diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,8388 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 500, + "global_step": 1190, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016806722689075631, + "grad_norm": 8.209144697154139, + "learning_rate": 9.99998257609161e-06, + "loss": 0.2445, + "step": 1 + }, + { + "epoch": 0.0033613445378151263, + "grad_norm": 7.387136868109575, + "learning_rate": 9.999930304487874e-06, + "loss": 0.232, + "step": 2 + }, + { + "epoch": 0.005042016806722689, + "grad_norm": 4.703733759728634, + "learning_rate": 9.999843185553106e-06, + "loss": 0.1499, + "step": 3 + }, + { + "epoch": 0.0067226890756302525, + "grad_norm": 4.59672828172244, + "learning_rate": 9.999721219894482e-06, + "loss": 0.1312, + "step": 4 + }, + { + "epoch": 0.008403361344537815, + "grad_norm": 7.869163994125804, + "learning_rate": 9.999564408362054e-06, + "loss": 0.2678, + "step": 5 + }, + { + "epoch": 0.010084033613445379, + "grad_norm": 5.069247100993632, + "learning_rate": 9.999372752048729e-06, + "loss": 0.1643, + "step": 6 + }, + { + "epoch": 0.011764705882352941, + "grad_norm": 4.971040093251534, + "learning_rate": 9.999146252290264e-06, + "loss": 0.1881, + "step": 7 + }, + { + "epoch": 0.013445378151260505, + "grad_norm": 4.612745404018306, + "learning_rate": 9.998884910665267e-06, + "loss": 0.2474, + "step": 8 + }, + { + "epoch": 0.015126050420168067, + "grad_norm": 5.582880543106663, + "learning_rate": 9.998588728995176e-06, + "loss": 0.2341, + "step": 9 + }, + { + "epoch": 0.01680672268907563, + "grad_norm": 5.10777402267266, + "learning_rate": 9.998257709344246e-06, + "loss": 0.2462, + "step": 10 + }, + { + "epoch": 0.018487394957983194, + "grad_norm": 3.912332856239327, + "learning_rate": 9.997891854019538e-06, + "loss": 0.2171, + "step": 11 + }, + { + "epoch": 0.020168067226890758, + "grad_norm": 4.514978773458917, + "learning_rate": 9.997491165570907e-06, + "loss": 0.2079, + "step": 12 + }, + { + "epoch": 0.021848739495798318, + "grad_norm": 4.982439292936562, + "learning_rate": 9.997055646790974e-06, + "loss": 0.2801, + "step": 13 + }, + { + "epoch": 0.023529411764705882, + "grad_norm": 6.199005044554472, + "learning_rate": 9.996585300715117e-06, + "loss": 0.2362, + "step": 14 + }, + { + "epoch": 0.025210084033613446, + "grad_norm": 4.2318843117354, + "learning_rate": 9.99608013062144e-06, + "loss": 0.2437, + "step": 15 + }, + { + "epoch": 0.02689075630252101, + "grad_norm": 4.678511423998445, + "learning_rate": 9.995540140030759e-06, + "loss": 0.2413, + "step": 16 + }, + { + "epoch": 0.02857142857142857, + "grad_norm": 5.019790515416029, + "learning_rate": 9.994965332706574e-06, + "loss": 0.2763, + "step": 17 + }, + { + "epoch": 0.030252100840336135, + "grad_norm": 3.7745510637286572, + "learning_rate": 9.99435571265504e-06, + "loss": 0.2413, + "step": 18 + }, + { + "epoch": 0.031932773109243695, + "grad_norm": 4.954591562450657, + "learning_rate": 9.993711284124943e-06, + "loss": 0.2748, + "step": 19 + }, + { + "epoch": 0.03361344537815126, + "grad_norm": 4.187965473046551, + "learning_rate": 9.99303205160767e-06, + "loss": 0.274, + "step": 20 + }, + { + "epoch": 0.03529411764705882, + "grad_norm": 4.6295824547220095, + "learning_rate": 9.992318019837171e-06, + "loss": 0.2448, + "step": 21 + }, + { + "epoch": 0.03697478991596639, + "grad_norm": 5.621884903621694, + "learning_rate": 9.991569193789938e-06, + "loss": 0.2579, + "step": 22 + }, + { + "epoch": 0.03865546218487395, + "grad_norm": 5.123406042716482, + "learning_rate": 9.990785578684963e-06, + "loss": 0.2391, + "step": 23 + }, + { + "epoch": 0.040336134453781515, + "grad_norm": 5.074868330604315, + "learning_rate": 9.989967179983699e-06, + "loss": 0.2584, + "step": 24 + }, + { + "epoch": 0.04201680672268908, + "grad_norm": 5.376525151914496, + "learning_rate": 9.989114003390028e-06, + "loss": 0.277, + "step": 25 + }, + { + "epoch": 0.043697478991596636, + "grad_norm": 4.0946093021497845, + "learning_rate": 9.988226054850218e-06, + "loss": 0.215, + "step": 26 + }, + { + "epoch": 0.0453781512605042, + "grad_norm": 4.552724543313885, + "learning_rate": 9.987303340552885e-06, + "loss": 0.2744, + "step": 27 + }, + { + "epoch": 0.047058823529411764, + "grad_norm": 5.012152565084418, + "learning_rate": 9.98634586692894e-06, + "loss": 0.385, + "step": 28 + }, + { + "epoch": 0.04873949579831933, + "grad_norm": 5.423054570055711, + "learning_rate": 9.985353640651563e-06, + "loss": 0.2702, + "step": 29 + }, + { + "epoch": 0.05042016806722689, + "grad_norm": 5.127889256940364, + "learning_rate": 9.984326668636131e-06, + "loss": 0.3168, + "step": 30 + }, + { + "epoch": 0.052100840336134456, + "grad_norm": 3.679450662065014, + "learning_rate": 9.983264958040194e-06, + "loss": 0.2874, + "step": 31 + }, + { + "epoch": 0.05378151260504202, + "grad_norm": 3.3004814033041585, + "learning_rate": 9.98216851626341e-06, + "loss": 0.2216, + "step": 32 + }, + { + "epoch": 0.05546218487394958, + "grad_norm": 3.6744435985276502, + "learning_rate": 9.981037350947503e-06, + "loss": 0.1946, + "step": 33 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 4.945817900667547, + "learning_rate": 9.979871469976197e-06, + "loss": 0.2352, + "step": 34 + }, + { + "epoch": 0.058823529411764705, + "grad_norm": 4.199740458372265, + "learning_rate": 9.978670881475173e-06, + "loss": 0.2658, + "step": 35 + }, + { + "epoch": 0.06050420168067227, + "grad_norm": 4.098117482797779, + "learning_rate": 9.977435593812013e-06, + "loss": 0.2907, + "step": 36 + }, + { + "epoch": 0.06218487394957983, + "grad_norm": 4.042885470534792, + "learning_rate": 9.976165615596128e-06, + "loss": 0.2, + "step": 37 + }, + { + "epoch": 0.06386554621848739, + "grad_norm": 4.617103574848525, + "learning_rate": 9.974860955678715e-06, + "loss": 0.2306, + "step": 38 + }, + { + "epoch": 0.06554621848739496, + "grad_norm": 4.4388689219941675, + "learning_rate": 9.973521623152682e-06, + "loss": 0.2407, + "step": 39 + }, + { + "epoch": 0.06722689075630252, + "grad_norm": 6.477467686130754, + "learning_rate": 9.972147627352593e-06, + "loss": 0.3914, + "step": 40 + }, + { + "epoch": 0.06890756302521009, + "grad_norm": 4.607889040112, + "learning_rate": 9.970738977854597e-06, + "loss": 0.2606, + "step": 41 + }, + { + "epoch": 0.07058823529411765, + "grad_norm": 5.8414772767780665, + "learning_rate": 9.96929568447637e-06, + "loss": 0.2878, + "step": 42 + }, + { + "epoch": 0.07226890756302522, + "grad_norm": 4.530793945019326, + "learning_rate": 9.967817757277031e-06, + "loss": 0.222, + "step": 43 + }, + { + "epoch": 0.07394957983193277, + "grad_norm": 5.7019752450601295, + "learning_rate": 9.966305206557092e-06, + "loss": 0.3215, + "step": 44 + }, + { + "epoch": 0.07563025210084033, + "grad_norm": 4.976614437008921, + "learning_rate": 9.964758042858368e-06, + "loss": 0.3344, + "step": 45 + }, + { + "epoch": 0.0773109243697479, + "grad_norm": 4.411005767374718, + "learning_rate": 9.963176276963916e-06, + "loss": 0.2107, + "step": 46 + }, + { + "epoch": 0.07899159663865546, + "grad_norm": 5.409158327460203, + "learning_rate": 9.961559919897954e-06, + "loss": 0.2534, + "step": 47 + }, + { + "epoch": 0.08067226890756303, + "grad_norm": 4.484574670660062, + "learning_rate": 9.959908982925783e-06, + "loss": 0.2586, + "step": 48 + }, + { + "epoch": 0.08235294117647059, + "grad_norm": 4.587259020764486, + "learning_rate": 9.958223477553715e-06, + "loss": 0.2533, + "step": 49 + }, + { + "epoch": 0.08403361344537816, + "grad_norm": 4.405358710771115, + "learning_rate": 9.956503415528984e-06, + "loss": 0.2559, + "step": 50 + }, + { + "epoch": 0.08571428571428572, + "grad_norm": 4.967732320368905, + "learning_rate": 9.954748808839675e-06, + "loss": 0.2503, + "step": 51 + }, + { + "epoch": 0.08739495798319327, + "grad_norm": 5.218196839992478, + "learning_rate": 9.952959669714627e-06, + "loss": 0.2862, + "step": 52 + }, + { + "epoch": 0.08907563025210084, + "grad_norm": 4.8935636677407475, + "learning_rate": 9.951136010623359e-06, + "loss": 0.3151, + "step": 53 + }, + { + "epoch": 0.0907563025210084, + "grad_norm": 3.9500115872361072, + "learning_rate": 9.94927784427598e-06, + "loss": 0.2711, + "step": 54 + }, + { + "epoch": 0.09243697478991597, + "grad_norm": 4.817134300002951, + "learning_rate": 9.947385183623099e-06, + "loss": 0.2697, + "step": 55 + }, + { + "epoch": 0.09411764705882353, + "grad_norm": 5.486354681435603, + "learning_rate": 9.945458041855732e-06, + "loss": 0.3244, + "step": 56 + }, + { + "epoch": 0.0957983193277311, + "grad_norm": 3.924865124104986, + "learning_rate": 9.943496432405213e-06, + "loss": 0.2126, + "step": 57 + }, + { + "epoch": 0.09747899159663866, + "grad_norm": 4.059739131992801, + "learning_rate": 9.941500368943111e-06, + "loss": 0.2209, + "step": 58 + }, + { + "epoch": 0.09915966386554621, + "grad_norm": 3.925936844915292, + "learning_rate": 9.939469865381111e-06, + "loss": 0.2238, + "step": 59 + }, + { + "epoch": 0.10084033613445378, + "grad_norm": 4.929638111711791, + "learning_rate": 9.937404935870938e-06, + "loss": 0.2876, + "step": 60 + }, + { + "epoch": 0.10252100840336134, + "grad_norm": 4.333643541029462, + "learning_rate": 9.935305594804247e-06, + "loss": 0.2622, + "step": 61 + }, + { + "epoch": 0.10420168067226891, + "grad_norm": 6.551835764229068, + "learning_rate": 9.933171856812533e-06, + "loss": 0.3608, + "step": 62 + }, + { + "epoch": 0.10588235294117647, + "grad_norm": 4.370528317389317, + "learning_rate": 9.931003736767013e-06, + "loss": 0.2189, + "step": 63 + }, + { + "epoch": 0.10756302521008404, + "grad_norm": 3.9034215243844557, + "learning_rate": 9.92880124977854e-06, + "loss": 0.2076, + "step": 64 + }, + { + "epoch": 0.1092436974789916, + "grad_norm": 6.462115529099049, + "learning_rate": 9.926564411197488e-06, + "loss": 0.2704, + "step": 65 + }, + { + "epoch": 0.11092436974789915, + "grad_norm": 4.974959833762651, + "learning_rate": 9.924293236613643e-06, + "loss": 0.2586, + "step": 66 + }, + { + "epoch": 0.11260504201680673, + "grad_norm": 5.397773235075291, + "learning_rate": 9.921987741856099e-06, + "loss": 0.3708, + "step": 67 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 4.290267842379414, + "learning_rate": 9.91964794299315e-06, + "loss": 0.297, + "step": 68 + }, + { + "epoch": 0.11596638655462185, + "grad_norm": 4.297436061861156, + "learning_rate": 9.91727385633217e-06, + "loss": 0.2723, + "step": 69 + }, + { + "epoch": 0.11764705882352941, + "grad_norm": 3.6926079145905097, + "learning_rate": 9.91486549841951e-06, + "loss": 0.2431, + "step": 70 + }, + { + "epoch": 0.11932773109243698, + "grad_norm": 4.305921925067097, + "learning_rate": 9.91242288604037e-06, + "loss": 0.2779, + "step": 71 + }, + { + "epoch": 0.12100840336134454, + "grad_norm": 4.517120872670056, + "learning_rate": 9.909946036218694e-06, + "loss": 0.2961, + "step": 72 + }, + { + "epoch": 0.1226890756302521, + "grad_norm": 4.785215176674839, + "learning_rate": 9.907434966217041e-06, + "loss": 0.2819, + "step": 73 + }, + { + "epoch": 0.12436974789915967, + "grad_norm": 4.214559462932771, + "learning_rate": 9.904889693536475e-06, + "loss": 0.2109, + "step": 74 + }, + { + "epoch": 0.12605042016806722, + "grad_norm": 4.389318542382428, + "learning_rate": 9.902310235916435e-06, + "loss": 0.2522, + "step": 75 + }, + { + "epoch": 0.12773109243697478, + "grad_norm": 3.8325771771754664, + "learning_rate": 9.899696611334612e-06, + "loss": 0.2598, + "step": 76 + }, + { + "epoch": 0.12941176470588237, + "grad_norm": 4.860313041028263, + "learning_rate": 9.89704883800683e-06, + "loss": 0.3235, + "step": 77 + }, + { + "epoch": 0.13109243697478992, + "grad_norm": 4.910636980849552, + "learning_rate": 9.894366934386913e-06, + "loss": 0.2905, + "step": 78 + }, + { + "epoch": 0.13277310924369748, + "grad_norm": 5.308926346498934, + "learning_rate": 9.891650919166558e-06, + "loss": 0.3174, + "step": 79 + }, + { + "epoch": 0.13445378151260504, + "grad_norm": 5.932647263725452, + "learning_rate": 9.888900811275205e-06, + "loss": 0.3704, + "step": 80 + }, + { + "epoch": 0.1361344537815126, + "grad_norm": 4.809268256405562, + "learning_rate": 9.886116629879906e-06, + "loss": 0.3038, + "step": 81 + }, + { + "epoch": 0.13781512605042018, + "grad_norm": 4.03544501605081, + "learning_rate": 9.883298394385186e-06, + "loss": 0.262, + "step": 82 + }, + { + "epoch": 0.13949579831932774, + "grad_norm": 4.712835324738283, + "learning_rate": 9.880446124432921e-06, + "loss": 0.2692, + "step": 83 + }, + { + "epoch": 0.1411764705882353, + "grad_norm": 4.563081327265746, + "learning_rate": 9.877559839902185e-06, + "loss": 0.3518, + "step": 84 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 4.028984316421756, + "learning_rate": 9.874639560909118e-06, + "loss": 0.277, + "step": 85 + }, + { + "epoch": 0.14453781512605043, + "grad_norm": 4.3533769418599055, + "learning_rate": 9.871685307806796e-06, + "loss": 0.2605, + "step": 86 + }, + { + "epoch": 0.146218487394958, + "grad_norm": 4.32320216003599, + "learning_rate": 9.868697101185066e-06, + "loss": 0.2483, + "step": 87 + }, + { + "epoch": 0.14789915966386555, + "grad_norm": 4.252161322978531, + "learning_rate": 9.865674961870428e-06, + "loss": 0.2292, + "step": 88 + }, + { + "epoch": 0.1495798319327731, + "grad_norm": 3.7472917316130707, + "learning_rate": 9.862618910925873e-06, + "loss": 0.2314, + "step": 89 + }, + { + "epoch": 0.15126050420168066, + "grad_norm": 3.4288032918623017, + "learning_rate": 9.859528969650739e-06, + "loss": 0.2511, + "step": 90 + }, + { + "epoch": 0.15294117647058825, + "grad_norm": 4.6329004322037335, + "learning_rate": 9.85640515958057e-06, + "loss": 0.2949, + "step": 91 + }, + { + "epoch": 0.1546218487394958, + "grad_norm": 3.5652517822054794, + "learning_rate": 9.853247502486957e-06, + "loss": 0.2423, + "step": 92 + }, + { + "epoch": 0.15630252100840336, + "grad_norm": 5.123085560371222, + "learning_rate": 9.850056020377392e-06, + "loss": 0.326, + "step": 93 + }, + { + "epoch": 0.15798319327731092, + "grad_norm": 4.810628133927044, + "learning_rate": 9.846830735495112e-06, + "loss": 0.2907, + "step": 94 + }, + { + "epoch": 0.15966386554621848, + "grad_norm": 4.024916887888118, + "learning_rate": 9.843571670318943e-06, + "loss": 0.3123, + "step": 95 + }, + { + "epoch": 0.16134453781512606, + "grad_norm": 7.985561798082157, + "learning_rate": 9.840278847563147e-06, + "loss": 0.3295, + "step": 96 + }, + { + "epoch": 0.16302521008403362, + "grad_norm": 3.7710475419246254, + "learning_rate": 9.836952290177261e-06, + "loss": 0.2212, + "step": 97 + }, + { + "epoch": 0.16470588235294117, + "grad_norm": 4.423657356806522, + "learning_rate": 9.833592021345938e-06, + "loss": 0.31, + "step": 98 + }, + { + "epoch": 0.16638655462184873, + "grad_norm": 4.222677175561096, + "learning_rate": 9.830198064488783e-06, + "loss": 0.2842, + "step": 99 + }, + { + "epoch": 0.16806722689075632, + "grad_norm": 4.168091860783304, + "learning_rate": 9.826770443260193e-06, + "loss": 0.252, + "step": 100 + }, + { + "epoch": 0.16974789915966387, + "grad_norm": 6.3620832971275085, + "learning_rate": 9.823309181549194e-06, + "loss": 0.2902, + "step": 101 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 6.17536019379638, + "learning_rate": 9.819814303479268e-06, + "loss": 0.2962, + "step": 102 + }, + { + "epoch": 0.173109243697479, + "grad_norm": 4.438145668562799, + "learning_rate": 9.816285833408185e-06, + "loss": 0.3056, + "step": 103 + }, + { + "epoch": 0.17478991596638654, + "grad_norm": 3.916870643757942, + "learning_rate": 9.812723795927848e-06, + "loss": 0.2413, + "step": 104 + }, + { + "epoch": 0.17647058823529413, + "grad_norm": 4.137088990144677, + "learning_rate": 9.809128215864096e-06, + "loss": 0.2926, + "step": 105 + }, + { + "epoch": 0.1781512605042017, + "grad_norm": 4.803813178194908, + "learning_rate": 9.805499118276555e-06, + "loss": 0.3213, + "step": 106 + }, + { + "epoch": 0.17983193277310924, + "grad_norm": 4.13461633288206, + "learning_rate": 9.801836528458453e-06, + "loss": 0.2581, + "step": 107 + }, + { + "epoch": 0.1815126050420168, + "grad_norm": 4.0431954077083, + "learning_rate": 9.798140471936437e-06, + "loss": 0.2011, + "step": 108 + }, + { + "epoch": 0.18319327731092436, + "grad_norm": 4.983771680315847, + "learning_rate": 9.79441097447041e-06, + "loss": 0.304, + "step": 109 + }, + { + "epoch": 0.18487394957983194, + "grad_norm": 4.001740838488305, + "learning_rate": 9.790648062053341e-06, + "loss": 0.3108, + "step": 110 + }, + { + "epoch": 0.1865546218487395, + "grad_norm": 4.286049724314065, + "learning_rate": 9.786851760911084e-06, + "loss": 0.222, + "step": 111 + }, + { + "epoch": 0.18823529411764706, + "grad_norm": 4.087795367081155, + "learning_rate": 9.783022097502204e-06, + "loss": 0.3072, + "step": 112 + }, + { + "epoch": 0.1899159663865546, + "grad_norm": 3.543885725592475, + "learning_rate": 9.779159098517781e-06, + "loss": 0.2562, + "step": 113 + }, + { + "epoch": 0.1915966386554622, + "grad_norm": 4.331258375709638, + "learning_rate": 9.77526279088123e-06, + "loss": 0.3, + "step": 114 + }, + { + "epoch": 0.19327731092436976, + "grad_norm": 5.020497063978092, + "learning_rate": 9.771333201748116e-06, + "loss": 0.2768, + "step": 115 + }, + { + "epoch": 0.1949579831932773, + "grad_norm": 3.3790105404739355, + "learning_rate": 9.767370358505958e-06, + "loss": 0.2275, + "step": 116 + }, + { + "epoch": 0.19663865546218487, + "grad_norm": 5.114492285734836, + "learning_rate": 9.763374288774043e-06, + "loss": 0.3194, + "step": 117 + }, + { + "epoch": 0.19831932773109243, + "grad_norm": 4.147457648794141, + "learning_rate": 9.759345020403233e-06, + "loss": 0.2384, + "step": 118 + }, + { + "epoch": 0.2, + "grad_norm": 4.002739168668194, + "learning_rate": 9.755282581475769e-06, + "loss": 0.2595, + "step": 119 + }, + { + "epoch": 0.20168067226890757, + "grad_norm": 3.665748728737787, + "learning_rate": 9.751187000305076e-06, + "loss": 0.2456, + "step": 120 + }, + { + "epoch": 0.20336134453781513, + "grad_norm": 4.317063513127341, + "learning_rate": 9.747058305435566e-06, + "loss": 0.3031, + "step": 121 + }, + { + "epoch": 0.20504201680672268, + "grad_norm": 4.491629737709359, + "learning_rate": 9.742896525642442e-06, + "loss": 0.3095, + "step": 122 + }, + { + "epoch": 0.20672268907563024, + "grad_norm": 4.5060463998958245, + "learning_rate": 9.738701689931488e-06, + "loss": 0.309, + "step": 123 + }, + { + "epoch": 0.20840336134453782, + "grad_norm": 3.8391374970872487, + "learning_rate": 9.734473827538881e-06, + "loss": 0.2701, + "step": 124 + }, + { + "epoch": 0.21008403361344538, + "grad_norm": 5.610803912721021, + "learning_rate": 9.730212967930974e-06, + "loss": 0.3006, + "step": 125 + }, + { + "epoch": 0.21176470588235294, + "grad_norm": 3.865567382675256, + "learning_rate": 9.7259191408041e-06, + "loss": 0.2494, + "step": 126 + }, + { + "epoch": 0.2134453781512605, + "grad_norm": 4.157071231774723, + "learning_rate": 9.721592376084355e-06, + "loss": 0.2929, + "step": 127 + }, + { + "epoch": 0.21512605042016808, + "grad_norm": 4.69641419523669, + "learning_rate": 9.717232703927402e-06, + "loss": 0.2922, + "step": 128 + }, + { + "epoch": 0.21680672268907564, + "grad_norm": 3.9492507253815137, + "learning_rate": 9.712840154718253e-06, + "loss": 0.2449, + "step": 129 + }, + { + "epoch": 0.2184873949579832, + "grad_norm": 4.438337881075212, + "learning_rate": 9.70841475907106e-06, + "loss": 0.2738, + "step": 130 + }, + { + "epoch": 0.22016806722689075, + "grad_norm": 4.515382778757333, + "learning_rate": 9.703956547828893e-06, + "loss": 0.264, + "step": 131 + }, + { + "epoch": 0.2218487394957983, + "grad_norm": 3.2827204648104185, + "learning_rate": 9.69946555206354e-06, + "loss": 0.2103, + "step": 132 + }, + { + "epoch": 0.2235294117647059, + "grad_norm": 6.877887492871786, + "learning_rate": 9.694941803075285e-06, + "loss": 0.4638, + "step": 133 + }, + { + "epoch": 0.22521008403361345, + "grad_norm": 3.826504365543723, + "learning_rate": 9.690385332392676e-06, + "loss": 0.2347, + "step": 134 + }, + { + "epoch": 0.226890756302521, + "grad_norm": 4.474707023359214, + "learning_rate": 9.685796171772327e-06, + "loss": 0.4061, + "step": 135 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 4.228391017498601, + "learning_rate": 9.681174353198687e-06, + "loss": 0.3177, + "step": 136 + }, + { + "epoch": 0.23025210084033612, + "grad_norm": 3.7515452786239196, + "learning_rate": 9.67651990888381e-06, + "loss": 0.3122, + "step": 137 + }, + { + "epoch": 0.2319327731092437, + "grad_norm": 3.8337555170576305, + "learning_rate": 9.67183287126714e-06, + "loss": 0.2236, + "step": 138 + }, + { + "epoch": 0.23361344537815126, + "grad_norm": 3.575500732595517, + "learning_rate": 9.667113273015283e-06, + "loss": 0.257, + "step": 139 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 4.960211439105124, + "learning_rate": 9.66236114702178e-06, + "loss": 0.3158, + "step": 140 + }, + { + "epoch": 0.23697478991596638, + "grad_norm": 3.7297597163549945, + "learning_rate": 9.657576526406872e-06, + "loss": 0.2612, + "step": 141 + }, + { + "epoch": 0.23865546218487396, + "grad_norm": 5.716132411582617, + "learning_rate": 9.652759444517276e-06, + "loss": 0.2984, + "step": 142 + }, + { + "epoch": 0.24033613445378152, + "grad_norm": 3.9938378467188556, + "learning_rate": 9.647909934925952e-06, + "loss": 0.2578, + "step": 143 + }, + { + "epoch": 0.24201680672268908, + "grad_norm": 4.768607552808638, + "learning_rate": 9.64302803143186e-06, + "loss": 0.3166, + "step": 144 + }, + { + "epoch": 0.24369747899159663, + "grad_norm": 4.36886602623328, + "learning_rate": 9.63811376805974e-06, + "loss": 0.2879, + "step": 145 + }, + { + "epoch": 0.2453781512605042, + "grad_norm": 5.159036155402597, + "learning_rate": 9.633167179059859e-06, + "loss": 0.3313, + "step": 146 + }, + { + "epoch": 0.24705882352941178, + "grad_norm": 3.9079551640082775, + "learning_rate": 9.628188298907782e-06, + "loss": 0.2307, + "step": 147 + }, + { + "epoch": 0.24873949579831933, + "grad_norm": 4.972815617641453, + "learning_rate": 9.623177162304132e-06, + "loss": 0.3627, + "step": 148 + }, + { + "epoch": 0.2504201680672269, + "grad_norm": 3.7182475542942637, + "learning_rate": 9.618133804174341e-06, + "loss": 0.2088, + "step": 149 + }, + { + "epoch": 0.25210084033613445, + "grad_norm": 5.105115930829285, + "learning_rate": 9.613058259668416e-06, + "loss": 0.3468, + "step": 150 + }, + { + "epoch": 0.253781512605042, + "grad_norm": 4.2243880410090515, + "learning_rate": 9.607950564160682e-06, + "loss": 0.278, + "step": 151 + }, + { + "epoch": 0.25546218487394956, + "grad_norm": 3.5788756892431417, + "learning_rate": 9.602810753249549e-06, + "loss": 0.2613, + "step": 152 + }, + { + "epoch": 0.2571428571428571, + "grad_norm": 3.6146625169924502, + "learning_rate": 9.597638862757255e-06, + "loss": 0.2197, + "step": 153 + }, + { + "epoch": 0.25882352941176473, + "grad_norm": 5.192851016189807, + "learning_rate": 9.592434928729617e-06, + "loss": 0.2943, + "step": 154 + }, + { + "epoch": 0.2605042016806723, + "grad_norm": 5.012984899286897, + "learning_rate": 9.587198987435782e-06, + "loss": 0.2742, + "step": 155 + }, + { + "epoch": 0.26218487394957984, + "grad_norm": 4.243030317874798, + "learning_rate": 9.581931075367979e-06, + "loss": 0.3106, + "step": 156 + }, + { + "epoch": 0.2638655462184874, + "grad_norm": 4.352999777089954, + "learning_rate": 9.576631229241248e-06, + "loss": 0.2602, + "step": 157 + }, + { + "epoch": 0.26554621848739496, + "grad_norm": 3.8746827574175016, + "learning_rate": 9.57129948599321e-06, + "loss": 0.2375, + "step": 158 + }, + { + "epoch": 0.2672268907563025, + "grad_norm": 4.370512006750166, + "learning_rate": 9.565935882783784e-06, + "loss": 0.3138, + "step": 159 + }, + { + "epoch": 0.2689075630252101, + "grad_norm": 4.530164031043821, + "learning_rate": 9.56054045699494e-06, + "loss": 0.2808, + "step": 160 + }, + { + "epoch": 0.27058823529411763, + "grad_norm": 4.240473901674803, + "learning_rate": 9.555113246230443e-06, + "loss": 0.2534, + "step": 161 + }, + { + "epoch": 0.2722689075630252, + "grad_norm": 4.518086654583603, + "learning_rate": 9.54965428831558e-06, + "loss": 0.338, + "step": 162 + }, + { + "epoch": 0.2739495798319328, + "grad_norm": 4.703674977488523, + "learning_rate": 9.544163621296906e-06, + "loss": 0.3264, + "step": 163 + }, + { + "epoch": 0.27563025210084036, + "grad_norm": 3.5692443136214327, + "learning_rate": 9.538641283441974e-06, + "loss": 0.3219, + "step": 164 + }, + { + "epoch": 0.2773109243697479, + "grad_norm": 3.6543200856527696, + "learning_rate": 9.533087313239065e-06, + "loss": 0.2259, + "step": 165 + }, + { + "epoch": 0.27899159663865547, + "grad_norm": 3.215902167038249, + "learning_rate": 9.527501749396924e-06, + "loss": 0.2681, + "step": 166 + }, + { + "epoch": 0.280672268907563, + "grad_norm": 3.63082759717829, + "learning_rate": 9.521884630844498e-06, + "loss": 0.3603, + "step": 167 + }, + { + "epoch": 0.2823529411764706, + "grad_norm": 4.395327561528559, + "learning_rate": 9.516235996730645e-06, + "loss": 0.3001, + "step": 168 + }, + { + "epoch": 0.28403361344537814, + "grad_norm": 3.402290374462376, + "learning_rate": 9.510555886423883e-06, + "loss": 0.2243, + "step": 169 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 4.229267629127821, + "learning_rate": 9.504844339512096e-06, + "loss": 0.2396, + "step": 170 + }, + { + "epoch": 0.28739495798319326, + "grad_norm": 5.626539496937702, + "learning_rate": 9.499101395802277e-06, + "loss": 0.3552, + "step": 171 + }, + { + "epoch": 0.28907563025210087, + "grad_norm": 4.261902788104125, + "learning_rate": 9.493327095320231e-06, + "loss": 0.2933, + "step": 172 + }, + { + "epoch": 0.2907563025210084, + "grad_norm": 4.315799314745412, + "learning_rate": 9.487521478310316e-06, + "loss": 0.337, + "step": 173 + }, + { + "epoch": 0.292436974789916, + "grad_norm": 3.2957622174585466, + "learning_rate": 9.481684585235145e-06, + "loss": 0.2044, + "step": 174 + }, + { + "epoch": 0.29411764705882354, + "grad_norm": 3.7840993226552264, + "learning_rate": 9.475816456775313e-06, + "loss": 0.355, + "step": 175 + }, + { + "epoch": 0.2957983193277311, + "grad_norm": 4.315595576291231, + "learning_rate": 9.469917133829114e-06, + "loss": 0.2866, + "step": 176 + }, + { + "epoch": 0.29747899159663865, + "grad_norm": 3.3039529791986375, + "learning_rate": 9.463986657512254e-06, + "loss": 0.2351, + "step": 177 + }, + { + "epoch": 0.2991596638655462, + "grad_norm": 3.501475550122351, + "learning_rate": 9.458025069157563e-06, + "loss": 0.2393, + "step": 178 + }, + { + "epoch": 0.30084033613445377, + "grad_norm": 2.96105228086886, + "learning_rate": 9.452032410314709e-06, + "loss": 0.1883, + "step": 179 + }, + { + "epoch": 0.3025210084033613, + "grad_norm": 2.680755297383461, + "learning_rate": 9.446008722749906e-06, + "loss": 0.1813, + "step": 180 + }, + { + "epoch": 0.3042016806722689, + "grad_norm": 4.321387391660858, + "learning_rate": 9.439954048445628e-06, + "loss": 0.2939, + "step": 181 + }, + { + "epoch": 0.3058823529411765, + "grad_norm": 3.0370917327024367, + "learning_rate": 9.43386842960031e-06, + "loss": 0.2264, + "step": 182 + }, + { + "epoch": 0.30756302521008405, + "grad_norm": 3.310352207725729, + "learning_rate": 9.427751908628059e-06, + "loss": 0.2324, + "step": 183 + }, + { + "epoch": 0.3092436974789916, + "grad_norm": 4.805046115785464, + "learning_rate": 9.421604528158355e-06, + "loss": 0.3299, + "step": 184 + }, + { + "epoch": 0.31092436974789917, + "grad_norm": 4.739559492527739, + "learning_rate": 9.415426331035754e-06, + "loss": 0.3299, + "step": 185 + }, + { + "epoch": 0.3126050420168067, + "grad_norm": 4.490880194826058, + "learning_rate": 9.409217360319594e-06, + "loss": 0.3312, + "step": 186 + }, + { + "epoch": 0.3142857142857143, + "grad_norm": 4.700636400248507, + "learning_rate": 9.40297765928369e-06, + "loss": 0.3269, + "step": 187 + }, + { + "epoch": 0.31596638655462184, + "grad_norm": 4.776516740186767, + "learning_rate": 9.396707271416035e-06, + "loss": 0.4301, + "step": 188 + }, + { + "epoch": 0.3176470588235294, + "grad_norm": 4.1685832195956465, + "learning_rate": 9.39040624041849e-06, + "loss": 0.2683, + "step": 189 + }, + { + "epoch": 0.31932773109243695, + "grad_norm": 3.746172051394499, + "learning_rate": 9.384074610206495e-06, + "loss": 0.2646, + "step": 190 + }, + { + "epoch": 0.32100840336134456, + "grad_norm": 4.379779974988923, + "learning_rate": 9.377712424908743e-06, + "loss": 0.2967, + "step": 191 + }, + { + "epoch": 0.3226890756302521, + "grad_norm": 4.319433814312415, + "learning_rate": 9.371319728866892e-06, + "loss": 0.3311, + "step": 192 + }, + { + "epoch": 0.3243697478991597, + "grad_norm": 4.861684100170174, + "learning_rate": 9.36489656663524e-06, + "loss": 0.354, + "step": 193 + }, + { + "epoch": 0.32605042016806723, + "grad_norm": 4.50859409722337, + "learning_rate": 9.35844298298042e-06, + "loss": 0.3593, + "step": 194 + }, + { + "epoch": 0.3277310924369748, + "grad_norm": 3.5195246430906493, + "learning_rate": 9.351959022881098e-06, + "loss": 0.2799, + "step": 195 + }, + { + "epoch": 0.32941176470588235, + "grad_norm": 4.0869900443685045, + "learning_rate": 9.345444731527642e-06, + "loss": 0.2529, + "step": 196 + }, + { + "epoch": 0.3310924369747899, + "grad_norm": 3.546421899021016, + "learning_rate": 9.338900154321818e-06, + "loss": 0.2463, + "step": 197 + }, + { + "epoch": 0.33277310924369746, + "grad_norm": 3.9484095318347627, + "learning_rate": 9.332325336876472e-06, + "loss": 0.2371, + "step": 198 + }, + { + "epoch": 0.334453781512605, + "grad_norm": 4.483956700532791, + "learning_rate": 9.325720325015211e-06, + "loss": 0.3235, + "step": 199 + }, + { + "epoch": 0.33613445378151263, + "grad_norm": 4.211527014803536, + "learning_rate": 9.319085164772082e-06, + "loss": 0.2591, + "step": 200 + }, + { + "epoch": 0.3378151260504202, + "grad_norm": 3.622584139095102, + "learning_rate": 9.312419902391256e-06, + "loss": 0.2483, + "step": 201 + }, + { + "epoch": 0.33949579831932775, + "grad_norm": 3.88188639438146, + "learning_rate": 9.305724584326702e-06, + "loss": 0.2586, + "step": 202 + }, + { + "epoch": 0.3411764705882353, + "grad_norm": 3.9405976888051475, + "learning_rate": 9.298999257241862e-06, + "loss": 0.2274, + "step": 203 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 4.146224854604681, + "learning_rate": 9.292243968009332e-06, + "loss": 0.2594, + "step": 204 + }, + { + "epoch": 0.3445378151260504, + "grad_norm": 3.989185944129595, + "learning_rate": 9.285458763710524e-06, + "loss": 0.2683, + "step": 205 + }, + { + "epoch": 0.346218487394958, + "grad_norm": 5.608784350311355, + "learning_rate": 9.278643691635352e-06, + "loss": 0.3709, + "step": 206 + }, + { + "epoch": 0.34789915966386553, + "grad_norm": 5.532967872431193, + "learning_rate": 9.271798799281893e-06, + "loss": 0.297, + "step": 207 + }, + { + "epoch": 0.3495798319327731, + "grad_norm": 4.155027058077883, + "learning_rate": 9.264924134356057e-06, + "loss": 0.2531, + "step": 208 + }, + { + "epoch": 0.35126050420168065, + "grad_norm": 3.754241120066514, + "learning_rate": 9.258019744771256e-06, + "loss": 0.2435, + "step": 209 + }, + { + "epoch": 0.35294117647058826, + "grad_norm": 4.052985466968748, + "learning_rate": 9.251085678648072e-06, + "loss": 0.3153, + "step": 210 + }, + { + "epoch": 0.3546218487394958, + "grad_norm": 4.148768432508325, + "learning_rate": 9.244121984313916e-06, + "loss": 0.2732, + "step": 211 + }, + { + "epoch": 0.3563025210084034, + "grad_norm": 5.068122975449146, + "learning_rate": 9.2371287103027e-06, + "loss": 0.3473, + "step": 212 + }, + { + "epoch": 0.35798319327731093, + "grad_norm": 4.006936512374535, + "learning_rate": 9.23010590535449e-06, + "loss": 0.315, + "step": 213 + }, + { + "epoch": 0.3596638655462185, + "grad_norm": 3.547781254170819, + "learning_rate": 9.223053618415168e-06, + "loss": 0.2365, + "step": 214 + }, + { + "epoch": 0.36134453781512604, + "grad_norm": 4.36880445829326, + "learning_rate": 9.215971898636094e-06, + "loss": 0.2967, + "step": 215 + }, + { + "epoch": 0.3630252100840336, + "grad_norm": 4.496167262693602, + "learning_rate": 9.208860795373765e-06, + "loss": 0.2878, + "step": 216 + }, + { + "epoch": 0.36470588235294116, + "grad_norm": 3.887663796616842, + "learning_rate": 9.201720358189464e-06, + "loss": 0.2525, + "step": 217 + }, + { + "epoch": 0.3663865546218487, + "grad_norm": 4.233661308098312, + "learning_rate": 9.194550636848923e-06, + "loss": 0.2519, + "step": 218 + }, + { + "epoch": 0.3680672268907563, + "grad_norm": 4.54484885962988, + "learning_rate": 9.187351681321965e-06, + "loss": 0.3203, + "step": 219 + }, + { + "epoch": 0.3697478991596639, + "grad_norm": 3.4662216920786424, + "learning_rate": 9.180123541782172e-06, + "loss": 0.2214, + "step": 220 + }, + { + "epoch": 0.37142857142857144, + "grad_norm": 3.5300962901500506, + "learning_rate": 9.172866268606514e-06, + "loss": 0.2963, + "step": 221 + }, + { + "epoch": 0.373109243697479, + "grad_norm": 4.206031375820797, + "learning_rate": 9.16557991237502e-06, + "loss": 0.2203, + "step": 222 + }, + { + "epoch": 0.37478991596638656, + "grad_norm": 3.757840416808929, + "learning_rate": 9.158264523870413e-06, + "loss": 0.3239, + "step": 223 + }, + { + "epoch": 0.3764705882352941, + "grad_norm": 3.597090675392685, + "learning_rate": 9.150920154077753e-06, + "loss": 0.2665, + "step": 224 + }, + { + "epoch": 0.37815126050420167, + "grad_norm": 4.446089735200772, + "learning_rate": 9.143546854184095e-06, + "loss": 0.2796, + "step": 225 + }, + { + "epoch": 0.3798319327731092, + "grad_norm": 4.073434401093653, + "learning_rate": 9.136144675578114e-06, + "loss": 0.3209, + "step": 226 + }, + { + "epoch": 0.3815126050420168, + "grad_norm": 5.293045962034339, + "learning_rate": 9.128713669849767e-06, + "loss": 0.2809, + "step": 227 + }, + { + "epoch": 0.3831932773109244, + "grad_norm": 3.680410107923094, + "learning_rate": 9.121253888789916e-06, + "loss": 0.2351, + "step": 228 + }, + { + "epoch": 0.38487394957983195, + "grad_norm": 4.088899341235213, + "learning_rate": 9.113765384389984e-06, + "loss": 0.241, + "step": 229 + }, + { + "epoch": 0.3865546218487395, + "grad_norm": 4.210493690445477, + "learning_rate": 9.106248208841568e-06, + "loss": 0.2192, + "step": 230 + }, + { + "epoch": 0.38823529411764707, + "grad_norm": 3.587405660068709, + "learning_rate": 9.098702414536107e-06, + "loss": 0.2711, + "step": 231 + }, + { + "epoch": 0.3899159663865546, + "grad_norm": 3.825397280710117, + "learning_rate": 9.091128054064487e-06, + "loss": 0.3151, + "step": 232 + }, + { + "epoch": 0.3915966386554622, + "grad_norm": 4.285084366810055, + "learning_rate": 9.083525180216697e-06, + "loss": 0.3844, + "step": 233 + }, + { + "epoch": 0.39327731092436974, + "grad_norm": 4.755201193219765, + "learning_rate": 9.075893845981445e-06, + "loss": 0.3369, + "step": 234 + }, + { + "epoch": 0.3949579831932773, + "grad_norm": 4.281159237147637, + "learning_rate": 9.0682341045458e-06, + "loss": 0.2796, + "step": 235 + }, + { + "epoch": 0.39663865546218485, + "grad_norm": 4.368844736645276, + "learning_rate": 9.060546009294818e-06, + "loss": 0.2487, + "step": 236 + }, + { + "epoch": 0.3983193277310924, + "grad_norm": 4.6683321618853135, + "learning_rate": 9.05282961381116e-06, + "loss": 0.3492, + "step": 237 + }, + { + "epoch": 0.4, + "grad_norm": 5.315996126677612, + "learning_rate": 9.045084971874738e-06, + "loss": 0.3656, + "step": 238 + }, + { + "epoch": 0.4016806722689076, + "grad_norm": 3.858049576778821, + "learning_rate": 9.037312137462323e-06, + "loss": 0.2944, + "step": 239 + }, + { + "epoch": 0.40336134453781514, + "grad_norm": 3.759649106675766, + "learning_rate": 9.029511164747175e-06, + "loss": 0.2459, + "step": 240 + }, + { + "epoch": 0.4050420168067227, + "grad_norm": 3.202243641118359, + "learning_rate": 9.021682108098671e-06, + "loss": 0.2363, + "step": 241 + }, + { + "epoch": 0.40672268907563025, + "grad_norm": 3.6191519137994645, + "learning_rate": 9.013825022081915e-06, + "loss": 0.2047, + "step": 242 + }, + { + "epoch": 0.4084033613445378, + "grad_norm": 3.6644452904006033, + "learning_rate": 9.005939961457366e-06, + "loss": 0.262, + "step": 243 + }, + { + "epoch": 0.41008403361344536, + "grad_norm": 3.7496680659825086, + "learning_rate": 8.998026981180454e-06, + "loss": 0.2771, + "step": 244 + }, + { + "epoch": 0.4117647058823529, + "grad_norm": 3.8508524533519846, + "learning_rate": 8.990086136401199e-06, + "loss": 0.2579, + "step": 245 + }, + { + "epoch": 0.4134453781512605, + "grad_norm": 3.7111410813135364, + "learning_rate": 8.982117482463817e-06, + "loss": 0.2416, + "step": 246 + }, + { + "epoch": 0.4151260504201681, + "grad_norm": 3.989630255920349, + "learning_rate": 8.97412107490635e-06, + "loss": 0.298, + "step": 247 + }, + { + "epoch": 0.41680672268907565, + "grad_norm": 4.732255647732948, + "learning_rate": 8.966096969460263e-06, + "loss": 0.32, + "step": 248 + }, + { + "epoch": 0.4184873949579832, + "grad_norm": 4.411373768817184, + "learning_rate": 8.958045222050073e-06, + "loss": 0.2691, + "step": 249 + }, + { + "epoch": 0.42016806722689076, + "grad_norm": 3.1311996398393442, + "learning_rate": 8.94996588879294e-06, + "loss": 0.221, + "step": 250 + }, + { + "epoch": 0.4218487394957983, + "grad_norm": 4.469878722633478, + "learning_rate": 8.94185902599829e-06, + "loss": 0.2651, + "step": 251 + }, + { + "epoch": 0.4235294117647059, + "grad_norm": 4.365076580209745, + "learning_rate": 8.933724690167417e-06, + "loss": 0.2595, + "step": 252 + }, + { + "epoch": 0.42521008403361343, + "grad_norm": 4.633047058160098, + "learning_rate": 8.92556293799309e-06, + "loss": 0.3017, + "step": 253 + }, + { + "epoch": 0.426890756302521, + "grad_norm": 5.2053731159390155, + "learning_rate": 8.917373826359156e-06, + "loss": 0.299, + "step": 254 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 4.447949765493168, + "learning_rate": 8.90915741234015e-06, + "loss": 0.3193, + "step": 255 + }, + { + "epoch": 0.43025210084033616, + "grad_norm": 3.8152143367175513, + "learning_rate": 8.900913753200887e-06, + "loss": 0.3044, + "step": 256 + }, + { + "epoch": 0.4319327731092437, + "grad_norm": 4.174809354409252, + "learning_rate": 8.892642906396076e-06, + "loss": 0.2842, + "step": 257 + }, + { + "epoch": 0.4336134453781513, + "grad_norm": 3.77208471800023, + "learning_rate": 8.884344929569905e-06, + "loss": 0.2609, + "step": 258 + }, + { + "epoch": 0.43529411764705883, + "grad_norm": 4.447034923610601, + "learning_rate": 8.87601988055565e-06, + "loss": 0.2726, + "step": 259 + }, + { + "epoch": 0.4369747899159664, + "grad_norm": 5.8600808097772195, + "learning_rate": 8.867667817375266e-06, + "loss": 0.346, + "step": 260 + }, + { + "epoch": 0.43865546218487395, + "grad_norm": 4.540496751153247, + "learning_rate": 8.859288798238988e-06, + "loss": 0.3182, + "step": 261 + }, + { + "epoch": 0.4403361344537815, + "grad_norm": 2.8122638765624397, + "learning_rate": 8.850882881544923e-06, + "loss": 0.1844, + "step": 262 + }, + { + "epoch": 0.44201680672268906, + "grad_norm": 3.5705823570496684, + "learning_rate": 8.842450125878634e-06, + "loss": 0.2554, + "step": 263 + }, + { + "epoch": 0.4436974789915966, + "grad_norm": 3.455189753673065, + "learning_rate": 8.833990590012749e-06, + "loss": 0.2733, + "step": 264 + }, + { + "epoch": 0.44537815126050423, + "grad_norm": 4.535794944046676, + "learning_rate": 8.825504332906542e-06, + "loss": 0.3123, + "step": 265 + }, + { + "epoch": 0.4470588235294118, + "grad_norm": 4.501300539384694, + "learning_rate": 8.816991413705515e-06, + "loss": 0.3499, + "step": 266 + }, + { + "epoch": 0.44873949579831934, + "grad_norm": 3.772508646537353, + "learning_rate": 8.808451891741001e-06, + "loss": 0.2276, + "step": 267 + }, + { + "epoch": 0.4504201680672269, + "grad_norm": 4.46019811065151, + "learning_rate": 8.799885826529736e-06, + "loss": 0.3519, + "step": 268 + }, + { + "epoch": 0.45210084033613446, + "grad_norm": 3.836911860530734, + "learning_rate": 8.79129327777346e-06, + "loss": 0.311, + "step": 269 + }, + { + "epoch": 0.453781512605042, + "grad_norm": 4.17176003502619, + "learning_rate": 8.782674305358481e-06, + "loss": 0.43, + "step": 270 + }, + { + "epoch": 0.45546218487394957, + "grad_norm": 4.563428438365214, + "learning_rate": 8.774028969355273e-06, + "loss": 0.3458, + "step": 271 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 3.1154650936868693, + "learning_rate": 8.765357330018056e-06, + "loss": 0.2571, + "step": 272 + }, + { + "epoch": 0.4588235294117647, + "grad_norm": 4.122139441670632, + "learning_rate": 8.756659447784367e-06, + "loss": 0.262, + "step": 273 + }, + { + "epoch": 0.46050420168067224, + "grad_norm": 3.8315294608965855, + "learning_rate": 8.74793538327465e-06, + "loss": 0.2821, + "step": 274 + }, + { + "epoch": 0.46218487394957986, + "grad_norm": 3.486277326170669, + "learning_rate": 8.739185197291824e-06, + "loss": 0.2609, + "step": 275 + }, + { + "epoch": 0.4638655462184874, + "grad_norm": 3.9161973496705804, + "learning_rate": 8.730408950820864e-06, + "loss": 0.2659, + "step": 276 + }, + { + "epoch": 0.46554621848739497, + "grad_norm": 3.9755393105316026, + "learning_rate": 8.721606705028376e-06, + "loss": 0.3386, + "step": 277 + }, + { + "epoch": 0.4672268907563025, + "grad_norm": 4.113828914872, + "learning_rate": 8.71277852126217e-06, + "loss": 0.3051, + "step": 278 + }, + { + "epoch": 0.4689075630252101, + "grad_norm": 4.36471703505004, + "learning_rate": 8.703924461050832e-06, + "loss": 0.3559, + "step": 279 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 3.2200963840010313, + "learning_rate": 8.695044586103297e-06, + "loss": 0.2411, + "step": 280 + }, + { + "epoch": 0.4722689075630252, + "grad_norm": 3.703533034697428, + "learning_rate": 8.686138958308415e-06, + "loss": 0.2494, + "step": 281 + }, + { + "epoch": 0.47394957983193275, + "grad_norm": 4.893742503645441, + "learning_rate": 8.67720763973452e-06, + "loss": 0.3223, + "step": 282 + }, + { + "epoch": 0.4756302521008403, + "grad_norm": 3.4635478444534953, + "learning_rate": 8.668250692629008e-06, + "loss": 0.2756, + "step": 283 + }, + { + "epoch": 0.4773109243697479, + "grad_norm": 3.6201745404997614, + "learning_rate": 8.659268179417886e-06, + "loss": 0.2908, + "step": 284 + }, + { + "epoch": 0.4789915966386555, + "grad_norm": 3.782243262021975, + "learning_rate": 8.65026016270535e-06, + "loss": 0.234, + "step": 285 + }, + { + "epoch": 0.48067226890756304, + "grad_norm": 3.5478171634722813, + "learning_rate": 8.641226705273344e-06, + "loss": 0.2861, + "step": 286 + }, + { + "epoch": 0.4823529411764706, + "grad_norm": 4.037626445390373, + "learning_rate": 8.632167870081122e-06, + "loss": 0.2432, + "step": 287 + }, + { + "epoch": 0.48403361344537815, + "grad_norm": 3.2184403921642213, + "learning_rate": 8.623083720264806e-06, + "loss": 0.2363, + "step": 288 + }, + { + "epoch": 0.4857142857142857, + "grad_norm": 5.052805269729211, + "learning_rate": 8.613974319136959e-06, + "loss": 0.3165, + "step": 289 + }, + { + "epoch": 0.48739495798319327, + "grad_norm": 3.218456415544736, + "learning_rate": 8.604839730186125e-06, + "loss": 0.2267, + "step": 290 + }, + { + "epoch": 0.4890756302521008, + "grad_norm": 4.252440000738233, + "learning_rate": 8.595680017076403e-06, + "loss": 0.2729, + "step": 291 + }, + { + "epoch": 0.4907563025210084, + "grad_norm": 3.1859679446297022, + "learning_rate": 8.586495243646992e-06, + "loss": 0.251, + "step": 292 + }, + { + "epoch": 0.492436974789916, + "grad_norm": 3.66415139207086, + "learning_rate": 8.577285473911753e-06, + "loss": 0.2324, + "step": 293 + }, + { + "epoch": 0.49411764705882355, + "grad_norm": 4.099894312537824, + "learning_rate": 8.568050772058763e-06, + "loss": 0.2386, + "step": 294 + }, + { + "epoch": 0.4957983193277311, + "grad_norm": 3.9209399542308607, + "learning_rate": 8.558791202449857e-06, + "loss": 0.3748, + "step": 295 + }, + { + "epoch": 0.49747899159663866, + "grad_norm": 3.3376430249824636, + "learning_rate": 8.549506829620193e-06, + "loss": 0.2286, + "step": 296 + }, + { + "epoch": 0.4991596638655462, + "grad_norm": 3.062020828679799, + "learning_rate": 8.540197718277797e-06, + "loss": 0.2708, + "step": 297 + }, + { + "epoch": 0.5008403361344538, + "grad_norm": 3.1937828441308334, + "learning_rate": 8.530863933303108e-06, + "loss": 0.2086, + "step": 298 + }, + { + "epoch": 0.5025210084033613, + "grad_norm": 3.521499964472593, + "learning_rate": 8.521505539748535e-06, + "loss": 0.2447, + "step": 299 + }, + { + "epoch": 0.5042016806722689, + "grad_norm": 4.642967511496677, + "learning_rate": 8.512122602837993e-06, + "loss": 0.2472, + "step": 300 + }, + { + "epoch": 0.5058823529411764, + "grad_norm": 4.284360904697484, + "learning_rate": 8.502715187966455e-06, + "loss": 0.3206, + "step": 301 + }, + { + "epoch": 0.507563025210084, + "grad_norm": 3.465500865378122, + "learning_rate": 8.493283360699496e-06, + "loss": 0.2294, + "step": 302 + }, + { + "epoch": 0.5092436974789916, + "grad_norm": 3.4457509437221177, + "learning_rate": 8.483827186772832e-06, + "loss": 0.2912, + "step": 303 + }, + { + "epoch": 0.5109243697478991, + "grad_norm": 3.808581617500537, + "learning_rate": 8.47434673209187e-06, + "loss": 0.2446, + "step": 304 + }, + { + "epoch": 0.5126050420168067, + "grad_norm": 3.3314215181937095, + "learning_rate": 8.464842062731235e-06, + "loss": 0.2212, + "step": 305 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 4.039124686139426, + "learning_rate": 8.455313244934324e-06, + "loss": 0.2455, + "step": 306 + }, + { + "epoch": 0.5159663865546219, + "grad_norm": 4.193722450586983, + "learning_rate": 8.445760345112836e-06, + "loss": 0.2481, + "step": 307 + }, + { + "epoch": 0.5176470588235295, + "grad_norm": 3.889796309337689, + "learning_rate": 8.436183429846314e-06, + "loss": 0.2859, + "step": 308 + }, + { + "epoch": 0.519327731092437, + "grad_norm": 3.9169255433081664, + "learning_rate": 8.426582565881674e-06, + "loss": 0.263, + "step": 309 + }, + { + "epoch": 0.5210084033613446, + "grad_norm": 3.418188058193833, + "learning_rate": 8.416957820132743e-06, + "loss": 0.2356, + "step": 310 + }, + { + "epoch": 0.5226890756302521, + "grad_norm": 3.2839702099179067, + "learning_rate": 8.407309259679801e-06, + "loss": 0.2555, + "step": 311 + }, + { + "epoch": 0.5243697478991597, + "grad_norm": 3.9981862118067957, + "learning_rate": 8.397636951769099e-06, + "loss": 0.2468, + "step": 312 + }, + { + "epoch": 0.5260504201680672, + "grad_norm": 4.1369216765373595, + "learning_rate": 8.387940963812398e-06, + "loss": 0.2481, + "step": 313 + }, + { + "epoch": 0.5277310924369748, + "grad_norm": 4.270925434895349, + "learning_rate": 8.378221363386506e-06, + "loss": 0.2558, + "step": 314 + }, + { + "epoch": 0.5294117647058824, + "grad_norm": 3.240664369248623, + "learning_rate": 8.368478218232787e-06, + "loss": 0.2583, + "step": 315 + }, + { + "epoch": 0.5310924369747899, + "grad_norm": 3.1933231590113142, + "learning_rate": 8.358711596256712e-06, + "loss": 0.215, + "step": 316 + }, + { + "epoch": 0.5327731092436975, + "grad_norm": 4.167983149331361, + "learning_rate": 8.348921565527373e-06, + "loss": 0.3324, + "step": 317 + }, + { + "epoch": 0.534453781512605, + "grad_norm": 3.4358411113634646, + "learning_rate": 8.339108194277006e-06, + "loss": 0.2589, + "step": 318 + }, + { + "epoch": 0.5361344537815126, + "grad_norm": 4.164325821445187, + "learning_rate": 8.329271550900528e-06, + "loss": 0.298, + "step": 319 + }, + { + "epoch": 0.5378151260504201, + "grad_norm": 3.5852432229900613, + "learning_rate": 8.319411703955042e-06, + "loss": 0.2575, + "step": 320 + }, + { + "epoch": 0.5394957983193277, + "grad_norm": 4.392211699734561, + "learning_rate": 8.309528722159383e-06, + "loss": 0.2755, + "step": 321 + }, + { + "epoch": 0.5411764705882353, + "grad_norm": 4.185705868785948, + "learning_rate": 8.299622674393615e-06, + "loss": 0.2893, + "step": 322 + }, + { + "epoch": 0.5428571428571428, + "grad_norm": 4.547143925765926, + "learning_rate": 8.289693629698564e-06, + "loss": 0.3375, + "step": 323 + }, + { + "epoch": 0.5445378151260504, + "grad_norm": 3.9462639969766977, + "learning_rate": 8.27974165727534e-06, + "loss": 0.3053, + "step": 324 + }, + { + "epoch": 0.5462184873949579, + "grad_norm": 3.972277144129639, + "learning_rate": 8.269766826484841e-06, + "loss": 0.2827, + "step": 325 + }, + { + "epoch": 0.5478991596638656, + "grad_norm": 3.7091300891124526, + "learning_rate": 8.259769206847286e-06, + "loss": 0.2751, + "step": 326 + }, + { + "epoch": 0.5495798319327732, + "grad_norm": 3.8728056075092736, + "learning_rate": 8.249748868041717e-06, + "loss": 0.2629, + "step": 327 + }, + { + "epoch": 0.5512605042016807, + "grad_norm": 3.230198623667927, + "learning_rate": 8.239705879905519e-06, + "loss": 0.2557, + "step": 328 + }, + { + "epoch": 0.5529411764705883, + "grad_norm": 4.162100326672927, + "learning_rate": 8.229640312433938e-06, + "loss": 0.2249, + "step": 329 + }, + { + "epoch": 0.5546218487394958, + "grad_norm": 3.030431828388766, + "learning_rate": 8.219552235779578e-06, + "loss": 0.2382, + "step": 330 + }, + { + "epoch": 0.5563025210084034, + "grad_norm": 3.2618326842590446, + "learning_rate": 8.209441720251934e-06, + "loss": 0.2562, + "step": 331 + }, + { + "epoch": 0.5579831932773109, + "grad_norm": 4.60003136005453, + "learning_rate": 8.199308836316883e-06, + "loss": 0.2635, + "step": 332 + }, + { + "epoch": 0.5596638655462185, + "grad_norm": 5.124504363614122, + "learning_rate": 8.189153654596199e-06, + "loss": 0.2848, + "step": 333 + }, + { + "epoch": 0.561344537815126, + "grad_norm": 3.093400315613663, + "learning_rate": 8.178976245867068e-06, + "loss": 0.2058, + "step": 334 + }, + { + "epoch": 0.5630252100840336, + "grad_norm": 3.393259880948687, + "learning_rate": 8.168776681061583e-06, + "loss": 0.2009, + "step": 335 + }, + { + "epoch": 0.5647058823529412, + "grad_norm": 3.9012793474429226, + "learning_rate": 8.158555031266255e-06, + "loss": 0.2728, + "step": 336 + }, + { + "epoch": 0.5663865546218487, + "grad_norm": 4.010363842008162, + "learning_rate": 8.148311367721524e-06, + "loss": 0.2487, + "step": 337 + }, + { + "epoch": 0.5680672268907563, + "grad_norm": 3.313090386559345, + "learning_rate": 8.138045761821252e-06, + "loss": 0.2392, + "step": 338 + }, + { + "epoch": 0.5697478991596638, + "grad_norm": 3.003737875930333, + "learning_rate": 8.127758285112226e-06, + "loss": 0.196, + "step": 339 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 4.813841107800995, + "learning_rate": 8.117449009293668e-06, + "loss": 0.2345, + "step": 340 + }, + { + "epoch": 0.573109243697479, + "grad_norm": 3.820262645541521, + "learning_rate": 8.107118006216732e-06, + "loss": 0.3774, + "step": 341 + }, + { + "epoch": 0.5747899159663865, + "grad_norm": 3.5798699267278256, + "learning_rate": 8.096765347883995e-06, + "loss": 0.323, + "step": 342 + }, + { + "epoch": 0.5764705882352941, + "grad_norm": 3.7918963101834695, + "learning_rate": 8.086391106448965e-06, + "loss": 0.2771, + "step": 343 + }, + { + "epoch": 0.5781512605042017, + "grad_norm": 3.710871164871179, + "learning_rate": 8.075995354215578e-06, + "loss": 0.2296, + "step": 344 + }, + { + "epoch": 0.5798319327731093, + "grad_norm": 3.081675164930532, + "learning_rate": 8.065578163637686e-06, + "loss": 0.239, + "step": 345 + }, + { + "epoch": 0.5815126050420169, + "grad_norm": 3.472850199568395, + "learning_rate": 8.055139607318558e-06, + "loss": 0.2581, + "step": 346 + }, + { + "epoch": 0.5831932773109244, + "grad_norm": 3.5895910105236988, + "learning_rate": 8.044679758010376e-06, + "loss": 0.25, + "step": 347 + }, + { + "epoch": 0.584873949579832, + "grad_norm": 5.03178448321192, + "learning_rate": 8.03419868861372e-06, + "loss": 0.1916, + "step": 348 + }, + { + "epoch": 0.5865546218487395, + "grad_norm": 3.821763309544136, + "learning_rate": 8.023696472177068e-06, + "loss": 0.2909, + "step": 349 + }, + { + "epoch": 0.5882352941176471, + "grad_norm": 4.0276823826082095, + "learning_rate": 8.013173181896283e-06, + "loss": 0.2604, + "step": 350 + }, + { + "epoch": 0.5899159663865546, + "grad_norm": 2.9398477769509705, + "learning_rate": 8.002628891114104e-06, + "loss": 0.2271, + "step": 351 + }, + { + "epoch": 0.5915966386554622, + "grad_norm": 4.554336270535579, + "learning_rate": 7.992063673319632e-06, + "loss": 0.3412, + "step": 352 + }, + { + "epoch": 0.5932773109243697, + "grad_norm": 4.154201529208768, + "learning_rate": 7.981477602147823e-06, + "loss": 0.265, + "step": 353 + }, + { + "epoch": 0.5949579831932773, + "grad_norm": 2.665436237812907, + "learning_rate": 7.97087075137897e-06, + "loss": 0.1938, + "step": 354 + }, + { + "epoch": 0.5966386554621849, + "grad_norm": 3.3626195738967577, + "learning_rate": 7.960243194938192e-06, + "loss": 0.228, + "step": 355 + }, + { + "epoch": 0.5983193277310924, + "grad_norm": 3.8286004476475335, + "learning_rate": 7.949595006894917e-06, + "loss": 0.257, + "step": 356 + }, + { + "epoch": 0.6, + "grad_norm": 3.0558429385007275, + "learning_rate": 7.938926261462366e-06, + "loss": 0.3105, + "step": 357 + }, + { + "epoch": 0.6016806722689075, + "grad_norm": 4.545539695509062, + "learning_rate": 7.928237032997037e-06, + "loss": 0.2787, + "step": 358 + }, + { + "epoch": 0.6033613445378151, + "grad_norm": 3.4144872596036326, + "learning_rate": 7.917527395998183e-06, + "loss": 0.2107, + "step": 359 + }, + { + "epoch": 0.6050420168067226, + "grad_norm": 3.197576947653472, + "learning_rate": 7.9067974251073e-06, + "loss": 0.2402, + "step": 360 + }, + { + "epoch": 0.6067226890756302, + "grad_norm": 4.953522954492192, + "learning_rate": 7.896047195107599e-06, + "loss": 0.4079, + "step": 361 + }, + { + "epoch": 0.6084033613445378, + "grad_norm": 3.913644693006596, + "learning_rate": 7.885276780923488e-06, + "loss": 0.2446, + "step": 362 + }, + { + "epoch": 0.6100840336134454, + "grad_norm": 4.162061152262566, + "learning_rate": 7.87448625762005e-06, + "loss": 0.2775, + "step": 363 + }, + { + "epoch": 0.611764705882353, + "grad_norm": 3.9118577002779915, + "learning_rate": 7.863675700402527e-06, + "loss": 0.2945, + "step": 364 + }, + { + "epoch": 0.6134453781512605, + "grad_norm": 3.835742001918319, + "learning_rate": 7.852845184615776e-06, + "loss": 0.3243, + "step": 365 + }, + { + "epoch": 0.6151260504201681, + "grad_norm": 3.811064930664159, + "learning_rate": 7.841994785743765e-06, + "loss": 0.275, + "step": 366 + }, + { + "epoch": 0.6168067226890757, + "grad_norm": 3.638636507424936, + "learning_rate": 7.831124579409036e-06, + "loss": 0.2385, + "step": 367 + }, + { + "epoch": 0.6184873949579832, + "grad_norm": 3.7590755941327263, + "learning_rate": 7.820234641372182e-06, + "loss": 0.2615, + "step": 368 + }, + { + "epoch": 0.6201680672268908, + "grad_norm": 4.262512697571422, + "learning_rate": 7.809325047531315e-06, + "loss": 0.2305, + "step": 369 + }, + { + "epoch": 0.6218487394957983, + "grad_norm": 3.5789208550854137, + "learning_rate": 7.798395873921542e-06, + "loss": 0.2829, + "step": 370 + }, + { + "epoch": 0.6235294117647059, + "grad_norm": 3.6770787617731657, + "learning_rate": 7.787447196714428e-06, + "loss": 0.3103, + "step": 371 + }, + { + "epoch": 0.6252100840336134, + "grad_norm": 4.681057793290432, + "learning_rate": 7.776479092217475e-06, + "loss": 0.3245, + "step": 372 + }, + { + "epoch": 0.626890756302521, + "grad_norm": 4.585623869647911, + "learning_rate": 7.76549163687358e-06, + "loss": 0.2892, + "step": 373 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 3.820748783778272, + "learning_rate": 7.754484907260513e-06, + "loss": 0.2256, + "step": 374 + }, + { + "epoch": 0.6302521008403361, + "grad_norm": 3.9111468680258086, + "learning_rate": 7.743458980090371e-06, + "loss": 0.2549, + "step": 375 + }, + { + "epoch": 0.6319327731092437, + "grad_norm": 3.1067953029914888, + "learning_rate": 7.73241393220905e-06, + "loss": 0.2132, + "step": 376 + }, + { + "epoch": 0.6336134453781512, + "grad_norm": 3.164293144904975, + "learning_rate": 7.721349840595713e-06, + "loss": 0.2092, + "step": 377 + }, + { + "epoch": 0.6352941176470588, + "grad_norm": 3.734593960662017, + "learning_rate": 7.710266782362248e-06, + "loss": 0.2976, + "step": 378 + }, + { + "epoch": 0.6369747899159663, + "grad_norm": 3.1220699118832407, + "learning_rate": 7.69916483475273e-06, + "loss": 0.2478, + "step": 379 + }, + { + "epoch": 0.6386554621848739, + "grad_norm": 3.2574034724313177, + "learning_rate": 7.688044075142888e-06, + "loss": 0.1939, + "step": 380 + }, + { + "epoch": 0.6403361344537815, + "grad_norm": 3.4284207068175907, + "learning_rate": 7.676904581039559e-06, + "loss": 0.2866, + "step": 381 + }, + { + "epoch": 0.6420168067226891, + "grad_norm": 3.040455959047564, + "learning_rate": 7.665746430080155e-06, + "loss": 0.217, + "step": 382 + }, + { + "epoch": 0.6436974789915967, + "grad_norm": 3.224524280206461, + "learning_rate": 7.654569700032112e-06, + "loss": 0.2332, + "step": 383 + }, + { + "epoch": 0.6453781512605042, + "grad_norm": 3.6681207305304744, + "learning_rate": 7.643374468792364e-06, + "loss": 0.23, + "step": 384 + }, + { + "epoch": 0.6470588235294118, + "grad_norm": 2.76803712271111, + "learning_rate": 7.63216081438678e-06, + "loss": 0.1768, + "step": 385 + }, + { + "epoch": 0.6487394957983194, + "grad_norm": 3.172049754177004, + "learning_rate": 7.620928814969636e-06, + "loss": 0.2458, + "step": 386 + }, + { + "epoch": 0.6504201680672269, + "grad_norm": 4.021324325221757, + "learning_rate": 7.609678548823065e-06, + "loss": 0.2878, + "step": 387 + }, + { + "epoch": 0.6521008403361345, + "grad_norm": 4.45631664976017, + "learning_rate": 7.5984100943565055e-06, + "loss": 0.3669, + "step": 388 + }, + { + "epoch": 0.653781512605042, + "grad_norm": 4.350536114923283, + "learning_rate": 7.587123530106171e-06, + "loss": 0.3114, + "step": 389 + }, + { + "epoch": 0.6554621848739496, + "grad_norm": 4.110763780984955, + "learning_rate": 7.57581893473448e-06, + "loss": 0.3002, + "step": 390 + }, + { + "epoch": 0.6571428571428571, + "grad_norm": 3.824638074318695, + "learning_rate": 7.564496387029532e-06, + "loss": 0.2555, + "step": 391 + }, + { + "epoch": 0.6588235294117647, + "grad_norm": 4.166305509765287, + "learning_rate": 7.553155965904535e-06, + "loss": 0.2617, + "step": 392 + }, + { + "epoch": 0.6605042016806723, + "grad_norm": 4.290626462011601, + "learning_rate": 7.541797750397277e-06, + "loss": 0.2565, + "step": 393 + }, + { + "epoch": 0.6621848739495798, + "grad_norm": 3.688859506019526, + "learning_rate": 7.530421819669558e-06, + "loss": 0.2639, + "step": 394 + }, + { + "epoch": 0.6638655462184874, + "grad_norm": 3.6789226419918446, + "learning_rate": 7.519028253006649e-06, + "loss": 0.2741, + "step": 395 + }, + { + "epoch": 0.6655462184873949, + "grad_norm": 3.3382701043360647, + "learning_rate": 7.507617129816733e-06, + "loss": 0.2206, + "step": 396 + }, + { + "epoch": 0.6672268907563025, + "grad_norm": 3.3066950479168153, + "learning_rate": 7.496188529630359e-06, + "loss": 0.2993, + "step": 397 + }, + { + "epoch": 0.66890756302521, + "grad_norm": 3.893635706930087, + "learning_rate": 7.484742532099878e-06, + "loss": 0.2326, + "step": 398 + }, + { + "epoch": 0.6705882352941176, + "grad_norm": 4.099621838646233, + "learning_rate": 7.473279216998896e-06, + "loss": 0.2828, + "step": 399 + }, + { + "epoch": 0.6722689075630253, + "grad_norm": 5.661766299326251, + "learning_rate": 7.461798664221711e-06, + "loss": 0.2377, + "step": 400 + }, + { + "epoch": 0.6739495798319328, + "grad_norm": 2.870347406882497, + "learning_rate": 7.450300953782768e-06, + "loss": 0.151, + "step": 401 + }, + { + "epoch": 0.6756302521008404, + "grad_norm": 3.530732621066666, + "learning_rate": 7.438786165816084e-06, + "loss": 0.3204, + "step": 402 + }, + { + "epoch": 0.6773109243697479, + "grad_norm": 4.951946205032062, + "learning_rate": 7.427254380574705e-06, + "loss": 0.3073, + "step": 403 + }, + { + "epoch": 0.6789915966386555, + "grad_norm": 4.425642984242483, + "learning_rate": 7.415705678430138e-06, + "loss": 0.2526, + "step": 404 + }, + { + "epoch": 0.680672268907563, + "grad_norm": 4.575800850868975, + "learning_rate": 7.404140139871797e-06, + "loss": 0.3725, + "step": 405 + }, + { + "epoch": 0.6823529411764706, + "grad_norm": 4.0365196047687535, + "learning_rate": 7.392557845506433e-06, + "loss": 0.3002, + "step": 406 + }, + { + "epoch": 0.6840336134453782, + "grad_norm": 5.981792145534908, + "learning_rate": 7.380958876057581e-06, + "loss": 0.2435, + "step": 407 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 4.363575458393627, + "learning_rate": 7.369343312364994e-06, + "loss": 0.2212, + "step": 408 + }, + { + "epoch": 0.6873949579831933, + "grad_norm": 4.340605315487809, + "learning_rate": 7.357711235384079e-06, + "loss": 0.2507, + "step": 409 + }, + { + "epoch": 0.6890756302521008, + "grad_norm": 3.6485734876832274, + "learning_rate": 7.346062726185332e-06, + "loss": 0.2597, + "step": 410 + }, + { + "epoch": 0.6907563025210084, + "grad_norm": 4.0943244771611, + "learning_rate": 7.3343978659537775e-06, + "loss": 0.2611, + "step": 411 + }, + { + "epoch": 0.692436974789916, + "grad_norm": 3.5101040552518206, + "learning_rate": 7.3227167359883964e-06, + "loss": 0.2051, + "step": 412 + }, + { + "epoch": 0.6941176470588235, + "grad_norm": 4.3439206313504535, + "learning_rate": 7.311019417701567e-06, + "loss": 0.2752, + "step": 413 + }, + { + "epoch": 0.6957983193277311, + "grad_norm": 3.9550057956306093, + "learning_rate": 7.299305992618488e-06, + "loss": 0.2956, + "step": 414 + }, + { + "epoch": 0.6974789915966386, + "grad_norm": 3.1277606523059927, + "learning_rate": 7.287576542376616e-06, + "loss": 0.2466, + "step": 415 + }, + { + "epoch": 0.6991596638655462, + "grad_norm": 3.1005601888199292, + "learning_rate": 7.275831148725101e-06, + "loss": 0.2144, + "step": 416 + }, + { + "epoch": 0.7008403361344537, + "grad_norm": 3.9059268145411625, + "learning_rate": 7.264069893524207e-06, + "loss": 0.3389, + "step": 417 + }, + { + "epoch": 0.7025210084033613, + "grad_norm": 3.377596564727831, + "learning_rate": 7.252292858744747e-06, + "loss": 0.2157, + "step": 418 + }, + { + "epoch": 0.704201680672269, + "grad_norm": 4.187645692769235, + "learning_rate": 7.24050012646751e-06, + "loss": 0.3017, + "step": 419 + }, + { + "epoch": 0.7058823529411765, + "grad_norm": 3.610783613896857, + "learning_rate": 7.2286917788826926e-06, + "loss": 0.2494, + "step": 420 + }, + { + "epoch": 0.7075630252100841, + "grad_norm": 3.048155560726766, + "learning_rate": 7.216867898289319e-06, + "loss": 0.2092, + "step": 421 + }, + { + "epoch": 0.7092436974789916, + "grad_norm": 3.416863835083985, + "learning_rate": 7.2050285670946776e-06, + "loss": 0.2012, + "step": 422 + }, + { + "epoch": 0.7109243697478992, + "grad_norm": 3.6205625981668983, + "learning_rate": 7.193173867813735e-06, + "loss": 0.1958, + "step": 423 + }, + { + "epoch": 0.7126050420168067, + "grad_norm": 3.2909907100719495, + "learning_rate": 7.181303883068569e-06, + "loss": 0.2534, + "step": 424 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 4.094896807207073, + "learning_rate": 7.169418695587791e-06, + "loss": 0.2834, + "step": 425 + }, + { + "epoch": 0.7159663865546219, + "grad_norm": 3.7570453300774975, + "learning_rate": 7.157518388205969e-06, + "loss": 0.2357, + "step": 426 + }, + { + "epoch": 0.7176470588235294, + "grad_norm": 4.34455688743933, + "learning_rate": 7.145603043863045e-06, + "loss": 0.4079, + "step": 427 + }, + { + "epoch": 0.719327731092437, + "grad_norm": 3.9605795228359453, + "learning_rate": 7.1336727456037716e-06, + "loss": 0.2724, + "step": 428 + }, + { + "epoch": 0.7210084033613445, + "grad_norm": 4.50690190414524, + "learning_rate": 7.121727576577116e-06, + "loss": 0.2715, + "step": 429 + }, + { + "epoch": 0.7226890756302521, + "grad_norm": 3.4305670733319635, + "learning_rate": 7.109767620035689e-06, + "loss": 0.2527, + "step": 430 + }, + { + "epoch": 0.7243697478991596, + "grad_norm": 3.7556549613232155, + "learning_rate": 7.097792959335169e-06, + "loss": 0.272, + "step": 431 + }, + { + "epoch": 0.7260504201680672, + "grad_norm": 4.331780366651023, + "learning_rate": 7.0858036779337095e-06, + "loss": 0.269, + "step": 432 + }, + { + "epoch": 0.7277310924369748, + "grad_norm": 2.9644967270995464, + "learning_rate": 7.0737998593913686e-06, + "loss": 0.2009, + "step": 433 + }, + { + "epoch": 0.7294117647058823, + "grad_norm": 3.424245481423386, + "learning_rate": 7.061781587369518e-06, + "loss": 0.1963, + "step": 434 + }, + { + "epoch": 0.7310924369747899, + "grad_norm": 3.5553808491409975, + "learning_rate": 7.049748945630269e-06, + "loss": 0.2204, + "step": 435 + }, + { + "epoch": 0.7327731092436974, + "grad_norm": 2.988384045794864, + "learning_rate": 7.037702018035878e-06, + "loss": 0.255, + "step": 436 + }, + { + "epoch": 0.7344537815126051, + "grad_norm": 3.877729077495976, + "learning_rate": 7.0256408885481694e-06, + "loss": 0.2379, + "step": 437 + }, + { + "epoch": 0.7361344537815127, + "grad_norm": 3.8152006801676293, + "learning_rate": 7.013565641227954e-06, + "loss": 0.2795, + "step": 438 + }, + { + "epoch": 0.7378151260504202, + "grad_norm": 3.8197950101640408, + "learning_rate": 7.001476360234428e-06, + "loss": 0.2441, + "step": 439 + }, + { + "epoch": 0.7394957983193278, + "grad_norm": 3.653213590928426, + "learning_rate": 6.989373129824605e-06, + "loss": 0.2333, + "step": 440 + }, + { + "epoch": 0.7411764705882353, + "grad_norm": 3.1988797850199986, + "learning_rate": 6.977256034352713e-06, + "loss": 0.1969, + "step": 441 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 3.6369472387733315, + "learning_rate": 6.965125158269619e-06, + "loss": 0.2039, + "step": 442 + }, + { + "epoch": 0.7445378151260504, + "grad_norm": 3.504132707054677, + "learning_rate": 6.952980586122231e-06, + "loss": 0.2294, + "step": 443 + }, + { + "epoch": 0.746218487394958, + "grad_norm": 3.9801905625390166, + "learning_rate": 6.940822402552914e-06, + "loss": 0.2557, + "step": 444 + }, + { + "epoch": 0.7478991596638656, + "grad_norm": 3.596393116732762, + "learning_rate": 6.928650692298898e-06, + "loss": 0.226, + "step": 445 + }, + { + "epoch": 0.7495798319327731, + "grad_norm": 3.0142886185199065, + "learning_rate": 6.916465540191692e-06, + "loss": 0.2098, + "step": 446 + }, + { + "epoch": 0.7512605042016807, + "grad_norm": 4.0997116691327316, + "learning_rate": 6.904267031156482e-06, + "loss": 0.2141, + "step": 447 + }, + { + "epoch": 0.7529411764705882, + "grad_norm": 3.2816445631622164, + "learning_rate": 6.892055250211552e-06, + "loss": 0.2118, + "step": 448 + }, + { + "epoch": 0.7546218487394958, + "grad_norm": 2.898884385242415, + "learning_rate": 6.879830282467682e-06, + "loss": 0.2249, + "step": 449 + }, + { + "epoch": 0.7563025210084033, + "grad_norm": 3.701834773955598, + "learning_rate": 6.867592213127559e-06, + "loss": 0.2975, + "step": 450 + }, + { + "epoch": 0.7579831932773109, + "grad_norm": 3.3905113341532824, + "learning_rate": 6.855341127485183e-06, + "loss": 0.2285, + "step": 451 + }, + { + "epoch": 0.7596638655462185, + "grad_norm": 4.432808727705175, + "learning_rate": 6.84307711092527e-06, + "loss": 0.2872, + "step": 452 + }, + { + "epoch": 0.761344537815126, + "grad_norm": 3.8277259460611974, + "learning_rate": 6.8308002489226645e-06, + "loss": 0.2294, + "step": 453 + }, + { + "epoch": 0.7630252100840336, + "grad_norm": 3.425696594498262, + "learning_rate": 6.81851062704173e-06, + "loss": 0.1943, + "step": 454 + }, + { + "epoch": 0.7647058823529411, + "grad_norm": 3.112228899870348, + "learning_rate": 6.806208330935766e-06, + "loss": 0.2217, + "step": 455 + }, + { + "epoch": 0.7663865546218488, + "grad_norm": 3.6268572961078602, + "learning_rate": 6.793893446346405e-06, + "loss": 0.207, + "step": 456 + }, + { + "epoch": 0.7680672268907563, + "grad_norm": 3.1270143931104, + "learning_rate": 6.7815660591030155e-06, + "loss": 0.2389, + "step": 457 + }, + { + "epoch": 0.7697478991596639, + "grad_norm": 3.343441373055428, + "learning_rate": 6.769226255122104e-06, + "loss": 0.1803, + "step": 458 + }, + { + "epoch": 0.7714285714285715, + "grad_norm": 3.881868079906461, + "learning_rate": 6.7568741204067145e-06, + "loss": 0.264, + "step": 459 + }, + { + "epoch": 0.773109243697479, + "grad_norm": 3.9170195071144396, + "learning_rate": 6.744509741045835e-06, + "loss": 0.2717, + "step": 460 + }, + { + "epoch": 0.7747899159663866, + "grad_norm": 3.597422471752963, + "learning_rate": 6.7321332032137875e-06, + "loss": 0.2373, + "step": 461 + }, + { + "epoch": 0.7764705882352941, + "grad_norm": 3.6326736789138248, + "learning_rate": 6.719744593169642e-06, + "loss": 0.2809, + "step": 462 + }, + { + "epoch": 0.7781512605042017, + "grad_norm": 3.2242024990634177, + "learning_rate": 6.7073439972565955e-06, + "loss": 0.2738, + "step": 463 + }, + { + "epoch": 0.7798319327731092, + "grad_norm": 4.097362370899167, + "learning_rate": 6.6949315019013895e-06, + "loss": 0.2422, + "step": 464 + }, + { + "epoch": 0.7815126050420168, + "grad_norm": 3.0129633820408532, + "learning_rate": 6.682507193613697e-06, + "loss": 0.2308, + "step": 465 + }, + { + "epoch": 0.7831932773109244, + "grad_norm": 3.3831749684902728, + "learning_rate": 6.670071158985521e-06, + "loss": 0.2536, + "step": 466 + }, + { + "epoch": 0.7848739495798319, + "grad_norm": 4.06561523313803, + "learning_rate": 6.657623484690593e-06, + "loss": 0.2635, + "step": 467 + }, + { + "epoch": 0.7865546218487395, + "grad_norm": 3.954232561509824, + "learning_rate": 6.645164257483767e-06, + "loss": 0.2617, + "step": 468 + }, + { + "epoch": 0.788235294117647, + "grad_norm": 3.2864328535168528, + "learning_rate": 6.6326935642004165e-06, + "loss": 0.2467, + "step": 469 + }, + { + "epoch": 0.7899159663865546, + "grad_norm": 4.756041222608582, + "learning_rate": 6.62021149175583e-06, + "loss": 0.3166, + "step": 470 + }, + { + "epoch": 0.7915966386554621, + "grad_norm": 3.2133137441191364, + "learning_rate": 6.607718127144601e-06, + "loss": 0.1957, + "step": 471 + }, + { + "epoch": 0.7932773109243697, + "grad_norm": 3.9946779468067057, + "learning_rate": 6.595213557440026e-06, + "loss": 0.2561, + "step": 472 + }, + { + "epoch": 0.7949579831932773, + "grad_norm": 4.178112022529024, + "learning_rate": 6.582697869793495e-06, + "loss": 0.2841, + "step": 473 + }, + { + "epoch": 0.7966386554621848, + "grad_norm": 3.831343610559717, + "learning_rate": 6.570171151433887e-06, + "loss": 0.204, + "step": 474 + }, + { + "epoch": 0.7983193277310925, + "grad_norm": 3.6832346179871545, + "learning_rate": 6.557633489666958e-06, + "loss": 0.189, + "step": 475 + }, + { + "epoch": 0.8, + "grad_norm": 2.917218404225396, + "learning_rate": 6.545084971874738e-06, + "loss": 0.1833, + "step": 476 + }, + { + "epoch": 0.8016806722689076, + "grad_norm": 2.9993597464552337, + "learning_rate": 6.532525685514915e-06, + "loss": 0.221, + "step": 477 + }, + { + "epoch": 0.8033613445378152, + "grad_norm": 5.077471303123534, + "learning_rate": 6.519955718120231e-06, + "loss": 0.3785, + "step": 478 + }, + { + "epoch": 0.8050420168067227, + "grad_norm": 3.44058389211428, + "learning_rate": 6.507375157297872e-06, + "loss": 0.2357, + "step": 479 + }, + { + "epoch": 0.8067226890756303, + "grad_norm": 3.0384174021471475, + "learning_rate": 6.494784090728852e-06, + "loss": 0.2069, + "step": 480 + }, + { + "epoch": 0.8084033613445378, + "grad_norm": 4.2110013520640335, + "learning_rate": 6.482182606167409e-06, + "loss": 0.2847, + "step": 481 + }, + { + "epoch": 0.8100840336134454, + "grad_norm": 3.4294578019501825, + "learning_rate": 6.469570791440385e-06, + "loss": 0.1842, + "step": 482 + }, + { + "epoch": 0.8117647058823529, + "grad_norm": 3.609094506614432, + "learning_rate": 6.456948734446624e-06, + "loss": 0.2636, + "step": 483 + }, + { + "epoch": 0.8134453781512605, + "grad_norm": 3.4896585628244656, + "learning_rate": 6.444316523156352e-06, + "loss": 0.2057, + "step": 484 + }, + { + "epoch": 0.8151260504201681, + "grad_norm": 3.79357908294218, + "learning_rate": 6.4316742456105645e-06, + "loss": 0.2387, + "step": 485 + }, + { + "epoch": 0.8168067226890756, + "grad_norm": 3.769325505088469, + "learning_rate": 6.419021989920416e-06, + "loss": 0.3014, + "step": 486 + }, + { + "epoch": 0.8184873949579832, + "grad_norm": 3.361574274750808, + "learning_rate": 6.406359844266607e-06, + "loss": 0.2333, + "step": 487 + }, + { + "epoch": 0.8201680672268907, + "grad_norm": 3.957312301077667, + "learning_rate": 6.393687896898759e-06, + "loss": 0.2738, + "step": 488 + }, + { + "epoch": 0.8218487394957983, + "grad_norm": 3.6962832913581245, + "learning_rate": 6.381006236134817e-06, + "loss": 0.202, + "step": 489 + }, + { + "epoch": 0.8235294117647058, + "grad_norm": 4.3843424567306695, + "learning_rate": 6.368314950360416e-06, + "loss": 0.3356, + "step": 490 + }, + { + "epoch": 0.8252100840336134, + "grad_norm": 3.988796144368344, + "learning_rate": 6.355614128028277e-06, + "loss": 0.2188, + "step": 491 + }, + { + "epoch": 0.826890756302521, + "grad_norm": 3.546665934297024, + "learning_rate": 6.342903857657585e-06, + "loss": 0.3062, + "step": 492 + }, + { + "epoch": 0.8285714285714286, + "grad_norm": 3.464405372369805, + "learning_rate": 6.330184227833376e-06, + "loss": 0.302, + "step": 493 + }, + { + "epoch": 0.8302521008403362, + "grad_norm": 3.2217174248829505, + "learning_rate": 6.317455327205916e-06, + "loss": 0.1587, + "step": 494 + }, + { + "epoch": 0.8319327731092437, + "grad_norm": 3.101235516793264, + "learning_rate": 6.304717244490084e-06, + "loss": 0.1999, + "step": 495 + }, + { + "epoch": 0.8336134453781513, + "grad_norm": 3.5091112695722333, + "learning_rate": 6.291970068464755e-06, + "loss": 0.2298, + "step": 496 + }, + { + "epoch": 0.8352941176470589, + "grad_norm": 3.327856429000738, + "learning_rate": 6.279213887972179e-06, + "loss": 0.2582, + "step": 497 + }, + { + "epoch": 0.8369747899159664, + "grad_norm": 3.62144154961139, + "learning_rate": 6.266448791917364e-06, + "loss": 0.3269, + "step": 498 + }, + { + "epoch": 0.838655462184874, + "grad_norm": 2.953403477082121, + "learning_rate": 6.253674869267457e-06, + "loss": 0.1917, + "step": 499 + }, + { + "epoch": 0.8403361344537815, + "grad_norm": 3.011235509544136, + "learning_rate": 6.24089220905112e-06, + "loss": 0.2309, + "step": 500 + }, + { + "epoch": 0.8403361344537815, + "eval_loss": 0.24230201542377472, + "eval_runtime": 7.6191, + "eval_samples_per_second": 6.431, + "eval_steps_per_second": 1.706, + "step": 500 + }, + { + "epoch": 0.8420168067226891, + "grad_norm": 3.64049833442948, + "learning_rate": 6.228100900357914e-06, + "loss": 0.277, + "step": 501 + }, + { + "epoch": 0.8436974789915966, + "grad_norm": 3.360168756218691, + "learning_rate": 6.215301032337674e-06, + "loss": 0.1947, + "step": 502 + }, + { + "epoch": 0.8453781512605042, + "grad_norm": 3.0764894114372976, + "learning_rate": 6.202492694199893e-06, + "loss": 0.2248, + "step": 503 + }, + { + "epoch": 0.8470588235294118, + "grad_norm": 3.425507539967475, + "learning_rate": 6.189675975213094e-06, + "loss": 0.211, + "step": 504 + }, + { + "epoch": 0.8487394957983193, + "grad_norm": 3.537768398824082, + "learning_rate": 6.176850964704213e-06, + "loss": 0.2887, + "step": 505 + }, + { + "epoch": 0.8504201680672269, + "grad_norm": 3.3295569480928515, + "learning_rate": 6.164017752057972e-06, + "loss": 0.2512, + "step": 506 + }, + { + "epoch": 0.8521008403361344, + "grad_norm": 2.8205057844309795, + "learning_rate": 6.151176426716261e-06, + "loss": 0.2132, + "step": 507 + }, + { + "epoch": 0.853781512605042, + "grad_norm": 3.313681962254624, + "learning_rate": 6.13832707817751e-06, + "loss": 0.2558, + "step": 508 + }, + { + "epoch": 0.8554621848739495, + "grad_norm": 2.8394597548886447, + "learning_rate": 6.125469795996065e-06, + "loss": 0.1857, + "step": 509 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 3.619219896462597, + "learning_rate": 6.112604669781572e-06, + "loss": 0.3292, + "step": 510 + }, + { + "epoch": 0.8588235294117647, + "grad_norm": 3.2255068003001677, + "learning_rate": 6.099731789198344e-06, + "loss": 0.1973, + "step": 511 + }, + { + "epoch": 0.8605042016806723, + "grad_norm": 3.226780905842978, + "learning_rate": 6.0868512439647345e-06, + "loss": 0.2322, + "step": 512 + }, + { + "epoch": 0.8621848739495799, + "grad_norm": 3.0735886445268297, + "learning_rate": 6.073963123852522e-06, + "loss": 0.2253, + "step": 513 + }, + { + "epoch": 0.8638655462184874, + "grad_norm": 3.2508167941737205, + "learning_rate": 6.061067518686277e-06, + "loss": 0.2204, + "step": 514 + }, + { + "epoch": 0.865546218487395, + "grad_norm": 3.48647312788438, + "learning_rate": 6.048164518342734e-06, + "loss": 0.2611, + "step": 515 + }, + { + "epoch": 0.8672268907563025, + "grad_norm": 3.9021723359330163, + "learning_rate": 6.035254212750172e-06, + "loss": 0.2766, + "step": 516 + }, + { + "epoch": 0.8689075630252101, + "grad_norm": 3.848856259653677, + "learning_rate": 6.022336691887785e-06, + "loss": 0.2467, + "step": 517 + }, + { + "epoch": 0.8705882352941177, + "grad_norm": 3.2427477527780995, + "learning_rate": 6.009412045785051e-06, + "loss": 0.2417, + "step": 518 + }, + { + "epoch": 0.8722689075630252, + "grad_norm": 3.186572269145269, + "learning_rate": 5.996480364521114e-06, + "loss": 0.2362, + "step": 519 + }, + { + "epoch": 0.8739495798319328, + "grad_norm": 3.4881108893153194, + "learning_rate": 5.983541738224141e-06, + "loss": 0.2492, + "step": 520 + }, + { + "epoch": 0.8756302521008403, + "grad_norm": 3.451309092320012, + "learning_rate": 5.970596257070711e-06, + "loss": 0.2459, + "step": 521 + }, + { + "epoch": 0.8773109243697479, + "grad_norm": 3.036808532165684, + "learning_rate": 5.957644011285173e-06, + "loss": 0.2303, + "step": 522 + }, + { + "epoch": 0.8789915966386554, + "grad_norm": 4.380816137152236, + "learning_rate": 5.944685091139026e-06, + "loss": 0.2671, + "step": 523 + }, + { + "epoch": 0.880672268907563, + "grad_norm": 3.5404271040433892, + "learning_rate": 5.931719586950286e-06, + "loss": 0.2766, + "step": 524 + }, + { + "epoch": 0.8823529411764706, + "grad_norm": 3.413794531222058, + "learning_rate": 5.918747589082853e-06, + "loss": 0.2001, + "step": 525 + }, + { + "epoch": 0.8840336134453781, + "grad_norm": 3.1862842821300834, + "learning_rate": 5.905769187945889e-06, + "loss": 0.2606, + "step": 526 + }, + { + "epoch": 0.8857142857142857, + "grad_norm": 3.0322653693123347, + "learning_rate": 5.892784473993184e-06, + "loss": 0.183, + "step": 527 + }, + { + "epoch": 0.8873949579831932, + "grad_norm": 3.160034382975318, + "learning_rate": 5.879793537722525e-06, + "loss": 0.2651, + "step": 528 + }, + { + "epoch": 0.8890756302521008, + "grad_norm": 3.8858225696631323, + "learning_rate": 5.8667964696750625e-06, + "loss": 0.2103, + "step": 529 + }, + { + "epoch": 0.8907563025210085, + "grad_norm": 4.046308603443172, + "learning_rate": 5.853793360434687e-06, + "loss": 0.2899, + "step": 530 + }, + { + "epoch": 0.892436974789916, + "grad_norm": 3.490280050594676, + "learning_rate": 5.840784300627396e-06, + "loss": 0.2757, + "step": 531 + }, + { + "epoch": 0.8941176470588236, + "grad_norm": 3.4886595872843698, + "learning_rate": 5.82776938092065e-06, + "loss": 0.2226, + "step": 532 + }, + { + "epoch": 0.8957983193277311, + "grad_norm": 3.459447687093064, + "learning_rate": 5.814748692022761e-06, + "loss": 0.2096, + "step": 533 + }, + { + "epoch": 0.8974789915966387, + "grad_norm": 3.4503094466900843, + "learning_rate": 5.801722324682243e-06, + "loss": 0.2201, + "step": 534 + }, + { + "epoch": 0.8991596638655462, + "grad_norm": 6.057225321208548, + "learning_rate": 5.788690369687188e-06, + "loss": 0.2534, + "step": 535 + }, + { + "epoch": 0.9008403361344538, + "grad_norm": 3.993936335248651, + "learning_rate": 5.775652917864633e-06, + "loss": 0.2468, + "step": 536 + }, + { + "epoch": 0.9025210084033614, + "grad_norm": 3.8240034384038957, + "learning_rate": 5.762610060079926e-06, + "loss": 0.2468, + "step": 537 + }, + { + "epoch": 0.9042016806722689, + "grad_norm": 3.506076848751469, + "learning_rate": 5.749561887236088e-06, + "loss": 0.2564, + "step": 538 + }, + { + "epoch": 0.9058823529411765, + "grad_norm": 3.4486725300170327, + "learning_rate": 5.736508490273189e-06, + "loss": 0.2775, + "step": 539 + }, + { + "epoch": 0.907563025210084, + "grad_norm": 3.6027114982004584, + "learning_rate": 5.723449960167703e-06, + "loss": 0.2172, + "step": 540 + }, + { + "epoch": 0.9092436974789916, + "grad_norm": 3.922464109691482, + "learning_rate": 5.710386387931886e-06, + "loss": 0.2995, + "step": 541 + }, + { + "epoch": 0.9109243697478991, + "grad_norm": 4.202065560048001, + "learning_rate": 5.697317864613127e-06, + "loss": 0.3745, + "step": 542 + }, + { + "epoch": 0.9126050420168067, + "grad_norm": 4.135603360147499, + "learning_rate": 5.684244481293335e-06, + "loss": 0.3435, + "step": 543 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 3.4093772473414066, + "learning_rate": 5.671166329088278e-06, + "loss": 0.2263, + "step": 544 + }, + { + "epoch": 0.9159663865546218, + "grad_norm": 3.5010536567581023, + "learning_rate": 5.658083499146968e-06, + "loss": 0.2573, + "step": 545 + }, + { + "epoch": 0.9176470588235294, + "grad_norm": 3.6193713170190582, + "learning_rate": 5.644996082651018e-06, + "loss": 0.2463, + "step": 546 + }, + { + "epoch": 0.9193277310924369, + "grad_norm": 3.4389315001827905, + "learning_rate": 5.6319041708140045e-06, + "loss": 0.1654, + "step": 547 + }, + { + "epoch": 0.9210084033613445, + "grad_norm": 3.269477559777336, + "learning_rate": 5.6188078548808366e-06, + "loss": 0.2372, + "step": 548 + }, + { + "epoch": 0.9226890756302522, + "grad_norm": 2.949744609372753, + "learning_rate": 5.6057072261271194e-06, + "loss": 0.1651, + "step": 549 + }, + { + "epoch": 0.9243697478991597, + "grad_norm": 3.284532781268528, + "learning_rate": 5.592602375858515e-06, + "loss": 0.3667, + "step": 550 + }, + { + "epoch": 0.9260504201680673, + "grad_norm": 3.033910900385078, + "learning_rate": 5.579493395410105e-06, + "loss": 0.2187, + "step": 551 + }, + { + "epoch": 0.9277310924369748, + "grad_norm": 3.8363812188814275, + "learning_rate": 5.566380376145762e-06, + "loss": 0.2711, + "step": 552 + }, + { + "epoch": 0.9294117647058824, + "grad_norm": 3.206650579866936, + "learning_rate": 5.553263409457504e-06, + "loss": 0.1997, + "step": 553 + }, + { + "epoch": 0.9310924369747899, + "grad_norm": 3.935436124623862, + "learning_rate": 5.540142586764862e-06, + "loss": 0.2661, + "step": 554 + }, + { + "epoch": 0.9327731092436975, + "grad_norm": 3.6180857295008395, + "learning_rate": 5.527017999514239e-06, + "loss": 0.2474, + "step": 555 + }, + { + "epoch": 0.934453781512605, + "grad_norm": 3.459755458614064, + "learning_rate": 5.51388973917828e-06, + "loss": 0.2821, + "step": 556 + }, + { + "epoch": 0.9361344537815126, + "grad_norm": 3.5773382583048527, + "learning_rate": 5.5007578972552246e-06, + "loss": 0.2927, + "step": 557 + }, + { + "epoch": 0.9378151260504202, + "grad_norm": 4.6363961630236785, + "learning_rate": 5.4876225652682776e-06, + "loss": 0.3551, + "step": 558 + }, + { + "epoch": 0.9394957983193277, + "grad_norm": 3.379098842911059, + "learning_rate": 5.474483834764968e-06, + "loss": 0.2059, + "step": 559 + }, + { + "epoch": 0.9411764705882353, + "grad_norm": 3.9316584340639245, + "learning_rate": 5.46134179731651e-06, + "loss": 0.2953, + "step": 560 + }, + { + "epoch": 0.9428571428571428, + "grad_norm": 3.3722167978017867, + "learning_rate": 5.448196544517168e-06, + "loss": 0.2451, + "step": 561 + }, + { + "epoch": 0.9445378151260504, + "grad_norm": 3.1185440228092887, + "learning_rate": 5.435048167983613e-06, + "loss": 0.2173, + "step": 562 + }, + { + "epoch": 0.946218487394958, + "grad_norm": 3.9485848008617523, + "learning_rate": 5.421896759354288e-06, + "loss": 0.2261, + "step": 563 + }, + { + "epoch": 0.9478991596638655, + "grad_norm": 2.9170922035357094, + "learning_rate": 5.408742410288769e-06, + "loss": 0.2287, + "step": 564 + }, + { + "epoch": 0.9495798319327731, + "grad_norm": 3.9244596118525674, + "learning_rate": 5.395585212467124e-06, + "loss": 0.2734, + "step": 565 + }, + { + "epoch": 0.9512605042016806, + "grad_norm": 4.603168201003037, + "learning_rate": 5.382425257589277e-06, + "loss": 0.3128, + "step": 566 + }, + { + "epoch": 0.9529411764705882, + "grad_norm": 3.818573488204182, + "learning_rate": 5.36926263737437e-06, + "loss": 0.3068, + "step": 567 + }, + { + "epoch": 0.9546218487394958, + "grad_norm": 2.467008222654659, + "learning_rate": 5.356097443560116e-06, + "loss": 0.2085, + "step": 568 + }, + { + "epoch": 0.9563025210084034, + "grad_norm": 3.8596036485613427, + "learning_rate": 5.342929767902168e-06, + "loss": 0.2432, + "step": 569 + }, + { + "epoch": 0.957983193277311, + "grad_norm": 4.001034423912602, + "learning_rate": 5.329759702173477e-06, + "loss": 0.2557, + "step": 570 + }, + { + "epoch": 0.9596638655462185, + "grad_norm": 3.4761405695641145, + "learning_rate": 5.316587338163649e-06, + "loss": 0.2165, + "step": 571 + }, + { + "epoch": 0.9613445378151261, + "grad_norm": 3.1581349516312316, + "learning_rate": 5.30341276767831e-06, + "loss": 0.2046, + "step": 572 + }, + { + "epoch": 0.9630252100840336, + "grad_norm": 3.717769612339884, + "learning_rate": 5.290236082538464e-06, + "loss": 0.2878, + "step": 573 + }, + { + "epoch": 0.9647058823529412, + "grad_norm": 3.8206293915711544, + "learning_rate": 5.27705737457985e-06, + "loss": 0.284, + "step": 574 + }, + { + "epoch": 0.9663865546218487, + "grad_norm": 3.5987474263840293, + "learning_rate": 5.2638767356523125e-06, + "loss": 0.1862, + "step": 575 + }, + { + "epoch": 0.9680672268907563, + "grad_norm": 2.628708627577784, + "learning_rate": 5.2506942576191466e-06, + "loss": 0.1986, + "step": 576 + }, + { + "epoch": 0.9697478991596639, + "grad_norm": 3.00948049798796, + "learning_rate": 5.23751003235647e-06, + "loss": 0.2272, + "step": 577 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 3.897035347258241, + "learning_rate": 5.224324151752575e-06, + "loss": 0.2377, + "step": 578 + }, + { + "epoch": 0.973109243697479, + "grad_norm": 3.1171554333300646, + "learning_rate": 5.211136707707293e-06, + "loss": 0.2222, + "step": 579 + }, + { + "epoch": 0.9747899159663865, + "grad_norm": 3.0039037163365983, + "learning_rate": 5.197947792131348e-06, + "loss": 0.1865, + "step": 580 + }, + { + "epoch": 0.9764705882352941, + "grad_norm": 6.257110222572669, + "learning_rate": 5.184757496945726e-06, + "loss": 0.2839, + "step": 581 + }, + { + "epoch": 0.9781512605042016, + "grad_norm": 3.821218218364182, + "learning_rate": 5.1715659140810225e-06, + "loss": 0.2529, + "step": 582 + }, + { + "epoch": 0.9798319327731092, + "grad_norm": 3.730030674136909, + "learning_rate": 5.158373135476811e-06, + "loss": 0.2524, + "step": 583 + }, + { + "epoch": 0.9815126050420168, + "grad_norm": 2.840325890140036, + "learning_rate": 5.145179253080997e-06, + "loss": 0.2176, + "step": 584 + }, + { + "epoch": 0.9831932773109243, + "grad_norm": 3.6607002570737817, + "learning_rate": 5.131984358849182e-06, + "loss": 0.2658, + "step": 585 + }, + { + "epoch": 0.984873949579832, + "grad_norm": 3.1743768713711322, + "learning_rate": 5.118788544744016e-06, + "loss": 0.2591, + "step": 586 + }, + { + "epoch": 0.9865546218487395, + "grad_norm": 3.2338897708714507, + "learning_rate": 5.105591902734561e-06, + "loss": 0.1775, + "step": 587 + }, + { + "epoch": 0.9882352941176471, + "grad_norm": 4.517369770013792, + "learning_rate": 5.09239452479565e-06, + "loss": 0.2587, + "step": 588 + }, + { + "epoch": 0.9899159663865547, + "grad_norm": 2.6496184847532733, + "learning_rate": 5.079196502907246e-06, + "loss": 0.1447, + "step": 589 + }, + { + "epoch": 0.9915966386554622, + "grad_norm": 2.7211633275819525, + "learning_rate": 5.065997929053795e-06, + "loss": 0.1926, + "step": 590 + }, + { + "epoch": 0.9932773109243698, + "grad_norm": 3.0952400745185, + "learning_rate": 5.052798895223597e-06, + "loss": 0.2134, + "step": 591 + }, + { + "epoch": 0.9949579831932773, + "grad_norm": 3.51837877201114, + "learning_rate": 5.039599493408154e-06, + "loss": 0.2745, + "step": 592 + }, + { + "epoch": 0.9966386554621849, + "grad_norm": 3.4866227884563026, + "learning_rate": 5.026399815601533e-06, + "loss": 0.2295, + "step": 593 + }, + { + "epoch": 0.9983193277310924, + "grad_norm": 3.3229672421715644, + "learning_rate": 5.0131999537997235e-06, + "loss": 0.1775, + "step": 594 + }, + { + "epoch": 1.0, + "grad_norm": 3.6363741518245245, + "learning_rate": 5e-06, + "loss": 0.1833, + "step": 595 + }, + { + "epoch": 1.0016806722689076, + "grad_norm": 3.0685430970214216, + "learning_rate": 4.986800046200278e-06, + "loss": 0.1295, + "step": 596 + }, + { + "epoch": 1.0033613445378151, + "grad_norm": 2.642341004416497, + "learning_rate": 4.97360018439847e-06, + "loss": 0.0994, + "step": 597 + }, + { + "epoch": 1.0050420168067227, + "grad_norm": 2.323376314615566, + "learning_rate": 4.960400506591848e-06, + "loss": 0.075, + "step": 598 + }, + { + "epoch": 1.0067226890756302, + "grad_norm": 2.1934167410322574, + "learning_rate": 4.947201104776404e-06, + "loss": 0.0863, + "step": 599 + }, + { + "epoch": 1.0084033613445378, + "grad_norm": 2.779595805074063, + "learning_rate": 4.934002070946206e-06, + "loss": 0.09, + "step": 600 + }, + { + "epoch": 1.0100840336134453, + "grad_norm": 2.642478794924014, + "learning_rate": 4.920803497092757e-06, + "loss": 0.1072, + "step": 601 + }, + { + "epoch": 1.011764705882353, + "grad_norm": 2.8206325875789786, + "learning_rate": 4.907605475204352e-06, + "loss": 0.1135, + "step": 602 + }, + { + "epoch": 1.0134453781512605, + "grad_norm": 2.601945372181, + "learning_rate": 4.894408097265441e-06, + "loss": 0.1026, + "step": 603 + }, + { + "epoch": 1.015126050420168, + "grad_norm": 2.651228941391384, + "learning_rate": 4.881211455255986e-06, + "loss": 0.0731, + "step": 604 + }, + { + "epoch": 1.0168067226890756, + "grad_norm": 2.3414792685919994, + "learning_rate": 4.86801564115082e-06, + "loss": 0.0886, + "step": 605 + }, + { + "epoch": 1.0184873949579831, + "grad_norm": 2.7273995366520243, + "learning_rate": 4.854820746919005e-06, + "loss": 0.0911, + "step": 606 + }, + { + "epoch": 1.0201680672268907, + "grad_norm": 2.8327315486397113, + "learning_rate": 4.8416268645231915e-06, + "loss": 0.1086, + "step": 607 + }, + { + "epoch": 1.0218487394957982, + "grad_norm": 2.7069106323936256, + "learning_rate": 4.82843408591898e-06, + "loss": 0.097, + "step": 608 + }, + { + "epoch": 1.0235294117647058, + "grad_norm": 2.442047706770237, + "learning_rate": 4.815242503054277e-06, + "loss": 0.0607, + "step": 609 + }, + { + "epoch": 1.0252100840336134, + "grad_norm": 3.033523413937561, + "learning_rate": 4.802052207868654e-06, + "loss": 0.1112, + "step": 610 + }, + { + "epoch": 1.026890756302521, + "grad_norm": 3.820262837255327, + "learning_rate": 4.78886329229271e-06, + "loss": 0.137, + "step": 611 + }, + { + "epoch": 1.0285714285714285, + "grad_norm": 2.733397229688364, + "learning_rate": 4.775675848247427e-06, + "loss": 0.0923, + "step": 612 + }, + { + "epoch": 1.030252100840336, + "grad_norm": 3.4691034150038034, + "learning_rate": 4.762489967643532e-06, + "loss": 0.118, + "step": 613 + }, + { + "epoch": 1.0319327731092436, + "grad_norm": 3.854013776962153, + "learning_rate": 4.749305742380853e-06, + "loss": 0.1301, + "step": 614 + }, + { + "epoch": 1.0336134453781514, + "grad_norm": 3.0184243827182833, + "learning_rate": 4.736123264347688e-06, + "loss": 0.0853, + "step": 615 + }, + { + "epoch": 1.035294117647059, + "grad_norm": 3.7517471305566517, + "learning_rate": 4.7229426254201504e-06, + "loss": 0.1193, + "step": 616 + }, + { + "epoch": 1.0369747899159665, + "grad_norm": 3.3080181370358566, + "learning_rate": 4.709763917461537e-06, + "loss": 0.1154, + "step": 617 + }, + { + "epoch": 1.038655462184874, + "grad_norm": 2.2629000826778856, + "learning_rate": 4.696587232321691e-06, + "loss": 0.0932, + "step": 618 + }, + { + "epoch": 1.0403361344537816, + "grad_norm": 3.1860739367441737, + "learning_rate": 4.683412661836351e-06, + "loss": 0.1039, + "step": 619 + }, + { + "epoch": 1.0420168067226891, + "grad_norm": 3.0601501008008256, + "learning_rate": 4.6702402978265235e-06, + "loss": 0.089, + "step": 620 + }, + { + "epoch": 1.0436974789915967, + "grad_norm": 2.603230804321902, + "learning_rate": 4.657070232097832e-06, + "loss": 0.0751, + "step": 621 + }, + { + "epoch": 1.0453781512605043, + "grad_norm": 2.8411688842957465, + "learning_rate": 4.643902556439885e-06, + "loss": 0.0817, + "step": 622 + }, + { + "epoch": 1.0470588235294118, + "grad_norm": 2.7986713354700763, + "learning_rate": 4.630737362625631e-06, + "loss": 0.1058, + "step": 623 + }, + { + "epoch": 1.0487394957983194, + "grad_norm": 3.5321634211727333, + "learning_rate": 4.6175747424107234e-06, + "loss": 0.1169, + "step": 624 + }, + { + "epoch": 1.050420168067227, + "grad_norm": 2.9302195577434462, + "learning_rate": 4.604414787532877e-06, + "loss": 0.1098, + "step": 625 + }, + { + "epoch": 1.0521008403361345, + "grad_norm": 2.2686346359628344, + "learning_rate": 4.591257589711233e-06, + "loss": 0.078, + "step": 626 + }, + { + "epoch": 1.053781512605042, + "grad_norm": 3.2727476625085266, + "learning_rate": 4.578103240645714e-06, + "loss": 0.1095, + "step": 627 + }, + { + "epoch": 1.0554621848739496, + "grad_norm": 2.341371455709515, + "learning_rate": 4.5649518320163885e-06, + "loss": 0.1126, + "step": 628 + }, + { + "epoch": 1.0571428571428572, + "grad_norm": 2.738651087640409, + "learning_rate": 4.551803455482833e-06, + "loss": 0.086, + "step": 629 + }, + { + "epoch": 1.0588235294117647, + "grad_norm": 2.3041964936668555, + "learning_rate": 4.53865820268349e-06, + "loss": 0.0822, + "step": 630 + }, + { + "epoch": 1.0605042016806723, + "grad_norm": 4.220651913549916, + "learning_rate": 4.525516165235034e-06, + "loss": 0.1141, + "step": 631 + }, + { + "epoch": 1.0621848739495798, + "grad_norm": 2.9322019645529904, + "learning_rate": 4.512377434731724e-06, + "loss": 0.1477, + "step": 632 + }, + { + "epoch": 1.0638655462184874, + "grad_norm": 2.5913843452442378, + "learning_rate": 4.499242102744777e-06, + "loss": 0.0929, + "step": 633 + }, + { + "epoch": 1.065546218487395, + "grad_norm": 2.445571807332056, + "learning_rate": 4.486110260821722e-06, + "loss": 0.0978, + "step": 634 + }, + { + "epoch": 1.0672268907563025, + "grad_norm": 2.5494116675856366, + "learning_rate": 4.4729820004857625e-06, + "loss": 0.1012, + "step": 635 + }, + { + "epoch": 1.06890756302521, + "grad_norm": 2.1101261139846783, + "learning_rate": 4.45985741323514e-06, + "loss": 0.0715, + "step": 636 + }, + { + "epoch": 1.0705882352941176, + "grad_norm": 1.9510646778084884, + "learning_rate": 4.446736590542497e-06, + "loss": 0.0549, + "step": 637 + }, + { + "epoch": 1.0722689075630252, + "grad_norm": 3.88179760628692, + "learning_rate": 4.4336196238542394e-06, + "loss": 0.1449, + "step": 638 + }, + { + "epoch": 1.0739495798319327, + "grad_norm": 2.541597748187505, + "learning_rate": 4.420506604589897e-06, + "loss": 0.0913, + "step": 639 + }, + { + "epoch": 1.0756302521008403, + "grad_norm": 3.519215369238923, + "learning_rate": 4.407397624141487e-06, + "loss": 0.1025, + "step": 640 + }, + { + "epoch": 1.0773109243697478, + "grad_norm": 2.123778059115166, + "learning_rate": 4.394292773872882e-06, + "loss": 0.0714, + "step": 641 + }, + { + "epoch": 1.0789915966386554, + "grad_norm": 5.1179373437116285, + "learning_rate": 4.381192145119165e-06, + "loss": 0.1282, + "step": 642 + }, + { + "epoch": 1.080672268907563, + "grad_norm": 2.8684696011439756, + "learning_rate": 4.368095829185999e-06, + "loss": 0.0791, + "step": 643 + }, + { + "epoch": 1.0823529411764705, + "grad_norm": 4.102832197816667, + "learning_rate": 4.355003917348985e-06, + "loss": 0.1225, + "step": 644 + }, + { + "epoch": 1.084033613445378, + "grad_norm": 3.6293736374607013, + "learning_rate": 4.341916500853034e-06, + "loss": 0.1105, + "step": 645 + }, + { + "epoch": 1.0857142857142856, + "grad_norm": 2.9599263387812873, + "learning_rate": 4.3288336709117246e-06, + "loss": 0.0765, + "step": 646 + }, + { + "epoch": 1.0873949579831932, + "grad_norm": 2.959898151373633, + "learning_rate": 4.315755518706667e-06, + "loss": 0.1221, + "step": 647 + }, + { + "epoch": 1.0890756302521007, + "grad_norm": 2.831223871678987, + "learning_rate": 4.302682135386873e-06, + "loss": 0.1047, + "step": 648 + }, + { + "epoch": 1.0907563025210083, + "grad_norm": 3.883463233494954, + "learning_rate": 4.289613612068118e-06, + "loss": 0.0778, + "step": 649 + }, + { + "epoch": 1.092436974789916, + "grad_norm": 3.3740664710443315, + "learning_rate": 4.276550039832299e-06, + "loss": 0.1244, + "step": 650 + }, + { + "epoch": 1.0941176470588236, + "grad_norm": 2.945414543907692, + "learning_rate": 4.263491509726812e-06, + "loss": 0.1062, + "step": 651 + }, + { + "epoch": 1.0957983193277312, + "grad_norm": 3.5755203026507107, + "learning_rate": 4.250438112763911e-06, + "loss": 0.0765, + "step": 652 + }, + { + "epoch": 1.0974789915966388, + "grad_norm": 2.864562468571205, + "learning_rate": 4.237389939920075e-06, + "loss": 0.0973, + "step": 653 + }, + { + "epoch": 1.0991596638655463, + "grad_norm": 3.309117703284289, + "learning_rate": 4.224347082135367e-06, + "loss": 0.1071, + "step": 654 + }, + { + "epoch": 1.1008403361344539, + "grad_norm": 4.444191677418515, + "learning_rate": 4.211309630312813e-06, + "loss": 0.1255, + "step": 655 + }, + { + "epoch": 1.1025210084033614, + "grad_norm": 2.6049652407237525, + "learning_rate": 4.198277675317758e-06, + "loss": 0.0953, + "step": 656 + }, + { + "epoch": 1.104201680672269, + "grad_norm": 2.9301571986545984, + "learning_rate": 4.185251307977241e-06, + "loss": 0.1391, + "step": 657 + }, + { + "epoch": 1.1058823529411765, + "grad_norm": 2.991359373447863, + "learning_rate": 4.17223061907935e-06, + "loss": 0.1011, + "step": 658 + }, + { + "epoch": 1.107563025210084, + "grad_norm": 2.908536947858802, + "learning_rate": 4.159215699372605e-06, + "loss": 0.0994, + "step": 659 + }, + { + "epoch": 1.1092436974789917, + "grad_norm": 3.0562253978122, + "learning_rate": 4.146206639565313e-06, + "loss": 0.0761, + "step": 660 + }, + { + "epoch": 1.1109243697478992, + "grad_norm": 2.3707071277066865, + "learning_rate": 4.133203530324938e-06, + "loss": 0.0791, + "step": 661 + }, + { + "epoch": 1.1126050420168068, + "grad_norm": 2.568827912231571, + "learning_rate": 4.120206462277478e-06, + "loss": 0.1079, + "step": 662 + }, + { + "epoch": 1.1142857142857143, + "grad_norm": 3.3675762016216346, + "learning_rate": 4.107215526006818e-06, + "loss": 0.1004, + "step": 663 + }, + { + "epoch": 1.1159663865546219, + "grad_norm": 2.711315362662935, + "learning_rate": 4.094230812054113e-06, + "loss": 0.087, + "step": 664 + }, + { + "epoch": 1.1176470588235294, + "grad_norm": 2.4363571585356296, + "learning_rate": 4.081252410917148e-06, + "loss": 0.0993, + "step": 665 + }, + { + "epoch": 1.119327731092437, + "grad_norm": 2.2853255955328327, + "learning_rate": 4.068280413049715e-06, + "loss": 0.0951, + "step": 666 + }, + { + "epoch": 1.1210084033613446, + "grad_norm": 3.1173097538332257, + "learning_rate": 4.0553149088609745e-06, + "loss": 0.1207, + "step": 667 + }, + { + "epoch": 1.122689075630252, + "grad_norm": 2.3961886874666645, + "learning_rate": 4.042355988714828e-06, + "loss": 0.0984, + "step": 668 + }, + { + "epoch": 1.1243697478991597, + "grad_norm": 2.6441840658169196, + "learning_rate": 4.02940374292929e-06, + "loss": 0.1055, + "step": 669 + }, + { + "epoch": 1.1260504201680672, + "grad_norm": 2.589296721112373, + "learning_rate": 4.01645826177586e-06, + "loss": 0.091, + "step": 670 + }, + { + "epoch": 1.1277310924369748, + "grad_norm": 3.101644326303377, + "learning_rate": 4.003519635478889e-06, + "loss": 0.1136, + "step": 671 + }, + { + "epoch": 1.1294117647058823, + "grad_norm": 2.085544584261255, + "learning_rate": 3.99058795421495e-06, + "loss": 0.0645, + "step": 672 + }, + { + "epoch": 1.13109243697479, + "grad_norm": 3.022884264669196, + "learning_rate": 3.977663308112216e-06, + "loss": 0.0982, + "step": 673 + }, + { + "epoch": 1.1327731092436975, + "grad_norm": 3.0918298593703177, + "learning_rate": 3.9647457872498295e-06, + "loss": 0.1309, + "step": 674 + }, + { + "epoch": 1.134453781512605, + "grad_norm": 3.097444725372621, + "learning_rate": 3.951835481657268e-06, + "loss": 0.0821, + "step": 675 + }, + { + "epoch": 1.1361344537815126, + "grad_norm": 3.480720174843856, + "learning_rate": 3.938932481313725e-06, + "loss": 0.1083, + "step": 676 + }, + { + "epoch": 1.1378151260504201, + "grad_norm": 3.167987925313736, + "learning_rate": 3.926036876147479e-06, + "loss": 0.0947, + "step": 677 + }, + { + "epoch": 1.1394957983193277, + "grad_norm": 2.9675142893422666, + "learning_rate": 3.913148756035266e-06, + "loss": 0.1074, + "step": 678 + }, + { + "epoch": 1.1411764705882352, + "grad_norm": 2.88904471583098, + "learning_rate": 3.9002682108016585e-06, + "loss": 0.0947, + "step": 679 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 2.3148945814778745, + "learning_rate": 3.887395330218429e-06, + "loss": 0.0543, + "step": 680 + }, + { + "epoch": 1.1445378151260504, + "grad_norm": 3.4421881750813714, + "learning_rate": 3.874530204003937e-06, + "loss": 0.1261, + "step": 681 + }, + { + "epoch": 1.146218487394958, + "grad_norm": 2.426782546704889, + "learning_rate": 3.861672921822493e-06, + "loss": 0.0701, + "step": 682 + }, + { + "epoch": 1.1478991596638655, + "grad_norm": 2.7112499258139673, + "learning_rate": 3.848823573283742e-06, + "loss": 0.0713, + "step": 683 + }, + { + "epoch": 1.149579831932773, + "grad_norm": 2.0739016767387057, + "learning_rate": 3.83598224794203e-06, + "loss": 0.0649, + "step": 684 + }, + { + "epoch": 1.1512605042016806, + "grad_norm": 4.360074547910969, + "learning_rate": 3.823149035295789e-06, + "loss": 0.1263, + "step": 685 + }, + { + "epoch": 1.1529411764705881, + "grad_norm": 3.8208016083874803, + "learning_rate": 3.8103240247869077e-06, + "loss": 0.1078, + "step": 686 + }, + { + "epoch": 1.1546218487394957, + "grad_norm": 3.136985722111122, + "learning_rate": 3.7975073058001092e-06, + "loss": 0.0801, + "step": 687 + }, + { + "epoch": 1.1563025210084033, + "grad_norm": 2.7226211242981035, + "learning_rate": 3.7846989676623265e-06, + "loss": 0.0937, + "step": 688 + }, + { + "epoch": 1.1579831932773108, + "grad_norm": 2.807183390633048, + "learning_rate": 3.7718990996420875e-06, + "loss": 0.0805, + "step": 689 + }, + { + "epoch": 1.1596638655462184, + "grad_norm": 2.7048742007757234, + "learning_rate": 3.7591077909488817e-06, + "loss": 0.0731, + "step": 690 + }, + { + "epoch": 1.1613445378151261, + "grad_norm": 2.4917292393625776, + "learning_rate": 3.7463251307325432e-06, + "loss": 0.0814, + "step": 691 + }, + { + "epoch": 1.1630252100840337, + "grad_norm": 2.863341194127507, + "learning_rate": 3.7335512080826363e-06, + "loss": 0.1151, + "step": 692 + }, + { + "epoch": 1.1647058823529413, + "grad_norm": 2.9183758099606245, + "learning_rate": 3.720786112027822e-06, + "loss": 0.0995, + "step": 693 + }, + { + "epoch": 1.1663865546218488, + "grad_norm": 2.5864607722539232, + "learning_rate": 3.708029931535246e-06, + "loss": 0.0796, + "step": 694 + }, + { + "epoch": 1.1680672268907564, + "grad_norm": 3.468324873932469, + "learning_rate": 3.695282755509917e-06, + "loss": 0.118, + "step": 695 + }, + { + "epoch": 1.169747899159664, + "grad_norm": 2.1915066606153846, + "learning_rate": 3.682544672794085e-06, + "loss": 0.0823, + "step": 696 + }, + { + "epoch": 1.1714285714285715, + "grad_norm": 2.8614797614523195, + "learning_rate": 3.669815772166625e-06, + "loss": 0.1117, + "step": 697 + }, + { + "epoch": 1.173109243697479, + "grad_norm": 3.0243317238966902, + "learning_rate": 3.6570961423424155e-06, + "loss": 0.1382, + "step": 698 + }, + { + "epoch": 1.1747899159663866, + "grad_norm": 2.5722750997684805, + "learning_rate": 3.644385871971725e-06, + "loss": 0.0901, + "step": 699 + }, + { + "epoch": 1.1764705882352942, + "grad_norm": 2.2211603476786945, + "learning_rate": 3.6316850496395863e-06, + "loss": 0.0707, + "step": 700 + }, + { + "epoch": 1.1781512605042017, + "grad_norm": 2.3653294803726754, + "learning_rate": 3.618993763865185e-06, + "loss": 0.0865, + "step": 701 + }, + { + "epoch": 1.1798319327731093, + "grad_norm": 3.4518859809747746, + "learning_rate": 3.6063121031012417e-06, + "loss": 0.0889, + "step": 702 + }, + { + "epoch": 1.1815126050420168, + "grad_norm": 3.1717027674030236, + "learning_rate": 3.5936401557333957e-06, + "loss": 0.1078, + "step": 703 + }, + { + "epoch": 1.1831932773109244, + "grad_norm": 2.5802159610520667, + "learning_rate": 3.5809780100795853e-06, + "loss": 0.0883, + "step": 704 + }, + { + "epoch": 1.184873949579832, + "grad_norm": 2.51669538204389, + "learning_rate": 3.568325754389438e-06, + "loss": 0.1018, + "step": 705 + }, + { + "epoch": 1.1865546218487395, + "grad_norm": 3.834199640369919, + "learning_rate": 3.5556834768436498e-06, + "loss": 0.1229, + "step": 706 + }, + { + "epoch": 1.188235294117647, + "grad_norm": 3.268611445606652, + "learning_rate": 3.5430512655533774e-06, + "loss": 0.1422, + "step": 707 + }, + { + "epoch": 1.1899159663865546, + "grad_norm": 2.569460634045658, + "learning_rate": 3.5304292085596166e-06, + "loss": 0.0842, + "step": 708 + }, + { + "epoch": 1.1915966386554622, + "grad_norm": 2.708337321883567, + "learning_rate": 3.517817393832593e-06, + "loss": 0.1071, + "step": 709 + }, + { + "epoch": 1.1932773109243697, + "grad_norm": 2.295889437868396, + "learning_rate": 3.505215909271149e-06, + "loss": 0.0842, + "step": 710 + }, + { + "epoch": 1.1949579831932773, + "grad_norm": 2.7494736128846315, + "learning_rate": 3.4926248427021293e-06, + "loss": 0.1067, + "step": 711 + }, + { + "epoch": 1.1966386554621848, + "grad_norm": 3.037610849894837, + "learning_rate": 3.48004428187977e-06, + "loss": 0.0995, + "step": 712 + }, + { + "epoch": 1.1983193277310924, + "grad_norm": 3.1209681259023343, + "learning_rate": 3.4674743144850865e-06, + "loss": 0.1054, + "step": 713 + }, + { + "epoch": 1.2, + "grad_norm": 2.7146088777741353, + "learning_rate": 3.4549150281252635e-06, + "loss": 0.0855, + "step": 714 + }, + { + "epoch": 1.2016806722689075, + "grad_norm": 2.579928115495318, + "learning_rate": 3.442366510333043e-06, + "loss": 0.1029, + "step": 715 + }, + { + "epoch": 1.203361344537815, + "grad_norm": 2.051251447279275, + "learning_rate": 3.4298288485661147e-06, + "loss": 0.0724, + "step": 716 + }, + { + "epoch": 1.2050420168067226, + "grad_norm": 2.25884749164061, + "learning_rate": 3.417302130206507e-06, + "loss": 0.0998, + "step": 717 + }, + { + "epoch": 1.2067226890756302, + "grad_norm": 2.5203879016401047, + "learning_rate": 3.404786442559976e-06, + "loss": 0.0804, + "step": 718 + }, + { + "epoch": 1.2084033613445377, + "grad_norm": 2.323136581363694, + "learning_rate": 3.3922818728554008e-06, + "loss": 0.0889, + "step": 719 + }, + { + "epoch": 1.2100840336134453, + "grad_norm": 2.2224533138933404, + "learning_rate": 3.3797885082441717e-06, + "loss": 0.0809, + "step": 720 + }, + { + "epoch": 1.2117647058823529, + "grad_norm": 3.4942626722992585, + "learning_rate": 3.3673064357995844e-06, + "loss": 0.1375, + "step": 721 + }, + { + "epoch": 1.2134453781512604, + "grad_norm": 2.7421302991103182, + "learning_rate": 3.3548357425162347e-06, + "loss": 0.1026, + "step": 722 + }, + { + "epoch": 1.2151260504201682, + "grad_norm": 2.405476921093656, + "learning_rate": 3.3423765153094097e-06, + "loss": 0.0875, + "step": 723 + }, + { + "epoch": 1.2168067226890757, + "grad_norm": 1.9474423499915685, + "learning_rate": 3.3299288410144813e-06, + "loss": 0.0645, + "step": 724 + }, + { + "epoch": 1.2184873949579833, + "grad_norm": 3.3943630595361127, + "learning_rate": 3.3174928063863054e-06, + "loss": 0.0822, + "step": 725 + }, + { + "epoch": 1.2201680672268909, + "grad_norm": 2.6533504550178177, + "learning_rate": 3.3050684980986105e-06, + "loss": 0.0727, + "step": 726 + }, + { + "epoch": 1.2218487394957984, + "grad_norm": 2.6778789026347813, + "learning_rate": 3.292656002743405e-06, + "loss": 0.0752, + "step": 727 + }, + { + "epoch": 1.223529411764706, + "grad_norm": 1.9406512061812329, + "learning_rate": 3.2802554068303595e-06, + "loss": 0.0715, + "step": 728 + }, + { + "epoch": 1.2252100840336135, + "grad_norm": 2.356227596188501, + "learning_rate": 3.267866796786212e-06, + "loss": 0.0677, + "step": 729 + }, + { + "epoch": 1.226890756302521, + "grad_norm": 1.937645206088155, + "learning_rate": 3.255490258954167e-06, + "loss": 0.0371, + "step": 730 + }, + { + "epoch": 1.2285714285714286, + "grad_norm": 3.012214547683442, + "learning_rate": 3.2431258795932863e-06, + "loss": 0.1026, + "step": 731 + }, + { + "epoch": 1.2302521008403362, + "grad_norm": 3.2105396902791328, + "learning_rate": 3.2307737448778977e-06, + "loss": 0.0909, + "step": 732 + }, + { + "epoch": 1.2319327731092438, + "grad_norm": 2.599649894232782, + "learning_rate": 3.2184339408969857e-06, + "loss": 0.0682, + "step": 733 + }, + { + "epoch": 1.2336134453781513, + "grad_norm": 1.761237551093203, + "learning_rate": 3.206106553653596e-06, + "loss": 0.0482, + "step": 734 + }, + { + "epoch": 1.2352941176470589, + "grad_norm": 3.0540661926673494, + "learning_rate": 3.1937916690642356e-06, + "loss": 0.074, + "step": 735 + }, + { + "epoch": 1.2369747899159664, + "grad_norm": 2.6423585592436742, + "learning_rate": 3.181489372958272e-06, + "loss": 0.079, + "step": 736 + }, + { + "epoch": 1.238655462184874, + "grad_norm": 3.5166766786686767, + "learning_rate": 3.1691997510773376e-06, + "loss": 0.0993, + "step": 737 + }, + { + "epoch": 1.2403361344537815, + "grad_norm": 3.0030664598658494, + "learning_rate": 3.1569228890747305e-06, + "loss": 0.1061, + "step": 738 + }, + { + "epoch": 1.242016806722689, + "grad_norm": 2.1826647597969906, + "learning_rate": 3.1446588725148186e-06, + "loss": 0.0559, + "step": 739 + }, + { + "epoch": 1.2436974789915967, + "grad_norm": 3.8308938156876025, + "learning_rate": 3.132407786872442e-06, + "loss": 0.146, + "step": 740 + }, + { + "epoch": 1.2453781512605042, + "grad_norm": 3.540659391096697, + "learning_rate": 3.120169717532319e-06, + "loss": 0.128, + "step": 741 + }, + { + "epoch": 1.2470588235294118, + "grad_norm": 2.572082033443637, + "learning_rate": 3.107944749788449e-06, + "loss": 0.1095, + "step": 742 + }, + { + "epoch": 1.2487394957983193, + "grad_norm": 3.1039972961858466, + "learning_rate": 3.095732968843519e-06, + "loss": 0.1072, + "step": 743 + }, + { + "epoch": 1.250420168067227, + "grad_norm": 3.433630247978511, + "learning_rate": 3.0835344598083095e-06, + "loss": 0.0995, + "step": 744 + }, + { + "epoch": 1.2521008403361344, + "grad_norm": 2.9037148814677027, + "learning_rate": 3.0713493077011027e-06, + "loss": 0.0988, + "step": 745 + }, + { + "epoch": 1.253781512605042, + "grad_norm": 1.6255984404646961, + "learning_rate": 3.059177597447087e-06, + "loss": 0.0472, + "step": 746 + }, + { + "epoch": 1.2554621848739496, + "grad_norm": 2.8035161644835784, + "learning_rate": 3.0470194138777707e-06, + "loss": 0.0901, + "step": 747 + }, + { + "epoch": 1.2571428571428571, + "grad_norm": 3.0114191790076346, + "learning_rate": 3.0348748417303826e-06, + "loss": 0.1108, + "step": 748 + }, + { + "epoch": 1.2588235294117647, + "grad_norm": 2.6244638367291415, + "learning_rate": 3.0227439656472878e-06, + "loss": 0.0622, + "step": 749 + }, + { + "epoch": 1.2605042016806722, + "grad_norm": 2.0291708665569934, + "learning_rate": 3.0106268701753967e-06, + "loss": 0.0566, + "step": 750 + }, + { + "epoch": 1.2621848739495798, + "grad_norm": 2.8532612744730907, + "learning_rate": 2.9985236397655726e-06, + "loss": 0.0888, + "step": 751 + }, + { + "epoch": 1.2638655462184873, + "grad_norm": 2.2036167732649052, + "learning_rate": 2.986434358772048e-06, + "loss": 0.0546, + "step": 752 + }, + { + "epoch": 1.265546218487395, + "grad_norm": 2.7784477191560573, + "learning_rate": 2.974359111451831e-06, + "loss": 0.1147, + "step": 753 + }, + { + "epoch": 1.2672268907563025, + "grad_norm": 3.1299626086221033, + "learning_rate": 2.962297981964124e-06, + "loss": 0.1018, + "step": 754 + }, + { + "epoch": 1.26890756302521, + "grad_norm": 3.1824195697491935, + "learning_rate": 2.950251054369733e-06, + "loss": 0.1057, + "step": 755 + }, + { + "epoch": 1.2705882352941176, + "grad_norm": 2.9740777866179, + "learning_rate": 2.9382184126304834e-06, + "loss": 0.1037, + "step": 756 + }, + { + "epoch": 1.2722689075630251, + "grad_norm": 2.6657894597056817, + "learning_rate": 2.926200140608634e-06, + "loss": 0.1039, + "step": 757 + }, + { + "epoch": 1.2739495798319327, + "grad_norm": 2.264102839980714, + "learning_rate": 2.9141963220662917e-06, + "loss": 0.076, + "step": 758 + }, + { + "epoch": 1.2756302521008402, + "grad_norm": 3.5306424078481626, + "learning_rate": 2.902207040664834e-06, + "loss": 0.1123, + "step": 759 + }, + { + "epoch": 1.2773109243697478, + "grad_norm": 3.1919340141274977, + "learning_rate": 2.8902323799643116e-06, + "loss": 0.0979, + "step": 760 + }, + { + "epoch": 1.2789915966386554, + "grad_norm": 2.3313048943399224, + "learning_rate": 2.8782724234228876e-06, + "loss": 0.0681, + "step": 761 + }, + { + "epoch": 1.280672268907563, + "grad_norm": 3.747945457345036, + "learning_rate": 2.8663272543962305e-06, + "loss": 0.1077, + "step": 762 + }, + { + "epoch": 1.2823529411764705, + "grad_norm": 2.8608755161882184, + "learning_rate": 2.8543969561369556e-06, + "loss": 0.0997, + "step": 763 + }, + { + "epoch": 1.284033613445378, + "grad_norm": 2.903410246959019, + "learning_rate": 2.842481611794032e-06, + "loss": 0.0793, + "step": 764 + }, + { + "epoch": 1.2857142857142856, + "grad_norm": 2.6778742078682223, + "learning_rate": 2.83058130441221e-06, + "loss": 0.0689, + "step": 765 + }, + { + "epoch": 1.2873949579831931, + "grad_norm": 2.7827741894771134, + "learning_rate": 2.818696116931431e-06, + "loss": 0.075, + "step": 766 + }, + { + "epoch": 1.289075630252101, + "grad_norm": 2.128764344141235, + "learning_rate": 2.8068261321862667e-06, + "loss": 0.09, + "step": 767 + }, + { + "epoch": 1.2907563025210085, + "grad_norm": 3.329548325521387, + "learning_rate": 2.794971432905323e-06, + "loss": 0.0795, + "step": 768 + }, + { + "epoch": 1.292436974789916, + "grad_norm": 2.2187604305649695, + "learning_rate": 2.7831321017106805e-06, + "loss": 0.0575, + "step": 769 + }, + { + "epoch": 1.2941176470588236, + "grad_norm": 3.3967436709413836, + "learning_rate": 2.771308221117309e-06, + "loss": 0.1082, + "step": 770 + }, + { + "epoch": 1.2957983193277312, + "grad_norm": 3.6080260280942316, + "learning_rate": 2.7594998735324905e-06, + "loss": 0.0971, + "step": 771 + }, + { + "epoch": 1.2974789915966387, + "grad_norm": 2.621544354495556, + "learning_rate": 2.7477071412552554e-06, + "loss": 0.0846, + "step": 772 + }, + { + "epoch": 1.2991596638655463, + "grad_norm": 2.5632423256672716, + "learning_rate": 2.735930106475794e-06, + "loss": 0.0786, + "step": 773 + }, + { + "epoch": 1.3008403361344538, + "grad_norm": 2.819748849096385, + "learning_rate": 2.724168851274901e-06, + "loss": 0.0913, + "step": 774 + }, + { + "epoch": 1.3025210084033614, + "grad_norm": 3.2483306380332384, + "learning_rate": 2.712423457623385e-06, + "loss": 0.1027, + "step": 775 + }, + { + "epoch": 1.304201680672269, + "grad_norm": 2.607112390166478, + "learning_rate": 2.7006940073815136e-06, + "loss": 0.0852, + "step": 776 + }, + { + "epoch": 1.3058823529411765, + "grad_norm": 2.550425691687378, + "learning_rate": 2.6889805822984348e-06, + "loss": 0.0782, + "step": 777 + }, + { + "epoch": 1.307563025210084, + "grad_norm": 2.4162221875946672, + "learning_rate": 2.6772832640116035e-06, + "loss": 0.0997, + "step": 778 + }, + { + "epoch": 1.3092436974789916, + "grad_norm": 2.8532159540534363, + "learning_rate": 2.6656021340462246e-06, + "loss": 0.1012, + "step": 779 + }, + { + "epoch": 1.3109243697478992, + "grad_norm": 3.2586342692184638, + "learning_rate": 2.6539372738146696e-06, + "loss": 0.1058, + "step": 780 + }, + { + "epoch": 1.3126050420168067, + "grad_norm": 2.752956076373071, + "learning_rate": 2.6422887646159234e-06, + "loss": 0.0999, + "step": 781 + }, + { + "epoch": 1.3142857142857143, + "grad_norm": 2.6282166566722163, + "learning_rate": 2.6306566876350072e-06, + "loss": 0.0904, + "step": 782 + }, + { + "epoch": 1.3159663865546218, + "grad_norm": 4.459511701783127, + "learning_rate": 2.619041123942419e-06, + "loss": 0.1412, + "step": 783 + }, + { + "epoch": 1.3176470588235294, + "grad_norm": 3.0171293567872057, + "learning_rate": 2.607442154493568e-06, + "loss": 0.0778, + "step": 784 + }, + { + "epoch": 1.319327731092437, + "grad_norm": 2.099785903495667, + "learning_rate": 2.5958598601282036e-06, + "loss": 0.0755, + "step": 785 + }, + { + "epoch": 1.3210084033613445, + "grad_norm": 2.20038597827373, + "learning_rate": 2.584294321569862e-06, + "loss": 0.0977, + "step": 786 + }, + { + "epoch": 1.322689075630252, + "grad_norm": 2.784330403766364, + "learning_rate": 2.572745619425296e-06, + "loss": 0.0736, + "step": 787 + }, + { + "epoch": 1.3243697478991596, + "grad_norm": 2.0879911769632886, + "learning_rate": 2.561213834183919e-06, + "loss": 0.0608, + "step": 788 + }, + { + "epoch": 1.3260504201680672, + "grad_norm": 3.9806688534541843, + "learning_rate": 2.5496990462172344e-06, + "loss": 0.0723, + "step": 789 + }, + { + "epoch": 1.3277310924369747, + "grad_norm": 2.7580721019333603, + "learning_rate": 2.538201335778289e-06, + "loss": 0.082, + "step": 790 + }, + { + "epoch": 1.3294117647058823, + "grad_norm": 2.8722823015793146, + "learning_rate": 2.526720783001107e-06, + "loss": 0.1008, + "step": 791 + }, + { + "epoch": 1.3310924369747898, + "grad_norm": 2.434260993789297, + "learning_rate": 2.5152574679001236e-06, + "loss": 0.0738, + "step": 792 + }, + { + "epoch": 1.3327731092436974, + "grad_norm": 2.80994760149212, + "learning_rate": 2.503811470369644e-06, + "loss": 0.0794, + "step": 793 + }, + { + "epoch": 1.334453781512605, + "grad_norm": 2.9346405462655465, + "learning_rate": 2.4923828701832682e-06, + "loss": 0.1039, + "step": 794 + }, + { + "epoch": 1.3361344537815127, + "grad_norm": 4.00969514669466, + "learning_rate": 2.4809717469933543e-06, + "loss": 0.0713, + "step": 795 + }, + { + "epoch": 1.3378151260504203, + "grad_norm": 3.3856635475000134, + "learning_rate": 2.469578180330444e-06, + "loss": 0.1224, + "step": 796 + }, + { + "epoch": 1.3394957983193279, + "grad_norm": 4.141002149954988, + "learning_rate": 2.458202249602726e-06, + "loss": 0.1295, + "step": 797 + }, + { + "epoch": 1.3411764705882354, + "grad_norm": 2.3909571975188566, + "learning_rate": 2.4468440340954664e-06, + "loss": 0.0858, + "step": 798 + }, + { + "epoch": 1.342857142857143, + "grad_norm": 2.732623939016431, + "learning_rate": 2.43550361297047e-06, + "loss": 0.1232, + "step": 799 + }, + { + "epoch": 1.3445378151260505, + "grad_norm": 2.27043539875792, + "learning_rate": 2.4241810652655197e-06, + "loss": 0.0646, + "step": 800 + }, + { + "epoch": 1.346218487394958, + "grad_norm": 2.800553021390132, + "learning_rate": 2.4128764698938297e-06, + "loss": 0.0856, + "step": 801 + }, + { + "epoch": 1.3478991596638656, + "grad_norm": 2.468540615484736, + "learning_rate": 2.4015899056434945e-06, + "loss": 0.1067, + "step": 802 + }, + { + "epoch": 1.3495798319327732, + "grad_norm": 4.025378528695785, + "learning_rate": 2.390321451176936e-06, + "loss": 0.105, + "step": 803 + }, + { + "epoch": 1.3512605042016808, + "grad_norm": 2.0577272684146872, + "learning_rate": 2.379071185030365e-06, + "loss": 0.067, + "step": 804 + }, + { + "epoch": 1.3529411764705883, + "grad_norm": 2.7859891346028185, + "learning_rate": 2.3678391856132203e-06, + "loss": 0.0715, + "step": 805 + }, + { + "epoch": 1.3546218487394959, + "grad_norm": 3.417663771387606, + "learning_rate": 2.356625531207638e-06, + "loss": 0.1134, + "step": 806 + }, + { + "epoch": 1.3563025210084034, + "grad_norm": 2.9934064584562807, + "learning_rate": 2.345430299967888e-06, + "loss": 0.099, + "step": 807 + }, + { + "epoch": 1.357983193277311, + "grad_norm": 3.0446906092550727, + "learning_rate": 2.334253569919846e-06, + "loss": 0.0881, + "step": 808 + }, + { + "epoch": 1.3596638655462185, + "grad_norm": 2.714244890028747, + "learning_rate": 2.323095418960442e-06, + "loss": 0.0651, + "step": 809 + }, + { + "epoch": 1.361344537815126, + "grad_norm": 2.7908745015689145, + "learning_rate": 2.311955924857113e-06, + "loss": 0.0928, + "step": 810 + }, + { + "epoch": 1.3630252100840337, + "grad_norm": 2.593550587022856, + "learning_rate": 2.3008351652472714e-06, + "loss": 0.0936, + "step": 811 + }, + { + "epoch": 1.3647058823529412, + "grad_norm": 2.6262874429380214, + "learning_rate": 2.289733217637753e-06, + "loss": 0.07, + "step": 812 + }, + { + "epoch": 1.3663865546218488, + "grad_norm": 2.5534127800889346, + "learning_rate": 2.278650159404289e-06, + "loss": 0.0778, + "step": 813 + }, + { + "epoch": 1.3680672268907563, + "grad_norm": 2.44366168179335, + "learning_rate": 2.267586067790952e-06, + "loss": 0.0825, + "step": 814 + }, + { + "epoch": 1.3697478991596639, + "grad_norm": 2.768515991865253, + "learning_rate": 2.2565410199096322e-06, + "loss": 0.086, + "step": 815 + }, + { + "epoch": 1.3714285714285714, + "grad_norm": 2.9765080509774546, + "learning_rate": 2.245515092739488e-06, + "loss": 0.0967, + "step": 816 + }, + { + "epoch": 1.373109243697479, + "grad_norm": 2.875743137704446, + "learning_rate": 2.234508363126419e-06, + "loss": 0.1058, + "step": 817 + }, + { + "epoch": 1.3747899159663866, + "grad_norm": 3.542235960340661, + "learning_rate": 2.2235209077825264e-06, + "loss": 0.153, + "step": 818 + }, + { + "epoch": 1.3764705882352941, + "grad_norm": 2.403232600671078, + "learning_rate": 2.2125528032855727e-06, + "loss": 0.0689, + "step": 819 + }, + { + "epoch": 1.3781512605042017, + "grad_norm": 2.7232278008796693, + "learning_rate": 2.2016041260784604e-06, + "loss": 0.0523, + "step": 820 + }, + { + "epoch": 1.3798319327731092, + "grad_norm": 2.854644874148179, + "learning_rate": 2.1906749524686856e-06, + "loss": 0.0783, + "step": 821 + }, + { + "epoch": 1.3815126050420168, + "grad_norm": 1.8526957378806435, + "learning_rate": 2.1797653586278193e-06, + "loss": 0.0658, + "step": 822 + }, + { + "epoch": 1.3831932773109243, + "grad_norm": 3.1776681065587606, + "learning_rate": 2.168875420590965e-06, + "loss": 0.1051, + "step": 823 + }, + { + "epoch": 1.384873949579832, + "grad_norm": 2.5512188793181623, + "learning_rate": 2.158005214256236e-06, + "loss": 0.0708, + "step": 824 + }, + { + "epoch": 1.3865546218487395, + "grad_norm": 2.6609634186618156, + "learning_rate": 2.147154815384226e-06, + "loss": 0.0768, + "step": 825 + }, + { + "epoch": 1.388235294117647, + "grad_norm": 2.3931424820403295, + "learning_rate": 2.136324299597474e-06, + "loss": 0.0715, + "step": 826 + }, + { + "epoch": 1.3899159663865546, + "grad_norm": 2.8679014750294125, + "learning_rate": 2.12551374237995e-06, + "loss": 0.1095, + "step": 827 + }, + { + "epoch": 1.3915966386554621, + "grad_norm": 2.180768007063552, + "learning_rate": 2.1147232190765137e-06, + "loss": 0.0675, + "step": 828 + }, + { + "epoch": 1.3932773109243697, + "grad_norm": 2.5066981219224775, + "learning_rate": 2.1039528048924043e-06, + "loss": 0.059, + "step": 829 + }, + { + "epoch": 1.3949579831932772, + "grad_norm": 3.3803755201148187, + "learning_rate": 2.0932025748927015e-06, + "loss": 0.1068, + "step": 830 + }, + { + "epoch": 1.3966386554621848, + "grad_norm": 2.371033561602867, + "learning_rate": 2.0824726040018174e-06, + "loss": 0.0728, + "step": 831 + }, + { + "epoch": 1.3983193277310924, + "grad_norm": 2.774208580231465, + "learning_rate": 2.0717629670029653e-06, + "loss": 0.0661, + "step": 832 + }, + { + "epoch": 1.4, + "grad_norm": 2.1336928525875836, + "learning_rate": 2.061073738537635e-06, + "loss": 0.0733, + "step": 833 + }, + { + "epoch": 1.4016806722689075, + "grad_norm": 3.1234467182623975, + "learning_rate": 2.050404993105085e-06, + "loss": 0.0831, + "step": 834 + }, + { + "epoch": 1.403361344537815, + "grad_norm": 3.091586518307653, + "learning_rate": 2.0397568050618095e-06, + "loss": 0.1086, + "step": 835 + }, + { + "epoch": 1.4050420168067226, + "grad_norm": 3.002925392888798, + "learning_rate": 2.0291292486210327e-06, + "loss": 0.0994, + "step": 836 + }, + { + "epoch": 1.4067226890756301, + "grad_norm": 2.4071605199504105, + "learning_rate": 2.018522397852178e-06, + "loss": 0.0642, + "step": 837 + }, + { + "epoch": 1.4084033613445377, + "grad_norm": 4.850453820458399, + "learning_rate": 2.0079363266803696e-06, + "loss": 0.1209, + "step": 838 + }, + { + "epoch": 1.4100840336134453, + "grad_norm": 2.5655766253111913, + "learning_rate": 1.9973711088858973e-06, + "loss": 0.0718, + "step": 839 + }, + { + "epoch": 1.4117647058823528, + "grad_norm": 2.7191226267335726, + "learning_rate": 1.9868268181037186e-06, + "loss": 0.0898, + "step": 840 + }, + { + "epoch": 1.4134453781512604, + "grad_norm": 2.636556426536022, + "learning_rate": 1.976303527822933e-06, + "loss": 0.0944, + "step": 841 + }, + { + "epoch": 1.4151260504201681, + "grad_norm": 3.427543761682362, + "learning_rate": 1.9658013113862806e-06, + "loss": 0.0629, + "step": 842 + }, + { + "epoch": 1.4168067226890757, + "grad_norm": 2.407314737887372, + "learning_rate": 1.9553202419896256e-06, + "loss": 0.0669, + "step": 843 + }, + { + "epoch": 1.4184873949579833, + "grad_norm": 2.3970562937015605, + "learning_rate": 1.944860392681442e-06, + "loss": 0.0842, + "step": 844 + }, + { + "epoch": 1.4201680672268908, + "grad_norm": 3.0111009657414436, + "learning_rate": 1.934421836362315e-06, + "loss": 0.0925, + "step": 845 + }, + { + "epoch": 1.4218487394957984, + "grad_norm": 2.3181211787455287, + "learning_rate": 1.9240046457844223e-06, + "loss": 0.0857, + "step": 846 + }, + { + "epoch": 1.423529411764706, + "grad_norm": 2.4618319020447057, + "learning_rate": 1.913608893551036e-06, + "loss": 0.0718, + "step": 847 + }, + { + "epoch": 1.4252100840336135, + "grad_norm": 2.4722571546574206, + "learning_rate": 1.9032346521160066e-06, + "loss": 0.0775, + "step": 848 + }, + { + "epoch": 1.426890756302521, + "grad_norm": 2.8957411522240957, + "learning_rate": 1.8928819937832689e-06, + "loss": 0.0981, + "step": 849 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 3.0058263449936424, + "learning_rate": 1.8825509907063328e-06, + "loss": 0.1281, + "step": 850 + }, + { + "epoch": 1.4302521008403362, + "grad_norm": 2.6662478992952585, + "learning_rate": 1.8722417148877752e-06, + "loss": 0.0808, + "step": 851 + }, + { + "epoch": 1.4319327731092437, + "grad_norm": 2.9792508981015247, + "learning_rate": 1.8619542381787508e-06, + "loss": 0.0999, + "step": 852 + }, + { + "epoch": 1.4336134453781513, + "grad_norm": 2.1089140232124155, + "learning_rate": 1.851688632278476e-06, + "loss": 0.0469, + "step": 853 + }, + { + "epoch": 1.4352941176470588, + "grad_norm": 3.5585819612230334, + "learning_rate": 1.8414449687337467e-06, + "loss": 0.0704, + "step": 854 + }, + { + "epoch": 1.4369747899159664, + "grad_norm": 2.24926411881091, + "learning_rate": 1.8312233189384194e-06, + "loss": 0.0735, + "step": 855 + }, + { + "epoch": 1.438655462184874, + "grad_norm": 2.6302186153665947, + "learning_rate": 1.821023754132933e-06, + "loss": 0.074, + "step": 856 + }, + { + "epoch": 1.4403361344537815, + "grad_norm": 2.8780342261679506, + "learning_rate": 1.8108463454038022e-06, + "loss": 0.0778, + "step": 857 + }, + { + "epoch": 1.442016806722689, + "grad_norm": 3.7396581161865172, + "learning_rate": 1.800691163683118e-06, + "loss": 0.1022, + "step": 858 + }, + { + "epoch": 1.4436974789915966, + "grad_norm": 3.60714342413831, + "learning_rate": 1.790558279748067e-06, + "loss": 0.1618, + "step": 859 + }, + { + "epoch": 1.4453781512605042, + "grad_norm": 2.9595163340771147, + "learning_rate": 1.780447764220422e-06, + "loss": 0.0904, + "step": 860 + }, + { + "epoch": 1.4470588235294117, + "grad_norm": 2.1689321130662886, + "learning_rate": 1.7703596875660645e-06, + "loss": 0.0719, + "step": 861 + }, + { + "epoch": 1.4487394957983193, + "grad_norm": 2.1941556772158264, + "learning_rate": 1.7602941200944812e-06, + "loss": 0.0635, + "step": 862 + }, + { + "epoch": 1.4504201680672268, + "grad_norm": 2.2798654285400737, + "learning_rate": 1.7502511319582855e-06, + "loss": 0.0875, + "step": 863 + }, + { + "epoch": 1.4521008403361344, + "grad_norm": 2.2334745972248298, + "learning_rate": 1.7402307931527157e-06, + "loss": 0.0687, + "step": 864 + }, + { + "epoch": 1.453781512605042, + "grad_norm": 2.8004894708871415, + "learning_rate": 1.7302331735151594e-06, + "loss": 0.1048, + "step": 865 + }, + { + "epoch": 1.4554621848739495, + "grad_norm": 3.108271318580809, + "learning_rate": 1.7202583427246633e-06, + "loss": 0.0753, + "step": 866 + }, + { + "epoch": 1.457142857142857, + "grad_norm": 2.312999197511707, + "learning_rate": 1.7103063703014372e-06, + "loss": 0.0746, + "step": 867 + }, + { + "epoch": 1.4588235294117646, + "grad_norm": 2.6351363101768954, + "learning_rate": 1.7003773256063882e-06, + "loss": 0.092, + "step": 868 + }, + { + "epoch": 1.4605042016806722, + "grad_norm": 2.183847704629778, + "learning_rate": 1.690471277840619e-06, + "loss": 0.0625, + "step": 869 + }, + { + "epoch": 1.46218487394958, + "grad_norm": 3.064581959699731, + "learning_rate": 1.6805882960449594e-06, + "loss": 0.1215, + "step": 870 + }, + { + "epoch": 1.4638655462184875, + "grad_norm": 3.1271728558378737, + "learning_rate": 1.6707284490994746e-06, + "loss": 0.0908, + "step": 871 + }, + { + "epoch": 1.465546218487395, + "grad_norm": 2.9326514691610455, + "learning_rate": 1.6608918057229944e-06, + "loss": 0.0798, + "step": 872 + }, + { + "epoch": 1.4672268907563026, + "grad_norm": 3.6801264170612193, + "learning_rate": 1.6510784344726294e-06, + "loss": 0.1136, + "step": 873 + }, + { + "epoch": 1.4689075630252102, + "grad_norm": 2.2001266838315057, + "learning_rate": 1.6412884037432875e-06, + "loss": 0.077, + "step": 874 + }, + { + "epoch": 1.4705882352941178, + "grad_norm": 2.018537490774913, + "learning_rate": 1.6315217817672142e-06, + "loss": 0.0577, + "step": 875 + }, + { + "epoch": 1.4722689075630253, + "grad_norm": 2.502467447738802, + "learning_rate": 1.6217786366134953e-06, + "loss": 0.0867, + "step": 876 + }, + { + "epoch": 1.4739495798319329, + "grad_norm": 2.404093931401216, + "learning_rate": 1.612059036187602e-06, + "loss": 0.0818, + "step": 877 + }, + { + "epoch": 1.4756302521008404, + "grad_norm": 2.7777757705585366, + "learning_rate": 1.6023630482309017e-06, + "loss": 0.0851, + "step": 878 + }, + { + "epoch": 1.477310924369748, + "grad_norm": 2.1744066253008025, + "learning_rate": 1.5926907403202001e-06, + "loss": 0.0667, + "step": 879 + }, + { + "epoch": 1.4789915966386555, + "grad_norm": 3.2942884352254542, + "learning_rate": 1.5830421798672568e-06, + "loss": 0.0714, + "step": 880 + }, + { + "epoch": 1.480672268907563, + "grad_norm": 3.4466527581467976, + "learning_rate": 1.5734174341183284e-06, + "loss": 0.1096, + "step": 881 + }, + { + "epoch": 1.4823529411764707, + "grad_norm": 2.530910684223825, + "learning_rate": 1.5638165701536866e-06, + "loss": 0.0786, + "step": 882 + }, + { + "epoch": 1.4840336134453782, + "grad_norm": 2.949069912774898, + "learning_rate": 1.554239654887163e-06, + "loss": 0.082, + "step": 883 + }, + { + "epoch": 1.4857142857142858, + "grad_norm": 2.918482254384839, + "learning_rate": 1.544686755065677e-06, + "loss": 0.1056, + "step": 884 + }, + { + "epoch": 1.4873949579831933, + "grad_norm": 2.9537497761417644, + "learning_rate": 1.5351579372687658e-06, + "loss": 0.0904, + "step": 885 + }, + { + "epoch": 1.4890756302521009, + "grad_norm": 3.0556323605290676, + "learning_rate": 1.525653267908132e-06, + "loss": 0.0871, + "step": 886 + }, + { + "epoch": 1.4907563025210084, + "grad_norm": 2.6043633400058006, + "learning_rate": 1.5161728132271674e-06, + "loss": 0.0744, + "step": 887 + }, + { + "epoch": 1.492436974789916, + "grad_norm": 2.0618324084427804, + "learning_rate": 1.5067166393005055e-06, + "loss": 0.0701, + "step": 888 + }, + { + "epoch": 1.4941176470588236, + "grad_norm": 2.9896346108157177, + "learning_rate": 1.4972848120335453e-06, + "loss": 0.1136, + "step": 889 + }, + { + "epoch": 1.495798319327731, + "grad_norm": 2.605228843859927, + "learning_rate": 1.4878773971620076e-06, + "loss": 0.0673, + "step": 890 + }, + { + "epoch": 1.4974789915966387, + "grad_norm": 2.7905166458094333, + "learning_rate": 1.4784944602514662e-06, + "loss": 0.0809, + "step": 891 + }, + { + "epoch": 1.4991596638655462, + "grad_norm": 3.1497593827029506, + "learning_rate": 1.4691360666968923e-06, + "loss": 0.1021, + "step": 892 + }, + { + "epoch": 1.5008403361344538, + "grad_norm": 2.1897838947360735, + "learning_rate": 1.4598022817222058e-06, + "loss": 0.0571, + "step": 893 + }, + { + "epoch": 1.5025210084033613, + "grad_norm": 2.302331646085917, + "learning_rate": 1.4504931703798086e-06, + "loss": 0.0576, + "step": 894 + }, + { + "epoch": 1.504201680672269, + "grad_norm": 2.7532766985557857, + "learning_rate": 1.4412087975501459e-06, + "loss": 0.076, + "step": 895 + }, + { + "epoch": 1.5058823529411764, + "grad_norm": 2.5202671032062582, + "learning_rate": 1.4319492279412388e-06, + "loss": 0.0678, + "step": 896 + }, + { + "epoch": 1.507563025210084, + "grad_norm": 3.3560869623465535, + "learning_rate": 1.4227145260882463e-06, + "loss": 0.0674, + "step": 897 + }, + { + "epoch": 1.5092436974789916, + "grad_norm": 2.7037346414852497, + "learning_rate": 1.413504756353009e-06, + "loss": 0.0877, + "step": 898 + }, + { + "epoch": 1.5109243697478991, + "grad_norm": 2.950076661717238, + "learning_rate": 1.4043199829235983e-06, + "loss": 0.1009, + "step": 899 + }, + { + "epoch": 1.5126050420168067, + "grad_norm": 2.9724993078538096, + "learning_rate": 1.3951602698138773e-06, + "loss": 0.0998, + "step": 900 + }, + { + "epoch": 1.5142857142857142, + "grad_norm": 2.171203981414814, + "learning_rate": 1.3860256808630429e-06, + "loss": 0.0649, + "step": 901 + }, + { + "epoch": 1.5159663865546218, + "grad_norm": 2.3717860449067407, + "learning_rate": 1.3769162797351953e-06, + "loss": 0.0628, + "step": 902 + }, + { + "epoch": 1.5176470588235293, + "grad_norm": 3.7600443429783352, + "learning_rate": 1.3678321299188802e-06, + "loss": 0.1002, + "step": 903 + }, + { + "epoch": 1.519327731092437, + "grad_norm": 3.0435496611540467, + "learning_rate": 1.3587732947266557e-06, + "loss": 0.095, + "step": 904 + }, + { + "epoch": 1.5210084033613445, + "grad_norm": 2.96613726072199, + "learning_rate": 1.34973983729465e-06, + "loss": 0.0914, + "step": 905 + }, + { + "epoch": 1.522689075630252, + "grad_norm": 2.8435196172045907, + "learning_rate": 1.340731820582114e-06, + "loss": 0.0699, + "step": 906 + }, + { + "epoch": 1.5243697478991596, + "grad_norm": 2.464476894025319, + "learning_rate": 1.3317493073709936e-06, + "loss": 0.0705, + "step": 907 + }, + { + "epoch": 1.5260504201680671, + "grad_norm": 1.890126330693869, + "learning_rate": 1.3227923602654808e-06, + "loss": 0.0561, + "step": 908 + }, + { + "epoch": 1.5277310924369747, + "grad_norm": 2.2602626027892176, + "learning_rate": 1.3138610416915887e-06, + "loss": 0.0733, + "step": 909 + }, + { + "epoch": 1.5294117647058822, + "grad_norm": 2.5239743120304468, + "learning_rate": 1.3049554138967052e-06, + "loss": 0.0658, + "step": 910 + }, + { + "epoch": 1.5310924369747898, + "grad_norm": 2.1639616773849344, + "learning_rate": 1.2960755389491703e-06, + "loss": 0.0644, + "step": 911 + }, + { + "epoch": 1.5327731092436974, + "grad_norm": 2.7172132336504373, + "learning_rate": 1.2872214787378306e-06, + "loss": 0.0806, + "step": 912 + }, + { + "epoch": 1.534453781512605, + "grad_norm": 2.433240574623036, + "learning_rate": 1.278393294971626e-06, + "loss": 0.0783, + "step": 913 + }, + { + "epoch": 1.5361344537815125, + "grad_norm": 3.0834984240711565, + "learning_rate": 1.269591049179138e-06, + "loss": 0.0617, + "step": 914 + }, + { + "epoch": 1.53781512605042, + "grad_norm": 2.397030675922682, + "learning_rate": 1.2608148027081773e-06, + "loss": 0.0818, + "step": 915 + }, + { + "epoch": 1.5394957983193276, + "grad_norm": 3.0216001414776374, + "learning_rate": 1.2520646167253514e-06, + "loss": 0.0821, + "step": 916 + }, + { + "epoch": 1.5411764705882351, + "grad_norm": 3.134391572738623, + "learning_rate": 1.2433405522156334e-06, + "loss": 0.0916, + "step": 917 + }, + { + "epoch": 1.5428571428571427, + "grad_norm": 2.3360210128403636, + "learning_rate": 1.234642669981946e-06, + "loss": 0.0641, + "step": 918 + }, + { + "epoch": 1.5445378151260503, + "grad_norm": 2.7163994508351217, + "learning_rate": 1.2259710306447275e-06, + "loss": 0.0886, + "step": 919 + }, + { + "epoch": 1.5462184873949578, + "grad_norm": 2.431875213779764, + "learning_rate": 1.2173256946415214e-06, + "loss": 0.0773, + "step": 920 + }, + { + "epoch": 1.5478991596638656, + "grad_norm": 2.5611857478832882, + "learning_rate": 1.2087067222265409e-06, + "loss": 0.0628, + "step": 921 + }, + { + "epoch": 1.5495798319327732, + "grad_norm": 2.9190300690138633, + "learning_rate": 1.2001141734702625e-06, + "loss": 0.077, + "step": 922 + }, + { + "epoch": 1.5512605042016807, + "grad_norm": 2.6935195739367526, + "learning_rate": 1.1915481082589998e-06, + "loss": 0.0651, + "step": 923 + }, + { + "epoch": 1.5529411764705883, + "grad_norm": 2.495024421506342, + "learning_rate": 1.1830085862944851e-06, + "loss": 0.0812, + "step": 924 + }, + { + "epoch": 1.5546218487394958, + "grad_norm": 2.3186901545314353, + "learning_rate": 1.17449566709346e-06, + "loss": 0.0682, + "step": 925 + }, + { + "epoch": 1.5563025210084034, + "grad_norm": 2.3237588553363944, + "learning_rate": 1.166009409987251e-06, + "loss": 0.0555, + "step": 926 + }, + { + "epoch": 1.557983193277311, + "grad_norm": 2.8044298597093107, + "learning_rate": 1.1575498741213682e-06, + "loss": 0.0947, + "step": 927 + }, + { + "epoch": 1.5596638655462185, + "grad_norm": 2.4178880085870396, + "learning_rate": 1.1491171184550799e-06, + "loss": 0.0629, + "step": 928 + }, + { + "epoch": 1.561344537815126, + "grad_norm": 1.976434008723005, + "learning_rate": 1.1407112017610134e-06, + "loss": 0.052, + "step": 929 + }, + { + "epoch": 1.5630252100840336, + "grad_norm": 3.483563565061037, + "learning_rate": 1.1323321826247347e-06, + "loss": 0.077, + "step": 930 + }, + { + "epoch": 1.5647058823529412, + "grad_norm": 1.8972048774489287, + "learning_rate": 1.1239801194443507e-06, + "loss": 0.0494, + "step": 931 + }, + { + "epoch": 1.5663865546218487, + "grad_norm": 3.1882538432651777, + "learning_rate": 1.115655070430096e-06, + "loss": 0.0779, + "step": 932 + }, + { + "epoch": 1.5680672268907563, + "grad_norm": 3.1426262134738603, + "learning_rate": 1.107357093603924e-06, + "loss": 0.076, + "step": 933 + }, + { + "epoch": 1.5697478991596638, + "grad_norm": 3.3491106407225555, + "learning_rate": 1.0990862467991132e-06, + "loss": 0.0939, + "step": 934 + }, + { + "epoch": 1.5714285714285714, + "grad_norm": 2.3555969578510196, + "learning_rate": 1.0908425876598512e-06, + "loss": 0.0683, + "step": 935 + }, + { + "epoch": 1.573109243697479, + "grad_norm": 2.4300759069980136, + "learning_rate": 1.082626173640846e-06, + "loss": 0.0788, + "step": 936 + }, + { + "epoch": 1.5747899159663865, + "grad_norm": 3.0685309864448636, + "learning_rate": 1.0744370620069122e-06, + "loss": 0.0852, + "step": 937 + }, + { + "epoch": 1.576470588235294, + "grad_norm": 2.568291131776804, + "learning_rate": 1.066275309832584e-06, + "loss": 0.074, + "step": 938 + }, + { + "epoch": 1.5781512605042018, + "grad_norm": 2.520877566474604, + "learning_rate": 1.0581409740017113e-06, + "loss": 0.0521, + "step": 939 + }, + { + "epoch": 1.5798319327731094, + "grad_norm": 2.792407855946344, + "learning_rate": 1.0500341112070605e-06, + "loss": 0.0789, + "step": 940 + }, + { + "epoch": 1.581512605042017, + "grad_norm": 3.1440477930637214, + "learning_rate": 1.0419547779499283e-06, + "loss": 0.111, + "step": 941 + }, + { + "epoch": 1.5831932773109245, + "grad_norm": 3.08626306661282, + "learning_rate": 1.0339030305397374e-06, + "loss": 0.0773, + "step": 942 + }, + { + "epoch": 1.584873949579832, + "grad_norm": 2.7818074011761613, + "learning_rate": 1.025878925093653e-06, + "loss": 0.1025, + "step": 943 + }, + { + "epoch": 1.5865546218487396, + "grad_norm": 2.0707670868256756, + "learning_rate": 1.0178825175361846e-06, + "loss": 0.0543, + "step": 944 + }, + { + "epoch": 1.5882352941176472, + "grad_norm": 3.277736077714651, + "learning_rate": 1.0099138635988026e-06, + "loss": 0.0886, + "step": 945 + }, + { + "epoch": 1.5899159663865547, + "grad_norm": 1.7856587713887242, + "learning_rate": 1.0019730188195464e-06, + "loss": 0.0498, + "step": 946 + }, + { + "epoch": 1.5915966386554623, + "grad_norm": 2.6385746957949787, + "learning_rate": 9.940600385426347e-07, + "loss": 0.0812, + "step": 947 + }, + { + "epoch": 1.5932773109243699, + "grad_norm": 2.776521241285141, + "learning_rate": 9.861749779180873e-07, + "loss": 0.087, + "step": 948 + }, + { + "epoch": 1.5949579831932774, + "grad_norm": 2.639723525487175, + "learning_rate": 9.783178919013297e-07, + "loss": 0.1119, + "step": 949 + }, + { + "epoch": 1.596638655462185, + "grad_norm": 2.7204981405013458, + "learning_rate": 9.704888352528257e-07, + "loss": 0.085, + "step": 950 + }, + { + "epoch": 1.5983193277310925, + "grad_norm": 2.5248398773851166, + "learning_rate": 9.626878625376784e-07, + "loss": 0.0695, + "step": 951 + }, + { + "epoch": 1.6, + "grad_norm": 2.670224920366225, + "learning_rate": 9.549150281252633e-07, + "loss": 0.0951, + "step": 952 + }, + { + "epoch": 1.6016806722689076, + "grad_norm": 2.7422961930525176, + "learning_rate": 9.471703861888398e-07, + "loss": 0.1044, + "step": 953 + }, + { + "epoch": 1.6033613445378152, + "grad_norm": 2.2888317772986673, + "learning_rate": 9.394539907051837e-07, + "loss": 0.0586, + "step": 954 + }, + { + "epoch": 1.6050420168067228, + "grad_norm": 2.1943114037634976, + "learning_rate": 9.317658954541992e-07, + "loss": 0.0687, + "step": 955 + }, + { + "epoch": 1.6067226890756303, + "grad_norm": 3.4427372615358003, + "learning_rate": 9.241061540185547e-07, + "loss": 0.0908, + "step": 956 + }, + { + "epoch": 1.6084033613445379, + "grad_norm": 2.817393993003906, + "learning_rate": 9.164748197833039e-07, + "loss": 0.0619, + "step": 957 + }, + { + "epoch": 1.6100840336134454, + "grad_norm": 2.9866238876292415, + "learning_rate": 9.088719459355133e-07, + "loss": 0.0672, + "step": 958 + }, + { + "epoch": 1.611764705882353, + "grad_norm": 2.785563725722871, + "learning_rate": 9.01297585463895e-07, + "loss": 0.0745, + "step": 959 + }, + { + "epoch": 1.6134453781512605, + "grad_norm": 2.3105959305373136, + "learning_rate": 8.937517911584321e-07, + "loss": 0.0705, + "step": 960 + }, + { + "epoch": 1.615126050420168, + "grad_norm": 1.99378117488201, + "learning_rate": 8.862346156100188e-07, + "loss": 0.0466, + "step": 961 + }, + { + "epoch": 1.6168067226890757, + "grad_norm": 2.6666955685587053, + "learning_rate": 8.787461112100837e-07, + "loss": 0.0656, + "step": 962 + }, + { + "epoch": 1.6184873949579832, + "grad_norm": 3.3249810625849956, + "learning_rate": 8.712863301502339e-07, + "loss": 0.0678, + "step": 963 + }, + { + "epoch": 1.6201680672268908, + "grad_norm": 3.353300594203047, + "learning_rate": 8.638553244218872e-07, + "loss": 0.0956, + "step": 964 + }, + { + "epoch": 1.6218487394957983, + "grad_norm": 2.8496488574356476, + "learning_rate": 8.56453145815907e-07, + "loss": 0.0838, + "step": 965 + }, + { + "epoch": 1.6235294117647059, + "grad_norm": 3.064310810621701, + "learning_rate": 8.490798459222477e-07, + "loss": 0.0865, + "step": 966 + }, + { + "epoch": 1.6252100840336134, + "grad_norm": 2.968813736122225, + "learning_rate": 8.417354761295876e-07, + "loss": 0.0728, + "step": 967 + }, + { + "epoch": 1.626890756302521, + "grad_norm": 2.049286780650219, + "learning_rate": 8.344200876249803e-07, + "loss": 0.0475, + "step": 968 + }, + { + "epoch": 1.6285714285714286, + "grad_norm": 3.100955143857725, + "learning_rate": 8.271337313934869e-07, + "loss": 0.0849, + "step": 969 + }, + { + "epoch": 1.6302521008403361, + "grad_norm": 2.9949248571700617, + "learning_rate": 8.198764582178303e-07, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 1.6319327731092437, + "grad_norm": 2.853687486256643, + "learning_rate": 8.12648318678036e-07, + "loss": 0.0653, + "step": 971 + }, + { + "epoch": 1.6336134453781512, + "grad_norm": 2.4758118064736685, + "learning_rate": 8.054493631510785e-07, + "loss": 0.0736, + "step": 972 + }, + { + "epoch": 1.6352941176470588, + "grad_norm": 2.480419073945819, + "learning_rate": 7.98279641810537e-07, + "loss": 0.0612, + "step": 973 + }, + { + "epoch": 1.6369747899159663, + "grad_norm": 2.6330025212546073, + "learning_rate": 7.911392046262367e-07, + "loss": 0.0665, + "step": 974 + }, + { + "epoch": 1.638655462184874, + "grad_norm": 2.6543901706240707, + "learning_rate": 7.840281013639078e-07, + "loss": 0.0508, + "step": 975 + }, + { + "epoch": 1.6403361344537815, + "grad_norm": 2.701255527964721, + "learning_rate": 7.769463815848344e-07, + "loss": 0.0906, + "step": 976 + }, + { + "epoch": 1.642016806722689, + "grad_norm": 3.0305303322182335, + "learning_rate": 7.698940946455125e-07, + "loss": 0.076, + "step": 977 + }, + { + "epoch": 1.6436974789915966, + "grad_norm": 2.560069381692694, + "learning_rate": 7.628712896973006e-07, + "loss": 0.0611, + "step": 978 + }, + { + "epoch": 1.6453781512605041, + "grad_norm": 2.126677934815069, + "learning_rate": 7.55878015686084e-07, + "loss": 0.0626, + "step": 979 + }, + { + "epoch": 1.6470588235294117, + "grad_norm": 3.320788842248769, + "learning_rate": 7.489143213519301e-07, + "loss": 0.066, + "step": 980 + }, + { + "epoch": 1.6487394957983192, + "grad_norm": 2.1324499834476893, + "learning_rate": 7.419802552287453e-07, + "loss": 0.0601, + "step": 981 + }, + { + "epoch": 1.6504201680672268, + "grad_norm": 1.6451700538742926, + "learning_rate": 7.350758656439455e-07, + "loss": 0.0358, + "step": 982 + }, + { + "epoch": 1.6521008403361344, + "grad_norm": 2.9089699245304526, + "learning_rate": 7.282012007181083e-07, + "loss": 0.091, + "step": 983 + }, + { + "epoch": 1.653781512605042, + "grad_norm": 3.3157789811498115, + "learning_rate": 7.213563083646497e-07, + "loss": 0.0808, + "step": 984 + }, + { + "epoch": 1.6554621848739495, + "grad_norm": 2.504068675080367, + "learning_rate": 7.145412362894771e-07, + "loss": 0.083, + "step": 985 + }, + { + "epoch": 1.657142857142857, + "grad_norm": 2.965188873541059, + "learning_rate": 7.077560319906696e-07, + "loss": 0.0761, + "step": 986 + }, + { + "epoch": 1.6588235294117646, + "grad_norm": 3.9651631993982632, + "learning_rate": 7.010007427581378e-07, + "loss": 0.0931, + "step": 987 + }, + { + "epoch": 1.6605042016806721, + "grad_norm": 2.6828484004077704, + "learning_rate": 6.942754156732978e-07, + "loss": 0.0899, + "step": 988 + }, + { + "epoch": 1.6621848739495797, + "grad_norm": 2.9178209309323617, + "learning_rate": 6.875800976087444e-07, + "loss": 0.0922, + "step": 989 + }, + { + "epoch": 1.6638655462184873, + "grad_norm": 2.167420070287885, + "learning_rate": 6.809148352279182e-07, + "loss": 0.0491, + "step": 990 + }, + { + "epoch": 1.6655462184873948, + "grad_norm": 2.1946934272492538, + "learning_rate": 6.742796749847908e-07, + "loss": 0.059, + "step": 991 + }, + { + "epoch": 1.6672268907563024, + "grad_norm": 2.312028197457495, + "learning_rate": 6.676746631235282e-07, + "loss": 0.06, + "step": 992 + }, + { + "epoch": 1.66890756302521, + "grad_norm": 2.6693038248515415, + "learning_rate": 6.61099845678183e-07, + "loss": 0.0644, + "step": 993 + }, + { + "epoch": 1.6705882352941175, + "grad_norm": 3.164886728441667, + "learning_rate": 6.545552684723583e-07, + "loss": 0.1167, + "step": 994 + }, + { + "epoch": 1.6722689075630253, + "grad_norm": 3.460096856141393, + "learning_rate": 6.480409771189027e-07, + "loss": 0.092, + "step": 995 + }, + { + "epoch": 1.6739495798319328, + "grad_norm": 2.0571715576847422, + "learning_rate": 6.415570170195801e-07, + "loss": 0.0583, + "step": 996 + }, + { + "epoch": 1.6756302521008404, + "grad_norm": 2.5258688350847662, + "learning_rate": 6.351034333647615e-07, + "loss": 0.0716, + "step": 997 + }, + { + "epoch": 1.677310924369748, + "grad_norm": 2.5066410840729385, + "learning_rate": 6.286802711331097e-07, + "loss": 0.0663, + "step": 998 + }, + { + "epoch": 1.6789915966386555, + "grad_norm": 2.686273803992896, + "learning_rate": 6.222875750912571e-07, + "loss": 0.0739, + "step": 999 + }, + { + "epoch": 1.680672268907563, + "grad_norm": 2.3800935639690572, + "learning_rate": 6.159253897935069e-07, + "loss": 0.0645, + "step": 1000 + }, + { + "epoch": 1.680672268907563, + "eval_loss": 0.21711505949497223, + "eval_runtime": 7.6127, + "eval_samples_per_second": 6.437, + "eval_steps_per_second": 1.708, + "step": 1000 + }, + { + "epoch": 1.6823529411764706, + "grad_norm": 2.3474068601495777, + "learning_rate": 6.095937595815104e-07, + "loss": 0.059, + "step": 1001 + }, + { + "epoch": 1.6840336134453782, + "grad_norm": 3.0027968700627143, + "learning_rate": 6.032927285839674e-07, + "loss": 0.11, + "step": 1002 + }, + { + "epoch": 1.6857142857142857, + "grad_norm": 2.4756615878732755, + "learning_rate": 5.9702234071631e-07, + "loss": 0.0652, + "step": 1003 + }, + { + "epoch": 1.6873949579831933, + "grad_norm": 2.4849167604857065, + "learning_rate": 5.907826396804062e-07, + "loss": 0.0707, + "step": 1004 + }, + { + "epoch": 1.6890756302521008, + "grad_norm": 2.527471856677461, + "learning_rate": 5.845736689642472e-07, + "loss": 0.0869, + "step": 1005 + }, + { + "epoch": 1.6907563025210084, + "grad_norm": 2.8221683255428434, + "learning_rate": 5.783954718416468e-07, + "loss": 0.0819, + "step": 1006 + }, + { + "epoch": 1.692436974789916, + "grad_norm": 2.631351359370838, + "learning_rate": 5.722480913719425e-07, + "loss": 0.0587, + "step": 1007 + }, + { + "epoch": 1.6941176470588235, + "grad_norm": 2.6033686025698515, + "learning_rate": 5.661315703996905e-07, + "loss": 0.0766, + "step": 1008 + }, + { + "epoch": 1.695798319327731, + "grad_norm": 2.543956882885212, + "learning_rate": 5.600459515543733e-07, + "loss": 0.065, + "step": 1009 + }, + { + "epoch": 1.6974789915966386, + "grad_norm": 3.650192626479405, + "learning_rate": 5.539912772500943e-07, + "loss": 0.0862, + "step": 1010 + }, + { + "epoch": 1.6991596638655462, + "grad_norm": 3.78077938487044, + "learning_rate": 5.47967589685292e-07, + "loss": 0.0892, + "step": 1011 + }, + { + "epoch": 1.7008403361344537, + "grad_norm": 2.7394887442627796, + "learning_rate": 5.419749308424377e-07, + "loss": 0.063, + "step": 1012 + }, + { + "epoch": 1.7025210084033613, + "grad_norm": 2.7156577152398538, + "learning_rate": 5.360133424877467e-07, + "loss": 0.0991, + "step": 1013 + }, + { + "epoch": 1.704201680672269, + "grad_norm": 3.596863078893733, + "learning_rate": 5.300828661708873e-07, + "loss": 0.124, + "step": 1014 + }, + { + "epoch": 1.7058823529411766, + "grad_norm": 2.702400026116516, + "learning_rate": 5.241835432246888e-07, + "loss": 0.0779, + "step": 1015 + }, + { + "epoch": 1.7075630252100842, + "grad_norm": 1.9594421326250522, + "learning_rate": 5.183154147648578e-07, + "loss": 0.0636, + "step": 1016 + }, + { + "epoch": 1.7092436974789917, + "grad_norm": 2.3716518580813006, + "learning_rate": 5.124785216896854e-07, + "loss": 0.065, + "step": 1017 + }, + { + "epoch": 1.7109243697478993, + "grad_norm": 1.8907181889521054, + "learning_rate": 5.066729046797692e-07, + "loss": 0.0473, + "step": 1018 + }, + { + "epoch": 1.7126050420168069, + "grad_norm": 2.4203847174896884, + "learning_rate": 5.008986041977254e-07, + "loss": 0.0544, + "step": 1019 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 3.053591851347738, + "learning_rate": 4.951556604879049e-07, + "loss": 0.084, + "step": 1020 + }, + { + "epoch": 1.715966386554622, + "grad_norm": 2.925477119081493, + "learning_rate": 4.894441135761197e-07, + "loss": 0.0962, + "step": 1021 + }, + { + "epoch": 1.7176470588235295, + "grad_norm": 2.2302715873385166, + "learning_rate": 4.837640032693558e-07, + "loss": 0.0966, + "step": 1022 + }, + { + "epoch": 1.719327731092437, + "grad_norm": 3.533404563017714, + "learning_rate": 4.781153691555035e-07, + "loss": 0.163, + "step": 1023 + }, + { + "epoch": 1.7210084033613446, + "grad_norm": 2.597578370940511, + "learning_rate": 4.724982506030762e-07, + "loss": 0.0756, + "step": 1024 + }, + { + "epoch": 1.7226890756302522, + "grad_norm": 3.1659295913472554, + "learning_rate": 4.669126867609375e-07, + "loss": 0.085, + "step": 1025 + }, + { + "epoch": 1.7243697478991598, + "grad_norm": 2.847957423266496, + "learning_rate": 4.613587165580269e-07, + "loss": 0.0917, + "step": 1026 + }, + { + "epoch": 1.7260504201680673, + "grad_norm": 2.4839081103193683, + "learning_rate": 4.5583637870309397e-07, + "loss": 0.0709, + "step": 1027 + }, + { + "epoch": 1.7277310924369749, + "grad_norm": 1.9119792986038078, + "learning_rate": 4.503457116844201e-07, + "loss": 0.0636, + "step": 1028 + }, + { + "epoch": 1.7294117647058824, + "grad_norm": 3.5996944839451395, + "learning_rate": 4.448867537695578e-07, + "loss": 0.0813, + "step": 1029 + }, + { + "epoch": 1.73109243697479, + "grad_norm": 1.7931626779021408, + "learning_rate": 4.394595430050613e-07, + "loss": 0.0553, + "step": 1030 + }, + { + "epoch": 1.7327731092436975, + "grad_norm": 2.3675109843091158, + "learning_rate": 4.34064117216218e-07, + "loss": 0.0555, + "step": 1031 + }, + { + "epoch": 1.734453781512605, + "grad_norm": 3.2733999141078143, + "learning_rate": 4.287005140067912e-07, + "loss": 0.1126, + "step": 1032 + }, + { + "epoch": 1.7361344537815127, + "grad_norm": 2.0302561800728944, + "learning_rate": 4.2336877075875136e-07, + "loss": 0.047, + "step": 1033 + }, + { + "epoch": 1.7378151260504202, + "grad_norm": 1.7878895517124131, + "learning_rate": 4.1806892463202353e-07, + "loss": 0.0403, + "step": 1034 + }, + { + "epoch": 1.7394957983193278, + "grad_norm": 2.517182395401572, + "learning_rate": 4.1280101256421903e-07, + "loss": 0.064, + "step": 1035 + }, + { + "epoch": 1.7411764705882353, + "grad_norm": 2.8819937038701005, + "learning_rate": 4.0756507127038494e-07, + "loss": 0.0839, + "step": 1036 + }, + { + "epoch": 1.7428571428571429, + "grad_norm": 2.382102175428615, + "learning_rate": 4.0236113724274716e-07, + "loss": 0.0583, + "step": 1037 + }, + { + "epoch": 1.7445378151260504, + "grad_norm": 2.9497304360347716, + "learning_rate": 3.971892467504518e-07, + "loss": 0.0689, + "step": 1038 + }, + { + "epoch": 1.746218487394958, + "grad_norm": 2.0320898349329886, + "learning_rate": 3.9204943583931953e-07, + "loss": 0.0682, + "step": 1039 + }, + { + "epoch": 1.7478991596638656, + "grad_norm": 3.004623695811176, + "learning_rate": 3.869417403315856e-07, + "loss": 0.0925, + "step": 1040 + }, + { + "epoch": 1.749579831932773, + "grad_norm": 2.747110324380183, + "learning_rate": 3.8186619582565974e-07, + "loss": 0.0586, + "step": 1041 + }, + { + "epoch": 1.7512605042016807, + "grad_norm": 2.094632480179336, + "learning_rate": 3.7682283769586883e-07, + "loss": 0.0538, + "step": 1042 + }, + { + "epoch": 1.7529411764705882, + "grad_norm": 2.7621486553756034, + "learning_rate": 3.71811701092219e-07, + "loss": 0.1123, + "step": 1043 + }, + { + "epoch": 1.7546218487394958, + "grad_norm": 2.7045998618097435, + "learning_rate": 3.6683282094014285e-07, + "loss": 0.0691, + "step": 1044 + }, + { + "epoch": 1.7563025210084033, + "grad_norm": 3.439936853763029, + "learning_rate": 3.6188623194026105e-07, + "loss": 0.085, + "step": 1045 + }, + { + "epoch": 1.757983193277311, + "grad_norm": 1.9441996285629553, + "learning_rate": 3.569719685681405e-07, + "loss": 0.0432, + "step": 1046 + }, + { + "epoch": 1.7596638655462185, + "grad_norm": 2.340946200287024, + "learning_rate": 3.5209006507404883e-07, + "loss": 0.0756, + "step": 1047 + }, + { + "epoch": 1.761344537815126, + "grad_norm": 1.868301185446278, + "learning_rate": 3.4724055548272406e-07, + "loss": 0.0589, + "step": 1048 + }, + { + "epoch": 1.7630252100840336, + "grad_norm": 2.7014525404333627, + "learning_rate": 3.4242347359312864e-07, + "loss": 0.0931, + "step": 1049 + }, + { + "epoch": 1.7647058823529411, + "grad_norm": 2.8475033114504877, + "learning_rate": 3.3763885297822153e-07, + "loss": 0.0735, + "step": 1050 + }, + { + "epoch": 1.7663865546218487, + "grad_norm": 2.9891854484974125, + "learning_rate": 3.3288672698471804e-07, + "loss": 0.0852, + "step": 1051 + }, + { + "epoch": 1.7680672268907562, + "grad_norm": 2.542773456492228, + "learning_rate": 3.281671287328614e-07, + "loss": 0.0732, + "step": 1052 + }, + { + "epoch": 1.7697478991596638, + "grad_norm": 3.6329591813631508, + "learning_rate": 3.2348009111619227e-07, + "loss": 0.0985, + "step": 1053 + }, + { + "epoch": 1.7714285714285714, + "grad_norm": 2.007617360456515, + "learning_rate": 3.18825646801314e-07, + "loss": 0.0477, + "step": 1054 + }, + { + "epoch": 1.773109243697479, + "grad_norm": 2.8286776456999094, + "learning_rate": 3.1420382822767326e-07, + "loss": 0.0835, + "step": 1055 + }, + { + "epoch": 1.7747899159663865, + "grad_norm": 2.8970226798307066, + "learning_rate": 3.096146676073253e-07, + "loss": 0.0989, + "step": 1056 + }, + { + "epoch": 1.776470588235294, + "grad_norm": 2.2847322270579453, + "learning_rate": 3.0505819692471797e-07, + "loss": 0.0629, + "step": 1057 + }, + { + "epoch": 1.7781512605042016, + "grad_norm": 3.247777179885162, + "learning_rate": 3.0053444793646024e-07, + "loss": 0.1072, + "step": 1058 + }, + { + "epoch": 1.7798319327731091, + "grad_norm": 2.8398776729370447, + "learning_rate": 2.960434521711086e-07, + "loss": 0.0636, + "step": 1059 + }, + { + "epoch": 1.7815126050420167, + "grad_norm": 1.7198170726412092, + "learning_rate": 2.915852409289421e-07, + "loss": 0.0459, + "step": 1060 + }, + { + "epoch": 1.7831932773109243, + "grad_norm": 2.47365582662474, + "learning_rate": 2.8715984528174757e-07, + "loss": 0.0565, + "step": 1061 + }, + { + "epoch": 1.7848739495798318, + "grad_norm": 2.7202776621889004, + "learning_rate": 2.827672960725991e-07, + "loss": 0.0868, + "step": 1062 + }, + { + "epoch": 1.7865546218487394, + "grad_norm": 2.7847895905936837, + "learning_rate": 2.7840762391564634e-07, + "loss": 0.0817, + "step": 1063 + }, + { + "epoch": 1.788235294117647, + "grad_norm": 2.2785885981744243, + "learning_rate": 2.7408085919590265e-07, + "loss": 0.057, + "step": 1064 + }, + { + "epoch": 1.7899159663865545, + "grad_norm": 2.5726828881196315, + "learning_rate": 2.697870320690266e-07, + "loss": 0.0738, + "step": 1065 + }, + { + "epoch": 1.791596638655462, + "grad_norm": 2.4462470824133793, + "learning_rate": 2.6552617246111966e-07, + "loss": 0.067, + "step": 1066 + }, + { + "epoch": 1.7932773109243696, + "grad_norm": 2.1874531530519956, + "learning_rate": 2.612983100685118e-07, + "loss": 0.0572, + "step": 1067 + }, + { + "epoch": 1.7949579831932772, + "grad_norm": 3.148057488521974, + "learning_rate": 2.5710347435755955e-07, + "loss": 0.0768, + "step": 1068 + }, + { + "epoch": 1.7966386554621847, + "grad_norm": 2.2736618720698942, + "learning_rate": 2.5294169456443416e-07, + "loss": 0.0607, + "step": 1069 + }, + { + "epoch": 1.7983193277310925, + "grad_norm": 2.408408179557469, + "learning_rate": 2.4881299969492514e-07, + "loss": 0.0839, + "step": 1070 + }, + { + "epoch": 1.8, + "grad_norm": 2.7245275157909683, + "learning_rate": 2.447174185242324e-07, + "loss": 0.0963, + "step": 1071 + }, + { + "epoch": 1.8016806722689076, + "grad_norm": 2.546372677098468, + "learning_rate": 2.406549795967678e-07, + "loss": 0.0501, + "step": 1072 + }, + { + "epoch": 1.8033613445378152, + "grad_norm": 2.2774525189800703, + "learning_rate": 2.366257112259579e-07, + "loss": 0.0506, + "step": 1073 + }, + { + "epoch": 1.8050420168067227, + "grad_norm": 1.9454534628385145, + "learning_rate": 2.3262964149404322e-07, + "loss": 0.0698, + "step": 1074 + }, + { + "epoch": 1.8067226890756303, + "grad_norm": 2.7533389388257405, + "learning_rate": 2.286667982518853e-07, + "loss": 0.0828, + "step": 1075 + }, + { + "epoch": 1.8084033613445378, + "grad_norm": 2.3409625134750476, + "learning_rate": 2.247372091187705e-07, + "loss": 0.0657, + "step": 1076 + }, + { + "epoch": 1.8100840336134454, + "grad_norm": 3.118804130005249, + "learning_rate": 2.2084090148221937e-07, + "loss": 0.1111, + "step": 1077 + }, + { + "epoch": 1.811764705882353, + "grad_norm": 2.401067953654103, + "learning_rate": 2.1697790249779638e-07, + "loss": 0.0748, + "step": 1078 + }, + { + "epoch": 1.8134453781512605, + "grad_norm": 3.0291208574888193, + "learning_rate": 2.1314823908891558e-07, + "loss": 0.0889, + "step": 1079 + }, + { + "epoch": 1.815126050420168, + "grad_norm": 2.75949156227135, + "learning_rate": 2.093519379466602e-07, + "loss": 0.0776, + "step": 1080 + }, + { + "epoch": 1.8168067226890756, + "grad_norm": 2.00422109514553, + "learning_rate": 2.0558902552959058e-07, + "loss": 0.0598, + "step": 1081 + }, + { + "epoch": 1.8184873949579832, + "grad_norm": 3.509921644126103, + "learning_rate": 2.018595280635638e-07, + "loss": 0.077, + "step": 1082 + }, + { + "epoch": 1.8201680672268907, + "grad_norm": 2.395269012866003, + "learning_rate": 1.981634715415487e-07, + "loss": 0.0476, + "step": 1083 + }, + { + "epoch": 1.8218487394957983, + "grad_norm": 3.1367885506904933, + "learning_rate": 1.945008817234445e-07, + "loss": 0.0899, + "step": 1084 + }, + { + "epoch": 1.8235294117647058, + "grad_norm": 2.831809081450393, + "learning_rate": 1.908717841359048e-07, + "loss": 0.0669, + "step": 1085 + }, + { + "epoch": 1.8252100840336134, + "grad_norm": 2.5459432634484083, + "learning_rate": 1.8727620407215375e-07, + "loss": 0.0558, + "step": 1086 + }, + { + "epoch": 1.826890756302521, + "grad_norm": 3.489221037352137, + "learning_rate": 1.837141665918152e-07, + "loss": 0.0928, + "step": 1087 + }, + { + "epoch": 1.8285714285714287, + "grad_norm": 2.668603090786725, + "learning_rate": 1.801856965207338e-07, + "loss": 0.066, + "step": 1088 + }, + { + "epoch": 1.8302521008403363, + "grad_norm": 2.9412325741168868, + "learning_rate": 1.7669081845080648e-07, + "loss": 0.0781, + "step": 1089 + }, + { + "epoch": 1.8319327731092439, + "grad_norm": 2.5595453190415802, + "learning_rate": 1.7322955673980678e-07, + "loss": 0.0681, + "step": 1090 + }, + { + "epoch": 1.8336134453781514, + "grad_norm": 2.656909922356702, + "learning_rate": 1.6980193551121848e-07, + "loss": 0.0535, + "step": 1091 + }, + { + "epoch": 1.835294117647059, + "grad_norm": 2.7002698095093534, + "learning_rate": 1.664079786540629e-07, + "loss": 0.0642, + "step": 1092 + }, + { + "epoch": 1.8369747899159665, + "grad_norm": 2.9100825480724506, + "learning_rate": 1.6304770982273898e-07, + "loss": 0.0979, + "step": 1093 + }, + { + "epoch": 1.838655462184874, + "grad_norm": 2.5089073185708646, + "learning_rate": 1.597211524368536e-07, + "loss": 0.0695, + "step": 1094 + }, + { + "epoch": 1.8403361344537816, + "grad_norm": 2.9673819680310936, + "learning_rate": 1.564283296810576e-07, + "loss": 0.0826, + "step": 1095 + }, + { + "epoch": 1.8420168067226892, + "grad_norm": 2.4463338389394402, + "learning_rate": 1.5316926450488878e-07, + "loss": 0.0619, + "step": 1096 + }, + { + "epoch": 1.8436974789915967, + "grad_norm": 1.9871122803788783, + "learning_rate": 1.499439796226082e-07, + "loss": 0.0536, + "step": 1097 + }, + { + "epoch": 1.8453781512605043, + "grad_norm": 3.047793126703823, + "learning_rate": 1.4675249751304353e-07, + "loss": 0.07, + "step": 1098 + }, + { + "epoch": 1.8470588235294119, + "grad_norm": 2.573964395132922, + "learning_rate": 1.435948404194304e-07, + "loss": 0.0546, + "step": 1099 + }, + { + "epoch": 1.8487394957983194, + "grad_norm": 2.157568300113264, + "learning_rate": 1.404710303492618e-07, + "loss": 0.0521, + "step": 1100 + }, + { + "epoch": 1.850420168067227, + "grad_norm": 2.5093947332693434, + "learning_rate": 1.373810890741284e-07, + "loss": 0.0591, + "step": 1101 + }, + { + "epoch": 1.8521008403361345, + "grad_norm": 2.7798941234564527, + "learning_rate": 1.3432503812957242e-07, + "loss": 0.0861, + "step": 1102 + }, + { + "epoch": 1.853781512605042, + "grad_norm": 3.3151152408491953, + "learning_rate": 1.3130289881493452e-07, + "loss": 0.1133, + "step": 1103 + }, + { + "epoch": 1.8554621848739496, + "grad_norm": 2.455051483279099, + "learning_rate": 1.2831469219320603e-07, + "loss": 0.0784, + "step": 1104 + }, + { + "epoch": 1.8571428571428572, + "grad_norm": 3.389083633842711, + "learning_rate": 1.253604390908819e-07, + "loss": 0.1172, + "step": 1105 + }, + { + "epoch": 1.8588235294117648, + "grad_norm": 2.0890124129209404, + "learning_rate": 1.22440160097817e-07, + "loss": 0.0597, + "step": 1106 + }, + { + "epoch": 1.8605042016806723, + "grad_norm": 2.5001823487473014, + "learning_rate": 1.1955387556708e-07, + "loss": 0.0484, + "step": 1107 + }, + { + "epoch": 1.8621848739495799, + "grad_norm": 1.972553815528427, + "learning_rate": 1.1670160561481458e-07, + "loss": 0.057, + "step": 1108 + }, + { + "epoch": 1.8638655462184874, + "grad_norm": 2.247555606030712, + "learning_rate": 1.1388337012009643e-07, + "loss": 0.067, + "step": 1109 + }, + { + "epoch": 1.865546218487395, + "grad_norm": 3.3951419821687034, + "learning_rate": 1.1109918872479642e-07, + "loss": 0.1168, + "step": 1110 + }, + { + "epoch": 1.8672268907563025, + "grad_norm": 2.561792162365592, + "learning_rate": 1.0834908083344253e-07, + "loss": 0.0737, + "step": 1111 + }, + { + "epoch": 1.86890756302521, + "grad_norm": 2.924693392143974, + "learning_rate": 1.0563306561308773e-07, + "loss": 0.0672, + "step": 1112 + }, + { + "epoch": 1.8705882352941177, + "grad_norm": 3.0666803684218835, + "learning_rate": 1.0295116199317057e-07, + "loss": 0.0758, + "step": 1113 + }, + { + "epoch": 1.8722689075630252, + "grad_norm": 3.1132381273055203, + "learning_rate": 1.0030338866538925e-07, + "loss": 0.0912, + "step": 1114 + }, + { + "epoch": 1.8739495798319328, + "grad_norm": 2.193473585701426, + "learning_rate": 9.768976408356667e-08, + "loss": 0.0737, + "step": 1115 + }, + { + "epoch": 1.8756302521008403, + "grad_norm": 2.316436170911881, + "learning_rate": 9.511030646352615e-08, + "loss": 0.0785, + "step": 1116 + }, + { + "epoch": 1.877310924369748, + "grad_norm": 3.2232498371051066, + "learning_rate": 9.256503378295978e-08, + "loss": 0.1079, + "step": 1117 + }, + { + "epoch": 1.8789915966386554, + "grad_norm": 2.989879633737669, + "learning_rate": 9.005396378130748e-08, + "loss": 0.0683, + "step": 1118 + }, + { + "epoch": 1.880672268907563, + "grad_norm": 2.897217749538415, + "learning_rate": 8.757711395963097e-08, + "loss": 0.104, + "step": 1119 + }, + { + "epoch": 1.8823529411764706, + "grad_norm": 5.9925913526073185, + "learning_rate": 8.513450158049109e-08, + "loss": 0.0856, + "step": 1120 + }, + { + "epoch": 1.8840336134453781, + "grad_norm": 3.437858233576137, + "learning_rate": 8.27261436678306e-08, + "loss": 0.083, + "step": 1121 + }, + { + "epoch": 1.8857142857142857, + "grad_norm": 2.1904465686897754, + "learning_rate": 8.035205700685167e-08, + "loss": 0.0551, + "step": 1122 + }, + { + "epoch": 1.8873949579831932, + "grad_norm": 3.574848102142179, + "learning_rate": 7.801225814390245e-08, + "loss": 0.0891, + "step": 1123 + }, + { + "epoch": 1.8890756302521008, + "grad_norm": 2.9774864616230166, + "learning_rate": 7.570676338635896e-08, + "loss": 0.0842, + "step": 1124 + }, + { + "epoch": 1.8907563025210083, + "grad_norm": 2.9240546637759635, + "learning_rate": 7.343558880251289e-08, + "loss": 0.0836, + "step": 1125 + }, + { + "epoch": 1.892436974789916, + "grad_norm": 2.536237081972416, + "learning_rate": 7.11987502214595e-08, + "loss": 0.0611, + "step": 1126 + }, + { + "epoch": 1.8941176470588235, + "grad_norm": 2.945869104329601, + "learning_rate": 6.899626323298714e-08, + "loss": 0.0661, + "step": 1127 + }, + { + "epoch": 1.895798319327731, + "grad_norm": 2.2128300706474624, + "learning_rate": 6.682814318746844e-08, + "loss": 0.0558, + "step": 1128 + }, + { + "epoch": 1.8974789915966386, + "grad_norm": 2.469504977662401, + "learning_rate": 6.46944051957532e-08, + "loss": 0.0971, + "step": 1129 + }, + { + "epoch": 1.8991596638655461, + "grad_norm": 2.336659774072773, + "learning_rate": 6.259506412906402e-08, + "loss": 0.0639, + "step": 1130 + }, + { + "epoch": 1.9008403361344537, + "grad_norm": 2.916145176561295, + "learning_rate": 6.053013461889023e-08, + "loss": 0.0748, + "step": 1131 + }, + { + "epoch": 1.9025210084033612, + "grad_norm": 2.7261524069835796, + "learning_rate": 5.849963105689027e-08, + "loss": 0.0767, + "step": 1132 + }, + { + "epoch": 1.9042016806722688, + "grad_norm": 2.9983934151605105, + "learning_rate": 5.65035675947867e-08, + "loss": 0.0721, + "step": 1133 + }, + { + "epoch": 1.9058823529411764, + "grad_norm": 2.1467336761652343, + "learning_rate": 5.454195814427021e-08, + "loss": 0.0576, + "step": 1134 + }, + { + "epoch": 1.907563025210084, + "grad_norm": 2.2227378360842436, + "learning_rate": 5.261481637690247e-08, + "loss": 0.0647, + "step": 1135 + }, + { + "epoch": 1.9092436974789915, + "grad_norm": 2.4269755333288194, + "learning_rate": 5.072215572402006e-08, + "loss": 0.0419, + "step": 1136 + }, + { + "epoch": 1.910924369747899, + "grad_norm": 3.0533438647127533, + "learning_rate": 4.886398937664127e-08, + "loss": 0.0734, + "step": 1137 + }, + { + "epoch": 1.9126050420168066, + "grad_norm": 3.2788567900077497, + "learning_rate": 4.704033028537391e-08, + "loss": 0.0902, + "step": 1138 + }, + { + "epoch": 1.9142857142857141, + "grad_norm": 2.551953448087741, + "learning_rate": 4.52511911603265e-08, + "loss": 0.0827, + "step": 1139 + }, + { + "epoch": 1.9159663865546217, + "grad_norm": 4.78760807070443, + "learning_rate": 4.3496584471016125e-08, + "loss": 0.0627, + "step": 1140 + }, + { + "epoch": 1.9176470588235293, + "grad_norm": 2.721593978748799, + "learning_rate": 4.177652244628627e-08, + "loss": 0.0628, + "step": 1141 + }, + { + "epoch": 1.9193277310924368, + "grad_norm": 2.3037527032294656, + "learning_rate": 4.009101707421803e-08, + "loss": 0.0668, + "step": 1142 + }, + { + "epoch": 1.9210084033613444, + "grad_norm": 2.480126673107869, + "learning_rate": 3.8440080102047364e-08, + "loss": 0.0724, + "step": 1143 + }, + { + "epoch": 1.9226890756302522, + "grad_norm": 3.1491790650313236, + "learning_rate": 3.6823723036084616e-08, + "loss": 0.0667, + "step": 1144 + }, + { + "epoch": 1.9243697478991597, + "grad_norm": 3.297054116842271, + "learning_rate": 3.5241957141632923e-08, + "loss": 0.1002, + "step": 1145 + }, + { + "epoch": 1.9260504201680673, + "grad_norm": 3.5015254517965873, + "learning_rate": 3.369479344290938e-08, + "loss": 0.1012, + "step": 1146 + }, + { + "epoch": 1.9277310924369748, + "grad_norm": 2.2369857549858376, + "learning_rate": 3.218224272296955e-08, + "loss": 0.0449, + "step": 1147 + }, + { + "epoch": 1.9294117647058824, + "grad_norm": 2.2243368132738723, + "learning_rate": 3.0704315523631956e-08, + "loss": 0.0528, + "step": 1148 + }, + { + "epoch": 1.93109243697479, + "grad_norm": 2.104823571308308, + "learning_rate": 2.926102214540316e-08, + "loss": 0.0474, + "step": 1149 + }, + { + "epoch": 1.9327731092436975, + "grad_norm": 2.832894345048546, + "learning_rate": 2.7852372647407812e-08, + "loss": 0.0747, + "step": 1150 + }, + { + "epoch": 1.934453781512605, + "grad_norm": 4.335292022183732, + "learning_rate": 2.6478376847318687e-08, + "loss": 0.0696, + "step": 1151 + }, + { + "epoch": 1.9361344537815126, + "grad_norm": 2.697120477019712, + "learning_rate": 2.5139044321286223e-08, + "loss": 0.0791, + "step": 1152 + }, + { + "epoch": 1.9378151260504202, + "grad_norm": 2.54968758188515, + "learning_rate": 2.383438440387298e-08, + "loss": 0.0745, + "step": 1153 + }, + { + "epoch": 1.9394957983193277, + "grad_norm": 3.02832569342735, + "learning_rate": 2.256440618798872e-08, + "loss": 0.0975, + "step": 1154 + }, + { + "epoch": 1.9411764705882353, + "grad_norm": 3.387365620438066, + "learning_rate": 2.1329118524827662e-08, + "loss": 0.0923, + "step": 1155 + }, + { + "epoch": 1.9428571428571428, + "grad_norm": 2.6767183811646413, + "learning_rate": 2.012853002380466e-08, + "loss": 0.0536, + "step": 1156 + }, + { + "epoch": 1.9445378151260504, + "grad_norm": 3.1445720922314337, + "learning_rate": 1.896264905249856e-08, + "loss": 0.1053, + "step": 1157 + }, + { + "epoch": 1.946218487394958, + "grad_norm": 3.2263045004178017, + "learning_rate": 1.783148373659005e-08, + "loss": 0.0798, + "step": 1158 + }, + { + "epoch": 1.9478991596638655, + "grad_norm": 2.176030962064042, + "learning_rate": 1.6735041959806686e-08, + "loss": 0.0512, + "step": 1159 + }, + { + "epoch": 1.949579831932773, + "grad_norm": 2.22081721828048, + "learning_rate": 1.567333136387017e-08, + "loss": 0.064, + "step": 1160 + }, + { + "epoch": 1.9512605042016806, + "grad_norm": 3.576068882581677, + "learning_rate": 1.4646359348439165e-08, + "loss": 0.1032, + "step": 1161 + }, + { + "epoch": 1.9529411764705882, + "grad_norm": 2.7548882247249846, + "learning_rate": 1.3654133071059894e-08, + "loss": 0.0958, + "step": 1162 + }, + { + "epoch": 1.954621848739496, + "grad_norm": 2.0537531997332277, + "learning_rate": 1.2696659447116732e-08, + "loss": 0.0555, + "step": 1163 + }, + { + "epoch": 1.9563025210084035, + "grad_norm": 2.576434155048857, + "learning_rate": 1.1773945149782805e-08, + "loss": 0.0681, + "step": 1164 + }, + { + "epoch": 1.957983193277311, + "grad_norm": 2.3313055510378513, + "learning_rate": 1.088599660997336e-08, + "loss": 0.0682, + "step": 1165 + }, + { + "epoch": 1.9596638655462186, + "grad_norm": 3.117362012224255, + "learning_rate": 1.0032820016302458e-08, + "loss": 0.0637, + "step": 1166 + }, + { + "epoch": 1.9613445378151262, + "grad_norm": 3.1378506905510215, + "learning_rate": 9.21442131503858e-09, + "loss": 0.0885, + "step": 1167 + }, + { + "epoch": 1.9630252100840337, + "grad_norm": 3.0313070424172217, + "learning_rate": 8.430806210062426e-09, + "loss": 0.0949, + "step": 1168 + }, + { + "epoch": 1.9647058823529413, + "grad_norm": 1.8993144925399335, + "learning_rate": 7.681980162830283e-09, + "loss": 0.0566, + "step": 1169 + }, + { + "epoch": 1.9663865546218489, + "grad_norm": 2.78106637596925, + "learning_rate": 6.9679483923318356e-09, + "loss": 0.0655, + "step": 1170 + }, + { + "epoch": 1.9680672268907564, + "grad_norm": 2.6179872514492843, + "learning_rate": 6.288715875057416e-09, + "loss": 0.0863, + "step": 1171 + }, + { + "epoch": 1.969747899159664, + "grad_norm": 2.85141007344782, + "learning_rate": 5.644287344960253e-09, + "loss": 0.0495, + "step": 1172 + }, + { + "epoch": 1.9714285714285715, + "grad_norm": 1.9037225399426012, + "learning_rate": 5.034667293427053e-09, + "loss": 0.0531, + "step": 1173 + }, + { + "epoch": 1.973109243697479, + "grad_norm": 2.8597867728725825, + "learning_rate": 4.45985996924192e-09, + "loss": 0.076, + "step": 1174 + }, + { + "epoch": 1.9747899159663866, + "grad_norm": 2.384741719443929, + "learning_rate": 3.919869378561925e-09, + "loss": 0.0781, + "step": 1175 + }, + { + "epoch": 1.9764705882352942, + "grad_norm": 3.3663742321659793, + "learning_rate": 3.41469928488547e-09, + "loss": 0.0829, + "step": 1176 + }, + { + "epoch": 1.9781512605042018, + "grad_norm": 2.5855062360633063, + "learning_rate": 2.9443532090273064e-09, + "loss": 0.068, + "step": 1177 + }, + { + "epoch": 1.9798319327731093, + "grad_norm": 2.349097606121586, + "learning_rate": 2.508834429094664e-09, + "loss": 0.0672, + "step": 1178 + }, + { + "epoch": 1.9815126050420169, + "grad_norm": 2.188284885883368, + "learning_rate": 2.108145980462828e-09, + "loss": 0.0587, + "step": 1179 + }, + { + "epoch": 1.9831932773109244, + "grad_norm": 2.603925500773801, + "learning_rate": 1.7422906557557074e-09, + "loss": 0.0898, + "step": 1180 + }, + { + "epoch": 1.984873949579832, + "grad_norm": 2.072728888868806, + "learning_rate": 1.4112710048247436e-09, + "loss": 0.0689, + "step": 1181 + }, + { + "epoch": 1.9865546218487395, + "grad_norm": 2.2130213181251284, + "learning_rate": 1.1150893347328107e-09, + "loss": 0.0537, + "step": 1182 + }, + { + "epoch": 1.988235294117647, + "grad_norm": 2.2290325180753983, + "learning_rate": 8.537477097364522e-10, + "loss": 0.0739, + "step": 1183 + }, + { + "epoch": 1.9899159663865547, + "grad_norm": 2.915171945856223, + "learning_rate": 6.272479512731133e-10, + "loss": 0.0693, + "step": 1184 + }, + { + "epoch": 1.9915966386554622, + "grad_norm": 3.387630014023736, + "learning_rate": 4.3559163794670844e-10, + "loss": 0.097, + "step": 1185 + }, + { + "epoch": 1.9932773109243698, + "grad_norm": 2.778376156187998, + "learning_rate": 2.787801055181838e-10, + "loss": 0.0861, + "step": 1186 + }, + { + "epoch": 1.9949579831932773, + "grad_norm": 2.6278202477958623, + "learning_rate": 1.568144468955257e-10, + "loss": 0.0779, + "step": 1187 + }, + { + "epoch": 1.9966386554621849, + "grad_norm": 2.434531776499178, + "learning_rate": 6.969551212598901e-11, + "loss": 0.0615, + "step": 1188 + }, + { + "epoch": 1.9983193277310924, + "grad_norm": 2.7044583414113026, + "learning_rate": 1.7423908390545862e-11, + "loss": 0.0862, + "step": 1189 + }, + { + "epoch": 2.0, + "grad_norm": 1.9872586148824984, + "learning_rate": 0.0, + "loss": 0.0517, + "step": 1190 + }, + { + "epoch": 2.0, + "step": 1190, + "total_flos": 6634871193600.0, + "train_loss": 0.17354622093819772, + "train_runtime": 2680.2801, + "train_samples_per_second": 3.549, + "train_steps_per_second": 0.444 + } + ], + "logging_steps": 1, + "max_steps": 1190, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 50000000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6634871193600.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}