{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.99835255354201, "eval_steps": 500, "global_step": 1365, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.002196595277320154, "grad_norm": 2047828.375, "learning_rate": 0.0, "loss": 1.3567, "step": 1 }, { "epoch": 0.004393190554640308, "grad_norm": 6674292.0, "learning_rate": 3.6496350364963505e-07, "loss": 1.5194, "step": 2 }, { "epoch": 0.006589785831960461, "grad_norm": 40031500.0, "learning_rate": 7.299270072992701e-07, "loss": 1.2716, "step": 3 }, { "epoch": 0.008786381109280615, "grad_norm": 4004006.25, "learning_rate": 1.0948905109489052e-06, "loss": 1.1705, "step": 4 }, { "epoch": 0.010982976386600769, "grad_norm": 11234545.0, "learning_rate": 1.4598540145985402e-06, "loss": 1.2851, "step": 5 }, { "epoch": 0.013179571663920923, "grad_norm": 7696731.0, "learning_rate": 1.824817518248175e-06, "loss": 1.1159, "step": 6 }, { "epoch": 0.015376166941241077, "grad_norm": 19380058.0, "learning_rate": 2.1897810218978103e-06, "loss": 1.0939, "step": 7 }, { "epoch": 0.01757276221856123, "grad_norm": 51641744.0, "learning_rate": 2.5547445255474454e-06, "loss": 1.1843, "step": 8 }, { "epoch": 0.019769357495881382, "grad_norm": 51203412.0, "learning_rate": 2.9197080291970804e-06, "loss": 1.4502, "step": 9 }, { "epoch": 0.021965952773201538, "grad_norm": 8740138.0, "learning_rate": 3.2846715328467155e-06, "loss": 1.1613, "step": 10 }, { "epoch": 0.02416254805052169, "grad_norm": 14772801.0, "learning_rate": 3.64963503649635e-06, "loss": 1.3874, "step": 11 }, { "epoch": 0.026359143327841845, "grad_norm": 11424318.0, "learning_rate": 4.014598540145985e-06, "loss": 1.2471, "step": 12 }, { "epoch": 0.028555738605161998, "grad_norm": 6222449.0, "learning_rate": 4.379562043795621e-06, "loss": 1.219, "step": 13 }, { "epoch": 0.030752333882482153, "grad_norm": 14838477.0, "learning_rate": 4.744525547445255e-06, "loss": 1.2428, "step": 14 }, { "epoch": 0.032948929159802305, "grad_norm": 34365772.0, "learning_rate": 5.109489051094891e-06, "loss": 1.3154, "step": 15 }, { "epoch": 0.03514552443712246, "grad_norm": 11343648.0, "learning_rate": 5.474452554744526e-06, "loss": 1.0682, "step": 16 }, { "epoch": 0.037342119714442616, "grad_norm": 16938996.0, "learning_rate": 5.839416058394161e-06, "loss": 1.1922, "step": 17 }, { "epoch": 0.039538714991762765, "grad_norm": 28895242.0, "learning_rate": 6.204379562043796e-06, "loss": 1.4344, "step": 18 }, { "epoch": 0.04173531026908292, "grad_norm": 11380300.0, "learning_rate": 6.569343065693431e-06, "loss": 1.3168, "step": 19 }, { "epoch": 0.043931905546403076, "grad_norm": 22187968.0, "learning_rate": 6.9343065693430655e-06, "loss": 1.2649, "step": 20 }, { "epoch": 0.04612850082372323, "grad_norm": 10284782.0, "learning_rate": 7.2992700729927e-06, "loss": 1.2709, "step": 21 }, { "epoch": 0.04832509610104338, "grad_norm": 14559371.0, "learning_rate": 7.664233576642336e-06, "loss": 1.2476, "step": 22 }, { "epoch": 0.050521691378363535, "grad_norm": 10865930.0, "learning_rate": 8.02919708029197e-06, "loss": 1.5553, "step": 23 }, { "epoch": 0.05271828665568369, "grad_norm": 12806353.0, "learning_rate": 8.394160583941606e-06, "loss": 1.2871, "step": 24 }, { "epoch": 0.054914881933003847, "grad_norm": 12299012.0, "learning_rate": 8.759124087591241e-06, "loss": 1.1707, "step": 25 }, { "epoch": 0.057111477210323995, "grad_norm": 7523549.0, "learning_rate": 9.124087591240877e-06, "loss": 1.2504, "step": 26 }, { "epoch": 0.05930807248764415, "grad_norm": 5730644.0, "learning_rate": 9.48905109489051e-06, "loss": 1.1693, "step": 27 }, { "epoch": 0.061504667764964306, "grad_norm": 5095684.5, "learning_rate": 9.854014598540148e-06, "loss": 1.0545, "step": 28 }, { "epoch": 0.06370126304228446, "grad_norm": 7149638.0, "learning_rate": 1.0218978102189781e-05, "loss": 1.1866, "step": 29 }, { "epoch": 0.06589785831960461, "grad_norm": 25770528.0, "learning_rate": 1.0583941605839417e-05, "loss": 1.1411, "step": 30 }, { "epoch": 0.06809445359692477, "grad_norm": 7752126.0, "learning_rate": 1.0948905109489052e-05, "loss": 1.209, "step": 31 }, { "epoch": 0.07029104887424492, "grad_norm": 11265648.0, "learning_rate": 1.1313868613138686e-05, "loss": 1.3958, "step": 32 }, { "epoch": 0.07248764415156507, "grad_norm": 21874444.0, "learning_rate": 1.1678832116788322e-05, "loss": 1.6249, "step": 33 }, { "epoch": 0.07468423942888523, "grad_norm": 11610835.0, "learning_rate": 1.2043795620437957e-05, "loss": 1.1728, "step": 34 }, { "epoch": 0.07688083470620538, "grad_norm": 5678601.5, "learning_rate": 1.2408759124087593e-05, "loss": 1.2732, "step": 35 }, { "epoch": 0.07907742998352553, "grad_norm": 14198422.0, "learning_rate": 1.2773722627737228e-05, "loss": 1.2813, "step": 36 }, { "epoch": 0.08127402526084569, "grad_norm": 8859336.0, "learning_rate": 1.3138686131386862e-05, "loss": 1.2603, "step": 37 }, { "epoch": 0.08347062053816584, "grad_norm": 5651872.0, "learning_rate": 1.3503649635036497e-05, "loss": 1.2841, "step": 38 }, { "epoch": 0.085667215815486, "grad_norm": 17596590.0, "learning_rate": 1.3868613138686131e-05, "loss": 1.4325, "step": 39 }, { "epoch": 0.08786381109280615, "grad_norm": 12388730.0, "learning_rate": 1.4233576642335767e-05, "loss": 1.1562, "step": 40 }, { "epoch": 0.0900604063701263, "grad_norm": 12075692.0, "learning_rate": 1.45985401459854e-05, "loss": 1.2676, "step": 41 }, { "epoch": 0.09225700164744646, "grad_norm": 27072708.0, "learning_rate": 1.496350364963504e-05, "loss": 1.4864, "step": 42 }, { "epoch": 0.09445359692476661, "grad_norm": 37403220.0, "learning_rate": 1.5328467153284673e-05, "loss": 1.2104, "step": 43 }, { "epoch": 0.09665019220208676, "grad_norm": 12227715.0, "learning_rate": 1.569343065693431e-05, "loss": 1.3041, "step": 44 }, { "epoch": 0.09884678747940692, "grad_norm": 12901322.0, "learning_rate": 1.605839416058394e-05, "loss": 1.3288, "step": 45 }, { "epoch": 0.10104338275672707, "grad_norm": 17107380.0, "learning_rate": 1.6423357664233576e-05, "loss": 1.2191, "step": 46 }, { "epoch": 0.10323997803404723, "grad_norm": 31481982.0, "learning_rate": 1.678832116788321e-05, "loss": 1.2633, "step": 47 }, { "epoch": 0.10543657331136738, "grad_norm": 15303975.0, "learning_rate": 1.715328467153285e-05, "loss": 1.2865, "step": 48 }, { "epoch": 0.10763316858868753, "grad_norm": 17270372.0, "learning_rate": 1.7518248175182482e-05, "loss": 1.4734, "step": 49 }, { "epoch": 0.10982976386600769, "grad_norm": 34183024.0, "learning_rate": 1.7883211678832118e-05, "loss": 1.303, "step": 50 }, { "epoch": 0.11202635914332784, "grad_norm": 14855132.0, "learning_rate": 1.8248175182481753e-05, "loss": 1.2756, "step": 51 }, { "epoch": 0.11422295442064799, "grad_norm": 18743198.0, "learning_rate": 1.861313868613139e-05, "loss": 1.1061, "step": 52 }, { "epoch": 0.11641954969796815, "grad_norm": 6949814.5, "learning_rate": 1.897810218978102e-05, "loss": 1.2949, "step": 53 }, { "epoch": 0.1186161449752883, "grad_norm": 7799824.0, "learning_rate": 1.934306569343066e-05, "loss": 1.0481, "step": 54 }, { "epoch": 0.12081274025260846, "grad_norm": 8161740.0, "learning_rate": 1.9708029197080295e-05, "loss": 1.1827, "step": 55 }, { "epoch": 0.12300933552992861, "grad_norm": 34570872.0, "learning_rate": 2.0072992700729927e-05, "loss": 1.4264, "step": 56 }, { "epoch": 0.12520593080724876, "grad_norm": 7429580.5, "learning_rate": 2.0437956204379563e-05, "loss": 1.0902, "step": 57 }, { "epoch": 0.12740252608456892, "grad_norm": 17569806.0, "learning_rate": 2.08029197080292e-05, "loss": 1.2842, "step": 58 }, { "epoch": 0.12959912136188906, "grad_norm": 15480506.0, "learning_rate": 2.1167883211678834e-05, "loss": 1.0864, "step": 59 }, { "epoch": 0.13179571663920922, "grad_norm": 12282927.0, "learning_rate": 2.1532846715328466e-05, "loss": 1.1316, "step": 60 }, { "epoch": 0.13399231191652938, "grad_norm": 4455538.0, "learning_rate": 2.1897810218978105e-05, "loss": 1.2465, "step": 61 }, { "epoch": 0.13618890719384955, "grad_norm": 15463966.0, "learning_rate": 2.226277372262774e-05, "loss": 1.099, "step": 62 }, { "epoch": 0.13838550247116968, "grad_norm": 22602364.0, "learning_rate": 2.2627737226277372e-05, "loss": 1.0459, "step": 63 }, { "epoch": 0.14058209774848984, "grad_norm": 10241717.0, "learning_rate": 2.2992700729927008e-05, "loss": 1.2806, "step": 64 }, { "epoch": 0.14277869302581, "grad_norm": 7588423.5, "learning_rate": 2.3357664233576643e-05, "loss": 1.114, "step": 65 }, { "epoch": 0.14497528830313014, "grad_norm": 10213540.0, "learning_rate": 2.372262773722628e-05, "loss": 0.9609, "step": 66 }, { "epoch": 0.1471718835804503, "grad_norm": 11998898.0, "learning_rate": 2.4087591240875914e-05, "loss": 1.1912, "step": 67 }, { "epoch": 0.14936847885777046, "grad_norm": 5615866.0, "learning_rate": 2.445255474452555e-05, "loss": 1.049, "step": 68 }, { "epoch": 0.1515650741350906, "grad_norm": 19550316.0, "learning_rate": 2.4817518248175185e-05, "loss": 1.2795, "step": 69 }, { "epoch": 0.15376166941241076, "grad_norm": 10943439.0, "learning_rate": 2.518248175182482e-05, "loss": 1.1303, "step": 70 }, { "epoch": 0.15595826468973092, "grad_norm": 19708706.0, "learning_rate": 2.5547445255474456e-05, "loss": 1.2344, "step": 71 }, { "epoch": 0.15815485996705106, "grad_norm": 17543686.0, "learning_rate": 2.591240875912409e-05, "loss": 1.1521, "step": 72 }, { "epoch": 0.16035145524437122, "grad_norm": 43875364.0, "learning_rate": 2.6277372262773724e-05, "loss": 1.3358, "step": 73 }, { "epoch": 0.16254805052169138, "grad_norm": 6051551.5, "learning_rate": 2.664233576642336e-05, "loss": 1.2323, "step": 74 }, { "epoch": 0.16474464579901152, "grad_norm": 32905656.0, "learning_rate": 2.7007299270072995e-05, "loss": 1.0926, "step": 75 }, { "epoch": 0.16694124107633168, "grad_norm": 8517439.0, "learning_rate": 2.737226277372263e-05, "loss": 1.1515, "step": 76 }, { "epoch": 0.16913783635365184, "grad_norm": 11192837.0, "learning_rate": 2.7737226277372262e-05, "loss": 1.0715, "step": 77 }, { "epoch": 0.171334431630972, "grad_norm": 9605860.0, "learning_rate": 2.8102189781021898e-05, "loss": 0.8551, "step": 78 }, { "epoch": 0.17353102690829214, "grad_norm": 8618499.0, "learning_rate": 2.8467153284671533e-05, "loss": 1.0788, "step": 79 }, { "epoch": 0.1757276221856123, "grad_norm": 14007595.0, "learning_rate": 2.883211678832117e-05, "loss": 1.2054, "step": 80 }, { "epoch": 0.17792421746293247, "grad_norm": 9647478.0, "learning_rate": 2.91970802919708e-05, "loss": 1.1001, "step": 81 }, { "epoch": 0.1801208127402526, "grad_norm": 38317572.0, "learning_rate": 2.9562043795620443e-05, "loss": 1.0888, "step": 82 }, { "epoch": 0.18231740801757276, "grad_norm": 16823494.0, "learning_rate": 2.992700729927008e-05, "loss": 1.0136, "step": 83 }, { "epoch": 0.18451400329489293, "grad_norm": 9234012.0, "learning_rate": 3.029197080291971e-05, "loss": 1.1833, "step": 84 }, { "epoch": 0.18671059857221306, "grad_norm": 13358546.0, "learning_rate": 3.0656934306569346e-05, "loss": 1.2703, "step": 85 }, { "epoch": 0.18890719384953322, "grad_norm": 22999784.0, "learning_rate": 3.102189781021898e-05, "loss": 1.4449, "step": 86 }, { "epoch": 0.19110378912685339, "grad_norm": 16768359.0, "learning_rate": 3.138686131386862e-05, "loss": 1.1601, "step": 87 }, { "epoch": 0.19330038440417352, "grad_norm": 31959344.0, "learning_rate": 3.175182481751825e-05, "loss": 1.2223, "step": 88 }, { "epoch": 0.19549697968149368, "grad_norm": 19150826.0, "learning_rate": 3.211678832116788e-05, "loss": 0.9576, "step": 89 }, { "epoch": 0.19769357495881384, "grad_norm": 13048818.0, "learning_rate": 3.248175182481752e-05, "loss": 1.3498, "step": 90 }, { "epoch": 0.19989017023613398, "grad_norm": 12262197.0, "learning_rate": 3.284671532846715e-05, "loss": 1.052, "step": 91 }, { "epoch": 0.20208676551345414, "grad_norm": 2876966.25, "learning_rate": 3.321167883211679e-05, "loss": 1.0189, "step": 92 }, { "epoch": 0.2042833607907743, "grad_norm": 18541894.0, "learning_rate": 3.357664233576642e-05, "loss": 1.0832, "step": 93 }, { "epoch": 0.20647995606809447, "grad_norm": 15970311.0, "learning_rate": 3.3941605839416055e-05, "loss": 1.2928, "step": 94 }, { "epoch": 0.2086765513454146, "grad_norm": 25580606.0, "learning_rate": 3.43065693430657e-05, "loss": 1.0362, "step": 95 }, { "epoch": 0.21087314662273476, "grad_norm": 14224102.0, "learning_rate": 3.467153284671533e-05, "loss": 1.2039, "step": 96 }, { "epoch": 0.21306974190005493, "grad_norm": 6323020.5, "learning_rate": 3.5036496350364965e-05, "loss": 1.237, "step": 97 }, { "epoch": 0.21526633717737506, "grad_norm": 4563988.5, "learning_rate": 3.5401459854014604e-05, "loss": 1.1155, "step": 98 }, { "epoch": 0.21746293245469522, "grad_norm": 11096799.0, "learning_rate": 3.5766423357664236e-05, "loss": 1.1138, "step": 99 }, { "epoch": 0.21965952773201539, "grad_norm": 5840709.5, "learning_rate": 3.613138686131387e-05, "loss": 1.1854, "step": 100 }, { "epoch": 0.22185612300933552, "grad_norm": 4187084.5, "learning_rate": 3.649635036496351e-05, "loss": 1.3305, "step": 101 }, { "epoch": 0.22405271828665568, "grad_norm": 6342277.5, "learning_rate": 3.686131386861314e-05, "loss": 1.1569, "step": 102 }, { "epoch": 0.22624931356397585, "grad_norm": 5076073.5, "learning_rate": 3.722627737226278e-05, "loss": 1.1453, "step": 103 }, { "epoch": 0.22844590884129598, "grad_norm": 12348493.0, "learning_rate": 3.759124087591241e-05, "loss": 1.0243, "step": 104 }, { "epoch": 0.23064250411861614, "grad_norm": 10997787.0, "learning_rate": 3.795620437956204e-05, "loss": 1.1016, "step": 105 }, { "epoch": 0.2328390993959363, "grad_norm": 6869693.0, "learning_rate": 3.832116788321168e-05, "loss": 1.1914, "step": 106 }, { "epoch": 0.23503569467325644, "grad_norm": 24705828.0, "learning_rate": 3.868613138686132e-05, "loss": 1.2834, "step": 107 }, { "epoch": 0.2372322899505766, "grad_norm": 13220064.0, "learning_rate": 3.905109489051095e-05, "loss": 0.9956, "step": 108 }, { "epoch": 0.23942888522789676, "grad_norm": 34857952.0, "learning_rate": 3.941605839416059e-05, "loss": 1.0755, "step": 109 }, { "epoch": 0.24162548050521693, "grad_norm": 17391034.0, "learning_rate": 3.978102189781022e-05, "loss": 1.0426, "step": 110 }, { "epoch": 0.24382207578253706, "grad_norm": 16179722.0, "learning_rate": 4.0145985401459855e-05, "loss": 1.1024, "step": 111 }, { "epoch": 0.24601867105985722, "grad_norm": 19842270.0, "learning_rate": 4.0510948905109494e-05, "loss": 1.0659, "step": 112 }, { "epoch": 0.2482152663371774, "grad_norm": 14679981.0, "learning_rate": 4.0875912408759126e-05, "loss": 1.0519, "step": 113 }, { "epoch": 0.2504118616144975, "grad_norm": 22071618.0, "learning_rate": 4.124087591240876e-05, "loss": 1.0097, "step": 114 }, { "epoch": 0.25260845689181766, "grad_norm": 22394396.0, "learning_rate": 4.16058394160584e-05, "loss": 1.1212, "step": 115 }, { "epoch": 0.25480505216913785, "grad_norm": 7273321.5, "learning_rate": 4.197080291970803e-05, "loss": 1.1518, "step": 116 }, { "epoch": 0.257001647446458, "grad_norm": 35136992.0, "learning_rate": 4.233576642335767e-05, "loss": 1.0966, "step": 117 }, { "epoch": 0.2591982427237781, "grad_norm": 20715392.0, "learning_rate": 4.27007299270073e-05, "loss": 1.0206, "step": 118 }, { "epoch": 0.2613948380010983, "grad_norm": 28416488.0, "learning_rate": 4.306569343065693e-05, "loss": 0.9672, "step": 119 }, { "epoch": 0.26359143327841844, "grad_norm": 12644584.0, "learning_rate": 4.343065693430657e-05, "loss": 0.9307, "step": 120 }, { "epoch": 0.26578802855573863, "grad_norm": 46953568.0, "learning_rate": 4.379562043795621e-05, "loss": 1.0041, "step": 121 }, { "epoch": 0.26798462383305877, "grad_norm": 18650110.0, "learning_rate": 4.416058394160584e-05, "loss": 0.9494, "step": 122 }, { "epoch": 0.2701812191103789, "grad_norm": 7411148.5, "learning_rate": 4.452554744525548e-05, "loss": 1.0256, "step": 123 }, { "epoch": 0.2723778143876991, "grad_norm": 14795885.0, "learning_rate": 4.489051094890511e-05, "loss": 1.006, "step": 124 }, { "epoch": 0.2745744096650192, "grad_norm": 7987120.0, "learning_rate": 4.5255474452554745e-05, "loss": 1.1112, "step": 125 }, { "epoch": 0.27677100494233936, "grad_norm": 3664041.5, "learning_rate": 4.5620437956204383e-05, "loss": 0.827, "step": 126 }, { "epoch": 0.27896760021965955, "grad_norm": 4188943.75, "learning_rate": 4.5985401459854016e-05, "loss": 1.0363, "step": 127 }, { "epoch": 0.2811641954969797, "grad_norm": 10032366.0, "learning_rate": 4.635036496350365e-05, "loss": 1.3081, "step": 128 }, { "epoch": 0.2833607907742998, "grad_norm": 19652358.0, "learning_rate": 4.6715328467153287e-05, "loss": 1.1723, "step": 129 }, { "epoch": 0.28555738605162, "grad_norm": 29400118.0, "learning_rate": 4.708029197080292e-05, "loss": 1.0401, "step": 130 }, { "epoch": 0.28775398132894014, "grad_norm": 6440731.0, "learning_rate": 4.744525547445256e-05, "loss": 0.9859, "step": 131 }, { "epoch": 0.2899505766062603, "grad_norm": 6155487.0, "learning_rate": 4.7810218978102196e-05, "loss": 0.8675, "step": 132 }, { "epoch": 0.29214717188358047, "grad_norm": 11451470.0, "learning_rate": 4.817518248175183e-05, "loss": 0.8143, "step": 133 }, { "epoch": 0.2943437671609006, "grad_norm": 8476629.0, "learning_rate": 4.854014598540147e-05, "loss": 0.9655, "step": 134 }, { "epoch": 0.29654036243822074, "grad_norm": 20757296.0, "learning_rate": 4.89051094890511e-05, "loss": 0.9572, "step": 135 }, { "epoch": 0.29873695771554093, "grad_norm": 17907030.0, "learning_rate": 4.927007299270073e-05, "loss": 1.1968, "step": 136 }, { "epoch": 0.30093355299286106, "grad_norm": 16662175.0, "learning_rate": 4.963503649635037e-05, "loss": 1.0537, "step": 137 }, { "epoch": 0.3031301482701812, "grad_norm": 10720323.0, "learning_rate": 5e-05, "loss": 1.032, "step": 138 }, { "epoch": 0.3053267435475014, "grad_norm": 6210951.5, "learning_rate": 4.995928338762215e-05, "loss": 0.9378, "step": 139 }, { "epoch": 0.3075233388248215, "grad_norm": 12067100.0, "learning_rate": 4.99185667752443e-05, "loss": 1.0416, "step": 140 }, { "epoch": 0.30971993410214166, "grad_norm": 10375612.0, "learning_rate": 4.9877850162866454e-05, "loss": 1.0868, "step": 141 }, { "epoch": 0.31191652937946185, "grad_norm": 19215752.0, "learning_rate": 4.9837133550488604e-05, "loss": 1.1573, "step": 142 }, { "epoch": 0.314113124656782, "grad_norm": 18632564.0, "learning_rate": 4.9796416938110755e-05, "loss": 1.0283, "step": 143 }, { "epoch": 0.3163097199341021, "grad_norm": 11247008.0, "learning_rate": 4.97557003257329e-05, "loss": 1.2683, "step": 144 }, { "epoch": 0.3185063152114223, "grad_norm": 10169539.0, "learning_rate": 4.971498371335505e-05, "loss": 1.0616, "step": 145 }, { "epoch": 0.32070291048874244, "grad_norm": 14896106.0, "learning_rate": 4.96742671009772e-05, "loss": 1.0025, "step": 146 }, { "epoch": 0.3228995057660626, "grad_norm": 11283311.0, "learning_rate": 4.963355048859935e-05, "loss": 0.919, "step": 147 }, { "epoch": 0.32509610104338277, "grad_norm": 5662458.5, "learning_rate": 4.95928338762215e-05, "loss": 0.8772, "step": 148 }, { "epoch": 0.3272926963207029, "grad_norm": 9590134.0, "learning_rate": 4.955211726384365e-05, "loss": 1.004, "step": 149 }, { "epoch": 0.32948929159802304, "grad_norm": 6902963.5, "learning_rate": 4.95114006514658e-05, "loss": 1.2467, "step": 150 }, { "epoch": 0.3316858868753432, "grad_norm": 33446322.0, "learning_rate": 4.947068403908795e-05, "loss": 1.0799, "step": 151 }, { "epoch": 0.33388248215266336, "grad_norm": 14731285.0, "learning_rate": 4.94299674267101e-05, "loss": 1.2127, "step": 152 }, { "epoch": 0.33607907742998355, "grad_norm": 15324599.0, "learning_rate": 4.938925081433225e-05, "loss": 0.8953, "step": 153 }, { "epoch": 0.3382756727073037, "grad_norm": 12761835.0, "learning_rate": 4.9348534201954396e-05, "loss": 0.9621, "step": 154 }, { "epoch": 0.3404722679846238, "grad_norm": 4396397.0, "learning_rate": 4.930781758957655e-05, "loss": 0.9655, "step": 155 }, { "epoch": 0.342668863261944, "grad_norm": 18684486.0, "learning_rate": 4.92671009771987e-05, "loss": 0.961, "step": 156 }, { "epoch": 0.34486545853926415, "grad_norm": 11647168.0, "learning_rate": 4.922638436482085e-05, "loss": 1.0236, "step": 157 }, { "epoch": 0.3470620538165843, "grad_norm": 7403086.5, "learning_rate": 4.9185667752443e-05, "loss": 0.921, "step": 158 }, { "epoch": 0.34925864909390447, "grad_norm": 9301290.0, "learning_rate": 4.914495114006515e-05, "loss": 1.073, "step": 159 }, { "epoch": 0.3514552443712246, "grad_norm": 6785378.5, "learning_rate": 4.91042345276873e-05, "loss": 0.9421, "step": 160 }, { "epoch": 0.35365183964854474, "grad_norm": 37305500.0, "learning_rate": 4.906351791530945e-05, "loss": 0.9959, "step": 161 }, { "epoch": 0.35584843492586493, "grad_norm": 12157369.0, "learning_rate": 4.90228013029316e-05, "loss": 0.8274, "step": 162 }, { "epoch": 0.35804503020318507, "grad_norm": 11638986.0, "learning_rate": 4.898208469055375e-05, "loss": 0.679, "step": 163 }, { "epoch": 0.3602416254805052, "grad_norm": 16125764.0, "learning_rate": 4.89413680781759e-05, "loss": 0.8744, "step": 164 }, { "epoch": 0.3624382207578254, "grad_norm": 5387654.5, "learning_rate": 4.8900651465798044e-05, "loss": 1.0482, "step": 165 }, { "epoch": 0.3646348160351455, "grad_norm": 23652066.0, "learning_rate": 4.8859934853420195e-05, "loss": 0.9726, "step": 166 }, { "epoch": 0.36683141131246566, "grad_norm": 9789622.0, "learning_rate": 4.8819218241042345e-05, "loss": 0.8733, "step": 167 }, { "epoch": 0.36902800658978585, "grad_norm": 30906436.0, "learning_rate": 4.8778501628664496e-05, "loss": 0.9609, "step": 168 }, { "epoch": 0.371224601867106, "grad_norm": 189772432.0, "learning_rate": 4.8737785016286646e-05, "loss": 1.0023, "step": 169 }, { "epoch": 0.3734211971444261, "grad_norm": 8754641.0, "learning_rate": 4.86970684039088e-05, "loss": 1.0592, "step": 170 }, { "epoch": 0.3756177924217463, "grad_norm": 17426248.0, "learning_rate": 4.865635179153095e-05, "loss": 0.8442, "step": 171 }, { "epoch": 0.37781438769906645, "grad_norm": 4542470.0, "learning_rate": 4.86156351791531e-05, "loss": 0.7911, "step": 172 }, { "epoch": 0.3800109829763866, "grad_norm": 8447095.0, "learning_rate": 4.857491856677525e-05, "loss": 0.9593, "step": 173 }, { "epoch": 0.38220757825370677, "grad_norm": 7352316.5, "learning_rate": 4.85342019543974e-05, "loss": 1.1872, "step": 174 }, { "epoch": 0.3844041735310269, "grad_norm": 5602028.5, "learning_rate": 4.849348534201954e-05, "loss": 0.7461, "step": 175 }, { "epoch": 0.38660076880834704, "grad_norm": 4677864.0, "learning_rate": 4.845276872964169e-05, "loss": 1.0743, "step": 176 }, { "epoch": 0.38879736408566723, "grad_norm": 4680797.5, "learning_rate": 4.841205211726384e-05, "loss": 0.6711, "step": 177 }, { "epoch": 0.39099395936298736, "grad_norm": 5534360.0, "learning_rate": 4.8371335504885994e-05, "loss": 0.9604, "step": 178 }, { "epoch": 0.3931905546403075, "grad_norm": 13123183.0, "learning_rate": 4.8330618892508144e-05, "loss": 1.017, "step": 179 }, { "epoch": 0.3953871499176277, "grad_norm": 6758069.0, "learning_rate": 4.8289902280130295e-05, "loss": 0.8691, "step": 180 }, { "epoch": 0.3975837451949478, "grad_norm": 8226452.0, "learning_rate": 4.8249185667752445e-05, "loss": 0.8608, "step": 181 }, { "epoch": 0.39978034047226796, "grad_norm": 3854469.5, "learning_rate": 4.8208469055374595e-05, "loss": 0.7598, "step": 182 }, { "epoch": 0.40197693574958815, "grad_norm": 18598116.0, "learning_rate": 4.8167752442996746e-05, "loss": 0.9759, "step": 183 }, { "epoch": 0.4041735310269083, "grad_norm": 11485942.0, "learning_rate": 4.8127035830618896e-05, "loss": 1.2585, "step": 184 }, { "epoch": 0.4063701263042284, "grad_norm": 11972959.0, "learning_rate": 4.808631921824105e-05, "loss": 1.0022, "step": 185 }, { "epoch": 0.4085667215815486, "grad_norm": 6080062.0, "learning_rate": 4.804560260586319e-05, "loss": 0.9132, "step": 186 }, { "epoch": 0.41076331685886874, "grad_norm": 8417478.0, "learning_rate": 4.800488599348534e-05, "loss": 0.7388, "step": 187 }, { "epoch": 0.41295991213618893, "grad_norm": 10634729.0, "learning_rate": 4.796416938110749e-05, "loss": 0.9308, "step": 188 }, { "epoch": 0.41515650741350907, "grad_norm": 11250071.0, "learning_rate": 4.792345276872964e-05, "loss": 1.033, "step": 189 }, { "epoch": 0.4173531026908292, "grad_norm": 5495279.0, "learning_rate": 4.788273615635179e-05, "loss": 0.8692, "step": 190 }, { "epoch": 0.4195496979681494, "grad_norm": 37268708.0, "learning_rate": 4.784201954397394e-05, "loss": 0.8936, "step": 191 }, { "epoch": 0.42174629324546953, "grad_norm": 6610986.5, "learning_rate": 4.780130293159609e-05, "loss": 0.9738, "step": 192 }, { "epoch": 0.42394288852278966, "grad_norm": 11809499.0, "learning_rate": 4.7760586319218244e-05, "loss": 0.9591, "step": 193 }, { "epoch": 0.42613948380010985, "grad_norm": 8254889.0, "learning_rate": 4.7719869706840394e-05, "loss": 0.9874, "step": 194 }, { "epoch": 0.42833607907743, "grad_norm": 7814906.0, "learning_rate": 4.7679153094462545e-05, "loss": 1.155, "step": 195 }, { "epoch": 0.4305326743547501, "grad_norm": 7317868.5, "learning_rate": 4.7638436482084695e-05, "loss": 1.3268, "step": 196 }, { "epoch": 0.4327292696320703, "grad_norm": 5934029.0, "learning_rate": 4.759771986970684e-05, "loss": 0.812, "step": 197 }, { "epoch": 0.43492586490939045, "grad_norm": 5832792.5, "learning_rate": 4.755700325732899e-05, "loss": 0.7935, "step": 198 }, { "epoch": 0.4371224601867106, "grad_norm": 8039267.0, "learning_rate": 4.751628664495114e-05, "loss": 0.8929, "step": 199 }, { "epoch": 0.43931905546403077, "grad_norm": 9420844.0, "learning_rate": 4.747557003257329e-05, "loss": 0.9899, "step": 200 }, { "epoch": 0.4415156507413509, "grad_norm": 2905903.25, "learning_rate": 4.743485342019544e-05, "loss": 0.7712, "step": 201 }, { "epoch": 0.44371224601867104, "grad_norm": 12287270.0, "learning_rate": 4.739413680781759e-05, "loss": 0.814, "step": 202 }, { "epoch": 0.44590884129599123, "grad_norm": 6525739.0, "learning_rate": 4.735342019543974e-05, "loss": 0.9559, "step": 203 }, { "epoch": 0.44810543657331137, "grad_norm": 10583591.0, "learning_rate": 4.731270358306189e-05, "loss": 1.0787, "step": 204 }, { "epoch": 0.4503020318506315, "grad_norm": 8354005.5, "learning_rate": 4.727198697068404e-05, "loss": 1.0828, "step": 205 }, { "epoch": 0.4524986271279517, "grad_norm": 10097118.0, "learning_rate": 4.723127035830619e-05, "loss": 1.1523, "step": 206 }, { "epoch": 0.4546952224052718, "grad_norm": 7784980.0, "learning_rate": 4.719055374592834e-05, "loss": 0.9164, "step": 207 }, { "epoch": 0.45689181768259196, "grad_norm": 2736766.75, "learning_rate": 4.714983713355049e-05, "loss": 0.8624, "step": 208 }, { "epoch": 0.45908841295991215, "grad_norm": 9288337.0, "learning_rate": 4.710912052117264e-05, "loss": 0.9828, "step": 209 }, { "epoch": 0.4612850082372323, "grad_norm": 7567845.0, "learning_rate": 4.706840390879479e-05, "loss": 0.9977, "step": 210 }, { "epoch": 0.4634816035145524, "grad_norm": 6014590.5, "learning_rate": 4.702768729641694e-05, "loss": 1.0515, "step": 211 }, { "epoch": 0.4656781987918726, "grad_norm": 4778235.0, "learning_rate": 4.698697068403909e-05, "loss": 0.9762, "step": 212 }, { "epoch": 0.46787479406919275, "grad_norm": 15021668.0, "learning_rate": 4.694625407166124e-05, "loss": 0.7264, "step": 213 }, { "epoch": 0.4700713893465129, "grad_norm": 14335472.0, "learning_rate": 4.690553745928339e-05, "loss": 0.8519, "step": 214 }, { "epoch": 0.47226798462383307, "grad_norm": 16014673.0, "learning_rate": 4.686482084690554e-05, "loss": 1.0334, "step": 215 }, { "epoch": 0.4744645799011532, "grad_norm": 7501092.0, "learning_rate": 4.682410423452769e-05, "loss": 0.8861, "step": 216 }, { "epoch": 0.47666117517847334, "grad_norm": 8373311.5, "learning_rate": 4.678338762214984e-05, "loss": 0.7788, "step": 217 }, { "epoch": 0.47885777045579353, "grad_norm": 4377398.0, "learning_rate": 4.6742671009771985e-05, "loss": 0.9944, "step": 218 }, { "epoch": 0.48105436573311366, "grad_norm": 42197748.0, "learning_rate": 4.6701954397394135e-05, "loss": 0.8525, "step": 219 }, { "epoch": 0.48325096101043385, "grad_norm": 5442925.0, "learning_rate": 4.6661237785016286e-05, "loss": 0.9012, "step": 220 }, { "epoch": 0.485447556287754, "grad_norm": 29140418.0, "learning_rate": 4.6620521172638436e-05, "loss": 0.9017, "step": 221 }, { "epoch": 0.4876441515650741, "grad_norm": 7130746.0, "learning_rate": 4.657980456026059e-05, "loss": 0.7604, "step": 222 }, { "epoch": 0.4898407468423943, "grad_norm": 5992578.0, "learning_rate": 4.653908794788274e-05, "loss": 0.931, "step": 223 }, { "epoch": 0.49203734211971445, "grad_norm": 17682494.0, "learning_rate": 4.649837133550489e-05, "loss": 0.9846, "step": 224 }, { "epoch": 0.4942339373970346, "grad_norm": 5666645.5, "learning_rate": 4.645765472312704e-05, "loss": 0.722, "step": 225 }, { "epoch": 0.4964305326743548, "grad_norm": 11430500.0, "learning_rate": 4.641693811074919e-05, "loss": 1.1372, "step": 226 }, { "epoch": 0.4986271279516749, "grad_norm": 7021308.5, "learning_rate": 4.637622149837134e-05, "loss": 0.7983, "step": 227 }, { "epoch": 0.500823723228995, "grad_norm": 10584033.0, "learning_rate": 4.633550488599348e-05, "loss": 0.835, "step": 228 }, { "epoch": 0.5030203185063152, "grad_norm": 16393499.0, "learning_rate": 4.629478827361563e-05, "loss": 0.8904, "step": 229 }, { "epoch": 0.5052169137836353, "grad_norm": 2858034.5, "learning_rate": 4.6254071661237784e-05, "loss": 1.0295, "step": 230 }, { "epoch": 0.5074135090609555, "grad_norm": 3826194.0, "learning_rate": 4.6213355048859934e-05, "loss": 0.9269, "step": 231 }, { "epoch": 0.5096101043382757, "grad_norm": 3889286.75, "learning_rate": 4.6172638436482085e-05, "loss": 0.882, "step": 232 }, { "epoch": 0.5118066996155958, "grad_norm": 11277812.0, "learning_rate": 4.6131921824104235e-05, "loss": 0.9674, "step": 233 }, { "epoch": 0.514003294892916, "grad_norm": 4758247.0, "learning_rate": 4.6091205211726385e-05, "loss": 0.9393, "step": 234 }, { "epoch": 0.5161998901702362, "grad_norm": 6524251.5, "learning_rate": 4.6050488599348536e-05, "loss": 0.9669, "step": 235 }, { "epoch": 0.5183964854475562, "grad_norm": 6961989.0, "learning_rate": 4.6009771986970686e-05, "loss": 0.8194, "step": 236 }, { "epoch": 0.5205930807248764, "grad_norm": 6262928.5, "learning_rate": 4.596905537459284e-05, "loss": 1.0925, "step": 237 }, { "epoch": 0.5227896760021966, "grad_norm": 11089486.0, "learning_rate": 4.592833876221499e-05, "loss": 0.9925, "step": 238 }, { "epoch": 0.5249862712795168, "grad_norm": 13835292.0, "learning_rate": 4.588762214983713e-05, "loss": 0.8165, "step": 239 }, { "epoch": 0.5271828665568369, "grad_norm": 4045450.0, "learning_rate": 4.584690553745928e-05, "loss": 0.9027, "step": 240 }, { "epoch": 0.5293794618341571, "grad_norm": 6143702.0, "learning_rate": 4.580618892508143e-05, "loss": 0.9063, "step": 241 }, { "epoch": 0.5315760571114773, "grad_norm": 13785544.0, "learning_rate": 4.576547231270358e-05, "loss": 0.8607, "step": 242 }, { "epoch": 0.5337726523887973, "grad_norm": 5596667.5, "learning_rate": 4.572475570032573e-05, "loss": 0.8006, "step": 243 }, { "epoch": 0.5359692476661175, "grad_norm": 8459262.0, "learning_rate": 4.568403908794788e-05, "loss": 0.9847, "step": 244 }, { "epoch": 0.5381658429434377, "grad_norm": 11312232.0, "learning_rate": 4.5643322475570034e-05, "loss": 0.9212, "step": 245 }, { "epoch": 0.5403624382207578, "grad_norm": 7100279.0, "learning_rate": 4.5602605863192184e-05, "loss": 0.8963, "step": 246 }, { "epoch": 0.542559033498078, "grad_norm": 9365418.0, "learning_rate": 4.5561889250814335e-05, "loss": 0.7869, "step": 247 }, { "epoch": 0.5447556287753982, "grad_norm": 21830076.0, "learning_rate": 4.5521172638436485e-05, "loss": 0.9359, "step": 248 }, { "epoch": 0.5469522240527183, "grad_norm": 4518256.5, "learning_rate": 4.548045602605863e-05, "loss": 0.8025, "step": 249 }, { "epoch": 0.5491488193300385, "grad_norm": 9094974.0, "learning_rate": 4.543973941368078e-05, "loss": 0.8602, "step": 250 }, { "epoch": 0.5513454146073586, "grad_norm": 15917480.0, "learning_rate": 4.539902280130293e-05, "loss": 0.9053, "step": 251 }, { "epoch": 0.5535420098846787, "grad_norm": 6764456.0, "learning_rate": 4.535830618892508e-05, "loss": 0.8822, "step": 252 }, { "epoch": 0.5557386051619989, "grad_norm": 4736989.5, "learning_rate": 4.531758957654723e-05, "loss": 0.9517, "step": 253 }, { "epoch": 0.5579352004393191, "grad_norm": 3093551.25, "learning_rate": 4.527687296416938e-05, "loss": 0.8166, "step": 254 }, { "epoch": 0.5601317957166392, "grad_norm": 10738455.0, "learning_rate": 4.523615635179153e-05, "loss": 0.9489, "step": 255 }, { "epoch": 0.5623283909939594, "grad_norm": 7849633.0, "learning_rate": 4.519543973941368e-05, "loss": 0.9002, "step": 256 }, { "epoch": 0.5645249862712796, "grad_norm": 3027068.75, "learning_rate": 4.515472312703583e-05, "loss": 0.7536, "step": 257 }, { "epoch": 0.5667215815485996, "grad_norm": 2546966.0, "learning_rate": 4.511400651465798e-05, "loss": 0.9777, "step": 258 }, { "epoch": 0.5689181768259198, "grad_norm": 22920500.0, "learning_rate": 4.507328990228013e-05, "loss": 0.8078, "step": 259 }, { "epoch": 0.57111477210324, "grad_norm": 3398881.0, "learning_rate": 4.503257328990228e-05, "loss": 0.9109, "step": 260 }, { "epoch": 0.5733113673805601, "grad_norm": 10613815.0, "learning_rate": 4.499185667752443e-05, "loss": 0.9227, "step": 261 }, { "epoch": 0.5755079626578803, "grad_norm": 5875875.5, "learning_rate": 4.495114006514658e-05, "loss": 0.8826, "step": 262 }, { "epoch": 0.5777045579352005, "grad_norm": 9503300.0, "learning_rate": 4.491042345276873e-05, "loss": 0.7595, "step": 263 }, { "epoch": 0.5799011532125206, "grad_norm": 14739306.0, "learning_rate": 4.486970684039088e-05, "loss": 0.9293, "step": 264 }, { "epoch": 0.5820977484898407, "grad_norm": 30256120.0, "learning_rate": 4.482899022801303e-05, "loss": 1.227, "step": 265 }, { "epoch": 0.5842943437671609, "grad_norm": 4365349.5, "learning_rate": 4.478827361563518e-05, "loss": 0.9806, "step": 266 }, { "epoch": 0.586490939044481, "grad_norm": 32792546.0, "learning_rate": 4.474755700325733e-05, "loss": 0.9428, "step": 267 }, { "epoch": 0.5886875343218012, "grad_norm": 8507007.0, "learning_rate": 4.470684039087948e-05, "loss": 0.9196, "step": 268 }, { "epoch": 0.5908841295991214, "grad_norm": 3569782.25, "learning_rate": 4.466612377850163e-05, "loss": 1.0419, "step": 269 }, { "epoch": 0.5930807248764415, "grad_norm": 21276318.0, "learning_rate": 4.462540716612378e-05, "loss": 0.9971, "step": 270 }, { "epoch": 0.5952773201537617, "grad_norm": 6198542.5, "learning_rate": 4.4584690553745925e-05, "loss": 0.8048, "step": 271 }, { "epoch": 0.5974739154310819, "grad_norm": 14407476.0, "learning_rate": 4.4543973941368076e-05, "loss": 0.8347, "step": 272 }, { "epoch": 0.5996705107084019, "grad_norm": 6460475.5, "learning_rate": 4.4503257328990226e-05, "loss": 0.8249, "step": 273 }, { "epoch": 0.6018671059857221, "grad_norm": 1941072.0, "learning_rate": 4.446254071661238e-05, "loss": 0.9565, "step": 274 }, { "epoch": 0.6040637012630423, "grad_norm": 5871742.5, "learning_rate": 4.442182410423453e-05, "loss": 0.9288, "step": 275 }, { "epoch": 0.6062602965403624, "grad_norm": 4966977.0, "learning_rate": 4.438110749185668e-05, "loss": 0.6796, "step": 276 }, { "epoch": 0.6084568918176826, "grad_norm": 7954624.5, "learning_rate": 4.434039087947883e-05, "loss": 0.9218, "step": 277 }, { "epoch": 0.6106534870950028, "grad_norm": 4816445.0, "learning_rate": 4.429967426710098e-05, "loss": 1.042, "step": 278 }, { "epoch": 0.6128500823723229, "grad_norm": 3931317.75, "learning_rate": 4.425895765472313e-05, "loss": 0.8656, "step": 279 }, { "epoch": 0.615046677649643, "grad_norm": 8411777.0, "learning_rate": 4.421824104234528e-05, "loss": 0.7576, "step": 280 }, { "epoch": 0.6172432729269632, "grad_norm": 2942928.25, "learning_rate": 4.417752442996742e-05, "loss": 0.7349, "step": 281 }, { "epoch": 0.6194398682042833, "grad_norm": 9794862.0, "learning_rate": 4.4136807817589574e-05, "loss": 0.7857, "step": 282 }, { "epoch": 0.6216364634816035, "grad_norm": 11497258.0, "learning_rate": 4.4096091205211724e-05, "loss": 0.9657, "step": 283 }, { "epoch": 0.6238330587589237, "grad_norm": 7352097.0, "learning_rate": 4.4055374592833875e-05, "loss": 0.7933, "step": 284 }, { "epoch": 0.6260296540362438, "grad_norm": 4899710.0, "learning_rate": 4.4014657980456025e-05, "loss": 0.8233, "step": 285 }, { "epoch": 0.628226249313564, "grad_norm": 4511158.5, "learning_rate": 4.3973941368078175e-05, "loss": 0.7914, "step": 286 }, { "epoch": 0.6304228445908842, "grad_norm": 20783182.0, "learning_rate": 4.3933224755700326e-05, "loss": 0.9018, "step": 287 }, { "epoch": 0.6326194398682042, "grad_norm": 6158838.0, "learning_rate": 4.3892508143322476e-05, "loss": 0.8031, "step": 288 }, { "epoch": 0.6348160351455244, "grad_norm": 23449790.0, "learning_rate": 4.385179153094463e-05, "loss": 0.7383, "step": 289 }, { "epoch": 0.6370126304228446, "grad_norm": 6257906.0, "learning_rate": 4.381107491856678e-05, "loss": 0.7691, "step": 290 }, { "epoch": 0.6392092257001647, "grad_norm": 4492374.0, "learning_rate": 4.377035830618893e-05, "loss": 0.987, "step": 291 }, { "epoch": 0.6414058209774849, "grad_norm": 17023156.0, "learning_rate": 4.372964169381108e-05, "loss": 0.9187, "step": 292 }, { "epoch": 0.6436024162548051, "grad_norm": 6583997.5, "learning_rate": 4.368892508143323e-05, "loss": 0.8994, "step": 293 }, { "epoch": 0.6457990115321252, "grad_norm": 7650719.5, "learning_rate": 4.364820846905538e-05, "loss": 0.9232, "step": 294 }, { "epoch": 0.6479956068094453, "grad_norm": 12442975.0, "learning_rate": 4.360749185667753e-05, "loss": 0.8611, "step": 295 }, { "epoch": 0.6501922020867655, "grad_norm": 11185547.0, "learning_rate": 4.356677524429968e-05, "loss": 0.8816, "step": 296 }, { "epoch": 0.6523887973640856, "grad_norm": 2894880.5, "learning_rate": 4.352605863192183e-05, "loss": 0.7881, "step": 297 }, { "epoch": 0.6545853926414058, "grad_norm": 8969189.0, "learning_rate": 4.348534201954398e-05, "loss": 0.8634, "step": 298 }, { "epoch": 0.656781987918726, "grad_norm": 6360104.5, "learning_rate": 4.344462540716613e-05, "loss": 0.8561, "step": 299 }, { "epoch": 0.6589785831960461, "grad_norm": 15523615.0, "learning_rate": 4.3403908794788275e-05, "loss": 0.8584, "step": 300 }, { "epoch": 0.6611751784733663, "grad_norm": 18714836.0, "learning_rate": 4.3363192182410426e-05, "loss": 0.9762, "step": 301 }, { "epoch": 0.6633717737506865, "grad_norm": 11480799.0, "learning_rate": 4.3322475570032576e-05, "loss": 0.8664, "step": 302 }, { "epoch": 0.6655683690280065, "grad_norm": 9781725.0, "learning_rate": 4.3281758957654726e-05, "loss": 0.7704, "step": 303 }, { "epoch": 0.6677649643053267, "grad_norm": 13431326.0, "learning_rate": 4.324104234527688e-05, "loss": 0.9839, "step": 304 }, { "epoch": 0.6699615595826469, "grad_norm": 6600915.5, "learning_rate": 4.320032573289903e-05, "loss": 0.9311, "step": 305 }, { "epoch": 0.6721581548599671, "grad_norm": 7580225.0, "learning_rate": 4.315960912052118e-05, "loss": 0.8591, "step": 306 }, { "epoch": 0.6743547501372872, "grad_norm": 6451969.0, "learning_rate": 4.311889250814333e-05, "loss": 0.9092, "step": 307 }, { "epoch": 0.6765513454146074, "grad_norm": 4211526.0, "learning_rate": 4.307817589576548e-05, "loss": 0.9243, "step": 308 }, { "epoch": 0.6787479406919276, "grad_norm": 4953117.5, "learning_rate": 4.303745928338763e-05, "loss": 0.9456, "step": 309 }, { "epoch": 0.6809445359692476, "grad_norm": 10224394.0, "learning_rate": 4.299674267100978e-05, "loss": 0.9441, "step": 310 }, { "epoch": 0.6831411312465678, "grad_norm": 22817490.0, "learning_rate": 4.295602605863192e-05, "loss": 0.8026, "step": 311 }, { "epoch": 0.685337726523888, "grad_norm": 8647461.0, "learning_rate": 4.2915309446254074e-05, "loss": 0.9289, "step": 312 }, { "epoch": 0.6875343218012081, "grad_norm": 16536851.0, "learning_rate": 4.2874592833876224e-05, "loss": 0.8415, "step": 313 }, { "epoch": 0.6897309170785283, "grad_norm": 11579658.0, "learning_rate": 4.2833876221498375e-05, "loss": 0.6157, "step": 314 }, { "epoch": 0.6919275123558485, "grad_norm": 5614166.0, "learning_rate": 4.2793159609120525e-05, "loss": 0.6692, "step": 315 }, { "epoch": 0.6941241076331686, "grad_norm": 16915516.0, "learning_rate": 4.2752442996742676e-05, "loss": 0.9245, "step": 316 }, { "epoch": 0.6963207029104888, "grad_norm": 5778363.0, "learning_rate": 4.2711726384364826e-05, "loss": 1.0392, "step": 317 }, { "epoch": 0.6985172981878089, "grad_norm": 5275346.5, "learning_rate": 4.2671009771986977e-05, "loss": 0.9945, "step": 318 }, { "epoch": 0.700713893465129, "grad_norm": 19296752.0, "learning_rate": 4.263029315960913e-05, "loss": 0.793, "step": 319 }, { "epoch": 0.7029104887424492, "grad_norm": 4581366.5, "learning_rate": 4.258957654723128e-05, "loss": 0.825, "step": 320 }, { "epoch": 0.7051070840197694, "grad_norm": 3308216.25, "learning_rate": 4.254885993485342e-05, "loss": 0.8774, "step": 321 }, { "epoch": 0.7073036792970895, "grad_norm": 7703581.5, "learning_rate": 4.250814332247557e-05, "loss": 0.8918, "step": 322 }, { "epoch": 0.7095002745744097, "grad_norm": 6396188.0, "learning_rate": 4.246742671009772e-05, "loss": 0.7692, "step": 323 }, { "epoch": 0.7116968698517299, "grad_norm": 6489722.5, "learning_rate": 4.242671009771987e-05, "loss": 0.9539, "step": 324 }, { "epoch": 0.7138934651290499, "grad_norm": 22732088.0, "learning_rate": 4.238599348534202e-05, "loss": 0.8366, "step": 325 }, { "epoch": 0.7160900604063701, "grad_norm": 3959640.0, "learning_rate": 4.2345276872964173e-05, "loss": 0.8401, "step": 326 }, { "epoch": 0.7182866556836903, "grad_norm": 4880879.5, "learning_rate": 4.2304560260586324e-05, "loss": 0.8851, "step": 327 }, { "epoch": 0.7204832509610104, "grad_norm": 23368146.0, "learning_rate": 4.2263843648208474e-05, "loss": 1.0028, "step": 328 }, { "epoch": 0.7226798462383306, "grad_norm": 9643974.0, "learning_rate": 4.2223127035830625e-05, "loss": 0.8387, "step": 329 }, { "epoch": 0.7248764415156508, "grad_norm": 3867018.0, "learning_rate": 4.2182410423452775e-05, "loss": 0.8287, "step": 330 }, { "epoch": 0.7270730367929709, "grad_norm": 7396840.5, "learning_rate": 4.2141693811074926e-05, "loss": 1.0076, "step": 331 }, { "epoch": 0.729269632070291, "grad_norm": 4308874.5, "learning_rate": 4.210097719869707e-05, "loss": 0.8102, "step": 332 }, { "epoch": 0.7314662273476112, "grad_norm": 11172095.0, "learning_rate": 4.206026058631922e-05, "loss": 0.9213, "step": 333 }, { "epoch": 0.7336628226249313, "grad_norm": 16709707.0, "learning_rate": 4.201954397394137e-05, "loss": 0.9974, "step": 334 }, { "epoch": 0.7358594179022515, "grad_norm": 9412243.0, "learning_rate": 4.197882736156352e-05, "loss": 0.8619, "step": 335 }, { "epoch": 0.7380560131795717, "grad_norm": 4649884.0, "learning_rate": 4.193811074918567e-05, "loss": 0.8352, "step": 336 }, { "epoch": 0.7402526084568918, "grad_norm": 4795967.0, "learning_rate": 4.189739413680782e-05, "loss": 1.0452, "step": 337 }, { "epoch": 0.742449203734212, "grad_norm": 21222040.0, "learning_rate": 4.185667752442997e-05, "loss": 0.821, "step": 338 }, { "epoch": 0.7446457990115322, "grad_norm": 8605101.0, "learning_rate": 4.181596091205212e-05, "loss": 0.89, "step": 339 }, { "epoch": 0.7468423942888522, "grad_norm": 10975735.0, "learning_rate": 4.177524429967427e-05, "loss": 0.9122, "step": 340 }, { "epoch": 0.7490389895661724, "grad_norm": 5105589.0, "learning_rate": 4.1734527687296424e-05, "loss": 0.81, "step": 341 }, { "epoch": 0.7512355848434926, "grad_norm": 11751471.0, "learning_rate": 4.1693811074918574e-05, "loss": 0.7396, "step": 342 }, { "epoch": 0.7534321801208127, "grad_norm": 8476453.0, "learning_rate": 4.165309446254072e-05, "loss": 0.8053, "step": 343 }, { "epoch": 0.7556287753981329, "grad_norm": 5329847.0, "learning_rate": 4.161237785016287e-05, "loss": 0.8507, "step": 344 }, { "epoch": 0.7578253706754531, "grad_norm": 4554331.0, "learning_rate": 4.157166123778502e-05, "loss": 0.7138, "step": 345 }, { "epoch": 0.7600219659527732, "grad_norm": 9091578.0, "learning_rate": 4.153094462540717e-05, "loss": 0.8644, "step": 346 }, { "epoch": 0.7622185612300933, "grad_norm": 27390056.0, "learning_rate": 4.149022801302932e-05, "loss": 0.8909, "step": 347 }, { "epoch": 0.7644151565074135, "grad_norm": 9285642.0, "learning_rate": 4.144951140065147e-05, "loss": 1.078, "step": 348 }, { "epoch": 0.7666117517847336, "grad_norm": 16344011.0, "learning_rate": 4.140879478827362e-05, "loss": 0.6494, "step": 349 }, { "epoch": 0.7688083470620538, "grad_norm": 5823217.0, "learning_rate": 4.136807817589577e-05, "loss": 1.0222, "step": 350 }, { "epoch": 0.771004942339374, "grad_norm": 24863778.0, "learning_rate": 4.132736156351792e-05, "loss": 1.0486, "step": 351 }, { "epoch": 0.7732015376166941, "grad_norm": 8073097.0, "learning_rate": 4.128664495114007e-05, "loss": 0.8891, "step": 352 }, { "epoch": 0.7753981328940143, "grad_norm": 2275112.75, "learning_rate": 4.1245928338762215e-05, "loss": 0.687, "step": 353 }, { "epoch": 0.7775947281713345, "grad_norm": 29366184.0, "learning_rate": 4.1205211726384366e-05, "loss": 0.8847, "step": 354 }, { "epoch": 0.7797913234486545, "grad_norm": 8535258.0, "learning_rate": 4.1164495114006516e-05, "loss": 0.6604, "step": 355 }, { "epoch": 0.7819879187259747, "grad_norm": 5347244.5, "learning_rate": 4.112377850162867e-05, "loss": 0.7014, "step": 356 }, { "epoch": 0.7841845140032949, "grad_norm": 17013644.0, "learning_rate": 4.108306188925082e-05, "loss": 0.8369, "step": 357 }, { "epoch": 0.786381109280615, "grad_norm": 5505895.0, "learning_rate": 4.104234527687297e-05, "loss": 0.8252, "step": 358 }, { "epoch": 0.7885777045579352, "grad_norm": 47764372.0, "learning_rate": 4.100162866449512e-05, "loss": 0.822, "step": 359 }, { "epoch": 0.7907742998352554, "grad_norm": 4797409.5, "learning_rate": 4.096091205211727e-05, "loss": 0.694, "step": 360 }, { "epoch": 0.7929708951125755, "grad_norm": 5890969.5, "learning_rate": 4.092019543973942e-05, "loss": 0.9714, "step": 361 }, { "epoch": 0.7951674903898956, "grad_norm": 7213682.0, "learning_rate": 4.087947882736157e-05, "loss": 0.884, "step": 362 }, { "epoch": 0.7973640856672158, "grad_norm": 10531884.0, "learning_rate": 4.083876221498372e-05, "loss": 0.7182, "step": 363 }, { "epoch": 0.7995606809445359, "grad_norm": 20195876.0, "learning_rate": 4.0798045602605864e-05, "loss": 0.7789, "step": 364 }, { "epoch": 0.8017572762218561, "grad_norm": 1930498.0, "learning_rate": 4.0757328990228014e-05, "loss": 0.6679, "step": 365 }, { "epoch": 0.8039538714991763, "grad_norm": 7317309.0, "learning_rate": 4.0716612377850165e-05, "loss": 0.925, "step": 366 }, { "epoch": 0.8061504667764964, "grad_norm": 5059402.0, "learning_rate": 4.0675895765472315e-05, "loss": 0.843, "step": 367 }, { "epoch": 0.8083470620538166, "grad_norm": 4036758.0, "learning_rate": 4.0635179153094466e-05, "loss": 0.7063, "step": 368 }, { "epoch": 0.8105436573311368, "grad_norm": 8493004.0, "learning_rate": 4.0594462540716616e-05, "loss": 0.7426, "step": 369 }, { "epoch": 0.8127402526084568, "grad_norm": 7284358.5, "learning_rate": 4.0553745928338767e-05, "loss": 0.6303, "step": 370 }, { "epoch": 0.814936847885777, "grad_norm": 15478656.0, "learning_rate": 4.051302931596092e-05, "loss": 0.9083, "step": 371 }, { "epoch": 0.8171334431630972, "grad_norm": 2402556.75, "learning_rate": 4.047231270358307e-05, "loss": 0.8856, "step": 372 }, { "epoch": 0.8193300384404174, "grad_norm": 5181035.5, "learning_rate": 4.043159609120522e-05, "loss": 0.8156, "step": 373 }, { "epoch": 0.8215266337177375, "grad_norm": 12148220.0, "learning_rate": 4.039087947882736e-05, "loss": 0.8112, "step": 374 }, { "epoch": 0.8237232289950577, "grad_norm": 9295641.0, "learning_rate": 4.035016286644951e-05, "loss": 0.9305, "step": 375 }, { "epoch": 0.8259198242723779, "grad_norm": 3331442.0, "learning_rate": 4.030944625407166e-05, "loss": 0.8619, "step": 376 }, { "epoch": 0.828116419549698, "grad_norm": 11976113.0, "learning_rate": 4.026872964169381e-05, "loss": 0.9033, "step": 377 }, { "epoch": 0.8303130148270181, "grad_norm": 12333790.0, "learning_rate": 4.0228013029315963e-05, "loss": 0.8309, "step": 378 }, { "epoch": 0.8325096101043383, "grad_norm": 18306704.0, "learning_rate": 4.0187296416938114e-05, "loss": 0.6869, "step": 379 }, { "epoch": 0.8347062053816584, "grad_norm": 12500234.0, "learning_rate": 4.0146579804560264e-05, "loss": 0.9265, "step": 380 }, { "epoch": 0.8369028006589786, "grad_norm": 3638457.25, "learning_rate": 4.0105863192182415e-05, "loss": 0.8659, "step": 381 }, { "epoch": 0.8390993959362988, "grad_norm": 39964704.0, "learning_rate": 4.0065146579804565e-05, "loss": 0.8859, "step": 382 }, { "epoch": 0.8412959912136189, "grad_norm": 1792221.5, "learning_rate": 4.0024429967426716e-05, "loss": 0.8887, "step": 383 }, { "epoch": 0.8434925864909391, "grad_norm": 73207232.0, "learning_rate": 3.9983713355048866e-05, "loss": 0.903, "step": 384 }, { "epoch": 0.8456891817682592, "grad_norm": 3413364.75, "learning_rate": 3.994299674267101e-05, "loss": 0.9844, "step": 385 }, { "epoch": 0.8478857770455793, "grad_norm": 8311068.5, "learning_rate": 3.990228013029316e-05, "loss": 0.9515, "step": 386 }, { "epoch": 0.8500823723228995, "grad_norm": 2447756.75, "learning_rate": 3.986156351791531e-05, "loss": 0.7988, "step": 387 }, { "epoch": 0.8522789676002197, "grad_norm": 4686553.5, "learning_rate": 3.982084690553746e-05, "loss": 0.9065, "step": 388 }, { "epoch": 0.8544755628775398, "grad_norm": 5194229.0, "learning_rate": 3.978013029315961e-05, "loss": 0.6724, "step": 389 }, { "epoch": 0.85667215815486, "grad_norm": 3372298.0, "learning_rate": 3.973941368078176e-05, "loss": 0.7658, "step": 390 }, { "epoch": 0.8588687534321802, "grad_norm": 5034177.0, "learning_rate": 3.969869706840391e-05, "loss": 0.8142, "step": 391 }, { "epoch": 0.8610653487095002, "grad_norm": 5968269.5, "learning_rate": 3.965798045602606e-05, "loss": 1.0806, "step": 392 }, { "epoch": 0.8632619439868204, "grad_norm": 8287274.0, "learning_rate": 3.9617263843648214e-05, "loss": 1.1554, "step": 393 }, { "epoch": 0.8654585392641406, "grad_norm": 17271268.0, "learning_rate": 3.9576547231270364e-05, "loss": 0.8774, "step": 394 }, { "epoch": 0.8676551345414607, "grad_norm": 12984409.0, "learning_rate": 3.953583061889251e-05, "loss": 0.7632, "step": 395 }, { "epoch": 0.8698517298187809, "grad_norm": 4521503.5, "learning_rate": 3.949511400651466e-05, "loss": 0.8149, "step": 396 }, { "epoch": 0.8720483250961011, "grad_norm": 3361893.25, "learning_rate": 3.945439739413681e-05, "loss": 0.8628, "step": 397 }, { "epoch": 0.8742449203734212, "grad_norm": 10988202.0, "learning_rate": 3.941368078175896e-05, "loss": 0.7513, "step": 398 }, { "epoch": 0.8764415156507414, "grad_norm": 2900712.25, "learning_rate": 3.937296416938111e-05, "loss": 0.8525, "step": 399 }, { "epoch": 0.8786381109280615, "grad_norm": 5232877.5, "learning_rate": 3.933224755700326e-05, "loss": 0.928, "step": 400 }, { "epoch": 0.8808347062053816, "grad_norm": 53784552.0, "learning_rate": 3.929153094462541e-05, "loss": 0.6787, "step": 401 }, { "epoch": 0.8830313014827018, "grad_norm": 21809352.0, "learning_rate": 3.925081433224756e-05, "loss": 0.8371, "step": 402 }, { "epoch": 0.885227896760022, "grad_norm": 4536432.0, "learning_rate": 3.921009771986971e-05, "loss": 0.9664, "step": 403 }, { "epoch": 0.8874244920373421, "grad_norm": 6359187.5, "learning_rate": 3.916938110749186e-05, "loss": 0.9898, "step": 404 }, { "epoch": 0.8896210873146623, "grad_norm": 7551597.0, "learning_rate": 3.912866449511401e-05, "loss": 0.705, "step": 405 }, { "epoch": 0.8918176825919825, "grad_norm": 6143014.0, "learning_rate": 3.9087947882736156e-05, "loss": 0.7789, "step": 406 }, { "epoch": 0.8940142778693025, "grad_norm": 5476680.0, "learning_rate": 3.9047231270358306e-05, "loss": 0.6997, "step": 407 }, { "epoch": 0.8962108731466227, "grad_norm": 12279581.0, "learning_rate": 3.900651465798046e-05, "loss": 0.7033, "step": 408 }, { "epoch": 0.8984074684239429, "grad_norm": 3615422.5, "learning_rate": 3.896579804560261e-05, "loss": 0.7568, "step": 409 }, { "epoch": 0.900604063701263, "grad_norm": 7852239.5, "learning_rate": 3.892508143322476e-05, "loss": 0.9656, "step": 410 }, { "epoch": 0.9028006589785832, "grad_norm": 6948022.0, "learning_rate": 3.888436482084691e-05, "loss": 0.7215, "step": 411 }, { "epoch": 0.9049972542559034, "grad_norm": 2027938.75, "learning_rate": 3.884364820846906e-05, "loss": 0.743, "step": 412 }, { "epoch": 0.9071938495332235, "grad_norm": 9140767.0, "learning_rate": 3.880293159609121e-05, "loss": 0.8042, "step": 413 }, { "epoch": 0.9093904448105437, "grad_norm": 7185347.0, "learning_rate": 3.876221498371336e-05, "loss": 0.9513, "step": 414 }, { "epoch": 0.9115870400878638, "grad_norm": 15829996.0, "learning_rate": 3.872149837133551e-05, "loss": 0.7824, "step": 415 }, { "epoch": 0.9137836353651839, "grad_norm": 11386417.0, "learning_rate": 3.8680781758957654e-05, "loss": 0.9135, "step": 416 }, { "epoch": 0.9159802306425041, "grad_norm": 10734097.0, "learning_rate": 3.8640065146579804e-05, "loss": 0.85, "step": 417 }, { "epoch": 0.9181768259198243, "grad_norm": 3796483.75, "learning_rate": 3.8599348534201955e-05, "loss": 0.5997, "step": 418 }, { "epoch": 0.9203734211971444, "grad_norm": 11318456.0, "learning_rate": 3.8558631921824105e-05, "loss": 0.7123, "step": 419 }, { "epoch": 0.9225700164744646, "grad_norm": 6741748.5, "learning_rate": 3.8517915309446256e-05, "loss": 0.8653, "step": 420 }, { "epoch": 0.9247666117517848, "grad_norm": 18654650.0, "learning_rate": 3.8477198697068406e-05, "loss": 0.8113, "step": 421 }, { "epoch": 0.9269632070291048, "grad_norm": 6612527.5, "learning_rate": 3.8436482084690556e-05, "loss": 0.8149, "step": 422 }, { "epoch": 0.929159802306425, "grad_norm": 2267397.25, "learning_rate": 3.839576547231271e-05, "loss": 0.9744, "step": 423 }, { "epoch": 0.9313563975837452, "grad_norm": 3449570.5, "learning_rate": 3.835504885993486e-05, "loss": 0.8645, "step": 424 }, { "epoch": 0.9335529928610653, "grad_norm": 2820739.5, "learning_rate": 3.831433224755701e-05, "loss": 0.7483, "step": 425 }, { "epoch": 0.9357495881383855, "grad_norm": 4588689.5, "learning_rate": 3.827361563517916e-05, "loss": 0.7371, "step": 426 }, { "epoch": 0.9379461834157057, "grad_norm": 2832083.25, "learning_rate": 3.82328990228013e-05, "loss": 0.8102, "step": 427 }, { "epoch": 0.9401427786930258, "grad_norm": 15555140.0, "learning_rate": 3.819218241042345e-05, "loss": 0.6949, "step": 428 }, { "epoch": 0.942339373970346, "grad_norm": 7799302.5, "learning_rate": 3.81514657980456e-05, "loss": 0.6668, "step": 429 }, { "epoch": 0.9445359692476661, "grad_norm": 2267858.5, "learning_rate": 3.811074918566775e-05, "loss": 0.8553, "step": 430 }, { "epoch": 0.9467325645249862, "grad_norm": 5764214.0, "learning_rate": 3.8070032573289904e-05, "loss": 0.7074, "step": 431 }, { "epoch": 0.9489291598023064, "grad_norm": 9709723.0, "learning_rate": 3.8029315960912054e-05, "loss": 0.9449, "step": 432 }, { "epoch": 0.9511257550796266, "grad_norm": 3719678.75, "learning_rate": 3.7988599348534205e-05, "loss": 0.8546, "step": 433 }, { "epoch": 0.9533223503569467, "grad_norm": 10701060.0, "learning_rate": 3.7947882736156355e-05, "loss": 0.7293, "step": 434 }, { "epoch": 0.9555189456342669, "grad_norm": 8692140.0, "learning_rate": 3.7907166123778506e-05, "loss": 0.7623, "step": 435 }, { "epoch": 0.9577155409115871, "grad_norm": 1107094.75, "learning_rate": 3.7866449511400656e-05, "loss": 0.6784, "step": 436 }, { "epoch": 0.9599121361889071, "grad_norm": 3959699.5, "learning_rate": 3.7825732899022807e-05, "loss": 0.8284, "step": 437 }, { "epoch": 0.9621087314662273, "grad_norm": 3346395.75, "learning_rate": 3.778501628664495e-05, "loss": 0.7126, "step": 438 }, { "epoch": 0.9643053267435475, "grad_norm": 3931378.75, "learning_rate": 3.77442996742671e-05, "loss": 1.0596, "step": 439 }, { "epoch": 0.9665019220208677, "grad_norm": 2804710.0, "learning_rate": 3.770358306188925e-05, "loss": 0.9038, "step": 440 }, { "epoch": 0.9686985172981878, "grad_norm": 3163888.5, "learning_rate": 3.76628664495114e-05, "loss": 0.7675, "step": 441 }, { "epoch": 0.970895112575508, "grad_norm": 8885306.0, "learning_rate": 3.762214983713355e-05, "loss": 0.7497, "step": 442 }, { "epoch": 0.9730917078528282, "grad_norm": 2620851.5, "learning_rate": 3.75814332247557e-05, "loss": 0.824, "step": 443 }, { "epoch": 0.9752883031301482, "grad_norm": 4316842.0, "learning_rate": 3.754071661237785e-05, "loss": 0.9471, "step": 444 }, { "epoch": 0.9774848984074684, "grad_norm": 3267577.0, "learning_rate": 3.7500000000000003e-05, "loss": 0.9057, "step": 445 }, { "epoch": 0.9796814936847886, "grad_norm": 3122342.75, "learning_rate": 3.7459283387622154e-05, "loss": 0.92, "step": 446 }, { "epoch": 0.9818780889621087, "grad_norm": 19377328.0, "learning_rate": 3.7418566775244304e-05, "loss": 0.9873, "step": 447 }, { "epoch": 0.9840746842394289, "grad_norm": 3318042.25, "learning_rate": 3.737785016286645e-05, "loss": 0.7811, "step": 448 }, { "epoch": 0.9862712795167491, "grad_norm": 5037451.0, "learning_rate": 3.73371335504886e-05, "loss": 0.8696, "step": 449 }, { "epoch": 0.9884678747940692, "grad_norm": 9572285.0, "learning_rate": 3.729641693811075e-05, "loss": 0.8111, "step": 450 }, { "epoch": 0.9906644700713894, "grad_norm": 7891369.5, "learning_rate": 3.72557003257329e-05, "loss": 0.8349, "step": 451 }, { "epoch": 0.9928610653487095, "grad_norm": 3774431.0, "learning_rate": 3.721498371335505e-05, "loss": 0.8568, "step": 452 }, { "epoch": 0.9950576606260296, "grad_norm": 14723319.0, "learning_rate": 3.71742671009772e-05, "loss": 0.901, "step": 453 }, { "epoch": 0.9972542559033498, "grad_norm": 9706890.0, "learning_rate": 3.713355048859935e-05, "loss": 0.6967, "step": 454 }, { "epoch": 0.99945085118067, "grad_norm": 2838797.5, "learning_rate": 3.70928338762215e-05, "loss": 0.777, "step": 455 }, { "epoch": 1.00164744645799, "grad_norm": 8991332.0, "learning_rate": 3.705211726384365e-05, "loss": 0.9142, "step": 456 }, { "epoch": 1.0038440417353103, "grad_norm": 3369319.5, "learning_rate": 3.70114006514658e-05, "loss": 0.8587, "step": 457 }, { "epoch": 1.0060406370126305, "grad_norm": 3280881.75, "learning_rate": 3.697068403908795e-05, "loss": 0.8162, "step": 458 }, { "epoch": 1.0082372322899507, "grad_norm": 5707167.5, "learning_rate": 3.6929967426710096e-05, "loss": 0.7555, "step": 459 }, { "epoch": 1.0104338275672706, "grad_norm": 7033941.5, "learning_rate": 3.688925081433225e-05, "loss": 0.758, "step": 460 }, { "epoch": 1.0126304228445908, "grad_norm": 1708048.25, "learning_rate": 3.68485342019544e-05, "loss": 0.6466, "step": 461 }, { "epoch": 1.014827018121911, "grad_norm": 4216993.5, "learning_rate": 3.680781758957655e-05, "loss": 1.1216, "step": 462 }, { "epoch": 1.0170236133992312, "grad_norm": 11881433.0, "learning_rate": 3.67671009771987e-05, "loss": 0.9845, "step": 463 }, { "epoch": 1.0192202086765514, "grad_norm": 13909134.0, "learning_rate": 3.672638436482085e-05, "loss": 0.775, "step": 464 }, { "epoch": 1.0214168039538716, "grad_norm": 13205813.0, "learning_rate": 3.6685667752443e-05, "loss": 1.0079, "step": 465 }, { "epoch": 1.0236133992311915, "grad_norm": 8669044.0, "learning_rate": 3.664495114006515e-05, "loss": 1.0056, "step": 466 }, { "epoch": 1.0258099945085117, "grad_norm": 4657157.0, "learning_rate": 3.66042345276873e-05, "loss": 0.8558, "step": 467 }, { "epoch": 1.028006589785832, "grad_norm": 9291309.0, "learning_rate": 3.656351791530945e-05, "loss": 0.7488, "step": 468 }, { "epoch": 1.0302031850631521, "grad_norm": 5880277.5, "learning_rate": 3.6522801302931594e-05, "loss": 0.9239, "step": 469 }, { "epoch": 1.0323997803404723, "grad_norm": 1935226.75, "learning_rate": 3.6482084690553745e-05, "loss": 0.9432, "step": 470 }, { "epoch": 1.0345963756177925, "grad_norm": 2122449.75, "learning_rate": 3.6441368078175895e-05, "loss": 0.7108, "step": 471 }, { "epoch": 1.0367929708951125, "grad_norm": 10027997.0, "learning_rate": 3.6400651465798046e-05, "loss": 0.8021, "step": 472 }, { "epoch": 1.0389895661724327, "grad_norm": 6696744.5, "learning_rate": 3.6359934853420196e-05, "loss": 1.0112, "step": 473 }, { "epoch": 1.0411861614497528, "grad_norm": 12777209.0, "learning_rate": 3.6319218241042346e-05, "loss": 0.8327, "step": 474 }, { "epoch": 1.043382756727073, "grad_norm": 3747195.75, "learning_rate": 3.62785016286645e-05, "loss": 0.7982, "step": 475 }, { "epoch": 1.0455793520043932, "grad_norm": 4901371.5, "learning_rate": 3.623778501628665e-05, "loss": 1.0869, "step": 476 }, { "epoch": 1.0477759472817134, "grad_norm": 3937130.0, "learning_rate": 3.61970684039088e-05, "loss": 0.7382, "step": 477 }, { "epoch": 1.0499725425590336, "grad_norm": 3580537.0, "learning_rate": 3.615635179153095e-05, "loss": 0.8539, "step": 478 }, { "epoch": 1.0521691378363536, "grad_norm": 6540120.5, "learning_rate": 3.61156351791531e-05, "loss": 0.9064, "step": 479 }, { "epoch": 1.0543657331136738, "grad_norm": 11335932.0, "learning_rate": 3.607491856677524e-05, "loss": 0.6979, "step": 480 }, { "epoch": 1.056562328390994, "grad_norm": 4903040.0, "learning_rate": 3.603420195439739e-05, "loss": 0.7633, "step": 481 }, { "epoch": 1.0587589236683141, "grad_norm": 3875244.0, "learning_rate": 3.599348534201954e-05, "loss": 0.7527, "step": 482 }, { "epoch": 1.0609555189456343, "grad_norm": 3177955.0, "learning_rate": 3.5952768729641694e-05, "loss": 0.7824, "step": 483 }, { "epoch": 1.0631521142229543, "grad_norm": 10429066.0, "learning_rate": 3.5912052117263844e-05, "loss": 0.8659, "step": 484 }, { "epoch": 1.0653487095002745, "grad_norm": 4362183.5, "learning_rate": 3.5871335504885995e-05, "loss": 0.8113, "step": 485 }, { "epoch": 1.0675453047775947, "grad_norm": 20388008.0, "learning_rate": 3.5830618892508145e-05, "loss": 0.7389, "step": 486 }, { "epoch": 1.0697419000549149, "grad_norm": 1476416.125, "learning_rate": 3.5789902280130296e-05, "loss": 0.7578, "step": 487 }, { "epoch": 1.071938495332235, "grad_norm": 4040277.0, "learning_rate": 3.5749185667752446e-05, "loss": 0.914, "step": 488 }, { "epoch": 1.0741350906095553, "grad_norm": 7079791.5, "learning_rate": 3.5708469055374597e-05, "loss": 0.6894, "step": 489 }, { "epoch": 1.0763316858868754, "grad_norm": 3127609.75, "learning_rate": 3.566775244299674e-05, "loss": 0.6968, "step": 490 }, { "epoch": 1.0785282811641954, "grad_norm": 3809764.5, "learning_rate": 3.562703583061889e-05, "loss": 1.0094, "step": 491 }, { "epoch": 1.0807248764415156, "grad_norm": 3065993.25, "learning_rate": 3.558631921824104e-05, "loss": 0.7706, "step": 492 }, { "epoch": 1.0829214717188358, "grad_norm": 2871166.5, "learning_rate": 3.554560260586319e-05, "loss": 0.9104, "step": 493 }, { "epoch": 1.085118066996156, "grad_norm": 3959990.25, "learning_rate": 3.550488599348534e-05, "loss": 0.6892, "step": 494 }, { "epoch": 1.0873146622734762, "grad_norm": 2142621.25, "learning_rate": 3.546416938110749e-05, "loss": 0.8145, "step": 495 }, { "epoch": 1.0895112575507964, "grad_norm": 9961341.0, "learning_rate": 3.542345276872964e-05, "loss": 0.711, "step": 496 }, { "epoch": 1.0917078528281163, "grad_norm": 3682777.25, "learning_rate": 3.5382736156351793e-05, "loss": 0.7049, "step": 497 }, { "epoch": 1.0939044481054365, "grad_norm": 4144218.75, "learning_rate": 3.5342019543973944e-05, "loss": 0.8052, "step": 498 }, { "epoch": 1.0961010433827567, "grad_norm": 1452248.25, "learning_rate": 3.5301302931596094e-05, "loss": 0.8244, "step": 499 }, { "epoch": 1.098297638660077, "grad_norm": 4761271.5, "learning_rate": 3.5260586319218245e-05, "loss": 0.8815, "step": 500 }, { "epoch": 1.100494233937397, "grad_norm": 9929670.0, "learning_rate": 3.521986970684039e-05, "loss": 0.9855, "step": 501 }, { "epoch": 1.1026908292147173, "grad_norm": 5726086.0, "learning_rate": 3.517915309446254e-05, "loss": 0.6964, "step": 502 }, { "epoch": 1.1048874244920373, "grad_norm": 8523648.0, "learning_rate": 3.513843648208469e-05, "loss": 0.8744, "step": 503 }, { "epoch": 1.1070840197693574, "grad_norm": 2879091.0, "learning_rate": 3.509771986970684e-05, "loss": 0.7042, "step": 504 }, { "epoch": 1.1092806150466776, "grad_norm": 8812327.0, "learning_rate": 3.505700325732899e-05, "loss": 1.0488, "step": 505 }, { "epoch": 1.1114772103239978, "grad_norm": 6325029.5, "learning_rate": 3.501628664495114e-05, "loss": 0.8452, "step": 506 }, { "epoch": 1.113673805601318, "grad_norm": 4794634.5, "learning_rate": 3.497557003257329e-05, "loss": 0.929, "step": 507 }, { "epoch": 1.1158704008786382, "grad_norm": 7767060.0, "learning_rate": 3.493485342019544e-05, "loss": 0.9219, "step": 508 }, { "epoch": 1.1180669961559582, "grad_norm": 4710876.5, "learning_rate": 3.489413680781759e-05, "loss": 0.8301, "step": 509 }, { "epoch": 1.1202635914332784, "grad_norm": 7873708.5, "learning_rate": 3.485342019543974e-05, "loss": 0.9386, "step": 510 }, { "epoch": 1.1224601867105986, "grad_norm": 1433709.625, "learning_rate": 3.481270358306189e-05, "loss": 0.7807, "step": 511 }, { "epoch": 1.1246567819879187, "grad_norm": 4903263.5, "learning_rate": 3.477198697068404e-05, "loss": 0.801, "step": 512 }, { "epoch": 1.126853377265239, "grad_norm": 3248271.75, "learning_rate": 3.473127035830619e-05, "loss": 0.9455, "step": 513 }, { "epoch": 1.1290499725425591, "grad_norm": 5260664.5, "learning_rate": 3.469055374592834e-05, "loss": 0.7725, "step": 514 }, { "epoch": 1.131246567819879, "grad_norm": 5526199.5, "learning_rate": 3.464983713355049e-05, "loss": 0.788, "step": 515 }, { "epoch": 1.1334431630971993, "grad_norm": 6048993.0, "learning_rate": 3.460912052117264e-05, "loss": 0.8122, "step": 516 }, { "epoch": 1.1356397583745195, "grad_norm": 10826018.0, "learning_rate": 3.456840390879479e-05, "loss": 0.8729, "step": 517 }, { "epoch": 1.1378363536518397, "grad_norm": 6635749.0, "learning_rate": 3.452768729641694e-05, "loss": 0.6768, "step": 518 }, { "epoch": 1.1400329489291599, "grad_norm": 2388449.5, "learning_rate": 3.448697068403909e-05, "loss": 0.6119, "step": 519 }, { "epoch": 1.14222954420648, "grad_norm": 5985002.0, "learning_rate": 3.444625407166124e-05, "loss": 0.8503, "step": 520 }, { "epoch": 1.1444261394838002, "grad_norm": 4519367.5, "learning_rate": 3.440553745928339e-05, "loss": 0.6913, "step": 521 }, { "epoch": 1.1466227347611202, "grad_norm": 8761667.0, "learning_rate": 3.4364820846905535e-05, "loss": 0.7784, "step": 522 }, { "epoch": 1.1488193300384404, "grad_norm": 3156495.5, "learning_rate": 3.4324104234527685e-05, "loss": 0.6168, "step": 523 }, { "epoch": 1.1510159253157606, "grad_norm": 8132012.0, "learning_rate": 3.4283387622149836e-05, "loss": 0.9162, "step": 524 }, { "epoch": 1.1532125205930808, "grad_norm": 3212973.0, "learning_rate": 3.4242671009771986e-05, "loss": 0.793, "step": 525 }, { "epoch": 1.155409115870401, "grad_norm": 6705481.5, "learning_rate": 3.4201954397394136e-05, "loss": 1.0598, "step": 526 }, { "epoch": 1.157605711147721, "grad_norm": 12085641.0, "learning_rate": 3.416123778501629e-05, "loss": 0.722, "step": 527 }, { "epoch": 1.1598023064250411, "grad_norm": 4394856.5, "learning_rate": 3.412052117263844e-05, "loss": 0.9344, "step": 528 }, { "epoch": 1.1619989017023613, "grad_norm": 4306202.0, "learning_rate": 3.407980456026059e-05, "loss": 0.6416, "step": 529 }, { "epoch": 1.1641954969796815, "grad_norm": 12400364.0, "learning_rate": 3.403908794788274e-05, "loss": 0.6212, "step": 530 }, { "epoch": 1.1663920922570017, "grad_norm": 2410434.25, "learning_rate": 3.399837133550489e-05, "loss": 0.7261, "step": 531 }, { "epoch": 1.1685886875343219, "grad_norm": 15016189.0, "learning_rate": 3.395765472312704e-05, "loss": 0.9121, "step": 532 }, { "epoch": 1.170785282811642, "grad_norm": 4933925.0, "learning_rate": 3.391693811074918e-05, "loss": 0.9069, "step": 533 }, { "epoch": 1.172981878088962, "grad_norm": 7928337.0, "learning_rate": 3.387622149837133e-05, "loss": 0.9353, "step": 534 }, { "epoch": 1.1751784733662822, "grad_norm": 7127355.0, "learning_rate": 3.3835504885993484e-05, "loss": 0.8018, "step": 535 }, { "epoch": 1.1773750686436024, "grad_norm": 3091171.25, "learning_rate": 3.3794788273615634e-05, "loss": 0.7222, "step": 536 }, { "epoch": 1.1795716639209226, "grad_norm": 2942911.75, "learning_rate": 3.3754071661237785e-05, "loss": 0.7417, "step": 537 }, { "epoch": 1.1817682591982428, "grad_norm": 2696065.5, "learning_rate": 3.3713355048859935e-05, "loss": 0.723, "step": 538 }, { "epoch": 1.1839648544755628, "grad_norm": 4667742.0, "learning_rate": 3.3672638436482086e-05, "loss": 1.0653, "step": 539 }, { "epoch": 1.186161449752883, "grad_norm": 17610622.0, "learning_rate": 3.3631921824104236e-05, "loss": 0.7089, "step": 540 }, { "epoch": 1.1883580450302031, "grad_norm": 9259681.0, "learning_rate": 3.3591205211726387e-05, "loss": 0.8135, "step": 541 }, { "epoch": 1.1905546403075233, "grad_norm": 28884450.0, "learning_rate": 3.355048859934854e-05, "loss": 0.6974, "step": 542 }, { "epoch": 1.1927512355848435, "grad_norm": 7348413.5, "learning_rate": 3.350977198697068e-05, "loss": 0.8565, "step": 543 }, { "epoch": 1.1949478308621637, "grad_norm": 5243163.5, "learning_rate": 3.346905537459283e-05, "loss": 0.7723, "step": 544 }, { "epoch": 1.197144426139484, "grad_norm": 2949088.5, "learning_rate": 3.342833876221498e-05, "loss": 0.9163, "step": 545 }, { "epoch": 1.1993410214168039, "grad_norm": 4970167.5, "learning_rate": 3.338762214983713e-05, "loss": 0.8277, "step": 546 }, { "epoch": 1.201537616694124, "grad_norm": 6460347.0, "learning_rate": 3.334690553745928e-05, "loss": 1.11, "step": 547 }, { "epoch": 1.2037342119714443, "grad_norm": 3999363.25, "learning_rate": 3.330618892508143e-05, "loss": 0.791, "step": 548 }, { "epoch": 1.2059308072487644, "grad_norm": 6791791.0, "learning_rate": 3.3265472312703583e-05, "loss": 1.0351, "step": 549 }, { "epoch": 1.2081274025260846, "grad_norm": 8928464.0, "learning_rate": 3.3224755700325734e-05, "loss": 0.9303, "step": 550 }, { "epoch": 1.2103239978034046, "grad_norm": 10347070.0, "learning_rate": 3.3184039087947884e-05, "loss": 0.7622, "step": 551 }, { "epoch": 1.2125205930807248, "grad_norm": 7738571.0, "learning_rate": 3.3143322475570035e-05, "loss": 0.9321, "step": 552 }, { "epoch": 1.214717188358045, "grad_norm": 4177337.25, "learning_rate": 3.3102605863192185e-05, "loss": 0.8865, "step": 553 }, { "epoch": 1.2169137836353652, "grad_norm": 6501912.0, "learning_rate": 3.306188925081433e-05, "loss": 0.8529, "step": 554 }, { "epoch": 1.2191103789126854, "grad_norm": 6199077.5, "learning_rate": 3.302117263843648e-05, "loss": 0.9066, "step": 555 }, { "epoch": 1.2213069741900056, "grad_norm": 10434565.0, "learning_rate": 3.298045602605863e-05, "loss": 0.7102, "step": 556 }, { "epoch": 1.2235035694673257, "grad_norm": 8458470.0, "learning_rate": 3.293973941368078e-05, "loss": 0.8486, "step": 557 }, { "epoch": 1.2257001647446457, "grad_norm": 9179014.0, "learning_rate": 3.289902280130293e-05, "loss": 0.8103, "step": 558 }, { "epoch": 1.227896760021966, "grad_norm": 10963746.0, "learning_rate": 3.285830618892508e-05, "loss": 1.0779, "step": 559 }, { "epoch": 1.230093355299286, "grad_norm": 5221127.5, "learning_rate": 3.281758957654723e-05, "loss": 0.8836, "step": 560 }, { "epoch": 1.2322899505766063, "grad_norm": 5596379.0, "learning_rate": 3.277687296416938e-05, "loss": 0.8029, "step": 561 }, { "epoch": 1.2344865458539265, "grad_norm": 6915608.0, "learning_rate": 3.273615635179153e-05, "loss": 0.8501, "step": 562 }, { "epoch": 1.2366831411312464, "grad_norm": 4770158.5, "learning_rate": 3.269543973941368e-05, "loss": 1.0326, "step": 563 }, { "epoch": 1.2388797364085666, "grad_norm": 5102892.5, "learning_rate": 3.265472312703583e-05, "loss": 0.7409, "step": 564 }, { "epoch": 1.2410763316858868, "grad_norm": 8208687.5, "learning_rate": 3.261400651465798e-05, "loss": 0.9087, "step": 565 }, { "epoch": 1.243272926963207, "grad_norm": 8665397.0, "learning_rate": 3.257328990228013e-05, "loss": 0.7881, "step": 566 }, { "epoch": 1.2454695222405272, "grad_norm": 7068941.5, "learning_rate": 3.253257328990228e-05, "loss": 0.7957, "step": 567 }, { "epoch": 1.2476661175178474, "grad_norm": 6942070.5, "learning_rate": 3.249185667752443e-05, "loss": 0.8498, "step": 568 }, { "epoch": 1.2498627127951676, "grad_norm": 22498534.0, "learning_rate": 3.245114006514658e-05, "loss": 0.7478, "step": 569 }, { "epoch": 1.2520593080724876, "grad_norm": 3654512.25, "learning_rate": 3.241042345276873e-05, "loss": 0.568, "step": 570 }, { "epoch": 1.2542559033498077, "grad_norm": 3267536.25, "learning_rate": 3.236970684039088e-05, "loss": 0.9269, "step": 571 }, { "epoch": 1.256452498627128, "grad_norm": 6504901.0, "learning_rate": 3.232899022801303e-05, "loss": 0.8469, "step": 572 }, { "epoch": 1.2586490939044481, "grad_norm": 5276869.0, "learning_rate": 3.228827361563518e-05, "loss": 0.807, "step": 573 }, { "epoch": 1.2608456891817683, "grad_norm": 3400798.5, "learning_rate": 3.224755700325733e-05, "loss": 0.8742, "step": 574 }, { "epoch": 1.2630422844590883, "grad_norm": 3815133.25, "learning_rate": 3.2206840390879475e-05, "loss": 0.9133, "step": 575 }, { "epoch": 1.2652388797364087, "grad_norm": 4158373.0, "learning_rate": 3.2166123778501625e-05, "loss": 0.7596, "step": 576 }, { "epoch": 1.2674354750137287, "grad_norm": 4513312.5, "learning_rate": 3.2125407166123776e-05, "loss": 1.047, "step": 577 }, { "epoch": 1.2696320702910489, "grad_norm": 6753482.0, "learning_rate": 3.2084690553745926e-05, "loss": 0.585, "step": 578 }, { "epoch": 1.271828665568369, "grad_norm": 2310095.5, "learning_rate": 3.204397394136808e-05, "loss": 0.6412, "step": 579 }, { "epoch": 1.2740252608456892, "grad_norm": 6548274.5, "learning_rate": 3.200325732899023e-05, "loss": 0.6608, "step": 580 }, { "epoch": 1.2762218561230094, "grad_norm": 9589902.0, "learning_rate": 3.196254071661238e-05, "loss": 0.8523, "step": 581 }, { "epoch": 1.2784184514003294, "grad_norm": 13084248.0, "learning_rate": 3.192182410423453e-05, "loss": 0.7467, "step": 582 }, { "epoch": 1.2806150466776496, "grad_norm": 2030765.0, "learning_rate": 3.188110749185668e-05, "loss": 0.7413, "step": 583 }, { "epoch": 1.2828116419549698, "grad_norm": 75569048.0, "learning_rate": 3.184039087947883e-05, "loss": 0.7575, "step": 584 }, { "epoch": 1.28500823723229, "grad_norm": 4401401.5, "learning_rate": 3.179967426710097e-05, "loss": 0.8569, "step": 585 }, { "epoch": 1.2872048325096102, "grad_norm": 8539482.0, "learning_rate": 3.175895765472312e-05, "loss": 0.791, "step": 586 }, { "epoch": 1.2894014277869301, "grad_norm": 5570772.5, "learning_rate": 3.1718241042345274e-05, "loss": 0.8206, "step": 587 }, { "epoch": 1.2915980230642505, "grad_norm": 6221003.5, "learning_rate": 3.1677524429967424e-05, "loss": 0.8668, "step": 588 }, { "epoch": 1.2937946183415705, "grad_norm": 17060828.0, "learning_rate": 3.1636807817589575e-05, "loss": 0.7171, "step": 589 }, { "epoch": 1.2959912136188907, "grad_norm": 11605545.0, "learning_rate": 3.1596091205211725e-05, "loss": 0.833, "step": 590 }, { "epoch": 1.2981878088962109, "grad_norm": 4201617.5, "learning_rate": 3.1555374592833876e-05, "loss": 0.597, "step": 591 }, { "epoch": 1.300384404173531, "grad_norm": 3455147.0, "learning_rate": 3.1514657980456026e-05, "loss": 0.8203, "step": 592 }, { "epoch": 1.3025809994508513, "grad_norm": 2380072.75, "learning_rate": 3.1473941368078177e-05, "loss": 0.7031, "step": 593 }, { "epoch": 1.3047775947281712, "grad_norm": 4997146.0, "learning_rate": 3.143322475570033e-05, "loss": 0.8235, "step": 594 }, { "epoch": 1.3069741900054914, "grad_norm": 19033640.0, "learning_rate": 3.139250814332248e-05, "loss": 0.9772, "step": 595 }, { "epoch": 1.3091707852828116, "grad_norm": 9704992.0, "learning_rate": 3.135179153094462e-05, "loss": 0.75, "step": 596 }, { "epoch": 1.3113673805601318, "grad_norm": 3873534.5, "learning_rate": 3.131107491856677e-05, "loss": 0.7585, "step": 597 }, { "epoch": 1.313563975837452, "grad_norm": 9189857.0, "learning_rate": 3.127035830618892e-05, "loss": 0.6987, "step": 598 }, { "epoch": 1.3157605711147722, "grad_norm": 2504172.25, "learning_rate": 3.122964169381108e-05, "loss": 0.7699, "step": 599 }, { "epoch": 1.3179571663920924, "grad_norm": 6048699.0, "learning_rate": 3.118892508143323e-05, "loss": 0.982, "step": 600 }, { "epoch": 1.3201537616694123, "grad_norm": 5469415.0, "learning_rate": 3.114820846905538e-05, "loss": 0.8172, "step": 601 }, { "epoch": 1.3223503569467325, "grad_norm": 2182520.25, "learning_rate": 3.110749185667753e-05, "loss": 0.7697, "step": 602 }, { "epoch": 1.3245469522240527, "grad_norm": 2348266.5, "learning_rate": 3.106677524429968e-05, "loss": 0.8726, "step": 603 }, { "epoch": 1.326743547501373, "grad_norm": 5198596.0, "learning_rate": 3.102605863192183e-05, "loss": 0.8597, "step": 604 }, { "epoch": 1.328940142778693, "grad_norm": 3521929.75, "learning_rate": 3.0985342019543975e-05, "loss": 0.9489, "step": 605 }, { "epoch": 1.331136738056013, "grad_norm": 2628773.0, "learning_rate": 3.0944625407166126e-05, "loss": 0.6981, "step": 606 }, { "epoch": 1.3333333333333333, "grad_norm": 5645844.5, "learning_rate": 3.0903908794788276e-05, "loss": 0.622, "step": 607 }, { "epoch": 1.3355299286106534, "grad_norm": 6870191.5, "learning_rate": 3.086319218241043e-05, "loss": 0.8146, "step": 608 }, { "epoch": 1.3377265238879736, "grad_norm": 4888755.5, "learning_rate": 3.082247557003258e-05, "loss": 0.8402, "step": 609 }, { "epoch": 1.3399231191652938, "grad_norm": 4698551.0, "learning_rate": 3.078175895765473e-05, "loss": 0.862, "step": 610 }, { "epoch": 1.342119714442614, "grad_norm": 2885170.75, "learning_rate": 3.074104234527688e-05, "loss": 0.8313, "step": 611 }, { "epoch": 1.3443163097199342, "grad_norm": 7445346.0, "learning_rate": 3.070032573289903e-05, "loss": 0.7226, "step": 612 }, { "epoch": 1.3465129049972542, "grad_norm": 6588745.0, "learning_rate": 3.065960912052118e-05, "loss": 0.8745, "step": 613 }, { "epoch": 1.3487095002745744, "grad_norm": 9090884.0, "learning_rate": 3.061889250814333e-05, "loss": 0.7313, "step": 614 }, { "epoch": 1.3509060955518946, "grad_norm": 6291323.5, "learning_rate": 3.057817589576547e-05, "loss": 0.9667, "step": 615 }, { "epoch": 1.3531026908292147, "grad_norm": 1949737.125, "learning_rate": 3.0537459283387624e-05, "loss": 0.9007, "step": 616 }, { "epoch": 1.355299286106535, "grad_norm": 3990109.5, "learning_rate": 3.0496742671009777e-05, "loss": 0.8265, "step": 617 }, { "epoch": 1.357495881383855, "grad_norm": 9570753.0, "learning_rate": 3.0456026058631924e-05, "loss": 1.0475, "step": 618 }, { "epoch": 1.359692476661175, "grad_norm": 4849979.5, "learning_rate": 3.0415309446254075e-05, "loss": 0.6503, "step": 619 }, { "epoch": 1.3618890719384953, "grad_norm": 3035817.5, "learning_rate": 3.0374592833876225e-05, "loss": 0.7252, "step": 620 }, { "epoch": 1.3640856672158155, "grad_norm": 6198328.5, "learning_rate": 3.0333876221498376e-05, "loss": 1.0213, "step": 621 }, { "epoch": 1.3662822624931357, "grad_norm": 2671334.25, "learning_rate": 3.0293159609120526e-05, "loss": 0.7657, "step": 622 }, { "epoch": 1.3684788577704559, "grad_norm": 24189594.0, "learning_rate": 3.0252442996742673e-05, "loss": 0.6755, "step": 623 }, { "epoch": 1.370675453047776, "grad_norm": 8585594.0, "learning_rate": 3.0211726384364824e-05, "loss": 0.8251, "step": 624 }, { "epoch": 1.372872048325096, "grad_norm": 9984543.0, "learning_rate": 3.0171009771986974e-05, "loss": 0.9082, "step": 625 }, { "epoch": 1.3750686436024162, "grad_norm": 5325456.5, "learning_rate": 3.0130293159609125e-05, "loss": 0.7263, "step": 626 }, { "epoch": 1.3772652388797364, "grad_norm": 11045564.0, "learning_rate": 3.0089576547231275e-05, "loss": 0.738, "step": 627 }, { "epoch": 1.3794618341570566, "grad_norm": 12307985.0, "learning_rate": 3.0048859934853422e-05, "loss": 0.6282, "step": 628 }, { "epoch": 1.3816584294343768, "grad_norm": 5545169.0, "learning_rate": 3.0008143322475573e-05, "loss": 0.796, "step": 629 }, { "epoch": 1.3838550247116967, "grad_norm": 2571234.75, "learning_rate": 2.9967426710097723e-05, "loss": 0.7542, "step": 630 }, { "epoch": 1.3860516199890172, "grad_norm": 3870043.25, "learning_rate": 2.9926710097719874e-05, "loss": 0.7682, "step": 631 }, { "epoch": 1.3882482152663371, "grad_norm": 5821889.5, "learning_rate": 2.9885993485342024e-05, "loss": 0.9469, "step": 632 }, { "epoch": 1.3904448105436573, "grad_norm": 2501907.75, "learning_rate": 2.9845276872964175e-05, "loss": 0.804, "step": 633 }, { "epoch": 1.3926414058209775, "grad_norm": 6694513.5, "learning_rate": 2.980456026058632e-05, "loss": 0.7361, "step": 634 }, { "epoch": 1.3948380010982977, "grad_norm": 4213918.5, "learning_rate": 2.9763843648208472e-05, "loss": 0.7994, "step": 635 }, { "epoch": 1.3970345963756179, "grad_norm": 7716902.0, "learning_rate": 2.9723127035830623e-05, "loss": 0.9123, "step": 636 }, { "epoch": 1.3992311916529379, "grad_norm": 8129986.5, "learning_rate": 2.9682410423452773e-05, "loss": 0.7939, "step": 637 }, { "epoch": 1.401427786930258, "grad_norm": 5230363.5, "learning_rate": 2.9641693811074923e-05, "loss": 0.7561, "step": 638 }, { "epoch": 1.4036243822075782, "grad_norm": 15260243.0, "learning_rate": 2.960097719869707e-05, "loss": 0.7561, "step": 639 }, { "epoch": 1.4058209774848984, "grad_norm": 9165478.0, "learning_rate": 2.956026058631922e-05, "loss": 1.0408, "step": 640 }, { "epoch": 1.4080175727622186, "grad_norm": 2777815.25, "learning_rate": 2.951954397394137e-05, "loss": 0.7271, "step": 641 }, { "epoch": 1.4102141680395386, "grad_norm": 3931786.0, "learning_rate": 2.9478827361563522e-05, "loss": 0.9403, "step": 642 }, { "epoch": 1.412410763316859, "grad_norm": 9556869.0, "learning_rate": 2.9438110749185672e-05, "loss": 0.8069, "step": 643 }, { "epoch": 1.414607358594179, "grad_norm": 3883110.25, "learning_rate": 2.939739413680782e-05, "loss": 0.6838, "step": 644 }, { "epoch": 1.4168039538714992, "grad_norm": 4311441.0, "learning_rate": 2.935667752442997e-05, "loss": 0.7977, "step": 645 }, { "epoch": 1.4190005491488193, "grad_norm": 11853439.0, "learning_rate": 2.931596091205212e-05, "loss": 0.7791, "step": 646 }, { "epoch": 1.4211971444261395, "grad_norm": 5503657.0, "learning_rate": 2.927524429967427e-05, "loss": 0.7692, "step": 647 }, { "epoch": 1.4233937397034597, "grad_norm": 14219026.0, "learning_rate": 2.923452768729642e-05, "loss": 0.7668, "step": 648 }, { "epoch": 1.4255903349807797, "grad_norm": 3119570.5, "learning_rate": 2.9193811074918572e-05, "loss": 0.9098, "step": 649 }, { "epoch": 1.4277869302580999, "grad_norm": 6621255.5, "learning_rate": 2.915309446254072e-05, "loss": 0.6781, "step": 650 }, { "epoch": 1.42998352553542, "grad_norm": 23727058.0, "learning_rate": 2.911237785016287e-05, "loss": 0.7309, "step": 651 }, { "epoch": 1.4321801208127403, "grad_norm": 31922152.0, "learning_rate": 2.907166123778502e-05, "loss": 0.7827, "step": 652 }, { "epoch": 1.4343767160900605, "grad_norm": 6264115.5, "learning_rate": 2.903094462540717e-05, "loss": 0.8391, "step": 653 }, { "epoch": 1.4365733113673804, "grad_norm": 22005284.0, "learning_rate": 2.899022801302932e-05, "loss": 0.764, "step": 654 }, { "epoch": 1.4387699066447008, "grad_norm": 5722619.5, "learning_rate": 2.8949511400651468e-05, "loss": 0.7783, "step": 655 }, { "epoch": 1.4409665019220208, "grad_norm": 6130999.0, "learning_rate": 2.8908794788273618e-05, "loss": 0.8338, "step": 656 }, { "epoch": 1.443163097199341, "grad_norm": 3925606.0, "learning_rate": 2.886807817589577e-05, "loss": 0.9843, "step": 657 }, { "epoch": 1.4453596924766612, "grad_norm": 6457721.0, "learning_rate": 2.882736156351792e-05, "loss": 0.8933, "step": 658 }, { "epoch": 1.4475562877539814, "grad_norm": 4999463.0, "learning_rate": 2.878664495114007e-05, "loss": 0.9811, "step": 659 }, { "epoch": 1.4497528830313016, "grad_norm": 4058591.75, "learning_rate": 2.8745928338762217e-05, "loss": 0.7143, "step": 660 }, { "epoch": 1.4519494783086215, "grad_norm": 3496151.5, "learning_rate": 2.8705211726384367e-05, "loss": 0.7023, "step": 661 }, { "epoch": 1.4541460735859417, "grad_norm": 4563755.5, "learning_rate": 2.8664495114006518e-05, "loss": 0.8082, "step": 662 }, { "epoch": 1.456342668863262, "grad_norm": 2525402.75, "learning_rate": 2.8623778501628668e-05, "loss": 0.8271, "step": 663 }, { "epoch": 1.458539264140582, "grad_norm": 5428222.5, "learning_rate": 2.858306188925082e-05, "loss": 0.7275, "step": 664 }, { "epoch": 1.4607358594179023, "grad_norm": 7972065.5, "learning_rate": 2.8542345276872965e-05, "loss": 0.9807, "step": 665 }, { "epoch": 1.4629324546952225, "grad_norm": 8523531.0, "learning_rate": 2.8501628664495116e-05, "loss": 0.9444, "step": 666 }, { "epoch": 1.4651290499725427, "grad_norm": 4583850.0, "learning_rate": 2.8460912052117266e-05, "loss": 0.753, "step": 667 }, { "epoch": 1.4673256452498626, "grad_norm": 7885099.5, "learning_rate": 2.8420195439739417e-05, "loss": 0.933, "step": 668 }, { "epoch": 1.4695222405271828, "grad_norm": 4484214.5, "learning_rate": 2.8379478827361567e-05, "loss": 0.7095, "step": 669 }, { "epoch": 1.471718835804503, "grad_norm": 7593714.5, "learning_rate": 2.8338762214983718e-05, "loss": 0.8228, "step": 670 }, { "epoch": 1.4739154310818232, "grad_norm": 4510652.5, "learning_rate": 2.8298045602605865e-05, "loss": 0.9625, "step": 671 }, { "epoch": 1.4761120263591434, "grad_norm": 3340290.0, "learning_rate": 2.8257328990228015e-05, "loss": 0.818, "step": 672 }, { "epoch": 1.4783086216364634, "grad_norm": 14126548.0, "learning_rate": 2.8216612377850166e-05, "loss": 0.7649, "step": 673 }, { "epoch": 1.4805052169137836, "grad_norm": 5096661.5, "learning_rate": 2.8175895765472316e-05, "loss": 0.8626, "step": 674 }, { "epoch": 1.4827018121911038, "grad_norm": 18412062.0, "learning_rate": 2.8135179153094467e-05, "loss": 0.7369, "step": 675 }, { "epoch": 1.484898407468424, "grad_norm": 13129880.0, "learning_rate": 2.8094462540716614e-05, "loss": 0.7982, "step": 676 }, { "epoch": 1.4870950027457441, "grad_norm": 3733168.25, "learning_rate": 2.8053745928338764e-05, "loss": 0.7679, "step": 677 }, { "epoch": 1.4892915980230643, "grad_norm": 4655964.0, "learning_rate": 2.8013029315960915e-05, "loss": 0.8692, "step": 678 }, { "epoch": 1.4914881933003845, "grad_norm": 2379629.0, "learning_rate": 2.7972312703583065e-05, "loss": 0.8423, "step": 679 }, { "epoch": 1.4936847885777045, "grad_norm": 14175747.0, "learning_rate": 2.7931596091205216e-05, "loss": 1.0303, "step": 680 }, { "epoch": 1.4958813838550247, "grad_norm": 6592980.0, "learning_rate": 2.7890879478827363e-05, "loss": 0.8141, "step": 681 }, { "epoch": 1.4980779791323449, "grad_norm": 2083189.75, "learning_rate": 2.7850162866449513e-05, "loss": 0.9554, "step": 682 }, { "epoch": 1.500274574409665, "grad_norm": 3264328.0, "learning_rate": 2.7809446254071664e-05, "loss": 0.5439, "step": 683 }, { "epoch": 1.5024711696869852, "grad_norm": 8035727.0, "learning_rate": 2.7768729641693814e-05, "loss": 0.6945, "step": 684 }, { "epoch": 1.5046677649643052, "grad_norm": 24236826.0, "learning_rate": 2.7728013029315965e-05, "loss": 0.9112, "step": 685 }, { "epoch": 1.5068643602416256, "grad_norm": 16203580.0, "learning_rate": 2.768729641693811e-05, "loss": 1.038, "step": 686 }, { "epoch": 1.5090609555189456, "grad_norm": 20527648.0, "learning_rate": 2.7646579804560262e-05, "loss": 0.8679, "step": 687 }, { "epoch": 1.5112575507962658, "grad_norm": 1687664.875, "learning_rate": 2.7605863192182412e-05, "loss": 0.8992, "step": 688 }, { "epoch": 1.513454146073586, "grad_norm": 6100792.5, "learning_rate": 2.7565146579804563e-05, "loss": 0.6928, "step": 689 }, { "epoch": 1.515650741350906, "grad_norm": 1779872.25, "learning_rate": 2.7524429967426713e-05, "loss": 0.7264, "step": 690 }, { "epoch": 1.5178473366282264, "grad_norm": 5877993.0, "learning_rate": 2.7483713355048864e-05, "loss": 0.9053, "step": 691 }, { "epoch": 1.5200439319055463, "grad_norm": 3860265.25, "learning_rate": 2.744299674267101e-05, "loss": 0.8396, "step": 692 }, { "epoch": 1.5222405271828665, "grad_norm": 2222628.5, "learning_rate": 2.740228013029316e-05, "loss": 0.8158, "step": 693 }, { "epoch": 1.5244371224601867, "grad_norm": 6485382.0, "learning_rate": 2.7361563517915312e-05, "loss": 0.7263, "step": 694 }, { "epoch": 1.526633717737507, "grad_norm": 5865175.0, "learning_rate": 2.7320846905537462e-05, "loss": 0.7776, "step": 695 }, { "epoch": 1.528830313014827, "grad_norm": 4594936.5, "learning_rate": 2.7280130293159613e-05, "loss": 0.7184, "step": 696 }, { "epoch": 1.531026908292147, "grad_norm": 1345459.5, "learning_rate": 2.723941368078176e-05, "loss": 0.8008, "step": 697 }, { "epoch": 1.5332235035694675, "grad_norm": 5575137.5, "learning_rate": 2.719869706840391e-05, "loss": 0.7645, "step": 698 }, { "epoch": 1.5354200988467874, "grad_norm": 13658198.0, "learning_rate": 2.715798045602606e-05, "loss": 0.8793, "step": 699 }, { "epoch": 1.5376166941241076, "grad_norm": 6223910.5, "learning_rate": 2.711726384364821e-05, "loss": 0.932, "step": 700 }, { "epoch": 1.5398132894014278, "grad_norm": 3107052.25, "learning_rate": 2.707654723127036e-05, "loss": 0.7443, "step": 701 }, { "epoch": 1.5420098846787478, "grad_norm": 3807637.5, "learning_rate": 2.703583061889251e-05, "loss": 0.7836, "step": 702 }, { "epoch": 1.5442064799560682, "grad_norm": 8423850.0, "learning_rate": 2.699511400651466e-05, "loss": 0.6999, "step": 703 }, { "epoch": 1.5464030752333882, "grad_norm": 3934505.5, "learning_rate": 2.695439739413681e-05, "loss": 0.6424, "step": 704 }, { "epoch": 1.5485996705107083, "grad_norm": 4523469.5, "learning_rate": 2.691368078175896e-05, "loss": 0.8295, "step": 705 }, { "epoch": 1.5507962657880285, "grad_norm": 1500514.125, "learning_rate": 2.687296416938111e-05, "loss": 0.7464, "step": 706 }, { "epoch": 1.5529928610653487, "grad_norm": 4100195.25, "learning_rate": 2.683224755700326e-05, "loss": 0.8612, "step": 707 }, { "epoch": 1.555189456342669, "grad_norm": 6796936.5, "learning_rate": 2.6791530944625408e-05, "loss": 0.9161, "step": 708 }, { "epoch": 1.5573860516199889, "grad_norm": 11311848.0, "learning_rate": 2.675081433224756e-05, "loss": 0.983, "step": 709 }, { "epoch": 1.5595826468973093, "grad_norm": 5327868.0, "learning_rate": 2.671009771986971e-05, "loss": 0.5595, "step": 710 }, { "epoch": 1.5617792421746293, "grad_norm": 6154893.5, "learning_rate": 2.666938110749186e-05, "loss": 0.9245, "step": 711 }, { "epoch": 1.5639758374519495, "grad_norm": 19080946.0, "learning_rate": 2.662866449511401e-05, "loss": 0.6188, "step": 712 }, { "epoch": 1.5661724327292696, "grad_norm": 6060025.0, "learning_rate": 2.6587947882736157e-05, "loss": 0.7311, "step": 713 }, { "epoch": 1.5683690280065898, "grad_norm": 3273898.0, "learning_rate": 2.6547231270358307e-05, "loss": 0.6852, "step": 714 }, { "epoch": 1.57056562328391, "grad_norm": 4101042.25, "learning_rate": 2.6506514657980458e-05, "loss": 0.8457, "step": 715 }, { "epoch": 1.57276221856123, "grad_norm": 7110334.0, "learning_rate": 2.646579804560261e-05, "loss": 0.9775, "step": 716 }, { "epoch": 1.5749588138385504, "grad_norm": 8520013.0, "learning_rate": 2.642508143322476e-05, "loss": 0.8237, "step": 717 }, { "epoch": 1.5771554091158704, "grad_norm": 11102958.0, "learning_rate": 2.6384364820846906e-05, "loss": 0.7317, "step": 718 }, { "epoch": 1.5793520043931906, "grad_norm": 4241118.5, "learning_rate": 2.6343648208469056e-05, "loss": 0.7595, "step": 719 }, { "epoch": 1.5815485996705108, "grad_norm": 2514485.0, "learning_rate": 2.6302931596091207e-05, "loss": 0.7004, "step": 720 }, { "epoch": 1.5837451949478307, "grad_norm": 6694659.0, "learning_rate": 2.6262214983713357e-05, "loss": 0.9085, "step": 721 }, { "epoch": 1.5859417902251511, "grad_norm": 5642870.0, "learning_rate": 2.6221498371335508e-05, "loss": 0.9417, "step": 722 }, { "epoch": 1.588138385502471, "grad_norm": 6868730.5, "learning_rate": 2.6180781758957655e-05, "loss": 0.7569, "step": 723 }, { "epoch": 1.5903349807797913, "grad_norm": 28863254.0, "learning_rate": 2.6140065146579805e-05, "loss": 0.6815, "step": 724 }, { "epoch": 1.5925315760571115, "grad_norm": 12179046.0, "learning_rate": 2.6099348534201956e-05, "loss": 0.8387, "step": 725 }, { "epoch": 1.5947281713344317, "grad_norm": 5193878.0, "learning_rate": 2.6058631921824106e-05, "loss": 0.8018, "step": 726 }, { "epoch": 1.5969247666117519, "grad_norm": 5733950.0, "learning_rate": 2.6017915309446257e-05, "loss": 0.8555, "step": 727 }, { "epoch": 1.5991213618890718, "grad_norm": 3895866.75, "learning_rate": 2.5977198697068407e-05, "loss": 0.8569, "step": 728 }, { "epoch": 1.6013179571663922, "grad_norm": 5921540.0, "learning_rate": 2.5936482084690554e-05, "loss": 0.8158, "step": 729 }, { "epoch": 1.6035145524437122, "grad_norm": 2908632.75, "learning_rate": 2.5895765472312705e-05, "loss": 0.8134, "step": 730 }, { "epoch": 1.6057111477210324, "grad_norm": 9060295.0, "learning_rate": 2.5855048859934855e-05, "loss": 0.5383, "step": 731 }, { "epoch": 1.6079077429983526, "grad_norm": 20096192.0, "learning_rate": 2.5814332247557006e-05, "loss": 0.8042, "step": 732 }, { "epoch": 1.6101043382756726, "grad_norm": 7516175.5, "learning_rate": 2.5773615635179156e-05, "loss": 0.9844, "step": 733 }, { "epoch": 1.612300933552993, "grad_norm": 7557309.5, "learning_rate": 2.5732899022801303e-05, "loss": 0.8605, "step": 734 }, { "epoch": 1.614497528830313, "grad_norm": 7147080.0, "learning_rate": 2.5692182410423454e-05, "loss": 0.9023, "step": 735 }, { "epoch": 1.6166941241076331, "grad_norm": 10267037.0, "learning_rate": 2.5651465798045604e-05, "loss": 0.8682, "step": 736 }, { "epoch": 1.6188907193849533, "grad_norm": 6557988.0, "learning_rate": 2.5610749185667754e-05, "loss": 0.8464, "step": 737 }, { "epoch": 1.6210873146622735, "grad_norm": 7624202.5, "learning_rate": 2.5570032573289905e-05, "loss": 0.824, "step": 738 }, { "epoch": 1.6232839099395937, "grad_norm": 4876888.5, "learning_rate": 2.5529315960912052e-05, "loss": 0.7947, "step": 739 }, { "epoch": 1.6254805052169137, "grad_norm": 11985913.0, "learning_rate": 2.5488599348534202e-05, "loss": 0.8355, "step": 740 }, { "epoch": 1.627677100494234, "grad_norm": 5538659.0, "learning_rate": 2.5447882736156353e-05, "loss": 0.866, "step": 741 }, { "epoch": 1.629873695771554, "grad_norm": 3729755.5, "learning_rate": 2.5407166123778503e-05, "loss": 0.7835, "step": 742 }, { "epoch": 1.6320702910488742, "grad_norm": 3816142.5, "learning_rate": 2.5366449511400654e-05, "loss": 0.7141, "step": 743 }, { "epoch": 1.6342668863261944, "grad_norm": 6218193.0, "learning_rate": 2.5325732899022804e-05, "loss": 0.648, "step": 744 }, { "epoch": 1.6364634816035144, "grad_norm": 5415427.0, "learning_rate": 2.528501628664495e-05, "loss": 0.689, "step": 745 }, { "epoch": 1.6386600768808348, "grad_norm": 3454429.0, "learning_rate": 2.5244299674267102e-05, "loss": 0.5743, "step": 746 }, { "epoch": 1.6408566721581548, "grad_norm": 2751250.75, "learning_rate": 2.5203583061889252e-05, "loss": 0.84, "step": 747 }, { "epoch": 1.643053267435475, "grad_norm": 5668017.0, "learning_rate": 2.5162866449511403e-05, "loss": 0.6741, "step": 748 }, { "epoch": 1.6452498627127952, "grad_norm": 5624318.0, "learning_rate": 2.5122149837133553e-05, "loss": 0.6624, "step": 749 }, { "epoch": 1.6474464579901154, "grad_norm": 8126046.5, "learning_rate": 2.50814332247557e-05, "loss": 0.7089, "step": 750 }, { "epoch": 1.6496430532674355, "grad_norm": 21289788.0, "learning_rate": 2.504071661237785e-05, "loss": 0.826, "step": 751 }, { "epoch": 1.6518396485447555, "grad_norm": 10167456.0, "learning_rate": 2.5e-05, "loss": 0.921, "step": 752 }, { "epoch": 1.654036243822076, "grad_norm": 2350454.25, "learning_rate": 2.495928338762215e-05, "loss": 1.0177, "step": 753 }, { "epoch": 1.656232839099396, "grad_norm": 10937058.0, "learning_rate": 2.4918566775244302e-05, "loss": 0.6545, "step": 754 }, { "epoch": 1.658429434376716, "grad_norm": 4438251.5, "learning_rate": 2.487785016286645e-05, "loss": 0.7499, "step": 755 }, { "epoch": 1.6606260296540363, "grad_norm": 3436862.5, "learning_rate": 2.48371335504886e-05, "loss": 0.7578, "step": 756 }, { "epoch": 1.6628226249313562, "grad_norm": 3206228.0, "learning_rate": 2.479641693811075e-05, "loss": 0.8695, "step": 757 }, { "epoch": 1.6650192202086767, "grad_norm": 3531367.25, "learning_rate": 2.47557003257329e-05, "loss": 0.8113, "step": 758 }, { "epoch": 1.6672158154859966, "grad_norm": 5536763.0, "learning_rate": 2.471498371335505e-05, "loss": 0.9103, "step": 759 }, { "epoch": 1.6694124107633168, "grad_norm": 7986967.0, "learning_rate": 2.4674267100977198e-05, "loss": 0.9087, "step": 760 }, { "epoch": 1.671609006040637, "grad_norm": 5164309.0, "learning_rate": 2.463355048859935e-05, "loss": 0.7603, "step": 761 }, { "epoch": 1.6738056013179572, "grad_norm": 3190495.25, "learning_rate": 2.45928338762215e-05, "loss": 0.8386, "step": 762 }, { "epoch": 1.6760021965952774, "grad_norm": 10057574.0, "learning_rate": 2.455211726384365e-05, "loss": 0.7874, "step": 763 }, { "epoch": 1.6781987918725974, "grad_norm": 5266538.0, "learning_rate": 2.45114006514658e-05, "loss": 0.8817, "step": 764 }, { "epoch": 1.6803953871499178, "grad_norm": 5973479.0, "learning_rate": 2.447068403908795e-05, "loss": 0.8821, "step": 765 }, { "epoch": 1.6825919824272377, "grad_norm": 8201318.0, "learning_rate": 2.4429967426710097e-05, "loss": 0.6769, "step": 766 }, { "epoch": 1.684788577704558, "grad_norm": 18101764.0, "learning_rate": 2.4389250814332248e-05, "loss": 0.8242, "step": 767 }, { "epoch": 1.6869851729818781, "grad_norm": 12085227.0, "learning_rate": 2.43485342019544e-05, "loss": 0.7494, "step": 768 }, { "epoch": 1.689181768259198, "grad_norm": 5681829.0, "learning_rate": 2.430781758957655e-05, "loss": 0.7705, "step": 769 }, { "epoch": 1.6913783635365185, "grad_norm": 6344797.0, "learning_rate": 2.42671009771987e-05, "loss": 0.894, "step": 770 }, { "epoch": 1.6935749588138385, "grad_norm": 2183038.75, "learning_rate": 2.4226384364820846e-05, "loss": 0.9672, "step": 771 }, { "epoch": 1.6957715540911587, "grad_norm": 5401895.0, "learning_rate": 2.4185667752442997e-05, "loss": 0.8957, "step": 772 }, { "epoch": 1.6979681493684788, "grad_norm": 4046054.0, "learning_rate": 2.4144951140065147e-05, "loss": 0.7156, "step": 773 }, { "epoch": 1.700164744645799, "grad_norm": 5847228.0, "learning_rate": 2.4104234527687298e-05, "loss": 0.6937, "step": 774 }, { "epoch": 1.7023613399231192, "grad_norm": 2440498.5, "learning_rate": 2.4063517915309448e-05, "loss": 0.8356, "step": 775 }, { "epoch": 1.7045579352004392, "grad_norm": 5103590.5, "learning_rate": 2.4022801302931595e-05, "loss": 0.6546, "step": 776 }, { "epoch": 1.7067545304777596, "grad_norm": 5184302.0, "learning_rate": 2.3982084690553746e-05, "loss": 0.8807, "step": 777 }, { "epoch": 1.7089511257550796, "grad_norm": 3406332.5, "learning_rate": 2.3941368078175896e-05, "loss": 0.8972, "step": 778 }, { "epoch": 1.7111477210323998, "grad_norm": 8893194.0, "learning_rate": 2.3900651465798047e-05, "loss": 0.791, "step": 779 }, { "epoch": 1.71334431630972, "grad_norm": 28299830.0, "learning_rate": 2.3859934853420197e-05, "loss": 0.8584, "step": 780 }, { "epoch": 1.7155409115870401, "grad_norm": 12228590.0, "learning_rate": 2.3819218241042348e-05, "loss": 0.8137, "step": 781 }, { "epoch": 1.7177375068643603, "grad_norm": 5580711.5, "learning_rate": 2.3778501628664495e-05, "loss": 0.7505, "step": 782 }, { "epoch": 1.7199341021416803, "grad_norm": 5130976.0, "learning_rate": 2.3737785016286645e-05, "loss": 0.8373, "step": 783 }, { "epoch": 1.7221306974190007, "grad_norm": 3462598.25, "learning_rate": 2.3697068403908796e-05, "loss": 0.7679, "step": 784 }, { "epoch": 1.7243272926963207, "grad_norm": 11268651.0, "learning_rate": 2.3656351791530946e-05, "loss": 1.0337, "step": 785 }, { "epoch": 1.7265238879736409, "grad_norm": 7928887.5, "learning_rate": 2.3615635179153096e-05, "loss": 0.6929, "step": 786 }, { "epoch": 1.728720483250961, "grad_norm": 6228416.0, "learning_rate": 2.3574918566775244e-05, "loss": 0.7268, "step": 787 }, { "epoch": 1.730917078528281, "grad_norm": 3173557.0, "learning_rate": 2.3534201954397394e-05, "loss": 0.7475, "step": 788 }, { "epoch": 1.7331136738056014, "grad_norm": 1184311.75, "learning_rate": 2.3493485342019544e-05, "loss": 0.8507, "step": 789 }, { "epoch": 1.7353102690829214, "grad_norm": 11702628.0, "learning_rate": 2.3452768729641695e-05, "loss": 0.9999, "step": 790 }, { "epoch": 1.7375068643602416, "grad_norm": 7791823.0, "learning_rate": 2.3412052117263845e-05, "loss": 0.7915, "step": 791 }, { "epoch": 1.7397034596375618, "grad_norm": 6442666.5, "learning_rate": 2.3371335504885992e-05, "loss": 0.7302, "step": 792 }, { "epoch": 1.741900054914882, "grad_norm": 4107820.75, "learning_rate": 2.3330618892508143e-05, "loss": 0.7632, "step": 793 }, { "epoch": 1.7440966501922022, "grad_norm": 7459289.0, "learning_rate": 2.3289902280130293e-05, "loss": 0.9324, "step": 794 }, { "epoch": 1.7462932454695221, "grad_norm": 3513638.0, "learning_rate": 2.3249185667752444e-05, "loss": 0.7417, "step": 795 }, { "epoch": 1.7484898407468425, "grad_norm": 2193017.75, "learning_rate": 2.3208469055374594e-05, "loss": 0.7389, "step": 796 }, { "epoch": 1.7506864360241625, "grad_norm": 6493457.5, "learning_rate": 2.316775244299674e-05, "loss": 0.9408, "step": 797 }, { "epoch": 1.7528830313014827, "grad_norm": 3541514.5, "learning_rate": 2.3127035830618892e-05, "loss": 0.8113, "step": 798 }, { "epoch": 1.755079626578803, "grad_norm": 3392785.75, "learning_rate": 2.3086319218241042e-05, "loss": 0.9173, "step": 799 }, { "epoch": 1.7572762218561229, "grad_norm": 3237105.75, "learning_rate": 2.3045602605863193e-05, "loss": 0.9845, "step": 800 }, { "epoch": 1.7594728171334433, "grad_norm": 7631320.0, "learning_rate": 2.3004885993485343e-05, "loss": 0.8848, "step": 801 }, { "epoch": 1.7616694124107632, "grad_norm": 7963966.0, "learning_rate": 2.2964169381107494e-05, "loss": 1.0407, "step": 802 }, { "epoch": 1.7638660076880834, "grad_norm": 3586038.5, "learning_rate": 2.292345276872964e-05, "loss": 1.1671, "step": 803 }, { "epoch": 1.7660626029654036, "grad_norm": 9959085.0, "learning_rate": 2.288273615635179e-05, "loss": 1.066, "step": 804 }, { "epoch": 1.7682591982427238, "grad_norm": 3935806.0, "learning_rate": 2.284201954397394e-05, "loss": 0.8171, "step": 805 }, { "epoch": 1.770455793520044, "grad_norm": 4664130.0, "learning_rate": 2.2801302931596092e-05, "loss": 0.7434, "step": 806 }, { "epoch": 1.772652388797364, "grad_norm": 20482446.0, "learning_rate": 2.2760586319218243e-05, "loss": 0.8425, "step": 807 }, { "epoch": 1.7748489840746844, "grad_norm": 19744094.0, "learning_rate": 2.271986970684039e-05, "loss": 0.9734, "step": 808 }, { "epoch": 1.7770455793520044, "grad_norm": 1815085.375, "learning_rate": 2.267915309446254e-05, "loss": 0.9162, "step": 809 }, { "epoch": 1.7792421746293245, "grad_norm": 2197733.5, "learning_rate": 2.263843648208469e-05, "loss": 0.6199, "step": 810 }, { "epoch": 1.7814387699066447, "grad_norm": 7639382.0, "learning_rate": 2.259771986970684e-05, "loss": 0.709, "step": 811 }, { "epoch": 1.7836353651839647, "grad_norm": 4249810.0, "learning_rate": 2.255700325732899e-05, "loss": 0.7909, "step": 812 }, { "epoch": 1.7858319604612851, "grad_norm": 13327131.0, "learning_rate": 2.251628664495114e-05, "loss": 0.9038, "step": 813 }, { "epoch": 1.788028555738605, "grad_norm": 10513216.0, "learning_rate": 2.247557003257329e-05, "loss": 0.8613, "step": 814 }, { "epoch": 1.7902251510159253, "grad_norm": 3392477.0, "learning_rate": 2.243485342019544e-05, "loss": 0.8942, "step": 815 }, { "epoch": 1.7924217462932455, "grad_norm": 3174575.75, "learning_rate": 2.239413680781759e-05, "loss": 0.8228, "step": 816 }, { "epoch": 1.7946183415705657, "grad_norm": 5832620.0, "learning_rate": 2.235342019543974e-05, "loss": 0.7187, "step": 817 }, { "epoch": 1.7968149368478858, "grad_norm": 6747026.0, "learning_rate": 2.231270358306189e-05, "loss": 0.9073, "step": 818 }, { "epoch": 1.7990115321252058, "grad_norm": 16638837.0, "learning_rate": 2.2271986970684038e-05, "loss": 0.7966, "step": 819 }, { "epoch": 1.8012081274025262, "grad_norm": 2785806.0, "learning_rate": 2.223127035830619e-05, "loss": 0.8535, "step": 820 }, { "epoch": 1.8034047226798462, "grad_norm": 2478058.25, "learning_rate": 2.219055374592834e-05, "loss": 0.835, "step": 821 }, { "epoch": 1.8056013179571664, "grad_norm": 2685527.0, "learning_rate": 2.214983713355049e-05, "loss": 0.9156, "step": 822 }, { "epoch": 1.8077979132344866, "grad_norm": 9808544.0, "learning_rate": 2.210912052117264e-05, "loss": 0.8345, "step": 823 }, { "epoch": 1.8099945085118065, "grad_norm": 2587785.25, "learning_rate": 2.2068403908794787e-05, "loss": 0.746, "step": 824 }, { "epoch": 1.812191103789127, "grad_norm": 9017761.0, "learning_rate": 2.2027687296416937e-05, "loss": 0.6969, "step": 825 }, { "epoch": 1.814387699066447, "grad_norm": 5046416.0, "learning_rate": 2.1986970684039088e-05, "loss": 0.6424, "step": 826 }, { "epoch": 1.8165842943437671, "grad_norm": 8755801.0, "learning_rate": 2.1946254071661238e-05, "loss": 0.8777, "step": 827 }, { "epoch": 1.8187808896210873, "grad_norm": 3163870.75, "learning_rate": 2.190553745928339e-05, "loss": 0.7417, "step": 828 }, { "epoch": 1.8209774848984075, "grad_norm": 3030208.0, "learning_rate": 2.186482084690554e-05, "loss": 0.739, "step": 829 }, { "epoch": 1.8231740801757277, "grad_norm": 8461639.0, "learning_rate": 2.182410423452769e-05, "loss": 0.6393, "step": 830 }, { "epoch": 1.8253706754530477, "grad_norm": 4746518.0, "learning_rate": 2.178338762214984e-05, "loss": 0.5455, "step": 831 }, { "epoch": 1.827567270730368, "grad_norm": 15795456.0, "learning_rate": 2.174267100977199e-05, "loss": 0.89, "step": 832 }, { "epoch": 1.829763866007688, "grad_norm": 3614378.75, "learning_rate": 2.1701954397394138e-05, "loss": 0.8752, "step": 833 }, { "epoch": 1.8319604612850082, "grad_norm": 8447181.0, "learning_rate": 2.1661237785016288e-05, "loss": 0.9715, "step": 834 }, { "epoch": 1.8341570565623284, "grad_norm": 10205021.0, "learning_rate": 2.162052117263844e-05, "loss": 0.8767, "step": 835 }, { "epoch": 1.8363536518396484, "grad_norm": 6217193.0, "learning_rate": 2.157980456026059e-05, "loss": 0.6549, "step": 836 }, { "epoch": 1.8385502471169688, "grad_norm": 3376318.75, "learning_rate": 2.153908794788274e-05, "loss": 0.7556, "step": 837 }, { "epoch": 1.8407468423942888, "grad_norm": 3759376.5, "learning_rate": 2.149837133550489e-05, "loss": 0.8284, "step": 838 }, { "epoch": 1.842943437671609, "grad_norm": 5908501.0, "learning_rate": 2.1457654723127037e-05, "loss": 0.7409, "step": 839 }, { "epoch": 1.8451400329489291, "grad_norm": 4526432.5, "learning_rate": 2.1416938110749187e-05, "loss": 0.8601, "step": 840 }, { "epoch": 1.8473366282262493, "grad_norm": 2507999.5, "learning_rate": 2.1376221498371338e-05, "loss": 0.7499, "step": 841 }, { "epoch": 1.8495332235035695, "grad_norm": 15405622.0, "learning_rate": 2.1335504885993488e-05, "loss": 0.8036, "step": 842 }, { "epoch": 1.8517298187808895, "grad_norm": 7135560.5, "learning_rate": 2.129478827361564e-05, "loss": 1.0048, "step": 843 }, { "epoch": 1.85392641405821, "grad_norm": 1439257.625, "learning_rate": 2.1254071661237786e-05, "loss": 0.8738, "step": 844 }, { "epoch": 1.8561230093355299, "grad_norm": 8523340.0, "learning_rate": 2.1213355048859936e-05, "loss": 0.7504, "step": 845 }, { "epoch": 1.85831960461285, "grad_norm": 4241663.0, "learning_rate": 2.1172638436482087e-05, "loss": 0.7513, "step": 846 }, { "epoch": 1.8605161998901703, "grad_norm": 3059999.25, "learning_rate": 2.1131921824104237e-05, "loss": 0.9571, "step": 847 }, { "epoch": 1.8627127951674904, "grad_norm": 7952339.0, "learning_rate": 2.1091205211726388e-05, "loss": 0.7104, "step": 848 }, { "epoch": 1.8649093904448106, "grad_norm": 5356255.0, "learning_rate": 2.1050488599348535e-05, "loss": 0.7163, "step": 849 }, { "epoch": 1.8671059857221306, "grad_norm": 4578026.5, "learning_rate": 2.1009771986970685e-05, "loss": 0.7132, "step": 850 }, { "epoch": 1.869302580999451, "grad_norm": 5333593.5, "learning_rate": 2.0969055374592836e-05, "loss": 0.6446, "step": 851 }, { "epoch": 1.871499176276771, "grad_norm": 15180029.0, "learning_rate": 2.0928338762214986e-05, "loss": 0.8549, "step": 852 }, { "epoch": 1.8736957715540912, "grad_norm": 101827592.0, "learning_rate": 2.0887622149837137e-05, "loss": 1.0132, "step": 853 }, { "epoch": 1.8758923668314114, "grad_norm": 4923558.5, "learning_rate": 2.0846905537459287e-05, "loss": 0.9386, "step": 854 }, { "epoch": 1.8780889621087313, "grad_norm": 1697080.5, "learning_rate": 2.0806188925081434e-05, "loss": 0.7045, "step": 855 }, { "epoch": 1.8802855573860517, "grad_norm": 4818134.5, "learning_rate": 2.0765472312703585e-05, "loss": 0.9553, "step": 856 }, { "epoch": 1.8824821526633717, "grad_norm": 2097056.625, "learning_rate": 2.0724755700325735e-05, "loss": 0.7434, "step": 857 }, { "epoch": 1.884678747940692, "grad_norm": 5853286.0, "learning_rate": 2.0684039087947885e-05, "loss": 0.8532, "step": 858 }, { "epoch": 1.886875343218012, "grad_norm": 4985847.0, "learning_rate": 2.0643322475570036e-05, "loss": 0.7246, "step": 859 }, { "epoch": 1.8890719384953323, "grad_norm": 16572541.0, "learning_rate": 2.0602605863192183e-05, "loss": 0.733, "step": 860 }, { "epoch": 1.8912685337726525, "grad_norm": 15598193.0, "learning_rate": 2.0561889250814333e-05, "loss": 0.9523, "step": 861 }, { "epoch": 1.8934651290499724, "grad_norm": 4152683.0, "learning_rate": 2.0521172638436484e-05, "loss": 0.8332, "step": 862 }, { "epoch": 1.8956617243272929, "grad_norm": 7006605.0, "learning_rate": 2.0480456026058634e-05, "loss": 1.0624, "step": 863 }, { "epoch": 1.8978583196046128, "grad_norm": 9953751.0, "learning_rate": 2.0439739413680785e-05, "loss": 0.693, "step": 864 }, { "epoch": 1.900054914881933, "grad_norm": 4379904.5, "learning_rate": 2.0399022801302932e-05, "loss": 0.8718, "step": 865 }, { "epoch": 1.9022515101592532, "grad_norm": 5631609.5, "learning_rate": 2.0358306188925082e-05, "loss": 0.759, "step": 866 }, { "epoch": 1.9044481054365732, "grad_norm": 9908307.0, "learning_rate": 2.0317589576547233e-05, "loss": 0.7545, "step": 867 }, { "epoch": 1.9066447007138936, "grad_norm": 8658726.0, "learning_rate": 2.0276872964169383e-05, "loss": 1.0098, "step": 868 }, { "epoch": 1.9088412959912135, "grad_norm": 14695729.0, "learning_rate": 2.0236156351791534e-05, "loss": 1.0508, "step": 869 }, { "epoch": 1.9110378912685337, "grad_norm": 1182297.625, "learning_rate": 2.019543973941368e-05, "loss": 0.7606, "step": 870 }, { "epoch": 1.913234486545854, "grad_norm": 5388146.5, "learning_rate": 2.015472312703583e-05, "loss": 0.7966, "step": 871 }, { "epoch": 1.9154310818231741, "grad_norm": 2985355.75, "learning_rate": 2.0114006514657982e-05, "loss": 0.8855, "step": 872 }, { "epoch": 1.9176276771004943, "grad_norm": 9514274.0, "learning_rate": 2.0073289902280132e-05, "loss": 0.9375, "step": 873 }, { "epoch": 1.9198242723778143, "grad_norm": 7851527.5, "learning_rate": 2.0032573289902283e-05, "loss": 1.0427, "step": 874 }, { "epoch": 1.9220208676551347, "grad_norm": 6111110.5, "learning_rate": 1.9991856677524433e-05, "loss": 0.9939, "step": 875 }, { "epoch": 1.9242174629324547, "grad_norm": 8540786.0, "learning_rate": 1.995114006514658e-05, "loss": 0.9236, "step": 876 }, { "epoch": 1.9264140582097748, "grad_norm": 13982911.0, "learning_rate": 1.991042345276873e-05, "loss": 0.7019, "step": 877 }, { "epoch": 1.928610653487095, "grad_norm": 6914903.5, "learning_rate": 1.986970684039088e-05, "loss": 0.8671, "step": 878 }, { "epoch": 1.930807248764415, "grad_norm": 3630576.25, "learning_rate": 1.982899022801303e-05, "loss": 0.8166, "step": 879 }, { "epoch": 1.9330038440417354, "grad_norm": 9739586.0, "learning_rate": 1.9788273615635182e-05, "loss": 0.7128, "step": 880 }, { "epoch": 1.9352004393190554, "grad_norm": 13686392.0, "learning_rate": 1.974755700325733e-05, "loss": 0.8336, "step": 881 }, { "epoch": 1.9373970345963756, "grad_norm": 2490466.5, "learning_rate": 1.970684039087948e-05, "loss": 0.6191, "step": 882 }, { "epoch": 1.9395936298736958, "grad_norm": 7867994.5, "learning_rate": 1.966612377850163e-05, "loss": 0.9043, "step": 883 }, { "epoch": 1.941790225151016, "grad_norm": 1060045.375, "learning_rate": 1.962540716612378e-05, "loss": 0.862, "step": 884 }, { "epoch": 1.9439868204283361, "grad_norm": 4620791.0, "learning_rate": 1.958469055374593e-05, "loss": 1.2265, "step": 885 }, { "epoch": 1.9461834157056561, "grad_norm": 30233016.0, "learning_rate": 1.9543973941368078e-05, "loss": 0.847, "step": 886 }, { "epoch": 1.9483800109829765, "grad_norm": 14562146.0, "learning_rate": 1.950325732899023e-05, "loss": 0.722, "step": 887 }, { "epoch": 1.9505766062602965, "grad_norm": 13033622.0, "learning_rate": 1.946254071661238e-05, "loss": 0.7176, "step": 888 }, { "epoch": 1.9527732015376167, "grad_norm": 2603171.75, "learning_rate": 1.942182410423453e-05, "loss": 0.8335, "step": 889 }, { "epoch": 1.9549697968149369, "grad_norm": 4586265.0, "learning_rate": 1.938110749185668e-05, "loss": 0.8147, "step": 890 }, { "epoch": 1.9571663920922568, "grad_norm": 2921726.0, "learning_rate": 1.9340390879478827e-05, "loss": 0.9424, "step": 891 }, { "epoch": 1.9593629873695773, "grad_norm": 4601482.5, "learning_rate": 1.9299674267100977e-05, "loss": 0.7312, "step": 892 }, { "epoch": 1.9615595826468972, "grad_norm": 2894955.5, "learning_rate": 1.9258957654723128e-05, "loss": 0.7901, "step": 893 }, { "epoch": 1.9637561779242174, "grad_norm": 3576191.75, "learning_rate": 1.9218241042345278e-05, "loss": 0.8681, "step": 894 }, { "epoch": 1.9659527732015376, "grad_norm": 6953558.5, "learning_rate": 1.917752442996743e-05, "loss": 0.6646, "step": 895 }, { "epoch": 1.9681493684788578, "grad_norm": 6109067.5, "learning_rate": 1.913680781758958e-05, "loss": 0.941, "step": 896 }, { "epoch": 1.970345963756178, "grad_norm": 7972811.5, "learning_rate": 1.9096091205211726e-05, "loss": 0.6893, "step": 897 }, { "epoch": 1.972542559033498, "grad_norm": 4189912.5, "learning_rate": 1.9055374592833877e-05, "loss": 0.9582, "step": 898 }, { "epoch": 1.9747391543108184, "grad_norm": 2655152.25, "learning_rate": 1.9014657980456027e-05, "loss": 0.9548, "step": 899 }, { "epoch": 1.9769357495881383, "grad_norm": 2501240.5, "learning_rate": 1.8973941368078178e-05, "loss": 0.7463, "step": 900 }, { "epoch": 1.9791323448654585, "grad_norm": 5079035.0, "learning_rate": 1.8933224755700328e-05, "loss": 0.9083, "step": 901 }, { "epoch": 1.9813289401427787, "grad_norm": 7339288.0, "learning_rate": 1.8892508143322475e-05, "loss": 0.8887, "step": 902 }, { "epoch": 1.9835255354200987, "grad_norm": 4568381.5, "learning_rate": 1.8851791530944626e-05, "loss": 0.7833, "step": 903 }, { "epoch": 1.985722130697419, "grad_norm": 2933216.5, "learning_rate": 1.8811074918566776e-05, "loss": 0.7092, "step": 904 }, { "epoch": 1.987918725974739, "grad_norm": 3911176.25, "learning_rate": 1.8770358306188927e-05, "loss": 0.6684, "step": 905 }, { "epoch": 1.9901153212520593, "grad_norm": 4835380.0, "learning_rate": 1.8729641693811077e-05, "loss": 0.9137, "step": 906 }, { "epoch": 1.9923119165293794, "grad_norm": 3693694.25, "learning_rate": 1.8688925081433224e-05, "loss": 0.6428, "step": 907 }, { "epoch": 1.9945085118066996, "grad_norm": 5489355.0, "learning_rate": 1.8648208469055375e-05, "loss": 0.66, "step": 908 }, { "epoch": 1.9967051070840198, "grad_norm": 3017182.75, "learning_rate": 1.8607491856677525e-05, "loss": 0.6366, "step": 909 }, { "epoch": 1.9989017023613398, "grad_norm": 8326379.5, "learning_rate": 1.8566775244299675e-05, "loss": 0.8199, "step": 910 }, { "epoch": 2.00109829763866, "grad_norm": 3494578.5, "learning_rate": 1.8526058631921826e-05, "loss": 0.9508, "step": 911 }, { "epoch": 2.00329489291598, "grad_norm": 5448883.5, "learning_rate": 1.8485342019543976e-05, "loss": 0.7703, "step": 912 }, { "epoch": 2.0054914881933006, "grad_norm": 7030463.5, "learning_rate": 1.8444625407166123e-05, "loss": 0.6459, "step": 913 }, { "epoch": 2.0076880834706206, "grad_norm": 2744755.25, "learning_rate": 1.8403908794788274e-05, "loss": 0.7983, "step": 914 }, { "epoch": 2.0098846787479405, "grad_norm": 6295109.5, "learning_rate": 1.8363192182410424e-05, "loss": 0.7915, "step": 915 }, { "epoch": 2.012081274025261, "grad_norm": 7412736.5, "learning_rate": 1.8322475570032575e-05, "loss": 1.0068, "step": 916 }, { "epoch": 2.014277869302581, "grad_norm": 7046878.5, "learning_rate": 1.8281758957654725e-05, "loss": 0.9608, "step": 917 }, { "epoch": 2.0164744645799013, "grad_norm": 2421389.25, "learning_rate": 1.8241042345276872e-05, "loss": 0.9851, "step": 918 }, { "epoch": 2.0186710598572213, "grad_norm": 3188937.0, "learning_rate": 1.8200325732899023e-05, "loss": 0.7339, "step": 919 }, { "epoch": 2.0208676551345413, "grad_norm": 13980595.0, "learning_rate": 1.8159609120521173e-05, "loss": 0.7492, "step": 920 }, { "epoch": 2.0230642504118617, "grad_norm": 6551893.0, "learning_rate": 1.8118892508143324e-05, "loss": 0.7184, "step": 921 }, { "epoch": 2.0252608456891816, "grad_norm": 9451679.0, "learning_rate": 1.8078175895765474e-05, "loss": 0.823, "step": 922 }, { "epoch": 2.027457440966502, "grad_norm": 8318042.5, "learning_rate": 1.803745928338762e-05, "loss": 0.8667, "step": 923 }, { "epoch": 2.029654036243822, "grad_norm": 4752105.5, "learning_rate": 1.799674267100977e-05, "loss": 0.9472, "step": 924 }, { "epoch": 2.0318506315211424, "grad_norm": 2738853.0, "learning_rate": 1.7956026058631922e-05, "loss": 0.7482, "step": 925 }, { "epoch": 2.0340472267984624, "grad_norm": 4173811.75, "learning_rate": 1.7915309446254073e-05, "loss": 0.8398, "step": 926 }, { "epoch": 2.0362438220757824, "grad_norm": 5078139.0, "learning_rate": 1.7874592833876223e-05, "loss": 0.9674, "step": 927 }, { "epoch": 2.0384404173531028, "grad_norm": 8356471.5, "learning_rate": 1.783387622149837e-05, "loss": 0.771, "step": 928 }, { "epoch": 2.0406370126304227, "grad_norm": 4511216.0, "learning_rate": 1.779315960912052e-05, "loss": 0.6947, "step": 929 }, { "epoch": 2.042833607907743, "grad_norm": 4675797.0, "learning_rate": 1.775244299674267e-05, "loss": 0.708, "step": 930 }, { "epoch": 2.045030203185063, "grad_norm": 6009520.0, "learning_rate": 1.771172638436482e-05, "loss": 0.7586, "step": 931 }, { "epoch": 2.047226798462383, "grad_norm": 2811151.25, "learning_rate": 1.7671009771986972e-05, "loss": 0.8171, "step": 932 }, { "epoch": 2.0494233937397035, "grad_norm": 3328623.75, "learning_rate": 1.7630293159609122e-05, "loss": 0.8007, "step": 933 }, { "epoch": 2.0516199890170235, "grad_norm": 9727441.0, "learning_rate": 1.758957654723127e-05, "loss": 0.7732, "step": 934 }, { "epoch": 2.053816584294344, "grad_norm": 1385656.5, "learning_rate": 1.754885993485342e-05, "loss": 0.7453, "step": 935 }, { "epoch": 2.056013179571664, "grad_norm": 3496263.0, "learning_rate": 1.750814332247557e-05, "loss": 0.939, "step": 936 }, { "epoch": 2.0582097748489843, "grad_norm": 3648535.25, "learning_rate": 1.746742671009772e-05, "loss": 0.8441, "step": 937 }, { "epoch": 2.0604063701263042, "grad_norm": 3279550.0, "learning_rate": 1.742671009771987e-05, "loss": 0.6513, "step": 938 }, { "epoch": 2.062602965403624, "grad_norm": 6393275.0, "learning_rate": 1.738599348534202e-05, "loss": 1.0011, "step": 939 }, { "epoch": 2.0647995606809446, "grad_norm": 7883316.5, "learning_rate": 1.734527687296417e-05, "loss": 0.7555, "step": 940 }, { "epoch": 2.0669961559582646, "grad_norm": 7346948.0, "learning_rate": 1.730456026058632e-05, "loss": 0.8227, "step": 941 }, { "epoch": 2.069192751235585, "grad_norm": 2171400.25, "learning_rate": 1.726384364820847e-05, "loss": 0.6303, "step": 942 }, { "epoch": 2.071389346512905, "grad_norm": 3247883.0, "learning_rate": 1.722312703583062e-05, "loss": 0.8052, "step": 943 }, { "epoch": 2.073585941790225, "grad_norm": 2736027.25, "learning_rate": 1.7182410423452767e-05, "loss": 0.719, "step": 944 }, { "epoch": 2.0757825370675453, "grad_norm": 4661686.0, "learning_rate": 1.7141693811074918e-05, "loss": 0.8049, "step": 945 }, { "epoch": 2.0779791323448653, "grad_norm": 6777251.5, "learning_rate": 1.7100977198697068e-05, "loss": 0.802, "step": 946 }, { "epoch": 2.0801757276221857, "grad_norm": 12720982.0, "learning_rate": 1.706026058631922e-05, "loss": 0.6167, "step": 947 }, { "epoch": 2.0823723228995057, "grad_norm": 3066535.25, "learning_rate": 1.701954397394137e-05, "loss": 0.7019, "step": 948 }, { "epoch": 2.084568918176826, "grad_norm": 8993991.0, "learning_rate": 1.697882736156352e-05, "loss": 0.9715, "step": 949 }, { "epoch": 2.086765513454146, "grad_norm": 1970840.25, "learning_rate": 1.6938110749185667e-05, "loss": 0.8013, "step": 950 }, { "epoch": 2.088962108731466, "grad_norm": 14743399.0, "learning_rate": 1.6897394136807817e-05, "loss": 0.8416, "step": 951 }, { "epoch": 2.0911587040087865, "grad_norm": 7608622.5, "learning_rate": 1.6856677524429968e-05, "loss": 0.8608, "step": 952 }, { "epoch": 2.0933552992861064, "grad_norm": 6201172.5, "learning_rate": 1.6815960912052118e-05, "loss": 0.9443, "step": 953 }, { "epoch": 2.095551894563427, "grad_norm": 2341765.75, "learning_rate": 1.677524429967427e-05, "loss": 0.6956, "step": 954 }, { "epoch": 2.097748489840747, "grad_norm": 3916520.5, "learning_rate": 1.6734527687296416e-05, "loss": 0.902, "step": 955 }, { "epoch": 2.099945085118067, "grad_norm": 33477604.0, "learning_rate": 1.6693811074918566e-05, "loss": 0.7859, "step": 956 }, { "epoch": 2.102141680395387, "grad_norm": 3646212.75, "learning_rate": 1.6653094462540716e-05, "loss": 0.8155, "step": 957 }, { "epoch": 2.104338275672707, "grad_norm": 10441493.0, "learning_rate": 1.6612377850162867e-05, "loss": 0.7251, "step": 958 }, { "epoch": 2.1065348709500276, "grad_norm": 11191649.0, "learning_rate": 1.6571661237785017e-05, "loss": 0.8062, "step": 959 }, { "epoch": 2.1087314662273475, "grad_norm": 2551275.0, "learning_rate": 1.6530944625407164e-05, "loss": 0.7731, "step": 960 }, { "epoch": 2.110928061504668, "grad_norm": 16948220.0, "learning_rate": 1.6490228013029315e-05, "loss": 1.0092, "step": 961 }, { "epoch": 2.113124656781988, "grad_norm": 7067645.5, "learning_rate": 1.6449511400651465e-05, "loss": 0.7808, "step": 962 }, { "epoch": 2.115321252059308, "grad_norm": 7295557.5, "learning_rate": 1.6408794788273616e-05, "loss": 0.6985, "step": 963 }, { "epoch": 2.1175178473366283, "grad_norm": 4207166.5, "learning_rate": 1.6368078175895766e-05, "loss": 0.7756, "step": 964 }, { "epoch": 2.1197144426139483, "grad_norm": 3023842.25, "learning_rate": 1.6327361563517913e-05, "loss": 0.7299, "step": 965 }, { "epoch": 2.1219110378912687, "grad_norm": 2193285.0, "learning_rate": 1.6286644951140064e-05, "loss": 0.7421, "step": 966 }, { "epoch": 2.1241076331685886, "grad_norm": 4508207.5, "learning_rate": 1.6245928338762214e-05, "loss": 0.7343, "step": 967 }, { "epoch": 2.1263042284459086, "grad_norm": 3777765.5, "learning_rate": 1.6205211726384365e-05, "loss": 0.7332, "step": 968 }, { "epoch": 2.128500823723229, "grad_norm": 6247828.0, "learning_rate": 1.6164495114006515e-05, "loss": 0.7085, "step": 969 }, { "epoch": 2.130697419000549, "grad_norm": 2008213.25, "learning_rate": 1.6123778501628666e-05, "loss": 0.8822, "step": 970 }, { "epoch": 2.1328940142778694, "grad_norm": 5647030.0, "learning_rate": 1.6083061889250813e-05, "loss": 0.7882, "step": 971 }, { "epoch": 2.1350906095551894, "grad_norm": 3650408.5, "learning_rate": 1.6042345276872963e-05, "loss": 0.9182, "step": 972 }, { "epoch": 2.13728720483251, "grad_norm": 4611719.5, "learning_rate": 1.6001628664495114e-05, "loss": 0.8011, "step": 973 }, { "epoch": 2.1394838001098297, "grad_norm": 2211410.75, "learning_rate": 1.5960912052117264e-05, "loss": 0.8519, "step": 974 }, { "epoch": 2.1416803953871497, "grad_norm": 8847583.0, "learning_rate": 1.5920195439739415e-05, "loss": 0.8471, "step": 975 }, { "epoch": 2.14387699066447, "grad_norm": 2778647.75, "learning_rate": 1.587947882736156e-05, "loss": 0.8439, "step": 976 }, { "epoch": 2.14607358594179, "grad_norm": 2086659.75, "learning_rate": 1.5838762214983712e-05, "loss": 0.7953, "step": 977 }, { "epoch": 2.1482701812191105, "grad_norm": 3368784.0, "learning_rate": 1.5798045602605863e-05, "loss": 0.7257, "step": 978 }, { "epoch": 2.1504667764964305, "grad_norm": 6053738.5, "learning_rate": 1.5757328990228013e-05, "loss": 0.7569, "step": 979 }, { "epoch": 2.152663371773751, "grad_norm": 10296760.0, "learning_rate": 1.5716612377850163e-05, "loss": 0.8769, "step": 980 }, { "epoch": 2.154859967051071, "grad_norm": 7475247.0, "learning_rate": 1.567589576547231e-05, "loss": 0.6506, "step": 981 }, { "epoch": 2.157056562328391, "grad_norm": 3973627.0, "learning_rate": 1.563517915309446e-05, "loss": 0.6861, "step": 982 }, { "epoch": 2.1592531576057112, "grad_norm": 4072698.75, "learning_rate": 1.5594462540716615e-05, "loss": 0.9369, "step": 983 }, { "epoch": 2.161449752883031, "grad_norm": 3552805.0, "learning_rate": 1.5553745928338765e-05, "loss": 0.7046, "step": 984 }, { "epoch": 2.1636463481603516, "grad_norm": 8278152.0, "learning_rate": 1.5513029315960916e-05, "loss": 0.718, "step": 985 }, { "epoch": 2.1658429434376716, "grad_norm": 4112720.25, "learning_rate": 1.5472312703583063e-05, "loss": 0.7558, "step": 986 }, { "epoch": 2.168039538714992, "grad_norm": 2435579.0, "learning_rate": 1.5431596091205213e-05, "loss": 0.8916, "step": 987 }, { "epoch": 2.170236133992312, "grad_norm": 3046778.5, "learning_rate": 1.5390879478827364e-05, "loss": 0.6922, "step": 988 }, { "epoch": 2.172432729269632, "grad_norm": 6671465.0, "learning_rate": 1.5350162866449514e-05, "loss": 0.8092, "step": 989 }, { "epoch": 2.1746293245469523, "grad_norm": 5192924.0, "learning_rate": 1.5309446254071665e-05, "loss": 0.9859, "step": 990 }, { "epoch": 2.1768259198242723, "grad_norm": 9703150.0, "learning_rate": 1.5268729641693812e-05, "loss": 0.7553, "step": 991 }, { "epoch": 2.1790225151015927, "grad_norm": 3975486.0, "learning_rate": 1.5228013029315962e-05, "loss": 0.6931, "step": 992 }, { "epoch": 2.1812191103789127, "grad_norm": 9991519.0, "learning_rate": 1.5187296416938113e-05, "loss": 0.8509, "step": 993 }, { "epoch": 2.1834157056562327, "grad_norm": 1893980.5, "learning_rate": 1.5146579804560263e-05, "loss": 0.7236, "step": 994 }, { "epoch": 2.185612300933553, "grad_norm": 4502999.5, "learning_rate": 1.5105863192182412e-05, "loss": 0.728, "step": 995 }, { "epoch": 2.187808896210873, "grad_norm": 4736523.0, "learning_rate": 1.5065146579804562e-05, "loss": 0.9339, "step": 996 }, { "epoch": 2.1900054914881935, "grad_norm": 8523063.0, "learning_rate": 1.5024429967426711e-05, "loss": 0.85, "step": 997 }, { "epoch": 2.1922020867655134, "grad_norm": 1488145.75, "learning_rate": 1.4983713355048862e-05, "loss": 0.8031, "step": 998 }, { "epoch": 2.1943986820428334, "grad_norm": 7985575.5, "learning_rate": 1.4942996742671012e-05, "loss": 0.9585, "step": 999 }, { "epoch": 2.196595277320154, "grad_norm": 7183530.0, "learning_rate": 1.490228013029316e-05, "loss": 0.7347, "step": 1000 }, { "epoch": 2.1987918725974738, "grad_norm": 3536959.0, "learning_rate": 1.4861563517915311e-05, "loss": 0.5781, "step": 1001 }, { "epoch": 2.200988467874794, "grad_norm": 2099161.25, "learning_rate": 1.4820846905537462e-05, "loss": 0.7237, "step": 1002 }, { "epoch": 2.203185063152114, "grad_norm": 4888102.5, "learning_rate": 1.478013029315961e-05, "loss": 0.8234, "step": 1003 }, { "epoch": 2.2053816584294346, "grad_norm": 4817862.5, "learning_rate": 1.4739413680781761e-05, "loss": 0.7021, "step": 1004 }, { "epoch": 2.2075782537067545, "grad_norm": 1827390.375, "learning_rate": 1.469869706840391e-05, "loss": 0.7217, "step": 1005 }, { "epoch": 2.2097748489840745, "grad_norm": 3758607.0, "learning_rate": 1.465798045602606e-05, "loss": 0.8257, "step": 1006 }, { "epoch": 2.211971444261395, "grad_norm": 12939548.0, "learning_rate": 1.461726384364821e-05, "loss": 0.6608, "step": 1007 }, { "epoch": 2.214168039538715, "grad_norm": 18534588.0, "learning_rate": 1.457654723127036e-05, "loss": 1.042, "step": 1008 }, { "epoch": 2.2163646348160353, "grad_norm": 1720359.375, "learning_rate": 1.453583061889251e-05, "loss": 0.8151, "step": 1009 }, { "epoch": 2.2185612300933553, "grad_norm": 29221916.0, "learning_rate": 1.449511400651466e-05, "loss": 0.7917, "step": 1010 }, { "epoch": 2.2207578253706757, "grad_norm": 1805597.625, "learning_rate": 1.4454397394136809e-05, "loss": 0.7776, "step": 1011 }, { "epoch": 2.2229544206479956, "grad_norm": 46326092.0, "learning_rate": 1.441368078175896e-05, "loss": 1.0515, "step": 1012 }, { "epoch": 2.2251510159253156, "grad_norm": 10872767.0, "learning_rate": 1.4372964169381108e-05, "loss": 0.9426, "step": 1013 }, { "epoch": 2.227347611202636, "grad_norm": 7785878.0, "learning_rate": 1.4332247557003259e-05, "loss": 0.9596, "step": 1014 }, { "epoch": 2.229544206479956, "grad_norm": 36649992.0, "learning_rate": 1.429153094462541e-05, "loss": 0.9342, "step": 1015 }, { "epoch": 2.2317408017572764, "grad_norm": 4091651.5, "learning_rate": 1.4250814332247558e-05, "loss": 0.8276, "step": 1016 }, { "epoch": 2.2339373970345964, "grad_norm": 4844306.0, "learning_rate": 1.4210097719869708e-05, "loss": 0.9671, "step": 1017 }, { "epoch": 2.2361339923119163, "grad_norm": 7422064.5, "learning_rate": 1.4169381107491859e-05, "loss": 1.0135, "step": 1018 }, { "epoch": 2.2383305875892368, "grad_norm": 5165876.0, "learning_rate": 1.4128664495114008e-05, "loss": 0.8735, "step": 1019 }, { "epoch": 2.2405271828665567, "grad_norm": 7189243.5, "learning_rate": 1.4087947882736158e-05, "loss": 0.8116, "step": 1020 }, { "epoch": 2.242723778143877, "grad_norm": 8663355.0, "learning_rate": 1.4047231270358307e-05, "loss": 0.7846, "step": 1021 }, { "epoch": 2.244920373421197, "grad_norm": 4290221.0, "learning_rate": 1.4006514657980457e-05, "loss": 0.6436, "step": 1022 }, { "epoch": 2.247116968698517, "grad_norm": 4210670.0, "learning_rate": 1.3965798045602608e-05, "loss": 0.8688, "step": 1023 }, { "epoch": 2.2493135639758375, "grad_norm": 16253383.0, "learning_rate": 1.3925081433224757e-05, "loss": 0.7745, "step": 1024 }, { "epoch": 2.2515101592531575, "grad_norm": 5735170.5, "learning_rate": 1.3884364820846907e-05, "loss": 0.7289, "step": 1025 }, { "epoch": 2.253706754530478, "grad_norm": 3034424.0, "learning_rate": 1.3843648208469056e-05, "loss": 0.7259, "step": 1026 }, { "epoch": 2.255903349807798, "grad_norm": 3135050.0, "learning_rate": 1.3802931596091206e-05, "loss": 0.8013, "step": 1027 }, { "epoch": 2.2580999450851182, "grad_norm": 7517336.0, "learning_rate": 1.3762214983713357e-05, "loss": 0.897, "step": 1028 }, { "epoch": 2.260296540362438, "grad_norm": 5234165.5, "learning_rate": 1.3721498371335505e-05, "loss": 0.8335, "step": 1029 }, { "epoch": 2.262493135639758, "grad_norm": 2984769.5, "learning_rate": 1.3680781758957656e-05, "loss": 0.6761, "step": 1030 }, { "epoch": 2.2646897309170786, "grad_norm": 2842281.25, "learning_rate": 1.3640065146579806e-05, "loss": 0.7023, "step": 1031 }, { "epoch": 2.2668863261943986, "grad_norm": 4719035.0, "learning_rate": 1.3599348534201955e-05, "loss": 0.6832, "step": 1032 }, { "epoch": 2.269082921471719, "grad_norm": 2050836.125, "learning_rate": 1.3558631921824106e-05, "loss": 0.9524, "step": 1033 }, { "epoch": 2.271279516749039, "grad_norm": 3227051.75, "learning_rate": 1.3517915309446254e-05, "loss": 0.7612, "step": 1034 }, { "epoch": 2.2734761120263594, "grad_norm": 5535661.0, "learning_rate": 1.3477198697068405e-05, "loss": 0.7114, "step": 1035 }, { "epoch": 2.2756727073036793, "grad_norm": 1950398.625, "learning_rate": 1.3436482084690555e-05, "loss": 0.7762, "step": 1036 }, { "epoch": 2.2778693025809993, "grad_norm": 4733415.0, "learning_rate": 1.3395765472312704e-05, "loss": 0.684, "step": 1037 }, { "epoch": 2.2800658978583197, "grad_norm": 3833281.0, "learning_rate": 1.3355048859934855e-05, "loss": 0.7297, "step": 1038 }, { "epoch": 2.2822624931356397, "grad_norm": 6211348.5, "learning_rate": 1.3314332247557005e-05, "loss": 0.7067, "step": 1039 }, { "epoch": 2.28445908841296, "grad_norm": 1275267.75, "learning_rate": 1.3273615635179154e-05, "loss": 0.8348, "step": 1040 }, { "epoch": 2.28665568369028, "grad_norm": 3057475.25, "learning_rate": 1.3232899022801304e-05, "loss": 0.764, "step": 1041 }, { "epoch": 2.2888522789676005, "grad_norm": 5520224.5, "learning_rate": 1.3192182410423453e-05, "loss": 0.867, "step": 1042 }, { "epoch": 2.2910488742449204, "grad_norm": 3466532.25, "learning_rate": 1.3151465798045603e-05, "loss": 0.6492, "step": 1043 }, { "epoch": 2.2932454695222404, "grad_norm": 2652557.0, "learning_rate": 1.3110749185667754e-05, "loss": 0.7513, "step": 1044 }, { "epoch": 2.295442064799561, "grad_norm": 2464003.75, "learning_rate": 1.3070032573289903e-05, "loss": 0.8589, "step": 1045 }, { "epoch": 2.297638660076881, "grad_norm": 3981063.0, "learning_rate": 1.3029315960912053e-05, "loss": 0.7246, "step": 1046 }, { "epoch": 2.2998352553542007, "grad_norm": 2509091.0, "learning_rate": 1.2988599348534204e-05, "loss": 0.614, "step": 1047 }, { "epoch": 2.302031850631521, "grad_norm": 5999327.0, "learning_rate": 1.2947882736156352e-05, "loss": 0.7993, "step": 1048 }, { "epoch": 2.304228445908841, "grad_norm": 8384308.0, "learning_rate": 1.2907166123778503e-05, "loss": 0.7839, "step": 1049 }, { "epoch": 2.3064250411861615, "grad_norm": 6297944.5, "learning_rate": 1.2866449511400652e-05, "loss": 0.8027, "step": 1050 }, { "epoch": 2.3086216364634815, "grad_norm": 13187944.0, "learning_rate": 1.2825732899022802e-05, "loss": 0.6071, "step": 1051 }, { "epoch": 2.310818231740802, "grad_norm": 9754928.0, "learning_rate": 1.2785016286644952e-05, "loss": 0.7241, "step": 1052 }, { "epoch": 2.313014827018122, "grad_norm": 1552770.875, "learning_rate": 1.2744299674267101e-05, "loss": 0.8248, "step": 1053 }, { "epoch": 2.315211422295442, "grad_norm": 16734540.0, "learning_rate": 1.2703583061889252e-05, "loss": 0.7476, "step": 1054 }, { "epoch": 2.3174080175727623, "grad_norm": 4143281.75, "learning_rate": 1.2662866449511402e-05, "loss": 1.0132, "step": 1055 }, { "epoch": 2.3196046128500822, "grad_norm": 7862158.0, "learning_rate": 1.2622149837133551e-05, "loss": 0.7782, "step": 1056 }, { "epoch": 2.3218012081274026, "grad_norm": 4118543.25, "learning_rate": 1.2581433224755701e-05, "loss": 1.0039, "step": 1057 }, { "epoch": 2.3239978034047226, "grad_norm": 4323584.5, "learning_rate": 1.254071661237785e-05, "loss": 0.6, "step": 1058 }, { "epoch": 2.326194398682043, "grad_norm": 3603073.5, "learning_rate": 1.25e-05, "loss": 0.7142, "step": 1059 }, { "epoch": 2.328390993959363, "grad_norm": 6588454.0, "learning_rate": 1.2459283387622151e-05, "loss": 0.7712, "step": 1060 }, { "epoch": 2.330587589236683, "grad_norm": 2593079.5, "learning_rate": 1.24185667752443e-05, "loss": 0.7785, "step": 1061 }, { "epoch": 2.3327841845140034, "grad_norm": 1775102.625, "learning_rate": 1.237785016286645e-05, "loss": 0.7758, "step": 1062 }, { "epoch": 2.3349807797913233, "grad_norm": 2622416.25, "learning_rate": 1.2337133550488599e-05, "loss": 0.8841, "step": 1063 }, { "epoch": 2.3371773750686438, "grad_norm": 3231957.5, "learning_rate": 1.229641693811075e-05, "loss": 0.8539, "step": 1064 }, { "epoch": 2.3393739703459637, "grad_norm": 1965101.875, "learning_rate": 1.22557003257329e-05, "loss": 0.7902, "step": 1065 }, { "epoch": 2.341570565623284, "grad_norm": 9011023.0, "learning_rate": 1.2214983713355049e-05, "loss": 0.9112, "step": 1066 }, { "epoch": 2.343767160900604, "grad_norm": 12300077.0, "learning_rate": 1.21742671009772e-05, "loss": 0.6904, "step": 1067 }, { "epoch": 2.345963756177924, "grad_norm": 4213445.0, "learning_rate": 1.213355048859935e-05, "loss": 0.8308, "step": 1068 }, { "epoch": 2.3481603514552445, "grad_norm": 3941625.0, "learning_rate": 1.2092833876221498e-05, "loss": 0.9742, "step": 1069 }, { "epoch": 2.3503569467325645, "grad_norm": 5948733.0, "learning_rate": 1.2052117263843649e-05, "loss": 0.7223, "step": 1070 }, { "epoch": 2.352553542009885, "grad_norm": 5675687.5, "learning_rate": 1.2011400651465798e-05, "loss": 0.8359, "step": 1071 }, { "epoch": 2.354750137287205, "grad_norm": 1645119.875, "learning_rate": 1.1970684039087948e-05, "loss": 0.6769, "step": 1072 }, { "epoch": 2.356946732564525, "grad_norm": 7933894.5, "learning_rate": 1.1929967426710099e-05, "loss": 0.7246, "step": 1073 }, { "epoch": 2.359143327841845, "grad_norm": 13084234.0, "learning_rate": 1.1889250814332247e-05, "loss": 0.7177, "step": 1074 }, { "epoch": 2.361339923119165, "grad_norm": 4980314.0, "learning_rate": 1.1848534201954398e-05, "loss": 0.874, "step": 1075 }, { "epoch": 2.3635365183964856, "grad_norm": 11697443.0, "learning_rate": 1.1807817589576548e-05, "loss": 0.8653, "step": 1076 }, { "epoch": 2.3657331136738056, "grad_norm": 6242828.5, "learning_rate": 1.1767100977198697e-05, "loss": 0.771, "step": 1077 }, { "epoch": 2.3679297089511255, "grad_norm": 3590073.5, "learning_rate": 1.1726384364820847e-05, "loss": 0.8289, "step": 1078 }, { "epoch": 2.370126304228446, "grad_norm": 7172018.0, "learning_rate": 1.1685667752442996e-05, "loss": 0.613, "step": 1079 }, { "epoch": 2.372322899505766, "grad_norm": 3888274.0, "learning_rate": 1.1644951140065147e-05, "loss": 0.7947, "step": 1080 }, { "epoch": 2.3745194947830863, "grad_norm": 8613911.0, "learning_rate": 1.1604234527687297e-05, "loss": 1.0802, "step": 1081 }, { "epoch": 2.3767160900604063, "grad_norm": 4574738.5, "learning_rate": 1.1563517915309446e-05, "loss": 0.7241, "step": 1082 }, { "epoch": 2.3789126853377267, "grad_norm": 9096491.0, "learning_rate": 1.1522801302931596e-05, "loss": 0.7928, "step": 1083 }, { "epoch": 2.3811092806150467, "grad_norm": 6841468.5, "learning_rate": 1.1482084690553747e-05, "loss": 0.9741, "step": 1084 }, { "epoch": 2.3833058758923666, "grad_norm": 3922741.5, "learning_rate": 1.1441368078175896e-05, "loss": 1.0277, "step": 1085 }, { "epoch": 2.385502471169687, "grad_norm": 3562315.5, "learning_rate": 1.1400651465798046e-05, "loss": 0.7098, "step": 1086 }, { "epoch": 2.387699066447007, "grad_norm": 14663703.0, "learning_rate": 1.1359934853420195e-05, "loss": 1.0815, "step": 1087 }, { "epoch": 2.3898956617243274, "grad_norm": 15018353.0, "learning_rate": 1.1319218241042345e-05, "loss": 0.7351, "step": 1088 }, { "epoch": 2.3920922570016474, "grad_norm": 8490890.0, "learning_rate": 1.1278501628664496e-05, "loss": 0.9321, "step": 1089 }, { "epoch": 2.394288852278968, "grad_norm": 4437916.5, "learning_rate": 1.1237785016286644e-05, "loss": 0.7472, "step": 1090 }, { "epoch": 2.396485447556288, "grad_norm": 9657434.0, "learning_rate": 1.1197068403908795e-05, "loss": 0.7775, "step": 1091 }, { "epoch": 2.3986820428336078, "grad_norm": 6190163.5, "learning_rate": 1.1156351791530945e-05, "loss": 0.8102, "step": 1092 }, { "epoch": 2.400878638110928, "grad_norm": 8654396.0, "learning_rate": 1.1115635179153094e-05, "loss": 0.8412, "step": 1093 }, { "epoch": 2.403075233388248, "grad_norm": 3719385.0, "learning_rate": 1.1074918566775245e-05, "loss": 0.8662, "step": 1094 }, { "epoch": 2.4052718286655685, "grad_norm": 7327033.5, "learning_rate": 1.1034201954397393e-05, "loss": 0.7252, "step": 1095 }, { "epoch": 2.4074684239428885, "grad_norm": 8585433.0, "learning_rate": 1.0993485342019544e-05, "loss": 0.6554, "step": 1096 }, { "epoch": 2.409665019220209, "grad_norm": 3423707.5, "learning_rate": 1.0952768729641694e-05, "loss": 0.8809, "step": 1097 }, { "epoch": 2.411861614497529, "grad_norm": 3511080.0, "learning_rate": 1.0912052117263845e-05, "loss": 0.8029, "step": 1098 }, { "epoch": 2.414058209774849, "grad_norm": 3968857.5, "learning_rate": 1.0871335504885995e-05, "loss": 0.8949, "step": 1099 }, { "epoch": 2.4162548050521693, "grad_norm": 2289465.75, "learning_rate": 1.0830618892508144e-05, "loss": 0.8306, "step": 1100 }, { "epoch": 2.4184514003294892, "grad_norm": 20745104.0, "learning_rate": 1.0789902280130294e-05, "loss": 0.7254, "step": 1101 }, { "epoch": 2.420647995606809, "grad_norm": 2543496.75, "learning_rate": 1.0749185667752445e-05, "loss": 0.8705, "step": 1102 }, { "epoch": 2.4228445908841296, "grad_norm": 5394329.5, "learning_rate": 1.0708469055374594e-05, "loss": 0.8089, "step": 1103 }, { "epoch": 2.4250411861614496, "grad_norm": 1576959.875, "learning_rate": 1.0667752442996744e-05, "loss": 0.8676, "step": 1104 }, { "epoch": 2.42723778143877, "grad_norm": 3800304.0, "learning_rate": 1.0627035830618893e-05, "loss": 0.845, "step": 1105 }, { "epoch": 2.42943437671609, "grad_norm": 15098813.0, "learning_rate": 1.0586319218241043e-05, "loss": 0.7557, "step": 1106 }, { "epoch": 2.4316309719934104, "grad_norm": 4500141.5, "learning_rate": 1.0545602605863194e-05, "loss": 0.8855, "step": 1107 }, { "epoch": 2.4338275672707304, "grad_norm": 2867308.5, "learning_rate": 1.0504885993485343e-05, "loss": 0.9233, "step": 1108 }, { "epoch": 2.4360241625480503, "grad_norm": 4740615.5, "learning_rate": 1.0464169381107493e-05, "loss": 0.7984, "step": 1109 }, { "epoch": 2.4382207578253707, "grad_norm": 9119673.0, "learning_rate": 1.0423452768729644e-05, "loss": 0.8595, "step": 1110 }, { "epoch": 2.4404173531026907, "grad_norm": 9822895.0, "learning_rate": 1.0382736156351792e-05, "loss": 0.7805, "step": 1111 }, { "epoch": 2.442613948380011, "grad_norm": 1951774.125, "learning_rate": 1.0342019543973943e-05, "loss": 0.6044, "step": 1112 }, { "epoch": 2.444810543657331, "grad_norm": 7767419.0, "learning_rate": 1.0301302931596091e-05, "loss": 1.0171, "step": 1113 }, { "epoch": 2.4470071389346515, "grad_norm": 31730786.0, "learning_rate": 1.0260586319218242e-05, "loss": 0.9539, "step": 1114 }, { "epoch": 2.4492037342119715, "grad_norm": 32194368.0, "learning_rate": 1.0219869706840392e-05, "loss": 0.7512, "step": 1115 }, { "epoch": 2.4514003294892914, "grad_norm": 58858368.0, "learning_rate": 1.0179153094462541e-05, "loss": 0.7022, "step": 1116 }, { "epoch": 2.453596924766612, "grad_norm": 3722460.5, "learning_rate": 1.0138436482084692e-05, "loss": 0.7823, "step": 1117 }, { "epoch": 2.455793520043932, "grad_norm": 12120713.0, "learning_rate": 1.009771986970684e-05, "loss": 0.8108, "step": 1118 }, { "epoch": 2.4579901153212522, "grad_norm": 7420889.0, "learning_rate": 1.0057003257328991e-05, "loss": 0.8501, "step": 1119 }, { "epoch": 2.460186710598572, "grad_norm": 3466311.75, "learning_rate": 1.0016286644951141e-05, "loss": 0.7779, "step": 1120 }, { "epoch": 2.4623833058758926, "grad_norm": 2764851.0, "learning_rate": 9.97557003257329e-06, "loss": 0.8147, "step": 1121 }, { "epoch": 2.4645799011532126, "grad_norm": 8838615.0, "learning_rate": 9.93485342019544e-06, "loss": 1.1987, "step": 1122 }, { "epoch": 2.4667764964305325, "grad_norm": 4142787.75, "learning_rate": 9.894136807817591e-06, "loss": 0.8362, "step": 1123 }, { "epoch": 2.468973091707853, "grad_norm": 8983909.0, "learning_rate": 9.85342019543974e-06, "loss": 0.8383, "step": 1124 }, { "epoch": 2.471169686985173, "grad_norm": 5277952.5, "learning_rate": 9.81270358306189e-06, "loss": 0.6959, "step": 1125 }, { "epoch": 2.473366282262493, "grad_norm": 3108833.25, "learning_rate": 9.771986970684039e-06, "loss": 0.9406, "step": 1126 }, { "epoch": 2.4755628775398133, "grad_norm": 3648057.0, "learning_rate": 9.73127035830619e-06, "loss": 0.7256, "step": 1127 }, { "epoch": 2.4777594728171333, "grad_norm": 4605368.5, "learning_rate": 9.69055374592834e-06, "loss": 0.7745, "step": 1128 }, { "epoch": 2.4799560680944537, "grad_norm": 4494435.5, "learning_rate": 9.649837133550489e-06, "loss": 1.0054, "step": 1129 }, { "epoch": 2.4821526633717736, "grad_norm": 8113612.0, "learning_rate": 9.609120521172639e-06, "loss": 0.802, "step": 1130 }, { "epoch": 2.484349258649094, "grad_norm": 4283793.0, "learning_rate": 9.56840390879479e-06, "loss": 0.7634, "step": 1131 }, { "epoch": 2.486545853926414, "grad_norm": 4527398.0, "learning_rate": 9.527687296416938e-06, "loss": 0.7985, "step": 1132 }, { "epoch": 2.488742449203734, "grad_norm": 2424841.5, "learning_rate": 9.486970684039089e-06, "loss": 0.6853, "step": 1133 }, { "epoch": 2.4909390444810544, "grad_norm": 4183110.5, "learning_rate": 9.446254071661238e-06, "loss": 0.8485, "step": 1134 }, { "epoch": 2.4931356397583744, "grad_norm": 10011695.0, "learning_rate": 9.405537459283388e-06, "loss": 0.839, "step": 1135 }, { "epoch": 2.495332235035695, "grad_norm": 4811375.5, "learning_rate": 9.364820846905538e-06, "loss": 0.702, "step": 1136 }, { "epoch": 2.4975288303130148, "grad_norm": 3211023.5, "learning_rate": 9.324104234527687e-06, "loss": 0.5787, "step": 1137 }, { "epoch": 2.499725425590335, "grad_norm": 10911671.0, "learning_rate": 9.283387622149838e-06, "loss": 0.8402, "step": 1138 }, { "epoch": 2.501922020867655, "grad_norm": 6305678.0, "learning_rate": 9.242671009771988e-06, "loss": 0.9766, "step": 1139 }, { "epoch": 2.504118616144975, "grad_norm": 7081075.5, "learning_rate": 9.201954397394137e-06, "loss": 0.7916, "step": 1140 }, { "epoch": 2.5063152114222955, "grad_norm": 7760064.5, "learning_rate": 9.161237785016287e-06, "loss": 0.7072, "step": 1141 }, { "epoch": 2.5085118066996155, "grad_norm": 37980600.0, "learning_rate": 9.120521172638436e-06, "loss": 0.8516, "step": 1142 }, { "epoch": 2.510708401976936, "grad_norm": 3480184.5, "learning_rate": 9.079804560260587e-06, "loss": 0.9093, "step": 1143 }, { "epoch": 2.512904997254256, "grad_norm": 4379595.0, "learning_rate": 9.039087947882737e-06, "loss": 0.7566, "step": 1144 }, { "epoch": 2.5151015925315763, "grad_norm": 7421131.0, "learning_rate": 8.998371335504886e-06, "loss": 0.7599, "step": 1145 }, { "epoch": 2.5172981878088962, "grad_norm": 6436979.5, "learning_rate": 8.957654723127036e-06, "loss": 0.8429, "step": 1146 }, { "epoch": 2.519494783086216, "grad_norm": 2583326.25, "learning_rate": 8.916938110749185e-06, "loss": 0.7495, "step": 1147 }, { "epoch": 2.5216913783635366, "grad_norm": 13952491.0, "learning_rate": 8.876221498371336e-06, "loss": 0.7041, "step": 1148 }, { "epoch": 2.5238879736408566, "grad_norm": 5217487.5, "learning_rate": 8.835504885993486e-06, "loss": 0.8489, "step": 1149 }, { "epoch": 2.5260845689181766, "grad_norm": 7477479.5, "learning_rate": 8.794788273615635e-06, "loss": 1.0378, "step": 1150 }, { "epoch": 2.528281164195497, "grad_norm": 6661338.0, "learning_rate": 8.754071661237785e-06, "loss": 0.8179, "step": 1151 }, { "epoch": 2.5304777594728174, "grad_norm": 3381347.0, "learning_rate": 8.713355048859936e-06, "loss": 0.7556, "step": 1152 }, { "epoch": 2.5326743547501374, "grad_norm": 6420329.5, "learning_rate": 8.672638436482084e-06, "loss": 0.8233, "step": 1153 }, { "epoch": 2.5348709500274573, "grad_norm": 4672946.5, "learning_rate": 8.631921824104235e-06, "loss": 0.7382, "step": 1154 }, { "epoch": 2.5370675453047777, "grad_norm": 3747122.25, "learning_rate": 8.591205211726384e-06, "loss": 0.7522, "step": 1155 }, { "epoch": 2.5392641405820977, "grad_norm": 8256866.5, "learning_rate": 8.550488599348534e-06, "loss": 0.9296, "step": 1156 }, { "epoch": 2.5414607358594177, "grad_norm": 6203118.5, "learning_rate": 8.509771986970685e-06, "loss": 0.9205, "step": 1157 }, { "epoch": 2.543657331136738, "grad_norm": 16221756.0, "learning_rate": 8.469055374592833e-06, "loss": 0.7076, "step": 1158 }, { "epoch": 2.545853926414058, "grad_norm": 11897712.0, "learning_rate": 8.428338762214984e-06, "loss": 0.6916, "step": 1159 }, { "epoch": 2.5480505216913785, "grad_norm": 19553034.0, "learning_rate": 8.387622149837134e-06, "loss": 0.772, "step": 1160 }, { "epoch": 2.5502471169686984, "grad_norm": 2140742.25, "learning_rate": 8.346905537459283e-06, "loss": 0.6454, "step": 1161 }, { "epoch": 2.552443712246019, "grad_norm": 2892942.5, "learning_rate": 8.306188925081433e-06, "loss": 0.828, "step": 1162 }, { "epoch": 2.554640307523339, "grad_norm": 8463796.0, "learning_rate": 8.265472312703582e-06, "loss": 0.6992, "step": 1163 }, { "epoch": 2.556836902800659, "grad_norm": 3616309.0, "learning_rate": 8.224755700325733e-06, "loss": 0.8697, "step": 1164 }, { "epoch": 2.559033498077979, "grad_norm": 5422616.5, "learning_rate": 8.184039087947883e-06, "loss": 0.861, "step": 1165 }, { "epoch": 2.561230093355299, "grad_norm": 10063970.0, "learning_rate": 8.143322475570032e-06, "loss": 0.7409, "step": 1166 }, { "epoch": 2.5634266886326196, "grad_norm": 3730941.75, "learning_rate": 8.102605863192182e-06, "loss": 0.6844, "step": 1167 }, { "epoch": 2.5656232839099395, "grad_norm": 5220560.0, "learning_rate": 8.061889250814333e-06, "loss": 0.8594, "step": 1168 }, { "epoch": 2.56781987918726, "grad_norm": 7238725.5, "learning_rate": 8.021172638436482e-06, "loss": 0.7781, "step": 1169 }, { "epoch": 2.57001647446458, "grad_norm": 5563441.5, "learning_rate": 7.980456026058632e-06, "loss": 0.7106, "step": 1170 }, { "epoch": 2.5722130697419, "grad_norm": 4423179.0, "learning_rate": 7.93973941368078e-06, "loss": 0.6717, "step": 1171 }, { "epoch": 2.5744096650192203, "grad_norm": 2079958.125, "learning_rate": 7.899022801302931e-06, "loss": 0.7362, "step": 1172 }, { "epoch": 2.5766062602965403, "grad_norm": 2770067.5, "learning_rate": 7.858306188925082e-06, "loss": 0.7288, "step": 1173 }, { "epoch": 2.5788028555738602, "grad_norm": 2280089.25, "learning_rate": 7.81758957654723e-06, "loss": 0.8524, "step": 1174 }, { "epoch": 2.5809994508511807, "grad_norm": 5417975.0, "learning_rate": 7.776872964169383e-06, "loss": 0.8495, "step": 1175 }, { "epoch": 2.583196046128501, "grad_norm": 5998843.5, "learning_rate": 7.736156351791531e-06, "loss": 0.7334, "step": 1176 }, { "epoch": 2.585392641405821, "grad_norm": 9620786.0, "learning_rate": 7.695439739413682e-06, "loss": 0.9633, "step": 1177 }, { "epoch": 2.587589236683141, "grad_norm": 27155608.0, "learning_rate": 7.654723127035832e-06, "loss": 0.7478, "step": 1178 }, { "epoch": 2.5897858319604614, "grad_norm": 3008433.0, "learning_rate": 7.614006514657981e-06, "loss": 0.9599, "step": 1179 }, { "epoch": 2.5919824272377814, "grad_norm": 4526184.0, "learning_rate": 7.5732899022801316e-06, "loss": 0.6892, "step": 1180 }, { "epoch": 2.5941790225151014, "grad_norm": 39450368.0, "learning_rate": 7.532573289902281e-06, "loss": 0.941, "step": 1181 }, { "epoch": 2.5963756177924218, "grad_norm": 2991268.5, "learning_rate": 7.491856677524431e-06, "loss": 0.8136, "step": 1182 }, { "epoch": 2.598572213069742, "grad_norm": 8942134.0, "learning_rate": 7.45114006514658e-06, "loss": 0.7942, "step": 1183 }, { "epoch": 2.600768808347062, "grad_norm": 14095416.0, "learning_rate": 7.410423452768731e-06, "loss": 0.8661, "step": 1184 }, { "epoch": 2.602965403624382, "grad_norm": 5406536.0, "learning_rate": 7.3697068403908805e-06, "loss": 0.8327, "step": 1185 }, { "epoch": 2.6051619989017025, "grad_norm": 3708295.0, "learning_rate": 7.32899022801303e-06, "loss": 0.8411, "step": 1186 }, { "epoch": 2.6073585941790225, "grad_norm": 5899132.0, "learning_rate": 7.28827361563518e-06, "loss": 0.6926, "step": 1187 }, { "epoch": 2.6095551894563425, "grad_norm": 2464656.5, "learning_rate": 7.24755700325733e-06, "loss": 0.7658, "step": 1188 }, { "epoch": 2.611751784733663, "grad_norm": 6742926.5, "learning_rate": 7.20684039087948e-06, "loss": 0.7401, "step": 1189 }, { "epoch": 2.613948380010983, "grad_norm": 1958478.375, "learning_rate": 7.166123778501629e-06, "loss": 0.7855, "step": 1190 }, { "epoch": 2.6161449752883033, "grad_norm": 2084311.875, "learning_rate": 7.125407166123779e-06, "loss": 0.7656, "step": 1191 }, { "epoch": 2.618341570565623, "grad_norm": 6223765.0, "learning_rate": 7.0846905537459294e-06, "loss": 0.622, "step": 1192 }, { "epoch": 2.6205381658429436, "grad_norm": 4192605.75, "learning_rate": 7.043973941368079e-06, "loss": 1.2417, "step": 1193 }, { "epoch": 2.6227347611202636, "grad_norm": 8324554.0, "learning_rate": 7.003257328990229e-06, "loss": 0.9172, "step": 1194 }, { "epoch": 2.6249313563975836, "grad_norm": 4022413.0, "learning_rate": 6.962540716612378e-06, "loss": 0.5115, "step": 1195 }, { "epoch": 2.627127951674904, "grad_norm": 1934851.625, "learning_rate": 6.921824104234528e-06, "loss": 0.7719, "step": 1196 }, { "epoch": 2.629324546952224, "grad_norm": 2949400.5, "learning_rate": 6.881107491856678e-06, "loss": 0.8278, "step": 1197 }, { "epoch": 2.6315211422295444, "grad_norm": 2916673.25, "learning_rate": 6.840390879478828e-06, "loss": 0.8172, "step": 1198 }, { "epoch": 2.6337177375068643, "grad_norm": 5639826.5, "learning_rate": 6.799674267100978e-06, "loss": 0.6383, "step": 1199 }, { "epoch": 2.6359143327841847, "grad_norm": 2851869.0, "learning_rate": 6.758957654723127e-06, "loss": 0.7823, "step": 1200 }, { "epoch": 2.6381109280615047, "grad_norm": 7851859.5, "learning_rate": 6.718241042345278e-06, "loss": 0.6393, "step": 1201 }, { "epoch": 2.6403075233388247, "grad_norm": 14723333.0, "learning_rate": 6.677524429967427e-06, "loss": 0.6357, "step": 1202 }, { "epoch": 2.642504118616145, "grad_norm": 2337256.25, "learning_rate": 6.636807817589577e-06, "loss": 0.9949, "step": 1203 }, { "epoch": 2.644700713893465, "grad_norm": 8048820.5, "learning_rate": 6.5960912052117265e-06, "loss": 0.7724, "step": 1204 }, { "epoch": 2.646897309170785, "grad_norm": 6673396.5, "learning_rate": 6.555374592833877e-06, "loss": 0.8397, "step": 1205 }, { "epoch": 2.6490939044481054, "grad_norm": 16071678.0, "learning_rate": 6.5146579804560266e-06, "loss": 0.9185, "step": 1206 }, { "epoch": 2.651290499725426, "grad_norm": 2104632.75, "learning_rate": 6.473941368078176e-06, "loss": 0.8789, "step": 1207 }, { "epoch": 2.653487095002746, "grad_norm": 4341807.0, "learning_rate": 6.433224755700326e-06, "loss": 0.8184, "step": 1208 }, { "epoch": 2.655683690280066, "grad_norm": 7949819.0, "learning_rate": 6.392508143322476e-06, "loss": 0.6682, "step": 1209 }, { "epoch": 2.657880285557386, "grad_norm": 7754156.0, "learning_rate": 6.351791530944626e-06, "loss": 1.0141, "step": 1210 }, { "epoch": 2.660076880834706, "grad_norm": 4420963.0, "learning_rate": 6.3110749185667755e-06, "loss": 0.7257, "step": 1211 }, { "epoch": 2.662273476112026, "grad_norm": 4343279.0, "learning_rate": 6.270358306188925e-06, "loss": 0.7054, "step": 1212 }, { "epoch": 2.6644700713893466, "grad_norm": 25877124.0, "learning_rate": 6.2296416938110755e-06, "loss": 0.8773, "step": 1213 }, { "epoch": 2.6666666666666665, "grad_norm": 3126885.5, "learning_rate": 6.188925081433225e-06, "loss": 0.7401, "step": 1214 }, { "epoch": 2.668863261943987, "grad_norm": 15323114.0, "learning_rate": 6.148208469055375e-06, "loss": 0.7644, "step": 1215 }, { "epoch": 2.671059857221307, "grad_norm": 1805705.375, "learning_rate": 6.107491856677524e-06, "loss": 0.8049, "step": 1216 }, { "epoch": 2.6732564524986273, "grad_norm": 6889947.5, "learning_rate": 6.066775244299675e-06, "loss": 0.8436, "step": 1217 }, { "epoch": 2.6754530477759473, "grad_norm": 4288368.5, "learning_rate": 6.0260586319218244e-06, "loss": 0.7333, "step": 1218 }, { "epoch": 2.6776496430532672, "grad_norm": 4976939.0, "learning_rate": 5.985342019543974e-06, "loss": 0.7262, "step": 1219 }, { "epoch": 2.6798462383305877, "grad_norm": 2611674.5, "learning_rate": 5.944625407166124e-06, "loss": 0.7965, "step": 1220 }, { "epoch": 2.6820428336079076, "grad_norm": 3217127.25, "learning_rate": 5.903908794788274e-06, "loss": 0.9846, "step": 1221 }, { "epoch": 2.684239428885228, "grad_norm": 8207299.0, "learning_rate": 5.863192182410424e-06, "loss": 0.6034, "step": 1222 }, { "epoch": 2.686436024162548, "grad_norm": 14461778.0, "learning_rate": 5.822475570032573e-06, "loss": 0.8427, "step": 1223 }, { "epoch": 2.6886326194398684, "grad_norm": 1894958.375, "learning_rate": 5.781758957654723e-06, "loss": 0.7883, "step": 1224 }, { "epoch": 2.6908292147171884, "grad_norm": 6389778.0, "learning_rate": 5.741042345276873e-06, "loss": 0.9012, "step": 1225 }, { "epoch": 2.6930258099945084, "grad_norm": 18744804.0, "learning_rate": 5.700325732899023e-06, "loss": 0.6706, "step": 1226 }, { "epoch": 2.6952224052718288, "grad_norm": 5564625.5, "learning_rate": 5.659609120521173e-06, "loss": 0.8883, "step": 1227 }, { "epoch": 2.6974190005491487, "grad_norm": 8987305.0, "learning_rate": 5.618892508143322e-06, "loss": 0.7295, "step": 1228 }, { "epoch": 2.6996155958264687, "grad_norm": 12430103.0, "learning_rate": 5.578175895765473e-06, "loss": 0.8755, "step": 1229 }, { "epoch": 2.701812191103789, "grad_norm": 23815454.0, "learning_rate": 5.537459283387622e-06, "loss": 0.6697, "step": 1230 }, { "epoch": 2.7040087863811095, "grad_norm": 3578766.75, "learning_rate": 5.496742671009772e-06, "loss": 0.6827, "step": 1231 }, { "epoch": 2.7062053816584295, "grad_norm": 5508980.5, "learning_rate": 5.456026058631922e-06, "loss": 0.859, "step": 1232 }, { "epoch": 2.7084019769357495, "grad_norm": 8206292.0, "learning_rate": 5.415309446254072e-06, "loss": 0.7212, "step": 1233 }, { "epoch": 2.71059857221307, "grad_norm": 9189232.0, "learning_rate": 5.3745928338762225e-06, "loss": 0.7818, "step": 1234 }, { "epoch": 2.71279516749039, "grad_norm": 2708089.75, "learning_rate": 5.333876221498372e-06, "loss": 0.8605, "step": 1235 }, { "epoch": 2.71499176276771, "grad_norm": 7786324.5, "learning_rate": 5.293159609120522e-06, "loss": 0.9834, "step": 1236 }, { "epoch": 2.7171883580450302, "grad_norm": 3118259.0, "learning_rate": 5.252442996742671e-06, "loss": 0.8038, "step": 1237 }, { "epoch": 2.71938495332235, "grad_norm": 3731574.0, "learning_rate": 5.211726384364822e-06, "loss": 0.8072, "step": 1238 }, { "epoch": 2.7215815485996706, "grad_norm": 3523254.0, "learning_rate": 5.171009771986971e-06, "loss": 0.7805, "step": 1239 }, { "epoch": 2.7237781438769906, "grad_norm": 1341622.375, "learning_rate": 5.130293159609121e-06, "loss": 0.7965, "step": 1240 }, { "epoch": 2.725974739154311, "grad_norm": 3696195.25, "learning_rate": 5.089576547231271e-06, "loss": 0.6148, "step": 1241 }, { "epoch": 2.728171334431631, "grad_norm": 5802127.5, "learning_rate": 5.04885993485342e-06, "loss": 0.8052, "step": 1242 }, { "epoch": 2.730367929708951, "grad_norm": 3392009.75, "learning_rate": 5.008143322475571e-06, "loss": 0.8646, "step": 1243 }, { "epoch": 2.7325645249862713, "grad_norm": 1102332.625, "learning_rate": 4.96742671009772e-06, "loss": 0.8439, "step": 1244 }, { "epoch": 2.7347611202635913, "grad_norm": 8662749.0, "learning_rate": 4.92671009771987e-06, "loss": 0.8066, "step": 1245 }, { "epoch": 2.7369577155409117, "grad_norm": 15605783.0, "learning_rate": 4.8859934853420195e-06, "loss": 0.8639, "step": 1246 }, { "epoch": 2.7391543108182317, "grad_norm": 11085640.0, "learning_rate": 4.84527687296417e-06, "loss": 1.3125, "step": 1247 }, { "epoch": 2.741350906095552, "grad_norm": 7253258.0, "learning_rate": 4.8045602605863196e-06, "loss": 0.7922, "step": 1248 }, { "epoch": 2.743547501372872, "grad_norm": 2673229.0, "learning_rate": 4.763843648208469e-06, "loss": 0.9026, "step": 1249 }, { "epoch": 2.745744096650192, "grad_norm": 1597929.875, "learning_rate": 4.723127035830619e-06, "loss": 0.828, "step": 1250 }, { "epoch": 2.7479406919275124, "grad_norm": 3410672.75, "learning_rate": 4.682410423452769e-06, "loss": 0.9291, "step": 1251 }, { "epoch": 2.7501372872048324, "grad_norm": 11938232.0, "learning_rate": 4.641693811074919e-06, "loss": 0.859, "step": 1252 }, { "epoch": 2.752333882482153, "grad_norm": 4468953.5, "learning_rate": 4.6009771986970685e-06, "loss": 0.6758, "step": 1253 }, { "epoch": 2.754530477759473, "grad_norm": 2428296.0, "learning_rate": 4.560260586319218e-06, "loss": 0.703, "step": 1254 }, { "epoch": 2.756727073036793, "grad_norm": 15036388.0, "learning_rate": 4.5195439739413685e-06, "loss": 0.9319, "step": 1255 }, { "epoch": 2.758923668314113, "grad_norm": 7085497.0, "learning_rate": 4.478827361563518e-06, "loss": 0.6379, "step": 1256 }, { "epoch": 2.761120263591433, "grad_norm": 3165533.5, "learning_rate": 4.438110749185668e-06, "loss": 0.7783, "step": 1257 }, { "epoch": 2.7633168588687536, "grad_norm": 3842944.5, "learning_rate": 4.397394136807817e-06, "loss": 0.6631, "step": 1258 }, { "epoch": 2.7655134541460735, "grad_norm": 26449858.0, "learning_rate": 4.356677524429968e-06, "loss": 0.9221, "step": 1259 }, { "epoch": 2.7677100494233935, "grad_norm": 4904693.5, "learning_rate": 4.3159609120521174e-06, "loss": 0.8772, "step": 1260 }, { "epoch": 2.769906644700714, "grad_norm": 3348442.5, "learning_rate": 4.275244299674267e-06, "loss": 0.7637, "step": 1261 }, { "epoch": 2.7721032399780343, "grad_norm": 8484901.0, "learning_rate": 4.234527687296417e-06, "loss": 0.8237, "step": 1262 }, { "epoch": 2.7742998352553543, "grad_norm": 9053304.0, "learning_rate": 4.193811074918567e-06, "loss": 0.8507, "step": 1263 }, { "epoch": 2.7764964305326743, "grad_norm": 5076836.5, "learning_rate": 4.153094462540717e-06, "loss": 0.9401, "step": 1264 }, { "epoch": 2.7786930258099947, "grad_norm": 4565967.5, "learning_rate": 4.112377850162866e-06, "loss": 0.8543, "step": 1265 }, { "epoch": 2.7808896210873146, "grad_norm": 3831978.0, "learning_rate": 4.071661237785016e-06, "loss": 0.6769, "step": 1266 }, { "epoch": 2.7830862163646346, "grad_norm": 2044434.375, "learning_rate": 4.030944625407166e-06, "loss": 0.6888, "step": 1267 }, { "epoch": 2.785282811641955, "grad_norm": 3601185.0, "learning_rate": 3.990228013029316e-06, "loss": 0.7531, "step": 1268 }, { "epoch": 2.787479406919275, "grad_norm": 5314910.0, "learning_rate": 3.949511400651466e-06, "loss": 0.6995, "step": 1269 }, { "epoch": 2.7896760021965954, "grad_norm": 5030508.0, "learning_rate": 3.908794788273615e-06, "loss": 0.7457, "step": 1270 }, { "epoch": 2.7918725974739154, "grad_norm": 5685473.5, "learning_rate": 3.868078175895766e-06, "loss": 0.7194, "step": 1271 }, { "epoch": 2.7940691927512358, "grad_norm": 9892777.0, "learning_rate": 3.827361563517916e-06, "loss": 0.8124, "step": 1272 }, { "epoch": 2.7962657880285557, "grad_norm": 4952108.5, "learning_rate": 3.7866449511400658e-06, "loss": 0.8066, "step": 1273 }, { "epoch": 2.7984623833058757, "grad_norm": 6134618.0, "learning_rate": 3.7459283387622154e-06, "loss": 0.9912, "step": 1274 }, { "epoch": 2.800658978583196, "grad_norm": 2698271.75, "learning_rate": 3.7052117263843654e-06, "loss": 0.773, "step": 1275 }, { "epoch": 2.802855573860516, "grad_norm": 3984969.5, "learning_rate": 3.664495114006515e-06, "loss": 0.9424, "step": 1276 }, { "epoch": 2.8050521691378365, "grad_norm": 10884478.0, "learning_rate": 3.623778501628665e-06, "loss": 1.0783, "step": 1277 }, { "epoch": 2.8072487644151565, "grad_norm": 4614646.5, "learning_rate": 3.5830618892508147e-06, "loss": 0.865, "step": 1278 }, { "epoch": 2.809445359692477, "grad_norm": 12210489.0, "learning_rate": 3.5423452768729647e-06, "loss": 0.7308, "step": 1279 }, { "epoch": 2.811641954969797, "grad_norm": 9137895.0, "learning_rate": 3.5016286644951143e-06, "loss": 0.7726, "step": 1280 }, { "epoch": 2.813838550247117, "grad_norm": 9227584.0, "learning_rate": 3.460912052117264e-06, "loss": 0.8759, "step": 1281 }, { "epoch": 2.8160351455244372, "grad_norm": 6036039.5, "learning_rate": 3.420195439739414e-06, "loss": 0.8112, "step": 1282 }, { "epoch": 2.818231740801757, "grad_norm": 9014040.0, "learning_rate": 3.3794788273615636e-06, "loss": 0.9741, "step": 1283 }, { "epoch": 2.820428336079077, "grad_norm": 2439870.5, "learning_rate": 3.3387622149837136e-06, "loss": 0.8584, "step": 1284 }, { "epoch": 2.8226249313563976, "grad_norm": 1889212.875, "learning_rate": 3.2980456026058632e-06, "loss": 0.8445, "step": 1285 }, { "epoch": 2.824821526633718, "grad_norm": 6801981.0, "learning_rate": 3.2573289902280133e-06, "loss": 1.0537, "step": 1286 }, { "epoch": 2.827018121911038, "grad_norm": 2069030.75, "learning_rate": 3.216612377850163e-06, "loss": 0.728, "step": 1287 }, { "epoch": 2.829214717188358, "grad_norm": 10793646.0, "learning_rate": 3.175895765472313e-06, "loss": 0.8375, "step": 1288 }, { "epoch": 2.8314113124656783, "grad_norm": 5786920.0, "learning_rate": 3.1351791530944625e-06, "loss": 0.7214, "step": 1289 }, { "epoch": 2.8336079077429983, "grad_norm": 9872860.0, "learning_rate": 3.0944625407166126e-06, "loss": 0.8624, "step": 1290 }, { "epoch": 2.8358045030203183, "grad_norm": 61847104.0, "learning_rate": 3.053745928338762e-06, "loss": 0.6621, "step": 1291 }, { "epoch": 2.8380010982976387, "grad_norm": 2021296.5, "learning_rate": 3.0130293159609122e-06, "loss": 0.7126, "step": 1292 }, { "epoch": 2.8401976935749587, "grad_norm": 1506967.125, "learning_rate": 2.972312703583062e-06, "loss": 0.5763, "step": 1293 }, { "epoch": 2.842394288852279, "grad_norm": 5157691.5, "learning_rate": 2.931596091205212e-06, "loss": 0.8836, "step": 1294 }, { "epoch": 2.844590884129599, "grad_norm": 8408041.0, "learning_rate": 2.8908794788273615e-06, "loss": 0.9025, "step": 1295 }, { "epoch": 2.8467874794069195, "grad_norm": 2117217.75, "learning_rate": 2.8501628664495115e-06, "loss": 0.9261, "step": 1296 }, { "epoch": 2.8489840746842394, "grad_norm": 4311697.0, "learning_rate": 2.809446254071661e-06, "loss": 0.674, "step": 1297 }, { "epoch": 2.8511806699615594, "grad_norm": 1351027.875, "learning_rate": 2.768729641693811e-06, "loss": 1.0208, "step": 1298 }, { "epoch": 2.85337726523888, "grad_norm": 19619468.0, "learning_rate": 2.728013029315961e-06, "loss": 0.8283, "step": 1299 }, { "epoch": 2.8555738605161998, "grad_norm": 7202300.5, "learning_rate": 2.6872964169381112e-06, "loss": 0.9765, "step": 1300 }, { "epoch": 2.85777045579352, "grad_norm": 2850933.25, "learning_rate": 2.646579804560261e-06, "loss": 0.8268, "step": 1301 }, { "epoch": 2.85996705107084, "grad_norm": 3646593.25, "learning_rate": 2.605863192182411e-06, "loss": 0.6783, "step": 1302 }, { "epoch": 2.8621636463481606, "grad_norm": 3372742.0, "learning_rate": 2.5651465798045605e-06, "loss": 1.0352, "step": 1303 }, { "epoch": 2.8643602416254805, "grad_norm": 3541434.0, "learning_rate": 2.52442996742671e-06, "loss": 0.7158, "step": 1304 }, { "epoch": 2.8665568369028005, "grad_norm": 2451411.5, "learning_rate": 2.48371335504886e-06, "loss": 0.753, "step": 1305 }, { "epoch": 2.868753432180121, "grad_norm": 3310861.0, "learning_rate": 2.4429967426710097e-06, "loss": 0.7655, "step": 1306 }, { "epoch": 2.870950027457441, "grad_norm": 3677972.0, "learning_rate": 2.4022801302931598e-06, "loss": 0.831, "step": 1307 }, { "epoch": 2.873146622734761, "grad_norm": 2547505.5, "learning_rate": 2.3615635179153094e-06, "loss": 0.6619, "step": 1308 }, { "epoch": 2.8753432180120813, "grad_norm": 15637284.0, "learning_rate": 2.3208469055374594e-06, "loss": 0.8139, "step": 1309 }, { "epoch": 2.8775398132894017, "grad_norm": 4371933.5, "learning_rate": 2.280130293159609e-06, "loss": 0.8925, "step": 1310 }, { "epoch": 2.8797364085667216, "grad_norm": 6664858.0, "learning_rate": 2.239413680781759e-06, "loss": 0.8394, "step": 1311 }, { "epoch": 2.8819330038440416, "grad_norm": 6765905.5, "learning_rate": 2.1986970684039087e-06, "loss": 0.8293, "step": 1312 }, { "epoch": 2.884129599121362, "grad_norm": 4205558.5, "learning_rate": 2.1579804560260587e-06, "loss": 0.8106, "step": 1313 }, { "epoch": 2.886326194398682, "grad_norm": 2466667.75, "learning_rate": 2.1172638436482083e-06, "loss": 0.819, "step": 1314 }, { "epoch": 2.888522789676002, "grad_norm": 8145394.0, "learning_rate": 2.0765472312703584e-06, "loss": 0.7181, "step": 1315 }, { "epoch": 2.8907193849533224, "grad_norm": 4562730.0, "learning_rate": 2.035830618892508e-06, "loss": 0.7766, "step": 1316 }, { "epoch": 2.892915980230643, "grad_norm": 4878736.5, "learning_rate": 1.995114006514658e-06, "loss": 0.7264, "step": 1317 }, { "epoch": 2.8951125755079627, "grad_norm": 7035435.5, "learning_rate": 1.9543973941368076e-06, "loss": 0.6426, "step": 1318 }, { "epoch": 2.8973091707852827, "grad_norm": 4361831.5, "learning_rate": 1.913680781758958e-06, "loss": 0.6632, "step": 1319 }, { "epoch": 2.899505766062603, "grad_norm": 6432325.5, "learning_rate": 1.8729641693811077e-06, "loss": 0.8785, "step": 1320 }, { "epoch": 2.901702361339923, "grad_norm": 2889484.5, "learning_rate": 1.8322475570032575e-06, "loss": 0.6304, "step": 1321 }, { "epoch": 2.903898956617243, "grad_norm": 12899765.0, "learning_rate": 1.7915309446254073e-06, "loss": 0.8323, "step": 1322 }, { "epoch": 2.9060955518945635, "grad_norm": 8848465.0, "learning_rate": 1.7508143322475572e-06, "loss": 0.7776, "step": 1323 }, { "epoch": 2.9082921471718834, "grad_norm": 9044314.0, "learning_rate": 1.710097719869707e-06, "loss": 0.7768, "step": 1324 }, { "epoch": 2.910488742449204, "grad_norm": 6190195.0, "learning_rate": 1.6693811074918568e-06, "loss": 0.7254, "step": 1325 }, { "epoch": 2.912685337726524, "grad_norm": 6261659.0, "learning_rate": 1.6286644951140066e-06, "loss": 0.8139, "step": 1326 }, { "epoch": 2.9148819330038442, "grad_norm": 2689539.25, "learning_rate": 1.5879478827361565e-06, "loss": 0.8426, "step": 1327 }, { "epoch": 2.917078528281164, "grad_norm": 4591746.5, "learning_rate": 1.5472312703583063e-06, "loss": 0.6496, "step": 1328 }, { "epoch": 2.919275123558484, "grad_norm": 5625928.5, "learning_rate": 1.5065146579804561e-06, "loss": 0.571, "step": 1329 }, { "epoch": 2.9214717188358046, "grad_norm": 5071654.5, "learning_rate": 1.465798045602606e-06, "loss": 0.7503, "step": 1330 }, { "epoch": 2.9236683141131246, "grad_norm": 3913498.75, "learning_rate": 1.4250814332247558e-06, "loss": 0.8757, "step": 1331 }, { "epoch": 2.925864909390445, "grad_norm": 1437075.625, "learning_rate": 1.3843648208469056e-06, "loss": 0.475, "step": 1332 }, { "epoch": 2.928061504667765, "grad_norm": 4601580.0, "learning_rate": 1.3436482084690556e-06, "loss": 0.7994, "step": 1333 }, { "epoch": 2.9302580999450853, "grad_norm": 4288684.5, "learning_rate": 1.3029315960912054e-06, "loss": 0.7059, "step": 1334 }, { "epoch": 2.9324546952224053, "grad_norm": 9257111.0, "learning_rate": 1.262214983713355e-06, "loss": 0.6693, "step": 1335 }, { "epoch": 2.9346512904997253, "grad_norm": 2904731.5, "learning_rate": 1.2214983713355049e-06, "loss": 0.7124, "step": 1336 }, { "epoch": 2.9368478857770457, "grad_norm": 5157225.0, "learning_rate": 1.1807817589576547e-06, "loss": 0.6926, "step": 1337 }, { "epoch": 2.9390444810543657, "grad_norm": 8206613.5, "learning_rate": 1.1400651465798045e-06, "loss": 0.7442, "step": 1338 }, { "epoch": 2.9412410763316856, "grad_norm": 3170124.5, "learning_rate": 1.0993485342019543e-06, "loss": 0.72, "step": 1339 }, { "epoch": 2.943437671609006, "grad_norm": 2436431.5, "learning_rate": 1.0586319218241042e-06, "loss": 0.5681, "step": 1340 }, { "epoch": 2.9456342668863265, "grad_norm": 5246524.0, "learning_rate": 1.017915309446254e-06, "loss": 0.7163, "step": 1341 }, { "epoch": 2.9478308621636464, "grad_norm": 3337089.5, "learning_rate": 9.771986970684038e-07, "loss": 0.7938, "step": 1342 }, { "epoch": 2.9500274574409664, "grad_norm": 6755895.0, "learning_rate": 9.364820846905538e-07, "loss": 0.878, "step": 1343 }, { "epoch": 2.952224052718287, "grad_norm": 8914362.0, "learning_rate": 8.957654723127037e-07, "loss": 0.9456, "step": 1344 }, { "epoch": 2.9544206479956068, "grad_norm": 4952710.0, "learning_rate": 8.550488599348535e-07, "loss": 0.8511, "step": 1345 }, { "epoch": 2.9566172432729267, "grad_norm": 14665521.0, "learning_rate": 8.143322475570033e-07, "loss": 0.7337, "step": 1346 }, { "epoch": 2.958813838550247, "grad_norm": 7912421.5, "learning_rate": 7.736156351791531e-07, "loss": 0.7101, "step": 1347 }, { "epoch": 2.961010433827567, "grad_norm": 30088358.0, "learning_rate": 7.32899022801303e-07, "loss": 0.7023, "step": 1348 }, { "epoch": 2.9632070291048875, "grad_norm": 4725255.5, "learning_rate": 6.921824104234528e-07, "loss": 0.9731, "step": 1349 }, { "epoch": 2.9654036243822075, "grad_norm": 5158853.5, "learning_rate": 6.514657980456027e-07, "loss": 0.7145, "step": 1350 }, { "epoch": 2.967600219659528, "grad_norm": 5323983.5, "learning_rate": 6.107491856677524e-07, "loss": 0.8979, "step": 1351 }, { "epoch": 2.969796814936848, "grad_norm": 5994244.0, "learning_rate": 5.700325732899023e-07, "loss": 0.769, "step": 1352 }, { "epoch": 2.971993410214168, "grad_norm": 4088484.5, "learning_rate": 5.293159609120521e-07, "loss": 0.8527, "step": 1353 }, { "epoch": 2.9741900054914883, "grad_norm": 2901818.5, "learning_rate": 4.885993485342019e-07, "loss": 0.9364, "step": 1354 }, { "epoch": 2.9763866007688082, "grad_norm": 11335834.0, "learning_rate": 4.4788273615635184e-07, "loss": 0.6777, "step": 1355 }, { "epoch": 2.9785831960461286, "grad_norm": 2763095.75, "learning_rate": 4.0716612377850166e-07, "loss": 0.89, "step": 1356 }, { "epoch": 2.9807797913234486, "grad_norm": 3734213.0, "learning_rate": 3.664495114006515e-07, "loss": 0.8536, "step": 1357 }, { "epoch": 2.982976386600769, "grad_norm": 30005090.0, "learning_rate": 3.2573289902280136e-07, "loss": 0.8909, "step": 1358 }, { "epoch": 2.985172981878089, "grad_norm": 5646650.0, "learning_rate": 2.8501628664495113e-07, "loss": 0.9318, "step": 1359 }, { "epoch": 2.987369577155409, "grad_norm": 10859570.0, "learning_rate": 2.4429967426710095e-07, "loss": 0.6719, "step": 1360 }, { "epoch": 2.9895661724327294, "grad_norm": 6024375.0, "learning_rate": 2.0358306188925083e-07, "loss": 0.814, "step": 1361 }, { "epoch": 2.9917627677100493, "grad_norm": 3320985.25, "learning_rate": 1.6286644951140068e-07, "loss": 1.0307, "step": 1362 }, { "epoch": 2.9939593629873693, "grad_norm": 6593533.0, "learning_rate": 1.2214983713355048e-07, "loss": 0.8078, "step": 1363 }, { "epoch": 2.9961559582646897, "grad_norm": 4601075.5, "learning_rate": 8.143322475570034e-08, "loss": 0.8431, "step": 1364 }, { "epoch": 2.99835255354201, "grad_norm": 7609306.5, "learning_rate": 4.071661237785017e-08, "loss": 0.888, "step": 1365 }, { "epoch": 2.99835255354201, "step": 1365, "total_flos": 4.598491246242038e+17, "train_loss": 0.8648729579352634, "train_runtime": 64264.928, "train_samples_per_second": 0.34, "train_steps_per_second": 0.021 } ], "logging_steps": 1.0, "max_steps": 1365, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4.598491246242038e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }