| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.6635859519408502, | |
| "eval_steps": 80, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018484288354898336, | |
| "grad_norm": 40.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.321, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036968576709796672, | |
| "grad_norm": 36.25, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 1.9522, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005545286506469501, | |
| "grad_norm": 16.25, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 1.0777, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0073937153419593345, | |
| "grad_norm": 35.25, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 2.1435, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009242144177449169, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 1.1501, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011090573012939002, | |
| "grad_norm": 35.75, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.0715, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012939001848428836, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.9734, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014787430683918669, | |
| "grad_norm": 16.625, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 1.1015, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0166358595194085, | |
| "grad_norm": 16.625, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.1099, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018484288354898338, | |
| "grad_norm": 59.0, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 3.4796, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02033271719038817, | |
| "grad_norm": 20.75, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.2906, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.022181146025878003, | |
| "grad_norm": 45.75, | |
| "learning_rate": 1.1e-06, | |
| "loss": 2.696, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.024029574861367836, | |
| "grad_norm": 13.5, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.9347, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025878003696857672, | |
| "grad_norm": 32.5, | |
| "learning_rate": 1.3e-06, | |
| "loss": 2.1339, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.027726432532347505, | |
| "grad_norm": 72.5, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 4.4045, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029574861367837338, | |
| "grad_norm": 54.0, | |
| "learning_rate": 1.5e-06, | |
| "loss": 3.5853, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.031423290203327174, | |
| "grad_norm": 18.75, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 1.3962, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.033271719038817, | |
| "grad_norm": 33.75, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 2.0172, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03512014787430684, | |
| "grad_norm": 15.0, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 1.1185, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.036968576709796676, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 1.047, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.038817005545286505, | |
| "grad_norm": 32.75, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.5555, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04066543438077634, | |
| "grad_norm": 18.375, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 1.6626, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04251386321626617, | |
| "grad_norm": 12.875, | |
| "learning_rate": 2.2e-06, | |
| "loss": 1.0417, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04436229205175601, | |
| "grad_norm": 16.125, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 1.5335, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04621072088724584, | |
| "grad_norm": 17.0, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.7166, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04805914972273567, | |
| "grad_norm": 15.6875, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.6069, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04990757855822551, | |
| "grad_norm": 19.0, | |
| "learning_rate": 2.6e-06, | |
| "loss": 2.0973, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.051756007393715345, | |
| "grad_norm": 24.25, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 2.6387, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.053604436229205174, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.8924, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05545286506469501, | |
| "grad_norm": 22.625, | |
| "learning_rate": 2.9e-06, | |
| "loss": 2.3964, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05730129390018484, | |
| "grad_norm": 22.0, | |
| "learning_rate": 3e-06, | |
| "loss": 2.6245, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.059149722735674676, | |
| "grad_norm": 17.125, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 2.1851, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06099815157116451, | |
| "grad_norm": 18.5, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 2.1781, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06284658040665435, | |
| "grad_norm": 20.625, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 2.3016, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06469500924214418, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.8133, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.066543438077634, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.9642, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06839186691312385, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.4513, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07024029574861368, | |
| "grad_norm": 12.25, | |
| "learning_rate": 3.7e-06, | |
| "loss": 1.5899, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07208872458410351, | |
| "grad_norm": 31.875, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 2.9897, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07393715341959335, | |
| "grad_norm": 19.625, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 2.2359, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07578558225508318, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.7837, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07763401109057301, | |
| "grad_norm": 11.0, | |
| "learning_rate": 4.1e-06, | |
| "loss": 1.4583, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07948243992606285, | |
| "grad_norm": 9.625, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 1.4312, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08133086876155268, | |
| "grad_norm": 4.5, | |
| "learning_rate": 4.3e-06, | |
| "loss": 0.8753, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08317929759704251, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.9408, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08502772643253234, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.7432, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08687615526802218, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.7335, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08872458410351201, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.8767, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09057301293900184, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.5655, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09242144177449169, | |
| "grad_norm": 13.125, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 1.7886, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09426987060998152, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.8237, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09611829944547134, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 4.999988416227265e-06, | |
| "loss": 0.8633, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09796672828096119, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 4.999953665016405e-06, | |
| "loss": 0.8165, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09981515711645102, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.99989574668946e-06, | |
| "loss": 0.6063, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10166358595194085, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 4.999814661783161e-06, | |
| "loss": 1.0097, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10351201478743069, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 4.999710411048924e-06, | |
| "loss": 0.7601, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10536044362292052, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.999582995452842e-06, | |
| "loss": 1.3231, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10720887245841035, | |
| "grad_norm": 3.0, | |
| "learning_rate": 4.999432416175676e-06, | |
| "loss": 0.6672, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10905730129390019, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 4.999258674612849e-06, | |
| "loss": 0.7139, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11090573012939002, | |
| "grad_norm": 5.5, | |
| "learning_rate": 4.999061772374426e-06, | |
| "loss": 1.1064, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11275415896487985, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 4.998841711285104e-06, | |
| "loss": 0.6181, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11460258780036968, | |
| "grad_norm": 5.0, | |
| "learning_rate": 4.998598493384193e-06, | |
| "loss": 1.0482, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11645101663585952, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 4.998332120925598e-06, | |
| "loss": 0.8835, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11829944547134935, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 4.998042596377798e-06, | |
| "loss": 0.9458, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12014787430683918, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.997729922423821e-06, | |
| "loss": 0.6227, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12199630314232902, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 4.9973941019612235e-06, | |
| "loss": 0.63, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12384473197781885, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.997035138102057e-06, | |
| "loss": 1.3906, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1256931608133087, | |
| "grad_norm": 6.0, | |
| "learning_rate": 4.99665303417285e-06, | |
| "loss": 0.9835, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12754158964879853, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 4.996247793714565e-06, | |
| "loss": 0.6305, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12939001848428835, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.995819420482572e-06, | |
| "loss": 0.9368, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13123844731977818, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.995367918446613e-06, | |
| "loss": 1.1323, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.133086876155268, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 4.994893291790768e-06, | |
| "loss": 0.8496, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13493530499075784, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 4.994395544913409e-06, | |
| "loss": 0.5263, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1367837338262477, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 4.993874682427165e-06, | |
| "loss": 0.9357, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13863216266173753, | |
| "grad_norm": 4.5, | |
| "learning_rate": 4.993330709158879e-06, | |
| "loss": 0.8272, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14048059149722736, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 4.992763630149562e-06, | |
| "loss": 0.4897, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1423290203327172, | |
| "grad_norm": 7.875, | |
| "learning_rate": 4.992173450654344e-06, | |
| "loss": 1.2429, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14417744916820702, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.9915601761424304e-06, | |
| "loss": 0.8666, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14602587800369685, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 4.990923812297046e-06, | |
| "loss": 1.1015, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1478743068391867, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 4.990264365015386e-06, | |
| "loss": 0.908, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1478743068391867, | |
| "eval_loss": 0.9451218843460083, | |
| "eval_runtime": 19.8706, | |
| "eval_samples_per_second": 5.737, | |
| "eval_steps_per_second": 2.869, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14972273567467653, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.989581840408562e-06, | |
| "loss": 0.5932, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.15157116451016636, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.988876244801541e-06, | |
| "loss": 0.7823, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1534195933456562, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 4.98814758473309e-06, | |
| "loss": 1.0936, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15526802218114602, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 4.987395866955716e-06, | |
| "loss": 1.1613, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15711645101663585, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.986621098435599e-06, | |
| "loss": 1.1484, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1589648798521257, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 4.9858232863525366e-06, | |
| "loss": 0.6921, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.16081330868761554, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.9850024380998655e-06, | |
| "loss": 1.0746, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.16266173752310537, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 4.984158561284401e-06, | |
| "loss": 0.45, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1645101663585952, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 4.9832916637263665e-06, | |
| "loss": 1.005, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16635859519408502, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 4.982401753459317e-06, | |
| "loss": 0.5901, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16820702402957485, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 4.981488838730066e-06, | |
| "loss": 0.9943, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.17005545286506468, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 4.980552927998612e-06, | |
| "loss": 0.561, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.17190388170055454, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 4.979594029938058e-06, | |
| "loss": 0.6962, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.17375231053604437, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 4.978612153434527e-06, | |
| "loss": 0.9293, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1756007393715342, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.977607307587087e-06, | |
| "loss": 1.1104, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17744916820702403, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.976579501707665e-06, | |
| "loss": 0.9912, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17929759704251386, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 4.975528745320953e-06, | |
| "loss": 0.9826, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.18114602587800369, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.974455048164334e-06, | |
| "loss": 0.518, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.18299445471349354, | |
| "grad_norm": 3.5, | |
| "learning_rate": 4.973358420187776e-06, | |
| "loss": 0.4571, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18484288354898337, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.972238871553751e-06, | |
| "loss": 0.7805, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1866913123844732, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.971096412637137e-06, | |
| "loss": 0.4784, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.18853974121996303, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 4.969931054025122e-06, | |
| "loss": 0.4652, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.19038817005545286, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 4.968742806517104e-06, | |
| "loss": 0.6581, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1922365988909427, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.9675316811245955e-06, | |
| "loss": 0.9293, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.19408502772643252, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 4.966297689071117e-06, | |
| "loss": 0.6352, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19593345656192238, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.965040841792095e-06, | |
| "loss": 0.8504, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1977818853974122, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 4.963761150934757e-06, | |
| "loss": 0.5028, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.19963031423290203, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.962458628358021e-06, | |
| "loss": 0.8722, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.20147874306839186, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.9611332861323875e-06, | |
| "loss": 1.1018, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2033271719038817, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 4.959785136539827e-06, | |
| "loss": 0.3881, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20517560073937152, | |
| "grad_norm": 4.375, | |
| "learning_rate": 4.958414192073665e-06, | |
| "loss": 0.8694, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.20702402957486138, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.957020465438471e-06, | |
| "loss": 1.1025, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2088724584103512, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.955603969549934e-06, | |
| "loss": 0.7625, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.21072088724584104, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 4.954164717534748e-06, | |
| "loss": 0.5436, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.21256931608133087, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 4.952702722730486e-06, | |
| "loss": 0.3518, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2144177449168207, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.951217998685481e-06, | |
| "loss": 0.8437, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21626617375231053, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 4.949710559158699e-06, | |
| "loss": 0.9893, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21811460258780038, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.948180418119608e-06, | |
| "loss": 0.9009, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2199630314232902, | |
| "grad_norm": 4.625, | |
| "learning_rate": 4.946627589748053e-06, | |
| "loss": 0.7, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.22181146025878004, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.945052088434123e-06, | |
| "loss": 0.9304, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22365988909426987, | |
| "grad_norm": 5.125, | |
| "learning_rate": 4.943453928778018e-06, | |
| "loss": 0.8282, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2255083179297597, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 4.941833125589912e-06, | |
| "loss": 0.6033, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.22735674676524953, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 4.940189693889819e-06, | |
| "loss": 0.7976, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.22920517560073936, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.938523648907447e-06, | |
| "loss": 0.4154, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.23105360443622922, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.936835006082068e-06, | |
| "loss": 1.0393, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.23290203327171904, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 4.9351237810623655e-06, | |
| "loss": 0.9806, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.23475046210720887, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 4.933389989706292e-06, | |
| "loss": 0.9188, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2365988909426987, | |
| "grad_norm": 3.375, | |
| "learning_rate": 4.931633648080924e-06, | |
| "loss": 0.704, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23844731977818853, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.929854772462312e-06, | |
| "loss": 0.7756, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.24029574861367836, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.9280533793353275e-06, | |
| "loss": 0.7783, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.24214417744916822, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.926229485393513e-06, | |
| "loss": 1.0043, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.24399260628465805, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.924383107538929e-06, | |
| "loss": 1.0273, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24584103512014788, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.922514262881991e-06, | |
| "loss": 0.4076, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2476894639556377, | |
| "grad_norm": 4.875, | |
| "learning_rate": 4.920622968741317e-06, | |
| "loss": 0.973, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24953789279112754, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 4.918709242643563e-06, | |
| "loss": 0.7444, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2513863216266174, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 4.916773102323266e-06, | |
| "loss": 0.7349, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.2532347504621072, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 4.914814565722671e-06, | |
| "loss": 0.9455, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.25508317929759705, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 4.9128336509915746e-06, | |
| "loss": 0.9259, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25693160813308685, | |
| "grad_norm": 2.375, | |
| "learning_rate": 4.910830376487147e-06, | |
| "loss": 0.5286, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2587800369685767, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 4.908804760773771e-06, | |
| "loss": 0.6959, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26062846580406657, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.906756822622865e-06, | |
| "loss": 0.9512, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.26247689463955637, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.904686581012708e-06, | |
| "loss": 0.9307, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2643253234750462, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 4.902594055128267e-06, | |
| "loss": 0.3672, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.266173752310536, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 4.900479264361017e-06, | |
| "loss": 0.8701, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2680221811460259, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 4.8983422283087635e-06, | |
| "loss": 0.8548, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2698706099815157, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.896182966775458e-06, | |
| "loss": 0.9402, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.27171903881700554, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.894001499771015e-06, | |
| "loss": 0.8815, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2735674676524954, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 4.891797847511131e-06, | |
| "loss": 0.9296, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2754158964879852, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 4.889572030417091e-06, | |
| "loss": 1.0949, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.27726432532347506, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.887324069115582e-06, | |
| "loss": 1.1124, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.27911275415896486, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 4.885053984438502e-06, | |
| "loss": 0.9039, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2809611829944547, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.882761797422767e-06, | |
| "loss": 1.1055, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2828096118299446, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.880447529310118e-06, | |
| "loss": 0.7592, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2846580406654344, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 4.878111201546917e-06, | |
| "loss": 0.521, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.28650646950092423, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 4.875752835783957e-06, | |
| "loss": 0.9844, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28835489833641403, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 4.873372453876255e-06, | |
| "loss": 0.5843, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2902033271719039, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.870970077882856e-06, | |
| "loss": 0.6412, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2920517560073937, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 4.8685457300666205e-06, | |
| "loss": 0.7729, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.29390018484288355, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.866099432894023e-06, | |
| "loss": 0.8735, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2957486136783734, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 4.863631209034947e-06, | |
| "loss": 0.4581, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2957486136783734, | |
| "eval_loss": 0.8287791013717651, | |
| "eval_runtime": 19.8734, | |
| "eval_samples_per_second": 5.736, | |
| "eval_steps_per_second": 2.868, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2975970425138632, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 4.861141081362463e-06, | |
| "loss": 0.9825, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.29944547134935307, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 4.858629072952635e-06, | |
| "loss": 0.6221, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.30129390018484287, | |
| "grad_norm": 4.125, | |
| "learning_rate": 4.856095207084287e-06, | |
| "loss": 0.5451, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3031423290203327, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 4.8535395072388e-06, | |
| "loss": 0.7593, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3049907578558225, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.850961997099892e-06, | |
| "loss": 1.0049, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3068391866913124, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 4.848362700553397e-06, | |
| "loss": 0.4303, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.30868761552680224, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 4.845741641687043e-06, | |
| "loss": 0.6331, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.31053604436229204, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 4.843098844790228e-06, | |
| "loss": 0.4142, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3123844731977819, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.840434334353802e-06, | |
| "loss": 1.0196, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3142329020332717, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 4.83774813506983e-06, | |
| "loss": 0.35, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.31608133086876156, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.835040271831371e-06, | |
| "loss": 0.7789, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3179297597042514, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.83231076973224e-06, | |
| "loss": 0.8986, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3197781885397412, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 4.829559654066786e-06, | |
| "loss": 0.7184, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.32162661737523107, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.826786950329646e-06, | |
| "loss": 1.0167, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3234750462107209, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 4.823992684215516e-06, | |
| "loss": 0.5429, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.32532347504621073, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.821176881618912e-06, | |
| "loss": 0.8559, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.32717190388170053, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 4.818339568633926e-06, | |
| "loss": 0.4429, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3290203327171904, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 4.815480771553991e-06, | |
| "loss": 0.7874, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.33086876155268025, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 4.812600516871631e-06, | |
| "loss": 0.708, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.33271719038817005, | |
| "grad_norm": 4.75, | |
| "learning_rate": 4.809698831278217e-06, | |
| "loss": 0.7904, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3345656192236599, | |
| "grad_norm": 2.75, | |
| "learning_rate": 4.806775741663724e-06, | |
| "loss": 0.5476, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3364140480591497, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 4.803831275116477e-06, | |
| "loss": 0.6509, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.33826247689463956, | |
| "grad_norm": 11.125, | |
| "learning_rate": 4.800865458922899e-06, | |
| "loss": 0.9183, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.34011090573012936, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 4.797878320567264e-06, | |
| "loss": 0.8251, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3419593345656192, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.794869887731437e-06, | |
| "loss": 0.9493, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3438077634011091, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 4.79184018829462e-06, | |
| "loss": 1.102, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3456561922365989, | |
| "grad_norm": 7.125, | |
| "learning_rate": 4.7887892503330936e-06, | |
| "loss": 1.0035, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.34750462107208874, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 4.785717102119954e-06, | |
| "loss": 0.4152, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34935304990757854, | |
| "grad_norm": 3.0, | |
| "learning_rate": 4.782623772124854e-06, | |
| "loss": 0.4278, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3512014787430684, | |
| "grad_norm": 5.375, | |
| "learning_rate": 4.779509289013742e-06, | |
| "loss": 1.06, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.35304990757855825, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.776373681648586e-06, | |
| "loss": 0.5845, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.35489833641404805, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.77321697908712e-06, | |
| "loss": 0.7516, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3567467652495379, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 4.77003921058256e-06, | |
| "loss": 0.41, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3585951940850277, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 4.766840405583348e-06, | |
| "loss": 0.8026, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.36044362292051757, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.763620593732867e-06, | |
| "loss": 1.0103, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.36229205175600737, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 4.760379804869172e-06, | |
| "loss": 0.5553, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.36414048059149723, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.757118069024711e-06, | |
| "loss": 1.1879, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3659889094269871, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 4.7538354164260515e-06, | |
| "loss": 0.7766, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3678373382624769, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.750531877493594e-06, | |
| "loss": 0.6487, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 4.747207482841293e-06, | |
| "loss": 0.7008, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.37153419593345655, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 4.743862263276376e-06, | |
| "loss": 0.7592, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3733826247689464, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 4.740496249799053e-06, | |
| "loss": 0.8979, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3752310536044362, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 4.73710947360223e-06, | |
| "loss": 0.598, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.37707948243992606, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 4.733701966071226e-06, | |
| "loss": 0.6835, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3789279112754159, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 4.730273758783474e-06, | |
| "loss": 0.7821, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3807763401109057, | |
| "grad_norm": 5.125, | |
| "learning_rate": 4.726824883508235e-06, | |
| "loss": 0.7457, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3826247689463956, | |
| "grad_norm": 6.0, | |
| "learning_rate": 4.723355372206297e-06, | |
| "loss": 0.8121, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3844731977818854, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.719865257029686e-06, | |
| "loss": 0.9473, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.38632162661737524, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 4.716354570321361e-06, | |
| "loss": 0.8687, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.38817005545286504, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.712823344614921e-06, | |
| "loss": 0.8226, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3900184842883549, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.709271612634299e-06, | |
| "loss": 0.4422, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.39186691312384475, | |
| "grad_norm": 11.75, | |
| "learning_rate": 4.705699407293458e-06, | |
| "loss": 1.1324, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.39371534195933455, | |
| "grad_norm": 3.875, | |
| "learning_rate": 4.702106761696091e-06, | |
| "loss": 0.7663, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3955637707948244, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 4.698493709135311e-06, | |
| "loss": 0.3565, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3974121996303142, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 4.69486028309334e-06, | |
| "loss": 0.8618, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.39926062846580407, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.691206517241205e-06, | |
| "loss": 0.3152, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4011090573012939, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.6875324454384195e-06, | |
| "loss": 0.8727, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4029574861367837, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 4.683838101732674e-06, | |
| "loss": 0.6912, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4048059149722736, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 4.68012352035952e-06, | |
| "loss": 0.6973, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.4066543438077634, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 4.676388735742048e-06, | |
| "loss": 0.6748, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.40850277264325324, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.672633782490577e-06, | |
| "loss": 1.3242, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.41035120147874304, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 4.668858695402326e-06, | |
| "loss": 1.0964, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4121996303142329, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 4.665063509461098e-06, | |
| "loss": 0.967, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.41404805914972276, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.661248259836947e-06, | |
| "loss": 0.5724, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.41589648798521256, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 4.657412981885862e-06, | |
| "loss": 0.7476, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4177449168207024, | |
| "grad_norm": 5.0, | |
| "learning_rate": 4.653557711149434e-06, | |
| "loss": 0.6458, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.4195933456561922, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.6496824833545254e-06, | |
| "loss": 0.8271, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4214417744916821, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 4.645787334412945e-06, | |
| "loss": 0.6142, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4232902033271719, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 4.641872300421108e-06, | |
| "loss": 0.4389, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.42513863216266173, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 4.637937417659705e-06, | |
| "loss": 0.9207, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4269870609981516, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 4.633982722593367e-06, | |
| "loss": 0.7192, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.4288354898336414, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 4.630008251870324e-06, | |
| "loss": 0.3786, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.43068391866913125, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.626014042322069e-06, | |
| "loss": 0.9702, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.43253234750462105, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 4.622000130963015e-06, | |
| "loss": 0.4692, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4343807763401109, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.617966554990151e-06, | |
| "loss": 0.7393, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.43622920517560076, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 4.6139133517827e-06, | |
| "loss": 0.6702, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.43807763401109057, | |
| "grad_norm": 3.875, | |
| "learning_rate": 4.6098405589017685e-06, | |
| "loss": 0.5971, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.4399260628465804, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.605748214090002e-06, | |
| "loss": 0.6009, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.4417744916820702, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 4.601636355271235e-06, | |
| "loss": 0.3778, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.4436229205175601, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 4.597505020550138e-06, | |
| "loss": 0.5356, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4436229205175601, | |
| "eval_loss": 0.7416955232620239, | |
| "eval_runtime": 19.8749, | |
| "eval_samples_per_second": 5.736, | |
| "eval_steps_per_second": 2.868, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4454713493530499, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.593354248211866e-06, | |
| "loss": 0.779, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.44731977818853974, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 4.5891840767217e-06, | |
| "loss": 0.9026, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.4491682070240296, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.584994544724695e-06, | |
| "loss": 0.8965, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4510166358595194, | |
| "grad_norm": 6.0, | |
| "learning_rate": 4.5807856910453215e-06, | |
| "loss": 0.9645, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.45286506469500926, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.576557554687103e-06, | |
| "loss": 0.6508, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.45471349353049906, | |
| "grad_norm": 4.625, | |
| "learning_rate": 4.572310174832255e-06, | |
| "loss": 0.9041, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4565619223659889, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.568043590841325e-06, | |
| "loss": 0.661, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4584103512014787, | |
| "grad_norm": 7.875, | |
| "learning_rate": 4.563757842252824e-06, | |
| "loss": 0.828, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4602587800369686, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.5594529687828615e-06, | |
| "loss": 0.3424, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.46210720887245843, | |
| "grad_norm": 4.75, | |
| "learning_rate": 4.555129010324779e-06, | |
| "loss": 0.5807, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.46395563770794823, | |
| "grad_norm": 8.375, | |
| "learning_rate": 4.550786006948778e-06, | |
| "loss": 1.0949, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.4658040665434381, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 4.546423998901549e-06, | |
| "loss": 0.6161, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4676524953789279, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 4.542043026605902e-06, | |
| "loss": 0.7774, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.46950092421441775, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.537643130660385e-06, | |
| "loss": 0.8724, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.4713493530499076, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 4.533224351838914e-06, | |
| "loss": 0.7328, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.4731977818853974, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.528786731090394e-06, | |
| "loss": 0.822, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.47504621072088726, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 4.524330309538335e-06, | |
| "loss": 0.2581, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.47689463955637706, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 4.519855128480478e-06, | |
| "loss": 0.4583, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.4787430683918669, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.515361229388408e-06, | |
| "loss": 1.0834, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4805914972273567, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 4.5108486539071675e-06, | |
| "loss": 0.396, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4824399260628466, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 4.5063174438548775e-06, | |
| "loss": 0.8132, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.48428835489833644, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 4.501767641222343e-06, | |
| "loss": 0.6512, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.48613678373382624, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.4971992881726675e-06, | |
| "loss": 0.7899, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4879852125693161, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 4.492612427040864e-06, | |
| "loss": 0.6983, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4898336414048059, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 4.488007100333455e-06, | |
| "loss": 0.9323, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.49168207024029575, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 4.4833833507280884e-06, | |
| "loss": 0.9276, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.49353049907578556, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.478741221073136e-06, | |
| "loss": 0.376, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.4953789279112754, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 4.474080754387297e-06, | |
| "loss": 0.8011, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.49722735674676527, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.469401993859201e-06, | |
| "loss": 1.015, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.49907578558225507, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 4.464704982847008e-06, | |
| "loss": 0.3385, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5009242144177449, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.459989764878004e-06, | |
| "loss": 0.6179, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5027726432532348, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.455256383648199e-06, | |
| "loss": 0.3106, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5046210720887245, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.450504883021923e-06, | |
| "loss": 0.985, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5064695009242144, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 4.445735307031421e-06, | |
| "loss": 0.8767, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5083179297597042, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 4.4409476998764365e-06, | |
| "loss": 0.7676, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5101663585951941, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 4.436142105923814e-06, | |
| "loss": 0.3583, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.512014787430684, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.431318569707081e-06, | |
| "loss": 0.8458, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5138632162661737, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.426477135926035e-06, | |
| "loss": 0.5591, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5157116451016636, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.4216178494463305e-06, | |
| "loss": 0.8565, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5175600739371534, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 4.416740755299064e-06, | |
| "loss": 0.3491, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5194085027726433, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.411845898680354e-06, | |
| "loss": 0.2528, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5212569316081331, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.406933324950929e-06, | |
| "loss": 0.1933, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5231053604436229, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 4.402003079635695e-06, | |
| "loss": 0.7416, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5249537892791127, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.397055208423328e-06, | |
| "loss": 0.3184, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5268022181146026, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 4.392089757165841e-06, | |
| "loss": 0.6968, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5286506469500925, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.387106771878159e-06, | |
| "loss": 0.7838, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5304990757855823, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 4.3821062987376985e-06, | |
| "loss": 0.6749, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.532347504621072, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.377088384083935e-06, | |
| "loss": 1.0614, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5341959334565619, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.372053074417974e-06, | |
| "loss": 0.701, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5360443622920518, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 4.367000416402123e-06, | |
| "loss": 0.5594, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5378927911275416, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.361930456859455e-06, | |
| "loss": 0.6845, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5397412199630314, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.356843242773378e-06, | |
| "loss": 0.6825, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5415896487985212, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 4.351738821287196e-06, | |
| "loss": 0.3476, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5434380776340111, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.346617239703676e-06, | |
| "loss": 0.5582, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5452865064695009, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 4.34147854548461e-06, | |
| "loss": 0.6532, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5471349353049908, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 4.3363227862503674e-06, | |
| "loss": 0.6842, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5489833641404805, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.3311500097794655e-06, | |
| "loss": 1.2285, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5508317929759704, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.325960264008117e-06, | |
| "loss": 0.6185, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5526802218114603, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 4.32075359702979e-06, | |
| "loss": 0.4233, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5545286506469501, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 4.315530057094763e-06, | |
| "loss": 0.443, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.55637707948244, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.310289692609675e-06, | |
| "loss": 0.835, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5582255083179297, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.3050325521370785e-06, | |
| "loss": 0.8177, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5600739371534196, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 4.2997586843949905e-06, | |
| "loss": 0.5522, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5619223659889094, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 4.294468138256439e-06, | |
| "loss": 0.4428, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5637707948243993, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.289160962749012e-06, | |
| "loss": 0.496, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5656192236598891, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.2838372070544e-06, | |
| "loss": 0.3967, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5674676524953789, | |
| "grad_norm": 4.0, | |
| "learning_rate": 4.278496920507945e-06, | |
| "loss": 0.6934, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5693160813308688, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.2731401525981795e-06, | |
| "loss": 0.4632, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5711645101663586, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.267766952966369e-06, | |
| "loss": 0.2855, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5730129390018485, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.262377371406053e-06, | |
| "loss": 0.4477, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5748613678373382, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 4.2569714578625796e-06, | |
| "loss": 0.395, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5767097966728281, | |
| "grad_norm": 8.75, | |
| "learning_rate": 4.25154926243265e-06, | |
| "loss": 1.1819, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5785582255083179, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 4.2461108353638476e-06, | |
| "loss": 0.5877, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5804066543438078, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.240656227054174e-06, | |
| "loss": 0.8343, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5822550831792976, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.2351854880515856e-06, | |
| "loss": 0.3373, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5841035120147874, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 4.229698669053518e-06, | |
| "loss": 0.6325, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5859519408502772, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.2241958209064225e-06, | |
| "loss": 1.0372, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5878003696857671, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.218676994605295e-06, | |
| "loss": 0.9046, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.589648798521257, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 4.2131422412931985e-06, | |
| "loss": 0.2969, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5914972273567468, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 4.207591612260792e-06, | |
| "loss": 0.6589, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5914972273567468, | |
| "eval_loss": 0.7245264053344727, | |
| "eval_runtime": 19.8727, | |
| "eval_samples_per_second": 5.737, | |
| "eval_steps_per_second": 2.868, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5933456561922366, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 4.202025158945855e-06, | |
| "loss": 0.4466, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.5951940850277264, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 4.196442932932814e-06, | |
| "loss": 0.3829, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.5970425138632163, | |
| "grad_norm": 4.125, | |
| "learning_rate": 4.190844985952257e-06, | |
| "loss": 0.6488, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.5988909426987061, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.185231369880461e-06, | |
| "loss": 0.3511, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.600739371534196, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.179602136738909e-06, | |
| "loss": 0.8763, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6025878003696857, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 4.1739573386938075e-06, | |
| "loss": 0.6599, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.6044362292051756, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 4.168297028055599e-06, | |
| "loss": 0.7527, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6062846580406654, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 4.1626212572784884e-06, | |
| "loss": 0.3491, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6081330868761553, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.156930078959946e-06, | |
| "loss": 0.6552, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.609981515711645, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 4.151223545840225e-06, | |
| "loss": 0.7774, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6118299445471349, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 4.14550171080187e-06, | |
| "loss": 0.8564, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6136783733826248, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 4.139764626869233e-06, | |
| "loss": 0.7171, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6155268022181146, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 4.134012347207974e-06, | |
| "loss": 0.7456, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6173752310536045, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.128244925124572e-06, | |
| "loss": 0.2921, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6192236598890942, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 4.122462414065834e-06, | |
| "loss": 0.5071, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6210720887245841, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.116664867618395e-06, | |
| "loss": 0.8406, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6229205175600739, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 4.110852339508222e-06, | |
| "loss": 0.6328, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6247689463955638, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 4.10502488360012e-06, | |
| "loss": 0.6208, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6266173752310537, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 4.099182553897228e-06, | |
| "loss": 0.759, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6284658040665434, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 4.093325404540524e-06, | |
| "loss": 0.9401, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6303142329020333, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 4.087453489808314e-06, | |
| "loss": 0.7714, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6321626617375231, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.081566864115741e-06, | |
| "loss": 1.2085, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.634011090573013, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 4.075665582014272e-06, | |
| "loss": 0.8322, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6358595194085028, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 4.0697496981911955e-06, | |
| "loss": 0.4303, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6377079482439926, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 4.063819267469114e-06, | |
| "loss": 0.5353, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6395563770794824, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 4.057874344805437e-06, | |
| "loss": 0.8435, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6414048059149723, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 4.051914985291871e-06, | |
| "loss": 0.784, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6432532347504621, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 4.04594124415391e-06, | |
| "loss": 0.2081, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6451016635859519, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 4.039953176750321e-06, | |
| "loss": 0.2236, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6469500924214417, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 4.033950838572633e-06, | |
| "loss": 0.3375, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6487985212569316, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 4.027934285244624e-06, | |
| "loss": 0.751, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6506469500924215, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.021903572521802e-06, | |
| "loss": 0.8783, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6524953789279113, | |
| "grad_norm": 8.625, | |
| "learning_rate": 4.015858756290892e-06, | |
| "loss": 0.9485, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6543438077634011, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 4.009799892569317e-06, | |
| "loss": 0.7987, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6561922365988909, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 4.003727037504676e-06, | |
| "loss": 0.5482, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6580406654343808, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 3.997640247374228e-06, | |
| "loss": 0.7259, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6598890942698706, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.991539578584368e-06, | |
| "loss": 0.2106, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6617375231053605, | |
| "grad_norm": 5.5, | |
| "learning_rate": 3.985425087670104e-06, | |
| "loss": 0.7, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6635859519408502, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 3.979296831294535e-06, | |
| "loss": 0.6829, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6654343807763401, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 3.973154866248323e-06, | |
| "loss": 0.7398, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.66728280961183, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 3.966999249449172e-06, | |
| "loss": 0.6275, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6691312384473198, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 3.960830037941293e-06, | |
| "loss": 0.5373, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6709796672828097, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 3.9546472888948825e-06, | |
| "loss": 0.3239, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6728280961182994, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 3.948451059605588e-06, | |
| "loss": 0.1985, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6746765249537893, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 3.942241407493979e-06, | |
| "loss": 0.8387, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6765249537892791, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.936018390105013e-06, | |
| "loss": 0.9843, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.678373382624769, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.929782065107508e-06, | |
| "loss": 0.2061, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.6802218114602587, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 3.9235324902936e-06, | |
| "loss": 0.5996, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6820702402957486, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 3.917269723578212e-06, | |
| "loss": 0.6129, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.6839186691312384, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 3.910993822998518e-06, | |
| "loss": 0.645, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6857670979667283, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 3.904704846713402e-06, | |
| "loss": 0.6847, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6876155268022182, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 3.898402853002921e-06, | |
| "loss": 0.5214, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6894639556377079, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 3.8920879002677676e-06, | |
| "loss": 0.3786, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6913123844731978, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 3.885760047028722e-06, | |
| "loss": 0.6906, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6931608133086876, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.879419351926115e-06, | |
| "loss": 0.7821, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6950092421441775, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 3.873065873719285e-06, | |
| "loss": 0.8084, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6968576709796673, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 3.8666996712860286e-06, | |
| "loss": 0.6705, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.6987060998151571, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 3.86032080362206e-06, | |
| "loss": 0.5115, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7005545286506469, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 3.853929329840462e-06, | |
| "loss": 0.3842, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7024029574861368, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 3.847525309171138e-06, | |
| "loss": 0.7759, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7042513863216266, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 3.841108800960264e-06, | |
| "loss": 0.8197, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7060998151571165, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 3.83467986466974e-06, | |
| "loss": 0.3195, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7079482439926063, | |
| "grad_norm": 5.625, | |
| "learning_rate": 3.828238559876634e-06, | |
| "loss": 1.1011, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7097966728280961, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 3.8217849462726334e-06, | |
| "loss": 0.8037, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.711645101663586, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.815319083663496e-06, | |
| "loss": 0.3028, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7134935304990758, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 3.8088410319684853e-06, | |
| "loss": 0.5478, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7153419593345656, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.802350851219826e-06, | |
| "loss": 0.2858, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7171903881700554, | |
| "grad_norm": 5.25, | |
| "learning_rate": 3.79584860156214e-06, | |
| "loss": 0.5441, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7190388170055453, | |
| "grad_norm": 4.0, | |
| "learning_rate": 3.789334343251895e-06, | |
| "loss": 0.4993, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7208872458410351, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 3.7828081366568388e-06, | |
| "loss": 0.7737, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.722735674676525, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 3.776270042255448e-06, | |
| "loss": 0.7335, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7245841035120147, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 3.769720120636362e-06, | |
| "loss": 0.7952, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7264325323475046, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.763158432497824e-06, | |
| "loss": 0.8712, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7282809611829945, | |
| "grad_norm": 6.625, | |
| "learning_rate": 3.7565850386471157e-06, | |
| "loss": 0.835, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7301293900184843, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.4466, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7319778188539742, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 3.743403377580149e-06, | |
| "loss": 0.5889, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7338262476894639, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 3.7367952325185815e-06, | |
| "loss": 0.4429, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7356746765249538, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 3.730175626053099e-06, | |
| "loss": 0.6476, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7375231053604436, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 3.723544619527714e-06, | |
| "loss": 0.5749, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 3.7169022743920864e-06, | |
| "loss": 0.7849, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "eval_loss": 0.7030525803565979, | |
| "eval_runtime": 19.8774, | |
| "eval_samples_per_second": 5.735, | |
| "eval_steps_per_second": 2.868, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7412199630314233, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 3.7102486522009475e-06, | |
| "loss": 0.6911, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7430683918669131, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 3.703583814613536e-06, | |
| "loss": 0.7749, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.744916820702403, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 3.6969078233930235e-06, | |
| "loss": 0.7272, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7467652495378928, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 3.690220740405941e-06, | |
| "loss": 0.6549, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7486136783733827, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 3.6835226276216087e-06, | |
| "loss": 0.3819, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7504621072088724, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 3.67681354711156e-06, | |
| "loss": 0.8695, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7523105360443623, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 3.6700935610489653e-06, | |
| "loss": 0.3028, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.7541589648798521, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 3.663362731708059e-06, | |
| "loss": 1.0661, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.756007393715342, | |
| "grad_norm": 11.25, | |
| "learning_rate": 3.656621121463557e-06, | |
| "loss": 1.3266, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.7578558225508318, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 3.6498687927900855e-06, | |
| "loss": 0.885, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7597042513863216, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 3.6431058082615966e-06, | |
| "loss": 0.688, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7615526802218114, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.636332230550791e-06, | |
| "loss": 0.477, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.7634011090573013, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 3.6295481224285373e-06, | |
| "loss": 0.5917, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7652495378927912, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 3.6227535467632873e-06, | |
| "loss": 0.6099, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.767097966728281, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 3.6159485665204985e-06, | |
| "loss": 0.5766, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7689463955637708, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 3.6091332447620465e-06, | |
| "loss": 0.9963, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7707948243992606, | |
| "grad_norm": 6.625, | |
| "learning_rate": 3.6023076446456415e-06, | |
| "loss": 0.6713, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.7726432532347505, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 3.595471829424244e-06, | |
| "loss": 0.7673, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.7744916820702403, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 3.5886258624454786e-06, | |
| "loss": 0.3831, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.7763401109057301, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 3.581769807151044e-06, | |
| "loss": 0.596, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7781885397412199, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 3.574903727076133e-06, | |
| "loss": 0.2656, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.7800369685767098, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 3.5680276858488307e-06, | |
| "loss": 0.641, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.7818853974121996, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.561141747189538e-06, | |
| "loss": 0.6811, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.7837338262476895, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 3.5542459749103735e-06, | |
| "loss": 0.7878, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.7855822550831792, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 3.5473404329145843e-06, | |
| "loss": 0.5769, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7874306839186691, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 3.5404251851959537e-06, | |
| "loss": 0.8704, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.789279112754159, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.533500295838208e-06, | |
| "loss": 0.483, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.7911275415896488, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 3.5265658290144227e-06, | |
| "loss": 0.5987, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.7929759704251387, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 3.519621848986428e-06, | |
| "loss": 0.6344, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.7948243992606284, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.512668420104214e-06, | |
| "loss": 0.2874, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7966728280961183, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 3.5057056068053314e-06, | |
| "loss": 0.9778, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.7985212569316081, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 3.498733473614298e-06, | |
| "loss": 0.7903, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.800369685767098, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.4917520851419995e-06, | |
| "loss": 0.4643, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.8022181146025879, | |
| "grad_norm": 7.0, | |
| "learning_rate": 3.484761506085089e-06, | |
| "loss": 0.8675, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.8040665434380776, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 3.47776180122539e-06, | |
| "loss": 0.3105, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8059149722735675, | |
| "grad_norm": 4.875, | |
| "learning_rate": 3.470753035429296e-06, | |
| "loss": 0.7752, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8077634011090573, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.4637352736471653e-06, | |
| "loss": 0.2796, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8096118299445472, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 3.4567085809127247e-06, | |
| "loss": 0.8935, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8114602587800369, | |
| "grad_norm": 3.5, | |
| "learning_rate": 3.449673022342464e-06, | |
| "loss": 0.6599, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.8133086876155268, | |
| "grad_norm": 3.875, | |
| "learning_rate": 3.4426286631350312e-06, | |
| "loss": 0.8134, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8151571164510166, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 3.435575568570633e-06, | |
| "loss": 0.549, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8170055452865065, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 3.428513804010424e-06, | |
| "loss": 0.6928, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.8188539741219963, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 3.421443434895905e-06, | |
| "loss": 0.2479, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8207024029574861, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 3.4143645267483144e-06, | |
| "loss": 0.4003, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.822550831792976, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.4072771451680238e-06, | |
| "loss": 0.9285, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.8243992606284658, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 3.4001813558339263e-06, | |
| "loss": 0.8382, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8262476894639557, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 3.393077224502832e-06, | |
| "loss": 1.0314, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8280961182994455, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 3.3859648170088545e-06, | |
| "loss": 0.8829, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8299445471349353, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3.378844199262803e-06, | |
| "loss": 0.6821, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.8317929759704251, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 3.3717154372515716e-06, | |
| "loss": 0.711, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.833641404805915, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 3.3645785970375288e-06, | |
| "loss": 0.6403, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8354898336414048, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 3.357433744757902e-06, | |
| "loss": 0.24, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.8373382624768947, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.350280946624166e-06, | |
| "loss": 0.2786, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8391866913123844, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 3.3431202689214326e-06, | |
| "loss": 0.3571, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.8410351201478743, | |
| "grad_norm": 6.25, | |
| "learning_rate": 3.335951778007832e-06, | |
| "loss": 0.8584, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8428835489833642, | |
| "grad_norm": 15.8125, | |
| "learning_rate": 3.3287755403139007e-06, | |
| "loss": 1.2378, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.844731977818854, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 3.321591622341962e-06, | |
| "loss": 0.8754, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8465804066543438, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 3.3144000906655156e-06, | |
| "loss": 0.75, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8484288354898336, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 3.3072010119286156e-06, | |
| "loss": 0.8195, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8502772643253235, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 3.2999944528452575e-06, | |
| "loss": 0.6484, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8521256931608133, | |
| "grad_norm": 5.375, | |
| "learning_rate": 3.2927804801987533e-06, | |
| "loss": 0.8386, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.8539741219963032, | |
| "grad_norm": 5.0, | |
| "learning_rate": 3.2855591608411203e-06, | |
| "loss": 0.7973, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.8558225508317929, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 3.2783305616924543e-06, | |
| "loss": 0.4064, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8576709796672828, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.2710947497403176e-06, | |
| "loss": 0.2879, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8595194085027726, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 3.2638517920391095e-06, | |
| "loss": 0.3217, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8613678373382625, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 3.256601755709451e-06, | |
| "loss": 0.8764, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.8632162661737524, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.2493447079375605e-06, | |
| "loss": 0.7504, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8650646950092421, | |
| "grad_norm": 6.0, | |
| "learning_rate": 3.2420807159746333e-06, | |
| "loss": 0.8854, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.866913123844732, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 3.2348098471362132e-06, | |
| "loss": 0.555, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8687615526802218, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 3.227532168801574e-06, | |
| "loss": 0.8304, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8706099815157117, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 3.2202477484130947e-06, | |
| "loss": 0.6373, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.8724584103512015, | |
| "grad_norm": 5.875, | |
| "learning_rate": 3.2129566534756295e-06, | |
| "loss": 0.6689, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.8743068391866913, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 3.2056589515558894e-06, | |
| "loss": 0.3603, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.8761552680221811, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 3.1983547102818104e-06, | |
| "loss": 0.5722, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.878003696857671, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 3.1910439973419293e-06, | |
| "loss": 0.7527, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8798521256931608, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 3.183726880484755e-06, | |
| "loss": 0.6992, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.8817005545286506, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 3.1764034275181436e-06, | |
| "loss": 0.4991, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.8835489833641405, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 3.1690737063086664e-06, | |
| "loss": 0.3387, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.8853974121996303, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 3.161737784780984e-06, | |
| "loss": 0.6503, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.8872458410351202, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.1543957309172136e-06, | |
| "loss": 0.3008, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8872458410351202, | |
| "eval_loss": 0.7043923735618591, | |
| "eval_runtime": 19.8766, | |
| "eval_samples_per_second": 5.735, | |
| "eval_steps_per_second": 2.868, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.88909426987061, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 3.147047612756302e-06, | |
| "loss": 0.347, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.8909426987060998, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 3.139693498393395e-06, | |
| "loss": 0.3048, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.8927911275415896, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 3.132333455979202e-06, | |
| "loss": 0.4568, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.8946395563770795, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 3.124967553719372e-06, | |
| "loss": 0.5586, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.8964879852125693, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.117595859873854e-06, | |
| "loss": 0.8559, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.8983364140480592, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 3.1102184427562696e-06, | |
| "loss": 0.8834, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.9001848428835489, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.1028353707332776e-06, | |
| "loss": 0.9038, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9020332717190388, | |
| "grad_norm": 2.75, | |
| "learning_rate": 3.09544671222394e-06, | |
| "loss": 0.5715, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9038817005545287, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 3.0880525356990898e-06, | |
| "loss": 0.7079, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.9057301293900185, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.0806529096806953e-06, | |
| "loss": 0.8603, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9075785582255084, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 3.0732479027412254e-06, | |
| "loss": 0.9236, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9094269870609981, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 3.0658375835030148e-06, | |
| "loss": 0.5155, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.911275415896488, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 3.0584220206376253e-06, | |
| "loss": 0.9003, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9131238447319778, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 3.051001282865215e-06, | |
| "loss": 0.424, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9149722735674677, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 3.043575438953893e-06, | |
| "loss": 0.5802, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.9168207024029574, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 3.0361445577190918e-06, | |
| "loss": 0.5265, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.9186691312384473, | |
| "grad_norm": 4.0, | |
| "learning_rate": 3.0287087080229226e-06, | |
| "loss": 0.6223, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9205175600739371, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 3.02126795877354e-06, | |
| "loss": 0.6102, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.922365988909427, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.013822378924502e-06, | |
| "loss": 1.0314, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.9242144177449169, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 3.0063720374741324e-06, | |
| "loss": 0.542, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9260628465804066, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.9989170034648823e-06, | |
| "loss": 0.6886, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.9279112754158965, | |
| "grad_norm": 3.5, | |
| "learning_rate": 2.9914573459826855e-06, | |
| "loss": 0.6671, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.9297597042513863, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 2.9839931341563254e-06, | |
| "loss": 0.674, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.9316081330868762, | |
| "grad_norm": 4.875, | |
| "learning_rate": 2.9765244371567873e-06, | |
| "loss": 0.7132, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.933456561922366, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 2.9690513241966236e-06, | |
| "loss": 0.601, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.9353049907578558, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 2.9615738645293064e-06, | |
| "loss": 0.9621, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.9371534195933456, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 2.9540921274485913e-06, | |
| "loss": 0.7985, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9390018484288355, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.946606182287872e-06, | |
| "loss": 0.2716, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.9408502772643254, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 2.939116098419538e-06, | |
| "loss": 0.8653, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9426987060998152, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 2.9316219452543342e-06, | |
| "loss": 0.5268, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.944547134935305, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 2.924123792240712e-06, | |
| "loss": 0.8093, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9463955637707948, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 2.9166217088641935e-06, | |
| "loss": 0.7451, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9482439926062847, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 2.9091157646467205e-06, | |
| "loss": 1.0214, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9500924214417745, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 2.9016060291460162e-06, | |
| "loss": 0.7742, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9519408502772643, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 2.894092571954934e-06, | |
| "loss": 0.3196, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9537892791127541, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 2.886575462700821e-06, | |
| "loss": 0.7152, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.955637707948244, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 2.8790547710448623e-06, | |
| "loss": 0.6823, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.9574861367837338, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 2.871530566681447e-06, | |
| "loss": 0.3405, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.9593345656192237, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.864002919337513e-06, | |
| "loss": 0.7609, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9611829944547134, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.856471898771904e-06, | |
| "loss": 0.6978, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9630314232902033, | |
| "grad_norm": 5.375, | |
| "learning_rate": 2.848937574774725e-06, | |
| "loss": 0.6567, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9648798521256932, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 2.8414000171666952e-06, | |
| "loss": 0.6184, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.966728280961183, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 2.833859295798495e-06, | |
| "loss": 0.6687, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.9685767097966729, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.82631548055013e-06, | |
| "loss": 0.8295, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.9704251386321626, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.81876864133027e-06, | |
| "loss": 0.5869, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9722735674676525, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.811218848075614e-06, | |
| "loss": 0.357, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.9741219963031423, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 2.8036661707502317e-06, | |
| "loss": 0.8029, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.9759704251386322, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.7961106793449217e-06, | |
| "loss": 0.8667, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.977818853974122, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 2.7885524438765603e-06, | |
| "loss": 0.8206, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.9796672828096118, | |
| "grad_norm": 4.375, | |
| "learning_rate": 2.7809915343874533e-06, | |
| "loss": 0.7978, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9815157116451017, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.773428020944687e-06, | |
| "loss": 0.8059, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.9833641404805915, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 2.765861973639476e-06, | |
| "loss": 0.657, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.9852125693160814, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 2.7582934625865205e-06, | |
| "loss": 0.6126, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.9870609981515711, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 2.7507225579233487e-06, | |
| "loss": 0.6344, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.988909426987061, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.743149329809673e-06, | |
| "loss": 0.9873, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.9907578558225508, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 2.7355738484267345e-06, | |
| "loss": 0.503, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.9926062846580407, | |
| "grad_norm": 7.125, | |
| "learning_rate": 2.727996183976659e-06, | |
| "loss": 0.8659, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.9944547134935305, | |
| "grad_norm": 5.25, | |
| "learning_rate": 2.7204164066817995e-06, | |
| "loss": 0.5756, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.9963031423290203, | |
| "grad_norm": 3.75, | |
| "learning_rate": 2.7128345867840895e-06, | |
| "loss": 0.7666, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 2.705250794544393e-06, | |
| "loss": 0.2597, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.875, | |
| "learning_rate": 2.6976651002418504e-06, | |
| "loss": 0.8001, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.0018484288354899, | |
| "grad_norm": 2.25, | |
| "learning_rate": 2.690077574173228e-06, | |
| "loss": 0.5135, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.0036968576709797, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 2.682488286652269e-06, | |
| "loss": 0.4697, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.0055452865064696, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 2.6748973080090387e-06, | |
| "loss": 0.4738, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.0073937153419594, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 2.6673047085892728e-06, | |
| "loss": 0.8152, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.009242144177449, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 2.6597105587537307e-06, | |
| "loss": 0.4869, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.011090573012939, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 2.652114928877535e-06, | |
| "loss": 0.4794, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.0129390018484288, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 2.6445178893495267e-06, | |
| "loss": 0.6969, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.0147874306839186, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.6369195105716087e-06, | |
| "loss": 0.2867, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.0166358595194085, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.6293198629580974e-06, | |
| "loss": 0.3033, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0184842883548983, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 2.621719016935063e-06, | |
| "loss": 0.6519, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.0203327171903882, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 2.614117042939685e-06, | |
| "loss": 0.4737, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.022181146025878, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.6065140114195943e-06, | |
| "loss": 0.8458, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.024029574861368, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 2.598909992832222e-06, | |
| "loss": 0.6301, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.0258780036968578, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 2.591305057644148e-06, | |
| "loss": 0.8274, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.0277264325323474, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 2.583699276330443e-06, | |
| "loss": 0.5987, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.0295748613678373, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 2.576092719374023e-06, | |
| "loss": 0.88, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.0314232902033271, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 2.5684854572649876e-06, | |
| "loss": 0.6668, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.033271719038817, | |
| "grad_norm": 4.5, | |
| "learning_rate": 2.5608775604999746e-06, | |
| "loss": 0.5208, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.0351201478743068, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 2.553269099581501e-06, | |
| "loss": 0.5911, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0351201478743068, | |
| "eval_loss": 0.7039396166801453, | |
| "eval_runtime": 19.8706, | |
| "eval_samples_per_second": 5.737, | |
| "eval_steps_per_second": 2.869, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0369685767097967, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 2.5456601450173123e-06, | |
| "loss": 0.8642, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.0388170055452866, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 2.5380507673197297e-06, | |
| "loss": 0.7884, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.0406654343807764, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 2.530441037004994e-06, | |
| "loss": 0.6342, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.0425138632162663, | |
| "grad_norm": 3.125, | |
| "learning_rate": 2.522831024592615e-06, | |
| "loss": 0.5229, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.044362292051756, | |
| "grad_norm": 6.25, | |
| "learning_rate": 2.5152208006047156e-06, | |
| "loss": 0.5514, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.0462107208872458, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.5076104355653803e-06, | |
| "loss": 0.8105, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.0480591497227356, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.8325, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.0499075785582255, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 2.4923895644346205e-06, | |
| "loss": 0.317, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.0517560073937153, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 2.484779199395285e-06, | |
| "loss": 0.9237, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.0536044362292052, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 2.4771689754073856e-06, | |
| "loss": 0.2908, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.055452865064695, | |
| "grad_norm": 8.875, | |
| "learning_rate": 2.4695589629950067e-06, | |
| "loss": 0.7396, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.057301293900185, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 2.461949232680271e-06, | |
| "loss": 0.4392, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.0591497227356748, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 2.454339854982688e-06, | |
| "loss": 0.4072, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.0609981515711646, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.4467309004185e-06, | |
| "loss": 0.8076, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.0628465804066543, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.4391224395000263e-06, | |
| "loss": 0.9127, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.064695009242144, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 2.4315145427350132e-06, | |
| "loss": 0.7146, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.066543438077634, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 2.4239072806259776e-06, | |
| "loss": 0.7421, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.0683918669131238, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 2.4163007236695578e-06, | |
| "loss": 0.8132, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.0702402957486137, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 2.408694942355853e-06, | |
| "loss": 0.6022, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.0720887245841035, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 2.401090007167778e-06, | |
| "loss": 0.9218, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.0739371534195934, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 2.393485988580407e-06, | |
| "loss": 0.5004, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.0757855822550833, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 2.3858829570603157e-06, | |
| "loss": 0.6536, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.077634011090573, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.3782809830649377e-06, | |
| "loss": 0.5985, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.0794824399260627, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.370680137041903e-06, | |
| "loss": 0.5957, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.0813308687615526, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 2.363080489428391e-06, | |
| "loss": 0.6788, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.0831792975970425, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.3554821106504733e-06, | |
| "loss": 0.1703, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.0850277264325323, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 2.3478850711224655e-06, | |
| "loss": 0.8218, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.0868761552680222, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 2.3402894412462697e-06, | |
| "loss": 0.5959, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.088724584103512, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.3326952914107272e-06, | |
| "loss": 0.4873, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.0905730129390019, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 2.3251026919909617e-06, | |
| "loss": 0.5701, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.0924214417744917, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.317511713347731e-06, | |
| "loss": 0.8622, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.0942698706099816, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 2.3099224258267724e-06, | |
| "loss": 0.8753, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.0961182994454712, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 2.30233489975815e-06, | |
| "loss": 0.788, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.097966728280961, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.2947492054556075e-06, | |
| "loss": 0.8673, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.099815157116451, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 2.287165413215911e-06, | |
| "loss": 0.722, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.1016635859519408, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.2795835933182013e-06, | |
| "loss": 0.2898, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.1035120147874307, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 2.272003816023341e-06, | |
| "loss": 0.6431, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.1053604436229205, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 2.2644261515732655e-06, | |
| "loss": 0.7538, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.1072088724584104, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.2568506701903274e-06, | |
| "loss": 0.4134, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.1090573012939002, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.2492774420766517e-06, | |
| "loss": 0.6884, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.11090573012939, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 2.2417065374134808e-06, | |
| "loss": 0.2875, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.11275415896488, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.234138026360525e-06, | |
| "loss": 0.9249, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.1146025878003696, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 2.2265719790553147e-06, | |
| "loss": 0.7262, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.1164510166358594, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 2.219008465612548e-06, | |
| "loss": 0.535, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.1182994454713493, | |
| "grad_norm": 5.375, | |
| "learning_rate": 2.211447556123441e-06, | |
| "loss": 0.7473, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.1201478743068392, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.20388932065508e-06, | |
| "loss": 0.8215, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.121996303142329, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.1963338292497695e-06, | |
| "loss": 0.884, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.1238447319778189, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 2.1887811519243874e-06, | |
| "loss": 0.7523, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.1256931608133087, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 2.1812313586697307e-06, | |
| "loss": 0.942, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.1275415896487986, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 2.173684519449872e-06, | |
| "loss": 0.597, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.1293900184842884, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 2.1661407042015055e-06, | |
| "loss": 0.3094, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.1312384473197783, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.1585999828333065e-06, | |
| "loss": 0.8629, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.133086876155268, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 2.1510624252252756e-06, | |
| "loss": 0.5367, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.1349353049907578, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.143528101228097e-06, | |
| "loss": 0.2888, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.1367837338262476, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.1359970806624886e-06, | |
| "loss": 0.3376, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.1386321626617375, | |
| "grad_norm": 4.875, | |
| "learning_rate": 2.1284694333185534e-06, | |
| "loss": 0.6728, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.1404805914972274, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.120945228955138e-06, | |
| "loss": 0.3343, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.1423290203327172, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 2.11342453729918e-06, | |
| "loss": 0.7979, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.144177449168207, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 2.1059074280450663e-06, | |
| "loss": 0.5886, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.146025878003697, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.0983939708539846e-06, | |
| "loss": 0.3014, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.1478743068391868, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 2.0908842353532803e-06, | |
| "loss": 0.5744, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.1497227356746764, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 2.0833782911358073e-06, | |
| "loss": 0.7706, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.1515711645101663, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 2.0758762077592885e-06, | |
| "loss": 0.5108, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.1534195933456561, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 2.0683780547456666e-06, | |
| "loss": 0.5703, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.155268022181146, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 2.0608839015804623e-06, | |
| "loss": 0.2655, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.1571164510166358, | |
| "grad_norm": 6.875, | |
| "learning_rate": 2.0533938177121286e-06, | |
| "loss": 0.7283, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.1589648798521257, | |
| "grad_norm": 5.5, | |
| "learning_rate": 2.045907872551409e-06, | |
| "loss": 0.9952, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.1608133086876156, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 2.038426135470695e-06, | |
| "loss": 0.2742, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.1626617375231054, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 2.0309486758033777e-06, | |
| "loss": 0.7954, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.1645101663585953, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 2.0234755628432135e-06, | |
| "loss": 0.2734, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.166358595194085, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.0160068658436754e-06, | |
| "loss": 0.2676, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.1682070240295748, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.008542654017315e-06, | |
| "loss": 0.5901, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.1700554528650646, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.0010829965351185e-06, | |
| "loss": 0.6335, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.1719038817005545, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.993627962525868e-06, | |
| "loss": 0.725, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.1737523105360443, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.9861776210754988e-06, | |
| "loss": 0.6417, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.1756007393715342, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 1.978732041226461e-06, | |
| "loss": 0.6914, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.177449168207024, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 1.971291291977078e-06, | |
| "loss": 0.9675, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.179297597042514, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.963855442280909e-06, | |
| "loss": 0.6353, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.1811460258780038, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 1.956424561046108e-06, | |
| "loss": 0.8082, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.1829944547134936, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 1.9489987171347864e-06, | |
| "loss": 0.6143, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1829944547134936, | |
| "eval_loss": 0.6887955069541931, | |
| "eval_runtime": 19.8759, | |
| "eval_samples_per_second": 5.736, | |
| "eval_steps_per_second": 2.868, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1848428835489835, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 1.9415779793623755e-06, | |
| "loss": 0.7703, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.1866913123844731, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 1.934162416496986e-06, | |
| "loss": 0.7677, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.188539741219963, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 1.9267520972587754e-06, | |
| "loss": 0.6938, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.1903881700554528, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 1.9193470903193055e-06, | |
| "loss": 0.2752, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.1922365988909427, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.911947464300911e-06, | |
| "loss": 0.5809, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.1940850277264325, | |
| "grad_norm": 4.125, | |
| "learning_rate": 1.9045532877760605e-06, | |
| "loss": 0.7071, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.1959334565619224, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 1.8971646292667232e-06, | |
| "loss": 0.6245, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.1977818853974123, | |
| "grad_norm": 5.625, | |
| "learning_rate": 1.8897815572437304e-06, | |
| "loss": 0.7922, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.1996303142329021, | |
| "grad_norm": 7.0, | |
| "learning_rate": 1.8824041401261462e-06, | |
| "loss": 0.8811, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.201478743068392, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 1.8750324462806286e-06, | |
| "loss": 0.6595, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.2033271719038816, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.8676665440207982e-06, | |
| "loss": 0.885, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.2051756007393715, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 1.860306501606606e-06, | |
| "loss": 0.601, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.2070240295748613, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 1.852952387243698e-06, | |
| "loss": 0.5543, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.2088724584103512, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 1.845604269082787e-06, | |
| "loss": 0.5574, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.210720887245841, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.8382622152190161e-06, | |
| "loss": 0.233, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.212569316081331, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 1.8309262936913336e-06, | |
| "loss": 0.7057, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.2144177449168208, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.8235965724818563e-06, | |
| "loss": 1.037, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.2162661737523106, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.8162731195152454e-06, | |
| "loss": 0.3978, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.2181146025878005, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 1.808956002658071e-06, | |
| "loss": 0.6361, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.21996303142329, | |
| "grad_norm": 4.125, | |
| "learning_rate": 1.8016452897181902e-06, | |
| "loss": 0.725, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.22181146025878, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 1.7943410484441103e-06, | |
| "loss": 0.6095, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.2236598890942698, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.787043346524371e-06, | |
| "loss": 0.2469, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.2255083179297597, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 1.7797522515869062e-06, | |
| "loss": 1.0634, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.2273567467652495, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.7724678311984273e-06, | |
| "loss": 1.0534, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.2292051756007394, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 1.765190152863788e-06, | |
| "loss": 0.6844, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.2310536044362292, | |
| "grad_norm": 3.25, | |
| "learning_rate": 1.7579192840253678e-06, | |
| "loss": 0.3046, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.232902033271719, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 1.7506552920624397e-06, | |
| "loss": 0.693, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.234750462107209, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 1.74339824429055e-06, | |
| "loss": 0.3306, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.2365988909426986, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.7361482079608916e-06, | |
| "loss": 0.5415, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.2384473197781884, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.7289052502596836e-06, | |
| "loss": 0.3539, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.2402957486136783, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 1.721669438307546e-06, | |
| "loss": 0.6921, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.2421441774491682, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 1.7144408391588812e-06, | |
| "loss": 0.4945, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.243992606284658, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.7072195198012478e-06, | |
| "loss": 0.4828, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.2458410351201479, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 1.700005547154744e-06, | |
| "loss": 0.3758, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.2476894639556377, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.6927989880713852e-06, | |
| "loss": 0.6488, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.2495378927911276, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.6855999093344855e-06, | |
| "loss": 0.4919, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.2513863216266174, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 1.6784083776580395e-06, | |
| "loss": 0.8547, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.253234750462107, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 1.6712244596861008e-06, | |
| "loss": 0.6789, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.2550831792975972, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.6640482219921688e-06, | |
| "loss": 1.0684, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.2569316081330868, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 1.6568797310785678e-06, | |
| "loss": 0.6804, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.2587800369685767, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 1.6497190533758351e-06, | |
| "loss": 0.8177, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.2606284658040665, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 1.642566255242099e-06, | |
| "loss": 0.5538, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.2624768946395564, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 1.6354214029624719e-06, | |
| "loss": 0.7658, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.2643253234750462, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.628284562748429e-06, | |
| "loss": 0.8408, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.266173752310536, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 1.6211558007371976e-06, | |
| "loss": 0.9028, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.268022181146026, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 1.6140351829911466e-06, | |
| "loss": 0.4871, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.2698706099815156, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.6069227754971683e-06, | |
| "loss": 0.9585, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.2717190388170057, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.599818644166074e-06, | |
| "loss": 0.8345, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.2735674676524953, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 1.5927228548319768e-06, | |
| "loss": 0.6509, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.2754158964879851, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5856354732516865e-06, | |
| "loss": 0.252, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.277264325323475, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.5785565651040958e-06, | |
| "loss": 0.5776, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.2791127541589649, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.5714861959895766e-06, | |
| "loss": 0.3927, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.2809611829944547, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 1.5644244314293672e-06, | |
| "loss": 0.7105, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.2828096118299446, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.557371336864969e-06, | |
| "loss": 0.2498, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.2846580406654344, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 1.5503269776575364e-06, | |
| "loss": 0.6188, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.2865064695009243, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 1.5432914190872757e-06, | |
| "loss": 0.7849, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.2883548983364141, | |
| "grad_norm": 12.625, | |
| "learning_rate": 1.5362647263528358e-06, | |
| "loss": 0.9972, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.2902033271719038, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 1.5292469645707048e-06, | |
| "loss": 0.361, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.2920517560073936, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 1.5222381987746104e-06, | |
| "loss": 0.5836, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.2939001848428835, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 1.5152384939149117e-06, | |
| "loss": 0.6268, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2957486136783734, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 1.5082479148580016e-06, | |
| "loss": 0.7919, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.2975970425138632, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.5012665263857024e-06, | |
| "loss": 0.3666, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.299445471349353, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.4942943931946697e-06, | |
| "loss": 0.287, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.301293900184843, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.4873315798957865e-06, | |
| "loss": 0.3101, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.3031423290203328, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 1.4803781510135723e-06, | |
| "loss": 0.8061, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.3049907578558226, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4734341709855777e-06, | |
| "loss": 0.4358, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.3068391866913123, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 1.4664997041617929e-06, | |
| "loss": 0.6404, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.3086876155268024, | |
| "grad_norm": 4.5, | |
| "learning_rate": 1.4595748148040467e-06, | |
| "loss": 0.6323, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.310536044362292, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.452659567085416e-06, | |
| "loss": 0.2767, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.3123844731977818, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.445754025089627e-06, | |
| "loss": 0.5754, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.3142329020332717, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 1.4388582528104628e-06, | |
| "loss": 0.6333, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.3160813308687616, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.4319723141511693e-06, | |
| "loss": 0.2878, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.3179297597042514, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 1.4250962729238677e-06, | |
| "loss": 0.4207, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.3197781885397413, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.4182301928489556e-06, | |
| "loss": 0.7781, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.3216266173752311, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 1.4113741375545222e-06, | |
| "loss": 0.869, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.3234750462107208, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 1.4045281705757558e-06, | |
| "loss": 0.6826, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.3253234750462108, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 1.3976923553543587e-06, | |
| "loss": 0.6462, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.3271719038817005, | |
| "grad_norm": 4.375, | |
| "learning_rate": 1.390866755237954e-06, | |
| "loss": 0.718, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.3290203327171903, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.3840514334795019e-06, | |
| "loss": 0.7248, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.3308687615526802, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 1.3772464532367124e-06, | |
| "loss": 0.4895, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.3308687615526802, | |
| "eval_loss": 0.7117202281951904, | |
| "eval_runtime": 19.8765, | |
| "eval_samples_per_second": 5.735, | |
| "eval_steps_per_second": 2.868, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.33271719038817, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 1.3704518775714631e-06, | |
| "loss": 0.764, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.33456561922366, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.363667769449209e-06, | |
| "loss": 0.3025, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.3364140480591498, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 1.3568941917384038e-06, | |
| "loss": 0.4345, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.3382624768946396, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 1.3501312072099153e-06, | |
| "loss": 0.863, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.3401109057301293, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 1.3433788785364429e-06, | |
| "loss": 1.0277, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.3419593345656193, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 1.3366372682919415e-06, | |
| "loss": 0.6022, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.343807763401109, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.3299064389510346e-06, | |
| "loss": 0.7865, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.3456561922365988, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 1.3231864528884407e-06, | |
| "loss": 0.7712, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.3475046210720887, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 1.316477372378392e-06, | |
| "loss": 0.6921, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.3493530499075785, | |
| "grad_norm": 3.0, | |
| "learning_rate": 1.30977925959406e-06, | |
| "loss": 0.4245, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.3512014787430684, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 1.3030921766069782e-06, | |
| "loss": 0.7138, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.3530499075785583, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.2964161853864655e-06, | |
| "loss": 0.2775, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.354898336414048, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.2897513477990533e-06, | |
| "loss": 0.1933, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.356746765249538, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 1.2830977256079149e-06, | |
| "loss": 0.7493, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.3585951940850278, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 1.2764553804722868e-06, | |
| "loss": 0.7667, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.3604436229205175, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 1.2698243739469027e-06, | |
| "loss": 0.648, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.3622920517560073, | |
| "grad_norm": 6.625, | |
| "learning_rate": 1.263204767481419e-06, | |
| "loss": 0.8128, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.3641404805914972, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.256596622419852e-06, | |
| "loss": 0.3773, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.365988909426987, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.2500000000000007e-06, | |
| "loss": 0.3356, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.3678373382624769, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 1.2434149613528847e-06, | |
| "loss": 0.6262, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.3696857670979667, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.236841567502177e-06, | |
| "loss": 0.7075, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.3715341959334566, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.230279879363639e-06, | |
| "loss": 0.41, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.3733826247689465, | |
| "grad_norm": 6.0, | |
| "learning_rate": 1.223729957744553e-06, | |
| "loss": 0.8965, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.3752310536044363, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.2171918633431623e-06, | |
| "loss": 0.6955, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.377079482439926, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 1.2106656567481067e-06, | |
| "loss": 0.6845, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.378927911275416, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.20415139843786e-06, | |
| "loss": 0.6415, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.3807763401109057, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 1.1976491487801747e-06, | |
| "loss": 0.4861, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.3826247689463955, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.1911589680315155e-06, | |
| "loss": 0.4287, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.3844731977818854, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.1846809163365053e-06, | |
| "loss": 0.5578, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.3863216266173752, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.1782150537273665e-06, | |
| "loss": 0.8693, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.388170055452865, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.171761440123367e-06, | |
| "loss": 0.2463, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.390018484288355, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 1.1653201353302607e-06, | |
| "loss": 0.4324, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.3918669131238448, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 1.1588911990397364e-06, | |
| "loss": 0.8198, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.3937153419593344, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.1524746908288623e-06, | |
| "loss": 0.2055, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.3955637707948245, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 1.1460706701595387e-06, | |
| "loss": 0.5972, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.3974121996303142, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 1.139679196377941e-06, | |
| "loss": 0.5535, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.399260628465804, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 1.1333003287139727e-06, | |
| "loss": 0.6639, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.4011090573012939, | |
| "grad_norm": 7.75, | |
| "learning_rate": 1.1269341262807165e-06, | |
| "loss": 0.9163, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.4029574861367837, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.1205806480738852e-06, | |
| "loss": 0.3486, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.4048059149722736, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.1142399529712789e-06, | |
| "loss": 0.3102, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.4066543438077634, | |
| "grad_norm": 4.875, | |
| "learning_rate": 1.1079120997322333e-06, | |
| "loss": 0.7971, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.4085027726432533, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.1015971469970795e-06, | |
| "loss": 0.7212, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.410351201478743, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.095295153286599e-06, | |
| "loss": 0.4876, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.412199630314233, | |
| "grad_norm": 2.5, | |
| "learning_rate": 1.0890061770014829e-06, | |
| "loss": 0.3651, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.4140480591497226, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 1.0827302764217887e-06, | |
| "loss": 0.8161, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.4158964879852125, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 1.076467509706401e-06, | |
| "loss": 0.5648, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.4177449168207024, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.070217934892493e-06, | |
| "loss": 1.0064, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.4195933456561922, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.063981609894987e-06, | |
| "loss": 0.9541, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.421441774491682, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 1.0577585925060221e-06, | |
| "loss": 0.5169, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.423290203327172, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.051548940394413e-06, | |
| "loss": 0.5557, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.4251386321626618, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 1.0453527111051183e-06, | |
| "loss": 0.6249, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.4269870609981516, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.039169962058707e-06, | |
| "loss": 0.2433, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.4288354898336415, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 1.0330007505508286e-06, | |
| "loss": 0.9102, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.4306839186691311, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.0268451337516774e-06, | |
| "loss": 0.3431, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.432532347504621, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 1.0207031687054663e-06, | |
| "loss": 0.6644, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.4343807763401109, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.0145749123298965e-06, | |
| "loss": 0.2432, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.4362292051756007, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.0084604214156324e-06, | |
| "loss": 0.2645, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.4380776340110906, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 1.0023597526257723e-06, | |
| "loss": 0.6408, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.4399260628465804, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 9.962729624953244e-07, | |
| "loss": 0.609, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.4417744916820703, | |
| "grad_norm": 7.0, | |
| "learning_rate": 9.902001074306836e-07, | |
| "loss": 0.6692, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.4436229205175601, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.841412437091076e-07, | |
| "loss": 0.2694, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.44547134935305, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 9.780964274781984e-07, | |
| "loss": 0.8034, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.4473197781885396, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 9.72065714755377e-07, | |
| "loss": 0.6608, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.4491682070240297, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 9.660491614273678e-07, | |
| "loss": 0.7528, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.4510166358595193, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 9.600468232496793e-07, | |
| "loss": 0.7469, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.4528650646950092, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 9.5405875584609e-07, | |
| "loss": 0.3174, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.454713493530499, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 9.480850147081289e-07, | |
| "loss": 0.698, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.456561922365989, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 9.421256551945634e-07, | |
| "loss": 0.9137, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.4584103512014788, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 9.361807325308861e-07, | |
| "loss": 0.8854, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.4602587800369686, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 9.302503018088049e-07, | |
| "loss": 0.6859, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.4621072088724585, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 9.243344179857283e-07, | |
| "loss": 0.57, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.4639556377079481, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.184331358842594e-07, | |
| "loss": 0.3214, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.4658040665434382, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 9.125465101916866e-07, | |
| "loss": 0.9089, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.4676524953789278, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 9.066745954594775e-07, | |
| "loss": 0.4199, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.4695009242144177, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.008174461027724e-07, | |
| "loss": 0.6642, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.4713493530499075, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 8.949751163998812e-07, | |
| "loss": 0.6011, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.4731977818853974, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 8.891476604917787e-07, | |
| "loss": 0.4626, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.4750462107208873, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 8.833351323816064e-07, | |
| "loss": 0.5997, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.4768946395563771, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 8.775375859341669e-07, | |
| "loss": 0.5914, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.478743068391867, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 8.717550748754291e-07, | |
| "loss": 0.5436, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.478743068391867, | |
| "eval_loss": 0.6929174065589905, | |
| "eval_runtime": 19.8782, | |
| "eval_samples_per_second": 5.735, | |
| "eval_steps_per_second": 2.867, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4805914972273566, | |
| "grad_norm": 5.375, | |
| "learning_rate": 8.659876527920277e-07, | |
| "loss": 0.8208, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.4824399260628467, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 8.602353731307675e-07, | |
| "loss": 0.8082, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.4842883548983363, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 8.544982891981304e-07, | |
| "loss": 0.7467, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.4861367837338262, | |
| "grad_norm": 4.375, | |
| "learning_rate": 8.487764541597765e-07, | |
| "loss": 0.6383, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.487985212569316, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.430699210400553e-07, | |
| "loss": 0.153, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.489833641404806, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 8.373787427215122e-07, | |
| "loss": 0.7429, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.4916820702402958, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 8.317029719444017e-07, | |
| "loss": 0.2679, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.4935304990757856, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 8.260426613061944e-07, | |
| "loss": 0.8899, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.4953789279112755, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 8.203978632610915e-07, | |
| "loss": 0.6206, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.4972273567467653, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 8.147686301195384e-07, | |
| "loss": 0.5985, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.4990757855822552, | |
| "grad_norm": 4.375, | |
| "learning_rate": 8.09155014047743e-07, | |
| "loss": 0.638, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.5009242144177448, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.035570670671866e-07, | |
| "loss": 0.4361, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.502772643253235, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 7.979748410541452e-07, | |
| "loss": 0.8928, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.5046210720887245, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 7.924083877392094e-07, | |
| "loss": 0.2929, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.5064695009242144, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.868577587068021e-07, | |
| "loss": 0.1715, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.5083179297597042, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 7.813230053947054e-07, | |
| "loss": 0.3081, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.510166358595194, | |
| "grad_norm": 5.875, | |
| "learning_rate": 7.75804179093578e-07, | |
| "loss": 0.7596, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.512014787430684, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 7.703013309464835e-07, | |
| "loss": 0.5649, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.5138632162661736, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 7.648145119484152e-07, | |
| "loss": 0.2107, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.5157116451016637, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 7.593437729458262e-07, | |
| "loss": 0.6438, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.5175600739371533, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 7.538891646361529e-07, | |
| "loss": 0.6993, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.5194085027726434, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 7.484507375673505e-07, | |
| "loss": 0.6178, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.521256931608133, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 7.430285421374206e-07, | |
| "loss": 0.6061, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.5231053604436229, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 7.37622628593948e-07, | |
| "loss": 0.8917, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.5249537892791127, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 7.322330470336314e-07, | |
| "loss": 0.6046, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.5268022181146026, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 7.268598474018212e-07, | |
| "loss": 1.1106, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.5286506469500925, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 7.215030794920561e-07, | |
| "loss": 0.9603, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.5304990757855823, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.161627929456005e-07, | |
| "loss": 0.261, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.5323475046210722, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 7.108390372509894e-07, | |
| "loss": 0.8369, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.5341959334565618, | |
| "grad_norm": 3.375, | |
| "learning_rate": 7.05531861743562e-07, | |
| "loss": 0.627, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.5360443622920519, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 7.002413156050109e-07, | |
| "loss": 0.5524, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.5378927911275415, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 6.949674478629223e-07, | |
| "loss": 0.5247, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.5397412199630314, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 6.897103073903261e-07, | |
| "loss": 0.8861, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.5415896487985212, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 6.844699429052377e-07, | |
| "loss": 0.627, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.543438077634011, | |
| "grad_norm": 2.875, | |
| "learning_rate": 6.792464029702104e-07, | |
| "loss": 0.5021, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.545286506469501, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 6.74039735991883e-07, | |
| "loss": 0.6907, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.5471349353049908, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 6.688499902205345e-07, | |
| "loss": 0.9186, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.5489833641404807, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 6.636772137496328e-07, | |
| "loss": 0.6103, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.5508317929759703, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.585214545153912e-07, | |
| "loss": 0.2545, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.5526802218114604, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 6.533827602963244e-07, | |
| "loss": 0.7536, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.55452865064695, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.482611787128046e-07, | |
| "loss": 0.6925, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.55637707948244, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 6.431567572266231e-07, | |
| "loss": 0.4754, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.5582255083179297, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 6.380695431405453e-07, | |
| "loss": 0.4958, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.5600739371534196, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 6.329995835978775e-07, | |
| "loss": 0.8271, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.5619223659889094, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 6.279469255820256e-07, | |
| "loss": 0.2671, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.5637707948243993, | |
| "grad_norm": 3.75, | |
| "learning_rate": 6.229116159160653e-07, | |
| "loss": 0.7644, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.5656192236598891, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.178937012623015e-07, | |
| "loss": 0.6544, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.5674676524953788, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 6.128932281218411e-07, | |
| "loss": 0.7308, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.5693160813308689, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 6.079102428341588e-07, | |
| "loss": 1.0315, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.5711645101663585, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 6.029447915766715e-07, | |
| "loss": 0.6984, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.5730129390018486, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 5.979969203643049e-07, | |
| "loss": 0.7945, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.5748613678373382, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 5.930666750490724e-07, | |
| "loss": 0.6763, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.576709796672828, | |
| "grad_norm": 2.5, | |
| "learning_rate": 5.881541013196465e-07, | |
| "loss": 0.2988, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.578558225508318, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.832592447009367e-07, | |
| "loss": 0.5198, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.5804066543438078, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 5.783821505536696e-07, | |
| "loss": 0.5249, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.5822550831792976, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 5.735228640739652e-07, | |
| "loss": 0.8556, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.5841035120147873, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 5.686814302929188e-07, | |
| "loss": 0.7822, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.5859519408502774, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.63857894076186e-07, | |
| "loss": 0.8168, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.587800369685767, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 5.590523001235646e-07, | |
| "loss": 0.5047, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.589648798521257, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.542646929685811e-07, | |
| "loss": 0.2169, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.5914972273567467, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 5.494951169780777e-07, | |
| "loss": 0.6936, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.5933456561922366, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 5.447436163518017e-07, | |
| "loss": 0.4525, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.5951940850277264, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 5.400102351219969e-07, | |
| "loss": 0.6939, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.5970425138632163, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 5.352950171529928e-07, | |
| "loss": 0.8405, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.5988909426987061, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.305980061407998e-07, | |
| "loss": 0.7119, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.600739371534196, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 5.259192456127038e-07, | |
| "loss": 0.8154, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.6025878003696858, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 5.21258778926865e-07, | |
| "loss": 0.2969, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.6044362292051755, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 5.166166492719124e-07, | |
| "loss": 0.4302, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.6062846580406656, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 5.11992899666546e-07, | |
| "loss": 0.2097, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.6081330868761552, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 5.073875729591368e-07, | |
| "loss": 0.6771, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.609981515711645, | |
| "grad_norm": 2.5, | |
| "learning_rate": 5.028007118273324e-07, | |
| "loss": 0.3691, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.611829944547135, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.982323587776578e-07, | |
| "loss": 0.337, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.6136783733826248, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 4.936825561451236e-07, | |
| "loss": 0.9023, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.6155268022181146, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 4.891513460928335e-07, | |
| "loss": 0.272, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.6173752310536045, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.846387706115932e-07, | |
| "loss": 0.6996, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.6192236598890943, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.801448715195228e-07, | |
| "loss": 0.6802, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.621072088724584, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.7566969046166616e-07, | |
| "loss": 0.5023, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.622920517560074, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 4.712132689096077e-07, | |
| "loss": 0.3018, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.6247689463955637, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.667756481610866e-07, | |
| "loss": 1.0065, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.6266173752310538, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.6235686933961605e-07, | |
| "loss": 0.7803, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.6266173752310538, | |
| "eval_loss": 0.6892813444137573, | |
| "eval_runtime": 19.8692, | |
| "eval_samples_per_second": 5.738, | |
| "eval_steps_per_second": 2.869, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.6284658040665434, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 4.579569733940989e-07, | |
| "loss": 0.6694, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.6303142329020333, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.5357600109845134e-07, | |
| "loss": 0.5171, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.6321626617375231, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 4.4921399305122264e-07, | |
| "loss": 0.6084, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.634011090573013, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.448709896752215e-07, | |
| "loss": 0.8289, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.6358595194085028, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 4.405470312171392e-07, | |
| "loss": 0.8778, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.6377079482439925, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 4.3624215774717706e-07, | |
| "loss": 0.6691, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.6395563770794825, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 4.3195640915867585e-07, | |
| "loss": 0.7688, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.6414048059149722, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 4.2768982516774496e-07, | |
| "loss": 0.7971, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.6432532347504623, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.234424453128974e-07, | |
| "loss": 0.6229, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.645101663585952, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 4.192143089546788e-07, | |
| "loss": 0.6995, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.6469500924214417, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 4.150054552753055e-07, | |
| "loss": 0.7304, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.6487985212569316, | |
| "grad_norm": 7.5, | |
| "learning_rate": 4.108159232783007e-07, | |
| "loss": 0.9295, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.6506469500924215, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.0664575178813477e-07, | |
| "loss": 0.6466, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.6524953789279113, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.024949794498623e-07, | |
| "loss": 0.7287, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.654343807763401, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 3.983636447287656e-07, | |
| "loss": 0.701, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.656192236598891, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 3.9425178590999925e-07, | |
| "loss": 0.879, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.6580406654343807, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 3.901594410982326e-07, | |
| "loss": 0.7256, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.6598890942698707, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 3.860866482173009e-07, | |
| "loss": 0.7874, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.6617375231053604, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 3.8203344500984924e-07, | |
| "loss": 0.5719, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.6635859519408502, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 3.779998690369857e-07, | |
| "loss": 0.4106, | |
| "step": 900 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1082, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.2138058443155046e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |