diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,15309 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 2181, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0004585052728106373, + "grad_norm": 61.62693299208047, + "learning_rate": 1.5151515151515153e-05, + "loss": 7.7161, + "step": 1 + }, + { + "epoch": 0.0009170105456212746, + "grad_norm": 67.22710400484628, + "learning_rate": 3.0303030303030306e-05, + "loss": 7.6921, + "step": 2 + }, + { + "epoch": 0.001375515818431912, + "grad_norm": 47.13892446149569, + "learning_rate": 4.545454545454546e-05, + "loss": 7.2591, + "step": 3 + }, + { + "epoch": 0.0018340210912425492, + "grad_norm": 27.681156878509608, + "learning_rate": 6.060606060606061e-05, + "loss": 6.1936, + "step": 4 + }, + { + "epoch": 0.0022925263640531865, + "grad_norm": 17.715692400976526, + "learning_rate": 7.575757575757576e-05, + "loss": 5.615, + "step": 5 + }, + { + "epoch": 0.002751031636863824, + "grad_norm": 11.602911046832345, + "learning_rate": 9.090909090909092e-05, + "loss": 5.1284, + "step": 6 + }, + { + "epoch": 0.0032095369096744614, + "grad_norm": 6.708277018431127, + "learning_rate": 0.00010606060606060606, + "loss": 4.9254, + "step": 7 + }, + { + "epoch": 0.0036680421824850985, + "grad_norm": 5.2361920747082324, + "learning_rate": 0.00012121212121212122, + "loss": 4.6942, + "step": 8 + }, + { + "epoch": 0.0041265474552957355, + "grad_norm": 6.709596755330145, + "learning_rate": 0.00013636363636363637, + "loss": 4.6008, + "step": 9 + }, + { + "epoch": 0.004585052728106373, + "grad_norm": 4.0803834774420205, + "learning_rate": 0.00015151515151515152, + "loss": 4.4809, + "step": 10 + }, + { + "epoch": 0.0050435580009170105, + "grad_norm": 2.99498963682996, + "learning_rate": 0.00016666666666666666, + "loss": 4.3577, + "step": 11 + }, + { + "epoch": 0.005502063273727648, + "grad_norm": 2.927884494375966, + "learning_rate": 0.00018181818181818183, + "loss": 4.5176, + "step": 12 + }, + { + "epoch": 0.005960568546538285, + "grad_norm": 2.3332253023808702, + "learning_rate": 0.00019696969696969695, + "loss": 4.4292, + "step": 13 + }, + { + "epoch": 0.006419073819348923, + "grad_norm": 1.8390453088259915, + "learning_rate": 0.00021212121212121213, + "loss": 4.3496, + "step": 14 + }, + { + "epoch": 0.0068775790921595595, + "grad_norm": 1.8261539233952222, + "learning_rate": 0.00022727272727272727, + "loss": 4.2151, + "step": 15 + }, + { + "epoch": 0.007336084364970197, + "grad_norm": 1.4946942226341278, + "learning_rate": 0.00024242424242424245, + "loss": 4.2084, + "step": 16 + }, + { + "epoch": 0.007794589637780834, + "grad_norm": 1.7068840373654497, + "learning_rate": 0.00025757575757575756, + "loss": 4.1117, + "step": 17 + }, + { + "epoch": 0.008253094910591471, + "grad_norm": 1.2038243601956324, + "learning_rate": 0.00027272727272727274, + "loss": 4.0595, + "step": 18 + }, + { + "epoch": 0.00871160018340211, + "grad_norm": 0.9440289583248761, + "learning_rate": 0.0002878787878787879, + "loss": 4.0782, + "step": 19 + }, + { + "epoch": 0.009170105456212746, + "grad_norm": 0.9591676639820085, + "learning_rate": 0.00030303030303030303, + "loss": 4.214, + "step": 20 + }, + { + "epoch": 0.009628610729023384, + "grad_norm": 0.9189315124959249, + "learning_rate": 0.0003181818181818182, + "loss": 4.1936, + "step": 21 + }, + { + "epoch": 0.010087116001834021, + "grad_norm": 0.8471724633748985, + "learning_rate": 0.0003333333333333333, + "loss": 4.0434, + "step": 22 + }, + { + "epoch": 0.01054562127464466, + "grad_norm": 0.8513040068247689, + "learning_rate": 0.0003484848484848485, + "loss": 3.9915, + "step": 23 + }, + { + "epoch": 0.011004126547455296, + "grad_norm": 0.7322027250491603, + "learning_rate": 0.00036363636363636367, + "loss": 4.0654, + "step": 24 + }, + { + "epoch": 0.011462631820265932, + "grad_norm": 0.6709320894211065, + "learning_rate": 0.0003787878787878788, + "loss": 3.9753, + "step": 25 + }, + { + "epoch": 0.01192113709307657, + "grad_norm": 0.5988018585193845, + "learning_rate": 0.0003939393939393939, + "loss": 3.9206, + "step": 26 + }, + { + "epoch": 0.012379642365887207, + "grad_norm": 0.6858745057647339, + "learning_rate": 0.00040909090909090913, + "loss": 3.9863, + "step": 27 + }, + { + "epoch": 0.012838147638697846, + "grad_norm": 0.6836109745035435, + "learning_rate": 0.00042424242424242425, + "loss": 3.9099, + "step": 28 + }, + { + "epoch": 0.013296652911508482, + "grad_norm": 0.5649698496741163, + "learning_rate": 0.0004393939393939394, + "loss": 4.0176, + "step": 29 + }, + { + "epoch": 0.013755158184319119, + "grad_norm": 0.6149142825302876, + "learning_rate": 0.00045454545454545455, + "loss": 3.9409, + "step": 30 + }, + { + "epoch": 0.014213663457129757, + "grad_norm": 0.6034668930526498, + "learning_rate": 0.0004696969696969697, + "loss": 3.9003, + "step": 31 + }, + { + "epoch": 0.014672168729940394, + "grad_norm": 0.5506688241657357, + "learning_rate": 0.0004848484848484849, + "loss": 3.9581, + "step": 32 + }, + { + "epoch": 0.015130674002751032, + "grad_norm": 0.5552787660080769, + "learning_rate": 0.0005, + "loss": 3.8845, + "step": 33 + }, + { + "epoch": 0.015589179275561669, + "grad_norm": 0.5747253638676144, + "learning_rate": 0.0005151515151515151, + "loss": 3.9434, + "step": 34 + }, + { + "epoch": 0.016047684548372305, + "grad_norm": 0.6703215670417005, + "learning_rate": 0.0005303030303030302, + "loss": 3.8698, + "step": 35 + }, + { + "epoch": 0.016506189821182942, + "grad_norm": 0.5832388338678942, + "learning_rate": 0.0005454545454545455, + "loss": 3.7587, + "step": 36 + }, + { + "epoch": 0.016964695093993582, + "grad_norm": 0.6158042039951228, + "learning_rate": 0.0005606060606060606, + "loss": 3.8296, + "step": 37 + }, + { + "epoch": 0.01742320036680422, + "grad_norm": 0.6249906129430125, + "learning_rate": 0.0005757575757575758, + "loss": 3.8058, + "step": 38 + }, + { + "epoch": 0.017881705639614855, + "grad_norm": 0.6330565181515799, + "learning_rate": 0.0005909090909090909, + "loss": 3.8108, + "step": 39 + }, + { + "epoch": 0.018340210912425492, + "grad_norm": 0.6282744751415308, + "learning_rate": 0.0006060606060606061, + "loss": 3.7822, + "step": 40 + }, + { + "epoch": 0.01879871618523613, + "grad_norm": 0.6126602063231258, + "learning_rate": 0.0006212121212121212, + "loss": 3.671, + "step": 41 + }, + { + "epoch": 0.01925722145804677, + "grad_norm": 0.6540694403713017, + "learning_rate": 0.0006363636363636364, + "loss": 3.6944, + "step": 42 + }, + { + "epoch": 0.019715726730857405, + "grad_norm": 0.6649471298539525, + "learning_rate": 0.0006515151515151515, + "loss": 3.7147, + "step": 43 + }, + { + "epoch": 0.020174232003668042, + "grad_norm": 0.7102165431057467, + "learning_rate": 0.0006666666666666666, + "loss": 3.5843, + "step": 44 + }, + { + "epoch": 0.02063273727647868, + "grad_norm": 0.6271998749526863, + "learning_rate": 0.0006818181818181818, + "loss": 3.5796, + "step": 45 + }, + { + "epoch": 0.02109124254928932, + "grad_norm": 0.8217711473935139, + "learning_rate": 0.000696969696969697, + "loss": 3.6568, + "step": 46 + }, + { + "epoch": 0.021549747822099955, + "grad_norm": 0.7471746988857203, + "learning_rate": 0.0007121212121212122, + "loss": 3.561, + "step": 47 + }, + { + "epoch": 0.02200825309491059, + "grad_norm": 0.7140569659677037, + "learning_rate": 0.0007272727272727273, + "loss": 3.5824, + "step": 48 + }, + { + "epoch": 0.02246675836772123, + "grad_norm": 0.8114261947650983, + "learning_rate": 0.0007424242424242425, + "loss": 3.6559, + "step": 49 + }, + { + "epoch": 0.022925263640531865, + "grad_norm": 0.7338082537452666, + "learning_rate": 0.0007575757575757576, + "loss": 3.5688, + "step": 50 + }, + { + "epoch": 0.023383768913342505, + "grad_norm": 0.7682708582098703, + "learning_rate": 0.0007727272727272727, + "loss": 3.532, + "step": 51 + }, + { + "epoch": 0.02384227418615314, + "grad_norm": 0.721680134655681, + "learning_rate": 0.0007878787878787878, + "loss": 3.5962, + "step": 52 + }, + { + "epoch": 0.024300779458963778, + "grad_norm": 0.804166729472871, + "learning_rate": 0.000803030303030303, + "loss": 3.579, + "step": 53 + }, + { + "epoch": 0.024759284731774415, + "grad_norm": 0.8162388030957528, + "learning_rate": 0.0008181818181818183, + "loss": 3.5037, + "step": 54 + }, + { + "epoch": 0.02521779000458505, + "grad_norm": 0.7977439517145978, + "learning_rate": 0.0008333333333333334, + "loss": 3.4588, + "step": 55 + }, + { + "epoch": 0.02567629527739569, + "grad_norm": 0.7526024318738873, + "learning_rate": 0.0008484848484848485, + "loss": 3.5232, + "step": 56 + }, + { + "epoch": 0.026134800550206328, + "grad_norm": 0.7662405232948842, + "learning_rate": 0.0008636363636363636, + "loss": 3.4648, + "step": 57 + }, + { + "epoch": 0.026593305823016965, + "grad_norm": 0.8170563183610243, + "learning_rate": 0.0008787878787878789, + "loss": 3.5489, + "step": 58 + }, + { + "epoch": 0.0270518110958276, + "grad_norm": 0.789802344285853, + "learning_rate": 0.000893939393939394, + "loss": 3.391, + "step": 59 + }, + { + "epoch": 0.027510316368638238, + "grad_norm": 0.8123295896436565, + "learning_rate": 0.0009090909090909091, + "loss": 3.3914, + "step": 60 + }, + { + "epoch": 0.027968821641448878, + "grad_norm": 0.8555212140870131, + "learning_rate": 0.0009242424242424242, + "loss": 3.2414, + "step": 61 + }, + { + "epoch": 0.028427326914259515, + "grad_norm": 0.889375887240874, + "learning_rate": 0.0009393939393939394, + "loss": 3.3534, + "step": 62 + }, + { + "epoch": 0.02888583218707015, + "grad_norm": 0.895494537400719, + "learning_rate": 0.0009545454545454546, + "loss": 3.3293, + "step": 63 + }, + { + "epoch": 0.029344337459880788, + "grad_norm": 1.085941252942084, + "learning_rate": 0.0009696969696969698, + "loss": 3.4253, + "step": 64 + }, + { + "epoch": 0.029802842732691424, + "grad_norm": 0.8596832681586046, + "learning_rate": 0.000984848484848485, + "loss": 3.2295, + "step": 65 + }, + { + "epoch": 0.030261348005502064, + "grad_norm": 0.7870859001814557, + "learning_rate": 0.001, + "loss": 3.333, + "step": 66 + }, + { + "epoch": 0.0307198532783127, + "grad_norm": 0.7764331933519083, + "learning_rate": 0.0009999994484067654, + "loss": 3.3387, + "step": 67 + }, + { + "epoch": 0.031178358551123338, + "grad_norm": 0.7926574105228507, + "learning_rate": 0.0009999977936282788, + "loss": 3.2162, + "step": 68 + }, + { + "epoch": 0.03163686382393398, + "grad_norm": 0.856376855086844, + "learning_rate": 0.0009999950356681913, + "loss": 3.2564, + "step": 69 + }, + { + "epoch": 0.03209536909674461, + "grad_norm": 0.8052836242408958, + "learning_rate": 0.0009999911745325876, + "loss": 3.1471, + "step": 70 + }, + { + "epoch": 0.03255387436955525, + "grad_norm": 0.8335698033140377, + "learning_rate": 0.0009999862102299873, + "loss": 3.2115, + "step": 71 + }, + { + "epoch": 0.033012379642365884, + "grad_norm": 0.989073731522872, + "learning_rate": 0.0009999801427713433, + "loss": 3.2362, + "step": 72 + }, + { + "epoch": 0.033470884915176524, + "grad_norm": 0.8893593486995723, + "learning_rate": 0.0009999729721700424, + "loss": 3.0897, + "step": 73 + }, + { + "epoch": 0.033929390187987164, + "grad_norm": 0.8618765229176435, + "learning_rate": 0.000999964698441906, + "loss": 3.1702, + "step": 74 + }, + { + "epoch": 0.0343878954607978, + "grad_norm": 0.8765572268111518, + "learning_rate": 0.0009999553216051892, + "loss": 3.1909, + "step": 75 + }, + { + "epoch": 0.03484640073360844, + "grad_norm": 0.7595726506046908, + "learning_rate": 0.00099994484168058, + "loss": 3.122, + "step": 76 + }, + { + "epoch": 0.03530490600641907, + "grad_norm": 0.7374474062062896, + "learning_rate": 0.0009999332586912019, + "loss": 3.1254, + "step": 77 + }, + { + "epoch": 0.03576341127922971, + "grad_norm": 0.8245942240208148, + "learning_rate": 0.0009999205726626108, + "loss": 3.1166, + "step": 78 + }, + { + "epoch": 0.03622191655204035, + "grad_norm": 0.805092359631709, + "learning_rate": 0.000999906783622797, + "loss": 3.0746, + "step": 79 + }, + { + "epoch": 0.036680421824850984, + "grad_norm": 0.7819318501927193, + "learning_rate": 0.0009998918916021842, + "loss": 3.1306, + "step": 80 + }, + { + "epoch": 0.037138927097661624, + "grad_norm": 0.7393190699938926, + "learning_rate": 0.0009998758966336297, + "loss": 2.9788, + "step": 81 + }, + { + "epoch": 0.03759743237047226, + "grad_norm": 0.8450127105940195, + "learning_rate": 0.0009998587987524242, + "loss": 3.0226, + "step": 82 + }, + { + "epoch": 0.0380559376432829, + "grad_norm": 0.7564816913332578, + "learning_rate": 0.0009998405979962926, + "loss": 3.0503, + "step": 83 + }, + { + "epoch": 0.03851444291609354, + "grad_norm": 0.7956299808809639, + "learning_rate": 0.000999821294405392, + "loss": 3.0756, + "step": 84 + }, + { + "epoch": 0.03897294818890417, + "grad_norm": 0.8641811703249915, + "learning_rate": 0.0009998008880223134, + "loss": 3.1236, + "step": 85 + }, + { + "epoch": 0.03943145346171481, + "grad_norm": 0.8161553745750736, + "learning_rate": 0.000999779378892081, + "loss": 3.0318, + "step": 86 + }, + { + "epoch": 0.039889958734525444, + "grad_norm": 0.8267084750876362, + "learning_rate": 0.0009997567670621522, + "loss": 2.9168, + "step": 87 + }, + { + "epoch": 0.040348464007336084, + "grad_norm": 0.789819347351759, + "learning_rate": 0.0009997330525824165, + "loss": 3.0182, + "step": 88 + }, + { + "epoch": 0.040806969280146724, + "grad_norm": 0.7328130624693724, + "learning_rate": 0.0009997082355051976, + "loss": 3.0005, + "step": 89 + }, + { + "epoch": 0.04126547455295736, + "grad_norm": 0.7437959993762279, + "learning_rate": 0.000999682315885251, + "loss": 3.0358, + "step": 90 + }, + { + "epoch": 0.041723979825768, + "grad_norm": 0.7349992695953996, + "learning_rate": 0.0009996552937797645, + "loss": 2.9138, + "step": 91 + }, + { + "epoch": 0.04218248509857864, + "grad_norm": 0.8342997470595163, + "learning_rate": 0.0009996271692483596, + "loss": 2.9667, + "step": 92 + }, + { + "epoch": 0.04264099037138927, + "grad_norm": 0.7747182161592963, + "learning_rate": 0.0009995979423530893, + "loss": 2.9366, + "step": 93 + }, + { + "epoch": 0.04309949564419991, + "grad_norm": 0.7973298742090329, + "learning_rate": 0.000999567613158439, + "loss": 2.9722, + "step": 94 + }, + { + "epoch": 0.04355800091701054, + "grad_norm": 0.7593070965791381, + "learning_rate": 0.0009995361817313263, + "loss": 2.9439, + "step": 95 + }, + { + "epoch": 0.04401650618982118, + "grad_norm": 0.7744480750346329, + "learning_rate": 0.0009995036481411004, + "loss": 3.0013, + "step": 96 + }, + { + "epoch": 0.044475011462631824, + "grad_norm": 0.7865541845241293, + "learning_rate": 0.0009994700124595429, + "loss": 3.0724, + "step": 97 + }, + { + "epoch": 0.04493351673544246, + "grad_norm": 0.7433453857490718, + "learning_rate": 0.0009994352747608663, + "loss": 2.9739, + "step": 98 + }, + { + "epoch": 0.0453920220082531, + "grad_norm": 0.7980978917009903, + "learning_rate": 0.0009993994351217151, + "loss": 2.9607, + "step": 99 + }, + { + "epoch": 0.04585052728106373, + "grad_norm": 0.7820028974679176, + "learning_rate": 0.000999362493621165, + "loss": 2.9822, + "step": 100 + }, + { + "epoch": 0.04630903255387437, + "grad_norm": 0.8574937809834717, + "learning_rate": 0.0009993244503407226, + "loss": 2.9566, + "step": 101 + }, + { + "epoch": 0.04676753782668501, + "grad_norm": 0.7459585257013274, + "learning_rate": 0.0009992853053643258, + "loss": 2.8324, + "step": 102 + }, + { + "epoch": 0.04722604309949564, + "grad_norm": 0.7585613339812294, + "learning_rate": 0.0009992450587783426, + "loss": 2.865, + "step": 103 + }, + { + "epoch": 0.04768454837230628, + "grad_norm": 0.7751895523121589, + "learning_rate": 0.000999203710671572, + "loss": 2.929, + "step": 104 + }, + { + "epoch": 0.048143053645116916, + "grad_norm": 0.8216585888340207, + "learning_rate": 0.0009991612611352438, + "loss": 2.9098, + "step": 105 + }, + { + "epoch": 0.048601558917927556, + "grad_norm": 0.7610163515989358, + "learning_rate": 0.0009991177102630173, + "loss": 2.9307, + "step": 106 + }, + { + "epoch": 0.049060064190738197, + "grad_norm": 0.8121247410754002, + "learning_rate": 0.0009990730581509817, + "loss": 2.7693, + "step": 107 + }, + { + "epoch": 0.04951856946354883, + "grad_norm": 0.910311007849593, + "learning_rate": 0.0009990273048976566, + "loss": 2.8223, + "step": 108 + }, + { + "epoch": 0.04997707473635947, + "grad_norm": 0.844789308632093, + "learning_rate": 0.0009989804506039905, + "loss": 2.8641, + "step": 109 + }, + { + "epoch": 0.0504355800091701, + "grad_norm": 0.8803658704766711, + "learning_rate": 0.0009989324953733614, + "loss": 2.9267, + "step": 110 + }, + { + "epoch": 0.05089408528198074, + "grad_norm": 0.8643792061762, + "learning_rate": 0.0009988834393115766, + "loss": 2.9247, + "step": 111 + }, + { + "epoch": 0.05135259055479138, + "grad_norm": 0.7601083381540958, + "learning_rate": 0.000998833282526872, + "loss": 2.8857, + "step": 112 + }, + { + "epoch": 0.051811095827602016, + "grad_norm": 0.7471470980358736, + "learning_rate": 0.0009987820251299122, + "loss": 2.7725, + "step": 113 + }, + { + "epoch": 0.052269601100412656, + "grad_norm": 0.8256884963320346, + "learning_rate": 0.00099872966723379, + "loss": 2.8483, + "step": 114 + }, + { + "epoch": 0.05272810637322329, + "grad_norm": 0.8499016637184661, + "learning_rate": 0.0009986762089540266, + "loss": 2.8662, + "step": 115 + }, + { + "epoch": 0.05318661164603393, + "grad_norm": 0.8192409265920884, + "learning_rate": 0.0009986216504085709, + "loss": 2.8842, + "step": 116 + }, + { + "epoch": 0.05364511691884457, + "grad_norm": 0.826405680379096, + "learning_rate": 0.0009985659917177991, + "loss": 2.9386, + "step": 117 + }, + { + "epoch": 0.0541036221916552, + "grad_norm": 0.7585784734642335, + "learning_rate": 0.0009985092330045155, + "loss": 2.8199, + "step": 118 + }, + { + "epoch": 0.05456212746446584, + "grad_norm": 0.8711005106907375, + "learning_rate": 0.0009984513743939508, + "loss": 2.8613, + "step": 119 + }, + { + "epoch": 0.055020632737276476, + "grad_norm": 0.8084987864499593, + "learning_rate": 0.0009983924160137626, + "loss": 2.8392, + "step": 120 + }, + { + "epoch": 0.055479138010087116, + "grad_norm": 0.7516190802497104, + "learning_rate": 0.000998332357994035, + "loss": 2.7592, + "step": 121 + }, + { + "epoch": 0.055937643282897756, + "grad_norm": 0.738761614387082, + "learning_rate": 0.0009982712004672786, + "loss": 2.8041, + "step": 122 + }, + { + "epoch": 0.05639614855570839, + "grad_norm": 0.8019607425942012, + "learning_rate": 0.0009982089435684295, + "loss": 2.8011, + "step": 123 + }, + { + "epoch": 0.05685465382851903, + "grad_norm": 0.7723593626985255, + "learning_rate": 0.0009981455874348499, + "loss": 2.787, + "step": 124 + }, + { + "epoch": 0.05731315910132966, + "grad_norm": 0.7304650098786507, + "learning_rate": 0.0009980811322063269, + "loss": 2.7492, + "step": 125 + }, + { + "epoch": 0.0577716643741403, + "grad_norm": 0.7943195423514638, + "learning_rate": 0.0009980155780250728, + "loss": 2.7214, + "step": 126 + }, + { + "epoch": 0.05823016964695094, + "grad_norm": 0.7783386056011478, + "learning_rate": 0.0009979489250357243, + "loss": 2.6985, + "step": 127 + }, + { + "epoch": 0.058688674919761576, + "grad_norm": 0.7605653959550542, + "learning_rate": 0.0009978811733853431, + "loss": 2.7115, + "step": 128 + }, + { + "epoch": 0.059147180192572216, + "grad_norm": 0.7638947065434701, + "learning_rate": 0.0009978123232234147, + "loss": 2.7406, + "step": 129 + }, + { + "epoch": 0.05960568546538285, + "grad_norm": 0.7650268332151823, + "learning_rate": 0.000997742374701848, + "loss": 2.6748, + "step": 130 + }, + { + "epoch": 0.06006419073819349, + "grad_norm": 0.6629200623955231, + "learning_rate": 0.0009976713279749754, + "loss": 2.6281, + "step": 131 + }, + { + "epoch": 0.06052269601100413, + "grad_norm": 0.7500003169600492, + "learning_rate": 0.0009975991831995528, + "loss": 2.6884, + "step": 132 + }, + { + "epoch": 0.06098120128381476, + "grad_norm": 0.7889318375707616, + "learning_rate": 0.0009975259405347581, + "loss": 2.7015, + "step": 133 + }, + { + "epoch": 0.0614397065566254, + "grad_norm": 0.8060209969046405, + "learning_rate": 0.0009974516001421926, + "loss": 2.7438, + "step": 134 + }, + { + "epoch": 0.061898211829436035, + "grad_norm": 0.7786141481685369, + "learning_rate": 0.000997376162185878, + "loss": 2.637, + "step": 135 + }, + { + "epoch": 0.062356717102246675, + "grad_norm": 0.7673882496501128, + "learning_rate": 0.0009972996268322594, + "loss": 2.7028, + "step": 136 + }, + { + "epoch": 0.06281522237505731, + "grad_norm": 0.8038961045419274, + "learning_rate": 0.0009972219942502017, + "loss": 2.7388, + "step": 137 + }, + { + "epoch": 0.06327372764786796, + "grad_norm": 0.7368816463135941, + "learning_rate": 0.0009971432646109918, + "loss": 2.7096, + "step": 138 + }, + { + "epoch": 0.06373223292067859, + "grad_norm": 0.6737644224185774, + "learning_rate": 0.0009970634380883365, + "loss": 2.6984, + "step": 139 + }, + { + "epoch": 0.06419073819348922, + "grad_norm": 0.7122944582823613, + "learning_rate": 0.0009969825148583627, + "loss": 2.7472, + "step": 140 + }, + { + "epoch": 0.06464924346629987, + "grad_norm": 0.7367930070461869, + "learning_rate": 0.0009969004950996173, + "loss": 2.613, + "step": 141 + }, + { + "epoch": 0.0651077487391105, + "grad_norm": 0.7384043942792249, + "learning_rate": 0.0009968173789930668, + "loss": 2.7564, + "step": 142 + }, + { + "epoch": 0.06556625401192114, + "grad_norm": 0.767620293597036, + "learning_rate": 0.0009967331667220958, + "loss": 2.6088, + "step": 143 + }, + { + "epoch": 0.06602475928473177, + "grad_norm": 0.7240448356596734, + "learning_rate": 0.0009966478584725086, + "loss": 2.5833, + "step": 144 + }, + { + "epoch": 0.06648326455754242, + "grad_norm": 0.7605539889142524, + "learning_rate": 0.0009965614544325263, + "loss": 2.6971, + "step": 145 + }, + { + "epoch": 0.06694176983035305, + "grad_norm": 0.7752980683261473, + "learning_rate": 0.000996473954792789, + "loss": 2.7707, + "step": 146 + }, + { + "epoch": 0.06740027510316368, + "grad_norm": 0.7005869117519473, + "learning_rate": 0.0009963853597463532, + "loss": 2.6768, + "step": 147 + }, + { + "epoch": 0.06785878037597433, + "grad_norm": 0.713120060736132, + "learning_rate": 0.000996295669488693, + "loss": 2.6668, + "step": 148 + }, + { + "epoch": 0.06831728564878496, + "grad_norm": 0.7350794730800999, + "learning_rate": 0.0009962048842176979, + "loss": 2.7201, + "step": 149 + }, + { + "epoch": 0.0687757909215956, + "grad_norm": 0.6912952976589049, + "learning_rate": 0.0009961130041336748, + "loss": 2.6999, + "step": 150 + }, + { + "epoch": 0.06923429619440624, + "grad_norm": 0.6842722029402357, + "learning_rate": 0.0009960200294393449, + "loss": 2.7367, + "step": 151 + }, + { + "epoch": 0.06969280146721687, + "grad_norm": 0.6789175079414242, + "learning_rate": 0.0009959259603398453, + "loss": 2.7114, + "step": 152 + }, + { + "epoch": 0.07015130674002751, + "grad_norm": 0.7617632495279606, + "learning_rate": 0.0009958307970427275, + "loss": 2.5613, + "step": 153 + }, + { + "epoch": 0.07060981201283814, + "grad_norm": 0.6922731161314946, + "learning_rate": 0.0009957345397579572, + "loss": 2.6448, + "step": 154 + }, + { + "epoch": 0.07106831728564879, + "grad_norm": 0.7022988061559453, + "learning_rate": 0.0009956371886979138, + "loss": 2.5922, + "step": 155 + }, + { + "epoch": 0.07152682255845942, + "grad_norm": 0.687806583314749, + "learning_rate": 0.00099553874407739, + "loss": 2.7354, + "step": 156 + }, + { + "epoch": 0.07198532783127005, + "grad_norm": 0.654414411563606, + "learning_rate": 0.0009954392061135916, + "loss": 2.6333, + "step": 157 + }, + { + "epoch": 0.0724438331040807, + "grad_norm": 0.6964542783432541, + "learning_rate": 0.0009953385750261364, + "loss": 2.6398, + "step": 158 + }, + { + "epoch": 0.07290233837689133, + "grad_norm": 0.6616586530236905, + "learning_rate": 0.0009952368510370538, + "loss": 2.6082, + "step": 159 + }, + { + "epoch": 0.07336084364970197, + "grad_norm": 0.691976616068106, + "learning_rate": 0.0009951340343707852, + "loss": 2.6513, + "step": 160 + }, + { + "epoch": 0.07381934892251261, + "grad_norm": 0.7060409283385568, + "learning_rate": 0.0009950301252541823, + "loss": 2.6518, + "step": 161 + }, + { + "epoch": 0.07427785419532325, + "grad_norm": 0.6771366547334277, + "learning_rate": 0.0009949251239165075, + "loss": 2.5964, + "step": 162 + }, + { + "epoch": 0.07473635946813388, + "grad_norm": 0.6372964735945613, + "learning_rate": 0.000994819030589433, + "loss": 2.6566, + "step": 163 + }, + { + "epoch": 0.07519486474094451, + "grad_norm": 0.7220359424281946, + "learning_rate": 0.00099471184550704, + "loss": 2.5828, + "step": 164 + }, + { + "epoch": 0.07565337001375516, + "grad_norm": 0.7128759078897414, + "learning_rate": 0.0009946035689058189, + "loss": 2.6292, + "step": 165 + }, + { + "epoch": 0.0761118752865658, + "grad_norm": 0.7033128129940104, + "learning_rate": 0.0009944942010246681, + "loss": 2.7133, + "step": 166 + }, + { + "epoch": 0.07657038055937643, + "grad_norm": 0.6917000283830711, + "learning_rate": 0.0009943837421048942, + "loss": 2.6378, + "step": 167 + }, + { + "epoch": 0.07702888583218707, + "grad_norm": 0.6163952358395598, + "learning_rate": 0.0009942721923902106, + "loss": 2.5707, + "step": 168 + }, + { + "epoch": 0.07748739110499771, + "grad_norm": 0.6263725747256533, + "learning_rate": 0.0009941595521267377, + "loss": 2.5496, + "step": 169 + }, + { + "epoch": 0.07794589637780834, + "grad_norm": 0.6142427032474311, + "learning_rate": 0.0009940458215630017, + "loss": 2.6553, + "step": 170 + }, + { + "epoch": 0.07840440165061899, + "grad_norm": 0.641537442479825, + "learning_rate": 0.0009939310009499348, + "loss": 2.7016, + "step": 171 + }, + { + "epoch": 0.07886290692342962, + "grad_norm": 0.6641731783551235, + "learning_rate": 0.000993815090540874, + "loss": 2.714, + "step": 172 + }, + { + "epoch": 0.07932141219624025, + "grad_norm": 0.6232217230197011, + "learning_rate": 0.000993698090591561, + "loss": 2.7083, + "step": 173 + }, + { + "epoch": 0.07977991746905089, + "grad_norm": 0.6319551296511227, + "learning_rate": 0.0009935800013601416, + "loss": 2.615, + "step": 174 + }, + { + "epoch": 0.08023842274186153, + "grad_norm": 0.6538502813072825, + "learning_rate": 0.000993460823107164, + "loss": 2.6309, + "step": 175 + }, + { + "epoch": 0.08069692801467217, + "grad_norm": 0.6008147989908851, + "learning_rate": 0.0009933405560955803, + "loss": 2.6622, + "step": 176 + }, + { + "epoch": 0.0811554332874828, + "grad_norm": 0.6392984051320731, + "learning_rate": 0.0009932192005907446, + "loss": 2.6444, + "step": 177 + }, + { + "epoch": 0.08161393856029345, + "grad_norm": 0.6267566863371139, + "learning_rate": 0.0009930967568604118, + "loss": 2.7082, + "step": 178 + }, + { + "epoch": 0.08207244383310408, + "grad_norm": 0.6540599045141532, + "learning_rate": 0.000992973225174739, + "loss": 2.59, + "step": 179 + }, + { + "epoch": 0.08253094910591471, + "grad_norm": 0.5751047625777119, + "learning_rate": 0.0009928486058062827, + "loss": 2.5994, + "step": 180 + }, + { + "epoch": 0.08298945437872536, + "grad_norm": 0.6208688138210947, + "learning_rate": 0.0009927228990299999, + "loss": 2.6858, + "step": 181 + }, + { + "epoch": 0.083447959651536, + "grad_norm": 0.6821835495314102, + "learning_rate": 0.0009925961051232468, + "loss": 2.7239, + "step": 182 + }, + { + "epoch": 0.08390646492434663, + "grad_norm": 0.6590782673838924, + "learning_rate": 0.000992468224365778, + "loss": 2.6116, + "step": 183 + }, + { + "epoch": 0.08436497019715727, + "grad_norm": 0.6297179087568446, + "learning_rate": 0.000992339257039746, + "loss": 2.6503, + "step": 184 + }, + { + "epoch": 0.08482347546996791, + "grad_norm": 0.6141845745886244, + "learning_rate": 0.0009922092034297006, + "loss": 2.5864, + "step": 185 + }, + { + "epoch": 0.08528198074277854, + "grad_norm": 0.593295612858332, + "learning_rate": 0.0009920780638225891, + "loss": 2.5281, + "step": 186 + }, + { + "epoch": 0.08574048601558917, + "grad_norm": 0.6816813858436892, + "learning_rate": 0.0009919458385077538, + "loss": 2.5366, + "step": 187 + }, + { + "epoch": 0.08619899128839982, + "grad_norm": 0.58736035382407, + "learning_rate": 0.0009918125277769336, + "loss": 2.5814, + "step": 188 + }, + { + "epoch": 0.08665749656121045, + "grad_norm": 0.6246337372506041, + "learning_rate": 0.0009916781319242614, + "loss": 2.5973, + "step": 189 + }, + { + "epoch": 0.08711600183402109, + "grad_norm": 0.566855577737256, + "learning_rate": 0.0009915426512462646, + "loss": 2.6455, + "step": 190 + }, + { + "epoch": 0.08757450710683173, + "grad_norm": 0.608433828044013, + "learning_rate": 0.0009914060860418644, + "loss": 2.5669, + "step": 191 + }, + { + "epoch": 0.08803301237964237, + "grad_norm": 0.6286510985067574, + "learning_rate": 0.000991268436612374, + "loss": 2.6814, + "step": 192 + }, + { + "epoch": 0.088491517652453, + "grad_norm": 0.6288091346885321, + "learning_rate": 0.0009911297032614997, + "loss": 2.7118, + "step": 193 + }, + { + "epoch": 0.08895002292526365, + "grad_norm": 0.6547954104798998, + "learning_rate": 0.000990989886295339, + "loss": 2.5192, + "step": 194 + }, + { + "epoch": 0.08940852819807428, + "grad_norm": 0.6257593775705739, + "learning_rate": 0.0009908489860223804, + "loss": 2.6702, + "step": 195 + }, + { + "epoch": 0.08986703347088491, + "grad_norm": 0.6200072863203764, + "learning_rate": 0.000990707002753502, + "loss": 2.5861, + "step": 196 + }, + { + "epoch": 0.09032553874369555, + "grad_norm": 0.5851990616305708, + "learning_rate": 0.0009905639368019724, + "loss": 2.6084, + "step": 197 + }, + { + "epoch": 0.0907840440165062, + "grad_norm": 0.6434888176499868, + "learning_rate": 0.0009904197884834482, + "loss": 2.5862, + "step": 198 + }, + { + "epoch": 0.09124254928931683, + "grad_norm": 0.6087360451853181, + "learning_rate": 0.0009902745581159742, + "loss": 2.6399, + "step": 199 + }, + { + "epoch": 0.09170105456212746, + "grad_norm": 0.6109066548542731, + "learning_rate": 0.0009901282460199829, + "loss": 2.6762, + "step": 200 + }, + { + "epoch": 0.0921595598349381, + "grad_norm": 0.637654399581707, + "learning_rate": 0.0009899808525182935, + "loss": 2.6364, + "step": 201 + }, + { + "epoch": 0.09261806510774874, + "grad_norm": 0.6209584402146896, + "learning_rate": 0.0009898323779361107, + "loss": 2.5967, + "step": 202 + }, + { + "epoch": 0.09307657038055937, + "grad_norm": 0.6250414570082558, + "learning_rate": 0.000989682822601025, + "loss": 2.5635, + "step": 203 + }, + { + "epoch": 0.09353507565337002, + "grad_norm": 0.669390208509988, + "learning_rate": 0.0009895321868430113, + "loss": 2.6039, + "step": 204 + }, + { + "epoch": 0.09399358092618065, + "grad_norm": 0.6163543713335571, + "learning_rate": 0.0009893804709944281, + "loss": 2.5866, + "step": 205 + }, + { + "epoch": 0.09445208619899129, + "grad_norm": 0.5980418200765097, + "learning_rate": 0.0009892276753900174, + "loss": 2.6501, + "step": 206 + }, + { + "epoch": 0.09491059147180192, + "grad_norm": 0.7598004612009133, + "learning_rate": 0.0009890738003669028, + "loss": 2.6187, + "step": 207 + }, + { + "epoch": 0.09536909674461257, + "grad_norm": 0.5838937784812266, + "learning_rate": 0.0009889188462645904, + "loss": 2.6512, + "step": 208 + }, + { + "epoch": 0.0958276020174232, + "grad_norm": 0.6026678347664054, + "learning_rate": 0.0009887628134249667, + "loss": 2.6488, + "step": 209 + }, + { + "epoch": 0.09628610729023383, + "grad_norm": 0.6347867375310687, + "learning_rate": 0.0009886057021922983, + "loss": 2.705, + "step": 210 + }, + { + "epoch": 0.09674461256304448, + "grad_norm": 0.6194068888200545, + "learning_rate": 0.0009884475129132311, + "loss": 2.6149, + "step": 211 + }, + { + "epoch": 0.09720311783585511, + "grad_norm": 0.6170659209569226, + "learning_rate": 0.0009882882459367897, + "loss": 2.4506, + "step": 212 + }, + { + "epoch": 0.09766162310866575, + "grad_norm": 0.6494991945812304, + "learning_rate": 0.0009881279016143766, + "loss": 2.4766, + "step": 213 + }, + { + "epoch": 0.09812012838147639, + "grad_norm": 0.598252376254177, + "learning_rate": 0.0009879664802997707, + "loss": 2.5878, + "step": 214 + }, + { + "epoch": 0.09857863365428703, + "grad_norm": 0.6116175233157184, + "learning_rate": 0.000987803982349128, + "loss": 2.6725, + "step": 215 + }, + { + "epoch": 0.09903713892709766, + "grad_norm": 0.5776968041607143, + "learning_rate": 0.0009876404081209796, + "loss": 2.6445, + "step": 216 + }, + { + "epoch": 0.09949564419990829, + "grad_norm": 0.5725766866731347, + "learning_rate": 0.000987475757976231, + "loss": 2.5585, + "step": 217 + }, + { + "epoch": 0.09995414947271894, + "grad_norm": 0.5506704773670977, + "learning_rate": 0.000987310032278162, + "loss": 2.5715, + "step": 218 + }, + { + "epoch": 0.10041265474552957, + "grad_norm": 0.5592252437539471, + "learning_rate": 0.0009871432313924254, + "loss": 2.5515, + "step": 219 + }, + { + "epoch": 0.1008711600183402, + "grad_norm": 0.6091756039776418, + "learning_rate": 0.000986975355687046, + "loss": 2.6184, + "step": 220 + }, + { + "epoch": 0.10132966529115085, + "grad_norm": 0.6006174796435713, + "learning_rate": 0.0009868064055324204, + "loss": 2.6106, + "step": 221 + }, + { + "epoch": 0.10178817056396149, + "grad_norm": 0.6070705329919657, + "learning_rate": 0.0009866363813013153, + "loss": 2.6133, + "step": 222 + }, + { + "epoch": 0.10224667583677212, + "grad_norm": 0.6019221351711501, + "learning_rate": 0.0009864652833688676, + "loss": 2.5858, + "step": 223 + }, + { + "epoch": 0.10270518110958277, + "grad_norm": 0.6124743047197442, + "learning_rate": 0.0009862931121125836, + "loss": 2.5511, + "step": 224 + }, + { + "epoch": 0.1031636863823934, + "grad_norm": 0.5982733395128359, + "learning_rate": 0.000986119867912337, + "loss": 2.6024, + "step": 225 + }, + { + "epoch": 0.10362219165520403, + "grad_norm": 0.5931800917603413, + "learning_rate": 0.000985945551150369, + "loss": 2.5531, + "step": 226 + }, + { + "epoch": 0.10408069692801467, + "grad_norm": 0.5892312030265313, + "learning_rate": 0.0009857701622112876, + "loss": 2.6323, + "step": 227 + }, + { + "epoch": 0.10453920220082531, + "grad_norm": 0.5976048333281602, + "learning_rate": 0.000985593701482066, + "loss": 2.6601, + "step": 228 + }, + { + "epoch": 0.10499770747363595, + "grad_norm": 0.6483517279112491, + "learning_rate": 0.0009854161693520424, + "loss": 2.5432, + "step": 229 + }, + { + "epoch": 0.10545621274644658, + "grad_norm": 0.5756405034896843, + "learning_rate": 0.0009852375662129194, + "loss": 2.5948, + "step": 230 + }, + { + "epoch": 0.10591471801925723, + "grad_norm": 0.5845822219244076, + "learning_rate": 0.0009850578924587613, + "loss": 2.5679, + "step": 231 + }, + { + "epoch": 0.10637322329206786, + "grad_norm": 0.6276658573274776, + "learning_rate": 0.000984877148485996, + "loss": 2.6415, + "step": 232 + }, + { + "epoch": 0.10683172856487849, + "grad_norm": 0.6123907733706869, + "learning_rate": 0.000984695334693412, + "loss": 2.6868, + "step": 233 + }, + { + "epoch": 0.10729023383768914, + "grad_norm": 0.5770479288626241, + "learning_rate": 0.000984512451482158, + "loss": 2.5756, + "step": 234 + }, + { + "epoch": 0.10774873911049977, + "grad_norm": 0.6136530038717213, + "learning_rate": 0.0009843284992557431, + "loss": 2.5654, + "step": 235 + }, + { + "epoch": 0.1082072443833104, + "grad_norm": 0.5895473239848493, + "learning_rate": 0.000984143478420034, + "loss": 2.6142, + "step": 236 + }, + { + "epoch": 0.10866574965612105, + "grad_norm": 0.5980322347053713, + "learning_rate": 0.0009839573893832563, + "loss": 2.5312, + "step": 237 + }, + { + "epoch": 0.10912425492893169, + "grad_norm": 0.672953328740223, + "learning_rate": 0.000983770232555991, + "loss": 2.6849, + "step": 238 + }, + { + "epoch": 0.10958276020174232, + "grad_norm": 0.5766198528152916, + "learning_rate": 0.0009835820083511765, + "loss": 2.525, + "step": 239 + }, + { + "epoch": 0.11004126547455295, + "grad_norm": 0.5718968290598287, + "learning_rate": 0.0009833927171841055, + "loss": 2.6535, + "step": 240 + }, + { + "epoch": 0.1104997707473636, + "grad_norm": 0.5771599298872403, + "learning_rate": 0.0009832023594724246, + "loss": 2.6177, + "step": 241 + }, + { + "epoch": 0.11095827602017423, + "grad_norm": 0.6220741765906843, + "learning_rate": 0.0009830109356361344, + "loss": 2.607, + "step": 242 + }, + { + "epoch": 0.11141678129298486, + "grad_norm": 0.5731390524388003, + "learning_rate": 0.0009828184460975867, + "loss": 2.6105, + "step": 243 + }, + { + "epoch": 0.11187528656579551, + "grad_norm": 0.5752333628871463, + "learning_rate": 0.0009826248912814855, + "loss": 2.573, + "step": 244 + }, + { + "epoch": 0.11233379183860615, + "grad_norm": 0.568724149181011, + "learning_rate": 0.0009824302716148847, + "loss": 2.5519, + "step": 245 + }, + { + "epoch": 0.11279229711141678, + "grad_norm": 0.6103350690520521, + "learning_rate": 0.0009822345875271884, + "loss": 2.5199, + "step": 246 + }, + { + "epoch": 0.11325080238422743, + "grad_norm": 0.5968423306037046, + "learning_rate": 0.0009820378394501481, + "loss": 2.5546, + "step": 247 + }, + { + "epoch": 0.11370930765703806, + "grad_norm": 0.6039927272641011, + "learning_rate": 0.0009818400278178636, + "loss": 2.46, + "step": 248 + }, + { + "epoch": 0.11416781292984869, + "grad_norm": 0.5374269779724576, + "learning_rate": 0.0009816411530667814, + "loss": 2.5458, + "step": 249 + }, + { + "epoch": 0.11462631820265932, + "grad_norm": 0.563874408563703, + "learning_rate": 0.000981441215635693, + "loss": 2.646, + "step": 250 + }, + { + "epoch": 0.11508482347546997, + "grad_norm": 0.567710837862122, + "learning_rate": 0.0009812402159657353, + "loss": 2.65, + "step": 251 + }, + { + "epoch": 0.1155433287482806, + "grad_norm": 0.5703326660177015, + "learning_rate": 0.000981038154500388, + "loss": 2.6329, + "step": 252 + }, + { + "epoch": 0.11600183402109124, + "grad_norm": 0.5822679235271484, + "learning_rate": 0.0009808350316854746, + "loss": 2.5254, + "step": 253 + }, + { + "epoch": 0.11646033929390188, + "grad_norm": 0.5769639400060631, + "learning_rate": 0.0009806308479691594, + "loss": 2.6565, + "step": 254 + }, + { + "epoch": 0.11691884456671252, + "grad_norm": 0.5931650318430496, + "learning_rate": 0.0009804256038019482, + "loss": 2.6036, + "step": 255 + }, + { + "epoch": 0.11737734983952315, + "grad_norm": 0.5568408244238674, + "learning_rate": 0.0009802192996366857, + "loss": 2.6074, + "step": 256 + }, + { + "epoch": 0.1178358551123338, + "grad_norm": 0.5421508986269281, + "learning_rate": 0.0009800119359285563, + "loss": 2.5113, + "step": 257 + }, + { + "epoch": 0.11829436038514443, + "grad_norm": 0.5575380072853977, + "learning_rate": 0.0009798035131350813, + "loss": 2.5804, + "step": 258 + }, + { + "epoch": 0.11875286565795506, + "grad_norm": 0.5605650337697267, + "learning_rate": 0.0009795940317161194, + "loss": 2.5151, + "step": 259 + }, + { + "epoch": 0.1192113709307657, + "grad_norm": 0.5665470047055837, + "learning_rate": 0.0009793834921338646, + "loss": 2.6223, + "step": 260 + }, + { + "epoch": 0.11966987620357634, + "grad_norm": 0.601334277936254, + "learning_rate": 0.0009791718948528457, + "loss": 2.6182, + "step": 261 + }, + { + "epoch": 0.12012838147638698, + "grad_norm": 0.5856376071232455, + "learning_rate": 0.0009789592403399252, + "loss": 2.5557, + "step": 262 + }, + { + "epoch": 0.12058688674919761, + "grad_norm": 0.5560039576846926, + "learning_rate": 0.0009787455290642985, + "loss": 2.5425, + "step": 263 + }, + { + "epoch": 0.12104539202200826, + "grad_norm": 0.5986549334849482, + "learning_rate": 0.000978530761497492, + "loss": 2.5715, + "step": 264 + }, + { + "epoch": 0.12150389729481889, + "grad_norm": 0.6061374713031297, + "learning_rate": 0.0009783149381133633, + "loss": 2.542, + "step": 265 + }, + { + "epoch": 0.12196240256762952, + "grad_norm": 0.5638383124786797, + "learning_rate": 0.0009780980593880992, + "loss": 2.535, + "step": 266 + }, + { + "epoch": 0.12242090784044017, + "grad_norm": 0.5648568423094027, + "learning_rate": 0.0009778801258002153, + "loss": 2.5692, + "step": 267 + }, + { + "epoch": 0.1228794131132508, + "grad_norm": 0.5291556149163493, + "learning_rate": 0.000977661137830554, + "loss": 2.5181, + "step": 268 + }, + { + "epoch": 0.12333791838606144, + "grad_norm": 0.5515281821829071, + "learning_rate": 0.0009774410959622845, + "loss": 2.5925, + "step": 269 + }, + { + "epoch": 0.12379642365887207, + "grad_norm": 0.5549962471453032, + "learning_rate": 0.000977220000680901, + "loss": 2.4525, + "step": 270 + }, + { + "epoch": 0.12425492893168272, + "grad_norm": 0.547251879945252, + "learning_rate": 0.000976997852474223, + "loss": 2.466, + "step": 271 + }, + { + "epoch": 0.12471343420449335, + "grad_norm": 0.5754094172647423, + "learning_rate": 0.0009767746518323914, + "loss": 2.5096, + "step": 272 + }, + { + "epoch": 0.12517193947730398, + "grad_norm": 0.5844901997698709, + "learning_rate": 0.0009765503992478704, + "loss": 2.5299, + "step": 273 + }, + { + "epoch": 0.12563044475011462, + "grad_norm": 0.594468820796214, + "learning_rate": 0.0009763250952154449, + "loss": 2.5445, + "step": 274 + }, + { + "epoch": 0.12608895002292525, + "grad_norm": 0.5992243012571514, + "learning_rate": 0.0009760987402322195, + "loss": 2.5248, + "step": 275 + }, + { + "epoch": 0.1265474552957359, + "grad_norm": 0.5706384238644077, + "learning_rate": 0.0009758713347976178, + "loss": 2.6169, + "step": 276 + }, + { + "epoch": 0.12700596056854654, + "grad_norm": 0.5404344016187651, + "learning_rate": 0.000975642879413381, + "loss": 2.601, + "step": 277 + }, + { + "epoch": 0.12746446584135718, + "grad_norm": 0.5144206810737709, + "learning_rate": 0.0009754133745835665, + "loss": 2.547, + "step": 278 + }, + { + "epoch": 0.1279229711141678, + "grad_norm": 0.5739118987191536, + "learning_rate": 0.0009751828208145482, + "loss": 2.5798, + "step": 279 + }, + { + "epoch": 0.12838147638697844, + "grad_norm": 0.5748254342892767, + "learning_rate": 0.0009749512186150131, + "loss": 2.5758, + "step": 280 + }, + { + "epoch": 0.12883998165978908, + "grad_norm": 0.5757763565834444, + "learning_rate": 0.0009747185684959625, + "loss": 2.5645, + "step": 281 + }, + { + "epoch": 0.12929848693259974, + "grad_norm": 0.6078861473122287, + "learning_rate": 0.000974484870970709, + "loss": 2.5777, + "step": 282 + }, + { + "epoch": 0.12975699220541037, + "grad_norm": 0.6074095656536834, + "learning_rate": 0.0009742501265548767, + "loss": 2.5691, + "step": 283 + }, + { + "epoch": 0.130215497478221, + "grad_norm": 0.5798354821436761, + "learning_rate": 0.0009740143357663993, + "loss": 2.6626, + "step": 284 + }, + { + "epoch": 0.13067400275103164, + "grad_norm": 0.5576661143992973, + "learning_rate": 0.000973777499125519, + "loss": 2.5507, + "step": 285 + }, + { + "epoch": 0.13113250802384227, + "grad_norm": 0.587084673998335, + "learning_rate": 0.0009735396171547859, + "loss": 2.5111, + "step": 286 + }, + { + "epoch": 0.1315910132966529, + "grad_norm": 0.5737476390947697, + "learning_rate": 0.0009733006903790564, + "loss": 2.6256, + "step": 287 + }, + { + "epoch": 0.13204951856946354, + "grad_norm": 0.5808095613305292, + "learning_rate": 0.0009730607193254922, + "loss": 2.5666, + "step": 288 + }, + { + "epoch": 0.1325080238422742, + "grad_norm": 0.5651967191724799, + "learning_rate": 0.0009728197045235585, + "loss": 2.6118, + "step": 289 + }, + { + "epoch": 0.13296652911508483, + "grad_norm": 0.5886300860648199, + "learning_rate": 0.0009725776465050242, + "loss": 2.5369, + "step": 290 + }, + { + "epoch": 0.13342503438789546, + "grad_norm": 0.53545333237001, + "learning_rate": 0.0009723345458039594, + "loss": 2.5248, + "step": 291 + }, + { + "epoch": 0.1338835396607061, + "grad_norm": 0.5449175872941682, + "learning_rate": 0.000972090402956735, + "loss": 2.5906, + "step": 292 + }, + { + "epoch": 0.13434204493351673, + "grad_norm": 0.5095896355436754, + "learning_rate": 0.0009718452185020212, + "loss": 2.5173, + "step": 293 + }, + { + "epoch": 0.13480055020632736, + "grad_norm": 0.5574435938411332, + "learning_rate": 0.0009715989929807862, + "loss": 2.516, + "step": 294 + }, + { + "epoch": 0.13525905547913802, + "grad_norm": 0.526618071890048, + "learning_rate": 0.0009713517269362955, + "loss": 2.5834, + "step": 295 + }, + { + "epoch": 0.13571756075194866, + "grad_norm": 0.5533034927603759, + "learning_rate": 0.0009711034209141101, + "loss": 2.526, + "step": 296 + }, + { + "epoch": 0.1361760660247593, + "grad_norm": 0.5664116263134428, + "learning_rate": 0.0009708540754620856, + "loss": 2.6131, + "step": 297 + }, + { + "epoch": 0.13663457129756992, + "grad_norm": 0.5691228810629253, + "learning_rate": 0.0009706036911303713, + "loss": 2.65, + "step": 298 + }, + { + "epoch": 0.13709307657038056, + "grad_norm": 0.5579208753345004, + "learning_rate": 0.0009703522684714083, + "loss": 2.4637, + "step": 299 + }, + { + "epoch": 0.1375515818431912, + "grad_norm": 0.5190159341153917, + "learning_rate": 0.0009700998080399286, + "loss": 2.5604, + "step": 300 + }, + { + "epoch": 0.13801008711600182, + "grad_norm": 0.5375258136146636, + "learning_rate": 0.0009698463103929542, + "loss": 2.5747, + "step": 301 + }, + { + "epoch": 0.13846859238881248, + "grad_norm": 0.5554235359745026, + "learning_rate": 0.0009695917760897954, + "loss": 2.5373, + "step": 302 + }, + { + "epoch": 0.13892709766162312, + "grad_norm": 0.5520983558658098, + "learning_rate": 0.0009693362056920501, + "loss": 2.5997, + "step": 303 + }, + { + "epoch": 0.13938560293443375, + "grad_norm": 0.5453193579580007, + "learning_rate": 0.0009690795997636015, + "loss": 2.5561, + "step": 304 + }, + { + "epoch": 0.13984410820724438, + "grad_norm": 0.5226385128980934, + "learning_rate": 0.0009688219588706179, + "loss": 2.5535, + "step": 305 + }, + { + "epoch": 0.14030261348005502, + "grad_norm": 0.5354682863657569, + "learning_rate": 0.0009685632835815518, + "loss": 2.5406, + "step": 306 + }, + { + "epoch": 0.14076111875286565, + "grad_norm": 0.5221645925233275, + "learning_rate": 0.0009683035744671367, + "loss": 2.4932, + "step": 307 + }, + { + "epoch": 0.14121962402567628, + "grad_norm": 0.5378711032718523, + "learning_rate": 0.0009680428321003883, + "loss": 2.4842, + "step": 308 + }, + { + "epoch": 0.14167812929848694, + "grad_norm": 0.542344256269699, + "learning_rate": 0.000967781057056601, + "loss": 2.5497, + "step": 309 + }, + { + "epoch": 0.14213663457129758, + "grad_norm": 0.5609751585028458, + "learning_rate": 0.0009675182499133485, + "loss": 2.5906, + "step": 310 + }, + { + "epoch": 0.1425951398441082, + "grad_norm": 0.5312580093137661, + "learning_rate": 0.0009672544112504813, + "loss": 2.532, + "step": 311 + }, + { + "epoch": 0.14305364511691884, + "grad_norm": 0.5760449522250171, + "learning_rate": 0.0009669895416501257, + "loss": 2.4879, + "step": 312 + }, + { + "epoch": 0.14351215038972948, + "grad_norm": 0.568527467604185, + "learning_rate": 0.0009667236416966833, + "loss": 2.6042, + "step": 313 + }, + { + "epoch": 0.1439706556625401, + "grad_norm": 0.5452731215637787, + "learning_rate": 0.0009664567119768281, + "loss": 2.5207, + "step": 314 + }, + { + "epoch": 0.14442916093535077, + "grad_norm": 0.5690370074687068, + "learning_rate": 0.0009661887530795067, + "loss": 2.5093, + "step": 315 + }, + { + "epoch": 0.1448876662081614, + "grad_norm": 0.5229481201534326, + "learning_rate": 0.0009659197655959365, + "loss": 2.5368, + "step": 316 + }, + { + "epoch": 0.14534617148097204, + "grad_norm": 0.5492170918013664, + "learning_rate": 0.000965649750119604, + "loss": 2.5336, + "step": 317 + }, + { + "epoch": 0.14580467675378267, + "grad_norm": 0.5781341051418902, + "learning_rate": 0.0009653787072462643, + "loss": 2.5733, + "step": 318 + }, + { + "epoch": 0.1462631820265933, + "grad_norm": 0.5610789650428708, + "learning_rate": 0.0009651066375739388, + "loss": 2.6285, + "step": 319 + }, + { + "epoch": 0.14672168729940394, + "grad_norm": 0.5232316869562779, + "learning_rate": 0.000964833541702915, + "loss": 2.4971, + "step": 320 + }, + { + "epoch": 0.14718019257221457, + "grad_norm": 0.5160485281604494, + "learning_rate": 0.0009645594202357438, + "loss": 2.5608, + "step": 321 + }, + { + "epoch": 0.14763869784502523, + "grad_norm": 0.5307444742484372, + "learning_rate": 0.0009642842737772397, + "loss": 2.5195, + "step": 322 + }, + { + "epoch": 0.14809720311783586, + "grad_norm": 0.529528240456025, + "learning_rate": 0.0009640081029344782, + "loss": 2.5763, + "step": 323 + }, + { + "epoch": 0.1485557083906465, + "grad_norm": 0.5599084942332467, + "learning_rate": 0.0009637309083167956, + "loss": 2.5854, + "step": 324 + }, + { + "epoch": 0.14901421366345713, + "grad_norm": 0.5564398690064887, + "learning_rate": 0.0009634526905357859, + "loss": 2.5816, + "step": 325 + }, + { + "epoch": 0.14947271893626776, + "grad_norm": 0.5422827349972301, + "learning_rate": 0.000963173450205302, + "loss": 2.6253, + "step": 326 + }, + { + "epoch": 0.1499312242090784, + "grad_norm": 0.5337732988211527, + "learning_rate": 0.0009628931879414517, + "loss": 2.4659, + "step": 327 + }, + { + "epoch": 0.15038972948188903, + "grad_norm": 0.5161515465773284, + "learning_rate": 0.0009626119043625983, + "loss": 2.5288, + "step": 328 + }, + { + "epoch": 0.1508482347546997, + "grad_norm": 0.5311709997835761, + "learning_rate": 0.0009623296000893582, + "loss": 2.5035, + "step": 329 + }, + { + "epoch": 0.15130674002751032, + "grad_norm": 0.5231494780583055, + "learning_rate": 0.0009620462757446, + "loss": 2.5416, + "step": 330 + }, + { + "epoch": 0.15176524530032096, + "grad_norm": 0.48454434829020343, + "learning_rate": 0.0009617619319534428, + "loss": 2.5633, + "step": 331 + }, + { + "epoch": 0.1522237505731316, + "grad_norm": 0.5146099984202295, + "learning_rate": 0.000961476569343255, + "loss": 2.5492, + "step": 332 + }, + { + "epoch": 0.15268225584594222, + "grad_norm": 0.49166938000804344, + "learning_rate": 0.0009611901885436529, + "loss": 2.5809, + "step": 333 + }, + { + "epoch": 0.15314076111875286, + "grad_norm": 0.4783025654074074, + "learning_rate": 0.0009609027901864996, + "loss": 2.4909, + "step": 334 + }, + { + "epoch": 0.15359926639156352, + "grad_norm": 0.5089165567633791, + "learning_rate": 0.0009606143749059029, + "loss": 2.5651, + "step": 335 + }, + { + "epoch": 0.15405777166437415, + "grad_norm": 0.5468812678182273, + "learning_rate": 0.0009603249433382144, + "loss": 2.5859, + "step": 336 + }, + { + "epoch": 0.15451627693718478, + "grad_norm": 0.5130845862720415, + "learning_rate": 0.0009600344961220282, + "loss": 2.5498, + "step": 337 + }, + { + "epoch": 0.15497478220999542, + "grad_norm": 0.5360950187831669, + "learning_rate": 0.0009597430338981791, + "loss": 2.58, + "step": 338 + }, + { + "epoch": 0.15543328748280605, + "grad_norm": 0.5187407178310456, + "learning_rate": 0.0009594505573097414, + "loss": 2.5459, + "step": 339 + }, + { + "epoch": 0.15589179275561668, + "grad_norm": 0.5026814598150788, + "learning_rate": 0.0009591570670020277, + "loss": 2.5823, + "step": 340 + }, + { + "epoch": 0.15635029802842731, + "grad_norm": 0.5572460452192698, + "learning_rate": 0.0009588625636225871, + "loss": 2.547, + "step": 341 + }, + { + "epoch": 0.15680880330123798, + "grad_norm": 0.4999677533129889, + "learning_rate": 0.0009585670478212036, + "loss": 2.3897, + "step": 342 + }, + { + "epoch": 0.1572673085740486, + "grad_norm": 0.518854406633483, + "learning_rate": 0.0009582705202498956, + "loss": 2.5457, + "step": 343 + }, + { + "epoch": 0.15772581384685924, + "grad_norm": 0.5606148070947874, + "learning_rate": 0.0009579729815629133, + "loss": 2.6031, + "step": 344 + }, + { + "epoch": 0.15818431911966988, + "grad_norm": 0.5191812918581336, + "learning_rate": 0.0009576744324167379, + "loss": 2.6157, + "step": 345 + }, + { + "epoch": 0.1586428243924805, + "grad_norm": 0.5143831230136477, + "learning_rate": 0.0009573748734700804, + "loss": 2.4751, + "step": 346 + }, + { + "epoch": 0.15910132966529114, + "grad_norm": 0.49946222125314443, + "learning_rate": 0.0009570743053838796, + "loss": 2.5632, + "step": 347 + }, + { + "epoch": 0.15955983493810177, + "grad_norm": 0.5381662283774419, + "learning_rate": 0.0009567727288213005, + "loss": 2.4971, + "step": 348 + }, + { + "epoch": 0.16001834021091244, + "grad_norm": 0.544221760931853, + "learning_rate": 0.0009564701444477337, + "loss": 2.5955, + "step": 349 + }, + { + "epoch": 0.16047684548372307, + "grad_norm": 0.5052097413118806, + "learning_rate": 0.000956166552930793, + "loss": 2.5167, + "step": 350 + }, + { + "epoch": 0.1609353507565337, + "grad_norm": 0.4994099845279637, + "learning_rate": 0.0009558619549403147, + "loss": 2.6218, + "step": 351 + }, + { + "epoch": 0.16139385602934433, + "grad_norm": 0.504726915021702, + "learning_rate": 0.0009555563511483555, + "loss": 2.5321, + "step": 352 + }, + { + "epoch": 0.16185236130215497, + "grad_norm": 0.5370541925693408, + "learning_rate": 0.0009552497422291912, + "loss": 2.4896, + "step": 353 + }, + { + "epoch": 0.1623108665749656, + "grad_norm": 0.492856830320999, + "learning_rate": 0.0009549421288593157, + "loss": 2.5591, + "step": 354 + }, + { + "epoch": 0.16276937184777626, + "grad_norm": 0.5028330865231404, + "learning_rate": 0.0009546335117174385, + "loss": 2.5729, + "step": 355 + }, + { + "epoch": 0.1632278771205869, + "grad_norm": 0.5178782029538855, + "learning_rate": 0.0009543238914844843, + "loss": 2.5827, + "step": 356 + }, + { + "epoch": 0.16368638239339753, + "grad_norm": 0.5901501823535796, + "learning_rate": 0.0009540132688435907, + "loss": 2.555, + "step": 357 + }, + { + "epoch": 0.16414488766620816, + "grad_norm": 0.5490084136500466, + "learning_rate": 0.0009537016444801074, + "loss": 2.455, + "step": 358 + }, + { + "epoch": 0.1646033929390188, + "grad_norm": 0.5421270675193713, + "learning_rate": 0.0009533890190815935, + "loss": 2.673, + "step": 359 + }, + { + "epoch": 0.16506189821182943, + "grad_norm": 0.6075197528899413, + "learning_rate": 0.0009530753933378173, + "loss": 2.5391, + "step": 360 + }, + { + "epoch": 0.16552040348464006, + "grad_norm": 0.495822896383404, + "learning_rate": 0.0009527607679407545, + "loss": 2.5386, + "step": 361 + }, + { + "epoch": 0.16597890875745072, + "grad_norm": 0.48203328540254803, + "learning_rate": 0.0009524451435845857, + "loss": 2.5915, + "step": 362 + }, + { + "epoch": 0.16643741403026135, + "grad_norm": 0.5127007166259783, + "learning_rate": 0.0009521285209656963, + "loss": 2.5058, + "step": 363 + }, + { + "epoch": 0.166895919303072, + "grad_norm": 0.5399423774825034, + "learning_rate": 0.0009518109007826734, + "loss": 2.4627, + "step": 364 + }, + { + "epoch": 0.16735442457588262, + "grad_norm": 0.5523500618681902, + "learning_rate": 0.0009514922837363059, + "loss": 2.579, + "step": 365 + }, + { + "epoch": 0.16781292984869325, + "grad_norm": 0.49111259729303997, + "learning_rate": 0.0009511726705295817, + "loss": 2.5709, + "step": 366 + }, + { + "epoch": 0.1682714351215039, + "grad_norm": 0.5131912979469594, + "learning_rate": 0.000950852061867687, + "loss": 2.5322, + "step": 367 + }, + { + "epoch": 0.16872994039431455, + "grad_norm": 0.5509984470823034, + "learning_rate": 0.0009505304584580038, + "loss": 2.5317, + "step": 368 + }, + { + "epoch": 0.16918844566712518, + "grad_norm": 0.5246047924257722, + "learning_rate": 0.0009502078610101092, + "loss": 2.6448, + "step": 369 + }, + { + "epoch": 0.16964695093993581, + "grad_norm": 0.5033750000032712, + "learning_rate": 0.0009498842702357736, + "loss": 2.5984, + "step": 370 + }, + { + "epoch": 0.17010545621274645, + "grad_norm": 0.5235435810702732, + "learning_rate": 0.0009495596868489587, + "loss": 2.5638, + "step": 371 + }, + { + "epoch": 0.17056396148555708, + "grad_norm": 0.5009764760558294, + "learning_rate": 0.0009492341115658167, + "loss": 2.5667, + "step": 372 + }, + { + "epoch": 0.1710224667583677, + "grad_norm": 0.5011022918418896, + "learning_rate": 0.0009489075451046879, + "loss": 2.5104, + "step": 373 + }, + { + "epoch": 0.17148097203117835, + "grad_norm": 0.47598016910297947, + "learning_rate": 0.0009485799881861, + "loss": 2.4386, + "step": 374 + }, + { + "epoch": 0.171939477303989, + "grad_norm": 0.5069877570774532, + "learning_rate": 0.0009482514415327654, + "loss": 2.5567, + "step": 375 + }, + { + "epoch": 0.17239798257679964, + "grad_norm": 0.5121620186172534, + "learning_rate": 0.000947921905869581, + "loss": 2.5397, + "step": 376 + }, + { + "epoch": 0.17285648784961027, + "grad_norm": 0.5286869358667998, + "learning_rate": 0.0009475913819236248, + "loss": 2.6549, + "step": 377 + }, + { + "epoch": 0.1733149931224209, + "grad_norm": 0.4801490721373994, + "learning_rate": 0.0009472598704241561, + "loss": 2.5833, + "step": 378 + }, + { + "epoch": 0.17377349839523154, + "grad_norm": 0.5174828459653947, + "learning_rate": 0.0009469273721026131, + "loss": 2.4599, + "step": 379 + }, + { + "epoch": 0.17423200366804217, + "grad_norm": 0.5035328197539712, + "learning_rate": 0.0009465938876926111, + "loss": 2.4341, + "step": 380 + }, + { + "epoch": 0.1746905089408528, + "grad_norm": 0.49756340561730916, + "learning_rate": 0.0009462594179299406, + "loss": 2.6987, + "step": 381 + }, + { + "epoch": 0.17514901421366347, + "grad_norm": 0.5468794971519574, + "learning_rate": 0.0009459239635525672, + "loss": 2.5496, + "step": 382 + }, + { + "epoch": 0.1756075194864741, + "grad_norm": 0.5384442310014929, + "learning_rate": 0.0009455875253006281, + "loss": 2.561, + "step": 383 + }, + { + "epoch": 0.17606602475928473, + "grad_norm": 0.4987188434235622, + "learning_rate": 0.0009452501039164315, + "loss": 2.6456, + "step": 384 + }, + { + "epoch": 0.17652453003209537, + "grad_norm": 0.5345090961005148, + "learning_rate": 0.0009449117001444549, + "loss": 2.467, + "step": 385 + }, + { + "epoch": 0.176983035304906, + "grad_norm": 0.5084774333785659, + "learning_rate": 0.0009445723147313433, + "loss": 2.5938, + "step": 386 + }, + { + "epoch": 0.17744154057771663, + "grad_norm": 0.5202983874155489, + "learning_rate": 0.0009442319484259074, + "loss": 2.5367, + "step": 387 + }, + { + "epoch": 0.1779000458505273, + "grad_norm": 0.5065749889480968, + "learning_rate": 0.0009438906019791222, + "loss": 2.586, + "step": 388 + }, + { + "epoch": 0.17835855112333793, + "grad_norm": 0.5027433024553368, + "learning_rate": 0.0009435482761441251, + "loss": 2.5856, + "step": 389 + }, + { + "epoch": 0.17881705639614856, + "grad_norm": 0.5317606211926977, + "learning_rate": 0.000943204971676215, + "loss": 2.6271, + "step": 390 + }, + { + "epoch": 0.1792755616689592, + "grad_norm": 0.4744934111570975, + "learning_rate": 0.0009428606893328493, + "loss": 2.4723, + "step": 391 + }, + { + "epoch": 0.17973406694176983, + "grad_norm": 0.5035159541063142, + "learning_rate": 0.0009425154298736432, + "loss": 2.5731, + "step": 392 + }, + { + "epoch": 0.18019257221458046, + "grad_norm": 0.5474980749397559, + "learning_rate": 0.0009421691940603678, + "loss": 2.5151, + "step": 393 + }, + { + "epoch": 0.1806510774873911, + "grad_norm": 0.5265020390601378, + "learning_rate": 0.0009418219826569488, + "loss": 2.6679, + "step": 394 + }, + { + "epoch": 0.18110958276020175, + "grad_norm": 0.4955556423997725, + "learning_rate": 0.0009414737964294635, + "loss": 2.443, + "step": 395 + }, + { + "epoch": 0.1815680880330124, + "grad_norm": 0.5336906159561742, + "learning_rate": 0.000941124636146141, + "loss": 2.5222, + "step": 396 + }, + { + "epoch": 0.18202659330582302, + "grad_norm": 0.5525648068423951, + "learning_rate": 0.0009407745025773589, + "loss": 2.4272, + "step": 397 + }, + { + "epoch": 0.18248509857863365, + "grad_norm": 0.47901930915495705, + "learning_rate": 0.0009404233964956423, + "loss": 2.367, + "step": 398 + }, + { + "epoch": 0.1829436038514443, + "grad_norm": 0.5096002370733514, + "learning_rate": 0.0009400713186756625, + "loss": 2.5025, + "step": 399 + }, + { + "epoch": 0.18340210912425492, + "grad_norm": 0.5058920510980028, + "learning_rate": 0.0009397182698942342, + "loss": 2.5447, + "step": 400 + }, + { + "epoch": 0.18386061439706555, + "grad_norm": 0.4741198712086655, + "learning_rate": 0.0009393642509303149, + "loss": 2.5089, + "step": 401 + }, + { + "epoch": 0.1843191196698762, + "grad_norm": 0.5235058322650097, + "learning_rate": 0.0009390092625650023, + "loss": 2.5307, + "step": 402 + }, + { + "epoch": 0.18477762494268685, + "grad_norm": 0.49000049066192103, + "learning_rate": 0.0009386533055815332, + "loss": 2.477, + "step": 403 + }, + { + "epoch": 0.18523613021549748, + "grad_norm": 0.48660579739076976, + "learning_rate": 0.0009382963807652813, + "loss": 2.4918, + "step": 404 + }, + { + "epoch": 0.1856946354883081, + "grad_norm": 0.49990712989972147, + "learning_rate": 0.000937938488903756, + "loss": 2.5642, + "step": 405 + }, + { + "epoch": 0.18615314076111875, + "grad_norm": 0.521027387092141, + "learning_rate": 0.0009375796307866003, + "loss": 2.5828, + "step": 406 + }, + { + "epoch": 0.18661164603392938, + "grad_norm": 0.5274637297730356, + "learning_rate": 0.0009372198072055888, + "loss": 2.5753, + "step": 407 + }, + { + "epoch": 0.18707015130674004, + "grad_norm": 0.491468628120004, + "learning_rate": 0.0009368590189546268, + "loss": 2.4797, + "step": 408 + }, + { + "epoch": 0.18752865657955067, + "grad_norm": 0.5052236461694741, + "learning_rate": 0.0009364972668297474, + "loss": 2.5388, + "step": 409 + }, + { + "epoch": 0.1879871618523613, + "grad_norm": 0.4995670362551976, + "learning_rate": 0.0009361345516291111, + "loss": 2.4913, + "step": 410 + }, + { + "epoch": 0.18844566712517194, + "grad_norm": 0.5110394784648727, + "learning_rate": 0.0009357708741530024, + "loss": 2.5566, + "step": 411 + }, + { + "epoch": 0.18890417239798257, + "grad_norm": 0.5383997768253838, + "learning_rate": 0.00093540623520383, + "loss": 2.4932, + "step": 412 + }, + { + "epoch": 0.1893626776707932, + "grad_norm": 0.5203985379995043, + "learning_rate": 0.000935040635586123, + "loss": 2.5401, + "step": 413 + }, + { + "epoch": 0.18982118294360384, + "grad_norm": 0.48675392126746386, + "learning_rate": 0.0009346740761065305, + "loss": 2.4838, + "step": 414 + }, + { + "epoch": 0.1902796882164145, + "grad_norm": 0.5207676661589254, + "learning_rate": 0.0009343065575738197, + "loss": 2.4763, + "step": 415 + }, + { + "epoch": 0.19073819348922513, + "grad_norm": 0.4926618133621584, + "learning_rate": 0.0009339380807988733, + "loss": 2.543, + "step": 416 + }, + { + "epoch": 0.19119669876203577, + "grad_norm": 0.468957596992923, + "learning_rate": 0.0009335686465946887, + "loss": 2.5627, + "step": 417 + }, + { + "epoch": 0.1916552040348464, + "grad_norm": 0.5168762265590693, + "learning_rate": 0.0009331982557763754, + "loss": 2.5902, + "step": 418 + }, + { + "epoch": 0.19211370930765703, + "grad_norm": 0.4722151579841745, + "learning_rate": 0.0009328269091611537, + "loss": 2.413, + "step": 419 + }, + { + "epoch": 0.19257221458046767, + "grad_norm": 0.4885933374586764, + "learning_rate": 0.0009324546075683524, + "loss": 2.6201, + "step": 420 + }, + { + "epoch": 0.19303071985327833, + "grad_norm": 0.5134865037420794, + "learning_rate": 0.0009320813518194083, + "loss": 2.5602, + "step": 421 + }, + { + "epoch": 0.19348922512608896, + "grad_norm": 0.48227771543348197, + "learning_rate": 0.0009317071427378624, + "loss": 2.4689, + "step": 422 + }, + { + "epoch": 0.1939477303988996, + "grad_norm": 0.5076659646804255, + "learning_rate": 0.0009313319811493594, + "loss": 2.4871, + "step": 423 + }, + { + "epoch": 0.19440623567171023, + "grad_norm": 0.5235370085024751, + "learning_rate": 0.000930955867881646, + "loss": 2.6126, + "step": 424 + }, + { + "epoch": 0.19486474094452086, + "grad_norm": 0.4986631902013588, + "learning_rate": 0.0009305788037645681, + "loss": 2.4619, + "step": 425 + }, + { + "epoch": 0.1953232462173315, + "grad_norm": 0.4877750767661721, + "learning_rate": 0.0009302007896300697, + "loss": 2.547, + "step": 426 + }, + { + "epoch": 0.19578175149014213, + "grad_norm": 0.506609732069092, + "learning_rate": 0.0009298218263121911, + "loss": 2.5361, + "step": 427 + }, + { + "epoch": 0.19624025676295279, + "grad_norm": 0.49427495184220255, + "learning_rate": 0.0009294419146470668, + "loss": 2.5641, + "step": 428 + }, + { + "epoch": 0.19669876203576342, + "grad_norm": 0.4786905127241538, + "learning_rate": 0.0009290610554729234, + "loss": 2.4823, + "step": 429 + }, + { + "epoch": 0.19715726730857405, + "grad_norm": 0.5054402334030823, + "learning_rate": 0.0009286792496300784, + "loss": 2.5662, + "step": 430 + }, + { + "epoch": 0.19761577258138469, + "grad_norm": 0.4602203102272436, + "learning_rate": 0.0009282964979609379, + "loss": 2.4923, + "step": 431 + }, + { + "epoch": 0.19807427785419532, + "grad_norm": 0.47277276419741593, + "learning_rate": 0.0009279128013099947, + "loss": 2.4734, + "step": 432 + }, + { + "epoch": 0.19853278312700595, + "grad_norm": 0.5260293635979002, + "learning_rate": 0.0009275281605238268, + "loss": 2.5337, + "step": 433 + }, + { + "epoch": 0.19899128839981658, + "grad_norm": 0.504218725186957, + "learning_rate": 0.0009271425764510953, + "loss": 2.5808, + "step": 434 + }, + { + "epoch": 0.19944979367262725, + "grad_norm": 0.45655468400417876, + "learning_rate": 0.0009267560499425423, + "loss": 2.3943, + "step": 435 + }, + { + "epoch": 0.19990829894543788, + "grad_norm": 0.49755237774807953, + "learning_rate": 0.0009263685818509895, + "loss": 2.4422, + "step": 436 + }, + { + "epoch": 0.2003668042182485, + "grad_norm": 0.47903007665259395, + "learning_rate": 0.000925980173031336, + "loss": 2.5345, + "step": 437 + }, + { + "epoch": 0.20082530949105915, + "grad_norm": 0.48968432518044597, + "learning_rate": 0.0009255908243405567, + "loss": 2.4697, + "step": 438 + }, + { + "epoch": 0.20128381476386978, + "grad_norm": 0.5170605021230738, + "learning_rate": 0.0009252005366376996, + "loss": 2.6534, + "step": 439 + }, + { + "epoch": 0.2017423200366804, + "grad_norm": 0.49955925733899287, + "learning_rate": 0.0009248093107838852, + "loss": 2.4986, + "step": 440 + }, + { + "epoch": 0.20220082530949107, + "grad_norm": 0.48099725211379607, + "learning_rate": 0.0009244171476423036, + "loss": 2.4999, + "step": 441 + }, + { + "epoch": 0.2026593305823017, + "grad_norm": 0.47117264319897884, + "learning_rate": 0.0009240240480782129, + "loss": 2.4247, + "step": 442 + }, + { + "epoch": 0.20311783585511234, + "grad_norm": 0.4862692267781655, + "learning_rate": 0.0009236300129589376, + "loss": 2.6208, + "step": 443 + }, + { + "epoch": 0.20357634112792297, + "grad_norm": 0.48645210624711965, + "learning_rate": 0.0009232350431538657, + "loss": 2.616, + "step": 444 + }, + { + "epoch": 0.2040348464007336, + "grad_norm": 0.501730749024782, + "learning_rate": 0.0009228391395344482, + "loss": 2.4988, + "step": 445 + }, + { + "epoch": 0.20449335167354424, + "grad_norm": 0.4824719398211669, + "learning_rate": 0.000922442302974196, + "loss": 2.5737, + "step": 446 + }, + { + "epoch": 0.20495185694635487, + "grad_norm": 0.4924494908435174, + "learning_rate": 0.0009220445343486785, + "loss": 2.5171, + "step": 447 + }, + { + "epoch": 0.20541036221916553, + "grad_norm": 0.4956002021378178, + "learning_rate": 0.0009216458345355217, + "loss": 2.5263, + "step": 448 + }, + { + "epoch": 0.20586886749197617, + "grad_norm": 0.48506007541816504, + "learning_rate": 0.0009212462044144061, + "loss": 2.5925, + "step": 449 + }, + { + "epoch": 0.2063273727647868, + "grad_norm": 0.5054645396651274, + "learning_rate": 0.0009208456448670648, + "loss": 2.5059, + "step": 450 + }, + { + "epoch": 0.20678587803759743, + "grad_norm": 0.4650540654955784, + "learning_rate": 0.0009204441567772816, + "loss": 2.5828, + "step": 451 + }, + { + "epoch": 0.20724438331040806, + "grad_norm": 0.47907477427196954, + "learning_rate": 0.0009200417410308888, + "loss": 2.4955, + "step": 452 + }, + { + "epoch": 0.2077028885832187, + "grad_norm": 0.4908657271079527, + "learning_rate": 0.0009196383985157656, + "loss": 2.5467, + "step": 453 + }, + { + "epoch": 0.20816139385602933, + "grad_norm": 0.5018303615274841, + "learning_rate": 0.000919234130121836, + "loss": 2.4689, + "step": 454 + }, + { + "epoch": 0.20861989912884, + "grad_norm": 0.5256787652444656, + "learning_rate": 0.0009188289367410672, + "loss": 2.5006, + "step": 455 + }, + { + "epoch": 0.20907840440165062, + "grad_norm": 0.4868571822499403, + "learning_rate": 0.0009184228192674666, + "loss": 2.4687, + "step": 456 + }, + { + "epoch": 0.20953690967446126, + "grad_norm": 0.5097656108410421, + "learning_rate": 0.0009180157785970808, + "loss": 2.609, + "step": 457 + }, + { + "epoch": 0.2099954149472719, + "grad_norm": 0.4623857301831229, + "learning_rate": 0.0009176078156279932, + "loss": 2.4651, + "step": 458 + }, + { + "epoch": 0.21045392022008252, + "grad_norm": 0.48025935177781853, + "learning_rate": 0.0009171989312603226, + "loss": 2.4998, + "step": 459 + }, + { + "epoch": 0.21091242549289316, + "grad_norm": 0.47519319393110593, + "learning_rate": 0.0009167891263962202, + "loss": 2.5371, + "step": 460 + }, + { + "epoch": 0.21137093076570382, + "grad_norm": 0.49861970477838014, + "learning_rate": 0.0009163784019398685, + "loss": 2.5295, + "step": 461 + }, + { + "epoch": 0.21182943603851445, + "grad_norm": 0.516172063533237, + "learning_rate": 0.0009159667587974785, + "loss": 2.6199, + "step": 462 + }, + { + "epoch": 0.21228794131132508, + "grad_norm": 0.5031098154320722, + "learning_rate": 0.0009155541978772887, + "loss": 2.4601, + "step": 463 + }, + { + "epoch": 0.21274644658413572, + "grad_norm": 0.5249721312150872, + "learning_rate": 0.0009151407200895625, + "loss": 2.4217, + "step": 464 + }, + { + "epoch": 0.21320495185694635, + "grad_norm": 0.4868288086408481, + "learning_rate": 0.000914726326346586, + "loss": 2.4362, + "step": 465 + }, + { + "epoch": 0.21366345712975698, + "grad_norm": 0.5236789314271078, + "learning_rate": 0.0009143110175626661, + "loss": 2.5798, + "step": 466 + }, + { + "epoch": 0.21412196240256762, + "grad_norm": 0.5688187516117605, + "learning_rate": 0.0009138947946541291, + "loss": 2.6467, + "step": 467 + }, + { + "epoch": 0.21458046767537828, + "grad_norm": 0.5241552720867715, + "learning_rate": 0.0009134776585393181, + "loss": 2.5112, + "step": 468 + }, + { + "epoch": 0.2150389729481889, + "grad_norm": 0.4718207958767481, + "learning_rate": 0.0009130596101385906, + "loss": 2.5663, + "step": 469 + }, + { + "epoch": 0.21549747822099954, + "grad_norm": 0.45837289502732603, + "learning_rate": 0.0009126406503743174, + "loss": 2.5296, + "step": 470 + }, + { + "epoch": 0.21595598349381018, + "grad_norm": 0.4891681976917778, + "learning_rate": 0.0009122207801708802, + "loss": 2.4122, + "step": 471 + }, + { + "epoch": 0.2164144887666208, + "grad_norm": 0.4938666473272722, + "learning_rate": 0.0009118000004546689, + "loss": 2.5494, + "step": 472 + }, + { + "epoch": 0.21687299403943144, + "grad_norm": 0.46168366547395884, + "learning_rate": 0.0009113783121540807, + "loss": 2.468, + "step": 473 + }, + { + "epoch": 0.2173314993122421, + "grad_norm": 0.46627306752910636, + "learning_rate": 0.0009109557161995172, + "loss": 2.5601, + "step": 474 + }, + { + "epoch": 0.21779000458505274, + "grad_norm": 0.4960888642723818, + "learning_rate": 0.0009105322135233828, + "loss": 2.5554, + "step": 475 + }, + { + "epoch": 0.21824850985786337, + "grad_norm": 0.483492314490147, + "learning_rate": 0.0009101078050600821, + "loss": 2.4894, + "step": 476 + }, + { + "epoch": 0.218707015130674, + "grad_norm": 0.47605756265896215, + "learning_rate": 0.0009096824917460186, + "loss": 2.5397, + "step": 477 + }, + { + "epoch": 0.21916552040348464, + "grad_norm": 0.4928496843547731, + "learning_rate": 0.0009092562745195921, + "loss": 2.4888, + "step": 478 + }, + { + "epoch": 0.21962402567629527, + "grad_norm": 0.49304544585062693, + "learning_rate": 0.0009088291543211967, + "loss": 2.5285, + "step": 479 + }, + { + "epoch": 0.2200825309491059, + "grad_norm": 0.4756810085435481, + "learning_rate": 0.0009084011320932188, + "loss": 2.4919, + "step": 480 + }, + { + "epoch": 0.22054103622191656, + "grad_norm": 0.47168254752703787, + "learning_rate": 0.0009079722087800352, + "loss": 2.4515, + "step": 481 + }, + { + "epoch": 0.2209995414947272, + "grad_norm": 0.5060242534700281, + "learning_rate": 0.0009075423853280106, + "loss": 2.5381, + "step": 482 + }, + { + "epoch": 0.22145804676753783, + "grad_norm": 0.5069186780716055, + "learning_rate": 0.0009071116626854958, + "loss": 2.522, + "step": 483 + }, + { + "epoch": 0.22191655204034846, + "grad_norm": 0.5137032419498074, + "learning_rate": 0.0009066800418028256, + "loss": 2.5946, + "step": 484 + }, + { + "epoch": 0.2223750573131591, + "grad_norm": 0.44978040874193215, + "learning_rate": 0.0009062475236323168, + "loss": 2.5222, + "step": 485 + }, + { + "epoch": 0.22283356258596973, + "grad_norm": 0.5123801636534694, + "learning_rate": 0.0009058141091282656, + "loss": 2.5051, + "step": 486 + }, + { + "epoch": 0.22329206785878036, + "grad_norm": 0.49180042021846904, + "learning_rate": 0.0009053797992469461, + "loss": 2.5338, + "step": 487 + }, + { + "epoch": 0.22375057313159102, + "grad_norm": 0.4499977158117602, + "learning_rate": 0.0009049445949466078, + "loss": 2.4958, + "step": 488 + }, + { + "epoch": 0.22420907840440166, + "grad_norm": 0.46830147653004583, + "learning_rate": 0.0009045084971874737, + "loss": 2.4683, + "step": 489 + }, + { + "epoch": 0.2246675836772123, + "grad_norm": 0.5345617149541109, + "learning_rate": 0.0009040715069317382, + "loss": 2.4314, + "step": 490 + }, + { + "epoch": 0.22512608895002292, + "grad_norm": 0.48531910242798837, + "learning_rate": 0.0009036336251435648, + "loss": 2.4927, + "step": 491 + }, + { + "epoch": 0.22558459422283356, + "grad_norm": 0.5065144654504833, + "learning_rate": 0.0009031948527890839, + "loss": 2.4522, + "step": 492 + }, + { + "epoch": 0.2260430994956442, + "grad_norm": 0.4886650258113338, + "learning_rate": 0.000902755190836391, + "loss": 2.5254, + "step": 493 + }, + { + "epoch": 0.22650160476845485, + "grad_norm": 0.5158467030644189, + "learning_rate": 0.0009023146402555442, + "loss": 2.5143, + "step": 494 + }, + { + "epoch": 0.22696011004126548, + "grad_norm": 0.4833439314582239, + "learning_rate": 0.0009018732020185624, + "loss": 2.489, + "step": 495 + }, + { + "epoch": 0.22741861531407612, + "grad_norm": 0.4708719037510092, + "learning_rate": 0.0009014308770994235, + "loss": 2.5344, + "step": 496 + }, + { + "epoch": 0.22787712058688675, + "grad_norm": 0.4959434944423333, + "learning_rate": 0.0009009876664740605, + "loss": 2.5997, + "step": 497 + }, + { + "epoch": 0.22833562585969738, + "grad_norm": 0.4944899536599611, + "learning_rate": 0.0009005435711203618, + "loss": 2.5955, + "step": 498 + }, + { + "epoch": 0.22879413113250802, + "grad_norm": 0.4753069257460288, + "learning_rate": 0.000900098592018167, + "loss": 2.5362, + "step": 499 + }, + { + "epoch": 0.22925263640531865, + "grad_norm": 0.45285756315993814, + "learning_rate": 0.0008996527301492663, + "loss": 2.4092, + "step": 500 + }, + { + "epoch": 0.2297111416781293, + "grad_norm": 0.5070472804479449, + "learning_rate": 0.0008992059864973972, + "loss": 2.4733, + "step": 501 + }, + { + "epoch": 0.23016964695093994, + "grad_norm": 0.4900438819145011, + "learning_rate": 0.0008987583620482427, + "loss": 2.4073, + "step": 502 + }, + { + "epoch": 0.23062815222375058, + "grad_norm": 0.4987259692524487, + "learning_rate": 0.0008983098577894292, + "loss": 2.5047, + "step": 503 + }, + { + "epoch": 0.2310866574965612, + "grad_norm": 0.5000342607867347, + "learning_rate": 0.0008978604747105246, + "loss": 2.4661, + "step": 504 + }, + { + "epoch": 0.23154516276937184, + "grad_norm": 0.49299691384483413, + "learning_rate": 0.0008974102138030354, + "loss": 2.63, + "step": 505 + }, + { + "epoch": 0.23200366804218248, + "grad_norm": 0.4785645532894974, + "learning_rate": 0.000896959076060405, + "loss": 2.4694, + "step": 506 + }, + { + "epoch": 0.2324621733149931, + "grad_norm": 0.4475089836425941, + "learning_rate": 0.0008965070624780116, + "loss": 2.4682, + "step": 507 + }, + { + "epoch": 0.23292067858780377, + "grad_norm": 0.4421977744050373, + "learning_rate": 0.0008960541740531658, + "loss": 2.45, + "step": 508 + }, + { + "epoch": 0.2333791838606144, + "grad_norm": 0.4448303377691624, + "learning_rate": 0.0008956004117851083, + "loss": 2.4004, + "step": 509 + }, + { + "epoch": 0.23383768913342504, + "grad_norm": 0.48840138817071044, + "learning_rate": 0.0008951457766750079, + "loss": 2.6064, + "step": 510 + }, + { + "epoch": 0.23429619440623567, + "grad_norm": 0.4857804151830313, + "learning_rate": 0.0008946902697259593, + "loss": 2.4529, + "step": 511 + }, + { + "epoch": 0.2347546996790463, + "grad_norm": 0.4655407240464954, + "learning_rate": 0.0008942338919429805, + "loss": 2.4704, + "step": 512 + }, + { + "epoch": 0.23521320495185694, + "grad_norm": 0.4711130956691561, + "learning_rate": 0.0008937766443330113, + "loss": 2.4825, + "step": 513 + }, + { + "epoch": 0.2356717102246676, + "grad_norm": 0.5137635790986398, + "learning_rate": 0.0008933185279049103, + "loss": 2.5632, + "step": 514 + }, + { + "epoch": 0.23613021549747823, + "grad_norm": 0.456861808689729, + "learning_rate": 0.0008928595436694532, + "loss": 2.4475, + "step": 515 + }, + { + "epoch": 0.23658872077028886, + "grad_norm": 0.47150434049377193, + "learning_rate": 0.0008923996926393305, + "loss": 2.5082, + "step": 516 + }, + { + "epoch": 0.2370472260430995, + "grad_norm": 0.45688816753995054, + "learning_rate": 0.0008919389758291449, + "loss": 2.3896, + "step": 517 + }, + { + "epoch": 0.23750573131591013, + "grad_norm": 0.46546466439239775, + "learning_rate": 0.0008914773942554098, + "loss": 2.5071, + "step": 518 + }, + { + "epoch": 0.23796423658872076, + "grad_norm": 0.48271626879933005, + "learning_rate": 0.000891014948936546, + "loss": 2.5327, + "step": 519 + }, + { + "epoch": 0.2384227418615314, + "grad_norm": 0.4656591548989872, + "learning_rate": 0.0008905516408928804, + "loss": 2.521, + "step": 520 + }, + { + "epoch": 0.23888124713434206, + "grad_norm": 0.5002764569926602, + "learning_rate": 0.0008900874711466434, + "loss": 2.5655, + "step": 521 + }, + { + "epoch": 0.2393397524071527, + "grad_norm": 0.464908577551842, + "learning_rate": 0.0008896224407219666, + "loss": 2.4253, + "step": 522 + }, + { + "epoch": 0.23979825767996332, + "grad_norm": 0.46471608235961614, + "learning_rate": 0.0008891565506448804, + "loss": 2.4385, + "step": 523 + }, + { + "epoch": 0.24025676295277396, + "grad_norm": 0.4713636256532519, + "learning_rate": 0.0008886898019433122, + "loss": 2.4079, + "step": 524 + }, + { + "epoch": 0.2407152682255846, + "grad_norm": 0.5163834858351678, + "learning_rate": 0.0008882221956470836, + "loss": 2.5546, + "step": 525 + }, + { + "epoch": 0.24117377349839522, + "grad_norm": 0.4926257883157003, + "learning_rate": 0.0008877537327879086, + "loss": 2.4721, + "step": 526 + }, + { + "epoch": 0.24163227877120586, + "grad_norm": 0.4831101979066039, + "learning_rate": 0.0008872844143993908, + "loss": 2.4242, + "step": 527 + }, + { + "epoch": 0.24209078404401652, + "grad_norm": 0.48791318420179747, + "learning_rate": 0.0008868142415170218, + "loss": 2.4124, + "step": 528 + }, + { + "epoch": 0.24254928931682715, + "grad_norm": 0.48121655670376284, + "learning_rate": 0.0008863432151781781, + "loss": 2.5271, + "step": 529 + }, + { + "epoch": 0.24300779458963778, + "grad_norm": 0.4901381650874081, + "learning_rate": 0.0008858713364221195, + "loss": 2.4724, + "step": 530 + }, + { + "epoch": 0.24346629986244842, + "grad_norm": 0.48929715331506834, + "learning_rate": 0.0008853986062899868, + "loss": 2.5055, + "step": 531 + }, + { + "epoch": 0.24392480513525905, + "grad_norm": 0.4644637465465594, + "learning_rate": 0.0008849250258247986, + "loss": 2.5749, + "step": 532 + }, + { + "epoch": 0.24438331040806968, + "grad_norm": 0.4464974647181807, + "learning_rate": 0.0008844505960714503, + "loss": 2.3872, + "step": 533 + }, + { + "epoch": 0.24484181568088034, + "grad_norm": 0.4887745418253648, + "learning_rate": 0.0008839753180767108, + "loss": 2.5223, + "step": 534 + }, + { + "epoch": 0.24530032095369098, + "grad_norm": 0.486997870425633, + "learning_rate": 0.0008834991928892204, + "loss": 2.5835, + "step": 535 + }, + { + "epoch": 0.2457588262265016, + "grad_norm": 0.4721187294174993, + "learning_rate": 0.000883022221559489, + "loss": 2.5126, + "step": 536 + }, + { + "epoch": 0.24621733149931224, + "grad_norm": 0.4916578138565609, + "learning_rate": 0.0008825444051398934, + "loss": 2.5902, + "step": 537 + }, + { + "epoch": 0.24667583677212288, + "grad_norm": 0.475632593286815, + "learning_rate": 0.0008820657446846745, + "loss": 2.5482, + "step": 538 + }, + { + "epoch": 0.2471343420449335, + "grad_norm": 0.43822194408333887, + "learning_rate": 0.000881586241249936, + "loss": 2.3812, + "step": 539 + }, + { + "epoch": 0.24759284731774414, + "grad_norm": 0.46179608202803907, + "learning_rate": 0.0008811058958936411, + "loss": 2.4808, + "step": 540 + }, + { + "epoch": 0.2480513525905548, + "grad_norm": 0.46988708619216435, + "learning_rate": 0.000880624709675611, + "loss": 2.3701, + "step": 541 + }, + { + "epoch": 0.24850985786336544, + "grad_norm": 0.4458868532768844, + "learning_rate": 0.000880142683657522, + "loss": 2.5345, + "step": 542 + }, + { + "epoch": 0.24896836313617607, + "grad_norm": 0.4459694504636457, + "learning_rate": 0.0008796598189029029, + "loss": 2.4297, + "step": 543 + }, + { + "epoch": 0.2494268684089867, + "grad_norm": 0.48838754760513015, + "learning_rate": 0.0008791761164771338, + "loss": 2.4641, + "step": 544 + }, + { + "epoch": 0.24988537368179733, + "grad_norm": 0.4760913483888067, + "learning_rate": 0.0008786915774474424, + "loss": 2.5178, + "step": 545 + }, + { + "epoch": 0.25034387895460797, + "grad_norm": 0.502576516042263, + "learning_rate": 0.0008782062028829027, + "loss": 2.6284, + "step": 546 + }, + { + "epoch": 0.25080238422741863, + "grad_norm": 0.46154999347826386, + "learning_rate": 0.0008777199938544318, + "loss": 2.4388, + "step": 547 + }, + { + "epoch": 0.25126088950022923, + "grad_norm": 0.44694339058596894, + "learning_rate": 0.0008772329514347883, + "loss": 2.4387, + "step": 548 + }, + { + "epoch": 0.2517193947730399, + "grad_norm": 0.4899117789967993, + "learning_rate": 0.0008767450766985694, + "loss": 2.5931, + "step": 549 + }, + { + "epoch": 0.2521779000458505, + "grad_norm": 0.48035626238288576, + "learning_rate": 0.0008762563707222086, + "loss": 2.4594, + "step": 550 + }, + { + "epoch": 0.25263640531866116, + "grad_norm": 0.4751174428647926, + "learning_rate": 0.0008757668345839738, + "loss": 2.5322, + "step": 551 + }, + { + "epoch": 0.2530949105914718, + "grad_norm": 0.4633660543305586, + "learning_rate": 0.0008752764693639638, + "loss": 2.5045, + "step": 552 + }, + { + "epoch": 0.2535534158642824, + "grad_norm": 0.49349497443129703, + "learning_rate": 0.0008747852761441078, + "loss": 2.606, + "step": 553 + }, + { + "epoch": 0.2540119211370931, + "grad_norm": 0.4815679869267261, + "learning_rate": 0.0008742932560081607, + "loss": 2.5469, + "step": 554 + }, + { + "epoch": 0.2544704264099037, + "grad_norm": 0.4540596927290549, + "learning_rate": 0.0008738004100417025, + "loss": 2.5154, + "step": 555 + }, + { + "epoch": 0.25492893168271435, + "grad_norm": 0.47059419269751196, + "learning_rate": 0.0008733067393321355, + "loss": 2.534, + "step": 556 + }, + { + "epoch": 0.255387436955525, + "grad_norm": 0.4530945384334566, + "learning_rate": 0.000872812244968681, + "loss": 2.5327, + "step": 557 + }, + { + "epoch": 0.2558459422283356, + "grad_norm": 0.4430408191985952, + "learning_rate": 0.0008723169280423783, + "loss": 2.567, + "step": 558 + }, + { + "epoch": 0.2563044475011463, + "grad_norm": 0.4622300241325886, + "learning_rate": 0.0008718207896460811, + "loss": 2.5124, + "step": 559 + }, + { + "epoch": 0.2567629527739569, + "grad_norm": 0.4500233423636964, + "learning_rate": 0.0008713238308744557, + "loss": 2.473, + "step": 560 + }, + { + "epoch": 0.25722145804676755, + "grad_norm": 0.4906224687037954, + "learning_rate": 0.0008708260528239789, + "loss": 2.4349, + "step": 561 + }, + { + "epoch": 0.25767996331957815, + "grad_norm": 0.46870879424301615, + "learning_rate": 0.000870327456592934, + "loss": 2.4509, + "step": 562 + }, + { + "epoch": 0.2581384685923888, + "grad_norm": 0.48270567951459353, + "learning_rate": 0.0008698280432814107, + "loss": 2.5367, + "step": 563 + }, + { + "epoch": 0.2585969738651995, + "grad_norm": 0.49534379788021216, + "learning_rate": 0.000869327813991301, + "loss": 2.5553, + "step": 564 + }, + { + "epoch": 0.2590554791380101, + "grad_norm": 0.4758308439018621, + "learning_rate": 0.0008688267698262971, + "loss": 2.5314, + "step": 565 + }, + { + "epoch": 0.25951398441082074, + "grad_norm": 0.4396420242370547, + "learning_rate": 0.0008683249118918894, + "loss": 2.4994, + "step": 566 + }, + { + "epoch": 0.25997248968363135, + "grad_norm": 0.49645956277186476, + "learning_rate": 0.0008678222412953637, + "loss": 2.4951, + "step": 567 + }, + { + "epoch": 0.260430994956442, + "grad_norm": 0.4681969056805642, + "learning_rate": 0.0008673187591457987, + "loss": 2.6168, + "step": 568 + }, + { + "epoch": 0.2608895002292526, + "grad_norm": 0.47335853008535467, + "learning_rate": 0.0008668144665540639, + "loss": 2.5262, + "step": 569 + }, + { + "epoch": 0.2613480055020633, + "grad_norm": 0.45770215854602, + "learning_rate": 0.0008663093646328167, + "loss": 2.5386, + "step": 570 + }, + { + "epoch": 0.26180651077487394, + "grad_norm": 0.48867036121120216, + "learning_rate": 0.0008658034544965003, + "loss": 2.3764, + "step": 571 + }, + { + "epoch": 0.26226501604768454, + "grad_norm": 0.49573805285221906, + "learning_rate": 0.0008652967372613412, + "loss": 2.4647, + "step": 572 + }, + { + "epoch": 0.2627235213204952, + "grad_norm": 0.44703243404709625, + "learning_rate": 0.0008647892140453466, + "loss": 2.3894, + "step": 573 + }, + { + "epoch": 0.2631820265933058, + "grad_norm": 0.42943025537008583, + "learning_rate": 0.0008642808859683021, + "loss": 2.4149, + "step": 574 + }, + { + "epoch": 0.26364053186611647, + "grad_norm": 0.4767511909811112, + "learning_rate": 0.0008637717541517689, + "loss": 2.5284, + "step": 575 + }, + { + "epoch": 0.2640990371389271, + "grad_norm": 0.5066885648938176, + "learning_rate": 0.0008632618197190816, + "loss": 2.5514, + "step": 576 + }, + { + "epoch": 0.26455754241173773, + "grad_norm": 0.49866712459980167, + "learning_rate": 0.0008627510837953458, + "loss": 2.5333, + "step": 577 + }, + { + "epoch": 0.2650160476845484, + "grad_norm": 0.47831326993911194, + "learning_rate": 0.0008622395475074355, + "loss": 2.5342, + "step": 578 + }, + { + "epoch": 0.265474552957359, + "grad_norm": 0.5027598971722053, + "learning_rate": 0.0008617272119839903, + "loss": 2.4994, + "step": 579 + }, + { + "epoch": 0.26593305823016966, + "grad_norm": 0.48309023549523095, + "learning_rate": 0.0008612140783554136, + "loss": 2.5933, + "step": 580 + }, + { + "epoch": 0.26639156350298027, + "grad_norm": 0.45777696463040435, + "learning_rate": 0.0008607001477538696, + "loss": 2.4668, + "step": 581 + }, + { + "epoch": 0.2668500687757909, + "grad_norm": 0.4806634122882089, + "learning_rate": 0.0008601854213132807, + "loss": 2.5495, + "step": 582 + }, + { + "epoch": 0.26730857404860153, + "grad_norm": 0.456817832099134, + "learning_rate": 0.0008596699001693256, + "loss": 2.4875, + "step": 583 + }, + { + "epoch": 0.2677670793214122, + "grad_norm": 0.4820827054891331, + "learning_rate": 0.000859153585459436, + "loss": 2.4579, + "step": 584 + }, + { + "epoch": 0.26822558459422285, + "grad_norm": 0.44703973819333404, + "learning_rate": 0.0008586364783227949, + "loss": 2.4974, + "step": 585 + }, + { + "epoch": 0.26868408986703346, + "grad_norm": 0.44317390852103417, + "learning_rate": 0.0008581185799003332, + "loss": 2.5259, + "step": 586 + }, + { + "epoch": 0.2691425951398441, + "grad_norm": 0.4790895703327621, + "learning_rate": 0.0008575998913347283, + "loss": 2.5914, + "step": 587 + }, + { + "epoch": 0.2696011004126547, + "grad_norm": 0.44194519703552215, + "learning_rate": 0.0008570804137704004, + "loss": 2.4778, + "step": 588 + }, + { + "epoch": 0.2700596056854654, + "grad_norm": 0.4538620376044402, + "learning_rate": 0.0008565601483535108, + "loss": 2.4452, + "step": 589 + }, + { + "epoch": 0.27051811095827605, + "grad_norm": 0.44401889850722487, + "learning_rate": 0.0008560390962319591, + "loss": 2.4684, + "step": 590 + }, + { + "epoch": 0.27097661623108665, + "grad_norm": 0.46969771519665665, + "learning_rate": 0.0008555172585553804, + "loss": 2.5228, + "step": 591 + }, + { + "epoch": 0.2714351215038973, + "grad_norm": 0.45496365817435536, + "learning_rate": 0.0008549946364751435, + "loss": 2.4282, + "step": 592 + }, + { + "epoch": 0.2718936267767079, + "grad_norm": 0.4427858980976477, + "learning_rate": 0.0008544712311443475, + "loss": 2.505, + "step": 593 + }, + { + "epoch": 0.2723521320495186, + "grad_norm": 0.482965716395204, + "learning_rate": 0.0008539470437178196, + "loss": 2.4674, + "step": 594 + }, + { + "epoch": 0.2728106373223292, + "grad_norm": 0.41754134958412414, + "learning_rate": 0.000853422075352113, + "loss": 2.4879, + "step": 595 + }, + { + "epoch": 0.27326914259513985, + "grad_norm": 0.44276917056512627, + "learning_rate": 0.0008528963272055035, + "loss": 2.5304, + "step": 596 + }, + { + "epoch": 0.2737276478679505, + "grad_norm": 0.4443401726845757, + "learning_rate": 0.0008523698004379877, + "loss": 2.4556, + "step": 597 + }, + { + "epoch": 0.2741861531407611, + "grad_norm": 0.47022279022608665, + "learning_rate": 0.00085184249621128, + "loss": 2.4117, + "step": 598 + }, + { + "epoch": 0.2746446584135718, + "grad_norm": 0.48031231346457154, + "learning_rate": 0.0008513144156888101, + "loss": 2.4757, + "step": 599 + }, + { + "epoch": 0.2751031636863824, + "grad_norm": 0.47658617735694847, + "learning_rate": 0.0008507855600357207, + "loss": 2.5712, + "step": 600 + }, + { + "epoch": 0.27556166895919304, + "grad_norm": 0.4570667706954567, + "learning_rate": 0.0008502559304188644, + "loss": 2.4297, + "step": 601 + }, + { + "epoch": 0.27602017423200365, + "grad_norm": 0.5260633609576062, + "learning_rate": 0.0008497255280068019, + "loss": 2.5375, + "step": 602 + }, + { + "epoch": 0.2764786795048143, + "grad_norm": 0.47863285224719493, + "learning_rate": 0.0008491943539697986, + "loss": 2.491, + "step": 603 + }, + { + "epoch": 0.27693718477762497, + "grad_norm": 0.4665708194024154, + "learning_rate": 0.0008486624094798226, + "loss": 2.4507, + "step": 604 + }, + { + "epoch": 0.2773956900504356, + "grad_norm": 0.4696223902830353, + "learning_rate": 0.0008481296957105417, + "loss": 2.5268, + "step": 605 + }, + { + "epoch": 0.27785419532324623, + "grad_norm": 0.44408355949567724, + "learning_rate": 0.0008475962138373213, + "loss": 2.5693, + "step": 606 + }, + { + "epoch": 0.27831270059605684, + "grad_norm": 0.4661097681867018, + "learning_rate": 0.0008470619650372211, + "loss": 2.4654, + "step": 607 + }, + { + "epoch": 0.2787712058688675, + "grad_norm": 0.475323111201629, + "learning_rate": 0.0008465269504889934, + "loss": 2.4617, + "step": 608 + }, + { + "epoch": 0.2792297111416781, + "grad_norm": 0.4383136302118051, + "learning_rate": 0.0008459911713730799, + "loss": 2.4869, + "step": 609 + }, + { + "epoch": 0.27968821641448877, + "grad_norm": 0.4668198740951539, + "learning_rate": 0.0008454546288716089, + "loss": 2.5008, + "step": 610 + }, + { + "epoch": 0.2801467216872994, + "grad_norm": 0.45218775812016787, + "learning_rate": 0.0008449173241683935, + "loss": 2.4415, + "step": 611 + }, + { + "epoch": 0.28060522696011003, + "grad_norm": 0.4901374525409381, + "learning_rate": 0.0008443792584489281, + "loss": 2.6332, + "step": 612 + }, + { + "epoch": 0.2810637322329207, + "grad_norm": 0.44993173471362513, + "learning_rate": 0.0008438404329003863, + "loss": 2.4797, + "step": 613 + }, + { + "epoch": 0.2815222375057313, + "grad_norm": 0.4759098507718769, + "learning_rate": 0.0008433008487116183, + "loss": 2.4447, + "step": 614 + }, + { + "epoch": 0.28198074277854196, + "grad_norm": 0.4801141156513515, + "learning_rate": 0.0008427605070731481, + "loss": 2.405, + "step": 615 + }, + { + "epoch": 0.28243924805135256, + "grad_norm": 0.4528558986396663, + "learning_rate": 0.0008422194091771708, + "loss": 2.4736, + "step": 616 + }, + { + "epoch": 0.2828977533241632, + "grad_norm": 0.44219508474710273, + "learning_rate": 0.0008416775562175503, + "loss": 2.5298, + "step": 617 + }, + { + "epoch": 0.2833562585969739, + "grad_norm": 0.43754189793032655, + "learning_rate": 0.000841134949389816, + "loss": 2.4664, + "step": 618 + }, + { + "epoch": 0.2838147638697845, + "grad_norm": 0.4686457065947725, + "learning_rate": 0.0008405915898911611, + "loss": 2.5262, + "step": 619 + }, + { + "epoch": 0.28427326914259515, + "grad_norm": 0.45014248391122164, + "learning_rate": 0.0008400474789204396, + "loss": 2.5047, + "step": 620 + }, + { + "epoch": 0.28473177441540576, + "grad_norm": 0.4329805763911945, + "learning_rate": 0.0008395026176781626, + "loss": 2.4245, + "step": 621 + }, + { + "epoch": 0.2851902796882164, + "grad_norm": 0.469033060439781, + "learning_rate": 0.0008389570073664976, + "loss": 2.5145, + "step": 622 + }, + { + "epoch": 0.285648784961027, + "grad_norm": 0.47267166321426796, + "learning_rate": 0.0008384106491892642, + "loss": 2.5949, + "step": 623 + }, + { + "epoch": 0.2861072902338377, + "grad_norm": 0.44993018951589375, + "learning_rate": 0.0008378635443519327, + "loss": 2.5001, + "step": 624 + }, + { + "epoch": 0.28656579550664835, + "grad_norm": 0.45279167561876466, + "learning_rate": 0.0008373156940616199, + "loss": 2.4768, + "step": 625 + }, + { + "epoch": 0.28702430077945895, + "grad_norm": 0.45491627139761437, + "learning_rate": 0.0008367670995270882, + "loss": 2.5553, + "step": 626 + }, + { + "epoch": 0.2874828060522696, + "grad_norm": 0.43474344517640406, + "learning_rate": 0.0008362177619587416, + "loss": 2.4531, + "step": 627 + }, + { + "epoch": 0.2879413113250802, + "grad_norm": 0.4541004175766488, + "learning_rate": 0.0008356676825686238, + "loss": 2.4549, + "step": 628 + }, + { + "epoch": 0.2883998165978909, + "grad_norm": 0.463858397802096, + "learning_rate": 0.0008351168625704147, + "loss": 2.5975, + "step": 629 + }, + { + "epoch": 0.28885832187070154, + "grad_norm": 0.4537890702397275, + "learning_rate": 0.0008345653031794292, + "loss": 2.5208, + "step": 630 + }, + { + "epoch": 0.28931682714351215, + "grad_norm": 0.44011804848118924, + "learning_rate": 0.0008340130056126125, + "loss": 2.4055, + "step": 631 + }, + { + "epoch": 0.2897753324163228, + "grad_norm": 0.4157931947387841, + "learning_rate": 0.0008334599710885394, + "loss": 2.4771, + "step": 632 + }, + { + "epoch": 0.2902338376891334, + "grad_norm": 0.46535454987780894, + "learning_rate": 0.0008329062008274098, + "loss": 2.397, + "step": 633 + }, + { + "epoch": 0.2906923429619441, + "grad_norm": 0.449645113746727, + "learning_rate": 0.000832351696051048, + "loss": 2.4831, + "step": 634 + }, + { + "epoch": 0.2911508482347547, + "grad_norm": 0.4239995605303684, + "learning_rate": 0.000831796457982898, + "loss": 2.487, + "step": 635 + }, + { + "epoch": 0.29160935350756534, + "grad_norm": 0.4579079558278341, + "learning_rate": 0.0008312404878480222, + "loss": 2.4957, + "step": 636 + }, + { + "epoch": 0.292067858780376, + "grad_norm": 0.46614408886750003, + "learning_rate": 0.0008306837868730979, + "loss": 2.5016, + "step": 637 + }, + { + "epoch": 0.2925263640531866, + "grad_norm": 0.47284039789874993, + "learning_rate": 0.0008301263562864152, + "loss": 2.5188, + "step": 638 + }, + { + "epoch": 0.29298486932599727, + "grad_norm": 0.4884750072381276, + "learning_rate": 0.0008295681973178737, + "loss": 2.5578, + "step": 639 + }, + { + "epoch": 0.29344337459880787, + "grad_norm": 0.44703936007232076, + "learning_rate": 0.0008290093111989804, + "loss": 2.4807, + "step": 640 + }, + { + "epoch": 0.29390187987161853, + "grad_norm": 0.4528790939173257, + "learning_rate": 0.0008284496991628465, + "loss": 2.445, + "step": 641 + }, + { + "epoch": 0.29436038514442914, + "grad_norm": 0.4518089115195415, + "learning_rate": 0.0008278893624441847, + "loss": 2.4948, + "step": 642 + }, + { + "epoch": 0.2948188904172398, + "grad_norm": 0.45158441707566055, + "learning_rate": 0.000827328302279307, + "loss": 2.4139, + "step": 643 + }, + { + "epoch": 0.29527739569005046, + "grad_norm": 0.45828181430139403, + "learning_rate": 0.0008267665199061211, + "loss": 2.4489, + "step": 644 + }, + { + "epoch": 0.29573590096286106, + "grad_norm": 0.4495904975485981, + "learning_rate": 0.0008262040165641288, + "loss": 2.5082, + "step": 645 + }, + { + "epoch": 0.2961944062356717, + "grad_norm": 0.4408097873719146, + "learning_rate": 0.0008256407934944219, + "loss": 2.4688, + "step": 646 + }, + { + "epoch": 0.29665291150848233, + "grad_norm": 0.43453304514052055, + "learning_rate": 0.0008250768519396807, + "loss": 2.42, + "step": 647 + }, + { + "epoch": 0.297111416781293, + "grad_norm": 0.43823358713595467, + "learning_rate": 0.0008245121931441706, + "loss": 2.518, + "step": 648 + }, + { + "epoch": 0.2975699220541036, + "grad_norm": 0.4491538156349741, + "learning_rate": 0.0008239468183537393, + "loss": 2.5543, + "step": 649 + }, + { + "epoch": 0.29802842732691426, + "grad_norm": 0.4827072505095177, + "learning_rate": 0.0008233807288158146, + "loss": 2.4635, + "step": 650 + }, + { + "epoch": 0.2984869325997249, + "grad_norm": 0.4402995549038372, + "learning_rate": 0.0008228139257794012, + "loss": 2.5377, + "step": 651 + }, + { + "epoch": 0.2989454378725355, + "grad_norm": 0.44610974587749447, + "learning_rate": 0.0008222464104950778, + "loss": 2.4941, + "step": 652 + }, + { + "epoch": 0.2994039431453462, + "grad_norm": 0.4548220293753913, + "learning_rate": 0.000821678184214995, + "loss": 2.4543, + "step": 653 + }, + { + "epoch": 0.2998624484181568, + "grad_norm": 0.4645910214659261, + "learning_rate": 0.0008211092481928716, + "loss": 2.4288, + "step": 654 + }, + { + "epoch": 0.30032095369096745, + "grad_norm": 0.47286054756494195, + "learning_rate": 0.0008205396036839927, + "loss": 2.5357, + "step": 655 + }, + { + "epoch": 0.30077945896377806, + "grad_norm": 0.4750106894629853, + "learning_rate": 0.0008199692519452069, + "loss": 2.5972, + "step": 656 + }, + { + "epoch": 0.3012379642365887, + "grad_norm": 0.4991595326219104, + "learning_rate": 0.0008193981942349224, + "loss": 2.4304, + "step": 657 + }, + { + "epoch": 0.3016964695093994, + "grad_norm": 0.46363856278098975, + "learning_rate": 0.0008188264318131056, + "loss": 2.4605, + "step": 658 + }, + { + "epoch": 0.30215497478221, + "grad_norm": 0.45375231289610723, + "learning_rate": 0.0008182539659412776, + "loss": 2.4317, + "step": 659 + }, + { + "epoch": 0.30261348005502064, + "grad_norm": 0.507671084566265, + "learning_rate": 0.0008176807978825118, + "loss": 2.4357, + "step": 660 + }, + { + "epoch": 0.30307198532783125, + "grad_norm": 0.47331292961484694, + "learning_rate": 0.0008171069289014306, + "loss": 2.4499, + "step": 661 + }, + { + "epoch": 0.3035304906006419, + "grad_norm": 0.4494049774687332, + "learning_rate": 0.0008165323602642028, + "loss": 2.5682, + "step": 662 + }, + { + "epoch": 0.30398899587345257, + "grad_norm": 0.43472224751320016, + "learning_rate": 0.0008159570932385414, + "loss": 2.4843, + "step": 663 + }, + { + "epoch": 0.3044475011462632, + "grad_norm": 0.48294812767700684, + "learning_rate": 0.0008153811290936999, + "loss": 2.53, + "step": 664 + }, + { + "epoch": 0.30490600641907384, + "grad_norm": 0.46018714490398155, + "learning_rate": 0.0008148044691004698, + "loss": 2.5179, + "step": 665 + }, + { + "epoch": 0.30536451169188444, + "grad_norm": 0.4472487872819034, + "learning_rate": 0.0008142271145311783, + "loss": 2.513, + "step": 666 + }, + { + "epoch": 0.3058230169646951, + "grad_norm": 0.506771428765824, + "learning_rate": 0.000813649066659685, + "loss": 2.51, + "step": 667 + }, + { + "epoch": 0.3062815222375057, + "grad_norm": 0.4370475871534568, + "learning_rate": 0.0008130703267613787, + "loss": 2.43, + "step": 668 + }, + { + "epoch": 0.30674002751031637, + "grad_norm": 0.49673770473061146, + "learning_rate": 0.0008124908961131759, + "loss": 2.5203, + "step": 669 + }, + { + "epoch": 0.30719853278312703, + "grad_norm": 0.4572579453842779, + "learning_rate": 0.0008119107759935163, + "loss": 2.4385, + "step": 670 + }, + { + "epoch": 0.30765703805593764, + "grad_norm": 0.4628934530682827, + "learning_rate": 0.0008113299676823615, + "loss": 2.4973, + "step": 671 + }, + { + "epoch": 0.3081155433287483, + "grad_norm": 0.48733097494772515, + "learning_rate": 0.0008107484724611911, + "loss": 2.4566, + "step": 672 + }, + { + "epoch": 0.3085740486015589, + "grad_norm": 0.44796597576463193, + "learning_rate": 0.0008101662916130006, + "loss": 2.5119, + "step": 673 + }, + { + "epoch": 0.30903255387436956, + "grad_norm": 0.44028301635624845, + "learning_rate": 0.0008095834264222979, + "loss": 2.471, + "step": 674 + }, + { + "epoch": 0.30949105914718017, + "grad_norm": 0.42837448076567547, + "learning_rate": 0.0008089998781751009, + "loss": 2.5897, + "step": 675 + }, + { + "epoch": 0.30994956441999083, + "grad_norm": 0.412655922200955, + "learning_rate": 0.0008084156481589349, + "loss": 2.5103, + "step": 676 + }, + { + "epoch": 0.3104080696928015, + "grad_norm": 0.4180177766799232, + "learning_rate": 0.0008078307376628291, + "loss": 2.4547, + "step": 677 + }, + { + "epoch": 0.3108665749656121, + "grad_norm": 0.45798655233878793, + "learning_rate": 0.0008072451479773143, + "loss": 2.5544, + "step": 678 + }, + { + "epoch": 0.31132508023842276, + "grad_norm": 0.4173066878996162, + "learning_rate": 0.0008066588803944195, + "loss": 2.4604, + "step": 679 + }, + { + "epoch": 0.31178358551123336, + "grad_norm": 0.42848964980712995, + "learning_rate": 0.0008060719362076697, + "loss": 2.5339, + "step": 680 + }, + { + "epoch": 0.312242090784044, + "grad_norm": 0.4247430078353849, + "learning_rate": 0.0008054843167120826, + "loss": 2.454, + "step": 681 + }, + { + "epoch": 0.31270059605685463, + "grad_norm": 0.4301756175232488, + "learning_rate": 0.0008048960232041663, + "loss": 2.5367, + "step": 682 + }, + { + "epoch": 0.3131591013296653, + "grad_norm": 0.4758080283605412, + "learning_rate": 0.0008043070569819153, + "loss": 2.4213, + "step": 683 + }, + { + "epoch": 0.31361760660247595, + "grad_norm": 0.43183342750903375, + "learning_rate": 0.0008037174193448089, + "loss": 2.4872, + "step": 684 + }, + { + "epoch": 0.31407611187528656, + "grad_norm": 0.45283252684364733, + "learning_rate": 0.0008031271115938077, + "loss": 2.5004, + "step": 685 + }, + { + "epoch": 0.3145346171480972, + "grad_norm": 0.4904054380313948, + "learning_rate": 0.0008025361350313505, + "loss": 2.5492, + "step": 686 + }, + { + "epoch": 0.3149931224209078, + "grad_norm": 0.4480987737026053, + "learning_rate": 0.0008019444909613523, + "loss": 2.5017, + "step": 687 + }, + { + "epoch": 0.3154516276937185, + "grad_norm": 0.4582098475111062, + "learning_rate": 0.0008013521806892003, + "loss": 2.3986, + "step": 688 + }, + { + "epoch": 0.3159101329665291, + "grad_norm": 0.44250091748269615, + "learning_rate": 0.000800759205521752, + "loss": 2.5464, + "step": 689 + }, + { + "epoch": 0.31636863823933975, + "grad_norm": 0.4446609311998984, + "learning_rate": 0.0008001655667673318, + "loss": 2.5762, + "step": 690 + }, + { + "epoch": 0.3168271435121504, + "grad_norm": 0.43292005767703107, + "learning_rate": 0.0007995712657357279, + "loss": 2.4895, + "step": 691 + }, + { + "epoch": 0.317285648784961, + "grad_norm": 0.4386591316178465, + "learning_rate": 0.0007989763037381904, + "loss": 2.6616, + "step": 692 + }, + { + "epoch": 0.3177441540577717, + "grad_norm": 0.44702470503250535, + "learning_rate": 0.0007983806820874271, + "loss": 2.4596, + "step": 693 + }, + { + "epoch": 0.3182026593305823, + "grad_norm": 0.43001433440459497, + "learning_rate": 0.0007977844020976016, + "loss": 2.4394, + "step": 694 + }, + { + "epoch": 0.31866116460339294, + "grad_norm": 0.45530241112464076, + "learning_rate": 0.00079718746508433, + "loss": 2.5378, + "step": 695 + }, + { + "epoch": 0.31911966987620355, + "grad_norm": 0.44874473602020387, + "learning_rate": 0.0007965898723646776, + "loss": 2.5029, + "step": 696 + }, + { + "epoch": 0.3195781751490142, + "grad_norm": 0.4276742430228682, + "learning_rate": 0.0007959916252571573, + "loss": 2.4453, + "step": 697 + }, + { + "epoch": 0.32003668042182487, + "grad_norm": 0.42404989953320443, + "learning_rate": 0.000795392725081725, + "loss": 2.3968, + "step": 698 + }, + { + "epoch": 0.3204951856946355, + "grad_norm": 0.4706632831735465, + "learning_rate": 0.000794793173159778, + "loss": 2.5472, + "step": 699 + }, + { + "epoch": 0.32095369096744614, + "grad_norm": 0.46701876178014307, + "learning_rate": 0.0007941929708141513, + "loss": 2.5843, + "step": 700 + }, + { + "epoch": 0.32141219624025674, + "grad_norm": 0.4467637336200493, + "learning_rate": 0.0007935921193691153, + "loss": 2.4612, + "step": 701 + }, + { + "epoch": 0.3218707015130674, + "grad_norm": 0.45282120658806446, + "learning_rate": 0.0007929906201503722, + "loss": 2.473, + "step": 702 + }, + { + "epoch": 0.32232920678587806, + "grad_norm": 0.46085821344698386, + "learning_rate": 0.0007923884744850536, + "loss": 2.5398, + "step": 703 + }, + { + "epoch": 0.32278771205868867, + "grad_norm": 0.4746342356179217, + "learning_rate": 0.0007917856837017176, + "loss": 2.4518, + "step": 704 + }, + { + "epoch": 0.32324621733149933, + "grad_norm": 0.4616478506709981, + "learning_rate": 0.0007911822491303452, + "loss": 2.5308, + "step": 705 + }, + { + "epoch": 0.32370472260430994, + "grad_norm": 0.45031889048918816, + "learning_rate": 0.0007905781721023382, + "loss": 2.5212, + "step": 706 + }, + { + "epoch": 0.3241632278771206, + "grad_norm": 0.46674940933330383, + "learning_rate": 0.000789973453950516, + "loss": 2.3953, + "step": 707 + }, + { + "epoch": 0.3246217331499312, + "grad_norm": 0.42553817439205904, + "learning_rate": 0.000789368096009112, + "loss": 2.5229, + "step": 708 + }, + { + "epoch": 0.32508023842274186, + "grad_norm": 0.42920539778817257, + "learning_rate": 0.0007887620996137721, + "loss": 2.4171, + "step": 709 + }, + { + "epoch": 0.3255387436955525, + "grad_norm": 0.4389001857834567, + "learning_rate": 0.0007881554661015497, + "loss": 2.3904, + "step": 710 + }, + { + "epoch": 0.32599724896836313, + "grad_norm": 0.4708812609525895, + "learning_rate": 0.0007875481968109051, + "loss": 2.4852, + "step": 711 + }, + { + "epoch": 0.3264557542411738, + "grad_norm": 0.45432402791217485, + "learning_rate": 0.0007869402930817007, + "loss": 2.5534, + "step": 712 + }, + { + "epoch": 0.3269142595139844, + "grad_norm": 0.44934613634823606, + "learning_rate": 0.0007863317562551987, + "loss": 2.3794, + "step": 713 + }, + { + "epoch": 0.32737276478679506, + "grad_norm": 0.4174643011476095, + "learning_rate": 0.0007857225876740584, + "loss": 2.4426, + "step": 714 + }, + { + "epoch": 0.32783127005960566, + "grad_norm": 0.4443535979397841, + "learning_rate": 0.0007851127886823327, + "loss": 2.3663, + "step": 715 + }, + { + "epoch": 0.3282897753324163, + "grad_norm": 0.4289263975958363, + "learning_rate": 0.0007845023606254658, + "loss": 2.453, + "step": 716 + }, + { + "epoch": 0.328748280605227, + "grad_norm": 0.4602417630213102, + "learning_rate": 0.0007838913048502894, + "loss": 2.4316, + "step": 717 + }, + { + "epoch": 0.3292067858780376, + "grad_norm": 0.5095554645724768, + "learning_rate": 0.0007832796227050208, + "loss": 2.6374, + "step": 718 + }, + { + "epoch": 0.32966529115084825, + "grad_norm": 0.45093449795085355, + "learning_rate": 0.0007826673155392587, + "loss": 2.5193, + "step": 719 + }, + { + "epoch": 0.33012379642365886, + "grad_norm": 0.4261067176186917, + "learning_rate": 0.000782054384703981, + "loss": 2.3855, + "step": 720 + }, + { + "epoch": 0.3305823016964695, + "grad_norm": 0.4342086013966162, + "learning_rate": 0.0007814408315515418, + "loss": 2.4813, + "step": 721 + }, + { + "epoch": 0.3310408069692801, + "grad_norm": 0.43881367917237585, + "learning_rate": 0.0007808266574356683, + "loss": 2.524, + "step": 722 + }, + { + "epoch": 0.3314993122420908, + "grad_norm": 0.4608497184981568, + "learning_rate": 0.0007802118637114573, + "loss": 2.4343, + "step": 723 + }, + { + "epoch": 0.33195781751490144, + "grad_norm": 0.43407342261798904, + "learning_rate": 0.0007795964517353734, + "loss": 2.6304, + "step": 724 + }, + { + "epoch": 0.33241632278771205, + "grad_norm": 0.44098266204055864, + "learning_rate": 0.0007789804228652449, + "loss": 2.466, + "step": 725 + }, + { + "epoch": 0.3328748280605227, + "grad_norm": 0.4310821360476788, + "learning_rate": 0.0007783637784602609, + "loss": 2.4039, + "step": 726 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.4306596738998256, + "learning_rate": 0.0007777465198809692, + "loss": 2.4664, + "step": 727 + }, + { + "epoch": 0.333791838606144, + "grad_norm": 0.49793820130510813, + "learning_rate": 0.0007771286484892722, + "loss": 2.6201, + "step": 728 + }, + { + "epoch": 0.3342503438789546, + "grad_norm": 0.44063708846988925, + "learning_rate": 0.000776510165648425, + "loss": 2.5514, + "step": 729 + }, + { + "epoch": 0.33470884915176524, + "grad_norm": 0.43110258294800824, + "learning_rate": 0.0007758910727230311, + "loss": 2.4302, + "step": 730 + }, + { + "epoch": 0.3351673544245759, + "grad_norm": 0.47213751256800696, + "learning_rate": 0.0007752713710790404, + "loss": 2.4569, + "step": 731 + }, + { + "epoch": 0.3356258596973865, + "grad_norm": 0.46111743933488847, + "learning_rate": 0.0007746510620837459, + "loss": 2.5044, + "step": 732 + }, + { + "epoch": 0.33608436497019717, + "grad_norm": 0.4375751816065365, + "learning_rate": 0.0007740301471057807, + "loss": 2.4254, + "step": 733 + }, + { + "epoch": 0.3365428702430078, + "grad_norm": 0.4506902710997529, + "learning_rate": 0.0007734086275151146, + "loss": 2.4614, + "step": 734 + }, + { + "epoch": 0.33700137551581844, + "grad_norm": 0.42414858317458315, + "learning_rate": 0.0007727865046830517, + "loss": 2.3584, + "step": 735 + }, + { + "epoch": 0.3374598807886291, + "grad_norm": 0.4363193773730655, + "learning_rate": 0.0007721637799822269, + "loss": 2.5035, + "step": 736 + }, + { + "epoch": 0.3379183860614397, + "grad_norm": 0.42655432569593454, + "learning_rate": 0.0007715404547866032, + "loss": 2.449, + "step": 737 + }, + { + "epoch": 0.33837689133425036, + "grad_norm": 0.4324398312290806, + "learning_rate": 0.0007709165304714685, + "loss": 2.5452, + "step": 738 + }, + { + "epoch": 0.33883539660706097, + "grad_norm": 0.4369886964988613, + "learning_rate": 0.0007702920084134324, + "loss": 2.5097, + "step": 739 + }, + { + "epoch": 0.33929390187987163, + "grad_norm": 0.44719874488964606, + "learning_rate": 0.0007696668899904236, + "loss": 2.3994, + "step": 740 + }, + { + "epoch": 0.33975240715268223, + "grad_norm": 0.4592595079804714, + "learning_rate": 0.0007690411765816864, + "loss": 2.4843, + "step": 741 + }, + { + "epoch": 0.3402109124254929, + "grad_norm": 0.4463091554646953, + "learning_rate": 0.0007684148695677778, + "loss": 2.3813, + "step": 742 + }, + { + "epoch": 0.34066941769830356, + "grad_norm": 0.4432298716912301, + "learning_rate": 0.000767787970330565, + "loss": 2.4108, + "step": 743 + }, + { + "epoch": 0.34112792297111416, + "grad_norm": 0.4306609469634093, + "learning_rate": 0.000767160480253221, + "loss": 2.5686, + "step": 744 + }, + { + "epoch": 0.3415864282439248, + "grad_norm": 0.4522652956498799, + "learning_rate": 0.0007665324007202235, + "loss": 2.3987, + "step": 745 + }, + { + "epoch": 0.3420449335167354, + "grad_norm": 0.426406182319128, + "learning_rate": 0.0007659037331173498, + "loss": 2.4579, + "step": 746 + }, + { + "epoch": 0.3425034387895461, + "grad_norm": 0.42574153108451696, + "learning_rate": 0.0007652744788316752, + "loss": 2.3896, + "step": 747 + }, + { + "epoch": 0.3429619440623567, + "grad_norm": 0.43174620333433267, + "learning_rate": 0.0007646446392515692, + "loss": 2.4649, + "step": 748 + }, + { + "epoch": 0.34342044933516735, + "grad_norm": 0.43798257670945656, + "learning_rate": 0.000764014215766693, + "loss": 2.5154, + "step": 749 + }, + { + "epoch": 0.343878954607978, + "grad_norm": 0.46202532278136604, + "learning_rate": 0.0007633832097679958, + "loss": 2.4776, + "step": 750 + }, + { + "epoch": 0.3443374598807886, + "grad_norm": 0.4428167420191148, + "learning_rate": 0.0007627516226477122, + "loss": 2.4581, + "step": 751 + }, + { + "epoch": 0.3447959651535993, + "grad_norm": 0.44817202877092177, + "learning_rate": 0.0007621194557993589, + "loss": 2.568, + "step": 752 + }, + { + "epoch": 0.3452544704264099, + "grad_norm": 0.48752747576341166, + "learning_rate": 0.0007614867106177319, + "loss": 2.4794, + "step": 753 + }, + { + "epoch": 0.34571297569922055, + "grad_norm": 0.45756820320075436, + "learning_rate": 0.0007608533884989029, + "loss": 2.4787, + "step": 754 + }, + { + "epoch": 0.34617148097203115, + "grad_norm": 0.4497778264173633, + "learning_rate": 0.0007602194908402166, + "loss": 2.4376, + "step": 755 + }, + { + "epoch": 0.3466299862448418, + "grad_norm": 0.4285288839105894, + "learning_rate": 0.0007595850190402877, + "loss": 2.5216, + "step": 756 + }, + { + "epoch": 0.3470884915176525, + "grad_norm": 0.4526607173742684, + "learning_rate": 0.0007589499744989976, + "loss": 2.6144, + "step": 757 + }, + { + "epoch": 0.3475469967904631, + "grad_norm": 0.44051594190909515, + "learning_rate": 0.0007583143586174916, + "loss": 2.4775, + "step": 758 + }, + { + "epoch": 0.34800550206327374, + "grad_norm": 0.42791319074916806, + "learning_rate": 0.000757678172798175, + "loss": 2.5022, + "step": 759 + }, + { + "epoch": 0.34846400733608435, + "grad_norm": 0.4258079447985116, + "learning_rate": 0.0007570414184447112, + "loss": 2.4934, + "step": 760 + }, + { + "epoch": 0.348922512608895, + "grad_norm": 0.43665635402348274, + "learning_rate": 0.0007564040969620179, + "loss": 2.4765, + "step": 761 + }, + { + "epoch": 0.3493810178817056, + "grad_norm": 0.4369549765058256, + "learning_rate": 0.0007557662097562636, + "loss": 2.3396, + "step": 762 + }, + { + "epoch": 0.3498395231545163, + "grad_norm": 0.44257288811439627, + "learning_rate": 0.0007551277582348658, + "loss": 2.5649, + "step": 763 + }, + { + "epoch": 0.35029802842732694, + "grad_norm": 0.429009708189031, + "learning_rate": 0.0007544887438064862, + "loss": 2.4109, + "step": 764 + }, + { + "epoch": 0.35075653370013754, + "grad_norm": 0.440586062327402, + "learning_rate": 0.0007538491678810294, + "loss": 2.3519, + "step": 765 + }, + { + "epoch": 0.3512150389729482, + "grad_norm": 0.43621307926133845, + "learning_rate": 0.0007532090318696381, + "loss": 2.4999, + "step": 766 + }, + { + "epoch": 0.3516735442457588, + "grad_norm": 0.4282548620723116, + "learning_rate": 0.0007525683371846913, + "loss": 2.4798, + "step": 767 + }, + { + "epoch": 0.35213204951856947, + "grad_norm": 0.43723748292012826, + "learning_rate": 0.0007519270852398001, + "loss": 2.4885, + "step": 768 + }, + { + "epoch": 0.35259055479138013, + "grad_norm": 0.4114435603272793, + "learning_rate": 0.000751285277449806, + "loss": 2.4649, + "step": 769 + }, + { + "epoch": 0.35304906006419073, + "grad_norm": 0.4474425767489191, + "learning_rate": 0.0007506429152307756, + "loss": 2.4773, + "step": 770 + }, + { + "epoch": 0.3535075653370014, + "grad_norm": 0.41541451525993955, + "learning_rate": 0.00075, + "loss": 2.3963, + "step": 771 + }, + { + "epoch": 0.353966070609812, + "grad_norm": 0.39862418563746654, + "learning_rate": 0.00074935653317599, + "loss": 2.4583, + "step": 772 + }, + { + "epoch": 0.35442457588262266, + "grad_norm": 0.4397612103717388, + "learning_rate": 0.000748712516178473, + "loss": 2.4734, + "step": 773 + }, + { + "epoch": 0.35488308115543327, + "grad_norm": 0.43273847442685176, + "learning_rate": 0.0007480679504283911, + "loss": 2.4974, + "step": 774 + }, + { + "epoch": 0.3553415864282439, + "grad_norm": 0.4164213308874209, + "learning_rate": 0.0007474228373478964, + "loss": 2.4614, + "step": 775 + }, + { + "epoch": 0.3558000917010546, + "grad_norm": 0.44637092203736395, + "learning_rate": 0.0007467771783603492, + "loss": 2.3977, + "step": 776 + }, + { + "epoch": 0.3562585969738652, + "grad_norm": 0.4140621096175081, + "learning_rate": 0.0007461309748903138, + "loss": 2.5068, + "step": 777 + }, + { + "epoch": 0.35671710224667585, + "grad_norm": 0.426243884914397, + "learning_rate": 0.0007454842283635562, + "loss": 2.4687, + "step": 778 + }, + { + "epoch": 0.35717560751948646, + "grad_norm": 0.45498780272067524, + "learning_rate": 0.0007448369402070404, + "loss": 2.3415, + "step": 779 + }, + { + "epoch": 0.3576341127922971, + "grad_norm": 0.4613864172942696, + "learning_rate": 0.0007441891118489254, + "loss": 2.5426, + "step": 780 + }, + { + "epoch": 0.3580926180651077, + "grad_norm": 0.4397602281124493, + "learning_rate": 0.0007435407447185622, + "loss": 2.4884, + "step": 781 + }, + { + "epoch": 0.3585511233379184, + "grad_norm": 0.45752786081196234, + "learning_rate": 0.0007428918402464908, + "loss": 2.4652, + "step": 782 + }, + { + "epoch": 0.35900962861072905, + "grad_norm": 0.4377529666774645, + "learning_rate": 0.0007422423998644359, + "loss": 2.3831, + "step": 783 + }, + { + "epoch": 0.35946813388353965, + "grad_norm": 0.4180791052687565, + "learning_rate": 0.0007415924250053055, + "loss": 2.4932, + "step": 784 + }, + { + "epoch": 0.3599266391563503, + "grad_norm": 0.44901765878343014, + "learning_rate": 0.0007409419171031865, + "loss": 2.4551, + "step": 785 + }, + { + "epoch": 0.3603851444291609, + "grad_norm": 0.4378394407709021, + "learning_rate": 0.0007402908775933419, + "loss": 2.3933, + "step": 786 + }, + { + "epoch": 0.3608436497019716, + "grad_norm": 0.43713163496763424, + "learning_rate": 0.0007396393079122077, + "loss": 2.3368, + "step": 787 + }, + { + "epoch": 0.3613021549747822, + "grad_norm": 0.4329709502141535, + "learning_rate": 0.0007389872094973896, + "loss": 2.4125, + "step": 788 + }, + { + "epoch": 0.36176066024759285, + "grad_norm": 0.43224867370625486, + "learning_rate": 0.00073833458378766, + "loss": 2.4533, + "step": 789 + }, + { + "epoch": 0.3622191655204035, + "grad_norm": 0.45794213287670593, + "learning_rate": 0.0007376814322229544, + "loss": 2.5642, + "step": 790 + }, + { + "epoch": 0.3626776707932141, + "grad_norm": 0.4831813122336838, + "learning_rate": 0.0007370277562443688, + "loss": 2.4258, + "step": 791 + }, + { + "epoch": 0.3631361760660248, + "grad_norm": 0.4192744564902884, + "learning_rate": 0.0007363735572941564, + "loss": 2.5015, + "step": 792 + }, + { + "epoch": 0.3635946813388354, + "grad_norm": 0.4620781676548418, + "learning_rate": 0.0007357188368157236, + "loss": 2.58, + "step": 793 + }, + { + "epoch": 0.36405318661164604, + "grad_norm": 0.46197578837347736, + "learning_rate": 0.0007350635962536284, + "loss": 2.4341, + "step": 794 + }, + { + "epoch": 0.36451169188445665, + "grad_norm": 0.42593941244506267, + "learning_rate": 0.0007344078370535756, + "loss": 2.357, + "step": 795 + }, + { + "epoch": 0.3649701971572673, + "grad_norm": 0.41528932879738806, + "learning_rate": 0.0007337515606624148, + "loss": 2.4899, + "step": 796 + }, + { + "epoch": 0.36542870243007797, + "grad_norm": 0.41469250304827826, + "learning_rate": 0.0007330947685281362, + "loss": 2.3691, + "step": 797 + }, + { + "epoch": 0.3658872077028886, + "grad_norm": 0.4563577628205324, + "learning_rate": 0.0007324374620998682, + "loss": 2.4809, + "step": 798 + }, + { + "epoch": 0.36634571297569923, + "grad_norm": 0.4445407903490564, + "learning_rate": 0.000731779642827874, + "loss": 2.3933, + "step": 799 + }, + { + "epoch": 0.36680421824850984, + "grad_norm": 0.45542290728402274, + "learning_rate": 0.0007311213121635483, + "loss": 2.4177, + "step": 800 + }, + { + "epoch": 0.3672627235213205, + "grad_norm": 0.46127296748078495, + "learning_rate": 0.0007304624715594139, + "loss": 2.445, + "step": 801 + }, + { + "epoch": 0.3677212287941311, + "grad_norm": 0.4339105414299232, + "learning_rate": 0.0007298031224691193, + "loss": 2.465, + "step": 802 + }, + { + "epoch": 0.36817973406694177, + "grad_norm": 0.4206833545713446, + "learning_rate": 0.0007291432663474339, + "loss": 2.408, + "step": 803 + }, + { + "epoch": 0.3686382393397524, + "grad_norm": 0.4211854724877334, + "learning_rate": 0.0007284829046502467, + "loss": 2.4676, + "step": 804 + }, + { + "epoch": 0.36909674461256303, + "grad_norm": 0.4190206137699201, + "learning_rate": 0.0007278220388345619, + "loss": 2.4422, + "step": 805 + }, + { + "epoch": 0.3695552498853737, + "grad_norm": 0.40625555522953516, + "learning_rate": 0.0007271606703584958, + "loss": 2.4192, + "step": 806 + }, + { + "epoch": 0.3700137551581843, + "grad_norm": 0.41731364140460514, + "learning_rate": 0.000726498800681274, + "loss": 2.4128, + "step": 807 + }, + { + "epoch": 0.37047226043099496, + "grad_norm": 0.4405414378468179, + "learning_rate": 0.0007258364312632279, + "loss": 2.4614, + "step": 808 + }, + { + "epoch": 0.3709307657038056, + "grad_norm": 0.4398482059380941, + "learning_rate": 0.0007251735635657915, + "loss": 2.4299, + "step": 809 + }, + { + "epoch": 0.3713892709766162, + "grad_norm": 0.45791491006501656, + "learning_rate": 0.000724510199051498, + "loss": 2.4374, + "step": 810 + }, + { + "epoch": 0.3718477762494269, + "grad_norm": 0.46998220528733314, + "learning_rate": 0.0007238463391839769, + "loss": 2.434, + "step": 811 + }, + { + "epoch": 0.3723062815222375, + "grad_norm": 0.42615812154989063, + "learning_rate": 0.0007231819854279508, + "loss": 2.5267, + "step": 812 + }, + { + "epoch": 0.37276478679504815, + "grad_norm": 0.4233561883067928, + "learning_rate": 0.0007225171392492316, + "loss": 2.3653, + "step": 813 + }, + { + "epoch": 0.37322329206785876, + "grad_norm": 0.4288640205164911, + "learning_rate": 0.0007218518021147182, + "loss": 2.5245, + "step": 814 + }, + { + "epoch": 0.3736817973406694, + "grad_norm": 0.42823283513432353, + "learning_rate": 0.0007211859754923923, + "loss": 2.4616, + "step": 815 + }, + { + "epoch": 0.3741403026134801, + "grad_norm": 0.42956437465519737, + "learning_rate": 0.0007205196608513158, + "loss": 2.4016, + "step": 816 + }, + { + "epoch": 0.3745988078862907, + "grad_norm": 0.41136267365595486, + "learning_rate": 0.0007198528596616272, + "loss": 2.5093, + "step": 817 + }, + { + "epoch": 0.37505731315910135, + "grad_norm": 0.4178048171944708, + "learning_rate": 0.0007191855733945387, + "loss": 2.3694, + "step": 818 + }, + { + "epoch": 0.37551581843191195, + "grad_norm": 0.4176889414918142, + "learning_rate": 0.0007185178035223327, + "loss": 2.4184, + "step": 819 + }, + { + "epoch": 0.3759743237047226, + "grad_norm": 0.43002591272016377, + "learning_rate": 0.0007178495515183583, + "loss": 2.4523, + "step": 820 + }, + { + "epoch": 0.3764328289775332, + "grad_norm": 0.4183278931966018, + "learning_rate": 0.000717180818857029, + "loss": 2.4357, + "step": 821 + }, + { + "epoch": 0.3768913342503439, + "grad_norm": 0.4258477365244085, + "learning_rate": 0.0007165116070138182, + "loss": 2.3479, + "step": 822 + }, + { + "epoch": 0.37734983952315454, + "grad_norm": 0.40454263606233265, + "learning_rate": 0.0007158419174652569, + "loss": 2.3347, + "step": 823 + }, + { + "epoch": 0.37780834479596515, + "grad_norm": 0.4499772377867611, + "learning_rate": 0.00071517175168893, + "loss": 2.4183, + "step": 824 + }, + { + "epoch": 0.3782668500687758, + "grad_norm": 0.4142465256995089, + "learning_rate": 0.0007145011111634732, + "loss": 2.5192, + "step": 825 + }, + { + "epoch": 0.3787253553415864, + "grad_norm": 0.44971769757207647, + "learning_rate": 0.0007138299973685694, + "loss": 2.5544, + "step": 826 + }, + { + "epoch": 0.3791838606143971, + "grad_norm": 0.4598093053714518, + "learning_rate": 0.0007131584117849459, + "loss": 2.4664, + "step": 827 + }, + { + "epoch": 0.3796423658872077, + "grad_norm": 0.45680562689806714, + "learning_rate": 0.0007124863558943713, + "loss": 2.4764, + "step": 828 + }, + { + "epoch": 0.38010087116001834, + "grad_norm": 0.4407626678017755, + "learning_rate": 0.0007118138311796514, + "loss": 2.5195, + "step": 829 + }, + { + "epoch": 0.380559376432829, + "grad_norm": 0.42992747877230547, + "learning_rate": 0.0007111408391246262, + "loss": 2.4009, + "step": 830 + }, + { + "epoch": 0.3810178817056396, + "grad_norm": 0.427469739417204, + "learning_rate": 0.0007104673812141675, + "loss": 2.5214, + "step": 831 + }, + { + "epoch": 0.38147638697845027, + "grad_norm": 0.43549585551431147, + "learning_rate": 0.0007097934589341745, + "loss": 2.5023, + "step": 832 + }, + { + "epoch": 0.38193489225126087, + "grad_norm": 0.41644159985219253, + "learning_rate": 0.0007091190737715711, + "loss": 2.4909, + "step": 833 + }, + { + "epoch": 0.38239339752407153, + "grad_norm": 0.4061417766477284, + "learning_rate": 0.0007084442272143026, + "loss": 2.4417, + "step": 834 + }, + { + "epoch": 0.38285190279688214, + "grad_norm": 0.4139181736886135, + "learning_rate": 0.000707768920751332, + "loss": 2.5169, + "step": 835 + }, + { + "epoch": 0.3833104080696928, + "grad_norm": 0.4485741974645468, + "learning_rate": 0.0007070931558726373, + "loss": 2.5, + "step": 836 + }, + { + "epoch": 0.38376891334250346, + "grad_norm": 0.43095383388993747, + "learning_rate": 0.0007064169340692076, + "loss": 2.4349, + "step": 837 + }, + { + "epoch": 0.38422741861531406, + "grad_norm": 0.43828240590359585, + "learning_rate": 0.0007057402568330407, + "loss": 2.3731, + "step": 838 + }, + { + "epoch": 0.3846859238881247, + "grad_norm": 0.40236894027545284, + "learning_rate": 0.0007050631256571389, + "loss": 2.3738, + "step": 839 + }, + { + "epoch": 0.38514442916093533, + "grad_norm": 0.4286329975458471, + "learning_rate": 0.000704385542035506, + "loss": 2.4262, + "step": 840 + }, + { + "epoch": 0.385602934433746, + "grad_norm": 0.431805779586095, + "learning_rate": 0.000703707507463144, + "loss": 2.3928, + "step": 841 + }, + { + "epoch": 0.38606143970655665, + "grad_norm": 0.42961881641332955, + "learning_rate": 0.0007030290234360505, + "loss": 2.5564, + "step": 842 + }, + { + "epoch": 0.38651994497936726, + "grad_norm": 0.44818443090379195, + "learning_rate": 0.0007023500914512139, + "loss": 2.4697, + "step": 843 + }, + { + "epoch": 0.3869784502521779, + "grad_norm": 0.4208497847131704, + "learning_rate": 0.0007016707130066116, + "loss": 2.3898, + "step": 844 + }, + { + "epoch": 0.3874369555249885, + "grad_norm": 0.4105946404782101, + "learning_rate": 0.0007009908896012055, + "loss": 2.4586, + "step": 845 + }, + { + "epoch": 0.3878954607977992, + "grad_norm": 0.41481463170500693, + "learning_rate": 0.0007003106227349399, + "loss": 2.5498, + "step": 846 + }, + { + "epoch": 0.3883539660706098, + "grad_norm": 0.4541831873781609, + "learning_rate": 0.000699629913908737, + "loss": 2.4803, + "step": 847 + }, + { + "epoch": 0.38881247134342045, + "grad_norm": 0.46358476925727804, + "learning_rate": 0.0006989487646244943, + "loss": 2.4916, + "step": 848 + }, + { + "epoch": 0.3892709766162311, + "grad_norm": 0.45811054566025383, + "learning_rate": 0.0006982671763850814, + "loss": 2.4275, + "step": 849 + }, + { + "epoch": 0.3897294818890417, + "grad_norm": 0.42394255019702487, + "learning_rate": 0.0006975851506943359, + "loss": 2.3945, + "step": 850 + }, + { + "epoch": 0.3901879871618524, + "grad_norm": 0.47315019716348455, + "learning_rate": 0.0006969026890570611, + "loss": 2.4849, + "step": 851 + }, + { + "epoch": 0.390646492434663, + "grad_norm": 0.4409269073758902, + "learning_rate": 0.0006962197929790216, + "loss": 2.3856, + "step": 852 + }, + { + "epoch": 0.39110499770747365, + "grad_norm": 0.43956131265673604, + "learning_rate": 0.0006955364639669409, + "loss": 2.5497, + "step": 853 + }, + { + "epoch": 0.39156350298028425, + "grad_norm": 0.42676015466968276, + "learning_rate": 0.0006948527035284978, + "loss": 2.4851, + "step": 854 + }, + { + "epoch": 0.3920220082530949, + "grad_norm": 0.40591901999136787, + "learning_rate": 0.0006941685131723225, + "loss": 2.4902, + "step": 855 + }, + { + "epoch": 0.39248051352590557, + "grad_norm": 0.4347587150262814, + "learning_rate": 0.0006934838944079943, + "loss": 2.4904, + "step": 856 + }, + { + "epoch": 0.3929390187987162, + "grad_norm": 0.43472822031426034, + "learning_rate": 0.0006927988487460378, + "loss": 2.4276, + "step": 857 + }, + { + "epoch": 0.39339752407152684, + "grad_norm": 0.4433727137674392, + "learning_rate": 0.0006921133776979186, + "loss": 2.458, + "step": 858 + }, + { + "epoch": 0.39385602934433744, + "grad_norm": 0.4410286668841732, + "learning_rate": 0.0006914274827760418, + "loss": 2.3764, + "step": 859 + }, + { + "epoch": 0.3943145346171481, + "grad_norm": 0.4349111122129766, + "learning_rate": 0.0006907411654937475, + "loss": 2.494, + "step": 860 + }, + { + "epoch": 0.3947730398899587, + "grad_norm": 0.4630650511592329, + "learning_rate": 0.0006900544273653075, + "loss": 2.5819, + "step": 861 + }, + { + "epoch": 0.39523154516276937, + "grad_norm": 0.4446226883306221, + "learning_rate": 0.000689367269905922, + "loss": 2.4876, + "step": 862 + }, + { + "epoch": 0.39569005043558003, + "grad_norm": 0.43672264570595365, + "learning_rate": 0.0006886796946317168, + "loss": 2.4318, + "step": 863 + }, + { + "epoch": 0.39614855570839064, + "grad_norm": 0.42739033307845925, + "learning_rate": 0.0006879917030597397, + "loss": 2.4408, + "step": 864 + }, + { + "epoch": 0.3966070609812013, + "grad_norm": 0.43088764082177894, + "learning_rate": 0.0006873032967079561, + "loss": 2.4003, + "step": 865 + }, + { + "epoch": 0.3970655662540119, + "grad_norm": 0.4295586284371126, + "learning_rate": 0.0006866144770952474, + "loss": 2.4426, + "step": 866 + }, + { + "epoch": 0.39752407152682256, + "grad_norm": 0.442503973861287, + "learning_rate": 0.0006859252457414067, + "loss": 2.5291, + "step": 867 + }, + { + "epoch": 0.39798257679963317, + "grad_norm": 0.4290249221620293, + "learning_rate": 0.0006852356041671351, + "loss": 2.3394, + "step": 868 + }, + { + "epoch": 0.39844108207244383, + "grad_norm": 0.4403818875317827, + "learning_rate": 0.0006845455538940394, + "loss": 2.4979, + "step": 869 + }, + { + "epoch": 0.3988995873452545, + "grad_norm": 0.41172948752929345, + "learning_rate": 0.0006838550964446276, + "loss": 2.3943, + "step": 870 + }, + { + "epoch": 0.3993580926180651, + "grad_norm": 0.4434155190205296, + "learning_rate": 0.0006831642333423067, + "loss": 2.5523, + "step": 871 + }, + { + "epoch": 0.39981659789087576, + "grad_norm": 0.4609839491251962, + "learning_rate": 0.000682472966111378, + "loss": 2.4477, + "step": 872 + }, + { + "epoch": 0.40027510316368636, + "grad_norm": 0.42043074918503204, + "learning_rate": 0.0006817812962770348, + "loss": 2.4795, + "step": 873 + }, + { + "epoch": 0.400733608436497, + "grad_norm": 0.42926542863963524, + "learning_rate": 0.0006810892253653589, + "loss": 2.4108, + "step": 874 + }, + { + "epoch": 0.40119211370930763, + "grad_norm": 0.43788195784589734, + "learning_rate": 0.0006803967549033167, + "loss": 2.4406, + "step": 875 + }, + { + "epoch": 0.4016506189821183, + "grad_norm": 0.4390783994473238, + "learning_rate": 0.0006797038864187564, + "loss": 2.4931, + "step": 876 + }, + { + "epoch": 0.40210912425492895, + "grad_norm": 0.5017492860538838, + "learning_rate": 0.0006790106214404043, + "loss": 2.5406, + "step": 877 + }, + { + "epoch": 0.40256762952773956, + "grad_norm": 0.424029192579047, + "learning_rate": 0.0006783169614978614, + "loss": 2.4622, + "step": 878 + }, + { + "epoch": 0.4030261348005502, + "grad_norm": 0.4146620110607149, + "learning_rate": 0.0006776229081216001, + "loss": 2.4081, + "step": 879 + }, + { + "epoch": 0.4034846400733608, + "grad_norm": 0.4146674354880038, + "learning_rate": 0.0006769284628429611, + "loss": 2.4034, + "step": 880 + }, + { + "epoch": 0.4039431453461715, + "grad_norm": 0.44457863635144135, + "learning_rate": 0.0006762336271941498, + "loss": 2.4609, + "step": 881 + }, + { + "epoch": 0.40440165061898214, + "grad_norm": 0.40707024883507337, + "learning_rate": 0.0006755384027082326, + "loss": 2.4217, + "step": 882 + }, + { + "epoch": 0.40486015589179275, + "grad_norm": 0.4310811604776173, + "learning_rate": 0.0006748427909191342, + "loss": 2.4075, + "step": 883 + }, + { + "epoch": 0.4053186611646034, + "grad_norm": 0.4581123616023857, + "learning_rate": 0.0006741467933616335, + "loss": 2.4426, + "step": 884 + }, + { + "epoch": 0.405777166437414, + "grad_norm": 0.42365384680044466, + "learning_rate": 0.0006734504115713604, + "loss": 2.427, + "step": 885 + }, + { + "epoch": 0.4062356717102247, + "grad_norm": 0.4191949279685354, + "learning_rate": 0.0006727536470847932, + "loss": 2.4167, + "step": 886 + }, + { + "epoch": 0.4066941769830353, + "grad_norm": 0.4613844923690668, + "learning_rate": 0.000672056501439254, + "loss": 2.3901, + "step": 887 + }, + { + "epoch": 0.40715268225584594, + "grad_norm": 0.4171527809035184, + "learning_rate": 0.0006713589761729063, + "loss": 2.4183, + "step": 888 + }, + { + "epoch": 0.4076111875286566, + "grad_norm": 0.4645083436012071, + "learning_rate": 0.0006706610728247508, + "loss": 2.4985, + "step": 889 + }, + { + "epoch": 0.4080696928014672, + "grad_norm": 0.452106813261534, + "learning_rate": 0.0006699627929346227, + "loss": 2.5378, + "step": 890 + }, + { + "epoch": 0.40852819807427787, + "grad_norm": 0.44742712286896724, + "learning_rate": 0.0006692641380431879, + "loss": 2.4446, + "step": 891 + }, + { + "epoch": 0.4089867033470885, + "grad_norm": 0.41134885839182744, + "learning_rate": 0.0006685651096919393, + "loss": 2.4366, + "step": 892 + }, + { + "epoch": 0.40944520861989914, + "grad_norm": 0.44252707484930276, + "learning_rate": 0.0006678657094231944, + "loss": 2.2476, + "step": 893 + }, + { + "epoch": 0.40990371389270974, + "grad_norm": 0.4218713801060894, + "learning_rate": 0.0006671659387800909, + "loss": 2.5496, + "step": 894 + }, + { + "epoch": 0.4103622191655204, + "grad_norm": 0.4220393055985501, + "learning_rate": 0.000666465799306584, + "loss": 2.4379, + "step": 895 + }, + { + "epoch": 0.41082072443833106, + "grad_norm": 0.443053679749063, + "learning_rate": 0.0006657652925474423, + "loss": 2.4187, + "step": 896 + }, + { + "epoch": 0.41127922971114167, + "grad_norm": 0.45435628617314866, + "learning_rate": 0.000665064420048245, + "loss": 2.508, + "step": 897 + }, + { + "epoch": 0.41173773498395233, + "grad_norm": 0.4133220749005298, + "learning_rate": 0.0006643631833553785, + "loss": 2.3778, + "step": 898 + }, + { + "epoch": 0.41219624025676294, + "grad_norm": 0.4314758527851054, + "learning_rate": 0.000663661584016032, + "loss": 2.4103, + "step": 899 + }, + { + "epoch": 0.4126547455295736, + "grad_norm": 0.4334230477605464, + "learning_rate": 0.0006629596235781957, + "loss": 2.5258, + "step": 900 + }, + { + "epoch": 0.4131132508023842, + "grad_norm": 0.4390937758506919, + "learning_rate": 0.0006622573035906556, + "loss": 2.4303, + "step": 901 + }, + { + "epoch": 0.41357175607519486, + "grad_norm": 0.4329190892066562, + "learning_rate": 0.0006615546256029921, + "loss": 2.489, + "step": 902 + }, + { + "epoch": 0.4140302613480055, + "grad_norm": 0.44538484945585727, + "learning_rate": 0.0006608515911655743, + "loss": 2.6258, + "step": 903 + }, + { + "epoch": 0.41448876662081613, + "grad_norm": 0.4272863892217783, + "learning_rate": 0.0006601482018295591, + "loss": 2.5002, + "step": 904 + }, + { + "epoch": 0.4149472718936268, + "grad_norm": 0.393336719805924, + "learning_rate": 0.0006594444591468851, + "loss": 2.4659, + "step": 905 + }, + { + "epoch": 0.4154057771664374, + "grad_norm": 0.4026264223706243, + "learning_rate": 0.0006587403646702713, + "loss": 2.483, + "step": 906 + }, + { + "epoch": 0.41586428243924806, + "grad_norm": 0.41110460225363177, + "learning_rate": 0.0006580359199532126, + "loss": 2.4093, + "step": 907 + }, + { + "epoch": 0.41632278771205866, + "grad_norm": 0.4276444720856623, + "learning_rate": 0.000657331126549977, + "loss": 2.4007, + "step": 908 + }, + { + "epoch": 0.4167812929848693, + "grad_norm": 0.4062459467206545, + "learning_rate": 0.0006566259860156014, + "loss": 2.4106, + "step": 909 + }, + { + "epoch": 0.41723979825768, + "grad_norm": 0.4471564406908179, + "learning_rate": 0.0006559204999058888, + "loss": 2.4858, + "step": 910 + }, + { + "epoch": 0.4176983035304906, + "grad_norm": 0.46280690358545007, + "learning_rate": 0.0006552146697774049, + "loss": 2.482, + "step": 911 + }, + { + "epoch": 0.41815680880330125, + "grad_norm": 0.4661881595891513, + "learning_rate": 0.0006545084971874737, + "loss": 2.4685, + "step": 912 + }, + { + "epoch": 0.41861531407611186, + "grad_norm": 0.4392072181655545, + "learning_rate": 0.0006538019836941758, + "loss": 2.4431, + "step": 913 + }, + { + "epoch": 0.4190738193489225, + "grad_norm": 0.4503756361623562, + "learning_rate": 0.0006530951308563431, + "loss": 2.5086, + "step": 914 + }, + { + "epoch": 0.4195323246217332, + "grad_norm": 0.40844863556759253, + "learning_rate": 0.0006523879402335567, + "loss": 2.46, + "step": 915 + }, + { + "epoch": 0.4199908298945438, + "grad_norm": 0.4216650990066973, + "learning_rate": 0.0006516804133861429, + "loss": 2.4471, + "step": 916 + }, + { + "epoch": 0.42044933516735444, + "grad_norm": 0.3970895843688889, + "learning_rate": 0.0006509725518751698, + "loss": 2.377, + "step": 917 + }, + { + "epoch": 0.42090784044016505, + "grad_norm": 0.42657301254689134, + "learning_rate": 0.0006502643572624438, + "loss": 2.4522, + "step": 918 + }, + { + "epoch": 0.4213663457129757, + "grad_norm": 0.41655362822991565, + "learning_rate": 0.0006495558311105064, + "loss": 2.3646, + "step": 919 + }, + { + "epoch": 0.4218248509857863, + "grad_norm": 0.4147287750209825, + "learning_rate": 0.0006488469749826305, + "loss": 2.4422, + "step": 920 + }, + { + "epoch": 0.422283356258597, + "grad_norm": 0.4015234239375801, + "learning_rate": 0.000648137790442817, + "loss": 2.4634, + "step": 921 + }, + { + "epoch": 0.42274186153140764, + "grad_norm": 0.42084323098929893, + "learning_rate": 0.0006474282790557916, + "loss": 2.3934, + "step": 922 + }, + { + "epoch": 0.42320036680421824, + "grad_norm": 0.42435160404218253, + "learning_rate": 0.000646718442387001, + "loss": 2.4299, + "step": 923 + }, + { + "epoch": 0.4236588720770289, + "grad_norm": 0.4057650184382419, + "learning_rate": 0.0006460082820026094, + "loss": 2.4018, + "step": 924 + }, + { + "epoch": 0.4241173773498395, + "grad_norm": 0.42026437525769567, + "learning_rate": 0.0006452977994694959, + "loss": 2.3711, + "step": 925 + }, + { + "epoch": 0.42457588262265017, + "grad_norm": 0.42264311800584115, + "learning_rate": 0.0006445869963552496, + "loss": 2.3838, + "step": 926 + }, + { + "epoch": 0.4250343878954608, + "grad_norm": 0.4087904627197788, + "learning_rate": 0.0006438758742281672, + "loss": 2.4333, + "step": 927 + }, + { + "epoch": 0.42549289316827144, + "grad_norm": 0.4303648957311923, + "learning_rate": 0.0006431644346572495, + "loss": 2.4221, + "step": 928 + }, + { + "epoch": 0.4259513984410821, + "grad_norm": 0.4430355504797853, + "learning_rate": 0.0006424526792121974, + "loss": 2.3425, + "step": 929 + }, + { + "epoch": 0.4264099037138927, + "grad_norm": 0.3996330824549835, + "learning_rate": 0.0006417406094634089, + "loss": 2.5221, + "step": 930 + }, + { + "epoch": 0.42686840898670336, + "grad_norm": 0.41436522591687447, + "learning_rate": 0.0006410282269819756, + "loss": 2.4113, + "step": 931 + }, + { + "epoch": 0.42732691425951397, + "grad_norm": 0.43699425190910846, + "learning_rate": 0.0006403155333396787, + "loss": 2.5029, + "step": 932 + }, + { + "epoch": 0.42778541953232463, + "grad_norm": 0.42082779190785885, + "learning_rate": 0.0006396025301089863, + "loss": 2.4924, + "step": 933 + }, + { + "epoch": 0.42824392480513523, + "grad_norm": 0.42417189669091726, + "learning_rate": 0.0006388892188630493, + "loss": 2.3887, + "step": 934 + }, + { + "epoch": 0.4287024300779459, + "grad_norm": 0.411756716730507, + "learning_rate": 0.0006381756011756982, + "loss": 2.432, + "step": 935 + }, + { + "epoch": 0.42916093535075656, + "grad_norm": 0.42563643201908996, + "learning_rate": 0.0006374616786214403, + "loss": 2.5264, + "step": 936 + }, + { + "epoch": 0.42961944062356716, + "grad_norm": 0.41197515060016193, + "learning_rate": 0.0006367474527754544, + "loss": 2.4584, + "step": 937 + }, + { + "epoch": 0.4300779458963778, + "grad_norm": 0.40121161025324886, + "learning_rate": 0.0006360329252135894, + "loss": 2.4463, + "step": 938 + }, + { + "epoch": 0.4305364511691884, + "grad_norm": 0.4473599001120281, + "learning_rate": 0.0006353180975123595, + "loss": 2.4925, + "step": 939 + }, + { + "epoch": 0.4309949564419991, + "grad_norm": 0.41796659322543156, + "learning_rate": 0.0006346029712489413, + "loss": 2.4409, + "step": 940 + }, + { + "epoch": 0.4314534617148097, + "grad_norm": 0.41381500518474723, + "learning_rate": 0.0006338875480011698, + "loss": 2.5233, + "step": 941 + }, + { + "epoch": 0.43191196698762035, + "grad_norm": 0.42683883152714014, + "learning_rate": 0.0006331718293475357, + "loss": 2.3276, + "step": 942 + }, + { + "epoch": 0.432370472260431, + "grad_norm": 0.42604509679693797, + "learning_rate": 0.0006324558168671811, + "loss": 2.4186, + "step": 943 + }, + { + "epoch": 0.4328289775332416, + "grad_norm": 0.4408437324732849, + "learning_rate": 0.0006317395121398968, + "loss": 2.5047, + "step": 944 + }, + { + "epoch": 0.4332874828060523, + "grad_norm": 0.4251881503378964, + "learning_rate": 0.0006310229167461179, + "loss": 2.4376, + "step": 945 + }, + { + "epoch": 0.4337459880788629, + "grad_norm": 0.4332850205280687, + "learning_rate": 0.0006303060322669214, + "loss": 2.3081, + "step": 946 + }, + { + "epoch": 0.43420449335167355, + "grad_norm": 0.4631455753997376, + "learning_rate": 0.0006295888602840214, + "loss": 2.6062, + "step": 947 + }, + { + "epoch": 0.4346629986244842, + "grad_norm": 0.4293641462728073, + "learning_rate": 0.0006288714023797671, + "loss": 2.4223, + "step": 948 + }, + { + "epoch": 0.4351215038972948, + "grad_norm": 0.45025330292079613, + "learning_rate": 0.000628153660137138, + "loss": 2.5619, + "step": 949 + }, + { + "epoch": 0.4355800091701055, + "grad_norm": 0.4430133035033749, + "learning_rate": 0.0006274356351397413, + "loss": 2.4555, + "step": 950 + }, + { + "epoch": 0.4360385144429161, + "grad_norm": 0.4122775894683527, + "learning_rate": 0.0006267173289718079, + "loss": 2.3888, + "step": 951 + }, + { + "epoch": 0.43649701971572674, + "grad_norm": 0.4380038029015926, + "learning_rate": 0.000625998743218189, + "loss": 2.4985, + "step": 952 + }, + { + "epoch": 0.43695552498853735, + "grad_norm": 0.43381527327603897, + "learning_rate": 0.000625279879464353, + "loss": 2.4184, + "step": 953 + }, + { + "epoch": 0.437414030261348, + "grad_norm": 0.4459554355933446, + "learning_rate": 0.000624560739296381, + "loss": 2.4471, + "step": 954 + }, + { + "epoch": 0.43787253553415867, + "grad_norm": 0.43022944319185824, + "learning_rate": 0.0006238413243009648, + "loss": 2.3303, + "step": 955 + }, + { + "epoch": 0.4383310408069693, + "grad_norm": 0.41901609544188495, + "learning_rate": 0.000623121636065402, + "loss": 2.4529, + "step": 956 + }, + { + "epoch": 0.43878954607977994, + "grad_norm": 0.41610328048598777, + "learning_rate": 0.0006224016761775933, + "loss": 2.4084, + "step": 957 + }, + { + "epoch": 0.43924805135259054, + "grad_norm": 0.43742798234840885, + "learning_rate": 0.0006216814462260386, + "loss": 2.5043, + "step": 958 + }, + { + "epoch": 0.4397065566254012, + "grad_norm": 0.4145624168400819, + "learning_rate": 0.0006209609477998338, + "loss": 2.4344, + "step": 959 + }, + { + "epoch": 0.4401650618982118, + "grad_norm": 0.4315225796599489, + "learning_rate": 0.0006202401824886674, + "loss": 2.4558, + "step": 960 + }, + { + "epoch": 0.44062356717102247, + "grad_norm": 0.4434635808188439, + "learning_rate": 0.0006195191518828162, + "loss": 2.4585, + "step": 961 + }, + { + "epoch": 0.44108207244383313, + "grad_norm": 0.41721799948597815, + "learning_rate": 0.0006187978575731427, + "loss": 2.5101, + "step": 962 + }, + { + "epoch": 0.44154057771664373, + "grad_norm": 0.42897196480280647, + "learning_rate": 0.0006180763011510911, + "loss": 2.4697, + "step": 963 + }, + { + "epoch": 0.4419990829894544, + "grad_norm": 0.439334042627212, + "learning_rate": 0.000617354484208684, + "loss": 2.3677, + "step": 964 + }, + { + "epoch": 0.442457588262265, + "grad_norm": 0.415316062520926, + "learning_rate": 0.0006166324083385189, + "loss": 2.4499, + "step": 965 + }, + { + "epoch": 0.44291609353507566, + "grad_norm": 0.42263167307296734, + "learning_rate": 0.0006159100751337642, + "loss": 2.4567, + "step": 966 + }, + { + "epoch": 0.44337459880788627, + "grad_norm": 0.4184522966531005, + "learning_rate": 0.0006151874861881565, + "loss": 2.4919, + "step": 967 + }, + { + "epoch": 0.4438331040806969, + "grad_norm": 0.4157000190189219, + "learning_rate": 0.0006144646430959964, + "loss": 2.4148, + "step": 968 + }, + { + "epoch": 0.4442916093535076, + "grad_norm": 0.4183174155501703, + "learning_rate": 0.0006137415474521454, + "loss": 2.3947, + "step": 969 + }, + { + "epoch": 0.4447501146263182, + "grad_norm": 0.4262512762709655, + "learning_rate": 0.0006130182008520222, + "loss": 2.4391, + "step": 970 + }, + { + "epoch": 0.44520861989912885, + "grad_norm": 0.4076522038075935, + "learning_rate": 0.000612294604891599, + "loss": 2.4307, + "step": 971 + }, + { + "epoch": 0.44566712517193946, + "grad_norm": 0.43262603150104656, + "learning_rate": 0.0006115707611673986, + "loss": 2.4234, + "step": 972 + }, + { + "epoch": 0.4461256304447501, + "grad_norm": 0.4415934249437335, + "learning_rate": 0.0006108466712764902, + "loss": 2.5654, + "step": 973 + }, + { + "epoch": 0.4465841357175607, + "grad_norm": 0.4036737283686636, + "learning_rate": 0.0006101223368164858, + "loss": 2.482, + "step": 974 + }, + { + "epoch": 0.4470426409903714, + "grad_norm": 0.41097628908177264, + "learning_rate": 0.0006093977593855375, + "loss": 2.5321, + "step": 975 + }, + { + "epoch": 0.44750114626318205, + "grad_norm": 0.40538561888467367, + "learning_rate": 0.0006086729405823335, + "loss": 2.4253, + "step": 976 + }, + { + "epoch": 0.44795965153599265, + "grad_norm": 0.4406007558201177, + "learning_rate": 0.0006079478820060943, + "loss": 2.4442, + "step": 977 + }, + { + "epoch": 0.4484181568088033, + "grad_norm": 0.3955245029200891, + "learning_rate": 0.0006072225852565695, + "loss": 2.439, + "step": 978 + }, + { + "epoch": 0.4488766620816139, + "grad_norm": 0.4198640918065073, + "learning_rate": 0.0006064970519340341, + "loss": 2.4164, + "step": 979 + }, + { + "epoch": 0.4493351673544246, + "grad_norm": 0.4185686235292855, + "learning_rate": 0.0006057712836392856, + "loss": 2.4749, + "step": 980 + }, + { + "epoch": 0.4497936726272352, + "grad_norm": 0.4035696153662511, + "learning_rate": 0.0006050452819736389, + "loss": 2.3952, + "step": 981 + }, + { + "epoch": 0.45025217790004585, + "grad_norm": 0.4157923288882552, + "learning_rate": 0.000604319048538925, + "loss": 2.4154, + "step": 982 + }, + { + "epoch": 0.4507106831728565, + "grad_norm": 0.44801126065633123, + "learning_rate": 0.0006035925849374855, + "loss": 2.4328, + "step": 983 + }, + { + "epoch": 0.4511691884456671, + "grad_norm": 0.4211254851700999, + "learning_rate": 0.0006028658927721697, + "loss": 2.4538, + "step": 984 + }, + { + "epoch": 0.4516276937184778, + "grad_norm": 0.4034935442677631, + "learning_rate": 0.0006021389736463321, + "loss": 2.3944, + "step": 985 + }, + { + "epoch": 0.4520861989912884, + "grad_norm": 0.4222876559616626, + "learning_rate": 0.0006014118291638271, + "loss": 2.3888, + "step": 986 + }, + { + "epoch": 0.45254470426409904, + "grad_norm": 0.43733044088909556, + "learning_rate": 0.0006006844609290065, + "loss": 2.4716, + "step": 987 + }, + { + "epoch": 0.4530032095369097, + "grad_norm": 0.4212555462902067, + "learning_rate": 0.0005999568705467161, + "loss": 2.4524, + "step": 988 + }, + { + "epoch": 0.4534617148097203, + "grad_norm": 0.4055148296274541, + "learning_rate": 0.0005992290596222915, + "loss": 2.3858, + "step": 989 + }, + { + "epoch": 0.45392022008253097, + "grad_norm": 0.4168126763442859, + "learning_rate": 0.0005985010297615551, + "loss": 2.3576, + "step": 990 + }, + { + "epoch": 0.4543787253553416, + "grad_norm": 0.43312973058796905, + "learning_rate": 0.0005977727825708123, + "loss": 2.4994, + "step": 991 + }, + { + "epoch": 0.45483723062815223, + "grad_norm": 0.40119563840108674, + "learning_rate": 0.0005970443196568478, + "loss": 2.4728, + "step": 992 + }, + { + "epoch": 0.45529573590096284, + "grad_norm": 0.4766879998644203, + "learning_rate": 0.0005963156426269227, + "loss": 2.4909, + "step": 993 + }, + { + "epoch": 0.4557542411737735, + "grad_norm": 0.4258237659483852, + "learning_rate": 0.0005955867530887702, + "loss": 2.3791, + "step": 994 + }, + { + "epoch": 0.45621274644658416, + "grad_norm": 0.4199762163653014, + "learning_rate": 0.0005948576526505923, + "loss": 2.4072, + "step": 995 + }, + { + "epoch": 0.45667125171939477, + "grad_norm": 0.4125010629876395, + "learning_rate": 0.0005941283429210568, + "loss": 2.3402, + "step": 996 + }, + { + "epoch": 0.4571297569922054, + "grad_norm": 0.41189251692338746, + "learning_rate": 0.0005933988255092926, + "loss": 2.446, + "step": 997 + }, + { + "epoch": 0.45758826226501603, + "grad_norm": 0.4180986388373414, + "learning_rate": 0.0005926691020248874, + "loss": 2.4318, + "step": 998 + }, + { + "epoch": 0.4580467675378267, + "grad_norm": 0.43110241363234186, + "learning_rate": 0.0005919391740778833, + "loss": 2.3881, + "step": 999 + }, + { + "epoch": 0.4585052728106373, + "grad_norm": 0.4174689002692924, + "learning_rate": 0.0005912090432787736, + "loss": 2.4503, + "step": 1000 + }, + { + "epoch": 0.45896377808344796, + "grad_norm": 0.4305065298333914, + "learning_rate": 0.000590478711238499, + "loss": 2.4444, + "step": 1001 + }, + { + "epoch": 0.4594222833562586, + "grad_norm": 0.4172323006591254, + "learning_rate": 0.0005897481795684446, + "loss": 2.4105, + "step": 1002 + }, + { + "epoch": 0.4598807886290692, + "grad_norm": 0.415859684854009, + "learning_rate": 0.0005890174498804355, + "loss": 2.4303, + "step": 1003 + }, + { + "epoch": 0.4603392939018799, + "grad_norm": 0.41127451008820465, + "learning_rate": 0.0005882865237867339, + "loss": 2.3789, + "step": 1004 + }, + { + "epoch": 0.4607977991746905, + "grad_norm": 0.4048152439015935, + "learning_rate": 0.0005875554029000353, + "loss": 2.4635, + "step": 1005 + }, + { + "epoch": 0.46125630444750115, + "grad_norm": 0.4056469173049896, + "learning_rate": 0.0005868240888334653, + "loss": 2.409, + "step": 1006 + }, + { + "epoch": 0.46171480972031176, + "grad_norm": 0.4281850081111004, + "learning_rate": 0.0005860925832005753, + "loss": 2.4945, + "step": 1007 + }, + { + "epoch": 0.4621733149931224, + "grad_norm": 0.4387556841176759, + "learning_rate": 0.0005853608876153395, + "loss": 2.4755, + "step": 1008 + }, + { + "epoch": 0.4626318202659331, + "grad_norm": 0.4086676169664769, + "learning_rate": 0.0005846290036921512, + "loss": 2.3744, + "step": 1009 + }, + { + "epoch": 0.4630903255387437, + "grad_norm": 0.4194665090715104, + "learning_rate": 0.0005838969330458195, + "loss": 2.4307, + "step": 1010 + }, + { + "epoch": 0.46354883081155435, + "grad_norm": 0.43178707290717416, + "learning_rate": 0.0005831646772915651, + "loss": 2.426, + "step": 1011 + }, + { + "epoch": 0.46400733608436495, + "grad_norm": 0.41138806342224976, + "learning_rate": 0.0005824322380450173, + "loss": 2.4345, + "step": 1012 + }, + { + "epoch": 0.4644658413571756, + "grad_norm": 0.4333612435082081, + "learning_rate": 0.0005816996169222102, + "loss": 2.4367, + "step": 1013 + }, + { + "epoch": 0.4649243466299862, + "grad_norm": 0.4086058226797623, + "learning_rate": 0.0005809668155395793, + "loss": 2.4687, + "step": 1014 + }, + { + "epoch": 0.4653828519027969, + "grad_norm": 0.42318127807347705, + "learning_rate": 0.0005802338355139578, + "loss": 2.425, + "step": 1015 + }, + { + "epoch": 0.46584135717560754, + "grad_norm": 0.4514777178499627, + "learning_rate": 0.0005795006784625728, + "loss": 2.4771, + "step": 1016 + }, + { + "epoch": 0.46629986244841815, + "grad_norm": 0.4496035348047254, + "learning_rate": 0.0005787673460030423, + "loss": 2.4857, + "step": 1017 + }, + { + "epoch": 0.4667583677212288, + "grad_norm": 0.43447456564154663, + "learning_rate": 0.000578033839753371, + "loss": 2.5302, + "step": 1018 + }, + { + "epoch": 0.4672168729940394, + "grad_norm": 0.4382688682510541, + "learning_rate": 0.0005773001613319476, + "loss": 2.3352, + "step": 1019 + }, + { + "epoch": 0.4676753782668501, + "grad_norm": 0.439706130046969, + "learning_rate": 0.00057656631235754, + "loss": 2.4839, + "step": 1020 + }, + { + "epoch": 0.46813388353966073, + "grad_norm": 0.39640402948746606, + "learning_rate": 0.0005758322944492929, + "loss": 2.4817, + "step": 1021 + }, + { + "epoch": 0.46859238881247134, + "grad_norm": 0.4338168126762168, + "learning_rate": 0.0005750981092267237, + "loss": 2.4294, + "step": 1022 + }, + { + "epoch": 0.469050894085282, + "grad_norm": 0.43149294203915206, + "learning_rate": 0.0005743637583097183, + "loss": 2.3545, + "step": 1023 + }, + { + "epoch": 0.4695093993580926, + "grad_norm": 0.42741280627378525, + "learning_rate": 0.0005736292433185291, + "loss": 2.3914, + "step": 1024 + }, + { + "epoch": 0.46996790463090327, + "grad_norm": 0.42778138605643434, + "learning_rate": 0.0005728945658737699, + "loss": 2.5287, + "step": 1025 + }, + { + "epoch": 0.47042640990371387, + "grad_norm": 0.3985970183067096, + "learning_rate": 0.0005721597275964133, + "loss": 2.4419, + "step": 1026 + }, + { + "epoch": 0.47088491517652453, + "grad_norm": 0.4122815844368773, + "learning_rate": 0.0005714247301077865, + "loss": 2.4456, + "step": 1027 + }, + { + "epoch": 0.4713434204493352, + "grad_norm": 0.3929501592625896, + "learning_rate": 0.0005706895750295682, + "loss": 2.489, + "step": 1028 + }, + { + "epoch": 0.4718019257221458, + "grad_norm": 0.4210731953673899, + "learning_rate": 0.0005699542639837844, + "loss": 2.4164, + "step": 1029 + }, + { + "epoch": 0.47226043099495646, + "grad_norm": 0.39593692712650164, + "learning_rate": 0.0005692187985928055, + "loss": 2.5152, + "step": 1030 + }, + { + "epoch": 0.47271893626776706, + "grad_norm": 0.41604386740844285, + "learning_rate": 0.0005684831804793427, + "loss": 2.6957, + "step": 1031 + }, + { + "epoch": 0.4731774415405777, + "grad_norm": 0.4076212575715674, + "learning_rate": 0.0005677474112664438, + "loss": 2.3944, + "step": 1032 + }, + { + "epoch": 0.47363594681338833, + "grad_norm": 0.437012102350674, + "learning_rate": 0.0005670114925774899, + "loss": 2.4992, + "step": 1033 + }, + { + "epoch": 0.474094452086199, + "grad_norm": 0.4086462692974209, + "learning_rate": 0.0005662754260361924, + "loss": 2.4159, + "step": 1034 + }, + { + "epoch": 0.47455295735900965, + "grad_norm": 0.4339012609288313, + "learning_rate": 0.0005655392132665884, + "loss": 2.4536, + "step": 1035 + }, + { + "epoch": 0.47501146263182026, + "grad_norm": 0.4455226059547547, + "learning_rate": 0.000564802855893038, + "loss": 2.4403, + "step": 1036 + }, + { + "epoch": 0.4754699679046309, + "grad_norm": 0.4335570413820471, + "learning_rate": 0.0005640663555402198, + "loss": 2.5226, + "step": 1037 + }, + { + "epoch": 0.4759284731774415, + "grad_norm": 0.42028070840002757, + "learning_rate": 0.0005633297138331285, + "loss": 2.5152, + "step": 1038 + }, + { + "epoch": 0.4763869784502522, + "grad_norm": 0.41693122605770033, + "learning_rate": 0.0005625929323970705, + "loss": 2.6374, + "step": 1039 + }, + { + "epoch": 0.4768454837230628, + "grad_norm": 0.42951937930275125, + "learning_rate": 0.0005618560128576603, + "loss": 2.454, + "step": 1040 + }, + { + "epoch": 0.47730398899587345, + "grad_norm": 0.42076280401585087, + "learning_rate": 0.0005611189568408173, + "loss": 2.469, + "step": 1041 + }, + { + "epoch": 0.4777624942686841, + "grad_norm": 0.42348925908271984, + "learning_rate": 0.0005603817659727619, + "loss": 2.4013, + "step": 1042 + }, + { + "epoch": 0.4782209995414947, + "grad_norm": 0.42179333124820045, + "learning_rate": 0.0005596444418800121, + "loss": 2.3904, + "step": 1043 + }, + { + "epoch": 0.4786795048143054, + "grad_norm": 0.42154536294338973, + "learning_rate": 0.0005589069861893798, + "loss": 2.413, + "step": 1044 + }, + { + "epoch": 0.479138010087116, + "grad_norm": 0.4168781838257941, + "learning_rate": 0.0005581694005279673, + "loss": 2.3966, + "step": 1045 + }, + { + "epoch": 0.47959651535992665, + "grad_norm": 0.4142585435934598, + "learning_rate": 0.0005574316865231637, + "loss": 2.3694, + "step": 1046 + }, + { + "epoch": 0.48005502063273725, + "grad_norm": 0.43041026263615, + "learning_rate": 0.0005566938458026411, + "loss": 2.4224, + "step": 1047 + }, + { + "epoch": 0.4805135259055479, + "grad_norm": 0.4409984668805699, + "learning_rate": 0.0005559558799943514, + "loss": 2.3836, + "step": 1048 + }, + { + "epoch": 0.48097203117835857, + "grad_norm": 0.44083432909768117, + "learning_rate": 0.0005552177907265223, + "loss": 2.2774, + "step": 1049 + }, + { + "epoch": 0.4814305364511692, + "grad_norm": 0.4214104523334741, + "learning_rate": 0.000554479579627654, + "loss": 2.3719, + "step": 1050 + }, + { + "epoch": 0.48188904172397984, + "grad_norm": 0.44743333612346364, + "learning_rate": 0.0005537412483265157, + "loss": 2.4618, + "step": 1051 + }, + { + "epoch": 0.48234754699679044, + "grad_norm": 0.44216114571072834, + "learning_rate": 0.0005530027984521413, + "loss": 2.4233, + "step": 1052 + }, + { + "epoch": 0.4828060522696011, + "grad_norm": 0.4408158311269601, + "learning_rate": 0.0005522642316338268, + "loss": 2.373, + "step": 1053 + }, + { + "epoch": 0.4832645575424117, + "grad_norm": 0.43429509272032113, + "learning_rate": 0.0005515255495011259, + "loss": 2.4392, + "step": 1054 + }, + { + "epoch": 0.48372306281522237, + "grad_norm": 0.426799106836899, + "learning_rate": 0.0005507867536838472, + "loss": 2.3799, + "step": 1055 + }, + { + "epoch": 0.48418156808803303, + "grad_norm": 0.43613272928304603, + "learning_rate": 0.0005500478458120492, + "loss": 2.3589, + "step": 1056 + }, + { + "epoch": 0.48464007336084364, + "grad_norm": 0.4009922656618402, + "learning_rate": 0.0005493088275160387, + "loss": 2.3819, + "step": 1057 + }, + { + "epoch": 0.4850985786336543, + "grad_norm": 0.4028180033235332, + "learning_rate": 0.0005485697004263657, + "loss": 2.4534, + "step": 1058 + }, + { + "epoch": 0.4855570839064649, + "grad_norm": 0.4153851872199973, + "learning_rate": 0.0005478304661738199, + "loss": 2.5087, + "step": 1059 + }, + { + "epoch": 0.48601558917927556, + "grad_norm": 0.41024736956183533, + "learning_rate": 0.0005470911263894279, + "loss": 2.3933, + "step": 1060 + }, + { + "epoch": 0.4864740944520862, + "grad_norm": 0.43065837663889145, + "learning_rate": 0.0005463516827044491, + "loss": 2.4677, + "step": 1061 + }, + { + "epoch": 0.48693259972489683, + "grad_norm": 0.40847187315833844, + "learning_rate": 0.000545612136750372, + "loss": 2.3939, + "step": 1062 + }, + { + "epoch": 0.4873911049977075, + "grad_norm": 0.410831776276338, + "learning_rate": 0.0005448724901589107, + "loss": 2.5989, + "step": 1063 + }, + { + "epoch": 0.4878496102705181, + "grad_norm": 0.4102486599975156, + "learning_rate": 0.0005441327445620014, + "loss": 2.4334, + "step": 1064 + }, + { + "epoch": 0.48830811554332876, + "grad_norm": 0.43474796647350955, + "learning_rate": 0.0005433929015917988, + "loss": 2.4129, + "step": 1065 + }, + { + "epoch": 0.48876662081613936, + "grad_norm": 0.39287899477810795, + "learning_rate": 0.0005426529628806724, + "loss": 2.355, + "step": 1066 + }, + { + "epoch": 0.48922512608895, + "grad_norm": 0.4178330825724143, + "learning_rate": 0.0005419129300612029, + "loss": 2.4508, + "step": 1067 + }, + { + "epoch": 0.4896836313617607, + "grad_norm": 0.4254773463754292, + "learning_rate": 0.000541172804766179, + "loss": 2.5161, + "step": 1068 + }, + { + "epoch": 0.4901421366345713, + "grad_norm": 0.42459034897401177, + "learning_rate": 0.0005404325886285927, + "loss": 2.4052, + "step": 1069 + }, + { + "epoch": 0.49060064190738195, + "grad_norm": 0.43503293289261946, + "learning_rate": 0.000539692283281637, + "loss": 2.4615, + "step": 1070 + }, + { + "epoch": 0.49105914718019256, + "grad_norm": 0.4096261569772868, + "learning_rate": 0.0005389518903587017, + "loss": 2.4262, + "step": 1071 + }, + { + "epoch": 0.4915176524530032, + "grad_norm": 0.41422546442297886, + "learning_rate": 0.0005382114114933695, + "loss": 2.3768, + "step": 1072 + }, + { + "epoch": 0.4919761577258138, + "grad_norm": 0.42475551506981873, + "learning_rate": 0.0005374708483194132, + "loss": 2.4379, + "step": 1073 + }, + { + "epoch": 0.4924346629986245, + "grad_norm": 0.4430424927414206, + "learning_rate": 0.000536730202470791, + "loss": 2.3876, + "step": 1074 + }, + { + "epoch": 0.49289316827143514, + "grad_norm": 0.4435238674841608, + "learning_rate": 0.0005359894755816443, + "loss": 2.407, + "step": 1075 + }, + { + "epoch": 0.49335167354424575, + "grad_norm": 0.42004130544722584, + "learning_rate": 0.0005352486692862926, + "loss": 2.4557, + "step": 1076 + }, + { + "epoch": 0.4938101788170564, + "grad_norm": 0.4232923883681415, + "learning_rate": 0.0005345077852192307, + "loss": 2.4753, + "step": 1077 + }, + { + "epoch": 0.494268684089867, + "grad_norm": 0.42028384271629715, + "learning_rate": 0.0005337668250151254, + "loss": 2.3948, + "step": 1078 + }, + { + "epoch": 0.4947271893626777, + "grad_norm": 0.4180545270439937, + "learning_rate": 0.0005330257903088111, + "loss": 2.398, + "step": 1079 + }, + { + "epoch": 0.4951856946354883, + "grad_norm": 0.42069580733444367, + "learning_rate": 0.000532284682735287, + "loss": 2.3217, + "step": 1080 + }, + { + "epoch": 0.49564419990829894, + "grad_norm": 0.401480574034057, + "learning_rate": 0.0005315435039297124, + "loss": 2.4659, + "step": 1081 + }, + { + "epoch": 0.4961027051811096, + "grad_norm": 0.39200470808871507, + "learning_rate": 0.0005308022555274046, + "loss": 2.4004, + "step": 1082 + }, + { + "epoch": 0.4965612104539202, + "grad_norm": 0.3960445604580187, + "learning_rate": 0.0005300609391638336, + "loss": 2.3397, + "step": 1083 + }, + { + "epoch": 0.49701971572673087, + "grad_norm": 0.40631291327573205, + "learning_rate": 0.0005293195564746201, + "loss": 2.384, + "step": 1084 + }, + { + "epoch": 0.4974782209995415, + "grad_norm": 0.42455583345474124, + "learning_rate": 0.0005285781090955304, + "loss": 2.4216, + "step": 1085 + }, + { + "epoch": 0.49793672627235214, + "grad_norm": 0.4274050465947088, + "learning_rate": 0.0005278365986624743, + "loss": 2.3857, + "step": 1086 + }, + { + "epoch": 0.49839523154516274, + "grad_norm": 0.4090529421360174, + "learning_rate": 0.0005270950268115001, + "loss": 2.3778, + "step": 1087 + }, + { + "epoch": 0.4988537368179734, + "grad_norm": 0.3990351106611573, + "learning_rate": 0.0005263533951787919, + "loss": 2.4297, + "step": 1088 + }, + { + "epoch": 0.49931224209078406, + "grad_norm": 0.40989177129705745, + "learning_rate": 0.000525611705400666, + "loss": 2.4807, + "step": 1089 + }, + { + "epoch": 0.49977074736359467, + "grad_norm": 0.3989371959962871, + "learning_rate": 0.0005248699591135664, + "loss": 2.4152, + "step": 1090 + }, + { + "epoch": 0.5002292526364053, + "grad_norm": 0.4213791754833137, + "learning_rate": 0.0005241281579540618, + "loss": 2.3985, + "step": 1091 + }, + { + "epoch": 0.5006877579092159, + "grad_norm": 0.4204395178157146, + "learning_rate": 0.0005233863035588427, + "loss": 2.4836, + "step": 1092 + }, + { + "epoch": 0.5011462631820266, + "grad_norm": 0.42768598894130605, + "learning_rate": 0.0005226443975647161, + "loss": 2.4907, + "step": 1093 + }, + { + "epoch": 0.5016047684548373, + "grad_norm": 0.44396151890901986, + "learning_rate": 0.0005219024416086036, + "loss": 2.361, + "step": 1094 + }, + { + "epoch": 0.5020632737276479, + "grad_norm": 0.4074935978079562, + "learning_rate": 0.0005211604373275366, + "loss": 2.4634, + "step": 1095 + }, + { + "epoch": 0.5025217790004585, + "grad_norm": 0.43346120915165387, + "learning_rate": 0.0005204183863586533, + "loss": 2.3289, + "step": 1096 + }, + { + "epoch": 0.5029802842732691, + "grad_norm": 0.4355852239558619, + "learning_rate": 0.0005196762903391951, + "loss": 2.4448, + "step": 1097 + }, + { + "epoch": 0.5034387895460798, + "grad_norm": 0.4223224206685133, + "learning_rate": 0.0005189341509065023, + "loss": 2.4114, + "step": 1098 + }, + { + "epoch": 0.5038972948188905, + "grad_norm": 0.42651435000950355, + "learning_rate": 0.0005181919696980112, + "loss": 2.4598, + "step": 1099 + }, + { + "epoch": 0.504355800091701, + "grad_norm": 0.4167058271890261, + "learning_rate": 0.0005174497483512506, + "loss": 2.4507, + "step": 1100 + }, + { + "epoch": 0.5048143053645117, + "grad_norm": 0.4440844598397881, + "learning_rate": 0.0005167074885038374, + "loss": 2.3276, + "step": 1101 + }, + { + "epoch": 0.5052728106373223, + "grad_norm": 0.41065871365036555, + "learning_rate": 0.0005159651917934735, + "loss": 2.3806, + "step": 1102 + }, + { + "epoch": 0.505731315910133, + "grad_norm": 0.43739903378069966, + "learning_rate": 0.0005152228598579428, + "loss": 2.4906, + "step": 1103 + }, + { + "epoch": 0.5061898211829436, + "grad_norm": 0.43945756168245137, + "learning_rate": 0.000514480494335106, + "loss": 2.4337, + "step": 1104 + }, + { + "epoch": 0.5066483264557542, + "grad_norm": 0.4011718512219828, + "learning_rate": 0.0005137380968628983, + "loss": 2.4363, + "step": 1105 + }, + { + "epoch": 0.5071068317285649, + "grad_norm": 0.43228694428934294, + "learning_rate": 0.0005129956690793255, + "loss": 2.4041, + "step": 1106 + }, + { + "epoch": 0.5075653370013755, + "grad_norm": 0.40497768710220333, + "learning_rate": 0.0005122532126224601, + "loss": 2.4573, + "step": 1107 + }, + { + "epoch": 0.5080238422741862, + "grad_norm": 0.43540954857213454, + "learning_rate": 0.0005115107291304378, + "loss": 2.4599, + "step": 1108 + }, + { + "epoch": 0.5084823475469968, + "grad_norm": 0.42536343585824943, + "learning_rate": 0.0005107682202414544, + "loss": 2.4389, + "step": 1109 + }, + { + "epoch": 0.5089408528198074, + "grad_norm": 0.4168653742016338, + "learning_rate": 0.0005100256875937613, + "loss": 2.3794, + "step": 1110 + }, + { + "epoch": 0.509399358092618, + "grad_norm": 0.43692119289457904, + "learning_rate": 0.0005092831328256625, + "loss": 2.495, + "step": 1111 + }, + { + "epoch": 0.5098578633654287, + "grad_norm": 0.4183639185745104, + "learning_rate": 0.0005085405575755105, + "loss": 2.4262, + "step": 1112 + }, + { + "epoch": 0.5103163686382394, + "grad_norm": 0.42262479990440116, + "learning_rate": 0.0005077979634817034, + "loss": 2.3694, + "step": 1113 + }, + { + "epoch": 0.51077487391105, + "grad_norm": 0.4289990583311305, + "learning_rate": 0.0005070553521826808, + "loss": 2.4044, + "step": 1114 + }, + { + "epoch": 0.5112333791838606, + "grad_norm": 0.42160047806562606, + "learning_rate": 0.00050631272531692, + "loss": 2.5021, + "step": 1115 + }, + { + "epoch": 0.5116918844566712, + "grad_norm": 0.40915063823767395, + "learning_rate": 0.0005055700845229327, + "loss": 2.3841, + "step": 1116 + }, + { + "epoch": 0.5121503897294819, + "grad_norm": 0.4248349790077219, + "learning_rate": 0.000504827431439262, + "loss": 2.4921, + "step": 1117 + }, + { + "epoch": 0.5126088950022926, + "grad_norm": 0.43011410733312616, + "learning_rate": 0.000504084767704477, + "loss": 2.4655, + "step": 1118 + }, + { + "epoch": 0.5130674002751031, + "grad_norm": 0.4308485794962944, + "learning_rate": 0.0005033420949571712, + "loss": 2.384, + "step": 1119 + }, + { + "epoch": 0.5135259055479138, + "grad_norm": 0.4185162669801408, + "learning_rate": 0.0005025994148359574, + "loss": 2.4343, + "step": 1120 + }, + { + "epoch": 0.5139844108207244, + "grad_norm": 0.4528421130918301, + "learning_rate": 0.0005018567289794651, + "loss": 2.3319, + "step": 1121 + }, + { + "epoch": 0.5144429160935351, + "grad_norm": 0.44210096458602655, + "learning_rate": 0.0005011140390263362, + "loss": 2.4763, + "step": 1122 + }, + { + "epoch": 0.5149014213663458, + "grad_norm": 0.4228644190126658, + "learning_rate": 0.0005003713466152218, + "loss": 2.4128, + "step": 1123 + }, + { + "epoch": 0.5153599266391563, + "grad_norm": 0.4122145406156049, + "learning_rate": 0.0004996286533847783, + "loss": 2.4471, + "step": 1124 + }, + { + "epoch": 0.515818431911967, + "grad_norm": 0.43360559706676877, + "learning_rate": 0.000498885960973664, + "loss": 2.4475, + "step": 1125 + }, + { + "epoch": 0.5162769371847776, + "grad_norm": 0.44091862393855424, + "learning_rate": 0.000498143271020535, + "loss": 2.3701, + "step": 1126 + }, + { + "epoch": 0.5167354424575883, + "grad_norm": 0.4523302545865001, + "learning_rate": 0.0004974005851640428, + "loss": 2.4919, + "step": 1127 + }, + { + "epoch": 0.517193947730399, + "grad_norm": 0.42990914867581687, + "learning_rate": 0.000496657905042829, + "loss": 2.4582, + "step": 1128 + }, + { + "epoch": 0.5176524530032095, + "grad_norm": 0.4179733942622785, + "learning_rate": 0.0004959152322955232, + "loss": 2.3994, + "step": 1129 + }, + { + "epoch": 0.5181109582760202, + "grad_norm": 0.44623974338741984, + "learning_rate": 0.0004951725685607382, + "loss": 2.5244, + "step": 1130 + }, + { + "epoch": 0.5185694635488308, + "grad_norm": 0.42585407955353294, + "learning_rate": 0.0004944299154770673, + "loss": 2.3396, + "step": 1131 + }, + { + "epoch": 0.5190279688216415, + "grad_norm": 0.43046240183135454, + "learning_rate": 0.0004936872746830802, + "loss": 2.3915, + "step": 1132 + }, + { + "epoch": 0.519486474094452, + "grad_norm": 0.4175346818937369, + "learning_rate": 0.0004929446478173195, + "loss": 2.4253, + "step": 1133 + }, + { + "epoch": 0.5199449793672627, + "grad_norm": 0.43542514503707397, + "learning_rate": 0.0004922020365182968, + "loss": 2.3982, + "step": 1134 + }, + { + "epoch": 0.5204034846400734, + "grad_norm": 0.41953127806524004, + "learning_rate": 0.0004914594424244897, + "loss": 2.4347, + "step": 1135 + }, + { + "epoch": 0.520861989912884, + "grad_norm": 0.42443348352558, + "learning_rate": 0.0004907168671743376, + "loss": 2.4036, + "step": 1136 + }, + { + "epoch": 0.5213204951856947, + "grad_norm": 0.4155995627545679, + "learning_rate": 0.0004899743124062387, + "loss": 2.4231, + "step": 1137 + }, + { + "epoch": 0.5217790004585052, + "grad_norm": 0.40153502374236255, + "learning_rate": 0.0004892317797585456, + "loss": 2.3622, + "step": 1138 + }, + { + "epoch": 0.5222375057313159, + "grad_norm": 0.4054381320281407, + "learning_rate": 0.0004884892708695623, + "loss": 2.3076, + "step": 1139 + }, + { + "epoch": 0.5226960110041265, + "grad_norm": 0.425334292401327, + "learning_rate": 0.0004877467873775402, + "loss": 2.4406, + "step": 1140 + }, + { + "epoch": 0.5231545162769372, + "grad_norm": 0.42671575287172664, + "learning_rate": 0.00048700433092067473, + "loss": 2.5038, + "step": 1141 + }, + { + "epoch": 0.5236130215497479, + "grad_norm": 0.42171791168667006, + "learning_rate": 0.0004862619031371019, + "loss": 2.419, + "step": 1142 + }, + { + "epoch": 0.5240715268225584, + "grad_norm": 0.42577492516199, + "learning_rate": 0.0004855195056648942, + "loss": 2.3746, + "step": 1143 + }, + { + "epoch": 0.5245300320953691, + "grad_norm": 0.4112096981341729, + "learning_rate": 0.00048477714014205734, + "loss": 2.3891, + "step": 1144 + }, + { + "epoch": 0.5249885373681797, + "grad_norm": 0.4192744874492646, + "learning_rate": 0.00048403480820652644, + "loss": 2.4588, + "step": 1145 + }, + { + "epoch": 0.5254470426409904, + "grad_norm": 0.41708504817637826, + "learning_rate": 0.0004832925114961629, + "loss": 2.4067, + "step": 1146 + }, + { + "epoch": 0.5259055479138011, + "grad_norm": 0.40874694104514625, + "learning_rate": 0.0004825502516487497, + "loss": 2.4382, + "step": 1147 + }, + { + "epoch": 0.5263640531866116, + "grad_norm": 0.43008584325583776, + "learning_rate": 0.00048180803030198896, + "loss": 2.5152, + "step": 1148 + }, + { + "epoch": 0.5268225584594223, + "grad_norm": 0.41956377090954955, + "learning_rate": 0.0004810658490934979, + "loss": 2.4406, + "step": 1149 + }, + { + "epoch": 0.5272810637322329, + "grad_norm": 0.39951456281535036, + "learning_rate": 0.000480323709660805, + "loss": 2.5324, + "step": 1150 + }, + { + "epoch": 0.5277395690050436, + "grad_norm": 0.4232307857892946, + "learning_rate": 0.0004795816136413467, + "loss": 2.3351, + "step": 1151 + }, + { + "epoch": 0.5281980742778541, + "grad_norm": 0.3946280738313916, + "learning_rate": 0.00047883956267246353, + "loss": 2.4476, + "step": 1152 + }, + { + "epoch": 0.5286565795506648, + "grad_norm": 0.40632490638960994, + "learning_rate": 0.00047809755839139657, + "loss": 2.4882, + "step": 1153 + }, + { + "epoch": 0.5291150848234755, + "grad_norm": 0.4165042346693669, + "learning_rate": 0.0004773556024352841, + "loss": 2.4496, + "step": 1154 + }, + { + "epoch": 0.5295735900962861, + "grad_norm": 0.39717188767670414, + "learning_rate": 0.00047661369644115754, + "loss": 2.2918, + "step": 1155 + }, + { + "epoch": 0.5300320953690968, + "grad_norm": 0.4017724837082513, + "learning_rate": 0.0004758718420459383, + "loss": 2.3303, + "step": 1156 + }, + { + "epoch": 0.5304906006419073, + "grad_norm": 0.41705175997157823, + "learning_rate": 0.0004751300408864339, + "loss": 2.3637, + "step": 1157 + }, + { + "epoch": 0.530949105914718, + "grad_norm": 0.4270112991945995, + "learning_rate": 0.00047438829459933414, + "loss": 2.4248, + "step": 1158 + }, + { + "epoch": 0.5314076111875287, + "grad_norm": 0.4354755314847969, + "learning_rate": 0.0004736466048212082, + "loss": 2.4469, + "step": 1159 + }, + { + "epoch": 0.5318661164603393, + "grad_norm": 0.44137126794010595, + "learning_rate": 0.0004729049731885002, + "loss": 2.4079, + "step": 1160 + }, + { + "epoch": 0.53232462173315, + "grad_norm": 0.41461329261228347, + "learning_rate": 0.000472163401337526, + "loss": 2.3814, + "step": 1161 + }, + { + "epoch": 0.5327831270059605, + "grad_norm": 0.39881945340483743, + "learning_rate": 0.00047142189090446985, + "loss": 2.3639, + "step": 1162 + }, + { + "epoch": 0.5332416322787712, + "grad_norm": 0.39863977895758385, + "learning_rate": 0.0004706804435253802, + "loss": 2.5, + "step": 1163 + }, + { + "epoch": 0.5337001375515819, + "grad_norm": 0.3934421233384836, + "learning_rate": 0.0004699390608361665, + "loss": 2.5496, + "step": 1164 + }, + { + "epoch": 0.5341586428243925, + "grad_norm": 0.4211535925046809, + "learning_rate": 0.0004691977444725955, + "loss": 2.3872, + "step": 1165 + }, + { + "epoch": 0.5346171480972031, + "grad_norm": 0.41411846526042706, + "learning_rate": 0.0004684564960702877, + "loss": 2.442, + "step": 1166 + }, + { + "epoch": 0.5350756533700137, + "grad_norm": 0.41843466353977454, + "learning_rate": 0.0004677153172647131, + "loss": 2.4155, + "step": 1167 + }, + { + "epoch": 0.5355341586428244, + "grad_norm": 0.41949560200536595, + "learning_rate": 0.00046697420969118894, + "loss": 2.4839, + "step": 1168 + }, + { + "epoch": 0.535992663915635, + "grad_norm": 0.44226464491724465, + "learning_rate": 0.00046623317498487466, + "loss": 2.3646, + "step": 1169 + }, + { + "epoch": 0.5364511691884457, + "grad_norm": 0.4254267871553535, + "learning_rate": 0.0004654922147807694, + "loss": 2.5243, + "step": 1170 + }, + { + "epoch": 0.5369096744612563, + "grad_norm": 0.4005885590089012, + "learning_rate": 0.00046475133071370757, + "loss": 2.3623, + "step": 1171 + }, + { + "epoch": 0.5373681797340669, + "grad_norm": 0.4180630865435286, + "learning_rate": 0.00046401052441835574, + "loss": 2.4149, + "step": 1172 + }, + { + "epoch": 0.5378266850068776, + "grad_norm": 0.4228801615758844, + "learning_rate": 0.000463269797529209, + "loss": 2.3082, + "step": 1173 + }, + { + "epoch": 0.5382851902796882, + "grad_norm": 0.439208347178024, + "learning_rate": 0.00046252915168058697, + "loss": 2.4581, + "step": 1174 + }, + { + "epoch": 0.5387436955524989, + "grad_norm": 0.43667845792317467, + "learning_rate": 0.0004617885885066305, + "loss": 2.3654, + "step": 1175 + }, + { + "epoch": 0.5392022008253095, + "grad_norm": 0.4249750706835056, + "learning_rate": 0.0004610481096412984, + "loss": 2.4309, + "step": 1176 + }, + { + "epoch": 0.5396607060981201, + "grad_norm": 0.4384429346328608, + "learning_rate": 0.000460307716718363, + "loss": 2.4344, + "step": 1177 + }, + { + "epoch": 0.5401192113709308, + "grad_norm": 0.4204166368379327, + "learning_rate": 0.0004595674113714074, + "loss": 2.3821, + "step": 1178 + }, + { + "epoch": 0.5405777166437414, + "grad_norm": 0.388024988101524, + "learning_rate": 0.0004588271952338212, + "loss": 2.422, + "step": 1179 + }, + { + "epoch": 0.5410362219165521, + "grad_norm": 0.4484455643528539, + "learning_rate": 0.00045808706993879714, + "loss": 2.5296, + "step": 1180 + }, + { + "epoch": 0.5414947271893626, + "grad_norm": 0.4239130297669497, + "learning_rate": 0.00045734703711932767, + "loss": 2.4033, + "step": 1181 + }, + { + "epoch": 0.5419532324621733, + "grad_norm": 0.4006731615258789, + "learning_rate": 0.0004566070984082013, + "loss": 2.3155, + "step": 1182 + }, + { + "epoch": 0.542411737734984, + "grad_norm": 0.41905585508439563, + "learning_rate": 0.00045586725543799865, + "loss": 2.332, + "step": 1183 + }, + { + "epoch": 0.5428702430077946, + "grad_norm": 0.4255969906288959, + "learning_rate": 0.00045512750984108937, + "loss": 2.4608, + "step": 1184 + }, + { + "epoch": 0.5433287482806052, + "grad_norm": 0.43134932133364257, + "learning_rate": 0.000454387863249628, + "loss": 2.5146, + "step": 1185 + }, + { + "epoch": 0.5437872535534158, + "grad_norm": 0.4322181768396837, + "learning_rate": 0.00045364831729555096, + "loss": 2.3839, + "step": 1186 + }, + { + "epoch": 0.5442457588262265, + "grad_norm": 0.421509951532718, + "learning_rate": 0.0004529088736105721, + "loss": 2.4769, + "step": 1187 + }, + { + "epoch": 0.5447042640990372, + "grad_norm": 0.4314573004383821, + "learning_rate": 0.0004521695338261802, + "loss": 2.4953, + "step": 1188 + }, + { + "epoch": 0.5451627693718478, + "grad_norm": 0.41857907096927205, + "learning_rate": 0.0004514302995736344, + "loss": 2.4382, + "step": 1189 + }, + { + "epoch": 0.5456212746446584, + "grad_norm": 0.4161952318300972, + "learning_rate": 0.0004506911724839613, + "loss": 2.4326, + "step": 1190 + }, + { + "epoch": 0.546079779917469, + "grad_norm": 0.4088750229356972, + "learning_rate": 0.0004499521541879508, + "loss": 2.3829, + "step": 1191 + }, + { + "epoch": 0.5465382851902797, + "grad_norm": 0.42579957721358525, + "learning_rate": 0.00044921324631615303, + "loss": 2.4237, + "step": 1192 + }, + { + "epoch": 0.5469967904630904, + "grad_norm": 0.39580779752364453, + "learning_rate": 0.0004484744504988742, + "loss": 2.3794, + "step": 1193 + }, + { + "epoch": 0.547455295735901, + "grad_norm": 0.4339877159964755, + "learning_rate": 0.00044773576836617336, + "loss": 2.3709, + "step": 1194 + }, + { + "epoch": 0.5479138010087116, + "grad_norm": 0.4023027922907635, + "learning_rate": 0.0004469972015478588, + "loss": 2.3603, + "step": 1195 + }, + { + "epoch": 0.5483723062815222, + "grad_norm": 0.42462404825301236, + "learning_rate": 0.0004462587516734844, + "loss": 2.3944, + "step": 1196 + }, + { + "epoch": 0.5488308115543329, + "grad_norm": 0.42322526316676307, + "learning_rate": 0.00044552042037234596, + "loss": 2.4502, + "step": 1197 + }, + { + "epoch": 0.5492893168271435, + "grad_norm": 0.4084344315778444, + "learning_rate": 0.00044478220927347774, + "loss": 2.4418, + "step": 1198 + }, + { + "epoch": 0.5497478220999541, + "grad_norm": 0.39643569997499556, + "learning_rate": 0.00044404412000564875, + "loss": 2.3366, + "step": 1199 + }, + { + "epoch": 0.5502063273727648, + "grad_norm": 0.41330130874657267, + "learning_rate": 0.000443306154197359, + "loss": 2.4151, + "step": 1200 + }, + { + "epoch": 0.5506648326455754, + "grad_norm": 0.4265712863983102, + "learning_rate": 0.00044256831347683646, + "loss": 2.3608, + "step": 1201 + }, + { + "epoch": 0.5511233379183861, + "grad_norm": 0.4604132222054217, + "learning_rate": 0.0004418305994720328, + "loss": 2.4365, + "step": 1202 + }, + { + "epoch": 0.5515818431911967, + "grad_norm": 0.4138135323200747, + "learning_rate": 0.0004410930138106203, + "loss": 2.3795, + "step": 1203 + }, + { + "epoch": 0.5520403484640073, + "grad_norm": 0.40483723254839465, + "learning_rate": 0.000440355558119988, + "loss": 2.4753, + "step": 1204 + }, + { + "epoch": 0.552498853736818, + "grad_norm": 0.4207648200167955, + "learning_rate": 0.00043961823402723814, + "loss": 2.3962, + "step": 1205 + }, + { + "epoch": 0.5529573590096286, + "grad_norm": 0.414009348082704, + "learning_rate": 0.0004388810431591829, + "loss": 2.5117, + "step": 1206 + }, + { + "epoch": 0.5534158642824393, + "grad_norm": 0.41547366836684524, + "learning_rate": 0.0004381439871423398, + "loss": 2.3772, + "step": 1207 + }, + { + "epoch": 0.5538743695552499, + "grad_norm": 0.3988463265471703, + "learning_rate": 0.00043740706760292966, + "loss": 2.339, + "step": 1208 + }, + { + "epoch": 0.5543328748280605, + "grad_norm": 0.4017851463529806, + "learning_rate": 0.0004366702861668716, + "loss": 2.4588, + "step": 1209 + }, + { + "epoch": 0.5547913801008711, + "grad_norm": 0.4118577442607429, + "learning_rate": 0.00043593364445978036, + "loss": 2.4618, + "step": 1210 + }, + { + "epoch": 0.5552498853736818, + "grad_norm": 0.43889433593670624, + "learning_rate": 0.0004351971441069622, + "loss": 2.3171, + "step": 1211 + }, + { + "epoch": 0.5557083906464925, + "grad_norm": 0.4257246763754424, + "learning_rate": 0.0004344607867334116, + "loss": 2.35, + "step": 1212 + }, + { + "epoch": 0.556166895919303, + "grad_norm": 0.4226328606885362, + "learning_rate": 0.00043372457396380766, + "loss": 2.3452, + "step": 1213 + }, + { + "epoch": 0.5566254011921137, + "grad_norm": 0.44706290530771764, + "learning_rate": 0.00043298850742251013, + "loss": 2.5629, + "step": 1214 + }, + { + "epoch": 0.5570839064649243, + "grad_norm": 0.4409315832175849, + "learning_rate": 0.0004322525887335563, + "loss": 2.3899, + "step": 1215 + }, + { + "epoch": 0.557542411737735, + "grad_norm": 0.44613886586678364, + "learning_rate": 0.00043151681952065734, + "loss": 2.4833, + "step": 1216 + }, + { + "epoch": 0.5580009170105457, + "grad_norm": 0.4577766925231534, + "learning_rate": 0.00043078120140719456, + "loss": 2.4894, + "step": 1217 + }, + { + "epoch": 0.5584594222833562, + "grad_norm": 0.42629338773408365, + "learning_rate": 0.0004300457360162158, + "loss": 2.4081, + "step": 1218 + }, + { + "epoch": 0.5589179275561669, + "grad_norm": 0.43448824522823054, + "learning_rate": 0.0004293104249704319, + "loss": 2.3373, + "step": 1219 + }, + { + "epoch": 0.5593764328289775, + "grad_norm": 0.4142372339306823, + "learning_rate": 0.00042857526989221355, + "loss": 2.4908, + "step": 1220 + }, + { + "epoch": 0.5598349381017882, + "grad_norm": 0.416683512709731, + "learning_rate": 0.00042784027240358674, + "loss": 2.4563, + "step": 1221 + }, + { + "epoch": 0.5602934433745989, + "grad_norm": 0.41795018120878635, + "learning_rate": 0.0004271054341262301, + "loss": 2.3877, + "step": 1222 + }, + { + "epoch": 0.5607519486474094, + "grad_norm": 0.4083402050720735, + "learning_rate": 0.000426370756681471, + "loss": 2.3858, + "step": 1223 + }, + { + "epoch": 0.5612104539202201, + "grad_norm": 0.41601017071949714, + "learning_rate": 0.0004256362416902817, + "loss": 2.4481, + "step": 1224 + }, + { + "epoch": 0.5616689591930307, + "grad_norm": 0.41561083747662997, + "learning_rate": 0.00042490189077327637, + "loss": 2.3872, + "step": 1225 + }, + { + "epoch": 0.5621274644658414, + "grad_norm": 0.43190645402800076, + "learning_rate": 0.00042416770555070703, + "loss": 2.4675, + "step": 1226 + }, + { + "epoch": 0.562585969738652, + "grad_norm": 0.4043585818303126, + "learning_rate": 0.00042343368764246, + "loss": 2.3116, + "step": 1227 + }, + { + "epoch": 0.5630444750114626, + "grad_norm": 0.40927044919095185, + "learning_rate": 0.0004226998386680524, + "loss": 2.483, + "step": 1228 + }, + { + "epoch": 0.5635029802842733, + "grad_norm": 0.3951496333218402, + "learning_rate": 0.000421966160246629, + "loss": 2.5022, + "step": 1229 + }, + { + "epoch": 0.5639614855570839, + "grad_norm": 0.4374863974231965, + "learning_rate": 0.00042123265399695783, + "loss": 2.4126, + "step": 1230 + }, + { + "epoch": 0.5644199908298946, + "grad_norm": 0.4233160518221676, + "learning_rate": 0.0004204993215374273, + "loss": 2.5587, + "step": 1231 + }, + { + "epoch": 0.5648784961027051, + "grad_norm": 0.43094967677637336, + "learning_rate": 0.00041976616448604226, + "loss": 2.4828, + "step": 1232 + }, + { + "epoch": 0.5653370013755158, + "grad_norm": 0.4143553588852657, + "learning_rate": 0.00041903318446042076, + "loss": 2.4199, + "step": 1233 + }, + { + "epoch": 0.5657955066483265, + "grad_norm": 0.4272307229683105, + "learning_rate": 0.00041830038307778984, + "loss": 2.4608, + "step": 1234 + }, + { + "epoch": 0.5662540119211371, + "grad_norm": 0.4361725610182815, + "learning_rate": 0.0004175677619549828, + "loss": 2.5055, + "step": 1235 + }, + { + "epoch": 0.5667125171939478, + "grad_norm": 0.41184101813669627, + "learning_rate": 0.000416835322708435, + "loss": 2.3279, + "step": 1236 + }, + { + "epoch": 0.5671710224667583, + "grad_norm": 0.4229609640296688, + "learning_rate": 0.00041610306695418056, + "loss": 2.3885, + "step": 1237 + }, + { + "epoch": 0.567629527739569, + "grad_norm": 0.4268052663432935, + "learning_rate": 0.0004153709963078488, + "loss": 2.4666, + "step": 1238 + }, + { + "epoch": 0.5680880330123796, + "grad_norm": 0.42784200472549244, + "learning_rate": 0.0004146391123846606, + "loss": 2.4252, + "step": 1239 + }, + { + "epoch": 0.5685465382851903, + "grad_norm": 0.40482355865725245, + "learning_rate": 0.0004139074167994249, + "loss": 2.4389, + "step": 1240 + }, + { + "epoch": 0.569005043558001, + "grad_norm": 0.41060525012295135, + "learning_rate": 0.00041317591116653486, + "loss": 2.4367, + "step": 1241 + }, + { + "epoch": 0.5694635488308115, + "grad_norm": 0.40633017423415013, + "learning_rate": 0.0004124445970999648, + "loss": 2.4471, + "step": 1242 + }, + { + "epoch": 0.5699220541036222, + "grad_norm": 0.4552750660156522, + "learning_rate": 0.00041171347621326627, + "loss": 2.4187, + "step": 1243 + }, + { + "epoch": 0.5703805593764328, + "grad_norm": 0.4266806284168045, + "learning_rate": 0.00041098255011956465, + "loss": 2.4773, + "step": 1244 + }, + { + "epoch": 0.5708390646492435, + "grad_norm": 0.4169271924335219, + "learning_rate": 0.00041025182043155547, + "loss": 2.4733, + "step": 1245 + }, + { + "epoch": 0.571297569922054, + "grad_norm": 0.4254600881425234, + "learning_rate": 0.000409521288761501, + "loss": 2.4899, + "step": 1246 + }, + { + "epoch": 0.5717560751948647, + "grad_norm": 0.4118586716640665, + "learning_rate": 0.00040879095672122646, + "loss": 2.3842, + "step": 1247 + }, + { + "epoch": 0.5722145804676754, + "grad_norm": 0.47430190479870815, + "learning_rate": 0.0004080608259221167, + "loss": 2.4384, + "step": 1248 + }, + { + "epoch": 0.572673085740486, + "grad_norm": 0.41036873689132375, + "learning_rate": 0.0004073308979751126, + "loss": 2.3997, + "step": 1249 + }, + { + "epoch": 0.5731315910132967, + "grad_norm": 0.4169693846799596, + "learning_rate": 0.0004066011744907074, + "loss": 2.301, + "step": 1250 + }, + { + "epoch": 0.5735900962861072, + "grad_norm": 0.42013764424155825, + "learning_rate": 0.00040587165707894326, + "loss": 2.4164, + "step": 1251 + }, + { + "epoch": 0.5740486015589179, + "grad_norm": 0.40582969262004076, + "learning_rate": 0.0004051423473494076, + "loss": 2.378, + "step": 1252 + }, + { + "epoch": 0.5745071068317286, + "grad_norm": 0.41957031617871804, + "learning_rate": 0.0004044132469112299, + "loss": 2.4229, + "step": 1253 + }, + { + "epoch": 0.5749656121045392, + "grad_norm": 0.470967098567965, + "learning_rate": 0.00040368435737307733, + "loss": 2.4861, + "step": 1254 + }, + { + "epoch": 0.5754241173773499, + "grad_norm": 0.4123881321544328, + "learning_rate": 0.00040295568034315224, + "loss": 2.3506, + "step": 1255 + }, + { + "epoch": 0.5758826226501604, + "grad_norm": 0.4051546181148658, + "learning_rate": 0.0004022272174291878, + "loss": 2.4703, + "step": 1256 + }, + { + "epoch": 0.5763411279229711, + "grad_norm": 0.42942446113543065, + "learning_rate": 0.0004014989702384449, + "loss": 2.361, + "step": 1257 + }, + { + "epoch": 0.5767996331957818, + "grad_norm": 0.41087035389332716, + "learning_rate": 0.00040077094037770843, + "loss": 2.4388, + "step": 1258 + }, + { + "epoch": 0.5772581384685924, + "grad_norm": 0.43132719169929257, + "learning_rate": 0.0004000431294532838, + "loss": 2.3898, + "step": 1259 + }, + { + "epoch": 0.5777166437414031, + "grad_norm": 0.4027693040733909, + "learning_rate": 0.0003993155390709935, + "loss": 2.4323, + "step": 1260 + }, + { + "epoch": 0.5781751490142136, + "grad_norm": 0.3937627902394561, + "learning_rate": 0.0003985881708361729, + "loss": 2.4683, + "step": 1261 + }, + { + "epoch": 0.5786336542870243, + "grad_norm": 0.4044547112674934, + "learning_rate": 0.00039786102635366784, + "loss": 2.345, + "step": 1262 + }, + { + "epoch": 0.579092159559835, + "grad_norm": 0.422000563609545, + "learning_rate": 0.0003971341072278302, + "loss": 2.4324, + "step": 1263 + }, + { + "epoch": 0.5795506648326456, + "grad_norm": 0.4207468618711548, + "learning_rate": 0.00039640741506251457, + "loss": 2.482, + "step": 1264 + }, + { + "epoch": 0.5800091701054562, + "grad_norm": 0.43279918288020075, + "learning_rate": 0.00039568095146107495, + "loss": 2.3918, + "step": 1265 + }, + { + "epoch": 0.5804676753782668, + "grad_norm": 0.4240057307784889, + "learning_rate": 0.00039495471802636096, + "loss": 2.5448, + "step": 1266 + }, + { + "epoch": 0.5809261806510775, + "grad_norm": 0.439126793972811, + "learning_rate": 0.0003942287163607145, + "loss": 2.3337, + "step": 1267 + }, + { + "epoch": 0.5813846859238881, + "grad_norm": 0.4123717801445911, + "learning_rate": 0.0003935029480659658, + "loss": 2.4284, + "step": 1268 + }, + { + "epoch": 0.5818431911966988, + "grad_norm": 0.3839239664161491, + "learning_rate": 0.00039277741474343054, + "loss": 2.4935, + "step": 1269 + }, + { + "epoch": 0.5823016964695094, + "grad_norm": 0.41846023778677255, + "learning_rate": 0.0003920521179939057, + "loss": 2.4265, + "step": 1270 + }, + { + "epoch": 0.58276020174232, + "grad_norm": 0.4521402182923671, + "learning_rate": 0.00039132705941766644, + "loss": 2.5142, + "step": 1271 + }, + { + "epoch": 0.5832187070151307, + "grad_norm": 0.42256814034744067, + "learning_rate": 0.0003906022406144624, + "loss": 2.3238, + "step": 1272 + }, + { + "epoch": 0.5836772122879413, + "grad_norm": 0.40781536975943405, + "learning_rate": 0.0003898776631835143, + "loss": 2.4673, + "step": 1273 + }, + { + "epoch": 0.584135717560752, + "grad_norm": 0.42552664591257133, + "learning_rate": 0.00038915332872350994, + "loss": 2.3399, + "step": 1274 + }, + { + "epoch": 0.5845942228335625, + "grad_norm": 0.39562210253675034, + "learning_rate": 0.00038842923883260135, + "loss": 2.3177, + "step": 1275 + }, + { + "epoch": 0.5850527281063732, + "grad_norm": 0.42974305890969994, + "learning_rate": 0.00038770539510840093, + "loss": 2.4739, + "step": 1276 + }, + { + "epoch": 0.5855112333791839, + "grad_norm": 0.42083451931901594, + "learning_rate": 0.00038698179914797783, + "loss": 2.4349, + "step": 1277 + }, + { + "epoch": 0.5859697386519945, + "grad_norm": 0.4366778810160287, + "learning_rate": 0.0003862584525478545, + "loss": 2.4583, + "step": 1278 + }, + { + "epoch": 0.5864282439248051, + "grad_norm": 0.42033135914016534, + "learning_rate": 0.00038553535690400353, + "loss": 2.4812, + "step": 1279 + }, + { + "epoch": 0.5868867491976157, + "grad_norm": 0.40513056281850496, + "learning_rate": 0.00038481251381184355, + "loss": 2.372, + "step": 1280 + }, + { + "epoch": 0.5873452544704264, + "grad_norm": 0.41530634744369443, + "learning_rate": 0.00038408992486623584, + "loss": 2.4367, + "step": 1281 + }, + { + "epoch": 0.5878037597432371, + "grad_norm": 0.42789423934814497, + "learning_rate": 0.00038336759166148117, + "loss": 2.5016, + "step": 1282 + }, + { + "epoch": 0.5882622650160477, + "grad_norm": 0.4344261617481006, + "learning_rate": 0.0003826455157913159, + "loss": 2.2914, + "step": 1283 + }, + { + "epoch": 0.5887207702888583, + "grad_norm": 0.4189128360186991, + "learning_rate": 0.00038192369884890886, + "loss": 2.4532, + "step": 1284 + }, + { + "epoch": 0.5891792755616689, + "grad_norm": 0.43338209473718087, + "learning_rate": 0.00038120214242685723, + "loss": 2.3614, + "step": 1285 + }, + { + "epoch": 0.5896377808344796, + "grad_norm": 0.4441600034679179, + "learning_rate": 0.00038048084811718373, + "loss": 2.3829, + "step": 1286 + }, + { + "epoch": 0.5900962861072903, + "grad_norm": 0.4475008497478999, + "learning_rate": 0.0003797598175113327, + "loss": 2.3886, + "step": 1287 + }, + { + "epoch": 0.5905547913801009, + "grad_norm": 0.4180286066030608, + "learning_rate": 0.0003790390522001662, + "loss": 2.3727, + "step": 1288 + }, + { + "epoch": 0.5910132966529115, + "grad_norm": 0.4308500697822733, + "learning_rate": 0.0003783185537739615, + "loss": 2.482, + "step": 1289 + }, + { + "epoch": 0.5914718019257221, + "grad_norm": 0.4500762668018606, + "learning_rate": 0.00037759832382240697, + "loss": 2.444, + "step": 1290 + }, + { + "epoch": 0.5919303071985328, + "grad_norm": 0.41849405000875445, + "learning_rate": 0.00037687836393459826, + "loss": 2.415, + "step": 1291 + }, + { + "epoch": 0.5923888124713435, + "grad_norm": 0.43772601378908127, + "learning_rate": 0.0003761586756990354, + "loss": 2.4164, + "step": 1292 + }, + { + "epoch": 0.5928473177441541, + "grad_norm": 0.4352145148403022, + "learning_rate": 0.0003754392607036191, + "loss": 2.4701, + "step": 1293 + }, + { + "epoch": 0.5933058230169647, + "grad_norm": 0.4069612980413853, + "learning_rate": 0.0003747201205356472, + "loss": 2.4534, + "step": 1294 + }, + { + "epoch": 0.5937643282897753, + "grad_norm": 0.40808066400320403, + "learning_rate": 0.0003740012567818111, + "loss": 2.5546, + "step": 1295 + }, + { + "epoch": 0.594222833562586, + "grad_norm": 0.4163531679647669, + "learning_rate": 0.0003732826710281922, + "loss": 2.3603, + "step": 1296 + }, + { + "epoch": 0.5946813388353966, + "grad_norm": 0.4094368445488877, + "learning_rate": 0.0003725643648602588, + "loss": 2.3756, + "step": 1297 + }, + { + "epoch": 0.5951398441082072, + "grad_norm": 0.43929373049191484, + "learning_rate": 0.0003718463398628621, + "loss": 2.4589, + "step": 1298 + }, + { + "epoch": 0.5955983493810179, + "grad_norm": 0.4143089256156173, + "learning_rate": 0.0003711285976202331, + "loss": 2.445, + "step": 1299 + }, + { + "epoch": 0.5960568546538285, + "grad_norm": 0.40880830948879776, + "learning_rate": 0.0003704111397159787, + "loss": 2.3763, + "step": 1300 + }, + { + "epoch": 0.5965153599266392, + "grad_norm": 0.38385313563297896, + "learning_rate": 0.0003696939677330788, + "loss": 2.3807, + "step": 1301 + }, + { + "epoch": 0.5969738651994498, + "grad_norm": 0.38779739476062686, + "learning_rate": 0.00036897708325388213, + "loss": 2.3583, + "step": 1302 + }, + { + "epoch": 0.5974323704722604, + "grad_norm": 0.4046071730257965, + "learning_rate": 0.0003682604878601034, + "loss": 2.3673, + "step": 1303 + }, + { + "epoch": 0.597890875745071, + "grad_norm": 0.40754684553358134, + "learning_rate": 0.000367544183132819, + "loss": 2.3478, + "step": 1304 + }, + { + "epoch": 0.5983493810178817, + "grad_norm": 0.42093270421216683, + "learning_rate": 0.0003668281706524645, + "loss": 2.3075, + "step": 1305 + }, + { + "epoch": 0.5988078862906924, + "grad_norm": 0.39956536843118906, + "learning_rate": 0.0003661124519988304, + "loss": 2.4288, + "step": 1306 + }, + { + "epoch": 0.599266391563503, + "grad_norm": 0.4017088481805009, + "learning_rate": 0.00036539702875105893, + "loss": 2.3695, + "step": 1307 + }, + { + "epoch": 0.5997248968363136, + "grad_norm": 0.42570011159814264, + "learning_rate": 0.0003646819024876406, + "loss": 2.4064, + "step": 1308 + }, + { + "epoch": 0.6001834021091242, + "grad_norm": 0.4680998455767829, + "learning_rate": 0.0003639670747864107, + "loss": 2.4755, + "step": 1309 + }, + { + "epoch": 0.6006419073819349, + "grad_norm": 0.4109098293854661, + "learning_rate": 0.00036325254722454584, + "loss": 2.4408, + "step": 1310 + }, + { + "epoch": 0.6011004126547456, + "grad_norm": 0.441094477097312, + "learning_rate": 0.00036253832137855997, + "loss": 2.4379, + "step": 1311 + }, + { + "epoch": 0.6015589179275561, + "grad_norm": 0.4115741101643491, + "learning_rate": 0.00036182439882430183, + "loss": 2.4149, + "step": 1312 + }, + { + "epoch": 0.6020174232003668, + "grad_norm": 0.4052400351168226, + "learning_rate": 0.00036111078113695096, + "loss": 2.3079, + "step": 1313 + }, + { + "epoch": 0.6024759284731774, + "grad_norm": 0.4213763679965886, + "learning_rate": 0.0003603974698910139, + "loss": 2.2864, + "step": 1314 + }, + { + "epoch": 0.6029344337459881, + "grad_norm": 0.416681033435635, + "learning_rate": 0.0003596844666603214, + "loss": 2.3493, + "step": 1315 + }, + { + "epoch": 0.6033929390187988, + "grad_norm": 0.40738622976597777, + "learning_rate": 0.0003589717730180245, + "loss": 2.4247, + "step": 1316 + }, + { + "epoch": 0.6038514442916093, + "grad_norm": 0.4025303164456965, + "learning_rate": 0.00035825939053659117, + "loss": 2.5001, + "step": 1317 + }, + { + "epoch": 0.60430994956442, + "grad_norm": 0.3998650123244478, + "learning_rate": 0.00035754732078780273, + "loss": 2.3155, + "step": 1318 + }, + { + "epoch": 0.6047684548372306, + "grad_norm": 0.4371535960644618, + "learning_rate": 0.00035683556534275076, + "loss": 2.5214, + "step": 1319 + }, + { + "epoch": 0.6052269601100413, + "grad_norm": 0.4274160593375321, + "learning_rate": 0.00035612412577183303, + "loss": 2.3656, + "step": 1320 + }, + { + "epoch": 0.605685465382852, + "grad_norm": 0.41924046312697844, + "learning_rate": 0.00035541300364475063, + "loss": 2.3603, + "step": 1321 + }, + { + "epoch": 0.6061439706556625, + "grad_norm": 0.39177264783313626, + "learning_rate": 0.0003547022005305043, + "loss": 2.4712, + "step": 1322 + }, + { + "epoch": 0.6066024759284732, + "grad_norm": 0.4063797258473571, + "learning_rate": 0.0003539917179973907, + "loss": 2.4508, + "step": 1323 + }, + { + "epoch": 0.6070609812012838, + "grad_norm": 0.3842299056733971, + "learning_rate": 0.00035328155761299917, + "loss": 2.4123, + "step": 1324 + }, + { + "epoch": 0.6075194864740945, + "grad_norm": 0.43002895353803183, + "learning_rate": 0.0003525717209442085, + "loss": 2.3184, + "step": 1325 + }, + { + "epoch": 0.6079779917469051, + "grad_norm": 0.3923679562906387, + "learning_rate": 0.00035186220955718306, + "loss": 2.2652, + "step": 1326 + }, + { + "epoch": 0.6084364970197157, + "grad_norm": 0.4297368132405706, + "learning_rate": 0.0003511530250173696, + "loss": 2.4224, + "step": 1327 + }, + { + "epoch": 0.6088950022925264, + "grad_norm": 0.4135497436690042, + "learning_rate": 0.00035044416888949364, + "loss": 2.3659, + "step": 1328 + }, + { + "epoch": 0.609353507565337, + "grad_norm": 0.43736374278673, + "learning_rate": 0.0003497356427375562, + "loss": 2.4377, + "step": 1329 + }, + { + "epoch": 0.6098120128381477, + "grad_norm": 0.4352367830406737, + "learning_rate": 0.00034902744812483034, + "loss": 2.4434, + "step": 1330 + }, + { + "epoch": 0.6102705181109582, + "grad_norm": 0.43336620299040807, + "learning_rate": 0.00034831958661385714, + "loss": 2.4722, + "step": 1331 + }, + { + "epoch": 0.6107290233837689, + "grad_norm": 0.4046765077819017, + "learning_rate": 0.0003476120597664434, + "loss": 2.3435, + "step": 1332 + }, + { + "epoch": 0.6111875286565795, + "grad_norm": 0.40030253374598584, + "learning_rate": 0.00034690486914365704, + "loss": 2.4337, + "step": 1333 + }, + { + "epoch": 0.6116460339293902, + "grad_norm": 0.44048508699566674, + "learning_rate": 0.00034619801630582435, + "loss": 2.5067, + "step": 1334 + }, + { + "epoch": 0.6121045392022009, + "grad_norm": 0.41370840215297816, + "learning_rate": 0.00034549150281252633, + "loss": 2.3244, + "step": 1335 + }, + { + "epoch": 0.6125630444750114, + "grad_norm": 0.40586253031650543, + "learning_rate": 0.0003447853302225952, + "loss": 2.4128, + "step": 1336 + }, + { + "epoch": 0.6130215497478221, + "grad_norm": 0.41519460918882134, + "learning_rate": 0.00034407950009411126, + "loss": 2.2688, + "step": 1337 + }, + { + "epoch": 0.6134800550206327, + "grad_norm": 0.4163124960001266, + "learning_rate": 0.00034337401398439873, + "loss": 2.3534, + "step": 1338 + }, + { + "epoch": 0.6139385602934434, + "grad_norm": 0.4128604909498249, + "learning_rate": 0.00034266887345002305, + "loss": 2.3237, + "step": 1339 + }, + { + "epoch": 0.6143970655662541, + "grad_norm": 0.41138614522585826, + "learning_rate": 0.0003419640800467874, + "loss": 2.455, + "step": 1340 + }, + { + "epoch": 0.6148555708390646, + "grad_norm": 0.4280667714496267, + "learning_rate": 0.0003412596353297288, + "loss": 2.4458, + "step": 1341 + }, + { + "epoch": 0.6153140761118753, + "grad_norm": 0.44510745071978153, + "learning_rate": 0.00034055554085311493, + "loss": 2.3954, + "step": 1342 + }, + { + "epoch": 0.6157725813846859, + "grad_norm": 0.42728772750724625, + "learning_rate": 0.00033985179817044105, + "loss": 2.4521, + "step": 1343 + }, + { + "epoch": 0.6162310866574966, + "grad_norm": 0.42217906644575326, + "learning_rate": 0.0003391484088344257, + "loss": 2.3943, + "step": 1344 + }, + { + "epoch": 0.6166895919303071, + "grad_norm": 0.41971935583995323, + "learning_rate": 0.00033844537439700807, + "loss": 2.5016, + "step": 1345 + }, + { + "epoch": 0.6171480972031178, + "grad_norm": 0.40275339457225634, + "learning_rate": 0.00033774269640934445, + "loss": 2.4364, + "step": 1346 + }, + { + "epoch": 0.6176066024759285, + "grad_norm": 0.41556554037695864, + "learning_rate": 0.0003370403764218045, + "loss": 2.3681, + "step": 1347 + }, + { + "epoch": 0.6180651077487391, + "grad_norm": 0.4037901385194166, + "learning_rate": 0.000336338415983968, + "loss": 2.3383, + "step": 1348 + }, + { + "epoch": 0.6185236130215498, + "grad_norm": 0.3961373990154217, + "learning_rate": 0.00033563681664462155, + "loss": 2.3496, + "step": 1349 + }, + { + "epoch": 0.6189821182943603, + "grad_norm": 0.4476609976044065, + "learning_rate": 0.000334935579951755, + "loss": 2.5887, + "step": 1350 + }, + { + "epoch": 0.619440623567171, + "grad_norm": 0.41895461864072336, + "learning_rate": 0.0003342347074525578, + "loss": 2.3675, + "step": 1351 + }, + { + "epoch": 0.6198991288399817, + "grad_norm": 0.4028940703563436, + "learning_rate": 0.0003335342006934161, + "loss": 2.4493, + "step": 1352 + }, + { + "epoch": 0.6203576341127923, + "grad_norm": 0.4135124298032527, + "learning_rate": 0.00033283406121990914, + "loss": 2.4674, + "step": 1353 + }, + { + "epoch": 0.620816139385603, + "grad_norm": 0.43082617203135243, + "learning_rate": 0.0003321342905768057, + "loss": 2.3542, + "step": 1354 + }, + { + "epoch": 0.6212746446584135, + "grad_norm": 0.4373642304191737, + "learning_rate": 0.00033143489030806086, + "loss": 2.411, + "step": 1355 + }, + { + "epoch": 0.6217331499312242, + "grad_norm": 0.40117444012730924, + "learning_rate": 0.00033073586195681227, + "loss": 2.3707, + "step": 1356 + }, + { + "epoch": 0.6221916552040349, + "grad_norm": 0.4125074699973566, + "learning_rate": 0.00033003720706537736, + "loss": 2.29, + "step": 1357 + }, + { + "epoch": 0.6226501604768455, + "grad_norm": 0.44515907487315226, + "learning_rate": 0.0003293389271752492, + "loss": 2.3953, + "step": 1358 + }, + { + "epoch": 0.6231086657496562, + "grad_norm": 0.4318850070085816, + "learning_rate": 0.00032864102382709374, + "loss": 2.5076, + "step": 1359 + }, + { + "epoch": 0.6235671710224667, + "grad_norm": 0.4169697624291777, + "learning_rate": 0.000327943498560746, + "loss": 2.401, + "step": 1360 + }, + { + "epoch": 0.6240256762952774, + "grad_norm": 0.42055564103737963, + "learning_rate": 0.00032724635291520694, + "loss": 2.4204, + "step": 1361 + }, + { + "epoch": 0.624484181568088, + "grad_norm": 0.4479996857885999, + "learning_rate": 0.00032654958842863967, + "loss": 2.4087, + "step": 1362 + }, + { + "epoch": 0.6249426868408987, + "grad_norm": 0.4041491314261291, + "learning_rate": 0.0003258532066383667, + "loss": 2.3059, + "step": 1363 + }, + { + "epoch": 0.6254011921137093, + "grad_norm": 0.4033842738918856, + "learning_rate": 0.000325157209080866, + "loss": 2.3292, + "step": 1364 + }, + { + "epoch": 0.6258596973865199, + "grad_norm": 0.4198234268047018, + "learning_rate": 0.00032446159729176743, + "loss": 2.462, + "step": 1365 + }, + { + "epoch": 0.6263182026593306, + "grad_norm": 0.43331924950001516, + "learning_rate": 0.0003237663728058502, + "loss": 2.3238, + "step": 1366 + }, + { + "epoch": 0.6267767079321412, + "grad_norm": 0.41318410159084096, + "learning_rate": 0.0003230715371570389, + "loss": 2.5063, + "step": 1367 + }, + { + "epoch": 0.6272352132049519, + "grad_norm": 0.406122590136821, + "learning_rate": 0.00032237709187839996, + "loss": 2.3427, + "step": 1368 + }, + { + "epoch": 0.6276937184777625, + "grad_norm": 0.3921856503340904, + "learning_rate": 0.0003216830385021388, + "loss": 2.3129, + "step": 1369 + }, + { + "epoch": 0.6281522237505731, + "grad_norm": 0.42181826636078285, + "learning_rate": 0.0003209893785595959, + "loss": 2.5165, + "step": 1370 + }, + { + "epoch": 0.6286107290233838, + "grad_norm": 0.41906112248972505, + "learning_rate": 0.00032029611358124366, + "loss": 2.3509, + "step": 1371 + }, + { + "epoch": 0.6290692342961944, + "grad_norm": 0.45572342165096014, + "learning_rate": 0.00031960324509668336, + "loss": 2.3616, + "step": 1372 + }, + { + "epoch": 0.6295277395690051, + "grad_norm": 0.46268860064123385, + "learning_rate": 0.0003189107746346412, + "loss": 2.43, + "step": 1373 + }, + { + "epoch": 0.6299862448418156, + "grad_norm": 0.4175261996692696, + "learning_rate": 0.0003182187037229653, + "loss": 2.4108, + "step": 1374 + }, + { + "epoch": 0.6304447501146263, + "grad_norm": 0.43966983432127416, + "learning_rate": 0.0003175270338886221, + "loss": 2.5085, + "step": 1375 + }, + { + "epoch": 0.630903255387437, + "grad_norm": 0.45254006364598043, + "learning_rate": 0.00031683576665769345, + "loss": 2.4326, + "step": 1376 + }, + { + "epoch": 0.6313617606602476, + "grad_norm": 0.40285240945423695, + "learning_rate": 0.0003161449035553724, + "loss": 2.3268, + "step": 1377 + }, + { + "epoch": 0.6318202659330582, + "grad_norm": 0.4052573628857077, + "learning_rate": 0.00031545444610596077, + "loss": 2.444, + "step": 1378 + }, + { + "epoch": 0.6322787712058688, + "grad_norm": 0.4276387198062274, + "learning_rate": 0.000314764395832865, + "loss": 2.4216, + "step": 1379 + }, + { + "epoch": 0.6327372764786795, + "grad_norm": 0.4226848442123443, + "learning_rate": 0.0003140747542585934, + "loss": 2.3566, + "step": 1380 + }, + { + "epoch": 0.6331957817514902, + "grad_norm": 0.421927161877351, + "learning_rate": 0.00031338552290475266, + "loss": 2.4068, + "step": 1381 + }, + { + "epoch": 0.6336542870243008, + "grad_norm": 0.40284369697440797, + "learning_rate": 0.00031269670329204396, + "loss": 2.3372, + "step": 1382 + }, + { + "epoch": 0.6341127922971114, + "grad_norm": 0.42578832115745496, + "learning_rate": 0.0003120082969402604, + "loss": 2.4658, + "step": 1383 + }, + { + "epoch": 0.634571297569922, + "grad_norm": 0.423215365038528, + "learning_rate": 0.00031132030536828314, + "loss": 2.5206, + "step": 1384 + }, + { + "epoch": 0.6350298028427327, + "grad_norm": 0.43115602572176853, + "learning_rate": 0.00031063273009407805, + "loss": 2.4041, + "step": 1385 + }, + { + "epoch": 0.6354883081155434, + "grad_norm": 0.42858649318106024, + "learning_rate": 0.00030994557263469265, + "loss": 2.3833, + "step": 1386 + }, + { + "epoch": 0.635946813388354, + "grad_norm": 0.3972047602911259, + "learning_rate": 0.0003092588345062526, + "loss": 2.3627, + "step": 1387 + }, + { + "epoch": 0.6364053186611646, + "grad_norm": 0.43156875186195237, + "learning_rate": 0.0003085725172239582, + "loss": 2.4857, + "step": 1388 + }, + { + "epoch": 0.6368638239339752, + "grad_norm": 0.40738831096205463, + "learning_rate": 0.0003078866223020815, + "loss": 2.388, + "step": 1389 + }, + { + "epoch": 0.6373223292067859, + "grad_norm": 0.43041637546854733, + "learning_rate": 0.0003072011512539624, + "loss": 2.3935, + "step": 1390 + }, + { + "epoch": 0.6377808344795965, + "grad_norm": 0.4368832124873948, + "learning_rate": 0.00030651610559200574, + "loss": 2.4741, + "step": 1391 + }, + { + "epoch": 0.6382393397524071, + "grad_norm": 0.42179790276365803, + "learning_rate": 0.00030583148682767757, + "loss": 2.3813, + "step": 1392 + }, + { + "epoch": 0.6386978450252178, + "grad_norm": 0.41771160720842715, + "learning_rate": 0.00030514729647150243, + "loss": 2.3878, + "step": 1393 + }, + { + "epoch": 0.6391563502980284, + "grad_norm": 0.4120419693461303, + "learning_rate": 0.0003044635360330592, + "loss": 2.3753, + "step": 1394 + }, + { + "epoch": 0.6396148555708391, + "grad_norm": 0.40547679975131934, + "learning_rate": 0.00030378020702097845, + "loss": 2.4016, + "step": 1395 + }, + { + "epoch": 0.6400733608436497, + "grad_norm": 0.4294901189942919, + "learning_rate": 0.000303097310942939, + "loss": 2.5138, + "step": 1396 + }, + { + "epoch": 0.6405318661164603, + "grad_norm": 0.4129223439018124, + "learning_rate": 0.0003024148493056641, + "loss": 2.397, + "step": 1397 + }, + { + "epoch": 0.640990371389271, + "grad_norm": 0.4368452583585999, + "learning_rate": 0.00030173282361491865, + "loss": 2.374, + "step": 1398 + }, + { + "epoch": 0.6414488766620816, + "grad_norm": 0.4266300845009801, + "learning_rate": 0.0003010512353755057, + "loss": 2.3565, + "step": 1399 + }, + { + "epoch": 0.6419073819348923, + "grad_norm": 0.42856721706570455, + "learning_rate": 0.00030037008609126313, + "loss": 2.4002, + "step": 1400 + }, + { + "epoch": 0.6423658872077029, + "grad_norm": 0.4292185736166491, + "learning_rate": 0.0002996893772650602, + "loss": 2.3814, + "step": 1401 + }, + { + "epoch": 0.6428243924805135, + "grad_norm": 0.40657693861649774, + "learning_rate": 0.0002990091103987945, + "loss": 2.4948, + "step": 1402 + }, + { + "epoch": 0.6432828977533241, + "grad_norm": 0.4432960908316163, + "learning_rate": 0.0002983292869933886, + "loss": 2.4916, + "step": 1403 + }, + { + "epoch": 0.6437414030261348, + "grad_norm": 0.45264479841862426, + "learning_rate": 0.0002976499085487862, + "loss": 2.4216, + "step": 1404 + }, + { + "epoch": 0.6441999082989455, + "grad_norm": 0.41427683807367544, + "learning_rate": 0.00029697097656394963, + "loss": 2.4022, + "step": 1405 + }, + { + "epoch": 0.6446584135717561, + "grad_norm": 0.4120620563649848, + "learning_rate": 0.00029629249253685595, + "loss": 2.3532, + "step": 1406 + }, + { + "epoch": 0.6451169188445667, + "grad_norm": 0.4185930967563898, + "learning_rate": 0.00029561445796449416, + "loss": 2.4935, + "step": 1407 + }, + { + "epoch": 0.6455754241173773, + "grad_norm": 0.41958005668869147, + "learning_rate": 0.0002949368743428612, + "loss": 2.3473, + "step": 1408 + }, + { + "epoch": 0.646033929390188, + "grad_norm": 0.41934325668661787, + "learning_rate": 0.0002942597431669593, + "loss": 2.3799, + "step": 1409 + }, + { + "epoch": 0.6464924346629987, + "grad_norm": 0.4280788678810678, + "learning_rate": 0.0002935830659307924, + "loss": 2.4126, + "step": 1410 + }, + { + "epoch": 0.6469509399358092, + "grad_norm": 0.44995544243212565, + "learning_rate": 0.0002929068441273629, + "loss": 2.4091, + "step": 1411 + }, + { + "epoch": 0.6474094452086199, + "grad_norm": 0.42386710244799203, + "learning_rate": 0.0002922310792486681, + "loss": 2.3881, + "step": 1412 + }, + { + "epoch": 0.6478679504814305, + "grad_norm": 0.4105303797892493, + "learning_rate": 0.00029155577278569745, + "loss": 2.3509, + "step": 1413 + }, + { + "epoch": 0.6483264557542412, + "grad_norm": 0.415419580828065, + "learning_rate": 0.00029088092622842895, + "loss": 2.4507, + "step": 1414 + }, + { + "epoch": 0.6487849610270519, + "grad_norm": 0.4018142878694394, + "learning_rate": 0.00029020654106582544, + "loss": 2.4413, + "step": 1415 + }, + { + "epoch": 0.6492434662998624, + "grad_norm": 0.4294225207602916, + "learning_rate": 0.0002895326187858326, + "loss": 2.3764, + "step": 1416 + }, + { + "epoch": 0.6497019715726731, + "grad_norm": 0.4500599684075753, + "learning_rate": 0.00028885916087537377, + "loss": 2.4317, + "step": 1417 + }, + { + "epoch": 0.6501604768454837, + "grad_norm": 0.4173076151890919, + "learning_rate": 0.00028818616882034877, + "loss": 2.4445, + "step": 1418 + }, + { + "epoch": 0.6506189821182944, + "grad_norm": 0.40493533468453186, + "learning_rate": 0.0002875136441056286, + "loss": 2.3112, + "step": 1419 + }, + { + "epoch": 0.651077487391105, + "grad_norm": 0.4404695893501588, + "learning_rate": 0.000286841588215054, + "loss": 2.4722, + "step": 1420 + }, + { + "epoch": 0.6515359926639156, + "grad_norm": 0.4649322970386748, + "learning_rate": 0.0002861700026314308, + "loss": 2.4365, + "step": 1421 + }, + { + "epoch": 0.6519944979367263, + "grad_norm": 0.4309473400876179, + "learning_rate": 0.00028549888883652686, + "loss": 2.3646, + "step": 1422 + }, + { + "epoch": 0.6524530032095369, + "grad_norm": 0.42788883024085334, + "learning_rate": 0.00028482824831107, + "loss": 2.4505, + "step": 1423 + }, + { + "epoch": 0.6529115084823476, + "grad_norm": 0.44500888964930047, + "learning_rate": 0.000284158082534743, + "loss": 2.4174, + "step": 1424 + }, + { + "epoch": 0.6533700137551581, + "grad_norm": 0.4201803669038481, + "learning_rate": 0.00028348839298618177, + "loss": 2.3521, + "step": 1425 + }, + { + "epoch": 0.6538285190279688, + "grad_norm": 0.4204140510943559, + "learning_rate": 0.0002828191811429709, + "loss": 2.4016, + "step": 1426 + }, + { + "epoch": 0.6542870243007795, + "grad_norm": 0.4198481087432641, + "learning_rate": 0.00028215044848164164, + "loss": 2.4299, + "step": 1427 + }, + { + "epoch": 0.6547455295735901, + "grad_norm": 0.43729434797114347, + "learning_rate": 0.00028148219647766747, + "loss": 2.5593, + "step": 1428 + }, + { + "epoch": 0.6552040348464008, + "grad_norm": 0.43970904368925073, + "learning_rate": 0.00028081442660546124, + "loss": 2.3363, + "step": 1429 + }, + { + "epoch": 0.6556625401192113, + "grad_norm": 0.4430965909635184, + "learning_rate": 0.0002801471403383728, + "loss": 2.3368, + "step": 1430 + }, + { + "epoch": 0.656121045392022, + "grad_norm": 0.40358004637961103, + "learning_rate": 0.00027948033914868415, + "loss": 2.3816, + "step": 1431 + }, + { + "epoch": 0.6565795506648326, + "grad_norm": 0.4214442301995808, + "learning_rate": 0.00027881402450760775, + "loss": 2.4142, + "step": 1432 + }, + { + "epoch": 0.6570380559376433, + "grad_norm": 0.42644596094628323, + "learning_rate": 0.00027814819788528165, + "loss": 2.4663, + "step": 1433 + }, + { + "epoch": 0.657496561210454, + "grad_norm": 0.42411973022739785, + "learning_rate": 0.00027748286075076836, + "loss": 2.3543, + "step": 1434 + }, + { + "epoch": 0.6579550664832645, + "grad_norm": 0.42428006767226006, + "learning_rate": 0.00027681801457204937, + "loss": 2.3492, + "step": 1435 + }, + { + "epoch": 0.6584135717560752, + "grad_norm": 0.4240042351062888, + "learning_rate": 0.00027615366081602306, + "loss": 2.4951, + "step": 1436 + }, + { + "epoch": 0.6588720770288858, + "grad_norm": 0.41067990100820645, + "learning_rate": 0.0002754898009485021, + "loss": 2.3146, + "step": 1437 + }, + { + "epoch": 0.6593305823016965, + "grad_norm": 0.4054889915766571, + "learning_rate": 0.0002748264364342085, + "loss": 2.3535, + "step": 1438 + }, + { + "epoch": 0.6597890875745072, + "grad_norm": 0.4264677475550216, + "learning_rate": 0.00027416356873677204, + "loss": 2.3261, + "step": 1439 + }, + { + "epoch": 0.6602475928473177, + "grad_norm": 0.4037464421517066, + "learning_rate": 0.0002735011993187258, + "loss": 2.3399, + "step": 1440 + }, + { + "epoch": 0.6607060981201284, + "grad_norm": 0.42355354155712893, + "learning_rate": 0.0002728393296415042, + "loss": 2.4785, + "step": 1441 + }, + { + "epoch": 0.661164603392939, + "grad_norm": 0.41994949335938603, + "learning_rate": 0.00027217796116543817, + "loss": 2.4254, + "step": 1442 + }, + { + "epoch": 0.6616231086657497, + "grad_norm": 0.42382216559232827, + "learning_rate": 0.0002715170953497532, + "loss": 2.3212, + "step": 1443 + }, + { + "epoch": 0.6620816139385602, + "grad_norm": 0.40894674243120416, + "learning_rate": 0.00027085673365256614, + "loss": 2.3424, + "step": 1444 + }, + { + "epoch": 0.6625401192113709, + "grad_norm": 0.40907266766151085, + "learning_rate": 0.00027019687753088075, + "loss": 2.4555, + "step": 1445 + }, + { + "epoch": 0.6629986244841816, + "grad_norm": 0.4346419266028131, + "learning_rate": 0.00026953752844058597, + "loss": 2.4048, + "step": 1446 + }, + { + "epoch": 0.6634571297569922, + "grad_norm": 0.4250505849691299, + "learning_rate": 0.0002688786878364516, + "loss": 2.4444, + "step": 1447 + }, + { + "epoch": 0.6639156350298029, + "grad_norm": 0.43615209369324276, + "learning_rate": 0.00026822035717212597, + "loss": 2.4731, + "step": 1448 + }, + { + "epoch": 0.6643741403026134, + "grad_norm": 0.4117106278652555, + "learning_rate": 0.00026756253790013193, + "loss": 2.292, + "step": 1449 + }, + { + "epoch": 0.6648326455754241, + "grad_norm": 0.4222249716938658, + "learning_rate": 0.0002669052314718641, + "loss": 2.3421, + "step": 1450 + }, + { + "epoch": 0.6652911508482348, + "grad_norm": 0.4018844591717581, + "learning_rate": 0.0002662484393375855, + "loss": 2.3108, + "step": 1451 + }, + { + "epoch": 0.6657496561210454, + "grad_norm": 0.4493480993716058, + "learning_rate": 0.00026559216294642446, + "loss": 2.3535, + "step": 1452 + }, + { + "epoch": 0.6662081613938561, + "grad_norm": 0.4235565510586741, + "learning_rate": 0.0002649364037463718, + "loss": 2.4206, + "step": 1453 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.42697579580183764, + "learning_rate": 0.0002642811631842764, + "loss": 2.3949, + "step": 1454 + }, + { + "epoch": 0.6671251719394773, + "grad_norm": 0.4150150841910891, + "learning_rate": 0.0002636264427058439, + "loss": 2.3614, + "step": 1455 + }, + { + "epoch": 0.667583677212288, + "grad_norm": 0.4223028436042626, + "learning_rate": 0.00026297224375563123, + "loss": 2.4015, + "step": 1456 + }, + { + "epoch": 0.6680421824850986, + "grad_norm": 0.44087186825291147, + "learning_rate": 0.00026231856777704575, + "loss": 2.433, + "step": 1457 + }, + { + "epoch": 0.6685006877579092, + "grad_norm": 0.4152712758575585, + "learning_rate": 0.00026166541621234026, + "loss": 2.3177, + "step": 1458 + }, + { + "epoch": 0.6689591930307198, + "grad_norm": 0.410487427721036, + "learning_rate": 0.00026101279050261045, + "loss": 2.252, + "step": 1459 + }, + { + "epoch": 0.6694176983035305, + "grad_norm": 0.4304912584295408, + "learning_rate": 0.00026036069208779247, + "loss": 2.4212, + "step": 1460 + }, + { + "epoch": 0.6698762035763411, + "grad_norm": 0.42778633756809653, + "learning_rate": 0.0002597091224066581, + "loss": 2.3778, + "step": 1461 + }, + { + "epoch": 0.6703347088491518, + "grad_norm": 0.4230115574875622, + "learning_rate": 0.00025905808289681365, + "loss": 2.3547, + "step": 1462 + }, + { + "epoch": 0.6707932141219624, + "grad_norm": 0.41741621780662813, + "learning_rate": 0.0002584075749946946, + "loss": 2.4089, + "step": 1463 + }, + { + "epoch": 0.671251719394773, + "grad_norm": 0.4247592420768908, + "learning_rate": 0.00025775760013556424, + "loss": 2.4006, + "step": 1464 + }, + { + "epoch": 0.6717102246675837, + "grad_norm": 0.4060974842064876, + "learning_rate": 0.0002571081597535095, + "loss": 2.4586, + "step": 1465 + }, + { + "epoch": 0.6721687299403943, + "grad_norm": 0.4285738520543975, + "learning_rate": 0.00025645925528143776, + "loss": 2.4686, + "step": 1466 + }, + { + "epoch": 0.672627235213205, + "grad_norm": 0.47270773107775044, + "learning_rate": 0.0002558108881510747, + "loss": 2.5228, + "step": 1467 + }, + { + "epoch": 0.6730857404860155, + "grad_norm": 0.4380476673717588, + "learning_rate": 0.00025516305979295963, + "loss": 2.3176, + "step": 1468 + }, + { + "epoch": 0.6735442457588262, + "grad_norm": 0.40790489030475097, + "learning_rate": 0.0002545157716364439, + "loss": 2.4306, + "step": 1469 + }, + { + "epoch": 0.6740027510316369, + "grad_norm": 0.41657334162028997, + "learning_rate": 0.00025386902510968624, + "loss": 2.309, + "step": 1470 + }, + { + "epoch": 0.6744612563044475, + "grad_norm": 0.40659395730750375, + "learning_rate": 0.00025322282163965095, + "loss": 2.4564, + "step": 1471 + }, + { + "epoch": 0.6749197615772582, + "grad_norm": 0.4192496329029624, + "learning_rate": 0.00025257716265210384, + "loss": 2.3695, + "step": 1472 + }, + { + "epoch": 0.6753782668500687, + "grad_norm": 0.442979005096953, + "learning_rate": 0.0002519320495716091, + "loss": 2.3425, + "step": 1473 + }, + { + "epoch": 0.6758367721228794, + "grad_norm": 0.439663686110817, + "learning_rate": 0.00025128748382152716, + "loss": 2.4914, + "step": 1474 + }, + { + "epoch": 0.6762952773956901, + "grad_norm": 0.3940730435003999, + "learning_rate": 0.00025064346682401016, + "loss": 2.3696, + "step": 1475 + }, + { + "epoch": 0.6767537826685007, + "grad_norm": 0.4279017325145052, + "learning_rate": 0.0002500000000000001, + "loss": 2.3597, + "step": 1476 + }, + { + "epoch": 0.6772122879413113, + "grad_norm": 0.4426420851475869, + "learning_rate": 0.0002493570847692246, + "loss": 2.3681, + "step": 1477 + }, + { + "epoch": 0.6776707932141219, + "grad_norm": 0.4268780523647632, + "learning_rate": 0.00024871472255019424, + "loss": 2.3314, + "step": 1478 + }, + { + "epoch": 0.6781292984869326, + "grad_norm": 0.40460634998955547, + "learning_rate": 0.00024807291476019994, + "loss": 2.3864, + "step": 1479 + }, + { + "epoch": 0.6785878037597433, + "grad_norm": 0.42376346523400116, + "learning_rate": 0.00024743166281530877, + "loss": 2.3564, + "step": 1480 + }, + { + "epoch": 0.6790463090325539, + "grad_norm": 0.4248405705840611, + "learning_rate": 0.000246790968130362, + "loss": 2.4253, + "step": 1481 + }, + { + "epoch": 0.6795048143053645, + "grad_norm": 0.4427906339710954, + "learning_rate": 0.0002461508321189706, + "loss": 2.3625, + "step": 1482 + }, + { + "epoch": 0.6799633195781751, + "grad_norm": 0.4076233482774913, + "learning_rate": 0.00024551125619351385, + "loss": 2.3726, + "step": 1483 + }, + { + "epoch": 0.6804218248509858, + "grad_norm": 0.4154075357104816, + "learning_rate": 0.00024487224176513453, + "loss": 2.3664, + "step": 1484 + }, + { + "epoch": 0.6808803301237965, + "grad_norm": 0.45158280612405066, + "learning_rate": 0.0002442337902437365, + "loss": 2.4761, + "step": 1485 + }, + { + "epoch": 0.6813388353966071, + "grad_norm": 0.4296168804386725, + "learning_rate": 0.0002435959030379824, + "loss": 2.3398, + "step": 1486 + }, + { + "epoch": 0.6817973406694177, + "grad_norm": 0.41404128104623106, + "learning_rate": 0.00024295858155528888, + "loss": 2.3417, + "step": 1487 + }, + { + "epoch": 0.6822558459422283, + "grad_norm": 0.4193603717608296, + "learning_rate": 0.00024232182720182523, + "loss": 2.4278, + "step": 1488 + }, + { + "epoch": 0.682714351215039, + "grad_norm": 0.43828167505740906, + "learning_rate": 0.00024168564138250855, + "loss": 2.4187, + "step": 1489 + }, + { + "epoch": 0.6831728564878496, + "grad_norm": 0.4263370150253839, + "learning_rate": 0.00024105002550100246, + "loss": 2.4627, + "step": 1490 + }, + { + "epoch": 0.6836313617606602, + "grad_norm": 0.4074608405618422, + "learning_rate": 0.00024041498095971254, + "loss": 2.4032, + "step": 1491 + }, + { + "epoch": 0.6840898670334709, + "grad_norm": 0.4405805053062365, + "learning_rate": 0.0002397805091597835, + "loss": 2.4574, + "step": 1492 + }, + { + "epoch": 0.6845483723062815, + "grad_norm": 0.44936180894784034, + "learning_rate": 0.0002391466115010973, + "loss": 2.4154, + "step": 1493 + }, + { + "epoch": 0.6850068775790922, + "grad_norm": 0.4156334414613057, + "learning_rate": 0.00023851328938226808, + "loss": 2.4366, + "step": 1494 + }, + { + "epoch": 0.6854653828519028, + "grad_norm": 0.44250944963926764, + "learning_rate": 0.00023788054420064109, + "loss": 2.3564, + "step": 1495 + }, + { + "epoch": 0.6859238881247134, + "grad_norm": 0.42430709108722386, + "learning_rate": 0.00023724837735228773, + "loss": 2.3592, + "step": 1496 + }, + { + "epoch": 0.686382393397524, + "grad_norm": 0.4240574352359286, + "learning_rate": 0.00023661679023200422, + "loss": 2.4328, + "step": 1497 + }, + { + "epoch": 0.6868408986703347, + "grad_norm": 0.42514126984090717, + "learning_rate": 0.00023598578423330714, + "loss": 2.3839, + "step": 1498 + }, + { + "epoch": 0.6872994039431454, + "grad_norm": 0.42546728612382295, + "learning_rate": 0.00023535536074843083, + "loss": 2.3629, + "step": 1499 + }, + { + "epoch": 0.687757909215956, + "grad_norm": 0.42839512116820866, + "learning_rate": 0.00023472552116832502, + "loss": 2.4446, + "step": 1500 + }, + { + "epoch": 0.6882164144887666, + "grad_norm": 0.4086546234728548, + "learning_rate": 0.0002340962668826503, + "loss": 2.4197, + "step": 1501 + }, + { + "epoch": 0.6886749197615772, + "grad_norm": 0.4289744966548364, + "learning_rate": 0.00023346759927977663, + "loss": 2.4466, + "step": 1502 + }, + { + "epoch": 0.6891334250343879, + "grad_norm": 0.4221693976210663, + "learning_rate": 0.0002328395197467789, + "loss": 2.421, + "step": 1503 + }, + { + "epoch": 0.6895919303071986, + "grad_norm": 0.42166479034378507, + "learning_rate": 0.00023221202966943515, + "loss": 2.4391, + "step": 1504 + }, + { + "epoch": 0.6900504355800092, + "grad_norm": 0.4359041134015822, + "learning_rate": 0.0002315851304322223, + "loss": 2.4205, + "step": 1505 + }, + { + "epoch": 0.6905089408528198, + "grad_norm": 0.4198509572270831, + "learning_rate": 0.0002309588234183137, + "loss": 2.4657, + "step": 1506 + }, + { + "epoch": 0.6909674461256304, + "grad_norm": 0.4291227278505325, + "learning_rate": 0.00023033311000957653, + "loss": 2.3499, + "step": 1507 + }, + { + "epoch": 0.6914259513984411, + "grad_norm": 0.4353684826596712, + "learning_rate": 0.00022970799158656758, + "loss": 2.4395, + "step": 1508 + }, + { + "epoch": 0.6918844566712518, + "grad_norm": 0.43442641190262904, + "learning_rate": 0.0002290834695285316, + "loss": 2.376, + "step": 1509 + }, + { + "epoch": 0.6923429619440623, + "grad_norm": 0.4413123743064568, + "learning_rate": 0.00022845954521339678, + "loss": 2.4985, + "step": 1510 + }, + { + "epoch": 0.692801467216873, + "grad_norm": 0.4295642859001141, + "learning_rate": 0.0002278362200177732, + "loss": 2.4056, + "step": 1511 + }, + { + "epoch": 0.6932599724896836, + "grad_norm": 0.42180841649540335, + "learning_rate": 0.00022721349531694852, + "loss": 2.3462, + "step": 1512 + }, + { + "epoch": 0.6937184777624943, + "grad_norm": 0.42960522226980663, + "learning_rate": 0.0002265913724848855, + "loss": 2.3906, + "step": 1513 + }, + { + "epoch": 0.694176983035305, + "grad_norm": 0.4299927056977889, + "learning_rate": 0.00022596985289421946, + "loss": 2.3931, + "step": 1514 + }, + { + "epoch": 0.6946354883081155, + "grad_norm": 0.4169399864769771, + "learning_rate": 0.00022534893791625405, + "loss": 2.3557, + "step": 1515 + }, + { + "epoch": 0.6950939935809262, + "grad_norm": 0.4210078416435987, + "learning_rate": 0.00022472862892095968, + "loss": 2.383, + "step": 1516 + }, + { + "epoch": 0.6955524988537368, + "grad_norm": 0.42676567573963087, + "learning_rate": 0.00022410892727696896, + "loss": 2.5315, + "step": 1517 + }, + { + "epoch": 0.6960110041265475, + "grad_norm": 0.4067907372533715, + "learning_rate": 0.0002234898343515751, + "loss": 2.4185, + "step": 1518 + }, + { + "epoch": 0.6964695093993581, + "grad_norm": 0.41638178533208764, + "learning_rate": 0.00022287135151072792, + "loss": 2.3808, + "step": 1519 + }, + { + "epoch": 0.6969280146721687, + "grad_norm": 0.408269753134797, + "learning_rate": 0.00022225348011903096, + "loss": 2.3971, + "step": 1520 + }, + { + "epoch": 0.6973865199449794, + "grad_norm": 0.4326143962983188, + "learning_rate": 0.0002216362215397393, + "loss": 2.3353, + "step": 1521 + }, + { + "epoch": 0.69784502521779, + "grad_norm": 0.42458749924671396, + "learning_rate": 0.00022101957713475522, + "loss": 2.3771, + "step": 1522 + }, + { + "epoch": 0.6983035304906007, + "grad_norm": 0.39545433527663754, + "learning_rate": 0.00022040354826462666, + "loss": 2.3746, + "step": 1523 + }, + { + "epoch": 0.6987620357634112, + "grad_norm": 0.41957803790847104, + "learning_rate": 0.0002197881362885426, + "loss": 2.3894, + "step": 1524 + }, + { + "epoch": 0.6992205410362219, + "grad_norm": 0.4278650489736586, + "learning_rate": 0.0002191733425643318, + "loss": 2.2759, + "step": 1525 + }, + { + "epoch": 0.6996790463090325, + "grad_norm": 0.42804790313526386, + "learning_rate": 0.00021855916844845826, + "loss": 2.343, + "step": 1526 + }, + { + "epoch": 0.7001375515818432, + "grad_norm": 0.40918453650116443, + "learning_rate": 0.00021794561529601898, + "loss": 2.35, + "step": 1527 + }, + { + "epoch": 0.7005960568546539, + "grad_norm": 0.4431040252760064, + "learning_rate": 0.00021733268446074138, + "loss": 2.3347, + "step": 1528 + }, + { + "epoch": 0.7010545621274644, + "grad_norm": 0.42803062914867496, + "learning_rate": 0.00021672037729497917, + "loss": 2.3285, + "step": 1529 + }, + { + "epoch": 0.7015130674002751, + "grad_norm": 0.43836439238894137, + "learning_rate": 0.0002161086951497106, + "loss": 2.5039, + "step": 1530 + }, + { + "epoch": 0.7019715726730857, + "grad_norm": 0.42920465948716285, + "learning_rate": 0.00021549763937453442, + "loss": 2.5081, + "step": 1531 + }, + { + "epoch": 0.7024300779458964, + "grad_norm": 0.42237785554006785, + "learning_rate": 0.00021488721131766736, + "loss": 2.4413, + "step": 1532 + }, + { + "epoch": 0.7028885832187071, + "grad_norm": 0.43027014717574713, + "learning_rate": 0.00021427741232594183, + "loss": 2.3648, + "step": 1533 + }, + { + "epoch": 0.7033470884915176, + "grad_norm": 0.4319437460037222, + "learning_rate": 0.0002136682437448013, + "loss": 2.4359, + "step": 1534 + }, + { + "epoch": 0.7038055937643283, + "grad_norm": 0.4334665245516632, + "learning_rate": 0.0002130597069182994, + "loss": 2.3156, + "step": 1535 + }, + { + "epoch": 0.7042640990371389, + "grad_norm": 0.42819480046914504, + "learning_rate": 0.0002124518031890948, + "loss": 2.3674, + "step": 1536 + }, + { + "epoch": 0.7047226043099496, + "grad_norm": 0.4137995282392458, + "learning_rate": 0.0002118445338984502, + "loss": 2.3984, + "step": 1537 + }, + { + "epoch": 0.7051811095827603, + "grad_norm": 0.44027886588972576, + "learning_rate": 0.00021123790038622808, + "loss": 2.4352, + "step": 1538 + }, + { + "epoch": 0.7056396148555708, + "grad_norm": 0.4634230824286172, + "learning_rate": 0.0002106319039908879, + "loss": 2.5136, + "step": 1539 + }, + { + "epoch": 0.7060981201283815, + "grad_norm": 0.4179853070903575, + "learning_rate": 0.00021002654604948412, + "loss": 2.4014, + "step": 1540 + }, + { + "epoch": 0.7065566254011921, + "grad_norm": 0.42932417448027055, + "learning_rate": 0.00020942182789766172, + "loss": 2.3978, + "step": 1541 + }, + { + "epoch": 0.7070151306740028, + "grad_norm": 0.45338064992076604, + "learning_rate": 0.00020881775086965492, + "loss": 2.3825, + "step": 1542 + }, + { + "epoch": 0.7074736359468133, + "grad_norm": 0.44577193062382403, + "learning_rate": 0.00020821431629828246, + "loss": 2.3902, + "step": 1543 + }, + { + "epoch": 0.707932141219624, + "grad_norm": 0.4279300937853059, + "learning_rate": 0.00020761152551494643, + "loss": 2.3699, + "step": 1544 + }, + { + "epoch": 0.7083906464924347, + "grad_norm": 0.4110077920744941, + "learning_rate": 0.00020700937984962798, + "loss": 2.3324, + "step": 1545 + }, + { + "epoch": 0.7088491517652453, + "grad_norm": 0.4406636894729357, + "learning_rate": 0.0002064078806308848, + "loss": 2.4911, + "step": 1546 + }, + { + "epoch": 0.709307657038056, + "grad_norm": 0.4155066686186663, + "learning_rate": 0.00020580702918584882, + "loss": 2.4898, + "step": 1547 + }, + { + "epoch": 0.7097661623108665, + "grad_norm": 0.408492177909834, + "learning_rate": 0.000205206826840222, + "loss": 2.2451, + "step": 1548 + }, + { + "epoch": 0.7102246675836772, + "grad_norm": 0.4178541613529449, + "learning_rate": 0.0002046072749182751, + "loss": 2.463, + "step": 1549 + }, + { + "epoch": 0.7106831728564879, + "grad_norm": 0.41895601591251264, + "learning_rate": 0.00020400837474284273, + "loss": 2.4041, + "step": 1550 + }, + { + "epoch": 0.7111416781292985, + "grad_norm": 0.41334254090570566, + "learning_rate": 0.0002034101276353224, + "loss": 2.3917, + "step": 1551 + }, + { + "epoch": 0.7116001834021092, + "grad_norm": 0.40398089917021424, + "learning_rate": 0.00020281253491567027, + "loss": 2.4253, + "step": 1552 + }, + { + "epoch": 0.7120586886749197, + "grad_norm": 0.41021345085959543, + "learning_rate": 0.0002022155979023984, + "loss": 2.3547, + "step": 1553 + }, + { + "epoch": 0.7125171939477304, + "grad_norm": 0.42390704779786265, + "learning_rate": 0.000201619317912573, + "loss": 2.4041, + "step": 1554 + }, + { + "epoch": 0.712975699220541, + "grad_norm": 0.41881446098143543, + "learning_rate": 0.00020102369626180962, + "loss": 2.3522, + "step": 1555 + }, + { + "epoch": 0.7134342044933517, + "grad_norm": 0.41513826615155763, + "learning_rate": 0.0002004287342642721, + "loss": 2.3726, + "step": 1556 + }, + { + "epoch": 0.7138927097661623, + "grad_norm": 0.4287222800188132, + "learning_rate": 0.00019983443323266824, + "loss": 2.5177, + "step": 1557 + }, + { + "epoch": 0.7143512150389729, + "grad_norm": 0.4324062172870726, + "learning_rate": 0.00019924079447824805, + "loss": 2.3695, + "step": 1558 + }, + { + "epoch": 0.7148097203117836, + "grad_norm": 0.4242900173763143, + "learning_rate": 0.00019864781931079977, + "loss": 2.3118, + "step": 1559 + }, + { + "epoch": 0.7152682255845942, + "grad_norm": 0.409963486424083, + "learning_rate": 0.00019805550903864773, + "loss": 2.391, + "step": 1560 + }, + { + "epoch": 0.7157267308574049, + "grad_norm": 0.41601773746160103, + "learning_rate": 0.00019746386496864948, + "loss": 2.3636, + "step": 1561 + }, + { + "epoch": 0.7161852361302155, + "grad_norm": 0.425109237402067, + "learning_rate": 0.00019687288840619226, + "loss": 2.3756, + "step": 1562 + }, + { + "epoch": 0.7166437414030261, + "grad_norm": 0.42635170918973936, + "learning_rate": 0.0001962825806551911, + "loss": 2.3208, + "step": 1563 + }, + { + "epoch": 0.7171022466758368, + "grad_norm": 0.43152749863006185, + "learning_rate": 0.0001956929430180846, + "loss": 2.3116, + "step": 1564 + }, + { + "epoch": 0.7175607519486474, + "grad_norm": 0.4238210355587638, + "learning_rate": 0.00019510397679583374, + "loss": 2.3987, + "step": 1565 + }, + { + "epoch": 0.7180192572214581, + "grad_norm": 0.4312614597693188, + "learning_rate": 0.0001945156832879174, + "loss": 2.3119, + "step": 1566 + }, + { + "epoch": 0.7184777624942686, + "grad_norm": 0.43517061264714224, + "learning_rate": 0.00019392806379233036, + "loss": 2.3666, + "step": 1567 + }, + { + "epoch": 0.7189362677670793, + "grad_norm": 0.43127964641766864, + "learning_rate": 0.00019334111960558065, + "loss": 2.2725, + "step": 1568 + }, + { + "epoch": 0.71939477303989, + "grad_norm": 0.44590143397181176, + "learning_rate": 0.00019275485202268573, + "loss": 2.4829, + "step": 1569 + }, + { + "epoch": 0.7198532783127006, + "grad_norm": 0.4423363134333576, + "learning_rate": 0.00019216926233717085, + "loss": 2.2916, + "step": 1570 + }, + { + "epoch": 0.7203117835855112, + "grad_norm": 0.4120903146322442, + "learning_rate": 0.00019158435184106498, + "loss": 2.3964, + "step": 1571 + }, + { + "epoch": 0.7207702888583218, + "grad_norm": 0.44516260647469935, + "learning_rate": 0.00019100012182489905, + "loss": 2.3796, + "step": 1572 + }, + { + "epoch": 0.7212287941311325, + "grad_norm": 0.4282789946330222, + "learning_rate": 0.00019041657357770226, + "loss": 2.4083, + "step": 1573 + }, + { + "epoch": 0.7216872994039432, + "grad_norm": 0.4123369672931772, + "learning_rate": 0.00018983370838699943, + "loss": 2.3617, + "step": 1574 + }, + { + "epoch": 0.7221458046767538, + "grad_norm": 0.42155547799011533, + "learning_rate": 0.00018925152753880892, + "loss": 2.3017, + "step": 1575 + }, + { + "epoch": 0.7226043099495644, + "grad_norm": 0.40948820162526917, + "learning_rate": 0.00018867003231763847, + "loss": 2.4274, + "step": 1576 + }, + { + "epoch": 0.723062815222375, + "grad_norm": 0.4327810372224957, + "learning_rate": 0.00018808922400648375, + "loss": 2.3548, + "step": 1577 + }, + { + "epoch": 0.7235213204951857, + "grad_norm": 0.4073311998045277, + "learning_rate": 0.00018750910388682428, + "loss": 2.3844, + "step": 1578 + }, + { + "epoch": 0.7239798257679964, + "grad_norm": 0.42478218293047215, + "learning_rate": 0.00018692967323862125, + "loss": 2.338, + "step": 1579 + }, + { + "epoch": 0.724438331040807, + "grad_norm": 0.44457221220621335, + "learning_rate": 0.00018635093334031517, + "loss": 2.3564, + "step": 1580 + }, + { + "epoch": 0.7248968363136176, + "grad_norm": 0.4273121465304749, + "learning_rate": 0.00018577288546882165, + "loss": 2.303, + "step": 1581 + }, + { + "epoch": 0.7253553415864282, + "grad_norm": 0.43211393896141764, + "learning_rate": 0.00018519553089953023, + "loss": 2.385, + "step": 1582 + }, + { + "epoch": 0.7258138468592389, + "grad_norm": 0.4466265496888155, + "learning_rate": 0.0001846188709063001, + "loss": 2.4697, + "step": 1583 + }, + { + "epoch": 0.7262723521320495, + "grad_norm": 0.44030281283192846, + "learning_rate": 0.00018404290676145857, + "loss": 2.4741, + "step": 1584 + }, + { + "epoch": 0.7267308574048602, + "grad_norm": 0.42508491585312475, + "learning_rate": 0.00018346763973579722, + "loss": 2.361, + "step": 1585 + }, + { + "epoch": 0.7271893626776708, + "grad_norm": 0.45161150373859027, + "learning_rate": 0.00018289307109856939, + "loss": 2.3073, + "step": 1586 + }, + { + "epoch": 0.7276478679504814, + "grad_norm": 0.42968777309755835, + "learning_rate": 0.0001823192021174882, + "loss": 2.4098, + "step": 1587 + }, + { + "epoch": 0.7281063732232921, + "grad_norm": 0.4228903878439202, + "learning_rate": 0.0001817460340587223, + "loss": 2.3997, + "step": 1588 + }, + { + "epoch": 0.7285648784961027, + "grad_norm": 0.4286443919637372, + "learning_rate": 0.00018117356818689445, + "loss": 2.2772, + "step": 1589 + }, + { + "epoch": 0.7290233837689133, + "grad_norm": 0.4324037261247518, + "learning_rate": 0.00018060180576507756, + "loss": 2.4194, + "step": 1590 + }, + { + "epoch": 0.729481889041724, + "grad_norm": 0.43206022075007644, + "learning_rate": 0.00018003074805479313, + "loss": 2.4547, + "step": 1591 + }, + { + "epoch": 0.7299403943145346, + "grad_norm": 0.43038222155392525, + "learning_rate": 0.00017946039631600724, + "loss": 2.3352, + "step": 1592 + }, + { + "epoch": 0.7303988995873453, + "grad_norm": 0.43739755423040677, + "learning_rate": 0.00017889075180712837, + "loss": 2.4425, + "step": 1593 + }, + { + "epoch": 0.7308574048601559, + "grad_norm": 0.4280174790363362, + "learning_rate": 0.00017832181578500512, + "loss": 2.3693, + "step": 1594 + }, + { + "epoch": 0.7313159101329665, + "grad_norm": 0.44851589001623, + "learning_rate": 0.0001777535895049221, + "loss": 2.3091, + "step": 1595 + }, + { + "epoch": 0.7317744154057771, + "grad_norm": 0.42520746341633003, + "learning_rate": 0.0001771860742205988, + "loss": 2.4049, + "step": 1596 + }, + { + "epoch": 0.7322329206785878, + "grad_norm": 0.42427720694339316, + "learning_rate": 0.00017661927118418525, + "loss": 2.3282, + "step": 1597 + }, + { + "epoch": 0.7326914259513985, + "grad_norm": 0.4319944739695155, + "learning_rate": 0.00017605318164626066, + "loss": 2.3382, + "step": 1598 + }, + { + "epoch": 0.7331499312242091, + "grad_norm": 0.4369843820727463, + "learning_rate": 0.00017548780685582949, + "loss": 2.461, + "step": 1599 + }, + { + "epoch": 0.7336084364970197, + "grad_norm": 0.4232745259074503, + "learning_rate": 0.00017492314806031922, + "loss": 2.5144, + "step": 1600 + }, + { + "epoch": 0.7340669417698303, + "grad_norm": 0.4403486839467425, + "learning_rate": 0.00017435920650557806, + "loss": 2.2776, + "step": 1601 + }, + { + "epoch": 0.734525447042641, + "grad_norm": 0.40882496705171417, + "learning_rate": 0.00017379598343587112, + "loss": 2.3752, + "step": 1602 + }, + { + "epoch": 0.7349839523154517, + "grad_norm": 0.4209693903651445, + "learning_rate": 0.00017323348009387878, + "loss": 2.4547, + "step": 1603 + }, + { + "epoch": 0.7354424575882622, + "grad_norm": 0.43129583746085803, + "learning_rate": 0.0001726716977206929, + "loss": 2.3748, + "step": 1604 + }, + { + "epoch": 0.7359009628610729, + "grad_norm": 0.42846271317321, + "learning_rate": 0.00017211063755581525, + "loss": 2.3514, + "step": 1605 + }, + { + "epoch": 0.7363594681338835, + "grad_norm": 0.42804560771939215, + "learning_rate": 0.0001715503008371536, + "loss": 2.3795, + "step": 1606 + }, + { + "epoch": 0.7368179734066942, + "grad_norm": 0.42944838578935196, + "learning_rate": 0.0001709906888010196, + "loss": 2.44, + "step": 1607 + }, + { + "epoch": 0.7372764786795049, + "grad_norm": 0.42078614327318176, + "learning_rate": 0.00017043180268212638, + "loss": 2.3655, + "step": 1608 + }, + { + "epoch": 0.7377349839523154, + "grad_norm": 0.45131935432974024, + "learning_rate": 0.00016987364371358481, + "loss": 2.3885, + "step": 1609 + }, + { + "epoch": 0.7381934892251261, + "grad_norm": 0.4347546921657413, + "learning_rate": 0.00016931621312690214, + "loss": 2.4297, + "step": 1610 + }, + { + "epoch": 0.7386519944979367, + "grad_norm": 0.41991031264865525, + "learning_rate": 0.00016875951215197777, + "loss": 2.4138, + "step": 1611 + }, + { + "epoch": 0.7391104997707474, + "grad_norm": 0.4535256982373814, + "learning_rate": 0.00016820354201710214, + "loss": 2.3502, + "step": 1612 + }, + { + "epoch": 0.739569005043558, + "grad_norm": 0.4285515724444353, + "learning_rate": 0.00016764830394895203, + "loss": 2.4019, + "step": 1613 + }, + { + "epoch": 0.7400275103163686, + "grad_norm": 0.4509945591709699, + "learning_rate": 0.00016709379917259027, + "loss": 2.3855, + "step": 1614 + }, + { + "epoch": 0.7404860155891793, + "grad_norm": 0.4090981471918442, + "learning_rate": 0.00016654002891146091, + "loss": 2.3363, + "step": 1615 + }, + { + "epoch": 0.7409445208619899, + "grad_norm": 0.4473400022890688, + "learning_rate": 0.00016598699438738764, + "loss": 2.3477, + "step": 1616 + }, + { + "epoch": 0.7414030261348006, + "grad_norm": 0.4310454514736446, + "learning_rate": 0.00016543469682057105, + "loss": 2.4161, + "step": 1617 + }, + { + "epoch": 0.7418615314076112, + "grad_norm": 0.4246024210660052, + "learning_rate": 0.00016488313742958526, + "loss": 2.4296, + "step": 1618 + }, + { + "epoch": 0.7423200366804218, + "grad_norm": 0.4442241724637171, + "learning_rate": 0.00016433231743137646, + "loss": 2.4177, + "step": 1619 + }, + { + "epoch": 0.7427785419532325, + "grad_norm": 0.42412754891547094, + "learning_rate": 0.0001637822380412584, + "loss": 2.3436, + "step": 1620 + }, + { + "epoch": 0.7432370472260431, + "grad_norm": 0.42505612742309673, + "learning_rate": 0.00016323290047291195, + "loss": 2.3231, + "step": 1621 + }, + { + "epoch": 0.7436955524988538, + "grad_norm": 0.43454744688699365, + "learning_rate": 0.0001626843059383803, + "loss": 2.4488, + "step": 1622 + }, + { + "epoch": 0.7441540577716643, + "grad_norm": 0.4530816973402632, + "learning_rate": 0.00016213645564806752, + "loss": 2.4654, + "step": 1623 + }, + { + "epoch": 0.744612563044475, + "grad_norm": 0.43994032370659586, + "learning_rate": 0.0001615893508107359, + "loss": 2.244, + "step": 1624 + }, + { + "epoch": 0.7450710683172856, + "grad_norm": 0.40904860912070384, + "learning_rate": 0.00016104299263350252, + "loss": 2.3952, + "step": 1625 + }, + { + "epoch": 0.7455295735900963, + "grad_norm": 0.4121744043277294, + "learning_rate": 0.00016049738232183758, + "loss": 2.3498, + "step": 1626 + }, + { + "epoch": 0.745988078862907, + "grad_norm": 0.42194665249555585, + "learning_rate": 0.0001599525210795606, + "loss": 2.2748, + "step": 1627 + }, + { + "epoch": 0.7464465841357175, + "grad_norm": 0.41250044972695954, + "learning_rate": 0.00015940841010883889, + "loss": 2.3757, + "step": 1628 + }, + { + "epoch": 0.7469050894085282, + "grad_norm": 0.43422140905243667, + "learning_rate": 0.00015886505061018413, + "loss": 2.3633, + "step": 1629 + }, + { + "epoch": 0.7473635946813388, + "grad_norm": 0.41538094366087774, + "learning_rate": 0.0001583224437824498, + "loss": 2.343, + "step": 1630 + }, + { + "epoch": 0.7478220999541495, + "grad_norm": 0.43355988115696464, + "learning_rate": 0.0001577805908228293, + "loss": 2.4464, + "step": 1631 + }, + { + "epoch": 0.7482806052269602, + "grad_norm": 0.43855825042407076, + "learning_rate": 0.00015723949292685191, + "loss": 2.3157, + "step": 1632 + }, + { + "epoch": 0.7487391104997707, + "grad_norm": 0.45246506623351584, + "learning_rate": 0.0001566991512883818, + "loss": 2.3361, + "step": 1633 + }, + { + "epoch": 0.7491976157725814, + "grad_norm": 0.4380495441612108, + "learning_rate": 0.00015615956709961378, + "loss": 2.4926, + "step": 1634 + }, + { + "epoch": 0.749656121045392, + "grad_norm": 0.42510963735690827, + "learning_rate": 0.00015562074155107215, + "loss": 2.3684, + "step": 1635 + }, + { + "epoch": 0.7501146263182027, + "grad_norm": 0.46139009277142917, + "learning_rate": 0.0001550826758316068, + "loss": 2.4541, + "step": 1636 + }, + { + "epoch": 0.7505731315910132, + "grad_norm": 0.46573091544828554, + "learning_rate": 0.00015454537112839122, + "loss": 2.4722, + "step": 1637 + }, + { + "epoch": 0.7510316368638239, + "grad_norm": 0.4220135689418687, + "learning_rate": 0.00015400882862692033, + "loss": 2.4142, + "step": 1638 + }, + { + "epoch": 0.7514901421366346, + "grad_norm": 0.449632393310306, + "learning_rate": 0.00015347304951100665, + "loss": 2.407, + "step": 1639 + }, + { + "epoch": 0.7519486474094452, + "grad_norm": 0.43189822791096033, + "learning_rate": 0.00015293803496277907, + "loss": 2.4626, + "step": 1640 + }, + { + "epoch": 0.7524071526822559, + "grad_norm": 0.44626191117147207, + "learning_rate": 0.00015240378616267886, + "loss": 2.4306, + "step": 1641 + }, + { + "epoch": 0.7528656579550664, + "grad_norm": 0.44696655930829576, + "learning_rate": 0.00015187030428945843, + "loss": 2.264, + "step": 1642 + }, + { + "epoch": 0.7533241632278771, + "grad_norm": 0.4248971727423041, + "learning_rate": 0.0001513375905201776, + "loss": 2.4202, + "step": 1643 + }, + { + "epoch": 0.7537826685006878, + "grad_norm": 0.41520153455875025, + "learning_rate": 0.00015080564603020142, + "loss": 2.3807, + "step": 1644 + }, + { + "epoch": 0.7542411737734984, + "grad_norm": 0.43882739739148025, + "learning_rate": 0.0001502744719931982, + "loss": 2.2811, + "step": 1645 + }, + { + "epoch": 0.7546996790463091, + "grad_norm": 0.418549956327463, + "learning_rate": 0.00014974406958113558, + "loss": 2.4138, + "step": 1646 + }, + { + "epoch": 0.7551581843191196, + "grad_norm": 0.4536972296717646, + "learning_rate": 0.00014921443996427947, + "loss": 2.4105, + "step": 1647 + }, + { + "epoch": 0.7556166895919303, + "grad_norm": 0.4382804943856357, + "learning_rate": 0.0001486855843111901, + "loss": 2.463, + "step": 1648 + }, + { + "epoch": 0.756075194864741, + "grad_norm": 0.4290472659182229, + "learning_rate": 0.0001481575037887201, + "loss": 2.4379, + "step": 1649 + }, + { + "epoch": 0.7565337001375516, + "grad_norm": 0.41578382861818103, + "learning_rate": 0.00014763019956201253, + "loss": 2.4305, + "step": 1650 + }, + { + "epoch": 0.7569922054103623, + "grad_norm": 0.4344045405904627, + "learning_rate": 0.0001471036727944966, + "loss": 2.3306, + "step": 1651 + }, + { + "epoch": 0.7574507106831728, + "grad_norm": 0.4153861728887541, + "learning_rate": 0.0001465779246478872, + "loss": 2.3239, + "step": 1652 + }, + { + "epoch": 0.7579092159559835, + "grad_norm": 0.4149823102296612, + "learning_rate": 0.00014605295628218045, + "loss": 2.2755, + "step": 1653 + }, + { + "epoch": 0.7583677212287941, + "grad_norm": 0.4305013162051645, + "learning_rate": 0.0001455287688556527, + "loss": 2.3674, + "step": 1654 + }, + { + "epoch": 0.7588262265016048, + "grad_norm": 0.4204190446554248, + "learning_rate": 0.00014500536352485673, + "loss": 2.3282, + "step": 1655 + }, + { + "epoch": 0.7592847317744154, + "grad_norm": 0.4170709369235301, + "learning_rate": 0.00014448274144461965, + "loss": 2.3919, + "step": 1656 + }, + { + "epoch": 0.759743237047226, + "grad_norm": 0.41314913162611855, + "learning_rate": 0.00014396090376804112, + "loss": 2.4148, + "step": 1657 + }, + { + "epoch": 0.7602017423200367, + "grad_norm": 0.4027519301046483, + "learning_rate": 0.00014343985164648926, + "loss": 2.2627, + "step": 1658 + }, + { + "epoch": 0.7606602475928473, + "grad_norm": 0.42014747120985824, + "learning_rate": 0.00014291958622959973, + "loss": 2.5268, + "step": 1659 + }, + { + "epoch": 0.761118752865658, + "grad_norm": 0.439965151014772, + "learning_rate": 0.00014240010866527176, + "loss": 2.4, + "step": 1660 + }, + { + "epoch": 0.7615772581384685, + "grad_norm": 0.4231200379044763, + "learning_rate": 0.00014188142009966686, + "loss": 2.3816, + "step": 1661 + }, + { + "epoch": 0.7620357634112792, + "grad_norm": 0.43763127683055847, + "learning_rate": 0.0001413635216772053, + "loss": 2.3228, + "step": 1662 + }, + { + "epoch": 0.7624942686840899, + "grad_norm": 0.4399774768936926, + "learning_rate": 0.000140846414540564, + "loss": 2.4111, + "step": 1663 + }, + { + "epoch": 0.7629527739569005, + "grad_norm": 0.41917755743820734, + "learning_rate": 0.00014033009983067452, + "loss": 2.3795, + "step": 1664 + }, + { + "epoch": 0.7634112792297112, + "grad_norm": 0.4140037304291673, + "learning_rate": 0.00013981457868671927, + "loss": 2.3053, + "step": 1665 + }, + { + "epoch": 0.7638697845025217, + "grad_norm": 0.4202058500040669, + "learning_rate": 0.0001392998522461305, + "loss": 2.3555, + "step": 1666 + }, + { + "epoch": 0.7643282897753324, + "grad_norm": 0.4510593846894919, + "learning_rate": 0.00013878592164458635, + "loss": 2.4206, + "step": 1667 + }, + { + "epoch": 0.7647867950481431, + "grad_norm": 0.4440911964372599, + "learning_rate": 0.00013827278801600978, + "loss": 2.3627, + "step": 1668 + }, + { + "epoch": 0.7652453003209537, + "grad_norm": 0.4227409920269739, + "learning_rate": 0.0001377604524925647, + "loss": 2.3758, + "step": 1669 + }, + { + "epoch": 0.7657038055937643, + "grad_norm": 0.4341315819401016, + "learning_rate": 0.00013724891620465424, + "loss": 2.321, + "step": 1670 + }, + { + "epoch": 0.7661623108665749, + "grad_norm": 0.45006844831251375, + "learning_rate": 0.0001367381802809185, + "loss": 2.4225, + "step": 1671 + }, + { + "epoch": 0.7666208161393856, + "grad_norm": 0.4557560735945094, + "learning_rate": 0.00013622824584823113, + "loss": 2.3671, + "step": 1672 + }, + { + "epoch": 0.7670793214121963, + "grad_norm": 0.4312135160517491, + "learning_rate": 0.00013571911403169795, + "loss": 2.3927, + "step": 1673 + }, + { + "epoch": 0.7675378266850069, + "grad_norm": 0.44225493034273466, + "learning_rate": 0.0001352107859546533, + "loss": 2.4234, + "step": 1674 + }, + { + "epoch": 0.7679963319578175, + "grad_norm": 0.4240562683562441, + "learning_rate": 0.00013470326273865886, + "loss": 2.4026, + "step": 1675 + }, + { + "epoch": 0.7684548372306281, + "grad_norm": 0.4285738312743796, + "learning_rate": 0.00013419654550349985, + "loss": 2.3322, + "step": 1676 + }, + { + "epoch": 0.7689133425034388, + "grad_norm": 0.42450175144199803, + "learning_rate": 0.00013369063536718346, + "loss": 2.3151, + "step": 1677 + }, + { + "epoch": 0.7693718477762495, + "grad_norm": 0.4459842824405571, + "learning_rate": 0.00013318553344593632, + "loss": 2.3287, + "step": 1678 + }, + { + "epoch": 0.7698303530490601, + "grad_norm": 0.45904436657846176, + "learning_rate": 0.00013268124085420136, + "loss": 2.3499, + "step": 1679 + }, + { + "epoch": 0.7702888583218707, + "grad_norm": 0.43473874334707324, + "learning_rate": 0.0001321777587046364, + "loss": 2.377, + "step": 1680 + }, + { + "epoch": 0.7707473635946813, + "grad_norm": 0.44701668108073644, + "learning_rate": 0.00013167508810811059, + "loss": 2.3497, + "step": 1681 + }, + { + "epoch": 0.771205868867492, + "grad_norm": 0.43368400459183853, + "learning_rate": 0.0001311732301737029, + "loss": 2.4057, + "step": 1682 + }, + { + "epoch": 0.7716643741403026, + "grad_norm": 0.4187345699969937, + "learning_rate": 0.0001306721860086991, + "loss": 2.4097, + "step": 1683 + }, + { + "epoch": 0.7721228794131133, + "grad_norm": 0.41482178778982076, + "learning_rate": 0.00013017195671858928, + "loss": 2.3023, + "step": 1684 + }, + { + "epoch": 0.7725813846859239, + "grad_norm": 0.40896152385362694, + "learning_rate": 0.0001296725434070661, + "loss": 2.3164, + "step": 1685 + }, + { + "epoch": 0.7730398899587345, + "grad_norm": 0.4190537380749563, + "learning_rate": 0.00012917394717602121, + "loss": 2.3758, + "step": 1686 + }, + { + "epoch": 0.7734983952315452, + "grad_norm": 0.44013182443953447, + "learning_rate": 0.00012867616912554426, + "loss": 2.5028, + "step": 1687 + }, + { + "epoch": 0.7739569005043558, + "grad_norm": 0.42723732653205626, + "learning_rate": 0.00012817921035391882, + "loss": 2.3739, + "step": 1688 + }, + { + "epoch": 0.7744154057771664, + "grad_norm": 0.43959736682049094, + "learning_rate": 0.00012768307195762168, + "loss": 2.4331, + "step": 1689 + }, + { + "epoch": 0.774873911049977, + "grad_norm": 0.45294060377279666, + "learning_rate": 0.00012718775503131908, + "loss": 2.3379, + "step": 1690 + }, + { + "epoch": 0.7753324163227877, + "grad_norm": 0.40727139601802514, + "learning_rate": 0.0001266932606678646, + "loss": 2.3014, + "step": 1691 + }, + { + "epoch": 0.7757909215955984, + "grad_norm": 0.4340676947887724, + "learning_rate": 0.00012619958995829756, + "loss": 2.3623, + "step": 1692 + }, + { + "epoch": 0.776249426868409, + "grad_norm": 0.4322053576315099, + "learning_rate": 0.0001257067439918394, + "loss": 2.3125, + "step": 1693 + }, + { + "epoch": 0.7767079321412196, + "grad_norm": 0.43054744921417376, + "learning_rate": 0.00012521472385589234, + "loss": 2.4163, + "step": 1694 + }, + { + "epoch": 0.7771664374140302, + "grad_norm": 0.4458364778455876, + "learning_rate": 0.00012472353063603626, + "loss": 2.3796, + "step": 1695 + }, + { + "epoch": 0.7776249426868409, + "grad_norm": 0.4446022511545629, + "learning_rate": 0.0001242331654160263, + "loss": 2.3743, + "step": 1696 + }, + { + "epoch": 0.7780834479596516, + "grad_norm": 0.4461790620708075, + "learning_rate": 0.0001237436292777914, + "loss": 2.4548, + "step": 1697 + }, + { + "epoch": 0.7785419532324622, + "grad_norm": 0.41303609655562507, + "learning_rate": 0.00012325492330143061, + "loss": 2.4213, + "step": 1698 + }, + { + "epoch": 0.7790004585052728, + "grad_norm": 0.42974039776269335, + "learning_rate": 0.00012276704856521175, + "loss": 2.4788, + "step": 1699 + }, + { + "epoch": 0.7794589637780834, + "grad_norm": 0.41926053721410556, + "learning_rate": 0.00012228000614556816, + "loss": 2.2926, + "step": 1700 + }, + { + "epoch": 0.7799174690508941, + "grad_norm": 0.4271209391798469, + "learning_rate": 0.00012179379711709738, + "loss": 2.2344, + "step": 1701 + }, + { + "epoch": 0.7803759743237048, + "grad_norm": 0.43063775563933787, + "learning_rate": 0.0001213084225525577, + "loss": 2.5004, + "step": 1702 + }, + { + "epoch": 0.7808344795965153, + "grad_norm": 0.4119981692259131, + "learning_rate": 0.00012082388352286627, + "loss": 2.4077, + "step": 1703 + }, + { + "epoch": 0.781292984869326, + "grad_norm": 0.41443762959250446, + "learning_rate": 0.00012034018109709716, + "loss": 2.3722, + "step": 1704 + }, + { + "epoch": 0.7817514901421366, + "grad_norm": 0.425182375094494, + "learning_rate": 0.00011985731634247809, + "loss": 2.4388, + "step": 1705 + }, + { + "epoch": 0.7822099954149473, + "grad_norm": 0.4321029686282629, + "learning_rate": 0.00011937529032438904, + "loss": 2.3676, + "step": 1706 + }, + { + "epoch": 0.782668500687758, + "grad_norm": 0.4191680486443166, + "learning_rate": 0.00011889410410635887, + "loss": 2.3523, + "step": 1707 + }, + { + "epoch": 0.7831270059605685, + "grad_norm": 0.4299328205313405, + "learning_rate": 0.0001184137587500641, + "loss": 2.4224, + "step": 1708 + }, + { + "epoch": 0.7835855112333792, + "grad_norm": 0.4713042983767211, + "learning_rate": 0.00011793425531532564, + "loss": 2.3294, + "step": 1709 + }, + { + "epoch": 0.7840440165061898, + "grad_norm": 0.4290584400054199, + "learning_rate": 0.00011745559486010671, + "loss": 2.481, + "step": 1710 + }, + { + "epoch": 0.7845025217790005, + "grad_norm": 0.42321911716165345, + "learning_rate": 0.00011697777844051105, + "loss": 2.3152, + "step": 1711 + }, + { + "epoch": 0.7849610270518111, + "grad_norm": 0.42308715670252967, + "learning_rate": 0.00011650080711077964, + "loss": 2.3747, + "step": 1712 + }, + { + "epoch": 0.7854195323246217, + "grad_norm": 0.44302859681018714, + "learning_rate": 0.00011602468192328936, + "loss": 2.368, + "step": 1713 + }, + { + "epoch": 0.7858780375974324, + "grad_norm": 0.44070646397636487, + "learning_rate": 0.00011554940392854973, + "loss": 2.3619, + "step": 1714 + }, + { + "epoch": 0.786336542870243, + "grad_norm": 0.43606736159280013, + "learning_rate": 0.00011507497417520146, + "loss": 2.4271, + "step": 1715 + }, + { + "epoch": 0.7867950481430537, + "grad_norm": 0.44771694100442094, + "learning_rate": 0.00011460139371001339, + "loss": 2.4565, + "step": 1716 + }, + { + "epoch": 0.7872535534158643, + "grad_norm": 0.41573540250438173, + "learning_rate": 0.00011412866357788049, + "loss": 2.3506, + "step": 1717 + }, + { + "epoch": 0.7877120586886749, + "grad_norm": 0.4264442159930472, + "learning_rate": 0.00011365678482182207, + "loss": 2.3975, + "step": 1718 + }, + { + "epoch": 0.7881705639614855, + "grad_norm": 0.41855819738392097, + "learning_rate": 0.0001131857584829783, + "loss": 2.3919, + "step": 1719 + }, + { + "epoch": 0.7886290692342962, + "grad_norm": 0.4502021893495299, + "learning_rate": 0.0001127155856006093, + "loss": 2.3447, + "step": 1720 + }, + { + "epoch": 0.7890875745071069, + "grad_norm": 0.4263953892135246, + "learning_rate": 0.00011224626721209141, + "loss": 2.3838, + "step": 1721 + }, + { + "epoch": 0.7895460797799174, + "grad_norm": 0.4442767528270572, + "learning_rate": 0.0001117778043529164, + "loss": 2.4265, + "step": 1722 + }, + { + "epoch": 0.7900045850527281, + "grad_norm": 0.4400550841484901, + "learning_rate": 0.0001113101980566879, + "loss": 2.4087, + "step": 1723 + }, + { + "epoch": 0.7904630903255387, + "grad_norm": 0.4431040757195574, + "learning_rate": 0.00011084344935511958, + "loss": 2.3319, + "step": 1724 + }, + { + "epoch": 0.7909215955983494, + "grad_norm": 0.4170439557363895, + "learning_rate": 0.00011037755927803345, + "loss": 2.3545, + "step": 1725 + }, + { + "epoch": 0.7913801008711601, + "grad_norm": 0.41486627600417636, + "learning_rate": 0.00010991252885335651, + "loss": 2.3686, + "step": 1726 + }, + { + "epoch": 0.7918386061439706, + "grad_norm": 0.4289805222098861, + "learning_rate": 0.00010944835910711958, + "loss": 2.3469, + "step": 1727 + }, + { + "epoch": 0.7922971114167813, + "grad_norm": 0.46048993798321247, + "learning_rate": 0.00010898505106345396, + "loss": 2.4625, + "step": 1728 + }, + { + "epoch": 0.7927556166895919, + "grad_norm": 0.42217891257795975, + "learning_rate": 0.00010852260574459022, + "loss": 2.3048, + "step": 1729 + }, + { + "epoch": 0.7932141219624026, + "grad_norm": 0.4444867919458544, + "learning_rate": 0.00010806102417085512, + "loss": 2.3752, + "step": 1730 + }, + { + "epoch": 0.7936726272352133, + "grad_norm": 0.42547359706610105, + "learning_rate": 0.00010760030736066951, + "loss": 2.4423, + "step": 1731 + }, + { + "epoch": 0.7941311325080238, + "grad_norm": 0.42728575876581193, + "learning_rate": 0.00010714045633054687, + "loss": 2.3912, + "step": 1732 + }, + { + "epoch": 0.7945896377808345, + "grad_norm": 0.4214878485830814, + "learning_rate": 0.00010668147209508971, + "loss": 2.2925, + "step": 1733 + }, + { + "epoch": 0.7950481430536451, + "grad_norm": 0.43720643338185344, + "learning_rate": 0.00010622335566698877, + "loss": 2.5422, + "step": 1734 + }, + { + "epoch": 0.7955066483264558, + "grad_norm": 0.4194834002919776, + "learning_rate": 0.00010576610805701942, + "loss": 2.4158, + "step": 1735 + }, + { + "epoch": 0.7959651535992663, + "grad_norm": 0.449164680668503, + "learning_rate": 0.00010530973027404073, + "loss": 2.3622, + "step": 1736 + }, + { + "epoch": 0.796423658872077, + "grad_norm": 0.5140896483473295, + "learning_rate": 0.00010485422332499212, + "loss": 2.3913, + "step": 1737 + }, + { + "epoch": 0.7968821641448877, + "grad_norm": 0.432128413383289, + "learning_rate": 0.00010439958821489165, + "loss": 2.4097, + "step": 1738 + }, + { + "epoch": 0.7973406694176983, + "grad_norm": 0.4131573372711721, + "learning_rate": 0.00010394582594683428, + "loss": 2.4945, + "step": 1739 + }, + { + "epoch": 0.797799174690509, + "grad_norm": 0.41946230846696164, + "learning_rate": 0.0001034929375219884, + "loss": 2.3237, + "step": 1740 + }, + { + "epoch": 0.7982576799633195, + "grad_norm": 0.4477906615023577, + "learning_rate": 0.00010304092393959514, + "loss": 2.3019, + "step": 1741 + }, + { + "epoch": 0.7987161852361302, + "grad_norm": 0.4134057987318842, + "learning_rate": 0.00010258978619696468, + "loss": 2.3948, + "step": 1742 + }, + { + "epoch": 0.7991746905089409, + "grad_norm": 0.43022561907690354, + "learning_rate": 0.00010213952528947551, + "loss": 2.4408, + "step": 1743 + }, + { + "epoch": 0.7996331957817515, + "grad_norm": 0.4362143330485859, + "learning_rate": 0.00010169014221057089, + "loss": 2.3264, + "step": 1744 + }, + { + "epoch": 0.8000917010545622, + "grad_norm": 0.4480126349961008, + "learning_rate": 0.00010124163795175734, + "loss": 2.3754, + "step": 1745 + }, + { + "epoch": 0.8005502063273727, + "grad_norm": 0.4150378949485513, + "learning_rate": 0.00010079401350260287, + "loss": 2.4069, + "step": 1746 + }, + { + "epoch": 0.8010087116001834, + "grad_norm": 0.4277001975832102, + "learning_rate": 0.00010034726985073362, + "loss": 2.4384, + "step": 1747 + }, + { + "epoch": 0.801467216872994, + "grad_norm": 0.42401378329359546, + "learning_rate": 9.9901407981833e-05, + "loss": 2.415, + "step": 1748 + }, + { + "epoch": 0.8019257221458047, + "grad_norm": 0.42186304220056836, + "learning_rate": 9.94564288796384e-05, + "loss": 2.3111, + "step": 1749 + }, + { + "epoch": 0.8023842274186153, + "grad_norm": 0.40614954830976974, + "learning_rate": 9.901233352593953e-05, + "loss": 2.4017, + "step": 1750 + }, + { + "epoch": 0.8028427326914259, + "grad_norm": 0.42344526141037175, + "learning_rate": 9.856912290057668e-05, + "loss": 2.381, + "step": 1751 + }, + { + "epoch": 0.8033012379642366, + "grad_norm": 0.42258168097500465, + "learning_rate": 9.812679798143748e-05, + "loss": 2.3898, + "step": 1752 + }, + { + "epoch": 0.8037597432370472, + "grad_norm": 0.4208799781867097, + "learning_rate": 9.768535974445586e-05, + "loss": 2.3969, + "step": 1753 + }, + { + "epoch": 0.8042182485098579, + "grad_norm": 0.4345805756289245, + "learning_rate": 9.724480916360906e-05, + "loss": 2.3946, + "step": 1754 + }, + { + "epoch": 0.8046767537826685, + "grad_norm": 0.4471809713656317, + "learning_rate": 9.68051472109162e-05, + "loss": 2.4942, + "step": 1755 + }, + { + "epoch": 0.8051352590554791, + "grad_norm": 0.4323213053319276, + "learning_rate": 9.636637485643529e-05, + "loss": 2.3024, + "step": 1756 + }, + { + "epoch": 0.8055937643282898, + "grad_norm": 0.4156604869000618, + "learning_rate": 9.592849306826174e-05, + "loss": 2.3598, + "step": 1757 + }, + { + "epoch": 0.8060522696011004, + "grad_norm": 0.44003230882656547, + "learning_rate": 9.549150281252633e-05, + "loss": 2.495, + "step": 1758 + }, + { + "epoch": 0.8065107748739111, + "grad_norm": 0.4530183075854547, + "learning_rate": 9.505540505339223e-05, + "loss": 2.2902, + "step": 1759 + }, + { + "epoch": 0.8069692801467216, + "grad_norm": 0.44435344192588744, + "learning_rate": 9.4620200753054e-05, + "loss": 2.4008, + "step": 1760 + }, + { + "epoch": 0.8074277854195323, + "grad_norm": 0.4462961214861632, + "learning_rate": 9.418589087173441e-05, + "loss": 2.3534, + "step": 1761 + }, + { + "epoch": 0.807886290692343, + "grad_norm": 0.44147485858122826, + "learning_rate": 9.375247636768325e-05, + "loss": 2.3049, + "step": 1762 + }, + { + "epoch": 0.8083447959651536, + "grad_norm": 0.44289008534096275, + "learning_rate": 9.331995819717443e-05, + "loss": 2.3987, + "step": 1763 + }, + { + "epoch": 0.8088033012379643, + "grad_norm": 0.4303174056351489, + "learning_rate": 9.288833731450419e-05, + "loss": 2.3462, + "step": 1764 + }, + { + "epoch": 0.8092618065107748, + "grad_norm": 0.4340188384857429, + "learning_rate": 9.245761467198948e-05, + "loss": 2.3631, + "step": 1765 + }, + { + "epoch": 0.8097203117835855, + "grad_norm": 0.43722953615981125, + "learning_rate": 9.20277912199648e-05, + "loss": 2.2687, + "step": 1766 + }, + { + "epoch": 0.8101788170563962, + "grad_norm": 0.4390475519559125, + "learning_rate": 9.159886790678123e-05, + "loss": 2.4069, + "step": 1767 + }, + { + "epoch": 0.8106373223292068, + "grad_norm": 0.4216119370469256, + "learning_rate": 9.11708456788033e-05, + "loss": 2.3301, + "step": 1768 + }, + { + "epoch": 0.8110958276020174, + "grad_norm": 0.4311093617309789, + "learning_rate": 9.074372548040793e-05, + "loss": 2.3222, + "step": 1769 + }, + { + "epoch": 0.811554332874828, + "grad_norm": 0.4292942214298935, + "learning_rate": 9.031750825398145e-05, + "loss": 2.4259, + "step": 1770 + }, + { + "epoch": 0.8120128381476387, + "grad_norm": 0.433102595756274, + "learning_rate": 8.98921949399179e-05, + "loss": 2.4256, + "step": 1771 + }, + { + "epoch": 0.8124713434204494, + "grad_norm": 0.4313956559134939, + "learning_rate": 8.94677864766173e-05, + "loss": 2.3315, + "step": 1772 + }, + { + "epoch": 0.81292984869326, + "grad_norm": 0.47383452033375834, + "learning_rate": 8.904428380048269e-05, + "loss": 2.4092, + "step": 1773 + }, + { + "epoch": 0.8133883539660706, + "grad_norm": 0.41956581167824564, + "learning_rate": 8.862168784591929e-05, + "loss": 2.3817, + "step": 1774 + }, + { + "epoch": 0.8138468592388812, + "grad_norm": 0.4313590367303316, + "learning_rate": 8.819999954533115e-05, + "loss": 2.3385, + "step": 1775 + }, + { + "epoch": 0.8143053645116919, + "grad_norm": 0.4419524104850291, + "learning_rate": 8.777921982911996e-05, + "loss": 2.275, + "step": 1776 + }, + { + "epoch": 0.8147638697845025, + "grad_norm": 0.4367251499511904, + "learning_rate": 8.735934962568253e-05, + "loss": 2.3295, + "step": 1777 + }, + { + "epoch": 0.8152223750573132, + "grad_norm": 0.4447497039548884, + "learning_rate": 8.694038986140945e-05, + "loss": 2.3922, + "step": 1778 + }, + { + "epoch": 0.8156808803301238, + "grad_norm": 0.4389566256578111, + "learning_rate": 8.652234146068206e-05, + "loss": 2.3621, + "step": 1779 + }, + { + "epoch": 0.8161393856029344, + "grad_norm": 0.44275070583990417, + "learning_rate": 8.610520534587086e-05, + "loss": 2.3093, + "step": 1780 + }, + { + "epoch": 0.8165978908757451, + "grad_norm": 0.44585222368410743, + "learning_rate": 8.568898243733397e-05, + "loss": 2.3366, + "step": 1781 + }, + { + "epoch": 0.8170563961485557, + "grad_norm": 0.45092393747537324, + "learning_rate": 8.527367365341409e-05, + "loss": 2.4281, + "step": 1782 + }, + { + "epoch": 0.8175149014213663, + "grad_norm": 0.44979917382692663, + "learning_rate": 8.485927991043757e-05, + "loss": 2.4707, + "step": 1783 + }, + { + "epoch": 0.817973406694177, + "grad_norm": 0.4516621032879077, + "learning_rate": 8.444580212271125e-05, + "loss": 2.4223, + "step": 1784 + }, + { + "epoch": 0.8184319119669876, + "grad_norm": 0.42798259007439743, + "learning_rate": 8.403324120252159e-05, + "loss": 2.4051, + "step": 1785 + }, + { + "epoch": 0.8188904172397983, + "grad_norm": 0.44382705341651923, + "learning_rate": 8.362159806013175e-05, + "loss": 2.3233, + "step": 1786 + }, + { + "epoch": 0.8193489225126089, + "grad_norm": 0.43556100000262976, + "learning_rate": 8.321087360377988e-05, + "loss": 2.3412, + "step": 1787 + }, + { + "epoch": 0.8198074277854195, + "grad_norm": 0.4469530096376487, + "learning_rate": 8.280106873967752e-05, + "loss": 2.4658, + "step": 1788 + }, + { + "epoch": 0.8202659330582301, + "grad_norm": 0.4190939619256631, + "learning_rate": 8.239218437200679e-05, + "loss": 2.3612, + "step": 1789 + }, + { + "epoch": 0.8207244383310408, + "grad_norm": 0.4318856684878558, + "learning_rate": 8.198422140291939e-05, + "loss": 2.2942, + "step": 1790 + }, + { + "epoch": 0.8211829436038515, + "grad_norm": 0.5384460752426182, + "learning_rate": 8.157718073253351e-05, + "loss": 2.4419, + "step": 1791 + }, + { + "epoch": 0.8216414488766621, + "grad_norm": 0.4472409960210402, + "learning_rate": 8.117106325893287e-05, + "loss": 2.3095, + "step": 1792 + }, + { + "epoch": 0.8220999541494727, + "grad_norm": 0.4422700956252072, + "learning_rate": 8.076586987816404e-05, + "loss": 2.3634, + "step": 1793 + }, + { + "epoch": 0.8225584594222833, + "grad_norm": 0.44451612445125155, + "learning_rate": 8.036160148423449e-05, + "loss": 2.3702, + "step": 1794 + }, + { + "epoch": 0.823016964695094, + "grad_norm": 0.4315923840764446, + "learning_rate": 7.995825896911141e-05, + "loss": 2.36, + "step": 1795 + }, + { + "epoch": 0.8234754699679047, + "grad_norm": 0.4302893869095022, + "learning_rate": 7.955584322271853e-05, + "loss": 2.377, + "step": 1796 + }, + { + "epoch": 0.8239339752407153, + "grad_norm": 0.427218035730301, + "learning_rate": 7.915435513293523e-05, + "loss": 2.3895, + "step": 1797 + }, + { + "epoch": 0.8243924805135259, + "grad_norm": 0.4343302303326156, + "learning_rate": 7.875379558559387e-05, + "loss": 2.2154, + "step": 1798 + }, + { + "epoch": 0.8248509857863365, + "grad_norm": 0.43504335864307414, + "learning_rate": 7.835416546447838e-05, + "loss": 2.3735, + "step": 1799 + }, + { + "epoch": 0.8253094910591472, + "grad_norm": 0.42905323998380307, + "learning_rate": 7.795546565132167e-05, + "loss": 2.3797, + "step": 1800 + }, + { + "epoch": 0.8257679963319579, + "grad_norm": 0.4612498114360641, + "learning_rate": 7.755769702580412e-05, + "loss": 2.3704, + "step": 1801 + }, + { + "epoch": 0.8262265016047684, + "grad_norm": 0.43261984938411824, + "learning_rate": 7.716086046555193e-05, + "loss": 2.3543, + "step": 1802 + }, + { + "epoch": 0.8266850068775791, + "grad_norm": 0.43224631737646424, + "learning_rate": 7.676495684613432e-05, + "loss": 2.3875, + "step": 1803 + }, + { + "epoch": 0.8271435121503897, + "grad_norm": 0.43879788332293607, + "learning_rate": 7.636998704106252e-05, + "loss": 2.2905, + "step": 1804 + }, + { + "epoch": 0.8276020174232004, + "grad_norm": 0.4564744994895402, + "learning_rate": 7.597595192178702e-05, + "loss": 2.3105, + "step": 1805 + }, + { + "epoch": 0.828060522696011, + "grad_norm": 0.43689637873097525, + "learning_rate": 7.558285235769646e-05, + "loss": 2.3211, + "step": 1806 + }, + { + "epoch": 0.8285190279688216, + "grad_norm": 0.43512255172293623, + "learning_rate": 7.519068921611494e-05, + "loss": 2.3978, + "step": 1807 + }, + { + "epoch": 0.8289775332416323, + "grad_norm": 0.45419300758315945, + "learning_rate": 7.479946336230047e-05, + "loss": 2.3752, + "step": 1808 + }, + { + "epoch": 0.8294360385144429, + "grad_norm": 0.4518502590550213, + "learning_rate": 7.440917565944349e-05, + "loss": 2.4142, + "step": 1809 + }, + { + "epoch": 0.8298945437872536, + "grad_norm": 0.41769865229593256, + "learning_rate": 7.4019826968664e-05, + "loss": 2.3829, + "step": 1810 + }, + { + "epoch": 0.8303530490600642, + "grad_norm": 0.4528154742397934, + "learning_rate": 7.363141814901053e-05, + "loss": 2.3947, + "step": 1811 + }, + { + "epoch": 0.8308115543328748, + "grad_norm": 0.407994341274086, + "learning_rate": 7.32439500574577e-05, + "loss": 2.3662, + "step": 1812 + }, + { + "epoch": 0.8312700596056855, + "grad_norm": 0.41617615662744917, + "learning_rate": 7.285742354890473e-05, + "loss": 2.3292, + "step": 1813 + }, + { + "epoch": 0.8317285648784961, + "grad_norm": 0.44482613708635177, + "learning_rate": 7.247183947617325e-05, + "loss": 2.3408, + "step": 1814 + }, + { + "epoch": 0.8321870701513068, + "grad_norm": 0.4219762150336968, + "learning_rate": 7.20871986900053e-05, + "loss": 2.3107, + "step": 1815 + }, + { + "epoch": 0.8326455754241173, + "grad_norm": 0.4246383113726645, + "learning_rate": 7.170350203906218e-05, + "loss": 2.3322, + "step": 1816 + }, + { + "epoch": 0.833104080696928, + "grad_norm": 0.41475429204015163, + "learning_rate": 7.132075036992158e-05, + "loss": 2.3798, + "step": 1817 + }, + { + "epoch": 0.8335625859697386, + "grad_norm": 0.45124974329688006, + "learning_rate": 7.093894452707666e-05, + "loss": 2.317, + "step": 1818 + }, + { + "epoch": 0.8340210912425493, + "grad_norm": 0.4256849808964594, + "learning_rate": 7.055808535293334e-05, + "loss": 2.3041, + "step": 1819 + }, + { + "epoch": 0.83447959651536, + "grad_norm": 0.42712762906890367, + "learning_rate": 7.017817368780888e-05, + "loss": 2.39, + "step": 1820 + }, + { + "epoch": 0.8349381017881705, + "grad_norm": 0.406027945538829, + "learning_rate": 6.979921036993042e-05, + "loss": 2.3572, + "step": 1821 + }, + { + "epoch": 0.8353966070609812, + "grad_norm": 0.4392952769296799, + "learning_rate": 6.942119623543202e-05, + "loss": 2.4547, + "step": 1822 + }, + { + "epoch": 0.8358551123337918, + "grad_norm": 0.41708843049649746, + "learning_rate": 6.904413211835414e-05, + "loss": 2.3597, + "step": 1823 + }, + { + "epoch": 0.8363136176066025, + "grad_norm": 0.44693227306936506, + "learning_rate": 6.866801885064056e-05, + "loss": 2.3845, + "step": 1824 + }, + { + "epoch": 0.8367721228794132, + "grad_norm": 0.4353252079886047, + "learning_rate": 6.829285726213769e-05, + "loss": 2.3336, + "step": 1825 + }, + { + "epoch": 0.8372306281522237, + "grad_norm": 0.44663988328239573, + "learning_rate": 6.79186481805918e-05, + "loss": 2.3751, + "step": 1826 + }, + { + "epoch": 0.8376891334250344, + "grad_norm": 0.42621647134765317, + "learning_rate": 6.754539243164754e-05, + "loss": 2.3637, + "step": 1827 + }, + { + "epoch": 0.838147638697845, + "grad_norm": 0.4204052006587818, + "learning_rate": 6.717309083884654e-05, + "loss": 2.4032, + "step": 1828 + }, + { + "epoch": 0.8386061439706557, + "grad_norm": 0.4308073783633906, + "learning_rate": 6.680174422362468e-05, + "loss": 2.2715, + "step": 1829 + }, + { + "epoch": 0.8390646492434664, + "grad_norm": 0.4270377125060871, + "learning_rate": 6.643135340531136e-05, + "loss": 2.407, + "step": 1830 + }, + { + "epoch": 0.8395231545162769, + "grad_norm": 0.429600923388392, + "learning_rate": 6.606191920112664e-05, + "loss": 2.4558, + "step": 1831 + }, + { + "epoch": 0.8399816597890876, + "grad_norm": 0.4391054637894445, + "learning_rate": 6.569344242618036e-05, + "loss": 2.3155, + "step": 1832 + }, + { + "epoch": 0.8404401650618982, + "grad_norm": 0.4383310217696881, + "learning_rate": 6.532592389346958e-05, + "loss": 2.4433, + "step": 1833 + }, + { + "epoch": 0.8408986703347089, + "grad_norm": 0.428150631062745, + "learning_rate": 6.495936441387713e-05, + "loss": 2.3415, + "step": 1834 + }, + { + "epoch": 0.8413571756075194, + "grad_norm": 0.43161560691097733, + "learning_rate": 6.459376479617013e-05, + "loss": 2.3722, + "step": 1835 + }, + { + "epoch": 0.8418156808803301, + "grad_norm": 0.4269905396883278, + "learning_rate": 6.422912584699752e-05, + "loss": 2.4139, + "step": 1836 + }, + { + "epoch": 0.8422741861531408, + "grad_norm": 0.42719155699202926, + "learning_rate": 6.386544837088904e-05, + "loss": 2.3021, + "step": 1837 + }, + { + "epoch": 0.8427326914259514, + "grad_norm": 0.4235352321844141, + "learning_rate": 6.350273317025251e-05, + "loss": 2.27, + "step": 1838 + }, + { + "epoch": 0.8431911966987621, + "grad_norm": 0.4388833387988792, + "learning_rate": 6.314098104537324e-05, + "loss": 2.397, + "step": 1839 + }, + { + "epoch": 0.8436497019715726, + "grad_norm": 0.4330089420132666, + "learning_rate": 6.278019279441122e-05, + "loss": 2.4058, + "step": 1840 + }, + { + "epoch": 0.8441082072443833, + "grad_norm": 0.4338056712565015, + "learning_rate": 6.242036921339972e-05, + "loss": 2.3584, + "step": 1841 + }, + { + "epoch": 0.844566712517194, + "grad_norm": 0.4438180231624452, + "learning_rate": 6.206151109624402e-05, + "loss": 2.3616, + "step": 1842 + }, + { + "epoch": 0.8450252177900046, + "grad_norm": 0.43703389004988047, + "learning_rate": 6.170361923471868e-05, + "loss": 2.3012, + "step": 1843 + }, + { + "epoch": 0.8454837230628153, + "grad_norm": 0.4168904712840764, + "learning_rate": 6.134669441846691e-05, + "loss": 2.3426, + "step": 1844 + }, + { + "epoch": 0.8459422283356258, + "grad_norm": 0.4400036638909379, + "learning_rate": 6.099073743499772e-05, + "loss": 2.4094, + "step": 1845 + }, + { + "epoch": 0.8464007336084365, + "grad_norm": 0.4540034082938828, + "learning_rate": 6.063574906968511e-05, + "loss": 2.324, + "step": 1846 + }, + { + "epoch": 0.8468592388812471, + "grad_norm": 0.44579344400800197, + "learning_rate": 6.028173010576582e-05, + "loss": 2.2865, + "step": 1847 + }, + { + "epoch": 0.8473177441540578, + "grad_norm": 0.4416753876245607, + "learning_rate": 5.9928681324337544e-05, + "loss": 2.3978, + "step": 1848 + }, + { + "epoch": 0.8477762494268684, + "grad_norm": 0.44487677464154823, + "learning_rate": 5.957660350435773e-05, + "loss": 2.3238, + "step": 1849 + }, + { + "epoch": 0.848234754699679, + "grad_norm": 0.4403976533980101, + "learning_rate": 5.922549742264122e-05, + "loss": 2.3807, + "step": 1850 + }, + { + "epoch": 0.8486932599724897, + "grad_norm": 0.43387641867145127, + "learning_rate": 5.8875363853859166e-05, + "loss": 2.346, + "step": 1851 + }, + { + "epoch": 0.8491517652453003, + "grad_norm": 0.4165509744894734, + "learning_rate": 5.852620357053651e-05, + "loss": 2.4151, + "step": 1852 + }, + { + "epoch": 0.849610270518111, + "grad_norm": 0.43779421510355193, + "learning_rate": 5.8178017343051336e-05, + "loss": 2.3089, + "step": 1853 + }, + { + "epoch": 0.8500687757909215, + "grad_norm": 0.4125010309558686, + "learning_rate": 5.783080593963219e-05, + "loss": 2.3648, + "step": 1854 + }, + { + "epoch": 0.8505272810637322, + "grad_norm": 0.4440910162140296, + "learning_rate": 5.748457012635683e-05, + "loss": 2.3833, + "step": 1855 + }, + { + "epoch": 0.8509857863365429, + "grad_norm": 0.4160416935198964, + "learning_rate": 5.713931066715078e-05, + "loss": 2.2596, + "step": 1856 + }, + { + "epoch": 0.8514442916093535, + "grad_norm": 0.4070522284810349, + "learning_rate": 5.679502832378497e-05, + "loss": 2.4349, + "step": 1857 + }, + { + "epoch": 0.8519027968821642, + "grad_norm": 0.4220104532654952, + "learning_rate": 5.645172385587482e-05, + "loss": 2.2958, + "step": 1858 + }, + { + "epoch": 0.8523613021549747, + "grad_norm": 0.4135884830847816, + "learning_rate": 5.6109398020877834e-05, + "loss": 2.283, + "step": 1859 + }, + { + "epoch": 0.8528198074277854, + "grad_norm": 0.43206366703872756, + "learning_rate": 5.576805157409265e-05, + "loss": 2.3231, + "step": 1860 + }, + { + "epoch": 0.8532783127005961, + "grad_norm": 0.4325148658431918, + "learning_rate": 5.542768526865677e-05, + "loss": 2.3258, + "step": 1861 + }, + { + "epoch": 0.8537368179734067, + "grad_norm": 0.4457664210144167, + "learning_rate": 5.508829985554509e-05, + "loss": 2.4228, + "step": 1862 + }, + { + "epoch": 0.8541953232462174, + "grad_norm": 0.4387700990286473, + "learning_rate": 5.474989608356856e-05, + "loss": 2.4178, + "step": 1863 + }, + { + "epoch": 0.8546538285190279, + "grad_norm": 0.40871152264298755, + "learning_rate": 5.441247469937194e-05, + "loss": 2.2869, + "step": 1864 + }, + { + "epoch": 0.8551123337918386, + "grad_norm": 0.41243201590435524, + "learning_rate": 5.407603644743286e-05, + "loss": 2.2855, + "step": 1865 + }, + { + "epoch": 0.8555708390646493, + "grad_norm": 0.4186431042357051, + "learning_rate": 5.374058207005944e-05, + "loss": 2.2394, + "step": 1866 + }, + { + "epoch": 0.8560293443374599, + "grad_norm": 0.4471471579633281, + "learning_rate": 5.3406112307389066e-05, + "loss": 2.3544, + "step": 1867 + }, + { + "epoch": 0.8564878496102705, + "grad_norm": 0.4342784110189765, + "learning_rate": 5.3072627897386926e-05, + "loss": 2.378, + "step": 1868 + }, + { + "epoch": 0.8569463548830811, + "grad_norm": 0.4545008089957225, + "learning_rate": 5.27401295758439e-05, + "loss": 2.3386, + "step": 1869 + }, + { + "epoch": 0.8574048601558918, + "grad_norm": 0.4441525625857382, + "learning_rate": 5.2408618076375315e-05, + "loss": 2.4785, + "step": 1870 + }, + { + "epoch": 0.8578633654287025, + "grad_norm": 0.42908540015066055, + "learning_rate": 5.207809413041914e-05, + "loss": 2.4458, + "step": 1871 + }, + { + "epoch": 0.8583218707015131, + "grad_norm": 0.42883167014169915, + "learning_rate": 5.174855846723459e-05, + "loss": 2.3163, + "step": 1872 + }, + { + "epoch": 0.8587803759743237, + "grad_norm": 0.4317781681785938, + "learning_rate": 5.1420011813900104e-05, + "loss": 2.3593, + "step": 1873 + }, + { + "epoch": 0.8592388812471343, + "grad_norm": 0.4412488023603898, + "learning_rate": 5.109245489531211e-05, + "loss": 2.3605, + "step": 1874 + }, + { + "epoch": 0.859697386519945, + "grad_norm": 0.4213103775544417, + "learning_rate": 5.0765888434183446e-05, + "loss": 2.2579, + "step": 1875 + }, + { + "epoch": 0.8601558917927556, + "grad_norm": 0.43094378978221604, + "learning_rate": 5.0440313151041364e-05, + "loss": 2.3813, + "step": 1876 + }, + { + "epoch": 0.8606143970655663, + "grad_norm": 0.43253263024963334, + "learning_rate": 5.011572976422657e-05, + "loss": 2.3092, + "step": 1877 + }, + { + "epoch": 0.8610729023383769, + "grad_norm": 0.42658984983887255, + "learning_rate": 4.9792138989890825e-05, + "loss": 2.3513, + "step": 1878 + }, + { + "epoch": 0.8615314076111875, + "grad_norm": 0.4450312440455457, + "learning_rate": 4.9469541541996234e-05, + "loss": 2.3273, + "step": 1879 + }, + { + "epoch": 0.8619899128839982, + "grad_norm": 0.41967714810621715, + "learning_rate": 4.914793813231305e-05, + "loss": 2.4032, + "step": 1880 + }, + { + "epoch": 0.8624484181568088, + "grad_norm": 0.4369670296307905, + "learning_rate": 4.882732947041818e-05, + "loss": 2.3897, + "step": 1881 + }, + { + "epoch": 0.8629069234296194, + "grad_norm": 0.43462977431175515, + "learning_rate": 4.850771626369416e-05, + "loss": 2.3961, + "step": 1882 + }, + { + "epoch": 0.86336542870243, + "grad_norm": 0.44593749738900745, + "learning_rate": 4.818909921732662e-05, + "loss": 2.3587, + "step": 1883 + }, + { + "epoch": 0.8638239339752407, + "grad_norm": 0.43325184054992855, + "learning_rate": 4.787147903430383e-05, + "loss": 2.3049, + "step": 1884 + }, + { + "epoch": 0.8642824392480514, + "grad_norm": 0.4451261758524495, + "learning_rate": 4.755485641541424e-05, + "loss": 2.3307, + "step": 1885 + }, + { + "epoch": 0.864740944520862, + "grad_norm": 0.4501332521588564, + "learning_rate": 4.723923205924557e-05, + "loss": 2.3325, + "step": 1886 + }, + { + "epoch": 0.8651994497936726, + "grad_norm": 0.4453565668274609, + "learning_rate": 4.6924606662182736e-05, + "loss": 2.4369, + "step": 1887 + }, + { + "epoch": 0.8656579550664832, + "grad_norm": 0.4210872511525881, + "learning_rate": 4.6610980918406596e-05, + "loss": 2.3378, + "step": 1888 + }, + { + "epoch": 0.8661164603392939, + "grad_norm": 0.44488253574951403, + "learning_rate": 4.629835551989276e-05, + "loss": 2.3985, + "step": 1889 + }, + { + "epoch": 0.8665749656121046, + "grad_norm": 0.43272174522678974, + "learning_rate": 4.5986731156409224e-05, + "loss": 2.3556, + "step": 1890 + }, + { + "epoch": 0.8670334708849152, + "grad_norm": 0.4551227142526202, + "learning_rate": 4.567610851551568e-05, + "loss": 2.4905, + "step": 1891 + }, + { + "epoch": 0.8674919761577258, + "grad_norm": 0.4214606209240262, + "learning_rate": 4.536648828256146e-05, + "loss": 2.345, + "step": 1892 + }, + { + "epoch": 0.8679504814305364, + "grad_norm": 0.43120267845316534, + "learning_rate": 4.505787114068433e-05, + "loss": 2.3545, + "step": 1893 + }, + { + "epoch": 0.8684089867033471, + "grad_norm": 0.4332707158659077, + "learning_rate": 4.4750257770808764e-05, + "loss": 2.4098, + "step": 1894 + }, + { + "epoch": 0.8688674919761578, + "grad_norm": 0.43757964943471445, + "learning_rate": 4.444364885164448e-05, + "loss": 2.3348, + "step": 1895 + }, + { + "epoch": 0.8693259972489684, + "grad_norm": 0.439099300253302, + "learning_rate": 4.413804505968533e-05, + "loss": 2.3638, + "step": 1896 + }, + { + "epoch": 0.869784502521779, + "grad_norm": 0.44853846361687955, + "learning_rate": 4.3833447069206944e-05, + "loss": 2.4613, + "step": 1897 + }, + { + "epoch": 0.8702430077945896, + "grad_norm": 0.42790267506838525, + "learning_rate": 4.352985555226635e-05, + "loss": 2.3934, + "step": 1898 + }, + { + "epoch": 0.8707015130674003, + "grad_norm": 0.4300957654313624, + "learning_rate": 4.322727117869951e-05, + "loss": 2.3316, + "step": 1899 + }, + { + "epoch": 0.871160018340211, + "grad_norm": 0.416708861979887, + "learning_rate": 4.29256946161205e-05, + "loss": 2.3132, + "step": 1900 + }, + { + "epoch": 0.8716185236130215, + "grad_norm": 0.43090368138823487, + "learning_rate": 4.262512652991968e-05, + "loss": 2.2806, + "step": 1901 + }, + { + "epoch": 0.8720770288858322, + "grad_norm": 0.43857394806440075, + "learning_rate": 4.2325567583262113e-05, + "loss": 2.3208, + "step": 1902 + }, + { + "epoch": 0.8725355341586428, + "grad_norm": 0.45943480593158126, + "learning_rate": 4.2027018437086895e-05, + "loss": 2.3415, + "step": 1903 + }, + { + "epoch": 0.8729940394314535, + "grad_norm": 0.44419878897113385, + "learning_rate": 4.172947975010449e-05, + "loss": 2.301, + "step": 1904 + }, + { + "epoch": 0.8734525447042641, + "grad_norm": 0.44588842736845324, + "learning_rate": 4.143295217879645e-05, + "loss": 2.4807, + "step": 1905 + }, + { + "epoch": 0.8739110499770747, + "grad_norm": 0.42825442781321177, + "learning_rate": 4.113743637741296e-05, + "loss": 2.3477, + "step": 1906 + }, + { + "epoch": 0.8743695552498854, + "grad_norm": 0.42214810424818716, + "learning_rate": 4.084293299797226e-05, + "loss": 2.3848, + "step": 1907 + }, + { + "epoch": 0.874828060522696, + "grad_norm": 0.449128904635546, + "learning_rate": 4.054944269025862e-05, + "loss": 2.3018, + "step": 1908 + }, + { + "epoch": 0.8752865657955067, + "grad_norm": 0.4413165176910129, + "learning_rate": 4.025696610182095e-05, + "loss": 2.3437, + "step": 1909 + }, + { + "epoch": 0.8757450710683173, + "grad_norm": 0.45552403400150876, + "learning_rate": 3.996550387797187e-05, + "loss": 2.2647, + "step": 1910 + }, + { + "epoch": 0.8762035763411279, + "grad_norm": 0.43633002331576043, + "learning_rate": 3.9675056661785556e-05, + "loss": 2.2413, + "step": 1911 + }, + { + "epoch": 0.8766620816139385, + "grad_norm": 0.4256197739114342, + "learning_rate": 3.9385625094097154e-05, + "loss": 2.3989, + "step": 1912 + }, + { + "epoch": 0.8771205868867492, + "grad_norm": 0.4200085917746492, + "learning_rate": 3.909720981350034e-05, + "loss": 2.2602, + "step": 1913 + }, + { + "epoch": 0.8775790921595599, + "grad_norm": 0.4379076370474578, + "learning_rate": 3.880981145634704e-05, + "loss": 2.4365, + "step": 1914 + }, + { + "epoch": 0.8780375974323704, + "grad_norm": 0.4689962619374092, + "learning_rate": 3.852343065674507e-05, + "loss": 2.3918, + "step": 1915 + }, + { + "epoch": 0.8784961027051811, + "grad_norm": 0.440840639893045, + "learning_rate": 3.8238068046557276e-05, + "loss": 2.3276, + "step": 1916 + }, + { + "epoch": 0.8789546079779917, + "grad_norm": 0.4380184013584428, + "learning_rate": 3.795372425540006e-05, + "loss": 2.297, + "step": 1917 + }, + { + "epoch": 0.8794131132508024, + "grad_norm": 0.42906006620326165, + "learning_rate": 3.76703999106418e-05, + "loss": 2.2946, + "step": 1918 + }, + { + "epoch": 0.8798716185236131, + "grad_norm": 0.4201934129193374, + "learning_rate": 3.7388095637401754e-05, + "loss": 2.2683, + "step": 1919 + }, + { + "epoch": 0.8803301237964236, + "grad_norm": 0.40563510260271907, + "learning_rate": 3.7106812058548376e-05, + "loss": 2.3354, + "step": 1920 + }, + { + "epoch": 0.8807886290692343, + "grad_norm": 0.44463801543256926, + "learning_rate": 3.682654979469807e-05, + "loss": 2.355, + "step": 1921 + }, + { + "epoch": 0.8812471343420449, + "grad_norm": 0.4187403964973783, + "learning_rate": 3.654730946421403e-05, + "loss": 2.4009, + "step": 1922 + }, + { + "epoch": 0.8817056396148556, + "grad_norm": 0.43681093350458905, + "learning_rate": 3.6269091683204466e-05, + "loss": 2.3681, + "step": 1923 + }, + { + "epoch": 0.8821641448876663, + "grad_norm": 0.42916913884597724, + "learning_rate": 3.5991897065521693e-05, + "loss": 2.4, + "step": 1924 + }, + { + "epoch": 0.8826226501604768, + "grad_norm": 0.43810426716314266, + "learning_rate": 3.571572622276026e-05, + "loss": 2.3767, + "step": 1925 + }, + { + "epoch": 0.8830811554332875, + "grad_norm": 0.4689068554163599, + "learning_rate": 3.544057976425619e-05, + "loss": 2.5007, + "step": 1926 + }, + { + "epoch": 0.8835396607060981, + "grad_norm": 0.43661195235749667, + "learning_rate": 3.5166458297085146e-05, + "loss": 2.4232, + "step": 1927 + }, + { + "epoch": 0.8839981659789088, + "grad_norm": 0.42972453361465207, + "learning_rate": 3.489336242606111e-05, + "loss": 2.368, + "step": 1928 + }, + { + "epoch": 0.8844566712517193, + "grad_norm": 0.4187210267366841, + "learning_rate": 3.462129275373577e-05, + "loss": 2.332, + "step": 1929 + }, + { + "epoch": 0.88491517652453, + "grad_norm": 0.40794994548552616, + "learning_rate": 3.4350249880395924e-05, + "loss": 2.4004, + "step": 1930 + }, + { + "epoch": 0.8853736817973407, + "grad_norm": 0.4530873631136242, + "learning_rate": 3.408023440406355e-05, + "loss": 2.3299, + "step": 1931 + }, + { + "epoch": 0.8858321870701513, + "grad_norm": 0.4345306532422162, + "learning_rate": 3.381124692049331e-05, + "loss": 2.4243, + "step": 1932 + }, + { + "epoch": 0.886290692342962, + "grad_norm": 0.4154055000156589, + "learning_rate": 3.354328802317197e-05, + "loss": 2.3283, + "step": 1933 + }, + { + "epoch": 0.8867491976157725, + "grad_norm": 0.4425186270848397, + "learning_rate": 3.327635830331677e-05, + "loss": 2.2709, + "step": 1934 + }, + { + "epoch": 0.8872077028885832, + "grad_norm": 0.42619487143040014, + "learning_rate": 3.3010458349874206e-05, + "loss": 2.3329, + "step": 1935 + }, + { + "epoch": 0.8876662081613939, + "grad_norm": 0.423402905442138, + "learning_rate": 3.2745588749518775e-05, + "loss": 2.4183, + "step": 1936 + }, + { + "epoch": 0.8881247134342045, + "grad_norm": 0.43860808557929853, + "learning_rate": 3.248175008665161e-05, + "loss": 2.3122, + "step": 1937 + }, + { + "epoch": 0.8885832187070152, + "grad_norm": 0.4296063442696931, + "learning_rate": 3.221894294339911e-05, + "loss": 2.3639, + "step": 1938 + }, + { + "epoch": 0.8890417239798257, + "grad_norm": 0.42348638476487177, + "learning_rate": 3.1957167899611836e-05, + "loss": 2.3154, + "step": 1939 + }, + { + "epoch": 0.8895002292526364, + "grad_norm": 0.4331981933029681, + "learning_rate": 3.169642553286334e-05, + "loss": 2.4014, + "step": 1940 + }, + { + "epoch": 0.889958734525447, + "grad_norm": 0.43321811064378574, + "learning_rate": 3.143671641844831e-05, + "loss": 2.3661, + "step": 1941 + }, + { + "epoch": 0.8904172397982577, + "grad_norm": 0.44428520463207544, + "learning_rate": 3.117804112938205e-05, + "loss": 2.2765, + "step": 1942 + }, + { + "epoch": 0.8908757450710684, + "grad_norm": 0.43545046410700133, + "learning_rate": 3.092040023639869e-05, + "loss": 2.4305, + "step": 1943 + }, + { + "epoch": 0.8913342503438789, + "grad_norm": 0.42896983553339757, + "learning_rate": 3.066379430795002e-05, + "loss": 2.3787, + "step": 1944 + }, + { + "epoch": 0.8917927556166896, + "grad_norm": 0.4350330962631031, + "learning_rate": 3.040822391020459e-05, + "loss": 2.3742, + "step": 1945 + }, + { + "epoch": 0.8922512608895002, + "grad_norm": 0.4232594205765679, + "learning_rate": 3.0153689607045842e-05, + "loss": 2.3088, + "step": 1946 + }, + { + "epoch": 0.8927097661623109, + "grad_norm": 0.4321377273779936, + "learning_rate": 2.9900191960071545e-05, + "loss": 2.3187, + "step": 1947 + }, + { + "epoch": 0.8931682714351215, + "grad_norm": 0.44637211247563324, + "learning_rate": 2.9647731528591848e-05, + "loss": 2.3887, + "step": 1948 + }, + { + "epoch": 0.8936267767079321, + "grad_norm": 0.43426297094109295, + "learning_rate": 2.9396308869628795e-05, + "loss": 2.3632, + "step": 1949 + }, + { + "epoch": 0.8940852819807428, + "grad_norm": 0.4285856495092431, + "learning_rate": 2.914592453791448e-05, + "loss": 2.3427, + "step": 1950 + }, + { + "epoch": 0.8945437872535534, + "grad_norm": 0.43781173445339755, + "learning_rate": 2.8896579085889994e-05, + "loss": 2.3929, + "step": 1951 + }, + { + "epoch": 0.8950022925263641, + "grad_norm": 0.4205281880299689, + "learning_rate": 2.86482730637046e-05, + "loss": 2.3424, + "step": 1952 + }, + { + "epoch": 0.8954607977991746, + "grad_norm": 0.43683889344161975, + "learning_rate": 2.840100701921383e-05, + "loss": 2.4064, + "step": 1953 + }, + { + "epoch": 0.8959193030719853, + "grad_norm": 0.4270705945737499, + "learning_rate": 2.8154781497978898e-05, + "loss": 2.3871, + "step": 1954 + }, + { + "epoch": 0.896377808344796, + "grad_norm": 0.4530746447653266, + "learning_rate": 2.7909597043265013e-05, + "loss": 2.3511, + "step": 1955 + }, + { + "epoch": 0.8968363136176066, + "grad_norm": 0.4471315064136138, + "learning_rate": 2.7665454196040662e-05, + "loss": 2.376, + "step": 1956 + }, + { + "epoch": 0.8972948188904173, + "grad_norm": 0.4329654562442658, + "learning_rate": 2.7422353494975905e-05, + "loss": 2.3623, + "step": 1957 + }, + { + "epoch": 0.8977533241632278, + "grad_norm": 0.42942190428439886, + "learning_rate": 2.7180295476441573e-05, + "loss": 2.3515, + "step": 1958 + }, + { + "epoch": 0.8982118294360385, + "grad_norm": 0.43506407480073184, + "learning_rate": 2.6939280674508016e-05, + "loss": 2.3452, + "step": 1959 + }, + { + "epoch": 0.8986703347088492, + "grad_norm": 0.412377642766428, + "learning_rate": 2.669930962094358e-05, + "loss": 2.364, + "step": 1960 + }, + { + "epoch": 0.8991288399816598, + "grad_norm": 0.42991502695623246, + "learning_rate": 2.6460382845214126e-05, + "loss": 2.434, + "step": 1961 + }, + { + "epoch": 0.8995873452544704, + "grad_norm": 0.43120442237380474, + "learning_rate": 2.6222500874481025e-05, + "loss": 2.4068, + "step": 1962 + }, + { + "epoch": 0.900045850527281, + "grad_norm": 0.4328760268582132, + "learning_rate": 2.5985664233600827e-05, + "loss": 2.3838, + "step": 1963 + }, + { + "epoch": 0.9005043558000917, + "grad_norm": 0.44294514202626356, + "learning_rate": 2.574987344512336e-05, + "loss": 2.3455, + "step": 1964 + }, + { + "epoch": 0.9009628610729024, + "grad_norm": 0.43501461560830595, + "learning_rate": 2.5515129029290984e-05, + "loss": 2.3743, + "step": 1965 + }, + { + "epoch": 0.901421366345713, + "grad_norm": 0.4445430778449216, + "learning_rate": 2.5281431504037556e-05, + "loss": 2.3154, + "step": 1966 + }, + { + "epoch": 0.9018798716185236, + "grad_norm": 0.4317608107433222, + "learning_rate": 2.504878138498684e-05, + "loss": 2.2967, + "step": 1967 + }, + { + "epoch": 0.9023383768913342, + "grad_norm": 0.43411762442412266, + "learning_rate": 2.48171791854519e-05, + "loss": 2.4152, + "step": 1968 + }, + { + "epoch": 0.9027968821641449, + "grad_norm": 0.43279671872628955, + "learning_rate": 2.4586625416433473e-05, + "loss": 2.3608, + "step": 1969 + }, + { + "epoch": 0.9032553874369555, + "grad_norm": 0.42784817819594234, + "learning_rate": 2.435712058661921e-05, + "loss": 2.3662, + "step": 1970 + }, + { + "epoch": 0.9037138927097662, + "grad_norm": 0.4485985608222439, + "learning_rate": 2.4128665202382327e-05, + "loss": 2.3131, + "step": 1971 + }, + { + "epoch": 0.9041723979825768, + "grad_norm": 0.4259378247594785, + "learning_rate": 2.3901259767780515e-05, + "loss": 2.3231, + "step": 1972 + }, + { + "epoch": 0.9046309032553874, + "grad_norm": 0.4451521084784613, + "learning_rate": 2.367490478455514e-05, + "loss": 2.4501, + "step": 1973 + }, + { + "epoch": 0.9050894085281981, + "grad_norm": 0.42958204251377696, + "learning_rate": 2.3449600752129597e-05, + "loss": 2.3617, + "step": 1974 + }, + { + "epoch": 0.9055479138010087, + "grad_norm": 0.43573088484869427, + "learning_rate": 2.3225348167608685e-05, + "loss": 2.3191, + "step": 1975 + }, + { + "epoch": 0.9060064190738194, + "grad_norm": 0.43184659540144554, + "learning_rate": 2.3002147525777118e-05, + "loss": 2.444, + "step": 1976 + }, + { + "epoch": 0.90646492434663, + "grad_norm": 0.44502566680022854, + "learning_rate": 2.2779999319098856e-05, + "loss": 2.3327, + "step": 1977 + }, + { + "epoch": 0.9069234296194406, + "grad_norm": 0.43202769918150963, + "learning_rate": 2.255890403771571e-05, + "loss": 2.3371, + "step": 1978 + }, + { + "epoch": 0.9073819348922513, + "grad_norm": 0.4204549477582461, + "learning_rate": 2.233886216944614e-05, + "loss": 2.2657, + "step": 1979 + }, + { + "epoch": 0.9078404401650619, + "grad_norm": 0.43201487054868987, + "learning_rate": 2.211987419978484e-05, + "loss": 2.4265, + "step": 1980 + }, + { + "epoch": 0.9082989454378725, + "grad_norm": 0.4314808065530456, + "learning_rate": 2.1901940611900705e-05, + "loss": 2.3824, + "step": 1981 + }, + { + "epoch": 0.9087574507106831, + "grad_norm": 0.4299843770425036, + "learning_rate": 2.168506188663666e-05, + "loss": 2.4089, + "step": 1982 + }, + { + "epoch": 0.9092159559834938, + "grad_norm": 0.4237362239013383, + "learning_rate": 2.1469238502507925e-05, + "loss": 2.3662, + "step": 1983 + }, + { + "epoch": 0.9096744612563045, + "grad_norm": 0.4193672112657111, + "learning_rate": 2.125447093570154e-05, + "loss": 2.3052, + "step": 1984 + }, + { + "epoch": 0.9101329665291151, + "grad_norm": 0.44229593631812253, + "learning_rate": 2.1040759660074793e-05, + "loss": 2.3614, + "step": 1985 + }, + { + "epoch": 0.9105914718019257, + "grad_norm": 0.4475843487859251, + "learning_rate": 2.0828105147154273e-05, + "loss": 2.3548, + "step": 1986 + }, + { + "epoch": 0.9110499770747363, + "grad_norm": 0.44284849755872885, + "learning_rate": 2.061650786613545e-05, + "loss": 2.3767, + "step": 1987 + }, + { + "epoch": 0.911508482347547, + "grad_norm": 0.4666048372160949, + "learning_rate": 2.040596828388058e-05, + "loss": 2.3804, + "step": 1988 + }, + { + "epoch": 0.9119669876203577, + "grad_norm": 0.459074547716179, + "learning_rate": 2.019648686491865e-05, + "loss": 2.3992, + "step": 1989 + }, + { + "epoch": 0.9124254928931683, + "grad_norm": 0.44692283551803885, + "learning_rate": 1.9988064071443767e-05, + "loss": 2.289, + "step": 1990 + }, + { + "epoch": 0.9128839981659789, + "grad_norm": 0.43274887555818087, + "learning_rate": 1.9780700363314253e-05, + "loss": 2.3667, + "step": 1991 + }, + { + "epoch": 0.9133425034387895, + "grad_norm": 0.43225812232772176, + "learning_rate": 1.957439619805196e-05, + "loss": 2.4018, + "step": 1992 + }, + { + "epoch": 0.9138010087116002, + "grad_norm": 0.4531663005398075, + "learning_rate": 1.9369152030840554e-05, + "loss": 2.2692, + "step": 1993 + }, + { + "epoch": 0.9142595139844109, + "grad_norm": 0.4436965591822272, + "learning_rate": 1.916496831452552e-05, + "loss": 2.4088, + "step": 1994 + }, + { + "epoch": 0.9147180192572214, + "grad_norm": 0.45273529602710816, + "learning_rate": 1.8961845499611998e-05, + "loss": 2.2433, + "step": 1995 + }, + { + "epoch": 0.9151765245300321, + "grad_norm": 0.45971555704607026, + "learning_rate": 1.8759784034264925e-05, + "loss": 2.4139, + "step": 1996 + }, + { + "epoch": 0.9156350298028427, + "grad_norm": 0.41697047325567754, + "learning_rate": 1.855878436430708e-05, + "loss": 2.4522, + "step": 1997 + }, + { + "epoch": 0.9160935350756534, + "grad_norm": 0.4238152719012225, + "learning_rate": 1.835884693321871e-05, + "loss": 2.3284, + "step": 1998 + }, + { + "epoch": 0.916552040348464, + "grad_norm": 0.4394454654714852, + "learning_rate": 1.8159972182136386e-05, + "loss": 2.4554, + "step": 1999 + }, + { + "epoch": 0.9170105456212746, + "grad_norm": 0.43424813310330335, + "learning_rate": 1.7962160549851945e-05, + "loss": 2.3424, + "step": 2000 + }, + { + "epoch": 0.9174690508940853, + "grad_norm": 0.4337611784006883, + "learning_rate": 1.776541247281177e-05, + "loss": 2.4411, + "step": 2001 + }, + { + "epoch": 0.9179275561668959, + "grad_norm": 0.42900752754862603, + "learning_rate": 1.7569728385115224e-05, + "loss": 2.2886, + "step": 2002 + }, + { + "epoch": 0.9183860614397066, + "grad_norm": 0.439309864137302, + "learning_rate": 1.7375108718514665e-05, + "loss": 2.3085, + "step": 2003 + }, + { + "epoch": 0.9188445667125172, + "grad_norm": 0.43782777234683834, + "learning_rate": 1.7181553902413438e-05, + "loss": 2.3801, + "step": 2004 + }, + { + "epoch": 0.9193030719853278, + "grad_norm": 0.42696223638616215, + "learning_rate": 1.698906436386577e-05, + "loss": 2.3407, + "step": 2005 + }, + { + "epoch": 0.9197615772581385, + "grad_norm": 0.436063578796373, + "learning_rate": 1.679764052757532e-05, + "loss": 2.4369, + "step": 2006 + }, + { + "epoch": 0.9202200825309491, + "grad_norm": 0.43402182960242464, + "learning_rate": 1.6607282815894464e-05, + "loss": 2.3224, + "step": 2007 + }, + { + "epoch": 0.9206785878037598, + "grad_norm": 0.44209901770364884, + "learning_rate": 1.6417991648823405e-05, + "loss": 2.3225, + "step": 2008 + }, + { + "epoch": 0.9211370930765704, + "grad_norm": 0.4308418828054004, + "learning_rate": 1.6229767444008835e-05, + "loss": 2.3312, + "step": 2009 + }, + { + "epoch": 0.921595598349381, + "grad_norm": 0.4452485230200299, + "learning_rate": 1.604261061674378e-05, + "loss": 2.4437, + "step": 2010 + }, + { + "epoch": 0.9220541036221916, + "grad_norm": 0.4372091067328577, + "learning_rate": 1.5856521579965865e-05, + "loss": 2.2993, + "step": 2011 + }, + { + "epoch": 0.9225126088950023, + "grad_norm": 0.4471430345012066, + "learning_rate": 1.5671500744256938e-05, + "loss": 2.3875, + "step": 2012 + }, + { + "epoch": 0.922971114167813, + "grad_norm": 0.44526969840958514, + "learning_rate": 1.5487548517841953e-05, + "loss": 2.3121, + "step": 2013 + }, + { + "epoch": 0.9234296194406235, + "grad_norm": 0.4540665584834564, + "learning_rate": 1.530466530658814e-05, + "loss": 2.4197, + "step": 2014 + }, + { + "epoch": 0.9238881247134342, + "grad_norm": 0.4605811578303478, + "learning_rate": 1.5122851514004054e-05, + "loss": 2.4746, + "step": 2015 + }, + { + "epoch": 0.9243466299862448, + "grad_norm": 0.47353808669978964, + "learning_rate": 1.4942107541238703e-05, + "loss": 2.4049, + "step": 2016 + }, + { + "epoch": 0.9248051352590555, + "grad_norm": 0.4353102310570643, + "learning_rate": 1.4762433787080809e-05, + "loss": 2.2906, + "step": 2017 + }, + { + "epoch": 0.9252636405318662, + "grad_norm": 0.4516903640336729, + "learning_rate": 1.4583830647957541e-05, + "loss": 2.4356, + "step": 2018 + }, + { + "epoch": 0.9257221458046767, + "grad_norm": 0.4271335179651968, + "learning_rate": 1.4406298517934068e-05, + "loss": 2.2338, + "step": 2019 + }, + { + "epoch": 0.9261806510774874, + "grad_norm": 0.4316500050779121, + "learning_rate": 1.4229837788712562e-05, + "loss": 2.3077, + "step": 2020 + }, + { + "epoch": 0.926639156350298, + "grad_norm": 0.4251366541814639, + "learning_rate": 1.4054448849631085e-05, + "loss": 2.4227, + "step": 2021 + }, + { + "epoch": 0.9270976616231087, + "grad_norm": 0.4480625794729801, + "learning_rate": 1.3880132087663145e-05, + "loss": 2.3197, + "step": 2022 + }, + { + "epoch": 0.9275561668959194, + "grad_norm": 0.41651763172274797, + "learning_rate": 1.3706887887416419e-05, + "loss": 2.3365, + "step": 2023 + }, + { + "epoch": 0.9280146721687299, + "grad_norm": 0.43327919428229716, + "learning_rate": 1.3534716631132316e-05, + "loss": 2.286, + "step": 2024 + }, + { + "epoch": 0.9284731774415406, + "grad_norm": 0.43561598070238805, + "learning_rate": 1.3363618698684853e-05, + "loss": 2.315, + "step": 2025 + }, + { + "epoch": 0.9289316827143512, + "grad_norm": 0.41288794894578285, + "learning_rate": 1.3193594467579728e-05, + "loss": 2.3391, + "step": 2026 + }, + { + "epoch": 0.9293901879871619, + "grad_norm": 0.4270544232540947, + "learning_rate": 1.3024644312954026e-05, + "loss": 2.3345, + "step": 2027 + }, + { + "epoch": 0.9298486932599724, + "grad_norm": 0.44716178009189433, + "learning_rate": 1.2856768607574564e-05, + "loss": 2.3335, + "step": 2028 + }, + { + "epoch": 0.9303071985327831, + "grad_norm": 0.4227603904641448, + "learning_rate": 1.2689967721837947e-05, + "loss": 2.4125, + "step": 2029 + }, + { + "epoch": 0.9307657038055938, + "grad_norm": 0.4561266454335451, + "learning_rate": 1.2524242023769006e-05, + "loss": 2.434, + "step": 2030 + }, + { + "epoch": 0.9312242090784044, + "grad_norm": 0.4518770699668797, + "learning_rate": 1.2359591879020526e-05, + "loss": 2.4076, + "step": 2031 + }, + { + "epoch": 0.9316827143512151, + "grad_norm": 0.44273905076311343, + "learning_rate": 1.2196017650872081e-05, + "loss": 2.3679, + "step": 2032 + }, + { + "epoch": 0.9321412196240256, + "grad_norm": 0.4386402900516462, + "learning_rate": 1.2033519700229367e-05, + "loss": 2.274, + "step": 2033 + }, + { + "epoch": 0.9325997248968363, + "grad_norm": 0.45475897270272475, + "learning_rate": 1.1872098385623586e-05, + "loss": 2.431, + "step": 2034 + }, + { + "epoch": 0.933058230169647, + "grad_norm": 0.4241813435382978, + "learning_rate": 1.1711754063210289e-05, + "loss": 2.391, + "step": 2035 + }, + { + "epoch": 0.9335167354424576, + "grad_norm": 0.4378311915839905, + "learning_rate": 1.155248708676887e-05, + "loss": 2.3014, + "step": 2036 + }, + { + "epoch": 0.9339752407152683, + "grad_norm": 0.4350937432829679, + "learning_rate": 1.1394297807701737e-05, + "loss": 2.3245, + "step": 2037 + }, + { + "epoch": 0.9344337459880788, + "grad_norm": 0.42916052809905625, + "learning_rate": 1.1237186575033254e-05, + "loss": 2.4081, + "step": 2038 + }, + { + "epoch": 0.9348922512608895, + "grad_norm": 0.4467429029239166, + "learning_rate": 1.1081153735409522e-05, + "loss": 2.3334, + "step": 2039 + }, + { + "epoch": 0.9353507565337001, + "grad_norm": 0.4371451208887689, + "learning_rate": 1.0926199633097156e-05, + "loss": 2.4034, + "step": 2040 + }, + { + "epoch": 0.9358092618065108, + "grad_norm": 0.439163621141678, + "learning_rate": 1.0772324609982787e-05, + "loss": 2.3622, + "step": 2041 + }, + { + "epoch": 0.9362677670793215, + "grad_norm": 0.4309641381666847, + "learning_rate": 1.0619529005571893e-05, + "loss": 2.4079, + "step": 2042 + }, + { + "epoch": 0.936726272352132, + "grad_norm": 0.46118809681357137, + "learning_rate": 1.0467813156988748e-05, + "loss": 2.4319, + "step": 2043 + }, + { + "epoch": 0.9371847776249427, + "grad_norm": 0.41391554369263356, + "learning_rate": 1.0317177398975031e-05, + "loss": 2.3482, + "step": 2044 + }, + { + "epoch": 0.9376432828977533, + "grad_norm": 0.4188366616516629, + "learning_rate": 1.0167622063889326e-05, + "loss": 2.2686, + "step": 2045 + }, + { + "epoch": 0.938101788170564, + "grad_norm": 0.4109309584684844, + "learning_rate": 1.0019147481706625e-05, + "loss": 2.333, + "step": 2046 + }, + { + "epoch": 0.9385602934433745, + "grad_norm": 0.4524927818318839, + "learning_rate": 9.871753980017051e-06, + "loss": 2.3465, + "step": 2047 + }, + { + "epoch": 0.9390187987161852, + "grad_norm": 0.4390662283607588, + "learning_rate": 9.725441884025855e-06, + "loss": 2.4183, + "step": 2048 + }, + { + "epoch": 0.9394773039889959, + "grad_norm": 0.4321465009135439, + "learning_rate": 9.580211516551862e-06, + "loss": 2.3317, + "step": 2049 + }, + { + "epoch": 0.9399358092618065, + "grad_norm": 0.42684108120355346, + "learning_rate": 9.436063198027589e-06, + "loss": 2.3664, + "step": 2050 + }, + { + "epoch": 0.9403943145346172, + "grad_norm": 0.43631727542169224, + "learning_rate": 9.292997246497959e-06, + "loss": 2.3789, + "step": 2051 + }, + { + "epoch": 0.9408528198074277, + "grad_norm": 0.4431970374995743, + "learning_rate": 9.151013977619693e-06, + "loss": 2.2379, + "step": 2052 + }, + { + "epoch": 0.9413113250802384, + "grad_norm": 0.41852946847952727, + "learning_rate": 9.010113704661038e-06, + "loss": 2.2999, + "step": 2053 + }, + { + "epoch": 0.9417698303530491, + "grad_norm": 0.43381158972269784, + "learning_rate": 8.870296738500316e-06, + "loss": 2.3364, + "step": 2054 + }, + { + "epoch": 0.9422283356258597, + "grad_norm": 0.42099044042095185, + "learning_rate": 8.731563387626095e-06, + "loss": 2.3889, + "step": 2055 + }, + { + "epoch": 0.9426868408986704, + "grad_norm": 0.43101430024459786, + "learning_rate": 8.59391395813569e-06, + "loss": 2.3407, + "step": 2056 + }, + { + "epoch": 0.9431453461714809, + "grad_norm": 0.4274126259634198, + "learning_rate": 8.457348753735328e-06, + "loss": 2.3149, + "step": 2057 + }, + { + "epoch": 0.9436038514442916, + "grad_norm": 0.43044263635896685, + "learning_rate": 8.321868075738593e-06, + "loss": 2.3953, + "step": 2058 + }, + { + "epoch": 0.9440623567171023, + "grad_norm": 0.43315636355681414, + "learning_rate": 8.187472223066371e-06, + "loss": 2.3636, + "step": 2059 + }, + { + "epoch": 0.9445208619899129, + "grad_norm": 0.43533644636595353, + "learning_rate": 8.054161492246136e-06, + "loss": 2.3351, + "step": 2060 + }, + { + "epoch": 0.9449793672627235, + "grad_norm": 0.4517669785963701, + "learning_rate": 7.921936177411049e-06, + "loss": 2.4151, + "step": 2061 + }, + { + "epoch": 0.9454378725355341, + "grad_norm": 0.4479475124088369, + "learning_rate": 7.790796570299463e-06, + "loss": 2.355, + "step": 2062 + }, + { + "epoch": 0.9458963778083448, + "grad_norm": 0.4344929489645047, + "learning_rate": 7.660742960254207e-06, + "loss": 2.3879, + "step": 2063 + }, + { + "epoch": 0.9463548830811555, + "grad_norm": 0.45300231073676966, + "learning_rate": 7.531775634222138e-06, + "loss": 2.3224, + "step": 2064 + }, + { + "epoch": 0.9468133883539661, + "grad_norm": 0.4427222712364483, + "learning_rate": 7.403894876753192e-06, + "loss": 2.3044, + "step": 2065 + }, + { + "epoch": 0.9472718936267767, + "grad_norm": 0.4290283690897393, + "learning_rate": 7.277100970000061e-06, + "loss": 2.3116, + "step": 2066 + }, + { + "epoch": 0.9477303988995873, + "grad_norm": 0.43385060645398543, + "learning_rate": 7.151394193717408e-06, + "loss": 2.3508, + "step": 2067 + }, + { + "epoch": 0.948188904172398, + "grad_norm": 0.44097174932195016, + "learning_rate": 7.026774825261151e-06, + "loss": 2.3838, + "step": 2068 + }, + { + "epoch": 0.9486474094452086, + "grad_norm": 0.41046143485710956, + "learning_rate": 6.903243139588233e-06, + "loss": 2.3008, + "step": 2069 + }, + { + "epoch": 0.9491059147180193, + "grad_norm": 0.42977579074004446, + "learning_rate": 6.780799409255522e-06, + "loss": 2.3937, + "step": 2070 + }, + { + "epoch": 0.9495644199908299, + "grad_norm": 0.440919472923121, + "learning_rate": 6.659443904419637e-06, + "loss": 2.4903, + "step": 2071 + }, + { + "epoch": 0.9500229252636405, + "grad_norm": 0.4591908556063961, + "learning_rate": 6.539176892836008e-06, + "loss": 2.2895, + "step": 2072 + }, + { + "epoch": 0.9504814305364512, + "grad_norm": 0.42312383448151114, + "learning_rate": 6.4199986398585375e-06, + "loss": 2.4478, + "step": 2073 + }, + { + "epoch": 0.9509399358092618, + "grad_norm": 0.4419208534578447, + "learning_rate": 6.3019094084388884e-06, + "loss": 2.4146, + "step": 2074 + }, + { + "epoch": 0.9513984410820725, + "grad_norm": 0.43971052949376543, + "learning_rate": 6.18490945912592e-06, + "loss": 2.3058, + "step": 2075 + }, + { + "epoch": 0.951856946354883, + "grad_norm": 0.41861512117329025, + "learning_rate": 6.068999050065249e-06, + "loss": 2.3233, + "step": 2076 + }, + { + "epoch": 0.9523154516276937, + "grad_norm": 0.4326261782895952, + "learning_rate": 5.9541784369983586e-06, + "loss": 2.2734, + "step": 2077 + }, + { + "epoch": 0.9527739569005044, + "grad_norm": 0.46051241758532185, + "learning_rate": 5.840447873262433e-06, + "loss": 2.3594, + "step": 2078 + }, + { + "epoch": 0.953232462173315, + "grad_norm": 0.4217245424988234, + "learning_rate": 5.727807609789471e-06, + "loss": 2.3312, + "step": 2079 + }, + { + "epoch": 0.9536909674461256, + "grad_norm": 0.44054699443589573, + "learning_rate": 5.616257895105892e-06, + "loss": 2.3754, + "step": 2080 + }, + { + "epoch": 0.9541494727189362, + "grad_norm": 0.4211044411737424, + "learning_rate": 5.505798975331933e-06, + "loss": 2.4129, + "step": 2081 + }, + { + "epoch": 0.9546079779917469, + "grad_norm": 0.45730774261442714, + "learning_rate": 5.396431094181198e-06, + "loss": 2.4618, + "step": 2082 + }, + { + "epoch": 0.9550664832645576, + "grad_norm": 0.4267230016948954, + "learning_rate": 5.288154492960107e-06, + "loss": 2.3544, + "step": 2083 + }, + { + "epoch": 0.9555249885373682, + "grad_norm": 0.44108955330778704, + "learning_rate": 5.1809694105671155e-06, + "loss": 2.246, + "step": 2084 + }, + { + "epoch": 0.9559834938101788, + "grad_norm": 0.4264535636394948, + "learning_rate": 5.074876083492441e-06, + "loss": 2.3016, + "step": 2085 + }, + { + "epoch": 0.9564419990829894, + "grad_norm": 0.45212097630415593, + "learning_rate": 4.96987474581767e-06, + "loss": 2.4274, + "step": 2086 + }, + { + "epoch": 0.9569005043558001, + "grad_norm": 0.45250180825394437, + "learning_rate": 4.865965629214819e-06, + "loss": 2.3184, + "step": 2087 + }, + { + "epoch": 0.9573590096286108, + "grad_norm": 0.46296639365563125, + "learning_rate": 4.763148962946218e-06, + "loss": 2.3959, + "step": 2088 + }, + { + "epoch": 0.9578175149014214, + "grad_norm": 0.43656143187407237, + "learning_rate": 4.661424973863681e-06, + "loss": 2.4416, + "step": 2089 + }, + { + "epoch": 0.958276020174232, + "grad_norm": 0.4384813288544298, + "learning_rate": 4.560793886408398e-06, + "loss": 2.3262, + "step": 2090 + }, + { + "epoch": 0.9587345254470426, + "grad_norm": 0.48276121303456515, + "learning_rate": 4.461255922609986e-06, + "loss": 2.4144, + "step": 2091 + }, + { + "epoch": 0.9591930307198533, + "grad_norm": 0.43079821409416197, + "learning_rate": 4.362811302086267e-06, + "loss": 2.2733, + "step": 2092 + }, + { + "epoch": 0.959651535992664, + "grad_norm": 0.44467208927207624, + "learning_rate": 4.265460242042885e-06, + "loss": 2.3683, + "step": 2093 + }, + { + "epoch": 0.9601100412654745, + "grad_norm": 0.43377399389960997, + "learning_rate": 4.169202957272522e-06, + "loss": 2.3302, + "step": 2094 + }, + { + "epoch": 0.9605685465382852, + "grad_norm": 0.43134543211792986, + "learning_rate": 4.074039660154738e-06, + "loss": 2.2901, + "step": 2095 + }, + { + "epoch": 0.9610270518110958, + "grad_norm": 0.4283859073327877, + "learning_rate": 3.9799705606551325e-06, + "loss": 2.3924, + "step": 2096 + }, + { + "epoch": 0.9614855570839065, + "grad_norm": 0.4360714464261384, + "learning_rate": 3.886995866325294e-06, + "loss": 2.3621, + "step": 2097 + }, + { + "epoch": 0.9619440623567171, + "grad_norm": 0.4330210833339772, + "learning_rate": 3.795115782302072e-06, + "loss": 2.35, + "step": 2098 + }, + { + "epoch": 0.9624025676295277, + "grad_norm": 0.42642430467563547, + "learning_rate": 3.704330511307197e-06, + "loss": 2.3243, + "step": 2099 + }, + { + "epoch": 0.9628610729023384, + "grad_norm": 0.4246215790154623, + "learning_rate": 3.614640253646828e-06, + "loss": 2.3274, + "step": 2100 + }, + { + "epoch": 0.963319578175149, + "grad_norm": 0.44108292724559045, + "learning_rate": 3.5260452072110594e-06, + "loss": 2.4095, + "step": 2101 + }, + { + "epoch": 0.9637780834479597, + "grad_norm": 0.43807587696548445, + "learning_rate": 3.4385455674737498e-06, + "loss": 2.3708, + "step": 2102 + }, + { + "epoch": 0.9642365887207703, + "grad_norm": 0.4304839294614384, + "learning_rate": 3.3521415274915256e-06, + "loss": 2.3275, + "step": 2103 + }, + { + "epoch": 0.9646950939935809, + "grad_norm": 0.43901436625033174, + "learning_rate": 3.2668332779041133e-06, + "loss": 2.338, + "step": 2104 + }, + { + "epoch": 0.9651535992663915, + "grad_norm": 0.43142918709426, + "learning_rate": 3.1826210069332838e-06, + "loss": 2.3865, + "step": 2105 + }, + { + "epoch": 0.9656121045392022, + "grad_norm": 0.4402855664530684, + "learning_rate": 3.0995049003826324e-06, + "loss": 2.4497, + "step": 2106 + }, + { + "epoch": 0.9660706098120129, + "grad_norm": 0.4301563438782796, + "learning_rate": 3.017485141637355e-06, + "loss": 2.2642, + "step": 2107 + }, + { + "epoch": 0.9665291150848234, + "grad_norm": 0.4361375663622519, + "learning_rate": 2.9365619116636376e-06, + "loss": 2.303, + "step": 2108 + }, + { + "epoch": 0.9669876203576341, + "grad_norm": 0.4271898727211135, + "learning_rate": 2.856735389008269e-06, + "loss": 2.3432, + "step": 2109 + }, + { + "epoch": 0.9674461256304447, + "grad_norm": 0.41880893488267307, + "learning_rate": 2.778005749798307e-06, + "loss": 2.3136, + "step": 2110 + }, + { + "epoch": 0.9679046309032554, + "grad_norm": 0.4285986662081031, + "learning_rate": 2.700373167740744e-06, + "loss": 2.3748, + "step": 2111 + }, + { + "epoch": 0.9683631361760661, + "grad_norm": 0.4300849573216766, + "learning_rate": 2.62383781412201e-06, + "loss": 2.3251, + "step": 2112 + }, + { + "epoch": 0.9688216414488766, + "grad_norm": 0.44257368763027394, + "learning_rate": 2.5483998578076373e-06, + "loss": 2.3872, + "step": 2113 + }, + { + "epoch": 0.9692801467216873, + "grad_norm": 0.4467863919691622, + "learning_rate": 2.4740594652418736e-06, + "loss": 2.4003, + "step": 2114 + }, + { + "epoch": 0.9697386519944979, + "grad_norm": 0.4330935992664805, + "learning_rate": 2.4008168004472917e-06, + "loss": 2.3877, + "step": 2115 + }, + { + "epoch": 0.9701971572673086, + "grad_norm": 0.4176897202151024, + "learning_rate": 2.3286720250246253e-06, + "loss": 2.347, + "step": 2116 + }, + { + "epoch": 0.9706556625401193, + "grad_norm": 0.44352304671424314, + "learning_rate": 2.2576252981520994e-06, + "loss": 2.3449, + "step": 2117 + }, + { + "epoch": 0.9711141678129298, + "grad_norm": 0.43906168869200685, + "learning_rate": 2.1876767765853233e-06, + "loss": 2.3201, + "step": 2118 + }, + { + "epoch": 0.9715726730857405, + "grad_norm": 0.42617727621739726, + "learning_rate": 2.118826614656788e-06, + "loss": 2.3517, + "step": 2119 + }, + { + "epoch": 0.9720311783585511, + "grad_norm": 0.43590361345308265, + "learning_rate": 2.051074964275701e-06, + "loss": 2.3319, + "step": 2120 + }, + { + "epoch": 0.9724896836313618, + "grad_norm": 0.43091421974131433, + "learning_rate": 1.984421974927375e-06, + "loss": 2.3314, + "step": 2121 + }, + { + "epoch": 0.9729481889041725, + "grad_norm": 0.44942288813862913, + "learning_rate": 1.9188677936731734e-06, + "loss": 2.4371, + "step": 2122 + }, + { + "epoch": 0.973406694176983, + "grad_norm": 0.4476870348549295, + "learning_rate": 1.8544125651501208e-06, + "loss": 2.3685, + "step": 2123 + }, + { + "epoch": 0.9738651994497937, + "grad_norm": 0.4493483255449234, + "learning_rate": 1.7910564315704035e-06, + "loss": 2.4135, + "step": 2124 + }, + { + "epoch": 0.9743237047226043, + "grad_norm": 0.42611244659356945, + "learning_rate": 1.7287995327214257e-06, + "loss": 2.3557, + "step": 2125 + }, + { + "epoch": 0.974782209995415, + "grad_norm": 0.43861757503494986, + "learning_rate": 1.6676420059649754e-06, + "loss": 2.377, + "step": 2126 + }, + { + "epoch": 0.9752407152682255, + "grad_norm": 0.42503695235843, + "learning_rate": 1.6075839862374485e-06, + "loss": 2.2865, + "step": 2127 + }, + { + "epoch": 0.9756992205410362, + "grad_norm": 0.43441026987393105, + "learning_rate": 1.5486256060492366e-06, + "loss": 2.2969, + "step": 2128 + }, + { + "epoch": 0.9761577258138469, + "grad_norm": 0.44704125197625355, + "learning_rate": 1.4907669954844495e-06, + "loss": 2.3107, + "step": 2129 + }, + { + "epoch": 0.9766162310866575, + "grad_norm": 0.42817076648126956, + "learning_rate": 1.434008282200805e-06, + "loss": 2.3559, + "step": 2130 + }, + { + "epoch": 0.9770747363594682, + "grad_norm": 0.4491750496361067, + "learning_rate": 1.3783495914291844e-06, + "loss": 2.4119, + "step": 2131 + }, + { + "epoch": 0.9775332416322787, + "grad_norm": 0.44134043431268066, + "learning_rate": 1.3237910459734104e-06, + "loss": 2.4435, + "step": 2132 + }, + { + "epoch": 0.9779917469050894, + "grad_norm": 0.43835834511041205, + "learning_rate": 1.270332766210025e-06, + "loss": 2.4206, + "step": 2133 + }, + { + "epoch": 0.9784502521779, + "grad_norm": 0.4360548521700408, + "learning_rate": 1.2179748700879012e-06, + "loss": 2.4009, + "step": 2134 + }, + { + "epoch": 0.9789087574507107, + "grad_norm": 0.4405916750299509, + "learning_rate": 1.1667174731280205e-06, + "loss": 2.4506, + "step": 2135 + }, + { + "epoch": 0.9793672627235214, + "grad_norm": 0.42743351667381574, + "learning_rate": 1.1165606884234182e-06, + "loss": 2.3429, + "step": 2136 + }, + { + "epoch": 0.9798257679963319, + "grad_norm": 0.4430933286175603, + "learning_rate": 1.0675046266386268e-06, + "loss": 2.3759, + "step": 2137 + }, + { + "epoch": 0.9802842732691426, + "grad_norm": 0.44778264217663116, + "learning_rate": 1.019549396009567e-06, + "loss": 2.4363, + "step": 2138 + }, + { + "epoch": 0.9807427785419532, + "grad_norm": 0.4388240937566837, + "learning_rate": 9.726951023434348e-07, + "loss": 2.3545, + "step": 2139 + }, + { + "epoch": 0.9812012838147639, + "grad_norm": 0.4437888020396403, + "learning_rate": 9.269418490182591e-07, + "loss": 2.4498, + "step": 2140 + }, + { + "epoch": 0.9816597890875745, + "grad_norm": 0.435365103042124, + "learning_rate": 8.822897369827332e-07, + "loss": 2.3556, + "step": 2141 + }, + { + "epoch": 0.9821182943603851, + "grad_norm": 0.41676945636933943, + "learning_rate": 8.387388647561611e-07, + "loss": 2.2964, + "step": 2142 + }, + { + "epoch": 0.9825767996331958, + "grad_norm": 0.4469870647042109, + "learning_rate": 7.962893284279016e-07, + "loss": 2.222, + "step": 2143 + }, + { + "epoch": 0.9830353049060064, + "grad_norm": 0.45741410501300045, + "learning_rate": 7.549412216574791e-07, + "loss": 2.4316, + "step": 2144 + }, + { + "epoch": 0.9834938101788171, + "grad_norm": 0.4566826295664083, + "learning_rate": 7.146946356743067e-07, + "loss": 2.2753, + "step": 2145 + }, + { + "epoch": 0.9839523154516276, + "grad_norm": 0.45971053362817405, + "learning_rate": 6.755496592773524e-07, + "loss": 2.3326, + "step": 2146 + }, + { + "epoch": 0.9844108207244383, + "grad_norm": 0.46565965690101535, + "learning_rate": 6.375063788349733e-07, + "loss": 2.4064, + "step": 2147 + }, + { + "epoch": 0.984869325997249, + "grad_norm": 0.44016995325676456, + "learning_rate": 6.005648782848594e-07, + "loss": 2.3286, + "step": 2148 + }, + { + "epoch": 0.9853278312700596, + "grad_norm": 0.42942500359850144, + "learning_rate": 5.647252391337565e-07, + "loss": 2.4179, + "step": 2149 + }, + { + "epoch": 0.9857863365428703, + "grad_norm": 0.4670623550335772, + "learning_rate": 5.299875404572441e-07, + "loss": 2.3547, + "step": 2150 + }, + { + "epoch": 0.9862448418156808, + "grad_norm": 0.4324346621771281, + "learning_rate": 4.963518588996796e-07, + "loss": 2.3189, + "step": 2151 + }, + { + "epoch": 0.9867033470884915, + "grad_norm": 0.42672899779628115, + "learning_rate": 4.638182686738657e-07, + "loss": 2.3474, + "step": 2152 + }, + { + "epoch": 0.9871618523613022, + "grad_norm": 0.44192754229646986, + "learning_rate": 4.3238684156110543e-07, + "loss": 2.3058, + "step": 2153 + }, + { + "epoch": 0.9876203576341128, + "grad_norm": 0.4365858064383398, + "learning_rate": 4.020576469108139e-07, + "loss": 2.2662, + "step": 2154 + }, + { + "epoch": 0.9880788629069235, + "grad_norm": 0.43249068404619273, + "learning_rate": 3.7283075164046274e-07, + "loss": 2.3671, + "step": 2155 + }, + { + "epoch": 0.988537368179734, + "grad_norm": 0.4354541392774458, + "learning_rate": 3.4470622023557995e-07, + "loss": 2.3243, + "step": 2156 + }, + { + "epoch": 0.9889958734525447, + "grad_norm": 0.42861591431410195, + "learning_rate": 3.176841147492504e-07, + "loss": 2.3037, + "step": 2157 + }, + { + "epoch": 0.9894543787253554, + "grad_norm": 0.4189585275722092, + "learning_rate": 2.9176449480244895e-07, + "loss": 2.2643, + "step": 2158 + }, + { + "epoch": 0.989912883998166, + "grad_norm": 0.4473601391308684, + "learning_rate": 2.6694741758342967e-07, + "loss": 2.4023, + "step": 2159 + }, + { + "epoch": 0.9903713892709766, + "grad_norm": 0.4175072962523477, + "learning_rate": 2.432329378478926e-07, + "loss": 2.3647, + "step": 2160 + }, + { + "epoch": 0.9908298945437872, + "grad_norm": 0.43045639079948317, + "learning_rate": 2.2062110791892798e-07, + "loss": 2.3226, + "step": 2161 + }, + { + "epoch": 0.9912883998165979, + "grad_norm": 0.4288825989957075, + "learning_rate": 1.9911197768662792e-07, + "loss": 2.3459, + "step": 2162 + }, + { + "epoch": 0.9917469050894085, + "grad_norm": 0.4247843067443171, + "learning_rate": 1.7870559460814173e-07, + "loss": 2.4362, + "step": 2163 + }, + { + "epoch": 0.9922054103622192, + "grad_norm": 0.43756933586195884, + "learning_rate": 1.5940200370750947e-07, + "loss": 2.3744, + "step": 2164 + }, + { + "epoch": 0.9926639156350298, + "grad_norm": 0.4540119546750009, + "learning_rate": 1.4120124757577291e-07, + "loss": 2.4697, + "step": 2165 + }, + { + "epoch": 0.9931224209078404, + "grad_norm": 0.4439618629407356, + "learning_rate": 1.2410336637047603e-07, + "loss": 2.2743, + "step": 2166 + }, + { + "epoch": 0.9935809261806511, + "grad_norm": 0.43495856819954964, + "learning_rate": 1.081083978159425e-07, + "loss": 2.42, + "step": 2167 + }, + { + "epoch": 0.9940394314534617, + "grad_norm": 0.4279096509091773, + "learning_rate": 9.321637720310915e-08, + "loss": 2.3077, + "step": 2168 + }, + { + "epoch": 0.9944979367262724, + "grad_norm": 0.44509889500002975, + "learning_rate": 7.942733738924845e-08, + "loss": 2.4319, + "step": 2169 + }, + { + "epoch": 0.994956441999083, + "grad_norm": 0.43851439614398174, + "learning_rate": 6.6741308798135e-08, + "loss": 2.3582, + "step": 2170 + }, + { + "epoch": 0.9954149472718936, + "grad_norm": 0.4440232792122264, + "learning_rate": 5.5158319419934546e-08, + "loss": 2.4143, + "step": 2171 + }, + { + "epoch": 0.9958734525447043, + "grad_norm": 0.44477114913066335, + "learning_rate": 4.4678394810981906e-08, + "loss": 2.3458, + "step": 2172 + }, + { + "epoch": 0.9963319578175149, + "grad_norm": 0.4173415513388647, + "learning_rate": 3.5301558093947527e-08, + "loss": 2.4105, + "step": 2173 + }, + { + "epoch": 0.9967904630903255, + "grad_norm": 0.42155315371179747, + "learning_rate": 2.7027829957559925e-08, + "loss": 2.3531, + "step": 2174 + }, + { + "epoch": 0.9972489683631361, + "grad_norm": 0.45820698418378264, + "learning_rate": 1.985722865682771e-08, + "loss": 2.3673, + "step": 2175 + }, + { + "epoch": 0.9977074736359468, + "grad_norm": 0.4251975893913342, + "learning_rate": 1.3789770012762048e-08, + "loss": 2.2673, + "step": 2176 + }, + { + "epoch": 0.9981659789087575, + "grad_norm": 0.42451012801738225, + "learning_rate": 8.825467412376665e-09, + "loss": 2.3728, + "step": 2177 + }, + { + "epoch": 0.9986244841815681, + "grad_norm": 0.44103273071032423, + "learning_rate": 4.96433180879885e-09, + "loss": 2.3737, + "step": 2178 + }, + { + "epoch": 0.9990829894543787, + "grad_norm": 0.44160743048952145, + "learning_rate": 2.206371721158451e-09, + "loss": 2.3752, + "step": 2179 + }, + { + "epoch": 0.9995414947271893, + "grad_norm": 0.4339900388635622, + "learning_rate": 5.515932345323549e-10, + "loss": 2.2937, + "step": 2180 + }, + { + "epoch": 1.0, + "grad_norm": 0.4098020277124717, + "learning_rate": 0.0, + "loss": 2.3767, + "step": 2181 + }, + { + "epoch": 1.0, + "step": 2181, + "total_flos": 6.473628430872084e+17, + "train_loss": 2.506032412121024, + "train_runtime": 5166.7848, + "train_samples_per_second": 108.022, + "train_steps_per_second": 0.422 + } + ], + "logging_steps": 1.0, + "max_steps": 2181, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 24000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": false, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 6.473628430872084e+17, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}