diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,28089 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.5296367112810707, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0003824091778202677, + "grad_norm": 8.127729221609483, + "learning_rate": 0.0, + "loss": 0.7499, + "step": 1 + }, + { + "epoch": 0.0007648183556405354, + "grad_norm": 7.963595529643983, + "learning_rate": 9.658899046395883e-08, + "loss": 0.7745, + "step": 2 + }, + { + "epoch": 0.001147227533460803, + "grad_norm": 6.399453926631711, + "learning_rate": 1.530899278678881e-07, + "loss": 0.827, + "step": 3 + }, + { + "epoch": 0.0015296367112810707, + "grad_norm": 7.212906755583777, + "learning_rate": 1.9317798092791765e-07, + "loss": 0.9354, + "step": 4 + }, + { + "epoch": 0.0019120458891013384, + "grad_norm": 4.403583991640848, + "learning_rate": 2.2427269061507353e-07, + "loss": 0.8108, + "step": 5 + }, + { + "epoch": 0.002294455066921606, + "grad_norm": 9.30785115586022, + "learning_rate": 2.496789183318469e-07, + "loss": 0.8121, + "step": 6 + }, + { + "epoch": 0.0026768642447418736, + "grad_norm": 10.972546390514474, + "learning_rate": 2.711595777955698e-07, + "loss": 0.8276, + "step": 7 + }, + { + "epoch": 0.0030592734225621415, + "grad_norm": 7.964528600005935, + "learning_rate": 2.8976697139187646e-07, + "loss": 0.889, + "step": 8 + }, + { + "epoch": 0.0034416826003824093, + "grad_norm": 11.538619501059664, + "learning_rate": 3.061798557357762e-07, + "loss": 0.5456, + "step": 9 + }, + { + "epoch": 0.0038240917782026767, + "grad_norm": 9.923505802246833, + "learning_rate": 3.208616810790324e-07, + "loss": 0.5617, + "step": 10 + }, + { + "epoch": 0.004206500956022944, + "grad_norm": 13.868230594571852, + "learning_rate": 3.341430076232756e-07, + "loss": 0.8914, + "step": 11 + }, + { + "epoch": 0.004588910133843212, + "grad_norm": 15.82952470982823, + "learning_rate": 3.4626790879580574e-07, + "loss": 0.3943, + "step": 12 + }, + { + "epoch": 0.00497131931166348, + "grad_norm": 13.881207151660286, + "learning_rate": 3.5742173664798447e-07, + "loss": 0.9886, + "step": 13 + }, + { + "epoch": 0.005353728489483747, + "grad_norm": 9.351576777218634, + "learning_rate": 3.677485682595286e-07, + "loss": 1.2225, + "step": 14 + }, + { + "epoch": 0.0057361376673040155, + "grad_norm": 6.921096792456155, + "learning_rate": 3.773626184829616e-07, + "loss": 0.6324, + "step": 15 + }, + { + "epoch": 0.006118546845124283, + "grad_norm": 7.911741257096208, + "learning_rate": 3.863559618558353e-07, + "loss": 1.0263, + "step": 16 + }, + { + "epoch": 0.00650095602294455, + "grad_norm": 9.023578698841147, + "learning_rate": 3.948039093953151e-07, + "loss": 0.8379, + "step": 17 + }, + { + "epoch": 0.006883365200764819, + "grad_norm": 5.773447080611977, + "learning_rate": 4.02768846199735e-07, + "loss": 0.6313, + "step": 18 + }, + { + "epoch": 0.007265774378585086, + "grad_norm": 6.391540312564622, + "learning_rate": 4.103030300875908e-07, + "loss": 0.6562, + "step": 19 + }, + { + "epoch": 0.0076481835564053535, + "grad_norm": 4.734698092574861, + "learning_rate": 4.174506715429912e-07, + "loss": 0.9944, + "step": 20 + }, + { + "epoch": 0.00803059273422562, + "grad_norm": 5.543325744703233, + "learning_rate": 4.242495056634579e-07, + "loss": 0.8748, + "step": 21 + }, + { + "epoch": 0.008413001912045888, + "grad_norm": 6.139313458259769, + "learning_rate": 4.3073199808723443e-07, + "loss": 0.6497, + "step": 22 + }, + { + "epoch": 0.008795411089866157, + "grad_norm": 10.51882881230223, + "learning_rate": 4.369262826367178e-07, + "loss": 0.7235, + "step": 23 + }, + { + "epoch": 0.009177820267686425, + "grad_norm": 9.14158481873066, + "learning_rate": 4.428568992597646e-07, + "loss": 0.4482, + "step": 24 + }, + { + "epoch": 0.009560229445506692, + "grad_norm": 10.990275506131937, + "learning_rate": 4.4854538123014706e-07, + "loss": 0.8668, + "step": 25 + }, + { + "epoch": 0.00994263862332696, + "grad_norm": 11.815499706960994, + "learning_rate": 4.540107271119433e-07, + "loss": 0.8262, + "step": 26 + }, + { + "epoch": 0.010325047801147227, + "grad_norm": 7.542823857279405, + "learning_rate": 4.592697836036643e-07, + "loss": 0.9995, + "step": 27 + }, + { + "epoch": 0.010707456978967494, + "grad_norm": 3.439453125, + "learning_rate": 4.6433755872348744e-07, + "loss": 0.652, + "step": 28 + }, + { + "epoch": 0.011089866156787764, + "grad_norm": 4.447922408065247, + "learning_rate": 4.6922748001247033e-07, + "loss": 0.4116, + "step": 29 + }, + { + "epoch": 0.011472275334608031, + "grad_norm": 3.5419146432268502, + "learning_rate": 4.739516089469205e-07, + "loss": 0.603, + "step": 30 + }, + { + "epoch": 0.011854684512428298, + "grad_norm": 5.595453269399992, + "learning_rate": 4.78520820180538e-07, + "loss": 0.8754, + "step": 31 + }, + { + "epoch": 0.012237093690248566, + "grad_norm": 4.664053298711087, + "learning_rate": 4.829449523197942e-07, + "loss": 0.728, + "step": 32 + }, + { + "epoch": 0.012619502868068833, + "grad_norm": 4.184285994388638, + "learning_rate": 4.872329354911637e-07, + "loss": 0.7668, + "step": 33 + }, + { + "epoch": 0.0130019120458891, + "grad_norm": 3.8462487759245723, + "learning_rate": 4.91392899859274e-07, + "loss": 0.7642, + "step": 34 + }, + { + "epoch": 0.01338432122370937, + "grad_norm": 5.035267807203385, + "learning_rate": 4.954322684106434e-07, + "loss": 0.4231, + "step": 35 + }, + { + "epoch": 0.013766730401529637, + "grad_norm": 4.700427932737028, + "learning_rate": 4.993578366636938e-07, + "loss": 0.4554, + "step": 36 + }, + { + "epoch": 0.014149139579349905, + "grad_norm": 6.65272564780195, + "learning_rate": 5.031758414551729e-07, + "loss": 0.3467, + "step": 37 + }, + { + "epoch": 0.014531548757170172, + "grad_norm": 6.540097900964421, + "learning_rate": 5.068920205515497e-07, + "loss": 0.5521, + "step": 38 + }, + { + "epoch": 0.01491395793499044, + "grad_norm": 6.187611588521872, + "learning_rate": 5.105116645158726e-07, + "loss": 0.8477, + "step": 39 + }, + { + "epoch": 0.015296367112810707, + "grad_norm": 4.212933408814397, + "learning_rate": 5.140396620069501e-07, + "loss": 0.8013, + "step": 40 + }, + { + "epoch": 0.015678776290630976, + "grad_norm": 3.7306584334956354, + "learning_rate": 5.174805394842197e-07, + "loss": 0.34, + "step": 41 + }, + { + "epoch": 0.01606118546845124, + "grad_norm": 3.8129084634065613, + "learning_rate": 5.208384961274167e-07, + "loss": 0.7891, + "step": 42 + }, + { + "epoch": 0.01644359464627151, + "grad_norm": 3.521126067201788, + "learning_rate": 5.241174346468368e-07, + "loss": 0.8447, + "step": 43 + }, + { + "epoch": 0.016826003824091777, + "grad_norm": 5.279756052566196, + "learning_rate": 5.273209885511933e-07, + "loss": 0.6151, + "step": 44 + }, + { + "epoch": 0.017208413001912046, + "grad_norm": 4.009723289626729, + "learning_rate": 5.304525463508497e-07, + "loss": 0.6095, + "step": 45 + }, + { + "epoch": 0.017590822179732315, + "grad_norm": 3.7728758842220267, + "learning_rate": 5.335152731006765e-07, + "loss": 0.3871, + "step": 46 + }, + { + "epoch": 0.01797323135755258, + "grad_norm": 4.866870067648313, + "learning_rate": 5.365121296259033e-07, + "loss": 0.4796, + "step": 47 + }, + { + "epoch": 0.01835564053537285, + "grad_norm": 4.8744050665499286, + "learning_rate": 5.394458897237234e-07, + "loss": 0.5812, + "step": 48 + }, + { + "epoch": 0.018738049713193115, + "grad_norm": 9.108527378017467, + "learning_rate": 5.423191555911396e-07, + "loss": 0.319, + "step": 49 + }, + { + "epoch": 0.019120458891013385, + "grad_norm": 9.158412014105785, + "learning_rate": 5.45134371694106e-07, + "loss": 0.7643, + "step": 50 + }, + { + "epoch": 0.019502868068833654, + "grad_norm": 5.790040627239837, + "learning_rate": 5.478938372632032e-07, + "loss": 0.6109, + "step": 51 + }, + { + "epoch": 0.01988527724665392, + "grad_norm": 6.096683676341199, + "learning_rate": 5.505997175759021e-07, + "loss": 0.9369, + "step": 52 + }, + { + "epoch": 0.02026768642447419, + "grad_norm": 4.441546891668887, + "learning_rate": 5.532540541641197e-07, + "loss": 0.7858, + "step": 53 + }, + { + "epoch": 0.020650095602294454, + "grad_norm": 2.8696014470488382, + "learning_rate": 5.558587740676232e-07, + "loss": 0.8667, + "step": 54 + }, + { + "epoch": 0.021032504780114723, + "grad_norm": 3.2738861286399046, + "learning_rate": 5.584156982383491e-07, + "loss": 0.4673, + "step": 55 + }, + { + "epoch": 0.02141491395793499, + "grad_norm": 3.697994492835673, + "learning_rate": 5.609265491874463e-07, + "loss": 0.6328, + "step": 56 + }, + { + "epoch": 0.021797323135755258, + "grad_norm": 3.125139767382225, + "learning_rate": 5.633929579554789e-07, + "loss": 0.4692, + "step": 57 + }, + { + "epoch": 0.022179732313575527, + "grad_norm": 4.298443428237915, + "learning_rate": 5.658164704764291e-07, + "loss": 0.6768, + "step": 58 + }, + { + "epoch": 0.022562141491395793, + "grad_norm": 4.398677297665077, + "learning_rate": 5.681985533976847e-07, + "loss": 0.6619, + "step": 59 + }, + { + "epoch": 0.022944550669216062, + "grad_norm": 5.818330527160778, + "learning_rate": 5.705405994108792e-07, + "loss": 0.6268, + "step": 60 + }, + { + "epoch": 0.023326959847036328, + "grad_norm": 5.36324426425703, + "learning_rate": 5.728439321421063e-07, + "loss": 0.3157, + "step": 61 + }, + { + "epoch": 0.023709369024856597, + "grad_norm": 5.248781153565317, + "learning_rate": 5.751098106444967e-07, + "loss": 0.3525, + "step": 62 + }, + { + "epoch": 0.024091778202676863, + "grad_norm": 6.7358008908634925, + "learning_rate": 5.77339433531346e-07, + "loss": 0.6165, + "step": 63 + }, + { + "epoch": 0.024474187380497132, + "grad_norm": 5.135947096269937, + "learning_rate": 5.795339427837529e-07, + "loss": 0.9469, + "step": 64 + }, + { + "epoch": 0.0248565965583174, + "grad_norm": 4.721443670581346, + "learning_rate": 5.816944272630579e-07, + "loss": 1.0586, + "step": 65 + }, + { + "epoch": 0.025239005736137667, + "grad_norm": 4.560383135778482, + "learning_rate": 5.838219259551224e-07, + "loss": 0.6226, + "step": 66 + }, + { + "epoch": 0.025621414913957936, + "grad_norm": 3.0533395900689975, + "learning_rate": 5.859174309706495e-07, + "loss": 0.5508, + "step": 67 + }, + { + "epoch": 0.0260038240917782, + "grad_norm": 4.38574534050627, + "learning_rate": 5.879818903232327e-07, + "loss": 0.6519, + "step": 68 + }, + { + "epoch": 0.02638623326959847, + "grad_norm": 3.1760295850591347, + "learning_rate": 5.900162105046059e-07, + "loss": 0.7339, + "step": 69 + }, + { + "epoch": 0.02676864244741874, + "grad_norm": 3.243632755108415, + "learning_rate": 5.920212588746022e-07, + "loss": 0.4918, + "step": 70 + }, + { + "epoch": 0.027151051625239005, + "grad_norm": 3.432834024756851, + "learning_rate": 5.939978658815961e-07, + "loss": 0.5308, + "step": 71 + }, + { + "epoch": 0.027533460803059275, + "grad_norm": 3.677125723417059, + "learning_rate": 5.959468271276526e-07, + "loss": 0.3372, + "step": 72 + }, + { + "epoch": 0.02791586998087954, + "grad_norm": 2.8933168816281856, + "learning_rate": 5.978689052912401e-07, + "loss": 0.3834, + "step": 73 + }, + { + "epoch": 0.02829827915869981, + "grad_norm": 4.377616318007501, + "learning_rate": 5.997648319191317e-07, + "loss": 0.285, + "step": 74 + }, + { + "epoch": 0.028680688336520075, + "grad_norm": 5.329271657395167, + "learning_rate": 6.016353090980351e-07, + "loss": 0.4062, + "step": 75 + }, + { + "epoch": 0.029063097514340344, + "grad_norm": 7.870238302614566, + "learning_rate": 6.034810110155085e-07, + "loss": 0.7765, + "step": 76 + }, + { + "epoch": 0.029445506692160613, + "grad_norm": 4.2678637096949625, + "learning_rate": 6.053025854188454e-07, + "loss": 0.7735, + "step": 77 + }, + { + "epoch": 0.02982791586998088, + "grad_norm": 3.691418909121812, + "learning_rate": 6.071006549798314e-07, + "loss": 0.7857, + "step": 78 + }, + { + "epoch": 0.030210325047801148, + "grad_norm": 2.8591930560816983, + "learning_rate": 6.088758185725654e-07, + "loss": 0.7999, + "step": 79 + }, + { + "epoch": 0.030592734225621414, + "grad_norm": 3.9392785793477474, + "learning_rate": 6.106286524709088e-07, + "loss": 0.7652, + "step": 80 + }, + { + "epoch": 0.030975143403441683, + "grad_norm": 4.318070062532729, + "learning_rate": 6.123597114715524e-07, + "loss": 0.6564, + "step": 81 + }, + { + "epoch": 0.03135755258126195, + "grad_norm": 3.853141414610464, + "learning_rate": 6.140695299481785e-07, + "loss": 0.6589, + "step": 82 + }, + { + "epoch": 0.03173996175908222, + "grad_norm": 3.454451500377018, + "learning_rate": 6.157586228417297e-07, + "loss": 0.6253, + "step": 83 + }, + { + "epoch": 0.03212237093690248, + "grad_norm": 3.244831082924636, + "learning_rate": 6.174274865913755e-07, + "loss": 0.2258, + "step": 84 + }, + { + "epoch": 0.032504780114722756, + "grad_norm": 4.9841997360937835, + "learning_rate": 6.190766000103886e-07, + "loss": 0.4724, + "step": 85 + }, + { + "epoch": 0.03288718929254302, + "grad_norm": 4.358109922986119, + "learning_rate": 6.207064251107957e-07, + "loss": 0.3161, + "step": 86 + }, + { + "epoch": 0.03326959847036329, + "grad_norm": 3.752613174388839, + "learning_rate": 6.223174078803584e-07, + "loss": 0.2185, + "step": 87 + }, + { + "epoch": 0.03365200764818355, + "grad_norm": 6.035992748966501, + "learning_rate": 6.239099790151521e-07, + "loss": 0.7716, + "step": 88 + }, + { + "epoch": 0.034034416826003826, + "grad_norm": 5.726135487788577, + "learning_rate": 6.254845546107528e-07, + "loss": 1.1187, + "step": 89 + }, + { + "epoch": 0.03441682600382409, + "grad_norm": 4.188566157565, + "learning_rate": 6.270415368148085e-07, + "loss": 0.6178, + "step": 90 + }, + { + "epoch": 0.03479923518164436, + "grad_norm": 1.8973622357764504, + "learning_rate": 6.285813144435542e-07, + "loss": 0.82, + "step": 91 + }, + { + "epoch": 0.03518164435946463, + "grad_norm": 3.203917465804253, + "learning_rate": 6.301042635646355e-07, + "loss": 0.7247, + "step": 92 + }, + { + "epoch": 0.035564053537284895, + "grad_norm": 4.069173173828255, + "learning_rate": 6.316107480484261e-07, + "loss": 0.9489, + "step": 93 + }, + { + "epoch": 0.03594646271510516, + "grad_norm": 5.255571224914232, + "learning_rate": 6.331011200898622e-07, + "loss": 0.3546, + "step": 94 + }, + { + "epoch": 0.036328871892925434, + "grad_norm": 2.9958642426559994, + "learning_rate": 6.345757207026644e-07, + "loss": 0.5467, + "step": 95 + }, + { + "epoch": 0.0367112810707457, + "grad_norm": 3.6217326204556106, + "learning_rate": 6.360348801876822e-07, + "loss": 0.2342, + "step": 96 + }, + { + "epoch": 0.037093690248565965, + "grad_norm": 4.380818231179398, + "learning_rate": 6.374789185769719e-07, + "loss": 0.2475, + "step": 97 + }, + { + "epoch": 0.03747609942638623, + "grad_norm": 3.792650297804044, + "learning_rate": 6.389081460550985e-07, + "loss": 0.3622, + "step": 98 + }, + { + "epoch": 0.0378585086042065, + "grad_norm": 4.318067191395842, + "learning_rate": 6.403228633590517e-07, + "loss": 0.2876, + "step": 99 + }, + { + "epoch": 0.03824091778202677, + "grad_norm": 4.3852408306664685, + "learning_rate": 6.417233621580648e-07, + "loss": 0.3416, + "step": 100 + }, + { + "epoch": 0.038623326959847035, + "grad_norm": 5.55577400837663, + "learning_rate": 6.431099254145344e-07, + "loss": 0.6162, + "step": 101 + }, + { + "epoch": 0.03900573613766731, + "grad_norm": 4.679812829276982, + "learning_rate": 6.44482827727162e-07, + "loss": 0.8162, + "step": 102 + }, + { + "epoch": 0.03938814531548757, + "grad_norm": 4.147621318751239, + "learning_rate": 6.458423356573556e-07, + "loss": 1.1291, + "step": 103 + }, + { + "epoch": 0.03977055449330784, + "grad_norm": 3.2625495994568445, + "learning_rate": 6.471887080398609e-07, + "loss": 0.6366, + "step": 104 + }, + { + "epoch": 0.040152963671128104, + "grad_norm": 3.2876321396179984, + "learning_rate": 6.485221962785314e-07, + "loss": 0.877, + "step": 105 + }, + { + "epoch": 0.04053537284894838, + "grad_norm": 3.51584336344417, + "learning_rate": 6.498430446280784e-07, + "loss": 0.4054, + "step": 106 + }, + { + "epoch": 0.04091778202676864, + "grad_norm": 3.4273604229777384, + "learning_rate": 6.511514904625936e-07, + "loss": 0.4965, + "step": 107 + }, + { + "epoch": 0.04130019120458891, + "grad_norm": 3.827336942561945, + "learning_rate": 6.52447764531582e-07, + "loss": 0.5168, + "step": 108 + }, + { + "epoch": 0.04168260038240918, + "grad_norm": 3.217148613937907, + "learning_rate": 6.537320912041942e-07, + "loss": 0.5542, + "step": 109 + }, + { + "epoch": 0.04206500956022945, + "grad_norm": 5.320029812779465, + "learning_rate": 6.55004688702308e-07, + "loss": 0.4252, + "step": 110 + }, + { + "epoch": 0.04244741873804971, + "grad_norm": 3.4351211639757118, + "learning_rate": 6.562657693230609e-07, + "loss": 0.3598, + "step": 111 + }, + { + "epoch": 0.04282982791586998, + "grad_norm": 4.734678353137203, + "learning_rate": 6.575155396514051e-07, + "loss": 0.3292, + "step": 112 + }, + { + "epoch": 0.04321223709369025, + "grad_norm": 6.986725618584486, + "learning_rate": 6.587542007632132e-07, + "loss": 0.8562, + "step": 113 + }, + { + "epoch": 0.043594646271510516, + "grad_norm": 4.767342479966072, + "learning_rate": 6.599819484194378e-07, + "loss": 0.9069, + "step": 114 + }, + { + "epoch": 0.04397705544933078, + "grad_norm": 5.708434092428296, + "learning_rate": 6.611989732517913e-07, + "loss": 0.7799, + "step": 115 + }, + { + "epoch": 0.044359464627151055, + "grad_norm": 2.143719184142411, + "learning_rate": 6.62405460940388e-07, + "loss": 0.762, + "step": 116 + }, + { + "epoch": 0.04474187380497132, + "grad_norm": 3.075753164708183, + "learning_rate": 6.636015923837607e-07, + "loss": 0.6046, + "step": 117 + }, + { + "epoch": 0.045124282982791586, + "grad_norm": 4.041382586271456, + "learning_rate": 6.647875438616435e-07, + "loss": 0.9243, + "step": 118 + }, + { + "epoch": 0.04550669216061185, + "grad_norm": 4.0781175167555865, + "learning_rate": 6.659634871908849e-07, + "loss": 0.6291, + "step": 119 + }, + { + "epoch": 0.045889101338432124, + "grad_norm": 2.947617658316482, + "learning_rate": 6.671295898748382e-07, + "loss": 0.6508, + "step": 120 + }, + { + "epoch": 0.04627151051625239, + "grad_norm": 3.7795709081852134, + "learning_rate": 6.682860152465512e-07, + "loss": 0.4124, + "step": 121 + }, + { + "epoch": 0.046653919694072656, + "grad_norm": 3.5289342257995586, + "learning_rate": 6.694329226060651e-07, + "loss": 0.4992, + "step": 122 + }, + { + "epoch": 0.04703632887189293, + "grad_norm": 4.706878684962685, + "learning_rate": 6.705704673521077e-07, + "loss": 0.2933, + "step": 123 + }, + { + "epoch": 0.047418738049713194, + "grad_norm": 5.110256294049169, + "learning_rate": 6.716988011084555e-07, + "loss": 0.4063, + "step": 124 + }, + { + "epoch": 0.04780114722753346, + "grad_norm": 4.520100734904757, + "learning_rate": 6.728180718452207e-07, + "loss": 0.4, + "step": 125 + }, + { + "epoch": 0.048183556405353725, + "grad_norm": 7.432639962375598, + "learning_rate": 6.739284239953048e-07, + "loss": 0.695, + "step": 126 + }, + { + "epoch": 0.048565965583174, + "grad_norm": 4.290670013747284, + "learning_rate": 6.750299985662519e-07, + "loss": 0.4686, + "step": 127 + }, + { + "epoch": 0.048948374760994263, + "grad_norm": 4.121158313697451, + "learning_rate": 6.761229332477118e-07, + "loss": 0.6365, + "step": 128 + }, + { + "epoch": 0.04933078393881453, + "grad_norm": 2.135314479957041, + "learning_rate": 6.772073625147248e-07, + "loss": 0.6846, + "step": 129 + }, + { + "epoch": 0.0497131931166348, + "grad_norm": 2.342096991950194, + "learning_rate": 6.782834177270168e-07, + "loss": 0.6111, + "step": 130 + }, + { + "epoch": 0.05009560229445507, + "grad_norm": 3.647894547310256, + "learning_rate": 6.793512272244895e-07, + "loss": 0.4072, + "step": 131 + }, + { + "epoch": 0.05047801147227533, + "grad_norm": 4.354125440971228, + "learning_rate": 6.804109164190813e-07, + "loss": 0.6969, + "step": 132 + }, + { + "epoch": 0.050860420650095606, + "grad_norm": 2.7346670812193903, + "learning_rate": 6.814626078831606e-07, + "loss": 0.5829, + "step": 133 + }, + { + "epoch": 0.05124282982791587, + "grad_norm": 4.094534063602115, + "learning_rate": 6.825064214346084e-07, + "loss": 0.3749, + "step": 134 + }, + { + "epoch": 0.05162523900573614, + "grad_norm": 3.382758089469319, + "learning_rate": 6.835424742187379e-07, + "loss": 0.3332, + "step": 135 + }, + { + "epoch": 0.0520076481835564, + "grad_norm": 2.683021984242027, + "learning_rate": 6.845708807871916e-07, + "loss": 0.1764, + "step": 136 + }, + { + "epoch": 0.052390057361376675, + "grad_norm": 4.509963555368485, + "learning_rate": 6.855917531739482e-07, + "loss": 0.2143, + "step": 137 + }, + { + "epoch": 0.05277246653919694, + "grad_norm": 5.256642816586635, + "learning_rate": 6.866052009685647e-07, + "loss": 0.4351, + "step": 138 + }, + { + "epoch": 0.05315487571701721, + "grad_norm": 5.834538834758634, + "learning_rate": 6.876113313867756e-07, + "loss": 0.6942, + "step": 139 + }, + { + "epoch": 0.05353728489483748, + "grad_norm": 3.014295690206189, + "learning_rate": 6.88610249338561e-07, + "loss": 0.4463, + "step": 140 + }, + { + "epoch": 0.053919694072657745, + "grad_norm": 2.2857408607164182, + "learning_rate": 6.896020574937915e-07, + "loss": 0.5019, + "step": 141 + }, + { + "epoch": 0.05430210325047801, + "grad_norm": 2.8317586412436886, + "learning_rate": 6.905868563455549e-07, + "loss": 0.551, + "step": 142 + }, + { + "epoch": 0.054684512428298276, + "grad_norm": 3.590807962694632, + "learning_rate": 6.9156474427126e-07, + "loss": 0.5001, + "step": 143 + }, + { + "epoch": 0.05506692160611855, + "grad_norm": 3.3447864387888737, + "learning_rate": 6.925358175916115e-07, + "loss": 0.6405, + "step": 144 + }, + { + "epoch": 0.055449330783938815, + "grad_norm": 3.5812648007285097, + "learning_rate": 6.935001706275438e-07, + "loss": 0.3969, + "step": 145 + }, + { + "epoch": 0.05583173996175908, + "grad_norm": 2.588371307068287, + "learning_rate": 6.944578957551989e-07, + "loss": 0.4425, + "step": 146 + }, + { + "epoch": 0.05621414913957935, + "grad_norm": 3.8032063258431816, + "learning_rate": 6.954090834590278e-07, + "loss": 0.5201, + "step": 147 + }, + { + "epoch": 0.05659655831739962, + "grad_norm": 3.972727066954228, + "learning_rate": 6.963538223830905e-07, + "loss": 0.4198, + "step": 148 + }, + { + "epoch": 0.056978967495219884, + "grad_norm": 4.727081825163079, + "learning_rate": 6.972921993806316e-07, + "loss": 0.3914, + "step": 149 + }, + { + "epoch": 0.05736137667304015, + "grad_norm": 3.8566009806911032, + "learning_rate": 6.982242995619941e-07, + "loss": 0.3418, + "step": 150 + }, + { + "epoch": 0.05774378585086042, + "grad_norm": 9.883895151296194, + "learning_rate": 6.991502063409444e-07, + "loss": 0.4184, + "step": 151 + }, + { + "epoch": 0.05812619502868069, + "grad_norm": 3.9589817252241994, + "learning_rate": 7.000700014794673e-07, + "loss": 0.9268, + "step": 152 + }, + { + "epoch": 0.058508604206500954, + "grad_norm": 3.6920673081202664, + "learning_rate": 7.009837651310913e-07, + "loss": 0.7311, + "step": 153 + }, + { + "epoch": 0.05889101338432123, + "grad_norm": 2.8818112834107232, + "learning_rate": 7.018915758828043e-07, + "loss": 0.507, + "step": 154 + }, + { + "epoch": 0.05927342256214149, + "grad_norm": 4.095078580332929, + "learning_rate": 7.027935107956114e-07, + "loss": 0.6913, + "step": 155 + }, + { + "epoch": 0.05965583173996176, + "grad_norm": 4.097583635557744, + "learning_rate": 7.036896454437902e-07, + "loss": 0.7102, + "step": 156 + }, + { + "epoch": 0.060038240917782024, + "grad_norm": 2.6535932830808338, + "learning_rate": 7.045800539528896e-07, + "loss": 0.2475, + "step": 157 + }, + { + "epoch": 0.060420650095602296, + "grad_norm": 2.724522743078367, + "learning_rate": 7.054648090365243e-07, + "loss": 0.4209, + "step": 158 + }, + { + "epoch": 0.06080305927342256, + "grad_norm": 3.0125584007079067, + "learning_rate": 7.063439820320076e-07, + "loss": 0.4966, + "step": 159 + }, + { + "epoch": 0.06118546845124283, + "grad_norm": 3.511451787422904, + "learning_rate": 7.072176429348676e-07, + "loss": 0.4125, + "step": 160 + }, + { + "epoch": 0.0615678776290631, + "grad_norm": 4.837683638209349, + "learning_rate": 7.080858604322876e-07, + "loss": 0.6034, + "step": 161 + }, + { + "epoch": 0.061950286806883366, + "grad_norm": 3.7945186383087077, + "learning_rate": 7.089487019355111e-07, + "loss": 0.3376, + "step": 162 + }, + { + "epoch": 0.06233269598470363, + "grad_norm": 7.647899074910034, + "learning_rate": 7.098062336112514e-07, + "loss": 0.6145, + "step": 163 + }, + { + "epoch": 0.0627151051625239, + "grad_norm": 6.282933251606341, + "learning_rate": 7.106585204121373e-07, + "loss": 1.0463, + "step": 164 + }, + { + "epoch": 0.06309751434034416, + "grad_norm": 3.282497350571109, + "learning_rate": 7.115056261062372e-07, + "loss": 0.5178, + "step": 165 + }, + { + "epoch": 0.06347992351816444, + "grad_norm": 2.1352796433069017, + "learning_rate": 7.123476133056884e-07, + "loss": 0.6417, + "step": 166 + }, + { + "epoch": 0.06386233269598471, + "grad_norm": 2.5930365994625575, + "learning_rate": 7.131845434944682e-07, + "loss": 0.4984, + "step": 167 + }, + { + "epoch": 0.06424474187380497, + "grad_norm": 3.0331021971899714, + "learning_rate": 7.140164770553345e-07, + "loss": 0.454, + "step": 168 + }, + { + "epoch": 0.06462715105162524, + "grad_norm": 3.7143123017921016, + "learning_rate": 7.148434732959689e-07, + "loss": 0.5334, + "step": 169 + }, + { + "epoch": 0.06500956022944551, + "grad_norm": 2.653427329658854, + "learning_rate": 7.156655904743474e-07, + "loss": 0.3391, + "step": 170 + }, + { + "epoch": 0.06539196940726577, + "grad_norm": 3.9509925367882985, + "learning_rate": 7.164828858233672e-07, + "loss": 0.3772, + "step": 171 + }, + { + "epoch": 0.06577437858508604, + "grad_norm": 3.1080221175458673, + "learning_rate": 7.172954155747545e-07, + "loss": 0.1742, + "step": 172 + }, + { + "epoch": 0.06615678776290632, + "grad_norm": 3.1603834827334754, + "learning_rate": 7.181032349822828e-07, + "loss": 0.2136, + "step": 173 + }, + { + "epoch": 0.06653919694072657, + "grad_norm": 3.798146729538179, + "learning_rate": 7.189063983443172e-07, + "loss": 0.1902, + "step": 174 + }, + { + "epoch": 0.06692160611854685, + "grad_norm": 4.068236540992522, + "learning_rate": 7.197049590257169e-07, + "loss": 0.3806, + "step": 175 + }, + { + "epoch": 0.0673040152963671, + "grad_norm": 6.4222483282794265, + "learning_rate": 7.204989694791108e-07, + "loss": 0.3978, + "step": 176 + }, + { + "epoch": 0.06768642447418738, + "grad_norm": 4.017960993317785, + "learning_rate": 7.212884812655726e-07, + "loss": 0.78, + "step": 177 + }, + { + "epoch": 0.06806883365200765, + "grad_norm": 4.100137168636647, + "learning_rate": 7.220735450747116e-07, + "loss": 0.7534, + "step": 178 + }, + { + "epoch": 0.06845124282982791, + "grad_norm": 3.1819592357927955, + "learning_rate": 7.228542107442019e-07, + "loss": 0.4387, + "step": 179 + }, + { + "epoch": 0.06883365200764818, + "grad_norm": 3.426584008354179, + "learning_rate": 7.236305272787674e-07, + "loss": 0.5703, + "step": 180 + }, + { + "epoch": 0.06921606118546846, + "grad_norm": 5.465640857252583, + "learning_rate": 7.244025428686407e-07, + "loss": 0.6831, + "step": 181 + }, + { + "epoch": 0.06959847036328871, + "grad_norm": 4.072806796700166, + "learning_rate": 7.25170304907513e-07, + "loss": 0.6685, + "step": 182 + }, + { + "epoch": 0.06998087954110899, + "grad_norm": 3.824467496019909, + "learning_rate": 7.259338600099944e-07, + "loss": 0.2409, + "step": 183 + }, + { + "epoch": 0.07036328871892926, + "grad_norm": 2.8699744709221924, + "learning_rate": 7.266932540285943e-07, + "loss": 0.2914, + "step": 184 + }, + { + "epoch": 0.07074569789674952, + "grad_norm": 3.6637638189913524, + "learning_rate": 7.274485320702463e-07, + "loss": 0.311, + "step": 185 + }, + { + "epoch": 0.07112810707456979, + "grad_norm": 4.0116479081805005, + "learning_rate": 7.281997385123849e-07, + "loss": 0.4676, + "step": 186 + }, + { + "epoch": 0.07151051625239006, + "grad_norm": 5.0535021280835455, + "learning_rate": 7.289469170185906e-07, + "loss": 0.2098, + "step": 187 + }, + { + "epoch": 0.07189292543021032, + "grad_norm": 5.6749955097466405, + "learning_rate": 7.29690110553821e-07, + "loss": 0.4254, + "step": 188 + }, + { + "epoch": 0.0722753346080306, + "grad_norm": 4.620593059176047, + "learning_rate": 7.304293613992342e-07, + "loss": 0.8025, + "step": 189 + }, + { + "epoch": 0.07265774378585087, + "grad_norm": 4.98679630219835, + "learning_rate": 7.311647111666231e-07, + "loss": 0.495, + "step": 190 + }, + { + "epoch": 0.07304015296367113, + "grad_norm": 3.0927893177777492, + "learning_rate": 7.318962008124717e-07, + "loss": 1.122, + "step": 191 + }, + { + "epoch": 0.0734225621414914, + "grad_norm": 3.338170753541131, + "learning_rate": 7.326238706516411e-07, + "loss": 0.6298, + "step": 192 + }, + { + "epoch": 0.07380497131931166, + "grad_norm": 3.7093543600843475, + "learning_rate": 7.333477603707038e-07, + "loss": 0.6464, + "step": 193 + }, + { + "epoch": 0.07418738049713193, + "grad_norm": 3.032196998144923, + "learning_rate": 7.340679090409308e-07, + "loss": 0.355, + "step": 194 + }, + { + "epoch": 0.0745697896749522, + "grad_norm": 2.8795885703469266, + "learning_rate": 7.34784355130946e-07, + "loss": 0.446, + "step": 195 + }, + { + "epoch": 0.07495219885277246, + "grad_norm": 2.5962306649778797, + "learning_rate": 7.354971365190572e-07, + "loss": 0.5733, + "step": 196 + }, + { + "epoch": 0.07533460803059273, + "grad_norm": 2.92991291450522, + "learning_rate": 7.36206290505272e-07, + "loss": 0.452, + "step": 197 + }, + { + "epoch": 0.075717017208413, + "grad_norm": 3.6905001057964744, + "learning_rate": 7.369118538230106e-07, + "loss": 0.4618, + "step": 198 + }, + { + "epoch": 0.07609942638623327, + "grad_norm": 3.9153593735892276, + "learning_rate": 7.376138626505238e-07, + "loss": 0.2443, + "step": 199 + }, + { + "epoch": 0.07648183556405354, + "grad_norm": 3.9085249723017563, + "learning_rate": 7.383123526220235e-07, + "loss": 0.3602, + "step": 200 + }, + { + "epoch": 0.07686424474187381, + "grad_norm": 7.705678891740636, + "learning_rate": 7.390073588385376e-07, + "loss": 0.5742, + "step": 201 + }, + { + "epoch": 0.07724665391969407, + "grad_norm": 4.761261490106676, + "learning_rate": 7.396989158784932e-07, + "loss": 0.4602, + "step": 202 + }, + { + "epoch": 0.07762906309751434, + "grad_norm": 3.166197708607704, + "learning_rate": 7.403870578080401e-07, + "loss": 0.3944, + "step": 203 + }, + { + "epoch": 0.07801147227533461, + "grad_norm": 2.0182306761377085, + "learning_rate": 7.410718181911208e-07, + "loss": 0.8064, + "step": 204 + }, + { + "epoch": 0.07839388145315487, + "grad_norm": 3.179968819135469, + "learning_rate": 7.417532300992931e-07, + "loss": 0.4025, + "step": 205 + }, + { + "epoch": 0.07877629063097515, + "grad_norm": 4.173314221331269, + "learning_rate": 7.424313261213144e-07, + "loss": 0.7782, + "step": 206 + }, + { + "epoch": 0.0791586998087954, + "grad_norm": 3.1765811377343054, + "learning_rate": 7.431061383724939e-07, + "loss": 0.6397, + "step": 207 + }, + { + "epoch": 0.07954110898661568, + "grad_norm": 4.6575972892850315, + "learning_rate": 7.437776985038198e-07, + "loss": 0.4831, + "step": 208 + }, + { + "epoch": 0.07992351816443595, + "grad_norm": 3.76169593304996, + "learning_rate": 7.444460377108663e-07, + "loss": 0.228, + "step": 209 + }, + { + "epoch": 0.08030592734225621, + "grad_norm": 3.9842518506933544, + "learning_rate": 7.451111867424903e-07, + "loss": 0.4997, + "step": 210 + }, + { + "epoch": 0.08068833652007648, + "grad_norm": 3.313081798203436, + "learning_rate": 7.457731759093186e-07, + "loss": 0.3388, + "step": 211 + }, + { + "epoch": 0.08107074569789675, + "grad_norm": 4.346696691194836, + "learning_rate": 7.464320350920372e-07, + "loss": 0.2379, + "step": 212 + }, + { + "epoch": 0.08145315487571701, + "grad_norm": 5.402567938113109, + "learning_rate": 7.470877937494842e-07, + "loss": 0.4411, + "step": 213 + }, + { + "epoch": 0.08183556405353729, + "grad_norm": 4.638690789435122, + "learning_rate": 7.477404809265524e-07, + "loss": 0.6799, + "step": 214 + }, + { + "epoch": 0.08221797323135756, + "grad_norm": 4.500483169047832, + "learning_rate": 7.483901252619104e-07, + "loss": 0.6008, + "step": 215 + }, + { + "epoch": 0.08260038240917782, + "grad_norm": 2.7526394575028728, + "learning_rate": 7.490367549955409e-07, + "loss": 0.4236, + "step": 216 + }, + { + "epoch": 0.08298279158699809, + "grad_norm": 2.745932171383271, + "learning_rate": 7.496803979761077e-07, + "loss": 0.2916, + "step": 217 + }, + { + "epoch": 0.08336520076481836, + "grad_norm": 3.1189073297420293, + "learning_rate": 7.50321081668153e-07, + "loss": 0.5905, + "step": 218 + }, + { + "epoch": 0.08374760994263862, + "grad_norm": 4.533831262357185, + "learning_rate": 7.509588331591282e-07, + "loss": 0.5721, + "step": 219 + }, + { + "epoch": 0.0841300191204589, + "grad_norm": 2.9649262548735305, + "learning_rate": 7.515936791662668e-07, + "loss": 0.3887, + "step": 220 + }, + { + "epoch": 0.08451242829827915, + "grad_norm": 3.4466325994562816, + "learning_rate": 7.522256460432995e-07, + "loss": 0.5528, + "step": 221 + }, + { + "epoch": 0.08489483747609942, + "grad_norm": 3.6307072928242223, + "learning_rate": 7.528547597870197e-07, + "loss": 0.5246, + "step": 222 + }, + { + "epoch": 0.0852772466539197, + "grad_norm": 3.257998381144178, + "learning_rate": 7.534810460436997e-07, + "loss": 0.2334, + "step": 223 + }, + { + "epoch": 0.08565965583173996, + "grad_norm": 4.644430635868781, + "learning_rate": 7.541045301153639e-07, + "loss": 0.2095, + "step": 224 + }, + { + "epoch": 0.08604206500956023, + "grad_norm": 4.69510308685718, + "learning_rate": 7.547252369659232e-07, + "loss": 0.3575, + "step": 225 + }, + { + "epoch": 0.0864244741873805, + "grad_norm": 6.285179920519442, + "learning_rate": 7.553431912271721e-07, + "loss": 0.6316, + "step": 226 + }, + { + "epoch": 0.08680688336520076, + "grad_norm": 4.806351337927113, + "learning_rate": 7.559584172046536e-07, + "loss": 0.6541, + "step": 227 + }, + { + "epoch": 0.08718929254302103, + "grad_norm": 3.001050606188733, + "learning_rate": 7.565709388833966e-07, + "loss": 0.7255, + "step": 228 + }, + { + "epoch": 0.0875717017208413, + "grad_norm": 1.9583878069937495, + "learning_rate": 7.571807799335257e-07, + "loss": 0.5502, + "step": 229 + }, + { + "epoch": 0.08795411089866156, + "grad_norm": 4.884711154296954, + "learning_rate": 7.577879637157502e-07, + "loss": 0.9388, + "step": 230 + }, + { + "epoch": 0.08833652007648184, + "grad_norm": 2.907146356646814, + "learning_rate": 7.583925132867335e-07, + "loss": 0.4075, + "step": 231 + }, + { + "epoch": 0.08871892925430211, + "grad_norm": 2.7039529050754956, + "learning_rate": 7.589944514043468e-07, + "loss": 0.3678, + "step": 232 + }, + { + "epoch": 0.08910133843212237, + "grad_norm": 5.727635219381453, + "learning_rate": 7.595938005328094e-07, + "loss": 0.5911, + "step": 233 + }, + { + "epoch": 0.08948374760994264, + "grad_norm": 4.664435648088752, + "learning_rate": 7.601905828477195e-07, + "loss": 0.2719, + "step": 234 + }, + { + "epoch": 0.08986615678776291, + "grad_norm": 3.5945420428990595, + "learning_rate": 7.607848202409769e-07, + "loss": 0.366, + "step": 235 + }, + { + "epoch": 0.09024856596558317, + "grad_norm": 4.234453826958854, + "learning_rate": 7.613765343256024e-07, + "loss": 0.2863, + "step": 236 + }, + { + "epoch": 0.09063097514340344, + "grad_norm": 7.158668076219786, + "learning_rate": 7.619657464404537e-07, + "loss": 0.1991, + "step": 237 + }, + { + "epoch": 0.0910133843212237, + "grad_norm": 7.641112766415768, + "learning_rate": 7.625524776548438e-07, + "loss": 0.5713, + "step": 238 + }, + { + "epoch": 0.09139579349904398, + "grad_norm": 4.9158103757560285, + "learning_rate": 7.631367487730614e-07, + "loss": 0.6582, + "step": 239 + }, + { + "epoch": 0.09177820267686425, + "grad_norm": 4.465435372685059, + "learning_rate": 7.63718580338797e-07, + "loss": 0.3755, + "step": 240 + }, + { + "epoch": 0.09216061185468451, + "grad_norm": 1.2563248360581727, + "learning_rate": 7.642979926394772e-07, + "loss": 0.8277, + "step": 241 + }, + { + "epoch": 0.09254302103250478, + "grad_norm": 2.627476840656689, + "learning_rate": 7.6487500571051e-07, + "loss": 0.4431, + "step": 242 + }, + { + "epoch": 0.09292543021032505, + "grad_norm": 3.4558729742159238, + "learning_rate": 7.654496393394404e-07, + "loss": 0.7704, + "step": 243 + }, + { + "epoch": 0.09330783938814531, + "grad_norm": 2.711048310843166, + "learning_rate": 7.660219130700239e-07, + "loss": 0.3752, + "step": 244 + }, + { + "epoch": 0.09369024856596558, + "grad_norm": 2.259893705518403, + "learning_rate": 7.665918462062132e-07, + "loss": 0.4386, + "step": 245 + }, + { + "epoch": 0.09407265774378586, + "grad_norm": 2.9605598309499297, + "learning_rate": 7.671594578160665e-07, + "loss": 0.356, + "step": 246 + }, + { + "epoch": 0.09445506692160612, + "grad_norm": 4.510125953285779, + "learning_rate": 7.677247667355753e-07, + "loss": 0.3476, + "step": 247 + }, + { + "epoch": 0.09483747609942639, + "grad_norm": 4.195827573275409, + "learning_rate": 7.682877915724144e-07, + "loss": 0.3094, + "step": 248 + }, + { + "epoch": 0.09521988527724666, + "grad_norm": 3.3973305488866457, + "learning_rate": 7.688485507096177e-07, + "loss": 0.1854, + "step": 249 + }, + { + "epoch": 0.09560229445506692, + "grad_norm": 6.474794374242375, + "learning_rate": 7.694070623091794e-07, + "loss": 0.3838, + "step": 250 + }, + { + "epoch": 0.09598470363288719, + "grad_norm": 5.565976203234317, + "learning_rate": 7.699633443155836e-07, + "loss": 0.4187, + "step": 251 + }, + { + "epoch": 0.09636711281070745, + "grad_norm": 4.445144730605, + "learning_rate": 7.705174144592636e-07, + "loss": 0.5566, + "step": 252 + }, + { + "epoch": 0.09674952198852772, + "grad_norm": 4.125839841484762, + "learning_rate": 7.710692902599934e-07, + "loss": 0.3737, + "step": 253 + }, + { + "epoch": 0.097131931166348, + "grad_norm": 6.788624212693753, + "learning_rate": 7.716189890302107e-07, + "loss": 0.6117, + "step": 254 + }, + { + "epoch": 0.09751434034416825, + "grad_norm": 3.3993700173101784, + "learning_rate": 7.721665278782767e-07, + "loss": 0.5299, + "step": 255 + }, + { + "epoch": 0.09789674952198853, + "grad_norm": 3.5447109527471192, + "learning_rate": 7.727119237116706e-07, + "loss": 0.464, + "step": 256 + }, + { + "epoch": 0.0982791586998088, + "grad_norm": 3.451375250445069, + "learning_rate": 7.732551932401223e-07, + "loss": 0.3774, + "step": 257 + }, + { + "epoch": 0.09866156787762906, + "grad_norm": 2.985844594806625, + "learning_rate": 7.737963529786837e-07, + "loss": 0.4458, + "step": 258 + }, + { + "epoch": 0.09904397705544933, + "grad_norm": 3.433412792310995, + "learning_rate": 7.743354192507426e-07, + "loss": 0.4216, + "step": 259 + }, + { + "epoch": 0.0994263862332696, + "grad_norm": 3.1338503823562065, + "learning_rate": 7.748724081909757e-07, + "loss": 0.3232, + "step": 260 + }, + { + "epoch": 0.09980879541108986, + "grad_norm": 4.060734526613366, + "learning_rate": 7.754073357482466e-07, + "loss": 0.4153, + "step": 261 + }, + { + "epoch": 0.10019120458891013, + "grad_norm": 3.565755460928457, + "learning_rate": 7.759402176884483e-07, + "loss": 0.2684, + "step": 262 + }, + { + "epoch": 0.10057361376673041, + "grad_norm": 5.585943816921691, + "learning_rate": 7.764710695972918e-07, + "loss": 0.5857, + "step": 263 + }, + { + "epoch": 0.10095602294455067, + "grad_norm": 4.169284659304579, + "learning_rate": 7.769999068830402e-07, + "loss": 1.0044, + "step": 264 + }, + { + "epoch": 0.10133843212237094, + "grad_norm": 4.71342197779094, + "learning_rate": 7.775267447791931e-07, + "loss": 0.7013, + "step": 265 + }, + { + "epoch": 0.10172084130019121, + "grad_norm": 1.434724367100525, + "learning_rate": 7.780515983471195e-07, + "loss": 0.6375, + "step": 266 + }, + { + "epoch": 0.10210325047801147, + "grad_norm": 3.895493490862681, + "learning_rate": 7.785744824786409e-07, + "loss": 0.4048, + "step": 267 + }, + { + "epoch": 0.10248565965583174, + "grad_norm": 3.8662095808203607, + "learning_rate": 7.790954118985672e-07, + "loss": 0.7465, + "step": 268 + }, + { + "epoch": 0.102868068833652, + "grad_norm": 3.424467036971547, + "learning_rate": 7.796144011671844e-07, + "loss": 0.408, + "step": 269 + }, + { + "epoch": 0.10325047801147227, + "grad_norm": 3.550724310255783, + "learning_rate": 7.801314646826968e-07, + "loss": 0.3537, + "step": 270 + }, + { + "epoch": 0.10363288718929255, + "grad_norm": 2.89688389518549, + "learning_rate": 7.80646616683623e-07, + "loss": 0.3399, + "step": 271 + }, + { + "epoch": 0.1040152963671128, + "grad_norm": 4.119613367105412, + "learning_rate": 7.811598712511505e-07, + "loss": 0.3759, + "step": 272 + }, + { + "epoch": 0.10439770554493308, + "grad_norm": 2.973462029424571, + "learning_rate": 7.816712423114423e-07, + "loss": 0.2155, + "step": 273 + }, + { + "epoch": 0.10478011472275335, + "grad_norm": 3.3977701167614947, + "learning_rate": 7.82180743637907e-07, + "loss": 0.2368, + "step": 274 + }, + { + "epoch": 0.10516252390057361, + "grad_norm": 4.181511767850576, + "learning_rate": 7.826883888534227e-07, + "loss": 0.2863, + "step": 275 + }, + { + "epoch": 0.10554493307839388, + "grad_norm": 5.954453848161291, + "learning_rate": 7.831941914325235e-07, + "loss": 0.622, + "step": 276 + }, + { + "epoch": 0.10592734225621415, + "grad_norm": 4.171188698084672, + "learning_rate": 7.836981647035458e-07, + "loss": 0.9351, + "step": 277 + }, + { + "epoch": 0.10630975143403441, + "grad_norm": 2.846374809467165, + "learning_rate": 7.842003218507345e-07, + "loss": 0.6559, + "step": 278 + }, + { + "epoch": 0.10669216061185469, + "grad_norm": 2.124035560310232, + "learning_rate": 7.847006759163141e-07, + "loss": 0.5626, + "step": 279 + }, + { + "epoch": 0.10707456978967496, + "grad_norm": 3.5632323466003974, + "learning_rate": 7.851992398025198e-07, + "loss": 0.3027, + "step": 280 + }, + { + "epoch": 0.10745697896749522, + "grad_norm": 2.80981512133499, + "learning_rate": 7.856960262735946e-07, + "loss": 0.4338, + "step": 281 + }, + { + "epoch": 0.10783938814531549, + "grad_norm": 3.1493700033255347, + "learning_rate": 7.861910479577502e-07, + "loss": 0.3888, + "step": 282 + }, + { + "epoch": 0.10822179732313575, + "grad_norm": 3.199839325685757, + "learning_rate": 7.866843173490923e-07, + "loss": 0.581, + "step": 283 + }, + { + "epoch": 0.10860420650095602, + "grad_norm": 3.9744033797586047, + "learning_rate": 7.871758468095137e-07, + "loss": 0.3301, + "step": 284 + }, + { + "epoch": 0.1089866156787763, + "grad_norm": 4.082176563651574, + "learning_rate": 7.876656485705525e-07, + "loss": 0.2453, + "step": 285 + }, + { + "epoch": 0.10936902485659655, + "grad_norm": 3.762870572619933, + "learning_rate": 7.881537347352188e-07, + "loss": 0.1672, + "step": 286 + }, + { + "epoch": 0.10975143403441683, + "grad_norm": 4.497364119915088, + "learning_rate": 7.886401172797894e-07, + "loss": 0.1229, + "step": 287 + }, + { + "epoch": 0.1101338432122371, + "grad_norm": 6.455440373723637, + "learning_rate": 7.891248080555703e-07, + "loss": 0.2936, + "step": 288 + }, + { + "epoch": 0.11051625239005736, + "grad_norm": 3.5754464464259237, + "learning_rate": 7.896078187906302e-07, + "loss": 0.7403, + "step": 289 + }, + { + "epoch": 0.11089866156787763, + "grad_norm": 2.764753991163198, + "learning_rate": 7.900891610915027e-07, + "loss": 0.5118, + "step": 290 + }, + { + "epoch": 0.1112810707456979, + "grad_norm": 2.297678385254701, + "learning_rate": 7.9056884644486e-07, + "loss": 0.4936, + "step": 291 + }, + { + "epoch": 0.11166347992351816, + "grad_norm": 1.8158157537320423, + "learning_rate": 7.910468862191579e-07, + "loss": 0.7025, + "step": 292 + }, + { + "epoch": 0.11204588910133843, + "grad_norm": 3.552000535165901, + "learning_rate": 7.915232916662518e-07, + "loss": 0.706, + "step": 293 + }, + { + "epoch": 0.1124282982791587, + "grad_norm": 3.6530113548438585, + "learning_rate": 7.919980739229866e-07, + "loss": 0.3169, + "step": 294 + }, + { + "epoch": 0.11281070745697896, + "grad_norm": 3.2632719624729427, + "learning_rate": 7.924712440127582e-07, + "loss": 0.3265, + "step": 295 + }, + { + "epoch": 0.11319311663479924, + "grad_norm": 3.110373025338036, + "learning_rate": 7.929428128470493e-07, + "loss": 0.3102, + "step": 296 + }, + { + "epoch": 0.1135755258126195, + "grad_norm": 3.6801332499387422, + "learning_rate": 7.9341279122694e-07, + "loss": 0.423, + "step": 297 + }, + { + "epoch": 0.11395793499043977, + "grad_norm": 3.8900937862907234, + "learning_rate": 7.938811898445904e-07, + "loss": 0.2004, + "step": 298 + }, + { + "epoch": 0.11434034416826004, + "grad_norm": 3.7375127785760855, + "learning_rate": 7.943480192847023e-07, + "loss": 0.1989, + "step": 299 + }, + { + "epoch": 0.1147227533460803, + "grad_norm": 4.461046741971845, + "learning_rate": 7.948132900259528e-07, + "loss": 0.3909, + "step": 300 + }, + { + "epoch": 0.11510516252390057, + "grad_norm": 6.173622057914207, + "learning_rate": 7.952770124424067e-07, + "loss": 0.3894, + "step": 301 + }, + { + "epoch": 0.11548757170172085, + "grad_norm": 3.733296769394523, + "learning_rate": 7.957391968049033e-07, + "loss": 0.8673, + "step": 302 + }, + { + "epoch": 0.1158699808795411, + "grad_norm": 3.9489473123211307, + "learning_rate": 7.961998532824224e-07, + "loss": 0.6133, + "step": 303 + }, + { + "epoch": 0.11625239005736138, + "grad_norm": 4.475476091515576, + "learning_rate": 7.966589919434262e-07, + "loss": 0.7168, + "step": 304 + }, + { + "epoch": 0.11663479923518165, + "grad_norm": 2.3809739320778744, + "learning_rate": 7.971166227571798e-07, + "loss": 0.5665, + "step": 305 + }, + { + "epoch": 0.11701720841300191, + "grad_norm": 3.904736156855016, + "learning_rate": 7.975727555950501e-07, + "loss": 0.4922, + "step": 306 + }, + { + "epoch": 0.11739961759082218, + "grad_norm": 3.4877422125598634, + "learning_rate": 7.980274002317836e-07, + "loss": 0.3931, + "step": 307 + }, + { + "epoch": 0.11778202676864245, + "grad_norm": 3.089792822648413, + "learning_rate": 7.98480566346763e-07, + "loss": 0.3242, + "step": 308 + }, + { + "epoch": 0.11816443594646271, + "grad_norm": 3.1274358792520442, + "learning_rate": 7.989322635252437e-07, + "loss": 0.2631, + "step": 309 + }, + { + "epoch": 0.11854684512428298, + "grad_norm": 3.306093606693292, + "learning_rate": 7.993825012595703e-07, + "loss": 0.1887, + "step": 310 + }, + { + "epoch": 0.11892925430210326, + "grad_norm": 5.434712364074534, + "learning_rate": 7.998312889503738e-07, + "loss": 0.2773, + "step": 311 + }, + { + "epoch": 0.11931166347992352, + "grad_norm": 4.773654664339484, + "learning_rate": 8.00278635907749e-07, + "loss": 0.1587, + "step": 312 + }, + { + "epoch": 0.11969407265774379, + "grad_norm": 5.61964161283346, + "learning_rate": 8.007245513524136e-07, + "loss": 0.4011, + "step": 313 + }, + { + "epoch": 0.12007648183556405, + "grad_norm": 5.422191951403171, + "learning_rate": 8.011690444168484e-07, + "loss": 0.5186, + "step": 314 + }, + { + "epoch": 0.12045889101338432, + "grad_norm": 4.235650240037841, + "learning_rate": 8.016121241464195e-07, + "loss": 0.6452, + "step": 315 + }, + { + "epoch": 0.12084130019120459, + "grad_norm": 2.2959913772859717, + "learning_rate": 8.020537995004831e-07, + "loss": 0.3568, + "step": 316 + }, + { + "epoch": 0.12122370936902485, + "grad_norm": 2.7906570791555816, + "learning_rate": 8.024940793534721e-07, + "loss": 0.5597, + "step": 317 + }, + { + "epoch": 0.12160611854684512, + "grad_norm": 3.0943336321631882, + "learning_rate": 8.029329724959665e-07, + "loss": 0.5461, + "step": 318 + }, + { + "epoch": 0.1219885277246654, + "grad_norm": 3.4053237075530216, + "learning_rate": 8.033704876357459e-07, + "loss": 0.4284, + "step": 319 + }, + { + "epoch": 0.12237093690248566, + "grad_norm": 3.234592596281024, + "learning_rate": 8.038066333988265e-07, + "loss": 0.6019, + "step": 320 + }, + { + "epoch": 0.12275334608030593, + "grad_norm": 3.4472462596835673, + "learning_rate": 8.042414183304817e-07, + "loss": 0.5452, + "step": 321 + }, + { + "epoch": 0.1231357552581262, + "grad_norm": 3.6241358022865984, + "learning_rate": 8.046748508962464e-07, + "loss": 0.3125, + "step": 322 + }, + { + "epoch": 0.12351816443594646, + "grad_norm": 3.893943260136077, + "learning_rate": 8.051069394829059e-07, + "loss": 0.2509, + "step": 323 + }, + { + "epoch": 0.12390057361376673, + "grad_norm": 4.8189353026547135, + "learning_rate": 8.0553769239947e-07, + "loss": 0.1644, + "step": 324 + }, + { + "epoch": 0.124282982791587, + "grad_norm": 3.559991991055513, + "learning_rate": 8.059671178781316e-07, + "loss": 0.3537, + "step": 325 + }, + { + "epoch": 0.12466539196940726, + "grad_norm": 6.39953827384509, + "learning_rate": 8.063952240752102e-07, + "loss": 0.5631, + "step": 326 + }, + { + "epoch": 0.12504780114722752, + "grad_norm": 3.6130294464626025, + "learning_rate": 8.068220190720823e-07, + "loss": 0.7063, + "step": 327 + }, + { + "epoch": 0.1254302103250478, + "grad_norm": 3.1183942022317663, + "learning_rate": 8.072475108760961e-07, + "loss": 0.7305, + "step": 328 + }, + { + "epoch": 0.12581261950286807, + "grad_norm": 2.745672983244802, + "learning_rate": 8.076717074214732e-07, + "loss": 0.4498, + "step": 329 + }, + { + "epoch": 0.12619502868068833, + "grad_norm": 2.8516755382133048, + "learning_rate": 8.08094616570196e-07, + "loss": 0.356, + "step": 330 + }, + { + "epoch": 0.1265774378585086, + "grad_norm": 3.0513196560458202, + "learning_rate": 8.085162461128826e-07, + "loss": 0.2682, + "step": 331 + }, + { + "epoch": 0.12695984703632887, + "grad_norm": 4.121776361564982, + "learning_rate": 8.089366037696473e-07, + "loss": 0.4183, + "step": 332 + }, + { + "epoch": 0.12734225621414913, + "grad_norm": 2.9887883173753043, + "learning_rate": 8.093556971909491e-07, + "loss": 0.3487, + "step": 333 + }, + { + "epoch": 0.12772466539196942, + "grad_norm": 3.216439019351985, + "learning_rate": 8.09773533958427e-07, + "loss": 0.3038, + "step": 334 + }, + { + "epoch": 0.12810707456978968, + "grad_norm": 4.0139201660580275, + "learning_rate": 8.101901215857231e-07, + "loss": 0.2941, + "step": 335 + }, + { + "epoch": 0.12848948374760993, + "grad_norm": 3.350848924483089, + "learning_rate": 8.106054675192933e-07, + "loss": 0.175, + "step": 336 + }, + { + "epoch": 0.12887189292543022, + "grad_norm": 3.438037899587354, + "learning_rate": 8.110195791392056e-07, + "loss": 0.1396, + "step": 337 + }, + { + "epoch": 0.12925430210325048, + "grad_norm": 6.772774599013537, + "learning_rate": 8.114324637599278e-07, + "loss": 0.5966, + "step": 338 + }, + { + "epoch": 0.12963671128107074, + "grad_norm": 4.137346920627515, + "learning_rate": 8.118441286311013e-07, + "loss": 0.4748, + "step": 339 + }, + { + "epoch": 0.13001912045889102, + "grad_norm": 3.4568218620380566, + "learning_rate": 8.122545809383064e-07, + "loss": 0.5865, + "step": 340 + }, + { + "epoch": 0.13040152963671128, + "grad_norm": 3.0236478825988216, + "learning_rate": 8.126638278038134e-07, + "loss": 0.4805, + "step": 341 + }, + { + "epoch": 0.13078393881453154, + "grad_norm": 2.559726796175156, + "learning_rate": 8.130718762873258e-07, + "loss": 0.3147, + "step": 342 + }, + { + "epoch": 0.13116634799235183, + "grad_norm": 4.449152087749615, + "learning_rate": 8.134787333867094e-07, + "loss": 0.9307, + "step": 343 + }, + { + "epoch": 0.1315487571701721, + "grad_norm": 2.6546079160044753, + "learning_rate": 8.138844060387134e-07, + "loss": 0.2215, + "step": 344 + }, + { + "epoch": 0.13193116634799235, + "grad_norm": 2.47569283700409, + "learning_rate": 8.142889011196796e-07, + "loss": 0.4161, + "step": 345 + }, + { + "epoch": 0.13231357552581263, + "grad_norm": 4.2666821409978, + "learning_rate": 8.146922254462416e-07, + "loss": 0.2177, + "step": 346 + }, + { + "epoch": 0.1326959847036329, + "grad_norm": 3.236865425515548, + "learning_rate": 8.150943857760151e-07, + "loss": 0.2703, + "step": 347 + }, + { + "epoch": 0.13307839388145315, + "grad_norm": 4.03876947339817, + "learning_rate": 8.154953888082761e-07, + "loss": 0.4093, + "step": 348 + }, + { + "epoch": 0.13346080305927344, + "grad_norm": 5.638943831119226, + "learning_rate": 8.158952411846307e-07, + "loss": 0.3396, + "step": 349 + }, + { + "epoch": 0.1338432122370937, + "grad_norm": 4.114233821397829, + "learning_rate": 8.162939494896757e-07, + "loss": 0.346, + "step": 350 + }, + { + "epoch": 0.13422562141491395, + "grad_norm": 7.785144000084442, + "learning_rate": 8.166915202516487e-07, + "loss": 0.4865, + "step": 351 + }, + { + "epoch": 0.1346080305927342, + "grad_norm": 4.286640643865958, + "learning_rate": 8.170879599430697e-07, + "loss": 0.6353, + "step": 352 + }, + { + "epoch": 0.1349904397705545, + "grad_norm": 3.460942184410628, + "learning_rate": 8.174832749813731e-07, + "loss": 0.5312, + "step": 353 + }, + { + "epoch": 0.13537284894837476, + "grad_norm": 2.148699158427092, + "learning_rate": 8.178774717295314e-07, + "loss": 0.3499, + "step": 354 + }, + { + "epoch": 0.13575525812619502, + "grad_norm": 2.5779189518750436, + "learning_rate": 8.182705564966695e-07, + "loss": 0.3294, + "step": 355 + }, + { + "epoch": 0.1361376673040153, + "grad_norm": 3.7108314900565365, + "learning_rate": 8.186625355386705e-07, + "loss": 0.5441, + "step": 356 + }, + { + "epoch": 0.13652007648183556, + "grad_norm": 2.8954350222722463, + "learning_rate": 8.19053415058773e-07, + "loss": 0.4693, + "step": 357 + }, + { + "epoch": 0.13690248565965582, + "grad_norm": 3.8289599695272587, + "learning_rate": 8.194432012081608e-07, + "loss": 0.3562, + "step": 358 + }, + { + "epoch": 0.1372848948374761, + "grad_norm": 4.809977873753277, + "learning_rate": 8.198319000865427e-07, + "loss": 0.4452, + "step": 359 + }, + { + "epoch": 0.13766730401529637, + "grad_norm": 3.428323322356324, + "learning_rate": 8.202195177427262e-07, + "loss": 0.3554, + "step": 360 + }, + { + "epoch": 0.13804971319311662, + "grad_norm": 6.486314304317555, + "learning_rate": 8.206060601751816e-07, + "loss": 0.1975, + "step": 361 + }, + { + "epoch": 0.1384321223709369, + "grad_norm": 3.8071659483116043, + "learning_rate": 8.209915333325995e-07, + "loss": 0.2016, + "step": 362 + }, + { + "epoch": 0.13881453154875717, + "grad_norm": 7.472006881743372, + "learning_rate": 8.213759431144391e-07, + "loss": 0.4247, + "step": 363 + }, + { + "epoch": 0.13919694072657743, + "grad_norm": 4.879357371307577, + "learning_rate": 8.21759295371472e-07, + "loss": 0.6671, + "step": 364 + }, + { + "epoch": 0.13957934990439771, + "grad_norm": 5.342405808242789, + "learning_rate": 8.221415959063138e-07, + "loss": 0.5423, + "step": 365 + }, + { + "epoch": 0.13996175908221797, + "grad_norm": 4.174753631815001, + "learning_rate": 8.225228504739531e-07, + "loss": 0.269, + "step": 366 + }, + { + "epoch": 0.14034416826003823, + "grad_norm": 3.108287678673902, + "learning_rate": 8.229030647822711e-07, + "loss": 0.6411, + "step": 367 + }, + { + "epoch": 0.14072657743785852, + "grad_norm": 3.4124323184948455, + "learning_rate": 8.232822444925531e-07, + "loss": 0.4184, + "step": 368 + }, + { + "epoch": 0.14110898661567878, + "grad_norm": 3.5585165674944963, + "learning_rate": 8.236603952199958e-07, + "loss": 0.6019, + "step": 369 + }, + { + "epoch": 0.14149139579349904, + "grad_norm": 4.025666621193951, + "learning_rate": 8.240375225342052e-07, + "loss": 0.4895, + "step": 370 + }, + { + "epoch": 0.14187380497131932, + "grad_norm": 4.571471082115062, + "learning_rate": 8.244136319596894e-07, + "loss": 0.2911, + "step": 371 + }, + { + "epoch": 0.14225621414913958, + "grad_norm": 2.5329872104037254, + "learning_rate": 8.247887289763437e-07, + "loss": 0.205, + "step": 372 + }, + { + "epoch": 0.14263862332695984, + "grad_norm": 4.108219114705737, + "learning_rate": 8.251628190199301e-07, + "loss": 0.2088, + "step": 373 + }, + { + "epoch": 0.14302103250478013, + "grad_norm": 3.533705541211625, + "learning_rate": 8.255359074825496e-07, + "loss": 0.2682, + "step": 374 + }, + { + "epoch": 0.14340344168260039, + "grad_norm": 4.903078840243899, + "learning_rate": 8.259079997131087e-07, + "loss": 0.2899, + "step": 375 + }, + { + "epoch": 0.14378585086042064, + "grad_norm": 7.337460165874226, + "learning_rate": 8.262791010177799e-07, + "loss": 0.5831, + "step": 376 + }, + { + "epoch": 0.14416826003824093, + "grad_norm": 3.68396977986088, + "learning_rate": 8.266492166604547e-07, + "loss": 0.7252, + "step": 377 + }, + { + "epoch": 0.1445506692160612, + "grad_norm": 3.7188438596020146, + "learning_rate": 8.27018351863193e-07, + "loss": 0.473, + "step": 378 + }, + { + "epoch": 0.14493307839388145, + "grad_norm": 1.4870233300950009, + "learning_rate": 8.273865118066637e-07, + "loss": 0.5603, + "step": 379 + }, + { + "epoch": 0.14531548757170173, + "grad_norm": 2.2572929022034094, + "learning_rate": 8.27753701630582e-07, + "loss": 0.3373, + "step": 380 + }, + { + "epoch": 0.145697896749522, + "grad_norm": 3.0990770504221294, + "learning_rate": 8.2811992643414e-07, + "loss": 0.4191, + "step": 381 + }, + { + "epoch": 0.14608030592734225, + "grad_norm": 4.983702993982142, + "learning_rate": 8.284851912764304e-07, + "loss": 0.5062, + "step": 382 + }, + { + "epoch": 0.1464627151051625, + "grad_norm": 2.7700035650182153, + "learning_rate": 8.288495011768676e-07, + "loss": 0.3522, + "step": 383 + }, + { + "epoch": 0.1468451242829828, + "grad_norm": 2.503238963986765, + "learning_rate": 8.292128611156e-07, + "loss": 0.2989, + "step": 384 + }, + { + "epoch": 0.14722753346080306, + "grad_norm": 4.035654426834888, + "learning_rate": 8.29575276033919e-07, + "loss": 0.2093, + "step": 385 + }, + { + "epoch": 0.14760994263862331, + "grad_norm": 3.433500008719308, + "learning_rate": 8.299367508346627e-07, + "loss": 0.1824, + "step": 386 + }, + { + "epoch": 0.1479923518164436, + "grad_norm": 4.621602021590747, + "learning_rate": 8.302972903826131e-07, + "loss": 0.1898, + "step": 387 + }, + { + "epoch": 0.14837476099426386, + "grad_norm": 7.1558155444335645, + "learning_rate": 8.306568995048896e-07, + "loss": 0.5661, + "step": 388 + }, + { + "epoch": 0.14875717017208412, + "grad_norm": 4.818200438514181, + "learning_rate": 8.310155829913362e-07, + "loss": 0.7258, + "step": 389 + }, + { + "epoch": 0.1491395793499044, + "grad_norm": 4.729636663309238, + "learning_rate": 8.313733455949049e-07, + "loss": 0.7956, + "step": 390 + }, + { + "epoch": 0.14952198852772466, + "grad_norm": 4.556495373140328, + "learning_rate": 8.317301920320328e-07, + "loss": 0.4644, + "step": 391 + }, + { + "epoch": 0.14990439770554492, + "grad_norm": 2.829353529349588, + "learning_rate": 8.32086126983016e-07, + "loss": 0.3368, + "step": 392 + }, + { + "epoch": 0.1502868068833652, + "grad_norm": 2.855775662691332, + "learning_rate": 8.324411550923776e-07, + "loss": 0.445, + "step": 393 + }, + { + "epoch": 0.15066921606118547, + "grad_norm": 3.1955485734729185, + "learning_rate": 8.327952809692308e-07, + "loss": 0.4805, + "step": 394 + }, + { + "epoch": 0.15105162523900573, + "grad_norm": 2.628568675797958, + "learning_rate": 8.33148509187639e-07, + "loss": 0.3455, + "step": 395 + }, + { + "epoch": 0.151434034416826, + "grad_norm": 4.427946741019882, + "learning_rate": 8.335008442869694e-07, + "loss": 0.2197, + "step": 396 + }, + { + "epoch": 0.15181644359464627, + "grad_norm": 4.976054454603633, + "learning_rate": 8.338522907722437e-07, + "loss": 0.569, + "step": 397 + }, + { + "epoch": 0.15219885277246653, + "grad_norm": 3.0160754564089474, + "learning_rate": 8.342028531144826e-07, + "loss": 0.3384, + "step": 398 + }, + { + "epoch": 0.15258126195028682, + "grad_norm": 3.001475765915238, + "learning_rate": 8.345525357510488e-07, + "loss": 0.2957, + "step": 399 + }, + { + "epoch": 0.15296367112810708, + "grad_norm": 4.44835313476485, + "learning_rate": 8.349013430859824e-07, + "loss": 0.3145, + "step": 400 + }, + { + "epoch": 0.15334608030592733, + "grad_norm": 5.9358016948221834, + "learning_rate": 8.352492794903348e-07, + "loss": 0.5199, + "step": 401 + }, + { + "epoch": 0.15372848948374762, + "grad_norm": 3.8760502530236685, + "learning_rate": 8.355963493024964e-07, + "loss": 0.7258, + "step": 402 + }, + { + "epoch": 0.15411089866156788, + "grad_norm": 2.653734339821923, + "learning_rate": 8.359425568285223e-07, + "loss": 0.467, + "step": 403 + }, + { + "epoch": 0.15449330783938814, + "grad_norm": 3.2408897652245, + "learning_rate": 8.36287906342452e-07, + "loss": 0.5655, + "step": 404 + }, + { + "epoch": 0.15487571701720843, + "grad_norm": 5.064740097442185, + "learning_rate": 8.366324020866259e-07, + "loss": 0.4984, + "step": 405 + }, + { + "epoch": 0.15525812619502868, + "grad_norm": 4.151100582660372, + "learning_rate": 8.369760482719989e-07, + "loss": 0.3415, + "step": 406 + }, + { + "epoch": 0.15564053537284894, + "grad_norm": 2.1285019236890412, + "learning_rate": 8.373188490784484e-07, + "loss": 0.5143, + "step": 407 + }, + { + "epoch": 0.15602294455066923, + "grad_norm": 3.038817564991084, + "learning_rate": 8.376608086550797e-07, + "loss": 0.3227, + "step": 408 + }, + { + "epoch": 0.1564053537284895, + "grad_norm": 3.10377849908374, + "learning_rate": 8.380019311205272e-07, + "loss": 0.2325, + "step": 409 + }, + { + "epoch": 0.15678776290630975, + "grad_norm": 2.9242790246530648, + "learning_rate": 8.38342220563252e-07, + "loss": 0.2834, + "step": 410 + }, + { + "epoch": 0.15717017208413, + "grad_norm": 4.360945223969221, + "learning_rate": 8.386816810418363e-07, + "loss": 0.1797, + "step": 411 + }, + { + "epoch": 0.1575525812619503, + "grad_norm": 4.194239644376321, + "learning_rate": 8.390203165852732e-07, + "loss": 0.1658, + "step": 412 + }, + { + "epoch": 0.15793499043977055, + "grad_norm": 5.705579924332777, + "learning_rate": 8.393581311932545e-07, + "loss": 0.3086, + "step": 413 + }, + { + "epoch": 0.1583173996175908, + "grad_norm": 5.0412789608184765, + "learning_rate": 8.396951288364528e-07, + "loss": 0.5928, + "step": 414 + }, + { + "epoch": 0.1586998087954111, + "grad_norm": 4.067400983790442, + "learning_rate": 8.400313134568032e-07, + "loss": 0.4792, + "step": 415 + }, + { + "epoch": 0.15908221797323135, + "grad_norm": 2.325920546220163, + "learning_rate": 8.403666889677787e-07, + "loss": 0.3938, + "step": 416 + }, + { + "epoch": 0.1594646271510516, + "grad_norm": 2.529085434725153, + "learning_rate": 8.407012592546638e-07, + "loss": 0.4017, + "step": 417 + }, + { + "epoch": 0.1598470363288719, + "grad_norm": 4.022784667151783, + "learning_rate": 8.410350281748251e-07, + "loss": 0.4667, + "step": 418 + }, + { + "epoch": 0.16022944550669216, + "grad_norm": 5.081691581053613, + "learning_rate": 8.413679995579778e-07, + "loss": 0.3841, + "step": 419 + }, + { + "epoch": 0.16061185468451242, + "grad_norm": 2.8149274734721326, + "learning_rate": 8.417001772064491e-07, + "loss": 0.1482, + "step": 420 + }, + { + "epoch": 0.1609942638623327, + "grad_norm": 2.5876473942299856, + "learning_rate": 8.420315648954389e-07, + "loss": 0.219, + "step": 421 + }, + { + "epoch": 0.16137667304015296, + "grad_norm": 4.121041450349419, + "learning_rate": 8.423621663732774e-07, + "loss": 0.3829, + "step": 422 + }, + { + "epoch": 0.16175908221797322, + "grad_norm": 3.8963933260039, + "learning_rate": 8.426919853616795e-07, + "loss": 0.1564, + "step": 423 + }, + { + "epoch": 0.1621414913957935, + "grad_norm": 4.824550058208553, + "learning_rate": 8.430210255559961e-07, + "loss": 0.3041, + "step": 424 + }, + { + "epoch": 0.16252390057361377, + "grad_norm": 4.454444020661066, + "learning_rate": 8.433492906254623e-07, + "loss": 0.2036, + "step": 425 + }, + { + "epoch": 0.16290630975143403, + "grad_norm": 8.36002388690657, + "learning_rate": 8.43676784213443e-07, + "loss": 0.5876, + "step": 426 + }, + { + "epoch": 0.1632887189292543, + "grad_norm": 3.929175887732475, + "learning_rate": 8.440035099376761e-07, + "loss": 0.4755, + "step": 427 + }, + { + "epoch": 0.16367112810707457, + "grad_norm": 3.6463606680107525, + "learning_rate": 8.443294713905113e-07, + "loss": 0.5675, + "step": 428 + }, + { + "epoch": 0.16405353728489483, + "grad_norm": 3.6404536784059127, + "learning_rate": 8.446546721391481e-07, + "loss": 0.4856, + "step": 429 + }, + { + "epoch": 0.16443594646271512, + "grad_norm": 2.6217796917621112, + "learning_rate": 8.449791157258692e-07, + "loss": 0.6915, + "step": 430 + }, + { + "epoch": 0.16481835564053537, + "grad_norm": 3.42014551221246, + "learning_rate": 8.453028056682724e-07, + "loss": 0.627, + "step": 431 + }, + { + "epoch": 0.16520076481835563, + "grad_norm": 4.183283262725723, + "learning_rate": 8.456257454594997e-07, + "loss": 0.2805, + "step": 432 + }, + { + "epoch": 0.16558317399617592, + "grad_norm": 3.93192174657836, + "learning_rate": 8.459479385684625e-07, + "loss": 0.3485, + "step": 433 + }, + { + "epoch": 0.16596558317399618, + "grad_norm": 2.567881817531444, + "learning_rate": 8.462693884400665e-07, + "loss": 0.2407, + "step": 434 + }, + { + "epoch": 0.16634799235181644, + "grad_norm": 3.8513236755237674, + "learning_rate": 8.46590098495432e-07, + "loss": 0.1689, + "step": 435 + }, + { + "epoch": 0.16673040152963672, + "grad_norm": 5.43608618585279, + "learning_rate": 8.469100721321118e-07, + "loss": 0.3796, + "step": 436 + }, + { + "epoch": 0.16711281070745698, + "grad_norm": 3.427005492527403, + "learning_rate": 8.472293127243085e-07, + "loss": 0.1437, + "step": 437 + }, + { + "epoch": 0.16749521988527724, + "grad_norm": 7.971012527304747, + "learning_rate": 8.47547823623087e-07, + "loss": 0.4998, + "step": 438 + }, + { + "epoch": 0.16787762906309753, + "grad_norm": 4.308472231329761, + "learning_rate": 8.478656081565858e-07, + "loss": 0.6339, + "step": 439 + }, + { + "epoch": 0.1682600382409178, + "grad_norm": 4.791877255442337, + "learning_rate": 8.481826696302256e-07, + "loss": 0.6164, + "step": 440 + }, + { + "epoch": 0.16864244741873805, + "grad_norm": 3.8886846276787375, + "learning_rate": 8.484990113269158e-07, + "loss": 0.7065, + "step": 441 + }, + { + "epoch": 0.1690248565965583, + "grad_norm": 2.3271527404361185, + "learning_rate": 8.488146365072584e-07, + "loss": 0.4662, + "step": 442 + }, + { + "epoch": 0.1694072657743786, + "grad_norm": 2.7275549713665117, + "learning_rate": 8.491295484097495e-07, + "loss": 0.6759, + "step": 443 + }, + { + "epoch": 0.16978967495219885, + "grad_norm": 3.7458820621318125, + "learning_rate": 8.494437502509786e-07, + "loss": 0.4375, + "step": 444 + }, + { + "epoch": 0.1701720841300191, + "grad_norm": 4.34678554813099, + "learning_rate": 8.497572452258264e-07, + "loss": 0.3278, + "step": 445 + }, + { + "epoch": 0.1705544933078394, + "grad_norm": 3.6011246355051214, + "learning_rate": 8.500700365076586e-07, + "loss": 0.4652, + "step": 446 + }, + { + "epoch": 0.17093690248565965, + "grad_norm": 3.3932647976452524, + "learning_rate": 8.503821272485197e-07, + "loss": 0.2305, + "step": 447 + }, + { + "epoch": 0.1713193116634799, + "grad_norm": 3.7425874564043857, + "learning_rate": 8.506935205793227e-07, + "loss": 0.2994, + "step": 448 + }, + { + "epoch": 0.1717017208413002, + "grad_norm": 2.981305526120341, + "learning_rate": 8.510042196100389e-07, + "loss": 0.1297, + "step": 449 + }, + { + "epoch": 0.17208413001912046, + "grad_norm": 4.476331561724208, + "learning_rate": 8.513142274298821e-07, + "loss": 0.2766, + "step": 450 + }, + { + "epoch": 0.17246653919694072, + "grad_norm": 4.608373086952153, + "learning_rate": 8.516235471074952e-07, + "loss": 0.3614, + "step": 451 + }, + { + "epoch": 0.172848948374761, + "grad_norm": 4.827351233443686, + "learning_rate": 8.519321816911309e-07, + "loss": 0.5439, + "step": 452 + }, + { + "epoch": 0.17323135755258126, + "grad_norm": 2.511614617878112, + "learning_rate": 8.522401342088326e-07, + "loss": 0.5802, + "step": 453 + }, + { + "epoch": 0.17361376673040152, + "grad_norm": 3.2540966037790082, + "learning_rate": 8.525474076686125e-07, + "loss": 0.4176, + "step": 454 + }, + { + "epoch": 0.1739961759082218, + "grad_norm": 3.8871122994851324, + "learning_rate": 8.528540050586278e-07, + "loss": 0.6278, + "step": 455 + }, + { + "epoch": 0.17437858508604206, + "grad_norm": 4.048438517645086, + "learning_rate": 8.531599293473555e-07, + "loss": 0.3956, + "step": 456 + }, + { + "epoch": 0.17476099426386232, + "grad_norm": 3.11043051435314, + "learning_rate": 8.53465183483764e-07, + "loss": 0.2714, + "step": 457 + }, + { + "epoch": 0.1751434034416826, + "grad_norm": 4.921696929132415, + "learning_rate": 8.537697703974845e-07, + "loss": 0.5876, + "step": 458 + }, + { + "epoch": 0.17552581261950287, + "grad_norm": 4.903323521676653, + "learning_rate": 8.540736929989794e-07, + "loss": 0.4939, + "step": 459 + }, + { + "epoch": 0.17590822179732313, + "grad_norm": 4.8318538155200494, + "learning_rate": 8.54376954179709e-07, + "loss": 0.5429, + "step": 460 + }, + { + "epoch": 0.17629063097514341, + "grad_norm": 4.261696596026859, + "learning_rate": 8.546795568122967e-07, + "loss": 0.293, + "step": 461 + }, + { + "epoch": 0.17667304015296367, + "grad_norm": 4.632046460895395, + "learning_rate": 8.549815037506923e-07, + "loss": 0.2401, + "step": 462 + }, + { + "epoch": 0.17705544933078393, + "grad_norm": 8.216589582080363, + "learning_rate": 8.552827978303335e-07, + "loss": 0.6042, + "step": 463 + }, + { + "epoch": 0.17743785850860422, + "grad_norm": 4.647916363907478, + "learning_rate": 8.555834418683056e-07, + "loss": 0.7894, + "step": 464 + }, + { + "epoch": 0.17782026768642448, + "grad_norm": 2.2044464909614954, + "learning_rate": 8.558834386634994e-07, + "loss": 0.4387, + "step": 465 + }, + { + "epoch": 0.17820267686424474, + "grad_norm": 1.7303373058677423, + "learning_rate": 8.561827909967682e-07, + "loss": 0.4711, + "step": 466 + }, + { + "epoch": 0.17858508604206502, + "grad_norm": 2.3381052834893703, + "learning_rate": 8.564815016310815e-07, + "loss": 0.5498, + "step": 467 + }, + { + "epoch": 0.17896749521988528, + "grad_norm": 2.9010284441789858, + "learning_rate": 8.567795733116784e-07, + "loss": 0.2958, + "step": 468 + }, + { + "epoch": 0.17934990439770554, + "grad_norm": 2.861270463902799, + "learning_rate": 8.570770087662194e-07, + "loss": 0.3461, + "step": 469 + }, + { + "epoch": 0.17973231357552583, + "grad_norm": 2.3472607576215343, + "learning_rate": 8.573738107049358e-07, + "loss": 0.4536, + "step": 470 + }, + { + "epoch": 0.18011472275334608, + "grad_norm": 2.679119261168484, + "learning_rate": 8.576699818207777e-07, + "loss": 0.2337, + "step": 471 + }, + { + "epoch": 0.18049713193116634, + "grad_norm": 4.8104618264787, + "learning_rate": 8.579655247895611e-07, + "loss": 0.6163, + "step": 472 + }, + { + "epoch": 0.1808795411089866, + "grad_norm": 3.7416879722707956, + "learning_rate": 8.582604422701124e-07, + "loss": 0.3372, + "step": 473 + }, + { + "epoch": 0.1812619502868069, + "grad_norm": 3.6382333651998358, + "learning_rate": 8.585547369044125e-07, + "loss": 0.1598, + "step": 474 + }, + { + "epoch": 0.18164435946462715, + "grad_norm": 4.037690451650499, + "learning_rate": 8.588484113177379e-07, + "loss": 0.2873, + "step": 475 + }, + { + "epoch": 0.1820267686424474, + "grad_norm": 8.535887858375467, + "learning_rate": 8.591414681188027e-07, + "loss": 0.7283, + "step": 476 + }, + { + "epoch": 0.1824091778202677, + "grad_norm": 4.210883371999634, + "learning_rate": 8.594339098998958e-07, + "loss": 0.5389, + "step": 477 + }, + { + "epoch": 0.18279158699808795, + "grad_norm": 3.572301398296823, + "learning_rate": 8.597257392370203e-07, + "loss": 0.7289, + "step": 478 + }, + { + "epoch": 0.1831739961759082, + "grad_norm": 3.401284754868903, + "learning_rate": 8.60016958690028e-07, + "loss": 0.4667, + "step": 479 + }, + { + "epoch": 0.1835564053537285, + "grad_norm": 3.90239067163983, + "learning_rate": 8.603075708027559e-07, + "loss": 0.4377, + "step": 480 + }, + { + "epoch": 0.18393881453154876, + "grad_norm": 3.4348493586884308, + "learning_rate": 8.605975781031574e-07, + "loss": 0.4699, + "step": 481 + }, + { + "epoch": 0.18432122370936901, + "grad_norm": 2.926084046935673, + "learning_rate": 8.608869831034361e-07, + "loss": 0.3857, + "step": 482 + }, + { + "epoch": 0.1847036328871893, + "grad_norm": 3.147097621949093, + "learning_rate": 8.611757883001755e-07, + "loss": 0.2076, + "step": 483 + }, + { + "epoch": 0.18508604206500956, + "grad_norm": 2.9988211858901903, + "learning_rate": 8.614639961744689e-07, + "loss": 0.4364, + "step": 484 + }, + { + "epoch": 0.18546845124282982, + "grad_norm": 3.9578398681369062, + "learning_rate": 8.617516091920455e-07, + "loss": 0.2187, + "step": 485 + }, + { + "epoch": 0.1858508604206501, + "grad_norm": 2.580414743577251, + "learning_rate": 8.620386298033992e-07, + "loss": 0.1354, + "step": 486 + }, + { + "epoch": 0.18623326959847036, + "grad_norm": 5.465514528391503, + "learning_rate": 8.623250604439132e-07, + "loss": 0.1598, + "step": 487 + }, + { + "epoch": 0.18661567877629062, + "grad_norm": 5.688162628810543, + "learning_rate": 8.626109035339828e-07, + "loss": 0.5749, + "step": 488 + }, + { + "epoch": 0.1869980879541109, + "grad_norm": 4.06459812857032, + "learning_rate": 8.628961614791395e-07, + "loss": 0.4578, + "step": 489 + }, + { + "epoch": 0.18738049713193117, + "grad_norm": 4.059401753890258, + "learning_rate": 8.631808366701719e-07, + "loss": 0.5734, + "step": 490 + }, + { + "epoch": 0.18776290630975143, + "grad_norm": 3.6007972947017914, + "learning_rate": 8.634649314832464e-07, + "loss": 0.4735, + "step": 491 + }, + { + "epoch": 0.1881453154875717, + "grad_norm": 3.44509769384225, + "learning_rate": 8.637484482800254e-07, + "loss": 0.3236, + "step": 492 + }, + { + "epoch": 0.18852772466539197, + "grad_norm": 2.710251339310638, + "learning_rate": 8.640313894077854e-07, + "loss": 0.6083, + "step": 493 + }, + { + "epoch": 0.18891013384321223, + "grad_norm": 3.7640070941627126, + "learning_rate": 8.643137571995341e-07, + "loss": 0.4299, + "step": 494 + }, + { + "epoch": 0.18929254302103252, + "grad_norm": 11.060630511062472, + "learning_rate": 8.645955539741253e-07, + "loss": 0.2374, + "step": 495 + }, + { + "epoch": 0.18967495219885278, + "grad_norm": 2.9658908929801346, + "learning_rate": 8.648767820363733e-07, + "loss": 0.269, + "step": 496 + }, + { + "epoch": 0.19005736137667303, + "grad_norm": 4.718886215568088, + "learning_rate": 8.651574436771658e-07, + "loss": 0.3392, + "step": 497 + }, + { + "epoch": 0.19043977055449332, + "grad_norm": 4.07698996230184, + "learning_rate": 8.654375411735766e-07, + "loss": 0.2047, + "step": 498 + }, + { + "epoch": 0.19082217973231358, + "grad_norm": 4.322621506262301, + "learning_rate": 8.657170767889753e-07, + "loss": 0.1593, + "step": 499 + }, + { + "epoch": 0.19120458891013384, + "grad_norm": 3.6781333022238636, + "learning_rate": 8.659960527731382e-07, + "loss": 0.3298, + "step": 500 + }, + { + "epoch": 0.19120458891013384, + "eval_runtime": 822.3073, + "eval_samples_per_second": 1.865, + "eval_steps_per_second": 0.467, + "step": 500 + }, + { + "epoch": 0.19158699808795412, + "grad_norm": 5.341980758375264, + "learning_rate": 8.662744713623563e-07, + "loss": 0.5194, + "step": 501 + }, + { + "epoch": 0.19196940726577438, + "grad_norm": 3.8851356847921803, + "learning_rate": 8.665523347795424e-07, + "loss": 0.4366, + "step": 502 + }, + { + "epoch": 0.19235181644359464, + "grad_norm": 4.447133098733561, + "learning_rate": 8.668296452343392e-07, + "loss": 0.5872, + "step": 503 + }, + { + "epoch": 0.1927342256214149, + "grad_norm": 4.259308438470198, + "learning_rate": 8.671064049232225e-07, + "loss": 0.3353, + "step": 504 + }, + { + "epoch": 0.1931166347992352, + "grad_norm": 2.5481644625084, + "learning_rate": 8.673826160296078e-07, + "loss": 0.3876, + "step": 505 + }, + { + "epoch": 0.19349904397705545, + "grad_norm": 3.8753543814767766, + "learning_rate": 8.676582807239522e-07, + "loss": 0.6591, + "step": 506 + }, + { + "epoch": 0.1938814531548757, + "grad_norm": 2.9565461540677513, + "learning_rate": 8.679334011638571e-07, + "loss": 0.5829, + "step": 507 + }, + { + "epoch": 0.194263862332696, + "grad_norm": 2.8580497630694173, + "learning_rate": 8.682079794941694e-07, + "loss": 0.2155, + "step": 508 + }, + { + "epoch": 0.19464627151051625, + "grad_norm": 3.7126409535991756, + "learning_rate": 8.684820178470827e-07, + "loss": 0.3184, + "step": 509 + }, + { + "epoch": 0.1950286806883365, + "grad_norm": 3.961171280622527, + "learning_rate": 8.687555183422355e-07, + "loss": 0.368, + "step": 510 + }, + { + "epoch": 0.1954110898661568, + "grad_norm": 3.1431792391414968, + "learning_rate": 8.6902848308681e-07, + "loss": 0.1698, + "step": 511 + }, + { + "epoch": 0.19579349904397705, + "grad_norm": 3.4654473055387554, + "learning_rate": 8.693009141756294e-07, + "loss": 0.0964, + "step": 512 + }, + { + "epoch": 0.1961759082217973, + "grad_norm": 5.839838524245153, + "learning_rate": 8.695728136912551e-07, + "loss": 0.457, + "step": 513 + }, + { + "epoch": 0.1965583173996176, + "grad_norm": 4.7069374423354695, + "learning_rate": 8.698441837040811e-07, + "loss": 0.7124, + "step": 514 + }, + { + "epoch": 0.19694072657743786, + "grad_norm": 4.365320888404452, + "learning_rate": 8.701150262724292e-07, + "loss": 0.6243, + "step": 515 + }, + { + "epoch": 0.19732313575525812, + "grad_norm": 3.289324172386687, + "learning_rate": 8.703853434426426e-07, + "loss": 0.5249, + "step": 516 + }, + { + "epoch": 0.1977055449330784, + "grad_norm": 2.3283226390129723, + "learning_rate": 8.70655137249179e-07, + "loss": 0.5161, + "step": 517 + }, + { + "epoch": 0.19808795411089866, + "grad_norm": 2.5905756081726192, + "learning_rate": 8.709244097147014e-07, + "loss": 0.4175, + "step": 518 + }, + { + "epoch": 0.19847036328871892, + "grad_norm": 3.424691212788987, + "learning_rate": 8.711931628501709e-07, + "loss": 0.4771, + "step": 519 + }, + { + "epoch": 0.1988527724665392, + "grad_norm": 3.3766078651131006, + "learning_rate": 8.714613986549345e-07, + "loss": 0.4121, + "step": 520 + }, + { + "epoch": 0.19923518164435947, + "grad_norm": 3.7758399786743695, + "learning_rate": 8.717291191168164e-07, + "loss": 0.4434, + "step": 521 + }, + { + "epoch": 0.19961759082217972, + "grad_norm": 4.172649868787247, + "learning_rate": 8.719963262122055e-07, + "loss": 0.471, + "step": 522 + }, + { + "epoch": 0.2, + "grad_norm": 3.204457671034851, + "learning_rate": 8.722630219061423e-07, + "loss": 0.1419, + "step": 523 + }, + { + "epoch": 0.20038240917782027, + "grad_norm": 3.486618619725143, + "learning_rate": 8.725292081524071e-07, + "loss": 0.1515, + "step": 524 + }, + { + "epoch": 0.20076481835564053, + "grad_norm": 4.252158121924402, + "learning_rate": 8.72794886893605e-07, + "loss": 0.2632, + "step": 525 + }, + { + "epoch": 0.20114722753346082, + "grad_norm": 8.937775720998328, + "learning_rate": 8.730600600612507e-07, + "loss": 0.4893, + "step": 526 + }, + { + "epoch": 0.20152963671128107, + "grad_norm": 4.2345114823195935, + "learning_rate": 8.73324729575853e-07, + "loss": 0.527, + "step": 527 + }, + { + "epoch": 0.20191204588910133, + "grad_norm": 3.6419135829805787, + "learning_rate": 8.73588897346999e-07, + "loss": 0.6547, + "step": 528 + }, + { + "epoch": 0.20229445506692162, + "grad_norm": 1.8780114468031686, + "learning_rate": 8.738525652734356e-07, + "loss": 0.563, + "step": 529 + }, + { + "epoch": 0.20267686424474188, + "grad_norm": 3.003413959935241, + "learning_rate": 8.74115735243152e-07, + "loss": 0.6466, + "step": 530 + }, + { + "epoch": 0.20305927342256214, + "grad_norm": 3.2604255716773025, + "learning_rate": 8.743784091334607e-07, + "loss": 0.4341, + "step": 531 + }, + { + "epoch": 0.20344168260038242, + "grad_norm": 2.983473237094229, + "learning_rate": 8.746405888110783e-07, + "loss": 0.3732, + "step": 532 + }, + { + "epoch": 0.20382409177820268, + "grad_norm": 4.1953099994900604, + "learning_rate": 8.74902276132204e-07, + "loss": 0.4144, + "step": 533 + }, + { + "epoch": 0.20420650095602294, + "grad_norm": 3.45385513526699, + "learning_rate": 8.751634729425997e-07, + "loss": 0.1678, + "step": 534 + }, + { + "epoch": 0.2045889101338432, + "grad_norm": 3.767437257402485, + "learning_rate": 8.754241810776672e-07, + "loss": 0.2739, + "step": 535 + }, + { + "epoch": 0.20497131931166349, + "grad_norm": 4.0418594681414985, + "learning_rate": 8.756844023625261e-07, + "loss": 0.1897, + "step": 536 + }, + { + "epoch": 0.20535372848948374, + "grad_norm": 4.170464094943381, + "learning_rate": 8.7594413861209e-07, + "loss": 0.0853, + "step": 537 + }, + { + "epoch": 0.205736137667304, + "grad_norm": 7.11126053507429, + "learning_rate": 8.762033916311432e-07, + "loss": 0.2841, + "step": 538 + }, + { + "epoch": 0.2061185468451243, + "grad_norm": 8.038682400361811, + "learning_rate": 8.764621632144152e-07, + "loss": 0.5757, + "step": 539 + }, + { + "epoch": 0.20650095602294455, + "grad_norm": 3.8845616167680714, + "learning_rate": 8.767204551466555e-07, + "loss": 0.5021, + "step": 540 + }, + { + "epoch": 0.2068833652007648, + "grad_norm": 3.457770765270759, + "learning_rate": 8.769782692027076e-07, + "loss": 0.4517, + "step": 541 + }, + { + "epoch": 0.2072657743785851, + "grad_norm": 2.1231287402641406, + "learning_rate": 8.77235607147582e-07, + "loss": 0.406, + "step": 542 + }, + { + "epoch": 0.20764818355640535, + "grad_norm": 3.664290236042986, + "learning_rate": 8.774924707365287e-07, + "loss": 0.384, + "step": 543 + }, + { + "epoch": 0.2080305927342256, + "grad_norm": 3.8846193097131776, + "learning_rate": 8.777488617151092e-07, + "loss": 0.5264, + "step": 544 + }, + { + "epoch": 0.2084130019120459, + "grad_norm": 3.8864727613347774, + "learning_rate": 8.780047818192677e-07, + "loss": 0.3449, + "step": 545 + }, + { + "epoch": 0.20879541108986616, + "grad_norm": 3.074034475695364, + "learning_rate": 8.782602327754011e-07, + "loss": 0.4267, + "step": 546 + }, + { + "epoch": 0.20917782026768642, + "grad_norm": 4.650051420194106, + "learning_rate": 8.785152163004298e-07, + "loss": 0.2878, + "step": 547 + }, + { + "epoch": 0.2095602294455067, + "grad_norm": 3.9610133417757614, + "learning_rate": 8.787697341018658e-07, + "loss": 0.2085, + "step": 548 + }, + { + "epoch": 0.20994263862332696, + "grad_norm": 4.055132009142522, + "learning_rate": 8.790237878778824e-07, + "loss": 0.1718, + "step": 549 + }, + { + "epoch": 0.21032504780114722, + "grad_norm": 5.561179357908723, + "learning_rate": 8.792773793173814e-07, + "loss": 0.2292, + "step": 550 + }, + { + "epoch": 0.2107074569789675, + "grad_norm": 9.823343493565627, + "learning_rate": 8.795305101000612e-07, + "loss": 0.4747, + "step": 551 + }, + { + "epoch": 0.21108986615678776, + "grad_norm": 4.320240537516627, + "learning_rate": 8.797831818964823e-07, + "loss": 0.6607, + "step": 552 + }, + { + "epoch": 0.21147227533460802, + "grad_norm": 3.8821286495378406, + "learning_rate": 8.800353963681352e-07, + "loss": 0.6069, + "step": 553 + }, + { + "epoch": 0.2118546845124283, + "grad_norm": 3.456264535828526, + "learning_rate": 8.802871551675046e-07, + "loss": 0.5072, + "step": 554 + }, + { + "epoch": 0.21223709369024857, + "grad_norm": 4.004620506032858, + "learning_rate": 8.805384599381345e-07, + "loss": 0.3927, + "step": 555 + }, + { + "epoch": 0.21261950286806883, + "grad_norm": 3.7279097642824355, + "learning_rate": 8.807893123146934e-07, + "loss": 0.5367, + "step": 556 + }, + { + "epoch": 0.2130019120458891, + "grad_norm": 2.8182120069861316, + "learning_rate": 8.810397139230372e-07, + "loss": 0.4391, + "step": 557 + }, + { + "epoch": 0.21338432122370937, + "grad_norm": 2.7487648877915536, + "learning_rate": 8.812896663802729e-07, + "loss": 0.2057, + "step": 558 + }, + { + "epoch": 0.21376673040152963, + "grad_norm": 4.265964592351681, + "learning_rate": 8.815391712948212e-07, + "loss": 0.3774, + "step": 559 + }, + { + "epoch": 0.21414913957934992, + "grad_norm": 4.349966307761718, + "learning_rate": 8.817882302664786e-07, + "loss": 0.2399, + "step": 560 + }, + { + "epoch": 0.21453154875717018, + "grad_norm": 4.632009813011725, + "learning_rate": 8.820368448864787e-07, + "loss": 0.1874, + "step": 561 + }, + { + "epoch": 0.21491395793499043, + "grad_norm": 5.107100700943744, + "learning_rate": 8.822850167375535e-07, + "loss": 0.1995, + "step": 562 + }, + { + "epoch": 0.2152963671128107, + "grad_norm": 9.613688373896391, + "learning_rate": 8.82532747393994e-07, + "loss": 0.5337, + "step": 563 + }, + { + "epoch": 0.21567877629063098, + "grad_norm": 4.014300294858346, + "learning_rate": 8.827800384217091e-07, + "loss": 0.4894, + "step": 564 + }, + { + "epoch": 0.21606118546845124, + "grad_norm": 3.7285606417528623, + "learning_rate": 8.830268913782868e-07, + "loss": 0.564, + "step": 565 + }, + { + "epoch": 0.2164435946462715, + "grad_norm": 4.170282066923075, + "learning_rate": 8.832733078130512e-07, + "loss": 0.7278, + "step": 566 + }, + { + "epoch": 0.21682600382409178, + "grad_norm": 2.822889339600972, + "learning_rate": 8.835192892671222e-07, + "loss": 0.5958, + "step": 567 + }, + { + "epoch": 0.21720841300191204, + "grad_norm": 2.933753526097939, + "learning_rate": 8.837648372734725e-07, + "loss": 0.3823, + "step": 568 + }, + { + "epoch": 0.2175908221797323, + "grad_norm": 3.8063780614296845, + "learning_rate": 8.840099533569853e-07, + "loss": 0.4028, + "step": 569 + }, + { + "epoch": 0.2179732313575526, + "grad_norm": 5.715698094888678, + "learning_rate": 8.842546390345113e-07, + "loss": 0.4984, + "step": 570 + }, + { + "epoch": 0.21835564053537285, + "grad_norm": 3.3179041127519064, + "learning_rate": 8.844988958149243e-07, + "loss": 0.4208, + "step": 571 + }, + { + "epoch": 0.2187380497131931, + "grad_norm": 3.13872732648784, + "learning_rate": 8.847427251991777e-07, + "loss": 0.2319, + "step": 572 + }, + { + "epoch": 0.2191204588910134, + "grad_norm": 2.9904045191122965, + "learning_rate": 8.849861286803598e-07, + "loss": 0.3062, + "step": 573 + }, + { + "epoch": 0.21950286806883365, + "grad_norm": 4.134629771253638, + "learning_rate": 8.852291077437482e-07, + "loss": 0.1593, + "step": 574 + }, + { + "epoch": 0.2198852772466539, + "grad_norm": 5.5330836818095595, + "learning_rate": 8.854716638668649e-07, + "loss": 0.3453, + "step": 575 + }, + { + "epoch": 0.2202676864244742, + "grad_norm": 6.35525536721151, + "learning_rate": 8.857137985195292e-07, + "loss": 0.3971, + "step": 576 + }, + { + "epoch": 0.22065009560229445, + "grad_norm": 3.7363729200055866, + "learning_rate": 8.859555131639122e-07, + "loss": 0.4838, + "step": 577 + }, + { + "epoch": 0.2210325047801147, + "grad_norm": 2.884765790294934, + "learning_rate": 8.86196809254589e-07, + "loss": 0.3447, + "step": 578 + }, + { + "epoch": 0.221414913957935, + "grad_norm": 2.6882944041880656, + "learning_rate": 8.864376882385918e-07, + "loss": 0.6815, + "step": 579 + }, + { + "epoch": 0.22179732313575526, + "grad_norm": 2.6886467372301643, + "learning_rate": 8.866781515554614e-07, + "loss": 0.3703, + "step": 580 + }, + { + "epoch": 0.22217973231357552, + "grad_norm": 2.136208203008963, + "learning_rate": 8.869182006372994e-07, + "loss": 0.4558, + "step": 581 + }, + { + "epoch": 0.2225621414913958, + "grad_norm": 2.9403023845844443, + "learning_rate": 8.871578369088188e-07, + "loss": 0.2963, + "step": 582 + }, + { + "epoch": 0.22294455066921606, + "grad_norm": 2.875306486296151, + "learning_rate": 8.873970617873952e-07, + "loss": 0.4421, + "step": 583 + }, + { + "epoch": 0.22332695984703632, + "grad_norm": 4.129625905530537, + "learning_rate": 8.876358766831167e-07, + "loss": 0.2216, + "step": 584 + }, + { + "epoch": 0.2237093690248566, + "grad_norm": 3.3319302148918393, + "learning_rate": 8.878742829988343e-07, + "loss": 0.2468, + "step": 585 + }, + { + "epoch": 0.22409177820267687, + "grad_norm": 4.4162868600360365, + "learning_rate": 8.881122821302105e-07, + "loss": 0.1727, + "step": 586 + }, + { + "epoch": 0.22447418738049713, + "grad_norm": 3.5349965433538175, + "learning_rate": 8.883498754657699e-07, + "loss": 0.2095, + "step": 587 + }, + { + "epoch": 0.2248565965583174, + "grad_norm": 7.335889399719171, + "learning_rate": 8.885870643869454e-07, + "loss": 0.5801, + "step": 588 + }, + { + "epoch": 0.22523900573613767, + "grad_norm": 3.8911290589093004, + "learning_rate": 8.888238502681287e-07, + "loss": 0.4556, + "step": 589 + }, + { + "epoch": 0.22562141491395793, + "grad_norm": 2.9425051638704556, + "learning_rate": 8.890602344767169e-07, + "loss": 0.5179, + "step": 590 + }, + { + "epoch": 0.22600382409177822, + "grad_norm": 2.409984514495357, + "learning_rate": 8.892962183731602e-07, + "loss": 0.821, + "step": 591 + }, + { + "epoch": 0.22638623326959847, + "grad_norm": 4.32976382521085, + "learning_rate": 8.895318033110082e-07, + "loss": 0.6991, + "step": 592 + }, + { + "epoch": 0.22676864244741873, + "grad_norm": 6.219767271637266, + "learning_rate": 8.897669906369579e-07, + "loss": 0.5052, + "step": 593 + }, + { + "epoch": 0.227151051625239, + "grad_norm": 2.847185009555525, + "learning_rate": 8.900017816908987e-07, + "loss": 0.3373, + "step": 594 + }, + { + "epoch": 0.22753346080305928, + "grad_norm": 2.5647512641752654, + "learning_rate": 8.902361778059585e-07, + "loss": 0.3058, + "step": 595 + }, + { + "epoch": 0.22791586998087954, + "grad_norm": 3.2259829834349127, + "learning_rate": 8.904701803085492e-07, + "loss": 0.4217, + "step": 596 + }, + { + "epoch": 0.2282982791586998, + "grad_norm": 3.240345774202062, + "learning_rate": 8.907037905184118e-07, + "loss": 0.288, + "step": 597 + }, + { + "epoch": 0.22868068833652008, + "grad_norm": 2.8117471323067367, + "learning_rate": 8.90937009748661e-07, + "loss": 0.3434, + "step": 598 + }, + { + "epoch": 0.22906309751434034, + "grad_norm": 3.409382351064056, + "learning_rate": 8.911698393058294e-07, + "loss": 0.1568, + "step": 599 + }, + { + "epoch": 0.2294455066921606, + "grad_norm": 5.025906586355392, + "learning_rate": 8.914022804899116e-07, + "loss": 0.2689, + "step": 600 + }, + { + "epoch": 0.2298279158699809, + "grad_norm": 5.246780134732783, + "learning_rate": 8.916343345944075e-07, + "loss": 0.5604, + "step": 601 + }, + { + "epoch": 0.23021032504780115, + "grad_norm": 4.418125895193714, + "learning_rate": 8.918660029063655e-07, + "loss": 0.6224, + "step": 602 + }, + { + "epoch": 0.2305927342256214, + "grad_norm": 10.199199301015543, + "learning_rate": 8.920972867064258e-07, + "loss": 0.3815, + "step": 603 + }, + { + "epoch": 0.2309751434034417, + "grad_norm": 2.503913486607132, + "learning_rate": 8.923281872688621e-07, + "loss": 0.241, + "step": 604 + }, + { + "epoch": 0.23135755258126195, + "grad_norm": 2.652775185823977, + "learning_rate": 8.925587058616247e-07, + "loss": 0.4143, + "step": 605 + }, + { + "epoch": 0.2317399617590822, + "grad_norm": 2.9848565317305704, + "learning_rate": 8.927888437463812e-07, + "loss": 0.6514, + "step": 606 + }, + { + "epoch": 0.2321223709369025, + "grad_norm": 4.13808400622241, + "learning_rate": 8.930186021785589e-07, + "loss": 0.5333, + "step": 607 + }, + { + "epoch": 0.23250478011472275, + "grad_norm": 2.6761707774182173, + "learning_rate": 8.93247982407385e-07, + "loss": 0.4142, + "step": 608 + }, + { + "epoch": 0.232887189292543, + "grad_norm": 4.37456401287424, + "learning_rate": 8.934769856759282e-07, + "loss": 0.3991, + "step": 609 + }, + { + "epoch": 0.2332695984703633, + "grad_norm": 4.247755916759948, + "learning_rate": 8.937056132211385e-07, + "loss": 0.299, + "step": 610 + }, + { + "epoch": 0.23365200764818356, + "grad_norm": 2.9509372580052586, + "learning_rate": 8.939338662738878e-07, + "loss": 0.1705, + "step": 611 + }, + { + "epoch": 0.23403441682600382, + "grad_norm": 4.224394914083757, + "learning_rate": 8.94161746059009e-07, + "loss": 0.2091, + "step": 612 + }, + { + "epoch": 0.2344168260038241, + "grad_norm": 5.198523150312591, + "learning_rate": 8.943892537953359e-07, + "loss": 0.4931, + "step": 613 + }, + { + "epoch": 0.23479923518164436, + "grad_norm": 5.9789195081642825, + "learning_rate": 8.946163906957424e-07, + "loss": 0.4767, + "step": 614 + }, + { + "epoch": 0.23518164435946462, + "grad_norm": 3.501394266617619, + "learning_rate": 8.948431579671812e-07, + "loss": 0.7518, + "step": 615 + }, + { + "epoch": 0.2355640535372849, + "grad_norm": 2.534678080506824, + "learning_rate": 8.950695568107218e-07, + "loss": 0.6598, + "step": 616 + }, + { + "epoch": 0.23594646271510517, + "grad_norm": 3.2492748331786774, + "learning_rate": 8.952955884215896e-07, + "loss": 0.6163, + "step": 617 + }, + { + "epoch": 0.23632887189292542, + "grad_norm": 2.9826559338411354, + "learning_rate": 8.955212539892024e-07, + "loss": 0.3615, + "step": 618 + }, + { + "epoch": 0.2367112810707457, + "grad_norm": 3.318038053711697, + "learning_rate": 8.957465546972099e-07, + "loss": 0.5365, + "step": 619 + }, + { + "epoch": 0.23709369024856597, + "grad_norm": 3.4175041536921253, + "learning_rate": 8.959714917235292e-07, + "loss": 0.2957, + "step": 620 + }, + { + "epoch": 0.23747609942638623, + "grad_norm": 4.206538406165909, + "learning_rate": 8.96196066240382e-07, + "loss": 0.5224, + "step": 621 + }, + { + "epoch": 0.23785850860420651, + "grad_norm": 3.2186359274196215, + "learning_rate": 8.964202794143326e-07, + "loss": 0.1611, + "step": 622 + }, + { + "epoch": 0.23824091778202677, + "grad_norm": 3.1131388058497014, + "learning_rate": 8.966441324063227e-07, + "loss": 0.3956, + "step": 623 + }, + { + "epoch": 0.23862332695984703, + "grad_norm": 2.889521867739628, + "learning_rate": 8.968676263717078e-07, + "loss": 0.172, + "step": 624 + }, + { + "epoch": 0.2390057361376673, + "grad_norm": 5.782986436859959, + "learning_rate": 8.970907624602941e-07, + "loss": 0.183, + "step": 625 + }, + { + "epoch": 0.23938814531548758, + "grad_norm": 8.177253193316057, + "learning_rate": 8.973135418163724e-07, + "loss": 0.4219, + "step": 626 + }, + { + "epoch": 0.23977055449330784, + "grad_norm": 3.4656521822772333, + "learning_rate": 8.975359655787545e-07, + "loss": 0.6771, + "step": 627 + }, + { + "epoch": 0.2401529636711281, + "grad_norm": 3.9809862274031342, + "learning_rate": 8.977580348808072e-07, + "loss": 0.4621, + "step": 628 + }, + { + "epoch": 0.24053537284894838, + "grad_norm": 2.590580301854008, + "learning_rate": 8.97979750850488e-07, + "loss": 0.3399, + "step": 629 + }, + { + "epoch": 0.24091778202676864, + "grad_norm": 2.797762426317898, + "learning_rate": 8.982011146103783e-07, + "loss": 0.2546, + "step": 630 + }, + { + "epoch": 0.2413001912045889, + "grad_norm": 3.0114940752468007, + "learning_rate": 8.984221272777187e-07, + "loss": 0.1752, + "step": 631 + }, + { + "epoch": 0.24168260038240919, + "grad_norm": 3.0610270169202085, + "learning_rate": 8.986427899644419e-07, + "loss": 0.3983, + "step": 632 + }, + { + "epoch": 0.24206500956022944, + "grad_norm": 2.8067468080231506, + "learning_rate": 8.988631037772066e-07, + "loss": 0.2748, + "step": 633 + }, + { + "epoch": 0.2424474187380497, + "grad_norm": 4.000175233817278, + "learning_rate": 8.990830698174309e-07, + "loss": 0.1971, + "step": 634 + }, + { + "epoch": 0.24282982791587, + "grad_norm": 3.769745657560052, + "learning_rate": 8.993026891813253e-07, + "loss": 0.4406, + "step": 635 + }, + { + "epoch": 0.24321223709369025, + "grad_norm": 4.197854974765107, + "learning_rate": 8.995219629599254e-07, + "loss": 0.243, + "step": 636 + }, + { + "epoch": 0.2435946462715105, + "grad_norm": 2.9496964831922186, + "learning_rate": 8.997408922391241e-07, + "loss": 0.1551, + "step": 637 + }, + { + "epoch": 0.2439770554493308, + "grad_norm": 7.136065540371976, + "learning_rate": 8.999594780997046e-07, + "loss": 0.5281, + "step": 638 + }, + { + "epoch": 0.24435946462715105, + "grad_norm": 3.1328861675020034, + "learning_rate": 9.001777216173722e-07, + "loss": 0.4891, + "step": 639 + }, + { + "epoch": 0.2447418738049713, + "grad_norm": 3.851332837544597, + "learning_rate": 9.003956238627853e-07, + "loss": 0.478, + "step": 640 + }, + { + "epoch": 0.2451242829827916, + "grad_norm": 2.725577365010256, + "learning_rate": 9.00613185901588e-07, + "loss": 0.4471, + "step": 641 + }, + { + "epoch": 0.24550669216061186, + "grad_norm": 2.5746230877323146, + "learning_rate": 9.008304087944406e-07, + "loss": 0.3046, + "step": 642 + }, + { + "epoch": 0.24588910133843211, + "grad_norm": 4.590008308220286, + "learning_rate": 9.010472935970509e-07, + "loss": 0.343, + "step": 643 + }, + { + "epoch": 0.2462715105162524, + "grad_norm": 3.231391156948366, + "learning_rate": 9.012638413602053e-07, + "loss": 0.3504, + "step": 644 + }, + { + "epoch": 0.24665391969407266, + "grad_norm": 2.5125683523720896, + "learning_rate": 9.014800531297985e-07, + "loss": 0.2493, + "step": 645 + }, + { + "epoch": 0.24703632887189292, + "grad_norm": 2.79504368674542, + "learning_rate": 9.016959299468647e-07, + "loss": 0.2146, + "step": 646 + }, + { + "epoch": 0.2474187380497132, + "grad_norm": 3.8159029422788753, + "learning_rate": 9.019114728476074e-07, + "loss": 0.4076, + "step": 647 + }, + { + "epoch": 0.24780114722753346, + "grad_norm": 4.783488030319095, + "learning_rate": 9.021266828634289e-07, + "loss": 0.1647, + "step": 648 + }, + { + "epoch": 0.24818355640535372, + "grad_norm": 6.686053253263, + "learning_rate": 9.023415610209602e-07, + "loss": 0.1261, + "step": 649 + }, + { + "epoch": 0.248565965583174, + "grad_norm": 4.911620098644459, + "learning_rate": 9.025561083420904e-07, + "loss": 0.3726, + "step": 650 + }, + { + "epoch": 0.24894837476099427, + "grad_norm": 6.45514667607322, + "learning_rate": 9.027703258439958e-07, + "loss": 0.6213, + "step": 651 + }, + { + "epoch": 0.24933078393881453, + "grad_norm": 3.249407494161105, + "learning_rate": 9.029842145391691e-07, + "loss": 0.6847, + "step": 652 + }, + { + "epoch": 0.2497131931166348, + "grad_norm": 4.2437066998751725, + "learning_rate": 9.031977754354472e-07, + "loss": 0.5215, + "step": 653 + }, + { + "epoch": 0.25009560229445504, + "grad_norm": 1.7989403837715172, + "learning_rate": 9.034110095360411e-07, + "loss": 0.543, + "step": 654 + }, + { + "epoch": 0.25047801147227533, + "grad_norm": 2.760700388788634, + "learning_rate": 9.03623917839563e-07, + "loss": 0.3404, + "step": 655 + }, + { + "epoch": 0.2508604206500956, + "grad_norm": 3.1453054292526397, + "learning_rate": 9.038365013400548e-07, + "loss": 0.4307, + "step": 656 + }, + { + "epoch": 0.25124282982791585, + "grad_norm": 2.750521350338774, + "learning_rate": 9.040487610270163e-07, + "loss": 0.3821, + "step": 657 + }, + { + "epoch": 0.25162523900573613, + "grad_norm": 4.3433778452452625, + "learning_rate": 9.04260697885432e-07, + "loss": 0.4657, + "step": 658 + }, + { + "epoch": 0.2520076481835564, + "grad_norm": 3.038734869470121, + "learning_rate": 9.044723128957992e-07, + "loss": 0.4424, + "step": 659 + }, + { + "epoch": 0.25239005736137665, + "grad_norm": 3.6418428798847953, + "learning_rate": 9.046836070341548e-07, + "loss": 0.1612, + "step": 660 + }, + { + "epoch": 0.25277246653919694, + "grad_norm": 4.814089029716334, + "learning_rate": 9.04894581272103e-07, + "loss": 0.1493, + "step": 661 + }, + { + "epoch": 0.2531548757170172, + "grad_norm": 4.9591074534253785, + "learning_rate": 9.051052365768415e-07, + "loss": 0.1937, + "step": 662 + }, + { + "epoch": 0.25353728489483746, + "grad_norm": 3.944316957470118, + "learning_rate": 9.053155739111876e-07, + "loss": 0.3188, + "step": 663 + }, + { + "epoch": 0.25391969407265774, + "grad_norm": 4.223668375030919, + "learning_rate": 9.055255942336061e-07, + "loss": 0.6315, + "step": 664 + }, + { + "epoch": 0.25430210325047803, + "grad_norm": 3.4732364817739083, + "learning_rate": 9.057352984982342e-07, + "loss": 0.6295, + "step": 665 + }, + { + "epoch": 0.25468451242829826, + "grad_norm": 3.165842718713124, + "learning_rate": 9.05944687654908e-07, + "loss": 0.689, + "step": 666 + }, + { + "epoch": 0.25506692160611855, + "grad_norm": 2.3999509726920847, + "learning_rate": 9.06153762649188e-07, + "loss": 0.3787, + "step": 667 + }, + { + "epoch": 0.25544933078393883, + "grad_norm": 3.3319283544432743, + "learning_rate": 9.063625244223857e-07, + "loss": 0.4465, + "step": 668 + }, + { + "epoch": 0.25583173996175906, + "grad_norm": 2.5320497356188447, + "learning_rate": 9.065709739115878e-07, + "loss": 0.4041, + "step": 669 + }, + { + "epoch": 0.25621414913957935, + "grad_norm": 3.306812513990708, + "learning_rate": 9.067791120496818e-07, + "loss": 0.1911, + "step": 670 + }, + { + "epoch": 0.25659655831739964, + "grad_norm": 4.444888988091275, + "learning_rate": 9.069869397653817e-07, + "loss": 0.1971, + "step": 671 + }, + { + "epoch": 0.25697896749521987, + "grad_norm": 2.9716506440106363, + "learning_rate": 9.071944579832521e-07, + "loss": 0.1668, + "step": 672 + }, + { + "epoch": 0.25736137667304015, + "grad_norm": 4.030938661366508, + "learning_rate": 9.074016676237329e-07, + "loss": 0.2018, + "step": 673 + }, + { + "epoch": 0.25774378585086044, + "grad_norm": 3.297851128729243, + "learning_rate": 9.076085696031645e-07, + "loss": 0.1608, + "step": 674 + }, + { + "epoch": 0.25812619502868067, + "grad_norm": 4.587013339388978, + "learning_rate": 9.078151648338114e-07, + "loss": 0.3073, + "step": 675 + }, + { + "epoch": 0.25850860420650096, + "grad_norm": 5.998579174931337, + "learning_rate": 9.080214542238866e-07, + "loss": 0.3133, + "step": 676 + }, + { + "epoch": 0.25889101338432124, + "grad_norm": 3.426778824253407, + "learning_rate": 9.082274386775756e-07, + "loss": 0.6327, + "step": 677 + }, + { + "epoch": 0.2592734225621415, + "grad_norm": 2.835126944900558, + "learning_rate": 9.0843311909506e-07, + "loss": 0.7641, + "step": 678 + }, + { + "epoch": 0.25965583173996176, + "grad_norm": 3.2794105823608297, + "learning_rate": 9.086384963725418e-07, + "loss": 0.359, + "step": 679 + }, + { + "epoch": 0.26003824091778205, + "grad_norm": 2.7284902845883954, + "learning_rate": 9.088435714022651e-07, + "loss": 0.3456, + "step": 680 + }, + { + "epoch": 0.2604206500956023, + "grad_norm": 3.007036698599815, + "learning_rate": 9.090483450725417e-07, + "loss": 0.2821, + "step": 681 + }, + { + "epoch": 0.26080305927342257, + "grad_norm": 2.9265717481313267, + "learning_rate": 9.092528182677723e-07, + "loss": 0.5039, + "step": 682 + }, + { + "epoch": 0.26118546845124285, + "grad_norm": 2.9918622270880832, + "learning_rate": 9.094569918684705e-07, + "loss": 0.3882, + "step": 683 + }, + { + "epoch": 0.2615678776290631, + "grad_norm": 2.9372459464087823, + "learning_rate": 9.096608667512846e-07, + "loss": 0.2574, + "step": 684 + }, + { + "epoch": 0.26195028680688337, + "grad_norm": 2.7236732542824864, + "learning_rate": 9.098644437890217e-07, + "loss": 0.1896, + "step": 685 + }, + { + "epoch": 0.26233269598470366, + "grad_norm": 4.740230753629643, + "learning_rate": 9.100677238506682e-07, + "loss": 0.181, + "step": 686 + }, + { + "epoch": 0.2627151051625239, + "grad_norm": 4.5618603401664535, + "learning_rate": 9.102707078014137e-07, + "loss": 0.2754, + "step": 687 + }, + { + "epoch": 0.2630975143403442, + "grad_norm": 5.453959841069427, + "learning_rate": 9.104733965026722e-07, + "loss": 0.4325, + "step": 688 + }, + { + "epoch": 0.26347992351816446, + "grad_norm": 6.570727903318121, + "learning_rate": 9.106757908121041e-07, + "loss": 0.6136, + "step": 689 + }, + { + "epoch": 0.2638623326959847, + "grad_norm": 3.123896136824884, + "learning_rate": 9.108778915836382e-07, + "loss": 0.3504, + "step": 690 + }, + { + "epoch": 0.264244741873805, + "grad_norm": 3.4093210917126133, + "learning_rate": 9.110796996674936e-07, + "loss": 0.6091, + "step": 691 + }, + { + "epoch": 0.26462715105162526, + "grad_norm": 4.880015409789222, + "learning_rate": 9.112812159102003e-07, + "loss": 0.5309, + "step": 692 + }, + { + "epoch": 0.2650095602294455, + "grad_norm": 3.3460021233332453, + "learning_rate": 9.114824411546215e-07, + "loss": 0.4077, + "step": 693 + }, + { + "epoch": 0.2653919694072658, + "grad_norm": 2.398058833446472, + "learning_rate": 9.116833762399739e-07, + "loss": 0.3564, + "step": 694 + }, + { + "epoch": 0.26577437858508607, + "grad_norm": 3.449133037050817, + "learning_rate": 9.118840220018492e-07, + "loss": 0.1793, + "step": 695 + }, + { + "epoch": 0.2661567877629063, + "grad_norm": 3.9880952109829617, + "learning_rate": 9.120843792722349e-07, + "loss": 0.3157, + "step": 696 + }, + { + "epoch": 0.2665391969407266, + "grad_norm": 3.666445869964919, + "learning_rate": 9.122844488795347e-07, + "loss": 0.2338, + "step": 697 + }, + { + "epoch": 0.2669216061185469, + "grad_norm": 4.149391883508918, + "learning_rate": 9.124842316485896e-07, + "loss": 0.1581, + "step": 698 + }, + { + "epoch": 0.2673040152963671, + "grad_norm": 3.5721827310212104, + "learning_rate": 9.126837284006975e-07, + "loss": 0.1278, + "step": 699 + }, + { + "epoch": 0.2676864244741874, + "grad_norm": 4.706442844447054, + "learning_rate": 9.128829399536345e-07, + "loss": 0.2241, + "step": 700 + }, + { + "epoch": 0.2680688336520076, + "grad_norm": 10.142777271838952, + "learning_rate": 9.130818671216741e-07, + "loss": 0.5104, + "step": 701 + }, + { + "epoch": 0.2684512428298279, + "grad_norm": 4.933768492055857, + "learning_rate": 9.132805107156076e-07, + "loss": 0.5941, + "step": 702 + }, + { + "epoch": 0.2688336520076482, + "grad_norm": 2.7437742158346863, + "learning_rate": 9.134788715427637e-07, + "loss": 0.2277, + "step": 703 + }, + { + "epoch": 0.2692160611854684, + "grad_norm": 3.2670954993715995, + "learning_rate": 9.136769504070285e-07, + "loss": 0.5396, + "step": 704 + }, + { + "epoch": 0.2695984703632887, + "grad_norm": 2.0297961387309633, + "learning_rate": 9.13874748108865e-07, + "loss": 0.4631, + "step": 705 + }, + { + "epoch": 0.269980879541109, + "grad_norm": 6.472273452112172, + "learning_rate": 9.14072265445332e-07, + "loss": 0.481, + "step": 706 + }, + { + "epoch": 0.27036328871892923, + "grad_norm": 3.651355173071301, + "learning_rate": 9.142695032101042e-07, + "loss": 0.4247, + "step": 707 + }, + { + "epoch": 0.2707456978967495, + "grad_norm": 3.490073569104415, + "learning_rate": 9.144664621934904e-07, + "loss": 0.2685, + "step": 708 + }, + { + "epoch": 0.2711281070745698, + "grad_norm": 3.3061256255497162, + "learning_rate": 9.146631431824534e-07, + "loss": 0.4662, + "step": 709 + }, + { + "epoch": 0.27151051625239003, + "grad_norm": 4.023618112602692, + "learning_rate": 9.148595469606283e-07, + "loss": 0.243, + "step": 710 + }, + { + "epoch": 0.2718929254302103, + "grad_norm": 4.451440110939301, + "learning_rate": 9.150556743083417e-07, + "loss": 0.4252, + "step": 711 + }, + { + "epoch": 0.2722753346080306, + "grad_norm": 3.282564898834976, + "learning_rate": 9.152515260026294e-07, + "loss": 0.1568, + "step": 712 + }, + { + "epoch": 0.27265774378585084, + "grad_norm": 5.136206678967686, + "learning_rate": 9.154471028172556e-07, + "loss": 0.3462, + "step": 713 + }, + { + "epoch": 0.2730401529636711, + "grad_norm": 3.765053860100544, + "learning_rate": 9.156424055227317e-07, + "loss": 0.6971, + "step": 714 + }, + { + "epoch": 0.2734225621414914, + "grad_norm": 4.7913020561927375, + "learning_rate": 9.158374348863336e-07, + "loss": 0.6216, + "step": 715 + }, + { + "epoch": 0.27380497131931164, + "grad_norm": 2.896510879693719, + "learning_rate": 9.160321916721196e-07, + "loss": 0.5049, + "step": 716 + }, + { + "epoch": 0.2741873804971319, + "grad_norm": 3.0359787385017647, + "learning_rate": 9.162266766409494e-07, + "loss": 0.4884, + "step": 717 + }, + { + "epoch": 0.2745697896749522, + "grad_norm": 4.12477504232416, + "learning_rate": 9.164208905505015e-07, + "loss": 0.3276, + "step": 718 + }, + { + "epoch": 0.27495219885277244, + "grad_norm": 2.762041690834952, + "learning_rate": 9.166148341552905e-07, + "loss": 0.3178, + "step": 719 + }, + { + "epoch": 0.27533460803059273, + "grad_norm": 2.7617872949966977, + "learning_rate": 9.16808508206685e-07, + "loss": 0.2372, + "step": 720 + }, + { + "epoch": 0.275717017208413, + "grad_norm": 6.980629140487373, + "learning_rate": 9.170019134529255e-07, + "loss": 0.5347, + "step": 721 + }, + { + "epoch": 0.27609942638623325, + "grad_norm": 4.8218599923692045, + "learning_rate": 9.171950506391405e-07, + "loss": 0.3513, + "step": 722 + }, + { + "epoch": 0.27648183556405354, + "grad_norm": 4.577506274183089, + "learning_rate": 9.173879205073653e-07, + "loss": 0.1856, + "step": 723 + }, + { + "epoch": 0.2768642447418738, + "grad_norm": 3.8802146280054157, + "learning_rate": 9.175805237965581e-07, + "loss": 0.1468, + "step": 724 + }, + { + "epoch": 0.27724665391969405, + "grad_norm": 5.233321627626063, + "learning_rate": 9.177728612426173e-07, + "loss": 0.183, + "step": 725 + }, + { + "epoch": 0.27762906309751434, + "grad_norm": 4.871894434118751, + "learning_rate": 9.17964933578398e-07, + "loss": 0.5844, + "step": 726 + }, + { + "epoch": 0.2780114722753346, + "grad_norm": 3.7483749683375165, + "learning_rate": 9.181567415337294e-07, + "loss": 0.3946, + "step": 727 + }, + { + "epoch": 0.27839388145315486, + "grad_norm": 3.927752221141394, + "learning_rate": 9.183482858354308e-07, + "loss": 0.7132, + "step": 728 + }, + { + "epoch": 0.27877629063097514, + "grad_norm": 2.509795833658178, + "learning_rate": 9.185395672073286e-07, + "loss": 0.4222, + "step": 729 + }, + { + "epoch": 0.27915869980879543, + "grad_norm": 3.4221312544513647, + "learning_rate": 9.187305863702726e-07, + "loss": 0.501, + "step": 730 + }, + { + "epoch": 0.27954110898661566, + "grad_norm": 3.2600125355859806, + "learning_rate": 9.189213440421519e-07, + "loss": 0.4319, + "step": 731 + }, + { + "epoch": 0.27992351816443595, + "grad_norm": 3.6358730754710153, + "learning_rate": 9.191118409379119e-07, + "loss": 0.5501, + "step": 732 + }, + { + "epoch": 0.28030592734225623, + "grad_norm": 2.838876425692313, + "learning_rate": 9.193020777695696e-07, + "loss": 0.4226, + "step": 733 + }, + { + "epoch": 0.28068833652007646, + "grad_norm": 4.23248541643004, + "learning_rate": 9.194920552462298e-07, + "loss": 0.2216, + "step": 734 + }, + { + "epoch": 0.28107074569789675, + "grad_norm": 5.05975286098859, + "learning_rate": 9.196817740741012e-07, + "loss": 0.1942, + "step": 735 + }, + { + "epoch": 0.28145315487571704, + "grad_norm": 7.157943591816231, + "learning_rate": 9.19871234956512e-07, + "loss": 0.291, + "step": 736 + }, + { + "epoch": 0.28183556405353727, + "grad_norm": 6.540916189244941, + "learning_rate": 9.20060438593925e-07, + "loss": 0.1721, + "step": 737 + }, + { + "epoch": 0.28221797323135756, + "grad_norm": 6.142323188767702, + "learning_rate": 9.202493856839546e-07, + "loss": 0.3475, + "step": 738 + }, + { + "epoch": 0.28260038240917784, + "grad_norm": 5.9353717402798365, + "learning_rate": 9.204380769213804e-07, + "loss": 0.6243, + "step": 739 + }, + { + "epoch": 0.2829827915869981, + "grad_norm": 3.8146311322392403, + "learning_rate": 9.206265129981641e-07, + "loss": 0.659, + "step": 740 + }, + { + "epoch": 0.28336520076481836, + "grad_norm": 2.7360985255933583, + "learning_rate": 9.208146946034633e-07, + "loss": 0.3429, + "step": 741 + }, + { + "epoch": 0.28374760994263865, + "grad_norm": 3.501030089156713, + "learning_rate": 9.210026224236482e-07, + "loss": 0.4522, + "step": 742 + }, + { + "epoch": 0.2841300191204589, + "grad_norm": 2.9306435614494535, + "learning_rate": 9.211902971423152e-07, + "loss": 0.6043, + "step": 743 + }, + { + "epoch": 0.28451242829827916, + "grad_norm": 3.191075133909245, + "learning_rate": 9.213777194403025e-07, + "loss": 0.5403, + "step": 744 + }, + { + "epoch": 0.28489483747609945, + "grad_norm": 4.1194321016924285, + "learning_rate": 9.21564889995705e-07, + "loss": 0.1928, + "step": 745 + }, + { + "epoch": 0.2852772466539197, + "grad_norm": 2.9657019779034983, + "learning_rate": 9.217518094838887e-07, + "loss": 0.2501, + "step": 746 + }, + { + "epoch": 0.28565965583173997, + "grad_norm": 3.2172360609705466, + "learning_rate": 9.219384785775058e-07, + "loss": 0.4956, + "step": 747 + }, + { + "epoch": 0.28604206500956025, + "grad_norm": 2.7979749509223217, + "learning_rate": 9.221248979465084e-07, + "loss": 0.1676, + "step": 748 + }, + { + "epoch": 0.2864244741873805, + "grad_norm": 4.155212043281761, + "learning_rate": 9.223110682581634e-07, + "loss": 0.1349, + "step": 749 + }, + { + "epoch": 0.28680688336520077, + "grad_norm": 4.2952560426645, + "learning_rate": 9.224969901770675e-07, + "loss": 0.2733, + "step": 750 + }, + { + "epoch": 0.28718929254302106, + "grad_norm": 6.104755186628853, + "learning_rate": 9.226826643651602e-07, + "loss": 0.5484, + "step": 751 + }, + { + "epoch": 0.2875717017208413, + "grad_norm": 4.040718255733354, + "learning_rate": 9.228680914817387e-07, + "loss": 0.4754, + "step": 752 + }, + { + "epoch": 0.2879541108986616, + "grad_norm": 3.158374637813444, + "learning_rate": 9.230532721834717e-07, + "loss": 0.5389, + "step": 753 + }, + { + "epoch": 0.28833652007648186, + "grad_norm": 4.190667619708935, + "learning_rate": 9.232382071244136e-07, + "loss": 0.4755, + "step": 754 + }, + { + "epoch": 0.2887189292543021, + "grad_norm": 4.238531683629353, + "learning_rate": 9.234228969560181e-07, + "loss": 0.5469, + "step": 755 + }, + { + "epoch": 0.2891013384321224, + "grad_norm": 1.9615411931014568, + "learning_rate": 9.236073423271517e-07, + "loss": 0.3119, + "step": 756 + }, + { + "epoch": 0.28948374760994267, + "grad_norm": 2.6342224511339403, + "learning_rate": 9.237915438841085e-07, + "loss": 0.4063, + "step": 757 + }, + { + "epoch": 0.2898661567877629, + "grad_norm": 4.365004756614654, + "learning_rate": 9.239755022706225e-07, + "loss": 0.4622, + "step": 758 + }, + { + "epoch": 0.2902485659655832, + "grad_norm": 2.5054584994223874, + "learning_rate": 9.241592181278815e-07, + "loss": 0.312, + "step": 759 + }, + { + "epoch": 0.29063097514340347, + "grad_norm": 4.476758702956534, + "learning_rate": 9.243426920945409e-07, + "loss": 0.3864, + "step": 760 + }, + { + "epoch": 0.2910133843212237, + "grad_norm": 3.021955417639978, + "learning_rate": 9.245259248067367e-07, + "loss": 0.1333, + "step": 761 + }, + { + "epoch": 0.291395793499044, + "grad_norm": 4.690756315239575, + "learning_rate": 9.247089168980988e-07, + "loss": 0.1441, + "step": 762 + }, + { + "epoch": 0.2917782026768642, + "grad_norm": 7.829398959538781, + "learning_rate": 9.24891668999764e-07, + "loss": 0.2082, + "step": 763 + }, + { + "epoch": 0.2921606118546845, + "grad_norm": 5.258302209503068, + "learning_rate": 9.250741817403894e-07, + "loss": 0.6746, + "step": 764 + }, + { + "epoch": 0.2925430210325048, + "grad_norm": 4.864542185618209, + "learning_rate": 9.252564557461648e-07, + "loss": 0.6271, + "step": 765 + }, + { + "epoch": 0.292925430210325, + "grad_norm": 2.4464280384051205, + "learning_rate": 9.254384916408264e-07, + "loss": 0.3544, + "step": 766 + }, + { + "epoch": 0.2933078393881453, + "grad_norm": 1.737148919072545, + "learning_rate": 9.25620290045669e-07, + "loss": 0.4638, + "step": 767 + }, + { + "epoch": 0.2936902485659656, + "grad_norm": 3.298820030333467, + "learning_rate": 9.258018515795588e-07, + "loss": 0.3998, + "step": 768 + }, + { + "epoch": 0.2940726577437858, + "grad_norm": 2.7308884545045125, + "learning_rate": 9.25983176858946e-07, + "loss": 0.3025, + "step": 769 + }, + { + "epoch": 0.2944550669216061, + "grad_norm": 3.6917404104291522, + "learning_rate": 9.261642664978777e-07, + "loss": 0.3902, + "step": 770 + }, + { + "epoch": 0.2948374760994264, + "grad_norm": 3.438420536589146, + "learning_rate": 9.263451211080103e-07, + "loss": 0.1922, + "step": 771 + }, + { + "epoch": 0.29521988527724663, + "grad_norm": 2.690586446913425, + "learning_rate": 9.265257412986215e-07, + "loss": 0.1766, + "step": 772 + }, + { + "epoch": 0.2956022944550669, + "grad_norm": 3.889979297416141, + "learning_rate": 9.267061276766227e-07, + "loss": 0.1684, + "step": 773 + }, + { + "epoch": 0.2959847036328872, + "grad_norm": 5.210210883911897, + "learning_rate": 9.268862808465719e-07, + "loss": 0.143, + "step": 774 + }, + { + "epoch": 0.29636711281070743, + "grad_norm": 5.16822561719914, + "learning_rate": 9.270662014106851e-07, + "loss": 0.1774, + "step": 775 + }, + { + "epoch": 0.2967495219885277, + "grad_norm": 5.552930851914382, + "learning_rate": 9.272458899688485e-07, + "loss": 0.4315, + "step": 776 + }, + { + "epoch": 0.297131931166348, + "grad_norm": 3.574750216634461, + "learning_rate": 9.274253471186307e-07, + "loss": 0.6128, + "step": 777 + }, + { + "epoch": 0.29751434034416824, + "grad_norm": 3.457119249851013, + "learning_rate": 9.276045734552951e-07, + "loss": 0.6457, + "step": 778 + }, + { + "epoch": 0.2978967495219885, + "grad_norm": 3.1391020920925903, + "learning_rate": 9.277835695718105e-07, + "loss": 0.379, + "step": 779 + }, + { + "epoch": 0.2982791586998088, + "grad_norm": 3.245935686396853, + "learning_rate": 9.279623360588637e-07, + "loss": 0.3597, + "step": 780 + }, + { + "epoch": 0.29866156787762904, + "grad_norm": 4.948312629784024, + "learning_rate": 9.281408735048715e-07, + "loss": 0.5803, + "step": 781 + }, + { + "epoch": 0.29904397705544933, + "grad_norm": 3.5383041276653304, + "learning_rate": 9.283191824959918e-07, + "loss": 0.4158, + "step": 782 + }, + { + "epoch": 0.2994263862332696, + "grad_norm": 2.6961675642106813, + "learning_rate": 9.284972636161345e-07, + "loss": 0.234, + "step": 783 + }, + { + "epoch": 0.29980879541108985, + "grad_norm": 3.6581698862695102, + "learning_rate": 9.286751174469749e-07, + "loss": 0.2256, + "step": 784 + }, + { + "epoch": 0.30019120458891013, + "grad_norm": 3.658953979144392, + "learning_rate": 9.288527445679631e-07, + "loss": 0.2376, + "step": 785 + }, + { + "epoch": 0.3005736137667304, + "grad_norm": 4.144486839638947, + "learning_rate": 9.290301455563364e-07, + "loss": 0.1269, + "step": 786 + }, + { + "epoch": 0.30095602294455065, + "grad_norm": 4.947979200536853, + "learning_rate": 9.292073209871303e-07, + "loss": 0.1268, + "step": 787 + }, + { + "epoch": 0.30133843212237094, + "grad_norm": 6.4016710245636865, + "learning_rate": 9.293842714331897e-07, + "loss": 0.4139, + "step": 788 + }, + { + "epoch": 0.3017208413001912, + "grad_norm": 4.04470071597837, + "learning_rate": 9.295609974651798e-07, + "loss": 0.59, + "step": 789 + }, + { + "epoch": 0.30210325047801145, + "grad_norm": 4.723204474422228, + "learning_rate": 9.297374996515978e-07, + "loss": 0.6182, + "step": 790 + }, + { + "epoch": 0.30248565965583174, + "grad_norm": 3.057452343298742, + "learning_rate": 9.299137785587831e-07, + "loss": 0.537, + "step": 791 + }, + { + "epoch": 0.302868068833652, + "grad_norm": 2.19324306271355, + "learning_rate": 9.300898347509282e-07, + "loss": 0.6086, + "step": 792 + }, + { + "epoch": 0.30325047801147226, + "grad_norm": 2.9430732613303423, + "learning_rate": 9.302656687900906e-07, + "loss": 0.2686, + "step": 793 + }, + { + "epoch": 0.30363288718929254, + "grad_norm": 3.9081728665745197, + "learning_rate": 9.304412812362025e-07, + "loss": 0.444, + "step": 794 + }, + { + "epoch": 0.30401529636711283, + "grad_norm": 2.4089558237980917, + "learning_rate": 9.306166726470812e-07, + "loss": 0.3601, + "step": 795 + }, + { + "epoch": 0.30439770554493306, + "grad_norm": 3.4118547437366673, + "learning_rate": 9.307918435784414e-07, + "loss": 0.3465, + "step": 796 + }, + { + "epoch": 0.30478011472275335, + "grad_norm": 3.6235244641842153, + "learning_rate": 9.309667945839041e-07, + "loss": 0.1584, + "step": 797 + }, + { + "epoch": 0.30516252390057363, + "grad_norm": 3.437496948240833, + "learning_rate": 9.311415262150076e-07, + "loss": 0.2358, + "step": 798 + }, + { + "epoch": 0.30554493307839387, + "grad_norm": 3.4448784448801675, + "learning_rate": 9.313160390212184e-07, + "loss": 0.1215, + "step": 799 + }, + { + "epoch": 0.30592734225621415, + "grad_norm": 3.5191571611593258, + "learning_rate": 9.314903335499412e-07, + "loss": 0.1596, + "step": 800 + }, + { + "epoch": 0.30630975143403444, + "grad_norm": 6.999381174254604, + "learning_rate": 9.316644103465291e-07, + "loss": 0.5317, + "step": 801 + }, + { + "epoch": 0.30669216061185467, + "grad_norm": 3.616443532807995, + "learning_rate": 9.318382699542936e-07, + "loss": 0.3562, + "step": 802 + }, + { + "epoch": 0.30707456978967496, + "grad_norm": 2.3191912989439687, + "learning_rate": 9.320119129145158e-07, + "loss": 0.4623, + "step": 803 + }, + { + "epoch": 0.30745697896749524, + "grad_norm": 3.41795103232015, + "learning_rate": 9.321853397664554e-07, + "loss": 0.5882, + "step": 804 + }, + { + "epoch": 0.3078393881453155, + "grad_norm": 3.6543276697326195, + "learning_rate": 9.323585510473612e-07, + "loss": 0.4667, + "step": 805 + }, + { + "epoch": 0.30822179732313576, + "grad_norm": 2.31745390334883, + "learning_rate": 9.325315472924812e-07, + "loss": 0.2955, + "step": 806 + }, + { + "epoch": 0.30860420650095605, + "grad_norm": 4.031207417107943, + "learning_rate": 9.327043290350726e-07, + "loss": 0.3176, + "step": 807 + }, + { + "epoch": 0.3089866156787763, + "grad_norm": 4.097064125473531, + "learning_rate": 9.328768968064107e-07, + "loss": 0.3082, + "step": 808 + }, + { + "epoch": 0.30936902485659656, + "grad_norm": 2.4083338050005847, + "learning_rate": 9.330492511358006e-07, + "loss": 0.1999, + "step": 809 + }, + { + "epoch": 0.30975143403441685, + "grad_norm": 3.1727456086591035, + "learning_rate": 9.332213925505848e-07, + "loss": 0.1533, + "step": 810 + }, + { + "epoch": 0.3101338432122371, + "grad_norm": 4.600015117786278, + "learning_rate": 9.333933215761543e-07, + "loss": 0.2456, + "step": 811 + }, + { + "epoch": 0.31051625239005737, + "grad_norm": 7.192106412819565, + "learning_rate": 9.335650387359578e-07, + "loss": 0.2425, + "step": 812 + }, + { + "epoch": 0.31089866156787765, + "grad_norm": 7.02838320463472, + "learning_rate": 9.337365445515112e-07, + "loss": 0.2411, + "step": 813 + }, + { + "epoch": 0.3112810707456979, + "grad_norm": 4.490032601354969, + "learning_rate": 9.339078395424072e-07, + "loss": 0.6411, + "step": 814 + }, + { + "epoch": 0.31166347992351817, + "grad_norm": 3.2150438322932544, + "learning_rate": 9.340789242263248e-07, + "loss": 0.4324, + "step": 815 + }, + { + "epoch": 0.31204588910133846, + "grad_norm": 3.8721174778416825, + "learning_rate": 9.342497991190384e-07, + "loss": 0.3775, + "step": 816 + }, + { + "epoch": 0.3124282982791587, + "grad_norm": 4.4248550132662805, + "learning_rate": 9.344204647344277e-07, + "loss": 0.367, + "step": 817 + }, + { + "epoch": 0.312810707456979, + "grad_norm": 3.1897752906316335, + "learning_rate": 9.345909215844859e-07, + "loss": 0.2308, + "step": 818 + }, + { + "epoch": 0.31319311663479926, + "grad_norm": 3.450483597848423, + "learning_rate": 9.347611701793305e-07, + "loss": 0.4343, + "step": 819 + }, + { + "epoch": 0.3135755258126195, + "grad_norm": 2.481171179607701, + "learning_rate": 9.349312110272107e-07, + "loss": 0.3205, + "step": 820 + }, + { + "epoch": 0.3139579349904398, + "grad_norm": 2.788976245903436, + "learning_rate": 9.351010446345182e-07, + "loss": 0.2113, + "step": 821 + }, + { + "epoch": 0.31434034416826, + "grad_norm": 6.271641000165821, + "learning_rate": 9.352706715057951e-07, + "loss": 0.3858, + "step": 822 + }, + { + "epoch": 0.3147227533460803, + "grad_norm": 3.7436334607767705, + "learning_rate": 9.354400921437429e-07, + "loss": 0.163, + "step": 823 + }, + { + "epoch": 0.3151051625239006, + "grad_norm": 3.741192711347864, + "learning_rate": 9.356093070492322e-07, + "loss": 0.1277, + "step": 824 + }, + { + "epoch": 0.3154875717017208, + "grad_norm": 5.455032146299698, + "learning_rate": 9.357783167213107e-07, + "loss": 0.413, + "step": 825 + }, + { + "epoch": 0.3158699808795411, + "grad_norm": 7.935603838938422, + "learning_rate": 9.359471216572131e-07, + "loss": 0.5146, + "step": 826 + }, + { + "epoch": 0.3162523900573614, + "grad_norm": 5.691858654591456, + "learning_rate": 9.36115722352369e-07, + "loss": 0.7377, + "step": 827 + }, + { + "epoch": 0.3166347992351816, + "grad_norm": 3.392299537442618, + "learning_rate": 9.362841193004116e-07, + "loss": 0.5823, + "step": 828 + }, + { + "epoch": 0.3170172084130019, + "grad_norm": 3.2098505635401144, + "learning_rate": 9.36452312993187e-07, + "loss": 0.4236, + "step": 829 + }, + { + "epoch": 0.3173996175908222, + "grad_norm": 4.264147380292967, + "learning_rate": 9.36620303920762e-07, + "loss": 0.5482, + "step": 830 + }, + { + "epoch": 0.3177820267686424, + "grad_norm": 3.0817239702452555, + "learning_rate": 9.367880925714339e-07, + "loss": 0.3956, + "step": 831 + }, + { + "epoch": 0.3181644359464627, + "grad_norm": 3.747304837958074, + "learning_rate": 9.369556794317375e-07, + "loss": 0.581, + "step": 832 + }, + { + "epoch": 0.318546845124283, + "grad_norm": 3.7135952396817578, + "learning_rate": 9.371230649864548e-07, + "loss": 0.4357, + "step": 833 + }, + { + "epoch": 0.3189292543021032, + "grad_norm": 4.297254067086311, + "learning_rate": 9.372902497186226e-07, + "loss": 0.2272, + "step": 834 + }, + { + "epoch": 0.3193116634799235, + "grad_norm": 2.567065289328753, + "learning_rate": 9.374572341095416e-07, + "loss": 0.1196, + "step": 835 + }, + { + "epoch": 0.3196940726577438, + "grad_norm": 4.40650365653969, + "learning_rate": 9.376240186387841e-07, + "loss": 0.2121, + "step": 836 + }, + { + "epoch": 0.32007648183556403, + "grad_norm": 3.7109315892222665, + "learning_rate": 9.377906037842022e-07, + "loss": 0.248, + "step": 837 + }, + { + "epoch": 0.3204588910133843, + "grad_norm": 7.150674972152842, + "learning_rate": 9.379569900219368e-07, + "loss": 0.1896, + "step": 838 + }, + { + "epoch": 0.3208413001912046, + "grad_norm": 3.2211659077572907, + "learning_rate": 9.381231778264247e-07, + "loss": 0.5999, + "step": 839 + }, + { + "epoch": 0.32122370936902483, + "grad_norm": 4.1372208331105345, + "learning_rate": 9.382891676704079e-07, + "loss": 0.4782, + "step": 840 + }, + { + "epoch": 0.3216061185468451, + "grad_norm": 3.8547351452280036, + "learning_rate": 9.384549600249407e-07, + "loss": 0.3241, + "step": 841 + }, + { + "epoch": 0.3219885277246654, + "grad_norm": 2.3007511985383564, + "learning_rate": 9.386205553593977e-07, + "loss": 0.5596, + "step": 842 + }, + { + "epoch": 0.32237093690248564, + "grad_norm": 2.9696172150310205, + "learning_rate": 9.387859541414829e-07, + "loss": 0.6675, + "step": 843 + }, + { + "epoch": 0.3227533460803059, + "grad_norm": 3.513270153174881, + "learning_rate": 9.389511568372363e-07, + "loss": 0.4169, + "step": 844 + }, + { + "epoch": 0.3231357552581262, + "grad_norm": 2.6253018886813506, + "learning_rate": 9.391161639110424e-07, + "loss": 0.261, + "step": 845 + }, + { + "epoch": 0.32351816443594644, + "grad_norm": 3.17710482897548, + "learning_rate": 9.392809758256383e-07, + "loss": 0.2942, + "step": 846 + }, + { + "epoch": 0.32390057361376673, + "grad_norm": 7.235980455126496, + "learning_rate": 9.39445593042121e-07, + "loss": 0.1672, + "step": 847 + }, + { + "epoch": 0.324282982791587, + "grad_norm": 3.4921857249278907, + "learning_rate": 9.396100160199549e-07, + "loss": 0.2866, + "step": 848 + }, + { + "epoch": 0.32466539196940725, + "grad_norm": 3.557020021551161, + "learning_rate": 9.397742452169806e-07, + "loss": 0.1083, + "step": 849 + }, + { + "epoch": 0.32504780114722753, + "grad_norm": 4.516592588673517, + "learning_rate": 9.399382810894211e-07, + "loss": 0.1744, + "step": 850 + }, + { + "epoch": 0.3254302103250478, + "grad_norm": 4.88414591168416, + "learning_rate": 9.401021240918907e-07, + "loss": 0.6063, + "step": 851 + }, + { + "epoch": 0.32581261950286805, + "grad_norm": 3.997072937025769, + "learning_rate": 9.402657746774017e-07, + "loss": 0.5326, + "step": 852 + }, + { + "epoch": 0.32619502868068834, + "grad_norm": 2.682499513937132, + "learning_rate": 9.404292332973727e-07, + "loss": 0.7379, + "step": 853 + }, + { + "epoch": 0.3265774378585086, + "grad_norm": 1.6899059945545842, + "learning_rate": 9.405925004016348e-07, + "loss": 0.4393, + "step": 854 + }, + { + "epoch": 0.32695984703632885, + "grad_norm": 3.3084039374696004, + "learning_rate": 9.407555764384405e-07, + "loss": 0.5194, + "step": 855 + }, + { + "epoch": 0.32734225621414914, + "grad_norm": 4.0873745493638225, + "learning_rate": 9.409184618544701e-07, + "loss": 0.5545, + "step": 856 + }, + { + "epoch": 0.3277246653919694, + "grad_norm": 2.995847371103419, + "learning_rate": 9.410811570948394e-07, + "loss": 0.2915, + "step": 857 + }, + { + "epoch": 0.32810707456978966, + "grad_norm": 2.946807077379506, + "learning_rate": 9.412436626031069e-07, + "loss": 0.2918, + "step": 858 + }, + { + "epoch": 0.32848948374760994, + "grad_norm": 3.597120744279483, + "learning_rate": 9.414059788212811e-07, + "loss": 0.4846, + "step": 859 + }, + { + "epoch": 0.32887189292543023, + "grad_norm": 3.694052096178697, + "learning_rate": 9.415681061898281e-07, + "loss": 0.3207, + "step": 860 + }, + { + "epoch": 0.32925430210325046, + "grad_norm": 3.617068591542446, + "learning_rate": 9.417300451476775e-07, + "loss": 0.1891, + "step": 861 + }, + { + "epoch": 0.32963671128107075, + "grad_norm": 3.738340690243869, + "learning_rate": 9.418917961322313e-07, + "loss": 0.1135, + "step": 862 + }, + { + "epoch": 0.33001912045889104, + "grad_norm": 5.789680231781311, + "learning_rate": 9.420533595793696e-07, + "loss": 0.2475, + "step": 863 + }, + { + "epoch": 0.33040152963671127, + "grad_norm": 5.945408743882622, + "learning_rate": 9.422147359234583e-07, + "loss": 0.6057, + "step": 864 + }, + { + "epoch": 0.33078393881453155, + "grad_norm": 4.560439389124144, + "learning_rate": 9.423759255973562e-07, + "loss": 0.5982, + "step": 865 + }, + { + "epoch": 0.33116634799235184, + "grad_norm": 4.703064370002157, + "learning_rate": 9.425369290324213e-07, + "loss": 0.4356, + "step": 866 + }, + { + "epoch": 0.33154875717017207, + "grad_norm": 2.6385904634172532, + "learning_rate": 9.426977466585182e-07, + "loss": 0.3505, + "step": 867 + }, + { + "epoch": 0.33193116634799236, + "grad_norm": 3.0721824329168745, + "learning_rate": 9.428583789040255e-07, + "loss": 0.4258, + "step": 868 + }, + { + "epoch": 0.33231357552581264, + "grad_norm": 2.8782761193446076, + "learning_rate": 9.43018826195841e-07, + "loss": 0.3341, + "step": 869 + }, + { + "epoch": 0.3326959847036329, + "grad_norm": 4.241642307493842, + "learning_rate": 9.431790889593908e-07, + "loss": 0.2586, + "step": 870 + }, + { + "epoch": 0.33307839388145316, + "grad_norm": 2.5946874188452203, + "learning_rate": 9.433391676186339e-07, + "loss": 0.452, + "step": 871 + }, + { + "epoch": 0.33346080305927345, + "grad_norm": 2.8512966084297795, + "learning_rate": 9.434990625960707e-07, + "loss": 0.2671, + "step": 872 + }, + { + "epoch": 0.3338432122370937, + "grad_norm": 5.01618170099747, + "learning_rate": 9.436587743127481e-07, + "loss": 0.3723, + "step": 873 + }, + { + "epoch": 0.33422562141491396, + "grad_norm": 4.551976020065996, + "learning_rate": 9.438183031882673e-07, + "loss": 0.2959, + "step": 874 + }, + { + "epoch": 0.33460803059273425, + "grad_norm": 5.27782676852505, + "learning_rate": 9.439776496407904e-07, + "loss": 0.338, + "step": 875 + }, + { + "epoch": 0.3349904397705545, + "grad_norm": 5.48133248383455, + "learning_rate": 9.441368140870459e-07, + "loss": 0.29, + "step": 876 + }, + { + "epoch": 0.33537284894837477, + "grad_norm": 3.56000270650525, + "learning_rate": 9.442957969423364e-07, + "loss": 0.5127, + "step": 877 + }, + { + "epoch": 0.33575525812619506, + "grad_norm": 2.825018003288225, + "learning_rate": 9.444545986205447e-07, + "loss": 0.3841, + "step": 878 + }, + { + "epoch": 0.3361376673040153, + "grad_norm": 2.008479501093062, + "learning_rate": 9.446132195341398e-07, + "loss": 0.4476, + "step": 879 + }, + { + "epoch": 0.3365200764818356, + "grad_norm": 2.281587131619646, + "learning_rate": 9.447716600941844e-07, + "loss": 0.4282, + "step": 880 + }, + { + "epoch": 0.33690248565965586, + "grad_norm": 4.628300494383288, + "learning_rate": 9.449299207103402e-07, + "loss": 0.3542, + "step": 881 + }, + { + "epoch": 0.3372848948374761, + "grad_norm": 2.3056502061077704, + "learning_rate": 9.450880017908747e-07, + "loss": 0.4025, + "step": 882 + }, + { + "epoch": 0.3376673040152964, + "grad_norm": 2.4249371274682714, + "learning_rate": 9.452459037426675e-07, + "loss": 0.4326, + "step": 883 + }, + { + "epoch": 0.3380497131931166, + "grad_norm": 2.783489225974867, + "learning_rate": 9.454036269712172e-07, + "loss": 0.325, + "step": 884 + }, + { + "epoch": 0.3384321223709369, + "grad_norm": 4.7547218544653145, + "learning_rate": 9.455611718806462e-07, + "loss": 0.2726, + "step": 885 + }, + { + "epoch": 0.3388145315487572, + "grad_norm": 2.723020334953059, + "learning_rate": 9.457185388737084e-07, + "loss": 0.2322, + "step": 886 + }, + { + "epoch": 0.3391969407265774, + "grad_norm": 3.7076447665554233, + "learning_rate": 9.458757283517939e-07, + "loss": 0.1403, + "step": 887 + }, + { + "epoch": 0.3395793499043977, + "grad_norm": 6.507640163159305, + "learning_rate": 9.460327407149374e-07, + "loss": 0.2687, + "step": 888 + }, + { + "epoch": 0.339961759082218, + "grad_norm": 3.3387596467856, + "learning_rate": 9.461895763618216e-07, + "loss": 0.4304, + "step": 889 + }, + { + "epoch": 0.3403441682600382, + "grad_norm": 4.245030358591246, + "learning_rate": 9.463462356897853e-07, + "loss": 0.7139, + "step": 890 + }, + { + "epoch": 0.3407265774378585, + "grad_norm": 3.128485300081678, + "learning_rate": 9.46502719094828e-07, + "loss": 0.489, + "step": 891 + }, + { + "epoch": 0.3411089866156788, + "grad_norm": 2.747736953361468, + "learning_rate": 9.466590269716174e-07, + "loss": 0.4317, + "step": 892 + }, + { + "epoch": 0.341491395793499, + "grad_norm": 2.683976458450521, + "learning_rate": 9.468151597134941e-07, + "loss": 0.3164, + "step": 893 + }, + { + "epoch": 0.3418738049713193, + "grad_norm": 2.566947427066389, + "learning_rate": 9.469711177124784e-07, + "loss": 0.2012, + "step": 894 + }, + { + "epoch": 0.3422562141491396, + "grad_norm": 4.025599104560061, + "learning_rate": 9.471269013592754e-07, + "loss": 0.3256, + "step": 895 + }, + { + "epoch": 0.3426386233269598, + "grad_norm": 3.243968061052379, + "learning_rate": 9.472825110432816e-07, + "loss": 0.3572, + "step": 896 + }, + { + "epoch": 0.3430210325047801, + "grad_norm": 2.8867854832019284, + "learning_rate": 9.474379471525903e-07, + "loss": 0.2392, + "step": 897 + }, + { + "epoch": 0.3434034416826004, + "grad_norm": 4.063409322059927, + "learning_rate": 9.475932100739977e-07, + "loss": 0.1856, + "step": 898 + }, + { + "epoch": 0.3437858508604206, + "grad_norm": 4.12406679636957, + "learning_rate": 9.477483001930082e-07, + "loss": 0.1849, + "step": 899 + }, + { + "epoch": 0.3441682600382409, + "grad_norm": 5.161388818893121, + "learning_rate": 9.47903217893841e-07, + "loss": 0.1993, + "step": 900 + }, + { + "epoch": 0.3445506692160612, + "grad_norm": 5.069430658098168, + "learning_rate": 9.480579635594347e-07, + "loss": 0.6176, + "step": 901 + }, + { + "epoch": 0.34493307839388143, + "grad_norm": 4.224126032437366, + "learning_rate": 9.48212537571454e-07, + "loss": 0.651, + "step": 902 + }, + { + "epoch": 0.3453154875717017, + "grad_norm": 4.250506651354976, + "learning_rate": 9.483669403102948e-07, + "loss": 0.5044, + "step": 903 + }, + { + "epoch": 0.345697896749522, + "grad_norm": 2.465775931686004, + "learning_rate": 9.485211721550897e-07, + "loss": 0.3392, + "step": 904 + }, + { + "epoch": 0.34608030592734224, + "grad_norm": 2.3975292561230748, + "learning_rate": 9.48675233483714e-07, + "loss": 0.3344, + "step": 905 + }, + { + "epoch": 0.3464627151051625, + "grad_norm": 3.0526375294539094, + "learning_rate": 9.488291246727914e-07, + "loss": 0.4743, + "step": 906 + }, + { + "epoch": 0.3468451242829828, + "grad_norm": 2.336490425323715, + "learning_rate": 9.489828460976984e-07, + "loss": 0.5388, + "step": 907 + }, + { + "epoch": 0.34722753346080304, + "grad_norm": 2.5558834717862426, + "learning_rate": 9.491363981325713e-07, + "loss": 0.2754, + "step": 908 + }, + { + "epoch": 0.3476099426386233, + "grad_norm": 4.177412950760512, + "learning_rate": 9.492897811503105e-07, + "loss": 0.3361, + "step": 909 + }, + { + "epoch": 0.3479923518164436, + "grad_norm": 4.640426888395395, + "learning_rate": 9.494429955225867e-07, + "loss": 0.4094, + "step": 910 + }, + { + "epoch": 0.34837476099426384, + "grad_norm": 3.038139771953519, + "learning_rate": 9.495960416198456e-07, + "loss": 0.1049, + "step": 911 + }, + { + "epoch": 0.34875717017208413, + "grad_norm": 3.971266181679893, + "learning_rate": 9.497489198113142e-07, + "loss": 0.1444, + "step": 912 + }, + { + "epoch": 0.3491395793499044, + "grad_norm": 5.104859949894045, + "learning_rate": 9.499016304650052e-07, + "loss": 0.3623, + "step": 913 + }, + { + "epoch": 0.34952198852772465, + "grad_norm": 6.547232752382276, + "learning_rate": 9.500541739477227e-07, + "loss": 0.3693, + "step": 914 + }, + { + "epoch": 0.34990439770554493, + "grad_norm": 3.6828872796074594, + "learning_rate": 9.502065506250678e-07, + "loss": 0.5371, + "step": 915 + }, + { + "epoch": 0.3502868068833652, + "grad_norm": 3.6765337020842113, + "learning_rate": 9.503587608614433e-07, + "loss": 0.6631, + "step": 916 + }, + { + "epoch": 0.35066921606118545, + "grad_norm": 2.8021830903817624, + "learning_rate": 9.505108050200593e-07, + "loss": 0.3992, + "step": 917 + }, + { + "epoch": 0.35105162523900574, + "grad_norm": 3.703896615161285, + "learning_rate": 9.506626834629382e-07, + "loss": 0.5016, + "step": 918 + }, + { + "epoch": 0.351434034416826, + "grad_norm": 3.866072430385559, + "learning_rate": 9.508143965509201e-07, + "loss": 0.4969, + "step": 919 + }, + { + "epoch": 0.35181644359464626, + "grad_norm": 2.872815172718489, + "learning_rate": 9.509659446436679e-07, + "loss": 0.5259, + "step": 920 + }, + { + "epoch": 0.35219885277246654, + "grad_norm": 2.5012410897968893, + "learning_rate": 9.511173280996717e-07, + "loss": 0.3576, + "step": 921 + }, + { + "epoch": 0.35258126195028683, + "grad_norm": 4.912924727310679, + "learning_rate": 9.512685472762555e-07, + "loss": 0.2694, + "step": 922 + }, + { + "epoch": 0.35296367112810706, + "grad_norm": 3.868625812663427, + "learning_rate": 9.514196025295804e-07, + "loss": 0.1522, + "step": 923 + }, + { + "epoch": 0.35334608030592735, + "grad_norm": 3.2232447971096962, + "learning_rate": 9.515704942146511e-07, + "loss": 0.0933, + "step": 924 + }, + { + "epoch": 0.35372848948374763, + "grad_norm": 4.741836962242508, + "learning_rate": 9.5172122268532e-07, + "loss": 0.199, + "step": 925 + }, + { + "epoch": 0.35411089866156786, + "grad_norm": 11.242952979498702, + "learning_rate": 9.518717882942924e-07, + "loss": 0.3926, + "step": 926 + }, + { + "epoch": 0.35449330783938815, + "grad_norm": 3.748092420499065, + "learning_rate": 9.520221913931318e-07, + "loss": 0.3953, + "step": 927 + }, + { + "epoch": 0.35487571701720844, + "grad_norm": 2.366885628667578, + "learning_rate": 9.521724323322646e-07, + "loss": 0.574, + "step": 928 + }, + { + "epoch": 0.35525812619502867, + "grad_norm": 3.8608011552795505, + "learning_rate": 9.523225114609844e-07, + "loss": 0.3235, + "step": 929 + }, + { + "epoch": 0.35564053537284895, + "grad_norm": 3.9321756846858174, + "learning_rate": 9.524724291274583e-07, + "loss": 0.4958, + "step": 930 + }, + { + "epoch": 0.35602294455066924, + "grad_norm": 3.548843845632833, + "learning_rate": 9.526221856787305e-07, + "loss": 0.2713, + "step": 931 + }, + { + "epoch": 0.35640535372848947, + "grad_norm": 2.545063146620941, + "learning_rate": 9.527717814607271e-07, + "loss": 0.3894, + "step": 932 + }, + { + "epoch": 0.35678776290630976, + "grad_norm": 2.8256888672163294, + "learning_rate": 9.529212168182619e-07, + "loss": 0.3203, + "step": 933 + }, + { + "epoch": 0.35717017208413004, + "grad_norm": 3.8043380670836955, + "learning_rate": 9.530704920950403e-07, + "loss": 0.2826, + "step": 934 + }, + { + "epoch": 0.3575525812619503, + "grad_norm": 3.1976715557254938, + "learning_rate": 9.532196076336643e-07, + "loss": 0.2002, + "step": 935 + }, + { + "epoch": 0.35793499043977056, + "grad_norm": 4.822091588850192, + "learning_rate": 9.533685637756371e-07, + "loss": 0.3646, + "step": 936 + }, + { + "epoch": 0.35831739961759085, + "grad_norm": 6.43557710539182, + "learning_rate": 9.535173608613685e-07, + "loss": 0.1462, + "step": 937 + }, + { + "epoch": 0.3586998087954111, + "grad_norm": 5.112594663307877, + "learning_rate": 9.536659992301782e-07, + "loss": 0.3242, + "step": 938 + }, + { + "epoch": 0.35908221797323137, + "grad_norm": 4.220278314750709, + "learning_rate": 9.538144792203018e-07, + "loss": 0.4659, + "step": 939 + }, + { + "epoch": 0.35946462715105165, + "grad_norm": 3.2606053079423964, + "learning_rate": 9.539628011688945e-07, + "loss": 0.623, + "step": 940 + }, + { + "epoch": 0.3598470363288719, + "grad_norm": 2.972337021398215, + "learning_rate": 9.541109654120364e-07, + "loss": 0.4305, + "step": 941 + }, + { + "epoch": 0.36022944550669217, + "grad_norm": 3.5581539481074476, + "learning_rate": 9.542589722847364e-07, + "loss": 0.559, + "step": 942 + }, + { + "epoch": 0.3606118546845124, + "grad_norm": 4.023866500945532, + "learning_rate": 9.544068221209373e-07, + "loss": 0.3209, + "step": 943 + }, + { + "epoch": 0.3609942638623327, + "grad_norm": 2.35951869097576, + "learning_rate": 9.5455451525352e-07, + "loss": 0.4073, + "step": 944 + }, + { + "epoch": 0.361376673040153, + "grad_norm": 4.300513152619637, + "learning_rate": 9.547020520143076e-07, + "loss": 0.2415, + "step": 945 + }, + { + "epoch": 0.3617590822179732, + "grad_norm": 2.523667835789399, + "learning_rate": 9.548494327340713e-07, + "loss": 0.2748, + "step": 946 + }, + { + "epoch": 0.3621414913957935, + "grad_norm": 3.440650190075816, + "learning_rate": 9.54996657742533e-07, + "loss": 0.3607, + "step": 947 + }, + { + "epoch": 0.3625239005736138, + "grad_norm": 3.994097885238933, + "learning_rate": 9.551437273683711e-07, + "loss": 0.276, + "step": 948 + }, + { + "epoch": 0.362906309751434, + "grad_norm": 3.940188141218107, + "learning_rate": 9.552906419392247e-07, + "loss": 0.1379, + "step": 949 + }, + { + "epoch": 0.3632887189292543, + "grad_norm": 4.636109489642276, + "learning_rate": 9.554374017816967e-07, + "loss": 0.2433, + "step": 950 + }, + { + "epoch": 0.3636711281070746, + "grad_norm": 7.117983659417351, + "learning_rate": 9.555840072213603e-07, + "loss": 0.515, + "step": 951 + }, + { + "epoch": 0.3640535372848948, + "grad_norm": 4.1641341842006385, + "learning_rate": 9.557304585827614e-07, + "loss": 0.6473, + "step": 952 + }, + { + "epoch": 0.3644359464627151, + "grad_norm": 3.9707883871656287, + "learning_rate": 9.55876756189424e-07, + "loss": 0.6406, + "step": 953 + }, + { + "epoch": 0.3648183556405354, + "grad_norm": 3.0221837324920293, + "learning_rate": 9.560229003638547e-07, + "loss": 0.4578, + "step": 954 + }, + { + "epoch": 0.3652007648183556, + "grad_norm": 2.78359928992419, + "learning_rate": 9.561688914275451e-07, + "loss": 0.4396, + "step": 955 + }, + { + "epoch": 0.3655831739961759, + "grad_norm": 3.451375941237979, + "learning_rate": 9.56314729700979e-07, + "loss": 0.5369, + "step": 956 + }, + { + "epoch": 0.3659655831739962, + "grad_norm": 4.276913193584142, + "learning_rate": 9.56460415503634e-07, + "loss": 0.3295, + "step": 957 + }, + { + "epoch": 0.3663479923518164, + "grad_norm": 3.4730854605922308, + "learning_rate": 9.56605949153987e-07, + "loss": 0.4103, + "step": 958 + }, + { + "epoch": 0.3667304015296367, + "grad_norm": 4.399869153938256, + "learning_rate": 9.56751330969518e-07, + "loss": 0.24, + "step": 959 + }, + { + "epoch": 0.367112810707457, + "grad_norm": 3.7183967230189814, + "learning_rate": 9.568965612667146e-07, + "loss": 0.2586, + "step": 960 + }, + { + "epoch": 0.3674952198852772, + "grad_norm": 3.0025703227552922, + "learning_rate": 9.57041640361076e-07, + "loss": 0.1275, + "step": 961 + }, + { + "epoch": 0.3678776290630975, + "grad_norm": 2.641218130097246, + "learning_rate": 9.571865685671162e-07, + "loss": 0.1138, + "step": 962 + }, + { + "epoch": 0.3682600382409178, + "grad_norm": 6.87125641596111, + "learning_rate": 9.5733134619837e-07, + "loss": 0.3016, + "step": 963 + }, + { + "epoch": 0.36864244741873803, + "grad_norm": 2.9957571226002058, + "learning_rate": 9.574759735673949e-07, + "loss": 0.6492, + "step": 964 + }, + { + "epoch": 0.3690248565965583, + "grad_norm": 2.803405723846682, + "learning_rate": 9.576204509857772e-07, + "loss": 0.529, + "step": 965 + }, + { + "epoch": 0.3694072657743786, + "grad_norm": 3.5725217699224485, + "learning_rate": 9.577647787641344e-07, + "loss": 0.5717, + "step": 966 + }, + { + "epoch": 0.36978967495219883, + "grad_norm": 3.079030324956933, + "learning_rate": 9.5790895721212e-07, + "loss": 0.2471, + "step": 967 + }, + { + "epoch": 0.3701720841300191, + "grad_norm": 3.102122988851968, + "learning_rate": 9.580529866384277e-07, + "loss": 0.2238, + "step": 968 + }, + { + "epoch": 0.3705544933078394, + "grad_norm": 2.5352702298867085, + "learning_rate": 9.58196867350794e-07, + "loss": 0.2231, + "step": 969 + }, + { + "epoch": 0.37093690248565964, + "grad_norm": 3.6590706142420597, + "learning_rate": 9.583405996560044e-07, + "loss": 0.4481, + "step": 970 + }, + { + "epoch": 0.3713193116634799, + "grad_norm": 2.4556048062747413, + "learning_rate": 9.584841838598948e-07, + "loss": 0.3815, + "step": 971 + }, + { + "epoch": 0.3717017208413002, + "grad_norm": 3.831446860702458, + "learning_rate": 9.586276202673582e-07, + "loss": 0.1587, + "step": 972 + }, + { + "epoch": 0.37208413001912044, + "grad_norm": 4.014302670550158, + "learning_rate": 9.587709091823455e-07, + "loss": 0.2652, + "step": 973 + }, + { + "epoch": 0.3724665391969407, + "grad_norm": 4.001061298719788, + "learning_rate": 9.589140509078721e-07, + "loss": 0.1145, + "step": 974 + }, + { + "epoch": 0.372848948374761, + "grad_norm": 5.591551790985613, + "learning_rate": 9.590570457460196e-07, + "loss": 0.1634, + "step": 975 + }, + { + "epoch": 0.37323135755258124, + "grad_norm": 7.88177658850859, + "learning_rate": 9.591998939979414e-07, + "loss": 0.5488, + "step": 976 + }, + { + "epoch": 0.37361376673040153, + "grad_norm": 3.544611944096069, + "learning_rate": 9.593425959638656e-07, + "loss": 0.564, + "step": 977 + }, + { + "epoch": 0.3739961759082218, + "grad_norm": 3.1957798548422, + "learning_rate": 9.594851519430982e-07, + "loss": 0.5497, + "step": 978 + }, + { + "epoch": 0.37437858508604205, + "grad_norm": 2.273287109434571, + "learning_rate": 9.596275622340283e-07, + "loss": 0.4016, + "step": 979 + }, + { + "epoch": 0.37476099426386233, + "grad_norm": 5.093270308485771, + "learning_rate": 9.597698271341309e-07, + "loss": 0.3785, + "step": 980 + }, + { + "epoch": 0.3751434034416826, + "grad_norm": 3.8876978882103375, + "learning_rate": 9.599119469399705e-07, + "loss": 0.38, + "step": 981 + }, + { + "epoch": 0.37552581261950285, + "grad_norm": 3.890440051726995, + "learning_rate": 9.600539219472053e-07, + "loss": 0.3075, + "step": 982 + }, + { + "epoch": 0.37590822179732314, + "grad_norm": 3.4391686290984627, + "learning_rate": 9.60195752450591e-07, + "loss": 0.5038, + "step": 983 + }, + { + "epoch": 0.3762906309751434, + "grad_norm": 5.500985837495509, + "learning_rate": 9.603374387439842e-07, + "loss": 0.3613, + "step": 984 + }, + { + "epoch": 0.37667304015296366, + "grad_norm": 5.084852064713341, + "learning_rate": 9.604789811203456e-07, + "loss": 0.153, + "step": 985 + }, + { + "epoch": 0.37705544933078394, + "grad_norm": 3.5089018559660863, + "learning_rate": 9.606203798717442e-07, + "loss": 0.2237, + "step": 986 + }, + { + "epoch": 0.37743785850860423, + "grad_norm": 5.30939060543432, + "learning_rate": 9.607616352893613e-07, + "loss": 0.304, + "step": 987 + }, + { + "epoch": 0.37782026768642446, + "grad_norm": 10.545777214279783, + "learning_rate": 9.60902747663493e-07, + "loss": 0.3001, + "step": 988 + }, + { + "epoch": 0.37820267686424475, + "grad_norm": 3.1998032628617197, + "learning_rate": 9.610437172835547e-07, + "loss": 0.4488, + "step": 989 + }, + { + "epoch": 0.37858508604206503, + "grad_norm": 3.828534038255384, + "learning_rate": 9.611845444380842e-07, + "loss": 0.5062, + "step": 990 + }, + { + "epoch": 0.37896749521988526, + "grad_norm": 3.1512350900517463, + "learning_rate": 9.613252294147454e-07, + "loss": 0.4848, + "step": 991 + }, + { + "epoch": 0.37934990439770555, + "grad_norm": 2.554872465291332, + "learning_rate": 9.61465772500332e-07, + "loss": 0.4056, + "step": 992 + }, + { + "epoch": 0.37973231357552584, + "grad_norm": 2.996642777372816, + "learning_rate": 9.616061739807709e-07, + "loss": 0.6107, + "step": 993 + }, + { + "epoch": 0.38011472275334607, + "grad_norm": 2.9023946806391008, + "learning_rate": 9.617464341411247e-07, + "loss": 0.3574, + "step": 994 + }, + { + "epoch": 0.38049713193116635, + "grad_norm": 2.5186035340842863, + "learning_rate": 9.618865532655972e-07, + "loss": 0.3863, + "step": 995 + }, + { + "epoch": 0.38087954110898664, + "grad_norm": 2.5047887713828496, + "learning_rate": 9.620265316375355e-07, + "loss": 0.342, + "step": 996 + }, + { + "epoch": 0.38126195028680687, + "grad_norm": 4.023505526958285, + "learning_rate": 9.621663695394329e-07, + "loss": 0.1651, + "step": 997 + }, + { + "epoch": 0.38164435946462716, + "grad_norm": 2.9972760231516817, + "learning_rate": 9.623060672529342e-07, + "loss": 0.1481, + "step": 998 + }, + { + "epoch": 0.38202676864244745, + "grad_norm": 3.7625935489012625, + "learning_rate": 9.624456250588371e-07, + "loss": 0.1253, + "step": 999 + }, + { + "epoch": 0.3824091778202677, + "grad_norm": 4.143639960345454, + "learning_rate": 9.625850432370972e-07, + "loss": 0.2301, + "step": 1000 + }, + { + "epoch": 0.3824091778202677, + "eval_runtime": 822.5233, + "eval_samples_per_second": 1.865, + "eval_steps_per_second": 0.467, + "step": 1000 + }, + { + "epoch": 0.38279158699808796, + "grad_norm": 13.168096794653332, + "learning_rate": 9.627243220668299e-07, + "loss": 0.3114, + "step": 1001 + }, + { + "epoch": 0.38317399617590825, + "grad_norm": 4.6670873089084814, + "learning_rate": 9.62863461826315e-07, + "loss": 0.5301, + "step": 1002 + }, + { + "epoch": 0.3835564053537285, + "grad_norm": 3.4543609478607573, + "learning_rate": 9.630024627929996e-07, + "loss": 0.445, + "step": 1003 + }, + { + "epoch": 0.38393881453154877, + "grad_norm": 2.9601279890896595, + "learning_rate": 9.631413252435012e-07, + "loss": 0.4941, + "step": 1004 + }, + { + "epoch": 0.384321223709369, + "grad_norm": 4.258867325618021, + "learning_rate": 9.632800494536112e-07, + "loss": 0.4981, + "step": 1005 + }, + { + "epoch": 0.3847036328871893, + "grad_norm": 2.933275959869867, + "learning_rate": 9.634186356982979e-07, + "loss": 0.3513, + "step": 1006 + }, + { + "epoch": 0.38508604206500957, + "grad_norm": 2.9809046835212243, + "learning_rate": 9.635570842517104e-07, + "loss": 0.4706, + "step": 1007 + }, + { + "epoch": 0.3854684512428298, + "grad_norm": 2.8025353975808414, + "learning_rate": 9.636953953871814e-07, + "loss": 0.2774, + "step": 1008 + }, + { + "epoch": 0.3858508604206501, + "grad_norm": 3.7234322698750906, + "learning_rate": 9.638335693772303e-07, + "loss": 0.3433, + "step": 1009 + }, + { + "epoch": 0.3862332695984704, + "grad_norm": 3.5164420979708093, + "learning_rate": 9.639716064935667e-07, + "loss": 0.3517, + "step": 1010 + }, + { + "epoch": 0.3866156787762906, + "grad_norm": 3.5137337577930254, + "learning_rate": 9.641095070070938e-07, + "loss": 0.0892, + "step": 1011 + }, + { + "epoch": 0.3869980879541109, + "grad_norm": 2.5491696183768577, + "learning_rate": 9.64247271187911e-07, + "loss": 0.1025, + "step": 1012 + }, + { + "epoch": 0.3873804971319312, + "grad_norm": 5.791968724266462, + "learning_rate": 9.643848993053177e-07, + "loss": 0.4488, + "step": 1013 + }, + { + "epoch": 0.3877629063097514, + "grad_norm": 4.08149153504387, + "learning_rate": 9.64522391627816e-07, + "loss": 0.7407, + "step": 1014 + }, + { + "epoch": 0.3881453154875717, + "grad_norm": 2.888440435602648, + "learning_rate": 9.646597484231137e-07, + "loss": 0.3919, + "step": 1015 + }, + { + "epoch": 0.388527724665392, + "grad_norm": 2.506209672332284, + "learning_rate": 9.647969699581283e-07, + "loss": 0.3635, + "step": 1016 + }, + { + "epoch": 0.3889101338432122, + "grad_norm": 4.320370554749521, + "learning_rate": 9.649340564989894e-07, + "loss": 0.4528, + "step": 1017 + }, + { + "epoch": 0.3892925430210325, + "grad_norm": 3.345833672839789, + "learning_rate": 9.650710083110417e-07, + "loss": 0.3671, + "step": 1018 + }, + { + "epoch": 0.3896749521988528, + "grad_norm": 3.1623782349315905, + "learning_rate": 9.652078256588485e-07, + "loss": 0.3302, + "step": 1019 + }, + { + "epoch": 0.390057361376673, + "grad_norm": 2.825148728768212, + "learning_rate": 9.653445088061943e-07, + "loss": 0.3511, + "step": 1020 + }, + { + "epoch": 0.3904397705544933, + "grad_norm": 3.1126792641610357, + "learning_rate": 9.654810580160888e-07, + "loss": 0.3865, + "step": 1021 + }, + { + "epoch": 0.3908221797323136, + "grad_norm": 3.0898730714277214, + "learning_rate": 9.656174735507689e-07, + "loss": 0.1243, + "step": 1022 + }, + { + "epoch": 0.3912045889101338, + "grad_norm": 3.4606426449243703, + "learning_rate": 9.657537556717016e-07, + "loss": 0.1593, + "step": 1023 + }, + { + "epoch": 0.3915869980879541, + "grad_norm": 3.169414599544937, + "learning_rate": 9.658899046395884e-07, + "loss": 0.1735, + "step": 1024 + }, + { + "epoch": 0.3919694072657744, + "grad_norm": 3.807031179859044, + "learning_rate": 9.660259207143668e-07, + "loss": 0.2093, + "step": 1025 + }, + { + "epoch": 0.3923518164435946, + "grad_norm": 7.122380628459166, + "learning_rate": 9.66161804155214e-07, + "loss": 0.4193, + "step": 1026 + }, + { + "epoch": 0.3927342256214149, + "grad_norm": 3.9833003969839265, + "learning_rate": 9.6629755522055e-07, + "loss": 0.7618, + "step": 1027 + }, + { + "epoch": 0.3931166347992352, + "grad_norm": 3.7067925041665486, + "learning_rate": 9.664331741680397e-07, + "loss": 0.3657, + "step": 1028 + }, + { + "epoch": 0.39349904397705543, + "grad_norm": 3.1871963057109762, + "learning_rate": 9.665686612545975e-07, + "loss": 0.4019, + "step": 1029 + }, + { + "epoch": 0.3938814531548757, + "grad_norm": 3.2304683071814235, + "learning_rate": 9.66704016736388e-07, + "loss": 0.5361, + "step": 1030 + }, + { + "epoch": 0.394263862332696, + "grad_norm": 5.435379491785883, + "learning_rate": 9.668392408688305e-07, + "loss": 0.3405, + "step": 1031 + }, + { + "epoch": 0.39464627151051623, + "grad_norm": 2.7119694485364176, + "learning_rate": 9.669743339066014e-07, + "loss": 0.368, + "step": 1032 + }, + { + "epoch": 0.3950286806883365, + "grad_norm": 2.539139121440049, + "learning_rate": 9.671092961036375e-07, + "loss": 0.2497, + "step": 1033 + }, + { + "epoch": 0.3954110898661568, + "grad_norm": 5.451613252184311, + "learning_rate": 9.672441277131377e-07, + "loss": 0.3053, + "step": 1034 + }, + { + "epoch": 0.39579349904397704, + "grad_norm": 3.4615202471262028, + "learning_rate": 9.673788289875675e-07, + "loss": 0.2826, + "step": 1035 + }, + { + "epoch": 0.3961759082217973, + "grad_norm": 3.0198202253984565, + "learning_rate": 9.675134001786604e-07, + "loss": 0.1403, + "step": 1036 + }, + { + "epoch": 0.3965583173996176, + "grad_norm": 3.1099402139197587, + "learning_rate": 9.676478415374213e-07, + "loss": 0.1186, + "step": 1037 + }, + { + "epoch": 0.39694072657743784, + "grad_norm": 6.277490506121325, + "learning_rate": 9.677821533141296e-07, + "loss": 0.3295, + "step": 1038 + }, + { + "epoch": 0.39732313575525813, + "grad_norm": 4.740829851606653, + "learning_rate": 9.679163357583416e-07, + "loss": 0.6511, + "step": 1039 + }, + { + "epoch": 0.3977055449330784, + "grad_norm": 3.8249444945054254, + "learning_rate": 9.680503891188934e-07, + "loss": 0.4757, + "step": 1040 + }, + { + "epoch": 0.39808795411089865, + "grad_norm": 4.29025279831927, + "learning_rate": 9.681843136439033e-07, + "loss": 0.4677, + "step": 1041 + }, + { + "epoch": 0.39847036328871893, + "grad_norm": 3.8375520435494663, + "learning_rate": 9.683181095807752e-07, + "loss": 0.4587, + "step": 1042 + }, + { + "epoch": 0.3988527724665392, + "grad_norm": 3.2469451926081643, + "learning_rate": 9.684517771762013e-07, + "loss": 0.3822, + "step": 1043 + }, + { + "epoch": 0.39923518164435945, + "grad_norm": 2.8476832091595097, + "learning_rate": 9.68585316676164e-07, + "loss": 0.4699, + "step": 1044 + }, + { + "epoch": 0.39961759082217974, + "grad_norm": 2.4952240146529494, + "learning_rate": 9.687187283259399e-07, + "loss": 0.3083, + "step": 1045 + }, + { + "epoch": 0.4, + "grad_norm": 3.6329542563068378, + "learning_rate": 9.688520123701012e-07, + "loss": 0.2969, + "step": 1046 + }, + { + "epoch": 0.40038240917782025, + "grad_norm": 3.894775382973128, + "learning_rate": 9.689851690525188e-07, + "loss": 0.1538, + "step": 1047 + }, + { + "epoch": 0.40076481835564054, + "grad_norm": 3.796723288185588, + "learning_rate": 9.69118198616366e-07, + "loss": 0.1505, + "step": 1048 + }, + { + "epoch": 0.4011472275334608, + "grad_norm": 3.1142443384194247, + "learning_rate": 9.692511013041197e-07, + "loss": 0.0952, + "step": 1049 + }, + { + "epoch": 0.40152963671128106, + "grad_norm": 5.055412323055709, + "learning_rate": 9.693838773575638e-07, + "loss": 0.3218, + "step": 1050 + }, + { + "epoch": 0.40191204588910134, + "grad_norm": 6.485575296006208, + "learning_rate": 9.695165270177918e-07, + "loss": 0.3999, + "step": 1051 + }, + { + "epoch": 0.40229445506692163, + "grad_norm": 2.793858441119319, + "learning_rate": 9.696490505252096e-07, + "loss": 0.2352, + "step": 1052 + }, + { + "epoch": 0.40267686424474186, + "grad_norm": 4.063080496762148, + "learning_rate": 9.69781448119537e-07, + "loss": 0.5957, + "step": 1053 + }, + { + "epoch": 0.40305927342256215, + "grad_norm": 1.6902282700627846, + "learning_rate": 9.699137200398118e-07, + "loss": 0.4523, + "step": 1054 + }, + { + "epoch": 0.40344168260038243, + "grad_norm": 3.47420780290818, + "learning_rate": 9.700458665243921e-07, + "loss": 0.4985, + "step": 1055 + }, + { + "epoch": 0.40382409177820267, + "grad_norm": 5.289364692022554, + "learning_rate": 9.701778878109578e-07, + "loss": 0.6845, + "step": 1056 + }, + { + "epoch": 0.40420650095602295, + "grad_norm": 2.7646218759762546, + "learning_rate": 9.703097841365144e-07, + "loss": 0.5403, + "step": 1057 + }, + { + "epoch": 0.40458891013384324, + "grad_norm": 3.1681689329670193, + "learning_rate": 9.704415557373944e-07, + "loss": 0.1867, + "step": 1058 + }, + { + "epoch": 0.40497131931166347, + "grad_norm": 3.386309571500823, + "learning_rate": 9.705732028492613e-07, + "loss": 0.174, + "step": 1059 + }, + { + "epoch": 0.40535372848948376, + "grad_norm": 4.397510274991105, + "learning_rate": 9.70704725707111e-07, + "loss": 0.2867, + "step": 1060 + }, + { + "epoch": 0.40573613766730404, + "grad_norm": 3.4382151466677286, + "learning_rate": 9.70836124545274e-07, + "loss": 0.1454, + "step": 1061 + }, + { + "epoch": 0.4061185468451243, + "grad_norm": 4.147605683289184, + "learning_rate": 9.709673995974196e-07, + "loss": 0.166, + "step": 1062 + }, + { + "epoch": 0.40650095602294456, + "grad_norm": 5.851581079430617, + "learning_rate": 9.710985510965567e-07, + "loss": 0.442, + "step": 1063 + }, + { + "epoch": 0.40688336520076485, + "grad_norm": 5.406232006947117, + "learning_rate": 9.71229579275037e-07, + "loss": 0.3363, + "step": 1064 + }, + { + "epoch": 0.4072657743785851, + "grad_norm": 3.3648067120705054, + "learning_rate": 9.713604843645576e-07, + "loss": 0.4594, + "step": 1065 + }, + { + "epoch": 0.40764818355640536, + "grad_norm": 2.5631868325853686, + "learning_rate": 9.71491266596163e-07, + "loss": 0.3384, + "step": 1066 + }, + { + "epoch": 0.4080305927342256, + "grad_norm": 3.177223994575539, + "learning_rate": 9.716219262002475e-07, + "loss": 0.2959, + "step": 1067 + }, + { + "epoch": 0.4084130019120459, + "grad_norm": 3.2975564406093048, + "learning_rate": 9.717524634065587e-07, + "loss": 0.259, + "step": 1068 + }, + { + "epoch": 0.40879541108986617, + "grad_norm": 3.528287201659267, + "learning_rate": 9.718828784441982e-07, + "loss": 0.3725, + "step": 1069 + }, + { + "epoch": 0.4091778202676864, + "grad_norm": 2.679782343039616, + "learning_rate": 9.72013171541626e-07, + "loss": 0.3988, + "step": 1070 + }, + { + "epoch": 0.4095602294455067, + "grad_norm": 2.4261372131515095, + "learning_rate": 9.721433429266611e-07, + "loss": 0.3528, + "step": 1071 + }, + { + "epoch": 0.40994263862332697, + "grad_norm": 3.072739280279131, + "learning_rate": 9.722733928264848e-07, + "loss": 0.2398, + "step": 1072 + }, + { + "epoch": 0.4103250478011472, + "grad_norm": 3.8742556933860697, + "learning_rate": 9.724033214676432e-07, + "loss": 0.3672, + "step": 1073 + }, + { + "epoch": 0.4107074569789675, + "grad_norm": 3.8825253662538373, + "learning_rate": 9.72533129076049e-07, + "loss": 0.2031, + "step": 1074 + }, + { + "epoch": 0.4110898661567878, + "grad_norm": 4.078706780221329, + "learning_rate": 9.72662815876984e-07, + "loss": 0.141, + "step": 1075 + }, + { + "epoch": 0.411472275334608, + "grad_norm": 5.947445861390411, + "learning_rate": 9.72792382095102e-07, + "loss": 0.5128, + "step": 1076 + }, + { + "epoch": 0.4118546845124283, + "grad_norm": 4.128468991082184, + "learning_rate": 9.729218279544307e-07, + "loss": 0.3957, + "step": 1077 + }, + { + "epoch": 0.4122370936902486, + "grad_norm": 3.9534924000929577, + "learning_rate": 9.73051153678374e-07, + "loss": 0.5413, + "step": 1078 + }, + { + "epoch": 0.4126195028680688, + "grad_norm": 2.126150605100412, + "learning_rate": 9.73180359489714e-07, + "loss": 0.3293, + "step": 1079 + }, + { + "epoch": 0.4130019120458891, + "grad_norm": 3.3000665195580985, + "learning_rate": 9.733094456106143e-07, + "loss": 0.5077, + "step": 1080 + }, + { + "epoch": 0.4133843212237094, + "grad_norm": 4.306372670944154, + "learning_rate": 9.734384122626212e-07, + "loss": 0.5619, + "step": 1081 + }, + { + "epoch": 0.4137667304015296, + "grad_norm": 2.64362007344908, + "learning_rate": 9.735672596666664e-07, + "loss": 0.4506, + "step": 1082 + }, + { + "epoch": 0.4141491395793499, + "grad_norm": 2.485214763241518, + "learning_rate": 9.736959880430697e-07, + "loss": 0.277, + "step": 1083 + }, + { + "epoch": 0.4145315487571702, + "grad_norm": 3.7446658025324835, + "learning_rate": 9.738245976115407e-07, + "loss": 0.2825, + "step": 1084 + }, + { + "epoch": 0.4149139579349904, + "grad_norm": 2.290043435163736, + "learning_rate": 9.739530885911813e-07, + "loss": 0.1555, + "step": 1085 + }, + { + "epoch": 0.4152963671128107, + "grad_norm": 3.027236011187704, + "learning_rate": 9.740814612004875e-07, + "loss": 0.2064, + "step": 1086 + }, + { + "epoch": 0.415678776290631, + "grad_norm": 3.660505117362001, + "learning_rate": 9.742097156573525e-07, + "loss": 0.1534, + "step": 1087 + }, + { + "epoch": 0.4160611854684512, + "grad_norm": 5.677949369209507, + "learning_rate": 9.74337852179068e-07, + "loss": 0.1978, + "step": 1088 + }, + { + "epoch": 0.4164435946462715, + "grad_norm": 7.4228592069466295, + "learning_rate": 9.744658709823274e-07, + "loss": 0.6633, + "step": 1089 + }, + { + "epoch": 0.4168260038240918, + "grad_norm": 4.292387349378489, + "learning_rate": 9.745937722832267e-07, + "loss": 0.4571, + "step": 1090 + }, + { + "epoch": 0.417208413001912, + "grad_norm": 2.427752641285365, + "learning_rate": 9.747215562972675e-07, + "loss": 0.3619, + "step": 1091 + }, + { + "epoch": 0.4175908221797323, + "grad_norm": 2.5606948262759954, + "learning_rate": 9.7484922323936e-07, + "loss": 0.635, + "step": 1092 + }, + { + "epoch": 0.4179732313575526, + "grad_norm": 3.664221266049891, + "learning_rate": 9.749767733238231e-07, + "loss": 0.5136, + "step": 1093 + }, + { + "epoch": 0.41835564053537283, + "grad_norm": 5.36261973569705, + "learning_rate": 9.751042067643886e-07, + "loss": 0.6432, + "step": 1094 + }, + { + "epoch": 0.4187380497131931, + "grad_norm": 2.8878135411838866, + "learning_rate": 9.752315237742018e-07, + "loss": 0.404, + "step": 1095 + }, + { + "epoch": 0.4191204588910134, + "grad_norm": 2.406785682401822, + "learning_rate": 9.753587245658248e-07, + "loss": 0.2421, + "step": 1096 + }, + { + "epoch": 0.41950286806883363, + "grad_norm": 2.4421919634113785, + "learning_rate": 9.754858093512376e-07, + "loss": 0.3162, + "step": 1097 + }, + { + "epoch": 0.4198852772466539, + "grad_norm": 3.9177153515856413, + "learning_rate": 9.756127783418411e-07, + "loss": 0.5863, + "step": 1098 + }, + { + "epoch": 0.4202676864244742, + "grad_norm": 3.9391362257114078, + "learning_rate": 9.757396317484594e-07, + "loss": 0.142, + "step": 1099 + }, + { + "epoch": 0.42065009560229444, + "grad_norm": 4.385574205095433, + "learning_rate": 9.758663697813404e-07, + "loss": 0.2256, + "step": 1100 + }, + { + "epoch": 0.4210325047801147, + "grad_norm": 5.305099538719576, + "learning_rate": 9.75992992650159e-07, + "loss": 0.6464, + "step": 1101 + }, + { + "epoch": 0.421414913957935, + "grad_norm": 4.7650198130080605, + "learning_rate": 9.761195005640198e-07, + "loss": 0.5413, + "step": 1102 + }, + { + "epoch": 0.42179732313575524, + "grad_norm": 3.604277078043537, + "learning_rate": 9.762458937314578e-07, + "loss": 0.3032, + "step": 1103 + }, + { + "epoch": 0.42217973231357553, + "grad_norm": 2.8768596439751475, + "learning_rate": 9.76372172360441e-07, + "loss": 0.4792, + "step": 1104 + }, + { + "epoch": 0.4225621414913958, + "grad_norm": 4.166289630361482, + "learning_rate": 9.76498336658373e-07, + "loss": 0.3036, + "step": 1105 + }, + { + "epoch": 0.42294455066921605, + "grad_norm": 3.4417821395258987, + "learning_rate": 9.76624386832094e-07, + "loss": 0.4794, + "step": 1106 + }, + { + "epoch": 0.42332695984703633, + "grad_norm": 6.911003306707519, + "learning_rate": 9.76750323087884e-07, + "loss": 0.2857, + "step": 1107 + }, + { + "epoch": 0.4237093690248566, + "grad_norm": 4.189974836396834, + "learning_rate": 9.768761456314634e-07, + "loss": 0.3152, + "step": 1108 + }, + { + "epoch": 0.42409177820267685, + "grad_norm": 3.0886379353453357, + "learning_rate": 9.770018546679967e-07, + "loss": 0.2533, + "step": 1109 + }, + { + "epoch": 0.42447418738049714, + "grad_norm": 4.58699649885338, + "learning_rate": 9.771274504020933e-07, + "loss": 0.3245, + "step": 1110 + }, + { + "epoch": 0.4248565965583174, + "grad_norm": 4.591641934992285, + "learning_rate": 9.772529330378098e-07, + "loss": 0.2323, + "step": 1111 + }, + { + "epoch": 0.42523900573613765, + "grad_norm": 5.879606672872391, + "learning_rate": 9.773783027786522e-07, + "loss": 0.1734, + "step": 1112 + }, + { + "epoch": 0.42562141491395794, + "grad_norm": 5.900640821408424, + "learning_rate": 9.775035598275775e-07, + "loss": 0.1883, + "step": 1113 + }, + { + "epoch": 0.4260038240917782, + "grad_norm": 4.3154834986532435, + "learning_rate": 9.776287043869961e-07, + "loss": 0.6946, + "step": 1114 + }, + { + "epoch": 0.42638623326959846, + "grad_norm": 3.7595181787484857, + "learning_rate": 9.777537366587732e-07, + "loss": 0.7462, + "step": 1115 + }, + { + "epoch": 0.42676864244741874, + "grad_norm": 3.384034673851337, + "learning_rate": 9.778786568442317e-07, + "loss": 0.4953, + "step": 1116 + }, + { + "epoch": 0.42715105162523903, + "grad_norm": 5.47317848736149, + "learning_rate": 9.780034651441531e-07, + "loss": 0.2367, + "step": 1117 + }, + { + "epoch": 0.42753346080305926, + "grad_norm": 3.441935781036068, + "learning_rate": 9.781281617587801e-07, + "loss": 0.3293, + "step": 1118 + }, + { + "epoch": 0.42791586998087955, + "grad_norm": 3.590290560639132, + "learning_rate": 9.782527468878181e-07, + "loss": 0.5157, + "step": 1119 + }, + { + "epoch": 0.42829827915869984, + "grad_norm": 3.6094221632216916, + "learning_rate": 9.783772207304376e-07, + "loss": 0.2926, + "step": 1120 + }, + { + "epoch": 0.42868068833652007, + "grad_norm": 3.451588560725235, + "learning_rate": 9.785015834852755e-07, + "loss": 0.236, + "step": 1121 + }, + { + "epoch": 0.42906309751434035, + "grad_norm": 3.577192293232052, + "learning_rate": 9.786258353504375e-07, + "loss": 0.1973, + "step": 1122 + }, + { + "epoch": 0.42944550669216064, + "grad_norm": 2.5122485514213464, + "learning_rate": 9.787499765235003e-07, + "loss": 0.264, + "step": 1123 + }, + { + "epoch": 0.42982791586998087, + "grad_norm": 3.255338392584085, + "learning_rate": 9.788740072015124e-07, + "loss": 0.123, + "step": 1124 + }, + { + "epoch": 0.43021032504780116, + "grad_norm": 4.154827362036261, + "learning_rate": 9.78997927580997e-07, + "loss": 0.1447, + "step": 1125 + }, + { + "epoch": 0.4305927342256214, + "grad_norm": 4.182479353795052, + "learning_rate": 9.791217378579527e-07, + "loss": 0.4986, + "step": 1126 + }, + { + "epoch": 0.4309751434034417, + "grad_norm": 3.758631119416443, + "learning_rate": 9.792454382278575e-07, + "loss": 0.4975, + "step": 1127 + }, + { + "epoch": 0.43135755258126196, + "grad_norm": 3.017336503018636, + "learning_rate": 9.79369028885668e-07, + "loss": 0.6907, + "step": 1128 + }, + { + "epoch": 0.4317399617590822, + "grad_norm": 2.400174583601951, + "learning_rate": 9.794925100258231e-07, + "loss": 0.2594, + "step": 1129 + }, + { + "epoch": 0.4321223709369025, + "grad_norm": 4.02000882617494, + "learning_rate": 9.796158818422455e-07, + "loss": 0.4644, + "step": 1130 + }, + { + "epoch": 0.43250478011472276, + "grad_norm": 3.448245921169389, + "learning_rate": 9.79739144528343e-07, + "loss": 0.5918, + "step": 1131 + }, + { + "epoch": 0.432887189292543, + "grad_norm": 3.401682459862951, + "learning_rate": 9.7986229827701e-07, + "loss": 0.4152, + "step": 1132 + }, + { + "epoch": 0.4332695984703633, + "grad_norm": 2.734233394770841, + "learning_rate": 9.799853432806311e-07, + "loss": 0.3426, + "step": 1133 + }, + { + "epoch": 0.43365200764818357, + "grad_norm": 2.4236323717916033, + "learning_rate": 9.80108279731081e-07, + "loss": 0.5017, + "step": 1134 + }, + { + "epoch": 0.4340344168260038, + "grad_norm": 3.5109505327608437, + "learning_rate": 9.802311078197272e-07, + "loss": 0.1354, + "step": 1135 + }, + { + "epoch": 0.4344168260038241, + "grad_norm": 3.139030241432602, + "learning_rate": 9.803538277374314e-07, + "loss": 0.1875, + "step": 1136 + }, + { + "epoch": 0.4347992351816444, + "grad_norm": 2.7639194552862896, + "learning_rate": 9.804764396745518e-07, + "loss": 0.1821, + "step": 1137 + }, + { + "epoch": 0.4351816443594646, + "grad_norm": 4.877676791335944, + "learning_rate": 9.805989438209442e-07, + "loss": 0.234, + "step": 1138 + }, + { + "epoch": 0.4355640535372849, + "grad_norm": 4.27220233592374, + "learning_rate": 9.807213403659645e-07, + "loss": 0.6155, + "step": 1139 + }, + { + "epoch": 0.4359464627151052, + "grad_norm": 2.3128417510023125, + "learning_rate": 9.808436294984702e-07, + "loss": 0.5823, + "step": 1140 + }, + { + "epoch": 0.4363288718929254, + "grad_norm": 2.6664943838097916, + "learning_rate": 9.80965811406821e-07, + "loss": 0.5333, + "step": 1141 + }, + { + "epoch": 0.4367112810707457, + "grad_norm": 3.5690670053012563, + "learning_rate": 9.810878862788832e-07, + "loss": 0.4513, + "step": 1142 + }, + { + "epoch": 0.437093690248566, + "grad_norm": 2.8073558811824375, + "learning_rate": 9.81209854302028e-07, + "loss": 0.7191, + "step": 1143 + }, + { + "epoch": 0.4374760994263862, + "grad_norm": 3.482803469304339, + "learning_rate": 9.813317156631366e-07, + "loss": 0.4091, + "step": 1144 + }, + { + "epoch": 0.4378585086042065, + "grad_norm": 3.0622469739124307, + "learning_rate": 9.814534705485993e-07, + "loss": 0.298, + "step": 1145 + }, + { + "epoch": 0.4382409177820268, + "grad_norm": 3.5218069020874525, + "learning_rate": 9.815751191443186e-07, + "loss": 0.1966, + "step": 1146 + }, + { + "epoch": 0.438623326959847, + "grad_norm": 3.3552436442245703, + "learning_rate": 9.816966616357108e-07, + "loss": 0.2007, + "step": 1147 + }, + { + "epoch": 0.4390057361376673, + "grad_norm": 5.2585727997837015, + "learning_rate": 9.81818098207707e-07, + "loss": 0.2871, + "step": 1148 + }, + { + "epoch": 0.4393881453154876, + "grad_norm": 4.389194241831785, + "learning_rate": 9.819394290447558e-07, + "loss": 0.1631, + "step": 1149 + }, + { + "epoch": 0.4397705544933078, + "grad_norm": 5.51060607623087, + "learning_rate": 9.820606543308238e-07, + "loss": 0.1877, + "step": 1150 + }, + { + "epoch": 0.4401529636711281, + "grad_norm": 5.541953555069094, + "learning_rate": 9.82181774249398e-07, + "loss": 0.5161, + "step": 1151 + }, + { + "epoch": 0.4405353728489484, + "grad_norm": 4.145049412413277, + "learning_rate": 9.82302788983488e-07, + "loss": 0.325, + "step": 1152 + }, + { + "epoch": 0.4409177820267686, + "grad_norm": 4.833186487181417, + "learning_rate": 9.824236987156263e-07, + "loss": 0.7207, + "step": 1153 + }, + { + "epoch": 0.4413001912045889, + "grad_norm": 2.173152630125367, + "learning_rate": 9.82544503627871e-07, + "loss": 0.5159, + "step": 1154 + }, + { + "epoch": 0.4416826003824092, + "grad_norm": 2.1875687724610686, + "learning_rate": 9.82665203901807e-07, + "loss": 0.5773, + "step": 1155 + }, + { + "epoch": 0.4420650095602294, + "grad_norm": 2.490327816874326, + "learning_rate": 9.82785799718548e-07, + "loss": 0.5218, + "step": 1156 + }, + { + "epoch": 0.4424474187380497, + "grad_norm": 2.9280410412585467, + "learning_rate": 9.829062912587372e-07, + "loss": 0.3466, + "step": 1157 + }, + { + "epoch": 0.44282982791587, + "grad_norm": 2.2559649554229417, + "learning_rate": 9.830266787025507e-07, + "loss": 0.495, + "step": 1158 + }, + { + "epoch": 0.44321223709369023, + "grad_norm": 2.5825451499633303, + "learning_rate": 9.83146962229697e-07, + "loss": 0.2198, + "step": 1159 + }, + { + "epoch": 0.4435946462715105, + "grad_norm": 2.5606514380526155, + "learning_rate": 9.832671420194203e-07, + "loss": 0.126, + "step": 1160 + }, + { + "epoch": 0.4439770554493308, + "grad_norm": 3.403153017767686, + "learning_rate": 9.83387218250501e-07, + "loss": 0.1227, + "step": 1161 + }, + { + "epoch": 0.44435946462715104, + "grad_norm": 3.6106131935818038, + "learning_rate": 9.835071911012583e-07, + "loss": 0.1078, + "step": 1162 + }, + { + "epoch": 0.4447418738049713, + "grad_norm": 5.662654728070213, + "learning_rate": 9.836270607495506e-07, + "loss": 0.3093, + "step": 1163 + }, + { + "epoch": 0.4451242829827916, + "grad_norm": 4.739358928803586, + "learning_rate": 9.837468273727776e-07, + "loss": 0.3841, + "step": 1164 + }, + { + "epoch": 0.44550669216061184, + "grad_norm": 4.695340326261659, + "learning_rate": 9.83866491147883e-07, + "loss": 0.678, + "step": 1165 + }, + { + "epoch": 0.4458891013384321, + "grad_norm": 3.343741443658737, + "learning_rate": 9.83986052251354e-07, + "loss": 0.5099, + "step": 1166 + }, + { + "epoch": 0.4462715105162524, + "grad_norm": 4.670818707014738, + "learning_rate": 9.841055108592244e-07, + "loss": 0.4155, + "step": 1167 + }, + { + "epoch": 0.44665391969407264, + "grad_norm": 3.5820297856884276, + "learning_rate": 9.842248671470756e-07, + "loss": 0.5674, + "step": 1168 + }, + { + "epoch": 0.44703632887189293, + "grad_norm": 3.620177909315062, + "learning_rate": 9.84344121290038e-07, + "loss": 0.5238, + "step": 1169 + }, + { + "epoch": 0.4474187380497132, + "grad_norm": 2.889158629597049, + "learning_rate": 9.84463273462793e-07, + "loss": 0.1835, + "step": 1170 + }, + { + "epoch": 0.44780114722753345, + "grad_norm": 3.3970537547478528, + "learning_rate": 9.845823238395744e-07, + "loss": 0.4899, + "step": 1171 + }, + { + "epoch": 0.44818355640535373, + "grad_norm": 3.7631936989471986, + "learning_rate": 9.847012725941693e-07, + "loss": 0.3491, + "step": 1172 + }, + { + "epoch": 0.448565965583174, + "grad_norm": 4.2082091712890515, + "learning_rate": 9.84820119899921e-07, + "loss": 0.2193, + "step": 1173 + }, + { + "epoch": 0.44894837476099425, + "grad_norm": 3.8214286936618596, + "learning_rate": 9.849388659297287e-07, + "loss": 0.1701, + "step": 1174 + }, + { + "epoch": 0.44933078393881454, + "grad_norm": 3.445321634501071, + "learning_rate": 9.850575108560506e-07, + "loss": 0.1214, + "step": 1175 + }, + { + "epoch": 0.4497131931166348, + "grad_norm": 7.539568726932249, + "learning_rate": 9.851760548509042e-07, + "loss": 0.6822, + "step": 1176 + }, + { + "epoch": 0.45009560229445505, + "grad_norm": 4.843640726148957, + "learning_rate": 9.85294498085869e-07, + "loss": 0.7945, + "step": 1177 + }, + { + "epoch": 0.45047801147227534, + "grad_norm": 3.354664041040037, + "learning_rate": 9.854128407320875e-07, + "loss": 0.5656, + "step": 1178 + }, + { + "epoch": 0.45086042065009563, + "grad_norm": 3.114434960615066, + "learning_rate": 9.855310829602656e-07, + "loss": 0.5782, + "step": 1179 + }, + { + "epoch": 0.45124282982791586, + "grad_norm": 2.023679976497725, + "learning_rate": 9.856492249406758e-07, + "loss": 0.4287, + "step": 1180 + }, + { + "epoch": 0.45162523900573615, + "grad_norm": 2.9935946430201184, + "learning_rate": 9.857672668431576e-07, + "loss": 0.4449, + "step": 1181 + }, + { + "epoch": 0.45200764818355643, + "grad_norm": 3.0199307551245393, + "learning_rate": 9.85885208837119e-07, + "loss": 0.2413, + "step": 1182 + }, + { + "epoch": 0.45239005736137666, + "grad_norm": 3.2386902848763635, + "learning_rate": 9.860030510915388e-07, + "loss": 0.2867, + "step": 1183 + }, + { + "epoch": 0.45277246653919695, + "grad_norm": 1.7672623325899193, + "learning_rate": 9.86120793774967e-07, + "loss": 0.3667, + "step": 1184 + }, + { + "epoch": 0.45315487571701724, + "grad_norm": 2.526398707585068, + "learning_rate": 9.86238437055527e-07, + "loss": 0.2138, + "step": 1185 + }, + { + "epoch": 0.45353728489483747, + "grad_norm": 3.4035631134247284, + "learning_rate": 9.863559811009169e-07, + "loss": 0.1627, + "step": 1186 + }, + { + "epoch": 0.45391969407265775, + "grad_norm": 4.542402132079084, + "learning_rate": 9.864734260784098e-07, + "loss": 0.1862, + "step": 1187 + }, + { + "epoch": 0.454302103250478, + "grad_norm": 6.694197393829867, + "learning_rate": 9.865907721548575e-07, + "loss": 0.2665, + "step": 1188 + }, + { + "epoch": 0.45468451242829827, + "grad_norm": 4.0984265355302725, + "learning_rate": 9.8670801949669e-07, + "loss": 0.5322, + "step": 1189 + }, + { + "epoch": 0.45506692160611856, + "grad_norm": 4.028227863522493, + "learning_rate": 9.868251682699173e-07, + "loss": 0.5913, + "step": 1190 + }, + { + "epoch": 0.4554493307839388, + "grad_norm": 2.6755965343120414, + "learning_rate": 9.869422186401317e-07, + "loss": 0.2668, + "step": 1191 + }, + { + "epoch": 0.4558317399617591, + "grad_norm": 3.1639273155438374, + "learning_rate": 9.87059170772508e-07, + "loss": 0.3166, + "step": 1192 + }, + { + "epoch": 0.45621414913957936, + "grad_norm": 2.9884248586816313, + "learning_rate": 9.87176024831806e-07, + "loss": 0.3124, + "step": 1193 + }, + { + "epoch": 0.4565965583173996, + "grad_norm": 2.3033321209843156, + "learning_rate": 9.872927809823707e-07, + "loss": 0.1689, + "step": 1194 + }, + { + "epoch": 0.4569789674952199, + "grad_norm": 3.2703745493900755, + "learning_rate": 9.874094393881347e-07, + "loss": 0.2984, + "step": 1195 + }, + { + "epoch": 0.45736137667304017, + "grad_norm": 2.8477148565428396, + "learning_rate": 9.875260002126198e-07, + "loss": 0.2651, + "step": 1196 + }, + { + "epoch": 0.4577437858508604, + "grad_norm": 3.408230686896076, + "learning_rate": 9.876424636189368e-07, + "loss": 0.3628, + "step": 1197 + }, + { + "epoch": 0.4581261950286807, + "grad_norm": 3.6999295769252885, + "learning_rate": 9.877588297697883e-07, + "loss": 0.2163, + "step": 1198 + }, + { + "epoch": 0.45850860420650097, + "grad_norm": 2.585867705080345, + "learning_rate": 9.878750988274699e-07, + "loss": 0.1576, + "step": 1199 + }, + { + "epoch": 0.4588910133843212, + "grad_norm": 3.341464108030177, + "learning_rate": 9.879912709538705e-07, + "loss": 0.1122, + "step": 1200 + }, + { + "epoch": 0.4592734225621415, + "grad_norm": 4.997080522787351, + "learning_rate": 9.881073463104754e-07, + "loss": 0.5105, + "step": 1201 + }, + { + "epoch": 0.4596558317399618, + "grad_norm": 4.627752361454242, + "learning_rate": 9.882233250583664e-07, + "loss": 0.5844, + "step": 1202 + }, + { + "epoch": 0.460038240917782, + "grad_norm": 2.9673988881677205, + "learning_rate": 9.883392073582227e-07, + "loss": 0.4366, + "step": 1203 + }, + { + "epoch": 0.4604206500956023, + "grad_norm": 1.9153244529546578, + "learning_rate": 9.884549933703243e-07, + "loss": 0.3027, + "step": 1204 + }, + { + "epoch": 0.4608030592734226, + "grad_norm": 4.689452921634211, + "learning_rate": 9.885706832545508e-07, + "loss": 0.6043, + "step": 1205 + }, + { + "epoch": 0.4611854684512428, + "grad_norm": 3.31635650130163, + "learning_rate": 9.886862771703845e-07, + "loss": 0.4529, + "step": 1206 + }, + { + "epoch": 0.4615678776290631, + "grad_norm": 3.965577787654834, + "learning_rate": 9.88801775276911e-07, + "loss": 0.6734, + "step": 1207 + }, + { + "epoch": 0.4619502868068834, + "grad_norm": 2.7480814049780675, + "learning_rate": 9.88917177732821e-07, + "loss": 0.3801, + "step": 1208 + }, + { + "epoch": 0.4623326959847036, + "grad_norm": 4.818329092362784, + "learning_rate": 9.890324846964104e-07, + "loss": 0.3093, + "step": 1209 + }, + { + "epoch": 0.4627151051625239, + "grad_norm": 3.9353637955632736, + "learning_rate": 9.891476963255836e-07, + "loss": 0.1566, + "step": 1210 + }, + { + "epoch": 0.4630975143403442, + "grad_norm": 3.958589779761676, + "learning_rate": 9.892628127778526e-07, + "loss": 0.3374, + "step": 1211 + }, + { + "epoch": 0.4634799235181644, + "grad_norm": 5.670833775812039, + "learning_rate": 9.8937783421034e-07, + "loss": 0.219, + "step": 1212 + }, + { + "epoch": 0.4638623326959847, + "grad_norm": 10.89393948397313, + "learning_rate": 9.894927607797798e-07, + "loss": 0.3929, + "step": 1213 + }, + { + "epoch": 0.464244741873805, + "grad_norm": 3.952266675322374, + "learning_rate": 9.896075926425177e-07, + "loss": 0.4048, + "step": 1214 + }, + { + "epoch": 0.4646271510516252, + "grad_norm": 4.457437561932895, + "learning_rate": 9.89722329954514e-07, + "loss": 0.6058, + "step": 1215 + }, + { + "epoch": 0.4650095602294455, + "grad_norm": 2.677637322445704, + "learning_rate": 9.898369728713437e-07, + "loss": 0.345, + "step": 1216 + }, + { + "epoch": 0.4653919694072658, + "grad_norm": 3.192019680278025, + "learning_rate": 9.899515215481983e-07, + "loss": 0.3367, + "step": 1217 + }, + { + "epoch": 0.465774378585086, + "grad_norm": 3.4701386842304975, + "learning_rate": 9.900659761398872e-07, + "loss": 0.5685, + "step": 1218 + }, + { + "epoch": 0.4661567877629063, + "grad_norm": 4.125531480246737, + "learning_rate": 9.901803368008373e-07, + "loss": 0.5088, + "step": 1219 + }, + { + "epoch": 0.4665391969407266, + "grad_norm": 4.025808757774617, + "learning_rate": 9.902946036850975e-07, + "loss": 0.4209, + "step": 1220 + }, + { + "epoch": 0.46692160611854683, + "grad_norm": 3.4272727720703533, + "learning_rate": 9.904087769463365e-07, + "loss": 0.286, + "step": 1221 + }, + { + "epoch": 0.4673040152963671, + "grad_norm": 2.8436544213634063, + "learning_rate": 9.905228567378466e-07, + "loss": 0.2031, + "step": 1222 + }, + { + "epoch": 0.4676864244741874, + "grad_norm": 4.01348321104643, + "learning_rate": 9.906368432125435e-07, + "loss": 0.264, + "step": 1223 + }, + { + "epoch": 0.46806883365200763, + "grad_norm": 5.429052170282707, + "learning_rate": 9.907507365229678e-07, + "loss": 0.1528, + "step": 1224 + }, + { + "epoch": 0.4684512428298279, + "grad_norm": 5.180266887314177, + "learning_rate": 9.908645368212868e-07, + "loss": 0.1564, + "step": 1225 + }, + { + "epoch": 0.4688336520076482, + "grad_norm": 5.246787405276859, + "learning_rate": 9.909782442592948e-07, + "loss": 0.5546, + "step": 1226 + }, + { + "epoch": 0.46921606118546844, + "grad_norm": 3.9574744380238647, + "learning_rate": 9.910918589884153e-07, + "loss": 0.5016, + "step": 1227 + }, + { + "epoch": 0.4695984703632887, + "grad_norm": 3.5726202720433826, + "learning_rate": 9.912053811597014e-07, + "loss": 0.2558, + "step": 1228 + }, + { + "epoch": 0.469980879541109, + "grad_norm": 2.7987347605240522, + "learning_rate": 9.913188109238372e-07, + "loss": 0.2549, + "step": 1229 + }, + { + "epoch": 0.47036328871892924, + "grad_norm": 3.427817284431774, + "learning_rate": 9.9143214843114e-07, + "loss": 0.5589, + "step": 1230 + }, + { + "epoch": 0.4707456978967495, + "grad_norm": 3.163584129057733, + "learning_rate": 9.915453938315595e-07, + "loss": 0.2749, + "step": 1231 + }, + { + "epoch": 0.4711281070745698, + "grad_norm": 3.300248431477948, + "learning_rate": 9.916585472746808e-07, + "loss": 0.3889, + "step": 1232 + }, + { + "epoch": 0.47151051625239004, + "grad_norm": 2.516873446763163, + "learning_rate": 9.917716089097243e-07, + "loss": 0.3703, + "step": 1233 + }, + { + "epoch": 0.47189292543021033, + "grad_norm": 16.50220821516207, + "learning_rate": 9.918845788855485e-07, + "loss": 0.1553, + "step": 1234 + }, + { + "epoch": 0.4722753346080306, + "grad_norm": 3.361109653889434, + "learning_rate": 9.91997457350649e-07, + "loss": 0.2914, + "step": 1235 + }, + { + "epoch": 0.47265774378585085, + "grad_norm": 3.522181386406725, + "learning_rate": 9.921102444531612e-07, + "loss": 0.1333, + "step": 1236 + }, + { + "epoch": 0.47304015296367113, + "grad_norm": 5.504639489394662, + "learning_rate": 9.92222940340862e-07, + "loss": 0.1475, + "step": 1237 + }, + { + "epoch": 0.4734225621414914, + "grad_norm": 5.981770479044214, + "learning_rate": 9.923355451611689e-07, + "loss": 0.305, + "step": 1238 + }, + { + "epoch": 0.47380497131931165, + "grad_norm": 4.278875207431444, + "learning_rate": 9.924480590611424e-07, + "loss": 0.7047, + "step": 1239 + }, + { + "epoch": 0.47418738049713194, + "grad_norm": 3.7687764889622843, + "learning_rate": 9.92560482187488e-07, + "loss": 0.3426, + "step": 1240 + }, + { + "epoch": 0.4745697896749522, + "grad_norm": 2.821927862844503, + "learning_rate": 9.926728146865552e-07, + "loss": 0.5761, + "step": 1241 + }, + { + "epoch": 0.47495219885277246, + "grad_norm": 2.7631673291650007, + "learning_rate": 9.927850567043408e-07, + "loss": 0.4746, + "step": 1242 + }, + { + "epoch": 0.47533460803059274, + "grad_norm": 2.434910499134197, + "learning_rate": 9.928972083864887e-07, + "loss": 0.4313, + "step": 1243 + }, + { + "epoch": 0.47571701720841303, + "grad_norm": 2.7586319438112827, + "learning_rate": 9.930092698782916e-07, + "loss": 0.4159, + "step": 1244 + }, + { + "epoch": 0.47609942638623326, + "grad_norm": 3.0803940828055265, + "learning_rate": 9.931212413246912e-07, + "loss": 0.2955, + "step": 1245 + }, + { + "epoch": 0.47648183556405355, + "grad_norm": 3.7571820465812333, + "learning_rate": 9.932331228702814e-07, + "loss": 0.2781, + "step": 1246 + }, + { + "epoch": 0.4768642447418738, + "grad_norm": 4.616070762396008, + "learning_rate": 9.933449146593071e-07, + "loss": 0.3388, + "step": 1247 + }, + { + "epoch": 0.47724665391969406, + "grad_norm": 2.935917529514803, + "learning_rate": 9.934566168356667e-07, + "loss": 0.1672, + "step": 1248 + }, + { + "epoch": 0.47762906309751435, + "grad_norm": 5.0556371819656745, + "learning_rate": 9.935682295429127e-07, + "loss": 0.1328, + "step": 1249 + }, + { + "epoch": 0.4780114722753346, + "grad_norm": 4.746674427569042, + "learning_rate": 9.93679752924253e-07, + "loss": 0.1172, + "step": 1250 + }, + { + "epoch": 0.47839388145315487, + "grad_norm": 5.159086429836909, + "learning_rate": 9.93791187122552e-07, + "loss": 0.3563, + "step": 1251 + }, + { + "epoch": 0.47877629063097515, + "grad_norm": 3.385802042618818, + "learning_rate": 9.939025322803313e-07, + "loss": 0.5295, + "step": 1252 + }, + { + "epoch": 0.4791586998087954, + "grad_norm": 3.145228414079556, + "learning_rate": 9.940137885397717e-07, + "loss": 0.3932, + "step": 1253 + }, + { + "epoch": 0.47954110898661567, + "grad_norm": 2.2029955670367993, + "learning_rate": 9.941249560427134e-07, + "loss": 0.2603, + "step": 1254 + }, + { + "epoch": 0.47992351816443596, + "grad_norm": 2.91587793493996, + "learning_rate": 9.942360349306571e-07, + "loss": 0.4738, + "step": 1255 + }, + { + "epoch": 0.4803059273422562, + "grad_norm": 4.883861996587228, + "learning_rate": 9.94347025344766e-07, + "loss": 0.5368, + "step": 1256 + }, + { + "epoch": 0.4806883365200765, + "grad_norm": 2.9874481835121296, + "learning_rate": 9.94457927425866e-07, + "loss": 0.2862, + "step": 1257 + }, + { + "epoch": 0.48107074569789676, + "grad_norm": 3.496120482614276, + "learning_rate": 9.945687413144468e-07, + "loss": 0.3032, + "step": 1258 + }, + { + "epoch": 0.481453154875717, + "grad_norm": 2.92530042621577, + "learning_rate": 9.946794671506636e-07, + "loss": 0.4655, + "step": 1259 + }, + { + "epoch": 0.4818355640535373, + "grad_norm": 3.2392137977017943, + "learning_rate": 9.947901050743373e-07, + "loss": 0.2499, + "step": 1260 + }, + { + "epoch": 0.48221797323135757, + "grad_norm": 5.070935603384918, + "learning_rate": 9.949006552249564e-07, + "loss": 0.1595, + "step": 1261 + }, + { + "epoch": 0.4826003824091778, + "grad_norm": 3.7492990474444645, + "learning_rate": 9.950111177416775e-07, + "loss": 0.1476, + "step": 1262 + }, + { + "epoch": 0.4829827915869981, + "grad_norm": 4.9714620614137415, + "learning_rate": 9.951214927633269e-07, + "loss": 0.3147, + "step": 1263 + }, + { + "epoch": 0.48336520076481837, + "grad_norm": 3.678668940147185, + "learning_rate": 9.952317804284008e-07, + "loss": 0.3862, + "step": 1264 + }, + { + "epoch": 0.4837476099426386, + "grad_norm": 4.033745045311636, + "learning_rate": 9.95341980875067e-07, + "loss": 0.4374, + "step": 1265 + }, + { + "epoch": 0.4841300191204589, + "grad_norm": 2.5253120758279097, + "learning_rate": 9.954520942411654e-07, + "loss": 0.3187, + "step": 1266 + }, + { + "epoch": 0.4845124282982792, + "grad_norm": 2.629254027626614, + "learning_rate": 9.955621206642103e-07, + "loss": 0.4113, + "step": 1267 + }, + { + "epoch": 0.4848948374760994, + "grad_norm": 3.4575327364019297, + "learning_rate": 9.956720602813898e-07, + "loss": 0.3076, + "step": 1268 + }, + { + "epoch": 0.4852772466539197, + "grad_norm": 1.6920858167705082, + "learning_rate": 9.957819132295675e-07, + "loss": 0.3438, + "step": 1269 + }, + { + "epoch": 0.48565965583174, + "grad_norm": 2.9171764109992773, + "learning_rate": 9.958916796452842e-07, + "loss": 0.3557, + "step": 1270 + }, + { + "epoch": 0.4860420650095602, + "grad_norm": 3.131474311393124, + "learning_rate": 9.960013596647574e-07, + "loss": 0.203, + "step": 1271 + }, + { + "epoch": 0.4864244741873805, + "grad_norm": 2.664857896662694, + "learning_rate": 9.961109534238842e-07, + "loss": 0.2681, + "step": 1272 + }, + { + "epoch": 0.4868068833652008, + "grad_norm": 3.2868231537134185, + "learning_rate": 9.962204610582404e-07, + "loss": 0.1471, + "step": 1273 + }, + { + "epoch": 0.487189292543021, + "grad_norm": 5.023350550253528, + "learning_rate": 9.96329882703083e-07, + "loss": 0.1718, + "step": 1274 + }, + { + "epoch": 0.4875717017208413, + "grad_norm": 4.181604591011255, + "learning_rate": 9.964392184933502e-07, + "loss": 0.1601, + "step": 1275 + }, + { + "epoch": 0.4879541108986616, + "grad_norm": 4.9286020961146955, + "learning_rate": 9.965484685636635e-07, + "loss": 0.3611, + "step": 1276 + }, + { + "epoch": 0.4883365200764818, + "grad_norm": 3.785761892014254, + "learning_rate": 9.966576330483273e-07, + "loss": 0.5685, + "step": 1277 + }, + { + "epoch": 0.4887189292543021, + "grad_norm": 3.2837802395628066, + "learning_rate": 9.96766712081331e-07, + "loss": 0.5583, + "step": 1278 + }, + { + "epoch": 0.4891013384321224, + "grad_norm": 3.5432906672246642, + "learning_rate": 9.968757057963495e-07, + "loss": 0.4116, + "step": 1279 + }, + { + "epoch": 0.4894837476099426, + "grad_norm": 3.013914898253767, + "learning_rate": 9.969846143267443e-07, + "loss": 0.4147, + "step": 1280 + }, + { + "epoch": 0.4898661567877629, + "grad_norm": 4.501812252066544, + "learning_rate": 9.97093437805564e-07, + "loss": 0.5485, + "step": 1281 + }, + { + "epoch": 0.4902485659655832, + "grad_norm": 2.827393157856154, + "learning_rate": 9.972021763655468e-07, + "loss": 0.3098, + "step": 1282 + }, + { + "epoch": 0.4906309751434034, + "grad_norm": 3.0924155700062674, + "learning_rate": 9.973108301391195e-07, + "loss": 0.2388, + "step": 1283 + }, + { + "epoch": 0.4910133843212237, + "grad_norm": 2.9308757355975787, + "learning_rate": 9.974193992583994e-07, + "loss": 0.1811, + "step": 1284 + }, + { + "epoch": 0.491395793499044, + "grad_norm": 3.1683061939771453, + "learning_rate": 9.975278838551958e-07, + "loss": 0.259, + "step": 1285 + }, + { + "epoch": 0.49177820267686423, + "grad_norm": 5.305806869138074, + "learning_rate": 9.976362840610098e-07, + "loss": 0.1055, + "step": 1286 + }, + { + "epoch": 0.4921606118546845, + "grad_norm": 5.969263823596937, + "learning_rate": 9.977446000070363e-07, + "loss": 0.2077, + "step": 1287 + }, + { + "epoch": 0.4925430210325048, + "grad_norm": 4.696538527080613, + "learning_rate": 9.978528318241642e-07, + "loss": 0.3235, + "step": 1288 + }, + { + "epoch": 0.49292543021032503, + "grad_norm": 3.6845788538530737, + "learning_rate": 9.979609796429774e-07, + "loss": 0.4144, + "step": 1289 + }, + { + "epoch": 0.4933078393881453, + "grad_norm": 4.982568108398919, + "learning_rate": 9.980690435937572e-07, + "loss": 0.6038, + "step": 1290 + }, + { + "epoch": 0.4936902485659656, + "grad_norm": 2.4051702479577983, + "learning_rate": 9.981770238064806e-07, + "loss": 0.4925, + "step": 1291 + }, + { + "epoch": 0.49407265774378584, + "grad_norm": 2.7142936753034363, + "learning_rate": 9.982849204108237e-07, + "loss": 0.4256, + "step": 1292 + }, + { + "epoch": 0.4944550669216061, + "grad_norm": 3.632771415888318, + "learning_rate": 9.983927335361607e-07, + "loss": 0.7238, + "step": 1293 + }, + { + "epoch": 0.4948374760994264, + "grad_norm": 2.7053264848194076, + "learning_rate": 9.985004633115662e-07, + "loss": 0.3043, + "step": 1294 + }, + { + "epoch": 0.49521988527724664, + "grad_norm": 2.6085726480924345, + "learning_rate": 9.986081098658163e-07, + "loss": 0.4205, + "step": 1295 + }, + { + "epoch": 0.4956022944550669, + "grad_norm": 3.171201136373861, + "learning_rate": 9.987156733273876e-07, + "loss": 0.2772, + "step": 1296 + }, + { + "epoch": 0.4959847036328872, + "grad_norm": 3.872751875899766, + "learning_rate": 9.988231538244609e-07, + "loss": 0.2868, + "step": 1297 + }, + { + "epoch": 0.49636711281070744, + "grad_norm": 2.6882358696856294, + "learning_rate": 9.989305514849188e-07, + "loss": 0.1262, + "step": 1298 + }, + { + "epoch": 0.49674952198852773, + "grad_norm": 4.098589649932069, + "learning_rate": 9.990378664363505e-07, + "loss": 0.1473, + "step": 1299 + }, + { + "epoch": 0.497131931166348, + "grad_norm": 3.4145372003988643, + "learning_rate": 9.991450988060493e-07, + "loss": 0.1244, + "step": 1300 + }, + { + "epoch": 0.49751434034416825, + "grad_norm": 4.759577080988201, + "learning_rate": 9.992522487210147e-07, + "loss": 0.3778, + "step": 1301 + }, + { + "epoch": 0.49789674952198854, + "grad_norm": 3.4069575703506247, + "learning_rate": 9.993593163079546e-07, + "loss": 0.2946, + "step": 1302 + }, + { + "epoch": 0.4982791586998088, + "grad_norm": 2.3063970916899077, + "learning_rate": 9.994663016932842e-07, + "loss": 0.4804, + "step": 1303 + }, + { + "epoch": 0.49866156787762905, + "grad_norm": 2.736609543154999, + "learning_rate": 9.99573205003128e-07, + "loss": 0.2697, + "step": 1304 + }, + { + "epoch": 0.49904397705544934, + "grad_norm": 3.491110548163302, + "learning_rate": 9.9968002636332e-07, + "loss": 0.4484, + "step": 1305 + }, + { + "epoch": 0.4994263862332696, + "grad_norm": 3.0667836339730554, + "learning_rate": 9.99786765899406e-07, + "loss": 0.2802, + "step": 1306 + }, + { + "epoch": 0.49980879541108986, + "grad_norm": 2.063057217320862, + "learning_rate": 9.998934237366427e-07, + "loss": 0.4247, + "step": 1307 + }, + { + "epoch": 0.5001912045889101, + "grad_norm": 2.096118327710105, + "learning_rate": 1e-06, + "loss": 0.3153, + "step": 1308 + }, + { + "epoch": 0.5005736137667304, + "grad_norm": 3.693911264581626, + "learning_rate": 1e-06, + "loss": 0.2241, + "step": 1309 + }, + { + "epoch": 0.5009560229445507, + "grad_norm": 3.4515142351940358, + "learning_rate": 1e-06, + "loss": 0.4212, + "step": 1310 + }, + { + "epoch": 0.501338432122371, + "grad_norm": 4.489151912533015, + "learning_rate": 1e-06, + "loss": 0.1585, + "step": 1311 + }, + { + "epoch": 0.5017208413001912, + "grad_norm": 4.848606751094361, + "learning_rate": 1e-06, + "loss": 0.1356, + "step": 1312 + }, + { + "epoch": 0.5021032504780115, + "grad_norm": 4.52506228240761, + "learning_rate": 1e-06, + "loss": 0.1422, + "step": 1313 + }, + { + "epoch": 0.5024856596558317, + "grad_norm": 3.254764146126662, + "learning_rate": 1e-06, + "loss": 0.5845, + "step": 1314 + }, + { + "epoch": 0.502868068833652, + "grad_norm": 4.1668970934412455, + "learning_rate": 1e-06, + "loss": 0.5553, + "step": 1315 + }, + { + "epoch": 0.5032504780114723, + "grad_norm": 3.452117634896171, + "learning_rate": 1e-06, + "loss": 0.5817, + "step": 1316 + }, + { + "epoch": 0.5036328871892926, + "grad_norm": 3.2847363070902205, + "learning_rate": 1e-06, + "loss": 0.3663, + "step": 1317 + }, + { + "epoch": 0.5040152963671128, + "grad_norm": 3.2185864453288633, + "learning_rate": 1e-06, + "loss": 0.3871, + "step": 1318 + }, + { + "epoch": 0.5043977055449331, + "grad_norm": 2.2661042857469327, + "learning_rate": 1e-06, + "loss": 0.3103, + "step": 1319 + }, + { + "epoch": 0.5047801147227533, + "grad_norm": 3.6130270708739736, + "learning_rate": 1e-06, + "loss": 0.2834, + "step": 1320 + }, + { + "epoch": 0.5051625239005736, + "grad_norm": 3.068598067179928, + "learning_rate": 1e-06, + "loss": 0.1962, + "step": 1321 + }, + { + "epoch": 0.5055449330783939, + "grad_norm": 2.531267896047366, + "learning_rate": 1e-06, + "loss": 0.2529, + "step": 1322 + }, + { + "epoch": 0.5059273422562142, + "grad_norm": 5.057852790207729, + "learning_rate": 1e-06, + "loss": 0.1403, + "step": 1323 + }, + { + "epoch": 0.5063097514340344, + "grad_norm": 4.646924402504079, + "learning_rate": 1e-06, + "loss": 0.0901, + "step": 1324 + }, + { + "epoch": 0.5066921606118547, + "grad_norm": 3.614905937522274, + "learning_rate": 1e-06, + "loss": 0.1286, + "step": 1325 + }, + { + "epoch": 0.5070745697896749, + "grad_norm": 5.897274564579732, + "learning_rate": 1e-06, + "loss": 0.3708, + "step": 1326 + }, + { + "epoch": 0.5074569789674952, + "grad_norm": 3.4151865071294427, + "learning_rate": 1e-06, + "loss": 0.45, + "step": 1327 + }, + { + "epoch": 0.5078393881453155, + "grad_norm": 2.4104035434568982, + "learning_rate": 1e-06, + "loss": 0.3933, + "step": 1328 + }, + { + "epoch": 0.5082217973231358, + "grad_norm": 2.700161406319671, + "learning_rate": 1e-06, + "loss": 0.3618, + "step": 1329 + }, + { + "epoch": 0.5086042065009561, + "grad_norm": 3.2288464459211887, + "learning_rate": 1e-06, + "loss": 0.3783, + "step": 1330 + }, + { + "epoch": 0.5089866156787763, + "grad_norm": 2.586615804814528, + "learning_rate": 1e-06, + "loss": 0.1344, + "step": 1331 + }, + { + "epoch": 0.5093690248565965, + "grad_norm": 3.315912180695968, + "learning_rate": 1e-06, + "loss": 0.3309, + "step": 1332 + }, + { + "epoch": 0.5097514340344168, + "grad_norm": 3.367855704933834, + "learning_rate": 1e-06, + "loss": 0.2817, + "step": 1333 + }, + { + "epoch": 0.5101338432122371, + "grad_norm": 3.674560084579588, + "learning_rate": 1e-06, + "loss": 0.1843, + "step": 1334 + }, + { + "epoch": 0.5105162523900574, + "grad_norm": 4.558705881761429, + "learning_rate": 1e-06, + "loss": 0.2367, + "step": 1335 + }, + { + "epoch": 0.5108986615678777, + "grad_norm": 4.22902800735144, + "learning_rate": 1e-06, + "loss": 0.1465, + "step": 1336 + }, + { + "epoch": 0.511281070745698, + "grad_norm": 5.283314318377014, + "learning_rate": 1e-06, + "loss": 0.1821, + "step": 1337 + }, + { + "epoch": 0.5116634799235181, + "grad_norm": 9.455108182339645, + "learning_rate": 1e-06, + "loss": 0.231, + "step": 1338 + }, + { + "epoch": 0.5120458891013384, + "grad_norm": 4.439166118912293, + "learning_rate": 1e-06, + "loss": 0.5254, + "step": 1339 + }, + { + "epoch": 0.5124282982791587, + "grad_norm": 3.2336246523102847, + "learning_rate": 1e-06, + "loss": 0.7046, + "step": 1340 + }, + { + "epoch": 0.512810707456979, + "grad_norm": 2.6178188188833382, + "learning_rate": 1e-06, + "loss": 0.5535, + "step": 1341 + }, + { + "epoch": 0.5131931166347993, + "grad_norm": 2.8595616013971545, + "learning_rate": 1e-06, + "loss": 0.2644, + "step": 1342 + }, + { + "epoch": 0.5135755258126194, + "grad_norm": 2.3391660495381705, + "learning_rate": 1e-06, + "loss": 0.2714, + "step": 1343 + }, + { + "epoch": 0.5139579349904397, + "grad_norm": 2.3435370793584385, + "learning_rate": 1e-06, + "loss": 0.2043, + "step": 1344 + }, + { + "epoch": 0.51434034416826, + "grad_norm": 2.6533094398183104, + "learning_rate": 1e-06, + "loss": 0.1981, + "step": 1345 + }, + { + "epoch": 0.5147227533460803, + "grad_norm": 2.4914698030745965, + "learning_rate": 1e-06, + "loss": 0.3858, + "step": 1346 + }, + { + "epoch": 0.5151051625239006, + "grad_norm": 2.8747949941769626, + "learning_rate": 1e-06, + "loss": 0.2616, + "step": 1347 + }, + { + "epoch": 0.5154875717017209, + "grad_norm": 3.3837906125988515, + "learning_rate": 1e-06, + "loss": 0.1699, + "step": 1348 + }, + { + "epoch": 0.5158699808795411, + "grad_norm": 3.262904151642769, + "learning_rate": 1e-06, + "loss": 0.1372, + "step": 1349 + }, + { + "epoch": 0.5162523900573613, + "grad_norm": 2.9114792242420964, + "learning_rate": 1e-06, + "loss": 0.075, + "step": 1350 + }, + { + "epoch": 0.5166347992351816, + "grad_norm": 6.668526517203372, + "learning_rate": 1e-06, + "loss": 0.3453, + "step": 1351 + }, + { + "epoch": 0.5170172084130019, + "grad_norm": 5.45561340743341, + "learning_rate": 1e-06, + "loss": 0.7369, + "step": 1352 + }, + { + "epoch": 0.5173996175908222, + "grad_norm": 2.50805814971923, + "learning_rate": 1e-06, + "loss": 0.4992, + "step": 1353 + }, + { + "epoch": 0.5177820267686425, + "grad_norm": 2.969272125909418, + "learning_rate": 1e-06, + "loss": 0.4825, + "step": 1354 + }, + { + "epoch": 0.5181644359464627, + "grad_norm": 2.791645467497392, + "learning_rate": 1e-06, + "loss": 0.5708, + "step": 1355 + }, + { + "epoch": 0.518546845124283, + "grad_norm": 4.315351345216982, + "learning_rate": 1e-06, + "loss": 0.3612, + "step": 1356 + }, + { + "epoch": 0.5189292543021032, + "grad_norm": 2.652557679033688, + "learning_rate": 1e-06, + "loss": 0.2474, + "step": 1357 + }, + { + "epoch": 0.5193116634799235, + "grad_norm": 3.283614405753792, + "learning_rate": 1e-06, + "loss": 0.2597, + "step": 1358 + }, + { + "epoch": 0.5196940726577438, + "grad_norm": 3.046856219894045, + "learning_rate": 1e-06, + "loss": 0.1428, + "step": 1359 + }, + { + "epoch": 0.5200764818355641, + "grad_norm": 4.348918227812207, + "learning_rate": 1e-06, + "loss": 0.2275, + "step": 1360 + }, + { + "epoch": 0.5204588910133843, + "grad_norm": 3.8700391947809294, + "learning_rate": 1e-06, + "loss": 0.1018, + "step": 1361 + }, + { + "epoch": 0.5208413001912046, + "grad_norm": 3.9402543379897286, + "learning_rate": 1e-06, + "loss": 0.1486, + "step": 1362 + }, + { + "epoch": 0.5212237093690248, + "grad_norm": 5.106967370431914, + "learning_rate": 1e-06, + "loss": 0.2618, + "step": 1363 + }, + { + "epoch": 0.5216061185468451, + "grad_norm": 5.296478526713344, + "learning_rate": 1e-06, + "loss": 0.6094, + "step": 1364 + }, + { + "epoch": 0.5219885277246654, + "grad_norm": 3.660781790612689, + "learning_rate": 1e-06, + "loss": 0.4412, + "step": 1365 + }, + { + "epoch": 0.5223709369024857, + "grad_norm": 1.9007224315694646, + "learning_rate": 1e-06, + "loss": 0.2527, + "step": 1366 + }, + { + "epoch": 0.5227533460803059, + "grad_norm": 2.956085658596427, + "learning_rate": 1e-06, + "loss": 0.231, + "step": 1367 + }, + { + "epoch": 0.5231357552581262, + "grad_norm": 2.4687345480133516, + "learning_rate": 1e-06, + "loss": 0.411, + "step": 1368 + }, + { + "epoch": 0.5235181644359465, + "grad_norm": 4.025379489992484, + "learning_rate": 1e-06, + "loss": 0.3887, + "step": 1369 + }, + { + "epoch": 0.5239005736137667, + "grad_norm": 3.6222658040311884, + "learning_rate": 1e-06, + "loss": 0.2028, + "step": 1370 + }, + { + "epoch": 0.524282982791587, + "grad_norm": 3.7365587305909913, + "learning_rate": 1e-06, + "loss": 0.2103, + "step": 1371 + }, + { + "epoch": 0.5246653919694073, + "grad_norm": 4.312172255293865, + "learning_rate": 1e-06, + "loss": 0.1642, + "step": 1372 + }, + { + "epoch": 0.5250478011472275, + "grad_norm": 3.3071052238397027, + "learning_rate": 1e-06, + "loss": 0.1017, + "step": 1373 + }, + { + "epoch": 0.5254302103250478, + "grad_norm": 4.405403651125026, + "learning_rate": 1e-06, + "loss": 0.1677, + "step": 1374 + }, + { + "epoch": 0.5258126195028681, + "grad_norm": 5.855008464126268, + "learning_rate": 1e-06, + "loss": 0.1313, + "step": 1375 + }, + { + "epoch": 0.5261950286806883, + "grad_norm": 10.89078227166302, + "learning_rate": 1e-06, + "loss": 0.5938, + "step": 1376 + }, + { + "epoch": 0.5265774378585086, + "grad_norm": 3.8335694502936306, + "learning_rate": 1e-06, + "loss": 0.3765, + "step": 1377 + }, + { + "epoch": 0.5269598470363289, + "grad_norm": 3.2972544953386977, + "learning_rate": 1e-06, + "loss": 0.6239, + "step": 1378 + }, + { + "epoch": 0.5273422562141491, + "grad_norm": 2.458371232166294, + "learning_rate": 1e-06, + "loss": 0.3414, + "step": 1379 + }, + { + "epoch": 0.5277246653919694, + "grad_norm": 2.390335564139875, + "learning_rate": 1e-06, + "loss": 0.5026, + "step": 1380 + }, + { + "epoch": 0.5281070745697897, + "grad_norm": 2.967322076414318, + "learning_rate": 1e-06, + "loss": 0.444, + "step": 1381 + }, + { + "epoch": 0.52848948374761, + "grad_norm": 2.8882291195297185, + "learning_rate": 1e-06, + "loss": 0.4309, + "step": 1382 + }, + { + "epoch": 0.5288718929254302, + "grad_norm": 3.138768800521931, + "learning_rate": 1e-06, + "loss": 0.2174, + "step": 1383 + }, + { + "epoch": 0.5292543021032505, + "grad_norm": 4.084695835252402, + "learning_rate": 1e-06, + "loss": 0.3231, + "step": 1384 + }, + { + "epoch": 0.5296367112810707, + "grad_norm": 6.997493840222035, + "learning_rate": 1e-06, + "loss": 0.3456, + "step": 1385 + }, + { + "epoch": 0.530019120458891, + "grad_norm": 4.038094085353025, + "learning_rate": 1e-06, + "loss": 0.17, + "step": 1386 + }, + { + "epoch": 0.5304015296367113, + "grad_norm": 5.436952234498637, + "learning_rate": 1e-06, + "loss": 0.1275, + "step": 1387 + }, + { + "epoch": 0.5307839388145316, + "grad_norm": 5.222542304291769, + "learning_rate": 1e-06, + "loss": 0.4724, + "step": 1388 + }, + { + "epoch": 0.5311663479923519, + "grad_norm": 5.941772850910443, + "learning_rate": 1e-06, + "loss": 0.4343, + "step": 1389 + }, + { + "epoch": 0.5315487571701721, + "grad_norm": 2.8879997910792734, + "learning_rate": 1e-06, + "loss": 0.3274, + "step": 1390 + }, + { + "epoch": 0.5319311663479923, + "grad_norm": 3.1758467138471143, + "learning_rate": 1e-06, + "loss": 0.5694, + "step": 1391 + }, + { + "epoch": 0.5323135755258126, + "grad_norm": 1.874729391279537, + "learning_rate": 1e-06, + "loss": 0.281, + "step": 1392 + }, + { + "epoch": 0.5326959847036329, + "grad_norm": 2.1564058993828286, + "learning_rate": 1e-06, + "loss": 0.3753, + "step": 1393 + }, + { + "epoch": 0.5330783938814532, + "grad_norm": 3.7039091028549707, + "learning_rate": 1e-06, + "loss": 0.4773, + "step": 1394 + }, + { + "epoch": 0.5334608030592735, + "grad_norm": 2.860529765435142, + "learning_rate": 1e-06, + "loss": 0.2912, + "step": 1395 + }, + { + "epoch": 0.5338432122370937, + "grad_norm": 2.727603047047569, + "learning_rate": 1e-06, + "loss": 0.2058, + "step": 1396 + }, + { + "epoch": 0.5342256214149139, + "grad_norm": 3.342870008236601, + "learning_rate": 1e-06, + "loss": 0.1445, + "step": 1397 + }, + { + "epoch": 0.5346080305927342, + "grad_norm": 6.050217288820573, + "learning_rate": 1e-06, + "loss": 0.1494, + "step": 1398 + }, + { + "epoch": 0.5349904397705545, + "grad_norm": 3.790319615679061, + "learning_rate": 1e-06, + "loss": 0.1409, + "step": 1399 + }, + { + "epoch": 0.5353728489483748, + "grad_norm": 5.62295698575686, + "learning_rate": 1e-06, + "loss": 0.2005, + "step": 1400 + }, + { + "epoch": 0.5357552581261951, + "grad_norm": 4.8012770543407575, + "learning_rate": 1e-06, + "loss": 0.335, + "step": 1401 + }, + { + "epoch": 0.5361376673040152, + "grad_norm": 3.8272160910788013, + "learning_rate": 1e-06, + "loss": 0.6847, + "step": 1402 + }, + { + "epoch": 0.5365200764818355, + "grad_norm": 1.7962285081271852, + "learning_rate": 1e-06, + "loss": 0.4374, + "step": 1403 + }, + { + "epoch": 0.5369024856596558, + "grad_norm": 2.3516397558556457, + "learning_rate": 1e-06, + "loss": 0.3764, + "step": 1404 + }, + { + "epoch": 0.5372848948374761, + "grad_norm": 4.221537255614015, + "learning_rate": 1e-06, + "loss": 0.603, + "step": 1405 + }, + { + "epoch": 0.5376673040152964, + "grad_norm": 3.0694035162481716, + "learning_rate": 1e-06, + "loss": 0.2605, + "step": 1406 + }, + { + "epoch": 0.5380497131931167, + "grad_norm": 3.531148318286807, + "learning_rate": 1e-06, + "loss": 0.2123, + "step": 1407 + }, + { + "epoch": 0.5384321223709368, + "grad_norm": 3.2479461635957287, + "learning_rate": 1e-06, + "loss": 0.3191, + "step": 1408 + }, + { + "epoch": 0.5388145315487571, + "grad_norm": 3.652223113816529, + "learning_rate": 1e-06, + "loss": 0.1757, + "step": 1409 + }, + { + "epoch": 0.5391969407265774, + "grad_norm": 3.214449185044371, + "learning_rate": 1e-06, + "loss": 0.2905, + "step": 1410 + }, + { + "epoch": 0.5395793499043977, + "grad_norm": 4.39707348340983, + "learning_rate": 1e-06, + "loss": 0.1222, + "step": 1411 + }, + { + "epoch": 0.539961759082218, + "grad_norm": 3.814813381045284, + "learning_rate": 1e-06, + "loss": 0.1384, + "step": 1412 + }, + { + "epoch": 0.5403441682600383, + "grad_norm": 5.64874882981047, + "learning_rate": 1e-06, + "loss": 0.2985, + "step": 1413 + }, + { + "epoch": 0.5407265774378585, + "grad_norm": 5.740154627933632, + "learning_rate": 1e-06, + "loss": 0.8276, + "step": 1414 + }, + { + "epoch": 0.5411089866156787, + "grad_norm": 3.4403655159623447, + "learning_rate": 1e-06, + "loss": 0.6347, + "step": 1415 + }, + { + "epoch": 0.541491395793499, + "grad_norm": 2.4875120114750535, + "learning_rate": 1e-06, + "loss": 0.3806, + "step": 1416 + }, + { + "epoch": 0.5418738049713193, + "grad_norm": 3.758326758966102, + "learning_rate": 1e-06, + "loss": 0.4696, + "step": 1417 + }, + { + "epoch": 0.5422562141491396, + "grad_norm": 5.878757979338721, + "learning_rate": 1e-06, + "loss": 0.4185, + "step": 1418 + }, + { + "epoch": 0.5426386233269599, + "grad_norm": 3.576439072998601, + "learning_rate": 1e-06, + "loss": 0.342, + "step": 1419 + }, + { + "epoch": 0.5430210325047801, + "grad_norm": 2.7456398123435597, + "learning_rate": 1e-06, + "loss": 0.2941, + "step": 1420 + }, + { + "epoch": 0.5434034416826004, + "grad_norm": 4.197590300160669, + "learning_rate": 1e-06, + "loss": 0.3272, + "step": 1421 + }, + { + "epoch": 0.5437858508604206, + "grad_norm": 3.2487043585805786, + "learning_rate": 1e-06, + "loss": 0.182, + "step": 1422 + }, + { + "epoch": 0.5441682600382409, + "grad_norm": 3.94322200571778, + "learning_rate": 1e-06, + "loss": 0.168, + "step": 1423 + }, + { + "epoch": 0.5445506692160612, + "grad_norm": 2.604046709476756, + "learning_rate": 1e-06, + "loss": 0.0688, + "step": 1424 + }, + { + "epoch": 0.5449330783938815, + "grad_norm": 3.660151038883697, + "learning_rate": 1e-06, + "loss": 0.1562, + "step": 1425 + }, + { + "epoch": 0.5453154875717017, + "grad_norm": 4.966302616849774, + "learning_rate": 1e-06, + "loss": 0.419, + "step": 1426 + }, + { + "epoch": 0.545697896749522, + "grad_norm": 4.336279283404338, + "learning_rate": 1e-06, + "loss": 0.5309, + "step": 1427 + }, + { + "epoch": 0.5460803059273422, + "grad_norm": 2.8745609445764884, + "learning_rate": 1e-06, + "loss": 0.5821, + "step": 1428 + }, + { + "epoch": 0.5464627151051625, + "grad_norm": 3.690098251233938, + "learning_rate": 1e-06, + "loss": 0.4779, + "step": 1429 + }, + { + "epoch": 0.5468451242829828, + "grad_norm": 3.2592457058974755, + "learning_rate": 1e-06, + "loss": 0.4654, + "step": 1430 + }, + { + "epoch": 0.5472275334608031, + "grad_norm": 2.274986761704379, + "learning_rate": 1e-06, + "loss": 0.2362, + "step": 1431 + }, + { + "epoch": 0.5476099426386233, + "grad_norm": 2.4617258898051313, + "learning_rate": 1e-06, + "loss": 0.3788, + "step": 1432 + }, + { + "epoch": 0.5479923518164436, + "grad_norm": 3.181641074614379, + "learning_rate": 1e-06, + "loss": 0.1856, + "step": 1433 + }, + { + "epoch": 0.5483747609942639, + "grad_norm": 3.587572709785552, + "learning_rate": 1e-06, + "loss": 0.2723, + "step": 1434 + }, + { + "epoch": 0.5487571701720841, + "grad_norm": 6.8930961347213024, + "learning_rate": 1e-06, + "loss": 0.1598, + "step": 1435 + }, + { + "epoch": 0.5491395793499044, + "grad_norm": 3.597531129067457, + "learning_rate": 1e-06, + "loss": 0.158, + "step": 1436 + }, + { + "epoch": 0.5495219885277247, + "grad_norm": 3.356841369471808, + "learning_rate": 1e-06, + "loss": 0.1753, + "step": 1437 + }, + { + "epoch": 0.5499043977055449, + "grad_norm": 5.611603445647346, + "learning_rate": 1e-06, + "loss": 0.3639, + "step": 1438 + }, + { + "epoch": 0.5502868068833652, + "grad_norm": 3.989556029436323, + "learning_rate": 1e-06, + "loss": 0.6409, + "step": 1439 + }, + { + "epoch": 0.5506692160611855, + "grad_norm": 3.3816313410443484, + "learning_rate": 1e-06, + "loss": 0.5594, + "step": 1440 + }, + { + "epoch": 0.5510516252390057, + "grad_norm": 2.4438288761297637, + "learning_rate": 1e-06, + "loss": 0.547, + "step": 1441 + }, + { + "epoch": 0.551434034416826, + "grad_norm": 2.419172193954764, + "learning_rate": 1e-06, + "loss": 0.4859, + "step": 1442 + }, + { + "epoch": 0.5518164435946463, + "grad_norm": 2.7785293473648034, + "learning_rate": 1e-06, + "loss": 0.3713, + "step": 1443 + }, + { + "epoch": 0.5521988527724665, + "grad_norm": 4.100156008845883, + "learning_rate": 1e-06, + "loss": 0.4, + "step": 1444 + }, + { + "epoch": 0.5525812619502868, + "grad_norm": 2.379213310180867, + "learning_rate": 1e-06, + "loss": 0.2778, + "step": 1445 + }, + { + "epoch": 0.5529636711281071, + "grad_norm": 3.3810549921292683, + "learning_rate": 1e-06, + "loss": 0.3165, + "step": 1446 + }, + { + "epoch": 0.5533460803059274, + "grad_norm": 3.1252262796494, + "learning_rate": 1e-06, + "loss": 0.2668, + "step": 1447 + }, + { + "epoch": 0.5537284894837476, + "grad_norm": 3.3292930594949097, + "learning_rate": 1e-06, + "loss": 0.1549, + "step": 1448 + }, + { + "epoch": 0.5541108986615679, + "grad_norm": 3.138692536490799, + "learning_rate": 1e-06, + "loss": 0.1964, + "step": 1449 + }, + { + "epoch": 0.5544933078393881, + "grad_norm": 3.1526886138033787, + "learning_rate": 1e-06, + "loss": 0.1309, + "step": 1450 + }, + { + "epoch": 0.5548757170172084, + "grad_norm": 5.321098101587102, + "learning_rate": 1e-06, + "loss": 0.4587, + "step": 1451 + }, + { + "epoch": 0.5552581261950287, + "grad_norm": 4.287469731438118, + "learning_rate": 1e-06, + "loss": 0.7825, + "step": 1452 + }, + { + "epoch": 0.555640535372849, + "grad_norm": 2.4409792595356885, + "learning_rate": 1e-06, + "loss": 0.274, + "step": 1453 + }, + { + "epoch": 0.5560229445506693, + "grad_norm": 3.1741668989869627, + "learning_rate": 1e-06, + "loss": 0.346, + "step": 1454 + }, + { + "epoch": 0.5564053537284895, + "grad_norm": 2.319172280410221, + "learning_rate": 1e-06, + "loss": 0.3755, + "step": 1455 + }, + { + "epoch": 0.5567877629063097, + "grad_norm": 3.0541576501592655, + "learning_rate": 1e-06, + "loss": 0.3353, + "step": 1456 + }, + { + "epoch": 0.55717017208413, + "grad_norm": 2.044743483667686, + "learning_rate": 1e-06, + "loss": 0.2576, + "step": 1457 + }, + { + "epoch": 0.5575525812619503, + "grad_norm": 3.212134367919686, + "learning_rate": 1e-06, + "loss": 0.2496, + "step": 1458 + }, + { + "epoch": 0.5579349904397706, + "grad_norm": 2.270169245590621, + "learning_rate": 1e-06, + "loss": 0.0757, + "step": 1459 + }, + { + "epoch": 0.5583173996175909, + "grad_norm": 3.051400760362656, + "learning_rate": 1e-06, + "loss": 0.1447, + "step": 1460 + }, + { + "epoch": 0.558699808795411, + "grad_norm": 4.732923832941204, + "learning_rate": 1e-06, + "loss": 0.1583, + "step": 1461 + }, + { + "epoch": 0.5590822179732313, + "grad_norm": 3.7772486930176847, + "learning_rate": 1e-06, + "loss": 0.1173, + "step": 1462 + }, + { + "epoch": 0.5594646271510516, + "grad_norm": 6.13316050714446, + "learning_rate": 1e-06, + "loss": 0.2376, + "step": 1463 + }, + { + "epoch": 0.5598470363288719, + "grad_norm": 3.706321270468106, + "learning_rate": 1e-06, + "loss": 0.842, + "step": 1464 + }, + { + "epoch": 0.5602294455066922, + "grad_norm": 3.6925066564211964, + "learning_rate": 1e-06, + "loss": 0.4967, + "step": 1465 + }, + { + "epoch": 0.5606118546845125, + "grad_norm": 3.049498382342764, + "learning_rate": 1e-06, + "loss": 0.3982, + "step": 1466 + }, + { + "epoch": 0.5609942638623326, + "grad_norm": 3.5211664226608517, + "learning_rate": 1e-06, + "loss": 0.5354, + "step": 1467 + }, + { + "epoch": 0.5613766730401529, + "grad_norm": 2.583751941933015, + "learning_rate": 1e-06, + "loss": 0.3475, + "step": 1468 + }, + { + "epoch": 0.5617590822179732, + "grad_norm": 3.5314702328223095, + "learning_rate": 1e-06, + "loss": 0.2992, + "step": 1469 + }, + { + "epoch": 0.5621414913957935, + "grad_norm": 2.7819103357165913, + "learning_rate": 1e-06, + "loss": 0.3423, + "step": 1470 + }, + { + "epoch": 0.5625239005736138, + "grad_norm": 2.6250383283450858, + "learning_rate": 1e-06, + "loss": 0.3856, + "step": 1471 + }, + { + "epoch": 0.5629063097514341, + "grad_norm": 3.511457898186983, + "learning_rate": 1e-06, + "loss": 0.3121, + "step": 1472 + }, + { + "epoch": 0.5632887189292543, + "grad_norm": 3.1110922513874555, + "learning_rate": 1e-06, + "loss": 0.1538, + "step": 1473 + }, + { + "epoch": 0.5636711281070745, + "grad_norm": 3.4822119597680246, + "learning_rate": 1e-06, + "loss": 0.1029, + "step": 1474 + }, + { + "epoch": 0.5640535372848948, + "grad_norm": 3.7911484472481773, + "learning_rate": 1e-06, + "loss": 0.1442, + "step": 1475 + }, + { + "epoch": 0.5644359464627151, + "grad_norm": 4.900222672054585, + "learning_rate": 1e-06, + "loss": 0.3698, + "step": 1476 + }, + { + "epoch": 0.5648183556405354, + "grad_norm": 3.617631856101012, + "learning_rate": 1e-06, + "loss": 0.7431, + "step": 1477 + }, + { + "epoch": 0.5652007648183557, + "grad_norm": 3.8977123495787365, + "learning_rate": 1e-06, + "loss": 0.3348, + "step": 1478 + }, + { + "epoch": 0.5655831739961759, + "grad_norm": 2.2151442766239327, + "learning_rate": 1e-06, + "loss": 0.4633, + "step": 1479 + }, + { + "epoch": 0.5659655831739961, + "grad_norm": 4.813007278459072, + "learning_rate": 1e-06, + "loss": 0.4659, + "step": 1480 + }, + { + "epoch": 0.5663479923518164, + "grad_norm": 3.7812468473563174, + "learning_rate": 1e-06, + "loss": 0.2921, + "step": 1481 + }, + { + "epoch": 0.5667304015296367, + "grad_norm": 2.1776800721929224, + "learning_rate": 1e-06, + "loss": 0.3112, + "step": 1482 + }, + { + "epoch": 0.567112810707457, + "grad_norm": 5.087591649385154, + "learning_rate": 1e-06, + "loss": 0.2409, + "step": 1483 + }, + { + "epoch": 0.5674952198852773, + "grad_norm": 3.8126374516958377, + "learning_rate": 1e-06, + "loss": 0.2257, + "step": 1484 + }, + { + "epoch": 0.5678776290630975, + "grad_norm": 4.73382625610079, + "learning_rate": 1e-06, + "loss": 0.2601, + "step": 1485 + }, + { + "epoch": 0.5682600382409178, + "grad_norm": 4.026684918983304, + "learning_rate": 1e-06, + "loss": 0.1744, + "step": 1486 + }, + { + "epoch": 0.568642447418738, + "grad_norm": 4.549533132558017, + "learning_rate": 1e-06, + "loss": 0.1276, + "step": 1487 + }, + { + "epoch": 0.5690248565965583, + "grad_norm": 5.288982441463886, + "learning_rate": 1e-06, + "loss": 0.3607, + "step": 1488 + }, + { + "epoch": 0.5694072657743786, + "grad_norm": 2.8286032930866507, + "learning_rate": 1e-06, + "loss": 0.3267, + "step": 1489 + }, + { + "epoch": 0.5697896749521989, + "grad_norm": 3.6822937241925757, + "learning_rate": 1e-06, + "loss": 0.4347, + "step": 1490 + }, + { + "epoch": 0.5701720841300191, + "grad_norm": 2.3247244148589967, + "learning_rate": 1e-06, + "loss": 0.554, + "step": 1491 + }, + { + "epoch": 0.5705544933078394, + "grad_norm": 2.3272486323767416, + "learning_rate": 1e-06, + "loss": 0.3814, + "step": 1492 + }, + { + "epoch": 0.5709369024856596, + "grad_norm": 3.7883676422709254, + "learning_rate": 1e-06, + "loss": 0.5862, + "step": 1493 + }, + { + "epoch": 0.5713193116634799, + "grad_norm": 3.1826315707138937, + "learning_rate": 1e-06, + "loss": 0.4448, + "step": 1494 + }, + { + "epoch": 0.5717017208413002, + "grad_norm": 3.4616438099733, + "learning_rate": 1e-06, + "loss": 0.3137, + "step": 1495 + }, + { + "epoch": 0.5720841300191205, + "grad_norm": 2.8849867811511705, + "learning_rate": 1e-06, + "loss": 0.2362, + "step": 1496 + }, + { + "epoch": 0.5724665391969407, + "grad_norm": 3.38657766926089, + "learning_rate": 1e-06, + "loss": 0.3042, + "step": 1497 + }, + { + "epoch": 0.572848948374761, + "grad_norm": 3.1036333142214962, + "learning_rate": 1e-06, + "loss": 0.3551, + "step": 1498 + }, + { + "epoch": 0.5732313575525813, + "grad_norm": 5.5639005462292435, + "learning_rate": 1e-06, + "loss": 0.2236, + "step": 1499 + }, + { + "epoch": 0.5736137667304015, + "grad_norm": 6.187496609157057, + "learning_rate": 1e-06, + "loss": 0.2325, + "step": 1500 + }, + { + "epoch": 0.5736137667304015, + "eval_runtime": 833.3537, + "eval_samples_per_second": 1.841, + "eval_steps_per_second": 0.461, + "step": 1500 + }, + { + "epoch": 0.5739961759082218, + "grad_norm": 7.648043223925274, + "learning_rate": 1e-06, + "loss": 0.4667, + "step": 1501 + }, + { + "epoch": 0.5743785850860421, + "grad_norm": 3.2906155577956144, + "learning_rate": 1e-06, + "loss": 0.4217, + "step": 1502 + }, + { + "epoch": 0.5747609942638623, + "grad_norm": 3.190112670056694, + "learning_rate": 1e-06, + "loss": 0.6407, + "step": 1503 + }, + { + "epoch": 0.5751434034416826, + "grad_norm": 2.420269046046126, + "learning_rate": 1e-06, + "loss": 0.2823, + "step": 1504 + }, + { + "epoch": 0.5755258126195029, + "grad_norm": 2.518074596333273, + "learning_rate": 1e-06, + "loss": 0.3212, + "step": 1505 + }, + { + "epoch": 0.5759082217973231, + "grad_norm": 3.2015445438773242, + "learning_rate": 1e-06, + "loss": 0.2174, + "step": 1506 + }, + { + "epoch": 0.5762906309751434, + "grad_norm": 3.1915377306147423, + "learning_rate": 1e-06, + "loss": 0.3889, + "step": 1507 + }, + { + "epoch": 0.5766730401529637, + "grad_norm": 2.5112891890343283, + "learning_rate": 1e-06, + "loss": 0.3093, + "step": 1508 + }, + { + "epoch": 0.5770554493307839, + "grad_norm": 2.5316295574683827, + "learning_rate": 1e-06, + "loss": 0.2095, + "step": 1509 + }, + { + "epoch": 0.5774378585086042, + "grad_norm": 5.745471331963637, + "learning_rate": 1e-06, + "loss": 0.332, + "step": 1510 + }, + { + "epoch": 0.5778202676864245, + "grad_norm": 4.534498852143188, + "learning_rate": 1e-06, + "loss": 0.1538, + "step": 1511 + }, + { + "epoch": 0.5782026768642448, + "grad_norm": 4.822514802189205, + "learning_rate": 1e-06, + "loss": 0.2601, + "step": 1512 + }, + { + "epoch": 0.578585086042065, + "grad_norm": 4.8454374265880755, + "learning_rate": 1e-06, + "loss": 0.3081, + "step": 1513 + }, + { + "epoch": 0.5789674952198853, + "grad_norm": 4.366637603041822, + "learning_rate": 1e-06, + "loss": 0.5815, + "step": 1514 + }, + { + "epoch": 0.5793499043977055, + "grad_norm": 4.221327156986129, + "learning_rate": 1e-06, + "loss": 0.5176, + "step": 1515 + }, + { + "epoch": 0.5797323135755258, + "grad_norm": 3.141811772915536, + "learning_rate": 1e-06, + "loss": 0.2951, + "step": 1516 + }, + { + "epoch": 0.5801147227533461, + "grad_norm": 2.4990048335144266, + "learning_rate": 1e-06, + "loss": 0.202, + "step": 1517 + }, + { + "epoch": 0.5804971319311664, + "grad_norm": 3.9050183605197164, + "learning_rate": 1e-06, + "loss": 0.4118, + "step": 1518 + }, + { + "epoch": 0.5808795411089867, + "grad_norm": 4.53393096537423, + "learning_rate": 1e-06, + "loss": 0.3508, + "step": 1519 + }, + { + "epoch": 0.5812619502868069, + "grad_norm": 2.961584176578633, + "learning_rate": 1e-06, + "loss": 0.2947, + "step": 1520 + }, + { + "epoch": 0.5816443594646271, + "grad_norm": 3.441416641895136, + "learning_rate": 1e-06, + "loss": 0.186, + "step": 1521 + }, + { + "epoch": 0.5820267686424474, + "grad_norm": 2.72530329355094, + "learning_rate": 1e-06, + "loss": 0.1401, + "step": 1522 + }, + { + "epoch": 0.5824091778202677, + "grad_norm": 2.818957227714633, + "learning_rate": 1e-06, + "loss": 0.2435, + "step": 1523 + }, + { + "epoch": 0.582791586998088, + "grad_norm": 4.097375792877866, + "learning_rate": 1e-06, + "loss": 0.1301, + "step": 1524 + }, + { + "epoch": 0.5831739961759083, + "grad_norm": 3.5816197548964563, + "learning_rate": 1e-06, + "loss": 0.1609, + "step": 1525 + }, + { + "epoch": 0.5835564053537284, + "grad_norm": 5.585264450820185, + "learning_rate": 1e-06, + "loss": 0.3654, + "step": 1526 + }, + { + "epoch": 0.5839388145315487, + "grad_norm": 4.122925670293326, + "learning_rate": 1e-06, + "loss": 0.5015, + "step": 1527 + }, + { + "epoch": 0.584321223709369, + "grad_norm": 2.4922913915331124, + "learning_rate": 1e-06, + "loss": 0.4, + "step": 1528 + }, + { + "epoch": 0.5847036328871893, + "grad_norm": 2.615770189637759, + "learning_rate": 1e-06, + "loss": 0.3547, + "step": 1529 + }, + { + "epoch": 0.5850860420650096, + "grad_norm": 2.193498398136122, + "learning_rate": 1e-06, + "loss": 0.2621, + "step": 1530 + }, + { + "epoch": 0.5854684512428299, + "grad_norm": 5.002734962142251, + "learning_rate": 1e-06, + "loss": 0.421, + "step": 1531 + }, + { + "epoch": 0.58585086042065, + "grad_norm": 3.724211841126314, + "learning_rate": 1e-06, + "loss": 0.3183, + "step": 1532 + }, + { + "epoch": 0.5862332695984703, + "grad_norm": 2.7640993899681088, + "learning_rate": 1e-06, + "loss": 0.4786, + "step": 1533 + }, + { + "epoch": 0.5866156787762906, + "grad_norm": 3.5616439326168727, + "learning_rate": 1e-06, + "loss": 0.241, + "step": 1534 + }, + { + "epoch": 0.5869980879541109, + "grad_norm": 2.4489084926408275, + "learning_rate": 1e-06, + "loss": 0.0949, + "step": 1535 + }, + { + "epoch": 0.5873804971319312, + "grad_norm": 3.6020111665102528, + "learning_rate": 1e-06, + "loss": 0.1548, + "step": 1536 + }, + { + "epoch": 0.5877629063097515, + "grad_norm": 3.941590379510361, + "learning_rate": 1e-06, + "loss": 0.1511, + "step": 1537 + }, + { + "epoch": 0.5881453154875717, + "grad_norm": 5.340405583213029, + "learning_rate": 1e-06, + "loss": 0.2523, + "step": 1538 + }, + { + "epoch": 0.5885277246653919, + "grad_norm": 4.8752568495162265, + "learning_rate": 1e-06, + "loss": 0.3552, + "step": 1539 + }, + { + "epoch": 0.5889101338432122, + "grad_norm": 3.6597323014379723, + "learning_rate": 1e-06, + "loss": 0.6053, + "step": 1540 + }, + { + "epoch": 0.5892925430210325, + "grad_norm": 3.8974402610404315, + "learning_rate": 1e-06, + "loss": 0.5765, + "step": 1541 + }, + { + "epoch": 0.5896749521988528, + "grad_norm": 2.729337749991348, + "learning_rate": 1e-06, + "loss": 0.2775, + "step": 1542 + }, + { + "epoch": 0.5900573613766731, + "grad_norm": 3.1248181099409362, + "learning_rate": 1e-06, + "loss": 0.2691, + "step": 1543 + }, + { + "epoch": 0.5904397705544933, + "grad_norm": 3.088190342872511, + "learning_rate": 1e-06, + "loss": 0.2857, + "step": 1544 + }, + { + "epoch": 0.5908221797323135, + "grad_norm": 2.6967299120653387, + "learning_rate": 1e-06, + "loss": 0.3731, + "step": 1545 + }, + { + "epoch": 0.5912045889101338, + "grad_norm": 2.647025392981962, + "learning_rate": 1e-06, + "loss": 0.2182, + "step": 1546 + }, + { + "epoch": 0.5915869980879541, + "grad_norm": 3.2712816226817005, + "learning_rate": 1e-06, + "loss": 0.2024, + "step": 1547 + }, + { + "epoch": 0.5919694072657744, + "grad_norm": 5.930421484752699, + "learning_rate": 1e-06, + "loss": 0.3072, + "step": 1548 + }, + { + "epoch": 0.5923518164435947, + "grad_norm": 3.686741508450827, + "learning_rate": 1e-06, + "loss": 0.1599, + "step": 1549 + }, + { + "epoch": 0.5927342256214149, + "grad_norm": 5.403013462307581, + "learning_rate": 1e-06, + "loss": 0.1649, + "step": 1550 + }, + { + "epoch": 0.5931166347992352, + "grad_norm": 5.0398586383552955, + "learning_rate": 1e-06, + "loss": 0.3195, + "step": 1551 + }, + { + "epoch": 0.5934990439770554, + "grad_norm": 4.363840748595389, + "learning_rate": 1e-06, + "loss": 0.6514, + "step": 1552 + }, + { + "epoch": 0.5938814531548757, + "grad_norm": 3.7887919221205557, + "learning_rate": 1e-06, + "loss": 0.5849, + "step": 1553 + }, + { + "epoch": 0.594263862332696, + "grad_norm": 3.225302831344963, + "learning_rate": 1e-06, + "loss": 0.2775, + "step": 1554 + }, + { + "epoch": 0.5946462715105163, + "grad_norm": 2.861391117587321, + "learning_rate": 1e-06, + "loss": 0.5969, + "step": 1555 + }, + { + "epoch": 0.5950286806883365, + "grad_norm": 2.9745260165601777, + "learning_rate": 1e-06, + "loss": 0.4259, + "step": 1556 + }, + { + "epoch": 0.5954110898661568, + "grad_norm": 3.123675408970642, + "learning_rate": 1e-06, + "loss": 0.4054, + "step": 1557 + }, + { + "epoch": 0.595793499043977, + "grad_norm": 2.1075426973879976, + "learning_rate": 1e-06, + "loss": 0.3542, + "step": 1558 + }, + { + "epoch": 0.5961759082217973, + "grad_norm": 2.500735556159103, + "learning_rate": 1e-06, + "loss": 0.2346, + "step": 1559 + }, + { + "epoch": 0.5965583173996176, + "grad_norm": 4.499995337589816, + "learning_rate": 1e-06, + "loss": 0.3501, + "step": 1560 + }, + { + "epoch": 0.5969407265774379, + "grad_norm": 3.784316655787347, + "learning_rate": 1e-06, + "loss": 0.215, + "step": 1561 + }, + { + "epoch": 0.5973231357552581, + "grad_norm": 2.7468269422057237, + "learning_rate": 1e-06, + "loss": 0.084, + "step": 1562 + }, + { + "epoch": 0.5977055449330784, + "grad_norm": 6.792289783312273, + "learning_rate": 1e-06, + "loss": 0.2943, + "step": 1563 + }, + { + "epoch": 0.5980879541108987, + "grad_norm": 3.762110926386478, + "learning_rate": 1e-06, + "loss": 0.4925, + "step": 1564 + }, + { + "epoch": 0.5984703632887189, + "grad_norm": 4.271277626312197, + "learning_rate": 1e-06, + "loss": 0.6514, + "step": 1565 + }, + { + "epoch": 0.5988527724665392, + "grad_norm": 4.796524681197209, + "learning_rate": 1e-06, + "loss": 0.17, + "step": 1566 + }, + { + "epoch": 0.5992351816443595, + "grad_norm": 2.6530873937379136, + "learning_rate": 1e-06, + "loss": 0.4564, + "step": 1567 + }, + { + "epoch": 0.5996175908221797, + "grad_norm": 3.284529146448116, + "learning_rate": 1e-06, + "loss": 0.4594, + "step": 1568 + }, + { + "epoch": 0.6, + "grad_norm": 4.297701669413807, + "learning_rate": 1e-06, + "loss": 0.5233, + "step": 1569 + }, + { + "epoch": 0.6003824091778203, + "grad_norm": 2.7848361608503187, + "learning_rate": 1e-06, + "loss": 0.2024, + "step": 1570 + }, + { + "epoch": 0.6007648183556406, + "grad_norm": 3.1810191971967994, + "learning_rate": 1e-06, + "loss": 0.2397, + "step": 1571 + }, + { + "epoch": 0.6011472275334608, + "grad_norm": 3.462055516547677, + "learning_rate": 1e-06, + "loss": 0.2973, + "step": 1572 + }, + { + "epoch": 0.6015296367112811, + "grad_norm": 3.9998092605890285, + "learning_rate": 1e-06, + "loss": 0.2522, + "step": 1573 + }, + { + "epoch": 0.6019120458891013, + "grad_norm": 3.0841793411792917, + "learning_rate": 1e-06, + "loss": 0.0995, + "step": 1574 + }, + { + "epoch": 0.6022944550669216, + "grad_norm": 4.263103256718209, + "learning_rate": 1e-06, + "loss": 0.122, + "step": 1575 + }, + { + "epoch": 0.6026768642447419, + "grad_norm": 6.395907357084697, + "learning_rate": 1e-06, + "loss": 0.4544, + "step": 1576 + }, + { + "epoch": 0.6030592734225622, + "grad_norm": 4.479918920941317, + "learning_rate": 1e-06, + "loss": 0.592, + "step": 1577 + }, + { + "epoch": 0.6034416826003824, + "grad_norm": 3.0270337542897297, + "learning_rate": 1e-06, + "loss": 0.3419, + "step": 1578 + }, + { + "epoch": 0.6038240917782027, + "grad_norm": 1.8534895128314122, + "learning_rate": 1e-06, + "loss": 0.3151, + "step": 1579 + }, + { + "epoch": 0.6042065009560229, + "grad_norm": 2.6555878655433514, + "learning_rate": 1e-06, + "loss": 0.6539, + "step": 1580 + }, + { + "epoch": 0.6045889101338432, + "grad_norm": 2.056482727423432, + "learning_rate": 1e-06, + "loss": 0.3516, + "step": 1581 + }, + { + "epoch": 0.6049713193116635, + "grad_norm": 3.275965913625526, + "learning_rate": 1e-06, + "loss": 0.1933, + "step": 1582 + }, + { + "epoch": 0.6053537284894838, + "grad_norm": 3.0842731864844897, + "learning_rate": 1e-06, + "loss": 0.3483, + "step": 1583 + }, + { + "epoch": 0.605736137667304, + "grad_norm": 2.7037476280085597, + "learning_rate": 1e-06, + "loss": 0.2797, + "step": 1584 + }, + { + "epoch": 0.6061185468451242, + "grad_norm": 3.385057792226577, + "learning_rate": 1e-06, + "loss": 0.1719, + "step": 1585 + }, + { + "epoch": 0.6065009560229445, + "grad_norm": 3.4903293248748812, + "learning_rate": 1e-06, + "loss": 0.1536, + "step": 1586 + }, + { + "epoch": 0.6068833652007648, + "grad_norm": 4.144413895163444, + "learning_rate": 1e-06, + "loss": 0.1619, + "step": 1587 + }, + { + "epoch": 0.6072657743785851, + "grad_norm": 4.818634680964997, + "learning_rate": 1e-06, + "loss": 0.2627, + "step": 1588 + }, + { + "epoch": 0.6076481835564054, + "grad_norm": 4.040560829968868, + "learning_rate": 1e-06, + "loss": 0.4934, + "step": 1589 + }, + { + "epoch": 0.6080305927342257, + "grad_norm": 4.085364219462454, + "learning_rate": 1e-06, + "loss": 0.5564, + "step": 1590 + }, + { + "epoch": 0.6084130019120458, + "grad_norm": 2.268925965265979, + "learning_rate": 1e-06, + "loss": 0.1791, + "step": 1591 + }, + { + "epoch": 0.6087954110898661, + "grad_norm": 3.289065254554592, + "learning_rate": 1e-06, + "loss": 0.2348, + "step": 1592 + }, + { + "epoch": 0.6091778202676864, + "grad_norm": 1.9265604357928798, + "learning_rate": 1e-06, + "loss": 0.3429, + "step": 1593 + }, + { + "epoch": 0.6095602294455067, + "grad_norm": 6.16191158309742, + "learning_rate": 1e-06, + "loss": 0.6237, + "step": 1594 + }, + { + "epoch": 0.609942638623327, + "grad_norm": 2.846177793984969, + "learning_rate": 1e-06, + "loss": 0.3524, + "step": 1595 + }, + { + "epoch": 0.6103250478011473, + "grad_norm": 3.2424341314797536, + "learning_rate": 1e-06, + "loss": 0.1905, + "step": 1596 + }, + { + "epoch": 0.6107074569789674, + "grad_norm": 2.762463071412313, + "learning_rate": 1e-06, + "loss": 0.309, + "step": 1597 + }, + { + "epoch": 0.6110898661567877, + "grad_norm": 5.149475085719637, + "learning_rate": 1e-06, + "loss": 0.1613, + "step": 1598 + }, + { + "epoch": 0.611472275334608, + "grad_norm": 2.5937691193760686, + "learning_rate": 1e-06, + "loss": 0.0886, + "step": 1599 + }, + { + "epoch": 0.6118546845124283, + "grad_norm": 4.107008335919648, + "learning_rate": 1e-06, + "loss": 0.102, + "step": 1600 + }, + { + "epoch": 0.6122370936902486, + "grad_norm": 5.038598232912599, + "learning_rate": 1e-06, + "loss": 0.4016, + "step": 1601 + }, + { + "epoch": 0.6126195028680689, + "grad_norm": 3.6673026978221492, + "learning_rate": 1e-06, + "loss": 0.5224, + "step": 1602 + }, + { + "epoch": 0.613001912045889, + "grad_norm": 1.9355971313480531, + "learning_rate": 1e-06, + "loss": 0.4279, + "step": 1603 + }, + { + "epoch": 0.6133843212237093, + "grad_norm": 3.028508942851441, + "learning_rate": 1e-06, + "loss": 0.3999, + "step": 1604 + }, + { + "epoch": 0.6137667304015296, + "grad_norm": 3.490329461491545, + "learning_rate": 1e-06, + "loss": 0.4828, + "step": 1605 + }, + { + "epoch": 0.6141491395793499, + "grad_norm": 3.799259143687563, + "learning_rate": 1e-06, + "loss": 0.4321, + "step": 1606 + }, + { + "epoch": 0.6145315487571702, + "grad_norm": 3.293860915341487, + "learning_rate": 1e-06, + "loss": 0.3982, + "step": 1607 + }, + { + "epoch": 0.6149139579349905, + "grad_norm": 3.2708085822527555, + "learning_rate": 1e-06, + "loss": 0.2293, + "step": 1608 + }, + { + "epoch": 0.6152963671128107, + "grad_norm": 3.449351600862056, + "learning_rate": 1e-06, + "loss": 0.358, + "step": 1609 + }, + { + "epoch": 0.615678776290631, + "grad_norm": 4.529874816579575, + "learning_rate": 1e-06, + "loss": 0.1249, + "step": 1610 + }, + { + "epoch": 0.6160611854684512, + "grad_norm": 3.6972732342351042, + "learning_rate": 1e-06, + "loss": 0.1118, + "step": 1611 + }, + { + "epoch": 0.6164435946462715, + "grad_norm": 5.522058468562662, + "learning_rate": 1e-06, + "loss": 0.1054, + "step": 1612 + }, + { + "epoch": 0.6168260038240918, + "grad_norm": 7.399685167368398, + "learning_rate": 1e-06, + "loss": 0.4509, + "step": 1613 + }, + { + "epoch": 0.6172084130019121, + "grad_norm": 4.929145021409884, + "learning_rate": 1e-06, + "loss": 0.5219, + "step": 1614 + }, + { + "epoch": 0.6175908221797323, + "grad_norm": 3.370637546758573, + "learning_rate": 1e-06, + "loss": 0.5646, + "step": 1615 + }, + { + "epoch": 0.6179732313575526, + "grad_norm": 2.3358637054932956, + "learning_rate": 1e-06, + "loss": 0.3913, + "step": 1616 + }, + { + "epoch": 0.6183556405353728, + "grad_norm": 3.445634407350294, + "learning_rate": 1e-06, + "loss": 0.4011, + "step": 1617 + }, + { + "epoch": 0.6187380497131931, + "grad_norm": 3.4811639744318743, + "learning_rate": 1e-06, + "loss": 0.2578, + "step": 1618 + }, + { + "epoch": 0.6191204588910134, + "grad_norm": 4.069633441886181, + "learning_rate": 1e-06, + "loss": 0.299, + "step": 1619 + }, + { + "epoch": 0.6195028680688337, + "grad_norm": 2.0427357994877333, + "learning_rate": 1e-06, + "loss": 0.124, + "step": 1620 + }, + { + "epoch": 0.6198852772466539, + "grad_norm": 3.318965432166625, + "learning_rate": 1e-06, + "loss": 0.3652, + "step": 1621 + }, + { + "epoch": 0.6202676864244742, + "grad_norm": 3.1058845289668113, + "learning_rate": 1e-06, + "loss": 0.1816, + "step": 1622 + }, + { + "epoch": 0.6206500956022944, + "grad_norm": 3.494146083701717, + "learning_rate": 1e-06, + "loss": 0.2191, + "step": 1623 + }, + { + "epoch": 0.6210325047801147, + "grad_norm": 2.911256968586936, + "learning_rate": 1e-06, + "loss": 0.0933, + "step": 1624 + }, + { + "epoch": 0.621414913957935, + "grad_norm": 4.57964187854441, + "learning_rate": 1e-06, + "loss": 0.2255, + "step": 1625 + }, + { + "epoch": 0.6217973231357553, + "grad_norm": 7.234922581952464, + "learning_rate": 1e-06, + "loss": 0.4494, + "step": 1626 + }, + { + "epoch": 0.6221797323135755, + "grad_norm": 2.7171413880571014, + "learning_rate": 1e-06, + "loss": 0.5141, + "step": 1627 + }, + { + "epoch": 0.6225621414913958, + "grad_norm": 3.4219758876751425, + "learning_rate": 1e-06, + "loss": 0.3405, + "step": 1628 + }, + { + "epoch": 0.6229445506692161, + "grad_norm": 2.5694710417966085, + "learning_rate": 1e-06, + "loss": 0.395, + "step": 1629 + }, + { + "epoch": 0.6233269598470363, + "grad_norm": 3.0536773650517013, + "learning_rate": 1e-06, + "loss": 0.2961, + "step": 1630 + }, + { + "epoch": 0.6237093690248566, + "grad_norm": 3.8657433481687753, + "learning_rate": 1e-06, + "loss": 0.3545, + "step": 1631 + }, + { + "epoch": 0.6240917782026769, + "grad_norm": 3.1711600865062755, + "learning_rate": 1e-06, + "loss": 0.3322, + "step": 1632 + }, + { + "epoch": 0.6244741873804971, + "grad_norm": 2.5303667199251354, + "learning_rate": 1e-06, + "loss": 0.3185, + "step": 1633 + }, + { + "epoch": 0.6248565965583174, + "grad_norm": 3.089084536792682, + "learning_rate": 1e-06, + "loss": 0.2415, + "step": 1634 + }, + { + "epoch": 0.6252390057361377, + "grad_norm": 5.177332093718016, + "learning_rate": 1e-06, + "loss": 0.196, + "step": 1635 + }, + { + "epoch": 0.625621414913958, + "grad_norm": 2.9635470121544407, + "learning_rate": 1e-06, + "loss": 0.1352, + "step": 1636 + }, + { + "epoch": 0.6260038240917782, + "grad_norm": 3.3452187993175597, + "learning_rate": 1e-06, + "loss": 0.1187, + "step": 1637 + }, + { + "epoch": 0.6263862332695985, + "grad_norm": 6.97572531808549, + "learning_rate": 1e-06, + "loss": 0.3908, + "step": 1638 + }, + { + "epoch": 0.6267686424474187, + "grad_norm": 5.440352820567418, + "learning_rate": 1e-06, + "loss": 0.6388, + "step": 1639 + }, + { + "epoch": 0.627151051625239, + "grad_norm": 2.9229332950379163, + "learning_rate": 1e-06, + "loss": 0.4336, + "step": 1640 + }, + { + "epoch": 0.6275334608030593, + "grad_norm": 2.539355451849907, + "learning_rate": 1e-06, + "loss": 0.3208, + "step": 1641 + }, + { + "epoch": 0.6279158699808796, + "grad_norm": 2.289203105683247, + "learning_rate": 1e-06, + "loss": 0.4489, + "step": 1642 + }, + { + "epoch": 0.6282982791586998, + "grad_norm": 3.1309029331615954, + "learning_rate": 1e-06, + "loss": 0.3362, + "step": 1643 + }, + { + "epoch": 0.62868068833652, + "grad_norm": 3.3206080406153653, + "learning_rate": 1e-06, + "loss": 0.1911, + "step": 1644 + }, + { + "epoch": 0.6290630975143403, + "grad_norm": 3.1214176243879708, + "learning_rate": 1e-06, + "loss": 0.3695, + "step": 1645 + }, + { + "epoch": 0.6294455066921606, + "grad_norm": 3.8474336666265994, + "learning_rate": 1e-06, + "loss": 0.1746, + "step": 1646 + }, + { + "epoch": 0.6298279158699809, + "grad_norm": 3.2207731536599136, + "learning_rate": 1e-06, + "loss": 0.1577, + "step": 1647 + }, + { + "epoch": 0.6302103250478012, + "grad_norm": 2.067440358308397, + "learning_rate": 1e-06, + "loss": 0.1543, + "step": 1648 + }, + { + "epoch": 0.6305927342256215, + "grad_norm": 2.2835514530593635, + "learning_rate": 1e-06, + "loss": 0.0666, + "step": 1649 + }, + { + "epoch": 0.6309751434034416, + "grad_norm": 4.402501669541691, + "learning_rate": 1e-06, + "loss": 0.1261, + "step": 1650 + }, + { + "epoch": 0.6313575525812619, + "grad_norm": 5.3180156124555795, + "learning_rate": 1e-06, + "loss": 0.34, + "step": 1651 + }, + { + "epoch": 0.6317399617590822, + "grad_norm": 4.015899292598844, + "learning_rate": 1e-06, + "loss": 0.3995, + "step": 1652 + }, + { + "epoch": 0.6321223709369025, + "grad_norm": 2.813014089435832, + "learning_rate": 1e-06, + "loss": 0.5354, + "step": 1653 + }, + { + "epoch": 0.6325047801147228, + "grad_norm": 1.4126292599974228, + "learning_rate": 1e-06, + "loss": 0.5314, + "step": 1654 + }, + { + "epoch": 0.6328871892925431, + "grad_norm": 2.7594537363019023, + "learning_rate": 1e-06, + "loss": 0.5301, + "step": 1655 + }, + { + "epoch": 0.6332695984703632, + "grad_norm": 2.8815974954834624, + "learning_rate": 1e-06, + "loss": 0.3094, + "step": 1656 + }, + { + "epoch": 0.6336520076481835, + "grad_norm": 2.462411106267252, + "learning_rate": 1e-06, + "loss": 0.2472, + "step": 1657 + }, + { + "epoch": 0.6340344168260038, + "grad_norm": 3.105835399935763, + "learning_rate": 1e-06, + "loss": 0.3324, + "step": 1658 + }, + { + "epoch": 0.6344168260038241, + "grad_norm": 5.450155679421611, + "learning_rate": 1e-06, + "loss": 0.1916, + "step": 1659 + }, + { + "epoch": 0.6347992351816444, + "grad_norm": 5.199921225904796, + "learning_rate": 1e-06, + "loss": 0.1978, + "step": 1660 + }, + { + "epoch": 0.6351816443594647, + "grad_norm": 3.691235476787247, + "learning_rate": 1e-06, + "loss": 0.1718, + "step": 1661 + }, + { + "epoch": 0.6355640535372848, + "grad_norm": 5.067963089637659, + "learning_rate": 1e-06, + "loss": 0.1273, + "step": 1662 + }, + { + "epoch": 0.6359464627151051, + "grad_norm": 5.3491196451746665, + "learning_rate": 1e-06, + "loss": 0.3786, + "step": 1663 + }, + { + "epoch": 0.6363288718929254, + "grad_norm": 4.173036106563145, + "learning_rate": 1e-06, + "loss": 0.4033, + "step": 1664 + }, + { + "epoch": 0.6367112810707457, + "grad_norm": 3.6322536900100064, + "learning_rate": 1e-06, + "loss": 0.5193, + "step": 1665 + }, + { + "epoch": 0.637093690248566, + "grad_norm": 2.4756148297130585, + "learning_rate": 1e-06, + "loss": 0.4113, + "step": 1666 + }, + { + "epoch": 0.6374760994263863, + "grad_norm": 3.255770914927566, + "learning_rate": 1e-06, + "loss": 0.3114, + "step": 1667 + }, + { + "epoch": 0.6378585086042065, + "grad_norm": 3.8464504773120454, + "learning_rate": 1e-06, + "loss": 0.4712, + "step": 1668 + }, + { + "epoch": 0.6382409177820267, + "grad_norm": 2.955989195467105, + "learning_rate": 1e-06, + "loss": 0.1994, + "step": 1669 + }, + { + "epoch": 0.638623326959847, + "grad_norm": 2.600741970124298, + "learning_rate": 1e-06, + "loss": 0.1038, + "step": 1670 + }, + { + "epoch": 0.6390057361376673, + "grad_norm": 3.1099354607855556, + "learning_rate": 1e-06, + "loss": 0.2166, + "step": 1671 + }, + { + "epoch": 0.6393881453154876, + "grad_norm": 3.5664288456379634, + "learning_rate": 1e-06, + "loss": 0.3254, + "step": 1672 + }, + { + "epoch": 0.6397705544933079, + "grad_norm": 3.727625281456364, + "learning_rate": 1e-06, + "loss": 0.2059, + "step": 1673 + }, + { + "epoch": 0.6401529636711281, + "grad_norm": 3.330111631521315, + "learning_rate": 1e-06, + "loss": 0.1298, + "step": 1674 + }, + { + "epoch": 0.6405353728489483, + "grad_norm": 5.150736872723286, + "learning_rate": 1e-06, + "loss": 0.1979, + "step": 1675 + }, + { + "epoch": 0.6409177820267686, + "grad_norm": 4.3827972683837375, + "learning_rate": 1e-06, + "loss": 0.4554, + "step": 1676 + }, + { + "epoch": 0.6413001912045889, + "grad_norm": 5.885968945902527, + "learning_rate": 1e-06, + "loss": 0.4346, + "step": 1677 + }, + { + "epoch": 0.6416826003824092, + "grad_norm": 2.1649212286805266, + "learning_rate": 1e-06, + "loss": 0.488, + "step": 1678 + }, + { + "epoch": 0.6420650095602295, + "grad_norm": 2.8184786511597806, + "learning_rate": 1e-06, + "loss": 0.3281, + "step": 1679 + }, + { + "epoch": 0.6424474187380497, + "grad_norm": 2.618282624192141, + "learning_rate": 1e-06, + "loss": 0.3105, + "step": 1680 + }, + { + "epoch": 0.64282982791587, + "grad_norm": 2.867418825552063, + "learning_rate": 1e-06, + "loss": 0.2497, + "step": 1681 + }, + { + "epoch": 0.6432122370936902, + "grad_norm": 2.421551885892965, + "learning_rate": 1e-06, + "loss": 0.3143, + "step": 1682 + }, + { + "epoch": 0.6435946462715105, + "grad_norm": 2.664223942940917, + "learning_rate": 1e-06, + "loss": 0.1734, + "step": 1683 + }, + { + "epoch": 0.6439770554493308, + "grad_norm": 2.357248118143426, + "learning_rate": 1e-06, + "loss": 0.1671, + "step": 1684 + }, + { + "epoch": 0.6443594646271511, + "grad_norm": 6.723963502329862, + "learning_rate": 1e-06, + "loss": 0.2437, + "step": 1685 + }, + { + "epoch": 0.6447418738049713, + "grad_norm": 5.0651138527850295, + "learning_rate": 1e-06, + "loss": 0.2618, + "step": 1686 + }, + { + "epoch": 0.6451242829827916, + "grad_norm": 4.32736849456671, + "learning_rate": 1e-06, + "loss": 0.1156, + "step": 1687 + }, + { + "epoch": 0.6455066921606119, + "grad_norm": 5.403851986058783, + "learning_rate": 1e-06, + "loss": 0.2925, + "step": 1688 + }, + { + "epoch": 0.6458891013384321, + "grad_norm": 5.4870437258588804, + "learning_rate": 1e-06, + "loss": 0.5704, + "step": 1689 + }, + { + "epoch": 0.6462715105162524, + "grad_norm": 2.9527269529089644, + "learning_rate": 1e-06, + "loss": 0.4319, + "step": 1690 + }, + { + "epoch": 0.6466539196940727, + "grad_norm": 2.7521387799722774, + "learning_rate": 1e-06, + "loss": 0.5618, + "step": 1691 + }, + { + "epoch": 0.6470363288718929, + "grad_norm": 3.840974131049391, + "learning_rate": 1e-06, + "loss": 0.3291, + "step": 1692 + }, + { + "epoch": 0.6474187380497132, + "grad_norm": 3.426352719670689, + "learning_rate": 1e-06, + "loss": 0.5471, + "step": 1693 + }, + { + "epoch": 0.6478011472275335, + "grad_norm": 4.976341733721796, + "learning_rate": 1e-06, + "loss": 0.4847, + "step": 1694 + }, + { + "epoch": 0.6481835564053537, + "grad_norm": 3.552568478981308, + "learning_rate": 1e-06, + "loss": 0.43, + "step": 1695 + }, + { + "epoch": 0.648565965583174, + "grad_norm": 2.934198512740815, + "learning_rate": 1e-06, + "loss": 0.3435, + "step": 1696 + }, + { + "epoch": 0.6489483747609943, + "grad_norm": 2.8276520222911943, + "learning_rate": 1e-06, + "loss": 0.2352, + "step": 1697 + }, + { + "epoch": 0.6493307839388145, + "grad_norm": 3.9218239343973753, + "learning_rate": 1e-06, + "loss": 0.2914, + "step": 1698 + }, + { + "epoch": 0.6497131931166348, + "grad_norm": 3.3779882694490158, + "learning_rate": 1e-06, + "loss": 0.1081, + "step": 1699 + }, + { + "epoch": 0.6500956022944551, + "grad_norm": 4.367757852123347, + "learning_rate": 1e-06, + "loss": 0.1843, + "step": 1700 + }, + { + "epoch": 0.6504780114722754, + "grad_norm": 5.006101890401848, + "learning_rate": 1e-06, + "loss": 0.7616, + "step": 1701 + }, + { + "epoch": 0.6508604206500956, + "grad_norm": 3.2721356919280042, + "learning_rate": 1e-06, + "loss": 0.2695, + "step": 1702 + }, + { + "epoch": 0.6512428298279158, + "grad_norm": 1.3918892760416124, + "learning_rate": 1e-06, + "loss": 0.5447, + "step": 1703 + }, + { + "epoch": 0.6516252390057361, + "grad_norm": 3.9619045544238265, + "learning_rate": 1e-06, + "loss": 0.7589, + "step": 1704 + }, + { + "epoch": 0.6520076481835564, + "grad_norm": 2.9728468083222777, + "learning_rate": 1e-06, + "loss": 0.2531, + "step": 1705 + }, + { + "epoch": 0.6523900573613767, + "grad_norm": 2.996392624332678, + "learning_rate": 1e-06, + "loss": 0.3013, + "step": 1706 + }, + { + "epoch": 0.652772466539197, + "grad_norm": 3.7618078615402455, + "learning_rate": 1e-06, + "loss": 0.2693, + "step": 1707 + }, + { + "epoch": 0.6531548757170172, + "grad_norm": 2.285208729198416, + "learning_rate": 1e-06, + "loss": 0.1972, + "step": 1708 + }, + { + "epoch": 0.6535372848948374, + "grad_norm": 3.874416184284152, + "learning_rate": 1e-06, + "loss": 0.1978, + "step": 1709 + }, + { + "epoch": 0.6539196940726577, + "grad_norm": 3.1600430801931734, + "learning_rate": 1e-06, + "loss": 0.1301, + "step": 1710 + }, + { + "epoch": 0.654302103250478, + "grad_norm": 5.052248524065872, + "learning_rate": 1e-06, + "loss": 0.3586, + "step": 1711 + }, + { + "epoch": 0.6546845124282983, + "grad_norm": 2.662989425449097, + "learning_rate": 1e-06, + "loss": 0.09, + "step": 1712 + }, + { + "epoch": 0.6550669216061186, + "grad_norm": 6.50430067544631, + "learning_rate": 1e-06, + "loss": 0.3388, + "step": 1713 + }, + { + "epoch": 0.6554493307839389, + "grad_norm": 3.6709607873296353, + "learning_rate": 1e-06, + "loss": 0.4649, + "step": 1714 + }, + { + "epoch": 0.655831739961759, + "grad_norm": 3.3219462605079326, + "learning_rate": 1e-06, + "loss": 0.4879, + "step": 1715 + }, + { + "epoch": 0.6562141491395793, + "grad_norm": 3.2695636269118244, + "learning_rate": 1e-06, + "loss": 0.5663, + "step": 1716 + }, + { + "epoch": 0.6565965583173996, + "grad_norm": 2.92610914279973, + "learning_rate": 1e-06, + "loss": 0.5111, + "step": 1717 + }, + { + "epoch": 0.6569789674952199, + "grad_norm": 4.1060225007137, + "learning_rate": 1e-06, + "loss": 0.5329, + "step": 1718 + }, + { + "epoch": 0.6573613766730402, + "grad_norm": 3.1458467563233303, + "learning_rate": 1e-06, + "loss": 0.3792, + "step": 1719 + }, + { + "epoch": 0.6577437858508605, + "grad_norm": 2.8146208502026533, + "learning_rate": 1e-06, + "loss": 0.2088, + "step": 1720 + }, + { + "epoch": 0.6581261950286806, + "grad_norm": 2.5008707437468076, + "learning_rate": 1e-06, + "loss": 0.375, + "step": 1721 + }, + { + "epoch": 0.6585086042065009, + "grad_norm": 4.107443700161939, + "learning_rate": 1e-06, + "loss": 0.17, + "step": 1722 + }, + { + "epoch": 0.6588910133843212, + "grad_norm": 3.5735843283007958, + "learning_rate": 1e-06, + "loss": 0.2203, + "step": 1723 + }, + { + "epoch": 0.6592734225621415, + "grad_norm": 3.2595638992124876, + "learning_rate": 1e-06, + "loss": 0.1531, + "step": 1724 + }, + { + "epoch": 0.6596558317399618, + "grad_norm": 3.4207497057215885, + "learning_rate": 1e-06, + "loss": 0.1183, + "step": 1725 + }, + { + "epoch": 0.6600382409177821, + "grad_norm": 4.449081994852851, + "learning_rate": 1e-06, + "loss": 0.4619, + "step": 1726 + }, + { + "epoch": 0.6604206500956022, + "grad_norm": 3.645980105397598, + "learning_rate": 1e-06, + "loss": 0.5698, + "step": 1727 + }, + { + "epoch": 0.6608030592734225, + "grad_norm": 3.17886019354846, + "learning_rate": 1e-06, + "loss": 0.576, + "step": 1728 + }, + { + "epoch": 0.6611854684512428, + "grad_norm": 2.309275982996253, + "learning_rate": 1e-06, + "loss": 0.26, + "step": 1729 + }, + { + "epoch": 0.6615678776290631, + "grad_norm": 3.1960452855402597, + "learning_rate": 1e-06, + "loss": 0.24, + "step": 1730 + }, + { + "epoch": 0.6619502868068834, + "grad_norm": 2.987791013709219, + "learning_rate": 1e-06, + "loss": 0.3228, + "step": 1731 + }, + { + "epoch": 0.6623326959847037, + "grad_norm": 4.365707774596113, + "learning_rate": 1e-06, + "loss": 0.3277, + "step": 1732 + }, + { + "epoch": 0.6627151051625239, + "grad_norm": 2.2909913773160038, + "learning_rate": 1e-06, + "loss": 0.1836, + "step": 1733 + }, + { + "epoch": 0.6630975143403441, + "grad_norm": 2.5442436058021807, + "learning_rate": 1e-06, + "loss": 0.2696, + "step": 1734 + }, + { + "epoch": 0.6634799235181644, + "grad_norm": 3.717153511036284, + "learning_rate": 1e-06, + "loss": 0.1953, + "step": 1735 + }, + { + "epoch": 0.6638623326959847, + "grad_norm": 3.883942898127248, + "learning_rate": 1e-06, + "loss": 0.1044, + "step": 1736 + }, + { + "epoch": 0.664244741873805, + "grad_norm": 4.682744984081185, + "learning_rate": 1e-06, + "loss": 0.185, + "step": 1737 + }, + { + "epoch": 0.6646271510516253, + "grad_norm": 6.51903592851396, + "learning_rate": 1e-06, + "loss": 0.2527, + "step": 1738 + }, + { + "epoch": 0.6650095602294455, + "grad_norm": 4.508646816747134, + "learning_rate": 1e-06, + "loss": 0.4237, + "step": 1739 + }, + { + "epoch": 0.6653919694072657, + "grad_norm": 4.032820994306484, + "learning_rate": 1e-06, + "loss": 0.3399, + "step": 1740 + }, + { + "epoch": 0.665774378585086, + "grad_norm": 2.9156831627182824, + "learning_rate": 1e-06, + "loss": 0.5586, + "step": 1741 + }, + { + "epoch": 0.6661567877629063, + "grad_norm": 2.5427943578595387, + "learning_rate": 1e-06, + "loss": 0.2631, + "step": 1742 + }, + { + "epoch": 0.6665391969407266, + "grad_norm": 3.71340905051251, + "learning_rate": 1e-06, + "loss": 0.6265, + "step": 1743 + }, + { + "epoch": 0.6669216061185469, + "grad_norm": 3.0503382635928866, + "learning_rate": 1e-06, + "loss": 0.322, + "step": 1744 + }, + { + "epoch": 0.6673040152963671, + "grad_norm": 3.1499401510699, + "learning_rate": 1e-06, + "loss": 0.3289, + "step": 1745 + }, + { + "epoch": 0.6676864244741874, + "grad_norm": 3.863479470316066, + "learning_rate": 1e-06, + "loss": 0.2829, + "step": 1746 + }, + { + "epoch": 0.6680688336520076, + "grad_norm": 3.4202907849106525, + "learning_rate": 1e-06, + "loss": 0.3704, + "step": 1747 + }, + { + "epoch": 0.6684512428298279, + "grad_norm": 2.5903809509547693, + "learning_rate": 1e-06, + "loss": 0.1806, + "step": 1748 + }, + { + "epoch": 0.6688336520076482, + "grad_norm": 3.983156264664412, + "learning_rate": 1e-06, + "loss": 0.0728, + "step": 1749 + }, + { + "epoch": 0.6692160611854685, + "grad_norm": 3.8396316423169607, + "learning_rate": 1e-06, + "loss": 0.1351, + "step": 1750 + }, + { + "epoch": 0.6695984703632887, + "grad_norm": 6.639303636918794, + "learning_rate": 1e-06, + "loss": 0.2754, + "step": 1751 + }, + { + "epoch": 0.669980879541109, + "grad_norm": 4.3634602536441, + "learning_rate": 1e-06, + "loss": 0.5403, + "step": 1752 + }, + { + "epoch": 0.6703632887189293, + "grad_norm": 2.234263063841419, + "learning_rate": 1e-06, + "loss": 0.4408, + "step": 1753 + }, + { + "epoch": 0.6707456978967495, + "grad_norm": 2.782095277044659, + "learning_rate": 1e-06, + "loss": 0.3659, + "step": 1754 + }, + { + "epoch": 0.6711281070745698, + "grad_norm": 3.3950645809183007, + "learning_rate": 1e-06, + "loss": 0.2794, + "step": 1755 + }, + { + "epoch": 0.6715105162523901, + "grad_norm": 3.319697783169732, + "learning_rate": 1e-06, + "loss": 0.4635, + "step": 1756 + }, + { + "epoch": 0.6718929254302103, + "grad_norm": 4.639040281423089, + "learning_rate": 1e-06, + "loss": 0.3632, + "step": 1757 + }, + { + "epoch": 0.6722753346080306, + "grad_norm": 2.7845421497937988, + "learning_rate": 1e-06, + "loss": 0.1857, + "step": 1758 + }, + { + "epoch": 0.6726577437858509, + "grad_norm": 4.246098017405284, + "learning_rate": 1e-06, + "loss": 0.2429, + "step": 1759 + }, + { + "epoch": 0.6730401529636711, + "grad_norm": 2.5775100436972322, + "learning_rate": 1e-06, + "loss": 0.1729, + "step": 1760 + }, + { + "epoch": 0.6734225621414914, + "grad_norm": 3.85421123307039, + "learning_rate": 1e-06, + "loss": 0.1487, + "step": 1761 + }, + { + "epoch": 0.6738049713193117, + "grad_norm": 4.20058843940997, + "learning_rate": 1e-06, + "loss": 0.1726, + "step": 1762 + }, + { + "epoch": 0.6741873804971319, + "grad_norm": 4.100446509261812, + "learning_rate": 1e-06, + "loss": 0.3263, + "step": 1763 + }, + { + "epoch": 0.6745697896749522, + "grad_norm": 3.6600236247816706, + "learning_rate": 1e-06, + "loss": 0.5095, + "step": 1764 + }, + { + "epoch": 0.6749521988527725, + "grad_norm": 3.121850372003473, + "learning_rate": 1e-06, + "loss": 0.371, + "step": 1765 + }, + { + "epoch": 0.6753346080305928, + "grad_norm": 2.2848251755027307, + "learning_rate": 1e-06, + "loss": 0.4956, + "step": 1766 + }, + { + "epoch": 0.675717017208413, + "grad_norm": 3.3260047140357583, + "learning_rate": 1e-06, + "loss": 0.4139, + "step": 1767 + }, + { + "epoch": 0.6760994263862332, + "grad_norm": 2.881912545840869, + "learning_rate": 1e-06, + "loss": 0.4155, + "step": 1768 + }, + { + "epoch": 0.6764818355640535, + "grad_norm": 3.464996231319255, + "learning_rate": 1e-06, + "loss": 0.2739, + "step": 1769 + }, + { + "epoch": 0.6768642447418738, + "grad_norm": 2.4101347837311367, + "learning_rate": 1e-06, + "loss": 0.2138, + "step": 1770 + }, + { + "epoch": 0.6772466539196941, + "grad_norm": 2.86264568108073, + "learning_rate": 1e-06, + "loss": 0.105, + "step": 1771 + }, + { + "epoch": 0.6776290630975144, + "grad_norm": 2.4440868096800394, + "learning_rate": 1e-06, + "loss": 0.1686, + "step": 1772 + }, + { + "epoch": 0.6780114722753346, + "grad_norm": 2.3747362693119, + "learning_rate": 1e-06, + "loss": 0.0947, + "step": 1773 + }, + { + "epoch": 0.6783938814531548, + "grad_norm": 7.99965380873262, + "learning_rate": 1e-06, + "loss": 0.222, + "step": 1774 + }, + { + "epoch": 0.6787762906309751, + "grad_norm": 4.487624635798562, + "learning_rate": 1e-06, + "loss": 0.2057, + "step": 1775 + }, + { + "epoch": 0.6791586998087954, + "grad_norm": 4.712107853042115, + "learning_rate": 1e-06, + "loss": 0.4001, + "step": 1776 + }, + { + "epoch": 0.6795411089866157, + "grad_norm": 2.910476728368903, + "learning_rate": 1e-06, + "loss": 0.6801, + "step": 1777 + }, + { + "epoch": 0.679923518164436, + "grad_norm": 2.90719490687544, + "learning_rate": 1e-06, + "loss": 0.4021, + "step": 1778 + }, + { + "epoch": 0.6803059273422563, + "grad_norm": 2.956284382077828, + "learning_rate": 1e-06, + "loss": 0.4785, + "step": 1779 + }, + { + "epoch": 0.6806883365200764, + "grad_norm": 3.7123519611063123, + "learning_rate": 1e-06, + "loss": 0.4861, + "step": 1780 + }, + { + "epoch": 0.6810707456978967, + "grad_norm": 2.3835400080491, + "learning_rate": 1e-06, + "loss": 0.5126, + "step": 1781 + }, + { + "epoch": 0.681453154875717, + "grad_norm": 2.0118450828809378, + "learning_rate": 1e-06, + "loss": 0.3863, + "step": 1782 + }, + { + "epoch": 0.6818355640535373, + "grad_norm": 2.476015337358834, + "learning_rate": 1e-06, + "loss": 0.223, + "step": 1783 + }, + { + "epoch": 0.6822179732313576, + "grad_norm": 2.9900879827204383, + "learning_rate": 1e-06, + "loss": 0.2357, + "step": 1784 + }, + { + "epoch": 0.6826003824091779, + "grad_norm": 4.134742790811418, + "learning_rate": 1e-06, + "loss": 0.2346, + "step": 1785 + }, + { + "epoch": 0.682982791586998, + "grad_norm": 4.008352142945461, + "learning_rate": 1e-06, + "loss": 0.1502, + "step": 1786 + }, + { + "epoch": 0.6833652007648183, + "grad_norm": 3.5316711444715336, + "learning_rate": 1e-06, + "loss": 0.0743, + "step": 1787 + }, + { + "epoch": 0.6837476099426386, + "grad_norm": 4.468546189148733, + "learning_rate": 1e-06, + "loss": 0.2104, + "step": 1788 + }, + { + "epoch": 0.6841300191204589, + "grad_norm": 4.987051600900471, + "learning_rate": 1e-06, + "loss": 0.6941, + "step": 1789 + }, + { + "epoch": 0.6845124282982792, + "grad_norm": 3.0797028346190394, + "learning_rate": 1e-06, + "loss": 0.508, + "step": 1790 + }, + { + "epoch": 0.6848948374760995, + "grad_norm": 3.0170605656983893, + "learning_rate": 1e-06, + "loss": 0.4026, + "step": 1791 + }, + { + "epoch": 0.6852772466539196, + "grad_norm": 3.1333803342445936, + "learning_rate": 1e-06, + "loss": 0.3969, + "step": 1792 + }, + { + "epoch": 0.6856596558317399, + "grad_norm": 2.9137190138234184, + "learning_rate": 1e-06, + "loss": 0.363, + "step": 1793 + }, + { + "epoch": 0.6860420650095602, + "grad_norm": 3.12110795840203, + "learning_rate": 1e-06, + "loss": 0.3825, + "step": 1794 + }, + { + "epoch": 0.6864244741873805, + "grad_norm": 2.0927769691626175, + "learning_rate": 1e-06, + "loss": 0.2357, + "step": 1795 + }, + { + "epoch": 0.6868068833652008, + "grad_norm": 3.637968870796287, + "learning_rate": 1e-06, + "loss": 0.3408, + "step": 1796 + }, + { + "epoch": 0.6871892925430211, + "grad_norm": 2.7240100707373522, + "learning_rate": 1e-06, + "loss": 0.222, + "step": 1797 + }, + { + "epoch": 0.6875717017208413, + "grad_norm": 4.31605781080897, + "learning_rate": 1e-06, + "loss": 0.1569, + "step": 1798 + }, + { + "epoch": 0.6879541108986615, + "grad_norm": 4.558239763730331, + "learning_rate": 1e-06, + "loss": 0.1587, + "step": 1799 + }, + { + "epoch": 0.6883365200764818, + "grad_norm": 4.345872538832599, + "learning_rate": 1e-06, + "loss": 0.1445, + "step": 1800 + }, + { + "epoch": 0.6887189292543021, + "grad_norm": 5.634707064795355, + "learning_rate": 1e-06, + "loss": 0.4121, + "step": 1801 + }, + { + "epoch": 0.6891013384321224, + "grad_norm": 3.095011906516201, + "learning_rate": 1e-06, + "loss": 0.471, + "step": 1802 + }, + { + "epoch": 0.6894837476099427, + "grad_norm": 6.63452931394001, + "learning_rate": 1e-06, + "loss": 0.4599, + "step": 1803 + }, + { + "epoch": 0.6898661567877629, + "grad_norm": 3.9377954765746828, + "learning_rate": 1e-06, + "loss": 0.2971, + "step": 1804 + }, + { + "epoch": 0.6902485659655831, + "grad_norm": 2.9167459931258204, + "learning_rate": 1e-06, + "loss": 0.546, + "step": 1805 + }, + { + "epoch": 0.6906309751434034, + "grad_norm": 3.3406928516773355, + "learning_rate": 1e-06, + "loss": 0.5512, + "step": 1806 + }, + { + "epoch": 0.6910133843212237, + "grad_norm": 3.420617138020482, + "learning_rate": 1e-06, + "loss": 0.2173, + "step": 1807 + }, + { + "epoch": 0.691395793499044, + "grad_norm": 2.6627656795135697, + "learning_rate": 1e-06, + "loss": 0.2218, + "step": 1808 + }, + { + "epoch": 0.6917782026768643, + "grad_norm": 3.961977489209145, + "learning_rate": 1e-06, + "loss": 0.1896, + "step": 1809 + }, + { + "epoch": 0.6921606118546845, + "grad_norm": 3.6100959883421306, + "learning_rate": 1e-06, + "loss": 0.0963, + "step": 1810 + }, + { + "epoch": 0.6925430210325048, + "grad_norm": 5.676831982382531, + "learning_rate": 1e-06, + "loss": 0.2514, + "step": 1811 + }, + { + "epoch": 0.692925430210325, + "grad_norm": 4.603937595014096, + "learning_rate": 1e-06, + "loss": 0.154, + "step": 1812 + }, + { + "epoch": 0.6933078393881453, + "grad_norm": 6.680118372890564, + "learning_rate": 1e-06, + "loss": 0.1824, + "step": 1813 + }, + { + "epoch": 0.6936902485659656, + "grad_norm": 4.771502762960885, + "learning_rate": 1e-06, + "loss": 0.75, + "step": 1814 + }, + { + "epoch": 0.6940726577437859, + "grad_norm": 3.6447441145215334, + "learning_rate": 1e-06, + "loss": 0.6686, + "step": 1815 + }, + { + "epoch": 0.6944550669216061, + "grad_norm": 2.5821162915328326, + "learning_rate": 1e-06, + "loss": 0.3767, + "step": 1816 + }, + { + "epoch": 0.6948374760994264, + "grad_norm": 3.5827351152104057, + "learning_rate": 1e-06, + "loss": 0.409, + "step": 1817 + }, + { + "epoch": 0.6952198852772467, + "grad_norm": 2.1990456114892005, + "learning_rate": 1e-06, + "loss": 0.3815, + "step": 1818 + }, + { + "epoch": 0.6956022944550669, + "grad_norm": 3.148419780066062, + "learning_rate": 1e-06, + "loss": 0.2984, + "step": 1819 + }, + { + "epoch": 0.6959847036328872, + "grad_norm": 3.385631628792315, + "learning_rate": 1e-06, + "loss": 0.2934, + "step": 1820 + }, + { + "epoch": 0.6963671128107075, + "grad_norm": 3.236165016682208, + "learning_rate": 1e-06, + "loss": 0.2187, + "step": 1821 + }, + { + "epoch": 0.6967495219885277, + "grad_norm": 2.7217993461877756, + "learning_rate": 1e-06, + "loss": 0.1829, + "step": 1822 + }, + { + "epoch": 0.697131931166348, + "grad_norm": 4.96599324843293, + "learning_rate": 1e-06, + "loss": 0.219, + "step": 1823 + }, + { + "epoch": 0.6975143403441683, + "grad_norm": 3.333908428808765, + "learning_rate": 1e-06, + "loss": 0.0983, + "step": 1824 + }, + { + "epoch": 0.6978967495219885, + "grad_norm": 5.570685033124853, + "learning_rate": 1e-06, + "loss": 0.263, + "step": 1825 + }, + { + "epoch": 0.6982791586998088, + "grad_norm": 6.411933573492392, + "learning_rate": 1e-06, + "loss": 0.2969, + "step": 1826 + }, + { + "epoch": 0.698661567877629, + "grad_norm": 3.775101636314174, + "learning_rate": 1e-06, + "loss": 0.5427, + "step": 1827 + }, + { + "epoch": 0.6990439770554493, + "grad_norm": 7.071409024518453, + "learning_rate": 1e-06, + "loss": 0.6173, + "step": 1828 + }, + { + "epoch": 0.6994263862332696, + "grad_norm": 3.762640439070742, + "learning_rate": 1e-06, + "loss": 0.3903, + "step": 1829 + }, + { + "epoch": 0.6998087954110899, + "grad_norm": 3.2679992513854197, + "learning_rate": 1e-06, + "loss": 0.3204, + "step": 1830 + }, + { + "epoch": 0.7001912045889102, + "grad_norm": 3.7052534079024153, + "learning_rate": 1e-06, + "loss": 0.577, + "step": 1831 + }, + { + "epoch": 0.7005736137667304, + "grad_norm": 2.291632033577672, + "learning_rate": 1e-06, + "loss": 0.3083, + "step": 1832 + }, + { + "epoch": 0.7009560229445506, + "grad_norm": 3.1704137283039517, + "learning_rate": 1e-06, + "loss": 0.2115, + "step": 1833 + }, + { + "epoch": 0.7013384321223709, + "grad_norm": 2.7626738238990645, + "learning_rate": 1e-06, + "loss": 0.2287, + "step": 1834 + }, + { + "epoch": 0.7017208413001912, + "grad_norm": 3.2028281469723106, + "learning_rate": 1e-06, + "loss": 0.2692, + "step": 1835 + }, + { + "epoch": 0.7021032504780115, + "grad_norm": 3.9638971410776977, + "learning_rate": 1e-06, + "loss": 0.121, + "step": 1836 + }, + { + "epoch": 0.7024856596558318, + "grad_norm": 2.8519179867245357, + "learning_rate": 1e-06, + "loss": 0.0548, + "step": 1837 + }, + { + "epoch": 0.702868068833652, + "grad_norm": 4.584365497311428, + "learning_rate": 1e-06, + "loss": 0.2794, + "step": 1838 + }, + { + "epoch": 0.7032504780114722, + "grad_norm": 5.193457985207564, + "learning_rate": 1e-06, + "loss": 0.4874, + "step": 1839 + }, + { + "epoch": 0.7036328871892925, + "grad_norm": 3.693068744555693, + "learning_rate": 1e-06, + "loss": 0.6087, + "step": 1840 + }, + { + "epoch": 0.7040152963671128, + "grad_norm": 2.6873466093069327, + "learning_rate": 1e-06, + "loss": 0.4215, + "step": 1841 + }, + { + "epoch": 0.7043977055449331, + "grad_norm": 4.868728516204405, + "learning_rate": 1e-06, + "loss": 0.4795, + "step": 1842 + }, + { + "epoch": 0.7047801147227534, + "grad_norm": 2.704295087056366, + "learning_rate": 1e-06, + "loss": 0.3664, + "step": 1843 + }, + { + "epoch": 0.7051625239005737, + "grad_norm": 2.0360407746734377, + "learning_rate": 1e-06, + "loss": 0.181, + "step": 1844 + }, + { + "epoch": 0.7055449330783938, + "grad_norm": 2.5505686929886457, + "learning_rate": 1e-06, + "loss": 0.1876, + "step": 1845 + }, + { + "epoch": 0.7059273422562141, + "grad_norm": 2.925389588323107, + "learning_rate": 1e-06, + "loss": 0.2242, + "step": 1846 + }, + { + "epoch": 0.7063097514340344, + "grad_norm": 3.9714455649276004, + "learning_rate": 1e-06, + "loss": 0.2156, + "step": 1847 + }, + { + "epoch": 0.7066921606118547, + "grad_norm": 2.8432737517589484, + "learning_rate": 1e-06, + "loss": 0.1292, + "step": 1848 + }, + { + "epoch": 0.707074569789675, + "grad_norm": 3.3187594023892824, + "learning_rate": 1e-06, + "loss": 0.0879, + "step": 1849 + }, + { + "epoch": 0.7074569789674953, + "grad_norm": 4.174831071736803, + "learning_rate": 1e-06, + "loss": 0.1524, + "step": 1850 + }, + { + "epoch": 0.7078393881453154, + "grad_norm": 4.211490516939054, + "learning_rate": 1e-06, + "loss": 0.3833, + "step": 1851 + }, + { + "epoch": 0.7082217973231357, + "grad_norm": 4.472239592355886, + "learning_rate": 1e-06, + "loss": 0.7896, + "step": 1852 + }, + { + "epoch": 0.708604206500956, + "grad_norm": 3.334509800882675, + "learning_rate": 1e-06, + "loss": 0.4439, + "step": 1853 + }, + { + "epoch": 0.7089866156787763, + "grad_norm": 2.205621049401282, + "learning_rate": 1e-06, + "loss": 0.5861, + "step": 1854 + }, + { + "epoch": 0.7093690248565966, + "grad_norm": 3.6115895533078644, + "learning_rate": 1e-06, + "loss": 0.5834, + "step": 1855 + }, + { + "epoch": 0.7097514340344169, + "grad_norm": 2.5330680628810085, + "learning_rate": 1e-06, + "loss": 0.41, + "step": 1856 + }, + { + "epoch": 0.710133843212237, + "grad_norm": 1.9168456104613885, + "learning_rate": 1e-06, + "loss": 0.3146, + "step": 1857 + }, + { + "epoch": 0.7105162523900573, + "grad_norm": 3.2846287361661197, + "learning_rate": 1e-06, + "loss": 0.3478, + "step": 1858 + }, + { + "epoch": 0.7108986615678776, + "grad_norm": 2.62089053416537, + "learning_rate": 1e-06, + "loss": 0.1584, + "step": 1859 + }, + { + "epoch": 0.7112810707456979, + "grad_norm": 4.912714301717813, + "learning_rate": 1e-06, + "loss": 0.2955, + "step": 1860 + }, + { + "epoch": 0.7116634799235182, + "grad_norm": 3.5978899122935775, + "learning_rate": 1e-06, + "loss": 0.0844, + "step": 1861 + }, + { + "epoch": 0.7120458891013385, + "grad_norm": 4.862546027193691, + "learning_rate": 1e-06, + "loss": 0.1187, + "step": 1862 + }, + { + "epoch": 0.7124282982791587, + "grad_norm": 6.103384061082042, + "learning_rate": 1e-06, + "loss": 0.1864, + "step": 1863 + }, + { + "epoch": 0.7128107074569789, + "grad_norm": 4.6811541518214534, + "learning_rate": 1e-06, + "loss": 0.565, + "step": 1864 + }, + { + "epoch": 0.7131931166347992, + "grad_norm": 3.4928539075510487, + "learning_rate": 1e-06, + "loss": 0.5202, + "step": 1865 + }, + { + "epoch": 0.7135755258126195, + "grad_norm": 3.4404708509015625, + "learning_rate": 1e-06, + "loss": 0.3804, + "step": 1866 + }, + { + "epoch": 0.7139579349904398, + "grad_norm": 2.1495928622815477, + "learning_rate": 1e-06, + "loss": 0.5346, + "step": 1867 + }, + { + "epoch": 0.7143403441682601, + "grad_norm": 2.6974655618670087, + "learning_rate": 1e-06, + "loss": 0.505, + "step": 1868 + }, + { + "epoch": 0.7147227533460803, + "grad_norm": 4.0097496898276885, + "learning_rate": 1e-06, + "loss": 0.3809, + "step": 1869 + }, + { + "epoch": 0.7151051625239006, + "grad_norm": 3.45407519493576, + "learning_rate": 1e-06, + "loss": 0.3726, + "step": 1870 + }, + { + "epoch": 0.7154875717017208, + "grad_norm": 3.2900777595397894, + "learning_rate": 1e-06, + "loss": 0.2127, + "step": 1871 + }, + { + "epoch": 0.7158699808795411, + "grad_norm": 4.368430736920466, + "learning_rate": 1e-06, + "loss": 0.1821, + "step": 1872 + }, + { + "epoch": 0.7162523900573614, + "grad_norm": 3.4827469242403564, + "learning_rate": 1e-06, + "loss": 0.2243, + "step": 1873 + }, + { + "epoch": 0.7166347992351817, + "grad_norm": 4.181869137118051, + "learning_rate": 1e-06, + "loss": 0.1036, + "step": 1874 + }, + { + "epoch": 0.7170172084130019, + "grad_norm": 6.611198300054725, + "learning_rate": 1e-06, + "loss": 0.2067, + "step": 1875 + }, + { + "epoch": 0.7173996175908222, + "grad_norm": 6.57649236067522, + "learning_rate": 1e-06, + "loss": 0.6051, + "step": 1876 + }, + { + "epoch": 0.7177820267686424, + "grad_norm": 4.121052326880088, + "learning_rate": 1e-06, + "loss": 0.6721, + "step": 1877 + }, + { + "epoch": 0.7181644359464627, + "grad_norm": 3.35239881916105, + "learning_rate": 1e-06, + "loss": 0.5213, + "step": 1878 + }, + { + "epoch": 0.718546845124283, + "grad_norm": 3.1891759318654165, + "learning_rate": 1e-06, + "loss": 0.3168, + "step": 1879 + }, + { + "epoch": 0.7189292543021033, + "grad_norm": 3.4945218948112817, + "learning_rate": 1e-06, + "loss": 0.2938, + "step": 1880 + }, + { + "epoch": 0.7193116634799235, + "grad_norm": 3.5559991784969105, + "learning_rate": 1e-06, + "loss": 0.2688, + "step": 1881 + }, + { + "epoch": 0.7196940726577438, + "grad_norm": 2.812100784790976, + "learning_rate": 1e-06, + "loss": 0.3084, + "step": 1882 + }, + { + "epoch": 0.720076481835564, + "grad_norm": 2.4886104539636467, + "learning_rate": 1e-06, + "loss": 0.1731, + "step": 1883 + }, + { + "epoch": 0.7204588910133843, + "grad_norm": 3.2822932446830215, + "learning_rate": 1e-06, + "loss": 0.1773, + "step": 1884 + }, + { + "epoch": 0.7208413001912046, + "grad_norm": 2.9886354722655595, + "learning_rate": 1e-06, + "loss": 0.1471, + "step": 1885 + }, + { + "epoch": 0.7212237093690248, + "grad_norm": 4.083413389290498, + "learning_rate": 1e-06, + "loss": 0.162, + "step": 1886 + }, + { + "epoch": 0.7216061185468451, + "grad_norm": 3.2348496830492492, + "learning_rate": 1e-06, + "loss": 0.1084, + "step": 1887 + }, + { + "epoch": 0.7219885277246654, + "grad_norm": 6.62174994443766, + "learning_rate": 1e-06, + "loss": 0.3879, + "step": 1888 + }, + { + "epoch": 0.7223709369024857, + "grad_norm": 4.169378212144267, + "learning_rate": 1e-06, + "loss": 0.5458, + "step": 1889 + }, + { + "epoch": 0.722753346080306, + "grad_norm": 4.000780506279222, + "learning_rate": 1e-06, + "loss": 0.3392, + "step": 1890 + }, + { + "epoch": 0.7231357552581262, + "grad_norm": 2.150841162498716, + "learning_rate": 1e-06, + "loss": 0.1693, + "step": 1891 + }, + { + "epoch": 0.7235181644359464, + "grad_norm": 2.410613030623257, + "learning_rate": 1e-06, + "loss": 0.5151, + "step": 1892 + }, + { + "epoch": 0.7239005736137667, + "grad_norm": 2.9877919712803735, + "learning_rate": 1e-06, + "loss": 0.5863, + "step": 1893 + }, + { + "epoch": 0.724282982791587, + "grad_norm": 4.180110503565843, + "learning_rate": 1e-06, + "loss": 0.4053, + "step": 1894 + }, + { + "epoch": 0.7246653919694073, + "grad_norm": 3.0684353669078828, + "learning_rate": 1e-06, + "loss": 0.4632, + "step": 1895 + }, + { + "epoch": 0.7250478011472276, + "grad_norm": 3.7202368576774356, + "learning_rate": 1e-06, + "loss": 0.2632, + "step": 1896 + }, + { + "epoch": 0.7254302103250478, + "grad_norm": 2.7200966492479637, + "learning_rate": 1e-06, + "loss": 0.2772, + "step": 1897 + }, + { + "epoch": 0.725812619502868, + "grad_norm": 2.4582363260386018, + "learning_rate": 1e-06, + "loss": 0.1286, + "step": 1898 + }, + { + "epoch": 0.7261950286806883, + "grad_norm": 4.900075927457515, + "learning_rate": 1e-06, + "loss": 0.1468, + "step": 1899 + }, + { + "epoch": 0.7265774378585086, + "grad_norm": 5.773966722403739, + "learning_rate": 1e-06, + "loss": 0.1803, + "step": 1900 + }, + { + "epoch": 0.7269598470363289, + "grad_norm": 4.439438517353882, + "learning_rate": 1e-06, + "loss": 0.4792, + "step": 1901 + }, + { + "epoch": 0.7273422562141492, + "grad_norm": 3.5652131574126233, + "learning_rate": 1e-06, + "loss": 0.5077, + "step": 1902 + }, + { + "epoch": 0.7277246653919694, + "grad_norm": 4.41994129400838, + "learning_rate": 1e-06, + "loss": 0.4999, + "step": 1903 + }, + { + "epoch": 0.7281070745697896, + "grad_norm": 2.0358063287410446, + "learning_rate": 1e-06, + "loss": 0.5988, + "step": 1904 + }, + { + "epoch": 0.7284894837476099, + "grad_norm": 2.6470563770108355, + "learning_rate": 1e-06, + "loss": 0.3909, + "step": 1905 + }, + { + "epoch": 0.7288718929254302, + "grad_norm": 3.8597156204460625, + "learning_rate": 1e-06, + "loss": 0.3542, + "step": 1906 + }, + { + "epoch": 0.7292543021032505, + "grad_norm": 3.1934985395899718, + "learning_rate": 1e-06, + "loss": 0.2437, + "step": 1907 + }, + { + "epoch": 0.7296367112810708, + "grad_norm": 3.2701656040639486, + "learning_rate": 1e-06, + "loss": 0.1474, + "step": 1908 + }, + { + "epoch": 0.7300191204588911, + "grad_norm": 2.4833295049787196, + "learning_rate": 1e-06, + "loss": 0.2791, + "step": 1909 + }, + { + "epoch": 0.7304015296367112, + "grad_norm": 3.7614979268161752, + "learning_rate": 1e-06, + "loss": 0.1431, + "step": 1910 + }, + { + "epoch": 0.7307839388145315, + "grad_norm": 4.869833139933135, + "learning_rate": 1e-06, + "loss": 0.2185, + "step": 1911 + }, + { + "epoch": 0.7311663479923518, + "grad_norm": 4.298374649525742, + "learning_rate": 1e-06, + "loss": 0.1067, + "step": 1912 + }, + { + "epoch": 0.7315487571701721, + "grad_norm": 6.476052700174656, + "learning_rate": 1e-06, + "loss": 0.3223, + "step": 1913 + }, + { + "epoch": 0.7319311663479924, + "grad_norm": 4.323621920618311, + "learning_rate": 1e-06, + "loss": 0.3189, + "step": 1914 + }, + { + "epoch": 0.7323135755258127, + "grad_norm": 2.657115211943629, + "learning_rate": 1e-06, + "loss": 0.3768, + "step": 1915 + }, + { + "epoch": 0.7326959847036328, + "grad_norm": 3.37992019875126, + "learning_rate": 1e-06, + "loss": 0.3918, + "step": 1916 + }, + { + "epoch": 0.7330783938814531, + "grad_norm": 3.1789435938762542, + "learning_rate": 1e-06, + "loss": 0.3355, + "step": 1917 + }, + { + "epoch": 0.7334608030592734, + "grad_norm": 3.519897035479995, + "learning_rate": 1e-06, + "loss": 0.3091, + "step": 1918 + }, + { + "epoch": 0.7338432122370937, + "grad_norm": 2.9668340021438433, + "learning_rate": 1e-06, + "loss": 0.3655, + "step": 1919 + }, + { + "epoch": 0.734225621414914, + "grad_norm": 4.146447115500284, + "learning_rate": 1e-06, + "loss": 0.2495, + "step": 1920 + }, + { + "epoch": 0.7346080305927343, + "grad_norm": 2.6198207077530102, + "learning_rate": 1e-06, + "loss": 0.1357, + "step": 1921 + }, + { + "epoch": 0.7349904397705544, + "grad_norm": 3.949068664657057, + "learning_rate": 1e-06, + "loss": 0.2524, + "step": 1922 + }, + { + "epoch": 0.7353728489483747, + "grad_norm": 4.445273669028006, + "learning_rate": 1e-06, + "loss": 0.152, + "step": 1923 + }, + { + "epoch": 0.735755258126195, + "grad_norm": 3.1655209794422565, + "learning_rate": 1e-06, + "loss": 0.1172, + "step": 1924 + }, + { + "epoch": 0.7361376673040153, + "grad_norm": 3.7767993806941504, + "learning_rate": 1e-06, + "loss": 0.1635, + "step": 1925 + }, + { + "epoch": 0.7365200764818356, + "grad_norm": 4.993509848767819, + "learning_rate": 1e-06, + "loss": 0.2869, + "step": 1926 + }, + { + "epoch": 0.7369024856596559, + "grad_norm": 3.545566401462217, + "learning_rate": 1e-06, + "loss": 0.5885, + "step": 1927 + }, + { + "epoch": 0.7372848948374761, + "grad_norm": 10.560810382843867, + "learning_rate": 1e-06, + "loss": 0.4926, + "step": 1928 + }, + { + "epoch": 0.7376673040152963, + "grad_norm": 2.330562058466115, + "learning_rate": 1e-06, + "loss": 0.4563, + "step": 1929 + }, + { + "epoch": 0.7380497131931166, + "grad_norm": 2.144996022929882, + "learning_rate": 1e-06, + "loss": 0.314, + "step": 1930 + }, + { + "epoch": 0.7384321223709369, + "grad_norm": 3.0278741974794325, + "learning_rate": 1e-06, + "loss": 0.3752, + "step": 1931 + }, + { + "epoch": 0.7388145315487572, + "grad_norm": 2.002602790927952, + "learning_rate": 1e-06, + "loss": 0.3564, + "step": 1932 + }, + { + "epoch": 0.7391969407265775, + "grad_norm": 3.743961558788568, + "learning_rate": 1e-06, + "loss": 0.399, + "step": 1933 + }, + { + "epoch": 0.7395793499043977, + "grad_norm": 3.8753553658246336, + "learning_rate": 1e-06, + "loss": 0.3506, + "step": 1934 + }, + { + "epoch": 0.739961759082218, + "grad_norm": 3.002976847952848, + "learning_rate": 1e-06, + "loss": 0.2142, + "step": 1935 + }, + { + "epoch": 0.7403441682600382, + "grad_norm": 2.0153218122643652, + "learning_rate": 1e-06, + "loss": 0.057, + "step": 1936 + }, + { + "epoch": 0.7407265774378585, + "grad_norm": 3.0885672265335886, + "learning_rate": 1e-06, + "loss": 0.0932, + "step": 1937 + }, + { + "epoch": 0.7411089866156788, + "grad_norm": 4.183245419088883, + "learning_rate": 1e-06, + "loss": 0.2852, + "step": 1938 + }, + { + "epoch": 0.7414913957934991, + "grad_norm": 4.42740417514626, + "learning_rate": 1e-06, + "loss": 0.556, + "step": 1939 + }, + { + "epoch": 0.7418738049713193, + "grad_norm": 3.603905699585369, + "learning_rate": 1e-06, + "loss": 0.4392, + "step": 1940 + }, + { + "epoch": 0.7422562141491396, + "grad_norm": 3.007430411506426, + "learning_rate": 1e-06, + "loss": 0.3079, + "step": 1941 + }, + { + "epoch": 0.7426386233269598, + "grad_norm": 1.5184469006913335, + "learning_rate": 1e-06, + "loss": 0.284, + "step": 1942 + }, + { + "epoch": 0.7430210325047801, + "grad_norm": 2.088187044118794, + "learning_rate": 1e-06, + "loss": 0.3736, + "step": 1943 + }, + { + "epoch": 0.7434034416826004, + "grad_norm": 3.5731852053310473, + "learning_rate": 1e-06, + "loss": 0.5477, + "step": 1944 + }, + { + "epoch": 0.7437858508604207, + "grad_norm": 2.961146525953916, + "learning_rate": 1e-06, + "loss": 0.3065, + "step": 1945 + }, + { + "epoch": 0.7441682600382409, + "grad_norm": 3.6100323231272373, + "learning_rate": 1e-06, + "loss": 0.2124, + "step": 1946 + }, + { + "epoch": 0.7445506692160612, + "grad_norm": 2.7169788774070143, + "learning_rate": 1e-06, + "loss": 0.252, + "step": 1947 + }, + { + "epoch": 0.7449330783938815, + "grad_norm": 4.019746201887364, + "learning_rate": 1e-06, + "loss": 0.2892, + "step": 1948 + }, + { + "epoch": 0.7453154875717017, + "grad_norm": 3.0840721963429876, + "learning_rate": 1e-06, + "loss": 0.1068, + "step": 1949 + }, + { + "epoch": 0.745697896749522, + "grad_norm": 4.265207347055815, + "learning_rate": 1e-06, + "loss": 0.2368, + "step": 1950 + }, + { + "epoch": 0.7460803059273422, + "grad_norm": 5.169757882209663, + "learning_rate": 1e-06, + "loss": 0.4368, + "step": 1951 + }, + { + "epoch": 0.7464627151051625, + "grad_norm": 3.5514206850845325, + "learning_rate": 1e-06, + "loss": 0.5124, + "step": 1952 + }, + { + "epoch": 0.7468451242829828, + "grad_norm": 3.3868851670293685, + "learning_rate": 1e-06, + "loss": 0.3539, + "step": 1953 + }, + { + "epoch": 0.7472275334608031, + "grad_norm": 3.3671307127516408, + "learning_rate": 1e-06, + "loss": 0.3289, + "step": 1954 + }, + { + "epoch": 0.7476099426386233, + "grad_norm": 3.7899307357377197, + "learning_rate": 1e-06, + "loss": 0.4966, + "step": 1955 + }, + { + "epoch": 0.7479923518164436, + "grad_norm": 2.3303891635442464, + "learning_rate": 1e-06, + "loss": 0.2466, + "step": 1956 + }, + { + "epoch": 0.7483747609942638, + "grad_norm": 4.211143587692316, + "learning_rate": 1e-06, + "loss": 0.468, + "step": 1957 + }, + { + "epoch": 0.7487571701720841, + "grad_norm": 2.4047094590719373, + "learning_rate": 1e-06, + "loss": 0.2326, + "step": 1958 + }, + { + "epoch": 0.7491395793499044, + "grad_norm": 3.3531454834041328, + "learning_rate": 1e-06, + "loss": 0.3103, + "step": 1959 + }, + { + "epoch": 0.7495219885277247, + "grad_norm": 3.2136104994727934, + "learning_rate": 1e-06, + "loss": 0.2424, + "step": 1960 + }, + { + "epoch": 0.749904397705545, + "grad_norm": 4.63095925431355, + "learning_rate": 1e-06, + "loss": 0.1649, + "step": 1961 + }, + { + "epoch": 0.7502868068833652, + "grad_norm": 4.552692899183663, + "learning_rate": 1e-06, + "loss": 0.1504, + "step": 1962 + }, + { + "epoch": 0.7506692160611854, + "grad_norm": 6.039571289106634, + "learning_rate": 1e-06, + "loss": 0.3453, + "step": 1963 + }, + { + "epoch": 0.7510516252390057, + "grad_norm": 6.0466352706032405, + "learning_rate": 1e-06, + "loss": 0.5867, + "step": 1964 + }, + { + "epoch": 0.751434034416826, + "grad_norm": 2.1601803107413398, + "learning_rate": 1e-06, + "loss": 0.3917, + "step": 1965 + }, + { + "epoch": 0.7518164435946463, + "grad_norm": 1.9370251965949365, + "learning_rate": 1e-06, + "loss": 0.1983, + "step": 1966 + }, + { + "epoch": 0.7521988527724666, + "grad_norm": 2.6129106335779917, + "learning_rate": 1e-06, + "loss": 0.1441, + "step": 1967 + }, + { + "epoch": 0.7525812619502869, + "grad_norm": 4.027613455975739, + "learning_rate": 1e-06, + "loss": 0.4051, + "step": 1968 + }, + { + "epoch": 0.752963671128107, + "grad_norm": 3.8408939326388887, + "learning_rate": 1e-06, + "loss": 0.4728, + "step": 1969 + }, + { + "epoch": 0.7533460803059273, + "grad_norm": 3.506971908478582, + "learning_rate": 1e-06, + "loss": 0.3481, + "step": 1970 + }, + { + "epoch": 0.7537284894837476, + "grad_norm": 3.4824632266719013, + "learning_rate": 1e-06, + "loss": 0.3397, + "step": 1971 + }, + { + "epoch": 0.7541108986615679, + "grad_norm": 3.8161371110799367, + "learning_rate": 1e-06, + "loss": 0.2475, + "step": 1972 + }, + { + "epoch": 0.7544933078393882, + "grad_norm": 3.2389161300894944, + "learning_rate": 1e-06, + "loss": 0.1609, + "step": 1973 + }, + { + "epoch": 0.7548757170172085, + "grad_norm": 3.676404650826159, + "learning_rate": 1e-06, + "loss": 0.1427, + "step": 1974 + }, + { + "epoch": 0.7552581261950286, + "grad_norm": 4.308149492976423, + "learning_rate": 1e-06, + "loss": 0.1601, + "step": 1975 + }, + { + "epoch": 0.7556405353728489, + "grad_norm": 4.053353915371422, + "learning_rate": 1e-06, + "loss": 0.5546, + "step": 1976 + }, + { + "epoch": 0.7560229445506692, + "grad_norm": 3.8180897493119867, + "learning_rate": 1e-06, + "loss": 0.4642, + "step": 1977 + }, + { + "epoch": 0.7564053537284895, + "grad_norm": 3.779026354982245, + "learning_rate": 1e-06, + "loss": 0.41, + "step": 1978 + }, + { + "epoch": 0.7567877629063098, + "grad_norm": 3.352605484021963, + "learning_rate": 1e-06, + "loss": 0.4069, + "step": 1979 + }, + { + "epoch": 0.7571701720841301, + "grad_norm": 2.5243564503217946, + "learning_rate": 1e-06, + "loss": 0.3045, + "step": 1980 + }, + { + "epoch": 0.7575525812619502, + "grad_norm": 3.1271164402913008, + "learning_rate": 1e-06, + "loss": 0.178, + "step": 1981 + }, + { + "epoch": 0.7579349904397705, + "grad_norm": 2.1511791145309562, + "learning_rate": 1e-06, + "loss": 0.281, + "step": 1982 + }, + { + "epoch": 0.7583173996175908, + "grad_norm": 3.538560709632089, + "learning_rate": 1e-06, + "loss": 0.3045, + "step": 1983 + }, + { + "epoch": 0.7586998087954111, + "grad_norm": 2.408642755802393, + "learning_rate": 1e-06, + "loss": 0.2651, + "step": 1984 + }, + { + "epoch": 0.7590822179732314, + "grad_norm": 5.043234631231656, + "learning_rate": 1e-06, + "loss": 0.295, + "step": 1985 + }, + { + "epoch": 0.7594646271510517, + "grad_norm": 2.754032299863494, + "learning_rate": 1e-06, + "loss": 0.0662, + "step": 1986 + }, + { + "epoch": 0.7598470363288718, + "grad_norm": 5.2173929791516915, + "learning_rate": 1e-06, + "loss": 0.1823, + "step": 1987 + }, + { + "epoch": 0.7602294455066921, + "grad_norm": 4.407626755352812, + "learning_rate": 1e-06, + "loss": 0.2591, + "step": 1988 + }, + { + "epoch": 0.7606118546845124, + "grad_norm": 3.548506575751531, + "learning_rate": 1e-06, + "loss": 0.4805, + "step": 1989 + }, + { + "epoch": 0.7609942638623327, + "grad_norm": 3.728761547964631, + "learning_rate": 1e-06, + "loss": 0.4266, + "step": 1990 + }, + { + "epoch": 0.761376673040153, + "grad_norm": 2.82515176686109, + "learning_rate": 1e-06, + "loss": 0.527, + "step": 1991 + }, + { + "epoch": 0.7617590822179733, + "grad_norm": 2.477642605985699, + "learning_rate": 1e-06, + "loss": 0.3361, + "step": 1992 + }, + { + "epoch": 0.7621414913957935, + "grad_norm": 2.260690827744086, + "learning_rate": 1e-06, + "loss": 0.4581, + "step": 1993 + }, + { + "epoch": 0.7625239005736137, + "grad_norm": 4.9252590595408146, + "learning_rate": 1e-06, + "loss": 0.2767, + "step": 1994 + }, + { + "epoch": 0.762906309751434, + "grad_norm": 2.867861967247307, + "learning_rate": 1e-06, + "loss": 0.1954, + "step": 1995 + }, + { + "epoch": 0.7632887189292543, + "grad_norm": 3.9892459789833667, + "learning_rate": 1e-06, + "loss": 0.173, + "step": 1996 + }, + { + "epoch": 0.7636711281070746, + "grad_norm": 2.8326691428688644, + "learning_rate": 1e-06, + "loss": 0.2761, + "step": 1997 + }, + { + "epoch": 0.7640535372848949, + "grad_norm": 5.023119309761381, + "learning_rate": 1e-06, + "loss": 0.2801, + "step": 1998 + }, + { + "epoch": 0.7644359464627151, + "grad_norm": 4.259385684497444, + "learning_rate": 1e-06, + "loss": 0.2594, + "step": 1999 + }, + { + "epoch": 0.7648183556405354, + "grad_norm": 4.156254359651731, + "learning_rate": 1e-06, + "loss": 0.1968, + "step": 2000 + }, + { + "epoch": 0.7648183556405354, + "eval_runtime": 805.9648, + "eval_samples_per_second": 1.903, + "eval_steps_per_second": 0.476, + "step": 2000 + }, + { + "epoch": 0.7652007648183556, + "grad_norm": 4.5302225987056355, + "learning_rate": 1e-06, + "loss": 0.4251, + "step": 2001 + }, + { + "epoch": 0.7655831739961759, + "grad_norm": 3.6469216311888584, + "learning_rate": 1e-06, + "loss": 0.5221, + "step": 2002 + }, + { + "epoch": 0.7659655831739962, + "grad_norm": 4.288490576907294, + "learning_rate": 1e-06, + "loss": 0.6978, + "step": 2003 + }, + { + "epoch": 0.7663479923518165, + "grad_norm": 2.6543820264349183, + "learning_rate": 1e-06, + "loss": 0.3381, + "step": 2004 + }, + { + "epoch": 0.7667304015296367, + "grad_norm": 1.9267132652921741, + "learning_rate": 1e-06, + "loss": 0.3503, + "step": 2005 + }, + { + "epoch": 0.767112810707457, + "grad_norm": 2.7465367617797094, + "learning_rate": 1e-06, + "loss": 0.366, + "step": 2006 + }, + { + "epoch": 0.7674952198852772, + "grad_norm": 5.328444306322995, + "learning_rate": 1e-06, + "loss": 0.49, + "step": 2007 + }, + { + "epoch": 0.7678776290630975, + "grad_norm": 3.017219398681553, + "learning_rate": 1e-06, + "loss": 0.2735, + "step": 2008 + }, + { + "epoch": 0.7682600382409178, + "grad_norm": 1.95948573625478, + "learning_rate": 1e-06, + "loss": 0.2887, + "step": 2009 + }, + { + "epoch": 0.768642447418738, + "grad_norm": 3.835460003814006, + "learning_rate": 1e-06, + "loss": 0.1644, + "step": 2010 + }, + { + "epoch": 0.7690248565965583, + "grad_norm": 2.9154317875633056, + "learning_rate": 1e-06, + "loss": 0.2051, + "step": 2011 + }, + { + "epoch": 0.7694072657743786, + "grad_norm": 3.836144342692119, + "learning_rate": 1e-06, + "loss": 0.1727, + "step": 2012 + }, + { + "epoch": 0.7697896749521989, + "grad_norm": 4.068723635727276, + "learning_rate": 1e-06, + "loss": 0.2228, + "step": 2013 + }, + { + "epoch": 0.7701720841300191, + "grad_norm": 5.441157021055651, + "learning_rate": 1e-06, + "loss": 0.8692, + "step": 2014 + }, + { + "epoch": 0.7705544933078394, + "grad_norm": 3.608137438759441, + "learning_rate": 1e-06, + "loss": 0.4858, + "step": 2015 + }, + { + "epoch": 0.7709369024856596, + "grad_norm": 1.690505283368894, + "learning_rate": 1e-06, + "loss": 0.4245, + "step": 2016 + }, + { + "epoch": 0.7713193116634799, + "grad_norm": 2.8006470034892073, + "learning_rate": 1e-06, + "loss": 0.3648, + "step": 2017 + }, + { + "epoch": 0.7717017208413002, + "grad_norm": 3.9321450044005655, + "learning_rate": 1e-06, + "loss": 0.4713, + "step": 2018 + }, + { + "epoch": 0.7720841300191205, + "grad_norm": 3.3301640384811915, + "learning_rate": 1e-06, + "loss": 0.2594, + "step": 2019 + }, + { + "epoch": 0.7724665391969407, + "grad_norm": 2.8530852967891436, + "learning_rate": 1e-06, + "loss": 0.4297, + "step": 2020 + }, + { + "epoch": 0.772848948374761, + "grad_norm": 3.9360004778915982, + "learning_rate": 1e-06, + "loss": 0.3659, + "step": 2021 + }, + { + "epoch": 0.7732313575525812, + "grad_norm": 3.245990333791879, + "learning_rate": 1e-06, + "loss": 0.2666, + "step": 2022 + }, + { + "epoch": 0.7736137667304015, + "grad_norm": 3.5073750314453194, + "learning_rate": 1e-06, + "loss": 0.2183, + "step": 2023 + }, + { + "epoch": 0.7739961759082218, + "grad_norm": 2.669386390727826, + "learning_rate": 1e-06, + "loss": 0.1312, + "step": 2024 + }, + { + "epoch": 0.7743785850860421, + "grad_norm": 4.800203144225261, + "learning_rate": 1e-06, + "loss": 0.2841, + "step": 2025 + }, + { + "epoch": 0.7747609942638624, + "grad_norm": 5.814659783865227, + "learning_rate": 1e-06, + "loss": 0.3287, + "step": 2026 + }, + { + "epoch": 0.7751434034416826, + "grad_norm": 3.2658903023307304, + "learning_rate": 1e-06, + "loss": 0.4673, + "step": 2027 + }, + { + "epoch": 0.7755258126195028, + "grad_norm": 3.1296906934503803, + "learning_rate": 1e-06, + "loss": 0.5212, + "step": 2028 + }, + { + "epoch": 0.7759082217973231, + "grad_norm": 2.535589195865614, + "learning_rate": 1e-06, + "loss": 0.4094, + "step": 2029 + }, + { + "epoch": 0.7762906309751434, + "grad_norm": 2.3952707265855557, + "learning_rate": 1e-06, + "loss": 0.2946, + "step": 2030 + }, + { + "epoch": 0.7766730401529637, + "grad_norm": 3.7295965163628253, + "learning_rate": 1e-06, + "loss": 0.3445, + "step": 2031 + }, + { + "epoch": 0.777055449330784, + "grad_norm": 2.110488597651997, + "learning_rate": 1e-06, + "loss": 0.2046, + "step": 2032 + }, + { + "epoch": 0.7774378585086043, + "grad_norm": 2.665524078998848, + "learning_rate": 1e-06, + "loss": 0.1469, + "step": 2033 + }, + { + "epoch": 0.7778202676864244, + "grad_norm": 3.196103620642146, + "learning_rate": 1e-06, + "loss": 0.2041, + "step": 2034 + }, + { + "epoch": 0.7782026768642447, + "grad_norm": 5.1756694047467375, + "learning_rate": 1e-06, + "loss": 0.2137, + "step": 2035 + }, + { + "epoch": 0.778585086042065, + "grad_norm": 5.961311538580228, + "learning_rate": 1e-06, + "loss": 0.1228, + "step": 2036 + }, + { + "epoch": 0.7789674952198853, + "grad_norm": 4.2007380518253825, + "learning_rate": 1e-06, + "loss": 0.0702, + "step": 2037 + }, + { + "epoch": 0.7793499043977056, + "grad_norm": 6.277067243637935, + "learning_rate": 1e-06, + "loss": 0.1584, + "step": 2038 + }, + { + "epoch": 0.7797323135755259, + "grad_norm": 3.8274154316873874, + "learning_rate": 1e-06, + "loss": 0.666, + "step": 2039 + }, + { + "epoch": 0.780114722753346, + "grad_norm": 4.8090131934626115, + "learning_rate": 1e-06, + "loss": 0.4777, + "step": 2040 + }, + { + "epoch": 0.7804971319311663, + "grad_norm": 2.7526655283536225, + "learning_rate": 1e-06, + "loss": 0.4588, + "step": 2041 + }, + { + "epoch": 0.7808795411089866, + "grad_norm": 3.68605530450721, + "learning_rate": 1e-06, + "loss": 0.4125, + "step": 2042 + }, + { + "epoch": 0.7812619502868069, + "grad_norm": 2.9814896141381855, + "learning_rate": 1e-06, + "loss": 0.24, + "step": 2043 + }, + { + "epoch": 0.7816443594646272, + "grad_norm": 2.883457349487923, + "learning_rate": 1e-06, + "loss": 0.2844, + "step": 2044 + }, + { + "epoch": 0.7820267686424475, + "grad_norm": 2.7928683603024536, + "learning_rate": 1e-06, + "loss": 0.2646, + "step": 2045 + }, + { + "epoch": 0.7824091778202676, + "grad_norm": 3.7991790687831086, + "learning_rate": 1e-06, + "loss": 0.4261, + "step": 2046 + }, + { + "epoch": 0.7827915869980879, + "grad_norm": 2.2591240581273455, + "learning_rate": 1e-06, + "loss": 0.1932, + "step": 2047 + }, + { + "epoch": 0.7831739961759082, + "grad_norm": 4.054189780085439, + "learning_rate": 1e-06, + "loss": 0.1903, + "step": 2048 + }, + { + "epoch": 0.7835564053537285, + "grad_norm": 2.4492055110786533, + "learning_rate": 1e-06, + "loss": 0.0825, + "step": 2049 + }, + { + "epoch": 0.7839388145315488, + "grad_norm": 4.351313969142728, + "learning_rate": 1e-06, + "loss": 0.0969, + "step": 2050 + }, + { + "epoch": 0.7843212237093691, + "grad_norm": 5.379664326799858, + "learning_rate": 1e-06, + "loss": 0.6871, + "step": 2051 + }, + { + "epoch": 0.7847036328871893, + "grad_norm": 2.9534803711545825, + "learning_rate": 1e-06, + "loss": 0.4627, + "step": 2052 + }, + { + "epoch": 0.7850860420650095, + "grad_norm": 2.7381981852036636, + "learning_rate": 1e-06, + "loss": 0.3545, + "step": 2053 + }, + { + "epoch": 0.7854684512428298, + "grad_norm": 2.803899202632896, + "learning_rate": 1e-06, + "loss": 0.3812, + "step": 2054 + }, + { + "epoch": 0.7858508604206501, + "grad_norm": 4.112794331044607, + "learning_rate": 1e-06, + "loss": 0.4038, + "step": 2055 + }, + { + "epoch": 0.7862332695984704, + "grad_norm": 2.702379950892697, + "learning_rate": 1e-06, + "loss": 0.1801, + "step": 2056 + }, + { + "epoch": 0.7866156787762907, + "grad_norm": 2.579214154999277, + "learning_rate": 1e-06, + "loss": 0.249, + "step": 2057 + }, + { + "epoch": 0.7869980879541109, + "grad_norm": 2.8267955711350563, + "learning_rate": 1e-06, + "loss": 0.1629, + "step": 2058 + }, + { + "epoch": 0.7873804971319311, + "grad_norm": 3.2930302913974288, + "learning_rate": 1e-06, + "loss": 0.0863, + "step": 2059 + }, + { + "epoch": 0.7877629063097514, + "grad_norm": 3.6787594153296386, + "learning_rate": 1e-06, + "loss": 0.145, + "step": 2060 + }, + { + "epoch": 0.7881453154875717, + "grad_norm": 3.5522703567148186, + "learning_rate": 1e-06, + "loss": 0.1292, + "step": 2061 + }, + { + "epoch": 0.788527724665392, + "grad_norm": 4.591228182164673, + "learning_rate": 1e-06, + "loss": 0.1309, + "step": 2062 + }, + { + "epoch": 0.7889101338432123, + "grad_norm": 3.541918682032985, + "learning_rate": 1e-06, + "loss": 0.2255, + "step": 2063 + }, + { + "epoch": 0.7892925430210325, + "grad_norm": 4.971798519407049, + "learning_rate": 1e-06, + "loss": 0.471, + "step": 2064 + }, + { + "epoch": 0.7896749521988528, + "grad_norm": 4.940052193740455, + "learning_rate": 1e-06, + "loss": 0.4142, + "step": 2065 + }, + { + "epoch": 0.790057361376673, + "grad_norm": 4.393978914942134, + "learning_rate": 1e-06, + "loss": 0.3067, + "step": 2066 + }, + { + "epoch": 0.7904397705544933, + "grad_norm": 4.102075860730455, + "learning_rate": 1e-06, + "loss": 0.3246, + "step": 2067 + }, + { + "epoch": 0.7908221797323136, + "grad_norm": 2.186783046397264, + "learning_rate": 1e-06, + "loss": 0.3369, + "step": 2068 + }, + { + "epoch": 0.7912045889101338, + "grad_norm": 2.822973459571022, + "learning_rate": 1e-06, + "loss": 0.2647, + "step": 2069 + }, + { + "epoch": 0.7915869980879541, + "grad_norm": 3.3141656682396077, + "learning_rate": 1e-06, + "loss": 0.3238, + "step": 2070 + }, + { + "epoch": 0.7919694072657744, + "grad_norm": 2.6746682184119144, + "learning_rate": 1e-06, + "loss": 0.194, + "step": 2071 + }, + { + "epoch": 0.7923518164435946, + "grad_norm": 8.472692268826158, + "learning_rate": 1e-06, + "loss": 0.3371, + "step": 2072 + }, + { + "epoch": 0.7927342256214149, + "grad_norm": 3.716701039267943, + "learning_rate": 1e-06, + "loss": 0.1934, + "step": 2073 + }, + { + "epoch": 0.7931166347992352, + "grad_norm": 4.896144689304383, + "learning_rate": 1e-06, + "loss": 0.1346, + "step": 2074 + }, + { + "epoch": 0.7934990439770554, + "grad_norm": 4.428809884678829, + "learning_rate": 1e-06, + "loss": 0.2384, + "step": 2075 + }, + { + "epoch": 0.7938814531548757, + "grad_norm": 6.608011307953772, + "learning_rate": 1e-06, + "loss": 0.4356, + "step": 2076 + }, + { + "epoch": 0.794263862332696, + "grad_norm": 3.8141769723306393, + "learning_rate": 1e-06, + "loss": 0.4045, + "step": 2077 + }, + { + "epoch": 0.7946462715105163, + "grad_norm": 2.7864923683060625, + "learning_rate": 1e-06, + "loss": 0.323, + "step": 2078 + }, + { + "epoch": 0.7950286806883365, + "grad_norm": 2.414149705614162, + "learning_rate": 1e-06, + "loss": 0.379, + "step": 2079 + }, + { + "epoch": 0.7954110898661568, + "grad_norm": 3.170569841930405, + "learning_rate": 1e-06, + "loss": 0.2751, + "step": 2080 + }, + { + "epoch": 0.795793499043977, + "grad_norm": 3.8556209947698514, + "learning_rate": 1e-06, + "loss": 0.4716, + "step": 2081 + }, + { + "epoch": 0.7961759082217973, + "grad_norm": 2.704039490642762, + "learning_rate": 1e-06, + "loss": 0.4642, + "step": 2082 + }, + { + "epoch": 0.7965583173996176, + "grad_norm": 2.478727339171121, + "learning_rate": 1e-06, + "loss": 0.2617, + "step": 2083 + }, + { + "epoch": 0.7969407265774379, + "grad_norm": 3.615397923096984, + "learning_rate": 1e-06, + "loss": 0.2749, + "step": 2084 + }, + { + "epoch": 0.7973231357552581, + "grad_norm": 3.7581430396873516, + "learning_rate": 1e-06, + "loss": 0.2536, + "step": 2085 + }, + { + "epoch": 0.7977055449330784, + "grad_norm": 3.1040276788578613, + "learning_rate": 1e-06, + "loss": 0.1833, + "step": 2086 + }, + { + "epoch": 0.7980879541108986, + "grad_norm": 3.7896379486113414, + "learning_rate": 1e-06, + "loss": 0.0956, + "step": 2087 + }, + { + "epoch": 0.7984703632887189, + "grad_norm": 6.420381172724574, + "learning_rate": 1e-06, + "loss": 0.1931, + "step": 2088 + }, + { + "epoch": 0.7988527724665392, + "grad_norm": 5.77144670038465, + "learning_rate": 1e-06, + "loss": 0.6335, + "step": 2089 + }, + { + "epoch": 0.7992351816443595, + "grad_norm": 4.792139157205931, + "learning_rate": 1e-06, + "loss": 0.3828, + "step": 2090 + }, + { + "epoch": 0.7996175908221798, + "grad_norm": 2.162194061688356, + "learning_rate": 1e-06, + "loss": 0.3287, + "step": 2091 + }, + { + "epoch": 0.8, + "grad_norm": 2.8522792803652104, + "learning_rate": 1e-06, + "loss": 0.3452, + "step": 2092 + }, + { + "epoch": 0.8003824091778202, + "grad_norm": 3.708971775831023, + "learning_rate": 1e-06, + "loss": 0.3769, + "step": 2093 + }, + { + "epoch": 0.8007648183556405, + "grad_norm": 3.4264057421019594, + "learning_rate": 1e-06, + "loss": 0.5301, + "step": 2094 + }, + { + "epoch": 0.8011472275334608, + "grad_norm": 2.1893933413099647, + "learning_rate": 1e-06, + "loss": 0.1319, + "step": 2095 + }, + { + "epoch": 0.8015296367112811, + "grad_norm": 1.9048559375667495, + "learning_rate": 1e-06, + "loss": 0.1312, + "step": 2096 + }, + { + "epoch": 0.8019120458891014, + "grad_norm": 3.8689251935544466, + "learning_rate": 1e-06, + "loss": 0.3397, + "step": 2097 + }, + { + "epoch": 0.8022944550669217, + "grad_norm": 5.43743352739766, + "learning_rate": 1e-06, + "loss": 0.3189, + "step": 2098 + }, + { + "epoch": 0.8026768642447418, + "grad_norm": 3.5664840639614823, + "learning_rate": 1e-06, + "loss": 0.1821, + "step": 2099 + }, + { + "epoch": 0.8030592734225621, + "grad_norm": 4.467345116934534, + "learning_rate": 1e-06, + "loss": 0.1286, + "step": 2100 + }, + { + "epoch": 0.8034416826003824, + "grad_norm": 5.400395393024568, + "learning_rate": 1e-06, + "loss": 0.4901, + "step": 2101 + }, + { + "epoch": 0.8038240917782027, + "grad_norm": 2.9126148053901044, + "learning_rate": 1e-06, + "loss": 0.3685, + "step": 2102 + }, + { + "epoch": 0.804206500956023, + "grad_norm": 3.0631492082331566, + "learning_rate": 1e-06, + "loss": 0.371, + "step": 2103 + }, + { + "epoch": 0.8045889101338433, + "grad_norm": 3.255761248615045, + "learning_rate": 1e-06, + "loss": 0.4381, + "step": 2104 + }, + { + "epoch": 0.8049713193116634, + "grad_norm": 2.3928423801771963, + "learning_rate": 1e-06, + "loss": 0.2927, + "step": 2105 + }, + { + "epoch": 0.8053537284894837, + "grad_norm": 5.051747523942017, + "learning_rate": 1e-06, + "loss": 0.5989, + "step": 2106 + }, + { + "epoch": 0.805736137667304, + "grad_norm": 3.504058664466598, + "learning_rate": 1e-06, + "loss": 0.198, + "step": 2107 + }, + { + "epoch": 0.8061185468451243, + "grad_norm": 3.1708870086604937, + "learning_rate": 1e-06, + "loss": 0.1881, + "step": 2108 + }, + { + "epoch": 0.8065009560229446, + "grad_norm": 2.917293118003835, + "learning_rate": 1e-06, + "loss": 0.2526, + "step": 2109 + }, + { + "epoch": 0.8068833652007649, + "grad_norm": 3.846004166258144, + "learning_rate": 1e-06, + "loss": 0.2796, + "step": 2110 + }, + { + "epoch": 0.807265774378585, + "grad_norm": 3.0444224341340176, + "learning_rate": 1e-06, + "loss": 0.1437, + "step": 2111 + }, + { + "epoch": 0.8076481835564053, + "grad_norm": 3.5297505393779964, + "learning_rate": 1e-06, + "loss": 0.1154, + "step": 2112 + }, + { + "epoch": 0.8080305927342256, + "grad_norm": 4.927851847527151, + "learning_rate": 1e-06, + "loss": 0.1744, + "step": 2113 + }, + { + "epoch": 0.8084130019120459, + "grad_norm": 4.281335314715068, + "learning_rate": 1e-06, + "loss": 0.3368, + "step": 2114 + }, + { + "epoch": 0.8087954110898662, + "grad_norm": 5.608802702460388, + "learning_rate": 1e-06, + "loss": 0.5034, + "step": 2115 + }, + { + "epoch": 0.8091778202676865, + "grad_norm": 2.787287898183835, + "learning_rate": 1e-06, + "loss": 0.406, + "step": 2116 + }, + { + "epoch": 0.8095602294455067, + "grad_norm": 1.6220800134229783, + "learning_rate": 1e-06, + "loss": 0.2047, + "step": 2117 + }, + { + "epoch": 0.8099426386233269, + "grad_norm": 2.822382370801178, + "learning_rate": 1e-06, + "loss": 0.2202, + "step": 2118 + }, + { + "epoch": 0.8103250478011472, + "grad_norm": 4.004621696750148, + "learning_rate": 1e-06, + "loss": 0.3022, + "step": 2119 + }, + { + "epoch": 0.8107074569789675, + "grad_norm": 4.1272121624366065, + "learning_rate": 1e-06, + "loss": 0.3423, + "step": 2120 + }, + { + "epoch": 0.8110898661567878, + "grad_norm": 2.2668283358120718, + "learning_rate": 1e-06, + "loss": 0.1929, + "step": 2121 + }, + { + "epoch": 0.8114722753346081, + "grad_norm": 4.1438659651742, + "learning_rate": 1e-06, + "loss": 0.2024, + "step": 2122 + }, + { + "epoch": 0.8118546845124283, + "grad_norm": 3.14419428755532, + "learning_rate": 1e-06, + "loss": 0.1916, + "step": 2123 + }, + { + "epoch": 0.8122370936902485, + "grad_norm": 3.3102456825100854, + "learning_rate": 1e-06, + "loss": 0.1257, + "step": 2124 + }, + { + "epoch": 0.8126195028680688, + "grad_norm": 3.8880439990972344, + "learning_rate": 1e-06, + "loss": 0.1195, + "step": 2125 + }, + { + "epoch": 0.8130019120458891, + "grad_norm": 7.244368635195714, + "learning_rate": 1e-06, + "loss": 0.6411, + "step": 2126 + }, + { + "epoch": 0.8133843212237094, + "grad_norm": 3.16347289066628, + "learning_rate": 1e-06, + "loss": 0.3795, + "step": 2127 + }, + { + "epoch": 0.8137667304015297, + "grad_norm": 3.2375683339098558, + "learning_rate": 1e-06, + "loss": 0.3666, + "step": 2128 + }, + { + "epoch": 0.8141491395793499, + "grad_norm": 2.5387238731523696, + "learning_rate": 1e-06, + "loss": 0.428, + "step": 2129 + }, + { + "epoch": 0.8145315487571702, + "grad_norm": 2.6485824320249045, + "learning_rate": 1e-06, + "loss": 0.3104, + "step": 2130 + }, + { + "epoch": 0.8149139579349904, + "grad_norm": 3.061018916513091, + "learning_rate": 1e-06, + "loss": 0.3034, + "step": 2131 + }, + { + "epoch": 0.8152963671128107, + "grad_norm": 2.5765817358109753, + "learning_rate": 1e-06, + "loss": 0.335, + "step": 2132 + }, + { + "epoch": 0.815678776290631, + "grad_norm": 3.6465646646054988, + "learning_rate": 1e-06, + "loss": 0.2214, + "step": 2133 + }, + { + "epoch": 0.8160611854684512, + "grad_norm": 3.659063446829306, + "learning_rate": 1e-06, + "loss": 0.2679, + "step": 2134 + }, + { + "epoch": 0.8164435946462715, + "grad_norm": 4.176751076793435, + "learning_rate": 1e-06, + "loss": 0.2274, + "step": 2135 + }, + { + "epoch": 0.8168260038240918, + "grad_norm": 2.9000141143455354, + "learning_rate": 1e-06, + "loss": 0.1416, + "step": 2136 + }, + { + "epoch": 0.817208413001912, + "grad_norm": 3.5229236033480564, + "learning_rate": 1e-06, + "loss": 0.1278, + "step": 2137 + }, + { + "epoch": 0.8175908221797323, + "grad_norm": 7.9440475252080205, + "learning_rate": 1e-06, + "loss": 0.242, + "step": 2138 + }, + { + "epoch": 0.8179732313575526, + "grad_norm": 5.1135330313058365, + "learning_rate": 1e-06, + "loss": 0.6503, + "step": 2139 + }, + { + "epoch": 0.8183556405353728, + "grad_norm": 4.578243527489235, + "learning_rate": 1e-06, + "loss": 0.5686, + "step": 2140 + }, + { + "epoch": 0.8187380497131931, + "grad_norm": 4.147331133365789, + "learning_rate": 1e-06, + "loss": 0.5776, + "step": 2141 + }, + { + "epoch": 0.8191204588910134, + "grad_norm": 1.9505210913813733, + "learning_rate": 1e-06, + "loss": 0.4936, + "step": 2142 + }, + { + "epoch": 0.8195028680688337, + "grad_norm": 3.321460911231512, + "learning_rate": 1e-06, + "loss": 0.2061, + "step": 2143 + }, + { + "epoch": 0.8198852772466539, + "grad_norm": 3.6699691696327217, + "learning_rate": 1e-06, + "loss": 0.2966, + "step": 2144 + }, + { + "epoch": 0.8202676864244742, + "grad_norm": 3.001923103653967, + "learning_rate": 1e-06, + "loss": 0.2617, + "step": 2145 + }, + { + "epoch": 0.8206500956022944, + "grad_norm": 2.4060394393149123, + "learning_rate": 1e-06, + "loss": 0.2851, + "step": 2146 + }, + { + "epoch": 0.8210325047801147, + "grad_norm": 3.218550592560278, + "learning_rate": 1e-06, + "loss": 0.2154, + "step": 2147 + }, + { + "epoch": 0.821414913957935, + "grad_norm": 4.372564237197992, + "learning_rate": 1e-06, + "loss": 0.2861, + "step": 2148 + }, + { + "epoch": 0.8217973231357553, + "grad_norm": 3.943080882818625, + "learning_rate": 1e-06, + "loss": 0.2615, + "step": 2149 + }, + { + "epoch": 0.8221797323135756, + "grad_norm": 4.047737412632037, + "learning_rate": 1e-06, + "loss": 0.2142, + "step": 2150 + }, + { + "epoch": 0.8225621414913958, + "grad_norm": 7.329343259359967, + "learning_rate": 1e-06, + "loss": 0.6134, + "step": 2151 + }, + { + "epoch": 0.822944550669216, + "grad_norm": 4.629531908204858, + "learning_rate": 1e-06, + "loss": 0.6396, + "step": 2152 + }, + { + "epoch": 0.8233269598470363, + "grad_norm": 4.350650054767431, + "learning_rate": 1e-06, + "loss": 0.4429, + "step": 2153 + }, + { + "epoch": 0.8237093690248566, + "grad_norm": 3.1065780865506847, + "learning_rate": 1e-06, + "loss": 0.248, + "step": 2154 + }, + { + "epoch": 0.8240917782026769, + "grad_norm": 3.943895746551621, + "learning_rate": 1e-06, + "loss": 0.3949, + "step": 2155 + }, + { + "epoch": 0.8244741873804972, + "grad_norm": 3.179876148523129, + "learning_rate": 1e-06, + "loss": 0.5421, + "step": 2156 + }, + { + "epoch": 0.8248565965583174, + "grad_norm": 3.589649946524929, + "learning_rate": 1e-06, + "loss": 0.2262, + "step": 2157 + }, + { + "epoch": 0.8252390057361376, + "grad_norm": 3.8537420640357034, + "learning_rate": 1e-06, + "loss": 0.3552, + "step": 2158 + }, + { + "epoch": 0.8256214149139579, + "grad_norm": 2.8540038644808594, + "learning_rate": 1e-06, + "loss": 0.2076, + "step": 2159 + }, + { + "epoch": 0.8260038240917782, + "grad_norm": 3.918334578742821, + "learning_rate": 1e-06, + "loss": 0.3502, + "step": 2160 + }, + { + "epoch": 0.8263862332695985, + "grad_norm": 4.105354459933907, + "learning_rate": 1e-06, + "loss": 0.1463, + "step": 2161 + }, + { + "epoch": 0.8267686424474188, + "grad_norm": 4.794980547948051, + "learning_rate": 1e-06, + "loss": 0.1324, + "step": 2162 + }, + { + "epoch": 0.827151051625239, + "grad_norm": 5.959562410509337, + "learning_rate": 1e-06, + "loss": 0.2033, + "step": 2163 + }, + { + "epoch": 0.8275334608030592, + "grad_norm": 4.1375193339008485, + "learning_rate": 1e-06, + "loss": 0.5293, + "step": 2164 + }, + { + "epoch": 0.8279158699808795, + "grad_norm": 3.7650827357711742, + "learning_rate": 1e-06, + "loss": 0.3524, + "step": 2165 + }, + { + "epoch": 0.8282982791586998, + "grad_norm": 2.258227247535923, + "learning_rate": 1e-06, + "loss": 0.1928, + "step": 2166 + }, + { + "epoch": 0.8286806883365201, + "grad_norm": 2.73280594630298, + "learning_rate": 1e-06, + "loss": 0.3709, + "step": 2167 + }, + { + "epoch": 0.8290630975143404, + "grad_norm": 3.8576784922553324, + "learning_rate": 1e-06, + "loss": 0.4781, + "step": 2168 + }, + { + "epoch": 0.8294455066921607, + "grad_norm": 2.629339355343209, + "learning_rate": 1e-06, + "loss": 0.384, + "step": 2169 + }, + { + "epoch": 0.8298279158699808, + "grad_norm": 3.0564081755663697, + "learning_rate": 1e-06, + "loss": 0.3269, + "step": 2170 + }, + { + "epoch": 0.8302103250478011, + "grad_norm": 2.679405798159991, + "learning_rate": 1e-06, + "loss": 0.2193, + "step": 2171 + }, + { + "epoch": 0.8305927342256214, + "grad_norm": 3.185608096677568, + "learning_rate": 1e-06, + "loss": 0.2305, + "step": 2172 + }, + { + "epoch": 0.8309751434034417, + "grad_norm": 4.0877777098595525, + "learning_rate": 1e-06, + "loss": 0.2141, + "step": 2173 + }, + { + "epoch": 0.831357552581262, + "grad_norm": 3.0768842181172853, + "learning_rate": 1e-06, + "loss": 0.1719, + "step": 2174 + }, + { + "epoch": 0.8317399617590823, + "grad_norm": 3.5653037030946315, + "learning_rate": 1e-06, + "loss": 0.1642, + "step": 2175 + }, + { + "epoch": 0.8321223709369024, + "grad_norm": 3.9048649888382423, + "learning_rate": 1e-06, + "loss": 0.3858, + "step": 2176 + }, + { + "epoch": 0.8325047801147227, + "grad_norm": 3.1362740663446607, + "learning_rate": 1e-06, + "loss": 0.5031, + "step": 2177 + }, + { + "epoch": 0.832887189292543, + "grad_norm": 3.297428898135157, + "learning_rate": 1e-06, + "loss": 0.6066, + "step": 2178 + }, + { + "epoch": 0.8332695984703633, + "grad_norm": 2.465040776377811, + "learning_rate": 1e-06, + "loss": 0.3532, + "step": 2179 + }, + { + "epoch": 0.8336520076481836, + "grad_norm": 2.667095348551791, + "learning_rate": 1e-06, + "loss": 0.3922, + "step": 2180 + }, + { + "epoch": 0.8340344168260039, + "grad_norm": 2.9862716000822846, + "learning_rate": 1e-06, + "loss": 0.2624, + "step": 2181 + }, + { + "epoch": 0.834416826003824, + "grad_norm": 2.5867503752715075, + "learning_rate": 1e-06, + "loss": 0.2768, + "step": 2182 + }, + { + "epoch": 0.8347992351816443, + "grad_norm": 3.99836637993455, + "learning_rate": 1e-06, + "loss": 0.3965, + "step": 2183 + }, + { + "epoch": 0.8351816443594646, + "grad_norm": 3.9710708200234355, + "learning_rate": 1e-06, + "loss": 0.2693, + "step": 2184 + }, + { + "epoch": 0.8355640535372849, + "grad_norm": 3.4491208711692822, + "learning_rate": 1e-06, + "loss": 0.2509, + "step": 2185 + }, + { + "epoch": 0.8359464627151052, + "grad_norm": 2.6952109497400447, + "learning_rate": 1e-06, + "loss": 0.1052, + "step": 2186 + }, + { + "epoch": 0.8363288718929255, + "grad_norm": 3.271212237981573, + "learning_rate": 1e-06, + "loss": 0.1229, + "step": 2187 + }, + { + "epoch": 0.8367112810707457, + "grad_norm": 5.315335762277588, + "learning_rate": 1e-06, + "loss": 0.3858, + "step": 2188 + }, + { + "epoch": 0.837093690248566, + "grad_norm": 4.039271690514878, + "learning_rate": 1e-06, + "loss": 0.3017, + "step": 2189 + }, + { + "epoch": 0.8374760994263862, + "grad_norm": 4.276716742000643, + "learning_rate": 1e-06, + "loss": 0.4611, + "step": 2190 + }, + { + "epoch": 0.8378585086042065, + "grad_norm": 2.3208846597343618, + "learning_rate": 1e-06, + "loss": 0.2205, + "step": 2191 + }, + { + "epoch": 0.8382409177820268, + "grad_norm": 2.910222118067495, + "learning_rate": 1e-06, + "loss": 0.3472, + "step": 2192 + }, + { + "epoch": 0.838623326959847, + "grad_norm": 3.7821854624348363, + "learning_rate": 1e-06, + "loss": 0.4182, + "step": 2193 + }, + { + "epoch": 0.8390057361376673, + "grad_norm": 3.645645412869907, + "learning_rate": 1e-06, + "loss": 0.4144, + "step": 2194 + }, + { + "epoch": 0.8393881453154876, + "grad_norm": 5.041450159557636, + "learning_rate": 1e-06, + "loss": 0.279, + "step": 2195 + }, + { + "epoch": 0.8397705544933078, + "grad_norm": 2.872907457528259, + "learning_rate": 1e-06, + "loss": 0.197, + "step": 2196 + }, + { + "epoch": 0.8401529636711281, + "grad_norm": 2.217400046141906, + "learning_rate": 1e-06, + "loss": 0.1678, + "step": 2197 + }, + { + "epoch": 0.8405353728489484, + "grad_norm": 5.82892511106803, + "learning_rate": 1e-06, + "loss": 0.256, + "step": 2198 + }, + { + "epoch": 0.8409177820267686, + "grad_norm": 7.5717419356407945, + "learning_rate": 1e-06, + "loss": 0.1019, + "step": 2199 + }, + { + "epoch": 0.8413001912045889, + "grad_norm": 3.061982091590941, + "learning_rate": 1e-06, + "loss": 0.0776, + "step": 2200 + }, + { + "epoch": 0.8416826003824092, + "grad_norm": 4.486253723670178, + "learning_rate": 1e-06, + "loss": 0.3442, + "step": 2201 + }, + { + "epoch": 0.8420650095602294, + "grad_norm": 3.257880267346353, + "learning_rate": 1e-06, + "loss": 0.4429, + "step": 2202 + }, + { + "epoch": 0.8424474187380497, + "grad_norm": 2.2060811656293344, + "learning_rate": 1e-06, + "loss": 0.3688, + "step": 2203 + }, + { + "epoch": 0.84282982791587, + "grad_norm": 2.597016265735945, + "learning_rate": 1e-06, + "loss": 0.2879, + "step": 2204 + }, + { + "epoch": 0.8432122370936902, + "grad_norm": 2.395171983632231, + "learning_rate": 1e-06, + "loss": 0.2489, + "step": 2205 + }, + { + "epoch": 0.8435946462715105, + "grad_norm": 3.1052566935755257, + "learning_rate": 1e-06, + "loss": 0.3973, + "step": 2206 + }, + { + "epoch": 0.8439770554493308, + "grad_norm": 3.2083941036925894, + "learning_rate": 1e-06, + "loss": 0.265, + "step": 2207 + }, + { + "epoch": 0.8443594646271511, + "grad_norm": 3.424640809318776, + "learning_rate": 1e-06, + "loss": 0.1639, + "step": 2208 + }, + { + "epoch": 0.8447418738049713, + "grad_norm": 4.030188132312924, + "learning_rate": 1e-06, + "loss": 0.2207, + "step": 2209 + }, + { + "epoch": 0.8451242829827916, + "grad_norm": 2.6657795225124086, + "learning_rate": 1e-06, + "loss": 0.101, + "step": 2210 + }, + { + "epoch": 0.8455066921606118, + "grad_norm": 3.655967766717767, + "learning_rate": 1e-06, + "loss": 0.1517, + "step": 2211 + }, + { + "epoch": 0.8458891013384321, + "grad_norm": 4.780504156044833, + "learning_rate": 1e-06, + "loss": 0.2003, + "step": 2212 + }, + { + "epoch": 0.8462715105162524, + "grad_norm": 7.961990661140336, + "learning_rate": 1e-06, + "loss": 0.1447, + "step": 2213 + }, + { + "epoch": 0.8466539196940727, + "grad_norm": 6.391086104716505, + "learning_rate": 1e-06, + "loss": 0.4777, + "step": 2214 + }, + { + "epoch": 0.847036328871893, + "grad_norm": 3.4270184326210464, + "learning_rate": 1e-06, + "loss": 0.3738, + "step": 2215 + }, + { + "epoch": 0.8474187380497132, + "grad_norm": 2.598951524459288, + "learning_rate": 1e-06, + "loss": 0.3055, + "step": 2216 + }, + { + "epoch": 0.8478011472275334, + "grad_norm": 2.271296276068281, + "learning_rate": 1e-06, + "loss": 0.3695, + "step": 2217 + }, + { + "epoch": 0.8481835564053537, + "grad_norm": 4.049813753719958, + "learning_rate": 1e-06, + "loss": 0.5607, + "step": 2218 + }, + { + "epoch": 0.848565965583174, + "grad_norm": 2.4192822760293224, + "learning_rate": 1e-06, + "loss": 0.2376, + "step": 2219 + }, + { + "epoch": 0.8489483747609943, + "grad_norm": 2.7489896131911116, + "learning_rate": 1e-06, + "loss": 0.4236, + "step": 2220 + }, + { + "epoch": 0.8493307839388146, + "grad_norm": 2.5501943850758444, + "learning_rate": 1e-06, + "loss": 0.2919, + "step": 2221 + }, + { + "epoch": 0.8497131931166348, + "grad_norm": 2.2434257251745597, + "learning_rate": 1e-06, + "loss": 0.1412, + "step": 2222 + }, + { + "epoch": 0.850095602294455, + "grad_norm": 4.330815439594046, + "learning_rate": 1e-06, + "loss": 0.2933, + "step": 2223 + }, + { + "epoch": 0.8504780114722753, + "grad_norm": 5.112670022613608, + "learning_rate": 1e-06, + "loss": 0.1171, + "step": 2224 + }, + { + "epoch": 0.8508604206500956, + "grad_norm": 3.269080419233297, + "learning_rate": 1e-06, + "loss": 0.1149, + "step": 2225 + }, + { + "epoch": 0.8512428298279159, + "grad_norm": 4.993048604088785, + "learning_rate": 1e-06, + "loss": 0.5126, + "step": 2226 + }, + { + "epoch": 0.8516252390057362, + "grad_norm": 3.5827526834504617, + "learning_rate": 1e-06, + "loss": 0.6183, + "step": 2227 + }, + { + "epoch": 0.8520076481835565, + "grad_norm": 4.147780659143189, + "learning_rate": 1e-06, + "loss": 0.3283, + "step": 2228 + }, + { + "epoch": 0.8523900573613766, + "grad_norm": 2.6242208686949526, + "learning_rate": 1e-06, + "loss": 0.4896, + "step": 2229 + }, + { + "epoch": 0.8527724665391969, + "grad_norm": 2.682157306947873, + "learning_rate": 1e-06, + "loss": 0.3766, + "step": 2230 + }, + { + "epoch": 0.8531548757170172, + "grad_norm": 3.256507374262415, + "learning_rate": 1e-06, + "loss": 0.3188, + "step": 2231 + }, + { + "epoch": 0.8535372848948375, + "grad_norm": 3.0808107693662503, + "learning_rate": 1e-06, + "loss": 0.2812, + "step": 2232 + }, + { + "epoch": 0.8539196940726578, + "grad_norm": 2.9379973294925694, + "learning_rate": 1e-06, + "loss": 0.3041, + "step": 2233 + }, + { + "epoch": 0.8543021032504781, + "grad_norm": 2.3469328567139156, + "learning_rate": 1e-06, + "loss": 0.1998, + "step": 2234 + }, + { + "epoch": 0.8546845124282982, + "grad_norm": 3.5654014684500237, + "learning_rate": 1e-06, + "loss": 0.1945, + "step": 2235 + }, + { + "epoch": 0.8550669216061185, + "grad_norm": 2.794885620479178, + "learning_rate": 1e-06, + "loss": 0.1304, + "step": 2236 + }, + { + "epoch": 0.8554493307839388, + "grad_norm": 3.892644151432938, + "learning_rate": 1e-06, + "loss": 0.1376, + "step": 2237 + }, + { + "epoch": 0.8558317399617591, + "grad_norm": 4.52290069610058, + "learning_rate": 1e-06, + "loss": 0.322, + "step": 2238 + }, + { + "epoch": 0.8562141491395794, + "grad_norm": 4.4033401295919665, + "learning_rate": 1e-06, + "loss": 0.4721, + "step": 2239 + }, + { + "epoch": 0.8565965583173997, + "grad_norm": 2.8414582460400957, + "learning_rate": 1e-06, + "loss": 0.2592, + "step": 2240 + }, + { + "epoch": 0.8569789674952198, + "grad_norm": 3.4026742067383045, + "learning_rate": 1e-06, + "loss": 0.283, + "step": 2241 + }, + { + "epoch": 0.8573613766730401, + "grad_norm": 2.563220783370359, + "learning_rate": 1e-06, + "loss": 0.2566, + "step": 2242 + }, + { + "epoch": 0.8577437858508604, + "grad_norm": 1.9607244733019165, + "learning_rate": 1e-06, + "loss": 0.4699, + "step": 2243 + }, + { + "epoch": 0.8581261950286807, + "grad_norm": 3.702986573798893, + "learning_rate": 1e-06, + "loss": 0.2381, + "step": 2244 + }, + { + "epoch": 0.858508604206501, + "grad_norm": 2.3633086904398657, + "learning_rate": 1e-06, + "loss": 0.3277, + "step": 2245 + }, + { + "epoch": 0.8588910133843213, + "grad_norm": 2.719397895171275, + "learning_rate": 1e-06, + "loss": 0.1863, + "step": 2246 + }, + { + "epoch": 0.8592734225621415, + "grad_norm": 4.029146577192611, + "learning_rate": 1e-06, + "loss": 0.2504, + "step": 2247 + }, + { + "epoch": 0.8596558317399617, + "grad_norm": 2.9593021436220557, + "learning_rate": 1e-06, + "loss": 0.2487, + "step": 2248 + }, + { + "epoch": 0.860038240917782, + "grad_norm": 4.2307703645077765, + "learning_rate": 1e-06, + "loss": 0.2118, + "step": 2249 + }, + { + "epoch": 0.8604206500956023, + "grad_norm": 5.035898276758321, + "learning_rate": 1e-06, + "loss": 0.2637, + "step": 2250 + }, + { + "epoch": 0.8608030592734226, + "grad_norm": 8.803036200658138, + "learning_rate": 1e-06, + "loss": 0.504, + "step": 2251 + }, + { + "epoch": 0.8611854684512428, + "grad_norm": 4.752974131219422, + "learning_rate": 1e-06, + "loss": 0.6026, + "step": 2252 + }, + { + "epoch": 0.8615678776290631, + "grad_norm": 3.3525969502812525, + "learning_rate": 1e-06, + "loss": 0.4438, + "step": 2253 + }, + { + "epoch": 0.8619502868068833, + "grad_norm": 3.616882179507573, + "learning_rate": 1e-06, + "loss": 0.2969, + "step": 2254 + }, + { + "epoch": 0.8623326959847036, + "grad_norm": 2.2632922561056605, + "learning_rate": 1e-06, + "loss": 0.4263, + "step": 2255 + }, + { + "epoch": 0.8627151051625239, + "grad_norm": 2.3040020456768597, + "learning_rate": 1e-06, + "loss": 0.1578, + "step": 2256 + }, + { + "epoch": 0.8630975143403442, + "grad_norm": 2.0374035865111786, + "learning_rate": 1e-06, + "loss": 0.2902, + "step": 2257 + }, + { + "epoch": 0.8634799235181644, + "grad_norm": 3.4553879447456906, + "learning_rate": 1e-06, + "loss": 0.1156, + "step": 2258 + }, + { + "epoch": 0.8638623326959847, + "grad_norm": 3.5985599339635344, + "learning_rate": 1e-06, + "loss": 0.2264, + "step": 2259 + }, + { + "epoch": 0.864244741873805, + "grad_norm": 3.138347198272023, + "learning_rate": 1e-06, + "loss": 0.1511, + "step": 2260 + }, + { + "epoch": 0.8646271510516252, + "grad_norm": 3.2196779857762614, + "learning_rate": 1e-06, + "loss": 0.1603, + "step": 2261 + }, + { + "epoch": 0.8650095602294455, + "grad_norm": 3.5260168137606565, + "learning_rate": 1e-06, + "loss": 0.1481, + "step": 2262 + }, + { + "epoch": 0.8653919694072658, + "grad_norm": 8.299671800857737, + "learning_rate": 1e-06, + "loss": 0.2358, + "step": 2263 + }, + { + "epoch": 0.865774378585086, + "grad_norm": 4.560272300018649, + "learning_rate": 1e-06, + "loss": 0.3339, + "step": 2264 + }, + { + "epoch": 0.8661567877629063, + "grad_norm": 3.5616443342604454, + "learning_rate": 1e-06, + "loss": 0.4314, + "step": 2265 + }, + { + "epoch": 0.8665391969407266, + "grad_norm": 2.1089186563301427, + "learning_rate": 1e-06, + "loss": 0.5122, + "step": 2266 + }, + { + "epoch": 0.8669216061185469, + "grad_norm": 2.3026888181986362, + "learning_rate": 1e-06, + "loss": 0.3382, + "step": 2267 + }, + { + "epoch": 0.8673040152963671, + "grad_norm": 3.6717550055193824, + "learning_rate": 1e-06, + "loss": 0.3318, + "step": 2268 + }, + { + "epoch": 0.8676864244741874, + "grad_norm": 4.314919718927125, + "learning_rate": 1e-06, + "loss": 0.3843, + "step": 2269 + }, + { + "epoch": 0.8680688336520076, + "grad_norm": 4.4882347164281615, + "learning_rate": 1e-06, + "loss": 0.3663, + "step": 2270 + }, + { + "epoch": 0.8684512428298279, + "grad_norm": 3.2509591081218883, + "learning_rate": 1e-06, + "loss": 0.3615, + "step": 2271 + }, + { + "epoch": 0.8688336520076482, + "grad_norm": 4.111744708704633, + "learning_rate": 1e-06, + "loss": 0.3047, + "step": 2272 + }, + { + "epoch": 0.8692160611854685, + "grad_norm": 2.742256098484252, + "learning_rate": 1e-06, + "loss": 0.0854, + "step": 2273 + }, + { + "epoch": 0.8695984703632887, + "grad_norm": 2.907756783978073, + "learning_rate": 1e-06, + "loss": 0.1346, + "step": 2274 + }, + { + "epoch": 0.869980879541109, + "grad_norm": 3.8665304839608385, + "learning_rate": 1e-06, + "loss": 0.1086, + "step": 2275 + }, + { + "epoch": 0.8703632887189292, + "grad_norm": 4.447655675349624, + "learning_rate": 1e-06, + "loss": 0.4916, + "step": 2276 + }, + { + "epoch": 0.8707456978967495, + "grad_norm": 3.674655721718259, + "learning_rate": 1e-06, + "loss": 0.5621, + "step": 2277 + }, + { + "epoch": 0.8711281070745698, + "grad_norm": 2.352703844049994, + "learning_rate": 1e-06, + "loss": 0.5073, + "step": 2278 + }, + { + "epoch": 0.8715105162523901, + "grad_norm": 2.4129684586374416, + "learning_rate": 1e-06, + "loss": 0.2702, + "step": 2279 + }, + { + "epoch": 0.8718929254302104, + "grad_norm": 2.4772123332611757, + "learning_rate": 1e-06, + "loss": 0.209, + "step": 2280 + }, + { + "epoch": 0.8722753346080306, + "grad_norm": 2.809172548121751, + "learning_rate": 1e-06, + "loss": 0.298, + "step": 2281 + }, + { + "epoch": 0.8726577437858508, + "grad_norm": 2.24072068780217, + "learning_rate": 1e-06, + "loss": 0.2951, + "step": 2282 + }, + { + "epoch": 0.8730401529636711, + "grad_norm": 2.7390198412117717, + "learning_rate": 1e-06, + "loss": 0.252, + "step": 2283 + }, + { + "epoch": 0.8734225621414914, + "grad_norm": 3.9408881732569134, + "learning_rate": 1e-06, + "loss": 0.4563, + "step": 2284 + }, + { + "epoch": 0.8738049713193117, + "grad_norm": 4.011042849052909, + "learning_rate": 1e-06, + "loss": 0.2125, + "step": 2285 + }, + { + "epoch": 0.874187380497132, + "grad_norm": 2.976018099924646, + "learning_rate": 1e-06, + "loss": 0.1403, + "step": 2286 + }, + { + "epoch": 0.8745697896749522, + "grad_norm": 4.713013047842722, + "learning_rate": 1e-06, + "loss": 0.1347, + "step": 2287 + }, + { + "epoch": 0.8749521988527724, + "grad_norm": 4.383286394416696, + "learning_rate": 1e-06, + "loss": 0.295, + "step": 2288 + }, + { + "epoch": 0.8753346080305927, + "grad_norm": 3.5377202804510763, + "learning_rate": 1e-06, + "loss": 0.5638, + "step": 2289 + }, + { + "epoch": 0.875717017208413, + "grad_norm": 2.5127476888002107, + "learning_rate": 1e-06, + "loss": 0.3803, + "step": 2290 + }, + { + "epoch": 0.8760994263862333, + "grad_norm": 3.7291951666796046, + "learning_rate": 1e-06, + "loss": 0.4362, + "step": 2291 + }, + { + "epoch": 0.8764818355640536, + "grad_norm": 2.9703860944513365, + "learning_rate": 1e-06, + "loss": 0.2383, + "step": 2292 + }, + { + "epoch": 0.8768642447418739, + "grad_norm": 3.187304920884007, + "learning_rate": 1e-06, + "loss": 0.3238, + "step": 2293 + }, + { + "epoch": 0.877246653919694, + "grad_norm": 2.690191916812979, + "learning_rate": 1e-06, + "loss": 0.3012, + "step": 2294 + }, + { + "epoch": 0.8776290630975143, + "grad_norm": 1.8738393370114195, + "learning_rate": 1e-06, + "loss": 0.2375, + "step": 2295 + }, + { + "epoch": 0.8780114722753346, + "grad_norm": 2.5574251491000695, + "learning_rate": 1e-06, + "loss": 0.2422, + "step": 2296 + }, + { + "epoch": 0.8783938814531549, + "grad_norm": 3.842897149970192, + "learning_rate": 1e-06, + "loss": 0.2012, + "step": 2297 + }, + { + "epoch": 0.8787762906309752, + "grad_norm": 4.2444895755793945, + "learning_rate": 1e-06, + "loss": 0.1554, + "step": 2298 + }, + { + "epoch": 0.8791586998087955, + "grad_norm": 3.3474364592466648, + "learning_rate": 1e-06, + "loss": 0.0742, + "step": 2299 + }, + { + "epoch": 0.8795411089866156, + "grad_norm": 4.249059011967546, + "learning_rate": 1e-06, + "loss": 0.146, + "step": 2300 + }, + { + "epoch": 0.8799235181644359, + "grad_norm": 8.497555549532043, + "learning_rate": 1e-06, + "loss": 0.4556, + "step": 2301 + }, + { + "epoch": 0.8803059273422562, + "grad_norm": 4.212337339073676, + "learning_rate": 1e-06, + "loss": 0.5643, + "step": 2302 + }, + { + "epoch": 0.8806883365200765, + "grad_norm": 3.865618023137525, + "learning_rate": 1e-06, + "loss": 0.508, + "step": 2303 + }, + { + "epoch": 0.8810707456978968, + "grad_norm": 2.7602690651256534, + "learning_rate": 1e-06, + "loss": 0.3783, + "step": 2304 + }, + { + "epoch": 0.8814531548757171, + "grad_norm": 1.8248915156689647, + "learning_rate": 1e-06, + "loss": 0.233, + "step": 2305 + }, + { + "epoch": 0.8818355640535372, + "grad_norm": 2.1531514557872695, + "learning_rate": 1e-06, + "loss": 0.3302, + "step": 2306 + }, + { + "epoch": 0.8822179732313575, + "grad_norm": 3.267504575532075, + "learning_rate": 1e-06, + "loss": 0.3309, + "step": 2307 + }, + { + "epoch": 0.8826003824091778, + "grad_norm": 3.3718184557425874, + "learning_rate": 1e-06, + "loss": 0.2072, + "step": 2308 + }, + { + "epoch": 0.8829827915869981, + "grad_norm": 2.665566565173866, + "learning_rate": 1e-06, + "loss": 0.0938, + "step": 2309 + }, + { + "epoch": 0.8833652007648184, + "grad_norm": 3.2153290281520714, + "learning_rate": 1e-06, + "loss": 0.2025, + "step": 2310 + }, + { + "epoch": 0.8837476099426386, + "grad_norm": 3.020818004350026, + "learning_rate": 1e-06, + "loss": 0.1081, + "step": 2311 + }, + { + "epoch": 0.8841300191204589, + "grad_norm": 5.98992423593946, + "learning_rate": 1e-06, + "loss": 0.1368, + "step": 2312 + }, + { + "epoch": 0.8845124282982791, + "grad_norm": 7.505415677752273, + "learning_rate": 1e-06, + "loss": 0.2759, + "step": 2313 + }, + { + "epoch": 0.8848948374760994, + "grad_norm": 4.147387700414422, + "learning_rate": 1e-06, + "loss": 0.382, + "step": 2314 + }, + { + "epoch": 0.8852772466539197, + "grad_norm": 2.1308112624179936, + "learning_rate": 1e-06, + "loss": 0.475, + "step": 2315 + }, + { + "epoch": 0.88565965583174, + "grad_norm": 3.5914450509956697, + "learning_rate": 1e-06, + "loss": 0.1266, + "step": 2316 + }, + { + "epoch": 0.8860420650095602, + "grad_norm": 3.7600621331945296, + "learning_rate": 1e-06, + "loss": 0.2785, + "step": 2317 + }, + { + "epoch": 0.8864244741873805, + "grad_norm": 2.826946455570886, + "learning_rate": 1e-06, + "loss": 0.3301, + "step": 2318 + }, + { + "epoch": 0.8868068833652007, + "grad_norm": 2.580528017911607, + "learning_rate": 1e-06, + "loss": 0.4009, + "step": 2319 + }, + { + "epoch": 0.887189292543021, + "grad_norm": 4.593738789447418, + "learning_rate": 1e-06, + "loss": 0.3341, + "step": 2320 + }, + { + "epoch": 0.8875717017208413, + "grad_norm": 3.9482181537961933, + "learning_rate": 1e-06, + "loss": 0.3531, + "step": 2321 + }, + { + "epoch": 0.8879541108986616, + "grad_norm": 3.688090778160383, + "learning_rate": 1e-06, + "loss": 0.1288, + "step": 2322 + }, + { + "epoch": 0.8883365200764818, + "grad_norm": 2.5997625315867006, + "learning_rate": 1e-06, + "loss": 0.0716, + "step": 2323 + }, + { + "epoch": 0.8887189292543021, + "grad_norm": 4.739358727579254, + "learning_rate": 1e-06, + "loss": 0.1075, + "step": 2324 + }, + { + "epoch": 0.8891013384321224, + "grad_norm": 3.958546535706509, + "learning_rate": 1e-06, + "loss": 0.1491, + "step": 2325 + }, + { + "epoch": 0.8894837476099426, + "grad_norm": 8.595346531387138, + "learning_rate": 1e-06, + "loss": 0.4131, + "step": 2326 + }, + { + "epoch": 0.8898661567877629, + "grad_norm": 3.994591871667418, + "learning_rate": 1e-06, + "loss": 0.3402, + "step": 2327 + }, + { + "epoch": 0.8902485659655832, + "grad_norm": 5.926634151911645, + "learning_rate": 1e-06, + "loss": 0.4476, + "step": 2328 + }, + { + "epoch": 0.8906309751434034, + "grad_norm": 3.0411437294890624, + "learning_rate": 1e-06, + "loss": 0.3332, + "step": 2329 + }, + { + "epoch": 0.8910133843212237, + "grad_norm": 5.351350432043598, + "learning_rate": 1e-06, + "loss": 0.3868, + "step": 2330 + }, + { + "epoch": 0.891395793499044, + "grad_norm": 4.472179670638214, + "learning_rate": 1e-06, + "loss": 0.3971, + "step": 2331 + }, + { + "epoch": 0.8917782026768643, + "grad_norm": 3.865687347131359, + "learning_rate": 1e-06, + "loss": 0.1836, + "step": 2332 + }, + { + "epoch": 0.8921606118546845, + "grad_norm": 2.8248239715557286, + "learning_rate": 1e-06, + "loss": 0.2198, + "step": 2333 + }, + { + "epoch": 0.8925430210325048, + "grad_norm": 2.7267736664679947, + "learning_rate": 1e-06, + "loss": 0.1597, + "step": 2334 + }, + { + "epoch": 0.892925430210325, + "grad_norm": 3.587110672403936, + "learning_rate": 1e-06, + "loss": 0.3088, + "step": 2335 + }, + { + "epoch": 0.8933078393881453, + "grad_norm": 3.054779597696523, + "learning_rate": 1e-06, + "loss": 0.0827, + "step": 2336 + }, + { + "epoch": 0.8936902485659656, + "grad_norm": 5.234545895647738, + "learning_rate": 1e-06, + "loss": 0.2402, + "step": 2337 + }, + { + "epoch": 0.8940726577437859, + "grad_norm": 5.033355651368099, + "learning_rate": 1e-06, + "loss": 0.1738, + "step": 2338 + }, + { + "epoch": 0.8944550669216061, + "grad_norm": 5.520875846651122, + "learning_rate": 1e-06, + "loss": 0.5797, + "step": 2339 + }, + { + "epoch": 0.8948374760994264, + "grad_norm": 3.5991780296358877, + "learning_rate": 1e-06, + "loss": 0.4866, + "step": 2340 + }, + { + "epoch": 0.8952198852772466, + "grad_norm": 2.7136232994721263, + "learning_rate": 1e-06, + "loss": 0.5725, + "step": 2341 + }, + { + "epoch": 0.8956022944550669, + "grad_norm": 2.5631846001892287, + "learning_rate": 1e-06, + "loss": 0.313, + "step": 2342 + }, + { + "epoch": 0.8959847036328872, + "grad_norm": 3.3386419621407724, + "learning_rate": 1e-06, + "loss": 0.3095, + "step": 2343 + }, + { + "epoch": 0.8963671128107075, + "grad_norm": 3.245795244247259, + "learning_rate": 1e-06, + "loss": 0.4238, + "step": 2344 + }, + { + "epoch": 0.8967495219885278, + "grad_norm": 2.2980130029879846, + "learning_rate": 1e-06, + "loss": 0.2659, + "step": 2345 + }, + { + "epoch": 0.897131931166348, + "grad_norm": 2.595626703634238, + "learning_rate": 1e-06, + "loss": 0.2533, + "step": 2346 + }, + { + "epoch": 0.8975143403441682, + "grad_norm": 2.7846215204267755, + "learning_rate": 1e-06, + "loss": 0.2174, + "step": 2347 + }, + { + "epoch": 0.8978967495219885, + "grad_norm": 3.3081022615874676, + "learning_rate": 1e-06, + "loss": 0.0999, + "step": 2348 + }, + { + "epoch": 0.8982791586998088, + "grad_norm": 3.0943818651222195, + "learning_rate": 1e-06, + "loss": 0.1473, + "step": 2349 + }, + { + "epoch": 0.8986615678776291, + "grad_norm": 4.190605947520891, + "learning_rate": 1e-06, + "loss": 0.1731, + "step": 2350 + }, + { + "epoch": 0.8990439770554494, + "grad_norm": 4.72621948006337, + "learning_rate": 1e-06, + "loss": 0.4197, + "step": 2351 + }, + { + "epoch": 0.8994263862332696, + "grad_norm": 3.5235686351792337, + "learning_rate": 1e-06, + "loss": 0.5667, + "step": 2352 + }, + { + "epoch": 0.8998087954110898, + "grad_norm": 2.469793075202874, + "learning_rate": 1e-06, + "loss": 0.2133, + "step": 2353 + }, + { + "epoch": 0.9001912045889101, + "grad_norm": 2.807314351874271, + "learning_rate": 1e-06, + "loss": 0.4659, + "step": 2354 + }, + { + "epoch": 0.9005736137667304, + "grad_norm": 3.795765153347245, + "learning_rate": 1e-06, + "loss": 0.4197, + "step": 2355 + }, + { + "epoch": 0.9009560229445507, + "grad_norm": 4.375152585229796, + "learning_rate": 1e-06, + "loss": 0.4471, + "step": 2356 + }, + { + "epoch": 0.901338432122371, + "grad_norm": 3.7510101229460644, + "learning_rate": 1e-06, + "loss": 0.2843, + "step": 2357 + }, + { + "epoch": 0.9017208413001913, + "grad_norm": 3.349252326998123, + "learning_rate": 1e-06, + "loss": 0.334, + "step": 2358 + }, + { + "epoch": 0.9021032504780114, + "grad_norm": 2.9929499439115674, + "learning_rate": 1e-06, + "loss": 0.2409, + "step": 2359 + }, + { + "epoch": 0.9024856596558317, + "grad_norm": 4.8274300578190426, + "learning_rate": 1e-06, + "loss": 0.1843, + "step": 2360 + }, + { + "epoch": 0.902868068833652, + "grad_norm": 3.04919407937432, + "learning_rate": 1e-06, + "loss": 0.12, + "step": 2361 + }, + { + "epoch": 0.9032504780114723, + "grad_norm": 3.614721920350499, + "learning_rate": 1e-06, + "loss": 0.1164, + "step": 2362 + }, + { + "epoch": 0.9036328871892926, + "grad_norm": 5.657648577523865, + "learning_rate": 1e-06, + "loss": 0.151, + "step": 2363 + }, + { + "epoch": 0.9040152963671129, + "grad_norm": 3.7782924214287825, + "learning_rate": 1e-06, + "loss": 0.5837, + "step": 2364 + }, + { + "epoch": 0.904397705544933, + "grad_norm": 3.29169842346582, + "learning_rate": 1e-06, + "loss": 0.4038, + "step": 2365 + }, + { + "epoch": 0.9047801147227533, + "grad_norm": 2.804095957650072, + "learning_rate": 1e-06, + "loss": 0.4941, + "step": 2366 + }, + { + "epoch": 0.9051625239005736, + "grad_norm": 2.6948585017325772, + "learning_rate": 1e-06, + "loss": 0.399, + "step": 2367 + }, + { + "epoch": 0.9055449330783939, + "grad_norm": 2.8048690694622755, + "learning_rate": 1e-06, + "loss": 0.3361, + "step": 2368 + }, + { + "epoch": 0.9059273422562142, + "grad_norm": 3.4714349201689294, + "learning_rate": 1e-06, + "loss": 0.4146, + "step": 2369 + }, + { + "epoch": 0.9063097514340345, + "grad_norm": 3.056515666125883, + "learning_rate": 1e-06, + "loss": 0.2881, + "step": 2370 + }, + { + "epoch": 0.9066921606118546, + "grad_norm": 3.8923882468271436, + "learning_rate": 1e-06, + "loss": 0.3238, + "step": 2371 + }, + { + "epoch": 0.9070745697896749, + "grad_norm": 3.9766025503723563, + "learning_rate": 1e-06, + "loss": 0.3097, + "step": 2372 + }, + { + "epoch": 0.9074569789674952, + "grad_norm": 3.991801204020184, + "learning_rate": 1e-06, + "loss": 0.2089, + "step": 2373 + }, + { + "epoch": 0.9078393881453155, + "grad_norm": 3.8195215822387363, + "learning_rate": 1e-06, + "loss": 0.1582, + "step": 2374 + }, + { + "epoch": 0.9082217973231358, + "grad_norm": 3.1414569113516273, + "learning_rate": 1e-06, + "loss": 0.0923, + "step": 2375 + }, + { + "epoch": 0.908604206500956, + "grad_norm": 4.024326974268517, + "learning_rate": 1e-06, + "loss": 0.3416, + "step": 2376 + }, + { + "epoch": 0.9089866156787763, + "grad_norm": 3.5697654121323223, + "learning_rate": 1e-06, + "loss": 0.6562, + "step": 2377 + }, + { + "epoch": 0.9093690248565965, + "grad_norm": 3.3032838261728106, + "learning_rate": 1e-06, + "loss": 0.3549, + "step": 2378 + }, + { + "epoch": 0.9097514340344168, + "grad_norm": 3.729689975157589, + "learning_rate": 1e-06, + "loss": 0.5737, + "step": 2379 + }, + { + "epoch": 0.9101338432122371, + "grad_norm": 3.084084874569733, + "learning_rate": 1e-06, + "loss": 0.4262, + "step": 2380 + }, + { + "epoch": 0.9105162523900574, + "grad_norm": 2.7431866599727837, + "learning_rate": 1e-06, + "loss": 0.2823, + "step": 2381 + }, + { + "epoch": 0.9108986615678776, + "grad_norm": 3.3325277944609093, + "learning_rate": 1e-06, + "loss": 0.1517, + "step": 2382 + }, + { + "epoch": 0.9112810707456979, + "grad_norm": 3.2485280738522864, + "learning_rate": 1e-06, + "loss": 0.3246, + "step": 2383 + }, + { + "epoch": 0.9116634799235181, + "grad_norm": 2.91043118195861, + "learning_rate": 1e-06, + "loss": 0.1999, + "step": 2384 + }, + { + "epoch": 0.9120458891013384, + "grad_norm": 2.892826705853231, + "learning_rate": 1e-06, + "loss": 0.1659, + "step": 2385 + }, + { + "epoch": 0.9124282982791587, + "grad_norm": 2.7669380648525954, + "learning_rate": 1e-06, + "loss": 0.1941, + "step": 2386 + }, + { + "epoch": 0.912810707456979, + "grad_norm": 3.397510761752007, + "learning_rate": 1e-06, + "loss": 0.0964, + "step": 2387 + }, + { + "epoch": 0.9131931166347992, + "grad_norm": 6.884696918983418, + "learning_rate": 1e-06, + "loss": 0.1733, + "step": 2388 + }, + { + "epoch": 0.9135755258126195, + "grad_norm": 4.374020276044012, + "learning_rate": 1e-06, + "loss": 0.4105, + "step": 2389 + }, + { + "epoch": 0.9139579349904398, + "grad_norm": 2.4249034036490626, + "learning_rate": 1e-06, + "loss": 0.3799, + "step": 2390 + }, + { + "epoch": 0.91434034416826, + "grad_norm": 1.7467237867717293, + "learning_rate": 1e-06, + "loss": 0.5461, + "step": 2391 + }, + { + "epoch": 0.9147227533460803, + "grad_norm": 3.769540990312869, + "learning_rate": 1e-06, + "loss": 0.3656, + "step": 2392 + }, + { + "epoch": 0.9151051625239006, + "grad_norm": 3.051950618909361, + "learning_rate": 1e-06, + "loss": 0.2499, + "step": 2393 + }, + { + "epoch": 0.9154875717017208, + "grad_norm": 2.3185247347830433, + "learning_rate": 1e-06, + "loss": 0.2253, + "step": 2394 + }, + { + "epoch": 0.9158699808795411, + "grad_norm": 3.6773879201524555, + "learning_rate": 1e-06, + "loss": 0.2363, + "step": 2395 + }, + { + "epoch": 0.9162523900573614, + "grad_norm": 2.7726956946995234, + "learning_rate": 1e-06, + "loss": 0.2691, + "step": 2396 + }, + { + "epoch": 0.9166347992351817, + "grad_norm": 4.293276636053358, + "learning_rate": 1e-06, + "loss": 0.1415, + "step": 2397 + }, + { + "epoch": 0.9170172084130019, + "grad_norm": 3.2937914272070086, + "learning_rate": 1e-06, + "loss": 0.1545, + "step": 2398 + }, + { + "epoch": 0.9173996175908222, + "grad_norm": 5.068736061989545, + "learning_rate": 1e-06, + "loss": 0.1117, + "step": 2399 + }, + { + "epoch": 0.9177820267686424, + "grad_norm": 4.144171811551991, + "learning_rate": 1e-06, + "loss": 0.211, + "step": 2400 + }, + { + "epoch": 0.9181644359464627, + "grad_norm": 4.271453564275025, + "learning_rate": 1e-06, + "loss": 0.4232, + "step": 2401 + }, + { + "epoch": 0.918546845124283, + "grad_norm": 4.142120897238968, + "learning_rate": 1e-06, + "loss": 0.3927, + "step": 2402 + }, + { + "epoch": 0.9189292543021033, + "grad_norm": 2.4200139924898028, + "learning_rate": 1e-06, + "loss": 0.3811, + "step": 2403 + }, + { + "epoch": 0.9193116634799235, + "grad_norm": 4.018243907902251, + "learning_rate": 1e-06, + "loss": 0.4066, + "step": 2404 + }, + { + "epoch": 0.9196940726577438, + "grad_norm": 2.7654656833084443, + "learning_rate": 1e-06, + "loss": 0.3045, + "step": 2405 + }, + { + "epoch": 0.920076481835564, + "grad_norm": 3.8460711162179333, + "learning_rate": 1e-06, + "loss": 0.276, + "step": 2406 + }, + { + "epoch": 0.9204588910133843, + "grad_norm": 3.2811800994693674, + "learning_rate": 1e-06, + "loss": 0.3471, + "step": 2407 + }, + { + "epoch": 0.9208413001912046, + "grad_norm": 2.816157188460834, + "learning_rate": 1e-06, + "loss": 0.271, + "step": 2408 + }, + { + "epoch": 0.9212237093690249, + "grad_norm": 4.139433145830336, + "learning_rate": 1e-06, + "loss": 0.321, + "step": 2409 + }, + { + "epoch": 0.9216061185468452, + "grad_norm": 3.250587116908922, + "learning_rate": 1e-06, + "loss": 0.1285, + "step": 2410 + }, + { + "epoch": 0.9219885277246654, + "grad_norm": 2.511674990354371, + "learning_rate": 1e-06, + "loss": 0.1159, + "step": 2411 + }, + { + "epoch": 0.9223709369024856, + "grad_norm": 2.8328083524342222, + "learning_rate": 1e-06, + "loss": 0.0856, + "step": 2412 + }, + { + "epoch": 0.9227533460803059, + "grad_norm": 4.623206950914736, + "learning_rate": 1e-06, + "loss": 0.2897, + "step": 2413 + }, + { + "epoch": 0.9231357552581262, + "grad_norm": 5.678655769427854, + "learning_rate": 1e-06, + "loss": 0.5478, + "step": 2414 + }, + { + "epoch": 0.9235181644359465, + "grad_norm": 3.1158680047222096, + "learning_rate": 1e-06, + "loss": 0.6466, + "step": 2415 + }, + { + "epoch": 0.9239005736137668, + "grad_norm": 2.4494728066125724, + "learning_rate": 1e-06, + "loss": 0.2526, + "step": 2416 + }, + { + "epoch": 0.924282982791587, + "grad_norm": 2.9669818632152016, + "learning_rate": 1e-06, + "loss": 0.3885, + "step": 2417 + }, + { + "epoch": 0.9246653919694072, + "grad_norm": 3.402103525083524, + "learning_rate": 1e-06, + "loss": 0.2849, + "step": 2418 + }, + { + "epoch": 0.9250478011472275, + "grad_norm": 3.453406274999731, + "learning_rate": 1e-06, + "loss": 0.3591, + "step": 2419 + }, + { + "epoch": 0.9254302103250478, + "grad_norm": 2.0307213902395898, + "learning_rate": 1e-06, + "loss": 0.1947, + "step": 2420 + }, + { + "epoch": 0.9258126195028681, + "grad_norm": 2.5965596802903117, + "learning_rate": 1e-06, + "loss": 0.1767, + "step": 2421 + }, + { + "epoch": 0.9261950286806884, + "grad_norm": 2.2293907005413836, + "learning_rate": 1e-06, + "loss": 0.0941, + "step": 2422 + }, + { + "epoch": 0.9265774378585087, + "grad_norm": 3.9204266674070825, + "learning_rate": 1e-06, + "loss": 0.2149, + "step": 2423 + }, + { + "epoch": 0.9269598470363288, + "grad_norm": 4.919483197059165, + "learning_rate": 1e-06, + "loss": 0.1295, + "step": 2424 + }, + { + "epoch": 0.9273422562141491, + "grad_norm": 7.121928573614454, + "learning_rate": 1e-06, + "loss": 0.306, + "step": 2425 + }, + { + "epoch": 0.9277246653919694, + "grad_norm": 6.867140560402005, + "learning_rate": 1e-06, + "loss": 0.3018, + "step": 2426 + }, + { + "epoch": 0.9281070745697897, + "grad_norm": 3.502978419998348, + "learning_rate": 1e-06, + "loss": 0.4303, + "step": 2427 + }, + { + "epoch": 0.92848948374761, + "grad_norm": 2.5172368921493087, + "learning_rate": 1e-06, + "loss": 0.3832, + "step": 2428 + }, + { + "epoch": 0.9288718929254303, + "grad_norm": 2.401462399463372, + "learning_rate": 1e-06, + "loss": 0.2555, + "step": 2429 + }, + { + "epoch": 0.9292543021032504, + "grad_norm": 3.824808357773478, + "learning_rate": 1e-06, + "loss": 0.1962, + "step": 2430 + }, + { + "epoch": 0.9296367112810707, + "grad_norm": 2.7748829688060845, + "learning_rate": 1e-06, + "loss": 0.3494, + "step": 2431 + }, + { + "epoch": 0.930019120458891, + "grad_norm": 2.8105509468351326, + "learning_rate": 1e-06, + "loss": 0.2075, + "step": 2432 + }, + { + "epoch": 0.9304015296367113, + "grad_norm": 2.543391648153547, + "learning_rate": 1e-06, + "loss": 0.1866, + "step": 2433 + }, + { + "epoch": 0.9307839388145316, + "grad_norm": 3.660156771111222, + "learning_rate": 1e-06, + "loss": 0.3947, + "step": 2434 + }, + { + "epoch": 0.9311663479923518, + "grad_norm": 2.6722867771087198, + "learning_rate": 1e-06, + "loss": 0.1001, + "step": 2435 + }, + { + "epoch": 0.931548757170172, + "grad_norm": 4.882081585919296, + "learning_rate": 1e-06, + "loss": 0.1693, + "step": 2436 + }, + { + "epoch": 0.9319311663479923, + "grad_norm": 3.8891033582736587, + "learning_rate": 1e-06, + "loss": 0.0689, + "step": 2437 + }, + { + "epoch": 0.9323135755258126, + "grad_norm": 5.030727762164894, + "learning_rate": 1e-06, + "loss": 0.1524, + "step": 2438 + }, + { + "epoch": 0.9326959847036329, + "grad_norm": 2.806366824158863, + "learning_rate": 1e-06, + "loss": 0.3798, + "step": 2439 + }, + { + "epoch": 0.9330783938814532, + "grad_norm": 3.993293385080311, + "learning_rate": 1e-06, + "loss": 0.4549, + "step": 2440 + }, + { + "epoch": 0.9334608030592734, + "grad_norm": 2.6540438297737485, + "learning_rate": 1e-06, + "loss": 0.4144, + "step": 2441 + }, + { + "epoch": 0.9338432122370937, + "grad_norm": 3.9939570079279467, + "learning_rate": 1e-06, + "loss": 0.2668, + "step": 2442 + }, + { + "epoch": 0.9342256214149139, + "grad_norm": 3.8146261321492796, + "learning_rate": 1e-06, + "loss": 0.4552, + "step": 2443 + }, + { + "epoch": 0.9346080305927342, + "grad_norm": 3.880128973164252, + "learning_rate": 1e-06, + "loss": 0.1813, + "step": 2444 + }, + { + "epoch": 0.9349904397705545, + "grad_norm": 4.808416243994762, + "learning_rate": 1e-06, + "loss": 0.2827, + "step": 2445 + }, + { + "epoch": 0.9353728489483748, + "grad_norm": 3.4106847609138664, + "learning_rate": 1e-06, + "loss": 0.1591, + "step": 2446 + }, + { + "epoch": 0.935755258126195, + "grad_norm": 2.4377755229460027, + "learning_rate": 1e-06, + "loss": 0.1474, + "step": 2447 + }, + { + "epoch": 0.9361376673040153, + "grad_norm": 3.229732406444275, + "learning_rate": 1e-06, + "loss": 0.1188, + "step": 2448 + }, + { + "epoch": 0.9365200764818356, + "grad_norm": 3.598057296881959, + "learning_rate": 1e-06, + "loss": 0.1297, + "step": 2449 + }, + { + "epoch": 0.9369024856596558, + "grad_norm": 4.248837031398682, + "learning_rate": 1e-06, + "loss": 0.1598, + "step": 2450 + }, + { + "epoch": 0.9372848948374761, + "grad_norm": 5.293401338699338, + "learning_rate": 1e-06, + "loss": 0.5099, + "step": 2451 + }, + { + "epoch": 0.9376673040152964, + "grad_norm": 4.239485131078506, + "learning_rate": 1e-06, + "loss": 0.4603, + "step": 2452 + }, + { + "epoch": 0.9380497131931166, + "grad_norm": 3.178237621971313, + "learning_rate": 1e-06, + "loss": 0.3857, + "step": 2453 + }, + { + "epoch": 0.9384321223709369, + "grad_norm": 2.820377761212562, + "learning_rate": 1e-06, + "loss": 0.4409, + "step": 2454 + }, + { + "epoch": 0.9388145315487572, + "grad_norm": 4.224050173562621, + "learning_rate": 1e-06, + "loss": 0.4657, + "step": 2455 + }, + { + "epoch": 0.9391969407265774, + "grad_norm": 3.480768501226541, + "learning_rate": 1e-06, + "loss": 0.2117, + "step": 2456 + }, + { + "epoch": 0.9395793499043977, + "grad_norm": 3.3157551444179916, + "learning_rate": 1e-06, + "loss": 0.2881, + "step": 2457 + }, + { + "epoch": 0.939961759082218, + "grad_norm": 2.677380605933176, + "learning_rate": 1e-06, + "loss": 0.2783, + "step": 2458 + }, + { + "epoch": 0.9403441682600382, + "grad_norm": 3.4601604889710007, + "learning_rate": 1e-06, + "loss": 0.1185, + "step": 2459 + }, + { + "epoch": 0.9407265774378585, + "grad_norm": 2.7606470167868156, + "learning_rate": 1e-06, + "loss": 0.1084, + "step": 2460 + }, + { + "epoch": 0.9411089866156788, + "grad_norm": 4.2092945524484815, + "learning_rate": 1e-06, + "loss": 0.1193, + "step": 2461 + }, + { + "epoch": 0.941491395793499, + "grad_norm": 4.857380252133691, + "learning_rate": 1e-06, + "loss": 0.2438, + "step": 2462 + }, + { + "epoch": 0.9418738049713193, + "grad_norm": 4.812435991926709, + "learning_rate": 1e-06, + "loss": 0.1943, + "step": 2463 + }, + { + "epoch": 0.9422562141491396, + "grad_norm": 3.0303294563586163, + "learning_rate": 1e-06, + "loss": 0.3405, + "step": 2464 + }, + { + "epoch": 0.9426386233269598, + "grad_norm": 3.4931876779237387, + "learning_rate": 1e-06, + "loss": 0.3735, + "step": 2465 + }, + { + "epoch": 0.9430210325047801, + "grad_norm": 3.105325947483179, + "learning_rate": 1e-06, + "loss": 0.193, + "step": 2466 + }, + { + "epoch": 0.9434034416826004, + "grad_norm": 3.3599543581996127, + "learning_rate": 1e-06, + "loss": 0.4249, + "step": 2467 + }, + { + "epoch": 0.9437858508604207, + "grad_norm": 3.9436596117981617, + "learning_rate": 1e-06, + "loss": 0.4425, + "step": 2468 + }, + { + "epoch": 0.944168260038241, + "grad_norm": 2.914177531817628, + "learning_rate": 1e-06, + "loss": 0.4259, + "step": 2469 + }, + { + "epoch": 0.9445506692160612, + "grad_norm": 3.3477942706469674, + "learning_rate": 1e-06, + "loss": 0.1776, + "step": 2470 + }, + { + "epoch": 0.9449330783938814, + "grad_norm": 3.404480500775937, + "learning_rate": 1e-06, + "loss": 0.3247, + "step": 2471 + }, + { + "epoch": 0.9453154875717017, + "grad_norm": 4.0445641949383795, + "learning_rate": 1e-06, + "loss": 0.1704, + "step": 2472 + }, + { + "epoch": 0.945697896749522, + "grad_norm": 4.140376706146293, + "learning_rate": 1e-06, + "loss": 0.1252, + "step": 2473 + }, + { + "epoch": 0.9460803059273423, + "grad_norm": 4.425210618001349, + "learning_rate": 1e-06, + "loss": 0.166, + "step": 2474 + }, + { + "epoch": 0.9464627151051626, + "grad_norm": 3.84246047804393, + "learning_rate": 1e-06, + "loss": 0.1434, + "step": 2475 + }, + { + "epoch": 0.9468451242829828, + "grad_norm": 4.633907503919647, + "learning_rate": 1e-06, + "loss": 0.5574, + "step": 2476 + }, + { + "epoch": 0.947227533460803, + "grad_norm": 3.4089993921191524, + "learning_rate": 1e-06, + "loss": 0.3722, + "step": 2477 + }, + { + "epoch": 0.9476099426386233, + "grad_norm": 2.981419562869889, + "learning_rate": 1e-06, + "loss": 0.3436, + "step": 2478 + }, + { + "epoch": 0.9479923518164436, + "grad_norm": 3.10857530594632, + "learning_rate": 1e-06, + "loss": 0.4303, + "step": 2479 + }, + { + "epoch": 0.9483747609942639, + "grad_norm": 2.475556467095296, + "learning_rate": 1e-06, + "loss": 0.4754, + "step": 2480 + }, + { + "epoch": 0.9487571701720842, + "grad_norm": 3.4656399367950486, + "learning_rate": 1e-06, + "loss": 0.3549, + "step": 2481 + }, + { + "epoch": 0.9491395793499044, + "grad_norm": 2.481484416967314, + "learning_rate": 1e-06, + "loss": 0.2656, + "step": 2482 + }, + { + "epoch": 0.9495219885277246, + "grad_norm": 3.2459105659146714, + "learning_rate": 1e-06, + "loss": 0.1877, + "step": 2483 + }, + { + "epoch": 0.9499043977055449, + "grad_norm": 2.9867060480189456, + "learning_rate": 1e-06, + "loss": 0.2577, + "step": 2484 + }, + { + "epoch": 0.9502868068833652, + "grad_norm": 2.9665059492603536, + "learning_rate": 1e-06, + "loss": 0.1622, + "step": 2485 + }, + { + "epoch": 0.9506692160611855, + "grad_norm": 3.2158824406103794, + "learning_rate": 1e-06, + "loss": 0.1046, + "step": 2486 + }, + { + "epoch": 0.9510516252390058, + "grad_norm": 7.309422032668531, + "learning_rate": 1e-06, + "loss": 0.2142, + "step": 2487 + }, + { + "epoch": 0.9514340344168261, + "grad_norm": 6.018173985632012, + "learning_rate": 1e-06, + "loss": 0.129, + "step": 2488 + }, + { + "epoch": 0.9518164435946462, + "grad_norm": 3.9050742859557754, + "learning_rate": 1e-06, + "loss": 0.2745, + "step": 2489 + }, + { + "epoch": 0.9521988527724665, + "grad_norm": 3.0204257207766267, + "learning_rate": 1e-06, + "loss": 0.5162, + "step": 2490 + }, + { + "epoch": 0.9525812619502868, + "grad_norm": 2.9900339210142612, + "learning_rate": 1e-06, + "loss": 0.2981, + "step": 2491 + }, + { + "epoch": 0.9529636711281071, + "grad_norm": 2.2315165769186533, + "learning_rate": 1e-06, + "loss": 0.2601, + "step": 2492 + }, + { + "epoch": 0.9533460803059274, + "grad_norm": 4.215404625202046, + "learning_rate": 1e-06, + "loss": 0.5005, + "step": 2493 + }, + { + "epoch": 0.9537284894837476, + "grad_norm": 4.286058607347795, + "learning_rate": 1e-06, + "loss": 0.3125, + "step": 2494 + }, + { + "epoch": 0.9541108986615678, + "grad_norm": 2.9477273365773953, + "learning_rate": 1e-06, + "loss": 0.1585, + "step": 2495 + }, + { + "epoch": 0.9544933078393881, + "grad_norm": 2.871610052694963, + "learning_rate": 1e-06, + "loss": 0.3119, + "step": 2496 + }, + { + "epoch": 0.9548757170172084, + "grad_norm": 4.013999045743957, + "learning_rate": 1e-06, + "loss": 0.1336, + "step": 2497 + }, + { + "epoch": 0.9552581261950287, + "grad_norm": 5.1393778021386245, + "learning_rate": 1e-06, + "loss": 0.1823, + "step": 2498 + }, + { + "epoch": 0.955640535372849, + "grad_norm": 4.134935378246604, + "learning_rate": 1e-06, + "loss": 0.1288, + "step": 2499 + }, + { + "epoch": 0.9560229445506692, + "grad_norm": 3.8629845176424045, + "learning_rate": 1e-06, + "loss": 0.2066, + "step": 2500 + }, + { + "epoch": 0.9560229445506692, + "eval_runtime": 799.3556, + "eval_samples_per_second": 1.919, + "eval_steps_per_second": 0.48, + "step": 2500 + }, + { + "epoch": 0.9564053537284894, + "grad_norm": 5.8565165553706136, + "learning_rate": 1e-06, + "loss": 0.2651, + "step": 2501 + }, + { + "epoch": 0.9567877629063097, + "grad_norm": 4.526794347596287, + "learning_rate": 1e-06, + "loss": 0.5482, + "step": 2502 + }, + { + "epoch": 0.95717017208413, + "grad_norm": 3.848471618079834, + "learning_rate": 1e-06, + "loss": 0.3268, + "step": 2503 + }, + { + "epoch": 0.9575525812619503, + "grad_norm": 2.5777081701684113, + "learning_rate": 1e-06, + "loss": 0.375, + "step": 2504 + }, + { + "epoch": 0.9579349904397706, + "grad_norm": 3.2761608075834734, + "learning_rate": 1e-06, + "loss": 0.4698, + "step": 2505 + }, + { + "epoch": 0.9583173996175908, + "grad_norm": 3.1059146201145658, + "learning_rate": 1e-06, + "loss": 0.3876, + "step": 2506 + }, + { + "epoch": 0.9586998087954111, + "grad_norm": 4.5630668392531994, + "learning_rate": 1e-06, + "loss": 0.293, + "step": 2507 + }, + { + "epoch": 0.9590822179732313, + "grad_norm": 2.5173547141041355, + "learning_rate": 1e-06, + "loss": 0.4098, + "step": 2508 + }, + { + "epoch": 0.9594646271510516, + "grad_norm": 4.096442116360893, + "learning_rate": 1e-06, + "loss": 0.3183, + "step": 2509 + }, + { + "epoch": 0.9598470363288719, + "grad_norm": 3.040329539105895, + "learning_rate": 1e-06, + "loss": 0.1837, + "step": 2510 + }, + { + "epoch": 0.9602294455066922, + "grad_norm": 4.060086046631855, + "learning_rate": 1e-06, + "loss": 0.1759, + "step": 2511 + }, + { + "epoch": 0.9606118546845124, + "grad_norm": 3.202650974833437, + "learning_rate": 1e-06, + "loss": 0.1191, + "step": 2512 + }, + { + "epoch": 0.9609942638623327, + "grad_norm": 5.244336297559902, + "learning_rate": 1e-06, + "loss": 0.221, + "step": 2513 + }, + { + "epoch": 0.961376673040153, + "grad_norm": 4.087860530153794, + "learning_rate": 1e-06, + "loss": 0.4887, + "step": 2514 + }, + { + "epoch": 0.9617590822179732, + "grad_norm": 4.008707106538918, + "learning_rate": 1e-06, + "loss": 0.2974, + "step": 2515 + }, + { + "epoch": 0.9621414913957935, + "grad_norm": 2.8413734985847023, + "learning_rate": 1e-06, + "loss": 0.2436, + "step": 2516 + }, + { + "epoch": 0.9625239005736138, + "grad_norm": 2.1263401069363788, + "learning_rate": 1e-06, + "loss": 0.4188, + "step": 2517 + }, + { + "epoch": 0.962906309751434, + "grad_norm": 2.3591884514025065, + "learning_rate": 1e-06, + "loss": 0.3366, + "step": 2518 + }, + { + "epoch": 0.9632887189292543, + "grad_norm": 3.4527239674509396, + "learning_rate": 1e-06, + "loss": 0.3246, + "step": 2519 + }, + { + "epoch": 0.9636711281070746, + "grad_norm": 3.1959436814522637, + "learning_rate": 1e-06, + "loss": 0.4524, + "step": 2520 + }, + { + "epoch": 0.9640535372848948, + "grad_norm": 2.6434737870598064, + "learning_rate": 1e-06, + "loss": 0.1332, + "step": 2521 + }, + { + "epoch": 0.9644359464627151, + "grad_norm": 2.9888744687942252, + "learning_rate": 1e-06, + "loss": 0.2597, + "step": 2522 + }, + { + "epoch": 0.9648183556405354, + "grad_norm": 2.8214012855189923, + "learning_rate": 1e-06, + "loss": 0.095, + "step": 2523 + }, + { + "epoch": 0.9652007648183556, + "grad_norm": 3.6102344100802566, + "learning_rate": 1e-06, + "loss": 0.2253, + "step": 2524 + }, + { + "epoch": 0.9655831739961759, + "grad_norm": 3.7707971532501547, + "learning_rate": 1e-06, + "loss": 0.1292, + "step": 2525 + }, + { + "epoch": 0.9659655831739962, + "grad_norm": 6.197414892919504, + "learning_rate": 1e-06, + "loss": 0.4118, + "step": 2526 + }, + { + "epoch": 0.9663479923518165, + "grad_norm": 4.232979970486748, + "learning_rate": 1e-06, + "loss": 0.525, + "step": 2527 + }, + { + "epoch": 0.9667304015296367, + "grad_norm": 2.516070213963481, + "learning_rate": 1e-06, + "loss": 0.2949, + "step": 2528 + }, + { + "epoch": 0.967112810707457, + "grad_norm": 2.5454068311021585, + "learning_rate": 1e-06, + "loss": 0.3649, + "step": 2529 + }, + { + "epoch": 0.9674952198852772, + "grad_norm": 3.0030260718779824, + "learning_rate": 1e-06, + "loss": 0.3655, + "step": 2530 + }, + { + "epoch": 0.9678776290630975, + "grad_norm": 2.587838203331088, + "learning_rate": 1e-06, + "loss": 0.1599, + "step": 2531 + }, + { + "epoch": 0.9682600382409178, + "grad_norm": 2.7870652342992006, + "learning_rate": 1e-06, + "loss": 0.2055, + "step": 2532 + }, + { + "epoch": 0.9686424474187381, + "grad_norm": 2.2474966958502223, + "learning_rate": 1e-06, + "loss": 0.2339, + "step": 2533 + }, + { + "epoch": 0.9690248565965583, + "grad_norm": 3.6074896838595163, + "learning_rate": 1e-06, + "loss": 0.2102, + "step": 2534 + }, + { + "epoch": 0.9694072657743786, + "grad_norm": 4.83622817242849, + "learning_rate": 1e-06, + "loss": 0.3518, + "step": 2535 + }, + { + "epoch": 0.9697896749521988, + "grad_norm": 4.1927200348811215, + "learning_rate": 1e-06, + "loss": 0.1836, + "step": 2536 + }, + { + "epoch": 0.9701720841300191, + "grad_norm": 2.8666457826016307, + "learning_rate": 1e-06, + "loss": 0.1056, + "step": 2537 + }, + { + "epoch": 0.9705544933078394, + "grad_norm": 4.817647979633222, + "learning_rate": 1e-06, + "loss": 0.16, + "step": 2538 + }, + { + "epoch": 0.9709369024856597, + "grad_norm": 4.854718396616762, + "learning_rate": 1e-06, + "loss": 0.3504, + "step": 2539 + }, + { + "epoch": 0.97131931166348, + "grad_norm": 4.5204929393590385, + "learning_rate": 1e-06, + "loss": 0.464, + "step": 2540 + }, + { + "epoch": 0.9717017208413002, + "grad_norm": 2.168201770803353, + "learning_rate": 1e-06, + "loss": 0.3012, + "step": 2541 + }, + { + "epoch": 0.9720841300191204, + "grad_norm": 2.494170162590611, + "learning_rate": 1e-06, + "loss": 0.3649, + "step": 2542 + }, + { + "epoch": 0.9724665391969407, + "grad_norm": 3.1576684464974107, + "learning_rate": 1e-06, + "loss": 0.3593, + "step": 2543 + }, + { + "epoch": 0.972848948374761, + "grad_norm": 2.856905678713925, + "learning_rate": 1e-06, + "loss": 0.4119, + "step": 2544 + }, + { + "epoch": 0.9732313575525813, + "grad_norm": 2.727520531175859, + "learning_rate": 1e-06, + "loss": 0.2192, + "step": 2545 + }, + { + "epoch": 0.9736137667304016, + "grad_norm": 4.18918313888206, + "learning_rate": 1e-06, + "loss": 0.2203, + "step": 2546 + }, + { + "epoch": 0.9739961759082219, + "grad_norm": 2.975542990234044, + "learning_rate": 1e-06, + "loss": 0.1945, + "step": 2547 + }, + { + "epoch": 0.974378585086042, + "grad_norm": 3.5107795384400453, + "learning_rate": 1e-06, + "loss": 0.2166, + "step": 2548 + }, + { + "epoch": 0.9747609942638623, + "grad_norm": 4.385758387404672, + "learning_rate": 1e-06, + "loss": 0.1169, + "step": 2549 + }, + { + "epoch": 0.9751434034416826, + "grad_norm": 4.209472401376355, + "learning_rate": 1e-06, + "loss": 0.0965, + "step": 2550 + }, + { + "epoch": 0.9755258126195029, + "grad_norm": 5.427274186872016, + "learning_rate": 1e-06, + "loss": 0.367, + "step": 2551 + }, + { + "epoch": 0.9759082217973232, + "grad_norm": 3.2644648339131286, + "learning_rate": 1e-06, + "loss": 0.2814, + "step": 2552 + }, + { + "epoch": 0.9762906309751435, + "grad_norm": 2.6137838065735473, + "learning_rate": 1e-06, + "loss": 0.4892, + "step": 2553 + }, + { + "epoch": 0.9766730401529636, + "grad_norm": 4.326270632337801, + "learning_rate": 1e-06, + "loss": 0.2991, + "step": 2554 + }, + { + "epoch": 0.9770554493307839, + "grad_norm": 3.050565548352456, + "learning_rate": 1e-06, + "loss": 0.1965, + "step": 2555 + }, + { + "epoch": 0.9774378585086042, + "grad_norm": 2.7712793684363803, + "learning_rate": 1e-06, + "loss": 0.4099, + "step": 2556 + }, + { + "epoch": 0.9778202676864245, + "grad_norm": 2.672851116499552, + "learning_rate": 1e-06, + "loss": 0.3241, + "step": 2557 + }, + { + "epoch": 0.9782026768642448, + "grad_norm": 3.88321062038747, + "learning_rate": 1e-06, + "loss": 0.2394, + "step": 2558 + }, + { + "epoch": 0.978585086042065, + "grad_norm": 2.5404347628870267, + "learning_rate": 1e-06, + "loss": 0.1967, + "step": 2559 + }, + { + "epoch": 0.9789674952198852, + "grad_norm": 4.196580748422142, + "learning_rate": 1e-06, + "loss": 0.1953, + "step": 2560 + }, + { + "epoch": 0.9793499043977055, + "grad_norm": 4.321526410325028, + "learning_rate": 1e-06, + "loss": 0.1945, + "step": 2561 + }, + { + "epoch": 0.9797323135755258, + "grad_norm": 3.620238498397353, + "learning_rate": 1e-06, + "loss": 0.128, + "step": 2562 + }, + { + "epoch": 0.9801147227533461, + "grad_norm": 5.664534261064663, + "learning_rate": 1e-06, + "loss": 0.3241, + "step": 2563 + }, + { + "epoch": 0.9804971319311664, + "grad_norm": 6.847996060806995, + "learning_rate": 1e-06, + "loss": 0.5073, + "step": 2564 + }, + { + "epoch": 0.9808795411089866, + "grad_norm": 3.571970882113405, + "learning_rate": 1e-06, + "loss": 0.513, + "step": 2565 + }, + { + "epoch": 0.9812619502868068, + "grad_norm": 2.4624811084622946, + "learning_rate": 1e-06, + "loss": 0.2175, + "step": 2566 + }, + { + "epoch": 0.9816443594646271, + "grad_norm": 3.7807054363969903, + "learning_rate": 1e-06, + "loss": 0.3091, + "step": 2567 + }, + { + "epoch": 0.9820267686424474, + "grad_norm": 2.8580470936274107, + "learning_rate": 1e-06, + "loss": 0.3945, + "step": 2568 + }, + { + "epoch": 0.9824091778202677, + "grad_norm": 2.351385673895787, + "learning_rate": 1e-06, + "loss": 0.1745, + "step": 2569 + }, + { + "epoch": 0.982791586998088, + "grad_norm": 2.1665177171811596, + "learning_rate": 1e-06, + "loss": 0.3091, + "step": 2570 + }, + { + "epoch": 0.9831739961759082, + "grad_norm": 2.3260906983631826, + "learning_rate": 1e-06, + "loss": 0.107, + "step": 2571 + }, + { + "epoch": 0.9835564053537285, + "grad_norm": 3.1163465087244333, + "learning_rate": 1e-06, + "loss": 0.2531, + "step": 2572 + }, + { + "epoch": 0.9839388145315487, + "grad_norm": 5.852612634199883, + "learning_rate": 1e-06, + "loss": 0.2657, + "step": 2573 + }, + { + "epoch": 0.984321223709369, + "grad_norm": 3.8884425648894876, + "learning_rate": 1e-06, + "loss": 0.2904, + "step": 2574 + }, + { + "epoch": 0.9847036328871893, + "grad_norm": 4.19222505362827, + "learning_rate": 1e-06, + "loss": 0.1286, + "step": 2575 + }, + { + "epoch": 0.9850860420650096, + "grad_norm": 5.189746600012265, + "learning_rate": 1e-06, + "loss": 0.4261, + "step": 2576 + }, + { + "epoch": 0.9854684512428298, + "grad_norm": 4.696277385973995, + "learning_rate": 1e-06, + "loss": 0.5642, + "step": 2577 + }, + { + "epoch": 0.9858508604206501, + "grad_norm": 3.4502829338930416, + "learning_rate": 1e-06, + "loss": 0.5265, + "step": 2578 + }, + { + "epoch": 0.9862332695984704, + "grad_norm": 3.274044007990694, + "learning_rate": 1e-06, + "loss": 0.318, + "step": 2579 + }, + { + "epoch": 0.9866156787762906, + "grad_norm": 2.9082146023740947, + "learning_rate": 1e-06, + "loss": 0.2627, + "step": 2580 + }, + { + "epoch": 0.9869980879541109, + "grad_norm": 3.030790156456422, + "learning_rate": 1e-06, + "loss": 0.3913, + "step": 2581 + }, + { + "epoch": 0.9873804971319312, + "grad_norm": 3.283557189694305, + "learning_rate": 1e-06, + "loss": 0.2163, + "step": 2582 + }, + { + "epoch": 0.9877629063097514, + "grad_norm": 3.117959445603992, + "learning_rate": 1e-06, + "loss": 0.2683, + "step": 2583 + }, + { + "epoch": 0.9881453154875717, + "grad_norm": 3.148377524436508, + "learning_rate": 1e-06, + "loss": 0.1586, + "step": 2584 + }, + { + "epoch": 0.988527724665392, + "grad_norm": 4.6813955612878, + "learning_rate": 1e-06, + "loss": 0.1509, + "step": 2585 + }, + { + "epoch": 0.9889101338432122, + "grad_norm": 3.899554809200781, + "learning_rate": 1e-06, + "loss": 0.1506, + "step": 2586 + }, + { + "epoch": 0.9892925430210325, + "grad_norm": 2.4800554066283054, + "learning_rate": 1e-06, + "loss": 0.0729, + "step": 2587 + }, + { + "epoch": 0.9896749521988528, + "grad_norm": 7.238309181847078, + "learning_rate": 1e-06, + "loss": 0.2698, + "step": 2588 + }, + { + "epoch": 0.990057361376673, + "grad_norm": 4.662581881297602, + "learning_rate": 1e-06, + "loss": 0.3281, + "step": 2589 + }, + { + "epoch": 0.9904397705544933, + "grad_norm": 2.739755187548956, + "learning_rate": 1e-06, + "loss": 0.5258, + "step": 2590 + }, + { + "epoch": 0.9908221797323136, + "grad_norm": 2.564721423929086, + "learning_rate": 1e-06, + "loss": 0.3786, + "step": 2591 + }, + { + "epoch": 0.9912045889101339, + "grad_norm": 2.5972957963319847, + "learning_rate": 1e-06, + "loss": 0.3551, + "step": 2592 + }, + { + "epoch": 0.9915869980879541, + "grad_norm": 3.8746655227526423, + "learning_rate": 1e-06, + "loss": 0.4603, + "step": 2593 + }, + { + "epoch": 0.9919694072657744, + "grad_norm": 2.229383749210781, + "learning_rate": 1e-06, + "loss": 0.2426, + "step": 2594 + }, + { + "epoch": 0.9923518164435946, + "grad_norm": 3.4840260147016933, + "learning_rate": 1e-06, + "loss": 0.3521, + "step": 2595 + }, + { + "epoch": 0.9927342256214149, + "grad_norm": 3.3125443365620715, + "learning_rate": 1e-06, + "loss": 0.3144, + "step": 2596 + }, + { + "epoch": 0.9931166347992352, + "grad_norm": 3.0452079711950035, + "learning_rate": 1e-06, + "loss": 0.1457, + "step": 2597 + }, + { + "epoch": 0.9934990439770555, + "grad_norm": 3.5410711049500967, + "learning_rate": 1e-06, + "loss": 0.2189, + "step": 2598 + }, + { + "epoch": 0.9938814531548757, + "grad_norm": 2.5759976944883465, + "learning_rate": 1e-06, + "loss": 0.063, + "step": 2599 + }, + { + "epoch": 0.994263862332696, + "grad_norm": 4.732681626241745, + "learning_rate": 1e-06, + "loss": 0.1459, + "step": 2600 + }, + { + "epoch": 0.9946462715105162, + "grad_norm": 5.7252164041855975, + "learning_rate": 1e-06, + "loss": 0.2892, + "step": 2601 + }, + { + "epoch": 0.9950286806883365, + "grad_norm": 3.9012106165781852, + "learning_rate": 1e-06, + "loss": 0.5871, + "step": 2602 + }, + { + "epoch": 0.9954110898661568, + "grad_norm": 4.077861675077326, + "learning_rate": 1e-06, + "loss": 0.3088, + "step": 2603 + }, + { + "epoch": 0.9957934990439771, + "grad_norm": 3.0754863447451473, + "learning_rate": 1e-06, + "loss": 0.3728, + "step": 2604 + }, + { + "epoch": 0.9961759082217974, + "grad_norm": 3.5602005347016674, + "learning_rate": 1e-06, + "loss": 0.3391, + "step": 2605 + }, + { + "epoch": 0.9965583173996176, + "grad_norm": 2.9985209633808587, + "learning_rate": 1e-06, + "loss": 0.3218, + "step": 2606 + }, + { + "epoch": 0.9969407265774378, + "grad_norm": 2.896954015417547, + "learning_rate": 1e-06, + "loss": 0.2512, + "step": 2607 + }, + { + "epoch": 0.9973231357552581, + "grad_norm": 2.8445699997682503, + "learning_rate": 1e-06, + "loss": 0.1779, + "step": 2608 + }, + { + "epoch": 0.9977055449330784, + "grad_norm": 3.448586359514333, + "learning_rate": 1e-06, + "loss": 0.2165, + "step": 2609 + }, + { + "epoch": 0.9980879541108987, + "grad_norm": 3.796851138444182, + "learning_rate": 1e-06, + "loss": 0.1629, + "step": 2610 + }, + { + "epoch": 0.998470363288719, + "grad_norm": 3.3308363781163437, + "learning_rate": 1e-06, + "loss": 0.1305, + "step": 2611 + }, + { + "epoch": 0.9988527724665393, + "grad_norm": 3.9866357709863354, + "learning_rate": 1e-06, + "loss": 0.1635, + "step": 2612 + }, + { + "epoch": 0.9992351816443594, + "grad_norm": 2.2663411028183136, + "learning_rate": 1e-06, + "loss": 0.2236, + "step": 2613 + }, + { + "epoch": 0.9996175908221797, + "grad_norm": 7.850867215017207, + "learning_rate": 1e-06, + "loss": 0.3879, + "step": 2614 + }, + { + "epoch": 1.0, + "grad_norm": 7.3600381928987115, + "learning_rate": 1e-06, + "loss": 0.2976, + "step": 2615 + }, + { + "epoch": 1.0003824091778202, + "grad_norm": 7.022242494277558, + "learning_rate": 1e-06, + "loss": 0.3882, + "step": 2616 + }, + { + "epoch": 1.0007648183556406, + "grad_norm": 3.2276398197207765, + "learning_rate": 1e-06, + "loss": 0.4019, + "step": 2617 + }, + { + "epoch": 1.0011472275334607, + "grad_norm": 4.711575339994919, + "learning_rate": 1e-06, + "loss": 0.384, + "step": 2618 + }, + { + "epoch": 1.0015296367112811, + "grad_norm": 3.964034996794371, + "learning_rate": 1e-06, + "loss": 0.3542, + "step": 2619 + }, + { + "epoch": 1.0019120458891013, + "grad_norm": 2.4735523305662164, + "learning_rate": 1e-06, + "loss": 0.367, + "step": 2620 + }, + { + "epoch": 1.0022944550669215, + "grad_norm": 2.984105921892883, + "learning_rate": 1e-06, + "loss": 0.2827, + "step": 2621 + }, + { + "epoch": 1.002676864244742, + "grad_norm": 2.690558711175746, + "learning_rate": 1e-06, + "loss": 0.2447, + "step": 2622 + }, + { + "epoch": 1.003059273422562, + "grad_norm": 2.195602777913192, + "learning_rate": 1e-06, + "loss": 0.2517, + "step": 2623 + }, + { + "epoch": 1.0034416826003825, + "grad_norm": 2.950277903769856, + "learning_rate": 1e-06, + "loss": 0.2829, + "step": 2624 + }, + { + "epoch": 1.0038240917782026, + "grad_norm": 2.2185522716692763, + "learning_rate": 1e-06, + "loss": 0.1924, + "step": 2625 + }, + { + "epoch": 1.004206500956023, + "grad_norm": 2.369061423893835, + "learning_rate": 1e-06, + "loss": 0.0956, + "step": 2626 + }, + { + "epoch": 1.0045889101338432, + "grad_norm": 3.1811822850303346, + "learning_rate": 1e-06, + "loss": 0.0704, + "step": 2627 + }, + { + "epoch": 1.0049713193116634, + "grad_norm": 4.695017368987337, + "learning_rate": 1e-06, + "loss": 0.2643, + "step": 2628 + }, + { + "epoch": 1.0053537284894838, + "grad_norm": 3.5123435749233964, + "learning_rate": 1e-06, + "loss": 0.3135, + "step": 2629 + }, + { + "epoch": 1.005736137667304, + "grad_norm": 4.023118682377531, + "learning_rate": 1e-06, + "loss": 0.4817, + "step": 2630 + }, + { + "epoch": 1.0061185468451244, + "grad_norm": 3.1225779493267387, + "learning_rate": 1e-06, + "loss": 0.5105, + "step": 2631 + }, + { + "epoch": 1.0065009560229445, + "grad_norm": 2.145657492040758, + "learning_rate": 1e-06, + "loss": 0.2007, + "step": 2632 + }, + { + "epoch": 1.0068833652007647, + "grad_norm": 3.395665794561428, + "learning_rate": 1e-06, + "loss": 0.2554, + "step": 2633 + }, + { + "epoch": 1.007265774378585, + "grad_norm": 2.738779237661642, + "learning_rate": 1e-06, + "loss": 0.2572, + "step": 2634 + }, + { + "epoch": 1.0076481835564053, + "grad_norm": 2.4865667883354314, + "learning_rate": 1e-06, + "loss": 0.3324, + "step": 2635 + }, + { + "epoch": 1.0080305927342257, + "grad_norm": 2.4270858328046465, + "learning_rate": 1e-06, + "loss": 0.199, + "step": 2636 + }, + { + "epoch": 1.0084130019120459, + "grad_norm": 2.83804453191572, + "learning_rate": 1e-06, + "loss": 0.2041, + "step": 2637 + }, + { + "epoch": 1.0087954110898663, + "grad_norm": 3.452598980500604, + "learning_rate": 1e-06, + "loss": 0.1231, + "step": 2638 + }, + { + "epoch": 1.0091778202676864, + "grad_norm": 3.1894734575285772, + "learning_rate": 1e-06, + "loss": 0.2265, + "step": 2639 + }, + { + "epoch": 1.0095602294455066, + "grad_norm": 4.684612147542248, + "learning_rate": 1e-06, + "loss": 0.1149, + "step": 2640 + }, + { + "epoch": 1.009942638623327, + "grad_norm": 6.372393505333487, + "learning_rate": 1e-06, + "loss": 0.3431, + "step": 2641 + }, + { + "epoch": 1.0103250478011472, + "grad_norm": 3.6672653807605475, + "learning_rate": 1e-06, + "loss": 0.4792, + "step": 2642 + }, + { + "epoch": 1.0107074569789676, + "grad_norm": 3.1671483192222225, + "learning_rate": 1e-06, + "loss": 0.1964, + "step": 2643 + }, + { + "epoch": 1.0110898661567878, + "grad_norm": 3.1009247169742444, + "learning_rate": 1e-06, + "loss": 0.3048, + "step": 2644 + }, + { + "epoch": 1.011472275334608, + "grad_norm": 2.561814565399431, + "learning_rate": 1e-06, + "loss": 0.4524, + "step": 2645 + }, + { + "epoch": 1.0118546845124283, + "grad_norm": 3.6103503738248652, + "learning_rate": 1e-06, + "loss": 0.4296, + "step": 2646 + }, + { + "epoch": 1.0122370936902485, + "grad_norm": 3.0457279491501494, + "learning_rate": 1e-06, + "loss": 0.2372, + "step": 2647 + }, + { + "epoch": 1.012619502868069, + "grad_norm": 2.6471071756222786, + "learning_rate": 1e-06, + "loss": 0.1504, + "step": 2648 + }, + { + "epoch": 1.013001912045889, + "grad_norm": 3.2796720070947787, + "learning_rate": 1e-06, + "loss": 0.3346, + "step": 2649 + }, + { + "epoch": 1.0133843212237095, + "grad_norm": 2.6879539993414476, + "learning_rate": 1e-06, + "loss": 0.1246, + "step": 2650 + }, + { + "epoch": 1.0137667304015296, + "grad_norm": 3.9806388541609574, + "learning_rate": 1e-06, + "loss": 0.2321, + "step": 2651 + }, + { + "epoch": 1.0141491395793498, + "grad_norm": 4.3342455612910635, + "learning_rate": 1e-06, + "loss": 0.1251, + "step": 2652 + }, + { + "epoch": 1.0145315487571702, + "grad_norm": 4.458886418246188, + "learning_rate": 1e-06, + "loss": 0.2768, + "step": 2653 + }, + { + "epoch": 1.0149139579349904, + "grad_norm": 3.829791768840152, + "learning_rate": 1e-06, + "loss": 0.7265, + "step": 2654 + }, + { + "epoch": 1.0152963671128108, + "grad_norm": 4.006958154705942, + "learning_rate": 1e-06, + "loss": 0.4493, + "step": 2655 + }, + { + "epoch": 1.015678776290631, + "grad_norm": 4.761598180256772, + "learning_rate": 1e-06, + "loss": 0.3316, + "step": 2656 + }, + { + "epoch": 1.0160611854684511, + "grad_norm": 2.4355635164452516, + "learning_rate": 1e-06, + "loss": 0.2961, + "step": 2657 + }, + { + "epoch": 1.0164435946462715, + "grad_norm": 3.4078103787408143, + "learning_rate": 1e-06, + "loss": 0.3313, + "step": 2658 + }, + { + "epoch": 1.0168260038240917, + "grad_norm": 2.336207345539462, + "learning_rate": 1e-06, + "loss": 0.266, + "step": 2659 + }, + { + "epoch": 1.0172084130019121, + "grad_norm": 2.483990525539346, + "learning_rate": 1e-06, + "loss": 0.2061, + "step": 2660 + }, + { + "epoch": 1.0175908221797323, + "grad_norm": 2.8088747183761535, + "learning_rate": 1e-06, + "loss": 0.2379, + "step": 2661 + }, + { + "epoch": 1.0179732313575527, + "grad_norm": 3.9472972749689803, + "learning_rate": 1e-06, + "loss": 0.1218, + "step": 2662 + }, + { + "epoch": 1.0183556405353729, + "grad_norm": 2.218365703129731, + "learning_rate": 1e-06, + "loss": 0.0802, + "step": 2663 + }, + { + "epoch": 1.018738049713193, + "grad_norm": 2.649953582645051, + "learning_rate": 1e-06, + "loss": 0.0778, + "step": 2664 + }, + { + "epoch": 1.0191204588910134, + "grad_norm": 4.448260732501134, + "learning_rate": 1e-06, + "loss": 0.1224, + "step": 2665 + }, + { + "epoch": 1.0195028680688336, + "grad_norm": 3.271824988098627, + "learning_rate": 1e-06, + "loss": 0.2896, + "step": 2666 + }, + { + "epoch": 1.019885277246654, + "grad_norm": 3.2242186671803053, + "learning_rate": 1e-06, + "loss": 0.3974, + "step": 2667 + }, + { + "epoch": 1.0202676864244742, + "grad_norm": 3.3364650478924704, + "learning_rate": 1e-06, + "loss": 0.3966, + "step": 2668 + }, + { + "epoch": 1.0206500956022944, + "grad_norm": 2.561851791645246, + "learning_rate": 1e-06, + "loss": 0.4702, + "step": 2669 + }, + { + "epoch": 1.0210325047801148, + "grad_norm": 2.296667180423467, + "learning_rate": 1e-06, + "loss": 0.1755, + "step": 2670 + }, + { + "epoch": 1.021414913957935, + "grad_norm": 2.5412507472527377, + "learning_rate": 1e-06, + "loss": 0.2024, + "step": 2671 + }, + { + "epoch": 1.0217973231357553, + "grad_norm": 3.1929982952396516, + "learning_rate": 1e-06, + "loss": 0.2564, + "step": 2672 + }, + { + "epoch": 1.0221797323135755, + "grad_norm": 3.12842570411532, + "learning_rate": 1e-06, + "loss": 0.1739, + "step": 2673 + }, + { + "epoch": 1.0225621414913957, + "grad_norm": 2.198754053160904, + "learning_rate": 1e-06, + "loss": 0.1205, + "step": 2674 + }, + { + "epoch": 1.022944550669216, + "grad_norm": 2.3468154188534847, + "learning_rate": 1e-06, + "loss": 0.099, + "step": 2675 + }, + { + "epoch": 1.0233269598470363, + "grad_norm": 2.9629685172276186, + "learning_rate": 1e-06, + "loss": 0.131, + "step": 2676 + }, + { + "epoch": 1.0237093690248567, + "grad_norm": 3.3928803005718886, + "learning_rate": 1e-06, + "loss": 0.0987, + "step": 2677 + }, + { + "epoch": 1.0240917782026768, + "grad_norm": 6.748226356603706, + "learning_rate": 1e-06, + "loss": 0.2138, + "step": 2678 + }, + { + "epoch": 1.0244741873804972, + "grad_norm": 4.177550609197598, + "learning_rate": 1e-06, + "loss": 0.4903, + "step": 2679 + }, + { + "epoch": 1.0248565965583174, + "grad_norm": 2.907307258180538, + "learning_rate": 1e-06, + "loss": 0.411, + "step": 2680 + }, + { + "epoch": 1.0252390057361376, + "grad_norm": 3.52547813581656, + "learning_rate": 1e-06, + "loss": 0.3489, + "step": 2681 + }, + { + "epoch": 1.025621414913958, + "grad_norm": 1.761864023922077, + "learning_rate": 1e-06, + "loss": 0.4432, + "step": 2682 + }, + { + "epoch": 1.0260038240917781, + "grad_norm": 4.062267120729046, + "learning_rate": 1e-06, + "loss": 0.5168, + "step": 2683 + }, + { + "epoch": 1.0263862332695985, + "grad_norm": 3.098152188290873, + "learning_rate": 1e-06, + "loss": 0.3486, + "step": 2684 + }, + { + "epoch": 1.0267686424474187, + "grad_norm": 2.888999358127268, + "learning_rate": 1e-06, + "loss": 0.3154, + "step": 2685 + }, + { + "epoch": 1.027151051625239, + "grad_norm": 2.1664383719064797, + "learning_rate": 1e-06, + "loss": 0.2393, + "step": 2686 + }, + { + "epoch": 1.0275334608030593, + "grad_norm": 2.937303577105565, + "learning_rate": 1e-06, + "loss": 0.1346, + "step": 2687 + }, + { + "epoch": 1.0279158699808795, + "grad_norm": 2.880227022759963, + "learning_rate": 1e-06, + "loss": 0.1251, + "step": 2688 + }, + { + "epoch": 1.0282982791586999, + "grad_norm": 5.532938898727842, + "learning_rate": 1e-06, + "loss": 0.1379, + "step": 2689 + }, + { + "epoch": 1.02868068833652, + "grad_norm": 4.385366963567743, + "learning_rate": 1e-06, + "loss": 0.1815, + "step": 2690 + }, + { + "epoch": 1.0290630975143404, + "grad_norm": 7.158234033869149, + "learning_rate": 1e-06, + "loss": 0.4647, + "step": 2691 + }, + { + "epoch": 1.0294455066921606, + "grad_norm": 3.0856505441641775, + "learning_rate": 1e-06, + "loss": 0.511, + "step": 2692 + }, + { + "epoch": 1.0298279158699808, + "grad_norm": 2.7459342552088613, + "learning_rate": 1e-06, + "loss": 0.396, + "step": 2693 + }, + { + "epoch": 1.0302103250478012, + "grad_norm": 2.322585408721161, + "learning_rate": 1e-06, + "loss": 0.2208, + "step": 2694 + }, + { + "epoch": 1.0305927342256214, + "grad_norm": 3.9542064770596155, + "learning_rate": 1e-06, + "loss": 0.2635, + "step": 2695 + }, + { + "epoch": 1.0309751434034418, + "grad_norm": 3.8928897501328565, + "learning_rate": 1e-06, + "loss": 0.3911, + "step": 2696 + }, + { + "epoch": 1.031357552581262, + "grad_norm": 2.718816822151605, + "learning_rate": 1e-06, + "loss": 0.2699, + "step": 2697 + }, + { + "epoch": 1.0317399617590821, + "grad_norm": 2.7116925070417888, + "learning_rate": 1e-06, + "loss": 0.1533, + "step": 2698 + }, + { + "epoch": 1.0321223709369025, + "grad_norm": 3.7386232098118484, + "learning_rate": 1e-06, + "loss": 0.2102, + "step": 2699 + }, + { + "epoch": 1.0325047801147227, + "grad_norm": 4.017373027397352, + "learning_rate": 1e-06, + "loss": 0.2086, + "step": 2700 + }, + { + "epoch": 1.032887189292543, + "grad_norm": 3.4537132072905083, + "learning_rate": 1e-06, + "loss": 0.0745, + "step": 2701 + }, + { + "epoch": 1.0332695984703633, + "grad_norm": 3.052307919169174, + "learning_rate": 1e-06, + "loss": 0.0833, + "step": 2702 + }, + { + "epoch": 1.0336520076481837, + "grad_norm": 5.773778758203473, + "learning_rate": 1e-06, + "loss": 0.2878, + "step": 2703 + }, + { + "epoch": 1.0340344168260038, + "grad_norm": 3.5658561558062556, + "learning_rate": 1e-06, + "loss": 0.3941, + "step": 2704 + }, + { + "epoch": 1.034416826003824, + "grad_norm": 3.8316777150803865, + "learning_rate": 1e-06, + "loss": 0.3779, + "step": 2705 + }, + { + "epoch": 1.0347992351816444, + "grad_norm": 3.0384828460897135, + "learning_rate": 1e-06, + "loss": 0.4879, + "step": 2706 + }, + { + "epoch": 1.0351816443594646, + "grad_norm": 2.343134786606256, + "learning_rate": 1e-06, + "loss": 0.176, + "step": 2707 + }, + { + "epoch": 1.035564053537285, + "grad_norm": 2.509989996354431, + "learning_rate": 1e-06, + "loss": 0.3084, + "step": 2708 + }, + { + "epoch": 1.0359464627151052, + "grad_norm": 4.5406688795706485, + "learning_rate": 1e-06, + "loss": 0.2898, + "step": 2709 + }, + { + "epoch": 1.0363288718929253, + "grad_norm": 2.519332617684988, + "learning_rate": 1e-06, + "loss": 0.3148, + "step": 2710 + }, + { + "epoch": 1.0367112810707457, + "grad_norm": 3.3118057963060408, + "learning_rate": 1e-06, + "loss": 0.1704, + "step": 2711 + }, + { + "epoch": 1.037093690248566, + "grad_norm": 3.1056092428420667, + "learning_rate": 1e-06, + "loss": 0.2319, + "step": 2712 + }, + { + "epoch": 1.0374760994263863, + "grad_norm": 2.8239605839709903, + "learning_rate": 1e-06, + "loss": 0.1413, + "step": 2713 + }, + { + "epoch": 1.0378585086042065, + "grad_norm": 3.6473681171255157, + "learning_rate": 1e-06, + "loss": 0.0879, + "step": 2714 + }, + { + "epoch": 1.0382409177820269, + "grad_norm": 4.901069773459302, + "learning_rate": 1e-06, + "loss": 0.1012, + "step": 2715 + }, + { + "epoch": 1.038623326959847, + "grad_norm": 5.90739502479764, + "learning_rate": 1e-06, + "loss": 0.3751, + "step": 2716 + }, + { + "epoch": 1.0390057361376672, + "grad_norm": 2.4297200780321946, + "learning_rate": 1e-06, + "loss": 0.3753, + "step": 2717 + }, + { + "epoch": 1.0393881453154876, + "grad_norm": 3.155282353796207, + "learning_rate": 1e-06, + "loss": 0.2652, + "step": 2718 + }, + { + "epoch": 1.0397705544933078, + "grad_norm": 1.8522253338930603, + "learning_rate": 1e-06, + "loss": 0.4514, + "step": 2719 + }, + { + "epoch": 1.0401529636711282, + "grad_norm": 3.8440830737692897, + "learning_rate": 1e-06, + "loss": 0.3758, + "step": 2720 + }, + { + "epoch": 1.0405353728489484, + "grad_norm": 3.6345675259188495, + "learning_rate": 1e-06, + "loss": 0.492, + "step": 2721 + }, + { + "epoch": 1.0409177820267685, + "grad_norm": 2.5366856174872265, + "learning_rate": 1e-06, + "loss": 0.3273, + "step": 2722 + }, + { + "epoch": 1.041300191204589, + "grad_norm": 2.9382334665978878, + "learning_rate": 1e-06, + "loss": 0.1969, + "step": 2723 + }, + { + "epoch": 1.0416826003824091, + "grad_norm": 2.463847447156355, + "learning_rate": 1e-06, + "loss": 0.1471, + "step": 2724 + }, + { + "epoch": 1.0420650095602295, + "grad_norm": 2.7102589926182095, + "learning_rate": 1e-06, + "loss": 0.0894, + "step": 2725 + }, + { + "epoch": 1.0424474187380497, + "grad_norm": 1.9479837117072258, + "learning_rate": 1e-06, + "loss": 0.0475, + "step": 2726 + }, + { + "epoch": 1.0428298279158699, + "grad_norm": 3.6407804660592453, + "learning_rate": 1e-06, + "loss": 0.0815, + "step": 2727 + }, + { + "epoch": 1.0432122370936903, + "grad_norm": 5.984271414634673, + "learning_rate": 1e-06, + "loss": 0.1927, + "step": 2728 + }, + { + "epoch": 1.0435946462715104, + "grad_norm": 3.93073151145546, + "learning_rate": 1e-06, + "loss": 0.5152, + "step": 2729 + }, + { + "epoch": 1.0439770554493308, + "grad_norm": 2.9375285695086757, + "learning_rate": 1e-06, + "loss": 0.434, + "step": 2730 + }, + { + "epoch": 1.044359464627151, + "grad_norm": 3.359699038468975, + "learning_rate": 1e-06, + "loss": 0.2864, + "step": 2731 + }, + { + "epoch": 1.0447418738049714, + "grad_norm": 3.754311181496516, + "learning_rate": 1e-06, + "loss": 0.2893, + "step": 2732 + }, + { + "epoch": 1.0451242829827916, + "grad_norm": 1.943514563723544, + "learning_rate": 1e-06, + "loss": 0.1743, + "step": 2733 + }, + { + "epoch": 1.0455066921606118, + "grad_norm": 3.242376188739394, + "learning_rate": 1e-06, + "loss": 0.2706, + "step": 2734 + }, + { + "epoch": 1.0458891013384322, + "grad_norm": 3.579687052426587, + "learning_rate": 1e-06, + "loss": 0.1136, + "step": 2735 + }, + { + "epoch": 1.0462715105162523, + "grad_norm": 2.2774925117913933, + "learning_rate": 1e-06, + "loss": 0.1565, + "step": 2736 + }, + { + "epoch": 1.0466539196940727, + "grad_norm": 2.907590495412597, + "learning_rate": 1e-06, + "loss": 0.1114, + "step": 2737 + }, + { + "epoch": 1.047036328871893, + "grad_norm": 3.855675039598945, + "learning_rate": 1e-06, + "loss": 0.1473, + "step": 2738 + }, + { + "epoch": 1.0474187380497133, + "grad_norm": 5.640318463309826, + "learning_rate": 1e-06, + "loss": 0.1222, + "step": 2739 + }, + { + "epoch": 1.0478011472275335, + "grad_norm": 3.7550896755030148, + "learning_rate": 1e-06, + "loss": 0.0826, + "step": 2740 + }, + { + "epoch": 1.0481835564053537, + "grad_norm": 8.14950996283073, + "learning_rate": 1e-06, + "loss": 0.4166, + "step": 2741 + }, + { + "epoch": 1.048565965583174, + "grad_norm": 3.7264255842412437, + "learning_rate": 1e-06, + "loss": 0.2951, + "step": 2742 + }, + { + "epoch": 1.0489483747609942, + "grad_norm": 3.1613406683671674, + "learning_rate": 1e-06, + "loss": 0.3623, + "step": 2743 + }, + { + "epoch": 1.0493307839388146, + "grad_norm": 2.42187224357202, + "learning_rate": 1e-06, + "loss": 0.5783, + "step": 2744 + }, + { + "epoch": 1.0497131931166348, + "grad_norm": 3.827291717144456, + "learning_rate": 1e-06, + "loss": 0.4027, + "step": 2745 + }, + { + "epoch": 1.050095602294455, + "grad_norm": 3.5942520454349443, + "learning_rate": 1e-06, + "loss": 0.2248, + "step": 2746 + }, + { + "epoch": 1.0504780114722754, + "grad_norm": 3.0847969361776753, + "learning_rate": 1e-06, + "loss": 0.2589, + "step": 2747 + }, + { + "epoch": 1.0508604206500956, + "grad_norm": 2.6449802755619127, + "learning_rate": 1e-06, + "loss": 0.2048, + "step": 2748 + }, + { + "epoch": 1.051242829827916, + "grad_norm": 3.955302607441138, + "learning_rate": 1e-06, + "loss": 0.2415, + "step": 2749 + }, + { + "epoch": 1.0516252390057361, + "grad_norm": 2.758995515643434, + "learning_rate": 1e-06, + "loss": 0.1459, + "step": 2750 + }, + { + "epoch": 1.0520076481835563, + "grad_norm": 4.547328441334444, + "learning_rate": 1e-06, + "loss": 0.2896, + "step": 2751 + }, + { + "epoch": 1.0523900573613767, + "grad_norm": 2.915663864678086, + "learning_rate": 1e-06, + "loss": 0.0652, + "step": 2752 + }, + { + "epoch": 1.0527724665391969, + "grad_norm": 5.872634736225061, + "learning_rate": 1e-06, + "loss": 0.1742, + "step": 2753 + }, + { + "epoch": 1.0531548757170173, + "grad_norm": 4.331522905743822, + "learning_rate": 1e-06, + "loss": 0.5289, + "step": 2754 + }, + { + "epoch": 1.0535372848948374, + "grad_norm": 3.6470304140580527, + "learning_rate": 1e-06, + "loss": 0.5093, + "step": 2755 + }, + { + "epoch": 1.0539196940726578, + "grad_norm": 2.853255597538155, + "learning_rate": 1e-06, + "loss": 0.3768, + "step": 2756 + }, + { + "epoch": 1.054302103250478, + "grad_norm": 2.253852407251978, + "learning_rate": 1e-06, + "loss": 0.3574, + "step": 2757 + }, + { + "epoch": 1.0546845124282982, + "grad_norm": 2.6423868475212187, + "learning_rate": 1e-06, + "loss": 0.2847, + "step": 2758 + }, + { + "epoch": 1.0550669216061186, + "grad_norm": 3.174076162182849, + "learning_rate": 1e-06, + "loss": 0.4799, + "step": 2759 + }, + { + "epoch": 1.0554493307839388, + "grad_norm": 2.041361831432893, + "learning_rate": 1e-06, + "loss": 0.1196, + "step": 2760 + }, + { + "epoch": 1.0558317399617592, + "grad_norm": 2.6674945460141313, + "learning_rate": 1e-06, + "loss": 0.155, + "step": 2761 + }, + { + "epoch": 1.0562141491395793, + "grad_norm": 2.9281701799976285, + "learning_rate": 1e-06, + "loss": 0.1407, + "step": 2762 + }, + { + "epoch": 1.0565965583173995, + "grad_norm": 3.137617198685308, + "learning_rate": 1e-06, + "loss": 0.1237, + "step": 2763 + }, + { + "epoch": 1.05697896749522, + "grad_norm": 2.9902279964007734, + "learning_rate": 1e-06, + "loss": 0.0829, + "step": 2764 + }, + { + "epoch": 1.05736137667304, + "grad_norm": 3.7043525817923886, + "learning_rate": 1e-06, + "loss": 0.1382, + "step": 2765 + }, + { + "epoch": 1.0577437858508605, + "grad_norm": 6.382018769497528, + "learning_rate": 1e-06, + "loss": 0.5412, + "step": 2766 + }, + { + "epoch": 1.0581261950286807, + "grad_norm": 3.774891701464836, + "learning_rate": 1e-06, + "loss": 0.4228, + "step": 2767 + }, + { + "epoch": 1.058508604206501, + "grad_norm": 3.79119988946345, + "learning_rate": 1e-06, + "loss": 0.4332, + "step": 2768 + }, + { + "epoch": 1.0588910133843212, + "grad_norm": 2.4406959904348207, + "learning_rate": 1e-06, + "loss": 0.2894, + "step": 2769 + }, + { + "epoch": 1.0592734225621414, + "grad_norm": 3.5431391329104103, + "learning_rate": 1e-06, + "loss": 0.3917, + "step": 2770 + }, + { + "epoch": 1.0596558317399618, + "grad_norm": 3.263546222708345, + "learning_rate": 1e-06, + "loss": 0.2839, + "step": 2771 + }, + { + "epoch": 1.060038240917782, + "grad_norm": 3.3798940989590522, + "learning_rate": 1e-06, + "loss": 0.2712, + "step": 2772 + }, + { + "epoch": 1.0604206500956024, + "grad_norm": 2.3786540782549745, + "learning_rate": 1e-06, + "loss": 0.1636, + "step": 2773 + }, + { + "epoch": 1.0608030592734226, + "grad_norm": 3.8724861604801575, + "learning_rate": 1e-06, + "loss": 0.268, + "step": 2774 + }, + { + "epoch": 1.0611854684512427, + "grad_norm": 3.4577965530462174, + "learning_rate": 1e-06, + "loss": 0.1355, + "step": 2775 + }, + { + "epoch": 1.0615678776290631, + "grad_norm": 2.6112711521458847, + "learning_rate": 1e-06, + "loss": 0.0885, + "step": 2776 + }, + { + "epoch": 1.0619502868068833, + "grad_norm": 3.3136666600623528, + "learning_rate": 1e-06, + "loss": 0.0643, + "step": 2777 + }, + { + "epoch": 1.0623326959847037, + "grad_norm": 4.084531699037168, + "learning_rate": 1e-06, + "loss": 0.1348, + "step": 2778 + }, + { + "epoch": 1.0627151051625239, + "grad_norm": 3.2009461970957367, + "learning_rate": 1e-06, + "loss": 0.2651, + "step": 2779 + }, + { + "epoch": 1.063097514340344, + "grad_norm": 3.2590256588801596, + "learning_rate": 1e-06, + "loss": 0.386, + "step": 2780 + }, + { + "epoch": 1.0634799235181644, + "grad_norm": 2.7223852999707043, + "learning_rate": 1e-06, + "loss": 0.463, + "step": 2781 + }, + { + "epoch": 1.0638623326959846, + "grad_norm": 2.9683073014908654, + "learning_rate": 1e-06, + "loss": 0.2778, + "step": 2782 + }, + { + "epoch": 1.064244741873805, + "grad_norm": 2.1593894770312243, + "learning_rate": 1e-06, + "loss": 0.2164, + "step": 2783 + }, + { + "epoch": 1.0646271510516252, + "grad_norm": 4.2599906179835205, + "learning_rate": 1e-06, + "loss": 0.3968, + "step": 2784 + }, + { + "epoch": 1.0650095602294456, + "grad_norm": 3.261442581637437, + "learning_rate": 1e-06, + "loss": 0.2604, + "step": 2785 + }, + { + "epoch": 1.0653919694072658, + "grad_norm": 2.456470419430958, + "learning_rate": 1e-06, + "loss": 0.2151, + "step": 2786 + }, + { + "epoch": 1.065774378585086, + "grad_norm": 5.335437041867739, + "learning_rate": 1e-06, + "loss": 0.2321, + "step": 2787 + }, + { + "epoch": 1.0661567877629063, + "grad_norm": 2.525007485860171, + "learning_rate": 1e-06, + "loss": 0.0643, + "step": 2788 + }, + { + "epoch": 1.0665391969407265, + "grad_norm": 3.073735239087343, + "learning_rate": 1e-06, + "loss": 0.0812, + "step": 2789 + }, + { + "epoch": 1.066921606118547, + "grad_norm": 3.9811808629522973, + "learning_rate": 1e-06, + "loss": 0.0931, + "step": 2790 + }, + { + "epoch": 1.067304015296367, + "grad_norm": 7.415904259517387, + "learning_rate": 1e-06, + "loss": 0.5791, + "step": 2791 + }, + { + "epoch": 1.0676864244741875, + "grad_norm": 3.223305006905413, + "learning_rate": 1e-06, + "loss": 0.3976, + "step": 2792 + }, + { + "epoch": 1.0680688336520077, + "grad_norm": 3.407348455696182, + "learning_rate": 1e-06, + "loss": 0.4962, + "step": 2793 + }, + { + "epoch": 1.0684512428298278, + "grad_norm": 2.289743784268278, + "learning_rate": 1e-06, + "loss": 0.1987, + "step": 2794 + }, + { + "epoch": 1.0688336520076482, + "grad_norm": 3.8234927430271717, + "learning_rate": 1e-06, + "loss": 0.3451, + "step": 2795 + }, + { + "epoch": 1.0692160611854684, + "grad_norm": 3.451374559652021, + "learning_rate": 1e-06, + "loss": 0.3375, + "step": 2796 + }, + { + "epoch": 1.0695984703632888, + "grad_norm": 1.776191354598207, + "learning_rate": 1e-06, + "loss": 0.3698, + "step": 2797 + }, + { + "epoch": 1.069980879541109, + "grad_norm": 2.7212399010039374, + "learning_rate": 1e-06, + "loss": 0.2403, + "step": 2798 + }, + { + "epoch": 1.0703632887189292, + "grad_norm": 3.5523219023805153, + "learning_rate": 1e-06, + "loss": 0.1581, + "step": 2799 + }, + { + "epoch": 1.0707456978967496, + "grad_norm": 3.3041342414204586, + "learning_rate": 1e-06, + "loss": 0.1054, + "step": 2800 + }, + { + "epoch": 1.0711281070745697, + "grad_norm": 3.322390778800794, + "learning_rate": 1e-06, + "loss": 0.1228, + "step": 2801 + }, + { + "epoch": 1.0715105162523901, + "grad_norm": 3.685726709228734, + "learning_rate": 1e-06, + "loss": 0.0915, + "step": 2802 + }, + { + "epoch": 1.0718929254302103, + "grad_norm": 3.7334297292616827, + "learning_rate": 1e-06, + "loss": 0.2692, + "step": 2803 + }, + { + "epoch": 1.0722753346080305, + "grad_norm": 4.571780863879412, + "learning_rate": 1e-06, + "loss": 0.4868, + "step": 2804 + }, + { + "epoch": 1.0726577437858509, + "grad_norm": 3.4253977579413846, + "learning_rate": 1e-06, + "loss": 0.3587, + "step": 2805 + }, + { + "epoch": 1.073040152963671, + "grad_norm": 5.506511994622715, + "learning_rate": 1e-06, + "loss": 0.322, + "step": 2806 + }, + { + "epoch": 1.0734225621414915, + "grad_norm": 1.916366747181914, + "learning_rate": 1e-06, + "loss": 0.2798, + "step": 2807 + }, + { + "epoch": 1.0738049713193116, + "grad_norm": 2.77367810562294, + "learning_rate": 1e-06, + "loss": 0.3029, + "step": 2808 + }, + { + "epoch": 1.074187380497132, + "grad_norm": 1.845260841622034, + "learning_rate": 1e-06, + "loss": 0.2007, + "step": 2809 + }, + { + "epoch": 1.0745697896749522, + "grad_norm": 2.767604485859373, + "learning_rate": 1e-06, + "loss": 0.1358, + "step": 2810 + }, + { + "epoch": 1.0749521988527724, + "grad_norm": 3.3942540871007716, + "learning_rate": 1e-06, + "loss": 0.1765, + "step": 2811 + }, + { + "epoch": 1.0753346080305928, + "grad_norm": 4.302997901385853, + "learning_rate": 1e-06, + "loss": 0.4149, + "step": 2812 + }, + { + "epoch": 1.075717017208413, + "grad_norm": 6.473597382733225, + "learning_rate": 1e-06, + "loss": 0.2666, + "step": 2813 + }, + { + "epoch": 1.0760994263862333, + "grad_norm": 2.8171827539455827, + "learning_rate": 1e-06, + "loss": 0.0674, + "step": 2814 + }, + { + "epoch": 1.0764818355640535, + "grad_norm": 5.56599882006727, + "learning_rate": 1e-06, + "loss": 0.1671, + "step": 2815 + }, + { + "epoch": 1.076864244741874, + "grad_norm": 7.30450915766582, + "learning_rate": 1e-06, + "loss": 0.3735, + "step": 2816 + }, + { + "epoch": 1.077246653919694, + "grad_norm": 3.423064560731871, + "learning_rate": 1e-06, + "loss": 0.4754, + "step": 2817 + }, + { + "epoch": 1.0776290630975143, + "grad_norm": 3.2174641531464228, + "learning_rate": 1e-06, + "loss": 0.4532, + "step": 2818 + }, + { + "epoch": 1.0780114722753347, + "grad_norm": 3.009919139751804, + "learning_rate": 1e-06, + "loss": 0.2953, + "step": 2819 + }, + { + "epoch": 1.0783938814531548, + "grad_norm": 3.5596707804751624, + "learning_rate": 1e-06, + "loss": 0.4202, + "step": 2820 + }, + { + "epoch": 1.0787762906309752, + "grad_norm": 3.4648525577678355, + "learning_rate": 1e-06, + "loss": 0.2884, + "step": 2821 + }, + { + "epoch": 1.0791586998087954, + "grad_norm": 2.6964693544564353, + "learning_rate": 1e-06, + "loss": 0.2034, + "step": 2822 + }, + { + "epoch": 1.0795411089866156, + "grad_norm": 2.2880146635826097, + "learning_rate": 1e-06, + "loss": 0.1792, + "step": 2823 + }, + { + "epoch": 1.079923518164436, + "grad_norm": 3.084925232016781, + "learning_rate": 1e-06, + "loss": 0.1477, + "step": 2824 + }, + { + "epoch": 1.0803059273422562, + "grad_norm": 3.3123235115682896, + "learning_rate": 1e-06, + "loss": 0.1193, + "step": 2825 + }, + { + "epoch": 1.0806883365200766, + "grad_norm": 3.385762890449034, + "learning_rate": 1e-06, + "loss": 0.0661, + "step": 2826 + }, + { + "epoch": 1.0810707456978967, + "grad_norm": 3.7251373060893673, + "learning_rate": 1e-06, + "loss": 0.1037, + "step": 2827 + }, + { + "epoch": 1.081453154875717, + "grad_norm": 5.675358482835371, + "learning_rate": 1e-06, + "loss": 0.2238, + "step": 2828 + }, + { + "epoch": 1.0818355640535373, + "grad_norm": 5.755246298364127, + "learning_rate": 1e-06, + "loss": 0.3204, + "step": 2829 + }, + { + "epoch": 1.0822179732313575, + "grad_norm": 4.0569034467401215, + "learning_rate": 1e-06, + "loss": 0.5256, + "step": 2830 + }, + { + "epoch": 1.0826003824091779, + "grad_norm": 2.688257775928962, + "learning_rate": 1e-06, + "loss": 0.2783, + "step": 2831 + }, + { + "epoch": 1.082982791586998, + "grad_norm": 2.166301794384953, + "learning_rate": 1e-06, + "loss": 0.2526, + "step": 2832 + }, + { + "epoch": 1.0833652007648185, + "grad_norm": 3.2912193348488654, + "learning_rate": 1e-06, + "loss": 0.6082, + "step": 2833 + }, + { + "epoch": 1.0837476099426386, + "grad_norm": 3.5575385094151377, + "learning_rate": 1e-06, + "loss": 0.4073, + "step": 2834 + }, + { + "epoch": 1.0841300191204588, + "grad_norm": 4.088115979365335, + "learning_rate": 1e-06, + "loss": 0.1996, + "step": 2835 + }, + { + "epoch": 1.0845124282982792, + "grad_norm": 2.1587060402799323, + "learning_rate": 1e-06, + "loss": 0.0898, + "step": 2836 + }, + { + "epoch": 1.0848948374760994, + "grad_norm": 3.0045968441273025, + "learning_rate": 1e-06, + "loss": 0.1477, + "step": 2837 + }, + { + "epoch": 1.0852772466539198, + "grad_norm": 4.259853608745767, + "learning_rate": 1e-06, + "loss": 0.1283, + "step": 2838 + }, + { + "epoch": 1.08565965583174, + "grad_norm": 4.144053756144276, + "learning_rate": 1e-06, + "loss": 0.1443, + "step": 2839 + }, + { + "epoch": 1.0860420650095601, + "grad_norm": 4.5304088995766465, + "learning_rate": 1e-06, + "loss": 0.1254, + "step": 2840 + }, + { + "epoch": 1.0864244741873805, + "grad_norm": 7.951371934660744, + "learning_rate": 1e-06, + "loss": 0.2675, + "step": 2841 + }, + { + "epoch": 1.0868068833652007, + "grad_norm": 4.102531275098238, + "learning_rate": 1e-06, + "loss": 0.4678, + "step": 2842 + }, + { + "epoch": 1.087189292543021, + "grad_norm": 2.8548635110113243, + "learning_rate": 1e-06, + "loss": 0.2682, + "step": 2843 + }, + { + "epoch": 1.0875717017208413, + "grad_norm": 2.550358174918483, + "learning_rate": 1e-06, + "loss": 0.1595, + "step": 2844 + }, + { + "epoch": 1.0879541108986617, + "grad_norm": 2.885750284199724, + "learning_rate": 1e-06, + "loss": 0.4214, + "step": 2845 + }, + { + "epoch": 1.0883365200764819, + "grad_norm": 2.413077045271455, + "learning_rate": 1e-06, + "loss": 0.3292, + "step": 2846 + }, + { + "epoch": 1.088718929254302, + "grad_norm": 2.977832913560461, + "learning_rate": 1e-06, + "loss": 0.2312, + "step": 2847 + }, + { + "epoch": 1.0891013384321224, + "grad_norm": 3.0670257137030372, + "learning_rate": 1e-06, + "loss": 0.2079, + "step": 2848 + }, + { + "epoch": 1.0894837476099426, + "grad_norm": 3.567873866935165, + "learning_rate": 1e-06, + "loss": 0.3071, + "step": 2849 + }, + { + "epoch": 1.089866156787763, + "grad_norm": 3.2884373014022796, + "learning_rate": 1e-06, + "loss": 0.216, + "step": 2850 + }, + { + "epoch": 1.0902485659655832, + "grad_norm": 3.167269515493883, + "learning_rate": 1e-06, + "loss": 0.0664, + "step": 2851 + }, + { + "epoch": 1.0906309751434033, + "grad_norm": 4.1067322330277465, + "learning_rate": 1e-06, + "loss": 0.1224, + "step": 2852 + }, + { + "epoch": 1.0910133843212237, + "grad_norm": 4.817414585815298, + "learning_rate": 1e-06, + "loss": 0.4915, + "step": 2853 + }, + { + "epoch": 1.091395793499044, + "grad_norm": 3.696390587484518, + "learning_rate": 1e-06, + "loss": 0.4389, + "step": 2854 + }, + { + "epoch": 1.0917782026768643, + "grad_norm": 3.057344105848634, + "learning_rate": 1e-06, + "loss": 0.4314, + "step": 2855 + }, + { + "epoch": 1.0921606118546845, + "grad_norm": 3.1848468584405674, + "learning_rate": 1e-06, + "loss": 0.3289, + "step": 2856 + }, + { + "epoch": 1.0925430210325047, + "grad_norm": 3.114695688387905, + "learning_rate": 1e-06, + "loss": 0.3441, + "step": 2857 + }, + { + "epoch": 1.092925430210325, + "grad_norm": 6.4186641341931985, + "learning_rate": 1e-06, + "loss": 0.4204, + "step": 2858 + }, + { + "epoch": 1.0933078393881452, + "grad_norm": 3.436542585730967, + "learning_rate": 1e-06, + "loss": 0.1804, + "step": 2859 + }, + { + "epoch": 1.0936902485659656, + "grad_norm": 2.5285077723571674, + "learning_rate": 1e-06, + "loss": 0.2002, + "step": 2860 + }, + { + "epoch": 1.0940726577437858, + "grad_norm": 2.581862820716856, + "learning_rate": 1e-06, + "loss": 0.1119, + "step": 2861 + }, + { + "epoch": 1.0944550669216062, + "grad_norm": 3.2284916982439156, + "learning_rate": 1e-06, + "loss": 0.1461, + "step": 2862 + }, + { + "epoch": 1.0948374760994264, + "grad_norm": 3.2432258138317156, + "learning_rate": 1e-06, + "loss": 0.1405, + "step": 2863 + }, + { + "epoch": 1.0952198852772466, + "grad_norm": 3.056057595903582, + "learning_rate": 1e-06, + "loss": 0.0567, + "step": 2864 + }, + { + "epoch": 1.095602294455067, + "grad_norm": 3.406085929068903, + "learning_rate": 1e-06, + "loss": 0.0868, + "step": 2865 + }, + { + "epoch": 1.0959847036328871, + "grad_norm": 5.2675466228896735, + "learning_rate": 1e-06, + "loss": 0.4209, + "step": 2866 + }, + { + "epoch": 1.0963671128107075, + "grad_norm": 3.495768577754933, + "learning_rate": 1e-06, + "loss": 0.5578, + "step": 2867 + }, + { + "epoch": 1.0967495219885277, + "grad_norm": 2.3411254429309367, + "learning_rate": 1e-06, + "loss": 0.3557, + "step": 2868 + }, + { + "epoch": 1.097131931166348, + "grad_norm": 3.0691303957486715, + "learning_rate": 1e-06, + "loss": 0.3543, + "step": 2869 + }, + { + "epoch": 1.0975143403441683, + "grad_norm": 3.065784639174185, + "learning_rate": 1e-06, + "loss": 0.3856, + "step": 2870 + }, + { + "epoch": 1.0978967495219885, + "grad_norm": 2.994955908652979, + "learning_rate": 1e-06, + "loss": 0.2518, + "step": 2871 + }, + { + "epoch": 1.0982791586998089, + "grad_norm": 4.563174472232909, + "learning_rate": 1e-06, + "loss": 0.1964, + "step": 2872 + }, + { + "epoch": 1.098661567877629, + "grad_norm": 3.4571755244435485, + "learning_rate": 1e-06, + "loss": 0.2262, + "step": 2873 + }, + { + "epoch": 1.0990439770554494, + "grad_norm": 2.5488111005777228, + "learning_rate": 1e-06, + "loss": 0.1324, + "step": 2874 + }, + { + "epoch": 1.0994263862332696, + "grad_norm": 3.945085394337583, + "learning_rate": 1e-06, + "loss": 0.1595, + "step": 2875 + }, + { + "epoch": 1.0998087954110898, + "grad_norm": 2.837069030836222, + "learning_rate": 1e-06, + "loss": 0.099, + "step": 2876 + }, + { + "epoch": 1.1001912045889102, + "grad_norm": 3.908802632757324, + "learning_rate": 1e-06, + "loss": 0.0899, + "step": 2877 + }, + { + "epoch": 1.1005736137667304, + "grad_norm": 5.114341818188961, + "learning_rate": 1e-06, + "loss": 0.3208, + "step": 2878 + }, + { + "epoch": 1.1009560229445507, + "grad_norm": 3.858759904409437, + "learning_rate": 1e-06, + "loss": 0.4088, + "step": 2879 + }, + { + "epoch": 1.101338432122371, + "grad_norm": 2.6757349163415456, + "learning_rate": 1e-06, + "loss": 0.2739, + "step": 2880 + }, + { + "epoch": 1.101720841300191, + "grad_norm": 2.6258524464265034, + "learning_rate": 1e-06, + "loss": 0.2561, + "step": 2881 + }, + { + "epoch": 1.1021032504780115, + "grad_norm": 4.256005700000015, + "learning_rate": 1e-06, + "loss": 0.2905, + "step": 2882 + }, + { + "epoch": 1.1024856596558317, + "grad_norm": 2.93229359998803, + "learning_rate": 1e-06, + "loss": 0.2834, + "step": 2883 + }, + { + "epoch": 1.102868068833652, + "grad_norm": 3.475864434613361, + "learning_rate": 1e-06, + "loss": 0.2758, + "step": 2884 + }, + { + "epoch": 1.1032504780114722, + "grad_norm": 2.295733836580188, + "learning_rate": 1e-06, + "loss": 0.1278, + "step": 2885 + }, + { + "epoch": 1.1036328871892926, + "grad_norm": 3.369627880300253, + "learning_rate": 1e-06, + "loss": 0.46, + "step": 2886 + }, + { + "epoch": 1.1040152963671128, + "grad_norm": 2.5749161604778497, + "learning_rate": 1e-06, + "loss": 0.0795, + "step": 2887 + }, + { + "epoch": 1.104397705544933, + "grad_norm": 5.368106058246976, + "learning_rate": 1e-06, + "loss": 0.3168, + "step": 2888 + }, + { + "epoch": 1.1047801147227534, + "grad_norm": 3.4822033328439526, + "learning_rate": 1e-06, + "loss": 0.0966, + "step": 2889 + }, + { + "epoch": 1.1051625239005736, + "grad_norm": 5.389917476496499, + "learning_rate": 1e-06, + "loss": 0.1366, + "step": 2890 + }, + { + "epoch": 1.105544933078394, + "grad_norm": 7.55366188827401, + "learning_rate": 1e-06, + "loss": 0.3474, + "step": 2891 + }, + { + "epoch": 1.1059273422562141, + "grad_norm": 4.755177135998886, + "learning_rate": 1e-06, + "loss": 0.627, + "step": 2892 + }, + { + "epoch": 1.1063097514340343, + "grad_norm": 2.0768261055438555, + "learning_rate": 1e-06, + "loss": 0.3692, + "step": 2893 + }, + { + "epoch": 1.1066921606118547, + "grad_norm": 3.225032268037455, + "learning_rate": 1e-06, + "loss": 0.3745, + "step": 2894 + }, + { + "epoch": 1.107074569789675, + "grad_norm": 3.9894524030723058, + "learning_rate": 1e-06, + "loss": 0.4802, + "step": 2895 + }, + { + "epoch": 1.1074569789674953, + "grad_norm": 3.2782512222295908, + "learning_rate": 1e-06, + "loss": 0.2156, + "step": 2896 + }, + { + "epoch": 1.1078393881453155, + "grad_norm": 3.825465185264008, + "learning_rate": 1e-06, + "loss": 0.2157, + "step": 2897 + }, + { + "epoch": 1.1082217973231359, + "grad_norm": 3.158481526651568, + "learning_rate": 1e-06, + "loss": 0.3175, + "step": 2898 + }, + { + "epoch": 1.108604206500956, + "grad_norm": 2.5433234979615165, + "learning_rate": 1e-06, + "loss": 0.0968, + "step": 2899 + }, + { + "epoch": 1.1089866156787762, + "grad_norm": 2.9339906548453296, + "learning_rate": 1e-06, + "loss": 0.0736, + "step": 2900 + }, + { + "epoch": 1.1093690248565966, + "grad_norm": 3.7096563115261763, + "learning_rate": 1e-06, + "loss": 0.1069, + "step": 2901 + }, + { + "epoch": 1.1097514340344168, + "grad_norm": 4.340461638900813, + "learning_rate": 1e-06, + "loss": 0.0848, + "step": 2902 + }, + { + "epoch": 1.1101338432122372, + "grad_norm": 5.716557984204094, + "learning_rate": 1e-06, + "loss": 0.2638, + "step": 2903 + }, + { + "epoch": 1.1105162523900574, + "grad_norm": 4.055877922402935, + "learning_rate": 1e-06, + "loss": 0.6603, + "step": 2904 + }, + { + "epoch": 1.1108986615678775, + "grad_norm": 2.792949030879908, + "learning_rate": 1e-06, + "loss": 0.3407, + "step": 2905 + }, + { + "epoch": 1.111281070745698, + "grad_norm": 3.4354327835127525, + "learning_rate": 1e-06, + "loss": 0.3378, + "step": 2906 + }, + { + "epoch": 1.111663479923518, + "grad_norm": 3.996718968379817, + "learning_rate": 1e-06, + "loss": 0.3919, + "step": 2907 + }, + { + "epoch": 1.1120458891013385, + "grad_norm": 2.982178363963938, + "learning_rate": 1e-06, + "loss": 0.3849, + "step": 2908 + }, + { + "epoch": 1.1124282982791587, + "grad_norm": 3.227012030296255, + "learning_rate": 1e-06, + "loss": 0.1708, + "step": 2909 + }, + { + "epoch": 1.1128107074569789, + "grad_norm": 2.6302030859161505, + "learning_rate": 1e-06, + "loss": 0.31, + "step": 2910 + }, + { + "epoch": 1.1131931166347993, + "grad_norm": 3.635419144215145, + "learning_rate": 1e-06, + "loss": 0.2621, + "step": 2911 + }, + { + "epoch": 1.1135755258126194, + "grad_norm": 2.9696596057171476, + "learning_rate": 1e-06, + "loss": 0.1988, + "step": 2912 + }, + { + "epoch": 1.1139579349904398, + "grad_norm": 4.434582248008877, + "learning_rate": 1e-06, + "loss": 0.1347, + "step": 2913 + }, + { + "epoch": 1.11434034416826, + "grad_norm": 3.670918181727296, + "learning_rate": 1e-06, + "loss": 0.082, + "step": 2914 + }, + { + "epoch": 1.1147227533460804, + "grad_norm": 4.325960685449046, + "learning_rate": 1e-06, + "loss": 0.1222, + "step": 2915 + }, + { + "epoch": 1.1151051625239006, + "grad_norm": 7.413527377948621, + "learning_rate": 1e-06, + "loss": 0.2725, + "step": 2916 + }, + { + "epoch": 1.1154875717017207, + "grad_norm": 4.250834214859102, + "learning_rate": 1e-06, + "loss": 0.784, + "step": 2917 + }, + { + "epoch": 1.1158699808795411, + "grad_norm": 3.401164607133041, + "learning_rate": 1e-06, + "loss": 0.4209, + "step": 2918 + }, + { + "epoch": 1.1162523900573613, + "grad_norm": 2.6646607921358387, + "learning_rate": 1e-06, + "loss": 0.4007, + "step": 2919 + }, + { + "epoch": 1.1166347992351817, + "grad_norm": 2.868380015331566, + "learning_rate": 1e-06, + "loss": 0.4022, + "step": 2920 + }, + { + "epoch": 1.117017208413002, + "grad_norm": 2.8090205392219025, + "learning_rate": 1e-06, + "loss": 0.2514, + "step": 2921 + }, + { + "epoch": 1.1173996175908223, + "grad_norm": 3.8349616835369784, + "learning_rate": 1e-06, + "loss": 0.3427, + "step": 2922 + }, + { + "epoch": 1.1177820267686425, + "grad_norm": 2.74083900442568, + "learning_rate": 1e-06, + "loss": 0.2086, + "step": 2923 + }, + { + "epoch": 1.1181644359464626, + "grad_norm": 2.8517920832772288, + "learning_rate": 1e-06, + "loss": 0.2163, + "step": 2924 + }, + { + "epoch": 1.118546845124283, + "grad_norm": 3.2410084982162966, + "learning_rate": 1e-06, + "loss": 0.2256, + "step": 2925 + }, + { + "epoch": 1.1189292543021032, + "grad_norm": 2.199672401485876, + "learning_rate": 1e-06, + "loss": 0.0648, + "step": 2926 + }, + { + "epoch": 1.1193116634799236, + "grad_norm": 2.89625339525968, + "learning_rate": 1e-06, + "loss": 0.0964, + "step": 2927 + }, + { + "epoch": 1.1196940726577438, + "grad_norm": 3.675327187849286, + "learning_rate": 1e-06, + "loss": 0.1628, + "step": 2928 + }, + { + "epoch": 1.120076481835564, + "grad_norm": 5.544467801282905, + "learning_rate": 1e-06, + "loss": 0.454, + "step": 2929 + }, + { + "epoch": 1.1204588910133844, + "grad_norm": 4.969388332989557, + "learning_rate": 1e-06, + "loss": 0.486, + "step": 2930 + }, + { + "epoch": 1.1208413001912045, + "grad_norm": 2.787810485380719, + "learning_rate": 1e-06, + "loss": 0.1456, + "step": 2931 + }, + { + "epoch": 1.121223709369025, + "grad_norm": 3.13911318094859, + "learning_rate": 1e-06, + "loss": 0.2817, + "step": 2932 + }, + { + "epoch": 1.121606118546845, + "grad_norm": 3.0418354318560086, + "learning_rate": 1e-06, + "loss": 0.2556, + "step": 2933 + }, + { + "epoch": 1.1219885277246653, + "grad_norm": 3.7041959220037035, + "learning_rate": 1e-06, + "loss": 0.2334, + "step": 2934 + }, + { + "epoch": 1.1223709369024857, + "grad_norm": 2.518353611350916, + "learning_rate": 1e-06, + "loss": 0.1713, + "step": 2935 + }, + { + "epoch": 1.1227533460803059, + "grad_norm": 2.864493702728919, + "learning_rate": 1e-06, + "loss": 0.2985, + "step": 2936 + }, + { + "epoch": 1.1231357552581263, + "grad_norm": 3.3934064434796807, + "learning_rate": 1e-06, + "loss": 0.0911, + "step": 2937 + }, + { + "epoch": 1.1235181644359464, + "grad_norm": 3.192170853461705, + "learning_rate": 1e-06, + "loss": 0.079, + "step": 2938 + }, + { + "epoch": 1.1239005736137668, + "grad_norm": 3.362388501215389, + "learning_rate": 1e-06, + "loss": 0.076, + "step": 2939 + }, + { + "epoch": 1.124282982791587, + "grad_norm": 5.310276957615031, + "learning_rate": 1e-06, + "loss": 0.1389, + "step": 2940 + }, + { + "epoch": 1.1246653919694072, + "grad_norm": 6.827063995262543, + "learning_rate": 1e-06, + "loss": 0.3134, + "step": 2941 + }, + { + "epoch": 1.1250478011472276, + "grad_norm": 3.67451674209873, + "learning_rate": 1e-06, + "loss": 0.4647, + "step": 2942 + }, + { + "epoch": 1.1254302103250478, + "grad_norm": 2.333746862097798, + "learning_rate": 1e-06, + "loss": 0.2564, + "step": 2943 + }, + { + "epoch": 1.1258126195028682, + "grad_norm": 3.0286945700995, + "learning_rate": 1e-06, + "loss": 0.5127, + "step": 2944 + }, + { + "epoch": 1.1261950286806883, + "grad_norm": 3.895260052991601, + "learning_rate": 1e-06, + "loss": 0.3841, + "step": 2945 + }, + { + "epoch": 1.1265774378585087, + "grad_norm": 4.60890159115522, + "learning_rate": 1e-06, + "loss": 0.2758, + "step": 2946 + }, + { + "epoch": 1.126959847036329, + "grad_norm": 2.299857558110427, + "learning_rate": 1e-06, + "loss": 0.1026, + "step": 2947 + }, + { + "epoch": 1.127342256214149, + "grad_norm": 2.301337752857731, + "learning_rate": 1e-06, + "loss": 0.0975, + "step": 2948 + }, + { + "epoch": 1.1277246653919695, + "grad_norm": 2.434746580938376, + "learning_rate": 1e-06, + "loss": 0.1672, + "step": 2949 + }, + { + "epoch": 1.1281070745697896, + "grad_norm": 3.3248961640400063, + "learning_rate": 1e-06, + "loss": 0.2036, + "step": 2950 + }, + { + "epoch": 1.12848948374761, + "grad_norm": 2.399819836212055, + "learning_rate": 1e-06, + "loss": 0.0503, + "step": 2951 + }, + { + "epoch": 1.1288718929254302, + "grad_norm": 8.43861980777982, + "learning_rate": 1e-06, + "loss": 0.1625, + "step": 2952 + }, + { + "epoch": 1.1292543021032504, + "grad_norm": 5.674700745073927, + "learning_rate": 1e-06, + "loss": 0.1837, + "step": 2953 + }, + { + "epoch": 1.1296367112810708, + "grad_norm": 2.8559599948293095, + "learning_rate": 1e-06, + "loss": 0.4173, + "step": 2954 + }, + { + "epoch": 1.130019120458891, + "grad_norm": 3.1598565675077217, + "learning_rate": 1e-06, + "loss": 0.3471, + "step": 2955 + }, + { + "epoch": 1.1304015296367114, + "grad_norm": 3.5022016140314807, + "learning_rate": 1e-06, + "loss": 0.2841, + "step": 2956 + }, + { + "epoch": 1.1307839388145315, + "grad_norm": 3.0422650798041744, + "learning_rate": 1e-06, + "loss": 0.3529, + "step": 2957 + }, + { + "epoch": 1.1311663479923517, + "grad_norm": 3.743453096516027, + "learning_rate": 1e-06, + "loss": 0.3814, + "step": 2958 + }, + { + "epoch": 1.1315487571701721, + "grad_norm": 3.170667446641155, + "learning_rate": 1e-06, + "loss": 0.3059, + "step": 2959 + }, + { + "epoch": 1.1319311663479923, + "grad_norm": 3.8093364910779806, + "learning_rate": 1e-06, + "loss": 0.1791, + "step": 2960 + }, + { + "epoch": 1.1323135755258127, + "grad_norm": 2.7752219034086587, + "learning_rate": 1e-06, + "loss": 0.2031, + "step": 2961 + }, + { + "epoch": 1.1326959847036329, + "grad_norm": 4.268290040018679, + "learning_rate": 1e-06, + "loss": 0.1393, + "step": 2962 + }, + { + "epoch": 1.133078393881453, + "grad_norm": 5.73511469173405, + "learning_rate": 1e-06, + "loss": 0.1168, + "step": 2963 + }, + { + "epoch": 1.1334608030592734, + "grad_norm": 5.463547853576047, + "learning_rate": 1e-06, + "loss": 0.1538, + "step": 2964 + }, + { + "epoch": 1.1338432122370936, + "grad_norm": 5.990980999178643, + "learning_rate": 1e-06, + "loss": 0.1144, + "step": 2965 + }, + { + "epoch": 1.134225621414914, + "grad_norm": 4.907622127247335, + "learning_rate": 1e-06, + "loss": 0.2499, + "step": 2966 + }, + { + "epoch": 1.1346080305927342, + "grad_norm": 2.6854682161775996, + "learning_rate": 1e-06, + "loss": 0.5883, + "step": 2967 + }, + { + "epoch": 1.1349904397705546, + "grad_norm": 3.890548644025143, + "learning_rate": 1e-06, + "loss": 0.4597, + "step": 2968 + }, + { + "epoch": 1.1353728489483748, + "grad_norm": 1.8231637260500944, + "learning_rate": 1e-06, + "loss": 0.3064, + "step": 2969 + }, + { + "epoch": 1.135755258126195, + "grad_norm": 3.5759270593086896, + "learning_rate": 1e-06, + "loss": 0.2339, + "step": 2970 + }, + { + "epoch": 1.1361376673040153, + "grad_norm": 3.260923807273525, + "learning_rate": 1e-06, + "loss": 0.3713, + "step": 2971 + }, + { + "epoch": 1.1365200764818355, + "grad_norm": 2.9478360402075094, + "learning_rate": 1e-06, + "loss": 0.1954, + "step": 2972 + }, + { + "epoch": 1.136902485659656, + "grad_norm": 2.9678518442464363, + "learning_rate": 1e-06, + "loss": 0.3386, + "step": 2973 + }, + { + "epoch": 1.137284894837476, + "grad_norm": 3.5506315126310657, + "learning_rate": 1e-06, + "loss": 0.1616, + "step": 2974 + }, + { + "epoch": 1.1376673040152965, + "grad_norm": 2.289140615334694, + "learning_rate": 1e-06, + "loss": 0.1095, + "step": 2975 + }, + { + "epoch": 1.1380497131931167, + "grad_norm": 3.952317830181921, + "learning_rate": 1e-06, + "loss": 0.25, + "step": 2976 + }, + { + "epoch": 1.1384321223709368, + "grad_norm": 3.027342174898784, + "learning_rate": 1e-06, + "loss": 0.0998, + "step": 2977 + }, + { + "epoch": 1.1388145315487572, + "grad_norm": 4.896933876954402, + "learning_rate": 1e-06, + "loss": 0.2392, + "step": 2978 + }, + { + "epoch": 1.1391969407265774, + "grad_norm": 4.925390144812214, + "learning_rate": 1e-06, + "loss": 0.6317, + "step": 2979 + }, + { + "epoch": 1.1395793499043978, + "grad_norm": 3.719461196723674, + "learning_rate": 1e-06, + "loss": 0.4982, + "step": 2980 + }, + { + "epoch": 1.139961759082218, + "grad_norm": 5.013784861285325, + "learning_rate": 1e-06, + "loss": 0.2064, + "step": 2981 + }, + { + "epoch": 1.1403441682600381, + "grad_norm": 4.531964686364737, + "learning_rate": 1e-06, + "loss": 0.27, + "step": 2982 + }, + { + "epoch": 1.1407265774378585, + "grad_norm": 2.65648317996501, + "learning_rate": 1e-06, + "loss": 0.3822, + "step": 2983 + }, + { + "epoch": 1.1411089866156787, + "grad_norm": 2.444294678068928, + "learning_rate": 1e-06, + "loss": 0.3012, + "step": 2984 + }, + { + "epoch": 1.1414913957934991, + "grad_norm": 3.6912378020428958, + "learning_rate": 1e-06, + "loss": 0.1829, + "step": 2985 + }, + { + "epoch": 1.1418738049713193, + "grad_norm": 3.5530446710288026, + "learning_rate": 1e-06, + "loss": 0.2182, + "step": 2986 + }, + { + "epoch": 1.1422562141491395, + "grad_norm": 3.4648946695983605, + "learning_rate": 1e-06, + "loss": 0.1276, + "step": 2987 + }, + { + "epoch": 1.1426386233269599, + "grad_norm": 3.051578275968885, + "learning_rate": 1e-06, + "loss": 0.1571, + "step": 2988 + }, + { + "epoch": 1.14302103250478, + "grad_norm": 3.6946144654454556, + "learning_rate": 1e-06, + "loss": 0.0636, + "step": 2989 + }, + { + "epoch": 1.1434034416826004, + "grad_norm": 3.6885851540309598, + "learning_rate": 1e-06, + "loss": 0.1163, + "step": 2990 + }, + { + "epoch": 1.1437858508604206, + "grad_norm": 8.472431579115465, + "learning_rate": 1e-06, + "loss": 0.4143, + "step": 2991 + }, + { + "epoch": 1.144168260038241, + "grad_norm": 4.0249433051434576, + "learning_rate": 1e-06, + "loss": 0.3931, + "step": 2992 + }, + { + "epoch": 1.1445506692160612, + "grad_norm": 5.038420691019398, + "learning_rate": 1e-06, + "loss": 0.3297, + "step": 2993 + }, + { + "epoch": 1.1449330783938814, + "grad_norm": 2.810695662811567, + "learning_rate": 1e-06, + "loss": 0.2456, + "step": 2994 + }, + { + "epoch": 1.1453154875717018, + "grad_norm": 2.493642256854354, + "learning_rate": 1e-06, + "loss": 0.2771, + "step": 2995 + }, + { + "epoch": 1.145697896749522, + "grad_norm": 3.446172697675935, + "learning_rate": 1e-06, + "loss": 0.2694, + "step": 2996 + }, + { + "epoch": 1.1460803059273423, + "grad_norm": 3.254186867729784, + "learning_rate": 1e-06, + "loss": 0.2904, + "step": 2997 + }, + { + "epoch": 1.1464627151051625, + "grad_norm": 3.111249349187137, + "learning_rate": 1e-06, + "loss": 0.27, + "step": 2998 + }, + { + "epoch": 1.146845124282983, + "grad_norm": 3.145490076061839, + "learning_rate": 1e-06, + "loss": 0.1915, + "step": 2999 + }, + { + "epoch": 1.147227533460803, + "grad_norm": 3.6650658638774907, + "learning_rate": 1e-06, + "loss": 0.1846, + "step": 3000 + }, + { + "epoch": 1.147227533460803, + "eval_runtime": 847.684, + "eval_samples_per_second": 1.81, + "eval_steps_per_second": 0.453, + "step": 3000 + }, + { + "epoch": 1.1476099426386233, + "grad_norm": 5.917576822830572, + "learning_rate": 1e-06, + "loss": 0.1033, + "step": 3001 + }, + { + "epoch": 1.1479923518164437, + "grad_norm": 4.405933774598463, + "learning_rate": 1e-06, + "loss": 0.0918, + "step": 3002 + }, + { + "epoch": 1.1483747609942638, + "grad_norm": 7.300066688964566, + "learning_rate": 1e-06, + "loss": 0.2163, + "step": 3003 + }, + { + "epoch": 1.1487571701720842, + "grad_norm": 3.231642596302886, + "learning_rate": 1e-06, + "loss": 0.4344, + "step": 3004 + }, + { + "epoch": 1.1491395793499044, + "grad_norm": 2.694351511485435, + "learning_rate": 1e-06, + "loss": 0.2757, + "step": 3005 + }, + { + "epoch": 1.1495219885277246, + "grad_norm": 2.3460577221116568, + "learning_rate": 1e-06, + "loss": 0.3138, + "step": 3006 + }, + { + "epoch": 1.149904397705545, + "grad_norm": 2.1963152241452626, + "learning_rate": 1e-06, + "loss": 0.1383, + "step": 3007 + }, + { + "epoch": 1.1502868068833652, + "grad_norm": 2.6964874802490995, + "learning_rate": 1e-06, + "loss": 0.1382, + "step": 3008 + }, + { + "epoch": 1.1506692160611856, + "grad_norm": 3.3874123818865773, + "learning_rate": 1e-06, + "loss": 0.2823, + "step": 3009 + }, + { + "epoch": 1.1510516252390057, + "grad_norm": 3.1424370088538383, + "learning_rate": 1e-06, + "loss": 0.1919, + "step": 3010 + }, + { + "epoch": 1.151434034416826, + "grad_norm": 3.4296154805401073, + "learning_rate": 1e-06, + "loss": 0.3296, + "step": 3011 + }, + { + "epoch": 1.1518164435946463, + "grad_norm": 2.6362655137077238, + "learning_rate": 1e-06, + "loss": 0.1152, + "step": 3012 + }, + { + "epoch": 1.1521988527724665, + "grad_norm": 2.797309202449541, + "learning_rate": 1e-06, + "loss": 0.0643, + "step": 3013 + }, + { + "epoch": 1.1525812619502869, + "grad_norm": 4.26345490529596, + "learning_rate": 1e-06, + "loss": 0.1267, + "step": 3014 + }, + { + "epoch": 1.152963671128107, + "grad_norm": 4.9750487512088055, + "learning_rate": 1e-06, + "loss": 0.0788, + "step": 3015 + }, + { + "epoch": 1.1533460803059272, + "grad_norm": 7.489272265458247, + "learning_rate": 1e-06, + "loss": 0.6298, + "step": 3016 + }, + { + "epoch": 1.1537284894837476, + "grad_norm": 3.8932575679836057, + "learning_rate": 1e-06, + "loss": 0.397, + "step": 3017 + }, + { + "epoch": 1.1541108986615678, + "grad_norm": 2.6774029571870215, + "learning_rate": 1e-06, + "loss": 0.2164, + "step": 3018 + }, + { + "epoch": 1.1544933078393882, + "grad_norm": 2.303157803057292, + "learning_rate": 1e-06, + "loss": 0.1739, + "step": 3019 + }, + { + "epoch": 1.1548757170172084, + "grad_norm": 3.7227835698485845, + "learning_rate": 1e-06, + "loss": 0.3657, + "step": 3020 + }, + { + "epoch": 1.1552581261950288, + "grad_norm": 3.15923417529325, + "learning_rate": 1e-06, + "loss": 0.3346, + "step": 3021 + }, + { + "epoch": 1.155640535372849, + "grad_norm": 2.8229112988710687, + "learning_rate": 1e-06, + "loss": 0.1501, + "step": 3022 + }, + { + "epoch": 1.1560229445506693, + "grad_norm": 2.5378042066908555, + "learning_rate": 1e-06, + "loss": 0.2735, + "step": 3023 + }, + { + "epoch": 1.1564053537284895, + "grad_norm": 3.1087593734348347, + "learning_rate": 1e-06, + "loss": 0.1523, + "step": 3024 + }, + { + "epoch": 1.1567877629063097, + "grad_norm": 3.8077057257996065, + "learning_rate": 1e-06, + "loss": 0.1377, + "step": 3025 + }, + { + "epoch": 1.15717017208413, + "grad_norm": 3.5215631815745247, + "learning_rate": 1e-06, + "loss": 0.1032, + "step": 3026 + }, + { + "epoch": 1.1575525812619503, + "grad_norm": 3.840747311710906, + "learning_rate": 1e-06, + "loss": 0.0706, + "step": 3027 + }, + { + "epoch": 1.1579349904397707, + "grad_norm": 6.1228736767769085, + "learning_rate": 1e-06, + "loss": 0.1465, + "step": 3028 + }, + { + "epoch": 1.1583173996175908, + "grad_norm": 3.953383365676891, + "learning_rate": 1e-06, + "loss": 0.5059, + "step": 3029 + }, + { + "epoch": 1.158699808795411, + "grad_norm": 3.1208554058711964, + "learning_rate": 1e-06, + "loss": 0.4097, + "step": 3030 + }, + { + "epoch": 1.1590822179732314, + "grad_norm": 3.3829022895578587, + "learning_rate": 1e-06, + "loss": 0.453, + "step": 3031 + }, + { + "epoch": 1.1594646271510516, + "grad_norm": 2.4288076297914083, + "learning_rate": 1e-06, + "loss": 0.2809, + "step": 3032 + }, + { + "epoch": 1.159847036328872, + "grad_norm": 2.6608344514507745, + "learning_rate": 1e-06, + "loss": 0.2809, + "step": 3033 + }, + { + "epoch": 1.1602294455066922, + "grad_norm": 4.02149955224188, + "learning_rate": 1e-06, + "loss": 0.2122, + "step": 3034 + }, + { + "epoch": 1.1606118546845123, + "grad_norm": 3.1588094172986207, + "learning_rate": 1e-06, + "loss": 0.2068, + "step": 3035 + }, + { + "epoch": 1.1609942638623327, + "grad_norm": 3.170947611115864, + "learning_rate": 1e-06, + "loss": 0.1697, + "step": 3036 + }, + { + "epoch": 1.161376673040153, + "grad_norm": 2.618345818476673, + "learning_rate": 1e-06, + "loss": 0.2284, + "step": 3037 + }, + { + "epoch": 1.1617590822179733, + "grad_norm": 3.863046482771306, + "learning_rate": 1e-06, + "loss": 0.1383, + "step": 3038 + }, + { + "epoch": 1.1621414913957935, + "grad_norm": 4.332092009549107, + "learning_rate": 1e-06, + "loss": 0.2665, + "step": 3039 + }, + { + "epoch": 1.1625239005736137, + "grad_norm": 3.7784688510718873, + "learning_rate": 1e-06, + "loss": 0.1089, + "step": 3040 + }, + { + "epoch": 1.162906309751434, + "grad_norm": 4.3595318509668814, + "learning_rate": 1e-06, + "loss": 0.3529, + "step": 3041 + }, + { + "epoch": 1.1632887189292542, + "grad_norm": 3.421964182627793, + "learning_rate": 1e-06, + "loss": 0.3063, + "step": 3042 + }, + { + "epoch": 1.1636711281070746, + "grad_norm": 2.8840746093919036, + "learning_rate": 1e-06, + "loss": 0.3455, + "step": 3043 + }, + { + "epoch": 1.1640535372848948, + "grad_norm": 3.5628827876517244, + "learning_rate": 1e-06, + "loss": 0.3037, + "step": 3044 + }, + { + "epoch": 1.1644359464627152, + "grad_norm": 4.2031746900379465, + "learning_rate": 1e-06, + "loss": 0.3853, + "step": 3045 + }, + { + "epoch": 1.1648183556405354, + "grad_norm": 3.292110670702286, + "learning_rate": 1e-06, + "loss": 0.1813, + "step": 3046 + }, + { + "epoch": 1.1652007648183555, + "grad_norm": 2.308336582479933, + "learning_rate": 1e-06, + "loss": 0.1946, + "step": 3047 + }, + { + "epoch": 1.165583173996176, + "grad_norm": 2.523929512551048, + "learning_rate": 1e-06, + "loss": 0.187, + "step": 3048 + }, + { + "epoch": 1.1659655831739961, + "grad_norm": 3.2102688652406965, + "learning_rate": 1e-06, + "loss": 0.1661, + "step": 3049 + }, + { + "epoch": 1.1663479923518165, + "grad_norm": 3.2124555948357485, + "learning_rate": 1e-06, + "loss": 0.1656, + "step": 3050 + }, + { + "epoch": 1.1667304015296367, + "grad_norm": 3.8921057399644803, + "learning_rate": 1e-06, + "loss": 0.1297, + "step": 3051 + }, + { + "epoch": 1.167112810707457, + "grad_norm": 3.6307160922256356, + "learning_rate": 1e-06, + "loss": 0.1017, + "step": 3052 + }, + { + "epoch": 1.1674952198852773, + "grad_norm": 7.806813360063579, + "learning_rate": 1e-06, + "loss": 0.2513, + "step": 3053 + }, + { + "epoch": 1.1678776290630974, + "grad_norm": 4.472717444655059, + "learning_rate": 1e-06, + "loss": 0.4653, + "step": 3054 + }, + { + "epoch": 1.1682600382409178, + "grad_norm": 3.798022438260248, + "learning_rate": 1e-06, + "loss": 0.4606, + "step": 3055 + }, + { + "epoch": 1.168642447418738, + "grad_norm": 3.1168322826859893, + "learning_rate": 1e-06, + "loss": 0.4191, + "step": 3056 + }, + { + "epoch": 1.1690248565965584, + "grad_norm": 2.7548288085144588, + "learning_rate": 1e-06, + "loss": 0.2374, + "step": 3057 + }, + { + "epoch": 1.1694072657743786, + "grad_norm": 3.4893264144529557, + "learning_rate": 1e-06, + "loss": 0.3185, + "step": 3058 + }, + { + "epoch": 1.1697896749521988, + "grad_norm": 2.09609705767973, + "learning_rate": 1e-06, + "loss": 0.199, + "step": 3059 + }, + { + "epoch": 1.1701720841300192, + "grad_norm": 2.7151776225119324, + "learning_rate": 1e-06, + "loss": 0.0903, + "step": 3060 + }, + { + "epoch": 1.1705544933078393, + "grad_norm": 3.4406712555786036, + "learning_rate": 1e-06, + "loss": 0.1539, + "step": 3061 + }, + { + "epoch": 1.1709369024856597, + "grad_norm": 3.926922350483711, + "learning_rate": 1e-06, + "loss": 0.2428, + "step": 3062 + }, + { + "epoch": 1.17131931166348, + "grad_norm": 3.4846850885877734, + "learning_rate": 1e-06, + "loss": 0.0542, + "step": 3063 + }, + { + "epoch": 1.1717017208413, + "grad_norm": 3.7392877800199655, + "learning_rate": 1e-06, + "loss": 0.0978, + "step": 3064 + }, + { + "epoch": 1.1720841300191205, + "grad_norm": 5.540720961719221, + "learning_rate": 1e-06, + "loss": 0.146, + "step": 3065 + }, + { + "epoch": 1.1724665391969407, + "grad_norm": 6.176756500123378, + "learning_rate": 1e-06, + "loss": 0.3402, + "step": 3066 + }, + { + "epoch": 1.172848948374761, + "grad_norm": 5.918927508454703, + "learning_rate": 1e-06, + "loss": 0.3864, + "step": 3067 + }, + { + "epoch": 1.1732313575525812, + "grad_norm": 2.693373184224026, + "learning_rate": 1e-06, + "loss": 0.2767, + "step": 3068 + }, + { + "epoch": 1.1736137667304014, + "grad_norm": 2.91300425605188, + "learning_rate": 1e-06, + "loss": 0.3393, + "step": 3069 + }, + { + "epoch": 1.1739961759082218, + "grad_norm": 3.2628369270654747, + "learning_rate": 1e-06, + "loss": 0.4239, + "step": 3070 + }, + { + "epoch": 1.174378585086042, + "grad_norm": 4.737414703003876, + "learning_rate": 1e-06, + "loss": 0.2154, + "step": 3071 + }, + { + "epoch": 1.1747609942638624, + "grad_norm": 4.059744501953611, + "learning_rate": 1e-06, + "loss": 0.2917, + "step": 3072 + }, + { + "epoch": 1.1751434034416826, + "grad_norm": 2.408483682271091, + "learning_rate": 1e-06, + "loss": 0.2515, + "step": 3073 + }, + { + "epoch": 1.175525812619503, + "grad_norm": 2.8249414556057793, + "learning_rate": 1e-06, + "loss": 0.1551, + "step": 3074 + }, + { + "epoch": 1.1759082217973231, + "grad_norm": 3.9065458872311956, + "learning_rate": 1e-06, + "loss": 0.132, + "step": 3075 + }, + { + "epoch": 1.1762906309751435, + "grad_norm": 3.04849106405762, + "learning_rate": 1e-06, + "loss": 0.0795, + "step": 3076 + }, + { + "epoch": 1.1766730401529637, + "grad_norm": 3.563581871260127, + "learning_rate": 1e-06, + "loss": 0.0695, + "step": 3077 + }, + { + "epoch": 1.1770554493307839, + "grad_norm": 5.442069402788394, + "learning_rate": 1e-06, + "loss": 0.2164, + "step": 3078 + }, + { + "epoch": 1.1774378585086043, + "grad_norm": 4.555951998034926, + "learning_rate": 1e-06, + "loss": 0.3949, + "step": 3079 + }, + { + "epoch": 1.1778202676864244, + "grad_norm": 3.025138908538679, + "learning_rate": 1e-06, + "loss": 0.2971, + "step": 3080 + }, + { + "epoch": 1.1782026768642448, + "grad_norm": 1.8902881416798445, + "learning_rate": 1e-06, + "loss": 0.184, + "step": 3081 + }, + { + "epoch": 1.178585086042065, + "grad_norm": 2.9900248308968744, + "learning_rate": 1e-06, + "loss": 0.3602, + "step": 3082 + }, + { + "epoch": 1.1789674952198852, + "grad_norm": 2.448414550362092, + "learning_rate": 1e-06, + "loss": 0.1353, + "step": 3083 + }, + { + "epoch": 1.1793499043977056, + "grad_norm": 2.5123728705739317, + "learning_rate": 1e-06, + "loss": 0.2672, + "step": 3084 + }, + { + "epoch": 1.1797323135755258, + "grad_norm": 2.6692734930754027, + "learning_rate": 1e-06, + "loss": 0.3763, + "step": 3085 + }, + { + "epoch": 1.1801147227533462, + "grad_norm": 3.5069497455686953, + "learning_rate": 1e-06, + "loss": 0.2568, + "step": 3086 + }, + { + "epoch": 1.1804971319311663, + "grad_norm": 3.536805905025983, + "learning_rate": 1e-06, + "loss": 0.314, + "step": 3087 + }, + { + "epoch": 1.1808795411089865, + "grad_norm": 3.8165563048241626, + "learning_rate": 1e-06, + "loss": 0.1105, + "step": 3088 + }, + { + "epoch": 1.181261950286807, + "grad_norm": 2.945043414950344, + "learning_rate": 1e-06, + "loss": 0.0548, + "step": 3089 + }, + { + "epoch": 1.181644359464627, + "grad_norm": 4.024186681235672, + "learning_rate": 1e-06, + "loss": 0.2052, + "step": 3090 + }, + { + "epoch": 1.1820267686424475, + "grad_norm": 4.989335512551899, + "learning_rate": 1e-06, + "loss": 0.3828, + "step": 3091 + }, + { + "epoch": 1.1824091778202677, + "grad_norm": 3.8864247886636702, + "learning_rate": 1e-06, + "loss": 0.392, + "step": 3092 + }, + { + "epoch": 1.1827915869980878, + "grad_norm": 2.6479801469502333, + "learning_rate": 1e-06, + "loss": 0.4046, + "step": 3093 + }, + { + "epoch": 1.1831739961759082, + "grad_norm": 3.7650680446681766, + "learning_rate": 1e-06, + "loss": 0.3263, + "step": 3094 + }, + { + "epoch": 1.1835564053537284, + "grad_norm": 2.9397296965506436, + "learning_rate": 1e-06, + "loss": 0.2835, + "step": 3095 + }, + { + "epoch": 1.1839388145315488, + "grad_norm": 3.034743350896505, + "learning_rate": 1e-06, + "loss": 0.2192, + "step": 3096 + }, + { + "epoch": 1.184321223709369, + "grad_norm": 3.721951172735419, + "learning_rate": 1e-06, + "loss": 0.321, + "step": 3097 + }, + { + "epoch": 1.1847036328871894, + "grad_norm": 2.5948500541040307, + "learning_rate": 1e-06, + "loss": 0.3201, + "step": 3098 + }, + { + "epoch": 1.1850860420650096, + "grad_norm": 3.331996617914249, + "learning_rate": 1e-06, + "loss": 0.2145, + "step": 3099 + }, + { + "epoch": 1.1854684512428297, + "grad_norm": 3.5142993377350598, + "learning_rate": 1e-06, + "loss": 0.2495, + "step": 3100 + }, + { + "epoch": 1.1858508604206501, + "grad_norm": 3.116332890662486, + "learning_rate": 1e-06, + "loss": 0.1489, + "step": 3101 + }, + { + "epoch": 1.1862332695984703, + "grad_norm": 5.030351262350556, + "learning_rate": 1e-06, + "loss": 0.1509, + "step": 3102 + }, + { + "epoch": 1.1866156787762907, + "grad_norm": 5.315250178538092, + "learning_rate": 1e-06, + "loss": 0.1946, + "step": 3103 + }, + { + "epoch": 1.1869980879541109, + "grad_norm": 2.920293206426201, + "learning_rate": 1e-06, + "loss": 0.3817, + "step": 3104 + }, + { + "epoch": 1.1873804971319313, + "grad_norm": 3.344548923202473, + "learning_rate": 1e-06, + "loss": 0.4088, + "step": 3105 + }, + { + "epoch": 1.1877629063097515, + "grad_norm": 2.7876508116018255, + "learning_rate": 1e-06, + "loss": 0.2356, + "step": 3106 + }, + { + "epoch": 1.1881453154875716, + "grad_norm": 3.4527489642980744, + "learning_rate": 1e-06, + "loss": 0.3825, + "step": 3107 + }, + { + "epoch": 1.188527724665392, + "grad_norm": 2.055256695830084, + "learning_rate": 1e-06, + "loss": 0.3209, + "step": 3108 + }, + { + "epoch": 1.1889101338432122, + "grad_norm": 6.191801175298961, + "learning_rate": 1e-06, + "loss": 0.2992, + "step": 3109 + }, + { + "epoch": 1.1892925430210326, + "grad_norm": 2.7793039092606873, + "learning_rate": 1e-06, + "loss": 0.2882, + "step": 3110 + }, + { + "epoch": 1.1896749521988528, + "grad_norm": 2.7949971122931507, + "learning_rate": 1e-06, + "loss": 0.1721, + "step": 3111 + }, + { + "epoch": 1.190057361376673, + "grad_norm": 2.5417224273401775, + "learning_rate": 1e-06, + "loss": 0.1158, + "step": 3112 + }, + { + "epoch": 1.1904397705544933, + "grad_norm": 3.445079423628347, + "learning_rate": 1e-06, + "loss": 0.1185, + "step": 3113 + }, + { + "epoch": 1.1908221797323135, + "grad_norm": 3.27700822133397, + "learning_rate": 1e-06, + "loss": 0.0558, + "step": 3114 + }, + { + "epoch": 1.191204588910134, + "grad_norm": 5.881778783265607, + "learning_rate": 1e-06, + "loss": 0.0909, + "step": 3115 + }, + { + "epoch": 1.191586998087954, + "grad_norm": 4.455296037387607, + "learning_rate": 1e-06, + "loss": 0.1491, + "step": 3116 + }, + { + "epoch": 1.1919694072657743, + "grad_norm": 4.650520229976839, + "learning_rate": 1e-06, + "loss": 0.7205, + "step": 3117 + }, + { + "epoch": 1.1923518164435947, + "grad_norm": 2.8509326814098337, + "learning_rate": 1e-06, + "loss": 0.3414, + "step": 3118 + }, + { + "epoch": 1.1927342256214148, + "grad_norm": 2.75630315266608, + "learning_rate": 1e-06, + "loss": 0.2999, + "step": 3119 + }, + { + "epoch": 1.1931166347992352, + "grad_norm": 2.8343341593128977, + "learning_rate": 1e-06, + "loss": 0.2819, + "step": 3120 + }, + { + "epoch": 1.1934990439770554, + "grad_norm": 5.30408160832426, + "learning_rate": 1e-06, + "loss": 0.2805, + "step": 3121 + }, + { + "epoch": 1.1938814531548756, + "grad_norm": 2.828519509129359, + "learning_rate": 1e-06, + "loss": 0.1974, + "step": 3122 + }, + { + "epoch": 1.194263862332696, + "grad_norm": 2.3947696038206474, + "learning_rate": 1e-06, + "loss": 0.1117, + "step": 3123 + }, + { + "epoch": 1.1946462715105162, + "grad_norm": 3.3063217703343417, + "learning_rate": 1e-06, + "loss": 0.2637, + "step": 3124 + }, + { + "epoch": 1.1950286806883366, + "grad_norm": 2.5935077783936054, + "learning_rate": 1e-06, + "loss": 0.0736, + "step": 3125 + }, + { + "epoch": 1.1954110898661567, + "grad_norm": 3.0294580570433327, + "learning_rate": 1e-06, + "loss": 0.078, + "step": 3126 + }, + { + "epoch": 1.1957934990439771, + "grad_norm": 2.656613762574868, + "learning_rate": 1e-06, + "loss": 0.0651, + "step": 3127 + }, + { + "epoch": 1.1961759082217973, + "grad_norm": 5.002093449076198, + "learning_rate": 1e-06, + "loss": 0.2206, + "step": 3128 + }, + { + "epoch": 1.1965583173996177, + "grad_norm": 5.949253695830555, + "learning_rate": 1e-06, + "loss": 0.3757, + "step": 3129 + }, + { + "epoch": 1.1969407265774379, + "grad_norm": 3.824105780299611, + "learning_rate": 1e-06, + "loss": 0.327, + "step": 3130 + }, + { + "epoch": 1.197323135755258, + "grad_norm": 1.6669698757294193, + "learning_rate": 1e-06, + "loss": 0.2487, + "step": 3131 + }, + { + "epoch": 1.1977055449330785, + "grad_norm": 3.2204468337705974, + "learning_rate": 1e-06, + "loss": 0.2952, + "step": 3132 + }, + { + "epoch": 1.1980879541108986, + "grad_norm": 2.9390804827645467, + "learning_rate": 1e-06, + "loss": 0.2063, + "step": 3133 + }, + { + "epoch": 1.198470363288719, + "grad_norm": 3.1498462942254966, + "learning_rate": 1e-06, + "loss": 0.1814, + "step": 3134 + }, + { + "epoch": 1.1988527724665392, + "grad_norm": 2.3931106908873043, + "learning_rate": 1e-06, + "loss": 0.1524, + "step": 3135 + }, + { + "epoch": 1.1992351816443594, + "grad_norm": 2.448078577625523, + "learning_rate": 1e-06, + "loss": 0.1638, + "step": 3136 + }, + { + "epoch": 1.1996175908221798, + "grad_norm": 2.7783528357883305, + "learning_rate": 1e-06, + "loss": 0.107, + "step": 3137 + }, + { + "epoch": 1.2, + "grad_norm": 4.39841516736393, + "learning_rate": 1e-06, + "loss": 0.3795, + "step": 3138 + }, + { + "epoch": 1.2003824091778204, + "grad_norm": 5.160358618465096, + "learning_rate": 1e-06, + "loss": 0.122, + "step": 3139 + }, + { + "epoch": 1.2007648183556405, + "grad_norm": 3.327718647251247, + "learning_rate": 1e-06, + "loss": 0.0418, + "step": 3140 + }, + { + "epoch": 1.2011472275334607, + "grad_norm": 3.2284810640731814, + "learning_rate": 1e-06, + "loss": 0.4536, + "step": 3141 + }, + { + "epoch": 1.201529636711281, + "grad_norm": 4.084538470078328, + "learning_rate": 1e-06, + "loss": 0.4799, + "step": 3142 + }, + { + "epoch": 1.2019120458891013, + "grad_norm": 3.1492961157662314, + "learning_rate": 1e-06, + "loss": 0.3956, + "step": 3143 + }, + { + "epoch": 1.2022944550669217, + "grad_norm": 3.3343670831530896, + "learning_rate": 1e-06, + "loss": 0.2856, + "step": 3144 + }, + { + "epoch": 1.2026768642447419, + "grad_norm": 3.375294001454161, + "learning_rate": 1e-06, + "loss": 0.4036, + "step": 3145 + }, + { + "epoch": 1.203059273422562, + "grad_norm": 3.1989175098082954, + "learning_rate": 1e-06, + "loss": 0.1995, + "step": 3146 + }, + { + "epoch": 1.2034416826003824, + "grad_norm": 3.151750132885931, + "learning_rate": 1e-06, + "loss": 0.1994, + "step": 3147 + }, + { + "epoch": 1.2038240917782026, + "grad_norm": 2.7971436781738004, + "learning_rate": 1e-06, + "loss": 0.2323, + "step": 3148 + }, + { + "epoch": 1.204206500956023, + "grad_norm": 3.129228249124635, + "learning_rate": 1e-06, + "loss": 0.1711, + "step": 3149 + }, + { + "epoch": 1.2045889101338432, + "grad_norm": 4.379543016536247, + "learning_rate": 1e-06, + "loss": 0.2515, + "step": 3150 + }, + { + "epoch": 1.2049713193116636, + "grad_norm": 3.2287229827920165, + "learning_rate": 1e-06, + "loss": 0.1618, + "step": 3151 + }, + { + "epoch": 1.2053537284894837, + "grad_norm": 3.0254380657695057, + "learning_rate": 1e-06, + "loss": 0.1024, + "step": 3152 + }, + { + "epoch": 1.205736137667304, + "grad_norm": 6.945429214166259, + "learning_rate": 1e-06, + "loss": 0.3903, + "step": 3153 + }, + { + "epoch": 1.2061185468451243, + "grad_norm": 3.033570649920902, + "learning_rate": 1e-06, + "loss": 0.5589, + "step": 3154 + }, + { + "epoch": 1.2065009560229445, + "grad_norm": 4.548874877322538, + "learning_rate": 1e-06, + "loss": 0.2562, + "step": 3155 + }, + { + "epoch": 1.206883365200765, + "grad_norm": 2.5369299751033494, + "learning_rate": 1e-06, + "loss": 0.1815, + "step": 3156 + }, + { + "epoch": 1.207265774378585, + "grad_norm": 2.8103183337956694, + "learning_rate": 1e-06, + "loss": 0.3868, + "step": 3157 + }, + { + "epoch": 1.2076481835564055, + "grad_norm": 2.536577340740925, + "learning_rate": 1e-06, + "loss": 0.2505, + "step": 3158 + }, + { + "epoch": 1.2080305927342256, + "grad_norm": 2.4715991419671997, + "learning_rate": 1e-06, + "loss": 0.2468, + "step": 3159 + }, + { + "epoch": 1.2084130019120458, + "grad_norm": 4.598208385363217, + "learning_rate": 1e-06, + "loss": 0.2992, + "step": 3160 + }, + { + "epoch": 1.2087954110898662, + "grad_norm": 3.0435948639799157, + "learning_rate": 1e-06, + "loss": 0.0999, + "step": 3161 + }, + { + "epoch": 1.2091778202676864, + "grad_norm": 2.9515441215431655, + "learning_rate": 1e-06, + "loss": 0.1144, + "step": 3162 + }, + { + "epoch": 1.2095602294455068, + "grad_norm": 3.805486011720499, + "learning_rate": 1e-06, + "loss": 0.0523, + "step": 3163 + }, + { + "epoch": 1.209942638623327, + "grad_norm": 3.691795821043919, + "learning_rate": 1e-06, + "loss": 0.1492, + "step": 3164 + }, + { + "epoch": 1.2103250478011471, + "grad_norm": 3.65279127020977, + "learning_rate": 1e-06, + "loss": 0.0802, + "step": 3165 + }, + { + "epoch": 1.2107074569789675, + "grad_norm": 5.169597573767907, + "learning_rate": 1e-06, + "loss": 0.4806, + "step": 3166 + }, + { + "epoch": 1.2110898661567877, + "grad_norm": 3.912581294863535, + "learning_rate": 1e-06, + "loss": 0.5567, + "step": 3167 + }, + { + "epoch": 1.211472275334608, + "grad_norm": 3.827730866987371, + "learning_rate": 1e-06, + "loss": 0.4558, + "step": 3168 + }, + { + "epoch": 1.2118546845124283, + "grad_norm": 1.7782685609060942, + "learning_rate": 1e-06, + "loss": 0.2566, + "step": 3169 + }, + { + "epoch": 1.2122370936902485, + "grad_norm": 1.924411743249389, + "learning_rate": 1e-06, + "loss": 0.2081, + "step": 3170 + }, + { + "epoch": 1.2126195028680689, + "grad_norm": 2.906422763734687, + "learning_rate": 1e-06, + "loss": 0.2273, + "step": 3171 + }, + { + "epoch": 1.213001912045889, + "grad_norm": 3.692293058574239, + "learning_rate": 1e-06, + "loss": 0.2721, + "step": 3172 + }, + { + "epoch": 1.2133843212237094, + "grad_norm": 2.9293873544167552, + "learning_rate": 1e-06, + "loss": 0.1852, + "step": 3173 + }, + { + "epoch": 1.2137667304015296, + "grad_norm": 3.8968379039574175, + "learning_rate": 1e-06, + "loss": 0.2908, + "step": 3174 + }, + { + "epoch": 1.21414913957935, + "grad_norm": 3.956550892842321, + "learning_rate": 1e-06, + "loss": 0.1222, + "step": 3175 + }, + { + "epoch": 1.2145315487571702, + "grad_norm": 2.752969785508569, + "learning_rate": 1e-06, + "loss": 0.0564, + "step": 3176 + }, + { + "epoch": 1.2149139579349904, + "grad_norm": 3.5963030163964476, + "learning_rate": 1e-06, + "loss": 0.095, + "step": 3177 + }, + { + "epoch": 1.2152963671128107, + "grad_norm": 6.239211575021387, + "learning_rate": 1e-06, + "loss": 0.3205, + "step": 3178 + }, + { + "epoch": 1.215678776290631, + "grad_norm": 3.5378499427487347, + "learning_rate": 1e-06, + "loss": 0.3933, + "step": 3179 + }, + { + "epoch": 1.2160611854684513, + "grad_norm": 4.331651263344768, + "learning_rate": 1e-06, + "loss": 0.569, + "step": 3180 + }, + { + "epoch": 1.2164435946462715, + "grad_norm": 2.1014254500503085, + "learning_rate": 1e-06, + "loss": 0.2221, + "step": 3181 + }, + { + "epoch": 1.216826003824092, + "grad_norm": 2.570308511377272, + "learning_rate": 1e-06, + "loss": 0.3164, + "step": 3182 + }, + { + "epoch": 1.217208413001912, + "grad_norm": 2.5685503174329316, + "learning_rate": 1e-06, + "loss": 0.2736, + "step": 3183 + }, + { + "epoch": 1.2175908221797322, + "grad_norm": 3.2785645167553654, + "learning_rate": 1e-06, + "loss": 0.1699, + "step": 3184 + }, + { + "epoch": 1.2179732313575526, + "grad_norm": 2.6871196566403186, + "learning_rate": 1e-06, + "loss": 0.146, + "step": 3185 + }, + { + "epoch": 1.2183556405353728, + "grad_norm": 3.3899147239019376, + "learning_rate": 1e-06, + "loss": 0.1565, + "step": 3186 + }, + { + "epoch": 1.2187380497131932, + "grad_norm": 3.341117321595433, + "learning_rate": 1e-06, + "loss": 0.1689, + "step": 3187 + }, + { + "epoch": 1.2191204588910134, + "grad_norm": 4.294075126498907, + "learning_rate": 1e-06, + "loss": 0.1261, + "step": 3188 + }, + { + "epoch": 1.2195028680688336, + "grad_norm": 6.297185468947398, + "learning_rate": 1e-06, + "loss": 0.1083, + "step": 3189 + }, + { + "epoch": 1.219885277246654, + "grad_norm": 6.680371629805088, + "learning_rate": 1e-06, + "loss": 0.2221, + "step": 3190 + }, + { + "epoch": 1.2202676864244741, + "grad_norm": 4.298948585247168, + "learning_rate": 1e-06, + "loss": 0.4022, + "step": 3191 + }, + { + "epoch": 1.2206500956022945, + "grad_norm": 4.209939530543174, + "learning_rate": 1e-06, + "loss": 0.4461, + "step": 3192 + }, + { + "epoch": 1.2210325047801147, + "grad_norm": 4.099530514144439, + "learning_rate": 1e-06, + "loss": 0.2386, + "step": 3193 + }, + { + "epoch": 1.221414913957935, + "grad_norm": 2.939929201220825, + "learning_rate": 1e-06, + "loss": 0.4194, + "step": 3194 + }, + { + "epoch": 1.2217973231357553, + "grad_norm": 3.1781842101516435, + "learning_rate": 1e-06, + "loss": 0.3398, + "step": 3195 + }, + { + "epoch": 1.2221797323135755, + "grad_norm": 3.066835720820562, + "learning_rate": 1e-06, + "loss": 0.4716, + "step": 3196 + }, + { + "epoch": 1.2225621414913959, + "grad_norm": 3.2932464735903735, + "learning_rate": 1e-06, + "loss": 0.373, + "step": 3197 + }, + { + "epoch": 1.222944550669216, + "grad_norm": 3.2954179337240603, + "learning_rate": 1e-06, + "loss": 0.3699, + "step": 3198 + }, + { + "epoch": 1.2233269598470362, + "grad_norm": 2.894484793718795, + "learning_rate": 1e-06, + "loss": 0.1878, + "step": 3199 + }, + { + "epoch": 1.2237093690248566, + "grad_norm": 3.8646033081744915, + "learning_rate": 1e-06, + "loss": 0.1205, + "step": 3200 + }, + { + "epoch": 1.2240917782026768, + "grad_norm": 3.157835307314965, + "learning_rate": 1e-06, + "loss": 0.2107, + "step": 3201 + }, + { + "epoch": 1.2244741873804972, + "grad_norm": 3.8377591719695143, + "learning_rate": 1e-06, + "loss": 0.1542, + "step": 3202 + }, + { + "epoch": 1.2248565965583174, + "grad_norm": 8.411278791056619, + "learning_rate": 1e-06, + "loss": 0.1974, + "step": 3203 + }, + { + "epoch": 1.2252390057361378, + "grad_norm": 4.466116175724537, + "learning_rate": 1e-06, + "loss": 0.3859, + "step": 3204 + }, + { + "epoch": 1.225621414913958, + "grad_norm": 4.452325688954857, + "learning_rate": 1e-06, + "loss": 0.4961, + "step": 3205 + }, + { + "epoch": 1.2260038240917783, + "grad_norm": 2.7106294044633, + "learning_rate": 1e-06, + "loss": 0.3824, + "step": 3206 + }, + { + "epoch": 1.2263862332695985, + "grad_norm": 2.5069241955879553, + "learning_rate": 1e-06, + "loss": 0.3542, + "step": 3207 + }, + { + "epoch": 1.2267686424474187, + "grad_norm": 4.087994438854399, + "learning_rate": 1e-06, + "loss": 0.4534, + "step": 3208 + }, + { + "epoch": 1.227151051625239, + "grad_norm": 3.07792267884815, + "learning_rate": 1e-06, + "loss": 0.2163, + "step": 3209 + }, + { + "epoch": 1.2275334608030593, + "grad_norm": 2.741479505123859, + "learning_rate": 1e-06, + "loss": 0.1893, + "step": 3210 + }, + { + "epoch": 1.2279158699808796, + "grad_norm": 2.9077849897736474, + "learning_rate": 1e-06, + "loss": 0.2899, + "step": 3211 + }, + { + "epoch": 1.2282982791586998, + "grad_norm": 3.499013352971792, + "learning_rate": 1e-06, + "loss": 0.1566, + "step": 3212 + }, + { + "epoch": 1.22868068833652, + "grad_norm": 2.333498086108299, + "learning_rate": 1e-06, + "loss": 0.0706, + "step": 3213 + }, + { + "epoch": 1.2290630975143404, + "grad_norm": 5.4533163217467315, + "learning_rate": 1e-06, + "loss": 0.0639, + "step": 3214 + }, + { + "epoch": 1.2294455066921606, + "grad_norm": 5.173192343380095, + "learning_rate": 1e-06, + "loss": 0.1286, + "step": 3215 + }, + { + "epoch": 1.229827915869981, + "grad_norm": 7.6937607493197655, + "learning_rate": 1e-06, + "loss": 0.3088, + "step": 3216 + }, + { + "epoch": 1.2302103250478011, + "grad_norm": 2.636512488315521, + "learning_rate": 1e-06, + "loss": 0.3282, + "step": 3217 + }, + { + "epoch": 1.2305927342256213, + "grad_norm": 3.1988446176420298, + "learning_rate": 1e-06, + "loss": 0.5207, + "step": 3218 + }, + { + "epoch": 1.2309751434034417, + "grad_norm": 2.15027620182632, + "learning_rate": 1e-06, + "loss": 0.2561, + "step": 3219 + }, + { + "epoch": 1.231357552581262, + "grad_norm": 2.987927943268552, + "learning_rate": 1e-06, + "loss": 0.2925, + "step": 3220 + }, + { + "epoch": 1.2317399617590823, + "grad_norm": 4.174506910981116, + "learning_rate": 1e-06, + "loss": 0.4875, + "step": 3221 + }, + { + "epoch": 1.2321223709369025, + "grad_norm": 2.3061472262508063, + "learning_rate": 1e-06, + "loss": 0.3425, + "step": 3222 + }, + { + "epoch": 1.2325047801147226, + "grad_norm": 2.5689735519514834, + "learning_rate": 1e-06, + "loss": 0.1534, + "step": 3223 + }, + { + "epoch": 1.232887189292543, + "grad_norm": 3.447316804241193, + "learning_rate": 1e-06, + "loss": 0.1273, + "step": 3224 + }, + { + "epoch": 1.2332695984703632, + "grad_norm": 3.474353285514223, + "learning_rate": 1e-06, + "loss": 0.1209, + "step": 3225 + }, + { + "epoch": 1.2336520076481836, + "grad_norm": 2.7847654434955316, + "learning_rate": 1e-06, + "loss": 0.0734, + "step": 3226 + }, + { + "epoch": 1.2340344168260038, + "grad_norm": 7.495555069151165, + "learning_rate": 1e-06, + "loss": 0.1253, + "step": 3227 + }, + { + "epoch": 1.2344168260038242, + "grad_norm": 6.239125366047068, + "learning_rate": 1e-06, + "loss": 0.2443, + "step": 3228 + }, + { + "epoch": 1.2347992351816444, + "grad_norm": 4.006409516632794, + "learning_rate": 1e-06, + "loss": 0.2613, + "step": 3229 + }, + { + "epoch": 1.2351816443594645, + "grad_norm": 3.460577056921942, + "learning_rate": 1e-06, + "loss": 0.3549, + "step": 3230 + }, + { + "epoch": 1.235564053537285, + "grad_norm": 2.0768008495023427, + "learning_rate": 1e-06, + "loss": 0.2616, + "step": 3231 + }, + { + "epoch": 1.235946462715105, + "grad_norm": 3.5882055884651836, + "learning_rate": 1e-06, + "loss": 0.2864, + "step": 3232 + }, + { + "epoch": 1.2363288718929255, + "grad_norm": 3.511522943216704, + "learning_rate": 1e-06, + "loss": 0.3731, + "step": 3233 + }, + { + "epoch": 1.2367112810707457, + "grad_norm": 15.408540690821539, + "learning_rate": 1e-06, + "loss": 0.2628, + "step": 3234 + }, + { + "epoch": 1.237093690248566, + "grad_norm": 3.8764653050309117, + "learning_rate": 1e-06, + "loss": 0.1392, + "step": 3235 + }, + { + "epoch": 1.2374760994263863, + "grad_norm": 2.5612972507242073, + "learning_rate": 1e-06, + "loss": 0.1312, + "step": 3236 + }, + { + "epoch": 1.2378585086042064, + "grad_norm": 2.7494724808096356, + "learning_rate": 1e-06, + "loss": 0.2313, + "step": 3237 + }, + { + "epoch": 1.2382409177820268, + "grad_norm": 3.3998463483760437, + "learning_rate": 1e-06, + "loss": 0.2531, + "step": 3238 + }, + { + "epoch": 1.238623326959847, + "grad_norm": 3.4332080750316316, + "learning_rate": 1e-06, + "loss": 0.0872, + "step": 3239 + }, + { + "epoch": 1.2390057361376674, + "grad_norm": 3.708626439087289, + "learning_rate": 1e-06, + "loss": 0.0635, + "step": 3240 + }, + { + "epoch": 1.2393881453154876, + "grad_norm": 7.655797579107328, + "learning_rate": 1e-06, + "loss": 0.3263, + "step": 3241 + }, + { + "epoch": 1.2397705544933078, + "grad_norm": 3.57863483378223, + "learning_rate": 1e-06, + "loss": 0.3405, + "step": 3242 + }, + { + "epoch": 1.2401529636711282, + "grad_norm": 3.602513781381543, + "learning_rate": 1e-06, + "loss": 0.4766, + "step": 3243 + }, + { + "epoch": 1.2405353728489483, + "grad_norm": 2.966494375933703, + "learning_rate": 1e-06, + "loss": 0.3667, + "step": 3244 + }, + { + "epoch": 1.2409177820267687, + "grad_norm": 3.3422614376098685, + "learning_rate": 1e-06, + "loss": 0.3988, + "step": 3245 + }, + { + "epoch": 1.241300191204589, + "grad_norm": 2.5268776889874465, + "learning_rate": 1e-06, + "loss": 0.1166, + "step": 3246 + }, + { + "epoch": 1.241682600382409, + "grad_norm": 4.3453519872183275, + "learning_rate": 1e-06, + "loss": 0.466, + "step": 3247 + }, + { + "epoch": 1.2420650095602295, + "grad_norm": 3.199253179902874, + "learning_rate": 1e-06, + "loss": 0.1567, + "step": 3248 + }, + { + "epoch": 1.2424474187380496, + "grad_norm": 2.5782818370855645, + "learning_rate": 1e-06, + "loss": 0.2876, + "step": 3249 + }, + { + "epoch": 1.24282982791587, + "grad_norm": 2.9079991481102874, + "learning_rate": 1e-06, + "loss": 0.2807, + "step": 3250 + }, + { + "epoch": 1.2432122370936902, + "grad_norm": 3.247128905721344, + "learning_rate": 1e-06, + "loss": 0.1605, + "step": 3251 + }, + { + "epoch": 1.2435946462715104, + "grad_norm": 3.7079759204086757, + "learning_rate": 1e-06, + "loss": 0.0931, + "step": 3252 + }, + { + "epoch": 1.2439770554493308, + "grad_norm": 5.32563826626586, + "learning_rate": 1e-06, + "loss": 0.1665, + "step": 3253 + }, + { + "epoch": 1.244359464627151, + "grad_norm": 4.559249347003769, + "learning_rate": 1e-06, + "loss": 0.4295, + "step": 3254 + }, + { + "epoch": 1.2447418738049714, + "grad_norm": 2.8274031081323483, + "learning_rate": 1e-06, + "loss": 0.3379, + "step": 3255 + }, + { + "epoch": 1.2451242829827915, + "grad_norm": 2.539665925140947, + "learning_rate": 1e-06, + "loss": 0.2315, + "step": 3256 + }, + { + "epoch": 1.245506692160612, + "grad_norm": 2.1872441823421767, + "learning_rate": 1e-06, + "loss": 0.2804, + "step": 3257 + }, + { + "epoch": 1.2458891013384321, + "grad_norm": 2.984422293882462, + "learning_rate": 1e-06, + "loss": 0.2912, + "step": 3258 + }, + { + "epoch": 1.2462715105162525, + "grad_norm": 3.5871001708627017, + "learning_rate": 1e-06, + "loss": 0.2893, + "step": 3259 + }, + { + "epoch": 1.2466539196940727, + "grad_norm": 2.5731002392062807, + "learning_rate": 1e-06, + "loss": 0.1797, + "step": 3260 + }, + { + "epoch": 1.2470363288718929, + "grad_norm": 2.6045865648618096, + "learning_rate": 1e-06, + "loss": 0.1515, + "step": 3261 + }, + { + "epoch": 1.2474187380497133, + "grad_norm": 2.938303898321132, + "learning_rate": 1e-06, + "loss": 0.1139, + "step": 3262 + }, + { + "epoch": 1.2478011472275334, + "grad_norm": 3.1936588998481197, + "learning_rate": 1e-06, + "loss": 0.0828, + "step": 3263 + }, + { + "epoch": 1.2481835564053538, + "grad_norm": 3.4553206009902238, + "learning_rate": 1e-06, + "loss": 0.1139, + "step": 3264 + }, + { + "epoch": 1.248565965583174, + "grad_norm": 3.466891779070712, + "learning_rate": 1e-06, + "loss": 0.0688, + "step": 3265 + }, + { + "epoch": 1.2489483747609942, + "grad_norm": 5.993548421483226, + "learning_rate": 1e-06, + "loss": 0.4569, + "step": 3266 + }, + { + "epoch": 1.2493307839388146, + "grad_norm": 3.3922338931488256, + "learning_rate": 1e-06, + "loss": 0.6279, + "step": 3267 + }, + { + "epoch": 1.2497131931166348, + "grad_norm": 2.6383454906627724, + "learning_rate": 1e-06, + "loss": 0.2484, + "step": 3268 + }, + { + "epoch": 1.2500956022944552, + "grad_norm": 4.660954908184366, + "learning_rate": 1e-06, + "loss": 0.3153, + "step": 3269 + }, + { + "epoch": 1.2504780114722753, + "grad_norm": 3.2982163435081064, + "learning_rate": 1e-06, + "loss": 0.2657, + "step": 3270 + }, + { + "epoch": 1.2508604206500955, + "grad_norm": 3.52731657809821, + "learning_rate": 1e-06, + "loss": 0.3095, + "step": 3271 + }, + { + "epoch": 1.251242829827916, + "grad_norm": 2.9784126839280836, + "learning_rate": 1e-06, + "loss": 0.1788, + "step": 3272 + }, + { + "epoch": 1.251625239005736, + "grad_norm": 3.3139081157315933, + "learning_rate": 1e-06, + "loss": 0.1776, + "step": 3273 + }, + { + "epoch": 1.2520076481835565, + "grad_norm": 3.064465106383923, + "learning_rate": 1e-06, + "loss": 0.1641, + "step": 3274 + }, + { + "epoch": 1.2523900573613767, + "grad_norm": 3.0953936593643636, + "learning_rate": 1e-06, + "loss": 0.0889, + "step": 3275 + }, + { + "epoch": 1.2527724665391968, + "grad_norm": 2.613790738970916, + "learning_rate": 1e-06, + "loss": 0.0905, + "step": 3276 + }, + { + "epoch": 1.2531548757170172, + "grad_norm": 2.995440037518959, + "learning_rate": 1e-06, + "loss": 0.063, + "step": 3277 + }, + { + "epoch": 1.2535372848948374, + "grad_norm": 4.153778029423889, + "learning_rate": 1e-06, + "loss": 0.1795, + "step": 3278 + }, + { + "epoch": 1.2539196940726578, + "grad_norm": 3.231124497711958, + "learning_rate": 1e-06, + "loss": 0.3398, + "step": 3279 + }, + { + "epoch": 1.254302103250478, + "grad_norm": 2.467408793733926, + "learning_rate": 1e-06, + "loss": 0.3896, + "step": 3280 + }, + { + "epoch": 1.2546845124282981, + "grad_norm": 2.5752837274770437, + "learning_rate": 1e-06, + "loss": 0.3662, + "step": 3281 + }, + { + "epoch": 1.2550669216061185, + "grad_norm": 3.266375587755585, + "learning_rate": 1e-06, + "loss": 0.3, + "step": 3282 + }, + { + "epoch": 1.255449330783939, + "grad_norm": 3.6222765985423666, + "learning_rate": 1e-06, + "loss": 0.2978, + "step": 3283 + }, + { + "epoch": 1.2558317399617591, + "grad_norm": 4.070443705898735, + "learning_rate": 1e-06, + "loss": 0.2446, + "step": 3284 + }, + { + "epoch": 1.2562141491395793, + "grad_norm": 2.598440960980999, + "learning_rate": 1e-06, + "loss": 0.1429, + "step": 3285 + }, + { + "epoch": 1.2565965583173997, + "grad_norm": 2.6561142830678044, + "learning_rate": 1e-06, + "loss": 0.1296, + "step": 3286 + }, + { + "epoch": 1.2569789674952199, + "grad_norm": 2.938971293046691, + "learning_rate": 1e-06, + "loss": 0.1215, + "step": 3287 + }, + { + "epoch": 1.2573613766730403, + "grad_norm": 4.553445270404203, + "learning_rate": 1e-06, + "loss": 0.2336, + "step": 3288 + }, + { + "epoch": 1.2577437858508604, + "grad_norm": 3.817733674357759, + "learning_rate": 1e-06, + "loss": 0.1953, + "step": 3289 + }, + { + "epoch": 1.2581261950286806, + "grad_norm": 4.205942110353462, + "learning_rate": 1e-06, + "loss": 0.1186, + "step": 3290 + }, + { + "epoch": 1.258508604206501, + "grad_norm": 5.425228961384696, + "learning_rate": 1e-06, + "loss": 0.29, + "step": 3291 + }, + { + "epoch": 1.2588910133843212, + "grad_norm": 3.728292707330749, + "learning_rate": 1e-06, + "loss": 0.3458, + "step": 3292 + }, + { + "epoch": 1.2592734225621416, + "grad_norm": 4.7855711714243245, + "learning_rate": 1e-06, + "loss": 0.7093, + "step": 3293 + }, + { + "epoch": 1.2596558317399618, + "grad_norm": 3.4524078314851363, + "learning_rate": 1e-06, + "loss": 0.3338, + "step": 3294 + }, + { + "epoch": 1.260038240917782, + "grad_norm": 4.576692011893362, + "learning_rate": 1e-06, + "loss": 0.5975, + "step": 3295 + }, + { + "epoch": 1.2604206500956023, + "grad_norm": 3.2486478486977677, + "learning_rate": 1e-06, + "loss": 0.2046, + "step": 3296 + }, + { + "epoch": 1.2608030592734225, + "grad_norm": 2.8974439866689035, + "learning_rate": 1e-06, + "loss": 0.1111, + "step": 3297 + }, + { + "epoch": 1.261185468451243, + "grad_norm": 2.3431041590632984, + "learning_rate": 1e-06, + "loss": 0.0771, + "step": 3298 + }, + { + "epoch": 1.261567877629063, + "grad_norm": 3.0383392493238692, + "learning_rate": 1e-06, + "loss": 0.2024, + "step": 3299 + }, + { + "epoch": 1.2619502868068833, + "grad_norm": 2.949298296733519, + "learning_rate": 1e-06, + "loss": 0.1405, + "step": 3300 + }, + { + "epoch": 1.2623326959847037, + "grad_norm": 3.538927088287955, + "learning_rate": 1e-06, + "loss": 0.1302, + "step": 3301 + }, + { + "epoch": 1.2627151051625238, + "grad_norm": 3.601970260607489, + "learning_rate": 1e-06, + "loss": 0.0826, + "step": 3302 + }, + { + "epoch": 1.2630975143403442, + "grad_norm": 5.1360362248554425, + "learning_rate": 1e-06, + "loss": 0.1543, + "step": 3303 + }, + { + "epoch": 1.2634799235181644, + "grad_norm": 4.114321672191617, + "learning_rate": 1e-06, + "loss": 0.2817, + "step": 3304 + }, + { + "epoch": 1.2638623326959846, + "grad_norm": 2.8486934076088635, + "learning_rate": 1e-06, + "loss": 0.2746, + "step": 3305 + }, + { + "epoch": 1.264244741873805, + "grad_norm": 2.444121927120025, + "learning_rate": 1e-06, + "loss": 0.2642, + "step": 3306 + }, + { + "epoch": 1.2646271510516254, + "grad_norm": 3.2443438976709884, + "learning_rate": 1e-06, + "loss": 0.6061, + "step": 3307 + }, + { + "epoch": 1.2650095602294456, + "grad_norm": 2.604470492123503, + "learning_rate": 1e-06, + "loss": 0.3553, + "step": 3308 + }, + { + "epoch": 1.2653919694072657, + "grad_norm": 3.1613936106023286, + "learning_rate": 1e-06, + "loss": 0.4285, + "step": 3309 + }, + { + "epoch": 1.2657743785850861, + "grad_norm": 2.5794572566869, + "learning_rate": 1e-06, + "loss": 0.1728, + "step": 3310 + }, + { + "epoch": 1.2661567877629063, + "grad_norm": 2.5218372290145044, + "learning_rate": 1e-06, + "loss": 0.1966, + "step": 3311 + }, + { + "epoch": 1.2665391969407267, + "grad_norm": 4.796943390697319, + "learning_rate": 1e-06, + "loss": 0.1027, + "step": 3312 + }, + { + "epoch": 1.2669216061185469, + "grad_norm": 2.730988503514679, + "learning_rate": 1e-06, + "loss": 0.1398, + "step": 3313 + }, + { + "epoch": 1.267304015296367, + "grad_norm": 3.111253487265852, + "learning_rate": 1e-06, + "loss": 0.0879, + "step": 3314 + }, + { + "epoch": 1.2676864244741874, + "grad_norm": 6.7800027429805825, + "learning_rate": 1e-06, + "loss": 0.156, + "step": 3315 + }, + { + "epoch": 1.2680688336520076, + "grad_norm": 7.963474337432703, + "learning_rate": 1e-06, + "loss": 0.495, + "step": 3316 + }, + { + "epoch": 1.268451242829828, + "grad_norm": 3.024389151697539, + "learning_rate": 1e-06, + "loss": 0.3465, + "step": 3317 + }, + { + "epoch": 1.2688336520076482, + "grad_norm": 2.860452084323208, + "learning_rate": 1e-06, + "loss": 0.3293, + "step": 3318 + }, + { + "epoch": 1.2692160611854684, + "grad_norm": 3.6980028742381386, + "learning_rate": 1e-06, + "loss": 0.2159, + "step": 3319 + }, + { + "epoch": 1.2695984703632888, + "grad_norm": 2.303032335671375, + "learning_rate": 1e-06, + "loss": 0.4376, + "step": 3320 + }, + { + "epoch": 1.269980879541109, + "grad_norm": 4.182387234113338, + "learning_rate": 1e-06, + "loss": 0.2782, + "step": 3321 + }, + { + "epoch": 1.2703632887189293, + "grad_norm": 2.692626676421947, + "learning_rate": 1e-06, + "loss": 0.1394, + "step": 3322 + }, + { + "epoch": 1.2707456978967495, + "grad_norm": 3.912974924141223, + "learning_rate": 1e-06, + "loss": 0.1964, + "step": 3323 + }, + { + "epoch": 1.2711281070745697, + "grad_norm": 3.876724474720001, + "learning_rate": 1e-06, + "loss": 0.1456, + "step": 3324 + }, + { + "epoch": 1.27151051625239, + "grad_norm": 4.701569055848703, + "learning_rate": 1e-06, + "loss": 0.2936, + "step": 3325 + }, + { + "epoch": 1.2718929254302103, + "grad_norm": 2.748696972178061, + "learning_rate": 1e-06, + "loss": 0.1832, + "step": 3326 + }, + { + "epoch": 1.2722753346080307, + "grad_norm": 3.8873667114565897, + "learning_rate": 1e-06, + "loss": 0.0835, + "step": 3327 + }, + { + "epoch": 1.2726577437858508, + "grad_norm": 5.0648131560708345, + "learning_rate": 1e-06, + "loss": 0.1069, + "step": 3328 + }, + { + "epoch": 1.273040152963671, + "grad_norm": 3.3261159644632983, + "learning_rate": 1e-06, + "loss": 0.3947, + "step": 3329 + }, + { + "epoch": 1.2734225621414914, + "grad_norm": 3.5728602428077467, + "learning_rate": 1e-06, + "loss": 0.5163, + "step": 3330 + }, + { + "epoch": 1.2738049713193116, + "grad_norm": 2.1417360415769147, + "learning_rate": 1e-06, + "loss": 0.2434, + "step": 3331 + }, + { + "epoch": 1.274187380497132, + "grad_norm": 2.986434465589483, + "learning_rate": 1e-06, + "loss": 0.2235, + "step": 3332 + }, + { + "epoch": 1.2745697896749522, + "grad_norm": 3.243911762650881, + "learning_rate": 1e-06, + "loss": 0.224, + "step": 3333 + }, + { + "epoch": 1.2749521988527723, + "grad_norm": 3.4212353252170034, + "learning_rate": 1e-06, + "loss": 0.2809, + "step": 3334 + }, + { + "epoch": 1.2753346080305927, + "grad_norm": 3.4021436104668235, + "learning_rate": 1e-06, + "loss": 0.181, + "step": 3335 + }, + { + "epoch": 1.2757170172084131, + "grad_norm": 3.3884305400824406, + "learning_rate": 1e-06, + "loss": 0.1349, + "step": 3336 + }, + { + "epoch": 1.2760994263862333, + "grad_norm": 2.9425503758710017, + "learning_rate": 1e-06, + "loss": 0.129, + "step": 3337 + }, + { + "epoch": 1.2764818355640535, + "grad_norm": 4.610561205285702, + "learning_rate": 1e-06, + "loss": 0.3488, + "step": 3338 + }, + { + "epoch": 1.2768642447418739, + "grad_norm": 4.070022426234139, + "learning_rate": 1e-06, + "loss": 0.0708, + "step": 3339 + }, + { + "epoch": 1.277246653919694, + "grad_norm": 3.9333845156771727, + "learning_rate": 1e-06, + "loss": 0.0555, + "step": 3340 + }, + { + "epoch": 1.2776290630975145, + "grad_norm": 5.3807590592963015, + "learning_rate": 1e-06, + "loss": 0.2507, + "step": 3341 + }, + { + "epoch": 1.2780114722753346, + "grad_norm": 3.451388928118952, + "learning_rate": 1e-06, + "loss": 0.3081, + "step": 3342 + }, + { + "epoch": 1.2783938814531548, + "grad_norm": 2.7262223258312663, + "learning_rate": 1e-06, + "loss": 0.2197, + "step": 3343 + }, + { + "epoch": 1.2787762906309752, + "grad_norm": 2.927656848331449, + "learning_rate": 1e-06, + "loss": 0.259, + "step": 3344 + }, + { + "epoch": 1.2791586998087954, + "grad_norm": 2.279239134069402, + "learning_rate": 1e-06, + "loss": 0.3078, + "step": 3345 + }, + { + "epoch": 1.2795411089866158, + "grad_norm": 3.1295168467792123, + "learning_rate": 1e-06, + "loss": 0.3413, + "step": 3346 + }, + { + "epoch": 1.279923518164436, + "grad_norm": 4.040598357806839, + "learning_rate": 1e-06, + "loss": 0.3243, + "step": 3347 + }, + { + "epoch": 1.2803059273422561, + "grad_norm": 3.267846333041392, + "learning_rate": 1e-06, + "loss": 0.1141, + "step": 3348 + }, + { + "epoch": 1.2806883365200765, + "grad_norm": 2.120686135370677, + "learning_rate": 1e-06, + "loss": 0.1233, + "step": 3349 + }, + { + "epoch": 1.2810707456978967, + "grad_norm": 4.662781505932946, + "learning_rate": 1e-06, + "loss": 0.2845, + "step": 3350 + }, + { + "epoch": 1.281453154875717, + "grad_norm": 2.194854797218163, + "learning_rate": 1e-06, + "loss": 0.0676, + "step": 3351 + }, + { + "epoch": 1.2818355640535373, + "grad_norm": 3.776868567416281, + "learning_rate": 1e-06, + "loss": 0.1338, + "step": 3352 + }, + { + "epoch": 1.2822179732313574, + "grad_norm": 5.714445568301108, + "learning_rate": 1e-06, + "loss": 0.2475, + "step": 3353 + }, + { + "epoch": 1.2826003824091778, + "grad_norm": 6.035578148177992, + "learning_rate": 1e-06, + "loss": 0.509, + "step": 3354 + }, + { + "epoch": 1.282982791586998, + "grad_norm": 4.109563902947867, + "learning_rate": 1e-06, + "loss": 0.4684, + "step": 3355 + }, + { + "epoch": 1.2833652007648184, + "grad_norm": 3.3846014987887463, + "learning_rate": 1e-06, + "loss": 0.1594, + "step": 3356 + }, + { + "epoch": 1.2837476099426386, + "grad_norm": 3.190726349030799, + "learning_rate": 1e-06, + "loss": 0.3597, + "step": 3357 + }, + { + "epoch": 1.2841300191204588, + "grad_norm": 3.3777993919562195, + "learning_rate": 1e-06, + "loss": 0.2764, + "step": 3358 + }, + { + "epoch": 1.2845124282982792, + "grad_norm": 4.926388563188322, + "learning_rate": 1e-06, + "loss": 0.2072, + "step": 3359 + }, + { + "epoch": 1.2848948374760996, + "grad_norm": 2.6804195549635703, + "learning_rate": 1e-06, + "loss": 0.1778, + "step": 3360 + }, + { + "epoch": 1.2852772466539197, + "grad_norm": 2.942750499502011, + "learning_rate": 1e-06, + "loss": 0.1543, + "step": 3361 + }, + { + "epoch": 1.28565965583174, + "grad_norm": 3.2003028905094935, + "learning_rate": 1e-06, + "loss": 0.2225, + "step": 3362 + }, + { + "epoch": 1.2860420650095603, + "grad_norm": 2.9642325312822067, + "learning_rate": 1e-06, + "loss": 0.1231, + "step": 3363 + }, + { + "epoch": 1.2864244741873805, + "grad_norm": 4.029409062150022, + "learning_rate": 1e-06, + "loss": 0.0952, + "step": 3364 + }, + { + "epoch": 1.2868068833652009, + "grad_norm": 3.274000315205367, + "learning_rate": 1e-06, + "loss": 0.0935, + "step": 3365 + }, + { + "epoch": 1.287189292543021, + "grad_norm": 4.361271490129927, + "learning_rate": 1e-06, + "loss": 0.374, + "step": 3366 + }, + { + "epoch": 1.2875717017208412, + "grad_norm": 3.3631651323324023, + "learning_rate": 1e-06, + "loss": 0.1535, + "step": 3367 + }, + { + "epoch": 1.2879541108986616, + "grad_norm": 4.472950275201895, + "learning_rate": 1e-06, + "loss": 0.6948, + "step": 3368 + }, + { + "epoch": 1.2883365200764818, + "grad_norm": 3.383041409330677, + "learning_rate": 1e-06, + "loss": 0.2226, + "step": 3369 + }, + { + "epoch": 1.2887189292543022, + "grad_norm": 3.066936782329665, + "learning_rate": 1e-06, + "loss": 0.2695, + "step": 3370 + }, + { + "epoch": 1.2891013384321224, + "grad_norm": 3.4226682257426444, + "learning_rate": 1e-06, + "loss": 0.2786, + "step": 3371 + }, + { + "epoch": 1.2894837476099426, + "grad_norm": 3.862396169705139, + "learning_rate": 1e-06, + "loss": 0.3995, + "step": 3372 + }, + { + "epoch": 1.289866156787763, + "grad_norm": 2.3260833185284504, + "learning_rate": 1e-06, + "loss": 0.1609, + "step": 3373 + }, + { + "epoch": 1.2902485659655831, + "grad_norm": 3.2342442287969453, + "learning_rate": 1e-06, + "loss": 0.1511, + "step": 3374 + }, + { + "epoch": 1.2906309751434035, + "grad_norm": 4.072343842923876, + "learning_rate": 1e-06, + "loss": 0.3727, + "step": 3375 + }, + { + "epoch": 1.2910133843212237, + "grad_norm": 3.900304435194107, + "learning_rate": 1e-06, + "loss": 0.1991, + "step": 3376 + }, + { + "epoch": 1.2913957934990439, + "grad_norm": 4.229406841705585, + "learning_rate": 1e-06, + "loss": 0.1204, + "step": 3377 + }, + { + "epoch": 1.2917782026768643, + "grad_norm": 5.929072774151254, + "learning_rate": 1e-06, + "loss": 0.2068, + "step": 3378 + }, + { + "epoch": 1.2921606118546844, + "grad_norm": 5.857656974691698, + "learning_rate": 1e-06, + "loss": 0.3528, + "step": 3379 + }, + { + "epoch": 1.2925430210325048, + "grad_norm": 3.4854770166462634, + "learning_rate": 1e-06, + "loss": 0.3787, + "step": 3380 + }, + { + "epoch": 1.292925430210325, + "grad_norm": 2.4692164113285253, + "learning_rate": 1e-06, + "loss": 0.2829, + "step": 3381 + }, + { + "epoch": 1.2933078393881452, + "grad_norm": 2.5575743063892857, + "learning_rate": 1e-06, + "loss": 0.2287, + "step": 3382 + }, + { + "epoch": 1.2936902485659656, + "grad_norm": 2.9849181954118578, + "learning_rate": 1e-06, + "loss": 0.2854, + "step": 3383 + }, + { + "epoch": 1.2940726577437858, + "grad_norm": 3.009305826027895, + "learning_rate": 1e-06, + "loss": 0.1549, + "step": 3384 + }, + { + "epoch": 1.2944550669216062, + "grad_norm": 2.844648785089339, + "learning_rate": 1e-06, + "loss": 0.2272, + "step": 3385 + }, + { + "epoch": 1.2948374760994263, + "grad_norm": 2.9873070818170366, + "learning_rate": 1e-06, + "loss": 0.1596, + "step": 3386 + }, + { + "epoch": 1.2952198852772465, + "grad_norm": 3.7417993523421957, + "learning_rate": 1e-06, + "loss": 0.2825, + "step": 3387 + }, + { + "epoch": 1.295602294455067, + "grad_norm": 3.6106958655900474, + "learning_rate": 1e-06, + "loss": 0.1479, + "step": 3388 + }, + { + "epoch": 1.2959847036328873, + "grad_norm": 5.045172720184043, + "learning_rate": 1e-06, + "loss": 0.1158, + "step": 3389 + }, + { + "epoch": 1.2963671128107075, + "grad_norm": 3.4616433967269815, + "learning_rate": 1e-06, + "loss": 0.0986, + "step": 3390 + }, + { + "epoch": 1.2967495219885277, + "grad_norm": 6.064327004022685, + "learning_rate": 1e-06, + "loss": 0.4425, + "step": 3391 + }, + { + "epoch": 1.297131931166348, + "grad_norm": 4.174517876661145, + "learning_rate": 1e-06, + "loss": 0.3251, + "step": 3392 + }, + { + "epoch": 1.2975143403441682, + "grad_norm": 3.382832234173097, + "learning_rate": 1e-06, + "loss": 0.2373, + "step": 3393 + }, + { + "epoch": 1.2978967495219886, + "grad_norm": 2.4883105696982386, + "learning_rate": 1e-06, + "loss": 0.1699, + "step": 3394 + }, + { + "epoch": 1.2982791586998088, + "grad_norm": 3.10111692672774, + "learning_rate": 1e-06, + "loss": 0.2498, + "step": 3395 + }, + { + "epoch": 1.298661567877629, + "grad_norm": 3.279655577789128, + "learning_rate": 1e-06, + "loss": 0.3954, + "step": 3396 + }, + { + "epoch": 1.2990439770554494, + "grad_norm": 3.242066898171103, + "learning_rate": 1e-06, + "loss": 0.3368, + "step": 3397 + }, + { + "epoch": 1.2994263862332696, + "grad_norm": 3.476942954516217, + "learning_rate": 1e-06, + "loss": 0.1567, + "step": 3398 + }, + { + "epoch": 1.29980879541109, + "grad_norm": 3.3651740123540197, + "learning_rate": 1e-06, + "loss": 0.1099, + "step": 3399 + }, + { + "epoch": 1.3001912045889101, + "grad_norm": 5.068972182666567, + "learning_rate": 1e-06, + "loss": 0.2837, + "step": 3400 + }, + { + "epoch": 1.3005736137667303, + "grad_norm": 2.556185874974078, + "learning_rate": 1e-06, + "loss": 0.0851, + "step": 3401 + }, + { + "epoch": 1.3009560229445507, + "grad_norm": 4.363005845437778, + "learning_rate": 1e-06, + "loss": 0.0745, + "step": 3402 + }, + { + "epoch": 1.3013384321223709, + "grad_norm": 6.656372355737953, + "learning_rate": 1e-06, + "loss": 0.1691, + "step": 3403 + }, + { + "epoch": 1.3017208413001913, + "grad_norm": 4.121266380418662, + "learning_rate": 1e-06, + "loss": 0.4311, + "step": 3404 + }, + { + "epoch": 1.3021032504780115, + "grad_norm": 2.798558075062855, + "learning_rate": 1e-06, + "loss": 0.2609, + "step": 3405 + }, + { + "epoch": 1.3024856596558316, + "grad_norm": 1.912774411629172, + "learning_rate": 1e-06, + "loss": 0.1919, + "step": 3406 + }, + { + "epoch": 1.302868068833652, + "grad_norm": 2.9925983037394017, + "learning_rate": 1e-06, + "loss": 0.2651, + "step": 3407 + }, + { + "epoch": 1.3032504780114722, + "grad_norm": 3.340020069130547, + "learning_rate": 1e-06, + "loss": 0.3107, + "step": 3408 + }, + { + "epoch": 1.3036328871892926, + "grad_norm": 3.84055772665031, + "learning_rate": 1e-06, + "loss": 0.2469, + "step": 3409 + }, + { + "epoch": 1.3040152963671128, + "grad_norm": 3.473565959742894, + "learning_rate": 1e-06, + "loss": 0.2675, + "step": 3410 + }, + { + "epoch": 1.304397705544933, + "grad_norm": 2.653459137453321, + "learning_rate": 1e-06, + "loss": 0.1012, + "step": 3411 + }, + { + "epoch": 1.3047801147227533, + "grad_norm": 2.996474101407398, + "learning_rate": 1e-06, + "loss": 0.1056, + "step": 3412 + }, + { + "epoch": 1.3051625239005737, + "grad_norm": 2.8112490097050937, + "learning_rate": 1e-06, + "loss": 0.0976, + "step": 3413 + }, + { + "epoch": 1.305544933078394, + "grad_norm": 3.459583439530436, + "learning_rate": 1e-06, + "loss": 0.0902, + "step": 3414 + }, + { + "epoch": 1.305927342256214, + "grad_norm": 6.510320539654921, + "learning_rate": 1e-06, + "loss": 0.2568, + "step": 3415 + }, + { + "epoch": 1.3063097514340345, + "grad_norm": 5.102605318798301, + "learning_rate": 1e-06, + "loss": 0.4318, + "step": 3416 + }, + { + "epoch": 1.3066921606118547, + "grad_norm": 3.021725191902132, + "learning_rate": 1e-06, + "loss": 0.2025, + "step": 3417 + }, + { + "epoch": 1.307074569789675, + "grad_norm": 3.7599091896291608, + "learning_rate": 1e-06, + "loss": 0.2958, + "step": 3418 + }, + { + "epoch": 1.3074569789674952, + "grad_norm": 3.176630523598782, + "learning_rate": 1e-06, + "loss": 0.2994, + "step": 3419 + }, + { + "epoch": 1.3078393881453154, + "grad_norm": 2.505285683519663, + "learning_rate": 1e-06, + "loss": 0.205, + "step": 3420 + }, + { + "epoch": 1.3082217973231358, + "grad_norm": 2.6696898688174446, + "learning_rate": 1e-06, + "loss": 0.2454, + "step": 3421 + }, + { + "epoch": 1.308604206500956, + "grad_norm": 1.8595218760718364, + "learning_rate": 1e-06, + "loss": 0.1584, + "step": 3422 + }, + { + "epoch": 1.3089866156787764, + "grad_norm": 4.18426183500529, + "learning_rate": 1e-06, + "loss": 0.2838, + "step": 3423 + }, + { + "epoch": 1.3093690248565966, + "grad_norm": 2.7474685635213083, + "learning_rate": 1e-06, + "loss": 0.1561, + "step": 3424 + }, + { + "epoch": 1.3097514340344167, + "grad_norm": 3.43407949715755, + "learning_rate": 1e-06, + "loss": 0.1478, + "step": 3425 + }, + { + "epoch": 1.3101338432122371, + "grad_norm": 2.822721429996645, + "learning_rate": 1e-06, + "loss": 0.0598, + "step": 3426 + }, + { + "epoch": 1.3105162523900573, + "grad_norm": 4.008866258962873, + "learning_rate": 1e-06, + "loss": 0.1668, + "step": 3427 + }, + { + "epoch": 1.3108986615678777, + "grad_norm": 5.787412256354303, + "learning_rate": 1e-06, + "loss": 0.2285, + "step": 3428 + }, + { + "epoch": 1.3112810707456979, + "grad_norm": 3.2702141598908003, + "learning_rate": 1e-06, + "loss": 0.4435, + "step": 3429 + }, + { + "epoch": 1.311663479923518, + "grad_norm": 2.8847752120905352, + "learning_rate": 1e-06, + "loss": 0.1651, + "step": 3430 + }, + { + "epoch": 1.3120458891013385, + "grad_norm": 2.979687909674854, + "learning_rate": 1e-06, + "loss": 0.3365, + "step": 3431 + }, + { + "epoch": 1.3124282982791586, + "grad_norm": 3.408940363941151, + "learning_rate": 1e-06, + "loss": 0.2235, + "step": 3432 + }, + { + "epoch": 1.312810707456979, + "grad_norm": 2.310862709057707, + "learning_rate": 1e-06, + "loss": 0.2305, + "step": 3433 + }, + { + "epoch": 1.3131931166347992, + "grad_norm": 3.007345425889769, + "learning_rate": 1e-06, + "loss": 0.2149, + "step": 3434 + }, + { + "epoch": 1.3135755258126194, + "grad_norm": 2.668927803638142, + "learning_rate": 1e-06, + "loss": 0.3073, + "step": 3435 + }, + { + "epoch": 1.3139579349904398, + "grad_norm": 2.104125321488783, + "learning_rate": 1e-06, + "loss": 0.1248, + "step": 3436 + }, + { + "epoch": 1.31434034416826, + "grad_norm": 2.6199810548636107, + "learning_rate": 1e-06, + "loss": 0.1346, + "step": 3437 + }, + { + "epoch": 1.3147227533460804, + "grad_norm": 5.242819462303336, + "learning_rate": 1e-06, + "loss": 0.1127, + "step": 3438 + }, + { + "epoch": 1.3151051625239005, + "grad_norm": 2.5933172025726394, + "learning_rate": 1e-06, + "loss": 0.0724, + "step": 3439 + }, + { + "epoch": 1.3154875717017207, + "grad_norm": 3.417160339666154, + "learning_rate": 1e-06, + "loss": 0.1142, + "step": 3440 + }, + { + "epoch": 1.315869980879541, + "grad_norm": 5.838230729366253, + "learning_rate": 1e-06, + "loss": 0.1927, + "step": 3441 + }, + { + "epoch": 1.3162523900573615, + "grad_norm": 3.554357089941153, + "learning_rate": 1e-06, + "loss": 0.2933, + "step": 3442 + }, + { + "epoch": 1.3166347992351817, + "grad_norm": 3.5519096503580494, + "learning_rate": 1e-06, + "loss": 0.319, + "step": 3443 + }, + { + "epoch": 1.3170172084130018, + "grad_norm": 2.831211155504948, + "learning_rate": 1e-06, + "loss": 0.2755, + "step": 3444 + }, + { + "epoch": 1.3173996175908222, + "grad_norm": 3.2813162660719373, + "learning_rate": 1e-06, + "loss": 0.2042, + "step": 3445 + }, + { + "epoch": 1.3177820267686424, + "grad_norm": 3.371527404121634, + "learning_rate": 1e-06, + "loss": 0.2246, + "step": 3446 + }, + { + "epoch": 1.3181644359464628, + "grad_norm": 2.7169618535907323, + "learning_rate": 1e-06, + "loss": 0.1714, + "step": 3447 + }, + { + "epoch": 1.318546845124283, + "grad_norm": 2.2601980522607117, + "learning_rate": 1e-06, + "loss": 0.0768, + "step": 3448 + }, + { + "epoch": 1.3189292543021032, + "grad_norm": 2.448373846554622, + "learning_rate": 1e-06, + "loss": 0.1442, + "step": 3449 + }, + { + "epoch": 1.3193116634799236, + "grad_norm": 4.5222297070623805, + "learning_rate": 1e-06, + "loss": 0.1773, + "step": 3450 + }, + { + "epoch": 1.3196940726577437, + "grad_norm": 2.897869866818496, + "learning_rate": 1e-06, + "loss": 0.08, + "step": 3451 + }, + { + "epoch": 1.3200764818355641, + "grad_norm": 3.647040743032006, + "learning_rate": 1e-06, + "loss": 0.1009, + "step": 3452 + }, + { + "epoch": 1.3204588910133843, + "grad_norm": 4.816123490027161, + "learning_rate": 1e-06, + "loss": 0.1705, + "step": 3453 + }, + { + "epoch": 1.3208413001912045, + "grad_norm": 4.641099693603532, + "learning_rate": 1e-06, + "loss": 0.5515, + "step": 3454 + }, + { + "epoch": 1.321223709369025, + "grad_norm": 4.405594797321751, + "learning_rate": 1e-06, + "loss": 0.4077, + "step": 3455 + }, + { + "epoch": 1.321606118546845, + "grad_norm": 4.125408730342099, + "learning_rate": 1e-06, + "loss": 0.3148, + "step": 3456 + }, + { + "epoch": 1.3219885277246655, + "grad_norm": 3.544837938055854, + "learning_rate": 1e-06, + "loss": 0.3524, + "step": 3457 + }, + { + "epoch": 1.3223709369024856, + "grad_norm": 3.337091393764473, + "learning_rate": 1e-06, + "loss": 0.4117, + "step": 3458 + }, + { + "epoch": 1.3227533460803058, + "grad_norm": 3.2285444254897344, + "learning_rate": 1e-06, + "loss": 0.2332, + "step": 3459 + }, + { + "epoch": 1.3231357552581262, + "grad_norm": 3.494300697886279, + "learning_rate": 1e-06, + "loss": 0.1988, + "step": 3460 + }, + { + "epoch": 1.3235181644359464, + "grad_norm": 3.257444511871968, + "learning_rate": 1e-06, + "loss": 0.204, + "step": 3461 + }, + { + "epoch": 1.3239005736137668, + "grad_norm": 3.4655272488098623, + "learning_rate": 1e-06, + "loss": 0.2455, + "step": 3462 + }, + { + "epoch": 1.324282982791587, + "grad_norm": 4.786664701654426, + "learning_rate": 1e-06, + "loss": 0.1595, + "step": 3463 + }, + { + "epoch": 1.3246653919694071, + "grad_norm": 3.0983968958115438, + "learning_rate": 1e-06, + "loss": 0.1049, + "step": 3464 + }, + { + "epoch": 1.3250478011472275, + "grad_norm": 4.1737502094827414, + "learning_rate": 1e-06, + "loss": 0.0904, + "step": 3465 + }, + { + "epoch": 1.325430210325048, + "grad_norm": 5.634086898572302, + "learning_rate": 1e-06, + "loss": 0.5333, + "step": 3466 + }, + { + "epoch": 1.325812619502868, + "grad_norm": 3.880113734515193, + "learning_rate": 1e-06, + "loss": 0.524, + "step": 3467 + }, + { + "epoch": 1.3261950286806883, + "grad_norm": 3.870828906256228, + "learning_rate": 1e-06, + "loss": 0.3742, + "step": 3468 + }, + { + "epoch": 1.3265774378585087, + "grad_norm": 2.1852502425555547, + "learning_rate": 1e-06, + "loss": 0.3189, + "step": 3469 + }, + { + "epoch": 1.3269598470363289, + "grad_norm": 3.554307317838956, + "learning_rate": 1e-06, + "loss": 0.1516, + "step": 3470 + }, + { + "epoch": 1.3273422562141493, + "grad_norm": 2.980956511376388, + "learning_rate": 1e-06, + "loss": 0.2794, + "step": 3471 + }, + { + "epoch": 1.3277246653919694, + "grad_norm": 2.975385938983526, + "learning_rate": 1e-06, + "loss": 0.1522, + "step": 3472 + }, + { + "epoch": 1.3281070745697896, + "grad_norm": 3.4128798612232973, + "learning_rate": 1e-06, + "loss": 0.1684, + "step": 3473 + }, + { + "epoch": 1.32848948374761, + "grad_norm": 3.013378037512129, + "learning_rate": 1e-06, + "loss": 0.1744, + "step": 3474 + }, + { + "epoch": 1.3288718929254302, + "grad_norm": 3.5179992796445525, + "learning_rate": 1e-06, + "loss": 0.2799, + "step": 3475 + }, + { + "epoch": 1.3292543021032506, + "grad_norm": 3.902518970118376, + "learning_rate": 1e-06, + "loss": 0.1032, + "step": 3476 + }, + { + "epoch": 1.3296367112810707, + "grad_norm": 2.980508586572323, + "learning_rate": 1e-06, + "loss": 0.0781, + "step": 3477 + }, + { + "epoch": 1.330019120458891, + "grad_norm": 4.325377325455937, + "learning_rate": 1e-06, + "loss": 0.1956, + "step": 3478 + }, + { + "epoch": 1.3304015296367113, + "grad_norm": 4.729012552626647, + "learning_rate": 1e-06, + "loss": 0.4777, + "step": 3479 + }, + { + "epoch": 1.3307839388145315, + "grad_norm": 3.6483533306337645, + "learning_rate": 1e-06, + "loss": 0.4687, + "step": 3480 + }, + { + "epoch": 1.331166347992352, + "grad_norm": 2.6517313501639608, + "learning_rate": 1e-06, + "loss": 0.2736, + "step": 3481 + }, + { + "epoch": 1.331548757170172, + "grad_norm": 2.55513038645958, + "learning_rate": 1e-06, + "loss": 0.1905, + "step": 3482 + }, + { + "epoch": 1.3319311663479922, + "grad_norm": 2.823382790757958, + "learning_rate": 1e-06, + "loss": 0.1812, + "step": 3483 + }, + { + "epoch": 1.3323135755258126, + "grad_norm": 3.1726080888231554, + "learning_rate": 1e-06, + "loss": 0.1316, + "step": 3484 + }, + { + "epoch": 1.3326959847036328, + "grad_norm": 3.4896985087200147, + "learning_rate": 1e-06, + "loss": 0.2107, + "step": 3485 + }, + { + "epoch": 1.3330783938814532, + "grad_norm": 3.3192770386801516, + "learning_rate": 1e-06, + "loss": 0.2316, + "step": 3486 + }, + { + "epoch": 1.3334608030592734, + "grad_norm": 3.5117125039050405, + "learning_rate": 1e-06, + "loss": 0.1562, + "step": 3487 + }, + { + "epoch": 1.3338432122370936, + "grad_norm": 4.913830386260947, + "learning_rate": 1e-06, + "loss": 0.0863, + "step": 3488 + }, + { + "epoch": 1.334225621414914, + "grad_norm": 2.962580645143604, + "learning_rate": 1e-06, + "loss": 0.0658, + "step": 3489 + }, + { + "epoch": 1.3346080305927344, + "grad_norm": 4.366650051832743, + "learning_rate": 1e-06, + "loss": 0.2274, + "step": 3490 + }, + { + "epoch": 1.3349904397705545, + "grad_norm": 5.46336980742746, + "learning_rate": 1e-06, + "loss": 0.3706, + "step": 3491 + }, + { + "epoch": 1.3353728489483747, + "grad_norm": 5.000078009950527, + "learning_rate": 1e-06, + "loss": 0.2676, + "step": 3492 + }, + { + "epoch": 1.335755258126195, + "grad_norm": 5.653906236505976, + "learning_rate": 1e-06, + "loss": 0.2413, + "step": 3493 + }, + { + "epoch": 1.3361376673040153, + "grad_norm": 3.2929180679855703, + "learning_rate": 1e-06, + "loss": 0.3646, + "step": 3494 + }, + { + "epoch": 1.3365200764818357, + "grad_norm": 3.7268240335047995, + "learning_rate": 1e-06, + "loss": 0.5141, + "step": 3495 + }, + { + "epoch": 1.3369024856596559, + "grad_norm": 3.816759575347373, + "learning_rate": 1e-06, + "loss": 0.225, + "step": 3496 + }, + { + "epoch": 1.337284894837476, + "grad_norm": 2.923530476823112, + "learning_rate": 1e-06, + "loss": 0.2098, + "step": 3497 + }, + { + "epoch": 1.3376673040152964, + "grad_norm": 2.427082984062028, + "learning_rate": 1e-06, + "loss": 0.1571, + "step": 3498 + }, + { + "epoch": 1.3380497131931166, + "grad_norm": 2.846076935543849, + "learning_rate": 1e-06, + "loss": 0.1271, + "step": 3499 + }, + { + "epoch": 1.338432122370937, + "grad_norm": 2.8243415774153005, + "learning_rate": 1e-06, + "loss": 0.1141, + "step": 3500 + }, + { + "epoch": 1.338432122370937, + "eval_runtime": 829.2431, + "eval_samples_per_second": 1.85, + "eval_steps_per_second": 0.463, + "step": 3500 + }, + { + "epoch": 1.3388145315487572, + "grad_norm": 3.799385904584203, + "learning_rate": 1e-06, + "loss": 0.1167, + "step": 3501 + }, + { + "epoch": 1.3391969407265774, + "grad_norm": 3.888971919732548, + "learning_rate": 1e-06, + "loss": 0.0944, + "step": 3502 + }, + { + "epoch": 1.3395793499043978, + "grad_norm": 3.5823894558746696, + "learning_rate": 1e-06, + "loss": 0.2139, + "step": 3503 + }, + { + "epoch": 1.339961759082218, + "grad_norm": 4.06310021294751, + "learning_rate": 1e-06, + "loss": 0.2905, + "step": 3504 + }, + { + "epoch": 1.3403441682600383, + "grad_norm": 3.404893237441377, + "learning_rate": 1e-06, + "loss": 0.2061, + "step": 3505 + }, + { + "epoch": 1.3407265774378585, + "grad_norm": 2.0965391342036903, + "learning_rate": 1e-06, + "loss": 0.286, + "step": 3506 + }, + { + "epoch": 1.3411089866156787, + "grad_norm": 2.977220037127773, + "learning_rate": 1e-06, + "loss": 0.1368, + "step": 3507 + }, + { + "epoch": 1.341491395793499, + "grad_norm": 3.686750174111812, + "learning_rate": 1e-06, + "loss": 0.345, + "step": 3508 + }, + { + "epoch": 1.3418738049713193, + "grad_norm": 3.084022565222293, + "learning_rate": 1e-06, + "loss": 0.2389, + "step": 3509 + }, + { + "epoch": 1.3422562141491396, + "grad_norm": 3.6780032699263425, + "learning_rate": 1e-06, + "loss": 0.2237, + "step": 3510 + }, + { + "epoch": 1.3426386233269598, + "grad_norm": 3.4518122184220217, + "learning_rate": 1e-06, + "loss": 0.1802, + "step": 3511 + }, + { + "epoch": 1.34302103250478, + "grad_norm": 3.9095442652174097, + "learning_rate": 1e-06, + "loss": 0.169, + "step": 3512 + }, + { + "epoch": 1.3434034416826004, + "grad_norm": 3.3225856763644526, + "learning_rate": 1e-06, + "loss": 0.0833, + "step": 3513 + }, + { + "epoch": 1.3437858508604206, + "grad_norm": 4.5201920905692194, + "learning_rate": 1e-06, + "loss": 0.0814, + "step": 3514 + }, + { + "epoch": 1.344168260038241, + "grad_norm": 4.5273921986999985, + "learning_rate": 1e-06, + "loss": 0.1546, + "step": 3515 + }, + { + "epoch": 1.3445506692160611, + "grad_norm": 5.1095078943917445, + "learning_rate": 1e-06, + "loss": 0.2919, + "step": 3516 + }, + { + "epoch": 1.3449330783938813, + "grad_norm": 3.249637730288223, + "learning_rate": 1e-06, + "loss": 0.5484, + "step": 3517 + }, + { + "epoch": 1.3453154875717017, + "grad_norm": 2.9656931347752957, + "learning_rate": 1e-06, + "loss": 0.269, + "step": 3518 + }, + { + "epoch": 1.3456978967495221, + "grad_norm": 3.563711127822158, + "learning_rate": 1e-06, + "loss": 0.1455, + "step": 3519 + }, + { + "epoch": 1.3460803059273423, + "grad_norm": 2.7947434129126334, + "learning_rate": 1e-06, + "loss": 0.328, + "step": 3520 + }, + { + "epoch": 1.3464627151051625, + "grad_norm": 2.752041145904254, + "learning_rate": 1e-06, + "loss": 0.2037, + "step": 3521 + }, + { + "epoch": 1.3468451242829829, + "grad_norm": 3.3458457867369633, + "learning_rate": 1e-06, + "loss": 0.1078, + "step": 3522 + }, + { + "epoch": 1.347227533460803, + "grad_norm": 2.7651796143950076, + "learning_rate": 1e-06, + "loss": 0.1771, + "step": 3523 + }, + { + "epoch": 1.3476099426386234, + "grad_norm": 3.3446008188042406, + "learning_rate": 1e-06, + "loss": 0.1971, + "step": 3524 + }, + { + "epoch": 1.3479923518164436, + "grad_norm": 3.924573479092709, + "learning_rate": 1e-06, + "loss": 0.2425, + "step": 3525 + }, + { + "epoch": 1.3483747609942638, + "grad_norm": 4.163756740430623, + "learning_rate": 1e-06, + "loss": 0.1407, + "step": 3526 + }, + { + "epoch": 1.3487571701720842, + "grad_norm": 2.5869219885330637, + "learning_rate": 1e-06, + "loss": 0.057, + "step": 3527 + }, + { + "epoch": 1.3491395793499044, + "grad_norm": 6.614779379835467, + "learning_rate": 1e-06, + "loss": 0.2318, + "step": 3528 + }, + { + "epoch": 1.3495219885277248, + "grad_norm": 6.1811222383353766, + "learning_rate": 1e-06, + "loss": 0.4225, + "step": 3529 + }, + { + "epoch": 1.349904397705545, + "grad_norm": 3.8939866093589983, + "learning_rate": 1e-06, + "loss": 0.53, + "step": 3530 + }, + { + "epoch": 1.350286806883365, + "grad_norm": 2.486690857371599, + "learning_rate": 1e-06, + "loss": 0.3089, + "step": 3531 + }, + { + "epoch": 1.3506692160611855, + "grad_norm": 3.0648838029899403, + "learning_rate": 1e-06, + "loss": 0.337, + "step": 3532 + }, + { + "epoch": 1.3510516252390057, + "grad_norm": 4.054914699649741, + "learning_rate": 1e-06, + "loss": 0.4405, + "step": 3533 + }, + { + "epoch": 1.351434034416826, + "grad_norm": 2.605171060784863, + "learning_rate": 1e-06, + "loss": 0.1541, + "step": 3534 + }, + { + "epoch": 1.3518164435946463, + "grad_norm": 2.86937046363919, + "learning_rate": 1e-06, + "loss": 0.1522, + "step": 3535 + }, + { + "epoch": 1.3521988527724664, + "grad_norm": 3.984496829096379, + "learning_rate": 1e-06, + "loss": 0.2619, + "step": 3536 + }, + { + "epoch": 1.3525812619502868, + "grad_norm": 3.5482169818093143, + "learning_rate": 1e-06, + "loss": 0.2178, + "step": 3537 + }, + { + "epoch": 1.352963671128107, + "grad_norm": 2.611337985560104, + "learning_rate": 1e-06, + "loss": 0.0677, + "step": 3538 + }, + { + "epoch": 1.3533460803059274, + "grad_norm": 3.1567322013306685, + "learning_rate": 1e-06, + "loss": 0.0957, + "step": 3539 + }, + { + "epoch": 1.3537284894837476, + "grad_norm": 8.004134540271641, + "learning_rate": 1e-06, + "loss": 0.1923, + "step": 3540 + }, + { + "epoch": 1.3541108986615678, + "grad_norm": 4.335753620819479, + "learning_rate": 1e-06, + "loss": 0.3444, + "step": 3541 + }, + { + "epoch": 1.3544933078393881, + "grad_norm": 3.0156895131538484, + "learning_rate": 1e-06, + "loss": 0.3191, + "step": 3542 + }, + { + "epoch": 1.3548757170172085, + "grad_norm": 3.65331430850215, + "learning_rate": 1e-06, + "loss": 0.2218, + "step": 3543 + }, + { + "epoch": 1.3552581261950287, + "grad_norm": 2.125882806844032, + "learning_rate": 1e-06, + "loss": 0.1969, + "step": 3544 + }, + { + "epoch": 1.355640535372849, + "grad_norm": 3.1973091733060524, + "learning_rate": 1e-06, + "loss": 0.3326, + "step": 3545 + }, + { + "epoch": 1.3560229445506693, + "grad_norm": 3.9776660159966237, + "learning_rate": 1e-06, + "loss": 0.3211, + "step": 3546 + }, + { + "epoch": 1.3564053537284895, + "grad_norm": 3.0101871776996854, + "learning_rate": 1e-06, + "loss": 0.2689, + "step": 3547 + }, + { + "epoch": 1.3567877629063099, + "grad_norm": 4.576216889183216, + "learning_rate": 1e-06, + "loss": 0.1728, + "step": 3548 + }, + { + "epoch": 1.35717017208413, + "grad_norm": 3.005881108037498, + "learning_rate": 1e-06, + "loss": 0.1002, + "step": 3549 + }, + { + "epoch": 1.3575525812619502, + "grad_norm": 3.1832495380444117, + "learning_rate": 1e-06, + "loss": 0.129, + "step": 3550 + }, + { + "epoch": 1.3579349904397706, + "grad_norm": 3.0458417656638703, + "learning_rate": 1e-06, + "loss": 0.1105, + "step": 3551 + }, + { + "epoch": 1.3583173996175908, + "grad_norm": 3.0338615886722966, + "learning_rate": 1e-06, + "loss": 0.0667, + "step": 3552 + }, + { + "epoch": 1.3586998087954112, + "grad_norm": 4.363334798868475, + "learning_rate": 1e-06, + "loss": 0.4191, + "step": 3553 + }, + { + "epoch": 1.3590822179732314, + "grad_norm": 3.7229429056573213, + "learning_rate": 1e-06, + "loss": 0.2664, + "step": 3554 + }, + { + "epoch": 1.3594646271510515, + "grad_norm": 2.870013016727716, + "learning_rate": 1e-06, + "loss": 0.3336, + "step": 3555 + }, + { + "epoch": 1.359847036328872, + "grad_norm": 2.168764700276357, + "learning_rate": 1e-06, + "loss": 0.1946, + "step": 3556 + }, + { + "epoch": 1.3602294455066921, + "grad_norm": 3.44598478993819, + "learning_rate": 1e-06, + "loss": 0.2641, + "step": 3557 + }, + { + "epoch": 1.3606118546845125, + "grad_norm": 4.045256655629172, + "learning_rate": 1e-06, + "loss": 0.4309, + "step": 3558 + }, + { + "epoch": 1.3609942638623327, + "grad_norm": 3.2178948016281863, + "learning_rate": 1e-06, + "loss": 0.3445, + "step": 3559 + }, + { + "epoch": 1.3613766730401529, + "grad_norm": 2.952558836580853, + "learning_rate": 1e-06, + "loss": 0.2832, + "step": 3560 + }, + { + "epoch": 1.3617590822179733, + "grad_norm": 2.876979643886604, + "learning_rate": 1e-06, + "loss": 0.209, + "step": 3561 + }, + { + "epoch": 1.3621414913957934, + "grad_norm": 3.4246323158433025, + "learning_rate": 1e-06, + "loss": 0.0917, + "step": 3562 + }, + { + "epoch": 1.3625239005736138, + "grad_norm": 3.293764500160801, + "learning_rate": 1e-06, + "loss": 0.0904, + "step": 3563 + }, + { + "epoch": 1.362906309751434, + "grad_norm": 3.631939526747678, + "learning_rate": 1e-06, + "loss": 0.0889, + "step": 3564 + }, + { + "epoch": 1.3632887189292542, + "grad_norm": 4.249874786327579, + "learning_rate": 1e-06, + "loss": 0.0662, + "step": 3565 + }, + { + "epoch": 1.3636711281070746, + "grad_norm": 8.425726468151769, + "learning_rate": 1e-06, + "loss": 0.2319, + "step": 3566 + }, + { + "epoch": 1.3640535372848948, + "grad_norm": 4.157774681426793, + "learning_rate": 1e-06, + "loss": 0.4404, + "step": 3567 + }, + { + "epoch": 1.3644359464627152, + "grad_norm": 3.2119620143596737, + "learning_rate": 1e-06, + "loss": 0.2398, + "step": 3568 + }, + { + "epoch": 1.3648183556405353, + "grad_norm": 2.3561767728643264, + "learning_rate": 1e-06, + "loss": 0.2127, + "step": 3569 + }, + { + "epoch": 1.3652007648183555, + "grad_norm": 3.921888860548839, + "learning_rate": 1e-06, + "loss": 0.5913, + "step": 3570 + }, + { + "epoch": 1.365583173996176, + "grad_norm": 3.304847659294973, + "learning_rate": 1e-06, + "loss": 0.1353, + "step": 3571 + }, + { + "epoch": 1.3659655831739963, + "grad_norm": 3.5156186252112342, + "learning_rate": 1e-06, + "loss": 0.1858, + "step": 3572 + }, + { + "epoch": 1.3663479923518165, + "grad_norm": 2.9621940100661, + "learning_rate": 1e-06, + "loss": 0.1808, + "step": 3573 + }, + { + "epoch": 1.3667304015296367, + "grad_norm": 3.2755091264155327, + "learning_rate": 1e-06, + "loss": 0.2464, + "step": 3574 + }, + { + "epoch": 1.367112810707457, + "grad_norm": 4.7476897393709425, + "learning_rate": 1e-06, + "loss": 0.297, + "step": 3575 + }, + { + "epoch": 1.3674952198852772, + "grad_norm": 3.513067781979689, + "learning_rate": 1e-06, + "loss": 0.1809, + "step": 3576 + }, + { + "epoch": 1.3678776290630976, + "grad_norm": 3.219846816358348, + "learning_rate": 1e-06, + "loss": 0.0732, + "step": 3577 + }, + { + "epoch": 1.3682600382409178, + "grad_norm": 6.085868864443686, + "learning_rate": 1e-06, + "loss": 0.1976, + "step": 3578 + }, + { + "epoch": 1.368642447418738, + "grad_norm": 4.61869093014494, + "learning_rate": 1e-06, + "loss": 0.6497, + "step": 3579 + }, + { + "epoch": 1.3690248565965584, + "grad_norm": 4.839491023895109, + "learning_rate": 1e-06, + "loss": 0.3992, + "step": 3580 + }, + { + "epoch": 1.3694072657743785, + "grad_norm": 2.65189525200031, + "learning_rate": 1e-06, + "loss": 0.2308, + "step": 3581 + }, + { + "epoch": 1.369789674952199, + "grad_norm": 4.629822562259575, + "learning_rate": 1e-06, + "loss": 0.3704, + "step": 3582 + }, + { + "epoch": 1.3701720841300191, + "grad_norm": 4.7582316096578845, + "learning_rate": 1e-06, + "loss": 0.4397, + "step": 3583 + }, + { + "epoch": 1.3705544933078393, + "grad_norm": 2.7484996344052743, + "learning_rate": 1e-06, + "loss": 0.174, + "step": 3584 + }, + { + "epoch": 1.3709369024856597, + "grad_norm": 4.355188909872203, + "learning_rate": 1e-06, + "loss": 0.1491, + "step": 3585 + }, + { + "epoch": 1.3713193116634799, + "grad_norm": 3.2913755135195517, + "learning_rate": 1e-06, + "loss": 0.1138, + "step": 3586 + }, + { + "epoch": 1.3717017208413003, + "grad_norm": 3.575477920278412, + "learning_rate": 1e-06, + "loss": 0.2969, + "step": 3587 + }, + { + "epoch": 1.3720841300191204, + "grad_norm": 3.3627597524155393, + "learning_rate": 1e-06, + "loss": 0.1533, + "step": 3588 + }, + { + "epoch": 1.3724665391969406, + "grad_norm": 3.8016096821658465, + "learning_rate": 1e-06, + "loss": 0.0731, + "step": 3589 + }, + { + "epoch": 1.372848948374761, + "grad_norm": 7.239808911072699, + "learning_rate": 1e-06, + "loss": 0.2692, + "step": 3590 + }, + { + "epoch": 1.3732313575525812, + "grad_norm": 7.992433784695021, + "learning_rate": 1e-06, + "loss": 0.5004, + "step": 3591 + }, + { + "epoch": 1.3736137667304016, + "grad_norm": 2.7557299480149937, + "learning_rate": 1e-06, + "loss": 0.3263, + "step": 3592 + }, + { + "epoch": 1.3739961759082218, + "grad_norm": 3.175777496298135, + "learning_rate": 1e-06, + "loss": 0.2712, + "step": 3593 + }, + { + "epoch": 1.374378585086042, + "grad_norm": 2.3902367170571446, + "learning_rate": 1e-06, + "loss": 0.3309, + "step": 3594 + }, + { + "epoch": 1.3747609942638623, + "grad_norm": 2.9567911310100063, + "learning_rate": 1e-06, + "loss": 0.2152, + "step": 3595 + }, + { + "epoch": 1.3751434034416827, + "grad_norm": 3.233958931666279, + "learning_rate": 1e-06, + "loss": 0.3663, + "step": 3596 + }, + { + "epoch": 1.375525812619503, + "grad_norm": 3.6868141959014014, + "learning_rate": 1e-06, + "loss": 0.2287, + "step": 3597 + }, + { + "epoch": 1.375908221797323, + "grad_norm": 2.6929733962463547, + "learning_rate": 1e-06, + "loss": 0.1352, + "step": 3598 + }, + { + "epoch": 1.3762906309751435, + "grad_norm": 3.3579409355072016, + "learning_rate": 1e-06, + "loss": 0.1688, + "step": 3599 + }, + { + "epoch": 1.3766730401529637, + "grad_norm": 4.225536623852819, + "learning_rate": 1e-06, + "loss": 0.1663, + "step": 3600 + }, + { + "epoch": 1.377055449330784, + "grad_norm": 6.74606886938087, + "learning_rate": 1e-06, + "loss": 0.3401, + "step": 3601 + }, + { + "epoch": 1.3774378585086042, + "grad_norm": 3.743459592837295, + "learning_rate": 1e-06, + "loss": 0.0814, + "step": 3602 + }, + { + "epoch": 1.3778202676864244, + "grad_norm": 9.690145414381327, + "learning_rate": 1e-06, + "loss": 0.2664, + "step": 3603 + }, + { + "epoch": 1.3782026768642448, + "grad_norm": 3.8949521678098136, + "learning_rate": 1e-06, + "loss": 0.3333, + "step": 3604 + }, + { + "epoch": 1.378585086042065, + "grad_norm": 3.4412114307973285, + "learning_rate": 1e-06, + "loss": 0.3546, + "step": 3605 + }, + { + "epoch": 1.3789674952198854, + "grad_norm": 2.817739226805909, + "learning_rate": 1e-06, + "loss": 0.3284, + "step": 3606 + }, + { + "epoch": 1.3793499043977056, + "grad_norm": 4.063424342708501, + "learning_rate": 1e-06, + "loss": 0.3598, + "step": 3607 + }, + { + "epoch": 1.3797323135755257, + "grad_norm": 2.7788390950734434, + "learning_rate": 1e-06, + "loss": 0.1863, + "step": 3608 + }, + { + "epoch": 1.3801147227533461, + "grad_norm": 2.5297091468417365, + "learning_rate": 1e-06, + "loss": 0.1796, + "step": 3609 + }, + { + "epoch": 1.3804971319311663, + "grad_norm": 2.5535086148641253, + "learning_rate": 1e-06, + "loss": 0.151, + "step": 3610 + }, + { + "epoch": 1.3808795411089867, + "grad_norm": 4.293329503154428, + "learning_rate": 1e-06, + "loss": 0.381, + "step": 3611 + }, + { + "epoch": 1.3812619502868069, + "grad_norm": 2.824774343204578, + "learning_rate": 1e-06, + "loss": 0.1622, + "step": 3612 + }, + { + "epoch": 1.381644359464627, + "grad_norm": 2.6020886688136997, + "learning_rate": 1e-06, + "loss": 0.073, + "step": 3613 + }, + { + "epoch": 1.3820267686424474, + "grad_norm": 2.7319932428279214, + "learning_rate": 1e-06, + "loss": 0.0816, + "step": 3614 + }, + { + "epoch": 1.3824091778202676, + "grad_norm": 3.8276916258599285, + "learning_rate": 1e-06, + "loss": 0.1063, + "step": 3615 + }, + { + "epoch": 1.382791586998088, + "grad_norm": 5.450492332688963, + "learning_rate": 1e-06, + "loss": 0.4124, + "step": 3616 + }, + { + "epoch": 1.3831739961759082, + "grad_norm": 4.239795776343202, + "learning_rate": 1e-06, + "loss": 0.5117, + "step": 3617 + }, + { + "epoch": 1.3835564053537284, + "grad_norm": 3.7588389178542854, + "learning_rate": 1e-06, + "loss": 0.3095, + "step": 3618 + }, + { + "epoch": 1.3839388145315488, + "grad_norm": 3.728928172975965, + "learning_rate": 1e-06, + "loss": 0.4486, + "step": 3619 + }, + { + "epoch": 1.384321223709369, + "grad_norm": 5.153491952946277, + "learning_rate": 1e-06, + "loss": 0.2694, + "step": 3620 + }, + { + "epoch": 1.3847036328871893, + "grad_norm": 3.255781167045849, + "learning_rate": 1e-06, + "loss": 0.2452, + "step": 3621 + }, + { + "epoch": 1.3850860420650095, + "grad_norm": 2.844236059883789, + "learning_rate": 1e-06, + "loss": 0.2148, + "step": 3622 + }, + { + "epoch": 1.3854684512428297, + "grad_norm": 2.263122439252642, + "learning_rate": 1e-06, + "loss": 0.2454, + "step": 3623 + }, + { + "epoch": 1.38585086042065, + "grad_norm": 2.4557283036738498, + "learning_rate": 1e-06, + "loss": 0.1102, + "step": 3624 + }, + { + "epoch": 1.3862332695984705, + "grad_norm": 3.1210013174184397, + "learning_rate": 1e-06, + "loss": 0.0913, + "step": 3625 + }, + { + "epoch": 1.3866156787762907, + "grad_norm": 2.220231756880406, + "learning_rate": 1e-06, + "loss": 0.0711, + "step": 3626 + }, + { + "epoch": 1.3869980879541108, + "grad_norm": 3.0206513097885277, + "learning_rate": 1e-06, + "loss": 0.106, + "step": 3627 + }, + { + "epoch": 1.3873804971319312, + "grad_norm": 3.5395434728529356, + "learning_rate": 1e-06, + "loss": 0.2138, + "step": 3628 + }, + { + "epoch": 1.3877629063097514, + "grad_norm": 14.14533653379434, + "learning_rate": 1e-06, + "loss": 0.3418, + "step": 3629 + }, + { + "epoch": 1.3881453154875718, + "grad_norm": 3.6628960092640654, + "learning_rate": 1e-06, + "loss": 0.5699, + "step": 3630 + }, + { + "epoch": 1.388527724665392, + "grad_norm": 2.6896311272339815, + "learning_rate": 1e-06, + "loss": 0.208, + "step": 3631 + }, + { + "epoch": 1.3889101338432122, + "grad_norm": 3.044524709505099, + "learning_rate": 1e-06, + "loss": 0.2483, + "step": 3632 + }, + { + "epoch": 1.3892925430210326, + "grad_norm": 3.045121377865358, + "learning_rate": 1e-06, + "loss": 0.2137, + "step": 3633 + }, + { + "epoch": 1.3896749521988527, + "grad_norm": 3.8082912540128318, + "learning_rate": 1e-06, + "loss": 0.314, + "step": 3634 + }, + { + "epoch": 1.3900573613766731, + "grad_norm": 2.786056052192463, + "learning_rate": 1e-06, + "loss": 0.202, + "step": 3635 + }, + { + "epoch": 1.3904397705544933, + "grad_norm": 2.5837768563626664, + "learning_rate": 1e-06, + "loss": 0.2055, + "step": 3636 + }, + { + "epoch": 1.3908221797323135, + "grad_norm": 3.4122340283827435, + "learning_rate": 1e-06, + "loss": 0.2463, + "step": 3637 + }, + { + "epoch": 1.3912045889101339, + "grad_norm": 4.804509254188609, + "learning_rate": 1e-06, + "loss": 0.1585, + "step": 3638 + }, + { + "epoch": 1.391586998087954, + "grad_norm": 3.546749700946221, + "learning_rate": 1e-06, + "loss": 0.0743, + "step": 3639 + }, + { + "epoch": 1.3919694072657744, + "grad_norm": 3.133098788985979, + "learning_rate": 1e-06, + "loss": 0.0622, + "step": 3640 + }, + { + "epoch": 1.3923518164435946, + "grad_norm": 6.5644693961788825, + "learning_rate": 1e-06, + "loss": 0.281, + "step": 3641 + }, + { + "epoch": 1.3927342256214148, + "grad_norm": 3.2124314000267664, + "learning_rate": 1e-06, + "loss": 0.3506, + "step": 3642 + }, + { + "epoch": 1.3931166347992352, + "grad_norm": 2.6462614859438487, + "learning_rate": 1e-06, + "loss": 0.3129, + "step": 3643 + }, + { + "epoch": 1.3934990439770554, + "grad_norm": 2.184482018464643, + "learning_rate": 1e-06, + "loss": 0.1818, + "step": 3644 + }, + { + "epoch": 1.3938814531548758, + "grad_norm": 2.879734122136413, + "learning_rate": 1e-06, + "loss": 0.2177, + "step": 3645 + }, + { + "epoch": 1.394263862332696, + "grad_norm": 2.2983385470605278, + "learning_rate": 1e-06, + "loss": 0.2533, + "step": 3646 + }, + { + "epoch": 1.3946462715105161, + "grad_norm": 2.7621492430731402, + "learning_rate": 1e-06, + "loss": 0.1919, + "step": 3647 + }, + { + "epoch": 1.3950286806883365, + "grad_norm": 3.2641662551948443, + "learning_rate": 1e-06, + "loss": 0.2589, + "step": 3648 + }, + { + "epoch": 1.395411089866157, + "grad_norm": 3.2110880795554464, + "learning_rate": 1e-06, + "loss": 0.1783, + "step": 3649 + }, + { + "epoch": 1.395793499043977, + "grad_norm": 4.443205483255379, + "learning_rate": 1e-06, + "loss": 0.1439, + "step": 3650 + }, + { + "epoch": 1.3961759082217973, + "grad_norm": 3.011841609442816, + "learning_rate": 1e-06, + "loss": 0.2, + "step": 3651 + }, + { + "epoch": 1.3965583173996177, + "grad_norm": 3.9484192353019756, + "learning_rate": 1e-06, + "loss": 0.1176, + "step": 3652 + }, + { + "epoch": 1.3969407265774378, + "grad_norm": 4.166531115552269, + "learning_rate": 1e-06, + "loss": 0.1474, + "step": 3653 + }, + { + "epoch": 1.3973231357552582, + "grad_norm": 5.26524301835712, + "learning_rate": 1e-06, + "loss": 0.5814, + "step": 3654 + }, + { + "epoch": 1.3977055449330784, + "grad_norm": 3.057540926717856, + "learning_rate": 1e-06, + "loss": 0.2639, + "step": 3655 + }, + { + "epoch": 1.3980879541108986, + "grad_norm": 2.015501269897273, + "learning_rate": 1e-06, + "loss": 0.1526, + "step": 3656 + }, + { + "epoch": 1.398470363288719, + "grad_norm": 2.9311632025582846, + "learning_rate": 1e-06, + "loss": 0.2517, + "step": 3657 + }, + { + "epoch": 1.3988527724665392, + "grad_norm": 2.72936683871947, + "learning_rate": 1e-06, + "loss": 0.1346, + "step": 3658 + }, + { + "epoch": 1.3992351816443596, + "grad_norm": 2.7550209332621054, + "learning_rate": 1e-06, + "loss": 0.1167, + "step": 3659 + }, + { + "epoch": 1.3996175908221797, + "grad_norm": 3.221825916941309, + "learning_rate": 1e-06, + "loss": 0.1641, + "step": 3660 + }, + { + "epoch": 1.4, + "grad_norm": 4.5738405353085545, + "learning_rate": 1e-06, + "loss": 0.2635, + "step": 3661 + }, + { + "epoch": 1.4003824091778203, + "grad_norm": 3.6624015508445966, + "learning_rate": 1e-06, + "loss": 0.1371, + "step": 3662 + }, + { + "epoch": 1.4007648183556405, + "grad_norm": 5.230158231832464, + "learning_rate": 1e-06, + "loss": 0.1611, + "step": 3663 + }, + { + "epoch": 1.4011472275334609, + "grad_norm": 2.4369477844717853, + "learning_rate": 1e-06, + "loss": 0.0432, + "step": 3664 + }, + { + "epoch": 1.401529636711281, + "grad_norm": 4.154970818335636, + "learning_rate": 1e-06, + "loss": 0.1246, + "step": 3665 + }, + { + "epoch": 1.4019120458891012, + "grad_norm": 6.541607834213285, + "learning_rate": 1e-06, + "loss": 0.6472, + "step": 3666 + }, + { + "epoch": 1.4022944550669216, + "grad_norm": 4.861430527622743, + "learning_rate": 1e-06, + "loss": 0.6617, + "step": 3667 + }, + { + "epoch": 1.4026768642447418, + "grad_norm": 2.6870531997099145, + "learning_rate": 1e-06, + "loss": 0.1698, + "step": 3668 + }, + { + "epoch": 1.4030592734225622, + "grad_norm": 2.079510105285526, + "learning_rate": 1e-06, + "loss": 0.2955, + "step": 3669 + }, + { + "epoch": 1.4034416826003824, + "grad_norm": 2.435760953621528, + "learning_rate": 1e-06, + "loss": 0.3218, + "step": 3670 + }, + { + "epoch": 1.4038240917782026, + "grad_norm": 4.586588356842911, + "learning_rate": 1e-06, + "loss": 0.2798, + "step": 3671 + }, + { + "epoch": 1.404206500956023, + "grad_norm": 3.088087969559685, + "learning_rate": 1e-06, + "loss": 0.2636, + "step": 3672 + }, + { + "epoch": 1.4045889101338433, + "grad_norm": 4.336824233652288, + "learning_rate": 1e-06, + "loss": 0.3802, + "step": 3673 + }, + { + "epoch": 1.4049713193116635, + "grad_norm": 3.0738729939280662, + "learning_rate": 1e-06, + "loss": 0.2326, + "step": 3674 + }, + { + "epoch": 1.4053537284894837, + "grad_norm": 3.5809442998820047, + "learning_rate": 1e-06, + "loss": 0.1055, + "step": 3675 + }, + { + "epoch": 1.405736137667304, + "grad_norm": 3.4578201342126524, + "learning_rate": 1e-06, + "loss": 0.1185, + "step": 3676 + }, + { + "epoch": 1.4061185468451243, + "grad_norm": 3.1254081459541228, + "learning_rate": 1e-06, + "loss": 0.1281, + "step": 3677 + }, + { + "epoch": 1.4065009560229447, + "grad_norm": 6.7148551119740985, + "learning_rate": 1e-06, + "loss": 0.3116, + "step": 3678 + }, + { + "epoch": 1.4068833652007648, + "grad_norm": 4.238429981909322, + "learning_rate": 1e-06, + "loss": 0.4395, + "step": 3679 + }, + { + "epoch": 1.407265774378585, + "grad_norm": 3.5441077103963172, + "learning_rate": 1e-06, + "loss": 0.2503, + "step": 3680 + }, + { + "epoch": 1.4076481835564054, + "grad_norm": 3.215104788916053, + "learning_rate": 1e-06, + "loss": 0.2262, + "step": 3681 + }, + { + "epoch": 1.4080305927342256, + "grad_norm": 4.2290451458289, + "learning_rate": 1e-06, + "loss": 0.2488, + "step": 3682 + }, + { + "epoch": 1.408413001912046, + "grad_norm": 3.231572065406559, + "learning_rate": 1e-06, + "loss": 0.1715, + "step": 3683 + }, + { + "epoch": 1.4087954110898662, + "grad_norm": 2.7355573441260024, + "learning_rate": 1e-06, + "loss": 0.2283, + "step": 3684 + }, + { + "epoch": 1.4091778202676863, + "grad_norm": 2.783879098283958, + "learning_rate": 1e-06, + "loss": 0.1159, + "step": 3685 + }, + { + "epoch": 1.4095602294455067, + "grad_norm": 3.3016736150440917, + "learning_rate": 1e-06, + "loss": 0.2454, + "step": 3686 + }, + { + "epoch": 1.409942638623327, + "grad_norm": 4.556744851867343, + "learning_rate": 1e-06, + "loss": 0.2448, + "step": 3687 + }, + { + "epoch": 1.4103250478011473, + "grad_norm": 3.576157207865198, + "learning_rate": 1e-06, + "loss": 0.1276, + "step": 3688 + }, + { + "epoch": 1.4107074569789675, + "grad_norm": 4.030599615579349, + "learning_rate": 1e-06, + "loss": 0.1065, + "step": 3689 + }, + { + "epoch": 1.4110898661567877, + "grad_norm": 3.85257161444581, + "learning_rate": 1e-06, + "loss": 0.1013, + "step": 3690 + }, + { + "epoch": 1.411472275334608, + "grad_norm": 6.152233691971946, + "learning_rate": 1e-06, + "loss": 0.3113, + "step": 3691 + }, + { + "epoch": 1.4118546845124282, + "grad_norm": 4.592085763921228, + "learning_rate": 1e-06, + "loss": 0.4697, + "step": 3692 + }, + { + "epoch": 1.4122370936902486, + "grad_norm": 4.530306171908304, + "learning_rate": 1e-06, + "loss": 0.3802, + "step": 3693 + }, + { + "epoch": 1.4126195028680688, + "grad_norm": 3.201830048596461, + "learning_rate": 1e-06, + "loss": 0.2451, + "step": 3694 + }, + { + "epoch": 1.413001912045889, + "grad_norm": 2.6634717717036365, + "learning_rate": 1e-06, + "loss": 0.1851, + "step": 3695 + }, + { + "epoch": 1.4133843212237094, + "grad_norm": 3.610814851851053, + "learning_rate": 1e-06, + "loss": 0.3227, + "step": 3696 + }, + { + "epoch": 1.4137667304015296, + "grad_norm": 2.707512002286083, + "learning_rate": 1e-06, + "loss": 0.2215, + "step": 3697 + }, + { + "epoch": 1.41414913957935, + "grad_norm": 2.7175176173168127, + "learning_rate": 1e-06, + "loss": 0.2505, + "step": 3698 + }, + { + "epoch": 1.4145315487571701, + "grad_norm": 3.070713628526035, + "learning_rate": 1e-06, + "loss": 0.2042, + "step": 3699 + }, + { + "epoch": 1.4149139579349903, + "grad_norm": 5.176987256310562, + "learning_rate": 1e-06, + "loss": 0.1853, + "step": 3700 + }, + { + "epoch": 1.4152963671128107, + "grad_norm": 3.5168723373172512, + "learning_rate": 1e-06, + "loss": 0.1126, + "step": 3701 + }, + { + "epoch": 1.415678776290631, + "grad_norm": 2.6169160161198914, + "learning_rate": 1e-06, + "loss": 0.0843, + "step": 3702 + }, + { + "epoch": 1.4160611854684513, + "grad_norm": 9.725306614452688, + "learning_rate": 1e-06, + "loss": 0.1681, + "step": 3703 + }, + { + "epoch": 1.4164435946462715, + "grad_norm": 5.82730579499387, + "learning_rate": 1e-06, + "loss": 0.5179, + "step": 3704 + }, + { + "epoch": 1.4168260038240919, + "grad_norm": 3.8021214034730724, + "learning_rate": 1e-06, + "loss": 0.4602, + "step": 3705 + }, + { + "epoch": 1.417208413001912, + "grad_norm": 3.5419228554611526, + "learning_rate": 1e-06, + "loss": 0.2668, + "step": 3706 + }, + { + "epoch": 1.4175908221797324, + "grad_norm": 2.426808506764268, + "learning_rate": 1e-06, + "loss": 0.1892, + "step": 3707 + }, + { + "epoch": 1.4179732313575526, + "grad_norm": 3.6122594099875975, + "learning_rate": 1e-06, + "loss": 0.2188, + "step": 3708 + }, + { + "epoch": 1.4183556405353728, + "grad_norm": 3.1404725602385586, + "learning_rate": 1e-06, + "loss": 0.3341, + "step": 3709 + }, + { + "epoch": 1.4187380497131932, + "grad_norm": 4.374365406333801, + "learning_rate": 1e-06, + "loss": 0.0922, + "step": 3710 + }, + { + "epoch": 1.4191204588910133, + "grad_norm": 3.265568344865903, + "learning_rate": 1e-06, + "loss": 0.1587, + "step": 3711 + }, + { + "epoch": 1.4195028680688337, + "grad_norm": 4.36933379413446, + "learning_rate": 1e-06, + "loss": 0.1202, + "step": 3712 + }, + { + "epoch": 1.419885277246654, + "grad_norm": 8.529275183522337, + "learning_rate": 1e-06, + "loss": 0.1127, + "step": 3713 + }, + { + "epoch": 1.420267686424474, + "grad_norm": 4.165621639173087, + "learning_rate": 1e-06, + "loss": 0.1621, + "step": 3714 + }, + { + "epoch": 1.4206500956022945, + "grad_norm": 3.330562361695334, + "learning_rate": 1e-06, + "loss": 0.1021, + "step": 3715 + }, + { + "epoch": 1.4210325047801147, + "grad_norm": 7.72024583415231, + "learning_rate": 1e-06, + "loss": 0.3048, + "step": 3716 + }, + { + "epoch": 1.421414913957935, + "grad_norm": 3.407402753442851, + "learning_rate": 1e-06, + "loss": 0.3945, + "step": 3717 + }, + { + "epoch": 1.4217973231357552, + "grad_norm": 3.9287353332808927, + "learning_rate": 1e-06, + "loss": 0.2428, + "step": 3718 + }, + { + "epoch": 1.4221797323135754, + "grad_norm": 3.744834903189784, + "learning_rate": 1e-06, + "loss": 0.1774, + "step": 3719 + }, + { + "epoch": 1.4225621414913958, + "grad_norm": 3.6131709231467375, + "learning_rate": 1e-06, + "loss": 0.2858, + "step": 3720 + }, + { + "epoch": 1.422944550669216, + "grad_norm": 3.777359403244454, + "learning_rate": 1e-06, + "loss": 0.3237, + "step": 3721 + }, + { + "epoch": 1.4233269598470364, + "grad_norm": 2.544264127995152, + "learning_rate": 1e-06, + "loss": 0.1212, + "step": 3722 + }, + { + "epoch": 1.4237093690248566, + "grad_norm": 2.447306931764326, + "learning_rate": 1e-06, + "loss": 0.2585, + "step": 3723 + }, + { + "epoch": 1.4240917782026767, + "grad_norm": 3.166217136265214, + "learning_rate": 1e-06, + "loss": 0.1585, + "step": 3724 + }, + { + "epoch": 1.4244741873804971, + "grad_norm": 2.9064360528273805, + "learning_rate": 1e-06, + "loss": 0.1129, + "step": 3725 + }, + { + "epoch": 1.4248565965583175, + "grad_norm": 3.5005435521533594, + "learning_rate": 1e-06, + "loss": 0.1485, + "step": 3726 + }, + { + "epoch": 1.4252390057361377, + "grad_norm": 3.010948072651472, + "learning_rate": 1e-06, + "loss": 0.0942, + "step": 3727 + }, + { + "epoch": 1.4256214149139579, + "grad_norm": 6.661490369952854, + "learning_rate": 1e-06, + "loss": 0.1345, + "step": 3728 + }, + { + "epoch": 1.4260038240917783, + "grad_norm": 4.087868695444046, + "learning_rate": 1e-06, + "loss": 0.3328, + "step": 3729 + }, + { + "epoch": 1.4263862332695985, + "grad_norm": 4.022493062697949, + "learning_rate": 1e-06, + "loss": 0.4384, + "step": 3730 + }, + { + "epoch": 1.4267686424474189, + "grad_norm": 3.9391085048895844, + "learning_rate": 1e-06, + "loss": 0.268, + "step": 3731 + }, + { + "epoch": 1.427151051625239, + "grad_norm": 2.7178987069135747, + "learning_rate": 1e-06, + "loss": 0.2054, + "step": 3732 + }, + { + "epoch": 1.4275334608030592, + "grad_norm": 2.9878210174595545, + "learning_rate": 1e-06, + "loss": 0.2353, + "step": 3733 + }, + { + "epoch": 1.4279158699808796, + "grad_norm": 2.6006159162924245, + "learning_rate": 1e-06, + "loss": 0.2965, + "step": 3734 + }, + { + "epoch": 1.4282982791586998, + "grad_norm": 2.152111006266795, + "learning_rate": 1e-06, + "loss": 0.192, + "step": 3735 + }, + { + "epoch": 1.4286806883365202, + "grad_norm": 3.0866092771573768, + "learning_rate": 1e-06, + "loss": 0.1834, + "step": 3736 + }, + { + "epoch": 1.4290630975143404, + "grad_norm": 3.2257135119431815, + "learning_rate": 1e-06, + "loss": 0.0982, + "step": 3737 + }, + { + "epoch": 1.4294455066921605, + "grad_norm": 2.867569983641896, + "learning_rate": 1e-06, + "loss": 0.1885, + "step": 3738 + }, + { + "epoch": 1.429827915869981, + "grad_norm": 2.979210024075797, + "learning_rate": 1e-06, + "loss": 0.0917, + "step": 3739 + }, + { + "epoch": 1.430210325047801, + "grad_norm": 4.089296903731755, + "learning_rate": 1e-06, + "loss": 0.1143, + "step": 3740 + }, + { + "epoch": 1.4305927342256215, + "grad_norm": 6.8203038306252175, + "learning_rate": 1e-06, + "loss": 0.387, + "step": 3741 + }, + { + "epoch": 1.4309751434034417, + "grad_norm": 3.697388015712296, + "learning_rate": 1e-06, + "loss": 0.2882, + "step": 3742 + }, + { + "epoch": 1.4313575525812618, + "grad_norm": 3.100641147392317, + "learning_rate": 1e-06, + "loss": 0.2431, + "step": 3743 + }, + { + "epoch": 1.4317399617590822, + "grad_norm": 2.276666399824016, + "learning_rate": 1e-06, + "loss": 0.2894, + "step": 3744 + }, + { + "epoch": 1.4321223709369024, + "grad_norm": 3.471309370632797, + "learning_rate": 1e-06, + "loss": 0.4578, + "step": 3745 + }, + { + "epoch": 1.4325047801147228, + "grad_norm": 2.513064864700208, + "learning_rate": 1e-06, + "loss": 0.1828, + "step": 3746 + }, + { + "epoch": 1.432887189292543, + "grad_norm": 3.0271788324535414, + "learning_rate": 1e-06, + "loss": 0.1807, + "step": 3747 + }, + { + "epoch": 1.4332695984703632, + "grad_norm": 2.9643592890999306, + "learning_rate": 1e-06, + "loss": 0.1636, + "step": 3748 + }, + { + "epoch": 1.4336520076481836, + "grad_norm": 4.422537278901826, + "learning_rate": 1e-06, + "loss": 0.1512, + "step": 3749 + }, + { + "epoch": 1.4340344168260037, + "grad_norm": 5.567163805358466, + "learning_rate": 1e-06, + "loss": 0.1327, + "step": 3750 + }, + { + "epoch": 1.4344168260038241, + "grad_norm": 3.1560433197336115, + "learning_rate": 1e-06, + "loss": 0.1617, + "step": 3751 + }, + { + "epoch": 1.4347992351816443, + "grad_norm": 3.695903448552923, + "learning_rate": 1e-06, + "loss": 0.0543, + "step": 3752 + }, + { + "epoch": 1.4351816443594645, + "grad_norm": 3.4197311304602107, + "learning_rate": 1e-06, + "loss": 0.2029, + "step": 3753 + }, + { + "epoch": 1.435564053537285, + "grad_norm": 5.646728932571445, + "learning_rate": 1e-06, + "loss": 0.3716, + "step": 3754 + }, + { + "epoch": 1.4359464627151053, + "grad_norm": 3.627177143727946, + "learning_rate": 1e-06, + "loss": 0.4611, + "step": 3755 + }, + { + "epoch": 1.4363288718929255, + "grad_norm": 3.729141718321022, + "learning_rate": 1e-06, + "loss": 0.4277, + "step": 3756 + }, + { + "epoch": 1.4367112810707456, + "grad_norm": 1.9529400547201448, + "learning_rate": 1e-06, + "loss": 0.1611, + "step": 3757 + }, + { + "epoch": 1.437093690248566, + "grad_norm": 3.5934846199569557, + "learning_rate": 1e-06, + "loss": 0.3111, + "step": 3758 + }, + { + "epoch": 1.4374760994263862, + "grad_norm": 3.3197003686657816, + "learning_rate": 1e-06, + "loss": 0.2611, + "step": 3759 + }, + { + "epoch": 1.4378585086042066, + "grad_norm": 3.3328892094064173, + "learning_rate": 1e-06, + "loss": 0.1534, + "step": 3760 + }, + { + "epoch": 1.4382409177820268, + "grad_norm": 2.6049055551698954, + "learning_rate": 1e-06, + "loss": 0.1336, + "step": 3761 + }, + { + "epoch": 1.438623326959847, + "grad_norm": 2.225003308925954, + "learning_rate": 1e-06, + "loss": 0.0527, + "step": 3762 + }, + { + "epoch": 1.4390057361376674, + "grad_norm": 2.6469452291312163, + "learning_rate": 1e-06, + "loss": 0.0788, + "step": 3763 + }, + { + "epoch": 1.4393881453154875, + "grad_norm": 4.325611693063195, + "learning_rate": 1e-06, + "loss": 0.1896, + "step": 3764 + }, + { + "epoch": 1.439770554493308, + "grad_norm": 4.960876752826891, + "learning_rate": 1e-06, + "loss": 0.1686, + "step": 3765 + }, + { + "epoch": 1.440152963671128, + "grad_norm": 29.15869287394734, + "learning_rate": 1e-06, + "loss": 0.4078, + "step": 3766 + }, + { + "epoch": 1.4405353728489483, + "grad_norm": 4.755322335566404, + "learning_rate": 1e-06, + "loss": 0.4174, + "step": 3767 + }, + { + "epoch": 1.4409177820267687, + "grad_norm": 3.6924374386414796, + "learning_rate": 1e-06, + "loss": 0.1956, + "step": 3768 + }, + { + "epoch": 1.4413001912045889, + "grad_norm": 2.2364999763391546, + "learning_rate": 1e-06, + "loss": 0.2799, + "step": 3769 + }, + { + "epoch": 1.4416826003824093, + "grad_norm": 2.291046844816055, + "learning_rate": 1e-06, + "loss": 0.2377, + "step": 3770 + }, + { + "epoch": 1.4420650095602294, + "grad_norm": 2.9311425423192157, + "learning_rate": 1e-06, + "loss": 0.3293, + "step": 3771 + }, + { + "epoch": 1.4424474187380496, + "grad_norm": 4.293245315320849, + "learning_rate": 1e-06, + "loss": 0.1789, + "step": 3772 + }, + { + "epoch": 1.44282982791587, + "grad_norm": 3.8070113900377227, + "learning_rate": 1e-06, + "loss": 0.2407, + "step": 3773 + }, + { + "epoch": 1.4432122370936902, + "grad_norm": 3.4656250770506922, + "learning_rate": 1e-06, + "loss": 0.0927, + "step": 3774 + }, + { + "epoch": 1.4435946462715106, + "grad_norm": 2.83877917121469, + "learning_rate": 1e-06, + "loss": 0.1849, + "step": 3775 + }, + { + "epoch": 1.4439770554493307, + "grad_norm": 2.9861681279695182, + "learning_rate": 1e-06, + "loss": 0.1312, + "step": 3776 + }, + { + "epoch": 1.444359464627151, + "grad_norm": 4.455346981968557, + "learning_rate": 1e-06, + "loss": 0.1102, + "step": 3777 + }, + { + "epoch": 1.4447418738049713, + "grad_norm": 3.6213307719757655, + "learning_rate": 1e-06, + "loss": 0.0999, + "step": 3778 + }, + { + "epoch": 1.4451242829827917, + "grad_norm": 3.864392682879823, + "learning_rate": 1e-06, + "loss": 0.4747, + "step": 3779 + }, + { + "epoch": 1.445506692160612, + "grad_norm": 3.804148171676907, + "learning_rate": 1e-06, + "loss": 0.3398, + "step": 3780 + }, + { + "epoch": 1.445889101338432, + "grad_norm": 2.6989292812121124, + "learning_rate": 1e-06, + "loss": 0.1673, + "step": 3781 + }, + { + "epoch": 1.4462715105162525, + "grad_norm": 2.8561388056707435, + "learning_rate": 1e-06, + "loss": 0.2778, + "step": 3782 + }, + { + "epoch": 1.4466539196940726, + "grad_norm": 3.971278549058097, + "learning_rate": 1e-06, + "loss": 0.402, + "step": 3783 + }, + { + "epoch": 1.447036328871893, + "grad_norm": 2.9940332840144133, + "learning_rate": 1e-06, + "loss": 0.274, + "step": 3784 + }, + { + "epoch": 1.4474187380497132, + "grad_norm": 3.310150464835929, + "learning_rate": 1e-06, + "loss": 0.1365, + "step": 3785 + }, + { + "epoch": 1.4478011472275334, + "grad_norm": 2.8408065486088017, + "learning_rate": 1e-06, + "loss": 0.0925, + "step": 3786 + }, + { + "epoch": 1.4481835564053538, + "grad_norm": 3.0926545640933805, + "learning_rate": 1e-06, + "loss": 0.1657, + "step": 3787 + }, + { + "epoch": 1.448565965583174, + "grad_norm": 2.6221871518719135, + "learning_rate": 1e-06, + "loss": 0.0585, + "step": 3788 + }, + { + "epoch": 1.4489483747609944, + "grad_norm": 2.4402629158765445, + "learning_rate": 1e-06, + "loss": 0.0521, + "step": 3789 + }, + { + "epoch": 1.4493307839388145, + "grad_norm": 3.840107378054394, + "learning_rate": 1e-06, + "loss": 0.1009, + "step": 3790 + }, + { + "epoch": 1.4497131931166347, + "grad_norm": 7.8042168002768, + "learning_rate": 1e-06, + "loss": 0.4154, + "step": 3791 + }, + { + "epoch": 1.450095602294455, + "grad_norm": 3.7882124429040727, + "learning_rate": 1e-06, + "loss": 0.3955, + "step": 3792 + }, + { + "epoch": 1.4504780114722753, + "grad_norm": 4.218763111235413, + "learning_rate": 1e-06, + "loss": 0.5727, + "step": 3793 + }, + { + "epoch": 1.4508604206500957, + "grad_norm": 2.3773355544667005, + "learning_rate": 1e-06, + "loss": 0.1635, + "step": 3794 + }, + { + "epoch": 1.4512428298279159, + "grad_norm": 2.944363145085735, + "learning_rate": 1e-06, + "loss": 0.2911, + "step": 3795 + }, + { + "epoch": 1.451625239005736, + "grad_norm": 3.8709171073944915, + "learning_rate": 1e-06, + "loss": 0.4245, + "step": 3796 + }, + { + "epoch": 1.4520076481835564, + "grad_norm": 3.4701117514608883, + "learning_rate": 1e-06, + "loss": 0.3219, + "step": 3797 + }, + { + "epoch": 1.4523900573613766, + "grad_norm": 4.621335820543163, + "learning_rate": 1e-06, + "loss": 0.2913, + "step": 3798 + }, + { + "epoch": 1.452772466539197, + "grad_norm": 2.8795663809181855, + "learning_rate": 1e-06, + "loss": 0.1822, + "step": 3799 + }, + { + "epoch": 1.4531548757170172, + "grad_norm": 2.873432561425201, + "learning_rate": 1e-06, + "loss": 0.1353, + "step": 3800 + }, + { + "epoch": 1.4535372848948374, + "grad_norm": 3.335319515404248, + "learning_rate": 1e-06, + "loss": 0.1563, + "step": 3801 + }, + { + "epoch": 1.4539196940726578, + "grad_norm": 4.237311101883541, + "learning_rate": 1e-06, + "loss": 0.0742, + "step": 3802 + }, + { + "epoch": 1.454302103250478, + "grad_norm": 6.744611496648465, + "learning_rate": 1e-06, + "loss": 0.2078, + "step": 3803 + }, + { + "epoch": 1.4546845124282983, + "grad_norm": 3.9523890115499265, + "learning_rate": 1e-06, + "loss": 0.391, + "step": 3804 + }, + { + "epoch": 1.4550669216061185, + "grad_norm": 4.015880769532698, + "learning_rate": 1e-06, + "loss": 0.5503, + "step": 3805 + }, + { + "epoch": 1.4554493307839387, + "grad_norm": 3.066642450845172, + "learning_rate": 1e-06, + "loss": 0.2996, + "step": 3806 + }, + { + "epoch": 1.455831739961759, + "grad_norm": 3.2684017951430984, + "learning_rate": 1e-06, + "loss": 0.4948, + "step": 3807 + }, + { + "epoch": 1.4562141491395795, + "grad_norm": 3.300095418070754, + "learning_rate": 1e-06, + "loss": 0.2781, + "step": 3808 + }, + { + "epoch": 1.4565965583173996, + "grad_norm": 3.419785092165752, + "learning_rate": 1e-06, + "loss": 0.1961, + "step": 3809 + }, + { + "epoch": 1.4569789674952198, + "grad_norm": 2.9591561547703815, + "learning_rate": 1e-06, + "loss": 0.3269, + "step": 3810 + }, + { + "epoch": 1.4573613766730402, + "grad_norm": 3.1471698944808435, + "learning_rate": 1e-06, + "loss": 0.1619, + "step": 3811 + }, + { + "epoch": 1.4577437858508604, + "grad_norm": 3.00136233232753, + "learning_rate": 1e-06, + "loss": 0.1578, + "step": 3812 + }, + { + "epoch": 1.4581261950286808, + "grad_norm": 3.375788561215613, + "learning_rate": 1e-06, + "loss": 0.2613, + "step": 3813 + }, + { + "epoch": 1.458508604206501, + "grad_norm": 2.9304037210464284, + "learning_rate": 1e-06, + "loss": 0.086, + "step": 3814 + }, + { + "epoch": 1.4588910133843211, + "grad_norm": 3.3510198264951336, + "learning_rate": 1e-06, + "loss": 0.1134, + "step": 3815 + }, + { + "epoch": 1.4592734225621415, + "grad_norm": 3.97487978003701, + "learning_rate": 1e-06, + "loss": 0.4668, + "step": 3816 + }, + { + "epoch": 1.4596558317399617, + "grad_norm": 3.71832054928221, + "learning_rate": 1e-06, + "loss": 0.4395, + "step": 3817 + }, + { + "epoch": 1.4600382409177821, + "grad_norm": 4.98342743001193, + "learning_rate": 1e-06, + "loss": 0.5187, + "step": 3818 + }, + { + "epoch": 1.4604206500956023, + "grad_norm": 3.080457239518899, + "learning_rate": 1e-06, + "loss": 0.2241, + "step": 3819 + }, + { + "epoch": 1.4608030592734225, + "grad_norm": 4.7729919525340785, + "learning_rate": 1e-06, + "loss": 0.2974, + "step": 3820 + }, + { + "epoch": 1.4611854684512429, + "grad_norm": 3.419843375428673, + "learning_rate": 1e-06, + "loss": 0.2616, + "step": 3821 + }, + { + "epoch": 1.461567877629063, + "grad_norm": 2.9801685987052786, + "learning_rate": 1e-06, + "loss": 0.2581, + "step": 3822 + }, + { + "epoch": 1.4619502868068834, + "grad_norm": 2.9065922361287666, + "learning_rate": 1e-06, + "loss": 0.168, + "step": 3823 + }, + { + "epoch": 1.4623326959847036, + "grad_norm": 3.7695722350427348, + "learning_rate": 1e-06, + "loss": 0.152, + "step": 3824 + }, + { + "epoch": 1.4627151051625238, + "grad_norm": 3.174334394717624, + "learning_rate": 1e-06, + "loss": 0.1804, + "step": 3825 + }, + { + "epoch": 1.4630975143403442, + "grad_norm": 4.837848045583718, + "learning_rate": 1e-06, + "loss": 0.1235, + "step": 3826 + }, + { + "epoch": 1.4634799235181644, + "grad_norm": 3.2905538263258594, + "learning_rate": 1e-06, + "loss": 0.0845, + "step": 3827 + }, + { + "epoch": 1.4638623326959848, + "grad_norm": 9.019931762843202, + "learning_rate": 1e-06, + "loss": 0.1512, + "step": 3828 + }, + { + "epoch": 1.464244741873805, + "grad_norm": 4.02266494592974, + "learning_rate": 1e-06, + "loss": 0.4447, + "step": 3829 + }, + { + "epoch": 1.464627151051625, + "grad_norm": 3.7410718811846464, + "learning_rate": 1e-06, + "loss": 0.2839, + "step": 3830 + }, + { + "epoch": 1.4650095602294455, + "grad_norm": 2.7484205218338853, + "learning_rate": 1e-06, + "loss": 0.1299, + "step": 3831 + }, + { + "epoch": 1.465391969407266, + "grad_norm": 3.654099573775783, + "learning_rate": 1e-06, + "loss": 0.3325, + "step": 3832 + }, + { + "epoch": 1.465774378585086, + "grad_norm": 2.777116940622431, + "learning_rate": 1e-06, + "loss": 0.2626, + "step": 3833 + }, + { + "epoch": 1.4661567877629063, + "grad_norm": 3.1262529531630423, + "learning_rate": 1e-06, + "loss": 0.1753, + "step": 3834 + }, + { + "epoch": 1.4665391969407267, + "grad_norm": 2.904293427147836, + "learning_rate": 1e-06, + "loss": 0.1693, + "step": 3835 + }, + { + "epoch": 1.4669216061185468, + "grad_norm": 2.8254094865133252, + "learning_rate": 1e-06, + "loss": 0.175, + "step": 3836 + }, + { + "epoch": 1.4673040152963672, + "grad_norm": 4.781284307219646, + "learning_rate": 1e-06, + "loss": 0.1232, + "step": 3837 + }, + { + "epoch": 1.4676864244741874, + "grad_norm": 3.1129096562075103, + "learning_rate": 1e-06, + "loss": 0.0863, + "step": 3838 + }, + { + "epoch": 1.4680688336520076, + "grad_norm": 4.842424568215032, + "learning_rate": 1e-06, + "loss": 0.1054, + "step": 3839 + }, + { + "epoch": 1.468451242829828, + "grad_norm": 3.693478667593567, + "learning_rate": 1e-06, + "loss": 0.105, + "step": 3840 + }, + { + "epoch": 1.4688336520076481, + "grad_norm": 8.429961729494517, + "learning_rate": 1e-06, + "loss": 0.3211, + "step": 3841 + }, + { + "epoch": 1.4692160611854685, + "grad_norm": 3.267691364594895, + "learning_rate": 1e-06, + "loss": 0.2938, + "step": 3842 + }, + { + "epoch": 1.4695984703632887, + "grad_norm": 3.8187200117416533, + "learning_rate": 1e-06, + "loss": 0.3805, + "step": 3843 + }, + { + "epoch": 1.469980879541109, + "grad_norm": 3.5728148658122003, + "learning_rate": 1e-06, + "loss": 0.2267, + "step": 3844 + }, + { + "epoch": 1.4703632887189293, + "grad_norm": 3.0617581461566554, + "learning_rate": 1e-06, + "loss": 0.2212, + "step": 3845 + }, + { + "epoch": 1.4707456978967495, + "grad_norm": 2.946903517454639, + "learning_rate": 1e-06, + "loss": 0.3188, + "step": 3846 + }, + { + "epoch": 1.4711281070745699, + "grad_norm": 2.6753001427941547, + "learning_rate": 1e-06, + "loss": 0.1129, + "step": 3847 + }, + { + "epoch": 1.47151051625239, + "grad_norm": 2.5027372633796934, + "learning_rate": 1e-06, + "loss": 0.0802, + "step": 3848 + }, + { + "epoch": 1.4718929254302102, + "grad_norm": 4.395501195041953, + "learning_rate": 1e-06, + "loss": 0.15, + "step": 3849 + }, + { + "epoch": 1.4722753346080306, + "grad_norm": 3.773209527436749, + "learning_rate": 1e-06, + "loss": 0.1679, + "step": 3850 + }, + { + "epoch": 1.4726577437858508, + "grad_norm": 3.860204445361068, + "learning_rate": 1e-06, + "loss": 0.1024, + "step": 3851 + }, + { + "epoch": 1.4730401529636712, + "grad_norm": 3.530051509059919, + "learning_rate": 1e-06, + "loss": 0.0615, + "step": 3852 + }, + { + "epoch": 1.4734225621414914, + "grad_norm": 5.866956033217151, + "learning_rate": 1e-06, + "loss": 0.1333, + "step": 3853 + }, + { + "epoch": 1.4738049713193115, + "grad_norm": 6.358621238633143, + "learning_rate": 1e-06, + "loss": 0.5757, + "step": 3854 + }, + { + "epoch": 1.474187380497132, + "grad_norm": 4.513029943399394, + "learning_rate": 1e-06, + "loss": 0.4743, + "step": 3855 + }, + { + "epoch": 1.4745697896749523, + "grad_norm": 2.8871868409602897, + "learning_rate": 1e-06, + "loss": 0.247, + "step": 3856 + }, + { + "epoch": 1.4749521988527725, + "grad_norm": 3.2121236795957935, + "learning_rate": 1e-06, + "loss": 0.1984, + "step": 3857 + }, + { + "epoch": 1.4753346080305927, + "grad_norm": 2.2729066838628866, + "learning_rate": 1e-06, + "loss": 0.2743, + "step": 3858 + }, + { + "epoch": 1.475717017208413, + "grad_norm": 4.660401407397884, + "learning_rate": 1e-06, + "loss": 0.334, + "step": 3859 + }, + { + "epoch": 1.4760994263862333, + "grad_norm": 3.0518740602859404, + "learning_rate": 1e-06, + "loss": 0.2125, + "step": 3860 + }, + { + "epoch": 1.4764818355640537, + "grad_norm": 2.2497827636964605, + "learning_rate": 1e-06, + "loss": 0.1154, + "step": 3861 + }, + { + "epoch": 1.4768642447418738, + "grad_norm": 2.755924951326017, + "learning_rate": 1e-06, + "loss": 0.0984, + "step": 3862 + }, + { + "epoch": 1.477246653919694, + "grad_norm": 3.4864116925821222, + "learning_rate": 1e-06, + "loss": 0.1778, + "step": 3863 + }, + { + "epoch": 1.4776290630975144, + "grad_norm": 3.5730174562939037, + "learning_rate": 1e-06, + "loss": 0.0803, + "step": 3864 + }, + { + "epoch": 1.4780114722753346, + "grad_norm": 3.129699225553734, + "learning_rate": 1e-06, + "loss": 0.0986, + "step": 3865 + }, + { + "epoch": 1.478393881453155, + "grad_norm": 3.894855328910445, + "learning_rate": 1e-06, + "loss": 0.3113, + "step": 3866 + }, + { + "epoch": 1.4787762906309752, + "grad_norm": 3.941234452555413, + "learning_rate": 1e-06, + "loss": 0.3866, + "step": 3867 + }, + { + "epoch": 1.4791586998087953, + "grad_norm": 5.609048222876385, + "learning_rate": 1e-06, + "loss": 0.3934, + "step": 3868 + }, + { + "epoch": 1.4795411089866157, + "grad_norm": 2.3695421248179196, + "learning_rate": 1e-06, + "loss": 0.2293, + "step": 3869 + }, + { + "epoch": 1.479923518164436, + "grad_norm": 3.562778261259576, + "learning_rate": 1e-06, + "loss": 0.259, + "step": 3870 + }, + { + "epoch": 1.4803059273422563, + "grad_norm": 3.198581506331975, + "learning_rate": 1e-06, + "loss": 0.229, + "step": 3871 + }, + { + "epoch": 1.4806883365200765, + "grad_norm": 2.1479781335964403, + "learning_rate": 1e-06, + "loss": 0.1967, + "step": 3872 + }, + { + "epoch": 1.4810707456978967, + "grad_norm": 3.325525549176048, + "learning_rate": 1e-06, + "loss": 0.1255, + "step": 3873 + }, + { + "epoch": 1.481453154875717, + "grad_norm": 4.47129439897385, + "learning_rate": 1e-06, + "loss": 0.1792, + "step": 3874 + }, + { + "epoch": 1.4818355640535372, + "grad_norm": 3.019859069163232, + "learning_rate": 1e-06, + "loss": 0.1168, + "step": 3875 + }, + { + "epoch": 1.4822179732313576, + "grad_norm": 4.957213245895642, + "learning_rate": 1e-06, + "loss": 0.1112, + "step": 3876 + }, + { + "epoch": 1.4826003824091778, + "grad_norm": 3.5225665248764537, + "learning_rate": 1e-06, + "loss": 0.0608, + "step": 3877 + }, + { + "epoch": 1.482982791586998, + "grad_norm": 5.737249417951869, + "learning_rate": 1e-06, + "loss": 0.1176, + "step": 3878 + }, + { + "epoch": 1.4833652007648184, + "grad_norm": 3.5995974527576573, + "learning_rate": 1e-06, + "loss": 0.3218, + "step": 3879 + }, + { + "epoch": 1.4837476099426385, + "grad_norm": 4.423747714156061, + "learning_rate": 1e-06, + "loss": 0.4048, + "step": 3880 + }, + { + "epoch": 1.484130019120459, + "grad_norm": 2.0583476257857614, + "learning_rate": 1e-06, + "loss": 0.2074, + "step": 3881 + }, + { + "epoch": 1.4845124282982791, + "grad_norm": 3.4366762041119294, + "learning_rate": 1e-06, + "loss": 0.1247, + "step": 3882 + }, + { + "epoch": 1.4848948374760993, + "grad_norm": 3.6979692195686207, + "learning_rate": 1e-06, + "loss": 0.298, + "step": 3883 + }, + { + "epoch": 1.4852772466539197, + "grad_norm": 2.3355698312108766, + "learning_rate": 1e-06, + "loss": 0.1487, + "step": 3884 + }, + { + "epoch": 1.48565965583174, + "grad_norm": 1.807286492449328, + "learning_rate": 1e-06, + "loss": 0.0802, + "step": 3885 + }, + { + "epoch": 1.4860420650095603, + "grad_norm": 3.6630470153807266, + "learning_rate": 1e-06, + "loss": 0.2576, + "step": 3886 + }, + { + "epoch": 1.4864244741873804, + "grad_norm": 2.801169566710892, + "learning_rate": 1e-06, + "loss": 0.0663, + "step": 3887 + }, + { + "epoch": 1.4868068833652008, + "grad_norm": 2.9192060224613656, + "learning_rate": 1e-06, + "loss": 0.1058, + "step": 3888 + }, + { + "epoch": 1.487189292543021, + "grad_norm": 2.5053213705995705, + "learning_rate": 1e-06, + "loss": 0.069, + "step": 3889 + }, + { + "epoch": 1.4875717017208414, + "grad_norm": 4.644347268223121, + "learning_rate": 1e-06, + "loss": 0.0938, + "step": 3890 + }, + { + "epoch": 1.4879541108986616, + "grad_norm": 5.329618272692444, + "learning_rate": 1e-06, + "loss": 0.4061, + "step": 3891 + }, + { + "epoch": 1.4883365200764818, + "grad_norm": 4.072459527757717, + "learning_rate": 1e-06, + "loss": 0.4351, + "step": 3892 + }, + { + "epoch": 1.4887189292543022, + "grad_norm": 3.5818813546370136, + "learning_rate": 1e-06, + "loss": 0.2766, + "step": 3893 + }, + { + "epoch": 1.4891013384321223, + "grad_norm": 2.5468588986502154, + "learning_rate": 1e-06, + "loss": 0.283, + "step": 3894 + }, + { + "epoch": 1.4894837476099427, + "grad_norm": 2.7777776580386666, + "learning_rate": 1e-06, + "loss": 0.4183, + "step": 3895 + }, + { + "epoch": 1.489866156787763, + "grad_norm": 3.3154771489607286, + "learning_rate": 1e-06, + "loss": 0.2987, + "step": 3896 + }, + { + "epoch": 1.490248565965583, + "grad_norm": 3.2799520195963985, + "learning_rate": 1e-06, + "loss": 0.1745, + "step": 3897 + }, + { + "epoch": 1.4906309751434035, + "grad_norm": 2.9072710110076274, + "learning_rate": 1e-06, + "loss": 0.1737, + "step": 3898 + }, + { + "epoch": 1.4910133843212237, + "grad_norm": 2.999697034160772, + "learning_rate": 1e-06, + "loss": 0.2543, + "step": 3899 + }, + { + "epoch": 1.491395793499044, + "grad_norm": 2.356610124423201, + "learning_rate": 1e-06, + "loss": 0.1466, + "step": 3900 + }, + { + "epoch": 1.4917782026768642, + "grad_norm": 3.3811958802240865, + "learning_rate": 1e-06, + "loss": 0.1331, + "step": 3901 + }, + { + "epoch": 1.4921606118546844, + "grad_norm": 3.9710000936123433, + "learning_rate": 1e-06, + "loss": 0.0954, + "step": 3902 + }, + { + "epoch": 1.4925430210325048, + "grad_norm": 6.990058105838466, + "learning_rate": 1e-06, + "loss": 0.2311, + "step": 3903 + }, + { + "epoch": 1.492925430210325, + "grad_norm": 4.267945716791568, + "learning_rate": 1e-06, + "loss": 0.4079, + "step": 3904 + }, + { + "epoch": 1.4933078393881454, + "grad_norm": 3.1010631091976446, + "learning_rate": 1e-06, + "loss": 0.3872, + "step": 3905 + }, + { + "epoch": 1.4936902485659656, + "grad_norm": 3.8123869331978844, + "learning_rate": 1e-06, + "loss": 0.2134, + "step": 3906 + }, + { + "epoch": 1.4940726577437857, + "grad_norm": 3.3967965910985645, + "learning_rate": 1e-06, + "loss": 0.1731, + "step": 3907 + }, + { + "epoch": 1.4944550669216061, + "grad_norm": 3.937442476367078, + "learning_rate": 1e-06, + "loss": 0.3576, + "step": 3908 + }, + { + "epoch": 1.4948374760994265, + "grad_norm": 2.028552332736425, + "learning_rate": 1e-06, + "loss": 0.1715, + "step": 3909 + }, + { + "epoch": 1.4952198852772467, + "grad_norm": 2.9499790449529675, + "learning_rate": 1e-06, + "loss": 0.1708, + "step": 3910 + }, + { + "epoch": 1.4956022944550669, + "grad_norm": 4.119911754737143, + "learning_rate": 1e-06, + "loss": 0.1522, + "step": 3911 + }, + { + "epoch": 1.4959847036328873, + "grad_norm": 3.274174500304965, + "learning_rate": 1e-06, + "loss": 0.1774, + "step": 3912 + }, + { + "epoch": 1.4963671128107074, + "grad_norm": 3.4899167181533137, + "learning_rate": 1e-06, + "loss": 0.1361, + "step": 3913 + }, + { + "epoch": 1.4967495219885278, + "grad_norm": 4.463253882918679, + "learning_rate": 1e-06, + "loss": 0.1868, + "step": 3914 + }, + { + "epoch": 1.497131931166348, + "grad_norm": 3.775710911789235, + "learning_rate": 1e-06, + "loss": 0.0805, + "step": 3915 + }, + { + "epoch": 1.4975143403441682, + "grad_norm": 6.183630455809408, + "learning_rate": 1e-06, + "loss": 0.2829, + "step": 3916 + }, + { + "epoch": 1.4978967495219886, + "grad_norm": 3.0684675346945762, + "learning_rate": 1e-06, + "loss": 0.4096, + "step": 3917 + }, + { + "epoch": 1.4982791586998088, + "grad_norm": 3.9605638075693372, + "learning_rate": 1e-06, + "loss": 0.3544, + "step": 3918 + }, + { + "epoch": 1.4986615678776292, + "grad_norm": 2.305883168558684, + "learning_rate": 1e-06, + "loss": 0.2506, + "step": 3919 + }, + { + "epoch": 1.4990439770554493, + "grad_norm": 2.8661289183856753, + "learning_rate": 1e-06, + "loss": 0.2661, + "step": 3920 + }, + { + "epoch": 1.4994263862332695, + "grad_norm": 3.0950222289438005, + "learning_rate": 1e-06, + "loss": 0.3595, + "step": 3921 + }, + { + "epoch": 1.49980879541109, + "grad_norm": 3.3450728322127183, + "learning_rate": 1e-06, + "loss": 0.194, + "step": 3922 + }, + { + "epoch": 1.50019120458891, + "grad_norm": 2.816971932371236, + "learning_rate": 1e-06, + "loss": 0.151, + "step": 3923 + }, + { + "epoch": 1.5005736137667305, + "grad_norm": 5.180212210060488, + "learning_rate": 1e-06, + "loss": 0.2595, + "step": 3924 + }, + { + "epoch": 1.5009560229445507, + "grad_norm": 4.4262759680472366, + "learning_rate": 1e-06, + "loss": 0.1916, + "step": 3925 + }, + { + "epoch": 1.5013384321223708, + "grad_norm": 5.305925497280416, + "learning_rate": 1e-06, + "loss": 0.1748, + "step": 3926 + }, + { + "epoch": 1.5017208413001912, + "grad_norm": 6.767331231968318, + "learning_rate": 1e-06, + "loss": 0.0759, + "step": 3927 + }, + { + "epoch": 1.5021032504780116, + "grad_norm": 4.925718907590221, + "learning_rate": 1e-06, + "loss": 0.1051, + "step": 3928 + }, + { + "epoch": 1.5024856596558318, + "grad_norm": 4.954566914159578, + "learning_rate": 1e-06, + "loss": 0.4665, + "step": 3929 + }, + { + "epoch": 1.502868068833652, + "grad_norm": 2.9394148409051746, + "learning_rate": 1e-06, + "loss": 0.3569, + "step": 3930 + }, + { + "epoch": 1.5032504780114722, + "grad_norm": 2.4671448909856215, + "learning_rate": 1e-06, + "loss": 0.1468, + "step": 3931 + }, + { + "epoch": 1.5036328871892926, + "grad_norm": 2.6880407454717656, + "learning_rate": 1e-06, + "loss": 0.2397, + "step": 3932 + }, + { + "epoch": 1.504015296367113, + "grad_norm": 4.153998891081867, + "learning_rate": 1e-06, + "loss": 0.2802, + "step": 3933 + }, + { + "epoch": 1.5043977055449331, + "grad_norm": 2.1257717750615743, + "learning_rate": 1e-06, + "loss": 0.1334, + "step": 3934 + }, + { + "epoch": 1.5047801147227533, + "grad_norm": 3.097527403875748, + "learning_rate": 1e-06, + "loss": 0.2852, + "step": 3935 + }, + { + "epoch": 1.5051625239005735, + "grad_norm": 2.4833219203644044, + "learning_rate": 1e-06, + "loss": 0.1797, + "step": 3936 + }, + { + "epoch": 1.5055449330783939, + "grad_norm": 1.8667316598590287, + "learning_rate": 1e-06, + "loss": 0.0644, + "step": 3937 + }, + { + "epoch": 1.5059273422562143, + "grad_norm": 3.546294177388295, + "learning_rate": 1e-06, + "loss": 0.161, + "step": 3938 + }, + { + "epoch": 1.5063097514340344, + "grad_norm": 2.9370517794571485, + "learning_rate": 1e-06, + "loss": 0.122, + "step": 3939 + }, + { + "epoch": 1.5066921606118546, + "grad_norm": 4.558320521883538, + "learning_rate": 1e-06, + "loss": 0.1002, + "step": 3940 + }, + { + "epoch": 1.5070745697896748, + "grad_norm": 4.809699891179583, + "learning_rate": 1e-06, + "loss": 0.54, + "step": 3941 + }, + { + "epoch": 1.5074569789674952, + "grad_norm": 3.744581376680773, + "learning_rate": 1e-06, + "loss": 0.4046, + "step": 3942 + }, + { + "epoch": 1.5078393881453156, + "grad_norm": 7.3905973857339555, + "learning_rate": 1e-06, + "loss": 0.4947, + "step": 3943 + }, + { + "epoch": 1.5082217973231358, + "grad_norm": 2.8259975798772667, + "learning_rate": 1e-06, + "loss": 0.2338, + "step": 3944 + }, + { + "epoch": 1.508604206500956, + "grad_norm": 2.8417804311165136, + "learning_rate": 1e-06, + "loss": 0.1986, + "step": 3945 + }, + { + "epoch": 1.5089866156787763, + "grad_norm": 2.783702840281398, + "learning_rate": 1e-06, + "loss": 0.3213, + "step": 3946 + }, + { + "epoch": 1.5093690248565965, + "grad_norm": 2.23992316454852, + "learning_rate": 1e-06, + "loss": 0.128, + "step": 3947 + }, + { + "epoch": 1.509751434034417, + "grad_norm": 2.86368157267148, + "learning_rate": 1e-06, + "loss": 0.2186, + "step": 3948 + }, + { + "epoch": 1.510133843212237, + "grad_norm": 2.986499289977013, + "learning_rate": 1e-06, + "loss": 0.1047, + "step": 3949 + }, + { + "epoch": 1.5105162523900573, + "grad_norm": 2.602839709123787, + "learning_rate": 1e-06, + "loss": 0.0556, + "step": 3950 + }, + { + "epoch": 1.5108986615678777, + "grad_norm": 3.4712862931999586, + "learning_rate": 1e-06, + "loss": 0.1932, + "step": 3951 + }, + { + "epoch": 1.511281070745698, + "grad_norm": 4.825550170523791, + "learning_rate": 1e-06, + "loss": 0.0981, + "step": 3952 + }, + { + "epoch": 1.5116634799235182, + "grad_norm": 5.906107582918369, + "learning_rate": 1e-06, + "loss": 0.1945, + "step": 3953 + }, + { + "epoch": 1.5120458891013384, + "grad_norm": 4.152000147472225, + "learning_rate": 1e-06, + "loss": 0.6537, + "step": 3954 + }, + { + "epoch": 1.5124282982791586, + "grad_norm": 2.9336616924763703, + "learning_rate": 1e-06, + "loss": 0.3252, + "step": 3955 + }, + { + "epoch": 1.512810707456979, + "grad_norm": 1.758289865736911, + "learning_rate": 1e-06, + "loss": 0.2387, + "step": 3956 + }, + { + "epoch": 1.5131931166347994, + "grad_norm": 4.006370478349455, + "learning_rate": 1e-06, + "loss": 0.2336, + "step": 3957 + }, + { + "epoch": 1.5135755258126196, + "grad_norm": 3.675672020743574, + "learning_rate": 1e-06, + "loss": 0.2642, + "step": 3958 + }, + { + "epoch": 1.5139579349904397, + "grad_norm": 3.4303843978487882, + "learning_rate": 1e-06, + "loss": 0.1566, + "step": 3959 + }, + { + "epoch": 1.51434034416826, + "grad_norm": 2.675526583170763, + "learning_rate": 1e-06, + "loss": 0.115, + "step": 3960 + }, + { + "epoch": 1.5147227533460803, + "grad_norm": 3.04067754121807, + "learning_rate": 1e-06, + "loss": 0.1148, + "step": 3961 + }, + { + "epoch": 1.5151051625239007, + "grad_norm": 4.519021416792543, + "learning_rate": 1e-06, + "loss": 0.1472, + "step": 3962 + }, + { + "epoch": 1.5154875717017209, + "grad_norm": 3.2548776850300896, + "learning_rate": 1e-06, + "loss": 0.1106, + "step": 3963 + }, + { + "epoch": 1.515869980879541, + "grad_norm": 3.9135161911344487, + "learning_rate": 1e-06, + "loss": 0.087, + "step": 3964 + }, + { + "epoch": 1.5162523900573612, + "grad_norm": 4.010244839773097, + "learning_rate": 1e-06, + "loss": 0.1205, + "step": 3965 + }, + { + "epoch": 1.5166347992351816, + "grad_norm": 4.56472932880087, + "learning_rate": 1e-06, + "loss": 0.4076, + "step": 3966 + }, + { + "epoch": 1.517017208413002, + "grad_norm": 5.479372008601221, + "learning_rate": 1e-06, + "loss": 0.4973, + "step": 3967 + }, + { + "epoch": 1.5173996175908222, + "grad_norm": 4.299789494529427, + "learning_rate": 1e-06, + "loss": 0.4959, + "step": 3968 + }, + { + "epoch": 1.5177820267686424, + "grad_norm": 2.925059821666413, + "learning_rate": 1e-06, + "loss": 0.1516, + "step": 3969 + }, + { + "epoch": 1.5181644359464626, + "grad_norm": 2.8787240497381847, + "learning_rate": 1e-06, + "loss": 0.2397, + "step": 3970 + }, + { + "epoch": 1.518546845124283, + "grad_norm": 3.7031709692310937, + "learning_rate": 1e-06, + "loss": 0.2515, + "step": 3971 + }, + { + "epoch": 1.5189292543021033, + "grad_norm": 1.955050686425526, + "learning_rate": 1e-06, + "loss": 0.1695, + "step": 3972 + }, + { + "epoch": 1.5193116634799235, + "grad_norm": 3.113865662290679, + "learning_rate": 1e-06, + "loss": 0.1802, + "step": 3973 + }, + { + "epoch": 1.5196940726577437, + "grad_norm": 3.2465922756667407, + "learning_rate": 1e-06, + "loss": 0.2266, + "step": 3974 + }, + { + "epoch": 1.520076481835564, + "grad_norm": 2.9431008046262033, + "learning_rate": 1e-06, + "loss": 0.1225, + "step": 3975 + }, + { + "epoch": 1.5204588910133843, + "grad_norm": 3.030820993169864, + "learning_rate": 1e-06, + "loss": 0.0523, + "step": 3976 + }, + { + "epoch": 1.5208413001912047, + "grad_norm": 3.3309977137094817, + "learning_rate": 1e-06, + "loss": 0.0825, + "step": 3977 + }, + { + "epoch": 1.5212237093690248, + "grad_norm": 35.00118231819998, + "learning_rate": 1e-06, + "loss": 0.2887, + "step": 3978 + }, + { + "epoch": 1.521606118546845, + "grad_norm": 5.631251950956932, + "learning_rate": 1e-06, + "loss": 0.3441, + "step": 3979 + }, + { + "epoch": 1.5219885277246654, + "grad_norm": 3.851298046511904, + "learning_rate": 1e-06, + "loss": 0.3802, + "step": 3980 + }, + { + "epoch": 1.5223709369024858, + "grad_norm": 4.1656829944416875, + "learning_rate": 1e-06, + "loss": 0.4271, + "step": 3981 + }, + { + "epoch": 1.522753346080306, + "grad_norm": 2.903704441089205, + "learning_rate": 1e-06, + "loss": 0.2593, + "step": 3982 + }, + { + "epoch": 1.5231357552581262, + "grad_norm": 2.0742226580854193, + "learning_rate": 1e-06, + "loss": 0.1707, + "step": 3983 + }, + { + "epoch": 1.5235181644359463, + "grad_norm": 3.663700175315848, + "learning_rate": 1e-06, + "loss": 0.2572, + "step": 3984 + }, + { + "epoch": 1.5239005736137667, + "grad_norm": 2.596237460580444, + "learning_rate": 1e-06, + "loss": 0.0764, + "step": 3985 + }, + { + "epoch": 1.5242829827915871, + "grad_norm": 3.001227445632857, + "learning_rate": 1e-06, + "loss": 0.1724, + "step": 3986 + }, + { + "epoch": 1.5246653919694073, + "grad_norm": 3.2115662045117355, + "learning_rate": 1e-06, + "loss": 0.1363, + "step": 3987 + }, + { + "epoch": 1.5250478011472275, + "grad_norm": 2.402127189489529, + "learning_rate": 1e-06, + "loss": 0.0842, + "step": 3988 + }, + { + "epoch": 1.5254302103250477, + "grad_norm": 3.184964836082194, + "learning_rate": 1e-06, + "loss": 0.0764, + "step": 3989 + }, + { + "epoch": 1.525812619502868, + "grad_norm": 4.7110279890640765, + "learning_rate": 1e-06, + "loss": 0.2332, + "step": 3990 + }, + { + "epoch": 1.5261950286806885, + "grad_norm": 6.3784233326273565, + "learning_rate": 1e-06, + "loss": 0.1823, + "step": 3991 + }, + { + "epoch": 1.5265774378585086, + "grad_norm": 5.270771533870147, + "learning_rate": 1e-06, + "loss": 0.5324, + "step": 3992 + }, + { + "epoch": 1.5269598470363288, + "grad_norm": 2.952825298945698, + "learning_rate": 1e-06, + "loss": 0.339, + "step": 3993 + }, + { + "epoch": 1.527342256214149, + "grad_norm": 4.203740298248357, + "learning_rate": 1e-06, + "loss": 0.2524, + "step": 3994 + }, + { + "epoch": 1.5277246653919694, + "grad_norm": 3.2305146551964388, + "learning_rate": 1e-06, + "loss": 0.2367, + "step": 3995 + }, + { + "epoch": 1.5281070745697898, + "grad_norm": 3.0066196206493974, + "learning_rate": 1e-06, + "loss": 0.4437, + "step": 3996 + }, + { + "epoch": 1.52848948374761, + "grad_norm": 3.164315823433579, + "learning_rate": 1e-06, + "loss": 0.1544, + "step": 3997 + }, + { + "epoch": 1.5288718929254301, + "grad_norm": 2.617455676558822, + "learning_rate": 1e-06, + "loss": 0.1725, + "step": 3998 + }, + { + "epoch": 1.5292543021032505, + "grad_norm": 3.616000085628137, + "learning_rate": 1e-06, + "loss": 0.2934, + "step": 3999 + }, + { + "epoch": 1.5296367112810707, + "grad_norm": 3.760712137352352, + "learning_rate": 1e-06, + "loss": 0.1458, + "step": 4000 + }, + { + "epoch": 1.5296367112810707, + "eval_runtime": 840.9805, + "eval_samples_per_second": 1.824, + "eval_steps_per_second": 0.457, + "step": 4000 + } + ], + "logging_steps": 1.0, + "max_steps": 13075, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.409350742875046e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}