{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.30353619669145543, "eval_steps": 500, "global_step": 1000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00030353619669145547, "grad_norm": 9.667811393737793, "learning_rate": 1e-05, "loss": 5.0202, "step": 1 }, { "epoch": 0.0006070723933829109, "grad_norm": 10.303421974182129, "learning_rate": 2e-05, "loss": 4.7469, "step": 2 }, { "epoch": 0.0009106085900743664, "grad_norm": 7.488056182861328, "learning_rate": 3e-05, "loss": 5.0105, "step": 3 }, { "epoch": 0.0012141447867658219, "grad_norm": 4.885837078094482, "learning_rate": 4e-05, "loss": 4.3945, "step": 4 }, { "epoch": 0.0015176809834572772, "grad_norm": 3.793656587600708, "learning_rate": 5e-05, "loss": 4.0574, "step": 5 }, { "epoch": 0.0018212171801487327, "grad_norm": 3.9249916076660156, "learning_rate": 6e-05, "loss": 3.8179, "step": 6 }, { "epoch": 0.002124753376840188, "grad_norm": 3.4937145709991455, "learning_rate": 7e-05, "loss": 3.5297, "step": 7 }, { "epoch": 0.0024282895735316438, "grad_norm": 2.499041795730591, "learning_rate": 8e-05, "loss": 3.15, "step": 8 }, { "epoch": 0.002731825770223099, "grad_norm": 2.0781290531158447, "learning_rate": 9e-05, "loss": 2.8658, "step": 9 }, { "epoch": 0.0030353619669145544, "grad_norm": 2.0124764442443848, "learning_rate": 0.0001, "loss": 2.6826, "step": 10 }, { "epoch": 0.00333889816360601, "grad_norm": 1.4209256172180176, "learning_rate": 9.99949377341298e-05, "loss": 2.5608, "step": 11 }, { "epoch": 0.0036424343602974654, "grad_norm": 3.176084041595459, "learning_rate": 9.99898754682596e-05, "loss": 2.2416, "step": 12 }, { "epoch": 0.003945970556988921, "grad_norm": 1.4457614421844482, "learning_rate": 9.998481320238939e-05, "loss": 2.1925, "step": 13 }, { "epoch": 0.004249506753680376, "grad_norm": 1.3989348411560059, "learning_rate": 9.997975093651918e-05, "loss": 2.2165, "step": 14 }, { "epoch": 0.004553042950371832, "grad_norm": 1.0647027492523193, "learning_rate": 9.997468867064899e-05, "loss": 2.3486, "step": 15 }, { "epoch": 0.0048565791470632875, "grad_norm": 1.0246940851211548, "learning_rate": 9.996962640477879e-05, "loss": 2.19, "step": 16 }, { "epoch": 0.005160115343754742, "grad_norm": 1.029646873474121, "learning_rate": 9.996456413890858e-05, "loss": 2.4052, "step": 17 }, { "epoch": 0.005463651540446198, "grad_norm": 1.322654128074646, "learning_rate": 9.995950187303838e-05, "loss": 2.1927, "step": 18 }, { "epoch": 0.005767187737137654, "grad_norm": 2.061326026916504, "learning_rate": 9.995443960716817e-05, "loss": 2.4574, "step": 19 }, { "epoch": 0.006070723933829109, "grad_norm": 1.1343607902526855, "learning_rate": 9.994937734129797e-05, "loss": 1.9598, "step": 20 }, { "epoch": 0.0063742601305205645, "grad_norm": 1.13712477684021, "learning_rate": 9.994431507542776e-05, "loss": 2.8643, "step": 21 }, { "epoch": 0.00667779632721202, "grad_norm": 0.8220421671867371, "learning_rate": 9.993925280955756e-05, "loss": 2.0474, "step": 22 }, { "epoch": 0.006981332523903475, "grad_norm": 0.8233473300933838, "learning_rate": 9.993419054368735e-05, "loss": 2.3597, "step": 23 }, { "epoch": 0.007284868720594931, "grad_norm": 0.8661925196647644, "learning_rate": 9.992912827781716e-05, "loss": 2.2163, "step": 24 }, { "epoch": 0.007588404917286387, "grad_norm": 0.7995729446411133, "learning_rate": 9.992406601194695e-05, "loss": 1.8051, "step": 25 }, { "epoch": 0.007891941113977842, "grad_norm": 0.810165286064148, "learning_rate": 9.991900374607675e-05, "loss": 1.9189, "step": 26 }, { "epoch": 0.008195477310669297, "grad_norm": 0.8240752220153809, "learning_rate": 9.991394148020654e-05, "loss": 1.7, "step": 27 }, { "epoch": 0.008499013507360752, "grad_norm": 1.0160635709762573, "learning_rate": 9.990887921433634e-05, "loss": 2.2964, "step": 28 }, { "epoch": 0.008802549704052209, "grad_norm": 0.794966995716095, "learning_rate": 9.990381694846613e-05, "loss": 1.7333, "step": 29 }, { "epoch": 0.009106085900743664, "grad_norm": 0.5594797134399414, "learning_rate": 9.989875468259593e-05, "loss": 2.0925, "step": 30 }, { "epoch": 0.009409622097435118, "grad_norm": 0.8100740909576416, "learning_rate": 9.989369241672572e-05, "loss": 2.1218, "step": 31 }, { "epoch": 0.009713158294126575, "grad_norm": 0.7057996392250061, "learning_rate": 9.988863015085552e-05, "loss": 2.005, "step": 32 }, { "epoch": 0.01001669449081803, "grad_norm": 0.8970999121665955, "learning_rate": 9.988356788498533e-05, "loss": 2.2414, "step": 33 }, { "epoch": 0.010320230687509485, "grad_norm": 0.6290627717971802, "learning_rate": 9.987850561911512e-05, "loss": 2.2422, "step": 34 }, { "epoch": 0.010623766884200941, "grad_norm": 0.5665722489356995, "learning_rate": 9.987344335324492e-05, "loss": 1.9342, "step": 35 }, { "epoch": 0.010927303080892396, "grad_norm": 0.5792561173439026, "learning_rate": 9.986838108737472e-05, "loss": 1.8733, "step": 36 }, { "epoch": 0.011230839277583851, "grad_norm": 0.5264159440994263, "learning_rate": 9.986331882150452e-05, "loss": 2.1739, "step": 37 }, { "epoch": 0.011534375474275308, "grad_norm": 0.5069584250450134, "learning_rate": 9.985825655563431e-05, "loss": 1.6235, "step": 38 }, { "epoch": 0.011837911670966763, "grad_norm": 0.7689110636711121, "learning_rate": 9.985319428976411e-05, "loss": 1.711, "step": 39 }, { "epoch": 0.012141447867658217, "grad_norm": 0.7001574635505676, "learning_rate": 9.98481320238939e-05, "loss": 1.651, "step": 40 }, { "epoch": 0.012444984064349674, "grad_norm": 0.5615801811218262, "learning_rate": 9.98430697580237e-05, "loss": 2.128, "step": 41 }, { "epoch": 0.012748520261041129, "grad_norm": 0.8766308426856995, "learning_rate": 9.983800749215349e-05, "loss": 2.4421, "step": 42 }, { "epoch": 0.013052056457732584, "grad_norm": 0.704547107219696, "learning_rate": 9.983294522628329e-05, "loss": 1.6921, "step": 43 }, { "epoch": 0.01335559265442404, "grad_norm": 0.5749143362045288, "learning_rate": 9.982788296041308e-05, "loss": 2.0173, "step": 44 }, { "epoch": 0.013659128851115495, "grad_norm": 0.7929263710975647, "learning_rate": 9.982282069454289e-05, "loss": 2.1755, "step": 45 }, { "epoch": 0.01396266504780695, "grad_norm": 1.6391934156417847, "learning_rate": 9.981775842867269e-05, "loss": 2.4995, "step": 46 }, { "epoch": 0.014266201244498407, "grad_norm": 0.49616461992263794, "learning_rate": 9.981269616280248e-05, "loss": 2.3363, "step": 47 }, { "epoch": 0.014569737441189862, "grad_norm": 0.614272952079773, "learning_rate": 9.980763389693227e-05, "loss": 2.0277, "step": 48 }, { "epoch": 0.014873273637881317, "grad_norm": 0.6181132197380066, "learning_rate": 9.980257163106207e-05, "loss": 2.2867, "step": 49 }, { "epoch": 0.015176809834572773, "grad_norm": 0.5342630743980408, "learning_rate": 9.979750936519186e-05, "loss": 1.7314, "step": 50 }, { "epoch": 0.015480346031264228, "grad_norm": 0.4582519233226776, "learning_rate": 9.979244709932166e-05, "loss": 1.9893, "step": 51 }, { "epoch": 0.015783882227955685, "grad_norm": 0.5448606014251709, "learning_rate": 9.978738483345145e-05, "loss": 2.3266, "step": 52 }, { "epoch": 0.01608741842464714, "grad_norm": 1.0823545455932617, "learning_rate": 9.978232256758125e-05, "loss": 2.1919, "step": 53 }, { "epoch": 0.016390954621338594, "grad_norm": 0.5506464838981628, "learning_rate": 9.977726030171106e-05, "loss": 2.0735, "step": 54 }, { "epoch": 0.01669449081803005, "grad_norm": 0.568626344203949, "learning_rate": 9.977219803584085e-05, "loss": 2.051, "step": 55 }, { "epoch": 0.016998027014721504, "grad_norm": 0.512907087802887, "learning_rate": 9.976713576997065e-05, "loss": 1.6473, "step": 56 }, { "epoch": 0.017301563211412962, "grad_norm": 0.5541898012161255, "learning_rate": 9.976207350410044e-05, "loss": 1.8184, "step": 57 }, { "epoch": 0.017605099408104417, "grad_norm": 0.5083638429641724, "learning_rate": 9.975701123823024e-05, "loss": 1.7573, "step": 58 }, { "epoch": 0.017908635604795872, "grad_norm": 0.4722895920276642, "learning_rate": 9.975194897236003e-05, "loss": 2.0311, "step": 59 }, { "epoch": 0.018212171801487327, "grad_norm": 0.5068002343177795, "learning_rate": 9.974688670648983e-05, "loss": 2.1245, "step": 60 }, { "epoch": 0.018515707998178782, "grad_norm": 0.5726852416992188, "learning_rate": 9.974182444061962e-05, "loss": 2.1017, "step": 61 }, { "epoch": 0.018819244194870237, "grad_norm": 0.5240160226821899, "learning_rate": 9.973676217474942e-05, "loss": 2.2665, "step": 62 }, { "epoch": 0.019122780391561695, "grad_norm": 0.4728144705295563, "learning_rate": 9.973169990887921e-05, "loss": 2.0537, "step": 63 }, { "epoch": 0.01942631658825315, "grad_norm": 0.47115418314933777, "learning_rate": 9.972663764300902e-05, "loss": 1.2815, "step": 64 }, { "epoch": 0.019729852784944605, "grad_norm": 0.7070208191871643, "learning_rate": 9.972157537713881e-05, "loss": 1.8514, "step": 65 }, { "epoch": 0.02003338898163606, "grad_norm": 0.529069185256958, "learning_rate": 9.971651311126861e-05, "loss": 1.7602, "step": 66 }, { "epoch": 0.020336925178327515, "grad_norm": 0.7532087564468384, "learning_rate": 9.97114508453984e-05, "loss": 2.2168, "step": 67 }, { "epoch": 0.02064046137501897, "grad_norm": 0.5654622912406921, "learning_rate": 9.97063885795282e-05, "loss": 1.9634, "step": 68 }, { "epoch": 0.020943997571710428, "grad_norm": 0.701452910900116, "learning_rate": 9.970132631365799e-05, "loss": 2.044, "step": 69 }, { "epoch": 0.021247533768401883, "grad_norm": 0.5750812888145447, "learning_rate": 9.969626404778779e-05, "loss": 1.8015, "step": 70 }, { "epoch": 0.021551069965093338, "grad_norm": 0.49930402636528015, "learning_rate": 9.969120178191758e-05, "loss": 1.7998, "step": 71 }, { "epoch": 0.021854606161784793, "grad_norm": 0.4348014295101166, "learning_rate": 9.968613951604738e-05, "loss": 1.9959, "step": 72 }, { "epoch": 0.022158142358476247, "grad_norm": 0.5268503427505493, "learning_rate": 9.968107725017719e-05, "loss": 1.8497, "step": 73 }, { "epoch": 0.022461678555167702, "grad_norm": 0.578822135925293, "learning_rate": 9.967601498430698e-05, "loss": 2.3277, "step": 74 }, { "epoch": 0.02276521475185916, "grad_norm": 0.52215975522995, "learning_rate": 9.967095271843677e-05, "loss": 2.1179, "step": 75 }, { "epoch": 0.023068750948550616, "grad_norm": 0.4557477533817291, "learning_rate": 9.966589045256657e-05, "loss": 2.0132, "step": 76 }, { "epoch": 0.02337228714524207, "grad_norm": 0.5032123327255249, "learning_rate": 9.966082818669638e-05, "loss": 1.8608, "step": 77 }, { "epoch": 0.023675823341933525, "grad_norm": 0.42689865827560425, "learning_rate": 9.965576592082617e-05, "loss": 2.0437, "step": 78 }, { "epoch": 0.02397935953862498, "grad_norm": 0.44310206174850464, "learning_rate": 9.965070365495597e-05, "loss": 2.1222, "step": 79 }, { "epoch": 0.024282895735316435, "grad_norm": 0.4377008378505707, "learning_rate": 9.964564138908576e-05, "loss": 2.0418, "step": 80 }, { "epoch": 0.024586431932007893, "grad_norm": 0.35174912214279175, "learning_rate": 9.964057912321556e-05, "loss": 1.6931, "step": 81 }, { "epoch": 0.024889968128699348, "grad_norm": 0.47877687215805054, "learning_rate": 9.963551685734535e-05, "loss": 1.7049, "step": 82 }, { "epoch": 0.025193504325390803, "grad_norm": 0.4063829183578491, "learning_rate": 9.963045459147515e-05, "loss": 1.8611, "step": 83 }, { "epoch": 0.025497040522082258, "grad_norm": 0.4149170219898224, "learning_rate": 9.962539232560496e-05, "loss": 1.9439, "step": 84 }, { "epoch": 0.025800576718773713, "grad_norm": 0.4882602393627167, "learning_rate": 9.962033005973475e-05, "loss": 1.5723, "step": 85 }, { "epoch": 0.026104112915465168, "grad_norm": 0.4600992202758789, "learning_rate": 9.961526779386454e-05, "loss": 2.0142, "step": 86 }, { "epoch": 0.026407649112156626, "grad_norm": 0.43366697430610657, "learning_rate": 9.961020552799434e-05, "loss": 1.9175, "step": 87 }, { "epoch": 0.02671118530884808, "grad_norm": 0.501487135887146, "learning_rate": 9.960514326212413e-05, "loss": 1.5043, "step": 88 }, { "epoch": 0.027014721505539536, "grad_norm": 0.43821993470191956, "learning_rate": 9.960008099625393e-05, "loss": 1.8622, "step": 89 }, { "epoch": 0.02731825770223099, "grad_norm": 0.4433805048465729, "learning_rate": 9.959501873038372e-05, "loss": 1.9459, "step": 90 }, { "epoch": 0.027621793898922446, "grad_norm": 0.4686216115951538, "learning_rate": 9.958995646451352e-05, "loss": 1.7405, "step": 91 }, { "epoch": 0.0279253300956139, "grad_norm": 0.48586198687553406, "learning_rate": 9.958489419864331e-05, "loss": 2.2233, "step": 92 }, { "epoch": 0.02822886629230536, "grad_norm": 0.4018734097480774, "learning_rate": 9.957983193277312e-05, "loss": 2.0027, "step": 93 }, { "epoch": 0.028532402488996814, "grad_norm": 0.4996435344219208, "learning_rate": 9.957476966690292e-05, "loss": 1.5949, "step": 94 }, { "epoch": 0.02883593868568827, "grad_norm": 0.45447826385498047, "learning_rate": 9.956970740103271e-05, "loss": 1.7636, "step": 95 }, { "epoch": 0.029139474882379723, "grad_norm": 0.4209904372692108, "learning_rate": 9.95646451351625e-05, "loss": 1.7523, "step": 96 }, { "epoch": 0.029443011079071178, "grad_norm": 0.3740164637565613, "learning_rate": 9.95595828692923e-05, "loss": 1.9136, "step": 97 }, { "epoch": 0.029746547275762633, "grad_norm": 0.4169963598251343, "learning_rate": 9.95545206034221e-05, "loss": 1.9136, "step": 98 }, { "epoch": 0.03005008347245409, "grad_norm": 0.4683006703853607, "learning_rate": 9.954945833755189e-05, "loss": 2.0657, "step": 99 }, { "epoch": 0.030353619669145546, "grad_norm": 0.4508633017539978, "learning_rate": 9.954439607168169e-05, "loss": 2.1099, "step": 100 }, { "epoch": 0.030657155865837, "grad_norm": 0.4136218726634979, "learning_rate": 9.953933380581148e-05, "loss": 2.0183, "step": 101 }, { "epoch": 0.030960692062528456, "grad_norm": 0.44510790705680847, "learning_rate": 9.953427153994127e-05, "loss": 1.9307, "step": 102 }, { "epoch": 0.031264228259219914, "grad_norm": 0.3713892698287964, "learning_rate": 9.952920927407108e-05, "loss": 1.7017, "step": 103 }, { "epoch": 0.03156776445591137, "grad_norm": 0.47902294993400574, "learning_rate": 9.952414700820088e-05, "loss": 2.1172, "step": 104 }, { "epoch": 0.031871300652602824, "grad_norm": 0.4492317736148834, "learning_rate": 9.951908474233067e-05, "loss": 1.9752, "step": 105 }, { "epoch": 0.03217483684929428, "grad_norm": 0.4096255302429199, "learning_rate": 9.951402247646047e-05, "loss": 1.5511, "step": 106 }, { "epoch": 0.032478373045985734, "grad_norm": 0.39630818367004395, "learning_rate": 9.950896021059026e-05, "loss": 2.11, "step": 107 }, { "epoch": 0.03278190924267719, "grad_norm": 0.42648032307624817, "learning_rate": 9.950389794472006e-05, "loss": 2.1784, "step": 108 }, { "epoch": 0.033085445439368644, "grad_norm": 0.4814178943634033, "learning_rate": 9.949883567884985e-05, "loss": 1.955, "step": 109 }, { "epoch": 0.0333889816360601, "grad_norm": 0.41600191593170166, "learning_rate": 9.949377341297965e-05, "loss": 1.9163, "step": 110 }, { "epoch": 0.03369251783275155, "grad_norm": 0.4610773026943207, "learning_rate": 9.948871114710944e-05, "loss": 1.7934, "step": 111 }, { "epoch": 0.03399605402944301, "grad_norm": 0.43061718344688416, "learning_rate": 9.948364888123925e-05, "loss": 1.9278, "step": 112 }, { "epoch": 0.03429959022613446, "grad_norm": 0.3907497227191925, "learning_rate": 9.947858661536904e-05, "loss": 1.996, "step": 113 }, { "epoch": 0.034603126422825925, "grad_norm": 0.3984166383743286, "learning_rate": 9.947352434949884e-05, "loss": 1.5936, "step": 114 }, { "epoch": 0.03490666261951738, "grad_norm": 0.43406423926353455, "learning_rate": 9.946846208362863e-05, "loss": 1.8866, "step": 115 }, { "epoch": 0.035210198816208835, "grad_norm": 0.45913639664649963, "learning_rate": 9.946339981775843e-05, "loss": 1.972, "step": 116 }, { "epoch": 0.03551373501290029, "grad_norm": 0.42077311873435974, "learning_rate": 9.945833755188822e-05, "loss": 2.0081, "step": 117 }, { "epoch": 0.035817271209591744, "grad_norm": 0.41479435563087463, "learning_rate": 9.945327528601802e-05, "loss": 2.0096, "step": 118 }, { "epoch": 0.0361208074062832, "grad_norm": 0.35669025778770447, "learning_rate": 9.944821302014781e-05, "loss": 2.0074, "step": 119 }, { "epoch": 0.036424343602974654, "grad_norm": 0.4088069796562195, "learning_rate": 9.944315075427761e-05, "loss": 1.817, "step": 120 }, { "epoch": 0.03672787979966611, "grad_norm": 0.49982163310050964, "learning_rate": 9.943808848840742e-05, "loss": 1.9218, "step": 121 }, { "epoch": 0.037031415996357564, "grad_norm": 0.39924055337905884, "learning_rate": 9.943302622253721e-05, "loss": 2.2463, "step": 122 }, { "epoch": 0.03733495219304902, "grad_norm": 0.40462177991867065, "learning_rate": 9.942796395666702e-05, "loss": 2.0844, "step": 123 }, { "epoch": 0.037638488389740474, "grad_norm": 0.43440741300582886, "learning_rate": 9.942290169079681e-05, "loss": 1.8808, "step": 124 }, { "epoch": 0.03794202458643193, "grad_norm": 0.4029730260372162, "learning_rate": 9.941783942492661e-05, "loss": 1.9427, "step": 125 }, { "epoch": 0.03824556078312339, "grad_norm": 0.7807103395462036, "learning_rate": 9.94127771590564e-05, "loss": 1.9072, "step": 126 }, { "epoch": 0.038549096979814845, "grad_norm": 0.5021561980247498, "learning_rate": 9.94077148931862e-05, "loss": 2.0582, "step": 127 }, { "epoch": 0.0388526331765063, "grad_norm": 0.5161197781562805, "learning_rate": 9.9402652627316e-05, "loss": 1.9861, "step": 128 }, { "epoch": 0.039156169373197755, "grad_norm": 0.5553935766220093, "learning_rate": 9.939759036144579e-05, "loss": 2.1893, "step": 129 }, { "epoch": 0.03945970556988921, "grad_norm": 0.4241655170917511, "learning_rate": 9.939252809557558e-05, "loss": 1.9722, "step": 130 }, { "epoch": 0.039763241766580665, "grad_norm": 0.43290001153945923, "learning_rate": 9.938746582970538e-05, "loss": 1.5364, "step": 131 }, { "epoch": 0.04006677796327212, "grad_norm": 0.40089091658592224, "learning_rate": 9.938240356383519e-05, "loss": 1.9686, "step": 132 }, { "epoch": 0.040370314159963575, "grad_norm": 0.4152032434940338, "learning_rate": 9.937734129796498e-05, "loss": 1.913, "step": 133 }, { "epoch": 0.04067385035665503, "grad_norm": 0.4443211555480957, "learning_rate": 9.937227903209478e-05, "loss": 2.2354, "step": 134 }, { "epoch": 0.040977386553346484, "grad_norm": 0.41355323791503906, "learning_rate": 9.936721676622457e-05, "loss": 2.1055, "step": 135 }, { "epoch": 0.04128092275003794, "grad_norm": 0.5837479829788208, "learning_rate": 9.936215450035437e-05, "loss": 1.9085, "step": 136 }, { "epoch": 0.041584458946729394, "grad_norm": 0.40269389748573303, "learning_rate": 9.935709223448416e-05, "loss": 2.0368, "step": 137 }, { "epoch": 0.041887995143420856, "grad_norm": 0.5898969769477844, "learning_rate": 9.935202996861396e-05, "loss": 1.7933, "step": 138 }, { "epoch": 0.04219153134011231, "grad_norm": 0.41117680072784424, "learning_rate": 9.934696770274375e-05, "loss": 1.7452, "step": 139 }, { "epoch": 0.042495067536803766, "grad_norm": 0.5090368390083313, "learning_rate": 9.934190543687354e-05, "loss": 2.0141, "step": 140 }, { "epoch": 0.04279860373349522, "grad_norm": 0.4821307957172394, "learning_rate": 9.933684317100334e-05, "loss": 1.9443, "step": 141 }, { "epoch": 0.043102139930186675, "grad_norm": 0.41939428448677063, "learning_rate": 9.933178090513315e-05, "loss": 1.7401, "step": 142 }, { "epoch": 0.04340567612687813, "grad_norm": 0.4531096816062927, "learning_rate": 9.932671863926294e-05, "loss": 1.9944, "step": 143 }, { "epoch": 0.043709212323569585, "grad_norm": 0.44440799951553345, "learning_rate": 9.932165637339274e-05, "loss": 1.9648, "step": 144 }, { "epoch": 0.04401274852026104, "grad_norm": 0.36847150325775146, "learning_rate": 9.931659410752253e-05, "loss": 2.0638, "step": 145 }, { "epoch": 0.044316284716952495, "grad_norm": 0.6394171118736267, "learning_rate": 9.931153184165233e-05, "loss": 1.9476, "step": 146 }, { "epoch": 0.04461982091364395, "grad_norm": 0.41597506403923035, "learning_rate": 9.930646957578212e-05, "loss": 1.535, "step": 147 }, { "epoch": 0.044923357110335405, "grad_norm": 0.5597077012062073, "learning_rate": 9.930140730991192e-05, "loss": 1.6826, "step": 148 }, { "epoch": 0.045226893307026866, "grad_norm": 0.5532084703445435, "learning_rate": 9.929634504404171e-05, "loss": 1.8063, "step": 149 }, { "epoch": 0.04553042950371832, "grad_norm": 0.467339426279068, "learning_rate": 9.92912827781715e-05, "loss": 2.017, "step": 150 }, { "epoch": 0.045833965700409776, "grad_norm": 0.4054040312767029, "learning_rate": 9.928622051230131e-05, "loss": 1.7582, "step": 151 }, { "epoch": 0.04613750189710123, "grad_norm": 1.2743823528289795, "learning_rate": 9.928115824643111e-05, "loss": 2.0202, "step": 152 }, { "epoch": 0.046441038093792686, "grad_norm": 0.4357397258281708, "learning_rate": 9.92760959805609e-05, "loss": 1.8788, "step": 153 }, { "epoch": 0.04674457429048414, "grad_norm": 2.8793208599090576, "learning_rate": 9.92710337146907e-05, "loss": 2.1204, "step": 154 }, { "epoch": 0.047048110487175596, "grad_norm": 0.9585952162742615, "learning_rate": 9.92659714488205e-05, "loss": 1.9356, "step": 155 }, { "epoch": 0.04735164668386705, "grad_norm": 0.7857603430747986, "learning_rate": 9.926090918295029e-05, "loss": 1.9097, "step": 156 }, { "epoch": 0.047655182880558505, "grad_norm": 0.5259221792221069, "learning_rate": 9.925584691708008e-05, "loss": 2.1589, "step": 157 }, { "epoch": 0.04795871907724996, "grad_norm": 2.793253183364868, "learning_rate": 9.925078465120988e-05, "loss": 1.7202, "step": 158 }, { "epoch": 0.048262255273941415, "grad_norm": 0.4432888627052307, "learning_rate": 9.924572238533967e-05, "loss": 1.9898, "step": 159 }, { "epoch": 0.04856579147063287, "grad_norm": 0.4347291588783264, "learning_rate": 9.924066011946948e-05, "loss": 1.8142, "step": 160 }, { "epoch": 0.04886932766732433, "grad_norm": 5.273514747619629, "learning_rate": 9.923559785359928e-05, "loss": 1.8665, "step": 161 }, { "epoch": 0.04917286386401579, "grad_norm": 0.47988301515579224, "learning_rate": 9.923053558772907e-05, "loss": 1.9439, "step": 162 }, { "epoch": 0.04947640006070724, "grad_norm": 0.3584117293357849, "learning_rate": 9.922547332185887e-05, "loss": 1.8109, "step": 163 }, { "epoch": 0.049779936257398696, "grad_norm": 0.4074074923992157, "learning_rate": 9.922041105598866e-05, "loss": 2.1056, "step": 164 }, { "epoch": 0.05008347245409015, "grad_norm": 3.159336566925049, "learning_rate": 9.921534879011846e-05, "loss": 1.8672, "step": 165 }, { "epoch": 0.050387008650781606, "grad_norm": 0.38132309913635254, "learning_rate": 9.921028652424826e-05, "loss": 1.8423, "step": 166 }, { "epoch": 0.05069054484747306, "grad_norm": 0.39241936802864075, "learning_rate": 9.920522425837806e-05, "loss": 1.5949, "step": 167 }, { "epoch": 0.050994081044164516, "grad_norm": 0.38212037086486816, "learning_rate": 9.920016199250785e-05, "loss": 1.9669, "step": 168 }, { "epoch": 0.05129761724085597, "grad_norm": 0.5353955030441284, "learning_rate": 9.919509972663765e-05, "loss": 2.1806, "step": 169 }, { "epoch": 0.051601153437547426, "grad_norm": 0.4129483699798584, "learning_rate": 9.919003746076744e-05, "loss": 1.8858, "step": 170 }, { "epoch": 0.05190468963423888, "grad_norm": 0.3832380771636963, "learning_rate": 9.918497519489725e-05, "loss": 2.0321, "step": 171 }, { "epoch": 0.052208225830930335, "grad_norm": 0.4078863859176636, "learning_rate": 9.917991292902705e-05, "loss": 1.6213, "step": 172 }, { "epoch": 0.0525117620276218, "grad_norm": 0.38865014910697937, "learning_rate": 9.917485066315684e-05, "loss": 2.0052, "step": 173 }, { "epoch": 0.05281529822431325, "grad_norm": 0.4339440166950226, "learning_rate": 9.916978839728664e-05, "loss": 2.2405, "step": 174 }, { "epoch": 0.05311883442100471, "grad_norm": 0.42063045501708984, "learning_rate": 9.916472613141643e-05, "loss": 1.6529, "step": 175 }, { "epoch": 0.05342237061769616, "grad_norm": 0.4765849709510803, "learning_rate": 9.915966386554623e-05, "loss": 1.9645, "step": 176 }, { "epoch": 0.05372590681438762, "grad_norm": 0.41431936621665955, "learning_rate": 9.915460159967602e-05, "loss": 1.9709, "step": 177 }, { "epoch": 0.05402944301107907, "grad_norm": 0.3591434359550476, "learning_rate": 9.914953933380581e-05, "loss": 1.685, "step": 178 }, { "epoch": 0.054332979207770526, "grad_norm": 0.45483240485191345, "learning_rate": 9.914447706793561e-05, "loss": 1.9362, "step": 179 }, { "epoch": 0.05463651540446198, "grad_norm": 0.5468000173568726, "learning_rate": 9.91394148020654e-05, "loss": 1.6984, "step": 180 }, { "epoch": 0.054940051601153436, "grad_norm": 0.4057190716266632, "learning_rate": 9.913435253619521e-05, "loss": 1.9887, "step": 181 }, { "epoch": 0.05524358779784489, "grad_norm": 0.383211612701416, "learning_rate": 9.912929027032501e-05, "loss": 1.7825, "step": 182 }, { "epoch": 0.055547123994536346, "grad_norm": 0.3480004668235779, "learning_rate": 9.91242280044548e-05, "loss": 1.8721, "step": 183 }, { "epoch": 0.0558506601912278, "grad_norm": 0.47680413722991943, "learning_rate": 9.91191657385846e-05, "loss": 1.8113, "step": 184 }, { "epoch": 0.05615419638791926, "grad_norm": 0.37727096676826477, "learning_rate": 9.911410347271439e-05, "loss": 1.7398, "step": 185 }, { "epoch": 0.05645773258461072, "grad_norm": 0.47738176584243774, "learning_rate": 9.910904120684419e-05, "loss": 1.4651, "step": 186 }, { "epoch": 0.05676126878130217, "grad_norm": 0.44533729553222656, "learning_rate": 9.910397894097398e-05, "loss": 1.5697, "step": 187 }, { "epoch": 0.05706480497799363, "grad_norm": 0.45051974058151245, "learning_rate": 9.909891667510378e-05, "loss": 2.1577, "step": 188 }, { "epoch": 0.05736834117468508, "grad_norm": 0.4709470272064209, "learning_rate": 9.909385440923357e-05, "loss": 2.0486, "step": 189 }, { "epoch": 0.05767187737137654, "grad_norm": 0.4063846170902252, "learning_rate": 9.908879214336338e-05, "loss": 1.5453, "step": 190 }, { "epoch": 0.05797541356806799, "grad_norm": 0.374362587928772, "learning_rate": 9.908372987749317e-05, "loss": 1.5611, "step": 191 }, { "epoch": 0.05827894976475945, "grad_norm": 0.4852111041545868, "learning_rate": 9.907866761162297e-05, "loss": 1.6234, "step": 192 }, { "epoch": 0.0585824859614509, "grad_norm": 0.6863122582435608, "learning_rate": 9.907360534575276e-05, "loss": 2.1612, "step": 193 }, { "epoch": 0.058886022158142357, "grad_norm": 0.6040588021278381, "learning_rate": 9.906854307988256e-05, "loss": 2.1092, "step": 194 }, { "epoch": 0.05918955835483381, "grad_norm": 0.4148467779159546, "learning_rate": 9.906348081401235e-05, "loss": 2.1108, "step": 195 }, { "epoch": 0.059493094551525266, "grad_norm": 0.36098209023475647, "learning_rate": 9.905841854814215e-05, "loss": 2.0002, "step": 196 }, { "epoch": 0.05979663074821673, "grad_norm": 0.42360183596611023, "learning_rate": 9.905335628227194e-05, "loss": 2.3124, "step": 197 }, { "epoch": 0.06010016694490818, "grad_norm": 0.3650914430618286, "learning_rate": 9.904829401640174e-05, "loss": 1.8778, "step": 198 }, { "epoch": 0.06040370314159964, "grad_norm": 0.392995148897171, "learning_rate": 9.904323175053155e-05, "loss": 2.16, "step": 199 }, { "epoch": 0.06070723933829109, "grad_norm": 0.46390387415885925, "learning_rate": 9.903816948466134e-05, "loss": 1.8695, "step": 200 }, { "epoch": 0.06101077553498255, "grad_norm": 0.3954870402812958, "learning_rate": 9.903310721879114e-05, "loss": 1.9233, "step": 201 }, { "epoch": 0.061314311731674, "grad_norm": 0.3650193214416504, "learning_rate": 9.902804495292093e-05, "loss": 2.2504, "step": 202 }, { "epoch": 0.06161784792836546, "grad_norm": 0.3582104742527008, "learning_rate": 9.902298268705073e-05, "loss": 1.9303, "step": 203 }, { "epoch": 0.06192138412505691, "grad_norm": 0.35688868165016174, "learning_rate": 9.901792042118052e-05, "loss": 1.7078, "step": 204 }, { "epoch": 0.06222492032174837, "grad_norm": 0.3666802942752838, "learning_rate": 9.901285815531031e-05, "loss": 1.941, "step": 205 }, { "epoch": 0.06252845651843983, "grad_norm": 0.42375093698501587, "learning_rate": 9.900779588944011e-05, "loss": 2.0858, "step": 206 }, { "epoch": 0.06283199271513128, "grad_norm": 0.3913770318031311, "learning_rate": 9.90027336235699e-05, "loss": 2.1423, "step": 207 }, { "epoch": 0.06313552891182274, "grad_norm": 0.4101809859275818, "learning_rate": 9.89976713576997e-05, "loss": 2.0497, "step": 208 }, { "epoch": 0.06343906510851419, "grad_norm": 0.3696439564228058, "learning_rate": 9.899260909182951e-05, "loss": 1.9692, "step": 209 }, { "epoch": 0.06374260130520565, "grad_norm": 0.3725574016571045, "learning_rate": 9.89875468259593e-05, "loss": 2.2053, "step": 210 }, { "epoch": 0.0640461375018971, "grad_norm": 0.4886903166770935, "learning_rate": 9.898248456008911e-05, "loss": 1.8981, "step": 211 }, { "epoch": 0.06434967369858856, "grad_norm": 0.4423249661922455, "learning_rate": 9.89774222942189e-05, "loss": 1.9058, "step": 212 }, { "epoch": 0.06465320989528, "grad_norm": 0.4045765697956085, "learning_rate": 9.89723600283487e-05, "loss": 1.8056, "step": 213 }, { "epoch": 0.06495674609197147, "grad_norm": 0.43866047263145447, "learning_rate": 9.89672977624785e-05, "loss": 1.6315, "step": 214 }, { "epoch": 0.06526028228866293, "grad_norm": 0.524714469909668, "learning_rate": 9.896223549660829e-05, "loss": 2.0156, "step": 215 }, { "epoch": 0.06556381848535438, "grad_norm": 0.3752996325492859, "learning_rate": 9.895717323073808e-05, "loss": 2.2768, "step": 216 }, { "epoch": 0.06586735468204584, "grad_norm": 0.4371670186519623, "learning_rate": 9.895211096486788e-05, "loss": 2.0755, "step": 217 }, { "epoch": 0.06617089087873729, "grad_norm": 0.3751063644886017, "learning_rate": 9.894704869899767e-05, "loss": 2.2451, "step": 218 }, { "epoch": 0.06647442707542875, "grad_norm": 0.6649600267410278, "learning_rate": 9.894198643312747e-05, "loss": 1.9835, "step": 219 }, { "epoch": 0.0667779632721202, "grad_norm": 0.3941735625267029, "learning_rate": 9.893692416725728e-05, "loss": 2.0203, "step": 220 }, { "epoch": 0.06708149946881166, "grad_norm": 0.41888293623924255, "learning_rate": 9.893186190138707e-05, "loss": 1.7572, "step": 221 }, { "epoch": 0.0673850356655031, "grad_norm": 0.4820149838924408, "learning_rate": 9.892679963551687e-05, "loss": 2.0591, "step": 222 }, { "epoch": 0.06768857186219457, "grad_norm": 0.3516736626625061, "learning_rate": 9.892173736964666e-05, "loss": 1.9398, "step": 223 }, { "epoch": 0.06799210805888602, "grad_norm": 0.3873218894004822, "learning_rate": 9.891667510377646e-05, "loss": 1.6389, "step": 224 }, { "epoch": 0.06829564425557748, "grad_norm": 0.3793487846851349, "learning_rate": 9.891161283790625e-05, "loss": 2.0075, "step": 225 }, { "epoch": 0.06859918045226893, "grad_norm": 0.38987675309181213, "learning_rate": 9.890655057203605e-05, "loss": 2.0903, "step": 226 }, { "epoch": 0.06890271664896039, "grad_norm": 0.4293549358844757, "learning_rate": 9.890148830616584e-05, "loss": 2.2099, "step": 227 }, { "epoch": 0.06920625284565185, "grad_norm": 0.39895692467689514, "learning_rate": 9.889642604029564e-05, "loss": 1.8615, "step": 228 }, { "epoch": 0.0695097890423433, "grad_norm": 0.4543936252593994, "learning_rate": 9.889136377442544e-05, "loss": 2.0828, "step": 229 }, { "epoch": 0.06981332523903476, "grad_norm": 0.448477566242218, "learning_rate": 9.888630150855524e-05, "loss": 1.5524, "step": 230 }, { "epoch": 0.07011686143572621, "grad_norm": 0.428975373506546, "learning_rate": 9.888123924268503e-05, "loss": 1.3828, "step": 231 }, { "epoch": 0.07042039763241767, "grad_norm": 0.42287349700927734, "learning_rate": 9.887617697681483e-05, "loss": 2.096, "step": 232 }, { "epoch": 0.07072393382910912, "grad_norm": 0.43614649772644043, "learning_rate": 9.887111471094462e-05, "loss": 1.8238, "step": 233 }, { "epoch": 0.07102747002580058, "grad_norm": 0.47309553623199463, "learning_rate": 9.886605244507442e-05, "loss": 2.3526, "step": 234 }, { "epoch": 0.07133100622249203, "grad_norm": 0.9558483362197876, "learning_rate": 9.886099017920421e-05, "loss": 1.9816, "step": 235 }, { "epoch": 0.07163454241918349, "grad_norm": 0.3529858887195587, "learning_rate": 9.885592791333401e-05, "loss": 2.0314, "step": 236 }, { "epoch": 0.07193807861587494, "grad_norm": 0.37652599811553955, "learning_rate": 9.88508656474638e-05, "loss": 1.9381, "step": 237 }, { "epoch": 0.0722416148125664, "grad_norm": 0.40783143043518066, "learning_rate": 9.884580338159361e-05, "loss": 1.966, "step": 238 }, { "epoch": 0.07254515100925786, "grad_norm": 0.4160328805446625, "learning_rate": 9.88407411157234e-05, "loss": 1.8176, "step": 239 }, { "epoch": 0.07284868720594931, "grad_norm": 0.4397304952144623, "learning_rate": 9.88356788498532e-05, "loss": 1.6766, "step": 240 }, { "epoch": 0.07315222340264077, "grad_norm": 0.42549702525138855, "learning_rate": 9.8830616583983e-05, "loss": 2.1176, "step": 241 }, { "epoch": 0.07345575959933222, "grad_norm": 0.3747939169406891, "learning_rate": 9.882555431811279e-05, "loss": 1.5494, "step": 242 }, { "epoch": 0.07375929579602368, "grad_norm": 3.4551990032196045, "learning_rate": 9.882049205224258e-05, "loss": 2.0336, "step": 243 }, { "epoch": 0.07406283199271513, "grad_norm": 1.5632964372634888, "learning_rate": 9.881542978637238e-05, "loss": 1.7452, "step": 244 }, { "epoch": 0.07436636818940659, "grad_norm": 0.41575855016708374, "learning_rate": 9.881036752050217e-05, "loss": 2.0243, "step": 245 }, { "epoch": 0.07466990438609804, "grad_norm": 0.44168713688850403, "learning_rate": 9.880530525463197e-05, "loss": 2.0022, "step": 246 }, { "epoch": 0.0749734405827895, "grad_norm": 0.46640321612358093, "learning_rate": 9.880024298876176e-05, "loss": 1.555, "step": 247 }, { "epoch": 0.07527697677948095, "grad_norm": 0.3622835576534271, "learning_rate": 9.879518072289157e-05, "loss": 1.876, "step": 248 }, { "epoch": 0.07558051297617241, "grad_norm": 0.6277987957000732, "learning_rate": 9.879011845702137e-05, "loss": 2.2753, "step": 249 }, { "epoch": 0.07588404917286386, "grad_norm": 0.40246644616127014, "learning_rate": 9.878505619115116e-05, "loss": 1.5991, "step": 250 }, { "epoch": 0.07618758536955532, "grad_norm": 0.38388529419898987, "learning_rate": 9.877999392528096e-05, "loss": 1.9226, "step": 251 }, { "epoch": 0.07649112156624678, "grad_norm": 0.39985090494155884, "learning_rate": 9.877493165941075e-05, "loss": 2.0722, "step": 252 }, { "epoch": 0.07679465776293823, "grad_norm": 0.3872128427028656, "learning_rate": 9.876986939354055e-05, "loss": 1.9132, "step": 253 }, { "epoch": 0.07709819395962969, "grad_norm": 0.3665171265602112, "learning_rate": 9.876480712767034e-05, "loss": 1.6244, "step": 254 }, { "epoch": 0.07740173015632114, "grad_norm": 0.4011310040950775, "learning_rate": 9.875974486180015e-05, "loss": 2.1289, "step": 255 }, { "epoch": 0.0777052663530126, "grad_norm": 0.35013166069984436, "learning_rate": 9.875468259592994e-05, "loss": 1.9738, "step": 256 }, { "epoch": 0.07800880254970405, "grad_norm": 0.48468607664108276, "learning_rate": 9.874962033005974e-05, "loss": 2.1368, "step": 257 }, { "epoch": 0.07831233874639551, "grad_norm": 0.5015551447868347, "learning_rate": 9.874455806418953e-05, "loss": 2.1218, "step": 258 }, { "epoch": 0.07861587494308696, "grad_norm": 0.41915133595466614, "learning_rate": 9.873949579831934e-05, "loss": 2.0052, "step": 259 }, { "epoch": 0.07891941113977842, "grad_norm": 0.4414760172367096, "learning_rate": 9.873443353244914e-05, "loss": 1.7249, "step": 260 }, { "epoch": 0.07922294733646987, "grad_norm": 0.47259169816970825, "learning_rate": 9.872937126657893e-05, "loss": 2.1041, "step": 261 }, { "epoch": 0.07952648353316133, "grad_norm": 0.3689124882221222, "learning_rate": 9.872430900070873e-05, "loss": 1.8956, "step": 262 }, { "epoch": 0.07983001972985279, "grad_norm": 0.3948320150375366, "learning_rate": 9.871924673483852e-05, "loss": 1.9211, "step": 263 }, { "epoch": 0.08013355592654424, "grad_norm": 0.4235248267650604, "learning_rate": 9.871418446896832e-05, "loss": 1.7115, "step": 264 }, { "epoch": 0.0804370921232357, "grad_norm": 0.48399198055267334, "learning_rate": 9.870912220309811e-05, "loss": 1.77, "step": 265 }, { "epoch": 0.08074062831992715, "grad_norm": 0.34047526121139526, "learning_rate": 9.87040599372279e-05, "loss": 1.7189, "step": 266 }, { "epoch": 0.08104416451661861, "grad_norm": 0.47203269600868225, "learning_rate": 9.86989976713577e-05, "loss": 1.7674, "step": 267 }, { "epoch": 0.08134770071331006, "grad_norm": 0.3752756118774414, "learning_rate": 9.869393540548751e-05, "loss": 1.8716, "step": 268 }, { "epoch": 0.08165123691000152, "grad_norm": 0.3437153697013855, "learning_rate": 9.86888731396173e-05, "loss": 1.9824, "step": 269 }, { "epoch": 0.08195477310669297, "grad_norm": 0.4854094088077545, "learning_rate": 9.86838108737471e-05, "loss": 1.4385, "step": 270 }, { "epoch": 0.08225830930338443, "grad_norm": 0.37674829363822937, "learning_rate": 9.86787486078769e-05, "loss": 1.7877, "step": 271 }, { "epoch": 0.08256184550007588, "grad_norm": 0.4215140640735626, "learning_rate": 9.867368634200669e-05, "loss": 2.1854, "step": 272 }, { "epoch": 0.08286538169676734, "grad_norm": 0.3680359423160553, "learning_rate": 9.866862407613648e-05, "loss": 2.104, "step": 273 }, { "epoch": 0.08316891789345879, "grad_norm": 0.4195649325847626, "learning_rate": 9.866356181026628e-05, "loss": 1.469, "step": 274 }, { "epoch": 0.08347245409015025, "grad_norm": 0.480640709400177, "learning_rate": 9.865849954439607e-05, "loss": 1.8329, "step": 275 }, { "epoch": 0.08377599028684171, "grad_norm": 0.34760695695877075, "learning_rate": 9.865343727852587e-05, "loss": 1.9495, "step": 276 }, { "epoch": 0.08407952648353316, "grad_norm": 0.3803161680698395, "learning_rate": 9.864837501265568e-05, "loss": 1.9294, "step": 277 }, { "epoch": 0.08438306268022462, "grad_norm": 0.41739675402641296, "learning_rate": 9.864331274678547e-05, "loss": 2.059, "step": 278 }, { "epoch": 0.08468659887691607, "grad_norm": 0.3807448744773865, "learning_rate": 9.863825048091527e-05, "loss": 1.9741, "step": 279 }, { "epoch": 0.08499013507360753, "grad_norm": 0.3610997200012207, "learning_rate": 9.863318821504506e-05, "loss": 1.9815, "step": 280 }, { "epoch": 0.08529367127029898, "grad_norm": 0.3797460198402405, "learning_rate": 9.862812594917485e-05, "loss": 2.1394, "step": 281 }, { "epoch": 0.08559720746699044, "grad_norm": 0.3922887444496155, "learning_rate": 9.862306368330465e-05, "loss": 2.184, "step": 282 }, { "epoch": 0.08590074366368189, "grad_norm": 0.38251930475234985, "learning_rate": 9.861800141743444e-05, "loss": 2.0186, "step": 283 }, { "epoch": 0.08620427986037335, "grad_norm": 0.35968562960624695, "learning_rate": 9.861293915156424e-05, "loss": 2.0, "step": 284 }, { "epoch": 0.0865078160570648, "grad_norm": 0.37149590253829956, "learning_rate": 9.860787688569403e-05, "loss": 1.7941, "step": 285 }, { "epoch": 0.08681135225375626, "grad_norm": 0.36890628933906555, "learning_rate": 9.860281461982383e-05, "loss": 1.906, "step": 286 }, { "epoch": 0.08711488845044772, "grad_norm": 0.36025917530059814, "learning_rate": 9.859775235395364e-05, "loss": 1.9655, "step": 287 }, { "epoch": 0.08741842464713917, "grad_norm": 0.3704364001750946, "learning_rate": 9.859269008808343e-05, "loss": 1.8657, "step": 288 }, { "epoch": 0.08772196084383063, "grad_norm": 0.5996513962745667, "learning_rate": 9.858762782221323e-05, "loss": 1.7448, "step": 289 }, { "epoch": 0.08802549704052208, "grad_norm": 0.3615630269050598, "learning_rate": 9.858256555634302e-05, "loss": 1.9007, "step": 290 }, { "epoch": 0.08832903323721354, "grad_norm": 0.36014246940612793, "learning_rate": 9.857750329047282e-05, "loss": 1.927, "step": 291 }, { "epoch": 0.08863256943390499, "grad_norm": 0.5038754940032959, "learning_rate": 9.857244102460261e-05, "loss": 1.6613, "step": 292 }, { "epoch": 0.08893610563059645, "grad_norm": 0.3880213797092438, "learning_rate": 9.85673787587324e-05, "loss": 1.5563, "step": 293 }, { "epoch": 0.0892396418272879, "grad_norm": 0.43225082755088806, "learning_rate": 9.85623164928622e-05, "loss": 1.5534, "step": 294 }, { "epoch": 0.08954317802397936, "grad_norm": 0.44342055916786194, "learning_rate": 9.8557254226992e-05, "loss": 1.6211, "step": 295 }, { "epoch": 0.08984671422067081, "grad_norm": 0.42114123702049255, "learning_rate": 9.85521919611218e-05, "loss": 1.9731, "step": 296 }, { "epoch": 0.09015025041736227, "grad_norm": 0.43151113390922546, "learning_rate": 9.85471296952516e-05, "loss": 1.9519, "step": 297 }, { "epoch": 0.09045378661405373, "grad_norm": 0.38092517852783203, "learning_rate": 9.85420674293814e-05, "loss": 2.0973, "step": 298 }, { "epoch": 0.09075732281074518, "grad_norm": 0.40729570388793945, "learning_rate": 9.853700516351119e-05, "loss": 1.4395, "step": 299 }, { "epoch": 0.09106085900743664, "grad_norm": 0.3631846308708191, "learning_rate": 9.8531942897641e-05, "loss": 1.2255, "step": 300 }, { "epoch": 0.09136439520412809, "grad_norm": 0.37764397263526917, "learning_rate": 9.852688063177079e-05, "loss": 1.9941, "step": 301 }, { "epoch": 0.09166793140081955, "grad_norm": 0.3755379319190979, "learning_rate": 9.852181836590059e-05, "loss": 1.7154, "step": 302 }, { "epoch": 0.091971467597511, "grad_norm": 0.39003854990005493, "learning_rate": 9.851675610003038e-05, "loss": 1.928, "step": 303 }, { "epoch": 0.09227500379420246, "grad_norm": 0.39592432975769043, "learning_rate": 9.851169383416018e-05, "loss": 2.1913, "step": 304 }, { "epoch": 0.09257853999089391, "grad_norm": 0.4315894842147827, "learning_rate": 9.850663156828997e-05, "loss": 1.6432, "step": 305 }, { "epoch": 0.09288207618758537, "grad_norm": 0.4103511571884155, "learning_rate": 9.850156930241977e-05, "loss": 1.9944, "step": 306 }, { "epoch": 0.09318561238427682, "grad_norm": 0.4236547350883484, "learning_rate": 9.849650703654957e-05, "loss": 1.875, "step": 307 }, { "epoch": 0.09348914858096828, "grad_norm": 0.41012468934059143, "learning_rate": 9.849144477067937e-05, "loss": 2.008, "step": 308 }, { "epoch": 0.09379268477765973, "grad_norm": 0.35538622736930847, "learning_rate": 9.848638250480916e-05, "loss": 1.7322, "step": 309 }, { "epoch": 0.09409622097435119, "grad_norm": 0.3874755799770355, "learning_rate": 9.848132023893896e-05, "loss": 1.9818, "step": 310 }, { "epoch": 0.09439975717104265, "grad_norm": 0.42444977164268494, "learning_rate": 9.847625797306875e-05, "loss": 2.1606, "step": 311 }, { "epoch": 0.0947032933677341, "grad_norm": 0.5855305194854736, "learning_rate": 9.847119570719855e-05, "loss": 1.4887, "step": 312 }, { "epoch": 0.09500682956442556, "grad_norm": 0.35223227739334106, "learning_rate": 9.846613344132834e-05, "loss": 2.0025, "step": 313 }, { "epoch": 0.09531036576111701, "grad_norm": 0.4013148844242096, "learning_rate": 9.846107117545814e-05, "loss": 1.9702, "step": 314 }, { "epoch": 0.09561390195780847, "grad_norm": 0.5038349032402039, "learning_rate": 9.845600890958793e-05, "loss": 2.1532, "step": 315 }, { "epoch": 0.09591743815449992, "grad_norm": 0.4826093018054962, "learning_rate": 9.845094664371774e-05, "loss": 2.0118, "step": 316 }, { "epoch": 0.09622097435119138, "grad_norm": 0.41135913133621216, "learning_rate": 9.844588437784754e-05, "loss": 2.0707, "step": 317 }, { "epoch": 0.09652451054788283, "grad_norm": 0.4353053569793701, "learning_rate": 9.844082211197733e-05, "loss": 2.104, "step": 318 }, { "epoch": 0.09682804674457429, "grad_norm": 0.4192908704280853, "learning_rate": 9.843575984610712e-05, "loss": 1.9489, "step": 319 }, { "epoch": 0.09713158294126574, "grad_norm": 0.380562424659729, "learning_rate": 9.843069758023692e-05, "loss": 1.3602, "step": 320 }, { "epoch": 0.0974351191379572, "grad_norm": 0.3394995331764221, "learning_rate": 9.842563531436671e-05, "loss": 2.2161, "step": 321 }, { "epoch": 0.09773865533464866, "grad_norm": 0.3419237434864044, "learning_rate": 9.842057304849651e-05, "loss": 1.7146, "step": 322 }, { "epoch": 0.09804219153134011, "grad_norm": 0.3590264618396759, "learning_rate": 9.84155107826263e-05, "loss": 1.8654, "step": 323 }, { "epoch": 0.09834572772803157, "grad_norm": 0.40006300806999207, "learning_rate": 9.84104485167561e-05, "loss": 1.5787, "step": 324 }, { "epoch": 0.09864926392472302, "grad_norm": 0.33313074707984924, "learning_rate": 9.84053862508859e-05, "loss": 1.8653, "step": 325 }, { "epoch": 0.09895280012141448, "grad_norm": 0.39681655168533325, "learning_rate": 9.84003239850157e-05, "loss": 2.178, "step": 326 }, { "epoch": 0.09925633631810593, "grad_norm": 0.41945868730545044, "learning_rate": 9.83952617191455e-05, "loss": 1.8324, "step": 327 }, { "epoch": 0.09955987251479739, "grad_norm": 0.3957304060459137, "learning_rate": 9.839019945327529e-05, "loss": 1.6468, "step": 328 }, { "epoch": 0.09986340871148884, "grad_norm": 0.35814937949180603, "learning_rate": 9.838513718740509e-05, "loss": 1.6492, "step": 329 }, { "epoch": 0.1001669449081803, "grad_norm": 0.38410916924476624, "learning_rate": 9.838007492153488e-05, "loss": 1.7223, "step": 330 }, { "epoch": 0.10047048110487175, "grad_norm": 0.38490885496139526, "learning_rate": 9.837501265566468e-05, "loss": 2.0166, "step": 331 }, { "epoch": 0.10077401730156321, "grad_norm": 0.38943415880203247, "learning_rate": 9.836995038979447e-05, "loss": 1.371, "step": 332 }, { "epoch": 0.10107755349825466, "grad_norm": 0.39741018414497375, "learning_rate": 9.836488812392427e-05, "loss": 1.6233, "step": 333 }, { "epoch": 0.10138108969494612, "grad_norm": 0.4663957357406616, "learning_rate": 9.835982585805406e-05, "loss": 1.746, "step": 334 }, { "epoch": 0.10168462589163758, "grad_norm": 0.37118905782699585, "learning_rate": 9.835476359218387e-05, "loss": 1.9684, "step": 335 }, { "epoch": 0.10198816208832903, "grad_norm": 0.40275588631629944, "learning_rate": 9.834970132631366e-05, "loss": 1.9551, "step": 336 }, { "epoch": 0.1022916982850205, "grad_norm": 0.4336283206939697, "learning_rate": 9.834463906044346e-05, "loss": 2.0711, "step": 337 }, { "epoch": 0.10259523448171194, "grad_norm": 0.35735735297203064, "learning_rate": 9.833957679457325e-05, "loss": 2.1397, "step": 338 }, { "epoch": 0.1028987706784034, "grad_norm": 0.37825390696525574, "learning_rate": 9.833451452870305e-05, "loss": 1.7494, "step": 339 }, { "epoch": 0.10320230687509485, "grad_norm": 0.3384961783885956, "learning_rate": 9.832945226283284e-05, "loss": 2.0197, "step": 340 }, { "epoch": 0.10350584307178631, "grad_norm": 0.46276888251304626, "learning_rate": 9.832438999696264e-05, "loss": 1.797, "step": 341 }, { "epoch": 0.10380937926847776, "grad_norm": 0.3685421347618103, "learning_rate": 9.831932773109243e-05, "loss": 1.9301, "step": 342 }, { "epoch": 0.10411291546516922, "grad_norm": 0.38931936025619507, "learning_rate": 9.831426546522223e-05, "loss": 1.9623, "step": 343 }, { "epoch": 0.10441645166186067, "grad_norm": 0.46678805351257324, "learning_rate": 9.830920319935204e-05, "loss": 1.6708, "step": 344 }, { "epoch": 0.10471998785855213, "grad_norm": 0.4199204444885254, "learning_rate": 9.830414093348183e-05, "loss": 1.8014, "step": 345 }, { "epoch": 0.1050235240552436, "grad_norm": 0.41024506092071533, "learning_rate": 9.829907866761164e-05, "loss": 1.8829, "step": 346 }, { "epoch": 0.10532706025193504, "grad_norm": 0.5271286368370056, "learning_rate": 9.829401640174143e-05, "loss": 1.7796, "step": 347 }, { "epoch": 0.1056305964486265, "grad_norm": 0.3593878448009491, "learning_rate": 9.828895413587123e-05, "loss": 2.0697, "step": 348 }, { "epoch": 0.10593413264531795, "grad_norm": 0.44404372572898865, "learning_rate": 9.828389187000102e-05, "loss": 2.3235, "step": 349 }, { "epoch": 0.10623766884200941, "grad_norm": 0.4072231650352478, "learning_rate": 9.827882960413082e-05, "loss": 1.5391, "step": 350 }, { "epoch": 0.10654120503870086, "grad_norm": 0.3924303352832794, "learning_rate": 9.827376733826061e-05, "loss": 2.0649, "step": 351 }, { "epoch": 0.10684474123539232, "grad_norm": 0.3815264105796814, "learning_rate": 9.826870507239041e-05, "loss": 1.5821, "step": 352 }, { "epoch": 0.10714827743208377, "grad_norm": 0.40832409262657166, "learning_rate": 9.82636428065202e-05, "loss": 2.1135, "step": 353 }, { "epoch": 0.10745181362877523, "grad_norm": 0.40270155668258667, "learning_rate": 9.825858054065e-05, "loss": 1.6561, "step": 354 }, { "epoch": 0.10775534982546668, "grad_norm": 0.38295283913612366, "learning_rate": 9.82535182747798e-05, "loss": 1.8938, "step": 355 }, { "epoch": 0.10805888602215814, "grad_norm": 0.41975417733192444, "learning_rate": 9.82484560089096e-05, "loss": 1.8605, "step": 356 }, { "epoch": 0.10836242221884959, "grad_norm": 0.41388946771621704, "learning_rate": 9.82433937430394e-05, "loss": 1.812, "step": 357 }, { "epoch": 0.10866595841554105, "grad_norm": 0.3470607101917267, "learning_rate": 9.823833147716919e-05, "loss": 2.1914, "step": 358 }, { "epoch": 0.10896949461223251, "grad_norm": 0.4417155385017395, "learning_rate": 9.823326921129898e-05, "loss": 1.7644, "step": 359 }, { "epoch": 0.10927303080892396, "grad_norm": 0.33910539746284485, "learning_rate": 9.822820694542878e-05, "loss": 1.8821, "step": 360 }, { "epoch": 0.10957656700561542, "grad_norm": 0.36742356419563293, "learning_rate": 9.822314467955857e-05, "loss": 1.9684, "step": 361 }, { "epoch": 0.10988010320230687, "grad_norm": 0.407844603061676, "learning_rate": 9.821808241368837e-05, "loss": 1.8797, "step": 362 }, { "epoch": 0.11018363939899833, "grad_norm": 0.4090898036956787, "learning_rate": 9.821302014781816e-05, "loss": 1.8401, "step": 363 }, { "epoch": 0.11048717559568978, "grad_norm": 0.3852720260620117, "learning_rate": 9.820795788194796e-05, "loss": 1.6887, "step": 364 }, { "epoch": 0.11079071179238124, "grad_norm": 0.4147186875343323, "learning_rate": 9.820289561607777e-05, "loss": 1.7263, "step": 365 }, { "epoch": 0.11109424798907269, "grad_norm": 0.7032086849212646, "learning_rate": 9.819783335020756e-05, "loss": 1.5382, "step": 366 }, { "epoch": 0.11139778418576415, "grad_norm": 0.3547534644603729, "learning_rate": 9.819277108433736e-05, "loss": 1.5988, "step": 367 }, { "epoch": 0.1117013203824556, "grad_norm": 0.45878785848617554, "learning_rate": 9.818770881846715e-05, "loss": 2.2467, "step": 368 }, { "epoch": 0.11200485657914706, "grad_norm": 0.39183077216148376, "learning_rate": 9.818264655259695e-05, "loss": 1.848, "step": 369 }, { "epoch": 0.11230839277583853, "grad_norm": 0.3735283315181732, "learning_rate": 9.817758428672674e-05, "loss": 1.6925, "step": 370 }, { "epoch": 0.11261192897252997, "grad_norm": 0.3878265917301178, "learning_rate": 9.817252202085654e-05, "loss": 2.04, "step": 371 }, { "epoch": 0.11291546516922144, "grad_norm": 0.38978812098503113, "learning_rate": 9.816745975498633e-05, "loss": 1.869, "step": 372 }, { "epoch": 0.11321900136591288, "grad_norm": 0.39212337136268616, "learning_rate": 9.816239748911613e-05, "loss": 2.0549, "step": 373 }, { "epoch": 0.11352253756260434, "grad_norm": 0.39528506994247437, "learning_rate": 9.815733522324593e-05, "loss": 1.5653, "step": 374 }, { "epoch": 0.11382607375929579, "grad_norm": 0.4226018786430359, "learning_rate": 9.815227295737573e-05, "loss": 1.6231, "step": 375 }, { "epoch": 0.11412960995598725, "grad_norm": 0.3577810823917389, "learning_rate": 9.814721069150552e-05, "loss": 1.9599, "step": 376 }, { "epoch": 0.1144331461526787, "grad_norm": 0.33580708503723145, "learning_rate": 9.814214842563532e-05, "loss": 2.0419, "step": 377 }, { "epoch": 0.11473668234937016, "grad_norm": 0.38860392570495605, "learning_rate": 9.813708615976511e-05, "loss": 1.7186, "step": 378 }, { "epoch": 0.11504021854606161, "grad_norm": 0.38994479179382324, "learning_rate": 9.813202389389491e-05, "loss": 2.1848, "step": 379 }, { "epoch": 0.11534375474275307, "grad_norm": 0.3947262763977051, "learning_rate": 9.81269616280247e-05, "loss": 2.1868, "step": 380 }, { "epoch": 0.11564729093944452, "grad_norm": 0.3112877607345581, "learning_rate": 9.81218993621545e-05, "loss": 1.8604, "step": 381 }, { "epoch": 0.11595082713613598, "grad_norm": 0.375689834356308, "learning_rate": 9.811683709628429e-05, "loss": 2.0418, "step": 382 }, { "epoch": 0.11625436333282745, "grad_norm": 0.34537243843078613, "learning_rate": 9.81117748304141e-05, "loss": 1.8874, "step": 383 }, { "epoch": 0.1165578995295189, "grad_norm": 0.5077370405197144, "learning_rate": 9.81067125645439e-05, "loss": 1.7497, "step": 384 }, { "epoch": 0.11686143572621036, "grad_norm": 0.3703441023826599, "learning_rate": 9.810165029867369e-05, "loss": 1.781, "step": 385 }, { "epoch": 0.1171649719229018, "grad_norm": 0.4386610984802246, "learning_rate": 9.809658803280348e-05, "loss": 1.8428, "step": 386 }, { "epoch": 0.11746850811959327, "grad_norm": 0.37781745195388794, "learning_rate": 9.809152576693328e-05, "loss": 2.0384, "step": 387 }, { "epoch": 0.11777204431628471, "grad_norm": 0.38956716656684875, "learning_rate": 9.808646350106307e-05, "loss": 2.3534, "step": 388 }, { "epoch": 0.11807558051297617, "grad_norm": 0.3444838523864746, "learning_rate": 9.808140123519288e-05, "loss": 1.921, "step": 389 }, { "epoch": 0.11837911670966762, "grad_norm": 0.39881742000579834, "learning_rate": 9.807633896932268e-05, "loss": 2.1758, "step": 390 }, { "epoch": 0.11868265290635908, "grad_norm": 0.384226530790329, "learning_rate": 9.807127670345247e-05, "loss": 1.7651, "step": 391 }, { "epoch": 0.11898618910305053, "grad_norm": 0.36255109310150146, "learning_rate": 9.806621443758227e-05, "loss": 1.8122, "step": 392 }, { "epoch": 0.119289725299742, "grad_norm": 0.3627421259880066, "learning_rate": 9.806115217171206e-05, "loss": 1.6304, "step": 393 }, { "epoch": 0.11959326149643346, "grad_norm": 0.8936781883239746, "learning_rate": 9.805608990584187e-05, "loss": 1.8827, "step": 394 }, { "epoch": 0.1198967976931249, "grad_norm": 0.5008642673492432, "learning_rate": 9.805102763997166e-05, "loss": 1.3597, "step": 395 }, { "epoch": 0.12020033388981637, "grad_norm": 0.4444289207458496, "learning_rate": 9.804596537410146e-05, "loss": 2.1768, "step": 396 }, { "epoch": 0.12050387008650781, "grad_norm": 0.3963356912136078, "learning_rate": 9.804090310823125e-05, "loss": 1.8373, "step": 397 }, { "epoch": 0.12080740628319928, "grad_norm": 0.44095271825790405, "learning_rate": 9.803584084236105e-05, "loss": 1.7893, "step": 398 }, { "epoch": 0.12111094247989072, "grad_norm": 0.4162418246269226, "learning_rate": 9.803077857649084e-05, "loss": 1.7482, "step": 399 }, { "epoch": 0.12141447867658219, "grad_norm": 0.3853035271167755, "learning_rate": 9.802571631062064e-05, "loss": 1.6274, "step": 400 }, { "epoch": 0.12171801487327363, "grad_norm": 1.1697463989257812, "learning_rate": 9.802065404475043e-05, "loss": 2.2254, "step": 401 }, { "epoch": 0.1220215510699651, "grad_norm": 0.3899803161621094, "learning_rate": 9.801559177888023e-05, "loss": 1.9754, "step": 402 }, { "epoch": 0.12232508726665654, "grad_norm": 0.43946412205696106, "learning_rate": 9.801052951301002e-05, "loss": 2.1184, "step": 403 }, { "epoch": 0.122628623463348, "grad_norm": 0.46882718801498413, "learning_rate": 9.800546724713983e-05, "loss": 1.4423, "step": 404 }, { "epoch": 0.12293215966003945, "grad_norm": 0.4379485547542572, "learning_rate": 9.800040498126963e-05, "loss": 2.0614, "step": 405 }, { "epoch": 0.12323569585673091, "grad_norm": 0.3837740123271942, "learning_rate": 9.799534271539942e-05, "loss": 1.9974, "step": 406 }, { "epoch": 0.12353923205342238, "grad_norm": 0.35403695702552795, "learning_rate": 9.799028044952922e-05, "loss": 1.5693, "step": 407 }, { "epoch": 0.12384276825011382, "grad_norm": 0.4070426821708679, "learning_rate": 9.798521818365901e-05, "loss": 1.8704, "step": 408 }, { "epoch": 0.12414630444680529, "grad_norm": 0.4301077425479889, "learning_rate": 9.79801559177888e-05, "loss": 1.077, "step": 409 }, { "epoch": 0.12444984064349673, "grad_norm": 0.37687429785728455, "learning_rate": 9.79750936519186e-05, "loss": 1.7323, "step": 410 }, { "epoch": 0.1247533768401882, "grad_norm": 0.37393873929977417, "learning_rate": 9.79700313860484e-05, "loss": 1.9532, "step": 411 }, { "epoch": 0.12505691303687966, "grad_norm": 0.4518846869468689, "learning_rate": 9.796496912017819e-05, "loss": 2.0123, "step": 412 }, { "epoch": 0.1253604492335711, "grad_norm": 0.39417609572410583, "learning_rate": 9.7959906854308e-05, "loss": 2.2669, "step": 413 }, { "epoch": 0.12566398543026255, "grad_norm": 0.3802976608276367, "learning_rate": 9.795484458843779e-05, "loss": 2.0506, "step": 414 }, { "epoch": 0.12596752162695402, "grad_norm": 1.3118431568145752, "learning_rate": 9.794978232256759e-05, "loss": 2.2551, "step": 415 }, { "epoch": 0.12627105782364548, "grad_norm": 0.9459638595581055, "learning_rate": 9.794472005669738e-05, "loss": 1.7829, "step": 416 }, { "epoch": 0.1265745940203369, "grad_norm": 0.571232795715332, "learning_rate": 9.793965779082718e-05, "loss": 1.7768, "step": 417 }, { "epoch": 0.12687813021702837, "grad_norm": 0.3973385989665985, "learning_rate": 9.793459552495697e-05, "loss": 1.88, "step": 418 }, { "epoch": 0.12718166641371983, "grad_norm": 0.3883122503757477, "learning_rate": 9.792953325908677e-05, "loss": 1.9592, "step": 419 }, { "epoch": 0.1274852026104113, "grad_norm": 0.40379586815834045, "learning_rate": 9.792447099321656e-05, "loss": 1.9697, "step": 420 }, { "epoch": 0.12778873880710276, "grad_norm": 0.3288556635379791, "learning_rate": 9.791940872734636e-05, "loss": 1.7282, "step": 421 }, { "epoch": 0.1280922750037942, "grad_norm": 0.3872746527194977, "learning_rate": 9.791434646147616e-05, "loss": 1.9348, "step": 422 }, { "epoch": 0.12839581120048565, "grad_norm": 0.37058207392692566, "learning_rate": 9.790928419560596e-05, "loss": 1.5684, "step": 423 }, { "epoch": 0.12869934739717712, "grad_norm": 0.37466561794281006, "learning_rate": 9.790422192973575e-05, "loss": 1.9535, "step": 424 }, { "epoch": 0.12900288359386858, "grad_norm": 0.32176846265792847, "learning_rate": 9.789915966386555e-05, "loss": 1.8537, "step": 425 }, { "epoch": 0.12930641979056, "grad_norm": 0.37653467059135437, "learning_rate": 9.789409739799534e-05, "loss": 2.0701, "step": 426 }, { "epoch": 0.12960995598725147, "grad_norm": 0.38768434524536133, "learning_rate": 9.788903513212514e-05, "loss": 1.731, "step": 427 }, { "epoch": 0.12991349218394294, "grad_norm": 0.5139635801315308, "learning_rate": 9.788397286625493e-05, "loss": 2.4437, "step": 428 }, { "epoch": 0.1302170283806344, "grad_norm": 0.3759630024433136, "learning_rate": 9.787891060038473e-05, "loss": 2.0918, "step": 429 }, { "epoch": 0.13052056457732586, "grad_norm": 0.3718818426132202, "learning_rate": 9.787384833451452e-05, "loss": 1.5854, "step": 430 }, { "epoch": 0.1308241007740173, "grad_norm": 0.6460405588150024, "learning_rate": 9.786878606864432e-05, "loss": 2.2442, "step": 431 }, { "epoch": 0.13112763697070876, "grad_norm": 0.40393388271331787, "learning_rate": 9.786372380277413e-05, "loss": 1.728, "step": 432 }, { "epoch": 0.13143117316740022, "grad_norm": 0.3772658407688141, "learning_rate": 9.785866153690393e-05, "loss": 1.668, "step": 433 }, { "epoch": 0.13173470936409168, "grad_norm": 2.5252649784088135, "learning_rate": 9.785359927103373e-05, "loss": 1.8864, "step": 434 }, { "epoch": 0.1320382455607831, "grad_norm": 0.42327219247817993, "learning_rate": 9.784853700516352e-05, "loss": 2.3174, "step": 435 }, { "epoch": 0.13234178175747457, "grad_norm": 0.3689473867416382, "learning_rate": 9.784347473929332e-05, "loss": 1.9671, "step": 436 }, { "epoch": 0.13264531795416604, "grad_norm": 0.37554243206977844, "learning_rate": 9.783841247342311e-05, "loss": 1.783, "step": 437 }, { "epoch": 0.1329488541508575, "grad_norm": 0.409587025642395, "learning_rate": 9.783335020755291e-05, "loss": 2.0385, "step": 438 }, { "epoch": 0.13325239034754893, "grad_norm": 0.349252849817276, "learning_rate": 9.78282879416827e-05, "loss": 1.8785, "step": 439 }, { "epoch": 0.1335559265442404, "grad_norm": 0.36687588691711426, "learning_rate": 9.78232256758125e-05, "loss": 2.1174, "step": 440 }, { "epoch": 0.13385946274093186, "grad_norm": 0.40221846103668213, "learning_rate": 9.781816340994229e-05, "loss": 1.8385, "step": 441 }, { "epoch": 0.13416299893762332, "grad_norm": 0.5634617805480957, "learning_rate": 9.781310114407209e-05, "loss": 1.9316, "step": 442 }, { "epoch": 0.13446653513431478, "grad_norm": 0.37704020738601685, "learning_rate": 9.78080388782019e-05, "loss": 1.8865, "step": 443 }, { "epoch": 0.1347700713310062, "grad_norm": 0.36043843626976013, "learning_rate": 9.780297661233169e-05, "loss": 1.585, "step": 444 }, { "epoch": 0.13507360752769768, "grad_norm": 0.33643844723701477, "learning_rate": 9.779791434646149e-05, "loss": 1.8098, "step": 445 }, { "epoch": 0.13537714372438914, "grad_norm": 0.6782101988792419, "learning_rate": 9.779285208059128e-05, "loss": 2.0468, "step": 446 }, { "epoch": 0.1356806799210806, "grad_norm": 0.38101980090141296, "learning_rate": 9.778778981472108e-05, "loss": 2.0624, "step": 447 }, { "epoch": 0.13598421611777203, "grad_norm": 0.399311900138855, "learning_rate": 9.778272754885087e-05, "loss": 2.1652, "step": 448 }, { "epoch": 0.1362877523144635, "grad_norm": 0.3491426706314087, "learning_rate": 9.777766528298066e-05, "loss": 1.9092, "step": 449 }, { "epoch": 0.13659128851115496, "grad_norm": 0.3654717803001404, "learning_rate": 9.777260301711046e-05, "loss": 1.9773, "step": 450 }, { "epoch": 0.13689482470784642, "grad_norm": 0.394699364900589, "learning_rate": 9.776754075124025e-05, "loss": 2.1568, "step": 451 }, { "epoch": 0.13719836090453785, "grad_norm": 0.3601212203502655, "learning_rate": 9.776247848537006e-05, "loss": 1.8744, "step": 452 }, { "epoch": 0.13750189710122931, "grad_norm": 0.40716952085494995, "learning_rate": 9.775741621949986e-05, "loss": 2.1052, "step": 453 }, { "epoch": 0.13780543329792078, "grad_norm": 0.37777504324913025, "learning_rate": 9.775235395362965e-05, "loss": 1.8896, "step": 454 }, { "epoch": 0.13810896949461224, "grad_norm": 0.368600994348526, "learning_rate": 9.774729168775945e-05, "loss": 1.8285, "step": 455 }, { "epoch": 0.1384125056913037, "grad_norm": 0.41742029786109924, "learning_rate": 9.774222942188924e-05, "loss": 1.8286, "step": 456 }, { "epoch": 0.13871604188799513, "grad_norm": 0.40132156014442444, "learning_rate": 9.773716715601904e-05, "loss": 1.9515, "step": 457 }, { "epoch": 0.1390195780846866, "grad_norm": 0.44473376870155334, "learning_rate": 9.773210489014883e-05, "loss": 1.8715, "step": 458 }, { "epoch": 0.13932311428137806, "grad_norm": 0.40146371722221375, "learning_rate": 9.772704262427863e-05, "loss": 2.1469, "step": 459 }, { "epoch": 0.13962665047806952, "grad_norm": 0.3863317370414734, "learning_rate": 9.772198035840842e-05, "loss": 1.9215, "step": 460 }, { "epoch": 0.13993018667476095, "grad_norm": 0.40235334634780884, "learning_rate": 9.771691809253823e-05, "loss": 2.1276, "step": 461 }, { "epoch": 0.14023372287145242, "grad_norm": 0.46011632680892944, "learning_rate": 9.771185582666802e-05, "loss": 1.244, "step": 462 }, { "epoch": 0.14053725906814388, "grad_norm": 0.3428272008895874, "learning_rate": 9.770679356079782e-05, "loss": 1.7991, "step": 463 }, { "epoch": 0.14084079526483534, "grad_norm": 0.39976757764816284, "learning_rate": 9.770173129492761e-05, "loss": 1.7166, "step": 464 }, { "epoch": 0.1411443314615268, "grad_norm": 0.3258446753025055, "learning_rate": 9.769666902905741e-05, "loss": 1.677, "step": 465 }, { "epoch": 0.14144786765821823, "grad_norm": 0.3950905501842499, "learning_rate": 9.76916067631872e-05, "loss": 2.0122, "step": 466 }, { "epoch": 0.1417514038549097, "grad_norm": 0.39712047576904297, "learning_rate": 9.7686544497317e-05, "loss": 1.7262, "step": 467 }, { "epoch": 0.14205494005160116, "grad_norm": 0.8331599235534668, "learning_rate": 9.768148223144679e-05, "loss": 1.9852, "step": 468 }, { "epoch": 0.14235847624829262, "grad_norm": 0.3578427731990814, "learning_rate": 9.767641996557659e-05, "loss": 1.8249, "step": 469 }, { "epoch": 0.14266201244498405, "grad_norm": 0.3736058473587036, "learning_rate": 9.767135769970638e-05, "loss": 1.43, "step": 470 }, { "epoch": 0.14296554864167552, "grad_norm": 0.48153185844421387, "learning_rate": 9.766629543383619e-05, "loss": 1.8667, "step": 471 }, { "epoch": 0.14326908483836698, "grad_norm": 0.3924524188041687, "learning_rate": 9.766123316796599e-05, "loss": 2.0385, "step": 472 }, { "epoch": 0.14357262103505844, "grad_norm": 0.38956940174102783, "learning_rate": 9.765617090209578e-05, "loss": 1.3157, "step": 473 }, { "epoch": 0.14387615723174987, "grad_norm": 0.4032903015613556, "learning_rate": 9.765110863622558e-05, "loss": 1.8793, "step": 474 }, { "epoch": 0.14417969342844134, "grad_norm": 0.5116568207740784, "learning_rate": 9.764604637035537e-05, "loss": 1.7658, "step": 475 }, { "epoch": 0.1444832296251328, "grad_norm": 0.3981756269931793, "learning_rate": 9.764098410448517e-05, "loss": 1.8087, "step": 476 }, { "epoch": 0.14478676582182426, "grad_norm": 0.43181854486465454, "learning_rate": 9.763592183861496e-05, "loss": 1.5241, "step": 477 }, { "epoch": 0.14509030201851572, "grad_norm": 0.4172961413860321, "learning_rate": 9.763085957274477e-05, "loss": 1.8318, "step": 478 }, { "epoch": 0.14539383821520716, "grad_norm": 0.4135033190250397, "learning_rate": 9.762579730687456e-05, "loss": 2.0783, "step": 479 }, { "epoch": 0.14569737441189862, "grad_norm": 0.36482739448547363, "learning_rate": 9.762073504100436e-05, "loss": 2.2524, "step": 480 }, { "epoch": 0.14600091060859008, "grad_norm": 0.3704656958580017, "learning_rate": 9.761567277513415e-05, "loss": 2.0369, "step": 481 }, { "epoch": 0.14630444680528154, "grad_norm": 1.588393211364746, "learning_rate": 9.761061050926396e-05, "loss": 1.8041, "step": 482 }, { "epoch": 0.14660798300197297, "grad_norm": 0.3309743404388428, "learning_rate": 9.760554824339376e-05, "loss": 1.8373, "step": 483 }, { "epoch": 0.14691151919866444, "grad_norm": 0.34598830342292786, "learning_rate": 9.760048597752355e-05, "loss": 1.6249, "step": 484 }, { "epoch": 0.1472150553953559, "grad_norm": 0.3433639109134674, "learning_rate": 9.759542371165335e-05, "loss": 1.9454, "step": 485 }, { "epoch": 0.14751859159204736, "grad_norm": 0.3801734149456024, "learning_rate": 9.759036144578314e-05, "loss": 2.1067, "step": 486 }, { "epoch": 0.1478221277887388, "grad_norm": 0.36811041831970215, "learning_rate": 9.758529917991293e-05, "loss": 1.8642, "step": 487 }, { "epoch": 0.14812566398543026, "grad_norm": 0.3999156355857849, "learning_rate": 9.758023691404273e-05, "loss": 2.1482, "step": 488 }, { "epoch": 0.14842920018212172, "grad_norm": 0.7651489973068237, "learning_rate": 9.757517464817252e-05, "loss": 1.8213, "step": 489 }, { "epoch": 0.14873273637881318, "grad_norm": 0.3491712808609009, "learning_rate": 9.757011238230232e-05, "loss": 2.1047, "step": 490 }, { "epoch": 0.14903627257550464, "grad_norm": 1.028256893157959, "learning_rate": 9.756505011643213e-05, "loss": 2.0519, "step": 491 }, { "epoch": 0.14933980877219608, "grad_norm": 0.5957101583480835, "learning_rate": 9.755998785056192e-05, "loss": 2.1236, "step": 492 }, { "epoch": 0.14964334496888754, "grad_norm": 0.40934717655181885, "learning_rate": 9.755492558469172e-05, "loss": 1.5391, "step": 493 }, { "epoch": 0.149946881165579, "grad_norm": 0.4403507709503174, "learning_rate": 9.754986331882151e-05, "loss": 1.8388, "step": 494 }, { "epoch": 0.15025041736227046, "grad_norm": 0.4258563220500946, "learning_rate": 9.754480105295131e-05, "loss": 1.8092, "step": 495 }, { "epoch": 0.1505539535589619, "grad_norm": 0.3594823181629181, "learning_rate": 9.75397387870811e-05, "loss": 1.7195, "step": 496 }, { "epoch": 0.15085748975565336, "grad_norm": 0.30373120307922363, "learning_rate": 9.75346765212109e-05, "loss": 1.9267, "step": 497 }, { "epoch": 0.15116102595234482, "grad_norm": 0.423096626996994, "learning_rate": 9.752961425534069e-05, "loss": 2.1559, "step": 498 }, { "epoch": 0.15146456214903628, "grad_norm": 0.36935552954673767, "learning_rate": 9.752455198947049e-05, "loss": 2.0357, "step": 499 }, { "epoch": 0.15176809834572771, "grad_norm": 0.7172725200653076, "learning_rate": 9.75194897236003e-05, "loss": 2.0973, "step": 500 }, { "epoch": 0.15207163454241918, "grad_norm": 0.36897605657577515, "learning_rate": 9.751442745773009e-05, "loss": 2.1672, "step": 501 }, { "epoch": 0.15237517073911064, "grad_norm": 0.35079488158226013, "learning_rate": 9.750936519185988e-05, "loss": 2.0808, "step": 502 }, { "epoch": 0.1526787069358021, "grad_norm": 0.37833186984062195, "learning_rate": 9.750430292598968e-05, "loss": 1.8393, "step": 503 }, { "epoch": 0.15298224313249356, "grad_norm": 0.3969264328479767, "learning_rate": 9.749924066011947e-05, "loss": 2.1213, "step": 504 }, { "epoch": 0.153285779329185, "grad_norm": 0.30432841181755066, "learning_rate": 9.749417839424927e-05, "loss": 1.6397, "step": 505 }, { "epoch": 0.15358931552587646, "grad_norm": 0.30847886204719543, "learning_rate": 9.748911612837906e-05, "loss": 1.6455, "step": 506 }, { "epoch": 0.15389285172256792, "grad_norm": 0.38480496406555176, "learning_rate": 9.748405386250886e-05, "loss": 1.803, "step": 507 }, { "epoch": 0.15419638791925938, "grad_norm": 0.48439183831214905, "learning_rate": 9.747899159663865e-05, "loss": 1.6892, "step": 508 }, { "epoch": 0.15449992411595082, "grad_norm": 0.5124354362487793, "learning_rate": 9.747392933076845e-05, "loss": 2.24, "step": 509 }, { "epoch": 0.15480346031264228, "grad_norm": 0.4051717221736908, "learning_rate": 9.746886706489826e-05, "loss": 1.8621, "step": 510 }, { "epoch": 0.15510699650933374, "grad_norm": 0.6452261209487915, "learning_rate": 9.746380479902805e-05, "loss": 1.7043, "step": 511 }, { "epoch": 0.1554105327060252, "grad_norm": 0.5453522801399231, "learning_rate": 9.745874253315785e-05, "loss": 1.7325, "step": 512 }, { "epoch": 0.15571406890271666, "grad_norm": 1.0983595848083496, "learning_rate": 9.745368026728764e-05, "loss": 2.169, "step": 513 }, { "epoch": 0.1560176050994081, "grad_norm": 0.3821035623550415, "learning_rate": 9.744861800141744e-05, "loss": 2.3305, "step": 514 }, { "epoch": 0.15632114129609956, "grad_norm": 0.3694508969783783, "learning_rate": 9.744355573554723e-05, "loss": 1.8453, "step": 515 }, { "epoch": 0.15662467749279102, "grad_norm": 0.3837510943412781, "learning_rate": 9.743849346967702e-05, "loss": 1.9679, "step": 516 }, { "epoch": 0.15692821368948248, "grad_norm": 0.41427966952323914, "learning_rate": 9.743343120380682e-05, "loss": 1.9331, "step": 517 }, { "epoch": 0.15723174988617392, "grad_norm": 0.34252259135246277, "learning_rate": 9.742836893793661e-05, "loss": 1.7938, "step": 518 }, { "epoch": 0.15753528608286538, "grad_norm": 0.4043283462524414, "learning_rate": 9.742330667206642e-05, "loss": 1.4037, "step": 519 }, { "epoch": 0.15783882227955684, "grad_norm": 0.4225389361381531, "learning_rate": 9.741824440619622e-05, "loss": 1.6224, "step": 520 }, { "epoch": 0.1581423584762483, "grad_norm": 0.377590537071228, "learning_rate": 9.741318214032601e-05, "loss": 2.0567, "step": 521 }, { "epoch": 0.15844589467293974, "grad_norm": 0.46170124411582947, "learning_rate": 9.740811987445582e-05, "loss": 2.0449, "step": 522 }, { "epoch": 0.1587494308696312, "grad_norm": 0.3752427399158478, "learning_rate": 9.740305760858562e-05, "loss": 1.8207, "step": 523 }, { "epoch": 0.15905296706632266, "grad_norm": 0.390803724527359, "learning_rate": 9.739799534271541e-05, "loss": 2.0781, "step": 524 }, { "epoch": 0.15935650326301412, "grad_norm": 0.38587453961372375, "learning_rate": 9.73929330768452e-05, "loss": 1.9932, "step": 525 }, { "epoch": 0.15966003945970558, "grad_norm": 0.4154350459575653, "learning_rate": 9.7387870810975e-05, "loss": 1.7649, "step": 526 }, { "epoch": 0.15996357565639702, "grad_norm": 0.3698589503765106, "learning_rate": 9.73828085451048e-05, "loss": 1.6921, "step": 527 }, { "epoch": 0.16026711185308848, "grad_norm": 0.4110312759876251, "learning_rate": 9.737774627923459e-05, "loss": 1.1834, "step": 528 }, { "epoch": 0.16057064804977994, "grad_norm": 0.4140758812427521, "learning_rate": 9.737268401336438e-05, "loss": 1.8354, "step": 529 }, { "epoch": 0.1608741842464714, "grad_norm": 0.38738423585891724, "learning_rate": 9.736762174749419e-05, "loss": 1.9223, "step": 530 }, { "epoch": 0.16117772044316284, "grad_norm": 0.4055260717868805, "learning_rate": 9.736255948162399e-05, "loss": 1.7802, "step": 531 }, { "epoch": 0.1614812566398543, "grad_norm": 0.44946524500846863, "learning_rate": 9.735749721575378e-05, "loss": 1.8654, "step": 532 }, { "epoch": 0.16178479283654576, "grad_norm": 0.43206432461738586, "learning_rate": 9.735243494988358e-05, "loss": 1.7607, "step": 533 }, { "epoch": 0.16208832903323722, "grad_norm": 0.5007991194725037, "learning_rate": 9.734737268401337e-05, "loss": 1.9378, "step": 534 }, { "epoch": 0.16239186522992866, "grad_norm": 0.48757919669151306, "learning_rate": 9.734231041814317e-05, "loss": 2.1829, "step": 535 }, { "epoch": 0.16269540142662012, "grad_norm": 0.4159701466560364, "learning_rate": 9.733724815227296e-05, "loss": 1.8847, "step": 536 }, { "epoch": 0.16299893762331158, "grad_norm": 0.40922749042510986, "learning_rate": 9.733218588640276e-05, "loss": 1.4376, "step": 537 }, { "epoch": 0.16330247382000304, "grad_norm": 0.33677083253860474, "learning_rate": 9.732712362053255e-05, "loss": 1.9568, "step": 538 }, { "epoch": 0.1636060100166945, "grad_norm": 0.3255022168159485, "learning_rate": 9.732206135466236e-05, "loss": 1.9949, "step": 539 }, { "epoch": 0.16390954621338594, "grad_norm": 0.3848338723182678, "learning_rate": 9.731699908879215e-05, "loss": 2.042, "step": 540 }, { "epoch": 0.1642130824100774, "grad_norm": 0.3888263404369354, "learning_rate": 9.731193682292195e-05, "loss": 1.885, "step": 541 }, { "epoch": 0.16451661860676886, "grad_norm": 0.40090805292129517, "learning_rate": 9.730687455705174e-05, "loss": 1.9093, "step": 542 }, { "epoch": 0.16482015480346032, "grad_norm": 0.4106220602989197, "learning_rate": 9.730181229118154e-05, "loss": 1.8392, "step": 543 }, { "epoch": 0.16512369100015176, "grad_norm": 0.3483395278453827, "learning_rate": 9.729675002531133e-05, "loss": 2.0235, "step": 544 }, { "epoch": 0.16542722719684322, "grad_norm": 0.3686208128929138, "learning_rate": 9.729168775944113e-05, "loss": 1.9218, "step": 545 }, { "epoch": 0.16573076339353468, "grad_norm": 0.36063849925994873, "learning_rate": 9.728662549357092e-05, "loss": 1.9334, "step": 546 }, { "epoch": 0.16603429959022614, "grad_norm": 0.39365142583847046, "learning_rate": 9.728156322770072e-05, "loss": 1.9825, "step": 547 }, { "epoch": 0.16633783578691758, "grad_norm": 0.4062787592411041, "learning_rate": 9.727650096183051e-05, "loss": 1.521, "step": 548 }, { "epoch": 0.16664137198360904, "grad_norm": 0.37347134947776794, "learning_rate": 9.727143869596032e-05, "loss": 1.9356, "step": 549 }, { "epoch": 0.1669449081803005, "grad_norm": 0.3538997173309326, "learning_rate": 9.726637643009012e-05, "loss": 1.845, "step": 550 }, { "epoch": 0.16724844437699196, "grad_norm": 0.3868335783481598, "learning_rate": 9.726131416421991e-05, "loss": 1.9803, "step": 551 }, { "epoch": 0.16755198057368342, "grad_norm": 0.34705451130867004, "learning_rate": 9.72562518983497e-05, "loss": 2.0866, "step": 552 }, { "epoch": 0.16785551677037486, "grad_norm": 0.3794872462749481, "learning_rate": 9.72511896324795e-05, "loss": 2.094, "step": 553 }, { "epoch": 0.16815905296706632, "grad_norm": 0.5801231861114502, "learning_rate": 9.72461273666093e-05, "loss": 1.7851, "step": 554 }, { "epoch": 0.16846258916375778, "grad_norm": 0.3076344132423401, "learning_rate": 9.724106510073909e-05, "loss": 1.5188, "step": 555 }, { "epoch": 0.16876612536044924, "grad_norm": 0.3552989363670349, "learning_rate": 9.723600283486888e-05, "loss": 2.1063, "step": 556 }, { "epoch": 0.16906966155714068, "grad_norm": 0.36939847469329834, "learning_rate": 9.723094056899868e-05, "loss": 1.7648, "step": 557 }, { "epoch": 0.16937319775383214, "grad_norm": 0.358634889125824, "learning_rate": 9.722587830312849e-05, "loss": 1.8007, "step": 558 }, { "epoch": 0.1696767339505236, "grad_norm": 0.39962029457092285, "learning_rate": 9.722081603725828e-05, "loss": 1.8845, "step": 559 }, { "epoch": 0.16998027014721506, "grad_norm": 0.4099076986312866, "learning_rate": 9.721575377138808e-05, "loss": 1.8894, "step": 560 }, { "epoch": 0.17028380634390652, "grad_norm": 0.3610551655292511, "learning_rate": 9.721069150551787e-05, "loss": 1.8089, "step": 561 }, { "epoch": 0.17058734254059796, "grad_norm": 0.5951200723648071, "learning_rate": 9.720562923964767e-05, "loss": 1.6966, "step": 562 }, { "epoch": 0.17089087873728942, "grad_norm": 0.562522292137146, "learning_rate": 9.720056697377746e-05, "loss": 1.7704, "step": 563 }, { "epoch": 0.17119441493398088, "grad_norm": 0.6662526726722717, "learning_rate": 9.719550470790726e-05, "loss": 1.7714, "step": 564 }, { "epoch": 0.17149795113067234, "grad_norm": 0.44034865498542786, "learning_rate": 9.719044244203705e-05, "loss": 2.1042, "step": 565 }, { "epoch": 0.17180148732736378, "grad_norm": 0.39868202805519104, "learning_rate": 9.718538017616685e-05, "loss": 1.952, "step": 566 }, { "epoch": 0.17210502352405524, "grad_norm": 0.3427380621433258, "learning_rate": 9.718031791029665e-05, "loss": 2.037, "step": 567 }, { "epoch": 0.1724085597207467, "grad_norm": 0.37980929017066956, "learning_rate": 9.717525564442645e-05, "loss": 1.5378, "step": 568 }, { "epoch": 0.17271209591743816, "grad_norm": 0.32314518094062805, "learning_rate": 9.717019337855626e-05, "loss": 1.6191, "step": 569 }, { "epoch": 0.1730156321141296, "grad_norm": 0.40600740909576416, "learning_rate": 9.716513111268605e-05, "loss": 1.6055, "step": 570 }, { "epoch": 0.17331916831082106, "grad_norm": 0.37318041920661926, "learning_rate": 9.716006884681585e-05, "loss": 1.8666, "step": 571 }, { "epoch": 0.17362270450751252, "grad_norm": 0.3656068444252014, "learning_rate": 9.715500658094564e-05, "loss": 1.5983, "step": 572 }, { "epoch": 0.17392624070420398, "grad_norm": 0.3546827733516693, "learning_rate": 9.714994431507544e-05, "loss": 2.2088, "step": 573 }, { "epoch": 0.17422977690089544, "grad_norm": 0.4293152689933777, "learning_rate": 9.714488204920523e-05, "loss": 1.803, "step": 574 }, { "epoch": 0.17453331309758688, "grad_norm": 0.3790314495563507, "learning_rate": 9.713981978333503e-05, "loss": 1.9874, "step": 575 }, { "epoch": 0.17483684929427834, "grad_norm": 0.37619829177856445, "learning_rate": 9.713475751746482e-05, "loss": 1.9061, "step": 576 }, { "epoch": 0.1751403854909698, "grad_norm": 0.36988991498947144, "learning_rate": 9.712969525159462e-05, "loss": 1.5463, "step": 577 }, { "epoch": 0.17544392168766126, "grad_norm": 0.367721825838089, "learning_rate": 9.712463298572442e-05, "loss": 1.6526, "step": 578 }, { "epoch": 0.1757474578843527, "grad_norm": 0.39620110392570496, "learning_rate": 9.711957071985422e-05, "loss": 2.056, "step": 579 }, { "epoch": 0.17605099408104416, "grad_norm": 0.41518276929855347, "learning_rate": 9.711450845398401e-05, "loss": 1.6847, "step": 580 }, { "epoch": 0.17635453027773562, "grad_norm": 0.3925170302391052, "learning_rate": 9.710944618811381e-05, "loss": 1.8476, "step": 581 }, { "epoch": 0.17665806647442708, "grad_norm": 0.36658090353012085, "learning_rate": 9.71043839222436e-05, "loss": 2.0699, "step": 582 }, { "epoch": 0.17696160267111852, "grad_norm": 0.3741433620452881, "learning_rate": 9.70993216563734e-05, "loss": 1.9645, "step": 583 }, { "epoch": 0.17726513886780998, "grad_norm": 0.3742316663265228, "learning_rate": 9.709425939050319e-05, "loss": 2.3717, "step": 584 }, { "epoch": 0.17756867506450144, "grad_norm": 0.3796440660953522, "learning_rate": 9.708919712463299e-05, "loss": 1.9356, "step": 585 }, { "epoch": 0.1778722112611929, "grad_norm": 0.3976511061191559, "learning_rate": 9.708413485876278e-05, "loss": 2.1889, "step": 586 }, { "epoch": 0.17817574745788436, "grad_norm": 0.34445542097091675, "learning_rate": 9.707907259289258e-05, "loss": 1.6535, "step": 587 }, { "epoch": 0.1784792836545758, "grad_norm": 0.3982098698616028, "learning_rate": 9.707401032702239e-05, "loss": 2.0542, "step": 588 }, { "epoch": 0.17878281985126726, "grad_norm": 0.42155295610427856, "learning_rate": 9.706894806115218e-05, "loss": 1.4605, "step": 589 }, { "epoch": 0.17908635604795872, "grad_norm": 0.36341744661331177, "learning_rate": 9.706388579528197e-05, "loss": 1.8069, "step": 590 }, { "epoch": 0.17938989224465018, "grad_norm": 0.3715178668498993, "learning_rate": 9.705882352941177e-05, "loss": 1.5512, "step": 591 }, { "epoch": 0.17969342844134162, "grad_norm": 0.376767635345459, "learning_rate": 9.705376126354156e-05, "loss": 1.6027, "step": 592 }, { "epoch": 0.17999696463803308, "grad_norm": 0.4033347964286804, "learning_rate": 9.704869899767136e-05, "loss": 1.5071, "step": 593 }, { "epoch": 0.18030050083472454, "grad_norm": 0.8200478553771973, "learning_rate": 9.704363673180115e-05, "loss": 1.924, "step": 594 }, { "epoch": 0.180604037031416, "grad_norm": 0.6224507093429565, "learning_rate": 9.703857446593095e-05, "loss": 1.9684, "step": 595 }, { "epoch": 0.18090757322810747, "grad_norm": 0.32032859325408936, "learning_rate": 9.703351220006074e-05, "loss": 1.9478, "step": 596 }, { "epoch": 0.1812111094247989, "grad_norm": 0.33331337571144104, "learning_rate": 9.702844993419055e-05, "loss": 1.8177, "step": 597 }, { "epoch": 0.18151464562149036, "grad_norm": 0.47399207949638367, "learning_rate": 9.702338766832035e-05, "loss": 2.07, "step": 598 }, { "epoch": 0.18181818181818182, "grad_norm": 0.30480411648750305, "learning_rate": 9.701832540245014e-05, "loss": 2.0407, "step": 599 }, { "epoch": 0.18212171801487329, "grad_norm": 0.40148988366127014, "learning_rate": 9.701326313657994e-05, "loss": 1.8774, "step": 600 }, { "epoch": 0.18242525421156472, "grad_norm": 0.3958423137664795, "learning_rate": 9.700820087070973e-05, "loss": 1.8462, "step": 601 }, { "epoch": 0.18272879040825618, "grad_norm": 0.34824639558792114, "learning_rate": 9.700313860483953e-05, "loss": 1.7839, "step": 602 }, { "epoch": 0.18303232660494764, "grad_norm": 0.38002872467041016, "learning_rate": 9.699807633896932e-05, "loss": 2.3237, "step": 603 }, { "epoch": 0.1833358628016391, "grad_norm": 0.37800419330596924, "learning_rate": 9.699301407309912e-05, "loss": 1.9375, "step": 604 }, { "epoch": 0.18363939899833054, "grad_norm": 0.4041115939617157, "learning_rate": 9.698795180722891e-05, "loss": 2.029, "step": 605 }, { "epoch": 0.183942935195022, "grad_norm": 0.3697315454483032, "learning_rate": 9.698288954135872e-05, "loss": 1.894, "step": 606 }, { "epoch": 0.18424647139171346, "grad_norm": 0.3809906542301178, "learning_rate": 9.697782727548851e-05, "loss": 1.8242, "step": 607 }, { "epoch": 0.18455000758840492, "grad_norm": 0.3997717499732971, "learning_rate": 9.697276500961831e-05, "loss": 2.0522, "step": 608 }, { "epoch": 0.18485354378509639, "grad_norm": 0.391699880361557, "learning_rate": 9.69677027437481e-05, "loss": 1.8521, "step": 609 }, { "epoch": 0.18515707998178782, "grad_norm": 0.3667858839035034, "learning_rate": 9.69626404778779e-05, "loss": 1.7613, "step": 610 }, { "epoch": 0.18546061617847928, "grad_norm": 0.3905411958694458, "learning_rate": 9.69575782120077e-05, "loss": 1.8285, "step": 611 }, { "epoch": 0.18576415237517074, "grad_norm": 0.4121951758861542, "learning_rate": 9.69525159461375e-05, "loss": 1.8104, "step": 612 }, { "epoch": 0.1860676885718622, "grad_norm": 0.34977591037750244, "learning_rate": 9.69474536802673e-05, "loss": 1.7737, "step": 613 }, { "epoch": 0.18637122476855364, "grad_norm": 0.34084367752075195, "learning_rate": 9.694239141439709e-05, "loss": 2.0407, "step": 614 }, { "epoch": 0.1866747609652451, "grad_norm": 0.35442525148391724, "learning_rate": 9.693732914852689e-05, "loss": 1.9152, "step": 615 }, { "epoch": 0.18697829716193656, "grad_norm": 0.34404149651527405, "learning_rate": 9.693226688265668e-05, "loss": 1.7621, "step": 616 }, { "epoch": 0.18728183335862802, "grad_norm": 0.4516477882862091, "learning_rate": 9.692720461678649e-05, "loss": 1.7624, "step": 617 }, { "epoch": 0.18758536955531946, "grad_norm": 0.3506614565849304, "learning_rate": 9.692214235091628e-05, "loss": 1.6627, "step": 618 }, { "epoch": 0.18788890575201092, "grad_norm": 0.9165719151496887, "learning_rate": 9.691708008504608e-05, "loss": 2.1926, "step": 619 }, { "epoch": 0.18819244194870238, "grad_norm": 0.3361871838569641, "learning_rate": 9.691201781917587e-05, "loss": 1.5229, "step": 620 }, { "epoch": 0.18849597814539384, "grad_norm": 0.32639381289482117, "learning_rate": 9.690695555330567e-05, "loss": 1.8778, "step": 621 }, { "epoch": 0.1887995143420853, "grad_norm": 0.44261273741722107, "learning_rate": 9.690189328743546e-05, "loss": 2.0903, "step": 622 }, { "epoch": 0.18910305053877674, "grad_norm": 0.4438890516757965, "learning_rate": 9.689683102156526e-05, "loss": 1.772, "step": 623 }, { "epoch": 0.1894065867354682, "grad_norm": 0.40160682797431946, "learning_rate": 9.689176875569505e-05, "loss": 2.0964, "step": 624 }, { "epoch": 0.18971012293215966, "grad_norm": 0.4022195637226105, "learning_rate": 9.688670648982485e-05, "loss": 1.7818, "step": 625 }, { "epoch": 0.19001365912885113, "grad_norm": 0.4233214855194092, "learning_rate": 9.688164422395464e-05, "loss": 1.922, "step": 626 }, { "epoch": 0.19031719532554256, "grad_norm": 0.3864254057407379, "learning_rate": 9.687658195808445e-05, "loss": 2.0279, "step": 627 }, { "epoch": 0.19062073152223402, "grad_norm": 0.36527585983276367, "learning_rate": 9.687151969221424e-05, "loss": 2.0732, "step": 628 }, { "epoch": 0.19092426771892548, "grad_norm": 0.399237722158432, "learning_rate": 9.686645742634404e-05, "loss": 1.8889, "step": 629 }, { "epoch": 0.19122780391561695, "grad_norm": 0.3860459625720978, "learning_rate": 9.686139516047383e-05, "loss": 1.968, "step": 630 }, { "epoch": 0.19153134011230838, "grad_norm": 0.32555973529815674, "learning_rate": 9.685633289460363e-05, "loss": 2.0722, "step": 631 }, { "epoch": 0.19183487630899984, "grad_norm": 0.6093998551368713, "learning_rate": 9.685127062873342e-05, "loss": 1.8553, "step": 632 }, { "epoch": 0.1921384125056913, "grad_norm": 0.4218057692050934, "learning_rate": 9.684620836286322e-05, "loss": 1.9647, "step": 633 }, { "epoch": 0.19244194870238276, "grad_norm": 0.3779148757457733, "learning_rate": 9.684114609699301e-05, "loss": 2.0681, "step": 634 }, { "epoch": 0.19274548489907423, "grad_norm": 0.3820381760597229, "learning_rate": 9.683608383112281e-05, "loss": 2.0603, "step": 635 }, { "epoch": 0.19304902109576566, "grad_norm": 0.29337063431739807, "learning_rate": 9.683102156525262e-05, "loss": 1.7516, "step": 636 }, { "epoch": 0.19335255729245712, "grad_norm": 0.4369249939918518, "learning_rate": 9.682595929938241e-05, "loss": 1.9822, "step": 637 }, { "epoch": 0.19365609348914858, "grad_norm": 0.3766214847564697, "learning_rate": 9.68208970335122e-05, "loss": 1.7229, "step": 638 }, { "epoch": 0.19395962968584005, "grad_norm": 0.4765011668205261, "learning_rate": 9.6815834767642e-05, "loss": 1.2865, "step": 639 }, { "epoch": 0.19426316588253148, "grad_norm": 0.34236472845077515, "learning_rate": 9.68107725017718e-05, "loss": 2.1024, "step": 640 }, { "epoch": 0.19456670207922294, "grad_norm": 0.398076593875885, "learning_rate": 9.680571023590159e-05, "loss": 1.8628, "step": 641 }, { "epoch": 0.1948702382759144, "grad_norm": 0.357099711894989, "learning_rate": 9.680064797003139e-05, "loss": 2.2163, "step": 642 }, { "epoch": 0.19517377447260587, "grad_norm": 0.3296545445919037, "learning_rate": 9.679558570416118e-05, "loss": 1.8227, "step": 643 }, { "epoch": 0.19547731066929733, "grad_norm": 0.36754927039146423, "learning_rate": 9.679052343829098e-05, "loss": 1.7179, "step": 644 }, { "epoch": 0.19578084686598876, "grad_norm": 0.37275364995002747, "learning_rate": 9.678546117242078e-05, "loss": 1.6782, "step": 645 }, { "epoch": 0.19608438306268022, "grad_norm": 0.3951006531715393, "learning_rate": 9.678039890655058e-05, "loss": 2.0756, "step": 646 }, { "epoch": 0.19638791925937168, "grad_norm": 0.3560970425605774, "learning_rate": 9.677533664068037e-05, "loss": 1.8093, "step": 647 }, { "epoch": 0.19669145545606315, "grad_norm": 0.31553730368614197, "learning_rate": 9.677027437481017e-05, "loss": 1.9174, "step": 648 }, { "epoch": 0.19699499165275458, "grad_norm": 0.39949625730514526, "learning_rate": 9.676521210893996e-05, "loss": 1.6687, "step": 649 }, { "epoch": 0.19729852784944604, "grad_norm": 0.37323635816574097, "learning_rate": 9.676014984306976e-05, "loss": 1.8149, "step": 650 }, { "epoch": 0.1976020640461375, "grad_norm": 0.43527746200561523, "learning_rate": 9.675508757719955e-05, "loss": 1.8744, "step": 651 }, { "epoch": 0.19790560024282897, "grad_norm": 0.39380425214767456, "learning_rate": 9.675002531132935e-05, "loss": 1.9721, "step": 652 }, { "epoch": 0.1982091364395204, "grad_norm": 0.3384545147418976, "learning_rate": 9.674496304545914e-05, "loss": 2.0122, "step": 653 }, { "epoch": 0.19851267263621186, "grad_norm": 0.39647915959358215, "learning_rate": 9.673990077958894e-05, "loss": 2.2419, "step": 654 }, { "epoch": 0.19881620883290332, "grad_norm": 0.3358941674232483, "learning_rate": 9.673483851371875e-05, "loss": 1.8758, "step": 655 }, { "epoch": 0.19911974502959479, "grad_norm": 0.3486049771308899, "learning_rate": 9.672977624784855e-05, "loss": 1.5762, "step": 656 }, { "epoch": 0.19942328122628625, "grad_norm": 2.3050696849823, "learning_rate": 9.672471398197835e-05, "loss": 2.0056, "step": 657 }, { "epoch": 0.19972681742297768, "grad_norm": 0.35023945569992065, "learning_rate": 9.671965171610814e-05, "loss": 1.619, "step": 658 }, { "epoch": 0.20003035361966914, "grad_norm": 0.513656735420227, "learning_rate": 9.671458945023794e-05, "loss": 1.5269, "step": 659 }, { "epoch": 0.2003338898163606, "grad_norm": 0.37498149275779724, "learning_rate": 9.670952718436773e-05, "loss": 1.8553, "step": 660 }, { "epoch": 0.20063742601305207, "grad_norm": 0.4101942479610443, "learning_rate": 9.670446491849753e-05, "loss": 2.1121, "step": 661 }, { "epoch": 0.2009409622097435, "grad_norm": 0.4265679717063904, "learning_rate": 9.669940265262732e-05, "loss": 2.1863, "step": 662 }, { "epoch": 0.20124449840643496, "grad_norm": 4.817168712615967, "learning_rate": 9.669434038675712e-05, "loss": 2.0906, "step": 663 }, { "epoch": 0.20154803460312642, "grad_norm": 7.518252849578857, "learning_rate": 9.668927812088691e-05, "loss": 1.8889, "step": 664 }, { "epoch": 0.2018515707998179, "grad_norm": 0.5480749011039734, "learning_rate": 9.66842158550167e-05, "loss": 1.8439, "step": 665 }, { "epoch": 0.20215510699650932, "grad_norm": 0.3578292429447174, "learning_rate": 9.667915358914651e-05, "loss": 1.8742, "step": 666 }, { "epoch": 0.20245864319320078, "grad_norm": 0.3799275755882263, "learning_rate": 9.667409132327631e-05, "loss": 1.994, "step": 667 }, { "epoch": 0.20276217938989224, "grad_norm": 0.3736335039138794, "learning_rate": 9.66690290574061e-05, "loss": 1.7933, "step": 668 }, { "epoch": 0.2030657155865837, "grad_norm": 0.3145211637020111, "learning_rate": 9.66639667915359e-05, "loss": 1.8193, "step": 669 }, { "epoch": 0.20336925178327517, "grad_norm": 0.4940774142742157, "learning_rate": 9.66589045256657e-05, "loss": 1.9238, "step": 670 }, { "epoch": 0.2036727879799666, "grad_norm": 0.431134968996048, "learning_rate": 9.665384225979549e-05, "loss": 1.5493, "step": 671 }, { "epoch": 0.20397632417665806, "grad_norm": 0.41438859701156616, "learning_rate": 9.664877999392528e-05, "loss": 1.2076, "step": 672 }, { "epoch": 0.20427986037334953, "grad_norm": 0.38191312551498413, "learning_rate": 9.664371772805508e-05, "loss": 1.8201, "step": 673 }, { "epoch": 0.204583396570041, "grad_norm": 0.3938577175140381, "learning_rate": 9.663865546218487e-05, "loss": 1.5166, "step": 674 }, { "epoch": 0.20488693276673242, "grad_norm": 0.46312233805656433, "learning_rate": 9.663359319631468e-05, "loss": 1.4652, "step": 675 }, { "epoch": 0.20519046896342388, "grad_norm": 0.4087234139442444, "learning_rate": 9.662853093044448e-05, "loss": 1.8288, "step": 676 }, { "epoch": 0.20549400516011535, "grad_norm": 0.37329304218292236, "learning_rate": 9.662346866457427e-05, "loss": 1.9084, "step": 677 }, { "epoch": 0.2057975413568068, "grad_norm": 0.37109607458114624, "learning_rate": 9.661840639870407e-05, "loss": 1.9674, "step": 678 }, { "epoch": 0.20610107755349824, "grad_norm": 0.3936561942100525, "learning_rate": 9.661334413283386e-05, "loss": 2.0342, "step": 679 }, { "epoch": 0.2064046137501897, "grad_norm": 0.4621008634567261, "learning_rate": 9.660828186696366e-05, "loss": 1.5157, "step": 680 }, { "epoch": 0.20670814994688116, "grad_norm": 0.3849358558654785, "learning_rate": 9.660321960109345e-05, "loss": 2.1513, "step": 681 }, { "epoch": 0.20701168614357263, "grad_norm": 0.4873330295085907, "learning_rate": 9.659815733522325e-05, "loss": 1.9116, "step": 682 }, { "epoch": 0.2073152223402641, "grad_norm": 0.4687885642051697, "learning_rate": 9.659309506935304e-05, "loss": 2.278, "step": 683 }, { "epoch": 0.20761875853695552, "grad_norm": 0.3966952860355377, "learning_rate": 9.658803280348285e-05, "loss": 1.4625, "step": 684 }, { "epoch": 0.20792229473364698, "grad_norm": 0.5782402157783508, "learning_rate": 9.658297053761264e-05, "loss": 2.2779, "step": 685 }, { "epoch": 0.20822583093033845, "grad_norm": 0.37465688586235046, "learning_rate": 9.657790827174244e-05, "loss": 1.8462, "step": 686 }, { "epoch": 0.2085293671270299, "grad_norm": 0.34408631920814514, "learning_rate": 9.657284600587223e-05, "loss": 1.9881, "step": 687 }, { "epoch": 0.20883290332372134, "grad_norm": 0.6892307996749878, "learning_rate": 9.656778374000203e-05, "loss": 1.9835, "step": 688 }, { "epoch": 0.2091364395204128, "grad_norm": 0.3698042631149292, "learning_rate": 9.656272147413182e-05, "loss": 2.0665, "step": 689 }, { "epoch": 0.20943997571710427, "grad_norm": 0.41265738010406494, "learning_rate": 9.655765920826162e-05, "loss": 2.0231, "step": 690 }, { "epoch": 0.20974351191379573, "grad_norm": 0.38251030445098877, "learning_rate": 9.655259694239141e-05, "loss": 1.7058, "step": 691 }, { "epoch": 0.2100470481104872, "grad_norm": 0.468905508518219, "learning_rate": 9.65475346765212e-05, "loss": 1.6182, "step": 692 }, { "epoch": 0.21035058430717862, "grad_norm": 1.0570484399795532, "learning_rate": 9.6542472410651e-05, "loss": 2.0165, "step": 693 }, { "epoch": 0.21065412050387008, "grad_norm": 0.3978007435798645, "learning_rate": 9.653741014478081e-05, "loss": 1.7859, "step": 694 }, { "epoch": 0.21095765670056155, "grad_norm": 0.42616939544677734, "learning_rate": 9.65323478789106e-05, "loss": 1.5197, "step": 695 }, { "epoch": 0.211261192897253, "grad_norm": 0.39380377531051636, "learning_rate": 9.65272856130404e-05, "loss": 1.3796, "step": 696 }, { "epoch": 0.21156472909394444, "grad_norm": 0.38581010699272156, "learning_rate": 9.65222233471702e-05, "loss": 1.8214, "step": 697 }, { "epoch": 0.2118682652906359, "grad_norm": 0.3610150218009949, "learning_rate": 9.651716108129999e-05, "loss": 1.897, "step": 698 }, { "epoch": 0.21217180148732737, "grad_norm": 0.44913700222969055, "learning_rate": 9.651209881542978e-05, "loss": 1.8873, "step": 699 }, { "epoch": 0.21247533768401883, "grad_norm": 1.9599745273590088, "learning_rate": 9.650703654955959e-05, "loss": 1.946, "step": 700 }, { "epoch": 0.21277887388071026, "grad_norm": 1.195716381072998, "learning_rate": 9.650197428368939e-05, "loss": 1.8749, "step": 701 }, { "epoch": 0.21308241007740172, "grad_norm": 0.3154665231704712, "learning_rate": 9.649691201781918e-05, "loss": 1.5924, "step": 702 }, { "epoch": 0.21338594627409319, "grad_norm": 0.3550672233104706, "learning_rate": 9.649184975194898e-05, "loss": 1.6094, "step": 703 }, { "epoch": 0.21368948247078465, "grad_norm": 0.33744126558303833, "learning_rate": 9.648678748607877e-05, "loss": 1.3399, "step": 704 }, { "epoch": 0.2139930186674761, "grad_norm": 0.33931079506874084, "learning_rate": 9.648172522020858e-05, "loss": 2.0096, "step": 705 }, { "epoch": 0.21429655486416754, "grad_norm": 0.38951364159584045, "learning_rate": 9.647666295433837e-05, "loss": 1.7676, "step": 706 }, { "epoch": 0.214600091060859, "grad_norm": 0.408087819814682, "learning_rate": 9.647160068846817e-05, "loss": 1.7948, "step": 707 }, { "epoch": 0.21490362725755047, "grad_norm": 0.37058812379837036, "learning_rate": 9.646653842259796e-05, "loss": 1.9891, "step": 708 }, { "epoch": 0.21520716345424193, "grad_norm": 0.4003254473209381, "learning_rate": 9.646147615672776e-05, "loss": 1.8895, "step": 709 }, { "epoch": 0.21551069965093336, "grad_norm": 0.38838204741477966, "learning_rate": 9.645641389085755e-05, "loss": 2.0121, "step": 710 }, { "epoch": 0.21581423584762482, "grad_norm": 0.41912707686424255, "learning_rate": 9.645135162498735e-05, "loss": 1.9804, "step": 711 }, { "epoch": 0.2161177720443163, "grad_norm": 0.353454053401947, "learning_rate": 9.644628935911714e-05, "loss": 2.0478, "step": 712 }, { "epoch": 0.21642130824100775, "grad_norm": 0.3825720548629761, "learning_rate": 9.644122709324694e-05, "loss": 1.6676, "step": 713 }, { "epoch": 0.21672484443769918, "grad_norm": 0.4197389781475067, "learning_rate": 9.643616482737675e-05, "loss": 1.9732, "step": 714 }, { "epoch": 0.21702838063439064, "grad_norm": 0.4452435076236725, "learning_rate": 9.643110256150654e-05, "loss": 2.0918, "step": 715 }, { "epoch": 0.2173319168310821, "grad_norm": 0.3366299271583557, "learning_rate": 9.642604029563634e-05, "loss": 1.7469, "step": 716 }, { "epoch": 0.21763545302777357, "grad_norm": 0.31280553340911865, "learning_rate": 9.642097802976613e-05, "loss": 2.0348, "step": 717 }, { "epoch": 0.21793898922446503, "grad_norm": 0.425503671169281, "learning_rate": 9.641591576389593e-05, "loss": 1.3629, "step": 718 }, { "epoch": 0.21824252542115646, "grad_norm": 0.3986441493034363, "learning_rate": 9.641085349802572e-05, "loss": 1.4703, "step": 719 }, { "epoch": 0.21854606161784793, "grad_norm": 0.34377026557922363, "learning_rate": 9.640579123215552e-05, "loss": 1.9788, "step": 720 }, { "epoch": 0.2188495978145394, "grad_norm": 0.3445621430873871, "learning_rate": 9.640072896628531e-05, "loss": 1.9137, "step": 721 }, { "epoch": 0.21915313401123085, "grad_norm": 0.40363574028015137, "learning_rate": 9.63956667004151e-05, "loss": 1.8911, "step": 722 }, { "epoch": 0.21945667020792228, "grad_norm": 0.36166059970855713, "learning_rate": 9.639060443454491e-05, "loss": 1.9176, "step": 723 }, { "epoch": 0.21976020640461374, "grad_norm": 0.7732321619987488, "learning_rate": 9.638554216867471e-05, "loss": 2.1942, "step": 724 }, { "epoch": 0.2200637426013052, "grad_norm": 0.4042604863643646, "learning_rate": 9.63804799028045e-05, "loss": 1.8964, "step": 725 }, { "epoch": 0.22036727879799667, "grad_norm": 0.3888862133026123, "learning_rate": 9.63754176369343e-05, "loss": 1.716, "step": 726 }, { "epoch": 0.22067081499468813, "grad_norm": 0.32185250520706177, "learning_rate": 9.637035537106409e-05, "loss": 2.1227, "step": 727 }, { "epoch": 0.22097435119137956, "grad_norm": 0.36421746015548706, "learning_rate": 9.636529310519389e-05, "loss": 1.3262, "step": 728 }, { "epoch": 0.22127788738807103, "grad_norm": 0.42780765891075134, "learning_rate": 9.636023083932368e-05, "loss": 1.806, "step": 729 }, { "epoch": 0.2215814235847625, "grad_norm": 0.3754510283470154, "learning_rate": 9.635516857345348e-05, "loss": 1.9286, "step": 730 }, { "epoch": 0.22188495978145395, "grad_norm": 0.35199174284935, "learning_rate": 9.635010630758327e-05, "loss": 1.9703, "step": 731 }, { "epoch": 0.22218849597814538, "grad_norm": 0.36272746324539185, "learning_rate": 9.634504404171307e-05, "loss": 1.7773, "step": 732 }, { "epoch": 0.22249203217483685, "grad_norm": 0.4233802556991577, "learning_rate": 9.633998177584287e-05, "loss": 2.0016, "step": 733 }, { "epoch": 0.2227955683715283, "grad_norm": 0.46138089895248413, "learning_rate": 9.633491950997267e-05, "loss": 1.764, "step": 734 }, { "epoch": 0.22309910456821977, "grad_norm": 0.37863031029701233, "learning_rate": 9.632985724410246e-05, "loss": 1.6493, "step": 735 }, { "epoch": 0.2234026407649112, "grad_norm": 0.4493837356567383, "learning_rate": 9.632479497823226e-05, "loss": 2.04, "step": 736 }, { "epoch": 0.22370617696160267, "grad_norm": 0.581119179725647, "learning_rate": 9.631973271236205e-05, "loss": 1.777, "step": 737 }, { "epoch": 0.22400971315829413, "grad_norm": 0.3730584979057312, "learning_rate": 9.631467044649185e-05, "loss": 1.8932, "step": 738 }, { "epoch": 0.2243132493549856, "grad_norm": 0.351421594619751, "learning_rate": 9.630960818062164e-05, "loss": 2.3182, "step": 739 }, { "epoch": 0.22461678555167705, "grad_norm": 0.4237976670265198, "learning_rate": 9.630454591475144e-05, "loss": 2.1315, "step": 740 }, { "epoch": 0.22492032174836848, "grad_norm": 0.38544562458992004, "learning_rate": 9.629948364888123e-05, "loss": 1.9596, "step": 741 }, { "epoch": 0.22522385794505995, "grad_norm": 0.407672256231308, "learning_rate": 9.629442138301104e-05, "loss": 1.8694, "step": 742 }, { "epoch": 0.2255273941417514, "grad_norm": 0.4415782690048218, "learning_rate": 9.628935911714084e-05, "loss": 1.8658, "step": 743 }, { "epoch": 0.22583093033844287, "grad_norm": 0.41300657391548157, "learning_rate": 9.628429685127063e-05, "loss": 2.0477, "step": 744 }, { "epoch": 0.2261344665351343, "grad_norm": 0.36000654101371765, "learning_rate": 9.627923458540044e-05, "loss": 1.9045, "step": 745 }, { "epoch": 0.22643800273182577, "grad_norm": 0.42653003334999084, "learning_rate": 9.627417231953023e-05, "loss": 1.2151, "step": 746 }, { "epoch": 0.22674153892851723, "grad_norm": 0.4157649874687195, "learning_rate": 9.626911005366003e-05, "loss": 1.9335, "step": 747 }, { "epoch": 0.2270450751252087, "grad_norm": 0.3805077373981476, "learning_rate": 9.626404778778982e-05, "loss": 2.0803, "step": 748 }, { "epoch": 0.22734861132190012, "grad_norm": 0.39710867404937744, "learning_rate": 9.625898552191962e-05, "loss": 2.2628, "step": 749 }, { "epoch": 0.22765214751859159, "grad_norm": 0.4012609124183655, "learning_rate": 9.625392325604941e-05, "loss": 1.9586, "step": 750 }, { "epoch": 0.22795568371528305, "grad_norm": 0.9281008243560791, "learning_rate": 9.624886099017921e-05, "loss": 1.168, "step": 751 }, { "epoch": 0.2282592199119745, "grad_norm": 0.36847764253616333, "learning_rate": 9.6243798724309e-05, "loss": 1.8907, "step": 752 }, { "epoch": 0.22856275610866597, "grad_norm": 0.4531751573085785, "learning_rate": 9.623873645843881e-05, "loss": 1.4511, "step": 753 }, { "epoch": 0.2288662923053574, "grad_norm": 0.36623820662498474, "learning_rate": 9.62336741925686e-05, "loss": 1.6707, "step": 754 }, { "epoch": 0.22916982850204887, "grad_norm": 0.3104342222213745, "learning_rate": 9.62286119266984e-05, "loss": 1.988, "step": 755 }, { "epoch": 0.22947336469874033, "grad_norm": 0.3790084421634674, "learning_rate": 9.62235496608282e-05, "loss": 1.979, "step": 756 }, { "epoch": 0.2297769008954318, "grad_norm": 0.3642970323562622, "learning_rate": 9.621848739495799e-05, "loss": 1.9998, "step": 757 }, { "epoch": 0.23008043709212322, "grad_norm": 0.34588292241096497, "learning_rate": 9.621342512908779e-05, "loss": 2.0511, "step": 758 }, { "epoch": 0.2303839732888147, "grad_norm": 0.3556496798992157, "learning_rate": 9.620836286321758e-05, "loss": 1.8785, "step": 759 }, { "epoch": 0.23068750948550615, "grad_norm": 0.4669034779071808, "learning_rate": 9.620330059734737e-05, "loss": 1.5027, "step": 760 }, { "epoch": 0.2309910456821976, "grad_norm": 0.39685994386672974, "learning_rate": 9.619823833147717e-05, "loss": 2.1644, "step": 761 }, { "epoch": 0.23129458187888904, "grad_norm": 0.39183005690574646, "learning_rate": 9.619317606560698e-05, "loss": 1.9615, "step": 762 }, { "epoch": 0.2315981180755805, "grad_norm": 0.36401331424713135, "learning_rate": 9.618811379973677e-05, "loss": 1.7535, "step": 763 }, { "epoch": 0.23190165427227197, "grad_norm": 0.43118295073509216, "learning_rate": 9.618305153386657e-05, "loss": 1.884, "step": 764 }, { "epoch": 0.23220519046896343, "grad_norm": 0.5061665177345276, "learning_rate": 9.617798926799636e-05, "loss": 2.0051, "step": 765 }, { "epoch": 0.2325087266656549, "grad_norm": 0.4487472474575043, "learning_rate": 9.617292700212616e-05, "loss": 1.6831, "step": 766 }, { "epoch": 0.23281226286234633, "grad_norm": 0.3660997450351715, "learning_rate": 9.616786473625595e-05, "loss": 1.9276, "step": 767 }, { "epoch": 0.2331157990590378, "grad_norm": 0.3823026716709137, "learning_rate": 9.616280247038575e-05, "loss": 1.9817, "step": 768 }, { "epoch": 0.23341933525572925, "grad_norm": 0.32568395137786865, "learning_rate": 9.615774020451554e-05, "loss": 1.508, "step": 769 }, { "epoch": 0.2337228714524207, "grad_norm": 0.34985265135765076, "learning_rate": 9.615267793864534e-05, "loss": 1.6793, "step": 770 }, { "epoch": 0.23402640764911214, "grad_norm": 0.38563957810401917, "learning_rate": 9.614761567277513e-05, "loss": 1.588, "step": 771 }, { "epoch": 0.2343299438458036, "grad_norm": 0.33572301268577576, "learning_rate": 9.614255340690494e-05, "loss": 1.9541, "step": 772 }, { "epoch": 0.23463348004249507, "grad_norm": 0.33936449885368347, "learning_rate": 9.613749114103473e-05, "loss": 1.9311, "step": 773 }, { "epoch": 0.23493701623918653, "grad_norm": 0.34984657168388367, "learning_rate": 9.613242887516453e-05, "loss": 1.9532, "step": 774 }, { "epoch": 0.235240552435878, "grad_norm": 0.3651373088359833, "learning_rate": 9.612736660929432e-05, "loss": 1.8815, "step": 775 }, { "epoch": 0.23554408863256943, "grad_norm": 0.4317852854728699, "learning_rate": 9.612230434342412e-05, "loss": 2.0262, "step": 776 }, { "epoch": 0.2358476248292609, "grad_norm": 0.375522255897522, "learning_rate": 9.611724207755391e-05, "loss": 1.9964, "step": 777 }, { "epoch": 0.23615116102595235, "grad_norm": 0.37290844321250916, "learning_rate": 9.611217981168371e-05, "loss": 1.7456, "step": 778 }, { "epoch": 0.2364546972226438, "grad_norm": 0.3768545985221863, "learning_rate": 9.61071175458135e-05, "loss": 1.9591, "step": 779 }, { "epoch": 0.23675823341933525, "grad_norm": 0.3147246837615967, "learning_rate": 9.61020552799433e-05, "loss": 1.4033, "step": 780 }, { "epoch": 0.2370617696160267, "grad_norm": 0.4480874240398407, "learning_rate": 9.60969930140731e-05, "loss": 1.9598, "step": 781 }, { "epoch": 0.23736530581271817, "grad_norm": 0.7287562489509583, "learning_rate": 9.60919307482029e-05, "loss": 2.0097, "step": 782 }, { "epoch": 0.23766884200940963, "grad_norm": 0.36199334263801575, "learning_rate": 9.60868684823327e-05, "loss": 1.8089, "step": 783 }, { "epoch": 0.23797237820610107, "grad_norm": 0.32855263352394104, "learning_rate": 9.608180621646249e-05, "loss": 2.0199, "step": 784 }, { "epoch": 0.23827591440279253, "grad_norm": 0.37182894349098206, "learning_rate": 9.607674395059229e-05, "loss": 1.7253, "step": 785 }, { "epoch": 0.238579450599484, "grad_norm": 0.3365595042705536, "learning_rate": 9.607168168472208e-05, "loss": 1.9308, "step": 786 }, { "epoch": 0.23888298679617545, "grad_norm": 0.400685578584671, "learning_rate": 9.606661941885187e-05, "loss": 1.8939, "step": 787 }, { "epoch": 0.2391865229928669, "grad_norm": 0.6354159116744995, "learning_rate": 9.606155715298167e-05, "loss": 2.1476, "step": 788 }, { "epoch": 0.23949005918955835, "grad_norm": 0.4196738600730896, "learning_rate": 9.605649488711148e-05, "loss": 1.8457, "step": 789 }, { "epoch": 0.2397935953862498, "grad_norm": 0.35839545726776123, "learning_rate": 9.605143262124127e-05, "loss": 1.824, "step": 790 }, { "epoch": 0.24009713158294127, "grad_norm": 0.3597940504550934, "learning_rate": 9.604637035537107e-05, "loss": 1.9583, "step": 791 }, { "epoch": 0.24040066777963273, "grad_norm": 0.5783160924911499, "learning_rate": 9.604130808950088e-05, "loss": 2.2, "step": 792 }, { "epoch": 0.24070420397632417, "grad_norm": 0.3544808030128479, "learning_rate": 9.603624582363067e-05, "loss": 2.1092, "step": 793 }, { "epoch": 0.24100774017301563, "grad_norm": 0.41170623898506165, "learning_rate": 9.603118355776047e-05, "loss": 1.6004, "step": 794 }, { "epoch": 0.2413112763697071, "grad_norm": 0.3832992613315582, "learning_rate": 9.602612129189026e-05, "loss": 1.4981, "step": 795 }, { "epoch": 0.24161481256639855, "grad_norm": 0.5239993333816528, "learning_rate": 9.602105902602006e-05, "loss": 1.6026, "step": 796 }, { "epoch": 0.24191834876308999, "grad_norm": 0.38445138931274414, "learning_rate": 9.601599676014985e-05, "loss": 1.5765, "step": 797 }, { "epoch": 0.24222188495978145, "grad_norm": 0.38520511984825134, "learning_rate": 9.601093449427964e-05, "loss": 2.1069, "step": 798 }, { "epoch": 0.2425254211564729, "grad_norm": 0.3519560694694519, "learning_rate": 9.600587222840944e-05, "loss": 1.8896, "step": 799 }, { "epoch": 0.24282895735316437, "grad_norm": 0.5392457246780396, "learning_rate": 9.600080996253923e-05, "loss": 1.6273, "step": 800 }, { "epoch": 0.24313249354985583, "grad_norm": 0.4213111996650696, "learning_rate": 9.599574769666904e-05, "loss": 1.489, "step": 801 }, { "epoch": 0.24343602974654727, "grad_norm": 0.4006531834602356, "learning_rate": 9.599068543079884e-05, "loss": 1.9842, "step": 802 }, { "epoch": 0.24373956594323873, "grad_norm": 0.3792324364185333, "learning_rate": 9.598562316492863e-05, "loss": 1.727, "step": 803 }, { "epoch": 0.2440431021399302, "grad_norm": 0.3555270731449127, "learning_rate": 9.598056089905843e-05, "loss": 1.68, "step": 804 }, { "epoch": 0.24434663833662165, "grad_norm": 0.33837342262268066, "learning_rate": 9.597549863318822e-05, "loss": 2.0709, "step": 805 }, { "epoch": 0.2446501745333131, "grad_norm": 0.3812510371208191, "learning_rate": 9.597043636731802e-05, "loss": 2.1211, "step": 806 }, { "epoch": 0.24495371073000455, "grad_norm": 0.33870792388916016, "learning_rate": 9.596537410144781e-05, "loss": 2.1047, "step": 807 }, { "epoch": 0.245257246926696, "grad_norm": 0.3948252201080322, "learning_rate": 9.59603118355776e-05, "loss": 1.7553, "step": 808 }, { "epoch": 0.24556078312338747, "grad_norm": 0.39410725235939026, "learning_rate": 9.59552495697074e-05, "loss": 1.9383, "step": 809 }, { "epoch": 0.2458643193200789, "grad_norm": 0.37794989347457886, "learning_rate": 9.59501873038372e-05, "loss": 1.9115, "step": 810 }, { "epoch": 0.24616785551677037, "grad_norm": 1.6270610094070435, "learning_rate": 9.5945125037967e-05, "loss": 1.8472, "step": 811 }, { "epoch": 0.24647139171346183, "grad_norm": 0.3724587559700012, "learning_rate": 9.59400627720968e-05, "loss": 1.9087, "step": 812 }, { "epoch": 0.2467749279101533, "grad_norm": 0.4097403585910797, "learning_rate": 9.59350005062266e-05, "loss": 1.8325, "step": 813 }, { "epoch": 0.24707846410684475, "grad_norm": 0.4052940905094147, "learning_rate": 9.592993824035639e-05, "loss": 2.0241, "step": 814 }, { "epoch": 0.2473820003035362, "grad_norm": 0.3887682557106018, "learning_rate": 9.592487597448618e-05, "loss": 1.6114, "step": 815 }, { "epoch": 0.24768553650022765, "grad_norm": 0.404450386762619, "learning_rate": 9.591981370861598e-05, "loss": 1.8384, "step": 816 }, { "epoch": 0.2479890726969191, "grad_norm": 0.7955893874168396, "learning_rate": 9.591475144274577e-05, "loss": 2.2149, "step": 817 }, { "epoch": 0.24829260889361057, "grad_norm": 4.355859279632568, "learning_rate": 9.590968917687557e-05, "loss": 2.3753, "step": 818 }, { "epoch": 0.248596145090302, "grad_norm": 0.3698444962501526, "learning_rate": 9.590462691100536e-05, "loss": 1.7354, "step": 819 }, { "epoch": 0.24889968128699347, "grad_norm": 0.3658899962902069, "learning_rate": 9.589956464513517e-05, "loss": 1.7803, "step": 820 }, { "epoch": 0.24920321748368493, "grad_norm": 0.405072957277298, "learning_rate": 9.589450237926497e-05, "loss": 1.7684, "step": 821 }, { "epoch": 0.2495067536803764, "grad_norm": 0.7590973973274231, "learning_rate": 9.588944011339476e-05, "loss": 1.9466, "step": 822 }, { "epoch": 0.24981028987706785, "grad_norm": 0.5217581987380981, "learning_rate": 9.588437784752456e-05, "loss": 2.1281, "step": 823 }, { "epoch": 0.2501138260737593, "grad_norm": 0.3716435134410858, "learning_rate": 9.587931558165435e-05, "loss": 2.114, "step": 824 }, { "epoch": 0.25041736227045075, "grad_norm": 0.44017624855041504, "learning_rate": 9.587425331578414e-05, "loss": 2.0445, "step": 825 }, { "epoch": 0.2507208984671422, "grad_norm": 0.370370090007782, "learning_rate": 9.586919104991394e-05, "loss": 1.8674, "step": 826 }, { "epoch": 0.2510244346638337, "grad_norm": 0.32125499844551086, "learning_rate": 9.586412878404373e-05, "loss": 1.4129, "step": 827 }, { "epoch": 0.2513279708605251, "grad_norm": 0.4143073856830597, "learning_rate": 9.585906651817353e-05, "loss": 1.9895, "step": 828 }, { "epoch": 0.2516315070572166, "grad_norm": 0.3492576777935028, "learning_rate": 9.585400425230334e-05, "loss": 2.0669, "step": 829 }, { "epoch": 0.25193504325390803, "grad_norm": 0.4044751524925232, "learning_rate": 9.584894198643313e-05, "loss": 1.5909, "step": 830 }, { "epoch": 0.25223857945059946, "grad_norm": 0.3410158157348633, "learning_rate": 9.584387972056293e-05, "loss": 1.7485, "step": 831 }, { "epoch": 0.25254211564729095, "grad_norm": 0.340320348739624, "learning_rate": 9.583881745469272e-05, "loss": 1.8897, "step": 832 }, { "epoch": 0.2528456518439824, "grad_norm": 0.35516276955604553, "learning_rate": 9.583375518882252e-05, "loss": 1.6332, "step": 833 }, { "epoch": 0.2531491880406738, "grad_norm": 0.4099842309951782, "learning_rate": 9.582869292295232e-05, "loss": 1.5617, "step": 834 }, { "epoch": 0.2534527242373653, "grad_norm": 0.38086098432540894, "learning_rate": 9.582363065708212e-05, "loss": 2.0837, "step": 835 }, { "epoch": 0.25375626043405675, "grad_norm": 0.8040663003921509, "learning_rate": 9.581856839121191e-05, "loss": 1.8587, "step": 836 }, { "epoch": 0.25405979663074824, "grad_norm": 0.41297683119773865, "learning_rate": 9.581350612534171e-05, "loss": 1.9602, "step": 837 }, { "epoch": 0.25436333282743967, "grad_norm": 0.38155442476272583, "learning_rate": 9.58084438594715e-05, "loss": 1.375, "step": 838 }, { "epoch": 0.2546668690241311, "grad_norm": 0.3956829905509949, "learning_rate": 9.58033815936013e-05, "loss": 1.9617, "step": 839 }, { "epoch": 0.2549704052208226, "grad_norm": 0.38675928115844727, "learning_rate": 9.579831932773111e-05, "loss": 1.8186, "step": 840 }, { "epoch": 0.255273941417514, "grad_norm": 0.33989018201828003, "learning_rate": 9.57932570618609e-05, "loss": 2.1734, "step": 841 }, { "epoch": 0.2555774776142055, "grad_norm": 0.3240448534488678, "learning_rate": 9.57881947959907e-05, "loss": 1.6238, "step": 842 }, { "epoch": 0.25588101381089695, "grad_norm": 0.6117075681686401, "learning_rate": 9.578313253012049e-05, "loss": 1.986, "step": 843 }, { "epoch": 0.2561845500075884, "grad_norm": 0.3781290650367737, "learning_rate": 9.577807026425029e-05, "loss": 2.0021, "step": 844 }, { "epoch": 0.2564880862042799, "grad_norm": 0.4373374879360199, "learning_rate": 9.577300799838008e-05, "loss": 2.0195, "step": 845 }, { "epoch": 0.2567916224009713, "grad_norm": 0.4125923216342926, "learning_rate": 9.576794573250988e-05, "loss": 1.9412, "step": 846 }, { "epoch": 0.2570951585976628, "grad_norm": 0.3557007908821106, "learning_rate": 9.576288346663967e-05, "loss": 1.8098, "step": 847 }, { "epoch": 0.25739869479435423, "grad_norm": 0.49475541710853577, "learning_rate": 9.575782120076947e-05, "loss": 1.5756, "step": 848 }, { "epoch": 0.25770223099104567, "grad_norm": 0.3507518768310547, "learning_rate": 9.575275893489926e-05, "loss": 1.6413, "step": 849 }, { "epoch": 0.25800576718773716, "grad_norm": 0.39508333802223206, "learning_rate": 9.574769666902907e-05, "loss": 1.9777, "step": 850 }, { "epoch": 0.2583093033844286, "grad_norm": 0.328807532787323, "learning_rate": 9.574263440315886e-05, "loss": 1.4948, "step": 851 }, { "epoch": 0.25861283958112, "grad_norm": 0.3154551386833191, "learning_rate": 9.573757213728866e-05, "loss": 1.7809, "step": 852 }, { "epoch": 0.2589163757778115, "grad_norm": 0.502554178237915, "learning_rate": 9.573250987141845e-05, "loss": 1.4369, "step": 853 }, { "epoch": 0.25921991197450295, "grad_norm": 0.4416670799255371, "learning_rate": 9.572744760554825e-05, "loss": 1.7364, "step": 854 }, { "epoch": 0.25952344817119444, "grad_norm": 0.43228060007095337, "learning_rate": 9.572238533967804e-05, "loss": 1.3281, "step": 855 }, { "epoch": 0.25982698436788587, "grad_norm": 0.3714723289012909, "learning_rate": 9.571732307380784e-05, "loss": 2.0893, "step": 856 }, { "epoch": 0.2601305205645773, "grad_norm": 0.3309679925441742, "learning_rate": 9.571226080793763e-05, "loss": 1.7982, "step": 857 }, { "epoch": 0.2604340567612688, "grad_norm": 0.3709767460823059, "learning_rate": 9.570719854206743e-05, "loss": 1.8628, "step": 858 }, { "epoch": 0.26073759295796023, "grad_norm": 0.6020816564559937, "learning_rate": 9.570213627619724e-05, "loss": 2.0077, "step": 859 }, { "epoch": 0.2610411291546517, "grad_norm": 0.30620431900024414, "learning_rate": 9.569707401032703e-05, "loss": 1.8834, "step": 860 }, { "epoch": 0.26134466535134315, "grad_norm": 0.41518962383270264, "learning_rate": 9.569201174445683e-05, "loss": 1.8025, "step": 861 }, { "epoch": 0.2616482015480346, "grad_norm": 0.3919786512851715, "learning_rate": 9.568694947858662e-05, "loss": 1.995, "step": 862 }, { "epoch": 0.2619517377447261, "grad_norm": 0.47429168224334717, "learning_rate": 9.568188721271641e-05, "loss": 1.9423, "step": 863 }, { "epoch": 0.2622552739414175, "grad_norm": 0.8941421508789062, "learning_rate": 9.567682494684621e-05, "loss": 1.5046, "step": 864 }, { "epoch": 0.26255881013810894, "grad_norm": 0.4357859194278717, "learning_rate": 9.5671762680976e-05, "loss": 2.0023, "step": 865 }, { "epoch": 0.26286234633480043, "grad_norm": 0.3873944878578186, "learning_rate": 9.56667004151058e-05, "loss": 2.0607, "step": 866 }, { "epoch": 0.26316588253149187, "grad_norm": 0.4355853497982025, "learning_rate": 9.56616381492356e-05, "loss": 1.8254, "step": 867 }, { "epoch": 0.26346941872818336, "grad_norm": 0.3882213234901428, "learning_rate": 9.56565758833654e-05, "loss": 1.7809, "step": 868 }, { "epoch": 0.2637729549248748, "grad_norm": 0.4021656811237335, "learning_rate": 9.56515136174952e-05, "loss": 2.1321, "step": 869 }, { "epoch": 0.2640764911215662, "grad_norm": 0.43587526679039, "learning_rate": 9.564645135162499e-05, "loss": 1.7865, "step": 870 }, { "epoch": 0.2643800273182577, "grad_norm": 0.364045649766922, "learning_rate": 9.564138908575479e-05, "loss": 1.8173, "step": 871 }, { "epoch": 0.26468356351494915, "grad_norm": 0.3956625461578369, "learning_rate": 9.563632681988458e-05, "loss": 1.4822, "step": 872 }, { "epoch": 0.26498709971164064, "grad_norm": 0.40755051374435425, "learning_rate": 9.563126455401438e-05, "loss": 1.9418, "step": 873 }, { "epoch": 0.2652906359083321, "grad_norm": 0.39405086636543274, "learning_rate": 9.562620228814417e-05, "loss": 1.4529, "step": 874 }, { "epoch": 0.2655941721050235, "grad_norm": 0.4400351047515869, "learning_rate": 9.562114002227397e-05, "loss": 2.1095, "step": 875 }, { "epoch": 0.265897708301715, "grad_norm": 0.40135496854782104, "learning_rate": 9.561607775640376e-05, "loss": 1.9462, "step": 876 }, { "epoch": 0.26620124449840643, "grad_norm": 0.5949604511260986, "learning_rate": 9.561101549053356e-05, "loss": 1.8797, "step": 877 }, { "epoch": 0.26650478069509786, "grad_norm": 0.38301005959510803, "learning_rate": 9.560595322466336e-05, "loss": 2.0887, "step": 878 }, { "epoch": 0.26680831689178935, "grad_norm": 0.6215627789497375, "learning_rate": 9.560089095879317e-05, "loss": 1.7846, "step": 879 }, { "epoch": 0.2671118530884808, "grad_norm": 0.4041058123111725, "learning_rate": 9.559582869292297e-05, "loss": 1.5127, "step": 880 }, { "epoch": 0.2674153892851723, "grad_norm": 0.30281975865364075, "learning_rate": 9.559076642705276e-05, "loss": 1.8487, "step": 881 }, { "epoch": 0.2677189254818637, "grad_norm": 0.34536200761795044, "learning_rate": 9.558570416118256e-05, "loss": 1.8976, "step": 882 }, { "epoch": 0.26802246167855515, "grad_norm": 0.367245614528656, "learning_rate": 9.558064189531235e-05, "loss": 1.9804, "step": 883 }, { "epoch": 0.26832599787524664, "grad_norm": 0.41750359535217285, "learning_rate": 9.557557962944215e-05, "loss": 1.5932, "step": 884 }, { "epoch": 0.26862953407193807, "grad_norm": 0.7777047157287598, "learning_rate": 9.557051736357194e-05, "loss": 1.8513, "step": 885 }, { "epoch": 0.26893307026862956, "grad_norm": 0.3720252215862274, "learning_rate": 9.556545509770174e-05, "loss": 2.1819, "step": 886 }, { "epoch": 0.269236606465321, "grad_norm": 0.7321712970733643, "learning_rate": 9.556039283183153e-05, "loss": 1.4653, "step": 887 }, { "epoch": 0.2695401426620124, "grad_norm": 0.4140429198741913, "learning_rate": 9.555533056596133e-05, "loss": 1.9816, "step": 888 }, { "epoch": 0.2698436788587039, "grad_norm": 0.40684935450553894, "learning_rate": 9.555026830009113e-05, "loss": 1.5866, "step": 889 }, { "epoch": 0.27014721505539535, "grad_norm": 0.4067225754261017, "learning_rate": 9.554520603422093e-05, "loss": 1.5951, "step": 890 }, { "epoch": 0.2704507512520868, "grad_norm": 0.34240391850471497, "learning_rate": 9.554014376835072e-05, "loss": 1.9076, "step": 891 }, { "epoch": 0.2707542874487783, "grad_norm": 0.4634522795677185, "learning_rate": 9.553508150248052e-05, "loss": 1.9856, "step": 892 }, { "epoch": 0.2710578236454697, "grad_norm": 0.408015638589859, "learning_rate": 9.553001923661031e-05, "loss": 1.7997, "step": 893 }, { "epoch": 0.2713613598421612, "grad_norm": 0.3894648253917694, "learning_rate": 9.552495697074011e-05, "loss": 1.8381, "step": 894 }, { "epoch": 0.27166489603885263, "grad_norm": 0.37494730949401855, "learning_rate": 9.55198947048699e-05, "loss": 2.0548, "step": 895 }, { "epoch": 0.27196843223554407, "grad_norm": 0.39796411991119385, "learning_rate": 9.55148324389997e-05, "loss": 1.9272, "step": 896 }, { "epoch": 0.27227196843223556, "grad_norm": 0.40153494477272034, "learning_rate": 9.550977017312949e-05, "loss": 1.7136, "step": 897 }, { "epoch": 0.272575504628927, "grad_norm": 0.39771386981010437, "learning_rate": 9.55047079072593e-05, "loss": 2.1017, "step": 898 }, { "epoch": 0.2728790408256185, "grad_norm": 0.4085974097251892, "learning_rate": 9.54996456413891e-05, "loss": 1.3951, "step": 899 }, { "epoch": 0.2731825770223099, "grad_norm": 0.39849239587783813, "learning_rate": 9.549458337551889e-05, "loss": 1.9988, "step": 900 }, { "epoch": 0.27348611321900135, "grad_norm": 0.38662001490592957, "learning_rate": 9.548952110964868e-05, "loss": 1.8491, "step": 901 }, { "epoch": 0.27378964941569284, "grad_norm": 0.38078710436820984, "learning_rate": 9.548445884377848e-05, "loss": 1.9, "step": 902 }, { "epoch": 0.27409318561238427, "grad_norm": 0.3548724949359894, "learning_rate": 9.547939657790827e-05, "loss": 1.8754, "step": 903 }, { "epoch": 0.2743967218090757, "grad_norm": 0.37712323665618896, "learning_rate": 9.547433431203807e-05, "loss": 1.5497, "step": 904 }, { "epoch": 0.2747002580057672, "grad_norm": 0.4060449004173279, "learning_rate": 9.546927204616786e-05, "loss": 1.7231, "step": 905 }, { "epoch": 0.27500379420245863, "grad_norm": 0.42080479860305786, "learning_rate": 9.546420978029766e-05, "loss": 2.1538, "step": 906 }, { "epoch": 0.2753073303991501, "grad_norm": 0.4034046232700348, "learning_rate": 9.545914751442747e-05, "loss": 1.7335, "step": 907 }, { "epoch": 0.27561086659584155, "grad_norm": 0.3676345646381378, "learning_rate": 9.545408524855726e-05, "loss": 1.6193, "step": 908 }, { "epoch": 0.275914402792533, "grad_norm": 0.3349851965904236, "learning_rate": 9.544902298268706e-05, "loss": 1.8997, "step": 909 }, { "epoch": 0.2762179389892245, "grad_norm": 0.3676302134990692, "learning_rate": 9.544396071681685e-05, "loss": 1.4031, "step": 910 }, { "epoch": 0.2765214751859159, "grad_norm": 0.36593666672706604, "learning_rate": 9.543889845094665e-05, "loss": 1.8838, "step": 911 }, { "epoch": 0.2768250113826074, "grad_norm": 0.3793712258338928, "learning_rate": 9.543383618507644e-05, "loss": 1.5949, "step": 912 }, { "epoch": 0.27712854757929883, "grad_norm": 0.47586631774902344, "learning_rate": 9.542877391920624e-05, "loss": 1.5687, "step": 913 }, { "epoch": 0.27743208377599027, "grad_norm": 0.38850024342536926, "learning_rate": 9.542371165333603e-05, "loss": 1.7336, "step": 914 }, { "epoch": 0.27773561997268176, "grad_norm": 0.4039680063724518, "learning_rate": 9.541864938746583e-05, "loss": 2.0476, "step": 915 }, { "epoch": 0.2780391561693732, "grad_norm": 0.40498992800712585, "learning_rate": 9.541358712159562e-05, "loss": 1.6699, "step": 916 }, { "epoch": 0.2783426923660646, "grad_norm": 0.39011168479919434, "learning_rate": 9.540852485572543e-05, "loss": 1.9935, "step": 917 }, { "epoch": 0.2786462285627561, "grad_norm": 0.3864549696445465, "learning_rate": 9.540346258985522e-05, "loss": 1.8271, "step": 918 }, { "epoch": 0.27894976475944755, "grad_norm": 0.33493247628211975, "learning_rate": 9.539840032398502e-05, "loss": 1.856, "step": 919 }, { "epoch": 0.27925330095613904, "grad_norm": 0.34132060408592224, "learning_rate": 9.539333805811481e-05, "loss": 1.8836, "step": 920 }, { "epoch": 0.2795568371528305, "grad_norm": 1.5312176942825317, "learning_rate": 9.538827579224461e-05, "loss": 2.0207, "step": 921 }, { "epoch": 0.2798603733495219, "grad_norm": 0.333932489156723, "learning_rate": 9.53832135263744e-05, "loss": 2.0908, "step": 922 }, { "epoch": 0.2801639095462134, "grad_norm": 0.3688269555568695, "learning_rate": 9.537815126050421e-05, "loss": 1.8464, "step": 923 }, { "epoch": 0.28046744574290483, "grad_norm": 0.4097294211387634, "learning_rate": 9.5373088994634e-05, "loss": 1.6891, "step": 924 }, { "epoch": 0.2807709819395963, "grad_norm": 0.3737453818321228, "learning_rate": 9.53680267287638e-05, "loss": 2.0549, "step": 925 }, { "epoch": 0.28107451813628775, "grad_norm": 0.6109428405761719, "learning_rate": 9.53629644628936e-05, "loss": 1.9437, "step": 926 }, { "epoch": 0.2813780543329792, "grad_norm": 0.46215322613716125, "learning_rate": 9.535790219702339e-05, "loss": 1.5133, "step": 927 }, { "epoch": 0.2816815905296707, "grad_norm": 0.8070108294487, "learning_rate": 9.53528399311532e-05, "loss": 1.8843, "step": 928 }, { "epoch": 0.2819851267263621, "grad_norm": 0.40304142236709595, "learning_rate": 9.534777766528299e-05, "loss": 1.9742, "step": 929 }, { "epoch": 0.2822886629230536, "grad_norm": 0.35046708583831787, "learning_rate": 9.534271539941279e-05, "loss": 1.8969, "step": 930 }, { "epoch": 0.28259219911974504, "grad_norm": 0.37241777777671814, "learning_rate": 9.533765313354258e-05, "loss": 1.8138, "step": 931 }, { "epoch": 0.28289573531643647, "grad_norm": 0.38689473271369934, "learning_rate": 9.533259086767238e-05, "loss": 1.669, "step": 932 }, { "epoch": 0.28319927151312796, "grad_norm": 0.3672066926956177, "learning_rate": 9.532752860180217e-05, "loss": 1.9093, "step": 933 }, { "epoch": 0.2835028077098194, "grad_norm": 0.4022217392921448, "learning_rate": 9.532246633593197e-05, "loss": 1.6959, "step": 934 }, { "epoch": 0.2838063439065108, "grad_norm": 0.3894721269607544, "learning_rate": 9.531740407006176e-05, "loss": 1.9898, "step": 935 }, { "epoch": 0.2841098801032023, "grad_norm": 0.4395015835762024, "learning_rate": 9.531234180419156e-05, "loss": 1.5538, "step": 936 }, { "epoch": 0.28441341629989375, "grad_norm": 0.8121886849403381, "learning_rate": 9.530727953832136e-05, "loss": 1.7403, "step": 937 }, { "epoch": 0.28471695249658524, "grad_norm": 0.40073227882385254, "learning_rate": 9.530221727245116e-05, "loss": 2.0544, "step": 938 }, { "epoch": 0.2850204886932767, "grad_norm": 0.3571331202983856, "learning_rate": 9.529715500658095e-05, "loss": 1.7157, "step": 939 }, { "epoch": 0.2853240248899681, "grad_norm": 0.485147625207901, "learning_rate": 9.529209274071075e-05, "loss": 2.1489, "step": 940 }, { "epoch": 0.2856275610866596, "grad_norm": 0.6882160305976868, "learning_rate": 9.528703047484054e-05, "loss": 1.8458, "step": 941 }, { "epoch": 0.28593109728335103, "grad_norm": 0.7156968116760254, "learning_rate": 9.528196820897034e-05, "loss": 1.9529, "step": 942 }, { "epoch": 0.2862346334800425, "grad_norm": 0.4198112487792969, "learning_rate": 9.527690594310013e-05, "loss": 2.0355, "step": 943 }, { "epoch": 0.28653816967673396, "grad_norm": 0.4178343117237091, "learning_rate": 9.527184367722993e-05, "loss": 1.5801, "step": 944 }, { "epoch": 0.2868417058734254, "grad_norm": 0.3721866011619568, "learning_rate": 9.526678141135972e-05, "loss": 2.1657, "step": 945 }, { "epoch": 0.2871452420701169, "grad_norm": 0.38586944341659546, "learning_rate": 9.526171914548953e-05, "loss": 1.4879, "step": 946 }, { "epoch": 0.2874487782668083, "grad_norm": 0.42727598547935486, "learning_rate": 9.525665687961933e-05, "loss": 1.8434, "step": 947 }, { "epoch": 0.28775231446349975, "grad_norm": 0.3686284124851227, "learning_rate": 9.525159461374912e-05, "loss": 1.9346, "step": 948 }, { "epoch": 0.28805585066019124, "grad_norm": 0.41984260082244873, "learning_rate": 9.524653234787892e-05, "loss": 1.4474, "step": 949 }, { "epoch": 0.28835938685688267, "grad_norm": 0.4530123174190521, "learning_rate": 9.524147008200871e-05, "loss": 1.6863, "step": 950 }, { "epoch": 0.28866292305357416, "grad_norm": 0.40047594904899597, "learning_rate": 9.52364078161385e-05, "loss": 1.908, "step": 951 }, { "epoch": 0.2889664592502656, "grad_norm": 0.3757762610912323, "learning_rate": 9.52313455502683e-05, "loss": 1.6235, "step": 952 }, { "epoch": 0.28926999544695703, "grad_norm": 0.4337126612663269, "learning_rate": 9.52262832843981e-05, "loss": 1.6229, "step": 953 }, { "epoch": 0.2895735316436485, "grad_norm": 0.4407886564731598, "learning_rate": 9.522122101852789e-05, "loss": 1.875, "step": 954 }, { "epoch": 0.28987706784033995, "grad_norm": 0.5278657674789429, "learning_rate": 9.521615875265768e-05, "loss": 1.7199, "step": 955 }, { "epoch": 0.29018060403703144, "grad_norm": 0.4441334307193756, "learning_rate": 9.521109648678749e-05, "loss": 1.1319, "step": 956 }, { "epoch": 0.2904841402337229, "grad_norm": 0.3992663025856018, "learning_rate": 9.520603422091729e-05, "loss": 1.6948, "step": 957 }, { "epoch": 0.2907876764304143, "grad_norm": 0.3979544937610626, "learning_rate": 9.520097195504708e-05, "loss": 1.8689, "step": 958 }, { "epoch": 0.2910912126271058, "grad_norm": 0.4011298418045044, "learning_rate": 9.519590968917688e-05, "loss": 1.9491, "step": 959 }, { "epoch": 0.29139474882379723, "grad_norm": 0.4377354383468628, "learning_rate": 9.519084742330667e-05, "loss": 1.7274, "step": 960 }, { "epoch": 0.29169828502048867, "grad_norm": 0.5056617856025696, "learning_rate": 9.518578515743647e-05, "loss": 2.006, "step": 961 }, { "epoch": 0.29200182121718016, "grad_norm": 0.36736002564430237, "learning_rate": 9.518072289156626e-05, "loss": 1.6558, "step": 962 }, { "epoch": 0.2923053574138716, "grad_norm": 0.37966540455818176, "learning_rate": 9.517566062569606e-05, "loss": 2.0098, "step": 963 }, { "epoch": 0.2926088936105631, "grad_norm": 0.4026505947113037, "learning_rate": 9.517059835982585e-05, "loss": 1.868, "step": 964 }, { "epoch": 0.2929124298072545, "grad_norm": 0.461910218000412, "learning_rate": 9.516553609395566e-05, "loss": 2.1131, "step": 965 }, { "epoch": 0.29321596600394595, "grad_norm": 0.4329175651073456, "learning_rate": 9.516047382808545e-05, "loss": 2.0068, "step": 966 }, { "epoch": 0.29351950220063744, "grad_norm": 0.7611956000328064, "learning_rate": 9.515541156221526e-05, "loss": 1.9177, "step": 967 }, { "epoch": 0.2938230383973289, "grad_norm": 0.6180218458175659, "learning_rate": 9.515034929634506e-05, "loss": 1.5603, "step": 968 }, { "epoch": 0.29412657459402036, "grad_norm": 0.6556726694107056, "learning_rate": 9.514528703047485e-05, "loss": 2.1081, "step": 969 }, { "epoch": 0.2944301107907118, "grad_norm": 0.3379404842853546, "learning_rate": 9.514022476460465e-05, "loss": 1.9701, "step": 970 }, { "epoch": 0.29473364698740323, "grad_norm": 0.42676112055778503, "learning_rate": 9.513516249873444e-05, "loss": 1.6116, "step": 971 }, { "epoch": 0.2950371831840947, "grad_norm": 0.35374894738197327, "learning_rate": 9.513010023286424e-05, "loss": 2.0621, "step": 972 }, { "epoch": 0.29534071938078615, "grad_norm": 0.33012476563453674, "learning_rate": 9.512503796699403e-05, "loss": 1.4534, "step": 973 }, { "epoch": 0.2956442555774776, "grad_norm": 0.37993383407592773, "learning_rate": 9.511997570112383e-05, "loss": 1.6306, "step": 974 }, { "epoch": 0.2959477917741691, "grad_norm": 0.47140204906463623, "learning_rate": 9.511491343525362e-05, "loss": 2.0465, "step": 975 }, { "epoch": 0.2962513279708605, "grad_norm": 0.40235936641693115, "learning_rate": 9.510985116938343e-05, "loss": 1.8247, "step": 976 }, { "epoch": 0.296554864167552, "grad_norm": 0.3992665112018585, "learning_rate": 9.510478890351322e-05, "loss": 1.5702, "step": 977 }, { "epoch": 0.29685840036424344, "grad_norm": 0.4469521641731262, "learning_rate": 9.509972663764302e-05, "loss": 1.8811, "step": 978 }, { "epoch": 0.29716193656093487, "grad_norm": 0.41400644183158875, "learning_rate": 9.509466437177281e-05, "loss": 1.5374, "step": 979 }, { "epoch": 0.29746547275762636, "grad_norm": 0.36348387598991394, "learning_rate": 9.508960210590261e-05, "loss": 1.9022, "step": 980 }, { "epoch": 0.2977690089543178, "grad_norm": 0.4069242477416992, "learning_rate": 9.50845398400324e-05, "loss": 2.0066, "step": 981 }, { "epoch": 0.2980725451510093, "grad_norm": 0.3684113323688507, "learning_rate": 9.50794775741622e-05, "loss": 1.8972, "step": 982 }, { "epoch": 0.2983760813477007, "grad_norm": 0.40827688574790955, "learning_rate": 9.5074415308292e-05, "loss": 2.0659, "step": 983 }, { "epoch": 0.29867961754439215, "grad_norm": 0.32065409421920776, "learning_rate": 9.506935304242179e-05, "loss": 2.0008, "step": 984 }, { "epoch": 0.29898315374108364, "grad_norm": 0.38805294036865234, "learning_rate": 9.50642907765516e-05, "loss": 1.5027, "step": 985 }, { "epoch": 0.2992866899377751, "grad_norm": 0.3656708896160126, "learning_rate": 9.505922851068139e-05, "loss": 1.7931, "step": 986 }, { "epoch": 0.2995902261344665, "grad_norm": 0.4354289770126343, "learning_rate": 9.505416624481119e-05, "loss": 2.1183, "step": 987 }, { "epoch": 0.299893762331158, "grad_norm": 0.3970641493797302, "learning_rate": 9.504910397894098e-05, "loss": 1.8188, "step": 988 }, { "epoch": 0.30019729852784943, "grad_norm": 0.35527995228767395, "learning_rate": 9.504404171307078e-05, "loss": 1.6329, "step": 989 }, { "epoch": 0.3005008347245409, "grad_norm": 0.4018630385398865, "learning_rate": 9.503897944720057e-05, "loss": 1.993, "step": 990 }, { "epoch": 0.30080437092123236, "grad_norm": 0.36514052748680115, "learning_rate": 9.503391718133037e-05, "loss": 2.0482, "step": 991 }, { "epoch": 0.3011079071179238, "grad_norm": 0.3790993094444275, "learning_rate": 9.502885491546016e-05, "loss": 2.0286, "step": 992 }, { "epoch": 0.3014114433146153, "grad_norm": 0.314779669046402, "learning_rate": 9.502379264958995e-05, "loss": 1.8135, "step": 993 }, { "epoch": 0.3017149795113067, "grad_norm": 0.42383378744125366, "learning_rate": 9.501873038371975e-05, "loss": 1.8783, "step": 994 }, { "epoch": 0.3020185157079982, "grad_norm": 0.4036683738231659, "learning_rate": 9.501366811784956e-05, "loss": 1.6091, "step": 995 }, { "epoch": 0.30232205190468964, "grad_norm": 0.3611324429512024, "learning_rate": 9.500860585197935e-05, "loss": 1.3388, "step": 996 }, { "epoch": 0.30262558810138107, "grad_norm": 0.44210389256477356, "learning_rate": 9.500354358610915e-05, "loss": 1.6133, "step": 997 }, { "epoch": 0.30292912429807256, "grad_norm": 0.37780526280403137, "learning_rate": 9.499848132023894e-05, "loss": 1.9993, "step": 998 }, { "epoch": 0.303232660494764, "grad_norm": 0.469959557056427, "learning_rate": 9.499341905436874e-05, "loss": 1.8094, "step": 999 }, { "epoch": 0.30353619669145543, "grad_norm": 0.38992664217948914, "learning_rate": 9.498835678849853e-05, "loss": 1.8975, "step": 1000 } ], "logging_steps": 1, "max_steps": 19764, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 6.448897543149158e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }