| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.04266712178263235, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.53342435652647e-05, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.8, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001706684871305294, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.8859, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0002560027306957941, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 3e-06, | |
| "loss": 1.7004, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0003413369742610588, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.6891, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0004266712178263235, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.6489, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0005120054613915882, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7294, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0005973397049568529, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 1.7988, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0006826739485221176, | |
| "grad_norm": 0.875, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6986, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0007680081920873823, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 9e-06, | |
| "loss": 1.9519, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.000853342435652647, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 1e-05, | |
| "loss": 1.8502, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0009386766792179117, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.6868, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0010240109227831764, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.0214, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.001109345166348441, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.7536, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0011946794099137058, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.8917, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0012800136534789704, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.758, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0013653478970442352, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.6823, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0014506821406094998, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 1.9613, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0015360163841747646, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.6926, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0016213506277400291, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.5389, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.001706684871305294, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 2e-05, | |
| "loss": 1.862, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0017920191148705585, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 2.1e-05, | |
| "loss": 1.7124, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0018773533584358233, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.8534, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.001962687602001088, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 2.0062, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0020480218455663527, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.6357, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0021333560891316173, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.7676, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.002218690332696882, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 2.0052, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.002304024576262147, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 1.9785, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0023893588198274115, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.7911, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.002474693063392676, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 2.9e-05, | |
| "loss": 1.8871, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0025600273069579407, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7074, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0026453615505232057, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 3.1e-05, | |
| "loss": 1.9118, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0027306957940884703, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.6336, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.002816030037653735, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 3.3e-05, | |
| "loss": 1.8685, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0029013642812189995, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.7546, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0029866985247842645, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.6461, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.003072032768349529, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.8946, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0031573670119147937, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 3.7e-05, | |
| "loss": 1.8116, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0032427012554800583, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.8445, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0033280354990453233, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.7648, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.003413369742610588, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 4e-05, | |
| "loss": 1.808, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0034987039861758525, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.6193, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.003584038229741117, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.7854, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.003669372473306382, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 4.3e-05, | |
| "loss": 1.8162, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0037547067168716467, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.8106, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0038400409604369113, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.6102, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.003925375204002176, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.6463, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.004010709447567441, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.546, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0040960436911327055, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.796, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.00418137793469797, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.9179, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.004266712178263235, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 5e-05, | |
| "loss": 1.6671, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004352046421828499, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 1.7568, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.004437380665393764, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.693, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.004522714908959029, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 1.8329, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.004608049152524294, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.9059, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0046933833960895585, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.6741, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.004778717639654823, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.8357, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.004864051883220088, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 1.733, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.004949386126785352, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.8006, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.005034720370350617, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 5.9e-05, | |
| "loss": 1.7329, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.005120054613915881, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 6e-05, | |
| "loss": 1.7769, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.005205388857481147, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 6.1e-05, | |
| "loss": 1.6315, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0052907231010464115, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.5063, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.005376057344611676, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 6.3e-05, | |
| "loss": 1.6361, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.005461391588176941, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.6185, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.005546725831742205, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.7212, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.00563206007530747, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.8681, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.005717394318872734, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 6.7e-05, | |
| "loss": 1.7063, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.005802728562437999, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.6248, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0058880628060032645, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 6.9e-05, | |
| "loss": 1.8105, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.005973397049568529, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 7e-05, | |
| "loss": 1.9044, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.006058731293133794, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 7.1e-05, | |
| "loss": 1.6803, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.006144065536699058, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.7565, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.006229399780264323, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 7.3e-05, | |
| "loss": 1.7511, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.006314734023829587, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.856, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.006400068267394852, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.6748, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.006485402510960117, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.6632, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.006570736754525381, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 7.7e-05, | |
| "loss": 1.8728, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.006656070998090647, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.6411, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.006741405241655911, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 1.6952, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.006826739485221176, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8e-05, | |
| "loss": 1.6167, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.00691207372878644, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.8736, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.006997407972351705, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.8119, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.00708274221591697, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 8.3e-05, | |
| "loss": 1.8269, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.007168076459482234, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.6436, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.007253410703047499, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.6116, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.007338744946612764, | |
| "grad_norm": 0.5, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.5625, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.007424079190178029, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 8.7e-05, | |
| "loss": 1.6858, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.007509413433743293, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.6378, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.007594747677308558, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 1.5291, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.007680081920873823, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9e-05, | |
| "loss": 1.6327, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.007765416164439087, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.1e-05, | |
| "loss": 1.7248, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.007850750408004353, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.6183, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.007936084651569617, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 1.532, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.008021418895134882, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.4675, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.008106753138700146, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.8124, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.008192087382265411, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.5886, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.008277421625830676, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.6661, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.00836275586939594, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.5448, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.008448090112961205, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.8449, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.00853342435652647, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5973, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.008618758600091734, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 9.999497487437187e-05, | |
| "loss": 1.4568, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.008704092843656999, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.998994974874373e-05, | |
| "loss": 1.9025, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.008789427087222263, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.998492462311558e-05, | |
| "loss": 1.4718, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.008874761330787528, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 9.997989949748744e-05, | |
| "loss": 1.5075, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.008960095574352792, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 9.99748743718593e-05, | |
| "loss": 1.5536, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.009045429817918059, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 9.996984924623116e-05, | |
| "loss": 1.4153, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.009130764061483323, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 9.996482412060301e-05, | |
| "loss": 1.7526, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.009216098305048588, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 9.995979899497487e-05, | |
| "loss": 1.6215, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.009301432548613852, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 9.995477386934674e-05, | |
| "loss": 1.6131, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.009386766792179117, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 9.99497487437186e-05, | |
| "loss": 1.7026, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.009472101035744382, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 9.994472361809045e-05, | |
| "loss": 1.5233, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.009557435279309646, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 9.993969849246232e-05, | |
| "loss": 1.4345, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.00964276952287491, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.993467336683417e-05, | |
| "loss": 1.5261, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.009728103766440175, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 9.992964824120603e-05, | |
| "loss": 1.7505, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.00981343801000544, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 9.99246231155779e-05, | |
| "loss": 1.534, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.009898772253570704, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 9.991959798994976e-05, | |
| "loss": 1.7211, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.009984106497135969, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 9.99145728643216e-05, | |
| "loss": 1.488, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.010069440740701234, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.990954773869348e-05, | |
| "loss": 1.6496, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.010154774984266498, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 9.990452261306533e-05, | |
| "loss": 1.5469, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.010240109227831763, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.989949748743719e-05, | |
| "loss": 1.5484, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.010325443471397027, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.989447236180905e-05, | |
| "loss": 1.5737, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.010410777714962294, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.988944723618092e-05, | |
| "loss": 1.4623, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.010496111958527558, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 9.988442211055276e-05, | |
| "loss": 1.5756, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.010581446202092823, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.987939698492463e-05, | |
| "loss": 1.4768, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.010666780445658088, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.987437185929649e-05, | |
| "loss": 1.5331, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.010752114689223352, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 9.986934673366835e-05, | |
| "loss": 1.5734, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.010837448932788617, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.98643216080402e-05, | |
| "loss": 1.5175, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.010922783176353881, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 9.985929648241207e-05, | |
| "loss": 1.4156, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.011008117419919146, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.985427135678392e-05, | |
| "loss": 1.5877, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.01109345166348441, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.984924623115578e-05, | |
| "loss": 1.4328, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.011178785907049675, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 9.984422110552765e-05, | |
| "loss": 1.6384, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.01126412015061494, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.983919597989951e-05, | |
| "loss": 1.5616, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.011349454394180204, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 9.983417085427136e-05, | |
| "loss": 1.4179, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.011434788637745469, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 9.982914572864322e-05, | |
| "loss": 1.5475, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.011520122881310733, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 9.982412060301508e-05, | |
| "loss": 1.5457, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.011605457124875998, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 9.981909547738694e-05, | |
| "loss": 1.5643, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.011690791368441263, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.98140703517588e-05, | |
| "loss": 1.703, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.011776125612006529, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.980904522613065e-05, | |
| "loss": 1.4245, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.011861459855571794, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.980402010050252e-05, | |
| "loss": 1.4476, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.011946794099137058, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.979899497487438e-05, | |
| "loss": 1.5824, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.012032128342702323, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 9.979396984924624e-05, | |
| "loss": 1.5277, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.012117462586267587, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.978894472361809e-05, | |
| "loss": 1.559, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.012202796829832852, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.978391959798995e-05, | |
| "loss": 1.3583, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.012288131073398116, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.977889447236181e-05, | |
| "loss": 1.5059, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.012373465316963381, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.977386934673367e-05, | |
| "loss": 1.5431, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.012458799560528646, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.976884422110552e-05, | |
| "loss": 1.5239, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.01254413380409391, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 9.97638190954774e-05, | |
| "loss": 1.5618, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.012629468047659175, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.975879396984925e-05, | |
| "loss": 1.6436, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.01271480229122444, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.975376884422111e-05, | |
| "loss": 1.4133, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.012800136534789704, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.974874371859297e-05, | |
| "loss": 1.5351, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.012885470778354969, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.974371859296483e-05, | |
| "loss": 1.5085, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.012970805021920233, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 9.973869346733668e-05, | |
| "loss": 1.3979, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.013056139265485498, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.973366834170856e-05, | |
| "loss": 1.476, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.013141473509050762, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.97286432160804e-05, | |
| "loss": 1.3851, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.013226807752616029, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.972361809045227e-05, | |
| "loss": 1.479, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.013312141996181293, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.971859296482412e-05, | |
| "loss": 1.6612, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.013397476239746558, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.971356783919599e-05, | |
| "loss": 1.4596, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.013482810483311822, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 9.970854271356784e-05, | |
| "loss": 1.4859, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.013568144726877087, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.97035175879397e-05, | |
| "loss": 1.514, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.013653478970442352, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 9.969849246231156e-05, | |
| "loss": 1.4963, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.013738813214007616, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.969346733668343e-05, | |
| "loss": 1.5223, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.01382414745757288, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.968844221105527e-05, | |
| "loss": 1.4844, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.013909481701138145, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.968341708542715e-05, | |
| "loss": 1.4614, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.01399481594470341, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.9678391959799e-05, | |
| "loss": 1.6443, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.014080150188268675, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.967336683417086e-05, | |
| "loss": 1.5618, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.01416548443183394, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.966834170854272e-05, | |
| "loss": 1.4337, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.014250818675399204, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.966331658291458e-05, | |
| "loss": 1.3675, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.014336152918964468, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.965829145728643e-05, | |
| "loss": 1.6303, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.014421487162529733, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.96532663316583e-05, | |
| "loss": 1.5298, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.014506821406094998, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.964824120603016e-05, | |
| "loss": 1.5157, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.014592155649660264, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.964321608040202e-05, | |
| "loss": 1.534, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.014677489893225528, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.963819095477387e-05, | |
| "loss": 1.4738, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.014762824136790793, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 9.963316582914573e-05, | |
| "loss": 1.3934, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.014848158380356058, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.962814070351759e-05, | |
| "loss": 1.4652, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.014933492623921322, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 9.962311557788945e-05, | |
| "loss": 1.4472, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.015018826867486587, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.961809045226132e-05, | |
| "loss": 1.4694, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.015104161111051851, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.961306532663316e-05, | |
| "loss": 1.5138, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.015189495354617116, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.960804020100503e-05, | |
| "loss": 1.5025, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.01527482959818238, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 9.960301507537689e-05, | |
| "loss": 1.49, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.015360163841747645, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.959798994974875e-05, | |
| "loss": 1.4287, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.01544549808531291, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.95929648241206e-05, | |
| "loss": 1.3334, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.015530832328878174, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 9.958793969849247e-05, | |
| "loss": 1.5699, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.015616166572443439, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 9.958291457286432e-05, | |
| "loss": 1.7028, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.015701500816008705, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.957788944723619e-05, | |
| "loss": 1.4966, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.01578683505957397, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.957286432160805e-05, | |
| "loss": 1.4278, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.015872169303139234, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.956783919597991e-05, | |
| "loss": 1.3241, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.0159575035467045, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.956281407035176e-05, | |
| "loss": 1.5802, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.016042837790269764, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.955778894472362e-05, | |
| "loss": 1.5102, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.016128172033835028, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.955276381909548e-05, | |
| "loss": 1.5297, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.016213506277400293, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.954773869346734e-05, | |
| "loss": 1.4609, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.016298840520965557, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.954271356783919e-05, | |
| "loss": 1.482, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.016384174764530822, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 9.953768844221107e-05, | |
| "loss": 1.5933, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.016469509008096087, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.953266331658292e-05, | |
| "loss": 1.6971, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.01655484325166135, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.952763819095478e-05, | |
| "loss": 1.7042, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.016640177495226616, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 9.952261306532664e-05, | |
| "loss": 1.5217, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.01672551173879188, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 9.95175879396985e-05, | |
| "loss": 1.5504, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.016810845982357145, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.951256281407035e-05, | |
| "loss": 1.3498, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.01689618022592241, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.950753768844223e-05, | |
| "loss": 1.495, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.016981514469487674, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.950251256281408e-05, | |
| "loss": 1.4104, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.01706684871305294, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.949748743718594e-05, | |
| "loss": 1.3059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.017152182956618203, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.94924623115578e-05, | |
| "loss": 1.6192, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.017237517200183468, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.948743718592966e-05, | |
| "loss": 1.5999, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.017322851443748732, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 9.948241206030151e-05, | |
| "loss": 1.4207, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.017408185687313997, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.947738693467337e-05, | |
| "loss": 1.4471, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.01749351993087926, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.947236180904523e-05, | |
| "loss": 1.5427, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.017578854174444526, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.94673366834171e-05, | |
| "loss": 1.4854, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.01766418841800979, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.946231155778894e-05, | |
| "loss": 1.4921, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.017749522661575055, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.94572864321608e-05, | |
| "loss": 1.4232, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.01783485690514032, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.945226130653267e-05, | |
| "loss": 1.3718, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.017920191148705585, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.944723618090453e-05, | |
| "loss": 1.4906, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.018005525392270853, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.944221105527639e-05, | |
| "loss": 1.3974, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.018090859635836117, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.943718592964824e-05, | |
| "loss": 1.4588, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.018176193879401382, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.94321608040201e-05, | |
| "loss": 1.5893, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.018261528122966646, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.942713567839197e-05, | |
| "loss": 1.594, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.01834686236653191, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.942211055276383e-05, | |
| "loss": 1.531, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.018432196610097176, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.941708542713568e-05, | |
| "loss": 1.4238, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.01851753085366244, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.941206030150754e-05, | |
| "loss": 1.3628, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.018602865097227705, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.94070351758794e-05, | |
| "loss": 1.48, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.01868819934079297, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.940201005025126e-05, | |
| "loss": 1.5223, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.018773533584358234, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.939698492462311e-05, | |
| "loss": 1.4358, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.0188588678279235, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.939195979899499e-05, | |
| "loss": 1.4313, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.018944202071488763, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.938693467336683e-05, | |
| "loss": 1.5912, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.019029536315054028, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.93819095477387e-05, | |
| "loss": 1.5051, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.019114870558619292, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.937688442211056e-05, | |
| "loss": 1.6389, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.019200204802184557, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.937185929648242e-05, | |
| "loss": 1.5974, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.01928553904574982, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.936683417085427e-05, | |
| "loss": 1.455, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.019370873289315086, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.936180904522614e-05, | |
| "loss": 1.4916, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.01945620753288035, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.935678391959799e-05, | |
| "loss": 1.5209, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.019541541776445615, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 9.935175879396985e-05, | |
| "loss": 1.5642, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.01962687602001088, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.934673366834172e-05, | |
| "loss": 1.4796, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.019712210263576144, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.934170854271358e-05, | |
| "loss": 1.4586, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.01979754450714141, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.933668341708543e-05, | |
| "loss": 1.5006, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.019882878750706674, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.933165829145729e-05, | |
| "loss": 1.572, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.019968212994271938, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.932663316582915e-05, | |
| "loss": 1.3313, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.020053547237837203, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.932160804020101e-05, | |
| "loss": 1.5505, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.020138881481402467, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.931658291457286e-05, | |
| "loss": 1.4026, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.020224215724967732, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.931155778894474e-05, | |
| "loss": 1.5383, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.020309549968532997, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.930653266331659e-05, | |
| "loss": 1.3315, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.02039488421209826, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.930150753768845e-05, | |
| "loss": 1.4409, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.020480218455663526, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.929648241206031e-05, | |
| "loss": 1.431, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.02056555269922879, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.929145728643217e-05, | |
| "loss": 1.4406, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.020650886942794055, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.928643216080402e-05, | |
| "loss": 1.3872, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.02073622118635932, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.928140703517588e-05, | |
| "loss": 1.6584, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.020821555429924588, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.927638190954774e-05, | |
| "loss": 1.4384, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.020906889673489852, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.927135678391961e-05, | |
| "loss": 1.4545, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.020992223917055117, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.926633165829147e-05, | |
| "loss": 1.4888, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.02107755816062038, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.926130653266332e-05, | |
| "loss": 1.5343, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.021162892404185646, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 9.925628140703518e-05, | |
| "loss": 1.4289, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.02124822664775091, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.925125628140703e-05, | |
| "loss": 1.5845, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.021333560891316175, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.92462311557789e-05, | |
| "loss": 1.3077, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02141889513488144, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.924120603015075e-05, | |
| "loss": 1.6101, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.021504229378446704, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.923618090452261e-05, | |
| "loss": 1.5453, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.02158956362201197, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.923115577889448e-05, | |
| "loss": 1.5328, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.021674897865577233, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.922613065326634e-05, | |
| "loss": 1.4348, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.021760232109142498, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.922110552763819e-05, | |
| "loss": 1.5544, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.021845566352707763, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.921608040201006e-05, | |
| "loss": 1.4347, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.021930900596273027, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.921105527638191e-05, | |
| "loss": 1.3953, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.022016234839838292, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.920603015075377e-05, | |
| "loss": 1.3102, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.022101569083403556, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.920100502512563e-05, | |
| "loss": 1.376, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.02218690332696882, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.91959798994975e-05, | |
| "loss": 1.3496, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.022272237570534086, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.919095477386935e-05, | |
| "loss": 1.3267, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.02235757181409935, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.918592964824122e-05, | |
| "loss": 1.2196, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.022442906057664615, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.918090452261307e-05, | |
| "loss": 1.419, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.02252824030122988, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.917587939698493e-05, | |
| "loss": 1.3407, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.022613574544795144, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.917085427135678e-05, | |
| "loss": 1.3234, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.02269890878836041, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.916582914572866e-05, | |
| "loss": 1.3534, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.022784243031925673, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.91608040201005e-05, | |
| "loss": 1.3959, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.022869577275490938, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.915577889447237e-05, | |
| "loss": 1.5845, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.022954911519056202, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.915075376884423e-05, | |
| "loss": 1.4996, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.023040245762621467, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 9.914572864321609e-05, | |
| "loss": 1.4105, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.02312558000618673, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.914070351758794e-05, | |
| "loss": 1.4621, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.023210914249751996, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.913567839195981e-05, | |
| "loss": 1.3192, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.02329624849331726, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.913065326633166e-05, | |
| "loss": 1.432, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.023381582736882525, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.912562814070352e-05, | |
| "loss": 1.4196, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.02346691698044779, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.912060301507539e-05, | |
| "loss": 1.3807, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.023552251224013058, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.911557788944725e-05, | |
| "loss": 1.4028, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.023637585467578322, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.91105527638191e-05, | |
| "loss": 1.4866, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.023722919711143587, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.910552763819096e-05, | |
| "loss": 1.314, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.02380825395470885, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.910050251256282e-05, | |
| "loss": 1.4455, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.023893588198274116, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.909547738693468e-05, | |
| "loss": 1.5878, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02397892244183938, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.909045226130653e-05, | |
| "loss": 1.5868, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.024064256685404645, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.90854271356784e-05, | |
| "loss": 1.5392, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.02414959092896991, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.908040201005026e-05, | |
| "loss": 1.3902, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.024234925172535175, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.90753768844221e-05, | |
| "loss": 1.4886, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.02432025941610044, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.907035175879398e-05, | |
| "loss": 1.353, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.024405593659665704, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 9.906532663316583e-05, | |
| "loss": 1.3988, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.02449092790323097, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.906030150753769e-05, | |
| "loss": 1.5054, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.024576262146796233, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.905527638190955e-05, | |
| "loss": 1.4862, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.024661596390361498, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.905025125628141e-05, | |
| "loss": 1.3854, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.024746930633926762, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.904522613065326e-05, | |
| "loss": 1.4284, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.024832264877492027, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.904020100502514e-05, | |
| "loss": 1.3654, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.02491759912105729, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.903517587939699e-05, | |
| "loss": 1.453, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.025002933364622556, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.903015075376885e-05, | |
| "loss": 1.4702, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.02508826760818782, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.902512562814071e-05, | |
| "loss": 1.4253, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.025173601851753085, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.902010050251257e-05, | |
| "loss": 1.2651, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.02525893609531835, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.901507537688442e-05, | |
| "loss": 1.4039, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.025344270338883614, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.901005025125628e-05, | |
| "loss": 1.522, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.02542960458244888, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.900502512562815e-05, | |
| "loss": 1.2836, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.025514938826014143, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.3898, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.025600273069579408, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.899497487437186e-05, | |
| "loss": 1.5715, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.025685607313144673, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.898994974874373e-05, | |
| "loss": 1.6434, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.025770941556709937, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.898492462311558e-05, | |
| "loss": 1.4651, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.025856275800275202, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.897989949748744e-05, | |
| "loss": 1.5617, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.025941610043840466, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.89748743718593e-05, | |
| "loss": 1.367, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.02602694428740573, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.896984924623117e-05, | |
| "loss": 1.3788, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.026112278530970996, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.896482412060301e-05, | |
| "loss": 1.9467, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.02619761277453626, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.895979899497489e-05, | |
| "loss": 1.3327, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.026282947018101525, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.895477386934674e-05, | |
| "loss": 1.455, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.026368281261666793, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.89497487437186e-05, | |
| "loss": 1.3412, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.026453615505232057, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.894472361809046e-05, | |
| "loss": 1.4269, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.026538949748797322, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.893969849246232e-05, | |
| "loss": 1.3771, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.026624283992362587, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.893467336683417e-05, | |
| "loss": 1.4056, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.02670961823592785, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.892964824120604e-05, | |
| "loss": 1.2114, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.026794952479493116, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9.89246231155779e-05, | |
| "loss": 1.4399, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.02688028672305838, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.891959798994975e-05, | |
| "loss": 1.3369, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.026965620966623645, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 9.891457286432161e-05, | |
| "loss": 1.3965, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.02705095521018891, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 9.890954773869347e-05, | |
| "loss": 1.2973, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.027136289453754174, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 9.890452261306533e-05, | |
| "loss": 1.5117, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.02722162369731944, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.889949748743718e-05, | |
| "loss": 1.4032, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.027306957940884703, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.889447236180906e-05, | |
| "loss": 1.3552, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.027392292184449968, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.88894472361809e-05, | |
| "loss": 1.4379, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.027477626428015232, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.888442211055277e-05, | |
| "loss": 1.3794, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.027562960671580497, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.887939698492463e-05, | |
| "loss": 1.4165, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.02764829491514576, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.887437185929649e-05, | |
| "loss": 1.6702, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.027733629158711026, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.886934673366834e-05, | |
| "loss": 1.5435, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.02781896340227629, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.886432160804021e-05, | |
| "loss": 1.5578, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.027904297645841555, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.885929648241206e-05, | |
| "loss": 1.5035, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.02798963188940682, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.885427135678393e-05, | |
| "loss": 1.5053, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.028074966132972085, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.884924623115577e-05, | |
| "loss": 1.4352, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.02816030037653735, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.884422110552765e-05, | |
| "loss": 1.2852, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.028245634620102614, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.88391959798995e-05, | |
| "loss": 1.5373, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.02833096886366788, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.883417085427136e-05, | |
| "loss": 1.4506, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.028416303107233143, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.882914572864322e-05, | |
| "loss": 1.4871, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.028501637350798407, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.882412060301508e-05, | |
| "loss": 1.5472, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.028586971594363672, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.881909547738693e-05, | |
| "loss": 1.3555, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.028672305837928937, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.881407035175881e-05, | |
| "loss": 1.4413, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0287576400814942, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.880904522613066e-05, | |
| "loss": 1.4191, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.028842974325059466, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.880402010050252e-05, | |
| "loss": 1.4605, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.02892830856862473, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.879899497487438e-05, | |
| "loss": 1.4177, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.029013642812189995, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.879396984924624e-05, | |
| "loss": 1.416, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.029098977055755263, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9.878894472361809e-05, | |
| "loss": 1.3267, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.029184311299320528, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.878391959798995e-05, | |
| "loss": 1.4578, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.029269645542885792, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.877889447236182e-05, | |
| "loss": 1.4612, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.029354979786451057, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.877386934673368e-05, | |
| "loss": 1.4726, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.02944031403001632, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.876884422110553e-05, | |
| "loss": 1.2706, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.029525648273581586, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.87638190954774e-05, | |
| "loss": 1.2087, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.02961098251714685, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.875879396984925e-05, | |
| "loss": 1.5016, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.029696316760712115, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.875376884422111e-05, | |
| "loss": 1.4597, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.02978165100427738, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.874874371859297e-05, | |
| "loss": 1.5274, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.029866985247842644, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.874371859296482e-05, | |
| "loss": 1.4724, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02995231949140791, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.873869346733668e-05, | |
| "loss": 1.4303, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.030037653734973174, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.873366834170855e-05, | |
| "loss": 1.4233, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.030122987978538438, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.872864321608041e-05, | |
| "loss": 1.3207, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.030208322222103703, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.872361809045226e-05, | |
| "loss": 1.5609, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.030293656465668967, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.871859296482413e-05, | |
| "loss": 1.4183, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.030378990709234232, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.871356783919598e-05, | |
| "loss": 1.3768, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.030464324952799497, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.870854271356784e-05, | |
| "loss": 1.4829, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.03054965919636476, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.870351758793969e-05, | |
| "loss": 1.4653, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.030634993439930026, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.869849246231157e-05, | |
| "loss": 1.3873, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.03072032768349529, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.869346733668342e-05, | |
| "loss": 1.5274, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.030805661927060555, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.868844221105528e-05, | |
| "loss": 1.4618, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.03089099617062582, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.868341708542714e-05, | |
| "loss": 1.532, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.030976330414191084, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.8678391959799e-05, | |
| "loss": 1.4401, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.03106166465775635, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.867336683417085e-05, | |
| "loss": 1.4295, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.031146998901321613, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.866834170854273e-05, | |
| "loss": 1.4831, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.031232333144886878, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.866331658291457e-05, | |
| "loss": 1.596, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.031317667388452146, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.865829145728644e-05, | |
| "loss": 1.4827, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.03140300163201741, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.86532663316583e-05, | |
| "loss": 1.498, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.031488335875582675, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.864824120603016e-05, | |
| "loss": 1.3065, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.03157367011914794, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.864321608040201e-05, | |
| "loss": 1.2668, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.031659004362713204, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.863819095477388e-05, | |
| "loss": 1.2517, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.03174433860627847, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.863316582914573e-05, | |
| "loss": 1.3486, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.03182967284984373, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.86281407035176e-05, | |
| "loss": 1.5, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.031915007093409, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.862311557788944e-05, | |
| "loss": 1.4812, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.03200034133697426, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.861809045226132e-05, | |
| "loss": 1.5226, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.03208567558053953, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.861306532663317e-05, | |
| "loss": 1.4164, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.03217100982410479, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.860804020100503e-05, | |
| "loss": 1.2662, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.032256344067670056, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.860301507537689e-05, | |
| "loss": 1.3327, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.03234167831123532, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.859798994974875e-05, | |
| "loss": 1.2236, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.032427012554800586, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.85929648241206e-05, | |
| "loss": 1.6013, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.03251234679836585, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.858793969849246e-05, | |
| "loss": 1.4355, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.032597681041931115, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.858291457286433e-05, | |
| "loss": 1.4302, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.03268301528549638, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.857788944723619e-05, | |
| "loss": 1.3834, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.032768349529061644, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.857286432160805e-05, | |
| "loss": 1.3019, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.03285368377262691, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.85678391959799e-05, | |
| "loss": 1.4871, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.03293901801619217, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.856281407035176e-05, | |
| "loss": 1.3936, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.03302435225975744, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 9.855778894472362e-05, | |
| "loss": 1.4884, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.0331096865033227, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.855276381909548e-05, | |
| "loss": 1.358, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.03319502074688797, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.854773869346733e-05, | |
| "loss": 1.3321, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.03328035499045323, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.85427135678392e-05, | |
| "loss": 1.3673, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.033365689234018496, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.853768844221106e-05, | |
| "loss": 1.4476, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.03345102347758376, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.853266331658292e-05, | |
| "loss": 1.3882, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.033536357721149025, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.852763819095477e-05, | |
| "loss": 1.2986, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.03362169196471429, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.852261306532664e-05, | |
| "loss": 1.5465, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.033707026208279554, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.851758793969849e-05, | |
| "loss": 1.3841, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.03379236045184482, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.851256281407035e-05, | |
| "loss": 1.5757, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.033877694695410084, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.850753768844222e-05, | |
| "loss": 1.3772, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.03396302893897535, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.850251256281408e-05, | |
| "loss": 1.3808, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.03404836318254061, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.849748743718593e-05, | |
| "loss": 1.4123, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.03413369742610588, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.84924623115578e-05, | |
| "loss": 1.5279, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.03421903166967114, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.848743718592965e-05, | |
| "loss": 1.5218, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.034304365913236406, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.848241206030151e-05, | |
| "loss": 1.5072, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.03438970015680167, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.847738693467337e-05, | |
| "loss": 1.3298, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.034475034400366936, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.847236180904524e-05, | |
| "loss": 1.3201, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.0345603686439322, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.846733668341709e-05, | |
| "loss": 1.3874, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.034645702887497465, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.846231155778895e-05, | |
| "loss": 1.4765, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.03473103713106273, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.845728643216081e-05, | |
| "loss": 1.3658, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.034816371374627994, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.845226130653267e-05, | |
| "loss": 1.3398, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.03490170561819326, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.844723618090452e-05, | |
| "loss": 1.486, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.03498703986175852, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.84422110552764e-05, | |
| "loss": 1.4888, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.03507237410532379, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.843718592964824e-05, | |
| "loss": 1.4893, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.03515770834888905, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.84321608040201e-05, | |
| "loss": 1.49, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.03524304259245432, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.842713567839197e-05, | |
| "loss": 1.2924, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.03532837683601958, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.842211055276383e-05, | |
| "loss": 1.4775, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.035413711079584846, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.841708542713568e-05, | |
| "loss": 1.3153, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.03549904532315011, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.841206030150754e-05, | |
| "loss": 1.2189, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.035584379566715375, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.84070351758794e-05, | |
| "loss": 1.3457, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.03566971381028064, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.840201005025126e-05, | |
| "loss": 1.3557, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.035755048053845905, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.839698492462313e-05, | |
| "loss": 1.3244, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.03584038229741117, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.839195979899497e-05, | |
| "loss": 1.492, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.035925716540976434, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.838693467336684e-05, | |
| "loss": 1.324, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.036011050784541705, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.83819095477387e-05, | |
| "loss": 1.4558, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.03609638502810697, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.837688442211056e-05, | |
| "loss": 1.2914, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.036181719271672234, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 9.837185929648241e-05, | |
| "loss": 1.5288, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.0362670535152375, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.836683417085427e-05, | |
| "loss": 1.4796, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.036352387758802764, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.836180904522613e-05, | |
| "loss": 1.3065, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.03643772200236803, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.8356783919598e-05, | |
| "loss": 1.5642, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.03652305624593329, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.835175879396984e-05, | |
| "loss": 1.4511, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.03660839048949856, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.834673366834172e-05, | |
| "loss": 1.3685, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.03669372473306382, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.834170854271357e-05, | |
| "loss": 1.4452, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.036779058976629087, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.833668341708543e-05, | |
| "loss": 1.4015, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.03686439322019435, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.833165829145729e-05, | |
| "loss": 1.3698, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.036949727463759616, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.832663316582915e-05, | |
| "loss": 1.2991, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.03703506170732488, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 9.8321608040201e-05, | |
| "loss": 1.4742, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.037120395950890145, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.831658291457288e-05, | |
| "loss": 1.3428, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.03720573019445541, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.831155778894473e-05, | |
| "loss": 1.3283, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.037291064438020674, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.830653266331659e-05, | |
| "loss": 1.4286, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.03737639868158594, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.830150753768844e-05, | |
| "loss": 1.5553, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.0374617329251512, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.829648241206031e-05, | |
| "loss": 1.3982, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.03754706716871647, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.829145728643216e-05, | |
| "loss": 1.4241, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.03763240141228173, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.828643216080402e-05, | |
| "loss": 1.389, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.037717735655847, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.828140703517589e-05, | |
| "loss": 1.5677, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.03780306989941226, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.827638190954775e-05, | |
| "loss": 1.3773, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.037888404142977526, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.82713567839196e-05, | |
| "loss": 1.4541, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.03797373838654279, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.826633165829147e-05, | |
| "loss": 1.4463, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.038059072630108055, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.826130653266332e-05, | |
| "loss": 1.3858, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.03814440687367332, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.825628140703518e-05, | |
| "loss": 1.4553, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.038229741117238585, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.825125628140704e-05, | |
| "loss": 1.44, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.03831507536080385, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.82462311557789e-05, | |
| "loss": 1.4376, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.038400409604369114, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 9.824120603015075e-05, | |
| "loss": 1.5416, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.03848574384793438, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.823618090452262e-05, | |
| "loss": 1.36, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.03857107809149964, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.823115577889448e-05, | |
| "loss": 1.279, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.03865641233506491, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.822613065326634e-05, | |
| "loss": 1.3215, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.03874174657863017, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.822110552763819e-05, | |
| "loss": 1.2748, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.03882708082219544, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.821608040201005e-05, | |
| "loss": 1.3368, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.0389124150657607, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.821105527638191e-05, | |
| "loss": 1.4237, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.038997749309325966, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.820603015075378e-05, | |
| "loss": 1.4056, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.03908308355289123, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.820100502512564e-05, | |
| "loss": 1.5238, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.039168417796456495, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.819597989949749e-05, | |
| "loss": 1.4108, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.03925375204002176, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.819095477386935e-05, | |
| "loss": 1.3971, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.039339086283587024, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.818592964824121e-05, | |
| "loss": 1.5662, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.03942442052715229, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.818090452261307e-05, | |
| "loss": 1.4605, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.03950975477071755, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.817587939698492e-05, | |
| "loss": 1.4673, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.03959508901428282, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.81708542713568e-05, | |
| "loss": 2.2422, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.03968042325784808, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.816582914572864e-05, | |
| "loss": 1.3987, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.03976575750141335, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.81608040201005e-05, | |
| "loss": 1.391, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.03985109174497861, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 9.815577889447236e-05, | |
| "loss": 1.3495, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.039936425988543876, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.815075376884423e-05, | |
| "loss": 1.2783, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.04002176023210914, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.814572864321608e-05, | |
| "loss": 1.4365, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.040107094475674405, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.814070351758794e-05, | |
| "loss": 1.2481, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.04019242871923967, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.81356783919598e-05, | |
| "loss": 1.2745, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.040277762962804935, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.813065326633167e-05, | |
| "loss": 1.1942, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.0403630972063702, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.812562814070351e-05, | |
| "loss": 1.3732, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.040448431449935464, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.812060301507539e-05, | |
| "loss": 1.4524, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.04053376569350073, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.811557788944724e-05, | |
| "loss": 1.5411, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.04061909993706599, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.81105527638191e-05, | |
| "loss": 1.4422, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.04070443418063126, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.810552763819096e-05, | |
| "loss": 1.3425, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.04078976842419652, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.810050251256282e-05, | |
| "loss": 1.413, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.04087510266776179, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.809547738693467e-05, | |
| "loss": 1.3434, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.04096043691132705, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.809045226130655e-05, | |
| "loss": 1.5095, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.041045771154892316, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.80854271356784e-05, | |
| "loss": 1.1851, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.04113110539845758, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.808040201005026e-05, | |
| "loss": 1.4115, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.041216439642022845, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.807537688442211e-05, | |
| "loss": 1.2593, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.04130177388558811, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.807035175879398e-05, | |
| "loss": 1.3393, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.041387108129153374, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.806532663316583e-05, | |
| "loss": 1.5251, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.04147244237271864, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.806030150753769e-05, | |
| "loss": 1.5754, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.04155777661628391, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.805527638190956e-05, | |
| "loss": 1.35, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.041643110859849175, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.805025125628142e-05, | |
| "loss": 1.3748, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.04172844510341444, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.804522613065327e-05, | |
| "loss": 1.3923, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.041813779346979704, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.804020100502513e-05, | |
| "loss": 1.3141, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.04189911359054497, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.803517587939699e-05, | |
| "loss": 1.2544, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.04198444783411023, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.803015075376885e-05, | |
| "loss": 1.3802, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.0420697820776755, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.802512562814071e-05, | |
| "loss": 1.488, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.04215511632124076, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.802010050251256e-05, | |
| "loss": 1.5492, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.04224045056480603, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.801507537688442e-05, | |
| "loss": 1.4367, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.04232578480837129, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.801005025125629e-05, | |
| "loss": 1.3595, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.042411119051936556, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.800502512562815e-05, | |
| "loss": 1.5332, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.04249645329550182, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.3893, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.042581787539067086, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.799497487437186e-05, | |
| "loss": 1.4769, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.04266712178263235, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.798994974874372e-05, | |
| "loss": 1.453, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.29157249892352e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |