diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,9492 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.995008319467554, + "eval_steps": 500, + "global_step": 1350, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0022185246810870773, + "grad_norm": 52.534149169921875, + "learning_rate": 3.703703703703704e-07, + "loss": 9.699, + "step": 1 + }, + { + "epoch": 0.004437049362174155, + "grad_norm": 54.63459014892578, + "learning_rate": 7.407407407407408e-07, + "loss": 9.5438, + "step": 2 + }, + { + "epoch": 0.0066555740432612314, + "grad_norm": 53.11503982543945, + "learning_rate": 1.1111111111111112e-06, + "loss": 9.5396, + "step": 3 + }, + { + "epoch": 0.00887409872434831, + "grad_norm": 51.103614807128906, + "learning_rate": 1.4814814814814817e-06, + "loss": 9.3102, + "step": 4 + }, + { + "epoch": 0.011092623405435386, + "grad_norm": 41.14444351196289, + "learning_rate": 1.8518518518518519e-06, + "loss": 9.1075, + "step": 5 + }, + { + "epoch": 0.013311148086522463, + "grad_norm": 42.219268798828125, + "learning_rate": 2.2222222222222225e-06, + "loss": 8.8976, + "step": 6 + }, + { + "epoch": 0.01552967276760954, + "grad_norm": 44.040130615234375, + "learning_rate": 2.5925925925925925e-06, + "loss": 8.5704, + "step": 7 + }, + { + "epoch": 0.01774819744869662, + "grad_norm": 53.685054779052734, + "learning_rate": 2.9629629629629633e-06, + "loss": 8.4233, + "step": 8 + }, + { + "epoch": 0.019966722129783693, + "grad_norm": Infinity, + "learning_rate": 2.9629629629629633e-06, + "loss": 7.652, + "step": 9 + }, + { + "epoch": 0.022185246810870772, + "grad_norm": 77.10890197753906, + "learning_rate": 3.3333333333333333e-06, + "loss": 8.3603, + "step": 10 + }, + { + "epoch": 0.024403771491957847, + "grad_norm": 37.35383987426758, + "learning_rate": 3.7037037037037037e-06, + "loss": 7.3239, + "step": 11 + }, + { + "epoch": 0.026622296173044926, + "grad_norm": 25.961170196533203, + "learning_rate": 4.074074074074075e-06, + "loss": 7.3322, + "step": 12 + }, + { + "epoch": 0.028840820854132, + "grad_norm": 20.448652267456055, + "learning_rate": 4.444444444444445e-06, + "loss": 7.1424, + "step": 13 + }, + { + "epoch": 0.03105934553521908, + "grad_norm": 19.287843704223633, + "learning_rate": 4.814814814814815e-06, + "loss": 7.0728, + "step": 14 + }, + { + "epoch": 0.033277870216306155, + "grad_norm": 19.270078659057617, + "learning_rate": 5.185185185185185e-06, + "loss": 7.0303, + "step": 15 + }, + { + "epoch": 0.03549639489739324, + "grad_norm": 15.79766845703125, + "learning_rate": 5.555555555555556e-06, + "loss": 7.1406, + "step": 16 + }, + { + "epoch": 0.03771491957848031, + "grad_norm": 13.498027801513672, + "learning_rate": 5.925925925925927e-06, + "loss": 6.7807, + "step": 17 + }, + { + "epoch": 0.03993344425956739, + "grad_norm": 18.08894920349121, + "learning_rate": 6.296296296296296e-06, + "loss": 6.9377, + "step": 18 + }, + { + "epoch": 0.04215196894065446, + "grad_norm": 13.833017349243164, + "learning_rate": 6.666666666666667e-06, + "loss": 6.7142, + "step": 19 + }, + { + "epoch": 0.044370493621741544, + "grad_norm": 8.692978858947754, + "learning_rate": 7.0370370370370375e-06, + "loss": 6.426, + "step": 20 + }, + { + "epoch": 0.04658901830282862, + "grad_norm": 12.649876594543457, + "learning_rate": 7.4074074074074075e-06, + "loss": 6.473, + "step": 21 + }, + { + "epoch": 0.048807542983915694, + "grad_norm": 11.425529479980469, + "learning_rate": 7.777777777777777e-06, + "loss": 6.4618, + "step": 22 + }, + { + "epoch": 0.051026067665002776, + "grad_norm": 8.402812004089355, + "learning_rate": 8.14814814814815e-06, + "loss": 6.4782, + "step": 23 + }, + { + "epoch": 0.05324459234608985, + "grad_norm": 12.67215633392334, + "learning_rate": 8.518518518518519e-06, + "loss": 6.4966, + "step": 24 + }, + { + "epoch": 0.05546311702717693, + "grad_norm": 10.478065490722656, + "learning_rate": 8.88888888888889e-06, + "loss": 6.128, + "step": 25 + }, + { + "epoch": 0.057681641708264, + "grad_norm": 11.001901626586914, + "learning_rate": 9.259259259259259e-06, + "loss": 6.2479, + "step": 26 + }, + { + "epoch": 0.059900166389351084, + "grad_norm": 12.435359954833984, + "learning_rate": 9.62962962962963e-06, + "loss": 6.3161, + "step": 27 + }, + { + "epoch": 0.06211869107043816, + "grad_norm": 8.93171501159668, + "learning_rate": 1e-05, + "loss": 6.2217, + "step": 28 + }, + { + "epoch": 0.06433721575152523, + "grad_norm": 11.941411972045898, + "learning_rate": 1.037037037037037e-05, + "loss": 6.5113, + "step": 29 + }, + { + "epoch": 0.06655574043261231, + "grad_norm": 12.103500366210938, + "learning_rate": 1.074074074074074e-05, + "loss": 6.3425, + "step": 30 + }, + { + "epoch": 0.06877426511369938, + "grad_norm": 11.753463745117188, + "learning_rate": 1.1111111111111112e-05, + "loss": 6.3389, + "step": 31 + }, + { + "epoch": 0.07099278979478647, + "grad_norm": 8.663858413696289, + "learning_rate": 1.1481481481481482e-05, + "loss": 6.12, + "step": 32 + }, + { + "epoch": 0.07321131447587355, + "grad_norm": 12.66290283203125, + "learning_rate": 1.1851851851851853e-05, + "loss": 6.1483, + "step": 33 + }, + { + "epoch": 0.07542983915696062, + "grad_norm": 12.855802536010742, + "learning_rate": 1.2222222222222222e-05, + "loss": 6.132, + "step": 34 + }, + { + "epoch": 0.0776483638380477, + "grad_norm": 13.951804161071777, + "learning_rate": 1.2592592592592592e-05, + "loss": 5.6273, + "step": 35 + }, + { + "epoch": 0.07986688851913477, + "grad_norm": 16.781387329101562, + "learning_rate": 1.2962962962962962e-05, + "loss": 6.0441, + "step": 36 + }, + { + "epoch": 0.08208541320022185, + "grad_norm": 18.647796630859375, + "learning_rate": 1.3333333333333333e-05, + "loss": 6.0212, + "step": 37 + }, + { + "epoch": 0.08430393788130892, + "grad_norm": 39.04359817504883, + "learning_rate": 1.3703703703703704e-05, + "loss": 5.7544, + "step": 38 + }, + { + "epoch": 0.08652246256239601, + "grad_norm": Infinity, + "learning_rate": 1.3703703703703704e-05, + "loss": 5.3053, + "step": 39 + }, + { + "epoch": 0.08874098724348309, + "grad_norm": 33.307411193847656, + "learning_rate": 1.4074074074074075e-05, + "loss": 5.6977, + "step": 40 + }, + { + "epoch": 0.09095951192457016, + "grad_norm": 33.078792572021484, + "learning_rate": 1.4444444444444444e-05, + "loss": 6.1941, + "step": 41 + }, + { + "epoch": 0.09317803660565724, + "grad_norm": 26.259971618652344, + "learning_rate": 1.4814814814814815e-05, + "loss": 5.4681, + "step": 42 + }, + { + "epoch": 0.09539656128674431, + "grad_norm": 40.13396453857422, + "learning_rate": 1.5185185185185186e-05, + "loss": 5.5606, + "step": 43 + }, + { + "epoch": 0.09761508596783139, + "grad_norm": 18.263961791992188, + "learning_rate": 1.5555555555555555e-05, + "loss": 5.407, + "step": 44 + }, + { + "epoch": 0.09983361064891846, + "grad_norm": 20.014732360839844, + "learning_rate": 1.5925925925925926e-05, + "loss": 5.3242, + "step": 45 + }, + { + "epoch": 0.10205213533000555, + "grad_norm": 20.575735092163086, + "learning_rate": 1.62962962962963e-05, + "loss": 5.5944, + "step": 46 + }, + { + "epoch": 0.10427066001109263, + "grad_norm": 54.92203903198242, + "learning_rate": 1.6666666666666667e-05, + "loss": 5.6131, + "step": 47 + }, + { + "epoch": 0.1064891846921797, + "grad_norm": 20.59406089782715, + "learning_rate": 1.7037037037037038e-05, + "loss": 5.1939, + "step": 48 + }, + { + "epoch": 0.10870770937326678, + "grad_norm": 45.853172302246094, + "learning_rate": 1.740740740740741e-05, + "loss": 5.7822, + "step": 49 + }, + { + "epoch": 0.11092623405435385, + "grad_norm": 21.450773239135742, + "learning_rate": 1.777777777777778e-05, + "loss": 5.2153, + "step": 50 + }, + { + "epoch": 0.11314475873544093, + "grad_norm": 48.41135787963867, + "learning_rate": 1.814814814814815e-05, + "loss": 5.4292, + "step": 51 + }, + { + "epoch": 0.115363283416528, + "grad_norm": 36.593502044677734, + "learning_rate": 1.8518518518518518e-05, + "loss": 5.5032, + "step": 52 + }, + { + "epoch": 0.11758180809761509, + "grad_norm": 25.414714813232422, + "learning_rate": 1.888888888888889e-05, + "loss": 5.4249, + "step": 53 + }, + { + "epoch": 0.11980033277870217, + "grad_norm": 21.491744995117188, + "learning_rate": 1.925925925925926e-05, + "loss": 5.5144, + "step": 54 + }, + { + "epoch": 0.12201885745978924, + "grad_norm": 18.893068313598633, + "learning_rate": 1.962962962962963e-05, + "loss": 5.2206, + "step": 55 + }, + { + "epoch": 0.12423738214087632, + "grad_norm": 16.005699157714844, + "learning_rate": 2e-05, + "loss": 5.0825, + "step": 56 + }, + { + "epoch": 0.1264559068219634, + "grad_norm": 19.90387535095215, + "learning_rate": 2.037037037037037e-05, + "loss": 5.1758, + "step": 57 + }, + { + "epoch": 0.12867443150305047, + "grad_norm": 12.69717788696289, + "learning_rate": 2.074074074074074e-05, + "loss": 5.0209, + "step": 58 + }, + { + "epoch": 0.13089295618413754, + "grad_norm": 21.71405792236328, + "learning_rate": 2.111111111111111e-05, + "loss": 5.0436, + "step": 59 + }, + { + "epoch": 0.13311148086522462, + "grad_norm": 17.426998138427734, + "learning_rate": 2.148148148148148e-05, + "loss": 4.7496, + "step": 60 + }, + { + "epoch": 0.1353300055463117, + "grad_norm": 13.18422794342041, + "learning_rate": 2.1851851851851852e-05, + "loss": 4.9545, + "step": 61 + }, + { + "epoch": 0.13754853022739877, + "grad_norm": 15.909061431884766, + "learning_rate": 2.2222222222222223e-05, + "loss": 5.0397, + "step": 62 + }, + { + "epoch": 0.13976705490848584, + "grad_norm": 13.034293174743652, + "learning_rate": 2.2592592592592594e-05, + "loss": 4.9677, + "step": 63 + }, + { + "epoch": 0.14198557958957295, + "grad_norm": 24.100385665893555, + "learning_rate": 2.2962962962962965e-05, + "loss": 4.78, + "step": 64 + }, + { + "epoch": 0.14420410427066002, + "grad_norm": 16.910551071166992, + "learning_rate": 2.3333333333333336e-05, + "loss": 4.8372, + "step": 65 + }, + { + "epoch": 0.1464226289517471, + "grad_norm": 16.66119956970215, + "learning_rate": 2.3703703703703707e-05, + "loss": 4.8177, + "step": 66 + }, + { + "epoch": 0.14864115363283417, + "grad_norm": 15.143113136291504, + "learning_rate": 2.4074074074074074e-05, + "loss": 4.8155, + "step": 67 + }, + { + "epoch": 0.15085967831392125, + "grad_norm": 14.052987098693848, + "learning_rate": 2.4444444444444445e-05, + "loss": 4.7078, + "step": 68 + }, + { + "epoch": 0.15307820299500832, + "grad_norm": 24.077512741088867, + "learning_rate": 2.4814814814814816e-05, + "loss": 4.6092, + "step": 69 + }, + { + "epoch": 0.1552967276760954, + "grad_norm": 21.495113372802734, + "learning_rate": 2.5185185185185183e-05, + "loss": 4.7682, + "step": 70 + }, + { + "epoch": 0.15751525235718247, + "grad_norm": 11.502049446105957, + "learning_rate": 2.5555555555555554e-05, + "loss": 4.6934, + "step": 71 + }, + { + "epoch": 0.15973377703826955, + "grad_norm": 17.933252334594727, + "learning_rate": 2.5925925925925925e-05, + "loss": 4.7504, + "step": 72 + }, + { + "epoch": 0.16195230171935662, + "grad_norm": 10.74679946899414, + "learning_rate": 2.6296296296296296e-05, + "loss": 4.5277, + "step": 73 + }, + { + "epoch": 0.1641708264004437, + "grad_norm": 15.563733100891113, + "learning_rate": 2.6666666666666667e-05, + "loss": 4.8872, + "step": 74 + }, + { + "epoch": 0.16638935108153077, + "grad_norm": 17.427799224853516, + "learning_rate": 2.7037037037037037e-05, + "loss": 4.4434, + "step": 75 + }, + { + "epoch": 0.16860787576261785, + "grad_norm": 14.798812866210938, + "learning_rate": 2.7407407407407408e-05, + "loss": 5.0677, + "step": 76 + }, + { + "epoch": 0.17082640044370492, + "grad_norm": 10.355371475219727, + "learning_rate": 2.777777777777778e-05, + "loss": 4.7111, + "step": 77 + }, + { + "epoch": 0.17304492512479203, + "grad_norm": 11.729372024536133, + "learning_rate": 2.814814814814815e-05, + "loss": 4.3926, + "step": 78 + }, + { + "epoch": 0.1752634498058791, + "grad_norm": 9.852009773254395, + "learning_rate": 2.851851851851852e-05, + "loss": 4.7114, + "step": 79 + }, + { + "epoch": 0.17748197448696618, + "grad_norm": 12.13970947265625, + "learning_rate": 2.8888888888888888e-05, + "loss": 4.5633, + "step": 80 + }, + { + "epoch": 0.17970049916805325, + "grad_norm": 9.55029010772705, + "learning_rate": 2.925925925925926e-05, + "loss": 4.5603, + "step": 81 + }, + { + "epoch": 0.18191902384914033, + "grad_norm": 9.343764305114746, + "learning_rate": 2.962962962962963e-05, + "loss": 4.5119, + "step": 82 + }, + { + "epoch": 0.1841375485302274, + "grad_norm": 9.35230827331543, + "learning_rate": 3e-05, + "loss": 4.437, + "step": 83 + }, + { + "epoch": 0.18635607321131448, + "grad_norm": 12.969839096069336, + "learning_rate": 3.037037037037037e-05, + "loss": 4.3512, + "step": 84 + }, + { + "epoch": 0.18857459789240155, + "grad_norm": 12.533326148986816, + "learning_rate": 3.074074074074074e-05, + "loss": 4.514, + "step": 85 + }, + { + "epoch": 0.19079312257348863, + "grad_norm": 11.187100410461426, + "learning_rate": 3.111111111111111e-05, + "loss": 4.2648, + "step": 86 + }, + { + "epoch": 0.1930116472545757, + "grad_norm": 8.709870338439941, + "learning_rate": 3.148148148148148e-05, + "loss": 4.6248, + "step": 87 + }, + { + "epoch": 0.19523017193566278, + "grad_norm": 8.719438552856445, + "learning_rate": 3.185185185185185e-05, + "loss": 4.4117, + "step": 88 + }, + { + "epoch": 0.19744869661674985, + "grad_norm": 10.545539855957031, + "learning_rate": 3.222222222222223e-05, + "loss": 4.469, + "step": 89 + }, + { + "epoch": 0.19966722129783693, + "grad_norm": 9.367774963378906, + "learning_rate": 3.25925925925926e-05, + "loss": 4.4419, + "step": 90 + }, + { + "epoch": 0.201885745978924, + "grad_norm": 8.216814041137695, + "learning_rate": 3.2962962962962964e-05, + "loss": 4.413, + "step": 91 + }, + { + "epoch": 0.2041042706600111, + "grad_norm": 10.238638877868652, + "learning_rate": 3.3333333333333335e-05, + "loss": 4.3629, + "step": 92 + }, + { + "epoch": 0.20632279534109818, + "grad_norm": 9.236513137817383, + "learning_rate": 3.3703703703703706e-05, + "loss": 4.4162, + "step": 93 + }, + { + "epoch": 0.20854132002218526, + "grad_norm": 8.21174430847168, + "learning_rate": 3.4074074074074077e-05, + "loss": 4.5541, + "step": 94 + }, + { + "epoch": 0.21075984470327233, + "grad_norm": 9.754289627075195, + "learning_rate": 3.444444444444445e-05, + "loss": 4.488, + "step": 95 + }, + { + "epoch": 0.2129783693843594, + "grad_norm": 7.130244255065918, + "learning_rate": 3.481481481481482e-05, + "loss": 4.3845, + "step": 96 + }, + { + "epoch": 0.21519689406544648, + "grad_norm": 9.086779594421387, + "learning_rate": 3.518518518518519e-05, + "loss": 4.2576, + "step": 97 + }, + { + "epoch": 0.21741541874653356, + "grad_norm": 8.378403663635254, + "learning_rate": 3.555555555555556e-05, + "loss": 4.313, + "step": 98 + }, + { + "epoch": 0.21963394342762063, + "grad_norm": 7.632753372192383, + "learning_rate": 3.592592592592593e-05, + "loss": 4.466, + "step": 99 + }, + { + "epoch": 0.2218524681087077, + "grad_norm": 7.975433349609375, + "learning_rate": 3.62962962962963e-05, + "loss": 4.1936, + "step": 100 + }, + { + "epoch": 0.22407099278979478, + "grad_norm": 8.657942771911621, + "learning_rate": 3.6666666666666666e-05, + "loss": 3.7853, + "step": 101 + }, + { + "epoch": 0.22628951747088186, + "grad_norm": 7.847908973693848, + "learning_rate": 3.7037037037037037e-05, + "loss": 4.2297, + "step": 102 + }, + { + "epoch": 0.22850804215196893, + "grad_norm": 10.147412300109863, + "learning_rate": 3.740740740740741e-05, + "loss": 4.3609, + "step": 103 + }, + { + "epoch": 0.230726566833056, + "grad_norm": 14.16489028930664, + "learning_rate": 3.777777777777778e-05, + "loss": 3.8695, + "step": 104 + }, + { + "epoch": 0.23294509151414308, + "grad_norm": 8.528613090515137, + "learning_rate": 3.814814814814815e-05, + "loss": 4.1844, + "step": 105 + }, + { + "epoch": 0.23516361619523019, + "grad_norm": 8.713404655456543, + "learning_rate": 3.851851851851852e-05, + "loss": 4.167, + "step": 106 + }, + { + "epoch": 0.23738214087631726, + "grad_norm": 8.042248725891113, + "learning_rate": 3.888888888888889e-05, + "loss": 4.1797, + "step": 107 + }, + { + "epoch": 0.23960066555740434, + "grad_norm": 8.301505088806152, + "learning_rate": 3.925925925925926e-05, + "loss": 4.2642, + "step": 108 + }, + { + "epoch": 0.2418191902384914, + "grad_norm": 7.887054443359375, + "learning_rate": 3.962962962962963e-05, + "loss": 4.3441, + "step": 109 + }, + { + "epoch": 0.24403771491957849, + "grad_norm": 13.462124824523926, + "learning_rate": 4e-05, + "loss": 4.446, + "step": 110 + }, + { + "epoch": 0.24625623960066556, + "grad_norm": 8.247685432434082, + "learning_rate": 4.0370370370370374e-05, + "loss": 4.3155, + "step": 111 + }, + { + "epoch": 0.24847476428175264, + "grad_norm": 8.609782218933105, + "learning_rate": 4.074074074074074e-05, + "loss": 4.0243, + "step": 112 + }, + { + "epoch": 0.2506932889628397, + "grad_norm": 6.894848823547363, + "learning_rate": 4.111111111111111e-05, + "loss": 3.989, + "step": 113 + }, + { + "epoch": 0.2529118136439268, + "grad_norm": 6.083759784698486, + "learning_rate": 4.148148148148148e-05, + "loss": 4.0013, + "step": 114 + }, + { + "epoch": 0.25513033832501386, + "grad_norm": 7.493844509124756, + "learning_rate": 4.185185185185185e-05, + "loss": 4.0726, + "step": 115 + }, + { + "epoch": 0.25734886300610094, + "grad_norm": 7.060963153839111, + "learning_rate": 4.222222222222222e-05, + "loss": 4.25, + "step": 116 + }, + { + "epoch": 0.259567387687188, + "grad_norm": 7.018922328948975, + "learning_rate": 4.259259259259259e-05, + "loss": 3.9978, + "step": 117 + }, + { + "epoch": 0.2617859123682751, + "grad_norm": 6.054490089416504, + "learning_rate": 4.296296296296296e-05, + "loss": 4.0494, + "step": 118 + }, + { + "epoch": 0.26400443704936216, + "grad_norm": 6.4319658279418945, + "learning_rate": 4.3333333333333334e-05, + "loss": 4.204, + "step": 119 + }, + { + "epoch": 0.26622296173044924, + "grad_norm": 5.93166971206665, + "learning_rate": 4.3703703703703705e-05, + "loss": 4.1812, + "step": 120 + }, + { + "epoch": 0.2684414864115363, + "grad_norm": 6.090641498565674, + "learning_rate": 4.4074074074074076e-05, + "loss": 4.0623, + "step": 121 + }, + { + "epoch": 0.2706600110926234, + "grad_norm": 6.934300899505615, + "learning_rate": 4.4444444444444447e-05, + "loss": 4.2747, + "step": 122 + }, + { + "epoch": 0.27287853577371046, + "grad_norm": 6.030778408050537, + "learning_rate": 4.481481481481482e-05, + "loss": 3.8953, + "step": 123 + }, + { + "epoch": 0.27509706045479754, + "grad_norm": 6.201318740844727, + "learning_rate": 4.518518518518519e-05, + "loss": 4.1609, + "step": 124 + }, + { + "epoch": 0.2773155851358846, + "grad_norm": 7.340195178985596, + "learning_rate": 4.555555555555556e-05, + "loss": 4.2279, + "step": 125 + }, + { + "epoch": 0.2795341098169717, + "grad_norm": 7.376201629638672, + "learning_rate": 4.592592592592593e-05, + "loss": 4.1653, + "step": 126 + }, + { + "epoch": 0.2817526344980588, + "grad_norm": 5.852590560913086, + "learning_rate": 4.62962962962963e-05, + "loss": 4.1048, + "step": 127 + }, + { + "epoch": 0.2839711591791459, + "grad_norm": 9.323796272277832, + "learning_rate": 4.666666666666667e-05, + "loss": 4.2577, + "step": 128 + }, + { + "epoch": 0.28618968386023297, + "grad_norm": 5.714437961578369, + "learning_rate": 4.703703703703704e-05, + "loss": 4.2329, + "step": 129 + }, + { + "epoch": 0.28840820854132004, + "grad_norm": 6.402439117431641, + "learning_rate": 4.740740740740741e-05, + "loss": 4.159, + "step": 130 + }, + { + "epoch": 0.2906267332224071, + "grad_norm": 6.096062183380127, + "learning_rate": 4.7777777777777784e-05, + "loss": 4.2683, + "step": 131 + }, + { + "epoch": 0.2928452579034942, + "grad_norm": 6.514817237854004, + "learning_rate": 4.814814814814815e-05, + "loss": 3.8523, + "step": 132 + }, + { + "epoch": 0.29506378258458127, + "grad_norm": 5.105497360229492, + "learning_rate": 4.851851851851852e-05, + "loss": 4.148, + "step": 133 + }, + { + "epoch": 0.29728230726566834, + "grad_norm": 6.290508270263672, + "learning_rate": 4.888888888888889e-05, + "loss": 4.1209, + "step": 134 + }, + { + "epoch": 0.2995008319467554, + "grad_norm": 7.905398845672607, + "learning_rate": 4.925925925925926e-05, + "loss": 3.8754, + "step": 135 + }, + { + "epoch": 0.3017193566278425, + "grad_norm": 5.988460063934326, + "learning_rate": 4.962962962962963e-05, + "loss": 3.81, + "step": 136 + }, + { + "epoch": 0.30393788130892957, + "grad_norm": 6.45888090133667, + "learning_rate": 5e-05, + "loss": 4.3125, + "step": 137 + }, + { + "epoch": 0.30615640599001664, + "grad_norm": 5.811912536621094, + "learning_rate": 4.995884773662552e-05, + "loss": 4.1226, + "step": 138 + }, + { + "epoch": 0.3083749306711037, + "grad_norm": 8.079099655151367, + "learning_rate": 4.991769547325103e-05, + "loss": 4.0868, + "step": 139 + }, + { + "epoch": 0.3105934553521908, + "grad_norm": 6.5362548828125, + "learning_rate": 4.987654320987655e-05, + "loss": 4.1231, + "step": 140 + }, + { + "epoch": 0.31281198003327787, + "grad_norm": 5.130688190460205, + "learning_rate": 4.983539094650206e-05, + "loss": 3.842, + "step": 141 + }, + { + "epoch": 0.31503050471436495, + "grad_norm": 6.628684043884277, + "learning_rate": 4.9794238683127575e-05, + "loss": 4.2446, + "step": 142 + }, + { + "epoch": 0.317249029395452, + "grad_norm": 7.863401412963867, + "learning_rate": 4.9753086419753084e-05, + "loss": 4.138, + "step": 143 + }, + { + "epoch": 0.3194675540765391, + "grad_norm": 5.999283313751221, + "learning_rate": 4.971193415637861e-05, + "loss": 4.2141, + "step": 144 + }, + { + "epoch": 0.32168607875762617, + "grad_norm": 5.943251132965088, + "learning_rate": 4.967078189300412e-05, + "loss": 4.1941, + "step": 145 + }, + { + "epoch": 0.32390460343871325, + "grad_norm": 5.245730876922607, + "learning_rate": 4.962962962962963e-05, + "loss": 4.0505, + "step": 146 + }, + { + "epoch": 0.3261231281198003, + "grad_norm": 5.412414073944092, + "learning_rate": 4.958847736625515e-05, + "loss": 3.9237, + "step": 147 + }, + { + "epoch": 0.3283416528008874, + "grad_norm": 5.193697452545166, + "learning_rate": 4.9547325102880656e-05, + "loss": 4.1242, + "step": 148 + }, + { + "epoch": 0.33056017748197447, + "grad_norm": 6.634949684143066, + "learning_rate": 4.950617283950618e-05, + "loss": 4.106, + "step": 149 + }, + { + "epoch": 0.33277870216306155, + "grad_norm": 6.305480003356934, + "learning_rate": 4.946502057613169e-05, + "loss": 4.1864, + "step": 150 + }, + { + "epoch": 0.3349972268441486, + "grad_norm": 6.962391376495361, + "learning_rate": 4.9423868312757204e-05, + "loss": 4.2511, + "step": 151 + }, + { + "epoch": 0.3372157515252357, + "grad_norm": 5.507792949676514, + "learning_rate": 4.938271604938271e-05, + "loss": 4.0817, + "step": 152 + }, + { + "epoch": 0.33943427620632277, + "grad_norm": 4.571567058563232, + "learning_rate": 4.9341563786008236e-05, + "loss": 4.0044, + "step": 153 + }, + { + "epoch": 0.34165280088740985, + "grad_norm": 4.95249605178833, + "learning_rate": 4.930041152263375e-05, + "loss": 4.0257, + "step": 154 + }, + { + "epoch": 0.343871325568497, + "grad_norm": 5.651655197143555, + "learning_rate": 4.925925925925926e-05, + "loss": 4.3137, + "step": 155 + }, + { + "epoch": 0.34608985024958405, + "grad_norm": 4.292090892791748, + "learning_rate": 4.9218106995884777e-05, + "loss": 4.0779, + "step": 156 + }, + { + "epoch": 0.3483083749306711, + "grad_norm": 6.156320095062256, + "learning_rate": 4.9176954732510286e-05, + "loss": 4.2033, + "step": 157 + }, + { + "epoch": 0.3505268996117582, + "grad_norm": 4.048032283782959, + "learning_rate": 4.913580246913581e-05, + "loss": 4.1121, + "step": 158 + }, + { + "epoch": 0.3527454242928453, + "grad_norm": 4.719449043273926, + "learning_rate": 4.909465020576132e-05, + "loss": 3.6145, + "step": 159 + }, + { + "epoch": 0.35496394897393235, + "grad_norm": 4.914869785308838, + "learning_rate": 4.905349794238683e-05, + "loss": 3.9358, + "step": 160 + }, + { + "epoch": 0.35718247365501943, + "grad_norm": 5.61899995803833, + "learning_rate": 4.901234567901235e-05, + "loss": 4.412, + "step": 161 + }, + { + "epoch": 0.3594009983361065, + "grad_norm": 5.874763488769531, + "learning_rate": 4.8971193415637865e-05, + "loss": 4.1115, + "step": 162 + }, + { + "epoch": 0.3616195230171936, + "grad_norm": 6.324469566345215, + "learning_rate": 4.893004115226338e-05, + "loss": 4.1024, + "step": 163 + }, + { + "epoch": 0.36383804769828065, + "grad_norm": 4.6851725578308105, + "learning_rate": 4.888888888888889e-05, + "loss": 4.0718, + "step": 164 + }, + { + "epoch": 0.36605657237936773, + "grad_norm": 5.07680606842041, + "learning_rate": 4.8847736625514406e-05, + "loss": 4.0501, + "step": 165 + }, + { + "epoch": 0.3682750970604548, + "grad_norm": 5.340401649475098, + "learning_rate": 4.8806584362139915e-05, + "loss": 4.1007, + "step": 166 + }, + { + "epoch": 0.3704936217415419, + "grad_norm": 5.174929618835449, + "learning_rate": 4.876543209876544e-05, + "loss": 3.9105, + "step": 167 + }, + { + "epoch": 0.37271214642262895, + "grad_norm": 5.105217933654785, + "learning_rate": 4.872427983539095e-05, + "loss": 3.8597, + "step": 168 + }, + { + "epoch": 0.37493067110371603, + "grad_norm": 7.893522262573242, + "learning_rate": 4.868312757201646e-05, + "loss": 3.7879, + "step": 169 + }, + { + "epoch": 0.3771491957848031, + "grad_norm": 6.1168012619018555, + "learning_rate": 4.864197530864198e-05, + "loss": 4.0409, + "step": 170 + }, + { + "epoch": 0.3793677204658902, + "grad_norm": 5.16205358505249, + "learning_rate": 4.860082304526749e-05, + "loss": 4.0682, + "step": 171 + }, + { + "epoch": 0.38158624514697725, + "grad_norm": 5.319783687591553, + "learning_rate": 4.855967078189301e-05, + "loss": 3.6806, + "step": 172 + }, + { + "epoch": 0.38380476982806433, + "grad_norm": 6.260192394256592, + "learning_rate": 4.851851851851852e-05, + "loss": 3.6691, + "step": 173 + }, + { + "epoch": 0.3860232945091514, + "grad_norm": 5.190227508544922, + "learning_rate": 4.8477366255144035e-05, + "loss": 3.8209, + "step": 174 + }, + { + "epoch": 0.3882418191902385, + "grad_norm": 5.380640983581543, + "learning_rate": 4.843621399176955e-05, + "loss": 3.9955, + "step": 175 + }, + { + "epoch": 0.39046034387132555, + "grad_norm": 5.388073444366455, + "learning_rate": 4.8395061728395067e-05, + "loss": 4.0229, + "step": 176 + }, + { + "epoch": 0.39267886855241263, + "grad_norm": 5.171314239501953, + "learning_rate": 4.835390946502058e-05, + "loss": 3.7463, + "step": 177 + }, + { + "epoch": 0.3948973932334997, + "grad_norm": 5.06121301651001, + "learning_rate": 4.831275720164609e-05, + "loss": 3.8546, + "step": 178 + }, + { + "epoch": 0.3971159179145868, + "grad_norm": 5.399997711181641, + "learning_rate": 4.827160493827161e-05, + "loss": 4.2523, + "step": 179 + }, + { + "epoch": 0.39933444259567386, + "grad_norm": 4.833516597747803, + "learning_rate": 4.8230452674897116e-05, + "loss": 3.9168, + "step": 180 + }, + { + "epoch": 0.40155296727676093, + "grad_norm": 5.467654705047607, + "learning_rate": 4.818930041152264e-05, + "loss": 3.8456, + "step": 181 + }, + { + "epoch": 0.403771491957848, + "grad_norm": 5.002521514892578, + "learning_rate": 4.814814814814815e-05, + "loss": 3.8746, + "step": 182 + }, + { + "epoch": 0.4059900166389351, + "grad_norm": 4.590516567230225, + "learning_rate": 4.8106995884773664e-05, + "loss": 3.7248, + "step": 183 + }, + { + "epoch": 0.4082085413200222, + "grad_norm": 5.944400787353516, + "learning_rate": 4.806584362139918e-05, + "loss": 4.0865, + "step": 184 + }, + { + "epoch": 0.4104270660011093, + "grad_norm": 5.781230449676514, + "learning_rate": 4.8024691358024696e-05, + "loss": 3.9351, + "step": 185 + }, + { + "epoch": 0.41264559068219636, + "grad_norm": 5.954226493835449, + "learning_rate": 4.798353909465021e-05, + "loss": 3.8865, + "step": 186 + }, + { + "epoch": 0.41486411536328344, + "grad_norm": 5.727841854095459, + "learning_rate": 4.794238683127572e-05, + "loss": 3.6115, + "step": 187 + }, + { + "epoch": 0.4170826400443705, + "grad_norm": 5.0357890129089355, + "learning_rate": 4.7901234567901237e-05, + "loss": 3.9431, + "step": 188 + }, + { + "epoch": 0.4193011647254576, + "grad_norm": 4.922348499298096, + "learning_rate": 4.7860082304526746e-05, + "loss": 3.9023, + "step": 189 + }, + { + "epoch": 0.42151968940654466, + "grad_norm": 4.735426425933838, + "learning_rate": 4.781893004115227e-05, + "loss": 3.7753, + "step": 190 + }, + { + "epoch": 0.42373821408763174, + "grad_norm": 5.203444004058838, + "learning_rate": 4.7777777777777784e-05, + "loss": 3.9787, + "step": 191 + }, + { + "epoch": 0.4259567387687188, + "grad_norm": 4.4607367515563965, + "learning_rate": 4.773662551440329e-05, + "loss": 3.7737, + "step": 192 + }, + { + "epoch": 0.4281752634498059, + "grad_norm": 3.815371036529541, + "learning_rate": 4.769547325102881e-05, + "loss": 3.7641, + "step": 193 + }, + { + "epoch": 0.43039378813089296, + "grad_norm": 5.5074357986450195, + "learning_rate": 4.7654320987654325e-05, + "loss": 3.8473, + "step": 194 + }, + { + "epoch": 0.43261231281198004, + "grad_norm": 4.774341106414795, + "learning_rate": 4.761316872427984e-05, + "loss": 3.8476, + "step": 195 + }, + { + "epoch": 0.4348308374930671, + "grad_norm": 4.429344654083252, + "learning_rate": 4.757201646090535e-05, + "loss": 3.843, + "step": 196 + }, + { + "epoch": 0.4370493621741542, + "grad_norm": 4.823075294494629, + "learning_rate": 4.7530864197530866e-05, + "loss": 3.8921, + "step": 197 + }, + { + "epoch": 0.43926788685524126, + "grad_norm": 4.532520294189453, + "learning_rate": 4.748971193415638e-05, + "loss": 4.0264, + "step": 198 + }, + { + "epoch": 0.44148641153632834, + "grad_norm": 4.486454963684082, + "learning_rate": 4.74485596707819e-05, + "loss": 3.7397, + "step": 199 + }, + { + "epoch": 0.4437049362174154, + "grad_norm": 4.371092319488525, + "learning_rate": 4.740740740740741e-05, + "loss": 3.6754, + "step": 200 + }, + { + "epoch": 0.4459234608985025, + "grad_norm": 3.9995057582855225, + "learning_rate": 4.736625514403292e-05, + "loss": 3.5851, + "step": 201 + }, + { + "epoch": 0.44814198557958956, + "grad_norm": 4.318523406982422, + "learning_rate": 4.732510288065844e-05, + "loss": 3.6055, + "step": 202 + }, + { + "epoch": 0.45036051026067664, + "grad_norm": 5.345367908477783, + "learning_rate": 4.7283950617283954e-05, + "loss": 3.8775, + "step": 203 + }, + { + "epoch": 0.4525790349417637, + "grad_norm": 5.964838981628418, + "learning_rate": 4.724279835390947e-05, + "loss": 4.1468, + "step": 204 + }, + { + "epoch": 0.4547975596228508, + "grad_norm": 5.4279632568359375, + "learning_rate": 4.7201646090534986e-05, + "loss": 3.9391, + "step": 205 + }, + { + "epoch": 0.45701608430393786, + "grad_norm": 4.6981635093688965, + "learning_rate": 4.7160493827160495e-05, + "loss": 3.5132, + "step": 206 + }, + { + "epoch": 0.45923460898502494, + "grad_norm": 4.1578497886657715, + "learning_rate": 4.711934156378601e-05, + "loss": 3.7166, + "step": 207 + }, + { + "epoch": 0.461453133666112, + "grad_norm": 4.3208441734313965, + "learning_rate": 4.7078189300411527e-05, + "loss": 3.4349, + "step": 208 + }, + { + "epoch": 0.4636716583471991, + "grad_norm": 4.67413330078125, + "learning_rate": 4.703703703703704e-05, + "loss": 3.717, + "step": 209 + }, + { + "epoch": 0.46589018302828616, + "grad_norm": 4.605541706085205, + "learning_rate": 4.699588477366255e-05, + "loss": 3.537, + "step": 210 + }, + { + "epoch": 0.46810870770937324, + "grad_norm": 4.660853385925293, + "learning_rate": 4.695473251028807e-05, + "loss": 3.6684, + "step": 211 + }, + { + "epoch": 0.47032723239046037, + "grad_norm": 4.483960151672363, + "learning_rate": 4.691358024691358e-05, + "loss": 3.3903, + "step": 212 + }, + { + "epoch": 0.47254575707154745, + "grad_norm": 5.170067310333252, + "learning_rate": 4.68724279835391e-05, + "loss": 3.8223, + "step": 213 + }, + { + "epoch": 0.4747642817526345, + "grad_norm": 4.827269077301025, + "learning_rate": 4.6831275720164615e-05, + "loss": 3.5872, + "step": 214 + }, + { + "epoch": 0.4769828064337216, + "grad_norm": 4.822846412658691, + "learning_rate": 4.6790123456790124e-05, + "loss": 3.7919, + "step": 215 + }, + { + "epoch": 0.47920133111480867, + "grad_norm": 4.748480319976807, + "learning_rate": 4.674897119341564e-05, + "loss": 3.6194, + "step": 216 + }, + { + "epoch": 0.48141985579589575, + "grad_norm": 4.786713600158691, + "learning_rate": 4.6707818930041156e-05, + "loss": 3.6185, + "step": 217 + }, + { + "epoch": 0.4836383804769828, + "grad_norm": 4.167690753936768, + "learning_rate": 4.666666666666667e-05, + "loss": 3.4432, + "step": 218 + }, + { + "epoch": 0.4858569051580699, + "grad_norm": 4.987302303314209, + "learning_rate": 4.662551440329218e-05, + "loss": 3.3928, + "step": 219 + }, + { + "epoch": 0.48807542983915697, + "grad_norm": 4.708268642425537, + "learning_rate": 4.6584362139917697e-05, + "loss": 3.6866, + "step": 220 + }, + { + "epoch": 0.49029395452024405, + "grad_norm": 4.866717338562012, + "learning_rate": 4.654320987654321e-05, + "loss": 3.8574, + "step": 221 + }, + { + "epoch": 0.4925124792013311, + "grad_norm": 4.52734899520874, + "learning_rate": 4.650205761316873e-05, + "loss": 3.7122, + "step": 222 + }, + { + "epoch": 0.4947310038824182, + "grad_norm": 4.690919399261475, + "learning_rate": 4.6460905349794244e-05, + "loss": 3.843, + "step": 223 + }, + { + "epoch": 0.49694952856350527, + "grad_norm": 4.603267192840576, + "learning_rate": 4.641975308641975e-05, + "loss": 3.5478, + "step": 224 + }, + { + "epoch": 0.49916805324459235, + "grad_norm": 4.186052322387695, + "learning_rate": 4.637860082304527e-05, + "loss": 3.507, + "step": 225 + }, + { + "epoch": 0.5013865779256794, + "grad_norm": 4.9482035636901855, + "learning_rate": 4.6337448559670785e-05, + "loss": 3.7335, + "step": 226 + }, + { + "epoch": 0.5036051026067665, + "grad_norm": 4.611546993255615, + "learning_rate": 4.62962962962963e-05, + "loss": 3.8908, + "step": 227 + }, + { + "epoch": 0.5058236272878536, + "grad_norm": 4.673487186431885, + "learning_rate": 4.625514403292182e-05, + "loss": 3.6157, + "step": 228 + }, + { + "epoch": 0.5080421519689406, + "grad_norm": 4.354834079742432, + "learning_rate": 4.6213991769547326e-05, + "loss": 3.7627, + "step": 229 + }, + { + "epoch": 0.5102606766500277, + "grad_norm": 3.939734697341919, + "learning_rate": 4.617283950617284e-05, + "loss": 3.6497, + "step": 230 + }, + { + "epoch": 0.5124792013311148, + "grad_norm": 4.662121772766113, + "learning_rate": 4.613168724279836e-05, + "loss": 3.5891, + "step": 231 + }, + { + "epoch": 0.5146977260122019, + "grad_norm": 4.295231819152832, + "learning_rate": 4.609053497942387e-05, + "loss": 3.5441, + "step": 232 + }, + { + "epoch": 0.516916250693289, + "grad_norm": 4.8317060470581055, + "learning_rate": 4.604938271604938e-05, + "loss": 3.4834, + "step": 233 + }, + { + "epoch": 0.519134775374376, + "grad_norm": 4.406210422515869, + "learning_rate": 4.60082304526749e-05, + "loss": 3.5203, + "step": 234 + }, + { + "epoch": 0.5213533000554631, + "grad_norm": 5.440707206726074, + "learning_rate": 4.5967078189300414e-05, + "loss": 3.3898, + "step": 235 + }, + { + "epoch": 0.5235718247365502, + "grad_norm": 4.552436828613281, + "learning_rate": 4.592592592592593e-05, + "loss": 3.6478, + "step": 236 + }, + { + "epoch": 0.5257903494176372, + "grad_norm": 4.949105262756348, + "learning_rate": 4.5884773662551446e-05, + "loss": 3.6882, + "step": 237 + }, + { + "epoch": 0.5280088740987243, + "grad_norm": 4.715395450592041, + "learning_rate": 4.5843621399176955e-05, + "loss": 3.2585, + "step": 238 + }, + { + "epoch": 0.5302273987798114, + "grad_norm": 4.920379161834717, + "learning_rate": 4.580246913580247e-05, + "loss": 3.236, + "step": 239 + }, + { + "epoch": 0.5324459234608985, + "grad_norm": 5.411755561828613, + "learning_rate": 4.5761316872427987e-05, + "loss": 3.6088, + "step": 240 + }, + { + "epoch": 0.5346644481419855, + "grad_norm": 4.764686107635498, + "learning_rate": 4.57201646090535e-05, + "loss": 3.2244, + "step": 241 + }, + { + "epoch": 0.5368829728230726, + "grad_norm": 4.309259414672852, + "learning_rate": 4.567901234567901e-05, + "loss": 3.3411, + "step": 242 + }, + { + "epoch": 0.5391014975041597, + "grad_norm": 4.65073823928833, + "learning_rate": 4.563786008230453e-05, + "loss": 3.3799, + "step": 243 + }, + { + "epoch": 0.5413200221852468, + "grad_norm": 5.38470983505249, + "learning_rate": 4.559670781893004e-05, + "loss": 3.544, + "step": 244 + }, + { + "epoch": 0.5435385468663338, + "grad_norm": 4.650628089904785, + "learning_rate": 4.555555555555556e-05, + "loss": 3.3204, + "step": 245 + }, + { + "epoch": 0.5457570715474209, + "grad_norm": 4.457350730895996, + "learning_rate": 4.5514403292181075e-05, + "loss": 3.3921, + "step": 246 + }, + { + "epoch": 0.547975596228508, + "grad_norm": 4.193648338317871, + "learning_rate": 4.5473251028806584e-05, + "loss": 3.5171, + "step": 247 + }, + { + "epoch": 0.5501941209095951, + "grad_norm": 4.395062446594238, + "learning_rate": 4.54320987654321e-05, + "loss": 3.3137, + "step": 248 + }, + { + "epoch": 0.5524126455906821, + "grad_norm": 4.102454662322998, + "learning_rate": 4.5390946502057616e-05, + "loss": 3.298, + "step": 249 + }, + { + "epoch": 0.5546311702717692, + "grad_norm": 4.101161956787109, + "learning_rate": 4.534979423868313e-05, + "loss": 3.3402, + "step": 250 + }, + { + "epoch": 0.5568496949528563, + "grad_norm": 4.484131813049316, + "learning_rate": 4.530864197530865e-05, + "loss": 3.3289, + "step": 251 + }, + { + "epoch": 0.5590682196339434, + "grad_norm": 4.731671333312988, + "learning_rate": 4.5267489711934157e-05, + "loss": 3.4629, + "step": 252 + }, + { + "epoch": 0.5612867443150305, + "grad_norm": 4.493623733520508, + "learning_rate": 4.522633744855967e-05, + "loss": 3.9369, + "step": 253 + }, + { + "epoch": 0.5635052689961176, + "grad_norm": 4.47726583480835, + "learning_rate": 4.518518518518519e-05, + "loss": 3.4003, + "step": 254 + }, + { + "epoch": 0.5657237936772047, + "grad_norm": 4.06917142868042, + "learning_rate": 4.5144032921810704e-05, + "loss": 3.3686, + "step": 255 + }, + { + "epoch": 0.5679423183582918, + "grad_norm": 4.803555011749268, + "learning_rate": 4.510288065843621e-05, + "loss": 3.3428, + "step": 256 + }, + { + "epoch": 0.5701608430393789, + "grad_norm": 4.667956829071045, + "learning_rate": 4.506172839506173e-05, + "loss": 3.3041, + "step": 257 + }, + { + "epoch": 0.5723793677204659, + "grad_norm": 4.465986251831055, + "learning_rate": 4.5020576131687245e-05, + "loss": 3.4169, + "step": 258 + }, + { + "epoch": 0.574597892401553, + "grad_norm": 4.457067012786865, + "learning_rate": 4.497942386831276e-05, + "loss": 3.5762, + "step": 259 + }, + { + "epoch": 0.5768164170826401, + "grad_norm": 4.306710243225098, + "learning_rate": 4.493827160493828e-05, + "loss": 3.6046, + "step": 260 + }, + { + "epoch": 0.5790349417637272, + "grad_norm": 4.341065406799316, + "learning_rate": 4.4897119341563786e-05, + "loss": 3.6239, + "step": 261 + }, + { + "epoch": 0.5812534664448142, + "grad_norm": 4.795947074890137, + "learning_rate": 4.48559670781893e-05, + "loss": 2.9047, + "step": 262 + }, + { + "epoch": 0.5834719911259013, + "grad_norm": 5.131918430328369, + "learning_rate": 4.481481481481482e-05, + "loss": 3.3283, + "step": 263 + }, + { + "epoch": 0.5856905158069884, + "grad_norm": 4.925499439239502, + "learning_rate": 4.477366255144033e-05, + "loss": 3.3819, + "step": 264 + }, + { + "epoch": 0.5879090404880755, + "grad_norm": 4.786744594573975, + "learning_rate": 4.473251028806584e-05, + "loss": 3.6277, + "step": 265 + }, + { + "epoch": 0.5901275651691625, + "grad_norm": 4.5993852615356445, + "learning_rate": 4.469135802469136e-05, + "loss": 3.472, + "step": 266 + }, + { + "epoch": 0.5923460898502496, + "grad_norm": 4.40986442565918, + "learning_rate": 4.4650205761316874e-05, + "loss": 3.4587, + "step": 267 + }, + { + "epoch": 0.5945646145313367, + "grad_norm": 4.476975917816162, + "learning_rate": 4.460905349794239e-05, + "loss": 3.4616, + "step": 268 + }, + { + "epoch": 0.5967831392124238, + "grad_norm": 3.7517433166503906, + "learning_rate": 4.4567901234567906e-05, + "loss": 3.7072, + "step": 269 + }, + { + "epoch": 0.5990016638935108, + "grad_norm": 3.9777588844299316, + "learning_rate": 4.4526748971193415e-05, + "loss": 3.3538, + "step": 270 + }, + { + "epoch": 0.6012201885745979, + "grad_norm": 5.085666179656982, + "learning_rate": 4.448559670781893e-05, + "loss": 3.3076, + "step": 271 + }, + { + "epoch": 0.603438713255685, + "grad_norm": 4.634458065032959, + "learning_rate": 4.4444444444444447e-05, + "loss": 3.0557, + "step": 272 + }, + { + "epoch": 0.6056572379367721, + "grad_norm": 4.459312915802002, + "learning_rate": 4.440329218106996e-05, + "loss": 3.3428, + "step": 273 + }, + { + "epoch": 0.6078757626178591, + "grad_norm": 4.53181266784668, + "learning_rate": 4.436213991769548e-05, + "loss": 3.33, + "step": 274 + }, + { + "epoch": 0.6100942872989462, + "grad_norm": 5.866949081420898, + "learning_rate": 4.432098765432099e-05, + "loss": 3.4035, + "step": 275 + }, + { + "epoch": 0.6123128119800333, + "grad_norm": 5.076850414276123, + "learning_rate": 4.42798353909465e-05, + "loss": 3.5642, + "step": 276 + }, + { + "epoch": 0.6145313366611204, + "grad_norm": 4.5255255699157715, + "learning_rate": 4.423868312757202e-05, + "loss": 3.5244, + "step": 277 + }, + { + "epoch": 0.6167498613422074, + "grad_norm": 4.63967752456665, + "learning_rate": 4.4197530864197535e-05, + "loss": 3.0194, + "step": 278 + }, + { + "epoch": 0.6189683860232945, + "grad_norm": 4.861447811126709, + "learning_rate": 4.4156378600823044e-05, + "loss": 3.5077, + "step": 279 + }, + { + "epoch": 0.6211869107043816, + "grad_norm": 3.981630325317383, + "learning_rate": 4.411522633744856e-05, + "loss": 3.3063, + "step": 280 + }, + { + "epoch": 0.6234054353854687, + "grad_norm": 3.9408223628997803, + "learning_rate": 4.4074074074074076e-05, + "loss": 3.2028, + "step": 281 + }, + { + "epoch": 0.6256239600665557, + "grad_norm": 4.115415096282959, + "learning_rate": 4.403292181069959e-05, + "loss": 2.8013, + "step": 282 + }, + { + "epoch": 0.6278424847476428, + "grad_norm": 4.497473239898682, + "learning_rate": 4.399176954732511e-05, + "loss": 3.5278, + "step": 283 + }, + { + "epoch": 0.6300610094287299, + "grad_norm": 4.794856071472168, + "learning_rate": 4.3950617283950617e-05, + "loss": 3.4567, + "step": 284 + }, + { + "epoch": 0.632279534109817, + "grad_norm": 4.686934471130371, + "learning_rate": 4.390946502057613e-05, + "loss": 3.3246, + "step": 285 + }, + { + "epoch": 0.634498058790904, + "grad_norm": 4.541492462158203, + "learning_rate": 4.386831275720165e-05, + "loss": 3.2917, + "step": 286 + }, + { + "epoch": 0.6367165834719911, + "grad_norm": 5.043437957763672, + "learning_rate": 4.3827160493827164e-05, + "loss": 3.3356, + "step": 287 + }, + { + "epoch": 0.6389351081530782, + "grad_norm": 5.479452610015869, + "learning_rate": 4.378600823045268e-05, + "loss": 2.9194, + "step": 288 + }, + { + "epoch": 0.6411536328341653, + "grad_norm": 4.429378509521484, + "learning_rate": 4.374485596707819e-05, + "loss": 3.3288, + "step": 289 + }, + { + "epoch": 0.6433721575152523, + "grad_norm": 7.362743854522705, + "learning_rate": 4.3703703703703705e-05, + "loss": 3.191, + "step": 290 + }, + { + "epoch": 0.6455906821963394, + "grad_norm": 4.778252124786377, + "learning_rate": 4.366255144032922e-05, + "loss": 3.4379, + "step": 291 + }, + { + "epoch": 0.6478092068774265, + "grad_norm": 3.964599132537842, + "learning_rate": 4.3621399176954737e-05, + "loss": 3.3816, + "step": 292 + }, + { + "epoch": 0.6500277315585136, + "grad_norm": 4.656654357910156, + "learning_rate": 4.3580246913580246e-05, + "loss": 3.6775, + "step": 293 + }, + { + "epoch": 0.6522462562396006, + "grad_norm": 3.938077449798584, + "learning_rate": 4.353909465020576e-05, + "loss": 3.271, + "step": 294 + }, + { + "epoch": 0.6544647809206877, + "grad_norm": 4.255359649658203, + "learning_rate": 4.349794238683128e-05, + "loss": 3.1789, + "step": 295 + }, + { + "epoch": 0.6566833056017748, + "grad_norm": 4.316828727722168, + "learning_rate": 4.345679012345679e-05, + "loss": 3.2222, + "step": 296 + }, + { + "epoch": 0.6589018302828619, + "grad_norm": 4.710544109344482, + "learning_rate": 4.341563786008231e-05, + "loss": 3.1336, + "step": 297 + }, + { + "epoch": 0.6611203549639489, + "grad_norm": 4.449007987976074, + "learning_rate": 4.337448559670782e-05, + "loss": 3.1638, + "step": 298 + }, + { + "epoch": 0.663338879645036, + "grad_norm": 4.722839832305908, + "learning_rate": 4.3333333333333334e-05, + "loss": 3.5576, + "step": 299 + }, + { + "epoch": 0.6655574043261231, + "grad_norm": 3.9031119346618652, + "learning_rate": 4.329218106995885e-05, + "loss": 3.3216, + "step": 300 + }, + { + "epoch": 0.6677759290072102, + "grad_norm": 3.7224013805389404, + "learning_rate": 4.3251028806584366e-05, + "loss": 2.9033, + "step": 301 + }, + { + "epoch": 0.6699944536882972, + "grad_norm": 3.85421085357666, + "learning_rate": 4.3209876543209875e-05, + "loss": 3.3348, + "step": 302 + }, + { + "epoch": 0.6722129783693843, + "grad_norm": 4.420081615447998, + "learning_rate": 4.316872427983539e-05, + "loss": 3.2536, + "step": 303 + }, + { + "epoch": 0.6744315030504714, + "grad_norm": 4.600344657897949, + "learning_rate": 4.3127572016460907e-05, + "loss": 3.2926, + "step": 304 + }, + { + "epoch": 0.6766500277315585, + "grad_norm": 3.715827703475952, + "learning_rate": 4.308641975308642e-05, + "loss": 3.2145, + "step": 305 + }, + { + "epoch": 0.6788685524126455, + "grad_norm": 4.9351630210876465, + "learning_rate": 4.304526748971194e-05, + "loss": 3.24, + "step": 306 + }, + { + "epoch": 0.6810870770937326, + "grad_norm": 4.325677871704102, + "learning_rate": 4.300411522633745e-05, + "loss": 3.3456, + "step": 307 + }, + { + "epoch": 0.6833056017748197, + "grad_norm": 4.247461795806885, + "learning_rate": 4.296296296296296e-05, + "loss": 3.1161, + "step": 308 + }, + { + "epoch": 0.6855241264559068, + "grad_norm": 3.8467044830322266, + "learning_rate": 4.292181069958848e-05, + "loss": 2.9523, + "step": 309 + }, + { + "epoch": 0.687742651136994, + "grad_norm": 4.189371109008789, + "learning_rate": 4.2880658436213995e-05, + "loss": 2.994, + "step": 310 + }, + { + "epoch": 0.689961175818081, + "grad_norm": 4.541285037994385, + "learning_rate": 4.283950617283951e-05, + "loss": 3.2566, + "step": 311 + }, + { + "epoch": 0.6921797004991681, + "grad_norm": 4.245055198669434, + "learning_rate": 4.279835390946502e-05, + "loss": 3.3166, + "step": 312 + }, + { + "epoch": 0.6943982251802552, + "grad_norm": 3.917374610900879, + "learning_rate": 4.2757201646090536e-05, + "loss": 3.1895, + "step": 313 + }, + { + "epoch": 0.6966167498613423, + "grad_norm": 4.36654806137085, + "learning_rate": 4.271604938271605e-05, + "loss": 3.0782, + "step": 314 + }, + { + "epoch": 0.6988352745424293, + "grad_norm": 3.973999500274658, + "learning_rate": 4.267489711934157e-05, + "loss": 3.1851, + "step": 315 + }, + { + "epoch": 0.7010537992235164, + "grad_norm": 4.2357563972473145, + "learning_rate": 4.2633744855967077e-05, + "loss": 3.0444, + "step": 316 + }, + { + "epoch": 0.7032723239046035, + "grad_norm": 3.9498109817504883, + "learning_rate": 4.259259259259259e-05, + "loss": 3.1961, + "step": 317 + }, + { + "epoch": 0.7054908485856906, + "grad_norm": 4.642027854919434, + "learning_rate": 4.255144032921811e-05, + "loss": 3.0797, + "step": 318 + }, + { + "epoch": 0.7077093732667776, + "grad_norm": 4.767258644104004, + "learning_rate": 4.2510288065843624e-05, + "loss": 3.2066, + "step": 319 + }, + { + "epoch": 0.7099278979478647, + "grad_norm": 4.588222980499268, + "learning_rate": 4.246913580246914e-05, + "loss": 3.2093, + "step": 320 + }, + { + "epoch": 0.7121464226289518, + "grad_norm": 4.097878932952881, + "learning_rate": 4.242798353909465e-05, + "loss": 3.0652, + "step": 321 + }, + { + "epoch": 0.7143649473100389, + "grad_norm": 4.219130039215088, + "learning_rate": 4.2386831275720165e-05, + "loss": 3.2443, + "step": 322 + }, + { + "epoch": 0.7165834719911259, + "grad_norm": 3.986713171005249, + "learning_rate": 4.234567901234568e-05, + "loss": 3.1975, + "step": 323 + }, + { + "epoch": 0.718801996672213, + "grad_norm": 4.343851089477539, + "learning_rate": 4.2304526748971197e-05, + "loss": 3.2402, + "step": 324 + }, + { + "epoch": 0.7210205213533001, + "grad_norm": 4.035735607147217, + "learning_rate": 4.2263374485596706e-05, + "loss": 2.9415, + "step": 325 + }, + { + "epoch": 0.7232390460343872, + "grad_norm": 4.227270603179932, + "learning_rate": 4.222222222222222e-05, + "loss": 3.3391, + "step": 326 + }, + { + "epoch": 0.7254575707154742, + "grad_norm": 4.317235469818115, + "learning_rate": 4.2181069958847744e-05, + "loss": 3.5274, + "step": 327 + }, + { + "epoch": 0.7276760953965613, + "grad_norm": 3.9255032539367676, + "learning_rate": 4.213991769547325e-05, + "loss": 3.1529, + "step": 328 + }, + { + "epoch": 0.7298946200776484, + "grad_norm": 4.6280694007873535, + "learning_rate": 4.209876543209877e-05, + "loss": 3.2162, + "step": 329 + }, + { + "epoch": 0.7321131447587355, + "grad_norm": 3.9828941822052, + "learning_rate": 4.205761316872428e-05, + "loss": 2.9742, + "step": 330 + }, + { + "epoch": 0.7343316694398225, + "grad_norm": 3.941335439682007, + "learning_rate": 4.2016460905349794e-05, + "loss": 3.326, + "step": 331 + }, + { + "epoch": 0.7365501941209096, + "grad_norm": 4.460665702819824, + "learning_rate": 4.197530864197531e-05, + "loss": 3.5276, + "step": 332 + }, + { + "epoch": 0.7387687188019967, + "grad_norm": 4.022602558135986, + "learning_rate": 4.1934156378600826e-05, + "loss": 2.8265, + "step": 333 + }, + { + "epoch": 0.7409872434830838, + "grad_norm": 3.9025964736938477, + "learning_rate": 4.189300411522634e-05, + "loss": 3.0678, + "step": 334 + }, + { + "epoch": 0.7432057681641708, + "grad_norm": 3.751948356628418, + "learning_rate": 4.185185185185185e-05, + "loss": 3.0311, + "step": 335 + }, + { + "epoch": 0.7454242928452579, + "grad_norm": 3.8363678455352783, + "learning_rate": 4.181069958847737e-05, + "loss": 3.0461, + "step": 336 + }, + { + "epoch": 0.747642817526345, + "grad_norm": 4.661734580993652, + "learning_rate": 4.176954732510288e-05, + "loss": 3.4654, + "step": 337 + }, + { + "epoch": 0.7498613422074321, + "grad_norm": 3.624544858932495, + "learning_rate": 4.17283950617284e-05, + "loss": 3.0499, + "step": 338 + }, + { + "epoch": 0.7520798668885191, + "grad_norm": 4.1967597007751465, + "learning_rate": 4.168724279835391e-05, + "loss": 3.357, + "step": 339 + }, + { + "epoch": 0.7542983915696062, + "grad_norm": 3.605306625366211, + "learning_rate": 4.164609053497942e-05, + "loss": 3.2803, + "step": 340 + }, + { + "epoch": 0.7565169162506933, + "grad_norm": 3.9251527786254883, + "learning_rate": 4.1604938271604946e-05, + "loss": 2.9314, + "step": 341 + }, + { + "epoch": 0.7587354409317804, + "grad_norm": 4.086531639099121, + "learning_rate": 4.1563786008230455e-05, + "loss": 2.7407, + "step": 342 + }, + { + "epoch": 0.7609539656128674, + "grad_norm": 4.0090861320495605, + "learning_rate": 4.152263374485597e-05, + "loss": 3.1164, + "step": 343 + }, + { + "epoch": 0.7631724902939545, + "grad_norm": 4.068595886230469, + "learning_rate": 4.148148148148148e-05, + "loss": 3.198, + "step": 344 + }, + { + "epoch": 0.7653910149750416, + "grad_norm": 3.9793202877044678, + "learning_rate": 4.1440329218106996e-05, + "loss": 3.3727, + "step": 345 + }, + { + "epoch": 0.7676095396561287, + "grad_norm": 3.915395498275757, + "learning_rate": 4.139917695473251e-05, + "loss": 3.2462, + "step": 346 + }, + { + "epoch": 0.7698280643372157, + "grad_norm": 4.488144397735596, + "learning_rate": 4.135802469135803e-05, + "loss": 3.2002, + "step": 347 + }, + { + "epoch": 0.7720465890183028, + "grad_norm": 4.146156311035156, + "learning_rate": 4.1316872427983537e-05, + "loss": 3.2987, + "step": 348 + }, + { + "epoch": 0.7742651136993899, + "grad_norm": 4.341185092926025, + "learning_rate": 4.127572016460905e-05, + "loss": 3.0765, + "step": 349 + }, + { + "epoch": 0.776483638380477, + "grad_norm": 4.122524261474609, + "learning_rate": 4.1234567901234575e-05, + "loss": 3.0699, + "step": 350 + }, + { + "epoch": 0.778702163061564, + "grad_norm": 4.303515911102295, + "learning_rate": 4.1193415637860084e-05, + "loss": 2.9912, + "step": 351 + }, + { + "epoch": 0.7809206877426511, + "grad_norm": 4.033647060394287, + "learning_rate": 4.11522633744856e-05, + "loss": 3.0208, + "step": 352 + }, + { + "epoch": 0.7831392124237382, + "grad_norm": 4.140043258666992, + "learning_rate": 4.111111111111111e-05, + "loss": 3.0409, + "step": 353 + }, + { + "epoch": 0.7853577371048253, + "grad_norm": 4.846273422241211, + "learning_rate": 4.1069958847736625e-05, + "loss": 3.2445, + "step": 354 + }, + { + "epoch": 0.7875762617859123, + "grad_norm": 4.226095199584961, + "learning_rate": 4.102880658436214e-05, + "loss": 3.2736, + "step": 355 + }, + { + "epoch": 0.7897947864669994, + "grad_norm": 4.134472846984863, + "learning_rate": 4.0987654320987657e-05, + "loss": 3.2875, + "step": 356 + }, + { + "epoch": 0.7920133111480865, + "grad_norm": 4.167330265045166, + "learning_rate": 4.094650205761317e-05, + "loss": 2.8983, + "step": 357 + }, + { + "epoch": 0.7942318358291736, + "grad_norm": 4.200919151306152, + "learning_rate": 4.090534979423868e-05, + "loss": 2.9499, + "step": 358 + }, + { + "epoch": 0.7964503605102606, + "grad_norm": 3.660663366317749, + "learning_rate": 4.0864197530864204e-05, + "loss": 2.9281, + "step": 359 + }, + { + "epoch": 0.7986688851913477, + "grad_norm": 4.0184431076049805, + "learning_rate": 4.082304526748971e-05, + "loss": 3.1832, + "step": 360 + }, + { + "epoch": 0.8008874098724348, + "grad_norm": 3.829444408416748, + "learning_rate": 4.078189300411523e-05, + "loss": 2.7744, + "step": 361 + }, + { + "epoch": 0.8031059345535219, + "grad_norm": 4.169424057006836, + "learning_rate": 4.074074074074074e-05, + "loss": 3.1612, + "step": 362 + }, + { + "epoch": 0.8053244592346089, + "grad_norm": 3.7797656059265137, + "learning_rate": 4.0699588477366254e-05, + "loss": 3.2244, + "step": 363 + }, + { + "epoch": 0.807542983915696, + "grad_norm": 4.050520896911621, + "learning_rate": 4.065843621399178e-05, + "loss": 3.165, + "step": 364 + }, + { + "epoch": 0.8097615085967831, + "grad_norm": 3.902958631515503, + "learning_rate": 4.0617283950617286e-05, + "loss": 3.049, + "step": 365 + }, + { + "epoch": 0.8119800332778702, + "grad_norm": 3.780456066131592, + "learning_rate": 4.05761316872428e-05, + "loss": 2.7271, + "step": 366 + }, + { + "epoch": 0.8141985579589573, + "grad_norm": 4.1589226722717285, + "learning_rate": 4.053497942386831e-05, + "loss": 3.4523, + "step": 367 + }, + { + "epoch": 0.8164170826400444, + "grad_norm": 3.9211738109588623, + "learning_rate": 4.049382716049383e-05, + "loss": 2.9924, + "step": 368 + }, + { + "epoch": 0.8186356073211315, + "grad_norm": 4.295964241027832, + "learning_rate": 4.045267489711934e-05, + "loss": 3.1542, + "step": 369 + }, + { + "epoch": 0.8208541320022186, + "grad_norm": 4.099763870239258, + "learning_rate": 4.041152263374486e-05, + "loss": 3.0062, + "step": 370 + }, + { + "epoch": 0.8230726566833056, + "grad_norm": 3.606642007827759, + "learning_rate": 4.0370370370370374e-05, + "loss": 3.0073, + "step": 371 + }, + { + "epoch": 0.8252911813643927, + "grad_norm": 4.5731306076049805, + "learning_rate": 4.032921810699588e-05, + "loss": 3.2897, + "step": 372 + }, + { + "epoch": 0.8275097060454798, + "grad_norm": 4.294490337371826, + "learning_rate": 4.0288065843621406e-05, + "loss": 3.2286, + "step": 373 + }, + { + "epoch": 0.8297282307265669, + "grad_norm": 4.153075695037842, + "learning_rate": 4.0246913580246915e-05, + "loss": 2.9182, + "step": 374 + }, + { + "epoch": 0.831946755407654, + "grad_norm": 4.124756336212158, + "learning_rate": 4.020576131687243e-05, + "loss": 3.0922, + "step": 375 + }, + { + "epoch": 0.834165280088741, + "grad_norm": 4.014266014099121, + "learning_rate": 4.016460905349794e-05, + "loss": 3.4029, + "step": 376 + }, + { + "epoch": 0.8363838047698281, + "grad_norm": 4.15941858291626, + "learning_rate": 4.012345679012346e-05, + "loss": 3.1866, + "step": 377 + }, + { + "epoch": 0.8386023294509152, + "grad_norm": 3.8265514373779297, + "learning_rate": 4.008230452674897e-05, + "loss": 3.2515, + "step": 378 + }, + { + "epoch": 0.8408208541320022, + "grad_norm": 3.6474077701568604, + "learning_rate": 4.004115226337449e-05, + "loss": 3.0033, + "step": 379 + }, + { + "epoch": 0.8430393788130893, + "grad_norm": 3.554481029510498, + "learning_rate": 4e-05, + "loss": 3.0461, + "step": 380 + }, + { + "epoch": 0.8452579034941764, + "grad_norm": 3.786874532699585, + "learning_rate": 3.995884773662551e-05, + "loss": 3.1073, + "step": 381 + }, + { + "epoch": 0.8474764281752635, + "grad_norm": 3.774604558944702, + "learning_rate": 3.9917695473251035e-05, + "loss": 2.8989, + "step": 382 + }, + { + "epoch": 0.8496949528563505, + "grad_norm": 3.8100473880767822, + "learning_rate": 3.9876543209876544e-05, + "loss": 2.778, + "step": 383 + }, + { + "epoch": 0.8519134775374376, + "grad_norm": 4.193554878234863, + "learning_rate": 3.983539094650206e-05, + "loss": 2.851, + "step": 384 + }, + { + "epoch": 0.8541320022185247, + "grad_norm": 3.4792885780334473, + "learning_rate": 3.979423868312757e-05, + "loss": 2.9311, + "step": 385 + }, + { + "epoch": 0.8563505268996118, + "grad_norm": 5.102469444274902, + "learning_rate": 3.975308641975309e-05, + "loss": 2.9832, + "step": 386 + }, + { + "epoch": 0.8585690515806988, + "grad_norm": 3.6804559230804443, + "learning_rate": 3.971193415637861e-05, + "loss": 3.1126, + "step": 387 + }, + { + "epoch": 0.8607875762617859, + "grad_norm": 4.578532695770264, + "learning_rate": 3.9670781893004117e-05, + "loss": 3.1337, + "step": 388 + }, + { + "epoch": 0.863006100942873, + "grad_norm": 4.2652668952941895, + "learning_rate": 3.962962962962963e-05, + "loss": 3.1119, + "step": 389 + }, + { + "epoch": 0.8652246256239601, + "grad_norm": 4.12747859954834, + "learning_rate": 3.958847736625514e-05, + "loss": 3.0991, + "step": 390 + }, + { + "epoch": 0.8674431503050472, + "grad_norm": 4.23831033706665, + "learning_rate": 3.9547325102880664e-05, + "loss": 3.0866, + "step": 391 + }, + { + "epoch": 0.8696616749861342, + "grad_norm": 3.7862677574157715, + "learning_rate": 3.950617283950617e-05, + "loss": 2.9735, + "step": 392 + }, + { + "epoch": 0.8718801996672213, + "grad_norm": 3.935744285583496, + "learning_rate": 3.946502057613169e-05, + "loss": 3.0124, + "step": 393 + }, + { + "epoch": 0.8740987243483084, + "grad_norm": 4.2084736824035645, + "learning_rate": 3.9423868312757205e-05, + "loss": 3.3819, + "step": 394 + }, + { + "epoch": 0.8763172490293955, + "grad_norm": 3.4937243461608887, + "learning_rate": 3.938271604938272e-05, + "loss": 2.841, + "step": 395 + }, + { + "epoch": 0.8785357737104825, + "grad_norm": 3.8536415100097656, + "learning_rate": 3.934156378600824e-05, + "loss": 2.9963, + "step": 396 + }, + { + "epoch": 0.8807542983915696, + "grad_norm": 3.809607744216919, + "learning_rate": 3.9300411522633746e-05, + "loss": 3.0783, + "step": 397 + }, + { + "epoch": 0.8829728230726567, + "grad_norm": 3.4915058612823486, + "learning_rate": 3.925925925925926e-05, + "loss": 2.9895, + "step": 398 + }, + { + "epoch": 0.8851913477537438, + "grad_norm": 3.998762369155884, + "learning_rate": 3.921810699588477e-05, + "loss": 3.3429, + "step": 399 + }, + { + "epoch": 0.8874098724348308, + "grad_norm": 4.4517951011657715, + "learning_rate": 3.917695473251029e-05, + "loss": 3.3334, + "step": 400 + }, + { + "epoch": 0.8896283971159179, + "grad_norm": 4.0959320068359375, + "learning_rate": 3.91358024691358e-05, + "loss": 2.5998, + "step": 401 + }, + { + "epoch": 0.891846921797005, + "grad_norm": 4.3074870109558105, + "learning_rate": 3.909465020576132e-05, + "loss": 3.1336, + "step": 402 + }, + { + "epoch": 0.894065446478092, + "grad_norm": 4.217771530151367, + "learning_rate": 3.9053497942386834e-05, + "loss": 3.0343, + "step": 403 + }, + { + "epoch": 0.8962839711591791, + "grad_norm": 3.8567652702331543, + "learning_rate": 3.901234567901234e-05, + "loss": 3.0218, + "step": 404 + }, + { + "epoch": 0.8985024958402662, + "grad_norm": 3.6335904598236084, + "learning_rate": 3.8971193415637866e-05, + "loss": 3.0051, + "step": 405 + }, + { + "epoch": 0.9007210205213533, + "grad_norm": 4.670261383056641, + "learning_rate": 3.8930041152263375e-05, + "loss": 2.8037, + "step": 406 + }, + { + "epoch": 0.9029395452024404, + "grad_norm": 3.909151792526245, + "learning_rate": 3.888888888888889e-05, + "loss": 3.211, + "step": 407 + }, + { + "epoch": 0.9051580698835274, + "grad_norm": 3.6557536125183105, + "learning_rate": 3.88477366255144e-05, + "loss": 2.9679, + "step": 408 + }, + { + "epoch": 0.9073765945646145, + "grad_norm": 3.7534427642822266, + "learning_rate": 3.880658436213992e-05, + "loss": 3.1407, + "step": 409 + }, + { + "epoch": 0.9095951192457016, + "grad_norm": 4.349733352661133, + "learning_rate": 3.876543209876544e-05, + "loss": 2.8153, + "step": 410 + }, + { + "epoch": 0.9118136439267887, + "grad_norm": 4.290287017822266, + "learning_rate": 3.872427983539095e-05, + "loss": 2.9201, + "step": 411 + }, + { + "epoch": 0.9140321686078757, + "grad_norm": 3.779778480529785, + "learning_rate": 3.868312757201646e-05, + "loss": 2.9133, + "step": 412 + }, + { + "epoch": 0.9162506932889628, + "grad_norm": 4.139045715332031, + "learning_rate": 3.864197530864197e-05, + "loss": 2.754, + "step": 413 + }, + { + "epoch": 0.9184692179700499, + "grad_norm": 4.156318664550781, + "learning_rate": 3.8600823045267495e-05, + "loss": 2.9966, + "step": 414 + }, + { + "epoch": 0.920687742651137, + "grad_norm": 3.8492860794067383, + "learning_rate": 3.8559670781893004e-05, + "loss": 2.841, + "step": 415 + }, + { + "epoch": 0.922906267332224, + "grad_norm": 3.6557183265686035, + "learning_rate": 3.851851851851852e-05, + "loss": 2.6662, + "step": 416 + }, + { + "epoch": 0.9251247920133111, + "grad_norm": 4.052201271057129, + "learning_rate": 3.8477366255144036e-05, + "loss": 3.2302, + "step": 417 + }, + { + "epoch": 0.9273433166943982, + "grad_norm": 3.795206069946289, + "learning_rate": 3.843621399176955e-05, + "loss": 2.9505, + "step": 418 + }, + { + "epoch": 0.9295618413754853, + "grad_norm": 4.3553876876831055, + "learning_rate": 3.839506172839507e-05, + "loss": 3.0911, + "step": 419 + }, + { + "epoch": 0.9317803660565723, + "grad_norm": 3.8547379970550537, + "learning_rate": 3.8353909465020577e-05, + "loss": 3.0471, + "step": 420 + }, + { + "epoch": 0.9339988907376594, + "grad_norm": 3.6057567596435547, + "learning_rate": 3.831275720164609e-05, + "loss": 2.8832, + "step": 421 + }, + { + "epoch": 0.9362174154187465, + "grad_norm": 3.6521315574645996, + "learning_rate": 3.82716049382716e-05, + "loss": 3.1121, + "step": 422 + }, + { + "epoch": 0.9384359400998337, + "grad_norm": 3.780792236328125, + "learning_rate": 3.8230452674897124e-05, + "loss": 3.1721, + "step": 423 + }, + { + "epoch": 0.9406544647809207, + "grad_norm": 3.777114152908325, + "learning_rate": 3.818930041152264e-05, + "loss": 3.0315, + "step": 424 + }, + { + "epoch": 0.9428729894620078, + "grad_norm": 3.5418272018432617, + "learning_rate": 3.814814814814815e-05, + "loss": 2.9704, + "step": 425 + }, + { + "epoch": 0.9450915141430949, + "grad_norm": 4.913465976715088, + "learning_rate": 3.8106995884773665e-05, + "loss": 3.2033, + "step": 426 + }, + { + "epoch": 0.947310038824182, + "grad_norm": 3.8711516857147217, + "learning_rate": 3.806584362139918e-05, + "loss": 2.98, + "step": 427 + }, + { + "epoch": 0.949528563505269, + "grad_norm": 4.114070415496826, + "learning_rate": 3.80246913580247e-05, + "loss": 3.0579, + "step": 428 + }, + { + "epoch": 0.9517470881863561, + "grad_norm": 3.9892330169677734, + "learning_rate": 3.7983539094650206e-05, + "loss": 2.9362, + "step": 429 + }, + { + "epoch": 0.9539656128674432, + "grad_norm": 4.3047003746032715, + "learning_rate": 3.794238683127572e-05, + "loss": 2.8793, + "step": 430 + }, + { + "epoch": 0.9561841375485303, + "grad_norm": 4.078428745269775, + "learning_rate": 3.790123456790123e-05, + "loss": 3.0196, + "step": 431 + }, + { + "epoch": 0.9584026622296173, + "grad_norm": 3.661656379699707, + "learning_rate": 3.786008230452675e-05, + "loss": 2.806, + "step": 432 + }, + { + "epoch": 0.9606211869107044, + "grad_norm": 3.8340139389038086, + "learning_rate": 3.781893004115227e-05, + "loss": 2.9195, + "step": 433 + }, + { + "epoch": 0.9628397115917915, + "grad_norm": 4.425356864929199, + "learning_rate": 3.777777777777778e-05, + "loss": 3.3246, + "step": 434 + }, + { + "epoch": 0.9650582362728786, + "grad_norm": 3.788599967956543, + "learning_rate": 3.7736625514403294e-05, + "loss": 2.5006, + "step": 435 + }, + { + "epoch": 0.9672767609539656, + "grad_norm": 3.6781575679779053, + "learning_rate": 3.769547325102881e-05, + "loss": 3.0682, + "step": 436 + }, + { + "epoch": 0.9694952856350527, + "grad_norm": 4.299561023712158, + "learning_rate": 3.7654320987654326e-05, + "loss": 2.9373, + "step": 437 + }, + { + "epoch": 0.9717138103161398, + "grad_norm": 4.0907487869262695, + "learning_rate": 3.7613168724279835e-05, + "loss": 2.7509, + "step": 438 + }, + { + "epoch": 0.9739323349972269, + "grad_norm": 4.051908493041992, + "learning_rate": 3.757201646090535e-05, + "loss": 2.8863, + "step": 439 + }, + { + "epoch": 0.9761508596783139, + "grad_norm": 3.7211685180664062, + "learning_rate": 3.7530864197530867e-05, + "loss": 3.0988, + "step": 440 + }, + { + "epoch": 0.978369384359401, + "grad_norm": 3.464411735534668, + "learning_rate": 3.748971193415638e-05, + "loss": 2.7387, + "step": 441 + }, + { + "epoch": 0.9805879090404881, + "grad_norm": 3.821256399154663, + "learning_rate": 3.74485596707819e-05, + "loss": 2.7261, + "step": 442 + }, + { + "epoch": 0.9828064337215752, + "grad_norm": 3.7412209510803223, + "learning_rate": 3.740740740740741e-05, + "loss": 3.1608, + "step": 443 + }, + { + "epoch": 0.9850249584026622, + "grad_norm": 4.151949405670166, + "learning_rate": 3.736625514403292e-05, + "loss": 2.8473, + "step": 444 + }, + { + "epoch": 0.9872434830837493, + "grad_norm": 4.479198932647705, + "learning_rate": 3.732510288065844e-05, + "loss": 2.8572, + "step": 445 + }, + { + "epoch": 0.9894620077648364, + "grad_norm": 3.9793031215667725, + "learning_rate": 3.7283950617283955e-05, + "loss": 2.8204, + "step": 446 + }, + { + "epoch": 0.9916805324459235, + "grad_norm": 3.8117809295654297, + "learning_rate": 3.724279835390947e-05, + "loss": 2.6009, + "step": 447 + }, + { + "epoch": 0.9938990571270105, + "grad_norm": 4.1536760330200195, + "learning_rate": 3.720164609053498e-05, + "loss": 3.0199, + "step": 448 + }, + { + "epoch": 0.9961175818080976, + "grad_norm": 4.665118217468262, + "learning_rate": 3.7160493827160496e-05, + "loss": 2.7474, + "step": 449 + }, + { + "epoch": 0.9983361064891847, + "grad_norm": 4.217140197753906, + "learning_rate": 3.711934156378601e-05, + "loss": 2.7188, + "step": 450 + }, + { + "epoch": 1.0005546311702718, + "grad_norm": 3.663736343383789, + "learning_rate": 3.707818930041153e-05, + "loss": 3.0799, + "step": 451 + }, + { + "epoch": 1.0027731558513588, + "grad_norm": 4.016757011413574, + "learning_rate": 3.7037037037037037e-05, + "loss": 2.3872, + "step": 452 + }, + { + "epoch": 1.004991680532446, + "grad_norm": 4.123344421386719, + "learning_rate": 3.699588477366255e-05, + "loss": 2.3915, + "step": 453 + }, + { + "epoch": 1.007210205213533, + "grad_norm": 4.707493305206299, + "learning_rate": 3.695473251028807e-05, + "loss": 2.6426, + "step": 454 + }, + { + "epoch": 1.00942872989462, + "grad_norm": 4.369646072387695, + "learning_rate": 3.6913580246913584e-05, + "loss": 2.5267, + "step": 455 + }, + { + "epoch": 1.0116472545757071, + "grad_norm": 4.436652183532715, + "learning_rate": 3.68724279835391e-05, + "loss": 2.5854, + "step": 456 + }, + { + "epoch": 1.0138657792567942, + "grad_norm": 4.110020637512207, + "learning_rate": 3.683127572016461e-05, + "loss": 2.2715, + "step": 457 + }, + { + "epoch": 1.0160843039378813, + "grad_norm": 4.102577209472656, + "learning_rate": 3.6790123456790125e-05, + "loss": 2.3777, + "step": 458 + }, + { + "epoch": 1.0183028286189684, + "grad_norm": 4.552696228027344, + "learning_rate": 3.674897119341564e-05, + "loss": 2.3946, + "step": 459 + }, + { + "epoch": 1.0205213533000554, + "grad_norm": 4.0455193519592285, + "learning_rate": 3.670781893004116e-05, + "loss": 2.1066, + "step": 460 + }, + { + "epoch": 1.0227398779811425, + "grad_norm": 4.069112300872803, + "learning_rate": 3.6666666666666666e-05, + "loss": 2.5035, + "step": 461 + }, + { + "epoch": 1.0249584026622296, + "grad_norm": 3.665438175201416, + "learning_rate": 3.662551440329218e-05, + "loss": 2.5676, + "step": 462 + }, + { + "epoch": 1.0271769273433167, + "grad_norm": 3.7036309242248535, + "learning_rate": 3.65843621399177e-05, + "loss": 2.3813, + "step": 463 + }, + { + "epoch": 1.0293954520244037, + "grad_norm": 4.176037311553955, + "learning_rate": 3.654320987654321e-05, + "loss": 2.2902, + "step": 464 + }, + { + "epoch": 1.0316139767054908, + "grad_norm": 4.588430404663086, + "learning_rate": 3.650205761316873e-05, + "loss": 2.4714, + "step": 465 + }, + { + "epoch": 1.033832501386578, + "grad_norm": 4.322487831115723, + "learning_rate": 3.646090534979424e-05, + "loss": 2.4925, + "step": 466 + }, + { + "epoch": 1.036051026067665, + "grad_norm": 4.5990495681762695, + "learning_rate": 3.6419753086419754e-05, + "loss": 2.5991, + "step": 467 + }, + { + "epoch": 1.038269550748752, + "grad_norm": 4.286531448364258, + "learning_rate": 3.637860082304527e-05, + "loss": 2.4002, + "step": 468 + }, + { + "epoch": 1.0404880754298391, + "grad_norm": 4.367201328277588, + "learning_rate": 3.6337448559670786e-05, + "loss": 2.4754, + "step": 469 + }, + { + "epoch": 1.0427066001109262, + "grad_norm": 3.8899192810058594, + "learning_rate": 3.62962962962963e-05, + "loss": 2.2955, + "step": 470 + }, + { + "epoch": 1.0449251247920133, + "grad_norm": 3.5898427963256836, + "learning_rate": 3.625514403292181e-05, + "loss": 2.1714, + "step": 471 + }, + { + "epoch": 1.0471436494731003, + "grad_norm": 4.042567729949951, + "learning_rate": 3.6213991769547327e-05, + "loss": 2.7085, + "step": 472 + }, + { + "epoch": 1.0493621741541874, + "grad_norm": 3.8549957275390625, + "learning_rate": 3.617283950617284e-05, + "loss": 2.4079, + "step": 473 + }, + { + "epoch": 1.0515806988352745, + "grad_norm": 3.8808138370513916, + "learning_rate": 3.613168724279836e-05, + "loss": 2.5593, + "step": 474 + }, + { + "epoch": 1.0537992235163616, + "grad_norm": 4.318870544433594, + "learning_rate": 3.609053497942387e-05, + "loss": 2.4303, + "step": 475 + }, + { + "epoch": 1.0560177481974486, + "grad_norm": 3.690723180770874, + "learning_rate": 3.604938271604938e-05, + "loss": 2.1645, + "step": 476 + }, + { + "epoch": 1.0582362728785357, + "grad_norm": 4.717057228088379, + "learning_rate": 3.60082304526749e-05, + "loss": 2.0229, + "step": 477 + }, + { + "epoch": 1.0604547975596228, + "grad_norm": 4.158778667449951, + "learning_rate": 3.5967078189300415e-05, + "loss": 2.2412, + "step": 478 + }, + { + "epoch": 1.0626733222407099, + "grad_norm": 4.444665908813477, + "learning_rate": 3.592592592592593e-05, + "loss": 2.5468, + "step": 479 + }, + { + "epoch": 1.064891846921797, + "grad_norm": 4.733316421508789, + "learning_rate": 3.588477366255144e-05, + "loss": 2.7055, + "step": 480 + }, + { + "epoch": 1.067110371602884, + "grad_norm": 4.197329998016357, + "learning_rate": 3.5843621399176956e-05, + "loss": 2.2208, + "step": 481 + }, + { + "epoch": 1.069328896283971, + "grad_norm": 3.8544178009033203, + "learning_rate": 3.580246913580247e-05, + "loss": 2.3809, + "step": 482 + }, + { + "epoch": 1.0715474209650582, + "grad_norm": 3.933499813079834, + "learning_rate": 3.576131687242799e-05, + "loss": 2.748, + "step": 483 + }, + { + "epoch": 1.0737659456461452, + "grad_norm": 4.320080757141113, + "learning_rate": 3.5720164609053497e-05, + "loss": 2.3988, + "step": 484 + }, + { + "epoch": 1.0759844703272323, + "grad_norm": 3.8032546043395996, + "learning_rate": 3.567901234567901e-05, + "loss": 2.2317, + "step": 485 + }, + { + "epoch": 1.0782029950083194, + "grad_norm": 3.877716302871704, + "learning_rate": 3.563786008230453e-05, + "loss": 2.7696, + "step": 486 + }, + { + "epoch": 1.0804215196894065, + "grad_norm": 4.630105495452881, + "learning_rate": 3.5596707818930044e-05, + "loss": 2.3231, + "step": 487 + }, + { + "epoch": 1.0826400443704935, + "grad_norm": 4.17884635925293, + "learning_rate": 3.555555555555556e-05, + "loss": 2.4731, + "step": 488 + }, + { + "epoch": 1.0848585690515806, + "grad_norm": 4.2930827140808105, + "learning_rate": 3.551440329218107e-05, + "loss": 2.7115, + "step": 489 + }, + { + "epoch": 1.0870770937326677, + "grad_norm": 4.076241493225098, + "learning_rate": 3.5473251028806585e-05, + "loss": 2.5439, + "step": 490 + }, + { + "epoch": 1.0892956184137548, + "grad_norm": 4.153909206390381, + "learning_rate": 3.54320987654321e-05, + "loss": 2.4563, + "step": 491 + }, + { + "epoch": 1.0915141430948418, + "grad_norm": 3.8423712253570557, + "learning_rate": 3.539094650205762e-05, + "loss": 2.7272, + "step": 492 + }, + { + "epoch": 1.093732667775929, + "grad_norm": 4.100058555603027, + "learning_rate": 3.534979423868313e-05, + "loss": 2.574, + "step": 493 + }, + { + "epoch": 1.095951192457016, + "grad_norm": 4.248688697814941, + "learning_rate": 3.530864197530864e-05, + "loss": 2.3599, + "step": 494 + }, + { + "epoch": 1.098169717138103, + "grad_norm": 4.449590682983398, + "learning_rate": 3.526748971193416e-05, + "loss": 2.5356, + "step": 495 + }, + { + "epoch": 1.1003882418191901, + "grad_norm": 4.3084940910339355, + "learning_rate": 3.522633744855967e-05, + "loss": 2.5411, + "step": 496 + }, + { + "epoch": 1.1026067665002772, + "grad_norm": 3.871274709701538, + "learning_rate": 3.518518518518519e-05, + "loss": 2.7009, + "step": 497 + }, + { + "epoch": 1.1048252911813643, + "grad_norm": 3.597825765609741, + "learning_rate": 3.51440329218107e-05, + "loss": 2.3427, + "step": 498 + }, + { + "epoch": 1.1070438158624514, + "grad_norm": 4.126097679138184, + "learning_rate": 3.5102880658436214e-05, + "loss": 2.4639, + "step": 499 + }, + { + "epoch": 1.1092623405435384, + "grad_norm": 3.9073421955108643, + "learning_rate": 3.506172839506173e-05, + "loss": 2.5326, + "step": 500 + }, + { + "epoch": 1.1114808652246255, + "grad_norm": 3.9562268257141113, + "learning_rate": 3.5020576131687246e-05, + "loss": 2.2201, + "step": 501 + }, + { + "epoch": 1.1136993899057126, + "grad_norm": 4.67470645904541, + "learning_rate": 3.497942386831276e-05, + "loss": 2.4851, + "step": 502 + }, + { + "epoch": 1.1159179145867997, + "grad_norm": 3.7237064838409424, + "learning_rate": 3.493827160493827e-05, + "loss": 2.6412, + "step": 503 + }, + { + "epoch": 1.1181364392678868, + "grad_norm": 3.978882074356079, + "learning_rate": 3.4897119341563787e-05, + "loss": 2.4756, + "step": 504 + }, + { + "epoch": 1.1203549639489738, + "grad_norm": 4.078076362609863, + "learning_rate": 3.48559670781893e-05, + "loss": 2.495, + "step": 505 + }, + { + "epoch": 1.122573488630061, + "grad_norm": 4.068276882171631, + "learning_rate": 3.481481481481482e-05, + "loss": 2.3662, + "step": 506 + }, + { + "epoch": 1.124792013311148, + "grad_norm": 4.234014987945557, + "learning_rate": 3.4773662551440334e-05, + "loss": 2.5452, + "step": 507 + }, + { + "epoch": 1.127010537992235, + "grad_norm": 4.388455390930176, + "learning_rate": 3.473251028806584e-05, + "loss": 2.0573, + "step": 508 + }, + { + "epoch": 1.1292290626733221, + "grad_norm": 3.3357832431793213, + "learning_rate": 3.469135802469136e-05, + "loss": 2.4277, + "step": 509 + }, + { + "epoch": 1.1314475873544092, + "grad_norm": 4.213736534118652, + "learning_rate": 3.4650205761316875e-05, + "loss": 2.4496, + "step": 510 + }, + { + "epoch": 1.1336661120354963, + "grad_norm": 4.35080099105835, + "learning_rate": 3.460905349794239e-05, + "loss": 2.7298, + "step": 511 + }, + { + "epoch": 1.1358846367165834, + "grad_norm": 4.051361560821533, + "learning_rate": 3.45679012345679e-05, + "loss": 2.4701, + "step": 512 + }, + { + "epoch": 1.1381031613976704, + "grad_norm": 4.115354537963867, + "learning_rate": 3.4526748971193416e-05, + "loss": 2.4155, + "step": 513 + }, + { + "epoch": 1.1403216860787575, + "grad_norm": 4.553377151489258, + "learning_rate": 3.448559670781893e-05, + "loss": 2.1426, + "step": 514 + }, + { + "epoch": 1.1425402107598448, + "grad_norm": 4.284748077392578, + "learning_rate": 3.444444444444445e-05, + "loss": 2.4332, + "step": 515 + }, + { + "epoch": 1.1447587354409319, + "grad_norm": 4.459059715270996, + "learning_rate": 3.440329218106996e-05, + "loss": 2.6648, + "step": 516 + }, + { + "epoch": 1.146977260122019, + "grad_norm": 4.4325432777404785, + "learning_rate": 3.436213991769547e-05, + "loss": 2.6446, + "step": 517 + }, + { + "epoch": 1.149195784803106, + "grad_norm": 4.3818769454956055, + "learning_rate": 3.432098765432099e-05, + "loss": 2.488, + "step": 518 + }, + { + "epoch": 1.151414309484193, + "grad_norm": 3.7928929328918457, + "learning_rate": 3.4279835390946504e-05, + "loss": 2.6927, + "step": 519 + }, + { + "epoch": 1.1536328341652802, + "grad_norm": 4.150576591491699, + "learning_rate": 3.423868312757202e-05, + "loss": 2.1619, + "step": 520 + }, + { + "epoch": 1.1558513588463672, + "grad_norm": 4.055851936340332, + "learning_rate": 3.419753086419753e-05, + "loss": 2.527, + "step": 521 + }, + { + "epoch": 1.1580698835274543, + "grad_norm": 4.087257385253906, + "learning_rate": 3.4156378600823045e-05, + "loss": 2.4962, + "step": 522 + }, + { + "epoch": 1.1602884082085414, + "grad_norm": 3.972928285598755, + "learning_rate": 3.411522633744856e-05, + "loss": 2.2517, + "step": 523 + }, + { + "epoch": 1.1625069328896285, + "grad_norm": 4.338064670562744, + "learning_rate": 3.4074074074074077e-05, + "loss": 2.4888, + "step": 524 + }, + { + "epoch": 1.1647254575707155, + "grad_norm": 3.753241539001465, + "learning_rate": 3.403292181069959e-05, + "loss": 2.2914, + "step": 525 + }, + { + "epoch": 1.1669439822518026, + "grad_norm": 3.8234362602233887, + "learning_rate": 3.39917695473251e-05, + "loss": 2.3571, + "step": 526 + }, + { + "epoch": 1.1691625069328897, + "grad_norm": 3.865476608276367, + "learning_rate": 3.395061728395062e-05, + "loss": 2.1054, + "step": 527 + }, + { + "epoch": 1.1713810316139768, + "grad_norm": 4.162647247314453, + "learning_rate": 3.390946502057613e-05, + "loss": 2.4552, + "step": 528 + }, + { + "epoch": 1.1735995562950639, + "grad_norm": 4.341167449951172, + "learning_rate": 3.386831275720165e-05, + "loss": 2.3191, + "step": 529 + }, + { + "epoch": 1.175818080976151, + "grad_norm": 4.435120582580566, + "learning_rate": 3.3827160493827165e-05, + "loss": 2.3835, + "step": 530 + }, + { + "epoch": 1.178036605657238, + "grad_norm": 4.9813432693481445, + "learning_rate": 3.3786008230452674e-05, + "loss": 2.4667, + "step": 531 + }, + { + "epoch": 1.180255130338325, + "grad_norm": 4.553356647491455, + "learning_rate": 3.374485596707819e-05, + "loss": 2.5243, + "step": 532 + }, + { + "epoch": 1.1824736550194122, + "grad_norm": 4.472057342529297, + "learning_rate": 3.3703703703703706e-05, + "loss": 2.1561, + "step": 533 + }, + { + "epoch": 1.1846921797004992, + "grad_norm": 3.5702545642852783, + "learning_rate": 3.366255144032922e-05, + "loss": 2.2841, + "step": 534 + }, + { + "epoch": 1.1869107043815863, + "grad_norm": 4.6420369148254395, + "learning_rate": 3.362139917695473e-05, + "loss": 2.5425, + "step": 535 + }, + { + "epoch": 1.1891292290626734, + "grad_norm": 4.727165222167969, + "learning_rate": 3.3580246913580247e-05, + "loss": 2.4011, + "step": 536 + }, + { + "epoch": 1.1913477537437605, + "grad_norm": 4.669632911682129, + "learning_rate": 3.353909465020576e-05, + "loss": 2.1578, + "step": 537 + }, + { + "epoch": 1.1935662784248475, + "grad_norm": 4.512310981750488, + "learning_rate": 3.349794238683128e-05, + "loss": 2.5912, + "step": 538 + }, + { + "epoch": 1.1957848031059346, + "grad_norm": 4.288442611694336, + "learning_rate": 3.3456790123456794e-05, + "loss": 2.4427, + "step": 539 + }, + { + "epoch": 1.1980033277870217, + "grad_norm": 4.098184585571289, + "learning_rate": 3.34156378600823e-05, + "loss": 2.4585, + "step": 540 + }, + { + "epoch": 1.2002218524681088, + "grad_norm": 3.859036684036255, + "learning_rate": 3.337448559670782e-05, + "loss": 2.2392, + "step": 541 + }, + { + "epoch": 1.2024403771491958, + "grad_norm": 4.416805267333984, + "learning_rate": 3.3333333333333335e-05, + "loss": 2.4253, + "step": 542 + }, + { + "epoch": 1.204658901830283, + "grad_norm": 4.012598991394043, + "learning_rate": 3.329218106995885e-05, + "loss": 2.3087, + "step": 543 + }, + { + "epoch": 1.20687742651137, + "grad_norm": 4.01534366607666, + "learning_rate": 3.325102880658436e-05, + "loss": 2.2264, + "step": 544 + }, + { + "epoch": 1.209095951192457, + "grad_norm": 4.603600978851318, + "learning_rate": 3.3209876543209876e-05, + "loss": 2.3535, + "step": 545 + }, + { + "epoch": 1.2113144758735441, + "grad_norm": 3.821659564971924, + "learning_rate": 3.316872427983539e-05, + "loss": 2.2621, + "step": 546 + }, + { + "epoch": 1.2135330005546312, + "grad_norm": 4.456361293792725, + "learning_rate": 3.312757201646091e-05, + "loss": 2.249, + "step": 547 + }, + { + "epoch": 1.2157515252357183, + "grad_norm": 4.629476547241211, + "learning_rate": 3.308641975308642e-05, + "loss": 2.1472, + "step": 548 + }, + { + "epoch": 1.2179700499168054, + "grad_norm": 4.362195014953613, + "learning_rate": 3.304526748971193e-05, + "loss": 2.4244, + "step": 549 + }, + { + "epoch": 1.2201885745978924, + "grad_norm": 4.740861415863037, + "learning_rate": 3.300411522633745e-05, + "loss": 2.5379, + "step": 550 + }, + { + "epoch": 1.2224070992789795, + "grad_norm": 4.2255120277404785, + "learning_rate": 3.2962962962962964e-05, + "loss": 2.4208, + "step": 551 + }, + { + "epoch": 1.2246256239600666, + "grad_norm": 4.779180526733398, + "learning_rate": 3.292181069958848e-05, + "loss": 2.4789, + "step": 552 + }, + { + "epoch": 1.2268441486411537, + "grad_norm": 3.9484500885009766, + "learning_rate": 3.2880658436213996e-05, + "loss": 2.2757, + "step": 553 + }, + { + "epoch": 1.2290626733222407, + "grad_norm": 3.9741711616516113, + "learning_rate": 3.2839506172839505e-05, + "loss": 2.4309, + "step": 554 + }, + { + "epoch": 1.2312811980033278, + "grad_norm": 3.9606285095214844, + "learning_rate": 3.279835390946502e-05, + "loss": 2.3854, + "step": 555 + }, + { + "epoch": 1.2334997226844149, + "grad_norm": 3.4910073280334473, + "learning_rate": 3.2757201646090537e-05, + "loss": 2.3084, + "step": 556 + }, + { + "epoch": 1.235718247365502, + "grad_norm": 4.2925825119018555, + "learning_rate": 3.271604938271605e-05, + "loss": 2.2042, + "step": 557 + }, + { + "epoch": 1.237936772046589, + "grad_norm": 3.9305036067962646, + "learning_rate": 3.267489711934156e-05, + "loss": 1.9932, + "step": 558 + }, + { + "epoch": 1.240155296727676, + "grad_norm": 3.976029634475708, + "learning_rate": 3.263374485596708e-05, + "loss": 2.5798, + "step": 559 + }, + { + "epoch": 1.2423738214087632, + "grad_norm": 3.9460747241973877, + "learning_rate": 3.25925925925926e-05, + "loss": 2.2117, + "step": 560 + }, + { + "epoch": 1.2445923460898503, + "grad_norm": 4.674043655395508, + "learning_rate": 3.255144032921811e-05, + "loss": 2.3407, + "step": 561 + }, + { + "epoch": 1.2468108707709373, + "grad_norm": 3.9236679077148438, + "learning_rate": 3.2510288065843625e-05, + "loss": 2.2961, + "step": 562 + }, + { + "epoch": 1.2490293954520244, + "grad_norm": 3.9175961017608643, + "learning_rate": 3.2469135802469134e-05, + "loss": 2.4009, + "step": 563 + }, + { + "epoch": 1.2512479201331115, + "grad_norm": 4.286099433898926, + "learning_rate": 3.242798353909465e-05, + "loss": 2.3604, + "step": 564 + }, + { + "epoch": 1.2534664448141986, + "grad_norm": 4.037940502166748, + "learning_rate": 3.2386831275720166e-05, + "loss": 2.4544, + "step": 565 + }, + { + "epoch": 1.2556849694952856, + "grad_norm": 3.7699592113494873, + "learning_rate": 3.234567901234568e-05, + "loss": 2.3568, + "step": 566 + }, + { + "epoch": 1.2579034941763727, + "grad_norm": 4.0947136878967285, + "learning_rate": 3.230452674897119e-05, + "loss": 2.5318, + "step": 567 + }, + { + "epoch": 1.2601220188574598, + "grad_norm": 3.550818681716919, + "learning_rate": 3.2263374485596707e-05, + "loss": 2.4885, + "step": 568 + }, + { + "epoch": 1.2623405435385469, + "grad_norm": 3.716627836227417, + "learning_rate": 3.222222222222223e-05, + "loss": 2.2967, + "step": 569 + }, + { + "epoch": 1.264559068219634, + "grad_norm": 3.7392661571502686, + "learning_rate": 3.218106995884774e-05, + "loss": 2.3745, + "step": 570 + }, + { + "epoch": 1.266777592900721, + "grad_norm": 3.5491394996643066, + "learning_rate": 3.2139917695473254e-05, + "loss": 2.0732, + "step": 571 + }, + { + "epoch": 1.268996117581808, + "grad_norm": 4.176416873931885, + "learning_rate": 3.209876543209876e-05, + "loss": 2.3932, + "step": 572 + }, + { + "epoch": 1.2712146422628952, + "grad_norm": 3.9201905727386475, + "learning_rate": 3.205761316872428e-05, + "loss": 2.4914, + "step": 573 + }, + { + "epoch": 1.2734331669439822, + "grad_norm": 3.999587297439575, + "learning_rate": 3.2016460905349795e-05, + "loss": 2.3094, + "step": 574 + }, + { + "epoch": 1.2756516916250693, + "grad_norm": 3.7487635612487793, + "learning_rate": 3.197530864197531e-05, + "loss": 2.571, + "step": 575 + }, + { + "epoch": 1.2778702163061564, + "grad_norm": 3.8658547401428223, + "learning_rate": 3.193415637860083e-05, + "loss": 2.048, + "step": 576 + }, + { + "epoch": 1.2800887409872435, + "grad_norm": 4.641408920288086, + "learning_rate": 3.1893004115226336e-05, + "loss": 2.3786, + "step": 577 + }, + { + "epoch": 1.2823072656683305, + "grad_norm": 4.479084491729736, + "learning_rate": 3.185185185185185e-05, + "loss": 2.4724, + "step": 578 + }, + { + "epoch": 1.2845257903494176, + "grad_norm": 3.9457547664642334, + "learning_rate": 3.181069958847737e-05, + "loss": 2.137, + "step": 579 + }, + { + "epoch": 1.2867443150305047, + "grad_norm": 4.635665416717529, + "learning_rate": 3.176954732510288e-05, + "loss": 2.4133, + "step": 580 + }, + { + "epoch": 1.2889628397115918, + "grad_norm": 3.9766151905059814, + "learning_rate": 3.172839506172839e-05, + "loss": 2.5152, + "step": 581 + }, + { + "epoch": 1.2911813643926788, + "grad_norm": 4.015515327453613, + "learning_rate": 3.168724279835391e-05, + "loss": 2.3287, + "step": 582 + }, + { + "epoch": 1.293399889073766, + "grad_norm": 3.9840633869171143, + "learning_rate": 3.164609053497943e-05, + "loss": 1.9578, + "step": 583 + }, + { + "epoch": 1.295618413754853, + "grad_norm": 4.282557964324951, + "learning_rate": 3.160493827160494e-05, + "loss": 2.3051, + "step": 584 + }, + { + "epoch": 1.29783693843594, + "grad_norm": 4.012889385223389, + "learning_rate": 3.1563786008230456e-05, + "loss": 2.4207, + "step": 585 + }, + { + "epoch": 1.3000554631170271, + "grad_norm": 4.529271602630615, + "learning_rate": 3.1522633744855965e-05, + "loss": 2.3237, + "step": 586 + }, + { + "epoch": 1.3022739877981142, + "grad_norm": 4.471197128295898, + "learning_rate": 3.148148148148148e-05, + "loss": 2.3808, + "step": 587 + }, + { + "epoch": 1.3044925124792013, + "grad_norm": 4.088883399963379, + "learning_rate": 3.1440329218106997e-05, + "loss": 2.336, + "step": 588 + }, + { + "epoch": 1.3067110371602884, + "grad_norm": 4.030023097991943, + "learning_rate": 3.139917695473251e-05, + "loss": 2.3506, + "step": 589 + }, + { + "epoch": 1.3089295618413754, + "grad_norm": 4.363866329193115, + "learning_rate": 3.135802469135803e-05, + "loss": 2.4201, + "step": 590 + }, + { + "epoch": 1.3111480865224625, + "grad_norm": 4.0475239753723145, + "learning_rate": 3.131687242798354e-05, + "loss": 2.1912, + "step": 591 + }, + { + "epoch": 1.3133666112035496, + "grad_norm": 4.5530171394348145, + "learning_rate": 3.127572016460906e-05, + "loss": 2.3913, + "step": 592 + }, + { + "epoch": 1.3155851358846367, + "grad_norm": 3.4183101654052734, + "learning_rate": 3.123456790123457e-05, + "loss": 2.3566, + "step": 593 + }, + { + "epoch": 1.3178036605657237, + "grad_norm": 4.089936256408691, + "learning_rate": 3.1193415637860085e-05, + "loss": 2.6309, + "step": 594 + }, + { + "epoch": 1.3200221852468108, + "grad_norm": 3.829469680786133, + "learning_rate": 3.1152263374485594e-05, + "loss": 2.4679, + "step": 595 + }, + { + "epoch": 1.3222407099278979, + "grad_norm": 4.022919178009033, + "learning_rate": 3.111111111111111e-05, + "loss": 2.3149, + "step": 596 + }, + { + "epoch": 1.324459234608985, + "grad_norm": 4.072815895080566, + "learning_rate": 3.1069958847736626e-05, + "loss": 2.1776, + "step": 597 + }, + { + "epoch": 1.326677759290072, + "grad_norm": 4.395301342010498, + "learning_rate": 3.102880658436214e-05, + "loss": 2.2288, + "step": 598 + }, + { + "epoch": 1.328896283971159, + "grad_norm": 4.749038219451904, + "learning_rate": 3.098765432098766e-05, + "loss": 2.3302, + "step": 599 + }, + { + "epoch": 1.3311148086522462, + "grad_norm": 4.089650630950928, + "learning_rate": 3.0946502057613167e-05, + "loss": 2.318, + "step": 600 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 3.4254374504089355, + "learning_rate": 3.090534979423869e-05, + "loss": 2.2864, + "step": 601 + }, + { + "epoch": 1.3355518580144203, + "grad_norm": 4.544743537902832, + "learning_rate": 3.08641975308642e-05, + "loss": 2.5758, + "step": 602 + }, + { + "epoch": 1.3377703826955074, + "grad_norm": 3.674410343170166, + "learning_rate": 3.0823045267489714e-05, + "loss": 2.208, + "step": 603 + }, + { + "epoch": 1.3399889073765945, + "grad_norm": 4.548892021179199, + "learning_rate": 3.078189300411522e-05, + "loss": 2.7019, + "step": 604 + }, + { + "epoch": 1.3422074320576816, + "grad_norm": 3.9281082153320312, + "learning_rate": 3.074074074074074e-05, + "loss": 2.4806, + "step": 605 + }, + { + "epoch": 1.3444259567387689, + "grad_norm": 4.0055389404296875, + "learning_rate": 3.069958847736626e-05, + "loss": 2.2918, + "step": 606 + }, + { + "epoch": 1.346644481419856, + "grad_norm": 3.7127838134765625, + "learning_rate": 3.065843621399177e-05, + "loss": 2.6629, + "step": 607 + }, + { + "epoch": 1.348863006100943, + "grad_norm": 3.489732503890991, + "learning_rate": 3.061728395061729e-05, + "loss": 2.2226, + "step": 608 + }, + { + "epoch": 1.35108153078203, + "grad_norm": 3.703979015350342, + "learning_rate": 3.0576131687242796e-05, + "loss": 2.6089, + "step": 609 + }, + { + "epoch": 1.3533000554631172, + "grad_norm": 4.246747970581055, + "learning_rate": 3.053497942386832e-05, + "loss": 2.1939, + "step": 610 + }, + { + "epoch": 1.3555185801442042, + "grad_norm": 3.8159308433532715, + "learning_rate": 3.0493827160493827e-05, + "loss": 2.1799, + "step": 611 + }, + { + "epoch": 1.3577371048252913, + "grad_norm": 4.390836238861084, + "learning_rate": 3.0452674897119343e-05, + "loss": 2.5376, + "step": 612 + }, + { + "epoch": 1.3599556295063784, + "grad_norm": 4.016605377197266, + "learning_rate": 3.041152263374486e-05, + "loss": 2.1765, + "step": 613 + }, + { + "epoch": 1.3621741541874655, + "grad_norm": 3.7472140789031982, + "learning_rate": 3.037037037037037e-05, + "loss": 2.2988, + "step": 614 + }, + { + "epoch": 1.3643926788685525, + "grad_norm": 4.373779773712158, + "learning_rate": 3.0329218106995887e-05, + "loss": 2.5876, + "step": 615 + }, + { + "epoch": 1.3666112035496396, + "grad_norm": 3.801569700241089, + "learning_rate": 3.02880658436214e-05, + "loss": 2.2484, + "step": 616 + }, + { + "epoch": 1.3688297282307267, + "grad_norm": 4.406134128570557, + "learning_rate": 3.0246913580246916e-05, + "loss": 2.5269, + "step": 617 + }, + { + "epoch": 1.3710482529118138, + "grad_norm": 4.064319610595703, + "learning_rate": 3.0205761316872428e-05, + "loss": 2.5371, + "step": 618 + }, + { + "epoch": 1.3732667775929008, + "grad_norm": 4.002112865447998, + "learning_rate": 3.0164609053497944e-05, + "loss": 2.2885, + "step": 619 + }, + { + "epoch": 1.375485302273988, + "grad_norm": 3.554884910583496, + "learning_rate": 3.012345679012346e-05, + "loss": 2.1652, + "step": 620 + }, + { + "epoch": 1.377703826955075, + "grad_norm": 3.713920831680298, + "learning_rate": 3.0082304526748972e-05, + "loss": 2.3239, + "step": 621 + }, + { + "epoch": 1.379922351636162, + "grad_norm": 3.7386646270751953, + "learning_rate": 3.0041152263374488e-05, + "loss": 2.1581, + "step": 622 + }, + { + "epoch": 1.3821408763172491, + "grad_norm": 4.203656196594238, + "learning_rate": 3e-05, + "loss": 2.2252, + "step": 623 + }, + { + "epoch": 1.3843594009983362, + "grad_norm": 3.8118200302124023, + "learning_rate": 2.9958847736625517e-05, + "loss": 2.164, + "step": 624 + }, + { + "epoch": 1.3865779256794233, + "grad_norm": 4.537066459655762, + "learning_rate": 2.991769547325103e-05, + "loss": 2.1578, + "step": 625 + }, + { + "epoch": 1.3887964503605104, + "grad_norm": 4.253244876861572, + "learning_rate": 2.9876543209876545e-05, + "loss": 2.1271, + "step": 626 + }, + { + "epoch": 1.3910149750415974, + "grad_norm": 3.6401307582855225, + "learning_rate": 2.9835390946502057e-05, + "loss": 2.0195, + "step": 627 + }, + { + "epoch": 1.3932334997226845, + "grad_norm": 4.0085625648498535, + "learning_rate": 2.9794238683127573e-05, + "loss": 2.1488, + "step": 628 + }, + { + "epoch": 1.3954520244037716, + "grad_norm": 4.079653739929199, + "learning_rate": 2.975308641975309e-05, + "loss": 2.2248, + "step": 629 + }, + { + "epoch": 1.3976705490848587, + "grad_norm": 3.7720460891723633, + "learning_rate": 2.97119341563786e-05, + "loss": 2.3619, + "step": 630 + }, + { + "epoch": 1.3998890737659457, + "grad_norm": 3.850748300552368, + "learning_rate": 2.9670781893004117e-05, + "loss": 2.3522, + "step": 631 + }, + { + "epoch": 1.4021075984470328, + "grad_norm": 3.7575185298919678, + "learning_rate": 2.962962962962963e-05, + "loss": 2.201, + "step": 632 + }, + { + "epoch": 1.4043261231281199, + "grad_norm": 3.5917763710021973, + "learning_rate": 2.9588477366255146e-05, + "loss": 2.4188, + "step": 633 + }, + { + "epoch": 1.406544647809207, + "grad_norm": 4.1729865074157715, + "learning_rate": 2.9547325102880658e-05, + "loss": 2.2417, + "step": 634 + }, + { + "epoch": 1.408763172490294, + "grad_norm": 3.9747564792633057, + "learning_rate": 2.9506172839506174e-05, + "loss": 2.1458, + "step": 635 + }, + { + "epoch": 1.410981697171381, + "grad_norm": 3.802412748336792, + "learning_rate": 2.946502057613169e-05, + "loss": 2.2767, + "step": 636 + }, + { + "epoch": 1.4132002218524682, + "grad_norm": 3.6346733570098877, + "learning_rate": 2.9423868312757202e-05, + "loss": 2.192, + "step": 637 + }, + { + "epoch": 1.4154187465335553, + "grad_norm": 4.105311393737793, + "learning_rate": 2.9382716049382718e-05, + "loss": 2.259, + "step": 638 + }, + { + "epoch": 1.4176372712146423, + "grad_norm": 3.9020328521728516, + "learning_rate": 2.934156378600823e-05, + "loss": 2.3168, + "step": 639 + }, + { + "epoch": 1.4198557958957294, + "grad_norm": 4.573899269104004, + "learning_rate": 2.9300411522633747e-05, + "loss": 2.3088, + "step": 640 + }, + { + "epoch": 1.4220743205768165, + "grad_norm": 4.5805535316467285, + "learning_rate": 2.925925925925926e-05, + "loss": 2.6135, + "step": 641 + }, + { + "epoch": 1.4242928452579036, + "grad_norm": 4.3353471755981445, + "learning_rate": 2.9218106995884775e-05, + "loss": 2.1512, + "step": 642 + }, + { + "epoch": 1.4265113699389906, + "grad_norm": 4.574662208557129, + "learning_rate": 2.917695473251029e-05, + "loss": 2.1382, + "step": 643 + }, + { + "epoch": 1.4287298946200777, + "grad_norm": 3.9994781017303467, + "learning_rate": 2.9135802469135803e-05, + "loss": 2.4249, + "step": 644 + }, + { + "epoch": 1.4309484193011648, + "grad_norm": 3.6273584365844727, + "learning_rate": 2.909465020576132e-05, + "loss": 2.2747, + "step": 645 + }, + { + "epoch": 1.4331669439822519, + "grad_norm": 3.8340954780578613, + "learning_rate": 2.905349794238683e-05, + "loss": 2.242, + "step": 646 + }, + { + "epoch": 1.435385468663339, + "grad_norm": 4.17929744720459, + "learning_rate": 2.9012345679012347e-05, + "loss": 2.3913, + "step": 647 + }, + { + "epoch": 1.437603993344426, + "grad_norm": 3.8408725261688232, + "learning_rate": 2.897119341563786e-05, + "loss": 1.9756, + "step": 648 + }, + { + "epoch": 1.439822518025513, + "grad_norm": 3.8463351726531982, + "learning_rate": 2.8930041152263376e-05, + "loss": 2.3436, + "step": 649 + }, + { + "epoch": 1.4420410427066002, + "grad_norm": 4.132676124572754, + "learning_rate": 2.8888888888888888e-05, + "loss": 2.2231, + "step": 650 + }, + { + "epoch": 1.4442595673876872, + "grad_norm": 3.964935064315796, + "learning_rate": 2.8847736625514404e-05, + "loss": 2.1741, + "step": 651 + }, + { + "epoch": 1.4464780920687743, + "grad_norm": 4.325876235961914, + "learning_rate": 2.880658436213992e-05, + "loss": 2.2319, + "step": 652 + }, + { + "epoch": 1.4486966167498614, + "grad_norm": 4.9623918533325195, + "learning_rate": 2.8765432098765432e-05, + "loss": 2.2155, + "step": 653 + }, + { + "epoch": 1.4509151414309485, + "grad_norm": 4.399862289428711, + "learning_rate": 2.8724279835390948e-05, + "loss": 2.029, + "step": 654 + }, + { + "epoch": 1.4531336661120355, + "grad_norm": 3.9154767990112305, + "learning_rate": 2.868312757201646e-05, + "loss": 2.1283, + "step": 655 + }, + { + "epoch": 1.4553521907931226, + "grad_norm": 4.250678062438965, + "learning_rate": 2.8641975308641977e-05, + "loss": 1.9629, + "step": 656 + }, + { + "epoch": 1.4575707154742097, + "grad_norm": 4.611120700836182, + "learning_rate": 2.860082304526749e-05, + "loss": 2.201, + "step": 657 + }, + { + "epoch": 1.4597892401552968, + "grad_norm": 3.887782573699951, + "learning_rate": 2.8559670781893005e-05, + "loss": 2.241, + "step": 658 + }, + { + "epoch": 1.4620077648363838, + "grad_norm": 4.356259822845459, + "learning_rate": 2.851851851851852e-05, + "loss": 2.5105, + "step": 659 + }, + { + "epoch": 1.464226289517471, + "grad_norm": 3.968994617462158, + "learning_rate": 2.8477366255144033e-05, + "loss": 1.6937, + "step": 660 + }, + { + "epoch": 1.466444814198558, + "grad_norm": 3.6068594455718994, + "learning_rate": 2.843621399176955e-05, + "loss": 2.5917, + "step": 661 + }, + { + "epoch": 1.468663338879645, + "grad_norm": 4.254590034484863, + "learning_rate": 2.839506172839506e-05, + "loss": 2.3225, + "step": 662 + }, + { + "epoch": 1.4708818635607321, + "grad_norm": 3.749721050262451, + "learning_rate": 2.8353909465020577e-05, + "loss": 2.2715, + "step": 663 + }, + { + "epoch": 1.4731003882418192, + "grad_norm": 3.8069770336151123, + "learning_rate": 2.831275720164609e-05, + "loss": 2.3337, + "step": 664 + }, + { + "epoch": 1.4753189129229063, + "grad_norm": 3.895416021347046, + "learning_rate": 2.8271604938271606e-05, + "loss": 2.2944, + "step": 665 + }, + { + "epoch": 1.4775374376039934, + "grad_norm": 3.8730530738830566, + "learning_rate": 2.823045267489712e-05, + "loss": 2.5907, + "step": 666 + }, + { + "epoch": 1.4797559622850804, + "grad_norm": 3.635863780975342, + "learning_rate": 2.8189300411522634e-05, + "loss": 2.5144, + "step": 667 + }, + { + "epoch": 1.4819744869661675, + "grad_norm": 4.339428424835205, + "learning_rate": 2.814814814814815e-05, + "loss": 2.1789, + "step": 668 + }, + { + "epoch": 1.4841930116472546, + "grad_norm": 4.260381698608398, + "learning_rate": 2.8106995884773662e-05, + "loss": 2.4531, + "step": 669 + }, + { + "epoch": 1.4864115363283417, + "grad_norm": 4.653558731079102, + "learning_rate": 2.8065843621399178e-05, + "loss": 2.4117, + "step": 670 + }, + { + "epoch": 1.4886300610094287, + "grad_norm": 3.8421361446380615, + "learning_rate": 2.802469135802469e-05, + "loss": 2.2148, + "step": 671 + }, + { + "epoch": 1.4908485856905158, + "grad_norm": 3.98817777633667, + "learning_rate": 2.7983539094650207e-05, + "loss": 2.3422, + "step": 672 + }, + { + "epoch": 1.493067110371603, + "grad_norm": 3.7659225463867188, + "learning_rate": 2.7942386831275726e-05, + "loss": 2.2374, + "step": 673 + }, + { + "epoch": 1.49528563505269, + "grad_norm": 4.270452976226807, + "learning_rate": 2.7901234567901235e-05, + "loss": 2.3291, + "step": 674 + }, + { + "epoch": 1.497504159733777, + "grad_norm": 4.088283538818359, + "learning_rate": 2.786008230452675e-05, + "loss": 2.2311, + "step": 675 + }, + { + "epoch": 1.4997226844148641, + "grad_norm": 4.297863006591797, + "learning_rate": 2.7818930041152263e-05, + "loss": 2.2877, + "step": 676 + }, + { + "epoch": 1.5019412090959512, + "grad_norm": 3.8774235248565674, + "learning_rate": 2.777777777777778e-05, + "loss": 2.2132, + "step": 677 + }, + { + "epoch": 1.5041597337770383, + "grad_norm": 4.21061897277832, + "learning_rate": 2.773662551440329e-05, + "loss": 1.9031, + "step": 678 + }, + { + "epoch": 1.5063782584581253, + "grad_norm": 4.13560152053833, + "learning_rate": 2.7695473251028807e-05, + "loss": 2.0203, + "step": 679 + }, + { + "epoch": 1.5085967831392124, + "grad_norm": 4.211042404174805, + "learning_rate": 2.765432098765432e-05, + "loss": 2.2563, + "step": 680 + }, + { + "epoch": 1.5108153078202995, + "grad_norm": 4.167869567871094, + "learning_rate": 2.7613168724279836e-05, + "loss": 2.463, + "step": 681 + }, + { + "epoch": 1.5130338325013866, + "grad_norm": 4.629598617553711, + "learning_rate": 2.757201646090535e-05, + "loss": 2.2994, + "step": 682 + }, + { + "epoch": 1.5152523571824736, + "grad_norm": 3.96327805519104, + "learning_rate": 2.7530864197530864e-05, + "loss": 2.0035, + "step": 683 + }, + { + "epoch": 1.5174708818635607, + "grad_norm": 4.349621772766113, + "learning_rate": 2.748971193415638e-05, + "loss": 2.2251, + "step": 684 + }, + { + "epoch": 1.5196894065446478, + "grad_norm": 4.3172712326049805, + "learning_rate": 2.7448559670781892e-05, + "loss": 2.2768, + "step": 685 + }, + { + "epoch": 1.5219079312257349, + "grad_norm": 4.351469993591309, + "learning_rate": 2.7407407407407408e-05, + "loss": 2.3002, + "step": 686 + }, + { + "epoch": 1.524126455906822, + "grad_norm": 3.774775266647339, + "learning_rate": 2.736625514403292e-05, + "loss": 2.1081, + "step": 687 + }, + { + "epoch": 1.526344980587909, + "grad_norm": 3.4512412548065186, + "learning_rate": 2.7325102880658437e-05, + "loss": 2.3601, + "step": 688 + }, + { + "epoch": 1.528563505268996, + "grad_norm": 4.177464008331299, + "learning_rate": 2.7283950617283956e-05, + "loss": 2.2514, + "step": 689 + }, + { + "epoch": 1.5307820299500832, + "grad_norm": 4.318098068237305, + "learning_rate": 2.7242798353909465e-05, + "loss": 2.2614, + "step": 690 + }, + { + "epoch": 1.5330005546311702, + "grad_norm": 3.631258964538574, + "learning_rate": 2.720164609053498e-05, + "loss": 2.3129, + "step": 691 + }, + { + "epoch": 1.5352190793122573, + "grad_norm": 4.141443252563477, + "learning_rate": 2.7160493827160493e-05, + "loss": 2.5664, + "step": 692 + }, + { + "epoch": 1.5374376039933444, + "grad_norm": 3.958188533782959, + "learning_rate": 2.711934156378601e-05, + "loss": 2.5954, + "step": 693 + }, + { + "epoch": 1.5396561286744315, + "grad_norm": 4.144777297973633, + "learning_rate": 2.707818930041152e-05, + "loss": 2.2577, + "step": 694 + }, + { + "epoch": 1.5418746533555185, + "grad_norm": 4.301990032196045, + "learning_rate": 2.7037037037037037e-05, + "loss": 1.9011, + "step": 695 + }, + { + "epoch": 1.5440931780366056, + "grad_norm": 3.9203460216522217, + "learning_rate": 2.6995884773662557e-05, + "loss": 2.1243, + "step": 696 + }, + { + "epoch": 1.5463117027176927, + "grad_norm": 4.094918727874756, + "learning_rate": 2.6954732510288066e-05, + "loss": 2.1672, + "step": 697 + }, + { + "epoch": 1.5485302273987798, + "grad_norm": 4.028655529022217, + "learning_rate": 2.6913580246913585e-05, + "loss": 2.1955, + "step": 698 + }, + { + "epoch": 1.5507487520798668, + "grad_norm": 4.240545749664307, + "learning_rate": 2.6872427983539094e-05, + "loss": 2.1937, + "step": 699 + }, + { + "epoch": 1.552967276760954, + "grad_norm": 4.467401027679443, + "learning_rate": 2.683127572016461e-05, + "loss": 2.1469, + "step": 700 + }, + { + "epoch": 1.555185801442041, + "grad_norm": 4.088786602020264, + "learning_rate": 2.6790123456790122e-05, + "loss": 2.1852, + "step": 701 + }, + { + "epoch": 1.557404326123128, + "grad_norm": 4.218378067016602, + "learning_rate": 2.6748971193415638e-05, + "loss": 2.2537, + "step": 702 + }, + { + "epoch": 1.5596228508042151, + "grad_norm": 4.318513870239258, + "learning_rate": 2.6707818930041158e-05, + "loss": 2.5191, + "step": 703 + }, + { + "epoch": 1.5618413754853022, + "grad_norm": 3.4546780586242676, + "learning_rate": 2.6666666666666667e-05, + "loss": 2.2809, + "step": 704 + }, + { + "epoch": 1.5640599001663893, + "grad_norm": 4.606024742126465, + "learning_rate": 2.6625514403292186e-05, + "loss": 2.0729, + "step": 705 + }, + { + "epoch": 1.5662784248474764, + "grad_norm": 3.872488260269165, + "learning_rate": 2.6584362139917695e-05, + "loss": 2.1367, + "step": 706 + }, + { + "epoch": 1.5684969495285634, + "grad_norm": 3.9035353660583496, + "learning_rate": 2.654320987654321e-05, + "loss": 1.9595, + "step": 707 + }, + { + "epoch": 1.5707154742096505, + "grad_norm": 3.989694356918335, + "learning_rate": 2.6502057613168723e-05, + "loss": 2.1013, + "step": 708 + }, + { + "epoch": 1.5729339988907376, + "grad_norm": 4.163475513458252, + "learning_rate": 2.646090534979424e-05, + "loss": 2.362, + "step": 709 + }, + { + "epoch": 1.5751525235718247, + "grad_norm": 4.137301921844482, + "learning_rate": 2.641975308641975e-05, + "loss": 2.3601, + "step": 710 + }, + { + "epoch": 1.5773710482529117, + "grad_norm": 4.229010581970215, + "learning_rate": 2.6378600823045267e-05, + "loss": 2.4772, + "step": 711 + }, + { + "epoch": 1.5795895729339988, + "grad_norm": 3.581212282180786, + "learning_rate": 2.6337448559670787e-05, + "loss": 2.234, + "step": 712 + }, + { + "epoch": 1.581808097615086, + "grad_norm": 4.449773788452148, + "learning_rate": 2.6296296296296296e-05, + "loss": 2.0019, + "step": 713 + }, + { + "epoch": 1.584026622296173, + "grad_norm": 3.8118135929107666, + "learning_rate": 2.6255144032921815e-05, + "loss": 2.177, + "step": 714 + }, + { + "epoch": 1.58624514697726, + "grad_norm": 4.020204067230225, + "learning_rate": 2.6213991769547324e-05, + "loss": 2.1385, + "step": 715 + }, + { + "epoch": 1.5884636716583471, + "grad_norm": 4.133056640625, + "learning_rate": 2.617283950617284e-05, + "loss": 2.3681, + "step": 716 + }, + { + "epoch": 1.5906821963394342, + "grad_norm": 3.714425802230835, + "learning_rate": 2.6131687242798352e-05, + "loss": 2.1759, + "step": 717 + }, + { + "epoch": 1.5929007210205213, + "grad_norm": 4.794270038604736, + "learning_rate": 2.6090534979423868e-05, + "loss": 2.3126, + "step": 718 + }, + { + "epoch": 1.5951192457016083, + "grad_norm": 4.016239166259766, + "learning_rate": 2.6049382716049388e-05, + "loss": 2.0764, + "step": 719 + }, + { + "epoch": 1.5973377703826954, + "grad_norm": 4.135544300079346, + "learning_rate": 2.6008230452674897e-05, + "loss": 2.3189, + "step": 720 + }, + { + "epoch": 1.5995562950637825, + "grad_norm": 3.8435475826263428, + "learning_rate": 2.5967078189300416e-05, + "loss": 2.3339, + "step": 721 + }, + { + "epoch": 1.6017748197448696, + "grad_norm": 3.8262276649475098, + "learning_rate": 2.5925925925925925e-05, + "loss": 2.3075, + "step": 722 + }, + { + "epoch": 1.6039933444259566, + "grad_norm": 4.241980075836182, + "learning_rate": 2.5884773662551444e-05, + "loss": 2.0394, + "step": 723 + }, + { + "epoch": 1.6062118691070437, + "grad_norm": 4.060568332672119, + "learning_rate": 2.5843621399176953e-05, + "loss": 2.4457, + "step": 724 + }, + { + "epoch": 1.6084303937881308, + "grad_norm": 3.8236539363861084, + "learning_rate": 2.580246913580247e-05, + "loss": 2.1317, + "step": 725 + }, + { + "epoch": 1.6106489184692179, + "grad_norm": 4.589874744415283, + "learning_rate": 2.576131687242799e-05, + "loss": 2.2723, + "step": 726 + }, + { + "epoch": 1.612867443150305, + "grad_norm": 3.71867299079895, + "learning_rate": 2.5720164609053497e-05, + "loss": 1.8852, + "step": 727 + }, + { + "epoch": 1.615085967831392, + "grad_norm": 4.523905277252197, + "learning_rate": 2.5679012345679017e-05, + "loss": 2.0833, + "step": 728 + }, + { + "epoch": 1.617304492512479, + "grad_norm": 4.160734176635742, + "learning_rate": 2.5637860082304526e-05, + "loss": 2.4889, + "step": 729 + }, + { + "epoch": 1.6195230171935662, + "grad_norm": 3.7809929847717285, + "learning_rate": 2.5596707818930045e-05, + "loss": 2.0417, + "step": 730 + }, + { + "epoch": 1.6217415418746532, + "grad_norm": 3.746743679046631, + "learning_rate": 2.5555555555555554e-05, + "loss": 2.1892, + "step": 731 + }, + { + "epoch": 1.6239600665557403, + "grad_norm": 4.11684513092041, + "learning_rate": 2.551440329218107e-05, + "loss": 1.9482, + "step": 732 + }, + { + "epoch": 1.6261785912368274, + "grad_norm": 3.924650192260742, + "learning_rate": 2.5473251028806582e-05, + "loss": 1.9622, + "step": 733 + }, + { + "epoch": 1.6283971159179145, + "grad_norm": 4.384554862976074, + "learning_rate": 2.5432098765432098e-05, + "loss": 2.1134, + "step": 734 + }, + { + "epoch": 1.6306156405990015, + "grad_norm": 3.9588046073913574, + "learning_rate": 2.5390946502057617e-05, + "loss": 2.2972, + "step": 735 + }, + { + "epoch": 1.6328341652800886, + "grad_norm": 4.203953266143799, + "learning_rate": 2.5349794238683127e-05, + "loss": 2.1272, + "step": 736 + }, + { + "epoch": 1.6350526899611757, + "grad_norm": 3.835453987121582, + "learning_rate": 2.5308641975308646e-05, + "loss": 2.0221, + "step": 737 + }, + { + "epoch": 1.6372712146422628, + "grad_norm": 3.986339807510376, + "learning_rate": 2.5267489711934155e-05, + "loss": 1.9596, + "step": 738 + }, + { + "epoch": 1.6394897393233498, + "grad_norm": 3.834696054458618, + "learning_rate": 2.5226337448559674e-05, + "loss": 2.3406, + "step": 739 + }, + { + "epoch": 1.641708264004437, + "grad_norm": 4.222187519073486, + "learning_rate": 2.5185185185185183e-05, + "loss": 2.35, + "step": 740 + }, + { + "epoch": 1.643926788685524, + "grad_norm": 3.8070225715637207, + "learning_rate": 2.51440329218107e-05, + "loss": 2.351, + "step": 741 + }, + { + "epoch": 1.646145313366611, + "grad_norm": 3.8583500385284424, + "learning_rate": 2.510288065843622e-05, + "loss": 2.1027, + "step": 742 + }, + { + "epoch": 1.6483638380476981, + "grad_norm": 3.6536943912506104, + "learning_rate": 2.5061728395061727e-05, + "loss": 2.1246, + "step": 743 + }, + { + "epoch": 1.6505823627287852, + "grad_norm": 4.509573459625244, + "learning_rate": 2.5020576131687247e-05, + "loss": 2.3402, + "step": 744 + }, + { + "epoch": 1.6528008874098723, + "grad_norm": 4.037731170654297, + "learning_rate": 2.497942386831276e-05, + "loss": 2.1476, + "step": 745 + }, + { + "epoch": 1.6550194120909594, + "grad_norm": 4.125838279724121, + "learning_rate": 2.4938271604938275e-05, + "loss": 2.3302, + "step": 746 + }, + { + "epoch": 1.6572379367720464, + "grad_norm": 4.298431396484375, + "learning_rate": 2.4897119341563787e-05, + "loss": 2.4729, + "step": 747 + }, + { + "epoch": 1.6594564614531335, + "grad_norm": 4.03386116027832, + "learning_rate": 2.4855967078189303e-05, + "loss": 2.2577, + "step": 748 + }, + { + "epoch": 1.6616749861342206, + "grad_norm": 4.168376922607422, + "learning_rate": 2.4814814814814816e-05, + "loss": 2.2789, + "step": 749 + }, + { + "epoch": 1.6638935108153077, + "grad_norm": 3.8156657218933105, + "learning_rate": 2.4773662551440328e-05, + "loss": 1.9957, + "step": 750 + }, + { + "epoch": 1.6661120354963947, + "grad_norm": 3.7365262508392334, + "learning_rate": 2.4732510288065844e-05, + "loss": 2.0378, + "step": 751 + }, + { + "epoch": 1.6683305601774818, + "grad_norm": 3.8020172119140625, + "learning_rate": 2.4691358024691357e-05, + "loss": 1.921, + "step": 752 + }, + { + "epoch": 1.670549084858569, + "grad_norm": 4.271883487701416, + "learning_rate": 2.4650205761316876e-05, + "loss": 2.1053, + "step": 753 + }, + { + "epoch": 1.672767609539656, + "grad_norm": 3.9889700412750244, + "learning_rate": 2.4609053497942388e-05, + "loss": 2.2561, + "step": 754 + }, + { + "epoch": 1.674986134220743, + "grad_norm": 3.9541916847229004, + "learning_rate": 2.4567901234567904e-05, + "loss": 2.4099, + "step": 755 + }, + { + "epoch": 1.6772046589018301, + "grad_norm": 3.5310137271881104, + "learning_rate": 2.4526748971193417e-05, + "loss": 2.4493, + "step": 756 + }, + { + "epoch": 1.6794231835829172, + "grad_norm": 3.7283899784088135, + "learning_rate": 2.4485596707818932e-05, + "loss": 2.2045, + "step": 757 + }, + { + "epoch": 1.6816417082640043, + "grad_norm": 4.096230506896973, + "learning_rate": 2.4444444444444445e-05, + "loss": 2.1767, + "step": 758 + }, + { + "epoch": 1.6838602329450914, + "grad_norm": 3.7775094509124756, + "learning_rate": 2.4403292181069957e-05, + "loss": 1.7769, + "step": 759 + }, + { + "epoch": 1.6860787576261784, + "grad_norm": 4.295887470245361, + "learning_rate": 2.4362139917695477e-05, + "loss": 2.0978, + "step": 760 + }, + { + "epoch": 1.6882972823072657, + "grad_norm": 3.7985684871673584, + "learning_rate": 2.432098765432099e-05, + "loss": 2.2047, + "step": 761 + }, + { + "epoch": 1.6905158069883528, + "grad_norm": 4.202483654022217, + "learning_rate": 2.4279835390946505e-05, + "loss": 1.9678, + "step": 762 + }, + { + "epoch": 1.6927343316694399, + "grad_norm": 4.425576686859131, + "learning_rate": 2.4238683127572017e-05, + "loss": 2.346, + "step": 763 + }, + { + "epoch": 1.694952856350527, + "grad_norm": 3.612640857696533, + "learning_rate": 2.4197530864197533e-05, + "loss": 2.1082, + "step": 764 + }, + { + "epoch": 1.697171381031614, + "grad_norm": 4.449995994567871, + "learning_rate": 2.4156378600823046e-05, + "loss": 2.2173, + "step": 765 + }, + { + "epoch": 1.699389905712701, + "grad_norm": 3.98574161529541, + "learning_rate": 2.4115226337448558e-05, + "loss": 2.0675, + "step": 766 + }, + { + "epoch": 1.7016084303937882, + "grad_norm": 4.2831501960754395, + "learning_rate": 2.4074074074074074e-05, + "loss": 2.2046, + "step": 767 + }, + { + "epoch": 1.7038269550748752, + "grad_norm": 3.977893829345703, + "learning_rate": 2.403292181069959e-05, + "loss": 2.1263, + "step": 768 + }, + { + "epoch": 1.7060454797559623, + "grad_norm": 4.289726257324219, + "learning_rate": 2.3991769547325106e-05, + "loss": 2.0441, + "step": 769 + }, + { + "epoch": 1.7082640044370494, + "grad_norm": 3.718662977218628, + "learning_rate": 2.3950617283950618e-05, + "loss": 2.2174, + "step": 770 + }, + { + "epoch": 1.7104825291181365, + "grad_norm": 3.5962371826171875, + "learning_rate": 2.3909465020576134e-05, + "loss": 2.223, + "step": 771 + }, + { + "epoch": 1.7127010537992235, + "grad_norm": 3.850080966949463, + "learning_rate": 2.3868312757201647e-05, + "loss": 1.9881, + "step": 772 + }, + { + "epoch": 1.7149195784803106, + "grad_norm": 3.8205456733703613, + "learning_rate": 2.3827160493827162e-05, + "loss": 2.1855, + "step": 773 + }, + { + "epoch": 1.7171381031613977, + "grad_norm": 3.4509408473968506, + "learning_rate": 2.3786008230452675e-05, + "loss": 2.1695, + "step": 774 + }, + { + "epoch": 1.7193566278424848, + "grad_norm": 4.670713901519775, + "learning_rate": 2.374485596707819e-05, + "loss": 2.4202, + "step": 775 + }, + { + "epoch": 1.7215751525235718, + "grad_norm": 3.682112693786621, + "learning_rate": 2.3703703703703707e-05, + "loss": 2.0528, + "step": 776 + }, + { + "epoch": 1.723793677204659, + "grad_norm": 3.5032169818878174, + "learning_rate": 2.366255144032922e-05, + "loss": 2.2287, + "step": 777 + }, + { + "epoch": 1.726012201885746, + "grad_norm": 3.9714601039886475, + "learning_rate": 2.3621399176954735e-05, + "loss": 2.3093, + "step": 778 + }, + { + "epoch": 1.728230726566833, + "grad_norm": 4.03189754486084, + "learning_rate": 2.3580246913580247e-05, + "loss": 2.393, + "step": 779 + }, + { + "epoch": 1.7304492512479202, + "grad_norm": 3.770035982131958, + "learning_rate": 2.3539094650205763e-05, + "loss": 2.1898, + "step": 780 + }, + { + "epoch": 1.7326677759290072, + "grad_norm": 3.8779125213623047, + "learning_rate": 2.3497942386831276e-05, + "loss": 2.0066, + "step": 781 + }, + { + "epoch": 1.7348863006100943, + "grad_norm": 3.9373600482940674, + "learning_rate": 2.345679012345679e-05, + "loss": 2.0249, + "step": 782 + }, + { + "epoch": 1.7371048252911814, + "grad_norm": 4.098447322845459, + "learning_rate": 2.3415637860082307e-05, + "loss": 2.1643, + "step": 783 + }, + { + "epoch": 1.7393233499722685, + "grad_norm": 4.4037394523620605, + "learning_rate": 2.337448559670782e-05, + "loss": 2.2192, + "step": 784 + }, + { + "epoch": 1.7415418746533555, + "grad_norm": 3.7593016624450684, + "learning_rate": 2.3333333333333336e-05, + "loss": 2.0716, + "step": 785 + }, + { + "epoch": 1.7437603993344426, + "grad_norm": 3.7431671619415283, + "learning_rate": 2.3292181069958848e-05, + "loss": 2.0171, + "step": 786 + }, + { + "epoch": 1.7459789240155297, + "grad_norm": 4.167656898498535, + "learning_rate": 2.3251028806584364e-05, + "loss": 2.0369, + "step": 787 + }, + { + "epoch": 1.7481974486966168, + "grad_norm": 4.351454257965088, + "learning_rate": 2.3209876543209877e-05, + "loss": 2.5613, + "step": 788 + }, + { + "epoch": 1.7504159733777038, + "grad_norm": 4.611893653869629, + "learning_rate": 2.3168724279835392e-05, + "loss": 1.9324, + "step": 789 + }, + { + "epoch": 1.752634498058791, + "grad_norm": 4.200326919555664, + "learning_rate": 2.312757201646091e-05, + "loss": 1.9861, + "step": 790 + }, + { + "epoch": 1.754853022739878, + "grad_norm": 4.2899651527404785, + "learning_rate": 2.308641975308642e-05, + "loss": 2.1611, + "step": 791 + }, + { + "epoch": 1.757071547420965, + "grad_norm": 4.159173488616943, + "learning_rate": 2.3045267489711937e-05, + "loss": 1.8266, + "step": 792 + }, + { + "epoch": 1.7592900721020521, + "grad_norm": 4.1094441413879395, + "learning_rate": 2.300411522633745e-05, + "loss": 1.8643, + "step": 793 + }, + { + "epoch": 1.7615085967831392, + "grad_norm": 4.472156524658203, + "learning_rate": 2.2962962962962965e-05, + "loss": 2.1325, + "step": 794 + }, + { + "epoch": 1.7637271214642263, + "grad_norm": 3.8675262928009033, + "learning_rate": 2.2921810699588477e-05, + "loss": 2.1682, + "step": 795 + }, + { + "epoch": 1.7659456461453134, + "grad_norm": 3.670234203338623, + "learning_rate": 2.2880658436213993e-05, + "loss": 2.1977, + "step": 796 + }, + { + "epoch": 1.7681641708264004, + "grad_norm": 4.147541046142578, + "learning_rate": 2.2839506172839506e-05, + "loss": 2.2869, + "step": 797 + }, + { + "epoch": 1.7703826955074875, + "grad_norm": 3.9078149795532227, + "learning_rate": 2.279835390946502e-05, + "loss": 2.292, + "step": 798 + }, + { + "epoch": 1.7726012201885746, + "grad_norm": 3.842430591583252, + "learning_rate": 2.2757201646090537e-05, + "loss": 1.9979, + "step": 799 + }, + { + "epoch": 1.7748197448696617, + "grad_norm": 3.866549015045166, + "learning_rate": 2.271604938271605e-05, + "loss": 2.1244, + "step": 800 + }, + { + "epoch": 1.7770382695507487, + "grad_norm": 3.99165940284729, + "learning_rate": 2.2674897119341566e-05, + "loss": 2.457, + "step": 801 + }, + { + "epoch": 1.7792567942318358, + "grad_norm": 4.039896488189697, + "learning_rate": 2.2633744855967078e-05, + "loss": 2.1683, + "step": 802 + }, + { + "epoch": 1.7814753189129229, + "grad_norm": 4.076860427856445, + "learning_rate": 2.2592592592592594e-05, + "loss": 2.1669, + "step": 803 + }, + { + "epoch": 1.78369384359401, + "grad_norm": 4.247653961181641, + "learning_rate": 2.2551440329218107e-05, + "loss": 2.206, + "step": 804 + }, + { + "epoch": 1.785912368275097, + "grad_norm": 4.260853290557861, + "learning_rate": 2.2510288065843622e-05, + "loss": 1.9445, + "step": 805 + }, + { + "epoch": 1.788130892956184, + "grad_norm": 3.9119644165039062, + "learning_rate": 2.246913580246914e-05, + "loss": 2.0339, + "step": 806 + }, + { + "epoch": 1.7903494176372712, + "grad_norm": 4.534612655639648, + "learning_rate": 2.242798353909465e-05, + "loss": 1.9204, + "step": 807 + }, + { + "epoch": 1.7925679423183583, + "grad_norm": 3.9563369750976562, + "learning_rate": 2.2386831275720167e-05, + "loss": 1.9938, + "step": 808 + }, + { + "epoch": 1.7947864669994453, + "grad_norm": 3.941863536834717, + "learning_rate": 2.234567901234568e-05, + "loss": 1.7996, + "step": 809 + }, + { + "epoch": 1.7970049916805324, + "grad_norm": 4.154852390289307, + "learning_rate": 2.2304526748971195e-05, + "loss": 2.0611, + "step": 810 + }, + { + "epoch": 1.7992235163616195, + "grad_norm": 4.835111618041992, + "learning_rate": 2.2263374485596707e-05, + "loss": 2.2442, + "step": 811 + }, + { + "epoch": 1.8014420410427066, + "grad_norm": 4.478168487548828, + "learning_rate": 2.2222222222222223e-05, + "loss": 2.1374, + "step": 812 + }, + { + "epoch": 1.8036605657237936, + "grad_norm": 3.561108350753784, + "learning_rate": 2.218106995884774e-05, + "loss": 2.0443, + "step": 813 + }, + { + "epoch": 1.8058790904048807, + "grad_norm": 3.8302249908447266, + "learning_rate": 2.213991769547325e-05, + "loss": 2.0072, + "step": 814 + }, + { + "epoch": 1.8080976150859678, + "grad_norm": 4.089479923248291, + "learning_rate": 2.2098765432098767e-05, + "loss": 2.0104, + "step": 815 + }, + { + "epoch": 1.8103161397670549, + "grad_norm": 3.7940220832824707, + "learning_rate": 2.205761316872428e-05, + "loss": 2.5663, + "step": 816 + }, + { + "epoch": 1.8125346644481422, + "grad_norm": 3.785043239593506, + "learning_rate": 2.2016460905349796e-05, + "loss": 2.1543, + "step": 817 + }, + { + "epoch": 1.8147531891292292, + "grad_norm": 3.2984659671783447, + "learning_rate": 2.1975308641975308e-05, + "loss": 2.047, + "step": 818 + }, + { + "epoch": 1.8169717138103163, + "grad_norm": 3.716160774230957, + "learning_rate": 2.1934156378600824e-05, + "loss": 1.8045, + "step": 819 + }, + { + "epoch": 1.8191902384914034, + "grad_norm": 4.3379435539245605, + "learning_rate": 2.189300411522634e-05, + "loss": 2.058, + "step": 820 + }, + { + "epoch": 1.8214087631724905, + "grad_norm": 3.8923938274383545, + "learning_rate": 2.1851851851851852e-05, + "loss": 1.9024, + "step": 821 + }, + { + "epoch": 1.8236272878535775, + "grad_norm": 3.9507992267608643, + "learning_rate": 2.1810699588477368e-05, + "loss": 2.0549, + "step": 822 + }, + { + "epoch": 1.8258458125346646, + "grad_norm": 3.9769158363342285, + "learning_rate": 2.176954732510288e-05, + "loss": 2.1329, + "step": 823 + }, + { + "epoch": 1.8280643372157517, + "grad_norm": 3.8494298458099365, + "learning_rate": 2.1728395061728397e-05, + "loss": 1.9619, + "step": 824 + }, + { + "epoch": 1.8302828618968388, + "grad_norm": 4.016391754150391, + "learning_rate": 2.168724279835391e-05, + "loss": 2.3344, + "step": 825 + }, + { + "epoch": 1.8325013865779258, + "grad_norm": 4.01151704788208, + "learning_rate": 2.1646090534979425e-05, + "loss": 2.119, + "step": 826 + }, + { + "epoch": 1.834719911259013, + "grad_norm": 3.8882858753204346, + "learning_rate": 2.1604938271604937e-05, + "loss": 2.3913, + "step": 827 + }, + { + "epoch": 1.8369384359401, + "grad_norm": 4.486508846282959, + "learning_rate": 2.1563786008230453e-05, + "loss": 2.1034, + "step": 828 + }, + { + "epoch": 1.839156960621187, + "grad_norm": 3.843597173690796, + "learning_rate": 2.152263374485597e-05, + "loss": 2.1329, + "step": 829 + }, + { + "epoch": 1.8413754853022741, + "grad_norm": 4.164466857910156, + "learning_rate": 2.148148148148148e-05, + "loss": 1.9764, + "step": 830 + }, + { + "epoch": 1.8435940099833612, + "grad_norm": 4.310817241668701, + "learning_rate": 2.1440329218106997e-05, + "loss": 2.1312, + "step": 831 + }, + { + "epoch": 1.8458125346644483, + "grad_norm": 3.9981887340545654, + "learning_rate": 2.139917695473251e-05, + "loss": 2.3264, + "step": 832 + }, + { + "epoch": 1.8480310593455354, + "grad_norm": 4.231433868408203, + "learning_rate": 2.1358024691358026e-05, + "loss": 2.0971, + "step": 833 + }, + { + "epoch": 1.8502495840266224, + "grad_norm": 3.7736361026763916, + "learning_rate": 2.1316872427983538e-05, + "loss": 1.9666, + "step": 834 + }, + { + "epoch": 1.8524681087077095, + "grad_norm": 4.19193696975708, + "learning_rate": 2.1275720164609054e-05, + "loss": 2.1026, + "step": 835 + }, + { + "epoch": 1.8546866333887966, + "grad_norm": 4.223895072937012, + "learning_rate": 2.123456790123457e-05, + "loss": 2.4588, + "step": 836 + }, + { + "epoch": 1.8569051580698837, + "grad_norm": 4.202375411987305, + "learning_rate": 2.1193415637860082e-05, + "loss": 2.099, + "step": 837 + }, + { + "epoch": 1.8591236827509707, + "grad_norm": 4.018009662628174, + "learning_rate": 2.1152263374485598e-05, + "loss": 1.9981, + "step": 838 + }, + { + "epoch": 1.8613422074320578, + "grad_norm": 4.160388469696045, + "learning_rate": 2.111111111111111e-05, + "loss": 1.9207, + "step": 839 + }, + { + "epoch": 1.8635607321131449, + "grad_norm": 3.643157482147217, + "learning_rate": 2.1069958847736627e-05, + "loss": 2.0207, + "step": 840 + }, + { + "epoch": 1.865779256794232, + "grad_norm": 3.6870994567871094, + "learning_rate": 2.102880658436214e-05, + "loss": 2.2176, + "step": 841 + }, + { + "epoch": 1.867997781475319, + "grad_norm": 3.5719969272613525, + "learning_rate": 2.0987654320987655e-05, + "loss": 2.2148, + "step": 842 + }, + { + "epoch": 1.870216306156406, + "grad_norm": 3.8579535484313965, + "learning_rate": 2.094650205761317e-05, + "loss": 1.8624, + "step": 843 + }, + { + "epoch": 1.8724348308374932, + "grad_norm": 4.128652572631836, + "learning_rate": 2.0905349794238687e-05, + "loss": 2.2909, + "step": 844 + }, + { + "epoch": 1.8746533555185803, + "grad_norm": 3.8186028003692627, + "learning_rate": 2.08641975308642e-05, + "loss": 2.3448, + "step": 845 + }, + { + "epoch": 1.8768718801996673, + "grad_norm": 3.622645139694214, + "learning_rate": 2.082304526748971e-05, + "loss": 2.0601, + "step": 846 + }, + { + "epoch": 1.8790904048807544, + "grad_norm": 3.5154318809509277, + "learning_rate": 2.0781893004115227e-05, + "loss": 2.0305, + "step": 847 + }, + { + "epoch": 1.8813089295618415, + "grad_norm": 3.8251326084136963, + "learning_rate": 2.074074074074074e-05, + "loss": 1.799, + "step": 848 + }, + { + "epoch": 1.8835274542429286, + "grad_norm": 3.9057698249816895, + "learning_rate": 2.0699588477366256e-05, + "loss": 2.3157, + "step": 849 + }, + { + "epoch": 1.8857459789240156, + "grad_norm": 4.118304252624512, + "learning_rate": 2.0658436213991768e-05, + "loss": 2.2835, + "step": 850 + }, + { + "epoch": 1.8879645036051027, + "grad_norm": 3.83282732963562, + "learning_rate": 2.0617283950617287e-05, + "loss": 2.1301, + "step": 851 + }, + { + "epoch": 1.8901830282861898, + "grad_norm": 3.5063726902008057, + "learning_rate": 2.05761316872428e-05, + "loss": 1.9577, + "step": 852 + }, + { + "epoch": 1.8924015529672769, + "grad_norm": 3.852447986602783, + "learning_rate": 2.0534979423868312e-05, + "loss": 2.152, + "step": 853 + }, + { + "epoch": 1.894620077648364, + "grad_norm": 4.305698394775391, + "learning_rate": 2.0493827160493828e-05, + "loss": 2.2891, + "step": 854 + }, + { + "epoch": 1.896838602329451, + "grad_norm": 3.4106321334838867, + "learning_rate": 2.045267489711934e-05, + "loss": 2.4543, + "step": 855 + }, + { + "epoch": 1.899057127010538, + "grad_norm": 3.7693469524383545, + "learning_rate": 2.0411522633744857e-05, + "loss": 2.1735, + "step": 856 + }, + { + "epoch": 1.9012756516916252, + "grad_norm": 3.9880661964416504, + "learning_rate": 2.037037037037037e-05, + "loss": 1.8565, + "step": 857 + }, + { + "epoch": 1.9034941763727122, + "grad_norm": 4.0608062744140625, + "learning_rate": 2.032921810699589e-05, + "loss": 1.9797, + "step": 858 + }, + { + "epoch": 1.9057127010537993, + "grad_norm": 3.865671157836914, + "learning_rate": 2.02880658436214e-05, + "loss": 2.3721, + "step": 859 + }, + { + "epoch": 1.9079312257348864, + "grad_norm": 4.145225524902344, + "learning_rate": 2.0246913580246917e-05, + "loss": 2.0478, + "step": 860 + }, + { + "epoch": 1.9101497504159735, + "grad_norm": 3.9357974529266357, + "learning_rate": 2.020576131687243e-05, + "loss": 2.4907, + "step": 861 + }, + { + "epoch": 1.9123682750970605, + "grad_norm": 3.9337215423583984, + "learning_rate": 2.016460905349794e-05, + "loss": 1.7458, + "step": 862 + }, + { + "epoch": 1.9145867997781476, + "grad_norm": 4.107207298278809, + "learning_rate": 2.0123456790123457e-05, + "loss": 1.8753, + "step": 863 + }, + { + "epoch": 1.9168053244592347, + "grad_norm": 4.043063163757324, + "learning_rate": 2.008230452674897e-05, + "loss": 2.0405, + "step": 864 + }, + { + "epoch": 1.9190238491403218, + "grad_norm": 3.8558523654937744, + "learning_rate": 2.0041152263374486e-05, + "loss": 2.1256, + "step": 865 + }, + { + "epoch": 1.9212423738214088, + "grad_norm": 4.55283260345459, + "learning_rate": 2e-05, + "loss": 2.1814, + "step": 866 + }, + { + "epoch": 1.923460898502496, + "grad_norm": 4.389678478240967, + "learning_rate": 1.9958847736625517e-05, + "loss": 1.714, + "step": 867 + }, + { + "epoch": 1.925679423183583, + "grad_norm": 4.455432415008545, + "learning_rate": 1.991769547325103e-05, + "loss": 2.4414, + "step": 868 + }, + { + "epoch": 1.92789794786467, + "grad_norm": 4.32024621963501, + "learning_rate": 1.9876543209876546e-05, + "loss": 2.1817, + "step": 869 + }, + { + "epoch": 1.9301164725457571, + "grad_norm": 3.663665771484375, + "learning_rate": 1.9835390946502058e-05, + "loss": 2.405, + "step": 870 + }, + { + "epoch": 1.9323349972268442, + "grad_norm": 4.360503196716309, + "learning_rate": 1.979423868312757e-05, + "loss": 2.0216, + "step": 871 + }, + { + "epoch": 1.9345535219079313, + "grad_norm": 4.406967639923096, + "learning_rate": 1.9753086419753087e-05, + "loss": 1.948, + "step": 872 + }, + { + "epoch": 1.9367720465890184, + "grad_norm": 4.175272464752197, + "learning_rate": 1.9711934156378602e-05, + "loss": 1.9618, + "step": 873 + }, + { + "epoch": 1.9389905712701054, + "grad_norm": 3.762576103210449, + "learning_rate": 1.967078189300412e-05, + "loss": 2.201, + "step": 874 + }, + { + "epoch": 1.9412090959511925, + "grad_norm": 3.8000175952911377, + "learning_rate": 1.962962962962963e-05, + "loss": 2.3686, + "step": 875 + }, + { + "epoch": 1.9434276206322796, + "grad_norm": 3.8225371837615967, + "learning_rate": 1.9588477366255147e-05, + "loss": 1.8603, + "step": 876 + }, + { + "epoch": 1.9456461453133667, + "grad_norm": 3.9086532592773438, + "learning_rate": 1.954732510288066e-05, + "loss": 1.9866, + "step": 877 + }, + { + "epoch": 1.9478646699944537, + "grad_norm": 3.963346242904663, + "learning_rate": 1.950617283950617e-05, + "loss": 1.8021, + "step": 878 + }, + { + "epoch": 1.9500831946755408, + "grad_norm": 4.227272987365723, + "learning_rate": 1.9465020576131687e-05, + "loss": 2.1856, + "step": 879 + }, + { + "epoch": 1.9523017193566279, + "grad_norm": 3.643037796020508, + "learning_rate": 1.94238683127572e-05, + "loss": 2.0394, + "step": 880 + }, + { + "epoch": 1.954520244037715, + "grad_norm": 3.7113544940948486, + "learning_rate": 1.938271604938272e-05, + "loss": 1.8597, + "step": 881 + }, + { + "epoch": 1.956738768718802, + "grad_norm": 3.7593636512756348, + "learning_rate": 1.934156378600823e-05, + "loss": 2.2038, + "step": 882 + }, + { + "epoch": 1.958957293399889, + "grad_norm": 4.492856502532959, + "learning_rate": 1.9300411522633747e-05, + "loss": 2.3998, + "step": 883 + }, + { + "epoch": 1.9611758180809762, + "grad_norm": 4.5002007484436035, + "learning_rate": 1.925925925925926e-05, + "loss": 2.0954, + "step": 884 + }, + { + "epoch": 1.9633943427620633, + "grad_norm": 3.941837787628174, + "learning_rate": 1.9218106995884776e-05, + "loss": 1.7954, + "step": 885 + }, + { + "epoch": 1.9656128674431503, + "grad_norm": 3.667760133743286, + "learning_rate": 1.9176954732510288e-05, + "loss": 1.8994, + "step": 886 + }, + { + "epoch": 1.9678313921242374, + "grad_norm": 3.961977481842041, + "learning_rate": 1.91358024691358e-05, + "loss": 2.416, + "step": 887 + }, + { + "epoch": 1.9700499168053245, + "grad_norm": 4.570499420166016, + "learning_rate": 1.909465020576132e-05, + "loss": 2.1653, + "step": 888 + }, + { + "epoch": 1.9722684414864116, + "grad_norm": 4.418700695037842, + "learning_rate": 1.9053497942386832e-05, + "loss": 2.0527, + "step": 889 + }, + { + "epoch": 1.9744869661674986, + "grad_norm": 3.7750182151794434, + "learning_rate": 1.901234567901235e-05, + "loss": 1.7754, + "step": 890 + }, + { + "epoch": 1.9767054908485857, + "grad_norm": 4.47550106048584, + "learning_rate": 1.897119341563786e-05, + "loss": 2.1947, + "step": 891 + }, + { + "epoch": 1.9789240155296728, + "grad_norm": 4.332265853881836, + "learning_rate": 1.8930041152263377e-05, + "loss": 1.9424, + "step": 892 + }, + { + "epoch": 1.9811425402107599, + "grad_norm": 4.29379940032959, + "learning_rate": 1.888888888888889e-05, + "loss": 2.0242, + "step": 893 + }, + { + "epoch": 1.983361064891847, + "grad_norm": 4.529409408569336, + "learning_rate": 1.8847736625514405e-05, + "loss": 2.2522, + "step": 894 + }, + { + "epoch": 1.985579589572934, + "grad_norm": 4.140018463134766, + "learning_rate": 1.8806584362139917e-05, + "loss": 2.2706, + "step": 895 + }, + { + "epoch": 1.987798114254021, + "grad_norm": 4.3104376792907715, + "learning_rate": 1.8765432098765433e-05, + "loss": 1.9383, + "step": 896 + }, + { + "epoch": 1.9900166389351082, + "grad_norm": 3.9684972763061523, + "learning_rate": 1.872427983539095e-05, + "loss": 1.7615, + "step": 897 + }, + { + "epoch": 1.9922351636161952, + "grad_norm": 4.0687408447265625, + "learning_rate": 1.868312757201646e-05, + "loss": 1.6791, + "step": 898 + }, + { + "epoch": 1.9944536882972823, + "grad_norm": 3.772174119949341, + "learning_rate": 1.8641975308641977e-05, + "loss": 2.0007, + "step": 899 + }, + { + "epoch": 1.9966722129783694, + "grad_norm": 3.9703104496002197, + "learning_rate": 1.860082304526749e-05, + "loss": 2.1051, + "step": 900 + }, + { + "epoch": 1.9988907376594565, + "grad_norm": 4.01883602142334, + "learning_rate": 1.8559670781893006e-05, + "loss": 2.0558, + "step": 901 + }, + { + "epoch": 2.0011092623405435, + "grad_norm": 3.3110318183898926, + "learning_rate": 1.8518518518518518e-05, + "loss": 1.7161, + "step": 902 + }, + { + "epoch": 2.0033277870216306, + "grad_norm": 3.125894069671631, + "learning_rate": 1.8477366255144034e-05, + "loss": 1.3673, + "step": 903 + }, + { + "epoch": 2.0055463117027177, + "grad_norm": 3.710397243499756, + "learning_rate": 1.843621399176955e-05, + "loss": 1.4557, + "step": 904 + }, + { + "epoch": 2.0077648363838048, + "grad_norm": 3.3064932823181152, + "learning_rate": 1.8395061728395062e-05, + "loss": 1.6007, + "step": 905 + }, + { + "epoch": 2.009983361064892, + "grad_norm": 4.337411403656006, + "learning_rate": 1.835390946502058e-05, + "loss": 1.2201, + "step": 906 + }, + { + "epoch": 2.012201885745979, + "grad_norm": 3.305535078048706, + "learning_rate": 1.831275720164609e-05, + "loss": 1.4731, + "step": 907 + }, + { + "epoch": 2.014420410427066, + "grad_norm": 4.008023738861084, + "learning_rate": 1.8271604938271607e-05, + "loss": 1.6144, + "step": 908 + }, + { + "epoch": 2.016638935108153, + "grad_norm": 3.83902907371521, + "learning_rate": 1.823045267489712e-05, + "loss": 1.4634, + "step": 909 + }, + { + "epoch": 2.01885745978924, + "grad_norm": 3.9539198875427246, + "learning_rate": 1.8189300411522635e-05, + "loss": 1.3503, + "step": 910 + }, + { + "epoch": 2.021075984470327, + "grad_norm": 3.8504693508148193, + "learning_rate": 1.814814814814815e-05, + "loss": 1.5008, + "step": 911 + }, + { + "epoch": 2.0232945091514143, + "grad_norm": 3.8806371688842773, + "learning_rate": 1.8106995884773663e-05, + "loss": 1.2818, + "step": 912 + }, + { + "epoch": 2.0255130338325014, + "grad_norm": 3.455753803253174, + "learning_rate": 1.806584362139918e-05, + "loss": 1.4619, + "step": 913 + }, + { + "epoch": 2.0277315585135884, + "grad_norm": 3.822279930114746, + "learning_rate": 1.802469135802469e-05, + "loss": 1.548, + "step": 914 + }, + { + "epoch": 2.0299500831946755, + "grad_norm": 4.107574939727783, + "learning_rate": 1.7983539094650207e-05, + "loss": 1.196, + "step": 915 + }, + { + "epoch": 2.0321686078757626, + "grad_norm": 3.8645405769348145, + "learning_rate": 1.794238683127572e-05, + "loss": 1.5128, + "step": 916 + }, + { + "epoch": 2.0343871325568497, + "grad_norm": 4.198647499084473, + "learning_rate": 1.7901234567901236e-05, + "loss": 1.323, + "step": 917 + }, + { + "epoch": 2.0366056572379367, + "grad_norm": 3.868516445159912, + "learning_rate": 1.7860082304526748e-05, + "loss": 1.2234, + "step": 918 + }, + { + "epoch": 2.038824181919024, + "grad_norm": 3.8612561225891113, + "learning_rate": 1.7818930041152264e-05, + "loss": 1.5588, + "step": 919 + }, + { + "epoch": 2.041042706600111, + "grad_norm": 3.9263014793395996, + "learning_rate": 1.777777777777778e-05, + "loss": 1.5178, + "step": 920 + }, + { + "epoch": 2.043261231281198, + "grad_norm": 3.5290560722351074, + "learning_rate": 1.7736625514403292e-05, + "loss": 1.4466, + "step": 921 + }, + { + "epoch": 2.045479755962285, + "grad_norm": 3.7670340538024902, + "learning_rate": 1.769547325102881e-05, + "loss": 1.4843, + "step": 922 + }, + { + "epoch": 2.047698280643372, + "grad_norm": 3.7114932537078857, + "learning_rate": 1.765432098765432e-05, + "loss": 1.6685, + "step": 923 + }, + { + "epoch": 2.049916805324459, + "grad_norm": 3.547624111175537, + "learning_rate": 1.7613168724279837e-05, + "loss": 1.3898, + "step": 924 + }, + { + "epoch": 2.0521353300055463, + "grad_norm": 3.970966100692749, + "learning_rate": 1.757201646090535e-05, + "loss": 1.5984, + "step": 925 + }, + { + "epoch": 2.0543538546866333, + "grad_norm": 3.54091739654541, + "learning_rate": 1.7530864197530865e-05, + "loss": 1.6613, + "step": 926 + }, + { + "epoch": 2.0565723793677204, + "grad_norm": 3.5774805545806885, + "learning_rate": 1.748971193415638e-05, + "loss": 1.6027, + "step": 927 + }, + { + "epoch": 2.0587909040488075, + "grad_norm": 3.988870859146118, + "learning_rate": 1.7448559670781893e-05, + "loss": 1.2207, + "step": 928 + }, + { + "epoch": 2.0610094287298946, + "grad_norm": 3.857330560684204, + "learning_rate": 1.740740740740741e-05, + "loss": 1.5249, + "step": 929 + }, + { + "epoch": 2.0632279534109816, + "grad_norm": 3.8848767280578613, + "learning_rate": 1.736625514403292e-05, + "loss": 1.5781, + "step": 930 + }, + { + "epoch": 2.0654464780920687, + "grad_norm": 3.7376761436462402, + "learning_rate": 1.7325102880658437e-05, + "loss": 1.7227, + "step": 931 + }, + { + "epoch": 2.067665002773156, + "grad_norm": 3.7883505821228027, + "learning_rate": 1.728395061728395e-05, + "loss": 1.3671, + "step": 932 + }, + { + "epoch": 2.069883527454243, + "grad_norm": 3.6929028034210205, + "learning_rate": 1.7242798353909466e-05, + "loss": 1.5789, + "step": 933 + }, + { + "epoch": 2.07210205213533, + "grad_norm": 4.370468616485596, + "learning_rate": 1.720164609053498e-05, + "loss": 1.6508, + "step": 934 + }, + { + "epoch": 2.074320576816417, + "grad_norm": 3.76127552986145, + "learning_rate": 1.7160493827160494e-05, + "loss": 1.2667, + "step": 935 + }, + { + "epoch": 2.076539101497504, + "grad_norm": 3.896806240081787, + "learning_rate": 1.711934156378601e-05, + "loss": 1.5758, + "step": 936 + }, + { + "epoch": 2.078757626178591, + "grad_norm": 3.4717047214508057, + "learning_rate": 1.7078189300411522e-05, + "loss": 1.4783, + "step": 937 + }, + { + "epoch": 2.0809761508596782, + "grad_norm": 3.5015692710876465, + "learning_rate": 1.7037037037037038e-05, + "loss": 1.3396, + "step": 938 + }, + { + "epoch": 2.0831946755407653, + "grad_norm": 3.873945474624634, + "learning_rate": 1.699588477366255e-05, + "loss": 1.2725, + "step": 939 + }, + { + "epoch": 2.0854132002218524, + "grad_norm": 3.7489960193634033, + "learning_rate": 1.6954732510288067e-05, + "loss": 1.5363, + "step": 940 + }, + { + "epoch": 2.0876317249029395, + "grad_norm": 4.113027572631836, + "learning_rate": 1.6913580246913582e-05, + "loss": 1.5029, + "step": 941 + }, + { + "epoch": 2.0898502495840265, + "grad_norm": 3.769381284713745, + "learning_rate": 1.6872427983539095e-05, + "loss": 1.4013, + "step": 942 + }, + { + "epoch": 2.0920687742651136, + "grad_norm": 4.095351219177246, + "learning_rate": 1.683127572016461e-05, + "loss": 1.3234, + "step": 943 + }, + { + "epoch": 2.0942872989462007, + "grad_norm": 4.020770072937012, + "learning_rate": 1.6790123456790123e-05, + "loss": 1.2809, + "step": 944 + }, + { + "epoch": 2.0965058236272878, + "grad_norm": 3.8478474617004395, + "learning_rate": 1.674897119341564e-05, + "loss": 1.8359, + "step": 945 + }, + { + "epoch": 2.098724348308375, + "grad_norm": 4.093289375305176, + "learning_rate": 1.670781893004115e-05, + "loss": 1.2023, + "step": 946 + }, + { + "epoch": 2.100942872989462, + "grad_norm": 3.7606143951416016, + "learning_rate": 1.6666666666666667e-05, + "loss": 1.6114, + "step": 947 + }, + { + "epoch": 2.103161397670549, + "grad_norm": 3.884422779083252, + "learning_rate": 1.662551440329218e-05, + "loss": 1.3754, + "step": 948 + }, + { + "epoch": 2.105379922351636, + "grad_norm": 4.038058757781982, + "learning_rate": 1.6584362139917696e-05, + "loss": 1.6409, + "step": 949 + }, + { + "epoch": 2.107598447032723, + "grad_norm": 3.9373056888580322, + "learning_rate": 1.654320987654321e-05, + "loss": 1.3775, + "step": 950 + }, + { + "epoch": 2.10981697171381, + "grad_norm": 3.717966318130493, + "learning_rate": 1.6502057613168724e-05, + "loss": 1.3088, + "step": 951 + }, + { + "epoch": 2.1120354963948973, + "grad_norm": 3.4117913246154785, + "learning_rate": 1.646090534979424e-05, + "loss": 1.4197, + "step": 952 + }, + { + "epoch": 2.1142540210759844, + "grad_norm": 4.042545795440674, + "learning_rate": 1.6419753086419752e-05, + "loss": 1.5152, + "step": 953 + }, + { + "epoch": 2.1164725457570714, + "grad_norm": 3.6911721229553223, + "learning_rate": 1.6378600823045268e-05, + "loss": 1.5418, + "step": 954 + }, + { + "epoch": 2.1186910704381585, + "grad_norm": 3.7374074459075928, + "learning_rate": 1.633744855967078e-05, + "loss": 1.2586, + "step": 955 + }, + { + "epoch": 2.1209095951192456, + "grad_norm": 3.812371015548706, + "learning_rate": 1.62962962962963e-05, + "loss": 1.5757, + "step": 956 + }, + { + "epoch": 2.1231281198003327, + "grad_norm": 4.081371307373047, + "learning_rate": 1.6255144032921812e-05, + "loss": 1.7645, + "step": 957 + }, + { + "epoch": 2.1253466444814197, + "grad_norm": 3.5569422245025635, + "learning_rate": 1.6213991769547325e-05, + "loss": 1.1991, + "step": 958 + }, + { + "epoch": 2.127565169162507, + "grad_norm": 3.908228874206543, + "learning_rate": 1.617283950617284e-05, + "loss": 1.2345, + "step": 959 + }, + { + "epoch": 2.129783693843594, + "grad_norm": 4.507850170135498, + "learning_rate": 1.6131687242798353e-05, + "loss": 1.7999, + "step": 960 + }, + { + "epoch": 2.132002218524681, + "grad_norm": 4.295609474182129, + "learning_rate": 1.609053497942387e-05, + "loss": 1.3914, + "step": 961 + }, + { + "epoch": 2.134220743205768, + "grad_norm": 4.323912620544434, + "learning_rate": 1.604938271604938e-05, + "loss": 1.8399, + "step": 962 + }, + { + "epoch": 2.136439267886855, + "grad_norm": 3.9559924602508545, + "learning_rate": 1.6008230452674897e-05, + "loss": 1.5007, + "step": 963 + }, + { + "epoch": 2.138657792567942, + "grad_norm": 3.775782346725464, + "learning_rate": 1.5967078189300413e-05, + "loss": 1.3043, + "step": 964 + }, + { + "epoch": 2.1408763172490293, + "grad_norm": 3.4541194438934326, + "learning_rate": 1.5925925925925926e-05, + "loss": 1.6098, + "step": 965 + }, + { + "epoch": 2.1430948419301163, + "grad_norm": 3.854104995727539, + "learning_rate": 1.588477366255144e-05, + "loss": 1.5001, + "step": 966 + }, + { + "epoch": 2.1453133666112034, + "grad_norm": 3.7088723182678223, + "learning_rate": 1.5843621399176954e-05, + "loss": 1.5027, + "step": 967 + }, + { + "epoch": 2.1475318912922905, + "grad_norm": 3.8169057369232178, + "learning_rate": 1.580246913580247e-05, + "loss": 1.5684, + "step": 968 + }, + { + "epoch": 2.1497504159733776, + "grad_norm": 3.7996973991394043, + "learning_rate": 1.5761316872427982e-05, + "loss": 1.6324, + "step": 969 + }, + { + "epoch": 2.1519689406544646, + "grad_norm": 3.9529306888580322, + "learning_rate": 1.5720164609053498e-05, + "loss": 1.4238, + "step": 970 + }, + { + "epoch": 2.1541874653355517, + "grad_norm": 4.147080898284912, + "learning_rate": 1.5679012345679014e-05, + "loss": 1.2799, + "step": 971 + }, + { + "epoch": 2.156405990016639, + "grad_norm": 3.765204906463623, + "learning_rate": 1.563786008230453e-05, + "loss": 1.2317, + "step": 972 + }, + { + "epoch": 2.158624514697726, + "grad_norm": 4.016109943389893, + "learning_rate": 1.5596707818930042e-05, + "loss": 1.4969, + "step": 973 + }, + { + "epoch": 2.160843039378813, + "grad_norm": 3.7870800495147705, + "learning_rate": 1.5555555555555555e-05, + "loss": 1.4568, + "step": 974 + }, + { + "epoch": 2.1630615640599, + "grad_norm": 3.7394566535949707, + "learning_rate": 1.551440329218107e-05, + "loss": 1.7338, + "step": 975 + }, + { + "epoch": 2.165280088740987, + "grad_norm": 3.6263229846954346, + "learning_rate": 1.5473251028806583e-05, + "loss": 1.3949, + "step": 976 + }, + { + "epoch": 2.167498613422074, + "grad_norm": 3.6777429580688477, + "learning_rate": 1.54320987654321e-05, + "loss": 1.3758, + "step": 977 + }, + { + "epoch": 2.1697171381031612, + "grad_norm": 3.6517012119293213, + "learning_rate": 1.539094650205761e-05, + "loss": 1.2905, + "step": 978 + }, + { + "epoch": 2.1719356627842483, + "grad_norm": 3.7644996643066406, + "learning_rate": 1.534979423868313e-05, + "loss": 1.6544, + "step": 979 + }, + { + "epoch": 2.1741541874653354, + "grad_norm": 3.7326903343200684, + "learning_rate": 1.5308641975308643e-05, + "loss": 1.3312, + "step": 980 + }, + { + "epoch": 2.1763727121464225, + "grad_norm": 3.796175718307495, + "learning_rate": 1.526748971193416e-05, + "loss": 1.1931, + "step": 981 + }, + { + "epoch": 2.1785912368275095, + "grad_norm": 3.6034348011016846, + "learning_rate": 1.5226337448559672e-05, + "loss": 1.0681, + "step": 982 + }, + { + "epoch": 2.1808097615085966, + "grad_norm": 4.1171040534973145, + "learning_rate": 1.5185185185185186e-05, + "loss": 1.8076, + "step": 983 + }, + { + "epoch": 2.1830282861896837, + "grad_norm": 3.762741804122925, + "learning_rate": 1.51440329218107e-05, + "loss": 1.2419, + "step": 984 + }, + { + "epoch": 2.1852468108707708, + "grad_norm": 3.983893632888794, + "learning_rate": 1.5102880658436214e-05, + "loss": 1.4032, + "step": 985 + }, + { + "epoch": 2.187465335551858, + "grad_norm": 4.113901138305664, + "learning_rate": 1.506172839506173e-05, + "loss": 1.5931, + "step": 986 + }, + { + "epoch": 2.189683860232945, + "grad_norm": 3.5009732246398926, + "learning_rate": 1.5020576131687244e-05, + "loss": 1.0297, + "step": 987 + }, + { + "epoch": 2.191902384914032, + "grad_norm": 3.7657742500305176, + "learning_rate": 1.4979423868312758e-05, + "loss": 1.0388, + "step": 988 + }, + { + "epoch": 2.194120909595119, + "grad_norm": 3.7497000694274902, + "learning_rate": 1.4938271604938272e-05, + "loss": 1.3122, + "step": 989 + }, + { + "epoch": 2.196339434276206, + "grad_norm": 3.6063804626464844, + "learning_rate": 1.4897119341563787e-05, + "loss": 1.2323, + "step": 990 + }, + { + "epoch": 2.1985579589572932, + "grad_norm": 4.0425262451171875, + "learning_rate": 1.48559670781893e-05, + "loss": 1.2266, + "step": 991 + }, + { + "epoch": 2.2007764836383803, + "grad_norm": 4.108901500701904, + "learning_rate": 1.4814814814814815e-05, + "loss": 1.4692, + "step": 992 + }, + { + "epoch": 2.2029950083194674, + "grad_norm": 4.147270202636719, + "learning_rate": 1.4773662551440329e-05, + "loss": 1.915, + "step": 993 + }, + { + "epoch": 2.2052135330005544, + "grad_norm": 4.131346225738525, + "learning_rate": 1.4732510288065845e-05, + "loss": 1.2982, + "step": 994 + }, + { + "epoch": 2.2074320576816415, + "grad_norm": 3.3308260440826416, + "learning_rate": 1.4691358024691359e-05, + "loss": 1.3111, + "step": 995 + }, + { + "epoch": 2.2096505823627286, + "grad_norm": 4.064446926116943, + "learning_rate": 1.4650205761316873e-05, + "loss": 1.4651, + "step": 996 + }, + { + "epoch": 2.2118691070438157, + "grad_norm": 3.8105697631835938, + "learning_rate": 1.4609053497942387e-05, + "loss": 1.3954, + "step": 997 + }, + { + "epoch": 2.2140876317249027, + "grad_norm": 3.992196798324585, + "learning_rate": 1.4567901234567902e-05, + "loss": 1.4688, + "step": 998 + }, + { + "epoch": 2.21630615640599, + "grad_norm": 3.73740553855896, + "learning_rate": 1.4526748971193416e-05, + "loss": 0.9741, + "step": 999 + }, + { + "epoch": 2.218524681087077, + "grad_norm": 3.579698324203491, + "learning_rate": 1.448559670781893e-05, + "loss": 1.2406, + "step": 1000 + }, + { + "epoch": 2.220743205768164, + "grad_norm": 4.043013095855713, + "learning_rate": 1.4444444444444444e-05, + "loss": 1.1739, + "step": 1001 + }, + { + "epoch": 2.222961730449251, + "grad_norm": 4.018491268157959, + "learning_rate": 1.440329218106996e-05, + "loss": 1.5736, + "step": 1002 + }, + { + "epoch": 2.225180255130338, + "grad_norm": 3.9588019847869873, + "learning_rate": 1.4362139917695474e-05, + "loss": 1.3537, + "step": 1003 + }, + { + "epoch": 2.227398779811425, + "grad_norm": 3.9010989665985107, + "learning_rate": 1.4320987654320988e-05, + "loss": 1.3429, + "step": 1004 + }, + { + "epoch": 2.2296173044925123, + "grad_norm": 3.7713825702667236, + "learning_rate": 1.4279835390946502e-05, + "loss": 1.5085, + "step": 1005 + }, + { + "epoch": 2.2318358291735994, + "grad_norm": 3.955414295196533, + "learning_rate": 1.4238683127572017e-05, + "loss": 1.5308, + "step": 1006 + }, + { + "epoch": 2.2340543538546864, + "grad_norm": 3.6655640602111816, + "learning_rate": 1.419753086419753e-05, + "loss": 1.3653, + "step": 1007 + }, + { + "epoch": 2.2362728785357735, + "grad_norm": 4.476771831512451, + "learning_rate": 1.4156378600823045e-05, + "loss": 1.6982, + "step": 1008 + }, + { + "epoch": 2.2384914032168606, + "grad_norm": 3.5810389518737793, + "learning_rate": 1.411522633744856e-05, + "loss": 1.6988, + "step": 1009 + }, + { + "epoch": 2.2407099278979477, + "grad_norm": 3.6955277919769287, + "learning_rate": 1.4074074074074075e-05, + "loss": 1.6186, + "step": 1010 + }, + { + "epoch": 2.2429284525790347, + "grad_norm": 3.8851161003112793, + "learning_rate": 1.4032921810699589e-05, + "loss": 1.3277, + "step": 1011 + }, + { + "epoch": 2.245146977260122, + "grad_norm": 3.9732959270477295, + "learning_rate": 1.3991769547325103e-05, + "loss": 1.4285, + "step": 1012 + }, + { + "epoch": 2.247365501941209, + "grad_norm": 3.7947592735290527, + "learning_rate": 1.3950617283950617e-05, + "loss": 1.2124, + "step": 1013 + }, + { + "epoch": 2.249584026622296, + "grad_norm": 4.125406265258789, + "learning_rate": 1.3909465020576132e-05, + "loss": 1.7062, + "step": 1014 + }, + { + "epoch": 2.251802551303383, + "grad_norm": 4.105384826660156, + "learning_rate": 1.3868312757201646e-05, + "loss": 1.8488, + "step": 1015 + }, + { + "epoch": 2.25402107598447, + "grad_norm": 3.9614810943603516, + "learning_rate": 1.382716049382716e-05, + "loss": 1.5319, + "step": 1016 + }, + { + "epoch": 2.256239600665557, + "grad_norm": 3.8662824630737305, + "learning_rate": 1.3786008230452676e-05, + "loss": 1.5196, + "step": 1017 + }, + { + "epoch": 2.2584581253466443, + "grad_norm": 4.39835786819458, + "learning_rate": 1.374485596707819e-05, + "loss": 1.7473, + "step": 1018 + }, + { + "epoch": 2.2606766500277313, + "grad_norm": 3.5864243507385254, + "learning_rate": 1.3703703703703704e-05, + "loss": 1.6467, + "step": 1019 + }, + { + "epoch": 2.2628951747088184, + "grad_norm": 3.591038465499878, + "learning_rate": 1.3662551440329218e-05, + "loss": 1.5338, + "step": 1020 + }, + { + "epoch": 2.2651136993899055, + "grad_norm": 3.775390863418579, + "learning_rate": 1.3621399176954732e-05, + "loss": 1.4184, + "step": 1021 + }, + { + "epoch": 2.2673322240709926, + "grad_norm": 3.8518340587615967, + "learning_rate": 1.3580246913580247e-05, + "loss": 1.1841, + "step": 1022 + }, + { + "epoch": 2.2695507487520796, + "grad_norm": 3.7525668144226074, + "learning_rate": 1.353909465020576e-05, + "loss": 1.3941, + "step": 1023 + }, + { + "epoch": 2.2717692734331667, + "grad_norm": 3.7586536407470703, + "learning_rate": 1.3497942386831278e-05, + "loss": 1.1819, + "step": 1024 + }, + { + "epoch": 2.2739877981142538, + "grad_norm": 3.5614144802093506, + "learning_rate": 1.3456790123456793e-05, + "loss": 1.4872, + "step": 1025 + }, + { + "epoch": 2.276206322795341, + "grad_norm": 4.390314102172852, + "learning_rate": 1.3415637860082305e-05, + "loss": 1.3354, + "step": 1026 + }, + { + "epoch": 2.278424847476428, + "grad_norm": 3.9189064502716064, + "learning_rate": 1.3374485596707819e-05, + "loss": 1.3888, + "step": 1027 + }, + { + "epoch": 2.280643372157515, + "grad_norm": 4.0031962394714355, + "learning_rate": 1.3333333333333333e-05, + "loss": 1.4519, + "step": 1028 + }, + { + "epoch": 2.2828618968386025, + "grad_norm": 3.5403802394866943, + "learning_rate": 1.3292181069958847e-05, + "loss": 1.3021, + "step": 1029 + }, + { + "epoch": 2.2850804215196896, + "grad_norm": 4.270956516265869, + "learning_rate": 1.3251028806584362e-05, + "loss": 1.548, + "step": 1030 + }, + { + "epoch": 2.2872989462007767, + "grad_norm": 4.335587024688721, + "learning_rate": 1.3209876543209876e-05, + "loss": 1.415, + "step": 1031 + }, + { + "epoch": 2.2895174708818637, + "grad_norm": 3.798656940460205, + "learning_rate": 1.3168724279835393e-05, + "loss": 1.5733, + "step": 1032 + }, + { + "epoch": 2.291735995562951, + "grad_norm": 4.109601020812988, + "learning_rate": 1.3127572016460907e-05, + "loss": 1.1627, + "step": 1033 + }, + { + "epoch": 2.293954520244038, + "grad_norm": 4.005099296569824, + "learning_rate": 1.308641975308642e-05, + "loss": 1.3211, + "step": 1034 + }, + { + "epoch": 2.296173044925125, + "grad_norm": 4.028604507446289, + "learning_rate": 1.3045267489711934e-05, + "loss": 1.403, + "step": 1035 + }, + { + "epoch": 2.298391569606212, + "grad_norm": 4.111745357513428, + "learning_rate": 1.3004115226337448e-05, + "loss": 1.5613, + "step": 1036 + }, + { + "epoch": 2.300610094287299, + "grad_norm": 3.9915003776550293, + "learning_rate": 1.2962962962962962e-05, + "loss": 1.3551, + "step": 1037 + }, + { + "epoch": 2.302828618968386, + "grad_norm": 4.25387716293335, + "learning_rate": 1.2921810699588477e-05, + "loss": 1.6476, + "step": 1038 + }, + { + "epoch": 2.3050471436494733, + "grad_norm": 3.9554834365844727, + "learning_rate": 1.2880658436213994e-05, + "loss": 1.3808, + "step": 1039 + }, + { + "epoch": 2.3072656683305603, + "grad_norm": 3.7731175422668457, + "learning_rate": 1.2839506172839508e-05, + "loss": 1.4033, + "step": 1040 + }, + { + "epoch": 2.3094841930116474, + "grad_norm": 3.7711591720581055, + "learning_rate": 1.2798353909465022e-05, + "loss": 1.7234, + "step": 1041 + }, + { + "epoch": 2.3117027176927345, + "grad_norm": 3.888958215713501, + "learning_rate": 1.2757201646090535e-05, + "loss": 1.3863, + "step": 1042 + }, + { + "epoch": 2.3139212423738216, + "grad_norm": 4.029396057128906, + "learning_rate": 1.2716049382716049e-05, + "loss": 1.2297, + "step": 1043 + }, + { + "epoch": 2.3161397670549086, + "grad_norm": 3.7822911739349365, + "learning_rate": 1.2674897119341563e-05, + "loss": 1.468, + "step": 1044 + }, + { + "epoch": 2.3183582917359957, + "grad_norm": 4.189430236816406, + "learning_rate": 1.2633744855967077e-05, + "loss": 1.4294, + "step": 1045 + }, + { + "epoch": 2.320576816417083, + "grad_norm": 3.677086114883423, + "learning_rate": 1.2592592592592592e-05, + "loss": 1.6808, + "step": 1046 + }, + { + "epoch": 2.32279534109817, + "grad_norm": 3.514536142349243, + "learning_rate": 1.255144032921811e-05, + "loss": 1.4354, + "step": 1047 + }, + { + "epoch": 2.325013865779257, + "grad_norm": 3.850598096847534, + "learning_rate": 1.2510288065843623e-05, + "loss": 1.4343, + "step": 1048 + }, + { + "epoch": 2.327232390460344, + "grad_norm": 4.873521327972412, + "learning_rate": 1.2469135802469137e-05, + "loss": 1.3326, + "step": 1049 + }, + { + "epoch": 2.329450915141431, + "grad_norm": 3.858448028564453, + "learning_rate": 1.2427983539094652e-05, + "loss": 1.2713, + "step": 1050 + }, + { + "epoch": 2.331669439822518, + "grad_norm": 3.9477570056915283, + "learning_rate": 1.2386831275720164e-05, + "loss": 1.5119, + "step": 1051 + }, + { + "epoch": 2.3338879645036053, + "grad_norm": 3.6840178966522217, + "learning_rate": 1.2345679012345678e-05, + "loss": 1.0818, + "step": 1052 + }, + { + "epoch": 2.3361064891846923, + "grad_norm": 4.025064468383789, + "learning_rate": 1.2304526748971194e-05, + "loss": 1.7836, + "step": 1053 + }, + { + "epoch": 2.3383250138657794, + "grad_norm": 3.750805377960205, + "learning_rate": 1.2263374485596708e-05, + "loss": 1.2293, + "step": 1054 + }, + { + "epoch": 2.3405435385468665, + "grad_norm": 3.4582715034484863, + "learning_rate": 1.2222222222222222e-05, + "loss": 1.2486, + "step": 1055 + }, + { + "epoch": 2.3427620632279536, + "grad_norm": 3.9408199787139893, + "learning_rate": 1.2181069958847738e-05, + "loss": 1.5427, + "step": 1056 + }, + { + "epoch": 2.3449805879090406, + "grad_norm": 4.038824081420898, + "learning_rate": 1.2139917695473252e-05, + "loss": 1.2267, + "step": 1057 + }, + { + "epoch": 2.3471991125901277, + "grad_norm": 4.156043529510498, + "learning_rate": 1.2098765432098767e-05, + "loss": 1.6785, + "step": 1058 + }, + { + "epoch": 2.3494176372712148, + "grad_norm": 4.085957050323486, + "learning_rate": 1.2057613168724279e-05, + "loss": 1.3797, + "step": 1059 + }, + { + "epoch": 2.351636161952302, + "grad_norm": 3.8693175315856934, + "learning_rate": 1.2016460905349795e-05, + "loss": 1.3051, + "step": 1060 + }, + { + "epoch": 2.353854686633389, + "grad_norm": 3.7445099353790283, + "learning_rate": 1.1975308641975309e-05, + "loss": 1.4274, + "step": 1061 + }, + { + "epoch": 2.356073211314476, + "grad_norm": 3.6459882259368896, + "learning_rate": 1.1934156378600823e-05, + "loss": 1.4436, + "step": 1062 + }, + { + "epoch": 2.358291735995563, + "grad_norm": 3.777609348297119, + "learning_rate": 1.1893004115226337e-05, + "loss": 1.3668, + "step": 1063 + }, + { + "epoch": 2.36051026067665, + "grad_norm": 3.9784533977508545, + "learning_rate": 1.1851851851851853e-05, + "loss": 1.4381, + "step": 1064 + }, + { + "epoch": 2.3627287853577372, + "grad_norm": 4.170048713684082, + "learning_rate": 1.1810699588477367e-05, + "loss": 1.2004, + "step": 1065 + }, + { + "epoch": 2.3649473100388243, + "grad_norm": 3.477304220199585, + "learning_rate": 1.1769547325102882e-05, + "loss": 1.358, + "step": 1066 + }, + { + "epoch": 2.3671658347199114, + "grad_norm": 4.022628307342529, + "learning_rate": 1.1728395061728396e-05, + "loss": 1.3585, + "step": 1067 + }, + { + "epoch": 2.3693843594009985, + "grad_norm": 3.8395752906799316, + "learning_rate": 1.168724279835391e-05, + "loss": 1.6334, + "step": 1068 + }, + { + "epoch": 2.3716028840820855, + "grad_norm": 3.6747024059295654, + "learning_rate": 1.1646090534979424e-05, + "loss": 1.3321, + "step": 1069 + }, + { + "epoch": 2.3738214087631726, + "grad_norm": 3.546860694885254, + "learning_rate": 1.1604938271604938e-05, + "loss": 1.3108, + "step": 1070 + }, + { + "epoch": 2.3760399334442597, + "grad_norm": 4.011758804321289, + "learning_rate": 1.1563786008230454e-05, + "loss": 1.73, + "step": 1071 + }, + { + "epoch": 2.3782584581253468, + "grad_norm": 4.049909591674805, + "learning_rate": 1.1522633744855968e-05, + "loss": 1.5038, + "step": 1072 + }, + { + "epoch": 2.380476982806434, + "grad_norm": 3.6497087478637695, + "learning_rate": 1.1481481481481482e-05, + "loss": 1.6093, + "step": 1073 + }, + { + "epoch": 2.382695507487521, + "grad_norm": 3.788050413131714, + "learning_rate": 1.1440329218106997e-05, + "loss": 1.5813, + "step": 1074 + }, + { + "epoch": 2.384914032168608, + "grad_norm": 3.833214044570923, + "learning_rate": 1.139917695473251e-05, + "loss": 1.1188, + "step": 1075 + }, + { + "epoch": 2.387132556849695, + "grad_norm": 4.049717903137207, + "learning_rate": 1.1358024691358025e-05, + "loss": 1.1781, + "step": 1076 + }, + { + "epoch": 2.389351081530782, + "grad_norm": 3.8376920223236084, + "learning_rate": 1.1316872427983539e-05, + "loss": 1.7855, + "step": 1077 + }, + { + "epoch": 2.391569606211869, + "grad_norm": 3.7818570137023926, + "learning_rate": 1.1275720164609053e-05, + "loss": 1.6846, + "step": 1078 + }, + { + "epoch": 2.3937881308929563, + "grad_norm": 4.380697727203369, + "learning_rate": 1.123456790123457e-05, + "loss": 1.5155, + "step": 1079 + }, + { + "epoch": 2.3960066555740434, + "grad_norm": 3.932436227798462, + "learning_rate": 1.1193415637860083e-05, + "loss": 1.2003, + "step": 1080 + }, + { + "epoch": 2.3982251802551304, + "grad_norm": 3.804105520248413, + "learning_rate": 1.1152263374485597e-05, + "loss": 1.5141, + "step": 1081 + }, + { + "epoch": 2.4004437049362175, + "grad_norm": 3.91374135017395, + "learning_rate": 1.1111111111111112e-05, + "loss": 1.3231, + "step": 1082 + }, + { + "epoch": 2.4026622296173046, + "grad_norm": 3.3297526836395264, + "learning_rate": 1.1069958847736626e-05, + "loss": 1.6887, + "step": 1083 + }, + { + "epoch": 2.4048807542983917, + "grad_norm": 3.404630422592163, + "learning_rate": 1.102880658436214e-05, + "loss": 1.2838, + "step": 1084 + }, + { + "epoch": 2.4070992789794787, + "grad_norm": 4.093790531158447, + "learning_rate": 1.0987654320987654e-05, + "loss": 1.3323, + "step": 1085 + }, + { + "epoch": 2.409317803660566, + "grad_norm": 3.7639834880828857, + "learning_rate": 1.094650205761317e-05, + "loss": 1.3049, + "step": 1086 + }, + { + "epoch": 2.411536328341653, + "grad_norm": 3.9918723106384277, + "learning_rate": 1.0905349794238684e-05, + "loss": 1.2269, + "step": 1087 + }, + { + "epoch": 2.41375485302274, + "grad_norm": 3.802032232284546, + "learning_rate": 1.0864197530864198e-05, + "loss": 1.2155, + "step": 1088 + }, + { + "epoch": 2.415973377703827, + "grad_norm": 3.6317007541656494, + "learning_rate": 1.0823045267489712e-05, + "loss": 1.2809, + "step": 1089 + }, + { + "epoch": 2.418191902384914, + "grad_norm": 4.001982688903809, + "learning_rate": 1.0781893004115227e-05, + "loss": 1.5621, + "step": 1090 + }, + { + "epoch": 2.420410427066001, + "grad_norm": 3.7673041820526123, + "learning_rate": 1.074074074074074e-05, + "loss": 1.5021, + "step": 1091 + }, + { + "epoch": 2.4226289517470883, + "grad_norm": 3.8592369556427, + "learning_rate": 1.0699588477366255e-05, + "loss": 1.4527, + "step": 1092 + }, + { + "epoch": 2.4248474764281753, + "grad_norm": 4.085554122924805, + "learning_rate": 1.0658436213991769e-05, + "loss": 1.3244, + "step": 1093 + }, + { + "epoch": 2.4270660011092624, + "grad_norm": 4.16439962387085, + "learning_rate": 1.0617283950617285e-05, + "loss": 1.423, + "step": 1094 + }, + { + "epoch": 2.4292845257903495, + "grad_norm": 3.798414468765259, + "learning_rate": 1.0576131687242799e-05, + "loss": 1.39, + "step": 1095 + }, + { + "epoch": 2.4315030504714366, + "grad_norm": 3.629950761795044, + "learning_rate": 1.0534979423868313e-05, + "loss": 1.5674, + "step": 1096 + }, + { + "epoch": 2.4337215751525236, + "grad_norm": 3.6328680515289307, + "learning_rate": 1.0493827160493827e-05, + "loss": 1.3634, + "step": 1097 + }, + { + "epoch": 2.4359400998336107, + "grad_norm": 3.9492409229278564, + "learning_rate": 1.0452674897119343e-05, + "loss": 1.3094, + "step": 1098 + }, + { + "epoch": 2.438158624514698, + "grad_norm": 3.667323589324951, + "learning_rate": 1.0411522633744856e-05, + "loss": 1.1942, + "step": 1099 + }, + { + "epoch": 2.440377149195785, + "grad_norm": 3.6013240814208984, + "learning_rate": 1.037037037037037e-05, + "loss": 1.2663, + "step": 1100 + }, + { + "epoch": 2.442595673876872, + "grad_norm": 3.7321043014526367, + "learning_rate": 1.0329218106995884e-05, + "loss": 1.3664, + "step": 1101 + }, + { + "epoch": 2.444814198557959, + "grad_norm": 3.5682711601257324, + "learning_rate": 1.02880658436214e-05, + "loss": 1.5059, + "step": 1102 + }, + { + "epoch": 2.447032723239046, + "grad_norm": 3.7535288333892822, + "learning_rate": 1.0246913580246914e-05, + "loss": 1.4421, + "step": 1103 + }, + { + "epoch": 2.449251247920133, + "grad_norm": 3.7439565658569336, + "learning_rate": 1.0205761316872428e-05, + "loss": 1.3316, + "step": 1104 + }, + { + "epoch": 2.4514697726012202, + "grad_norm": 3.4793806076049805, + "learning_rate": 1.0164609053497944e-05, + "loss": 1.1924, + "step": 1105 + }, + { + "epoch": 2.4536882972823073, + "grad_norm": 4.249213695526123, + "learning_rate": 1.0123456790123458e-05, + "loss": 1.2692, + "step": 1106 + }, + { + "epoch": 2.4559068219633944, + "grad_norm": 4.1428632736206055, + "learning_rate": 1.008230452674897e-05, + "loss": 1.4103, + "step": 1107 + }, + { + "epoch": 2.4581253466444815, + "grad_norm": 4.109711170196533, + "learning_rate": 1.0041152263374485e-05, + "loss": 1.2586, + "step": 1108 + }, + { + "epoch": 2.4603438713255685, + "grad_norm": 4.325128078460693, + "learning_rate": 1e-05, + "loss": 0.985, + "step": 1109 + }, + { + "epoch": 2.4625623960066556, + "grad_norm": 4.2221269607543945, + "learning_rate": 9.958847736625515e-06, + "loss": 1.398, + "step": 1110 + }, + { + "epoch": 2.4647809206877427, + "grad_norm": 3.656083822250366, + "learning_rate": 9.917695473251029e-06, + "loss": 1.0478, + "step": 1111 + }, + { + "epoch": 2.4669994453688298, + "grad_norm": 3.86702036857605, + "learning_rate": 9.876543209876543e-06, + "loss": 1.2366, + "step": 1112 + }, + { + "epoch": 2.469217970049917, + "grad_norm": 4.61433219909668, + "learning_rate": 9.83539094650206e-06, + "loss": 1.3806, + "step": 1113 + }, + { + "epoch": 2.471436494731004, + "grad_norm": 4.039587497711182, + "learning_rate": 9.794238683127573e-06, + "loss": 1.5498, + "step": 1114 + }, + { + "epoch": 2.473655019412091, + "grad_norm": 4.302928447723389, + "learning_rate": 9.753086419753086e-06, + "loss": 1.1752, + "step": 1115 + }, + { + "epoch": 2.475873544093178, + "grad_norm": 4.888288497924805, + "learning_rate": 9.7119341563786e-06, + "loss": 1.4821, + "step": 1116 + }, + { + "epoch": 2.478092068774265, + "grad_norm": 3.8774938583374023, + "learning_rate": 9.670781893004116e-06, + "loss": 1.5059, + "step": 1117 + }, + { + "epoch": 2.480310593455352, + "grad_norm": 4.046879291534424, + "learning_rate": 9.62962962962963e-06, + "loss": 1.2234, + "step": 1118 + }, + { + "epoch": 2.4825291181364393, + "grad_norm": 4.263421535491943, + "learning_rate": 9.588477366255144e-06, + "loss": 1.3343, + "step": 1119 + }, + { + "epoch": 2.4847476428175264, + "grad_norm": 3.679917097091675, + "learning_rate": 9.54732510288066e-06, + "loss": 1.5058, + "step": 1120 + }, + { + "epoch": 2.4869661674986134, + "grad_norm": 4.598412036895752, + "learning_rate": 9.506172839506174e-06, + "loss": 1.3729, + "step": 1121 + }, + { + "epoch": 2.4891846921797005, + "grad_norm": 3.9892165660858154, + "learning_rate": 9.465020576131688e-06, + "loss": 1.3406, + "step": 1122 + }, + { + "epoch": 2.4914032168607876, + "grad_norm": 3.717521905899048, + "learning_rate": 9.423868312757202e-06, + "loss": 1.2569, + "step": 1123 + }, + { + "epoch": 2.4936217415418747, + "grad_norm": 3.852292776107788, + "learning_rate": 9.382716049382717e-06, + "loss": 1.4192, + "step": 1124 + }, + { + "epoch": 2.4958402662229617, + "grad_norm": 3.787449598312378, + "learning_rate": 9.34156378600823e-06, + "loss": 1.6987, + "step": 1125 + }, + { + "epoch": 2.498058790904049, + "grad_norm": 4.288355350494385, + "learning_rate": 9.300411522633745e-06, + "loss": 1.4955, + "step": 1126 + }, + { + "epoch": 2.500277315585136, + "grad_norm": 3.4219627380371094, + "learning_rate": 9.259259259259259e-06, + "loss": 1.1955, + "step": 1127 + }, + { + "epoch": 2.502495840266223, + "grad_norm": 3.2673866748809814, + "learning_rate": 9.218106995884775e-06, + "loss": 1.3691, + "step": 1128 + }, + { + "epoch": 2.50471436494731, + "grad_norm": 4.100692272186279, + "learning_rate": 9.17695473251029e-06, + "loss": 1.557, + "step": 1129 + }, + { + "epoch": 2.506932889628397, + "grad_norm": 3.9575514793395996, + "learning_rate": 9.135802469135803e-06, + "loss": 1.1858, + "step": 1130 + }, + { + "epoch": 2.509151414309484, + "grad_norm": 4.138423442840576, + "learning_rate": 9.094650205761317e-06, + "loss": 1.3203, + "step": 1131 + }, + { + "epoch": 2.5113699389905713, + "grad_norm": 4.166389465332031, + "learning_rate": 9.053497942386832e-06, + "loss": 1.5005, + "step": 1132 + }, + { + "epoch": 2.5135884636716583, + "grad_norm": 3.5071444511413574, + "learning_rate": 9.012345679012346e-06, + "loss": 1.2941, + "step": 1133 + }, + { + "epoch": 2.5158069883527454, + "grad_norm": 4.00253963470459, + "learning_rate": 8.97119341563786e-06, + "loss": 1.5303, + "step": 1134 + }, + { + "epoch": 2.5180255130338325, + "grad_norm": 4.017101764678955, + "learning_rate": 8.930041152263374e-06, + "loss": 1.1583, + "step": 1135 + }, + { + "epoch": 2.5202440377149196, + "grad_norm": 3.860346555709839, + "learning_rate": 8.88888888888889e-06, + "loss": 1.4786, + "step": 1136 + }, + { + "epoch": 2.5224625623960066, + "grad_norm": 4.127157688140869, + "learning_rate": 8.847736625514404e-06, + "loss": 1.466, + "step": 1137 + }, + { + "epoch": 2.5246810870770937, + "grad_norm": 3.5574755668640137, + "learning_rate": 8.806584362139918e-06, + "loss": 1.4502, + "step": 1138 + }, + { + "epoch": 2.526899611758181, + "grad_norm": 3.9966347217559814, + "learning_rate": 8.765432098765432e-06, + "loss": 1.4984, + "step": 1139 + }, + { + "epoch": 2.529118136439268, + "grad_norm": 3.9290337562561035, + "learning_rate": 8.724279835390947e-06, + "loss": 1.3574, + "step": 1140 + }, + { + "epoch": 2.531336661120355, + "grad_norm": 3.7375710010528564, + "learning_rate": 8.68312757201646e-06, + "loss": 1.0199, + "step": 1141 + }, + { + "epoch": 2.533555185801442, + "grad_norm": 3.6255428791046143, + "learning_rate": 8.641975308641975e-06, + "loss": 1.2066, + "step": 1142 + }, + { + "epoch": 2.535773710482529, + "grad_norm": 3.889197587966919, + "learning_rate": 8.60082304526749e-06, + "loss": 1.5661, + "step": 1143 + }, + { + "epoch": 2.537992235163616, + "grad_norm": 3.5962159633636475, + "learning_rate": 8.559670781893005e-06, + "loss": 1.6541, + "step": 1144 + }, + { + "epoch": 2.5402107598447032, + "grad_norm": 4.1230692863464355, + "learning_rate": 8.518518518518519e-06, + "loss": 1.2713, + "step": 1145 + }, + { + "epoch": 2.5424292845257903, + "grad_norm": 3.928884506225586, + "learning_rate": 8.477366255144033e-06, + "loss": 1.464, + "step": 1146 + }, + { + "epoch": 2.5446478092068774, + "grad_norm": 3.9417717456817627, + "learning_rate": 8.436213991769547e-06, + "loss": 1.6556, + "step": 1147 + }, + { + "epoch": 2.5468663338879645, + "grad_norm": 3.7934725284576416, + "learning_rate": 8.395061728395062e-06, + "loss": 1.489, + "step": 1148 + }, + { + "epoch": 2.5490848585690515, + "grad_norm": 4.402404308319092, + "learning_rate": 8.353909465020576e-06, + "loss": 1.2869, + "step": 1149 + }, + { + "epoch": 2.5513033832501386, + "grad_norm": 4.070915699005127, + "learning_rate": 8.31275720164609e-06, + "loss": 1.166, + "step": 1150 + }, + { + "epoch": 2.5535219079312257, + "grad_norm": 3.5672740936279297, + "learning_rate": 8.271604938271606e-06, + "loss": 1.3355, + "step": 1151 + }, + { + "epoch": 2.5557404326123128, + "grad_norm": 3.982445240020752, + "learning_rate": 8.23045267489712e-06, + "loss": 1.2642, + "step": 1152 + }, + { + "epoch": 2.5579589572934, + "grad_norm": 3.516082286834717, + "learning_rate": 8.189300411522634e-06, + "loss": 1.3305, + "step": 1153 + }, + { + "epoch": 2.560177481974487, + "grad_norm": 3.841043472290039, + "learning_rate": 8.14814814814815e-06, + "loss": 1.3532, + "step": 1154 + }, + { + "epoch": 2.562396006655574, + "grad_norm": 3.8949553966522217, + "learning_rate": 8.106995884773662e-06, + "loss": 1.3692, + "step": 1155 + }, + { + "epoch": 2.564614531336661, + "grad_norm": 4.0778889656066895, + "learning_rate": 8.065843621399177e-06, + "loss": 1.2287, + "step": 1156 + }, + { + "epoch": 2.566833056017748, + "grad_norm": 3.6213083267211914, + "learning_rate": 8.02469135802469e-06, + "loss": 1.1447, + "step": 1157 + }, + { + "epoch": 2.569051580698835, + "grad_norm": 3.9147374629974365, + "learning_rate": 7.983539094650207e-06, + "loss": 1.6018, + "step": 1158 + }, + { + "epoch": 2.5712701053799223, + "grad_norm": 3.9777629375457764, + "learning_rate": 7.94238683127572e-06, + "loss": 1.2641, + "step": 1159 + }, + { + "epoch": 2.5734886300610094, + "grad_norm": 3.90348219871521, + "learning_rate": 7.901234567901235e-06, + "loss": 1.329, + "step": 1160 + }, + { + "epoch": 2.5757071547420964, + "grad_norm": 4.397609233856201, + "learning_rate": 7.860082304526749e-06, + "loss": 1.63, + "step": 1161 + }, + { + "epoch": 2.5779256794231835, + "grad_norm": 4.109441757202148, + "learning_rate": 7.818930041152265e-06, + "loss": 1.3309, + "step": 1162 + }, + { + "epoch": 2.5801442041042706, + "grad_norm": 4.050140380859375, + "learning_rate": 7.777777777777777e-06, + "loss": 1.5208, + "step": 1163 + }, + { + "epoch": 2.5823627287853577, + "grad_norm": 3.741532325744629, + "learning_rate": 7.736625514403292e-06, + "loss": 1.3165, + "step": 1164 + }, + { + "epoch": 2.5845812534664447, + "grad_norm": 3.6853437423706055, + "learning_rate": 7.695473251028806e-06, + "loss": 1.3556, + "step": 1165 + }, + { + "epoch": 2.586799778147532, + "grad_norm": 4.26183557510376, + "learning_rate": 7.654320987654322e-06, + "loss": 1.5284, + "step": 1166 + }, + { + "epoch": 2.589018302828619, + "grad_norm": 3.8578245639801025, + "learning_rate": 7.613168724279836e-06, + "loss": 1.4857, + "step": 1167 + }, + { + "epoch": 2.591236827509706, + "grad_norm": 3.890026569366455, + "learning_rate": 7.57201646090535e-06, + "loss": 1.2144, + "step": 1168 + }, + { + "epoch": 2.593455352190793, + "grad_norm": 3.6726675033569336, + "learning_rate": 7.530864197530865e-06, + "loss": 1.1495, + "step": 1169 + }, + { + "epoch": 2.59567387687188, + "grad_norm": 3.4492056369781494, + "learning_rate": 7.489711934156379e-06, + "loss": 1.3404, + "step": 1170 + }, + { + "epoch": 2.597892401552967, + "grad_norm": 3.8797051906585693, + "learning_rate": 7.448559670781893e-06, + "loss": 1.1676, + "step": 1171 + }, + { + "epoch": 2.6001109262340543, + "grad_norm": 4.247097015380859, + "learning_rate": 7.4074074074074075e-06, + "loss": 1.5959, + "step": 1172 + }, + { + "epoch": 2.6023294509151413, + "grad_norm": 3.860238790512085, + "learning_rate": 7.3662551440329225e-06, + "loss": 1.1935, + "step": 1173 + }, + { + "epoch": 2.6045479755962284, + "grad_norm": 3.6471588611602783, + "learning_rate": 7.325102880658437e-06, + "loss": 1.2864, + "step": 1174 + }, + { + "epoch": 2.6067665002773155, + "grad_norm": 4.316948413848877, + "learning_rate": 7.283950617283951e-06, + "loss": 1.3837, + "step": 1175 + }, + { + "epoch": 2.6089850249584026, + "grad_norm": 3.5124545097351074, + "learning_rate": 7.242798353909465e-06, + "loss": 1.4238, + "step": 1176 + }, + { + "epoch": 2.6112035496394896, + "grad_norm": 4.004002571105957, + "learning_rate": 7.20164609053498e-06, + "loss": 1.4178, + "step": 1177 + }, + { + "epoch": 2.6134220743205767, + "grad_norm": 3.9949088096618652, + "learning_rate": 7.160493827160494e-06, + "loss": 1.3406, + "step": 1178 + }, + { + "epoch": 2.615640599001664, + "grad_norm": 4.230592250823975, + "learning_rate": 7.119341563786008e-06, + "loss": 1.4301, + "step": 1179 + }, + { + "epoch": 2.617859123682751, + "grad_norm": 3.6380198001861572, + "learning_rate": 7.0781893004115225e-06, + "loss": 1.5242, + "step": 1180 + }, + { + "epoch": 2.620077648363838, + "grad_norm": 3.8554847240448, + "learning_rate": 7.0370370370370375e-06, + "loss": 1.5645, + "step": 1181 + }, + { + "epoch": 2.622296173044925, + "grad_norm": 3.908169746398926, + "learning_rate": 6.995884773662552e-06, + "loss": 1.5511, + "step": 1182 + }, + { + "epoch": 2.624514697726012, + "grad_norm": 4.089354515075684, + "learning_rate": 6.954732510288066e-06, + "loss": 1.0948, + "step": 1183 + }, + { + "epoch": 2.626733222407099, + "grad_norm": 4.142337322235107, + "learning_rate": 6.91358024691358e-06, + "loss": 1.2945, + "step": 1184 + }, + { + "epoch": 2.6289517470881862, + "grad_norm": 4.057214260101318, + "learning_rate": 6.872427983539095e-06, + "loss": 1.485, + "step": 1185 + }, + { + "epoch": 2.6311702717692733, + "grad_norm": 4.486181735992432, + "learning_rate": 6.831275720164609e-06, + "loss": 1.3249, + "step": 1186 + }, + { + "epoch": 2.6333887964503604, + "grad_norm": 4.087220191955566, + "learning_rate": 6.790123456790123e-06, + "loss": 1.5441, + "step": 1187 + }, + { + "epoch": 2.6356073211314475, + "grad_norm": 3.9394397735595703, + "learning_rate": 6.748971193415639e-06, + "loss": 1.3593, + "step": 1188 + }, + { + "epoch": 2.6378258458125345, + "grad_norm": 4.021302700042725, + "learning_rate": 6.7078189300411525e-06, + "loss": 1.4738, + "step": 1189 + }, + { + "epoch": 2.6400443704936216, + "grad_norm": 3.794167995452881, + "learning_rate": 6.666666666666667e-06, + "loss": 1.5093, + "step": 1190 + }, + { + "epoch": 2.6422628951747087, + "grad_norm": 3.8649303913116455, + "learning_rate": 6.625514403292181e-06, + "loss": 1.1667, + "step": 1191 + }, + { + "epoch": 2.6444814198557958, + "grad_norm": 3.9752180576324463, + "learning_rate": 6.584362139917697e-06, + "loss": 1.4928, + "step": 1192 + }, + { + "epoch": 2.646699944536883, + "grad_norm": 4.228023052215576, + "learning_rate": 6.54320987654321e-06, + "loss": 1.789, + "step": 1193 + }, + { + "epoch": 2.64891846921797, + "grad_norm": 3.921043872833252, + "learning_rate": 6.502057613168724e-06, + "loss": 1.2052, + "step": 1194 + }, + { + "epoch": 2.651136993899057, + "grad_norm": 3.966921329498291, + "learning_rate": 6.460905349794238e-06, + "loss": 1.4127, + "step": 1195 + }, + { + "epoch": 2.653355518580144, + "grad_norm": 4.2098774909973145, + "learning_rate": 6.419753086419754e-06, + "loss": 1.2768, + "step": 1196 + }, + { + "epoch": 2.655574043261231, + "grad_norm": 3.7732021808624268, + "learning_rate": 6.3786008230452675e-06, + "loss": 1.0774, + "step": 1197 + }, + { + "epoch": 2.657792567942318, + "grad_norm": 3.729081392288208, + "learning_rate": 6.337448559670782e-06, + "loss": 1.1856, + "step": 1198 + }, + { + "epoch": 2.6600110926234053, + "grad_norm": 3.564211845397949, + "learning_rate": 6.296296296296296e-06, + "loss": 1.2054, + "step": 1199 + }, + { + "epoch": 2.6622296173044924, + "grad_norm": 4.480288505554199, + "learning_rate": 6.255144032921812e-06, + "loss": 1.2905, + "step": 1200 + }, + { + "epoch": 2.6644481419855794, + "grad_norm": 3.799375534057617, + "learning_rate": 6.213991769547326e-06, + "loss": 1.2672, + "step": 1201 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 3.8796894550323486, + "learning_rate": 6.172839506172839e-06, + "loss": 1.5933, + "step": 1202 + }, + { + "epoch": 2.6688851913477536, + "grad_norm": 4.017917633056641, + "learning_rate": 6.131687242798354e-06, + "loss": 1.5456, + "step": 1203 + }, + { + "epoch": 2.6711037160288407, + "grad_norm": 4.277782917022705, + "learning_rate": 6.090534979423869e-06, + "loss": 1.5539, + "step": 1204 + }, + { + "epoch": 2.6733222407099277, + "grad_norm": 3.8942630290985107, + "learning_rate": 6.049382716049383e-06, + "loss": 1.3645, + "step": 1205 + }, + { + "epoch": 2.675540765391015, + "grad_norm": 3.851686477661133, + "learning_rate": 6.0082304526748975e-06, + "loss": 1.2811, + "step": 1206 + }, + { + "epoch": 2.677759290072102, + "grad_norm": 3.82453989982605, + "learning_rate": 5.967078189300412e-06, + "loss": 1.437, + "step": 1207 + }, + { + "epoch": 2.679977814753189, + "grad_norm": 3.883971929550171, + "learning_rate": 5.925925925925927e-06, + "loss": 1.4571, + "step": 1208 + }, + { + "epoch": 2.682196339434276, + "grad_norm": 3.6149773597717285, + "learning_rate": 5.884773662551441e-06, + "loss": 1.2119, + "step": 1209 + }, + { + "epoch": 2.684414864115363, + "grad_norm": 3.892890691757202, + "learning_rate": 5.843621399176955e-06, + "loss": 1.3441, + "step": 1210 + }, + { + "epoch": 2.68663338879645, + "grad_norm": 3.7986695766448975, + "learning_rate": 5.802469135802469e-06, + "loss": 1.4808, + "step": 1211 + }, + { + "epoch": 2.6888519134775377, + "grad_norm": 3.4227142333984375, + "learning_rate": 5.761316872427984e-06, + "loss": 1.4897, + "step": 1212 + }, + { + "epoch": 2.691070438158625, + "grad_norm": 3.9402198791503906, + "learning_rate": 5.720164609053498e-06, + "loss": 1.6328, + "step": 1213 + }, + { + "epoch": 2.693288962839712, + "grad_norm": 4.120849609375, + "learning_rate": 5.6790123456790125e-06, + "loss": 1.2544, + "step": 1214 + }, + { + "epoch": 2.695507487520799, + "grad_norm": 3.9833056926727295, + "learning_rate": 5.637860082304527e-06, + "loss": 1.3374, + "step": 1215 + }, + { + "epoch": 2.697726012201886, + "grad_norm": 4.012001991271973, + "learning_rate": 5.596707818930042e-06, + "loss": 1.4502, + "step": 1216 + }, + { + "epoch": 2.699944536882973, + "grad_norm": 4.146241664886475, + "learning_rate": 5.555555555555556e-06, + "loss": 1.2439, + "step": 1217 + }, + { + "epoch": 2.70216306156406, + "grad_norm": 4.025453090667725, + "learning_rate": 5.51440329218107e-06, + "loss": 1.1479, + "step": 1218 + }, + { + "epoch": 2.7043815862451472, + "grad_norm": 4.444920539855957, + "learning_rate": 5.473251028806585e-06, + "loss": 1.2896, + "step": 1219 + }, + { + "epoch": 2.7066001109262343, + "grad_norm": 3.9975016117095947, + "learning_rate": 5.432098765432099e-06, + "loss": 1.361, + "step": 1220 + }, + { + "epoch": 2.7088186356073214, + "grad_norm": 3.7240941524505615, + "learning_rate": 5.390946502057613e-06, + "loss": 1.4649, + "step": 1221 + }, + { + "epoch": 2.7110371602884085, + "grad_norm": 3.714324951171875, + "learning_rate": 5.3497942386831275e-06, + "loss": 1.497, + "step": 1222 + }, + { + "epoch": 2.7132556849694955, + "grad_norm": 3.7697701454162598, + "learning_rate": 5.3086419753086425e-06, + "loss": 1.531, + "step": 1223 + }, + { + "epoch": 2.7154742096505826, + "grad_norm": 4.123478889465332, + "learning_rate": 5.267489711934157e-06, + "loss": 1.619, + "step": 1224 + }, + { + "epoch": 2.7176927343316697, + "grad_norm": 4.4261555671691895, + "learning_rate": 5.226337448559672e-06, + "loss": 1.2703, + "step": 1225 + }, + { + "epoch": 2.7199112590127568, + "grad_norm": 4.132821083068848, + "learning_rate": 5.185185185185185e-06, + "loss": 1.2923, + "step": 1226 + }, + { + "epoch": 2.722129783693844, + "grad_norm": 3.842866897583008, + "learning_rate": 5.1440329218107e-06, + "loss": 1.3368, + "step": 1227 + }, + { + "epoch": 2.724348308374931, + "grad_norm": 4.058122634887695, + "learning_rate": 5.102880658436214e-06, + "loss": 1.6839, + "step": 1228 + }, + { + "epoch": 2.726566833056018, + "grad_norm": 4.331025123596191, + "learning_rate": 5.061728395061729e-06, + "loss": 1.4451, + "step": 1229 + }, + { + "epoch": 2.728785357737105, + "grad_norm": 3.842027187347412, + "learning_rate": 5.0205761316872425e-06, + "loss": 1.3365, + "step": 1230 + }, + { + "epoch": 2.731003882418192, + "grad_norm": 4.104088306427002, + "learning_rate": 4.9794238683127575e-06, + "loss": 1.6456, + "step": 1231 + }, + { + "epoch": 2.733222407099279, + "grad_norm": 4.311492919921875, + "learning_rate": 4.938271604938272e-06, + "loss": 0.9938, + "step": 1232 + }, + { + "epoch": 2.7354409317803663, + "grad_norm": 3.3146557807922363, + "learning_rate": 4.897119341563787e-06, + "loss": 1.0442, + "step": 1233 + }, + { + "epoch": 2.7376594564614534, + "grad_norm": 3.9576833248138428, + "learning_rate": 4.8559670781893e-06, + "loss": 1.2745, + "step": 1234 + }, + { + "epoch": 2.7398779811425404, + "grad_norm": 4.019186496734619, + "learning_rate": 4.814814814814815e-06, + "loss": 1.6008, + "step": 1235 + }, + { + "epoch": 2.7420965058236275, + "grad_norm": 3.473548173904419, + "learning_rate": 4.77366255144033e-06, + "loss": 1.3114, + "step": 1236 + }, + { + "epoch": 2.7443150305047146, + "grad_norm": 3.8294081687927246, + "learning_rate": 4.732510288065844e-06, + "loss": 1.6561, + "step": 1237 + }, + { + "epoch": 2.7465335551858017, + "grad_norm": 3.612860679626465, + "learning_rate": 4.691358024691358e-06, + "loss": 1.3411, + "step": 1238 + }, + { + "epoch": 2.7487520798668887, + "grad_norm": 4.271310806274414, + "learning_rate": 4.6502057613168725e-06, + "loss": 1.3586, + "step": 1239 + }, + { + "epoch": 2.750970604547976, + "grad_norm": 3.6449451446533203, + "learning_rate": 4.6090534979423875e-06, + "loss": 1.2219, + "step": 1240 + }, + { + "epoch": 2.753189129229063, + "grad_norm": 3.7549049854278564, + "learning_rate": 4.567901234567902e-06, + "loss": 1.55, + "step": 1241 + }, + { + "epoch": 2.75540765391015, + "grad_norm": 3.898547649383545, + "learning_rate": 4.526748971193416e-06, + "loss": 1.5591, + "step": 1242 + }, + { + "epoch": 2.757626178591237, + "grad_norm": 3.469655752182007, + "learning_rate": 4.48559670781893e-06, + "loss": 1.357, + "step": 1243 + }, + { + "epoch": 2.759844703272324, + "grad_norm": 3.8755788803100586, + "learning_rate": 4.444444444444445e-06, + "loss": 1.5631, + "step": 1244 + }, + { + "epoch": 2.762063227953411, + "grad_norm": 3.702204704284668, + "learning_rate": 4.403292181069959e-06, + "loss": 1.7423, + "step": 1245 + }, + { + "epoch": 2.7642817526344983, + "grad_norm": 4.185690402984619, + "learning_rate": 4.362139917695473e-06, + "loss": 1.5321, + "step": 1246 + }, + { + "epoch": 2.7665002773155853, + "grad_norm": 3.7786407470703125, + "learning_rate": 4.3209876543209875e-06, + "loss": 1.6, + "step": 1247 + }, + { + "epoch": 2.7687188019966724, + "grad_norm": 3.7824084758758545, + "learning_rate": 4.2798353909465025e-06, + "loss": 1.2423, + "step": 1248 + }, + { + "epoch": 2.7709373266777595, + "grad_norm": 3.3668673038482666, + "learning_rate": 4.238683127572017e-06, + "loss": 1.5037, + "step": 1249 + }, + { + "epoch": 2.7731558513588466, + "grad_norm": 4.137018203735352, + "learning_rate": 4.197530864197531e-06, + "loss": 1.3207, + "step": 1250 + }, + { + "epoch": 2.7753743760399336, + "grad_norm": 3.979236364364624, + "learning_rate": 4.156378600823045e-06, + "loss": 1.4319, + "step": 1251 + }, + { + "epoch": 2.7775929007210207, + "grad_norm": 3.8160438537597656, + "learning_rate": 4.11522633744856e-06, + "loss": 1.2583, + "step": 1252 + }, + { + "epoch": 2.779811425402108, + "grad_norm": 3.996751070022583, + "learning_rate": 4.074074074074075e-06, + "loss": 1.6085, + "step": 1253 + }, + { + "epoch": 2.782029950083195, + "grad_norm": 4.216793060302734, + "learning_rate": 4.032921810699588e-06, + "loss": 1.1344, + "step": 1254 + }, + { + "epoch": 2.784248474764282, + "grad_norm": 3.738990068435669, + "learning_rate": 3.991769547325103e-06, + "loss": 1.5784, + "step": 1255 + }, + { + "epoch": 2.786466999445369, + "grad_norm": 3.514416217803955, + "learning_rate": 3.9506172839506175e-06, + "loss": 1.2389, + "step": 1256 + }, + { + "epoch": 2.788685524126456, + "grad_norm": 4.052081108093262, + "learning_rate": 3.9094650205761325e-06, + "loss": 1.4035, + "step": 1257 + }, + { + "epoch": 2.790904048807543, + "grad_norm": 3.447850465774536, + "learning_rate": 3.868312757201646e-06, + "loss": 1.2103, + "step": 1258 + }, + { + "epoch": 2.7931225734886302, + "grad_norm": 3.84885311126709, + "learning_rate": 3.827160493827161e-06, + "loss": 1.5958, + "step": 1259 + }, + { + "epoch": 2.7953410981697173, + "grad_norm": 3.8603932857513428, + "learning_rate": 3.786008230452675e-06, + "loss": 1.3423, + "step": 1260 + }, + { + "epoch": 2.7975596228508044, + "grad_norm": 3.706003189086914, + "learning_rate": 3.7448559670781896e-06, + "loss": 1.1835, + "step": 1261 + }, + { + "epoch": 2.7997781475318915, + "grad_norm": 4.098482131958008, + "learning_rate": 3.7037037037037037e-06, + "loss": 1.4134, + "step": 1262 + }, + { + "epoch": 2.8019966722129785, + "grad_norm": 3.6399295330047607, + "learning_rate": 3.6625514403292183e-06, + "loss": 1.2167, + "step": 1263 + }, + { + "epoch": 2.8042151968940656, + "grad_norm": 4.077873229980469, + "learning_rate": 3.6213991769547325e-06, + "loss": 1.3574, + "step": 1264 + }, + { + "epoch": 2.8064337215751527, + "grad_norm": 4.1039018630981445, + "learning_rate": 3.580246913580247e-06, + "loss": 1.5386, + "step": 1265 + }, + { + "epoch": 2.8086522462562398, + "grad_norm": 4.231089115142822, + "learning_rate": 3.5390946502057612e-06, + "loss": 1.2051, + "step": 1266 + }, + { + "epoch": 2.810870770937327, + "grad_norm": 3.9101765155792236, + "learning_rate": 3.497942386831276e-06, + "loss": 1.4894, + "step": 1267 + }, + { + "epoch": 2.813089295618414, + "grad_norm": 4.410416603088379, + "learning_rate": 3.45679012345679e-06, + "loss": 1.4443, + "step": 1268 + }, + { + "epoch": 2.815307820299501, + "grad_norm": 3.7927825450897217, + "learning_rate": 3.4156378600823046e-06, + "loss": 1.4446, + "step": 1269 + }, + { + "epoch": 2.817526344980588, + "grad_norm": 3.9671130180358887, + "learning_rate": 3.3744855967078196e-06, + "loss": 1.3229, + "step": 1270 + }, + { + "epoch": 2.819744869661675, + "grad_norm": 3.8313026428222656, + "learning_rate": 3.3333333333333333e-06, + "loss": 1.4194, + "step": 1271 + }, + { + "epoch": 2.821963394342762, + "grad_norm": 4.051031589508057, + "learning_rate": 3.2921810699588483e-06, + "loss": 1.4894, + "step": 1272 + }, + { + "epoch": 2.8241819190238493, + "grad_norm": 4.0446085929870605, + "learning_rate": 3.251028806584362e-06, + "loss": 1.3977, + "step": 1273 + }, + { + "epoch": 2.8264004437049364, + "grad_norm": 3.634438991546631, + "learning_rate": 3.209876543209877e-06, + "loss": 1.3085, + "step": 1274 + }, + { + "epoch": 2.8286189683860234, + "grad_norm": 3.8940911293029785, + "learning_rate": 3.168724279835391e-06, + "loss": 1.5015, + "step": 1275 + }, + { + "epoch": 2.8308374930671105, + "grad_norm": 3.813113212585449, + "learning_rate": 3.127572016460906e-06, + "loss": 1.3032, + "step": 1276 + }, + { + "epoch": 2.8330560177481976, + "grad_norm": 3.8498361110687256, + "learning_rate": 3.0864197530864196e-06, + "loss": 1.5348, + "step": 1277 + }, + { + "epoch": 2.8352745424292847, + "grad_norm": 4.3953857421875, + "learning_rate": 3.0452674897119346e-06, + "loss": 1.2791, + "step": 1278 + }, + { + "epoch": 2.8374930671103717, + "grad_norm": 3.916227102279663, + "learning_rate": 3.0041152263374487e-06, + "loss": 1.3214, + "step": 1279 + }, + { + "epoch": 2.839711591791459, + "grad_norm": 3.862987518310547, + "learning_rate": 2.9629629629629633e-06, + "loss": 1.3262, + "step": 1280 + }, + { + "epoch": 2.841930116472546, + "grad_norm": 3.088654041290283, + "learning_rate": 2.9218106995884775e-06, + "loss": 1.2638, + "step": 1281 + }, + { + "epoch": 2.844148641153633, + "grad_norm": 3.841285467147827, + "learning_rate": 2.880658436213992e-06, + "loss": 1.5635, + "step": 1282 + }, + { + "epoch": 2.84636716583472, + "grad_norm": 3.6944167613983154, + "learning_rate": 2.8395061728395062e-06, + "loss": 1.2573, + "step": 1283 + }, + { + "epoch": 2.848585690515807, + "grad_norm": 4.018518447875977, + "learning_rate": 2.798353909465021e-06, + "loss": 1.3501, + "step": 1284 + }, + { + "epoch": 2.850804215196894, + "grad_norm": 3.793536424636841, + "learning_rate": 2.757201646090535e-06, + "loss": 1.2802, + "step": 1285 + }, + { + "epoch": 2.8530227398779813, + "grad_norm": 3.6093437671661377, + "learning_rate": 2.7160493827160496e-06, + "loss": 1.3591, + "step": 1286 + }, + { + "epoch": 2.8552412645590683, + "grad_norm": 3.442178726196289, + "learning_rate": 2.6748971193415637e-06, + "loss": 1.0028, + "step": 1287 + }, + { + "epoch": 2.8574597892401554, + "grad_norm": 3.800060749053955, + "learning_rate": 2.6337448559670783e-06, + "loss": 1.5151, + "step": 1288 + }, + { + "epoch": 2.8596783139212425, + "grad_norm": 4.474183559417725, + "learning_rate": 2.5925925925925925e-06, + "loss": 1.3412, + "step": 1289 + }, + { + "epoch": 2.8618968386023296, + "grad_norm": 3.6142868995666504, + "learning_rate": 2.551440329218107e-06, + "loss": 1.3159, + "step": 1290 + }, + { + "epoch": 2.8641153632834166, + "grad_norm": 3.766481399536133, + "learning_rate": 2.5102880658436212e-06, + "loss": 1.3485, + "step": 1291 + }, + { + "epoch": 2.8663338879645037, + "grad_norm": 3.918968677520752, + "learning_rate": 2.469135802469136e-06, + "loss": 1.1533, + "step": 1292 + }, + { + "epoch": 2.868552412645591, + "grad_norm": 4.321674346923828, + "learning_rate": 2.42798353909465e-06, + "loss": 1.5342, + "step": 1293 + }, + { + "epoch": 2.870770937326678, + "grad_norm": 3.647822618484497, + "learning_rate": 2.386831275720165e-06, + "loss": 1.6103, + "step": 1294 + }, + { + "epoch": 2.872989462007765, + "grad_norm": 4.214388370513916, + "learning_rate": 2.345679012345679e-06, + "loss": 1.5447, + "step": 1295 + }, + { + "epoch": 2.875207986688852, + "grad_norm": 3.4597132205963135, + "learning_rate": 2.3045267489711937e-06, + "loss": 1.4795, + "step": 1296 + }, + { + "epoch": 2.877426511369939, + "grad_norm": 3.623835563659668, + "learning_rate": 2.263374485596708e-06, + "loss": 1.5147, + "step": 1297 + }, + { + "epoch": 2.879645036051026, + "grad_norm": 3.927180051803589, + "learning_rate": 2.2222222222222225e-06, + "loss": 1.4242, + "step": 1298 + }, + { + "epoch": 2.8818635607321132, + "grad_norm": 3.7198712825775146, + "learning_rate": 2.1810699588477367e-06, + "loss": 1.579, + "step": 1299 + }, + { + "epoch": 2.8840820854132003, + "grad_norm": 3.6406099796295166, + "learning_rate": 2.1399176954732512e-06, + "loss": 1.1864, + "step": 1300 + }, + { + "epoch": 2.8863006100942874, + "grad_norm": 3.9649851322174072, + "learning_rate": 2.0987654320987654e-06, + "loss": 1.3111, + "step": 1301 + }, + { + "epoch": 2.8885191347753745, + "grad_norm": 3.9488768577575684, + "learning_rate": 2.05761316872428e-06, + "loss": 1.2698, + "step": 1302 + }, + { + "epoch": 2.8907376594564616, + "grad_norm": 3.4466288089752197, + "learning_rate": 2.016460905349794e-06, + "loss": 1.2526, + "step": 1303 + }, + { + "epoch": 2.8929561841375486, + "grad_norm": 3.979374885559082, + "learning_rate": 1.9753086419753087e-06, + "loss": 1.3333, + "step": 1304 + }, + { + "epoch": 2.8951747088186357, + "grad_norm": 3.6995716094970703, + "learning_rate": 1.934156378600823e-06, + "loss": 1.1513, + "step": 1305 + }, + { + "epoch": 2.8973932334997228, + "grad_norm": 3.46440052986145, + "learning_rate": 1.8930041152263375e-06, + "loss": 1.3233, + "step": 1306 + }, + { + "epoch": 2.89961175818081, + "grad_norm": 4.052750110626221, + "learning_rate": 1.8518518518518519e-06, + "loss": 1.3554, + "step": 1307 + }, + { + "epoch": 2.901830282861897, + "grad_norm": 3.550729990005493, + "learning_rate": 1.8106995884773662e-06, + "loss": 1.5419, + "step": 1308 + }, + { + "epoch": 2.904048807542984, + "grad_norm": 3.513901472091675, + "learning_rate": 1.7695473251028806e-06, + "loss": 1.3519, + "step": 1309 + }, + { + "epoch": 2.906267332224071, + "grad_norm": 3.817808151245117, + "learning_rate": 1.728395061728395e-06, + "loss": 1.0428, + "step": 1310 + }, + { + "epoch": 2.908485856905158, + "grad_norm": 3.4584131240844727, + "learning_rate": 1.6872427983539098e-06, + "loss": 1.4707, + "step": 1311 + }, + { + "epoch": 2.9107043815862452, + "grad_norm": 3.7010762691497803, + "learning_rate": 1.6460905349794242e-06, + "loss": 1.486, + "step": 1312 + }, + { + "epoch": 2.9129229062673323, + "grad_norm": 3.9752590656280518, + "learning_rate": 1.6049382716049385e-06, + "loss": 0.9255, + "step": 1313 + }, + { + "epoch": 2.9151414309484194, + "grad_norm": 3.8167366981506348, + "learning_rate": 1.563786008230453e-06, + "loss": 1.2917, + "step": 1314 + }, + { + "epoch": 2.9173599556295065, + "grad_norm": 3.7454349994659424, + "learning_rate": 1.5226337448559673e-06, + "loss": 1.3005, + "step": 1315 + }, + { + "epoch": 2.9195784803105935, + "grad_norm": 3.719059705734253, + "learning_rate": 1.4814814814814817e-06, + "loss": 1.5923, + "step": 1316 + }, + { + "epoch": 2.9217970049916806, + "grad_norm": 3.692495584487915, + "learning_rate": 1.440329218106996e-06, + "loss": 1.3219, + "step": 1317 + }, + { + "epoch": 2.9240155296727677, + "grad_norm": 4.059812068939209, + "learning_rate": 1.3991769547325104e-06, + "loss": 1.5102, + "step": 1318 + }, + { + "epoch": 2.9262340543538548, + "grad_norm": 3.960024356842041, + "learning_rate": 1.3580246913580248e-06, + "loss": 1.4501, + "step": 1319 + }, + { + "epoch": 2.928452579034942, + "grad_norm": 3.66017746925354, + "learning_rate": 1.3168724279835392e-06, + "loss": 1.4714, + "step": 1320 + }, + { + "epoch": 2.930671103716029, + "grad_norm": 3.466686248779297, + "learning_rate": 1.2757201646090535e-06, + "loss": 1.4209, + "step": 1321 + }, + { + "epoch": 2.932889628397116, + "grad_norm": 3.987457752227783, + "learning_rate": 1.234567901234568e-06, + "loss": 1.5396, + "step": 1322 + }, + { + "epoch": 2.935108153078203, + "grad_norm": 3.916424036026001, + "learning_rate": 1.1934156378600825e-06, + "loss": 1.2398, + "step": 1323 + }, + { + "epoch": 2.93732667775929, + "grad_norm": 3.8283638954162598, + "learning_rate": 1.1522633744855969e-06, + "loss": 1.2576, + "step": 1324 + }, + { + "epoch": 2.939545202440377, + "grad_norm": 3.9138474464416504, + "learning_rate": 1.1111111111111112e-06, + "loss": 1.3282, + "step": 1325 + }, + { + "epoch": 2.9417637271214643, + "grad_norm": 4.006441116333008, + "learning_rate": 1.0699588477366256e-06, + "loss": 1.3584, + "step": 1326 + }, + { + "epoch": 2.9439822518025514, + "grad_norm": 4.00734806060791, + "learning_rate": 1.02880658436214e-06, + "loss": 1.3813, + "step": 1327 + }, + { + "epoch": 2.9462007764836384, + "grad_norm": 3.838003396987915, + "learning_rate": 9.876543209876544e-07, + "loss": 1.6119, + "step": 1328 + }, + { + "epoch": 2.9484193011647255, + "grad_norm": 3.9395203590393066, + "learning_rate": 9.465020576131687e-07, + "loss": 1.1217, + "step": 1329 + }, + { + "epoch": 2.9506378258458126, + "grad_norm": 4.072542667388916, + "learning_rate": 9.053497942386831e-07, + "loss": 1.3267, + "step": 1330 + }, + { + "epoch": 2.9528563505268997, + "grad_norm": 3.7510476112365723, + "learning_rate": 8.641975308641975e-07, + "loss": 1.3094, + "step": 1331 + }, + { + "epoch": 2.9550748752079867, + "grad_norm": 3.277667760848999, + "learning_rate": 8.230452674897121e-07, + "loss": 1.5148, + "step": 1332 + }, + { + "epoch": 2.957293399889074, + "grad_norm": 4.075628757476807, + "learning_rate": 7.818930041152265e-07, + "loss": 1.1553, + "step": 1333 + }, + { + "epoch": 2.959511924570161, + "grad_norm": 4.107234477996826, + "learning_rate": 7.407407407407408e-07, + "loss": 1.2822, + "step": 1334 + }, + { + "epoch": 2.961730449251248, + "grad_norm": 3.8605690002441406, + "learning_rate": 6.995884773662552e-07, + "loss": 1.5609, + "step": 1335 + }, + { + "epoch": 2.963948973932335, + "grad_norm": 3.956441640853882, + "learning_rate": 6.584362139917696e-07, + "loss": 1.4206, + "step": 1336 + }, + { + "epoch": 2.966167498613422, + "grad_norm": 4.025729656219482, + "learning_rate": 6.17283950617284e-07, + "loss": 1.3708, + "step": 1337 + }, + { + "epoch": 2.968386023294509, + "grad_norm": 3.900634288787842, + "learning_rate": 5.761316872427984e-07, + "loss": 1.4221, + "step": 1338 + }, + { + "epoch": 2.9706045479755963, + "grad_norm": 3.565924644470215, + "learning_rate": 5.349794238683128e-07, + "loss": 1.2141, + "step": 1339 + }, + { + "epoch": 2.9728230726566833, + "grad_norm": 4.025356769561768, + "learning_rate": 4.938271604938272e-07, + "loss": 1.548, + "step": 1340 + }, + { + "epoch": 2.9750415973377704, + "grad_norm": 3.902052879333496, + "learning_rate": 4.5267489711934156e-07, + "loss": 1.3124, + "step": 1341 + }, + { + "epoch": 2.9772601220188575, + "grad_norm": 4.311855316162109, + "learning_rate": 4.1152263374485604e-07, + "loss": 1.2729, + "step": 1342 + }, + { + "epoch": 2.9794786466999446, + "grad_norm": 3.6897168159484863, + "learning_rate": 3.703703703703704e-07, + "loss": 1.3793, + "step": 1343 + }, + { + "epoch": 2.9816971713810316, + "grad_norm": 4.480618953704834, + "learning_rate": 3.292181069958848e-07, + "loss": 1.3326, + "step": 1344 + }, + { + "epoch": 2.9839156960621187, + "grad_norm": 3.3348441123962402, + "learning_rate": 2.880658436213992e-07, + "loss": 1.7949, + "step": 1345 + }, + { + "epoch": 2.986134220743206, + "grad_norm": 4.433835029602051, + "learning_rate": 2.469135802469136e-07, + "loss": 1.4746, + "step": 1346 + }, + { + "epoch": 2.988352745424293, + "grad_norm": 3.7093398571014404, + "learning_rate": 2.0576131687242802e-07, + "loss": 1.2032, + "step": 1347 + }, + { + "epoch": 2.99057127010538, + "grad_norm": 4.107085704803467, + "learning_rate": 1.646090534979424e-07, + "loss": 1.4183, + "step": 1348 + }, + { + "epoch": 2.992789794786467, + "grad_norm": 3.98372483253479, + "learning_rate": 1.234567901234568e-07, + "loss": 1.4408, + "step": 1349 + }, + { + "epoch": 2.995008319467554, + "grad_norm": 3.7591214179992676, + "learning_rate": 8.23045267489712e-08, + "loss": 1.2775, + "step": 1350 + }, + { + "epoch": 2.995008319467554, + "step": 1350, + "total_flos": 6.954972515663872e+16, + "train_loss": 2.5400382919664737, + "train_runtime": 2568.8907, + "train_samples_per_second": 8.421, + "train_steps_per_second": 0.526 + } + ], + "logging_steps": 1.0, + "max_steps": 1350, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.954972515663872e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}