norm": 1.6337559223175049, | |
| "learning_rate": 9.895743333948874e-05, | |
| "loss": 3.7759, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.009220925856228829, | |
| "grad_norm": 2.088834285736084, | |
| "learning_rate": 9.888191821396744e-05, | |
| "loss": 3.8653, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.00940910801656003, | |
| "grad_norm": 2.713131904602051, | |
| "learning_rate": 9.880379387779637e-05, | |
| "loss": 3.468, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.00959729017689123, | |
| "grad_norm": 1.8180787563323975, | |
| "learning_rate": 9.872306450108292e-05, | |
| "loss": 3.5826, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.00978547233722243, | |
| "grad_norm": 1.6677582263946533, | |
| "learning_rate": 9.863973439298597e-05, | |
| "loss": 3.7982, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.009973654497553632, | |
| "grad_norm": 1.569620966911316, | |
| "learning_rate": 9.855380800148572e-05, | |
| "loss": 3.7162, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.010161836657884832, | |
| "grad_norm": 1.6128700971603394, | |
| "learning_rate": 9.846528991314639e-05, | |
| "loss": 3.5721, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.010350018818216034, | |
| "grad_norm": 1.8484185934066772, | |
| "learning_rate": 9.837418485287127e-05, | |
| "loss": 3.9133, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.010538200978547234, | |
| "grad_norm": 1.706660270690918, | |
| "learning_rate": 9.828049768365068e-05, | |
| "loss": 3.7007, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.010726383138878434, | |
| "grad_norm": 1.7316683530807495, | |
| "learning_rate": 9.818423340630228e-05, | |
| "loss": 3.7366, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.010914565299209636, | |
| "grad_norm": 1.5684834718704224, | |
| "learning_rate": 9.808539715920414e-05, | |
| "loss": 3.6784, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.011102747459540836, | |
| "grad_norm": 1.5667593479156494, | |
| "learning_rate": 9.798399421802056e-05, | |
| "loss": 3.7003, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.011290929619872036, | |
| "grad_norm": 1.526808500289917, | |
| "learning_rate": 9.78800299954203e-05, | |
| "loss": 3.7078, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.011479111780203237, | |
| "grad_norm": 2.5877087116241455, | |
| "learning_rate": 9.777351004078783e-05, | |
| "loss": 3.6875, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.011667293940534437, | |
| "grad_norm": 1.545409917831421, | |
| "learning_rate": 9.766444003992703e-05, | |
| "loss": 3.8299, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.011855476100865637, | |
| "grad_norm": 1.5253973007202148, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.6405, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.012043658261196839, | |
| "grad_norm": 1.8160947561264038, | |
| "learning_rate": 9.743867332300478e-05, | |
| "loss": 3.7575, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.012231840421528039, | |
| "grad_norm": 1.62673819065094, | |
| "learning_rate": 9.732198865788047e-05, | |
| "loss": 3.736, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.012420022581859239, | |
| "grad_norm": 1.609445571899414, | |
| "learning_rate": 9.72027780477588e-05, | |
| "loss": 3.5029, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01260820474219044, | |
| "grad_norm": 1.640405535697937, | |
| "learning_rate": 9.708104785584323e-05, | |
| "loss": 3.6871, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01279638690252164, | |
| "grad_norm": 1.546713948249817, | |
| "learning_rate": 9.695680457982713e-05, | |
| "loss": 3.5635, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.012984569062852842, | |
| "grad_norm": 1.525071620941162, | |
| "learning_rate": 9.683005485154677e-05, | |
| "loss": 3.6238, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.013172751223184042, | |
| "grad_norm": 1.5092836618423462, | |
| "learning_rate": 9.67008054366274e-05, | |
| "loss": 3.6092, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.013360933383515242, | |
| "grad_norm": 1.6913105249404907, | |
| "learning_rate": 9.656906323412217e-05, | |
| "loss": 3.6286, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.013549115543846444, | |
| "grad_norm": 2.151688814163208, | |
| "learning_rate": 9.643483527614372e-05, | |
| "loss": 3.5817, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.013737297704177644, | |
| "grad_norm": 1.5763871669769287, | |
| "learning_rate": 9.629812872748901e-05, | |
| "loss": 3.5043, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.013925479864508844, | |
| "grad_norm": 1.5946530103683472, | |
| "learning_rate": 9.615895088525677e-05, | |
| "loss": 3.5902, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.014113662024840046, | |
| "grad_norm": 1.5125662088394165, | |
| "learning_rate": 9.601730917845797e-05, | |
| "loss": 3.5002, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.014301844185171246, | |
| "grad_norm": 1.952149510383606, | |
| "learning_rate": 9.587321116761938e-05, | |
| "loss": 3.5881, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.014490026345502446, | |
| "grad_norm": 1.5184606313705444, | |
| "learning_rate": 9.57266645443799e-05, | |
| "loss": 3.3213, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.014678208505833647, | |
| "grad_norm": 1.5643829107284546, | |
| "learning_rate": 9.557767713108009e-05, | |
| "loss": 3.7112, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.014866390666164847, | |
| "grad_norm": 1.5019704103469849, | |
| "learning_rate": 9.542625688034449e-05, | |
| "loss": 3.4665, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.015054572826496047, | |
| "grad_norm": 1.647053837776184, | |
| "learning_rate": 9.527241187465734e-05, | |
| "loss": 3.33, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.015242754986827249, | |
| "grad_norm": 1.527642846107483, | |
| "learning_rate": 9.511615032593096e-05, | |
| "loss": 3.6355, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.015430937147158449, | |
| "grad_norm": 1.6316813230514526, | |
| "learning_rate": 9.49574805750675e-05, | |
| "loss": 3.5475, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01561911930748965, | |
| "grad_norm": 1.6150362491607666, | |
| "learning_rate": 9.479641109151373e-05, | |
| "loss": 3.5272, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.01580730146782085, | |
| "grad_norm": 1.5907738208770752, | |
| "learning_rate": 9.463295047280891e-05, | |
| "loss": 3.6044, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.015995483628152053, | |
| "grad_norm": 1.5602425336837769, | |
| "learning_rate": 9.446710744412595e-05, | |
| "loss": 3.4846, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01618366578848325, | |
| "grad_norm": 1.5599199533462524, | |
| "learning_rate": 9.429889085780557e-05, | |
| "loss": 3.598, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.016371847948814452, | |
| "grad_norm": 1.4632290601730347, | |
| "learning_rate": 9.41283096928839e-05, | |
| "loss": 3.3955, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.016560030109145654, | |
| "grad_norm": 1.5269198417663574, | |
| "learning_rate": 9.395537305461311e-05, | |
| "loss": 3.5423, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.016748212269476852, | |
| "grad_norm": 1.6142536401748657, | |
| "learning_rate": 9.378009017397542e-05, | |
| "loss": 3.5366, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.016936394429808054, | |
| "grad_norm": 1.4717563390731812, | |
| "learning_rate": 9.360247040719039e-05, | |
| "loss": 3.4092, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.017124576590139256, | |
| "grad_norm": 2.3025355339050293, | |
| "learning_rate": 9.342252323521545e-05, | |
| "loss": 3.441, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.017312758750470454, | |
| "grad_norm": 1.6364383697509766, | |
| "learning_rate": 9.324025826323994e-05, | |
| "loss": 3.4081, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.017500940910801656, | |
| "grad_norm": 1.6637694835662842, | |
| "learning_rate": 9.305568522017227e-05, | |
| "loss": 3.4648, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.017689123071132858, | |
| "grad_norm": 1.5033453702926636, | |
| "learning_rate": 9.286881395812066e-05, | |
| "loss": 3.4444, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.017877305231464056, | |
| "grad_norm": 1.7990686893463135, | |
| "learning_rate": 9.267965445186733e-05, | |
| "loss": 3.4532, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.018065487391795258, | |
| "grad_norm": 1.5659180879592896, | |
| "learning_rate": 9.248821679833596e-05, | |
| "loss": 3.444, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01825366955212646, | |
| "grad_norm": 1.495245337486267, | |
| "learning_rate": 9.229451121605279e-05, | |
| "loss": 3.5251, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.018441851712457658, | |
| "grad_norm": 1.6818476915359497, | |
| "learning_rate": 9.209854804460121e-05, | |
| "loss": 3.323, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01863003387278886, | |
| "grad_norm": 1.9330247640609741, | |
| "learning_rate": 9.190033774406977e-05, | |
| "loss": 3.4345, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.01881821603312006, | |
| "grad_norm": 2.6322007179260254, | |
| "learning_rate": 9.16998908944939e-05, | |
| "loss": 3.2795, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01900639819345126, | |
| "grad_norm": 1.6533123254776, | |
| "learning_rate": 9.149721819529119e-05, | |
| "loss": 3.6009, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.01919458035378246, | |
| "grad_norm": 1.739073395729065, | |
| "learning_rate": 9.129233046469022e-05, | |
| "loss": 3.8017, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.019382762514113663, | |
| "grad_norm": 1.728940486907959, | |
| "learning_rate": 9.108523863915314e-05, | |
| "loss": 3.7084, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.01957094467444486, | |
| "grad_norm": 1.6274477243423462, | |
| "learning_rate": 9.087595377279192e-05, | |
| "loss": 3.5057, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.019759126834776063, | |
| "grad_norm": 1.6262691020965576, | |
| "learning_rate": 9.066448703677828e-05, | |
| "loss": 3.6988, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.019947308995107264, | |
| "grad_norm": 4.310033321380615, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 3.7893, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.020135491155438466, | |
| "grad_norm": 6.6696977615356445, | |
| "learning_rate": 9.023505322219536e-05, | |
| "loss": 3.5515, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.020323673315769664, | |
| "grad_norm": 1.6663572788238525, | |
| "learning_rate": 9.001710906587064e-05, | |
| "loss": 3.4112, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.020511855476100866, | |
| "grad_norm": 1.6214839220046997, | |
| "learning_rate": 8.9797028883159e-05, | |
| "loss": 3.3542, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.020700037636432068, | |
| "grad_norm": 1.540847897529602, | |
| "learning_rate": 8.957482442146272e-05, | |
| "loss": 3.6587, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.020888219796763266, | |
| "grad_norm": 1.5367134809494019, | |
| "learning_rate": 8.935050754157344e-05, | |
| "loss": 3.5199, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.021076401957094468, | |
| "grad_norm": 1.7050776481628418, | |
| "learning_rate": 8.912409021703913e-05, | |
| "loss": 3.6533, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.02126458411742567, | |
| "grad_norm": 1.5632902383804321, | |
| "learning_rate": 8.889558453352492e-05, | |
| "loss": 3.5123, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.021452766277756868, | |
| "grad_norm": 1.5934125185012817, | |
| "learning_rate": 8.866500268816803e-05, | |
| "loss": 3.5423, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02164094843808807, | |
| "grad_norm": 1.5730620622634888, | |
| "learning_rate": 8.84323569889266e-05, | |
| "loss": 3.3918, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.02182913059841927, | |
| "grad_norm": 1.5384756326675415, | |
| "learning_rate": 8.819765985392296e-05, | |
| "loss": 3.5834, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02201731275875047, | |
| "grad_norm": 1.4878556728363037, | |
| "learning_rate": 8.79609238107805e-05, | |
| "loss": 3.5273, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02220549491908167, | |
| "grad_norm": 1.8022403717041016, | |
| "learning_rate": 8.772216149595513e-05, | |
| "loss": 3.6865, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.022393677079412873, | |
| "grad_norm": 1.6074386835098267, | |
| "learning_rate": 8.748138565406081e-05, | |
| "loss": 3.35, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.02258185923974407, | |
| "grad_norm": 5.877849102020264, | |
| "learning_rate": 8.72386091371891e-05, | |
| "loss": 3.5387, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.022770041400075273, | |
| "grad_norm": 1.6074655055999756, | |
| "learning_rate": 8.699384490422331e-05, | |
| "loss": 3.4726, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.022958223560406475, | |
| "grad_norm": 1.5693819522857666, | |
| "learning_rate": 8.674710602014671e-05, | |
| "loss": 3.4279, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.023146405720737673, | |
| "grad_norm": 1.5131280422210693, | |
| "learning_rate": 8.649840565534513e-05, | |
| "loss": 3.4739, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.023334587881068874, | |
| "grad_norm": 1.5035758018493652, | |
| "learning_rate": 8.624775708490402e-05, | |
| "loss": 3.4492, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.023522770041400076, | |
| "grad_norm": 1.6009769439697266, | |
| "learning_rate": 8.59951736878998e-05, | |
| "loss": 3.6759, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.023710952201731274, | |
| "grad_norm": 1.539108395576477, | |
| "learning_rate": 8.574066894668573e-05, | |
| "loss": 3.5554, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.023899134362062476, | |
| "grad_norm": 1.8990849256515503, | |
| "learning_rate": 8.548425644617224e-05, | |
| "loss": 3.4451, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.024087316522393678, | |
| "grad_norm": 1.5071018934249878, | |
| "learning_rate": 8.522594987310184e-05, | |
| "loss": 3.4815, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.024275498682724876, | |
| "grad_norm": 1.5156155824661255, | |
| "learning_rate": 8.49657630153185e-05, | |
| "loss": 3.526, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.024463680843056078, | |
| "grad_norm": 1.4841188192367554, | |
| "learning_rate": 8.47037097610317e-05, | |
| "loss": 3.4503, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02465186300338728, | |
| "grad_norm": 1.8373347520828247, | |
| "learning_rate": 8.443980409807512e-05, | |
| "loss": 3.4492, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.024840045163718478, | |
| "grad_norm": 1.9632869958877563, | |
| "learning_rate": 8.417406011315998e-05, | |
| "loss": 3.5533, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02502822732404968, | |
| "grad_norm": 1.525312900543213, | |
| "learning_rate": 8.390649199112315e-05, | |
| "loss": 3.5785, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.02521640948438088, | |
| "grad_norm": 1.7048442363739014, | |
| "learning_rate": 8.363711401417e-05, | |
| "loss": 3.5644, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.025404591644712083, | |
| "grad_norm": 1.4897756576538086, | |
| "learning_rate": 8.336594056111197e-05, | |
| "loss": 3.4262, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02559277380504328, | |
| "grad_norm": 1.5408360958099365, | |
| "learning_rate": 8.309298610659916e-05, | |
| "loss": 3.5162, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.025780955965374483, | |
| "grad_norm": 1.613747239112854, | |
| "learning_rate": 8.281826522034764e-05, | |
| "loss": 3.6368, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.025969138125705685, | |
| "grad_norm": 1.5057644844055176, | |
| "learning_rate": 8.254179256636179e-05, | |
| "loss": 3.4841, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.026157320286036883, | |
| "grad_norm": 1.5167312622070312, | |
| "learning_rate": 8.226358290215151e-05, | |
| "loss": 3.3391, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.026345502446368085, | |
| "grad_norm": 1.5216038227081299, | |
| "learning_rate": 8.198365107794457e-05, | |
| "loss": 3.3654, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.026533684606699286, | |
| "grad_norm": 1.5505784749984741, | |
| "learning_rate": 8.17020120358939e-05, | |
| "loss": 3.461, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.026721866767030485, | |
| "grad_norm": 3.795452833175659, | |
| "learning_rate": 8.141868080927996e-05, | |
| "loss": 3.4515, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.026910048927361686, | |
| "grad_norm": 1.4928878545761108, | |
| "learning_rate": 8.113367252170844e-05, | |
| "loss": 3.4339, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.027098231087692888, | |
| "grad_norm": 1.5022090673446655, | |
| "learning_rate": 8.084700238630283e-05, | |
| "loss": 3.3639, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.027286413248024086, | |
| "grad_norm": 1.4917182922363281, | |
| "learning_rate": 8.055868570489247e-05, | |
| "loss": 3.2841, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.027474595408355288, | |
| "grad_norm": 1.5091995000839233, | |
| "learning_rate": 8.026873786719573e-05, | |
| "loss": 3.4067, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.02766277756868649, | |
| "grad_norm": 1.455728530883789, | |
| "learning_rate": 7.997717434999861e-05, | |
| "loss": 3.2996, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.027850959729017688, | |
| "grad_norm": 1.852761149406433, | |
| "learning_rate": 7.968401071632855e-05, | |
| "loss": 3.4227, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02803914188934889, | |
| "grad_norm": 1.8084666728973389, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.4068, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.02822732404968009, | |
| "grad_norm": 2.763946056365967, | |
| "learning_rate": 7.909294577789766e-05, | |
| "loss": 3.1156, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02822732404968009, | |
| "eval_loss": 0.8662680983543396, | |
| "eval_runtime": 982.7503, | |
| "eval_samples_per_second": 9.107, | |
| "eval_steps_per_second": 2.277, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02841550621001129, | |
| "grad_norm": 1.5687919855117798, | |
| "learning_rate": 7.879507602289979e-05, | |
| "loss": 3.6073, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.02860368837034249, | |
| "grad_norm": 1.6533427238464355, | |
| "learning_rate": 7.849566924927082e-05, | |
| "loss": 3.5083, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.028791870530673693, | |
| "grad_norm": 1.542561650276184, | |
| "learning_rate": 7.819474143869414e-05, | |
| "loss": 3.4967, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02898005269100489, | |
| "grad_norm": 1.516094446182251, | |
| "learning_rate": 7.789230865404287e-05, | |
| "loss": 3.3608, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.029168234851336093, | |
| "grad_norm": 1.6348127126693726, | |
| "learning_rate": 7.75883870385223e-05, | |
| "loss": 3.5607, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.029356417011667295, | |
| "grad_norm": 1.574805498123169, | |
| "learning_rate": 7.728299281480833e-05, | |
| "loss": 3.314, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.029544599171998493, | |
| "grad_norm": 1.5510149002075195, | |
| "learning_rate": 7.697614228418148e-05, | |
| "loss": 3.397, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.029732781332329695, | |
| "grad_norm": 1.4977482557296753, | |
| "learning_rate": 7.666785182565677e-05, | |
| "loss": 3.4013, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.029920963492660897, | |
| "grad_norm": 1.5679925680160522, | |
| "learning_rate": 7.635813789510941e-05, | |
| "loss": 3.3312, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.030109145652992095, | |
| "grad_norm": 1.5352201461791992, | |
| "learning_rate": 7.604701702439651e-05, | |
| "loss": 3.483, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.030297327813323296, | |
| "grad_norm": 1.52545964717865, | |
| "learning_rate": 7.573450582047457e-05, | |
| "loss": 3.5098, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.030485509973654498, | |
| "grad_norm": 1.5153985023498535, | |
| "learning_rate": 7.542062096451305e-05, | |
| "loss": 3.4748, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.030673692133985696, | |
| "grad_norm": 1.4865204095840454, | |
| "learning_rate": 7.510537921100398e-05, | |
| "loss": 3.3997, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.030861874294316898, | |
| "grad_norm": 1.5080010890960693, | |
| "learning_rate": 7.47887973868676e-05, | |
| "loss": 3.5227, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.0310500564546481, | |
| "grad_norm": 1.5261732339859009, | |
| "learning_rate": 7.447089239055428e-05, | |
| "loss": 3.7215, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0312382386149793, | |
| "grad_norm": 1.4631050825119019, | |
| "learning_rate": 7.41516811911424e-05, | |
| "loss": 3.3073, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.0314264207753105, | |
| "grad_norm": 1.5799752473831177, | |
| "learning_rate": 7.383118082743262e-05, | |
| "loss": 3.5134, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.0316146029356417, | |
| "grad_norm": 1.6088528633117676, | |
| "learning_rate": 7.350940840703842e-05, | |
| "loss": 3.4344, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.0318027850959729, | |
| "grad_norm": 1.5857129096984863, | |
| "learning_rate": 7.318638110547288e-05, | |
| "loss": 3.6729, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.031990967256304105, | |
| "grad_norm": 1.9101580381393433, | |
| "learning_rate": 7.286211616523193e-05, | |
| "loss": 3.4667, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0321791494166353, | |
| "grad_norm": 1.6800731420516968, | |
| "learning_rate": 7.253663089487395e-05, | |
| "loss": 3.3363, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.0323673315769665, | |
| "grad_norm": 1.5496375560760498, | |
| "learning_rate": 7.220994266809591e-05, | |
| "loss": 3.5252, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.0325555137372977, | |
| "grad_norm": 1.4972703456878662, | |
| "learning_rate": 7.188206892280594e-05, | |
| "loss": 3.43, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.032743695897628905, | |
| "grad_norm": 1.4559191465377808, | |
| "learning_rate": 7.155302716019263e-05, | |
| "loss": 3.2506, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03293187805796011, | |
| "grad_norm": 1.6048904657363892, | |
| "learning_rate": 7.122283494379076e-05, | |
| "loss": 3.6594, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03312006021829131, | |
| "grad_norm": 1.5769932270050049, | |
| "learning_rate": 7.089150989854385e-05, | |
| "loss": 3.3905, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.0333082423786225, | |
| "grad_norm": 1.508113980293274, | |
| "learning_rate": 7.055906970986336e-05, | |
| "loss": 3.4611, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.033496424538953705, | |
| "grad_norm": 1.4894671440124512, | |
| "learning_rate": 7.022553212268469e-05, | |
| "loss": 3.3249, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03368460669928491, | |
| "grad_norm": 1.5143218040466309, | |
| "learning_rate": 6.989091494051998e-05, | |
| "loss": 3.3934, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.03387278885961611, | |
| "grad_norm": 1.4756419658660889, | |
| "learning_rate": 6.95552360245078e-05, | |
| "loss": 3.2264, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.03406097101994731, | |
| "grad_norm": 1.5053011178970337, | |
| "learning_rate": 6.92185132924598e-05, | |
| "loss": 3.5413, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.03424915318027851, | |
| "grad_norm": 1.8715499639511108, | |
| "learning_rate": 6.888076471790424e-05, | |
| "loss": 3.4045, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.034437335340609714, | |
| "grad_norm": 1.8220783472061157, | |
| "learning_rate": 6.85420083291266e-05, | |
| "loss": 3.4066, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03462551750094091, | |
| "grad_norm": 1.5312058925628662, | |
| "learning_rate": 6.820226220820732e-05, | |
| "loss": 3.474, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03481369966127211, | |
| "grad_norm": 1.5189887285232544, | |
| "learning_rate": 6.786154449005665e-05, | |
| "loss": 3.2858, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03500188182160331, | |
| "grad_norm": 1.5011916160583496, | |
| "learning_rate": 6.751987336144648e-05, | |
| "loss": 3.3292, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03519006398193451, | |
| "grad_norm": 1.6477534770965576, | |
| "learning_rate": 6.717726706003974e-05, | |
| "loss": 3.4825, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.035378246142265715, | |
| "grad_norm": 1.5640369653701782, | |
| "learning_rate": 6.683374387341687e-05, | |
| "loss": 3.3886, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03556642830259692, | |
| "grad_norm": 1.5490057468414307, | |
| "learning_rate": 6.648932213809962e-05, | |
| "loss": 3.4445, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03575461046292811, | |
| "grad_norm": 1.5441166162490845, | |
| "learning_rate": 6.614402023857232e-05, | |
| "loss": 3.2839, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03594279262325931, | |
| "grad_norm": 1.5490269660949707, | |
| "learning_rate": 6.579785660630056e-05, | |
| "loss": 3.548, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.036130974783590515, | |
| "grad_norm": 1.4607436656951904, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.3022, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03631915694392172, | |
| "grad_norm": 1.5263044834136963, | |
| "learning_rate": 6.510301809838689e-05, | |
| "loss": 3.5151, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03650733910425292, | |
| "grad_norm": 1.7463645935058594, | |
| "learning_rate": 6.475438031171574e-05, | |
| "loss": 3.4085, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03669552126458412, | |
| "grad_norm": 1.4995930194854736, | |
| "learning_rate": 6.440495496826189e-05, | |
| "loss": 3.4332, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.036883703424915315, | |
| "grad_norm": 1.5385165214538574, | |
| "learning_rate": 6.405476071959143e-05, | |
| "loss": 3.432, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03707188558524652, | |
| "grad_norm": 1.5226582288742065, | |
| "learning_rate": 6.370381625831292e-05, | |
| "loss": 3.3812, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03726006774557772, | |
| "grad_norm": 1.5849664211273193, | |
| "learning_rate": 6.335214031707965e-05, | |
| "loss": 3.4423, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03744824990590892, | |
| "grad_norm": 1.8131951093673706, | |
| "learning_rate": 6.299975166758971e-05, | |
| "loss": 3.3169, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03763643206624012, | |
| "grad_norm": 2.523836851119995, | |
| "learning_rate": 6.264666911958404e-05, | |
| "loss": 3.1207, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.037824614226571324, | |
| "grad_norm": 1.6073590517044067, | |
| "learning_rate": 6.229291151984233e-05, | |
| "loss": 3.5545, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03801279638690252, | |
| "grad_norm": 3.597961187362671, | |
| "learning_rate": 6.19384977511771e-05, | |
| "loss": 3.3569, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03820097854723372, | |
| "grad_norm": 1.5076078176498413, | |
| "learning_rate": 6.158344673142573e-05, | |
| "loss": 3.5497, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.03838916070756492, | |
| "grad_norm": 1.5262771844863892, | |
| "learning_rate": 6.122777741244067e-05, | |
| "loss": 3.5109, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.038577342867896124, | |
| "grad_norm": 1.5841517448425293, | |
| "learning_rate": 6.0871508779077856e-05, | |
| "loss": 3.6458, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.038765525028227325, | |
| "grad_norm": 1.706384539604187, | |
| "learning_rate": 6.051465984818332e-05, | |
| "loss": 3.4581, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.03895370718855853, | |
| "grad_norm": 1.562947154045105, | |
| "learning_rate": 6.015724966757812e-05, | |
| "loss": 3.6251, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03914188934888972, | |
| "grad_norm": 1.5278199911117554, | |
| "learning_rate": 5.979929731504158e-05, | |
| "loss": 3.4025, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.039330071509220924, | |
| "grad_norm": 1.528238296508789, | |
| "learning_rate": 5.944082189729301e-05, | |
| "loss": 3.5747, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.039518253669552125, | |
| "grad_norm": 1.6547585725784302, | |
| "learning_rate": 5.908184254897182e-05, | |
| "loss": 3.4107, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03970643582988333, | |
| "grad_norm": 1.5595186948776245, | |
| "learning_rate": 5.872237843161612e-05, | |
| "loss": 3.4565, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.03989461799021453, | |
| "grad_norm": 1.5088775157928467, | |
| "learning_rate": 5.8362448732639894e-05, | |
| "loss": 3.3846, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.04008280015054573, | |
| "grad_norm": 1.543283224105835, | |
| "learning_rate": 5.800207266430895e-05, | |
| "loss": 3.5448, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.04027098231087693, | |
| "grad_norm": 1.4654241800308228, | |
| "learning_rate": 5.764126946271526e-05, | |
| "loss": 3.3228, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.04045916447120813, | |
| "grad_norm": 1.6389522552490234, | |
| "learning_rate": 5.7280058386750255e-05, | |
| "loss": 3.357, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.04064734663153933, | |
| "grad_norm": 1.4927722215652466, | |
| "learning_rate": 5.6918458717076815e-05, | |
| "loss": 3.434, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.04083552879187053, | |
| "grad_norm": 1.5620532035827637, | |
| "learning_rate": 5.655648975510014e-05, | |
| "loss": 3.5867, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.04102371095220173, | |
| "grad_norm": 1.6729021072387695, | |
| "learning_rate": 5.61941708219374e-05, | |
| "loss": 3.3869, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.041211893112532934, | |
| "grad_norm": 1.523036003112793, | |
| "learning_rate": 5.583152125738651e-05, | |
| "loss": 3.4844, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.041400075272864136, | |
| "grad_norm": 1.4695093631744385, | |
| "learning_rate": 5.546856041889373e-05, | |
| "loss": 3.4079, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.04158825743319533, | |
| "grad_norm": 1.525160789489746, | |
| "learning_rate": 5.510530768052047e-05, | |
| "loss": 3.358, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.04177643959352653, | |
| "grad_norm": 1.4674233198165894, | |
| "learning_rate": 5.4741782431909136e-05, | |
| "loss": 3.2476, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.041964621753857734, | |
| "grad_norm": 1.7390276193618774, | |
| "learning_rate": 5.437800407724812e-05, | |
| "loss": 3.3376, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.042152803914188935, | |
| "grad_norm": 1.5143543481826782, | |
| "learning_rate": 5.401399203423606e-05, | |
| "loss": 3.5849, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.04234098607452014, | |
| "grad_norm": 1.5307164192199707, | |
| "learning_rate": 5.364976573304538e-05, | |
| "loss": 3.2994, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.04252916823485134, | |
| "grad_norm": 1.5590108633041382, | |
| "learning_rate": 5.328534461528515e-05, | |
| "loss": 3.5251, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.042717350395182534, | |
| "grad_norm": 1.5407838821411133, | |
| "learning_rate": 5.29207481329633e-05, | |
| "loss": 3.3932, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.042905532555513735, | |
| "grad_norm": 1.5244678258895874, | |
| "learning_rate": 5.2555995747448364e-05, | |
| "loss": 3.3437, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.04309371471584494, | |
| "grad_norm": 1.4933329820632935, | |
| "learning_rate": 5.2191106928430644e-05, | |
| "loss": 3.399, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.04328189687617614, | |
| "grad_norm": 1.4906306266784668, | |
| "learning_rate": 5.182610115288295e-05, | |
| "loss": 3.2872, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.04347007903650734, | |
| "grad_norm": 1.647549033164978, | |
| "learning_rate": 5.1460997904021005e-05, | |
| "loss": 3.3683, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.04365826119683854, | |
| "grad_norm": 1.5551615953445435, | |
| "learning_rate": 5.109581667026341e-05, | |
| "loss": 3.4049, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.04384644335716974, | |
| "grad_norm": 1.5681029558181763, | |
| "learning_rate": 5.073057694419147e-05, | |
| "loss": 3.4455, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.04403462551750094, | |
| "grad_norm": 1.5506995916366577, | |
| "learning_rate": 5.036529822150865e-05, | |
| "loss": 3.3317, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.04422280767783214, | |
| "grad_norm": 2.233480215072632, | |
| "learning_rate": 5e-05, | |
| "loss": 3.4562, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.04441098983816334, | |
| "grad_norm": 1.4921319484710693, | |
| "learning_rate": 4.963470177849135e-05, | |
| "loss": 3.3868, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.044599171998494544, | |
| "grad_norm": 1.539383888244629, | |
| "learning_rate": 4.9269423055808544e-05, | |
| "loss": 3.3888, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.044787354158825746, | |
| "grad_norm": 1.494940161705017, | |
| "learning_rate": 4.8904183329736596e-05, | |
| "loss": 3.3123, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.04497553631915695, | |
| "grad_norm": 1.6059819459915161, | |
| "learning_rate": 4.853900209597903e-05, | |
| "loss": 3.5323, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.04516371847948814, | |
| "grad_norm": 1.5235280990600586, | |
| "learning_rate": 4.817389884711705e-05, | |
| "loss": 3.4477, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.045351900639819344, | |
| "grad_norm": 1.532889485359192, | |
| "learning_rate": 4.7808893071569374e-05, | |
| "loss": 3.4776, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.045540082800150546, | |
| "grad_norm": 1.5243102312088013, | |
| "learning_rate": 4.744400425255165e-05, | |
| "loss": 3.4302, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04572826496048175, | |
| "grad_norm": 1.4719178676605225, | |
| "learning_rate": 4.707925186703671e-05, | |
| "loss": 3.2717, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04591644712081295, | |
| "grad_norm": 1.5308884382247925, | |
| "learning_rate": 4.671465538471486e-05, | |
| "loss": 3.5024, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.04610462928114415, | |
| "grad_norm": 1.5471347570419312, | |
| "learning_rate": 4.6350234266954626e-05, | |
| "loss": 3.3279, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.046292811441475346, | |
| "grad_norm": 1.4882664680480957, | |
| "learning_rate": 4.598600796576395e-05, | |
| "loss": 3.2842, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.04648099360180655, | |
| "grad_norm": 1.546363115310669, | |
| "learning_rate": 4.562199592275188e-05, | |
| "loss": 3.7441, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.04666917576213775, | |
| "grad_norm": 1.5140584707260132, | |
| "learning_rate": 4.5258217568090876e-05, | |
| "loss": 3.337, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.04685735792246895, | |
| "grad_norm": 1.894798755645752, | |
| "learning_rate": 4.4894692319479544e-05, | |
| "loss": 3.2422, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04704554008280015, | |
| "grad_norm": 2.647658109664917, | |
| "learning_rate": 4.4531439581106295e-05, | |
| "loss": 3.1319, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.047233722243131354, | |
| "grad_norm": 1.497327446937561, | |
| "learning_rate": 4.4168478742613506e-05, | |
| "loss": 3.3551, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04742190440346255, | |
| "grad_norm": 1.5717421770095825, | |
| "learning_rate": 4.38058291780626e-05, | |
| "loss": 3.7432, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.04761008656379375, | |
| "grad_norm": 1.4942125082015991, | |
| "learning_rate": 4.3443510244899864e-05, | |
| "loss": 3.4546, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04779826872412495, | |
| "grad_norm": 1.5224213600158691, | |
| "learning_rate": 4.308154128292318e-05, | |
| "loss": 3.4569, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.047986450884456154, | |
| "grad_norm": 3.2993710041046143, | |
| "learning_rate": 4.271994161324977e-05, | |
| "loss": 3.4279, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.048174633044787356, | |
| "grad_norm": 1.776840090751648, | |
| "learning_rate": 4.235873053728475e-05, | |
| "loss": 3.4157, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.04836281520511856, | |
| "grad_norm": 1.5063600540161133, | |
| "learning_rate": 4.199792733569107e-05, | |
| "loss": 3.4676, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04855099736544975, | |
| "grad_norm": 1.5720107555389404, | |
| "learning_rate": 4.163755126736012e-05, | |
| "loss": 3.5201, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.048739179525780954, | |
| "grad_norm": 1.5020864009857178, | |
| "learning_rate": 4.127762156838389e-05, | |
| "loss": 3.382, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.048927361686112156, | |
| "grad_norm": 1.516310214996338, | |
| "learning_rate": 4.0918157451028185e-05, | |
| "loss": 3.4377, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.04911554384644336, | |
| "grad_norm": 1.50225031375885, | |
| "learning_rate": 4.055917810270698e-05, | |
| "loss": 3.4051, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.04930372600677456, | |
| "grad_norm": 1.5030755996704102, | |
| "learning_rate": 4.020070268495843e-05, | |
| "loss": 3.3375, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04949190816710576, | |
| "grad_norm": 1.4907896518707275, | |
| "learning_rate": 3.9842750332421896e-05, | |
| "loss": 3.4886, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.049680090327436956, | |
| "grad_norm": 1.4596471786499023, | |
| "learning_rate": 3.94853401518167e-05, | |
| "loss": 3.2683, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04986827248776816, | |
| "grad_norm": 1.4404525756835938, | |
| "learning_rate": 3.9128491220922156e-05, | |
| "loss": 3.3169, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.05005645464809936, | |
| "grad_norm": 1.473186731338501, | |
| "learning_rate": 3.877222258755935e-05, | |
| "loss": 3.1098, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.05024463680843056, | |
| "grad_norm": 1.5433677434921265, | |
| "learning_rate": 3.8416553268574285e-05, | |
| "loss": 3.4896, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.05043281896876176, | |
| "grad_norm": 1.4866740703582764, | |
| "learning_rate": 3.80615022488229e-05, | |
| "loss": 3.4388, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.050621001129092964, | |
| "grad_norm": 1.5249758958816528, | |
| "learning_rate": 3.770708848015768e-05, | |
| "loss": 3.3572, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.050809183289424166, | |
| "grad_norm": 1.5038138628005981, | |
| "learning_rate": 3.735333088041596e-05, | |
| "loss": 3.3581, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.05099736544975536, | |
| "grad_norm": 1.479463815689087, | |
| "learning_rate": 3.7000248332410304e-05, | |
| "loss": 3.2485, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.05118554761008656, | |
| "grad_norm": 1.5273246765136719, | |
| "learning_rate": 3.664785968292036e-05, | |
| "loss": 3.3607, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.051373729770417764, | |
| "grad_norm": 1.501825213432312, | |
| "learning_rate": 3.629618374168711e-05, | |
| "loss": 3.2803, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.051561911930748966, | |
| "grad_norm": 1.472866415977478, | |
| "learning_rate": 3.594523928040859e-05, | |
| "loss": 3.274, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.05175009409108017, | |
| "grad_norm": 1.490519404411316, | |
| "learning_rate": 3.5595045031738125e-05, | |
| "loss": 3.3221, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.05193827625141137, | |
| "grad_norm": 1.520186185836792, | |
| "learning_rate": 3.5245619688284274e-05, | |
| "loss": 3.4458, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.052126458411742564, | |
| "grad_norm": 1.5064996480941772, | |
| "learning_rate": 3.4896981901613104e-05, | |
| "loss": 3.3529, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.052314640572073766, | |
| "grad_norm": 1.519168734550476, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 3.6177, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.05250282273240497, | |
| "grad_norm": 1.5575183629989624, | |
| "learning_rate": 3.420214339369944e-05, | |
| "loss": 3.4754, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.05269100489273617, | |
| "grad_norm": 1.5328855514526367, | |
| "learning_rate": 3.38559797614277e-05, | |
| "loss": 3.2606, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.05287918705306737, | |
| "grad_norm": 1.585930347442627, | |
| "learning_rate": 3.351067786190038e-05, | |
| "loss": 3.3014, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.05306736921339857, | |
| "grad_norm": 1.5054539442062378, | |
| "learning_rate": 3.316625612658315e-05, | |
| "loss": 3.3176, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.05325555137372977, | |
| "grad_norm": 1.5619922876358032, | |
| "learning_rate": 3.282273293996027e-05, | |
| "loss": 3.4408, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.05344373353406097, | |
| "grad_norm": 1.470646858215332, | |
| "learning_rate": 3.248012663855353e-05, | |
| "loss": 3.2584, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.05363191569439217, | |
| "grad_norm": 1.5209406614303589, | |
| "learning_rate": 3.2138455509943366e-05, | |
| "loss": 3.3224, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.05382009785472337, | |
| "grad_norm": 1.4932469129562378, | |
| "learning_rate": 3.179773779179267e-05, | |
| "loss": 3.3588, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.054008280015054574, | |
| "grad_norm": 1.4727731943130493, | |
| "learning_rate": 3.145799167087342e-05, | |
| "loss": 3.2095, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.054196462175385776, | |
| "grad_norm": 1.501749038696289, | |
| "learning_rate": 3.111923528209577e-05, | |
| "loss": 3.3637, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.05438464433571697, | |
| "grad_norm": 1.5096619129180908, | |
| "learning_rate": 3.078148670754022e-05, | |
| "loss": 3.3815, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.05457282649604817, | |
| "grad_norm": 1.500936508178711, | |
| "learning_rate": 3.0444763975492208e-05, | |
| "loss": 3.1945, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.054761008656379374, | |
| "grad_norm": 1.4544003009796143, | |
| "learning_rate": 3.0109085059480017e-05, | |
| "loss": 3.3647, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.054949190816710576, | |
| "grad_norm": 1.552875280380249, | |
| "learning_rate": 2.977446787731532e-05, | |
| "loss": 3.39, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.05513737297704178, | |
| "grad_norm": 1.5364927053451538, | |
| "learning_rate": 2.944093029013664e-05, | |
| "loss": 3.4203, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.05532555513737298, | |
| "grad_norm": 1.5016520023345947, | |
| "learning_rate": 2.910849010145617e-05, | |
| "loss": 3.3805, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.055513737297704174, | |
| "grad_norm": 1.5016636848449707, | |
| "learning_rate": 2.8777165056209256e-05, | |
| "loss": 3.1924, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.055701919458035376, | |
| "grad_norm": 1.5240119695663452, | |
| "learning_rate": 2.8446972839807384e-05, | |
| "loss": 3.5986, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.05589010161836658, | |
| "grad_norm": 1.5180970430374146, | |
| "learning_rate": 2.8117931077194065e-05, | |
| "loss": 3.3258, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.05607828377869778, | |
| "grad_norm": 1.465223789215088, | |
| "learning_rate": 2.7790057331904117e-05, | |
| "loss": 3.2024, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.05626646593902898, | |
| "grad_norm": 1.7016164064407349, | |
| "learning_rate": 2.746336910512606e-05, | |
| "loss": 3.3025, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.05645464809936018, | |
| "grad_norm": 2.5566675662994385, | |
| "learning_rate": 2.7137883834768073e-05, | |
| "loss": 3.1934, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05645464809936018, | |
| "eval_loss": 0.8374229073524475, | |
| "eval_runtime": 982.5965, | |
| "eval_samples_per_second": 9.109, | |
| "eval_steps_per_second": 2.278, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.056642830259691385, | |
| "grad_norm": 1.5962673425674438, | |
| "learning_rate": 2.6813618894527138e-05, | |
| "loss": 3.5087, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.05683101242002258, | |
| "grad_norm": 1.5255632400512695, | |
| "learning_rate": 2.6490591592961578e-05, | |
| "loss": 3.4712, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.05701919458035378, | |
| "grad_norm": 1.542389154434204, | |
| "learning_rate": 2.6168819172567392e-05, | |
| "loss": 3.2852, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.05720737674068498, | |
| "grad_norm": 1.664982557296753, | |
| "learning_rate": 2.5848318808857606e-05, | |
| "loss": 3.5038, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.057395558901016185, | |
| "grad_norm": 1.4651890993118286, | |
| "learning_rate": 2.5529107609445733e-05, | |
| "loss": 3.1837, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.057583741061347386, | |
| "grad_norm": 1.486304521560669, | |
| "learning_rate": 2.521120261313241e-05, | |
| "loss": 3.2311, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.05777192322167859, | |
| "grad_norm": 1.4881970882415771, | |
| "learning_rate": 2.4894620788996037e-05, | |
| "loss": 3.419, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.05796010538200978, | |
| "grad_norm": 1.4943639039993286, | |
| "learning_rate": 2.457937903548695e-05, | |
| "loss": 3.3085, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.058148287542340985, | |
| "grad_norm": 1.5298432111740112, | |
| "learning_rate": 2.426549417952542e-05, | |
| "loss": 3.608, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.058336469702672186, | |
| "grad_norm": 1.5149232149124146, | |
| "learning_rate": 2.3952982975603496e-05, | |
| "loss": 3.3398, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05852465186300339, | |
| "grad_norm": 1.492951512336731, | |
| "learning_rate": 2.3641862104890595e-05, | |
| "loss": 3.3529, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.05871283402333459, | |
| "grad_norm": 1.448451280593872, | |
| "learning_rate": 2.3332148174343254e-05, | |
| "loss": 3.2021, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.05890101618366579, | |
| "grad_norm": 1.4983866214752197, | |
| "learning_rate": 2.3023857715818532e-05, | |
| "loss": 3.3579, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.059089198343996986, | |
| "grad_norm": 1.5230140686035156, | |
| "learning_rate": 2.2717007185191674e-05, | |
| "loss": 3.4931, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05927738050432819, | |
| "grad_norm": 1.5284234285354614, | |
| "learning_rate": 2.24116129614777e-05, | |
| "loss": 3.4331, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05946556266465939, | |
| "grad_norm": 1.4961637258529663, | |
| "learning_rate": 2.2107691345957133e-05, | |
| "loss": 3.3148, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05965374482499059, | |
| "grad_norm": 1.547951340675354, | |
| "learning_rate": 2.1805258561305862e-05, | |
| "loss": 3.3194, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05984192698532179, | |
| "grad_norm": 1.5121986865997314, | |
| "learning_rate": 2.1504330750729186e-05, | |
| "loss": 3.4504, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.060030109145652995, | |
| "grad_norm": 1.51499605178833, | |
| "learning_rate": 2.120492397710022e-05, | |
| "loss": 3.4349, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.06021829130598419, | |
| "grad_norm": 1.520011305809021, | |
| "learning_rate": 2.090705422210237e-05, | |
| "loss": 3.3619, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.06040647346631539, | |
| "grad_norm": 1.5512540340423584, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.38, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.06059465562664659, | |
| "grad_norm": 1.463521122932434, | |
| "learning_rate": 2.0315989283671473e-05, | |
| "loss": 3.1915, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.060782837786977795, | |
| "grad_norm": 1.5051774978637695, | |
| "learning_rate": 2.0022825650001387e-05, | |
| "loss": 3.4725, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.060971019947308996, | |
| "grad_norm": 1.4965063333511353, | |
| "learning_rate": 1.9731262132804274e-05, | |
| "loss": 3.3025, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.0611592021076402, | |
| "grad_norm": 1.4896841049194336, | |
| "learning_rate": 1.9441314295107537e-05, | |
| "loss": 3.4046, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.06134738426797139, | |
| "grad_norm": 1.4985949993133545, | |
| "learning_rate": 1.9152997613697183e-05, | |
| "loss": 3.4278, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.061535566428302595, | |
| "grad_norm": 1.5722819566726685, | |
| "learning_rate": 1.8866327478291546e-05, | |
| "loss": 3.4424, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.061723748588633796, | |
| "grad_norm": 1.457684874534607, | |
| "learning_rate": 1.8581319190720035e-05, | |
| "loss": 3.175, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.061911930748965, | |
| "grad_norm": 1.501202940940857, | |
| "learning_rate": 1.8297987964106115e-05, | |
| "loss": 3.2706, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.0621001129092962, | |
| "grad_norm": 1.4955599308013916, | |
| "learning_rate": 1.801634892205545e-05, | |
| "loss": 3.3183, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0622882950696274, | |
| "grad_norm": 1.52448308467865, | |
| "learning_rate": 1.7736417097848506e-05, | |
| "loss": 3.3863, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.0624764772299586, | |
| "grad_norm": 1.5245345830917358, | |
| "learning_rate": 1.7458207433638223e-05, | |
| "loss": 3.4718, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.0626646593902898, | |
| "grad_norm": 1.5418884754180908, | |
| "learning_rate": 1.718173477965236e-05, | |
| "loss": 3.3681, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.062852841550621, | |
| "grad_norm": 1.487151861190796, | |
| "learning_rate": 1.6907013893400837e-05, | |
| "loss": 3.3089, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.0630410237109522, | |
| "grad_norm": 1.4714837074279785, | |
| "learning_rate": 1.6634059438888033e-05, | |
| "loss": 3.181, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.0632292058712834, | |
| "grad_norm": 1.5032023191452026, | |
| "learning_rate": 1.636288598583e-05, | |
| "loss": 3.3265, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0634173880316146, | |
| "grad_norm": 1.5267105102539062, | |
| "learning_rate": 1.6093508008876857e-05, | |
| "loss": 3.1744, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.0636055701919458, | |
| "grad_norm": 1.517006516456604, | |
| "learning_rate": 1.5825939886840037e-05, | |
| "loss": 3.2575, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.06379375235227701, | |
| "grad_norm": 1.4955118894577026, | |
| "learning_rate": 1.5560195901924894e-05, | |
| "loss": 3.2745, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.06398193451260821, | |
| "grad_norm": 1.7153363227844238, | |
| "learning_rate": 1.5296290238968303e-05, | |
| "loss": 3.4614, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.06417011667293941, | |
| "grad_norm": 1.5209360122680664, | |
| "learning_rate": 1.50342369846815e-05, | |
| "loss": 3.3276, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.0643582988332706, | |
| "grad_norm": 1.5184190273284912, | |
| "learning_rate": 1.4774050126898164e-05, | |
| "loss": 3.3869, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.0645464809936018, | |
| "grad_norm": 1.5385315418243408, | |
| "learning_rate": 1.451574355382776e-05, | |
| "loss": 3.4677, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.064734663153933, | |
| "grad_norm": 1.4663329124450684, | |
| "learning_rate": 1.425933105331429e-05, | |
| "loss": 3.3069, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.0649228453142642, | |
| "grad_norm": 1.5196608304977417, | |
| "learning_rate": 1.4004826312100216e-05, | |
| "loss": 3.3985, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.0651110274745954, | |
| "grad_norm": 1.5887420177459717, | |
| "learning_rate": 1.3752242915095992e-05, | |
| "loss": 3.309, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.06529920963492661, | |
| "grad_norm": 1.5067335367202759, | |
| "learning_rate": 1.3501594344654884e-05, | |
| "loss": 3.186, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.06548739179525781, | |
| "grad_norm": 1.4992436170578003, | |
| "learning_rate": 1.3252893979853304e-05, | |
| "loss": 3.3428, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.06567557395558901, | |
| "grad_norm": 1.6664701700210571, | |
| "learning_rate": 1.3006155095776707e-05, | |
| "loss": 3.4061, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.06586375611592021, | |
| "grad_norm": 2.502802848815918, | |
| "learning_rate": 1.2761390862810907e-05, | |
| "loss": 3.0303, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.06605193827625142, | |
| "grad_norm": 1.500586748123169, | |
| "learning_rate": 1.2518614345939212e-05, | |
| "loss": 3.4372, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.06624012043658262, | |
| "grad_norm": 1.5125335454940796, | |
| "learning_rate": 1.227783850404487e-05, | |
| "loss": 3.3834, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.06642830259691382, | |
| "grad_norm": 1.5365911722183228, | |
| "learning_rate": 1.2039076189219517e-05, | |
| "loss": 3.4872, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.066616484757245, | |
| "grad_norm": 1.5221909284591675, | |
| "learning_rate": 1.1802340146077045e-05, | |
| "loss": 3.3993, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.06680466691757621, | |
| "grad_norm": 1.5232350826263428, | |
| "learning_rate": 1.1567643011073392e-05, | |
| "loss": 3.5355, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.06699284907790741, | |
| "grad_norm": 1.5335094928741455, | |
| "learning_rate": 1.1334997311832002e-05, | |
| "loss": 3.4007, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.06718103123823861, | |
| "grad_norm": 1.4804530143737793, | |
| "learning_rate": 1.1104415466475087e-05, | |
| "loss": 3.3991, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.06736921339856981, | |
| "grad_norm": 1.4774835109710693, | |
| "learning_rate": 1.0875909782960886e-05, | |
| "loss": 3.3081, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.06755739555890101, | |
| "grad_norm": 1.507690191268921, | |
| "learning_rate": 1.0649492458426564e-05, | |
| "loss": 3.5331, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.06774557771923222, | |
| "grad_norm": 1.4900294542312622, | |
| "learning_rate": 1.0425175578537299e-05, | |
| "loss": 3.4034, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.06793375987956342, | |
| "grad_norm": 1.536511778831482, | |
| "learning_rate": 1.020297111684101e-05, | |
| "loss": 3.3285, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.06812194203989462, | |
| "grad_norm": 1.4938386678695679, | |
| "learning_rate": 9.98289093412938e-06, | |
| "loss": 3.3401, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.06831012420022582, | |
| "grad_norm": 1.498859167098999, | |
| "learning_rate": 9.764946777804646e-06, | |
| "loss": 3.4705, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.06849830636055702, | |
| "grad_norm": 1.505251407623291, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 3.3769, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.06868648852088823, | |
| "grad_norm": 1.5352188348770142, | |
| "learning_rate": 9.335512963221732e-06, | |
| "loss": 3.4495, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.06887467068121943, | |
| "grad_norm": 1.5243737697601318, | |
| "learning_rate": 9.124046227208082e-06, | |
| "loss": 3.3756, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.06906285284155061, | |
| "grad_norm": 1.4866875410079956, | |
| "learning_rate": 8.914761360846869e-06, | |
| "loss": 3.2548, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.06925103500188182, | |
| "grad_norm": 1.70474374294281, | |
| "learning_rate": 8.707669535309793e-06, | |
| "loss": 3.3533, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.06943921716221302, | |
| "grad_norm": 1.4734491109848022, | |
| "learning_rate": 8.502781804708826e-06, | |
| "loss": 3.3556, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.06962739932254422, | |
| "grad_norm": 1.5178672075271606, | |
| "learning_rate": 8.30010910550611e-06, | |
| "loss": 3.4011, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.06981558148287542, | |
| "grad_norm": 1.5426009893417358, | |
| "learning_rate": 8.09966225593024e-06, | |
| "loss": 3.3262, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.07000376364320662, | |
| "grad_norm": 1.488108515739441, | |
| "learning_rate": 7.901451955398792e-06, | |
| "loss": 3.3038, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.07019194580353783, | |
| "grad_norm": 1.5539402961730957, | |
| "learning_rate": 7.705488783947202e-06, | |
| "loss": 3.4263, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.07038012796386903, | |
| "grad_norm": 1.5011879205703735, | |
| "learning_rate": 7.511783201664052e-06, | |
| "loss": 3.1954, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.07056831012420023, | |
| "grad_norm": 1.5092988014221191, | |
| "learning_rate": 7.320345548132679e-06, | |
| "loss": 3.328, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.07075649228453143, | |
| "grad_norm": 1.498982310295105, | |
| "learning_rate": 7.131186041879357e-06, | |
| "loss": 3.2402, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.07094467444486263, | |
| "grad_norm": 1.5231132507324219, | |
| "learning_rate": 6.944314779827749e-06, | |
| "loss": 3.4769, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.07113285660519383, | |
| "grad_norm": 1.5165536403656006, | |
| "learning_rate": 6.759741736760061e-06, | |
| "loss": 3.4978, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.07132103876552502, | |
| "grad_norm": 1.531844973564148, | |
| "learning_rate": 6.577476764784546e-06, | |
| "loss": 3.3847, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.07150922092585622, | |
| "grad_norm": 1.4768074750900269, | |
| "learning_rate": 6.397529592809614e-06, | |
| "loss": 3.1253, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.07169740308618743, | |
| "grad_norm": 1.5122772455215454, | |
| "learning_rate": 6.219909826024589e-06, | |
| "loss": 3.3401, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.07188558524651863, | |
| "grad_norm": 1.5160380601882935, | |
| "learning_rate": 6.0446269453868945e-06, | |
| "loss": 3.3382, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.07207376740684983, | |
| "grad_norm": 1.5226458311080933, | |
| "learning_rate": 5.871690307116107e-06, | |
| "loss": 3.431, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.07226194956718103, | |
| "grad_norm": 1.4903466701507568, | |
| "learning_rate": 5.701109142194422e-06, | |
| "loss": 3.2707, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.07245013172751223, | |
| "grad_norm": 1.5170961618423462, | |
| "learning_rate": 5.532892555874059e-06, | |
| "loss": 3.3027, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.07263831388784343, | |
| "grad_norm": 1.5108839273452759, | |
| "learning_rate": 5.3670495271910925e-06, | |
| "loss": 3.3152, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.07282649604817464, | |
| "grad_norm": 1.51710045337677, | |
| "learning_rate": 5.203588908486279e-06, | |
| "loss": 3.2884, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.07301467820850584, | |
| "grad_norm": 1.5793352127075195, | |
| "learning_rate": 5.042519424932513e-06, | |
| "loss": 3.5139, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.07320286036883704, | |
| "grad_norm": 1.481472373008728, | |
| "learning_rate": 4.883849674069058e-06, | |
| "loss": 3.2427, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.07339104252916824, | |
| "grad_norm": 1.5381510257720947, | |
| "learning_rate": 4.727588125342669e-06, | |
| "loss": 3.3485, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.07357922468949944, | |
| "grad_norm": 1.508355975151062, | |
| "learning_rate": 4.573743119655516e-06, | |
| "loss": 3.295, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.07376740684983063, | |
| "grad_norm": 1.527873158454895, | |
| "learning_rate": 4.422322868919937e-06, | |
| "loss": 3.2918, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.07395558901016183, | |
| "grad_norm": 1.5130548477172852, | |
| "learning_rate": 4.273335455620097e-06, | |
| "loss": 3.3859, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.07414377117049303, | |
| "grad_norm": 1.5391393899917603, | |
| "learning_rate": 4.126788832380629e-06, | |
| "loss": 3.3711, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.07433195333082424, | |
| "grad_norm": 1.5213444232940674, | |
| "learning_rate": 3.982690821542035e-06, | |
| "loss": 3.4224, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.07452013549115544, | |
| "grad_norm": 1.4749212265014648, | |
| "learning_rate": 3.8410491147432395e-06, | |
| "loss": 3.1882, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.07470831765148664, | |
| "grad_norm": 1.4860862493515015, | |
| "learning_rate": 3.7018712725109926e-06, | |
| "loss": 3.2488, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.07489649981181784, | |
| "grad_norm": 1.5500200986862183, | |
| "learning_rate": 3.5651647238562904e-06, | |
| "loss": 3.3369, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.07508468197214904, | |
| "grad_norm": 1.8019529581069946, | |
| "learning_rate": 3.430936765877857e-06, | |
| "loss": 3.1557, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.07527286413248024, | |
| "grad_norm": 2.5598955154418945, | |
| "learning_rate": 3.299194563372604e-06, | |
| "loss": 3.03, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07546104629281145, | |
| "grad_norm": 1.5204044580459595, | |
| "learning_rate": 3.1699451484532463e-06, | |
| "loss": 3.4419, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.07564922845314265, | |
| "grad_norm": 1.4903305768966675, | |
| "learning_rate": 3.0431954201728784e-06, | |
| "loss": 3.3295, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.07583741061347385, | |
| "grad_norm": 1.5254848003387451, | |
| "learning_rate": 2.9189521441567726e-06, | |
| "loss": 3.4107, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.07602559277380504, | |
| "grad_norm": 1.5176067352294922, | |
| "learning_rate": 2.797221952241219e-06, | |
| "loss": 3.3387, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.07621377493413624, | |
| "grad_norm": 1.495973825454712, | |
| "learning_rate": 2.6780113421195298e-06, | |
| "loss": 3.3161, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.07640195709446744, | |
| "grad_norm": 1.5124212503433228, | |
| "learning_rate": 2.561326676995218e-06, | |
| "loss": 3.3538, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.07659013925479864, | |
| "grad_norm": 1.4950138330459595, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 3.363, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.07677832141512984, | |
| "grad_norm": 1.5359971523284912, | |
| "learning_rate": 2.3355599600729915e-06, | |
| "loss": 3.4671, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.07696650357546105, | |
| "grad_norm": 1.4483258724212646, | |
| "learning_rate": 2.2264899592121744e-06, | |
| "loss": 3.1561, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.07715468573579225, | |
| "grad_norm": 1.483723521232605, | |
| "learning_rate": 2.1199700045797077e-06, | |
| "loss": 3.3924, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.07734286789612345, | |
| "grad_norm": 1.4614593982696533, | |
| "learning_rate": 2.0160057819794466e-06, | |
| "loss": 3.2969, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.07753105005645465, | |
| "grad_norm": 1.4894236326217651, | |
| "learning_rate": 1.9146028407958484e-06, | |
| "loss": 3.3376, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.07771923221678585, | |
| "grad_norm": 1.501068115234375, | |
| "learning_rate": 1.8157665936977263e-06, | |
| "loss": 3.3462, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.07790741437711705, | |
| "grad_norm": 1.4664418697357178, | |
| "learning_rate": 1.7195023163493252e-06, | |
| "loss": 3.1851, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.07809559653744826, | |
| "grad_norm": 1.5121281147003174, | |
| "learning_rate": 1.6258151471287396e-06, | |
| "loss": 3.3562, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.07828377869777944, | |
| "grad_norm": 1.5005747079849243, | |
| "learning_rate": 1.5347100868536246e-06, | |
| "loss": 3.276, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.07847196085811065, | |
| "grad_norm": 1.5016510486602783, | |
| "learning_rate": 1.4461919985142735e-06, | |
| "loss": 3.1685, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.07866014301844185, | |
| "grad_norm": 1.490427017211914, | |
| "learning_rate": 1.3602656070140275e-06, | |
| "loss": 3.3398, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.07884832517877305, | |
| "grad_norm": 1.5081381797790527, | |
| "learning_rate": 1.27693549891707e-06, | |
| "loss": 3.378, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.07903650733910425, | |
| "grad_norm": 1.5135114192962646, | |
| "learning_rate": 1.196206122203647e-06, | |
| "loss": 3.2584, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.07922468949943545, | |
| "grad_norm": 1.503239631652832, | |
| "learning_rate": 1.1180817860325599e-06, | |
| "loss": 3.3742, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.07941287165976665, | |
| "grad_norm": 1.5126641988754272, | |
| "learning_rate": 1.0425666605112517e-06, | |
| "loss": 3.4263, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.07960105382009786, | |
| "grad_norm": 1.5308884382247925, | |
| "learning_rate": 9.696647764731337e-07, | |
| "loss": 3.3315, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.07978923598042906, | |
| "grad_norm": 1.5186878442764282, | |
| "learning_rate": 8.993800252624862e-07, | |
| "loss": 3.3336, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.07997741814076026, | |
| "grad_norm": 1.5928887128829956, | |
| "learning_rate": 8.317161585266964e-07, | |
| "loss": 3.641, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.08016560030109146, | |
| "grad_norm": 1.5158437490463257, | |
| "learning_rate": 7.666767880160464e-07, | |
| "loss": 3.4028, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.08035378246142266, | |
| "grad_norm": 1.5123188495635986, | |
| "learning_rate": 7.042653853909064e-07, | |
| "loss": 3.4125, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.08054196462175386, | |
| "grad_norm": 1.4982678890228271, | |
| "learning_rate": 6.444852820364222e-07, | |
| "loss": 3.2116, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.08073014678208505, | |
| "grad_norm": 1.5639753341674805, | |
| "learning_rate": 5.87339668884701e-07, | |
| "loss": 3.1289, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.08091832894241625, | |
| "grad_norm": 1.572089672088623, | |
| "learning_rate": 5.328315962444874e-07, | |
| "loss": 3.3213, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.08110651110274746, | |
| "grad_norm": 1.5132265090942383, | |
| "learning_rate": 4.809639736383431e-07, | |
| "loss": 3.3827, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.08129469326307866, | |
| "grad_norm": 1.4733517169952393, | |
| "learning_rate": 4.317395696473214e-07, | |
| "loss": 3.2562, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.08148287542340986, | |
| "grad_norm": 1.548933982849121, | |
| "learning_rate": 3.851610117632354e-07, | |
| "loss": 3.4471, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.08167105758374106, | |
| "grad_norm": 1.4967180490493774, | |
| "learning_rate": 3.4123078624834216e-07, | |
| "loss": 3.0655, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.08185923974407226, | |
| "grad_norm": 1.6242047548294067, | |
| "learning_rate": 2.9995123800270476e-07, | |
| "loss": 3.2837, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.08204742190440346, | |
| "grad_norm": 1.4935169219970703, | |
| "learning_rate": 2.613245704389644e-07, | |
| "loss": 3.3263, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.08223560406473467, | |
| "grad_norm": 1.4887654781341553, | |
| "learning_rate": 2.2535284536476242e-07, | |
| "loss": 3.1296, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.08242378622506587, | |
| "grad_norm": 1.511015772819519, | |
| "learning_rate": 1.920379828726726e-07, | |
| "loss": 3.3212, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.08261196838539707, | |
| "grad_norm": 1.5536808967590332, | |
| "learning_rate": 1.6138176123770554e-07, | |
| "loss": 3.5203, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.08280015054572827, | |
| "grad_norm": 1.5585739612579346, | |
| "learning_rate": 1.333858168224178e-07, | |
| "loss": 3.2727, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.08298833270605946, | |
| "grad_norm": 1.5475882291793823, | |
| "learning_rate": 1.0805164398952072e-07, | |
| "loss": 3.4168, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.08317651486639066, | |
| "grad_norm": 1.5009870529174805, | |
| "learning_rate": 8.53805950221498e-08, | |
| "loss": 3.3267, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.08336469702672186, | |
| "grad_norm": 1.5358216762542725, | |
| "learning_rate": 6.537388005167233e-08, | |
| "loss": 3.4914, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.08355287918705306, | |
| "grad_norm": 1.5157291889190674, | |
| "learning_rate": 4.8032566993089225e-08, | |
| "loss": 3.3647, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.08374106134738427, | |
| "grad_norm": 1.5044530630111694, | |
| "learning_rate": 3.3357581488030475e-08, | |
| "loss": 3.2696, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.08392924350771547, | |
| "grad_norm": 1.5259320735931396, | |
| "learning_rate": 2.134970685536697e-08, | |
| "loss": 3.3305, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.08411742566804667, | |
| "grad_norm": 1.4801864624023438, | |
| "learning_rate": 1.200958404936059e-08, | |
| "loss": 3.1884, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.08430560782837787, | |
| "grad_norm": 1.4972193241119385, | |
| "learning_rate": 5.337711625497121e-09, | |
| "loss": 3.3929, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.08449378998870907, | |
| "grad_norm": 1.7666032314300537, | |
| "learning_rate": 1.3344457138297906e-09, | |
| "loss": 3.4298, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.08468197214904027, | |
| "grad_norm": 2.5446274280548096, | |
| "learning_rate": 0.0, | |
| "loss": 2.8081, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08468197214904027, | |
| "eval_loss": 0.8302884697914124, | |
| "eval_runtime": 982.1123, | |
| "eval_samples_per_second": 9.113, | |
| "eval_steps_per_second": 2.279, | |
| "step": 450 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 150, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 2, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0272928176694886e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |