| { | |
| "best_metric": 0.3032359897985208, | |
| "best_model_checkpoint": "outputs/checkpoint-400", | |
| "epoch": 4.25531914893617, | |
| "eval_steps": 50, | |
| "global_step": 450, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009456264775413711, | |
| "grad_norm": 6.37161922454834, | |
| "learning_rate": 9.995238095238095e-06, | |
| "loss": 1.7524, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.018912529550827423, | |
| "grad_norm": 4.851662635803223, | |
| "learning_rate": 9.990476190476191e-06, | |
| "loss": 1.6245, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.028368794326241134, | |
| "grad_norm": 4.380736827850342, | |
| "learning_rate": 9.985714285714286e-06, | |
| "loss": 1.4964, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.037825059101654845, | |
| "grad_norm": 3.163944959640503, | |
| "learning_rate": 9.980952380952382e-06, | |
| "loss": 1.4517, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04728132387706856, | |
| "grad_norm": 2.9784162044525146, | |
| "learning_rate": 9.976190476190477e-06, | |
| "loss": 1.3719, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05673758865248227, | |
| "grad_norm": 2.3548669815063477, | |
| "learning_rate": 9.971428571428571e-06, | |
| "loss": 1.3009, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06619385342789598, | |
| "grad_norm": 2.4280319213867188, | |
| "learning_rate": 9.966666666666667e-06, | |
| "loss": 1.2607, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07565011820330969, | |
| "grad_norm": 1.8799192905426025, | |
| "learning_rate": 9.961904761904763e-06, | |
| "loss": 1.2404, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 1.855432152748108, | |
| "learning_rate": 9.957142857142858e-06, | |
| "loss": 1.2647, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09456264775413711, | |
| "grad_norm": 2.1619343757629395, | |
| "learning_rate": 9.952380952380954e-06, | |
| "loss": 1.2139, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10401891252955082, | |
| "grad_norm": 1.4054124355316162, | |
| "learning_rate": 9.947619047619049e-06, | |
| "loss": 1.2353, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11347517730496454, | |
| "grad_norm": 1.2696382999420166, | |
| "learning_rate": 9.942857142857145e-06, | |
| "loss": 1.2791, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12293144208037825, | |
| "grad_norm": 1.0346633195877075, | |
| "learning_rate": 9.93809523809524e-06, | |
| "loss": 1.2959, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.13238770685579196, | |
| "grad_norm": 0.8732487559318542, | |
| "learning_rate": 9.933333333333334e-06, | |
| "loss": 1.223, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14184397163120568, | |
| "grad_norm": 0.7110472917556763, | |
| "learning_rate": 9.92857142857143e-06, | |
| "loss": 1.2079, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15130023640661938, | |
| "grad_norm": 1.2797213792800903, | |
| "learning_rate": 9.923809523809524e-06, | |
| "loss": 1.1188, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1607565011820331, | |
| "grad_norm": 0.8014526963233948, | |
| "learning_rate": 9.91904761904762e-06, | |
| "loss": 1.1624, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 2.1425223350524902, | |
| "learning_rate": 9.914285714285715e-06, | |
| "loss": 1.1996, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.17966903073286053, | |
| "grad_norm": 2.180694818496704, | |
| "learning_rate": 9.90952380952381e-06, | |
| "loss": 1.094, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.18912529550827423, | |
| "grad_norm": 1.3586962223052979, | |
| "learning_rate": 9.904761904761906e-06, | |
| "loss": 1.1439, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.19858156028368795, | |
| "grad_norm": 1.3756808042526245, | |
| "learning_rate": 9.9e-06, | |
| "loss": 1.1987, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.20803782505910165, | |
| "grad_norm": 1.4752682447433472, | |
| "learning_rate": 9.895238095238096e-06, | |
| "loss": 1.2424, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.21749408983451538, | |
| "grad_norm": 1.1806340217590332, | |
| "learning_rate": 9.89047619047619e-06, | |
| "loss": 1.1481, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22695035460992907, | |
| "grad_norm": 1.5014915466308594, | |
| "learning_rate": 9.885714285714287e-06, | |
| "loss": 1.1306, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2364066193853428, | |
| "grad_norm": 2.27945613861084, | |
| "learning_rate": 9.880952380952381e-06, | |
| "loss": 1.2185, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2458628841607565, | |
| "grad_norm": 2.4887917041778564, | |
| "learning_rate": 9.876190476190478e-06, | |
| "loss": 1.1919, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 1.334818959236145, | |
| "learning_rate": 9.871428571428572e-06, | |
| "loss": 1.1375, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2647754137115839, | |
| "grad_norm": 2.0502264499664307, | |
| "learning_rate": 9.866666666666668e-06, | |
| "loss": 1.167, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.27423167848699764, | |
| "grad_norm": 2.03678822517395, | |
| "learning_rate": 9.861904761904763e-06, | |
| "loss": 1.0997, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28368794326241137, | |
| "grad_norm": 1.9366700649261475, | |
| "learning_rate": 9.857142857142859e-06, | |
| "loss": 1.0612, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29314420803782504, | |
| "grad_norm": 1.8982603549957275, | |
| "learning_rate": 9.852380952380953e-06, | |
| "loss": 1.0902, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30260047281323876, | |
| "grad_norm": 1.7486457824707031, | |
| "learning_rate": 9.847619047619048e-06, | |
| "loss": 1.149, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3120567375886525, | |
| "grad_norm": 1.8188557624816895, | |
| "learning_rate": 9.842857142857144e-06, | |
| "loss": 1.0975, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3215130023640662, | |
| "grad_norm": 3.800814151763916, | |
| "learning_rate": 9.838095238095238e-06, | |
| "loss": 1.1611, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3309692671394799, | |
| "grad_norm": 2.2901618480682373, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 1.0969, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 2.420259475708008, | |
| "learning_rate": 9.828571428571429e-06, | |
| "loss": 1.0287, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.34988179669030733, | |
| "grad_norm": 2.751361846923828, | |
| "learning_rate": 9.823809523809524e-06, | |
| "loss": 1.0103, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.35933806146572106, | |
| "grad_norm": 2.2316462993621826, | |
| "learning_rate": 9.81904761904762e-06, | |
| "loss": 1.0957, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.36879432624113473, | |
| "grad_norm": 0.9498681426048279, | |
| "learning_rate": 9.814285714285716e-06, | |
| "loss": 1.0657, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.37825059101654845, | |
| "grad_norm": 1.658214807510376, | |
| "learning_rate": 9.80952380952381e-06, | |
| "loss": 1.1366, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3877068557919622, | |
| "grad_norm": 5.218255043029785, | |
| "learning_rate": 9.804761904761907e-06, | |
| "loss": 1.1998, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3971631205673759, | |
| "grad_norm": 3.6433420181274414, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.165, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4066193853427896, | |
| "grad_norm": 2.1015524864196777, | |
| "learning_rate": 9.795238095238097e-06, | |
| "loss": 1.0357, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4160756501182033, | |
| "grad_norm": 1.6407890319824219, | |
| "learning_rate": 9.790476190476192e-06, | |
| "loss": 1.1007, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 1.929078459739685, | |
| "learning_rate": 9.785714285714286e-06, | |
| "loss": 0.9683, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43498817966903075, | |
| "grad_norm": 1.3185120820999146, | |
| "learning_rate": 9.780952380952382e-06, | |
| "loss": 1.0848, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 3.1991031169891357, | |
| "learning_rate": 9.776190476190477e-06, | |
| "loss": 1.1025, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.45390070921985815, | |
| "grad_norm": 4.7339863777160645, | |
| "learning_rate": 9.771428571428571e-06, | |
| "loss": 1.0889, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.46335697399527187, | |
| "grad_norm": 2.6043481826782227, | |
| "learning_rate": 9.766666666666667e-06, | |
| "loss": 1.1063, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4728132387706856, | |
| "grad_norm": 3.4060988426208496, | |
| "learning_rate": 9.761904761904762e-06, | |
| "loss": 1.0141, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4728132387706856, | |
| "eval_f1_macro": 0.21460474338825342, | |
| "eval_loss": 1.0437264442443848, | |
| "eval_runtime": 14.0189, | |
| "eval_samples_per_second": 120.623, | |
| "eval_steps_per_second": 7.561, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48226950354609927, | |
| "grad_norm": 1.8903930187225342, | |
| "learning_rate": 9.757142857142858e-06, | |
| "loss": 1.0172, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.491725768321513, | |
| "grad_norm": 1.3195421695709229, | |
| "learning_rate": 9.752380952380953e-06, | |
| "loss": 1.0802, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5011820330969267, | |
| "grad_norm": 2.4457075595855713, | |
| "learning_rate": 9.747619047619049e-06, | |
| "loss": 1.1387, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 1.3236303329467773, | |
| "learning_rate": 9.742857142857143e-06, | |
| "loss": 0.9655, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5200945626477541, | |
| "grad_norm": 2.286327600479126, | |
| "learning_rate": 9.73809523809524e-06, | |
| "loss": 1.0168, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5295508274231678, | |
| "grad_norm": 1.6412752866744995, | |
| "learning_rate": 9.733333333333334e-06, | |
| "loss": 1.052, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5390070921985816, | |
| "grad_norm": 2.345207452774048, | |
| "learning_rate": 9.72857142857143e-06, | |
| "loss": 1.0871, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5484633569739953, | |
| "grad_norm": 1.5813645124435425, | |
| "learning_rate": 9.723809523809525e-06, | |
| "loss": 1.1483, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.557919621749409, | |
| "grad_norm": 2.1535708904266357, | |
| "learning_rate": 9.71904761904762e-06, | |
| "loss": 1.014, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5673758865248227, | |
| "grad_norm": 1.5568801164627075, | |
| "learning_rate": 9.714285714285715e-06, | |
| "loss": 0.9503, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5768321513002365, | |
| "grad_norm": 2.258251428604126, | |
| "learning_rate": 9.70952380952381e-06, | |
| "loss": 0.8972, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5862884160756501, | |
| "grad_norm": 1.8242671489715576, | |
| "learning_rate": 9.704761904761906e-06, | |
| "loss": 1.0092, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 4.071340084075928, | |
| "learning_rate": 9.7e-06, | |
| "loss": 1.1998, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6052009456264775, | |
| "grad_norm": 2.4138002395629883, | |
| "learning_rate": 9.695238095238096e-06, | |
| "loss": 1.1186, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6146572104018913, | |
| "grad_norm": 2.901859760284424, | |
| "learning_rate": 9.690476190476191e-06, | |
| "loss": 1.104, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.624113475177305, | |
| "grad_norm": 1.9879859685897827, | |
| "learning_rate": 9.685714285714285e-06, | |
| "loss": 1.022, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6335697399527187, | |
| "grad_norm": 1.620570182800293, | |
| "learning_rate": 9.680952380952382e-06, | |
| "loss": 1.109, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6430260047281324, | |
| "grad_norm": 2.608358144760132, | |
| "learning_rate": 9.676190476190476e-06, | |
| "loss": 1.1036, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6524822695035462, | |
| "grad_norm": 1.9834874868392944, | |
| "learning_rate": 9.671428571428572e-06, | |
| "loss": 0.9812, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6619385342789598, | |
| "grad_norm": 1.2722113132476807, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 1.044, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6713947990543735, | |
| "grad_norm": 4.8911333084106445, | |
| "learning_rate": 9.661904761904763e-06, | |
| "loss": 0.9368, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 5.012025356292725, | |
| "learning_rate": 9.657142857142859e-06, | |
| "loss": 1.0675, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6903073286052009, | |
| "grad_norm": 5.291039943695068, | |
| "learning_rate": 9.652380952380954e-06, | |
| "loss": 1.128, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6997635933806147, | |
| "grad_norm": 4.000207901000977, | |
| "learning_rate": 9.647619047619048e-06, | |
| "loss": 1.0296, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7092198581560284, | |
| "grad_norm": 2.474881410598755, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.9613, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7186761229314421, | |
| "grad_norm": 2.7820417881011963, | |
| "learning_rate": 9.638095238095239e-06, | |
| "loss": 1.0414, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7281323877068558, | |
| "grad_norm": 3.6213011741638184, | |
| "learning_rate": 9.633333333333335e-06, | |
| "loss": 1.0961, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7375886524822695, | |
| "grad_norm": 2.0980260372161865, | |
| "learning_rate": 9.62857142857143e-06, | |
| "loss": 0.9454, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7470449172576832, | |
| "grad_norm": 1.899154782295227, | |
| "learning_rate": 9.623809523809524e-06, | |
| "loss": 0.9763, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7565011820330969, | |
| "grad_norm": 2.9085607528686523, | |
| "learning_rate": 9.61904761904762e-06, | |
| "loss": 0.9829, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 2.3891396522521973, | |
| "learning_rate": 9.614285714285714e-06, | |
| "loss": 1.0338, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7754137115839244, | |
| "grad_norm": 2.1058549880981445, | |
| "learning_rate": 9.60952380952381e-06, | |
| "loss": 1.095, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7848699763593381, | |
| "grad_norm": 2.2750802040100098, | |
| "learning_rate": 9.604761904761905e-06, | |
| "loss": 0.8883, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7943262411347518, | |
| "grad_norm": 2.928985834121704, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.0404, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8037825059101655, | |
| "grad_norm": 3.526047706604004, | |
| "learning_rate": 9.595238095238096e-06, | |
| "loss": 0.9821, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8132387706855791, | |
| "grad_norm": 1.2852156162261963, | |
| "learning_rate": 9.590476190476192e-06, | |
| "loss": 0.9858, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8226950354609929, | |
| "grad_norm": 1.357054591178894, | |
| "learning_rate": 9.585714285714286e-06, | |
| "loss": 0.9973, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8321513002364066, | |
| "grad_norm": 1.4836820363998413, | |
| "learning_rate": 9.580952380952383e-06, | |
| "loss": 1.0258, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8416075650118203, | |
| "grad_norm": 1.3577030897140503, | |
| "learning_rate": 9.576190476190477e-06, | |
| "loss": 0.908, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 3.018012285232544, | |
| "learning_rate": 9.571428571428573e-06, | |
| "loss": 0.9476, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8605200945626478, | |
| "grad_norm": 2.8617868423461914, | |
| "learning_rate": 9.566666666666668e-06, | |
| "loss": 1.0086, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8699763593380615, | |
| "grad_norm": 2.0398824214935303, | |
| "learning_rate": 9.561904761904762e-06, | |
| "loss": 1.0729, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8794326241134752, | |
| "grad_norm": 4.715724468231201, | |
| "learning_rate": 9.557142857142858e-06, | |
| "loss": 0.9843, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.1730473041534424, | |
| "learning_rate": 9.552380952380953e-06, | |
| "loss": 0.985, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8983451536643026, | |
| "grad_norm": 4.119668483734131, | |
| "learning_rate": 9.547619047619049e-06, | |
| "loss": 1.0455, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9078014184397163, | |
| "grad_norm": 3.797478199005127, | |
| "learning_rate": 9.542857142857143e-06, | |
| "loss": 0.9879, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.91725768321513, | |
| "grad_norm": 5.911831378936768, | |
| "learning_rate": 9.538095238095238e-06, | |
| "loss": 1.1269, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9267139479905437, | |
| "grad_norm": 5.677768230438232, | |
| "learning_rate": 9.533333333333334e-06, | |
| "loss": 1.0706, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 6.479707717895508, | |
| "learning_rate": 9.528571428571429e-06, | |
| "loss": 1.0433, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9456264775413712, | |
| "grad_norm": 3.5368762016296387, | |
| "learning_rate": 9.523809523809525e-06, | |
| "loss": 0.8951, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9456264775413712, | |
| "eval_f1_macro": 0.280761247086368, | |
| "eval_loss": 0.9932119846343994, | |
| "eval_runtime": 13.9437, | |
| "eval_samples_per_second": 121.274, | |
| "eval_steps_per_second": 7.602, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9550827423167849, | |
| "grad_norm": 5.754081726074219, | |
| "learning_rate": 9.51904761904762e-06, | |
| "loss": 1.0286, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9645390070921985, | |
| "grad_norm": 3.2837677001953125, | |
| "learning_rate": 9.514285714285715e-06, | |
| "loss": 0.9234, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9739952718676123, | |
| "grad_norm": 3.854726791381836, | |
| "learning_rate": 9.50952380952381e-06, | |
| "loss": 0.907, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.983451536643026, | |
| "grad_norm": 6.779870510101318, | |
| "learning_rate": 9.504761904761906e-06, | |
| "loss": 1.0186, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9929078014184397, | |
| "grad_norm": 4.993762493133545, | |
| "learning_rate": 9.5e-06, | |
| "loss": 1.0855, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0023640661938533, | |
| "grad_norm": 5.992312908172607, | |
| "learning_rate": 9.495238095238097e-06, | |
| "loss": 1.0172, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.011820330969267, | |
| "grad_norm": 4.238219261169434, | |
| "learning_rate": 9.490476190476191e-06, | |
| "loss": 1.0305, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 2.7353601455688477, | |
| "learning_rate": 9.485714285714287e-06, | |
| "loss": 0.973, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0307328605200945, | |
| "grad_norm": 2.2366888523101807, | |
| "learning_rate": 9.480952380952382e-06, | |
| "loss": 1.002, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0401891252955082, | |
| "grad_norm": 2.0861849784851074, | |
| "learning_rate": 9.476190476190476e-06, | |
| "loss": 1.0294, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.049645390070922, | |
| "grad_norm": 5.589237213134766, | |
| "learning_rate": 9.471428571428572e-06, | |
| "loss": 0.9939, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0591016548463357, | |
| "grad_norm": 9.459518432617188, | |
| "learning_rate": 9.466666666666667e-06, | |
| "loss": 1.045, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0685579196217494, | |
| "grad_norm": 6.007238388061523, | |
| "learning_rate": 9.461904761904761e-06, | |
| "loss": 0.9692, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0780141843971631, | |
| "grad_norm": 4.863658428192139, | |
| "learning_rate": 9.457142857142858e-06, | |
| "loss": 1.2279, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0874704491725768, | |
| "grad_norm": 1.2910066843032837, | |
| "learning_rate": 9.452380952380952e-06, | |
| "loss": 1.0343, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0969267139479906, | |
| "grad_norm": 2.361463785171509, | |
| "learning_rate": 9.447619047619048e-06, | |
| "loss": 0.936, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 1.5398199558258057, | |
| "learning_rate": 9.442857142857144e-06, | |
| "loss": 0.98, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.115839243498818, | |
| "grad_norm": 5.576039791107178, | |
| "learning_rate": 9.438095238095239e-06, | |
| "loss": 0.9781, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1252955082742317, | |
| "grad_norm": 7.037659645080566, | |
| "learning_rate": 9.433333333333335e-06, | |
| "loss": 0.9037, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1347517730496455, | |
| "grad_norm": 3.028674364089966, | |
| "learning_rate": 9.42857142857143e-06, | |
| "loss": 0.9853, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1442080378250592, | |
| "grad_norm": 2.2026021480560303, | |
| "learning_rate": 9.423809523809526e-06, | |
| "loss": 0.903, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1536643026004727, | |
| "grad_norm": 2.6655337810516357, | |
| "learning_rate": 9.41904761904762e-06, | |
| "loss": 0.915, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1631205673758864, | |
| "grad_norm": 2.7415759563446045, | |
| "learning_rate": 9.414285714285715e-06, | |
| "loss": 1.0771, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1725768321513002, | |
| "grad_norm": 2.241513252258301, | |
| "learning_rate": 9.40952380952381e-06, | |
| "loss": 0.9635, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1820330969267139, | |
| "grad_norm": 1.8755624294281006, | |
| "learning_rate": 9.404761904761905e-06, | |
| "loss": 1.0371, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 7.259213924407959, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.9636, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2009456264775413, | |
| "grad_norm": 1.6004455089569092, | |
| "learning_rate": 9.395238095238096e-06, | |
| "loss": 0.9372, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.210401891252955, | |
| "grad_norm": 1.555505394935608, | |
| "learning_rate": 9.39047619047619e-06, | |
| "loss": 0.9623, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2198581560283688, | |
| "grad_norm": 3.08105731010437, | |
| "learning_rate": 9.385714285714287e-06, | |
| "loss": 1.0623, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2293144208037825, | |
| "grad_norm": 1.6849639415740967, | |
| "learning_rate": 9.380952380952381e-06, | |
| "loss": 0.9662, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2387706855791962, | |
| "grad_norm": 1.9433168172836304, | |
| "learning_rate": 9.376190476190477e-06, | |
| "loss": 1.0246, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.24822695035461, | |
| "grad_norm": 1.203316569328308, | |
| "learning_rate": 9.371428571428572e-06, | |
| "loss": 0.9972, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2576832151300237, | |
| "grad_norm": 3.589102029800415, | |
| "learning_rate": 9.366666666666668e-06, | |
| "loss": 1.0108, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2671394799054374, | |
| "grad_norm": 3.4743452072143555, | |
| "learning_rate": 9.361904761904762e-06, | |
| "loss": 0.8865, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 4.537514686584473, | |
| "learning_rate": 9.357142857142859e-06, | |
| "loss": 0.9915, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2860520094562649, | |
| "grad_norm": 2.8337080478668213, | |
| "learning_rate": 9.352380952380953e-06, | |
| "loss": 1.1094, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2955082742316786, | |
| "grad_norm": 3.608424663543701, | |
| "learning_rate": 9.34761904761905e-06, | |
| "loss": 0.9164, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3049645390070923, | |
| "grad_norm": 1.528435468673706, | |
| "learning_rate": 9.342857142857144e-06, | |
| "loss": 1.0314, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.314420803782506, | |
| "grad_norm": 2.550269842147827, | |
| "learning_rate": 9.338095238095238e-06, | |
| "loss": 0.967, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3238770685579198, | |
| "grad_norm": 2.2891745567321777, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.9615, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.508060097694397, | |
| "learning_rate": 9.328571428571429e-06, | |
| "loss": 0.8764, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.342789598108747, | |
| "grad_norm": 2.548435688018799, | |
| "learning_rate": 9.323809523809525e-06, | |
| "loss": 0.9724, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3522458628841607, | |
| "grad_norm": 1.7150108814239502, | |
| "learning_rate": 9.31904761904762e-06, | |
| "loss": 0.9821, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 3.745039224624634, | |
| "learning_rate": 9.314285714285714e-06, | |
| "loss": 0.9439, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3711583924349882, | |
| "grad_norm": 2.843501329421997, | |
| "learning_rate": 9.30952380952381e-06, | |
| "loss": 1.0294, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3806146572104019, | |
| "grad_norm": 1.8481038808822632, | |
| "learning_rate": 9.304761904761905e-06, | |
| "loss": 1.0332, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.3900709219858156, | |
| "grad_norm": 2.5383450984954834, | |
| "learning_rate": 9.3e-06, | |
| "loss": 1.0577, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.3995271867612293, | |
| "grad_norm": 2.205739736557007, | |
| "learning_rate": 9.295238095238095e-06, | |
| "loss": 1.0294, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.408983451536643, | |
| "grad_norm": 3.889091968536377, | |
| "learning_rate": 9.290476190476191e-06, | |
| "loss": 0.8781, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4184397163120568, | |
| "grad_norm": 3.154792308807373, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.8584, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4184397163120568, | |
| "eval_f1_macro": 0.283063644436965, | |
| "eval_loss": 0.9654386043548584, | |
| "eval_runtime": 13.9617, | |
| "eval_samples_per_second": 121.117, | |
| "eval_steps_per_second": 7.592, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4278959810874705, | |
| "grad_norm": 3.941533088684082, | |
| "learning_rate": 9.280952380952382e-06, | |
| "loss": 0.9382, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4373522458628842, | |
| "grad_norm": 1.6653056144714355, | |
| "learning_rate": 9.276190476190477e-06, | |
| "loss": 0.865, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 1.8307316303253174, | |
| "learning_rate": 9.271428571428573e-06, | |
| "loss": 0.9168, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4562647754137115, | |
| "grad_norm": 5.202136993408203, | |
| "learning_rate": 9.266666666666667e-06, | |
| "loss": 0.9884, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4657210401891252, | |
| "grad_norm": 2.868016481399536, | |
| "learning_rate": 9.261904761904763e-06, | |
| "loss": 0.9221, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.475177304964539, | |
| "grad_norm": 3.4823408126831055, | |
| "learning_rate": 9.257142857142858e-06, | |
| "loss": 0.9304, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.4846335697399526, | |
| "grad_norm": 2.435922861099243, | |
| "learning_rate": 9.252380952380952e-06, | |
| "loss": 0.8973, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.4940898345153664, | |
| "grad_norm": 1.9051096439361572, | |
| "learning_rate": 9.247619047619048e-06, | |
| "loss": 0.8849, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.50354609929078, | |
| "grad_norm": 1.81594717502594, | |
| "learning_rate": 9.242857142857143e-06, | |
| "loss": 0.9916, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5130023640661938, | |
| "grad_norm": 4.773565769195557, | |
| "learning_rate": 9.238095238095239e-06, | |
| "loss": 0.9752, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5224586288416075, | |
| "grad_norm": 4.862892150878906, | |
| "learning_rate": 9.233333333333334e-06, | |
| "loss": 0.9181, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 2.8068463802337646, | |
| "learning_rate": 9.22857142857143e-06, | |
| "loss": 1.0103, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.541371158392435, | |
| "grad_norm": 5.121504783630371, | |
| "learning_rate": 9.223809523809524e-06, | |
| "loss": 1.0227, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5508274231678487, | |
| "grad_norm": 2.266847610473633, | |
| "learning_rate": 9.21904761904762e-06, | |
| "loss": 0.8633, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5602836879432624, | |
| "grad_norm": 2.262877941131592, | |
| "learning_rate": 9.214285714285715e-06, | |
| "loss": 0.8789, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5697399527186762, | |
| "grad_norm": 4.233051300048828, | |
| "learning_rate": 9.209523809523811e-06, | |
| "loss": 0.9182, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.57919621749409, | |
| "grad_norm": 2.3418633937835693, | |
| "learning_rate": 9.204761904761906e-06, | |
| "loss": 1.0088, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.5886524822695036, | |
| "grad_norm": 4.228636741638184, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.0307, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.5981087470449173, | |
| "grad_norm": 5.86262845993042, | |
| "learning_rate": 9.195238095238096e-06, | |
| "loss": 0.9996, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.607565011820331, | |
| "grad_norm": 2.881577730178833, | |
| "learning_rate": 9.19047619047619e-06, | |
| "loss": 0.9544, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 1.8167017698287964, | |
| "learning_rate": 9.185714285714287e-06, | |
| "loss": 0.8761, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6264775413711585, | |
| "grad_norm": 4.837042331695557, | |
| "learning_rate": 9.180952380952381e-06, | |
| "loss": 0.9079, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6359338061465722, | |
| "grad_norm": 1.9222673177719116, | |
| "learning_rate": 9.176190476190477e-06, | |
| "loss": 1.0179, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.645390070921986, | |
| "grad_norm": 3.050659418106079, | |
| "learning_rate": 9.171428571428572e-06, | |
| "loss": 0.8789, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6548463356973995, | |
| "grad_norm": 1.841301679611206, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.9302, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6643026004728132, | |
| "grad_norm": 5.169556140899658, | |
| "learning_rate": 9.161904761904763e-06, | |
| "loss": 0.9374, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.673758865248227, | |
| "grad_norm": 3.1582770347595215, | |
| "learning_rate": 9.157142857142857e-06, | |
| "loss": 0.9497, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.6832151300236406, | |
| "grad_norm": 2.368600606918335, | |
| "learning_rate": 9.152380952380953e-06, | |
| "loss": 0.9879, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.6926713947990544, | |
| "grad_norm": 4.020886421203613, | |
| "learning_rate": 9.147619047619048e-06, | |
| "loss": 0.9077, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 4.776878356933594, | |
| "learning_rate": 9.142857142857144e-06, | |
| "loss": 0.9397, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7115839243498818, | |
| "grad_norm": 1.6028152704238892, | |
| "learning_rate": 9.13809523809524e-06, | |
| "loss": 0.8059, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7210401891252955, | |
| "grad_norm": 1.9851771593093872, | |
| "learning_rate": 9.133333333333335e-06, | |
| "loss": 0.8532, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7304964539007093, | |
| "grad_norm": 1.7983887195587158, | |
| "learning_rate": 9.128571428571429e-06, | |
| "loss": 0.9259, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7399527186761228, | |
| "grad_norm": 3.93186616897583, | |
| "learning_rate": 9.123809523809525e-06, | |
| "loss": 0.8489, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7494089834515365, | |
| "grad_norm": 4.8172478675842285, | |
| "learning_rate": 9.11904761904762e-06, | |
| "loss": 0.978, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7588652482269502, | |
| "grad_norm": 2.9614696502685547, | |
| "learning_rate": 9.114285714285716e-06, | |
| "loss": 0.8911, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.768321513002364, | |
| "grad_norm": 2.0468883514404297, | |
| "learning_rate": 9.10952380952381e-06, | |
| "loss": 0.8729, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 2.044738531112671, | |
| "learning_rate": 9.104761904761905e-06, | |
| "loss": 0.9597, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 2.1686525344848633, | |
| "learning_rate": 9.100000000000001e-06, | |
| "loss": 0.9128, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.7966903073286051, | |
| "grad_norm": 2.1187691688537598, | |
| "learning_rate": 9.095238095238095e-06, | |
| "loss": 0.8705, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8061465721040189, | |
| "grad_norm": 3.233659029006958, | |
| "learning_rate": 9.09047619047619e-06, | |
| "loss": 0.9908, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8156028368794326, | |
| "grad_norm": 2.1533398628234863, | |
| "learning_rate": 9.085714285714286e-06, | |
| "loss": 0.9562, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8250591016548463, | |
| "grad_norm": 3.7706165313720703, | |
| "learning_rate": 9.08095238095238e-06, | |
| "loss": 1.0129, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.83451536643026, | |
| "grad_norm": 1.9210500717163086, | |
| "learning_rate": 9.076190476190477e-06, | |
| "loss": 1.0186, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8439716312056738, | |
| "grad_norm": 3.554277181625366, | |
| "learning_rate": 9.071428571428573e-06, | |
| "loss": 0.9796, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8534278959810875, | |
| "grad_norm": 2.7373600006103516, | |
| "learning_rate": 9.066666666666667e-06, | |
| "loss": 0.9081, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8628841607565012, | |
| "grad_norm": 2.0324931144714355, | |
| "learning_rate": 9.061904761904764e-06, | |
| "loss": 0.8876, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 4.527245998382568, | |
| "learning_rate": 9.057142857142858e-06, | |
| "loss": 1.0367, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.8817966903073287, | |
| "grad_norm": 2.6082801818847656, | |
| "learning_rate": 9.052380952380954e-06, | |
| "loss": 0.9492, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.8912529550827424, | |
| "grad_norm": 2.471220016479492, | |
| "learning_rate": 9.047619047619049e-06, | |
| "loss": 0.9094, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.8912529550827424, | |
| "eval_f1_macro": 0.2837259159576206, | |
| "eval_loss": 0.9514940977096558, | |
| "eval_runtime": 14.0101, | |
| "eval_samples_per_second": 120.699, | |
| "eval_steps_per_second": 7.566, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.900709219858156, | |
| "grad_norm": 1.804741382598877, | |
| "learning_rate": 9.042857142857143e-06, | |
| "loss": 0.9649, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9101654846335698, | |
| "grad_norm": 2.066596508026123, | |
| "learning_rate": 9.03809523809524e-06, | |
| "loss": 0.8746, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9196217494089836, | |
| "grad_norm": 1.635953664779663, | |
| "learning_rate": 9.033333333333334e-06, | |
| "loss": 0.9072, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9290780141843973, | |
| "grad_norm": 5.4566850662231445, | |
| "learning_rate": 9.028571428571428e-06, | |
| "loss": 0.9648, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.938534278959811, | |
| "grad_norm": 2.198424816131592, | |
| "learning_rate": 9.023809523809524e-06, | |
| "loss": 0.8508, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9479905437352247, | |
| "grad_norm": 4.491063594818115, | |
| "learning_rate": 9.019047619047619e-06, | |
| "loss": 1.0851, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 2.8760032653808594, | |
| "learning_rate": 9.014285714285715e-06, | |
| "loss": 0.8874, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.966903073286052, | |
| "grad_norm": 7.4303765296936035, | |
| "learning_rate": 9.00952380952381e-06, | |
| "loss": 0.9912, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.9763593380614657, | |
| "grad_norm": 3.4187076091766357, | |
| "learning_rate": 9.004761904761906e-06, | |
| "loss": 0.8805, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.9858156028368794, | |
| "grad_norm": 4.913360118865967, | |
| "learning_rate": 9e-06, | |
| "loss": 0.8825, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.9952718676122931, | |
| "grad_norm": 3.889186143875122, | |
| "learning_rate": 8.995238095238096e-06, | |
| "loss": 1.0933, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0047281323877066, | |
| "grad_norm": 3.284080743789673, | |
| "learning_rate": 8.990476190476191e-06, | |
| "loss": 0.9464, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0141843971631204, | |
| "grad_norm": 3.6579349040985107, | |
| "learning_rate": 8.985714285714287e-06, | |
| "loss": 0.8692, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.023640661938534, | |
| "grad_norm": 2.684208631515503, | |
| "learning_rate": 8.980952380952382e-06, | |
| "loss": 0.964, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.033096926713948, | |
| "grad_norm": 2.114290475845337, | |
| "learning_rate": 8.976190476190478e-06, | |
| "loss": 0.8868, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 3.2135798931121826, | |
| "learning_rate": 8.971428571428572e-06, | |
| "loss": 0.8269, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0520094562647753, | |
| "grad_norm": 2.392505168914795, | |
| "learning_rate": 8.966666666666667e-06, | |
| "loss": 0.8804, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.061465721040189, | |
| "grad_norm": 3.2920775413513184, | |
| "learning_rate": 8.961904761904763e-06, | |
| "loss": 0.9756, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.0709219858156027, | |
| "grad_norm": 2.386057138442993, | |
| "learning_rate": 8.957142857142857e-06, | |
| "loss": 0.9884, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.0803782505910164, | |
| "grad_norm": 3.295609712600708, | |
| "learning_rate": 8.952380952380953e-06, | |
| "loss": 0.8619, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.08983451536643, | |
| "grad_norm": 4.821883201599121, | |
| "learning_rate": 8.947619047619048e-06, | |
| "loss": 0.9719, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.099290780141844, | |
| "grad_norm": 3.452568292617798, | |
| "learning_rate": 8.942857142857142e-06, | |
| "loss": 1.017, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1087470449172576, | |
| "grad_norm": 3.7388017177581787, | |
| "learning_rate": 8.938095238095239e-06, | |
| "loss": 0.9183, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1182033096926713, | |
| "grad_norm": 2.8859570026397705, | |
| "learning_rate": 8.933333333333333e-06, | |
| "loss": 0.9246, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 6.264657974243164, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 0.9333, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.137115839243499, | |
| "grad_norm": 2.137022018432617, | |
| "learning_rate": 8.923809523809525e-06, | |
| "loss": 0.8482, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1465721040189125, | |
| "grad_norm": 3.879429817199707, | |
| "learning_rate": 8.91904761904762e-06, | |
| "loss": 0.8743, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1560283687943262, | |
| "grad_norm": 2.706341505050659, | |
| "learning_rate": 8.914285714285716e-06, | |
| "loss": 0.8389, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.16548463356974, | |
| "grad_norm": 2.056065320968628, | |
| "learning_rate": 8.90952380952381e-06, | |
| "loss": 0.8362, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.1749408983451537, | |
| "grad_norm": 3.5867486000061035, | |
| "learning_rate": 8.904761904761905e-06, | |
| "loss": 0.8927, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.1843971631205674, | |
| "grad_norm": 2.523615598678589, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 0.9123, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.193853427895981, | |
| "grad_norm": 3.221954345703125, | |
| "learning_rate": 8.895238095238096e-06, | |
| "loss": 0.8026, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.203309692671395, | |
| "grad_norm": 5.576043128967285, | |
| "learning_rate": 8.890476190476192e-06, | |
| "loss": 1.0363, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 3.328744411468506, | |
| "learning_rate": 8.885714285714286e-06, | |
| "loss": 0.9239, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 4.1029372215271, | |
| "learning_rate": 8.88095238095238e-06, | |
| "loss": 0.9743, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.231678486997636, | |
| "grad_norm": 7.877633571624756, | |
| "learning_rate": 8.876190476190477e-06, | |
| "loss": 0.9407, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2411347517730498, | |
| "grad_norm": 6.3463311195373535, | |
| "learning_rate": 8.871428571428571e-06, | |
| "loss": 0.8537, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2505910165484635, | |
| "grad_norm": 9.1329345703125, | |
| "learning_rate": 8.866666666666668e-06, | |
| "loss": 0.8501, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.260047281323877, | |
| "grad_norm": 6.020025253295898, | |
| "learning_rate": 8.861904761904762e-06, | |
| "loss": 0.876, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.269503546099291, | |
| "grad_norm": 2.44421124458313, | |
| "learning_rate": 8.857142857142858e-06, | |
| "loss": 0.9192, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.2789598108747047, | |
| "grad_norm": 6.033252239227295, | |
| "learning_rate": 8.852380952380953e-06, | |
| "loss": 0.9477, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.2884160756501184, | |
| "grad_norm": 10.110373497009277, | |
| "learning_rate": 8.847619047619049e-06, | |
| "loss": 1.0997, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 7.680680274963379, | |
| "learning_rate": 8.842857142857143e-06, | |
| "loss": 0.8948, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3073286052009454, | |
| "grad_norm": 9.439537048339844, | |
| "learning_rate": 8.83809523809524e-06, | |
| "loss": 0.9058, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3167848699763596, | |
| "grad_norm": 8.714299201965332, | |
| "learning_rate": 8.833333333333334e-06, | |
| "loss": 0.8828, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.326241134751773, | |
| "grad_norm": 2.600574254989624, | |
| "learning_rate": 8.82857142857143e-06, | |
| "loss": 0.9187, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3356973995271866, | |
| "grad_norm": 4.401780605316162, | |
| "learning_rate": 8.823809523809525e-06, | |
| "loss": 0.8681, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3451536643026003, | |
| "grad_norm": 2.297525405883789, | |
| "learning_rate": 8.819047619047619e-06, | |
| "loss": 0.8172, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.354609929078014, | |
| "grad_norm": 4.044796943664551, | |
| "learning_rate": 8.814285714285715e-06, | |
| "loss": 0.8994, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3640661938534278, | |
| "grad_norm": 8.97205924987793, | |
| "learning_rate": 8.80952380952381e-06, | |
| "loss": 0.9875, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.3640661938534278, | |
| "eval_f1_macro": 0.2946908943355175, | |
| "eval_loss": 0.9593666791915894, | |
| "eval_runtime": 13.9875, | |
| "eval_samples_per_second": 120.893, | |
| "eval_steps_per_second": 7.578, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.3735224586288415, | |
| "grad_norm": 2.208160161972046, | |
| "learning_rate": 8.804761904761906e-06, | |
| "loss": 0.8831, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 5.02908182144165, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.9052, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.392434988179669, | |
| "grad_norm": 3.6499717235565186, | |
| "learning_rate": 8.795238095238095e-06, | |
| "loss": 0.8577, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4018912529550827, | |
| "grad_norm": 5.426293849945068, | |
| "learning_rate": 8.790476190476191e-06, | |
| "loss": 0.9268, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4113475177304964, | |
| "grad_norm": 8.641587257385254, | |
| "learning_rate": 8.785714285714286e-06, | |
| "loss": 0.9257, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.42080378250591, | |
| "grad_norm": 11.271356582641602, | |
| "learning_rate": 8.780952380952382e-06, | |
| "loss": 0.9812, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.430260047281324, | |
| "grad_norm": 8.257451057434082, | |
| "learning_rate": 8.776190476190476e-06, | |
| "loss": 0.9077, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4397163120567376, | |
| "grad_norm": 8.181621551513672, | |
| "learning_rate": 8.771428571428572e-06, | |
| "loss": 1.0347, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4491725768321513, | |
| "grad_norm": 2.263065814971924, | |
| "learning_rate": 8.766666666666669e-06, | |
| "loss": 0.8987, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.458628841607565, | |
| "grad_norm": 4.538362979888916, | |
| "learning_rate": 8.761904761904763e-06, | |
| "loss": 0.9275, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 10.531721115112305, | |
| "learning_rate": 8.757142857142858e-06, | |
| "loss": 0.8979, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.4775413711583925, | |
| "grad_norm": 4.876132011413574, | |
| "learning_rate": 8.752380952380954e-06, | |
| "loss": 0.8281, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.486997635933806, | |
| "grad_norm": 2.9724793434143066, | |
| "learning_rate": 8.747619047619048e-06, | |
| "loss": 0.7766, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.49645390070922, | |
| "grad_norm": 3.8613243103027344, | |
| "learning_rate": 8.742857142857144e-06, | |
| "loss": 0.939, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.5059101654846336, | |
| "grad_norm": 2.186553478240967, | |
| "learning_rate": 8.738095238095239e-06, | |
| "loss": 0.8761, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5153664302600474, | |
| "grad_norm": 3.3984861373901367, | |
| "learning_rate": 8.733333333333333e-06, | |
| "loss": 0.8638, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.524822695035461, | |
| "grad_norm": 6.405703067779541, | |
| "learning_rate": 8.72857142857143e-06, | |
| "loss": 0.8797, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.534278959810875, | |
| "grad_norm": 5.020063400268555, | |
| "learning_rate": 8.723809523809524e-06, | |
| "loss": 0.827, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5437352245862885, | |
| "grad_norm": 2.5324349403381348, | |
| "learning_rate": 8.71904761904762e-06, | |
| "loss": 0.9493, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 2.4035396575927734, | |
| "learning_rate": 8.714285714285715e-06, | |
| "loss": 0.929, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.562647754137116, | |
| "grad_norm": 3.447455644607544, | |
| "learning_rate": 8.70952380952381e-06, | |
| "loss": 0.9404, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.5721040189125297, | |
| "grad_norm": 2.383104085922241, | |
| "learning_rate": 8.704761904761905e-06, | |
| "loss": 0.8185, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.581560283687943, | |
| "grad_norm": 2.3809287548065186, | |
| "learning_rate": 8.700000000000001e-06, | |
| "loss": 0.9141, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.591016548463357, | |
| "grad_norm": 2.1600184440612793, | |
| "learning_rate": 8.695238095238096e-06, | |
| "loss": 0.8376, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.6004728132387704, | |
| "grad_norm": 3.5827438831329346, | |
| "learning_rate": 8.690476190476192e-06, | |
| "loss": 0.9028, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6099290780141846, | |
| "grad_norm": 2.5485596656799316, | |
| "learning_rate": 8.685714285714287e-06, | |
| "loss": 0.8242, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.619385342789598, | |
| "grad_norm": 5.618887901306152, | |
| "learning_rate": 8.680952380952383e-06, | |
| "loss": 0.9673, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.628841607565012, | |
| "grad_norm": 4.998472690582275, | |
| "learning_rate": 8.676190476190477e-06, | |
| "loss": 0.9069, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 5.121683597564697, | |
| "learning_rate": 8.671428571428572e-06, | |
| "loss": 0.9957, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6477541371158395, | |
| "grad_norm": 3.0463051795959473, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.772, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.657210401891253, | |
| "grad_norm": 2.2069895267486572, | |
| "learning_rate": 8.661904761904762e-06, | |
| "loss": 0.8402, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 3.6136813163757324, | |
| "learning_rate": 8.657142857142858e-06, | |
| "loss": 0.9181, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.6761229314420802, | |
| "grad_norm": 3.3590173721313477, | |
| "learning_rate": 8.652380952380953e-06, | |
| "loss": 0.9341, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.685579196217494, | |
| "grad_norm": 7.388670444488525, | |
| "learning_rate": 8.647619047619047e-06, | |
| "loss": 0.8949, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.6950354609929077, | |
| "grad_norm": 4.200950622558594, | |
| "learning_rate": 8.642857142857144e-06, | |
| "loss": 0.9149, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7044917257683214, | |
| "grad_norm": 4.610495090484619, | |
| "learning_rate": 8.638095238095238e-06, | |
| "loss": 0.918, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.713947990543735, | |
| "grad_norm": 3.4695537090301514, | |
| "learning_rate": 8.633333333333334e-06, | |
| "loss": 0.9051, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 5.142923355102539, | |
| "learning_rate": 8.628571428571429e-06, | |
| "loss": 0.9038, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7328605200945626, | |
| "grad_norm": 3.3170006275177, | |
| "learning_rate": 8.623809523809525e-06, | |
| "loss": 0.906, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.7423167848699763, | |
| "grad_norm": 2.6813061237335205, | |
| "learning_rate": 8.61904761904762e-06, | |
| "loss": 0.8809, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.75177304964539, | |
| "grad_norm": 3.356801748275757, | |
| "learning_rate": 8.614285714285716e-06, | |
| "loss": 1.1035, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7612293144208038, | |
| "grad_norm": 7.907340049743652, | |
| "learning_rate": 8.60952380952381e-06, | |
| "loss": 0.9155, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.7706855791962175, | |
| "grad_norm": 2.931204080581665, | |
| "learning_rate": 8.604761904761906e-06, | |
| "loss": 0.9312, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.780141843971631, | |
| "grad_norm": 4.311042308807373, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.0181, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.789598108747045, | |
| "grad_norm": 5.058907508850098, | |
| "learning_rate": 8.595238095238097e-06, | |
| "loss": 0.7962, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.7990543735224587, | |
| "grad_norm": 3.870861053466797, | |
| "learning_rate": 8.590476190476191e-06, | |
| "loss": 0.9256, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 6.868628025054932, | |
| "learning_rate": 8.585714285714286e-06, | |
| "loss": 0.9178, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.817966903073286, | |
| "grad_norm": 3.905592679977417, | |
| "learning_rate": 8.580952380952382e-06, | |
| "loss": 0.9549, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8274231678487, | |
| "grad_norm": 5.346381664276123, | |
| "learning_rate": 8.576190476190476e-06, | |
| "loss": 0.9538, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8368794326241136, | |
| "grad_norm": 4.774893283843994, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.972, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8368794326241136, | |
| "eval_f1_macro": 0.2760298327473914, | |
| "eval_loss": 0.9529216885566711, | |
| "eval_runtime": 13.9542, | |
| "eval_samples_per_second": 121.182, | |
| "eval_steps_per_second": 7.596, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8463356973995273, | |
| "grad_norm": 5.011666297912598, | |
| "learning_rate": 8.566666666666667e-06, | |
| "loss": 0.8952, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.855791962174941, | |
| "grad_norm": 4.010011196136475, | |
| "learning_rate": 8.561904761904762e-06, | |
| "loss": 0.8803, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.8652482269503547, | |
| "grad_norm": 3.0806877613067627, | |
| "learning_rate": 8.557142857142858e-06, | |
| "loss": 0.8768, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.8747044917257685, | |
| "grad_norm": 2.15812349319458, | |
| "learning_rate": 8.552380952380954e-06, | |
| "loss": 0.9884, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.884160756501182, | |
| "grad_norm": 4.941656112670898, | |
| "learning_rate": 8.547619047619048e-06, | |
| "loss": 0.9848, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 7.0383100509643555, | |
| "learning_rate": 8.542857142857145e-06, | |
| "loss": 0.9367, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9030732860520096, | |
| "grad_norm": 4.573930263519287, | |
| "learning_rate": 8.538095238095239e-06, | |
| "loss": 0.882, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.912529550827423, | |
| "grad_norm": 2.9290146827697754, | |
| "learning_rate": 8.533333333333335e-06, | |
| "loss": 0.9076, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.921985815602837, | |
| "grad_norm": 2.3926682472229004, | |
| "learning_rate": 8.52857142857143e-06, | |
| "loss": 0.9358, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9314420803782504, | |
| "grad_norm": 2.4800093173980713, | |
| "learning_rate": 8.523809523809524e-06, | |
| "loss": 0.8903, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9408983451536646, | |
| "grad_norm": 4.937615394592285, | |
| "learning_rate": 8.51904761904762e-06, | |
| "loss": 0.9122, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.950354609929078, | |
| "grad_norm": 3.547682285308838, | |
| "learning_rate": 8.514285714285715e-06, | |
| "loss": 0.8261, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.959810874704492, | |
| "grad_norm": 2.7049429416656494, | |
| "learning_rate": 8.50952380952381e-06, | |
| "loss": 0.9048, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.9692671394799053, | |
| "grad_norm": 3.8899621963500977, | |
| "learning_rate": 8.504761904761905e-06, | |
| "loss": 0.9234, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 6.804818153381348, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.9218, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.9881796690307327, | |
| "grad_norm": 5.924713611602783, | |
| "learning_rate": 8.495238095238096e-06, | |
| "loss": 0.8945, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.9976359338061465, | |
| "grad_norm": 6.0556254386901855, | |
| "learning_rate": 8.49047619047619e-06, | |
| "loss": 0.783, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.00709219858156, | |
| "grad_norm": 3.446620225906372, | |
| "learning_rate": 8.485714285714287e-06, | |
| "loss": 0.9164, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.016548463356974, | |
| "grad_norm": 2.734196901321411, | |
| "learning_rate": 8.480952380952381e-06, | |
| "loss": 0.8043, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.0260047281323876, | |
| "grad_norm": 3.3837802410125732, | |
| "learning_rate": 8.476190476190477e-06, | |
| "loss": 0.7815, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.0354609929078014, | |
| "grad_norm": 2.1560587882995605, | |
| "learning_rate": 8.471428571428572e-06, | |
| "loss": 0.7651, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.044917257683215, | |
| "grad_norm": 5.292918682098389, | |
| "learning_rate": 8.466666666666668e-06, | |
| "loss": 0.8416, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.054373522458629, | |
| "grad_norm": 3.834503412246704, | |
| "learning_rate": 8.461904761904763e-06, | |
| "loss": 0.8658, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.0638297872340425, | |
| "grad_norm": 5.551731109619141, | |
| "learning_rate": 8.457142857142859e-06, | |
| "loss": 0.8834, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.0732860520094563, | |
| "grad_norm": 8.869498252868652, | |
| "learning_rate": 8.452380952380953e-06, | |
| "loss": 0.9549, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.08274231678487, | |
| "grad_norm": 3.6207432746887207, | |
| "learning_rate": 8.447619047619048e-06, | |
| "loss": 0.8453, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.0921985815602837, | |
| "grad_norm": 5.913300037384033, | |
| "learning_rate": 8.442857142857144e-06, | |
| "loss": 0.973, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.1016548463356974, | |
| "grad_norm": 6.6758623123168945, | |
| "learning_rate": 8.438095238095238e-06, | |
| "loss": 0.9095, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 3.1101748943328857, | |
| "learning_rate": 8.433333333333334e-06, | |
| "loss": 0.8741, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.120567375886525, | |
| "grad_norm": 2.8428375720977783, | |
| "learning_rate": 8.428571428571429e-06, | |
| "loss": 0.8203, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.1300236406619386, | |
| "grad_norm": 3.4043846130371094, | |
| "learning_rate": 8.423809523809523e-06, | |
| "loss": 0.9088, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.1394799054373523, | |
| "grad_norm": 3.4590537548065186, | |
| "learning_rate": 8.41904761904762e-06, | |
| "loss": 0.8837, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.148936170212766, | |
| "grad_norm": 3.684206485748291, | |
| "learning_rate": 8.414285714285714e-06, | |
| "loss": 0.9075, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.15839243498818, | |
| "grad_norm": 6.9007720947265625, | |
| "learning_rate": 8.40952380952381e-06, | |
| "loss": 0.8656, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.1678486997635935, | |
| "grad_norm": 4.262526988983154, | |
| "learning_rate": 8.404761904761905e-06, | |
| "loss": 0.959, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.1773049645390072, | |
| "grad_norm": 6.345574378967285, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.8998, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.186761229314421, | |
| "grad_norm": 4.99959659576416, | |
| "learning_rate": 8.395238095238097e-06, | |
| "loss": 0.8619, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.1962174940898347, | |
| "grad_norm": 4.0740437507629395, | |
| "learning_rate": 8.390476190476192e-06, | |
| "loss": 0.7907, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.2056737588652484, | |
| "grad_norm": 3.4886386394500732, | |
| "learning_rate": 8.385714285714286e-06, | |
| "loss": 0.9112, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.215130023640662, | |
| "grad_norm": 3.8843610286712646, | |
| "learning_rate": 8.380952380952382e-06, | |
| "loss": 0.8705, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.2245862884160754, | |
| "grad_norm": 5.011359691619873, | |
| "learning_rate": 8.376190476190477e-06, | |
| "loss": 0.874, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.2340425531914896, | |
| "grad_norm": 4.989591121673584, | |
| "learning_rate": 8.371428571428573e-06, | |
| "loss": 0.9204, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.243498817966903, | |
| "grad_norm": 3.6091649532318115, | |
| "learning_rate": 8.366666666666667e-06, | |
| "loss": 0.8482, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.2529550827423166, | |
| "grad_norm": 2.7679734230041504, | |
| "learning_rate": 8.361904761904762e-06, | |
| "loss": 0.8709, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.2624113475177303, | |
| "grad_norm": 3.253309726715088, | |
| "learning_rate": 8.357142857142858e-06, | |
| "loss": 0.8301, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.271867612293144, | |
| "grad_norm": 3.695540189743042, | |
| "learning_rate": 8.352380952380952e-06, | |
| "loss": 0.7988, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.2813238770685578, | |
| "grad_norm": 6.305734157562256, | |
| "learning_rate": 8.347619047619049e-06, | |
| "loss": 0.8561, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.2907801418439715, | |
| "grad_norm": 4.090425491333008, | |
| "learning_rate": 8.342857142857143e-06, | |
| "loss": 0.8819, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.300236406619385, | |
| "grad_norm": 8.310733795166016, | |
| "learning_rate": 8.33809523809524e-06, | |
| "loss": 0.8018, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.309692671394799, | |
| "grad_norm": 7.966689109802246, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.9757, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.309692671394799, | |
| "eval_f1_macro": 0.2898114936552929, | |
| "eval_loss": 0.9974462985992432, | |
| "eval_runtime": 13.9296, | |
| "eval_samples_per_second": 121.396, | |
| "eval_steps_per_second": 7.61, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.3191489361702127, | |
| "grad_norm": 5.680312156677246, | |
| "learning_rate": 8.32857142857143e-06, | |
| "loss": 0.9264, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.3286052009456264, | |
| "grad_norm": 4.721024513244629, | |
| "learning_rate": 8.323809523809524e-06, | |
| "loss": 0.8284, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.33806146572104, | |
| "grad_norm": 3.5236101150512695, | |
| "learning_rate": 8.31904761904762e-06, | |
| "loss": 0.823, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.347517730496454, | |
| "grad_norm": 4.502352237701416, | |
| "learning_rate": 8.314285714285715e-06, | |
| "loss": 0.8426, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.3569739952718676, | |
| "grad_norm": 3.1332571506500244, | |
| "learning_rate": 8.309523809523811e-06, | |
| "loss": 0.7419, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.3664302600472813, | |
| "grad_norm": 11.218631744384766, | |
| "learning_rate": 8.304761904761906e-06, | |
| "loss": 0.8824, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.375886524822695, | |
| "grad_norm": 14.600545883178711, | |
| "learning_rate": 8.3e-06, | |
| "loss": 0.8601, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.3853427895981087, | |
| "grad_norm": 12.865662574768066, | |
| "learning_rate": 8.295238095238096e-06, | |
| "loss": 0.8797, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.3947990543735225, | |
| "grad_norm": 11.967954635620117, | |
| "learning_rate": 8.29047619047619e-06, | |
| "loss": 1.0035, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.404255319148936, | |
| "grad_norm": 4.384112358093262, | |
| "learning_rate": 8.285714285714287e-06, | |
| "loss": 0.9454, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.41371158392435, | |
| "grad_norm": 4.802806854248047, | |
| "learning_rate": 8.280952380952381e-06, | |
| "loss": 0.804, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.4231678486997636, | |
| "grad_norm": 5.697473526000977, | |
| "learning_rate": 8.276190476190476e-06, | |
| "loss": 0.936, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.4326241134751774, | |
| "grad_norm": 5.372440814971924, | |
| "learning_rate": 8.271428571428572e-06, | |
| "loss": 0.8373, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.442080378250591, | |
| "grad_norm": 6.925962924957275, | |
| "learning_rate": 8.266666666666667e-06, | |
| "loss": 0.8763, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.451536643026005, | |
| "grad_norm": 6.610288143157959, | |
| "learning_rate": 8.261904761904763e-06, | |
| "loss": 0.8084, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.4609929078014185, | |
| "grad_norm": 3.974923849105835, | |
| "learning_rate": 8.257142857142857e-06, | |
| "loss": 0.9025, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.4704491725768323, | |
| "grad_norm": 8.421207427978516, | |
| "learning_rate": 8.252380952380953e-06, | |
| "loss": 0.8086, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.479905437352246, | |
| "grad_norm": 11.004073143005371, | |
| "learning_rate": 8.24761904761905e-06, | |
| "loss": 0.9149, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.4893617021276597, | |
| "grad_norm": 6.909345626831055, | |
| "learning_rate": 8.242857142857144e-06, | |
| "loss": 0.7801, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.4988179669030735, | |
| "grad_norm": 4.8476386070251465, | |
| "learning_rate": 8.238095238095239e-06, | |
| "loss": 0.851, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.508274231678487, | |
| "grad_norm": 3.323458433151245, | |
| "learning_rate": 8.233333333333335e-06, | |
| "loss": 0.8203, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.5177304964539005, | |
| "grad_norm": 3.748945951461792, | |
| "learning_rate": 8.22857142857143e-06, | |
| "loss": 0.8316, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.5271867612293146, | |
| "grad_norm": 4.654240608215332, | |
| "learning_rate": 8.223809523809525e-06, | |
| "loss": 0.8346, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.536643026004728, | |
| "grad_norm": 2.6797022819519043, | |
| "learning_rate": 8.21904761904762e-06, | |
| "loss": 0.8317, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.546099290780142, | |
| "grad_norm": 7.341446876525879, | |
| "learning_rate": 8.214285714285714e-06, | |
| "loss": 0.7814, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 14.088275909423828, | |
| "learning_rate": 8.20952380952381e-06, | |
| "loss": 0.8655, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.5650118203309695, | |
| "grad_norm": 4.27020263671875, | |
| "learning_rate": 8.204761904761905e-06, | |
| "loss": 0.9051, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.574468085106383, | |
| "grad_norm": 4.922997951507568, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.7647, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.5839243498817965, | |
| "grad_norm": 3.8162479400634766, | |
| "learning_rate": 8.195238095238096e-06, | |
| "loss": 0.8543, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.5933806146572103, | |
| "grad_norm": 3.899261713027954, | |
| "learning_rate": 8.190476190476192e-06, | |
| "loss": 0.8353, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.602836879432624, | |
| "grad_norm": 8.860682487487793, | |
| "learning_rate": 8.185714285714286e-06, | |
| "loss": 0.8721, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.6122931442080377, | |
| "grad_norm": 6.590539455413818, | |
| "learning_rate": 8.180952380952382e-06, | |
| "loss": 0.8556, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.6217494089834514, | |
| "grad_norm": 8.594223976135254, | |
| "learning_rate": 8.176190476190477e-06, | |
| "loss": 0.8506, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.631205673758865, | |
| "grad_norm": 5.876694679260254, | |
| "learning_rate": 8.171428571428573e-06, | |
| "loss": 0.6882, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.640661938534279, | |
| "grad_norm": 10.495857238769531, | |
| "learning_rate": 8.166666666666668e-06, | |
| "loss": 0.8772, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.6501182033096926, | |
| "grad_norm": 10.53131103515625, | |
| "learning_rate": 8.161904761904764e-06, | |
| "loss": 0.8082, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.6595744680851063, | |
| "grad_norm": 9.257997512817383, | |
| "learning_rate": 8.157142857142858e-06, | |
| "loss": 0.9248, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.66903073286052, | |
| "grad_norm": 14.692602157592773, | |
| "learning_rate": 8.152380952380953e-06, | |
| "loss": 0.8357, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.678486997635934, | |
| "grad_norm": 7.023383617401123, | |
| "learning_rate": 8.147619047619049e-06, | |
| "loss": 0.83, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.6879432624113475, | |
| "grad_norm": 4.340334415435791, | |
| "learning_rate": 8.142857142857143e-06, | |
| "loss": 0.849, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.6973995271867612, | |
| "grad_norm": 5.631707191467285, | |
| "learning_rate": 8.138095238095238e-06, | |
| "loss": 0.8456, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.706855791962175, | |
| "grad_norm": 7.735780239105225, | |
| "learning_rate": 8.133333333333334e-06, | |
| "loss": 0.9483, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.7163120567375887, | |
| "grad_norm": 6.3525519371032715, | |
| "learning_rate": 8.128571428571428e-06, | |
| "loss": 0.8846, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.7257683215130024, | |
| "grad_norm": 9.425445556640625, | |
| "learning_rate": 8.123809523809525e-06, | |
| "loss": 0.8566, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.735224586288416, | |
| "grad_norm": 5.754733085632324, | |
| "learning_rate": 8.119047619047619e-06, | |
| "loss": 0.8363, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.74468085106383, | |
| "grad_norm": 3.9036519527435303, | |
| "learning_rate": 8.114285714285715e-06, | |
| "loss": 1.0731, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.7541371158392436, | |
| "grad_norm": 3.9182815551757812, | |
| "learning_rate": 8.10952380952381e-06, | |
| "loss": 1.0062, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.7635933806146573, | |
| "grad_norm": 10.651973724365234, | |
| "learning_rate": 8.104761904761906e-06, | |
| "loss": 0.8626, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.773049645390071, | |
| "grad_norm": 10.012125968933105, | |
| "learning_rate": 8.1e-06, | |
| "loss": 0.9117, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.7825059101654848, | |
| "grad_norm": 9.443008422851562, | |
| "learning_rate": 8.095238095238097e-06, | |
| "loss": 0.8736, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.7825059101654848, | |
| "eval_f1_macro": 0.3032359897985208, | |
| "eval_loss": 0.9756927490234375, | |
| "eval_runtime": 13.9326, | |
| "eval_samples_per_second": 121.37, | |
| "eval_steps_per_second": 7.608, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.7919621749408985, | |
| "grad_norm": 6.038490295410156, | |
| "learning_rate": 8.090476190476191e-06, | |
| "loss": 0.8401, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.801418439716312, | |
| "grad_norm": 4.123776912689209, | |
| "learning_rate": 8.085714285714287e-06, | |
| "loss": 0.8569, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.8108747044917255, | |
| "grad_norm": 3.2641398906707764, | |
| "learning_rate": 8.080952380952382e-06, | |
| "loss": 0.7882, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.8203309692671397, | |
| "grad_norm": 7.038326263427734, | |
| "learning_rate": 8.076190476190476e-06, | |
| "loss": 0.7609, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.829787234042553, | |
| "grad_norm": 10.111557960510254, | |
| "learning_rate": 8.071428571428572e-06, | |
| "loss": 0.853, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.839243498817967, | |
| "grad_norm": 8.650494575500488, | |
| "learning_rate": 8.066666666666667e-06, | |
| "loss": 0.8044, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.8486997635933804, | |
| "grad_norm": 10.700185775756836, | |
| "learning_rate": 8.061904761904763e-06, | |
| "loss": 0.9259, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.8581560283687946, | |
| "grad_norm": 3.3702189922332764, | |
| "learning_rate": 8.057142857142857e-06, | |
| "loss": 0.7111, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.867612293144208, | |
| "grad_norm": 4.435060024261475, | |
| "learning_rate": 8.052380952380952e-06, | |
| "loss": 0.7695, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.877068557919622, | |
| "grad_norm": 8.08366870880127, | |
| "learning_rate": 8.047619047619048e-06, | |
| "loss": 0.843, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.8865248226950353, | |
| "grad_norm": 6.264552116394043, | |
| "learning_rate": 8.042857142857143e-06, | |
| "loss": 0.8128, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.895981087470449, | |
| "grad_norm": 4.465488433837891, | |
| "learning_rate": 8.038095238095239e-06, | |
| "loss": 0.7762, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.9054373522458627, | |
| "grad_norm": 4.775088310241699, | |
| "learning_rate": 8.033333333333335e-06, | |
| "loss": 0.9124, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.9148936170212765, | |
| "grad_norm": 3.0204665660858154, | |
| "learning_rate": 8.02857142857143e-06, | |
| "loss": 0.7446, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.92434988179669, | |
| "grad_norm": 7.028670787811279, | |
| "learning_rate": 8.023809523809526e-06, | |
| "loss": 0.8979, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.933806146572104, | |
| "grad_norm": 7.230231285095215, | |
| "learning_rate": 8.01904761904762e-06, | |
| "loss": 0.8861, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.9432624113475176, | |
| "grad_norm": 6.362856388092041, | |
| "learning_rate": 8.014285714285715e-06, | |
| "loss": 0.8104, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.9527186761229314, | |
| "grad_norm": 4.598600387573242, | |
| "learning_rate": 8.00952380952381e-06, | |
| "loss": 0.6737, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.962174940898345, | |
| "grad_norm": 5.961977481842041, | |
| "learning_rate": 8.004761904761905e-06, | |
| "loss": 0.8154, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.971631205673759, | |
| "grad_norm": 4.175229072570801, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.7205, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.9810874704491725, | |
| "grad_norm": 6.275330543518066, | |
| "learning_rate": 7.995238095238096e-06, | |
| "loss": 0.8317, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.9905437352245863, | |
| "grad_norm": 9.173235893249512, | |
| "learning_rate": 7.99047619047619e-06, | |
| "loss": 0.76, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.785587310791016, | |
| "learning_rate": 7.985714285714286e-06, | |
| "loss": 0.9516, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 4.009456264775413, | |
| "grad_norm": 7.507620334625244, | |
| "learning_rate": 7.980952380952381e-06, | |
| "loss": 0.7683, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.0189125295508275, | |
| "grad_norm": 5.261313438415527, | |
| "learning_rate": 7.976190476190477e-06, | |
| "loss": 0.7983, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.028368794326241, | |
| "grad_norm": 6.864541053771973, | |
| "learning_rate": 7.971428571428572e-06, | |
| "loss": 0.9051, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 4.037825059101655, | |
| "grad_norm": 5.394888401031494, | |
| "learning_rate": 7.966666666666668e-06, | |
| "loss": 0.6639, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 4.047281323877068, | |
| "grad_norm": 7.618609428405762, | |
| "learning_rate": 7.961904761904762e-06, | |
| "loss": 0.8791, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 4.056737588652482, | |
| "grad_norm": 7.359228134155273, | |
| "learning_rate": 7.957142857142858e-06, | |
| "loss": 0.9106, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 4.066193853427896, | |
| "grad_norm": 3.4456787109375, | |
| "learning_rate": 7.952380952380953e-06, | |
| "loss": 0.7331, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.07565011820331, | |
| "grad_norm": 4.019674777984619, | |
| "learning_rate": 7.947619047619049e-06, | |
| "loss": 0.7135, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 4.085106382978723, | |
| "grad_norm": 6.423426151275635, | |
| "learning_rate": 7.942857142857144e-06, | |
| "loss": 0.7535, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 4.094562647754137, | |
| "grad_norm": 6.142768383026123, | |
| "learning_rate": 7.93809523809524e-06, | |
| "loss": 0.7536, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 4.1040189125295505, | |
| "grad_norm": 5.819328784942627, | |
| "learning_rate": 7.933333333333334e-06, | |
| "loss": 0.694, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 4.113475177304965, | |
| "grad_norm": 8.63968276977539, | |
| "learning_rate": 7.928571428571429e-06, | |
| "loss": 0.6545, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 4.122931442080378, | |
| "grad_norm": 3.8477113246917725, | |
| "learning_rate": 7.923809523809525e-06, | |
| "loss": 0.7708, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 4.132387706855792, | |
| "grad_norm": 7.744399070739746, | |
| "learning_rate": 7.91904761904762e-06, | |
| "loss": 0.6903, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 4.141843971631205, | |
| "grad_norm": 4.299695014953613, | |
| "learning_rate": 7.914285714285715e-06, | |
| "loss": 0.7932, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.15130023640662, | |
| "grad_norm": 4.58480978012085, | |
| "learning_rate": 7.90952380952381e-06, | |
| "loss": 0.7599, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 4.160756501182033, | |
| "grad_norm": 4.495441436767578, | |
| "learning_rate": 7.904761904761904e-06, | |
| "loss": 0.8248, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.170212765957447, | |
| "grad_norm": 9.265257835388184, | |
| "learning_rate": 7.9e-06, | |
| "loss": 0.8237, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 4.17966903073286, | |
| "grad_norm": 15.275799751281738, | |
| "learning_rate": 7.895238095238095e-06, | |
| "loss": 0.7988, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 4.1891252955082745, | |
| "grad_norm": 13.653985023498535, | |
| "learning_rate": 7.890476190476191e-06, | |
| "loss": 0.6772, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 4.198581560283688, | |
| "grad_norm": 7.599392890930176, | |
| "learning_rate": 7.885714285714286e-06, | |
| "loss": 0.8561, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.208037825059102, | |
| "grad_norm": 4.898796081542969, | |
| "learning_rate": 7.880952380952382e-06, | |
| "loss": 0.7731, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 4.217494089834515, | |
| "grad_norm": 10.565406799316406, | |
| "learning_rate": 7.876190476190478e-06, | |
| "loss": 0.7069, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.226950354609929, | |
| "grad_norm": 14.41064167022705, | |
| "learning_rate": 7.871428571428573e-06, | |
| "loss": 0.9041, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 4.236406619385343, | |
| "grad_norm": 10.0387544631958, | |
| "learning_rate": 7.866666666666667e-06, | |
| "loss": 0.7182, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.245862884160757, | |
| "grad_norm": 8.706192016601562, | |
| "learning_rate": 7.861904761904763e-06, | |
| "loss": 0.7909, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "grad_norm": 4.160794734954834, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.7255, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "eval_f1_macro": 0.2963613043601986, | |
| "eval_loss": 1.0044952630996704, | |
| "eval_runtime": 13.9352, | |
| "eval_samples_per_second": 121.347, | |
| "eval_steps_per_second": 7.607, | |
| "step": 450 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8945034676158464e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |