{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.08993897640450954, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 8.993897640450955e-06, "grad_norm": 640.7632446289062, "learning_rate": 1.0000000000000001e-07, "loss": 17.6667, "step": 1 }, { "epoch": 1.798779528090191e-05, "grad_norm": 729.949462890625, "learning_rate": 2.0000000000000002e-07, "loss": 17.1986, "step": 2 }, { "epoch": 2.6981692921352862e-05, "grad_norm": 1213.2032470703125, "learning_rate": 3.0000000000000004e-07, "loss": 17.0297, "step": 3 }, { "epoch": 3.597559056180382e-05, "grad_norm": 1249.3485107421875, "learning_rate": 4.0000000000000003e-07, "loss": 16.7016, "step": 4 }, { "epoch": 4.496948820225477e-05, "grad_norm": 462.6761169433594, "learning_rate": 5.000000000000001e-07, "loss": 17.4513, "step": 5 }, { "epoch": 5.3963385842705725e-05, "grad_norm": 907.276611328125, "learning_rate": 6.000000000000001e-07, "loss": 17.367, "step": 6 }, { "epoch": 6.295728348315668e-05, "grad_norm": 894.066162109375, "learning_rate": 7.000000000000001e-07, "loss": 17.8713, "step": 7 }, { "epoch": 7.195118112360764e-05, "grad_norm": 792.5089721679688, "learning_rate": 8.000000000000001e-07, "loss": 16.979, "step": 8 }, { "epoch": 8.094507876405858e-05, "grad_norm": 1808.308349609375, "learning_rate": 9e-07, "loss": 16.9944, "step": 9 }, { "epoch": 8.993897640450954e-05, "grad_norm": 1762.574951171875, "learning_rate": 1.0000000000000002e-06, "loss": 17.2708, "step": 10 }, { "epoch": 9.89328740449605e-05, "grad_norm": 1064.3702392578125, "learning_rate": 1.1e-06, "loss": 17.3214, "step": 11 }, { "epoch": 0.00010792677168541145, "grad_norm": 1691.4847412109375, "learning_rate": 1.2000000000000002e-06, "loss": 17.6209, "step": 12 }, { "epoch": 0.0001169206693258624, "grad_norm": 2237.672607421875, "learning_rate": 1.3e-06, "loss": 16.8936, "step": 13 }, { "epoch": 0.00012591456696631336, "grad_norm": 1722.5802001953125, "learning_rate": 1.4000000000000001e-06, "loss": 17.3657, "step": 14 }, { "epoch": 0.00013490846460676432, "grad_norm": 2044.4725341796875, "learning_rate": 1.5e-06, "loss": 16.7222, "step": 15 }, { "epoch": 0.00014390236224721528, "grad_norm": 3082.70751953125, "learning_rate": 1.6000000000000001e-06, "loss": 17.459, "step": 16 }, { "epoch": 0.0001528962598876662, "grad_norm": 1364.982666015625, "learning_rate": 1.7000000000000002e-06, "loss": 17.0347, "step": 17 }, { "epoch": 0.00016189015752811716, "grad_norm": 1626.4031982421875, "learning_rate": 1.8e-06, "loss": 17.2873, "step": 18 }, { "epoch": 0.00017088405516856812, "grad_norm": 1115.7916259765625, "learning_rate": 1.9e-06, "loss": 16.7047, "step": 19 }, { "epoch": 0.00017987795280901907, "grad_norm": 2770.2607421875, "learning_rate": 2.0000000000000003e-06, "loss": 16.9849, "step": 20 }, { "epoch": 0.00018887185044947003, "grad_norm": 901.8019409179688, "learning_rate": 2.1000000000000002e-06, "loss": 17.0128, "step": 21 }, { "epoch": 0.000197865748089921, "grad_norm": 2143.13818359375, "learning_rate": 2.2e-06, "loss": 16.5597, "step": 22 }, { "epoch": 0.00020685964573037194, "grad_norm": 1087.392578125, "learning_rate": 2.3e-06, "loss": 17.2394, "step": 23 }, { "epoch": 0.0002158535433708229, "grad_norm": 1091.6373291015625, "learning_rate": 2.4000000000000003e-06, "loss": 16.7669, "step": 24 }, { "epoch": 0.00022484744101127386, "grad_norm": 961.8850708007812, "learning_rate": 2.5e-06, "loss": 16.4527, "step": 25 }, { "epoch": 0.0002338413386517248, "grad_norm": 431.3570251464844, "learning_rate": 2.6e-06, "loss": 16.7052, "step": 26 }, { "epoch": 0.00024283523629217577, "grad_norm": 545.7597045898438, "learning_rate": 2.7e-06, "loss": 16.5546, "step": 27 }, { "epoch": 0.0002518291339326267, "grad_norm": 398.2252502441406, "learning_rate": 2.8000000000000003e-06, "loss": 16.9392, "step": 28 }, { "epoch": 0.0002608230315730777, "grad_norm": 969.6422119140625, "learning_rate": 2.9e-06, "loss": 16.5505, "step": 29 }, { "epoch": 0.00026981692921352864, "grad_norm": 484.8598937988281, "learning_rate": 3e-06, "loss": 16.5766, "step": 30 }, { "epoch": 0.0002788108268539796, "grad_norm": 1171.35595703125, "learning_rate": 3.1e-06, "loss": 16.6083, "step": 31 }, { "epoch": 0.00028780472449443055, "grad_norm": 431.4801330566406, "learning_rate": 3.2000000000000003e-06, "loss": 16.0039, "step": 32 }, { "epoch": 0.0002967986221348815, "grad_norm": 614.3028564453125, "learning_rate": 3.3e-06, "loss": 15.8999, "step": 33 }, { "epoch": 0.0003057925197753324, "grad_norm": 480.09381103515625, "learning_rate": 3.4000000000000005e-06, "loss": 16.0048, "step": 34 }, { "epoch": 0.00031478641741578337, "grad_norm": 325.46142578125, "learning_rate": 3.5000000000000004e-06, "loss": 15.6774, "step": 35 }, { "epoch": 0.0003237803150562343, "grad_norm": 1252.955810546875, "learning_rate": 3.6e-06, "loss": 16.4012, "step": 36 }, { "epoch": 0.0003327742126966853, "grad_norm": 1667.865234375, "learning_rate": 3.7e-06, "loss": 16.2314, "step": 37 }, { "epoch": 0.00034176811033713624, "grad_norm": 1876.4945068359375, "learning_rate": 3.8e-06, "loss": 16.3586, "step": 38 }, { "epoch": 0.0003507620079775872, "grad_norm": 345.5151672363281, "learning_rate": 3.9e-06, "loss": 15.5359, "step": 39 }, { "epoch": 0.00035975590561803815, "grad_norm": 661.9005126953125, "learning_rate": 4.000000000000001e-06, "loss": 15.7883, "step": 40 }, { "epoch": 0.0003687498032584891, "grad_norm": 585.4451293945312, "learning_rate": 4.1000000000000006e-06, "loss": 17.1201, "step": 41 }, { "epoch": 0.00037774370089894006, "grad_norm": 200.67575073242188, "learning_rate": 4.2000000000000004e-06, "loss": 15.4414, "step": 42 }, { "epoch": 0.000386737598539391, "grad_norm": 866.076416015625, "learning_rate": 4.2999999999999995e-06, "loss": 15.625, "step": 43 }, { "epoch": 0.000395731496179842, "grad_norm": 185.14735412597656, "learning_rate": 4.4e-06, "loss": 15.0003, "step": 44 }, { "epoch": 0.00040472539382029293, "grad_norm": 277.2205505371094, "learning_rate": 4.5e-06, "loss": 16.244, "step": 45 }, { "epoch": 0.0004137192914607439, "grad_norm": 377.3096008300781, "learning_rate": 4.6e-06, "loss": 15.6906, "step": 46 }, { "epoch": 0.00042271318910119484, "grad_norm": 773.2913208007812, "learning_rate": 4.7e-06, "loss": 15.9253, "step": 47 }, { "epoch": 0.0004317070867416458, "grad_norm": 445.2944641113281, "learning_rate": 4.800000000000001e-06, "loss": 14.8663, "step": 48 }, { "epoch": 0.00044070098438209676, "grad_norm": 614.1619262695312, "learning_rate": 4.9000000000000005e-06, "loss": 15.0696, "step": 49 }, { "epoch": 0.0004496948820225477, "grad_norm": 302.6289367675781, "learning_rate": 5e-06, "loss": 15.3708, "step": 50 }, { "epoch": 0.00045868877966299867, "grad_norm": 574.1361083984375, "learning_rate": 5.1e-06, "loss": 15.8231, "step": 51 }, { "epoch": 0.0004676826773034496, "grad_norm": 834.5971069335938, "learning_rate": 5.2e-06, "loss": 14.7522, "step": 52 }, { "epoch": 0.0004766765749439006, "grad_norm": 359.33636474609375, "learning_rate": 5.3e-06, "loss": 15.013, "step": 53 }, { "epoch": 0.00048567047258435154, "grad_norm": 531.15478515625, "learning_rate": 5.4e-06, "loss": 15.0103, "step": 54 }, { "epoch": 0.0004946643702248024, "grad_norm": 374.7421569824219, "learning_rate": 5.500000000000001e-06, "loss": 15.3783, "step": 55 }, { "epoch": 0.0005036582678652535, "grad_norm": 518.5064086914062, "learning_rate": 5.600000000000001e-06, "loss": 15.1811, "step": 56 }, { "epoch": 0.0005126521655057044, "grad_norm": 882.5928955078125, "learning_rate": 5.7000000000000005e-06, "loss": 14.5055, "step": 57 }, { "epoch": 0.0005216460631461554, "grad_norm": 722.7720336914062, "learning_rate": 5.8e-06, "loss": 14.5446, "step": 58 }, { "epoch": 0.0005306399607866063, "grad_norm": 398.8530578613281, "learning_rate": 5.9e-06, "loss": 14.8201, "step": 59 }, { "epoch": 0.0005396338584270573, "grad_norm": 246.3039093017578, "learning_rate": 6e-06, "loss": 14.49, "step": 60 }, { "epoch": 0.0005486277560675082, "grad_norm": 436.939697265625, "learning_rate": 6.1e-06, "loss": 14.2099, "step": 61 }, { "epoch": 0.0005576216537079592, "grad_norm": 278.00665283203125, "learning_rate": 6.2e-06, "loss": 13.7787, "step": 62 }, { "epoch": 0.0005666155513484101, "grad_norm": 375.7424011230469, "learning_rate": 6.300000000000001e-06, "loss": 14.2643, "step": 63 }, { "epoch": 0.0005756094489888611, "grad_norm": 737.5791015625, "learning_rate": 6.4000000000000006e-06, "loss": 13.9459, "step": 64 }, { "epoch": 0.000584603346629312, "grad_norm": 222.89804077148438, "learning_rate": 6.5000000000000004e-06, "loss": 13.9486, "step": 65 }, { "epoch": 0.000593597244269763, "grad_norm": 322.67864990234375, "learning_rate": 6.6e-06, "loss": 13.8074, "step": 66 }, { "epoch": 0.0006025911419102139, "grad_norm": 544.8624267578125, "learning_rate": 6.700000000000001e-06, "loss": 13.325, "step": 67 }, { "epoch": 0.0006115850395506648, "grad_norm": 526.0505981445312, "learning_rate": 6.800000000000001e-06, "loss": 13.3675, "step": 68 }, { "epoch": 0.0006205789371911158, "grad_norm": 607.5252075195312, "learning_rate": 6.900000000000001e-06, "loss": 13.6162, "step": 69 }, { "epoch": 0.0006295728348315667, "grad_norm": 405.1535949707031, "learning_rate": 7.000000000000001e-06, "loss": 13.0723, "step": 70 }, { "epoch": 0.0006385667324720177, "grad_norm": 450.2685852050781, "learning_rate": 7.1e-06, "loss": 13.0776, "step": 71 }, { "epoch": 0.0006475606301124686, "grad_norm": 293.61962890625, "learning_rate": 7.2e-06, "loss": 13.0886, "step": 72 }, { "epoch": 0.0006565545277529197, "grad_norm": 669.0880737304688, "learning_rate": 7.2999999999999996e-06, "loss": 13.0485, "step": 73 }, { "epoch": 0.0006655484253933706, "grad_norm": 428.8260192871094, "learning_rate": 7.4e-06, "loss": 12.522, "step": 74 }, { "epoch": 0.0006745423230338216, "grad_norm": 616.1552734375, "learning_rate": 7.5e-06, "loss": 12.7778, "step": 75 }, { "epoch": 0.0006835362206742725, "grad_norm": 315.3016357421875, "learning_rate": 7.6e-06, "loss": 12.4643, "step": 76 }, { "epoch": 0.0006925301183147235, "grad_norm": 460.06549072265625, "learning_rate": 7.7e-06, "loss": 12.5043, "step": 77 }, { "epoch": 0.0007015240159551744, "grad_norm": 575.2285766601562, "learning_rate": 7.8e-06, "loss": 12.2469, "step": 78 }, { "epoch": 0.0007105179135956254, "grad_norm": 417.6985778808594, "learning_rate": 7.9e-06, "loss": 12.1922, "step": 79 }, { "epoch": 0.0007195118112360763, "grad_norm": 245.1527862548828, "learning_rate": 8.000000000000001e-06, "loss": 11.8656, "step": 80 }, { "epoch": 0.0007285057088765273, "grad_norm": 387.66522216796875, "learning_rate": 8.1e-06, "loss": 11.8198, "step": 81 }, { "epoch": 0.0007374996065169782, "grad_norm": 534.48583984375, "learning_rate": 8.200000000000001e-06, "loss": 11.6834, "step": 82 }, { "epoch": 0.0007464935041574292, "grad_norm": 525.0757446289062, "learning_rate": 8.3e-06, "loss": 11.4226, "step": 83 }, { "epoch": 0.0007554874017978801, "grad_norm": 795.5214233398438, "learning_rate": 8.400000000000001e-06, "loss": 11.5924, "step": 84 }, { "epoch": 0.0007644812994383311, "grad_norm": 342.18475341796875, "learning_rate": 8.500000000000002e-06, "loss": 11.4092, "step": 85 }, { "epoch": 0.000773475197078782, "grad_norm": 568.600341796875, "learning_rate": 8.599999999999999e-06, "loss": 11.0694, "step": 86 }, { "epoch": 0.000782469094719233, "grad_norm": 409.3172607421875, "learning_rate": 8.7e-06, "loss": 10.9358, "step": 87 }, { "epoch": 0.000791462992359684, "grad_norm": 293.0856018066406, "learning_rate": 8.8e-06, "loss": 10.7946, "step": 88 }, { "epoch": 0.000800456890000135, "grad_norm": 292.7913818359375, "learning_rate": 8.9e-06, "loss": 10.5607, "step": 89 }, { "epoch": 0.0008094507876405859, "grad_norm": 562.8467407226562, "learning_rate": 9e-06, "loss": 10.5687, "step": 90 }, { "epoch": 0.0008184446852810368, "grad_norm": 400.8448181152344, "learning_rate": 9.100000000000001e-06, "loss": 10.4255, "step": 91 }, { "epoch": 0.0008274385829214878, "grad_norm": 501.48236083984375, "learning_rate": 9.2e-06, "loss": 10.1359, "step": 92 }, { "epoch": 0.0008364324805619387, "grad_norm": 338.1478271484375, "learning_rate": 9.3e-06, "loss": 10.0752, "step": 93 }, { "epoch": 0.0008454263782023897, "grad_norm": 275.2373352050781, "learning_rate": 9.4e-06, "loss": 9.8534, "step": 94 }, { "epoch": 0.0008544202758428406, "grad_norm": 225.74720764160156, "learning_rate": 9.5e-06, "loss": 9.6955, "step": 95 }, { "epoch": 0.0008634141734832916, "grad_norm": 808.6383056640625, "learning_rate": 9.600000000000001e-06, "loss": 9.54, "step": 96 }, { "epoch": 0.0008724080711237425, "grad_norm": 671.9873657226562, "learning_rate": 9.7e-06, "loss": 9.2056, "step": 97 }, { "epoch": 0.0008814019687641935, "grad_norm": 373.6921081542969, "learning_rate": 9.800000000000001e-06, "loss": 9.0752, "step": 98 }, { "epoch": 0.0008903958664046444, "grad_norm": 437.3731994628906, "learning_rate": 9.900000000000002e-06, "loss": 8.7667, "step": 99 }, { "epoch": 0.0008993897640450954, "grad_norm": 241.6179962158203, "learning_rate": 1e-05, "loss": 8.5467, "step": 100 }, { "epoch": 0.0009083836616855463, "grad_norm": 306.84820556640625, "learning_rate": 1.0100000000000002e-05, "loss": 8.2923, "step": 101 }, { "epoch": 0.0009173775593259973, "grad_norm": 362.672607421875, "learning_rate": 1.02e-05, "loss": 7.997, "step": 102 }, { "epoch": 0.0009263714569664482, "grad_norm": 509.5506286621094, "learning_rate": 1.03e-05, "loss": 7.7283, "step": 103 }, { "epoch": 0.0009353653546068993, "grad_norm": 425.3578186035156, "learning_rate": 1.04e-05, "loss": 7.5146, "step": 104 }, { "epoch": 0.0009443592522473502, "grad_norm": 256.6763610839844, "learning_rate": 1.05e-05, "loss": 7.084, "step": 105 }, { "epoch": 0.0009533531498878012, "grad_norm": 469.381103515625, "learning_rate": 1.06e-05, "loss": 6.786, "step": 106 }, { "epoch": 0.0009623470475282521, "grad_norm": 342.9939270019531, "learning_rate": 1.0700000000000001e-05, "loss": 6.3985, "step": 107 }, { "epoch": 0.0009713409451687031, "grad_norm": 277.4405517578125, "learning_rate": 1.08e-05, "loss": 6.3308, "step": 108 }, { "epoch": 0.000980334842809154, "grad_norm": 214.1918182373047, "learning_rate": 1.09e-05, "loss": 5.9799, "step": 109 }, { "epoch": 0.0009893287404496049, "grad_norm": 139.2164306640625, "learning_rate": 1.1000000000000001e-05, "loss": 5.6918, "step": 110 }, { "epoch": 0.000998322638090056, "grad_norm": 206.45428466796875, "learning_rate": 1.11e-05, "loss": 5.6447, "step": 111 }, { "epoch": 0.001007316535730507, "grad_norm": 203.53302001953125, "learning_rate": 1.1200000000000001e-05, "loss": 5.3883, "step": 112 }, { "epoch": 0.001016310433370958, "grad_norm": 179.51927185058594, "learning_rate": 1.13e-05, "loss": 4.9607, "step": 113 }, { "epoch": 0.0010253043310114087, "grad_norm": 343.1466979980469, "learning_rate": 1.1400000000000001e-05, "loss": 4.7796, "step": 114 }, { "epoch": 0.0010342982286518597, "grad_norm": 361.537841796875, "learning_rate": 1.1500000000000002e-05, "loss": 4.6183, "step": 115 }, { "epoch": 0.0010432921262923107, "grad_norm": 192.5137939453125, "learning_rate": 1.16e-05, "loss": 4.5968, "step": 116 }, { "epoch": 0.0010522860239327615, "grad_norm": 227.02706909179688, "learning_rate": 1.1700000000000001e-05, "loss": 4.5182, "step": 117 }, { "epoch": 0.0010612799215732125, "grad_norm": 176.24400329589844, "learning_rate": 1.18e-05, "loss": 4.3656, "step": 118 }, { "epoch": 0.0010702738192136635, "grad_norm": 147.75003051757812, "learning_rate": 1.19e-05, "loss": 4.2635, "step": 119 }, { "epoch": 0.0010792677168541146, "grad_norm": 312.3755798339844, "learning_rate": 1.2e-05, "loss": 4.1145, "step": 120 }, { "epoch": 0.0010882616144945653, "grad_norm": 115.83843994140625, "learning_rate": 1.2100000000000001e-05, "loss": 4.1342, "step": 121 }, { "epoch": 0.0010972555121350164, "grad_norm": 87.58048248291016, "learning_rate": 1.22e-05, "loss": 3.998, "step": 122 }, { "epoch": 0.0011062494097754674, "grad_norm": 155.33116149902344, "learning_rate": 1.23e-05, "loss": 4.0172, "step": 123 }, { "epoch": 0.0011152433074159184, "grad_norm": 138.20004272460938, "learning_rate": 1.24e-05, "loss": 3.903, "step": 124 }, { "epoch": 0.0011242372050563692, "grad_norm": 244.00262451171875, "learning_rate": 1.25e-05, "loss": 3.9346, "step": 125 }, { "epoch": 0.0011332311026968202, "grad_norm": 213.140869140625, "learning_rate": 1.2600000000000001e-05, "loss": 3.7249, "step": 126 }, { "epoch": 0.0011422250003372712, "grad_norm": 187.2421112060547, "learning_rate": 1.27e-05, "loss": 3.8301, "step": 127 }, { "epoch": 0.0011512188979777222, "grad_norm": 296.37298583984375, "learning_rate": 1.2800000000000001e-05, "loss": 3.8237, "step": 128 }, { "epoch": 0.001160212795618173, "grad_norm": 192.2098388671875, "learning_rate": 1.29e-05, "loss": 3.5661, "step": 129 }, { "epoch": 0.001169206693258624, "grad_norm": 101.5536880493164, "learning_rate": 1.3000000000000001e-05, "loss": 3.6685, "step": 130 }, { "epoch": 0.001178200590899075, "grad_norm": 110.81195831298828, "learning_rate": 1.3100000000000002e-05, "loss": 3.6712, "step": 131 }, { "epoch": 0.001187194488539526, "grad_norm": 66.6405029296875, "learning_rate": 1.32e-05, "loss": 3.6038, "step": 132 }, { "epoch": 0.0011961883861799768, "grad_norm": 208.43743896484375, "learning_rate": 1.3300000000000001e-05, "loss": 3.5777, "step": 133 }, { "epoch": 0.0012051822838204278, "grad_norm": 113.67240142822266, "learning_rate": 1.3400000000000002e-05, "loss": 3.4868, "step": 134 }, { "epoch": 0.0012141761814608788, "grad_norm": 135.07162475585938, "learning_rate": 1.3500000000000001e-05, "loss": 3.564, "step": 135 }, { "epoch": 0.0012231700791013296, "grad_norm": 169.67071533203125, "learning_rate": 1.3600000000000002e-05, "loss": 3.526, "step": 136 }, { "epoch": 0.0012321639767417807, "grad_norm": 98.35337829589844, "learning_rate": 1.3700000000000001e-05, "loss": 3.4719, "step": 137 }, { "epoch": 0.0012411578743822317, "grad_norm": 241.689453125, "learning_rate": 1.3800000000000002e-05, "loss": 3.4388, "step": 138 }, { "epoch": 0.0012501517720226827, "grad_norm": 194.14877319335938, "learning_rate": 1.3900000000000002e-05, "loss": 3.3171, "step": 139 }, { "epoch": 0.0012591456696631335, "grad_norm": 196.54966735839844, "learning_rate": 1.4000000000000001e-05, "loss": 3.3183, "step": 140 }, { "epoch": 0.0012681395673035845, "grad_norm": 88.80168151855469, "learning_rate": 1.4099999999999999e-05, "loss": 3.3443, "step": 141 }, { "epoch": 0.0012771334649440355, "grad_norm": 91.37580871582031, "learning_rate": 1.42e-05, "loss": 3.1437, "step": 142 }, { "epoch": 0.0012861273625844865, "grad_norm": 128.65420532226562, "learning_rate": 1.43e-05, "loss": 3.3068, "step": 143 }, { "epoch": 0.0012951212602249373, "grad_norm": 500.7449035644531, "learning_rate": 1.44e-05, "loss": 3.3677, "step": 144 }, { "epoch": 0.0013041151578653883, "grad_norm": 158.20843505859375, "learning_rate": 1.45e-05, "loss": 3.2618, "step": 145 }, { "epoch": 0.0013131090555058393, "grad_norm": 85.0632553100586, "learning_rate": 1.4599999999999999e-05, "loss": 3.2044, "step": 146 }, { "epoch": 0.0013221029531462903, "grad_norm": 278.6473693847656, "learning_rate": 1.47e-05, "loss": 3.4244, "step": 147 }, { "epoch": 0.0013310968507867411, "grad_norm": 259.0623474121094, "learning_rate": 1.48e-05, "loss": 3.2486, "step": 148 }, { "epoch": 0.0013400907484271921, "grad_norm": 198.1902618408203, "learning_rate": 1.49e-05, "loss": 3.3153, "step": 149 }, { "epoch": 0.0013490846460676431, "grad_norm": 289.02923583984375, "learning_rate": 1.5e-05, "loss": 3.3178, "step": 150 }, { "epoch": 0.0013580785437080941, "grad_norm": 211.2479248046875, "learning_rate": 1.51e-05, "loss": 3.2842, "step": 151 }, { "epoch": 0.001367072441348545, "grad_norm": 186.1503448486328, "learning_rate": 1.52e-05, "loss": 3.1704, "step": 152 }, { "epoch": 0.001376066338988996, "grad_norm": 127.63630676269531, "learning_rate": 1.53e-05, "loss": 3.0608, "step": 153 }, { "epoch": 0.001385060236629447, "grad_norm": 138.69873046875, "learning_rate": 1.54e-05, "loss": 3.1819, "step": 154 }, { "epoch": 0.001394054134269898, "grad_norm": 296.7915954589844, "learning_rate": 1.55e-05, "loss": 3.1122, "step": 155 }, { "epoch": 0.0014030480319103488, "grad_norm": 226.61485290527344, "learning_rate": 1.56e-05, "loss": 2.9698, "step": 156 }, { "epoch": 0.0014120419295507998, "grad_norm": 227.7154083251953, "learning_rate": 1.5700000000000002e-05, "loss": 3.0942, "step": 157 }, { "epoch": 0.0014210358271912508, "grad_norm": 179.65577697753906, "learning_rate": 1.58e-05, "loss": 3.0883, "step": 158 }, { "epoch": 0.0014300297248317016, "grad_norm": 151.16664123535156, "learning_rate": 1.59e-05, "loss": 3.0361, "step": 159 }, { "epoch": 0.0014390236224721526, "grad_norm": 148.95591735839844, "learning_rate": 1.6000000000000003e-05, "loss": 3.0741, "step": 160 }, { "epoch": 0.0014480175201126036, "grad_norm": 824.3826293945312, "learning_rate": 1.6100000000000002e-05, "loss": 3.1651, "step": 161 }, { "epoch": 0.0014570114177530546, "grad_norm": 237.74867248535156, "learning_rate": 1.62e-05, "loss": 3.1489, "step": 162 }, { "epoch": 0.0014660053153935054, "grad_norm": 327.0044250488281, "learning_rate": 1.63e-05, "loss": 3.2912, "step": 163 }, { "epoch": 0.0014749992130339564, "grad_norm": 168.18836975097656, "learning_rate": 1.6400000000000002e-05, "loss": 3.1273, "step": 164 }, { "epoch": 0.0014839931106744074, "grad_norm": 149.23497009277344, "learning_rate": 1.65e-05, "loss": 2.9333, "step": 165 }, { "epoch": 0.0014929870083148584, "grad_norm": 69.95569610595703, "learning_rate": 1.66e-05, "loss": 2.9598, "step": 166 }, { "epoch": 0.0015019809059553092, "grad_norm": 81.76969909667969, "learning_rate": 1.6700000000000003e-05, "loss": 2.9935, "step": 167 }, { "epoch": 0.0015109748035957602, "grad_norm": 237.5215301513672, "learning_rate": 1.6800000000000002e-05, "loss": 3.0589, "step": 168 }, { "epoch": 0.0015199687012362113, "grad_norm": 126.43791961669922, "learning_rate": 1.69e-05, "loss": 2.8997, "step": 169 }, { "epoch": 0.0015289625988766623, "grad_norm": 72.03694915771484, "learning_rate": 1.7000000000000003e-05, "loss": 2.796, "step": 170 }, { "epoch": 0.001537956496517113, "grad_norm": 202.87074279785156, "learning_rate": 1.7100000000000002e-05, "loss": 2.9639, "step": 171 }, { "epoch": 0.001546950394157564, "grad_norm": 193.0821533203125, "learning_rate": 1.7199999999999998e-05, "loss": 3.0195, "step": 172 }, { "epoch": 0.001555944291798015, "grad_norm": 141.74383544921875, "learning_rate": 1.73e-05, "loss": 2.9638, "step": 173 }, { "epoch": 0.001564938189438466, "grad_norm": 107.39008331298828, "learning_rate": 1.74e-05, "loss": 3.0687, "step": 174 }, { "epoch": 0.0015739320870789169, "grad_norm": 265.2167663574219, "learning_rate": 1.75e-05, "loss": 3.0564, "step": 175 }, { "epoch": 0.001582925984719368, "grad_norm": 246.4090118408203, "learning_rate": 1.76e-05, "loss": 2.9282, "step": 176 }, { "epoch": 0.001591919882359819, "grad_norm": 88.30968475341797, "learning_rate": 1.77e-05, "loss": 2.8467, "step": 177 }, { "epoch": 0.00160091378000027, "grad_norm": 186.2428741455078, "learning_rate": 1.78e-05, "loss": 2.9758, "step": 178 }, { "epoch": 0.0016099076776407207, "grad_norm": 345.4529724121094, "learning_rate": 1.79e-05, "loss": 2.9958, "step": 179 }, { "epoch": 0.0016189015752811717, "grad_norm": 357.7126770019531, "learning_rate": 1.8e-05, "loss": 3.1002, "step": 180 }, { "epoch": 0.0016278954729216227, "grad_norm": 331.7604064941406, "learning_rate": 1.81e-05, "loss": 2.8368, "step": 181 }, { "epoch": 0.0016368893705620735, "grad_norm": 792.7327880859375, "learning_rate": 1.8200000000000002e-05, "loss": 3.0491, "step": 182 }, { "epoch": 0.0016458832682025245, "grad_norm": 271.6466979980469, "learning_rate": 1.83e-05, "loss": 2.8546, "step": 183 }, { "epoch": 0.0016548771658429755, "grad_norm": 132.30126953125, "learning_rate": 1.84e-05, "loss": 2.9186, "step": 184 }, { "epoch": 0.0016638710634834266, "grad_norm": 186.4729766845703, "learning_rate": 1.85e-05, "loss": 2.8325, "step": 185 }, { "epoch": 0.0016728649611238774, "grad_norm": 153.6566925048828, "learning_rate": 1.86e-05, "loss": 2.7955, "step": 186 }, { "epoch": 0.0016818588587643284, "grad_norm": 138.53509521484375, "learning_rate": 1.87e-05, "loss": 2.8964, "step": 187 }, { "epoch": 0.0016908527564047794, "grad_norm": 115.81424713134766, "learning_rate": 1.88e-05, "loss": 2.7509, "step": 188 }, { "epoch": 0.0016998466540452304, "grad_norm": 106.81310272216797, "learning_rate": 1.8900000000000002e-05, "loss": 2.8412, "step": 189 }, { "epoch": 0.0017088405516856812, "grad_norm": 127.3940658569336, "learning_rate": 1.9e-05, "loss": 2.7875, "step": 190 }, { "epoch": 0.0017178344493261322, "grad_norm": 88.74758911132812, "learning_rate": 1.91e-05, "loss": 2.7421, "step": 191 }, { "epoch": 0.0017268283469665832, "grad_norm": 107.35409545898438, "learning_rate": 1.9200000000000003e-05, "loss": 2.7106, "step": 192 }, { "epoch": 0.0017358222446070342, "grad_norm": 125.40448760986328, "learning_rate": 1.93e-05, "loss": 2.7194, "step": 193 }, { "epoch": 0.001744816142247485, "grad_norm": 233.53756713867188, "learning_rate": 1.94e-05, "loss": 2.9024, "step": 194 }, { "epoch": 0.001753810039887936, "grad_norm": 106.69174194335938, "learning_rate": 1.9500000000000003e-05, "loss": 2.7936, "step": 195 }, { "epoch": 0.001762803937528387, "grad_norm": 114.40642547607422, "learning_rate": 1.9600000000000002e-05, "loss": 2.7192, "step": 196 }, { "epoch": 0.001771797835168838, "grad_norm": 172.35498046875, "learning_rate": 1.97e-05, "loss": 2.7224, "step": 197 }, { "epoch": 0.0017807917328092888, "grad_norm": 58.3707275390625, "learning_rate": 1.9800000000000004e-05, "loss": 2.8107, "step": 198 }, { "epoch": 0.0017897856304497398, "grad_norm": 157.778564453125, "learning_rate": 1.9900000000000003e-05, "loss": 2.7716, "step": 199 }, { "epoch": 0.0017987795280901909, "grad_norm": 200.3644256591797, "learning_rate": 2e-05, "loss": 2.7948, "step": 200 }, { "epoch": 0.0018077734257306419, "grad_norm": 99.80152893066406, "learning_rate": 2.01e-05, "loss": 2.8006, "step": 201 }, { "epoch": 0.0018167673233710927, "grad_norm": 159.87576293945312, "learning_rate": 2.0200000000000003e-05, "loss": 2.8626, "step": 202 }, { "epoch": 0.0018257612210115437, "grad_norm": 150.257568359375, "learning_rate": 2.0300000000000002e-05, "loss": 2.8707, "step": 203 }, { "epoch": 0.0018347551186519947, "grad_norm": 328.46307373046875, "learning_rate": 2.04e-05, "loss": 2.783, "step": 204 }, { "epoch": 0.0018437490162924455, "grad_norm": 177.53663635253906, "learning_rate": 2.05e-05, "loss": 2.7279, "step": 205 }, { "epoch": 0.0018527429139328965, "grad_norm": 97.39747619628906, "learning_rate": 2.06e-05, "loss": 2.7084, "step": 206 }, { "epoch": 0.0018617368115733475, "grad_norm": 109.01382446289062, "learning_rate": 2.07e-05, "loss": 2.8108, "step": 207 }, { "epoch": 0.0018707307092137985, "grad_norm": 165.84030151367188, "learning_rate": 2.08e-05, "loss": 2.6219, "step": 208 }, { "epoch": 0.0018797246068542493, "grad_norm": 114.0225601196289, "learning_rate": 2.09e-05, "loss": 2.7262, "step": 209 }, { "epoch": 0.0018887185044947003, "grad_norm": 256.5831298828125, "learning_rate": 2.1e-05, "loss": 2.7321, "step": 210 }, { "epoch": 0.0018977124021351513, "grad_norm": 504.25469970703125, "learning_rate": 2.11e-05, "loss": 2.7968, "step": 211 }, { "epoch": 0.0019067062997756023, "grad_norm": 110.62533569335938, "learning_rate": 2.12e-05, "loss": 2.6275, "step": 212 }, { "epoch": 0.0019157001974160531, "grad_norm": 90.07482147216797, "learning_rate": 2.13e-05, "loss": 2.5958, "step": 213 }, { "epoch": 0.0019246940950565041, "grad_norm": 276.80059814453125, "learning_rate": 2.1400000000000002e-05, "loss": 2.6725, "step": 214 }, { "epoch": 0.0019336879926969551, "grad_norm": 229.91360473632812, "learning_rate": 2.15e-05, "loss": 2.8595, "step": 215 }, { "epoch": 0.0019426818903374062, "grad_norm": 309.5966491699219, "learning_rate": 2.16e-05, "loss": 2.788, "step": 216 }, { "epoch": 0.001951675787977857, "grad_norm": 110.78538513183594, "learning_rate": 2.1700000000000002e-05, "loss": 2.7053, "step": 217 }, { "epoch": 0.001960669685618308, "grad_norm": 254.88717651367188, "learning_rate": 2.18e-05, "loss": 2.742, "step": 218 }, { "epoch": 0.001969663583258759, "grad_norm": 306.1717224121094, "learning_rate": 2.19e-05, "loss": 2.7021, "step": 219 }, { "epoch": 0.0019786574808992098, "grad_norm": 248.65711975097656, "learning_rate": 2.2000000000000003e-05, "loss": 2.6783, "step": 220 }, { "epoch": 0.001987651378539661, "grad_norm": 202.61465454101562, "learning_rate": 2.2100000000000002e-05, "loss": 2.7195, "step": 221 }, { "epoch": 0.001996645276180112, "grad_norm": 114.76556396484375, "learning_rate": 2.22e-05, "loss": 2.7639, "step": 222 }, { "epoch": 0.0020056391738205626, "grad_norm": 92.22837829589844, "learning_rate": 2.23e-05, "loss": 2.7361, "step": 223 }, { "epoch": 0.002014633071461014, "grad_norm": 61.617618560791016, "learning_rate": 2.2400000000000002e-05, "loss": 2.5691, "step": 224 }, { "epoch": 0.0020236269691014646, "grad_norm": 80.87197875976562, "learning_rate": 2.25e-05, "loss": 2.6413, "step": 225 }, { "epoch": 0.002032620866741916, "grad_norm": 86.947509765625, "learning_rate": 2.26e-05, "loss": 2.7277, "step": 226 }, { "epoch": 0.0020416147643823666, "grad_norm": 73.94273376464844, "learning_rate": 2.2700000000000003e-05, "loss": 2.6458, "step": 227 }, { "epoch": 0.0020506086620228174, "grad_norm": 84.89952087402344, "learning_rate": 2.2800000000000002e-05, "loss": 2.6706, "step": 228 }, { "epoch": 0.0020596025596632686, "grad_norm": 44.78229522705078, "learning_rate": 2.29e-05, "loss": 2.6158, "step": 229 }, { "epoch": 0.0020685964573037194, "grad_norm": 88.20100402832031, "learning_rate": 2.3000000000000003e-05, "loss": 2.6753, "step": 230 }, { "epoch": 0.0020775903549441702, "grad_norm": 189.51966857910156, "learning_rate": 2.3100000000000002e-05, "loss": 2.6145, "step": 231 }, { "epoch": 0.0020865842525846215, "grad_norm": 125.79097747802734, "learning_rate": 2.32e-05, "loss": 2.7237, "step": 232 }, { "epoch": 0.0020955781502250722, "grad_norm": 143.98294067382812, "learning_rate": 2.3300000000000004e-05, "loss": 2.5877, "step": 233 }, { "epoch": 0.002104572047865523, "grad_norm": 109.03492736816406, "learning_rate": 2.3400000000000003e-05, "loss": 2.6428, "step": 234 }, { "epoch": 0.0021135659455059743, "grad_norm": 228.28143310546875, "learning_rate": 2.35e-05, "loss": 2.6777, "step": 235 }, { "epoch": 0.002122559843146425, "grad_norm": 287.90496826171875, "learning_rate": 2.36e-05, "loss": 2.8239, "step": 236 }, { "epoch": 0.0021315537407868763, "grad_norm": 100.6318359375, "learning_rate": 2.37e-05, "loss": 2.7823, "step": 237 }, { "epoch": 0.002140547638427327, "grad_norm": 322.9976501464844, "learning_rate": 2.38e-05, "loss": 2.7157, "step": 238 }, { "epoch": 0.002149541536067778, "grad_norm": 478.9703674316406, "learning_rate": 2.39e-05, "loss": 2.7828, "step": 239 }, { "epoch": 0.002158535433708229, "grad_norm": 361.4254455566406, "learning_rate": 2.4e-05, "loss": 2.7122, "step": 240 }, { "epoch": 0.00216752933134868, "grad_norm": 170.66314697265625, "learning_rate": 2.41e-05, "loss": 2.6334, "step": 241 }, { "epoch": 0.0021765232289891307, "grad_norm": 220.52621459960938, "learning_rate": 2.4200000000000002e-05, "loss": 2.6762, "step": 242 }, { "epoch": 0.002185517126629582, "grad_norm": 282.06072998046875, "learning_rate": 2.43e-05, "loss": 2.7003, "step": 243 }, { "epoch": 0.0021945110242700327, "grad_norm": 194.8751983642578, "learning_rate": 2.44e-05, "loss": 2.6338, "step": 244 }, { "epoch": 0.002203504921910484, "grad_norm": 141.71151733398438, "learning_rate": 2.45e-05, "loss": 2.6159, "step": 245 }, { "epoch": 0.0022124988195509347, "grad_norm": 103.0807876586914, "learning_rate": 2.46e-05, "loss": 2.622, "step": 246 }, { "epoch": 0.0022214927171913855, "grad_norm": 85.0677261352539, "learning_rate": 2.47e-05, "loss": 2.6077, "step": 247 }, { "epoch": 0.0022304866148318368, "grad_norm": 76.68571472167969, "learning_rate": 2.48e-05, "loss": 2.5066, "step": 248 }, { "epoch": 0.0022394805124722876, "grad_norm": 40.317840576171875, "learning_rate": 2.4900000000000002e-05, "loss": 2.7911, "step": 249 }, { "epoch": 0.0022484744101127383, "grad_norm": 44.18238067626953, "learning_rate": 2.5e-05, "loss": 2.546, "step": 250 }, { "epoch": 0.0022574683077531896, "grad_norm": 102.31619262695312, "learning_rate": 2.51e-05, "loss": 2.6973, "step": 251 }, { "epoch": 0.0022664622053936404, "grad_norm": 118.46347045898438, "learning_rate": 2.5200000000000003e-05, "loss": 2.6742, "step": 252 }, { "epoch": 0.002275456103034091, "grad_norm": 56.0390510559082, "learning_rate": 2.5300000000000002e-05, "loss": 2.6121, "step": 253 }, { "epoch": 0.0022844500006745424, "grad_norm": 68.61161041259766, "learning_rate": 2.54e-05, "loss": 2.5346, "step": 254 }, { "epoch": 0.002293443898314993, "grad_norm": 109.55023193359375, "learning_rate": 2.5500000000000003e-05, "loss": 2.6332, "step": 255 }, { "epoch": 0.0023024377959554444, "grad_norm": 54.4196662902832, "learning_rate": 2.5600000000000002e-05, "loss": 2.5449, "step": 256 }, { "epoch": 0.002311431693595895, "grad_norm": 74.51066589355469, "learning_rate": 2.57e-05, "loss": 2.6683, "step": 257 }, { "epoch": 0.002320425591236346, "grad_norm": 78.13221740722656, "learning_rate": 2.58e-05, "loss": 2.5201, "step": 258 }, { "epoch": 0.0023294194888767972, "grad_norm": 138.5928955078125, "learning_rate": 2.5900000000000003e-05, "loss": 2.6847, "step": 259 }, { "epoch": 0.002338413386517248, "grad_norm": 253.73043823242188, "learning_rate": 2.6000000000000002e-05, "loss": 2.6163, "step": 260 }, { "epoch": 0.002347407284157699, "grad_norm": 100.11920928955078, "learning_rate": 2.61e-05, "loss": 2.5562, "step": 261 }, { "epoch": 0.00235640118179815, "grad_norm": 81.5299301147461, "learning_rate": 2.6200000000000003e-05, "loss": 2.4997, "step": 262 }, { "epoch": 0.002365395079438601, "grad_norm": 61.319828033447266, "learning_rate": 2.6300000000000002e-05, "loss": 2.497, "step": 263 }, { "epoch": 0.002374388977079052, "grad_norm": 49.23329544067383, "learning_rate": 2.64e-05, "loss": 2.5356, "step": 264 }, { "epoch": 0.002383382874719503, "grad_norm": 70.633056640625, "learning_rate": 2.6500000000000004e-05, "loss": 2.4633, "step": 265 }, { "epoch": 0.0023923767723599536, "grad_norm": 67.63239288330078, "learning_rate": 2.6600000000000003e-05, "loss": 2.5644, "step": 266 }, { "epoch": 0.002401370670000405, "grad_norm": 92.0289077758789, "learning_rate": 2.6700000000000002e-05, "loss": 2.5142, "step": 267 }, { "epoch": 0.0024103645676408557, "grad_norm": 54.603271484375, "learning_rate": 2.6800000000000004e-05, "loss": 2.5737, "step": 268 }, { "epoch": 0.0024193584652813065, "grad_norm": 162.80580139160156, "learning_rate": 2.6900000000000003e-05, "loss": 2.5982, "step": 269 }, { "epoch": 0.0024283523629217577, "grad_norm": 76.19132232666016, "learning_rate": 2.7000000000000002e-05, "loss": 2.6659, "step": 270 }, { "epoch": 0.0024373462605622085, "grad_norm": 94.15829467773438, "learning_rate": 2.7100000000000005e-05, "loss": 2.6133, "step": 271 }, { "epoch": 0.0024463401582026593, "grad_norm": 51.162899017333984, "learning_rate": 2.7200000000000004e-05, "loss": 2.4114, "step": 272 }, { "epoch": 0.0024553340558431105, "grad_norm": 58.24684524536133, "learning_rate": 2.7300000000000003e-05, "loss": 2.5119, "step": 273 }, { "epoch": 0.0024643279534835613, "grad_norm": 37.575984954833984, "learning_rate": 2.7400000000000002e-05, "loss": 2.3714, "step": 274 }, { "epoch": 0.0024733218511240125, "grad_norm": 44.4429817199707, "learning_rate": 2.7500000000000004e-05, "loss": 2.4675, "step": 275 }, { "epoch": 0.0024823157487644633, "grad_norm": 67.38363647460938, "learning_rate": 2.7600000000000003e-05, "loss": 2.5588, "step": 276 }, { "epoch": 0.002491309646404914, "grad_norm": 196.1378173828125, "learning_rate": 2.7700000000000002e-05, "loss": 2.5487, "step": 277 }, { "epoch": 0.0025003035440453653, "grad_norm": 242.487060546875, "learning_rate": 2.7800000000000005e-05, "loss": 2.5146, "step": 278 }, { "epoch": 0.002509297441685816, "grad_norm": 132.01617431640625, "learning_rate": 2.7900000000000004e-05, "loss": 2.4936, "step": 279 }, { "epoch": 0.002518291339326267, "grad_norm": 31.95110511779785, "learning_rate": 2.8000000000000003e-05, "loss": 2.4226, "step": 280 }, { "epoch": 0.002527285236966718, "grad_norm": 138.77349853515625, "learning_rate": 2.8100000000000005e-05, "loss": 2.5149, "step": 281 }, { "epoch": 0.002536279134607169, "grad_norm": 316.9434509277344, "learning_rate": 2.8199999999999998e-05, "loss": 2.5951, "step": 282 }, { "epoch": 0.00254527303224762, "grad_norm": 271.0513916015625, "learning_rate": 2.83e-05, "loss": 2.6479, "step": 283 }, { "epoch": 0.002554266929888071, "grad_norm": 82.91835021972656, "learning_rate": 2.84e-05, "loss": 2.5587, "step": 284 }, { "epoch": 0.0025632608275285218, "grad_norm": 217.08937072753906, "learning_rate": 2.8499999999999998e-05, "loss": 2.5427, "step": 285 }, { "epoch": 0.002572254725168973, "grad_norm": 206.92457580566406, "learning_rate": 2.86e-05, "loss": 2.5698, "step": 286 }, { "epoch": 0.002581248622809424, "grad_norm": 124.58306121826172, "learning_rate": 2.87e-05, "loss": 2.4237, "step": 287 }, { "epoch": 0.0025902425204498746, "grad_norm": 93.71749877929688, "learning_rate": 2.88e-05, "loss": 2.4248, "step": 288 }, { "epoch": 0.002599236418090326, "grad_norm": 149.79690551757812, "learning_rate": 2.8899999999999998e-05, "loss": 2.483, "step": 289 }, { "epoch": 0.0026082303157307766, "grad_norm": 35.76045227050781, "learning_rate": 2.9e-05, "loss": 2.5009, "step": 290 }, { "epoch": 0.002617224213371228, "grad_norm": 205.15260314941406, "learning_rate": 2.91e-05, "loss": 2.5552, "step": 291 }, { "epoch": 0.0026262181110116786, "grad_norm": 121.84813690185547, "learning_rate": 2.9199999999999998e-05, "loss": 2.4251, "step": 292 }, { "epoch": 0.0026352120086521294, "grad_norm": 85.95841979980469, "learning_rate": 2.93e-05, "loss": 2.5326, "step": 293 }, { "epoch": 0.0026442059062925806, "grad_norm": 187.80039978027344, "learning_rate": 2.94e-05, "loss": 2.4564, "step": 294 }, { "epoch": 0.0026531998039330314, "grad_norm": 192.32235717773438, "learning_rate": 2.95e-05, "loss": 2.4292, "step": 295 }, { "epoch": 0.0026621937015734822, "grad_norm": 82.99427795410156, "learning_rate": 2.96e-05, "loss": 2.3344, "step": 296 }, { "epoch": 0.0026711875992139335, "grad_norm": 87.15036010742188, "learning_rate": 2.97e-05, "loss": 2.4989, "step": 297 }, { "epoch": 0.0026801814968543843, "grad_norm": 170.1896514892578, "learning_rate": 2.98e-05, "loss": 2.4031, "step": 298 }, { "epoch": 0.002689175394494835, "grad_norm": 102.06433868408203, "learning_rate": 2.9900000000000002e-05, "loss": 2.4162, "step": 299 }, { "epoch": 0.0026981692921352863, "grad_norm": 74.31095886230469, "learning_rate": 3e-05, "loss": 2.4387, "step": 300 }, { "epoch": 0.002707163189775737, "grad_norm": 54.91065979003906, "learning_rate": 3.01e-05, "loss": 2.3692, "step": 301 }, { "epoch": 0.0027161570874161883, "grad_norm": 52.14082717895508, "learning_rate": 3.02e-05, "loss": 2.4454, "step": 302 }, { "epoch": 0.002725150985056639, "grad_norm": 35.328407287597656, "learning_rate": 3.03e-05, "loss": 2.443, "step": 303 }, { "epoch": 0.00273414488269709, "grad_norm": 66.14423370361328, "learning_rate": 3.04e-05, "loss": 2.3643, "step": 304 }, { "epoch": 0.002743138780337541, "grad_norm": 153.15097045898438, "learning_rate": 3.05e-05, "loss": 2.2905, "step": 305 }, { "epoch": 0.002752132677977992, "grad_norm": 64.53681182861328, "learning_rate": 3.06e-05, "loss": 2.3349, "step": 306 }, { "epoch": 0.0027611265756184427, "grad_norm": 107.87183380126953, "learning_rate": 3.07e-05, "loss": 2.3087, "step": 307 }, { "epoch": 0.002770120473258894, "grad_norm": 62.940032958984375, "learning_rate": 3.08e-05, "loss": 2.3641, "step": 308 }, { "epoch": 0.0027791143708993447, "grad_norm": 100.97374725341797, "learning_rate": 3.09e-05, "loss": 2.3249, "step": 309 }, { "epoch": 0.002788108268539796, "grad_norm": 56.232513427734375, "learning_rate": 3.1e-05, "loss": 2.3902, "step": 310 }, { "epoch": 0.0027971021661802467, "grad_norm": 26.233421325683594, "learning_rate": 3.1100000000000004e-05, "loss": 2.3103, "step": 311 }, { "epoch": 0.0028060960638206975, "grad_norm": 150.95115661621094, "learning_rate": 3.12e-05, "loss": 2.3902, "step": 312 }, { "epoch": 0.0028150899614611488, "grad_norm": 269.62030029296875, "learning_rate": 3.13e-05, "loss": 2.4161, "step": 313 }, { "epoch": 0.0028240838591015996, "grad_norm": 62.62043380737305, "learning_rate": 3.1400000000000004e-05, "loss": 2.3591, "step": 314 }, { "epoch": 0.0028330777567420504, "grad_norm": 32.369964599609375, "learning_rate": 3.15e-05, "loss": 2.4911, "step": 315 }, { "epoch": 0.0028420716543825016, "grad_norm": 52.81867218017578, "learning_rate": 3.16e-05, "loss": 2.4109, "step": 316 }, { "epoch": 0.0028510655520229524, "grad_norm": 85.33836364746094, "learning_rate": 3.1700000000000005e-05, "loss": 2.3174, "step": 317 }, { "epoch": 0.002860059449663403, "grad_norm": 40.911502838134766, "learning_rate": 3.18e-05, "loss": 2.3609, "step": 318 }, { "epoch": 0.0028690533473038544, "grad_norm": 26.358247756958008, "learning_rate": 3.19e-05, "loss": 2.4348, "step": 319 }, { "epoch": 0.002878047244944305, "grad_norm": 70.71055603027344, "learning_rate": 3.2000000000000005e-05, "loss": 2.3038, "step": 320 }, { "epoch": 0.0028870411425847564, "grad_norm": 60.44719696044922, "learning_rate": 3.21e-05, "loss": 2.3875, "step": 321 }, { "epoch": 0.002896035040225207, "grad_norm": 187.33462524414062, "learning_rate": 3.2200000000000003e-05, "loss": 2.4148, "step": 322 }, { "epoch": 0.002905028937865658, "grad_norm": 122.424560546875, "learning_rate": 3.2300000000000006e-05, "loss": 2.316, "step": 323 }, { "epoch": 0.0029140228355061092, "grad_norm": 65.63255310058594, "learning_rate": 3.24e-05, "loss": 2.3456, "step": 324 }, { "epoch": 0.00292301673314656, "grad_norm": 43.2950553894043, "learning_rate": 3.2500000000000004e-05, "loss": 2.3846, "step": 325 }, { "epoch": 0.002932010630787011, "grad_norm": 130.36334228515625, "learning_rate": 3.26e-05, "loss": 2.3822, "step": 326 }, { "epoch": 0.002941004528427462, "grad_norm": 59.44133758544922, "learning_rate": 3.27e-05, "loss": 2.3209, "step": 327 }, { "epoch": 0.002949998426067913, "grad_norm": 82.3916244506836, "learning_rate": 3.2800000000000004e-05, "loss": 2.3491, "step": 328 }, { "epoch": 0.002958992323708364, "grad_norm": 105.73941040039062, "learning_rate": 3.29e-05, "loss": 2.4125, "step": 329 }, { "epoch": 0.002967986221348815, "grad_norm": 51.09678649902344, "learning_rate": 3.3e-05, "loss": 2.4083, "step": 330 }, { "epoch": 0.0029769801189892657, "grad_norm": 36.366546630859375, "learning_rate": 3.3100000000000005e-05, "loss": 2.2821, "step": 331 }, { "epoch": 0.002985974016629717, "grad_norm": 95.54082489013672, "learning_rate": 3.32e-05, "loss": 2.2432, "step": 332 }, { "epoch": 0.0029949679142701677, "grad_norm": 48.38098907470703, "learning_rate": 3.33e-05, "loss": 2.4007, "step": 333 }, { "epoch": 0.0030039618119106185, "grad_norm": 106.1603012084961, "learning_rate": 3.3400000000000005e-05, "loss": 2.4442, "step": 334 }, { "epoch": 0.0030129557095510697, "grad_norm": 111.94950103759766, "learning_rate": 3.35e-05, "loss": 2.3875, "step": 335 }, { "epoch": 0.0030219496071915205, "grad_norm": 126.90693664550781, "learning_rate": 3.3600000000000004e-05, "loss": 2.3551, "step": 336 }, { "epoch": 0.0030309435048319717, "grad_norm": 79.47080993652344, "learning_rate": 3.3700000000000006e-05, "loss": 2.2794, "step": 337 }, { "epoch": 0.0030399374024724225, "grad_norm": 28.410152435302734, "learning_rate": 3.38e-05, "loss": 2.3043, "step": 338 }, { "epoch": 0.0030489313001128733, "grad_norm": 19.942371368408203, "learning_rate": 3.3900000000000004e-05, "loss": 2.2162, "step": 339 }, { "epoch": 0.0030579251977533245, "grad_norm": 37.186851501464844, "learning_rate": 3.4000000000000007e-05, "loss": 2.434, "step": 340 }, { "epoch": 0.0030669190953937753, "grad_norm": 35.834983825683594, "learning_rate": 3.41e-05, "loss": 2.3768, "step": 341 }, { "epoch": 0.003075912993034226, "grad_norm": 201.9110870361328, "learning_rate": 3.4200000000000005e-05, "loss": 2.3252, "step": 342 }, { "epoch": 0.0030849068906746773, "grad_norm": 179.37872314453125, "learning_rate": 3.430000000000001e-05, "loss": 2.3229, "step": 343 }, { "epoch": 0.003093900788315128, "grad_norm": 127.170654296875, "learning_rate": 3.4399999999999996e-05, "loss": 2.3756, "step": 344 }, { "epoch": 0.003102894685955579, "grad_norm": 29.71765899658203, "learning_rate": 3.45e-05, "loss": 2.2425, "step": 345 }, { "epoch": 0.00311188858359603, "grad_norm": 127.28743743896484, "learning_rate": 3.46e-05, "loss": 2.2983, "step": 346 }, { "epoch": 0.003120882481236481, "grad_norm": 549.4426879882812, "learning_rate": 3.4699999999999996e-05, "loss": 2.6578, "step": 347 }, { "epoch": 0.003129876378876932, "grad_norm": 566.208251953125, "learning_rate": 3.48e-05, "loss": 2.6818, "step": 348 }, { "epoch": 0.003138870276517383, "grad_norm": 619.4710693359375, "learning_rate": 3.49e-05, "loss": 2.4366, "step": 349 }, { "epoch": 0.0031478641741578338, "grad_norm": 24.49801254272461, "learning_rate": 3.5e-05, "loss": 2.2666, "step": 350 }, { "epoch": 0.003156858071798285, "grad_norm": 144.26083374023438, "learning_rate": 3.51e-05, "loss": 2.3359, "step": 351 }, { "epoch": 0.003165851969438736, "grad_norm": 321.1274719238281, "learning_rate": 3.52e-05, "loss": 2.3761, "step": 352 }, { "epoch": 0.0031748458670791866, "grad_norm": 211.29849243164062, "learning_rate": 3.53e-05, "loss": 2.4916, "step": 353 }, { "epoch": 0.003183839764719638, "grad_norm": 70.34711456298828, "learning_rate": 3.54e-05, "loss": 2.4465, "step": 354 }, { "epoch": 0.0031928336623600886, "grad_norm": 53.300262451171875, "learning_rate": 3.55e-05, "loss": 2.3951, "step": 355 }, { "epoch": 0.00320182756000054, "grad_norm": 46.30522155761719, "learning_rate": 3.56e-05, "loss": 2.334, "step": 356 }, { "epoch": 0.0032108214576409906, "grad_norm": 42.729610443115234, "learning_rate": 3.57e-05, "loss": 2.3471, "step": 357 }, { "epoch": 0.0032198153552814414, "grad_norm": 79.16353607177734, "learning_rate": 3.58e-05, "loss": 2.333, "step": 358 }, { "epoch": 0.0032288092529218927, "grad_norm": 25.005178451538086, "learning_rate": 3.59e-05, "loss": 2.2444, "step": 359 }, { "epoch": 0.0032378031505623434, "grad_norm": 32.197269439697266, "learning_rate": 3.6e-05, "loss": 2.3087, "step": 360 }, { "epoch": 0.0032467970482027942, "grad_norm": 18.117626190185547, "learning_rate": 3.61e-05, "loss": 2.3144, "step": 361 }, { "epoch": 0.0032557909458432455, "grad_norm": 29.421463012695312, "learning_rate": 3.62e-05, "loss": 2.2462, "step": 362 }, { "epoch": 0.0032647848434836963, "grad_norm": 77.9415283203125, "learning_rate": 3.63e-05, "loss": 2.27, "step": 363 }, { "epoch": 0.003273778741124147, "grad_norm": 202.46302795410156, "learning_rate": 3.6400000000000004e-05, "loss": 2.3681, "step": 364 }, { "epoch": 0.0032827726387645983, "grad_norm": 44.71063232421875, "learning_rate": 3.65e-05, "loss": 2.2395, "step": 365 }, { "epoch": 0.003291766536405049, "grad_norm": 19.18665313720703, "learning_rate": 3.66e-05, "loss": 2.3578, "step": 366 }, { "epoch": 0.0033007604340455003, "grad_norm": 228.6815948486328, "learning_rate": 3.6700000000000004e-05, "loss": 2.2739, "step": 367 }, { "epoch": 0.003309754331685951, "grad_norm": 46.618995666503906, "learning_rate": 3.68e-05, "loss": 2.24, "step": 368 }, { "epoch": 0.003318748229326402, "grad_norm": 26.15361976623535, "learning_rate": 3.69e-05, "loss": 2.2431, "step": 369 }, { "epoch": 0.003327742126966853, "grad_norm": 159.9843292236328, "learning_rate": 3.7e-05, "loss": 2.2679, "step": 370 }, { "epoch": 0.003336736024607304, "grad_norm": 242.16363525390625, "learning_rate": 3.71e-05, "loss": 2.3197, "step": 371 }, { "epoch": 0.0033457299222477547, "grad_norm": 161.19627380371094, "learning_rate": 3.72e-05, "loss": 2.1869, "step": 372 }, { "epoch": 0.003354723819888206, "grad_norm": 43.067405700683594, "learning_rate": 3.73e-05, "loss": 2.3733, "step": 373 }, { "epoch": 0.0033637177175286567, "grad_norm": 55.32299041748047, "learning_rate": 3.74e-05, "loss": 2.3045, "step": 374 }, { "epoch": 0.003372711615169108, "grad_norm": 59.884891510009766, "learning_rate": 3.7500000000000003e-05, "loss": 2.2802, "step": 375 }, { "epoch": 0.0033817055128095587, "grad_norm": 28.28325653076172, "learning_rate": 3.76e-05, "loss": 2.2184, "step": 376 }, { "epoch": 0.0033906994104500095, "grad_norm": 15.199847221374512, "learning_rate": 3.77e-05, "loss": 2.2706, "step": 377 }, { "epoch": 0.0033996933080904608, "grad_norm": 102.14122772216797, "learning_rate": 3.7800000000000004e-05, "loss": 2.3356, "step": 378 }, { "epoch": 0.0034086872057309116, "grad_norm": 76.07613372802734, "learning_rate": 3.79e-05, "loss": 2.3538, "step": 379 }, { "epoch": 0.0034176811033713624, "grad_norm": 17.340497970581055, "learning_rate": 3.8e-05, "loss": 2.3015, "step": 380 }, { "epoch": 0.0034266750010118136, "grad_norm": 21.418004989624023, "learning_rate": 3.8100000000000005e-05, "loss": 2.1243, "step": 381 }, { "epoch": 0.0034356688986522644, "grad_norm": 40.507408142089844, "learning_rate": 3.82e-05, "loss": 2.2708, "step": 382 }, { "epoch": 0.0034446627962927156, "grad_norm": 64.36341094970703, "learning_rate": 3.83e-05, "loss": 2.2871, "step": 383 }, { "epoch": 0.0034536566939331664, "grad_norm": 35.082298278808594, "learning_rate": 3.8400000000000005e-05, "loss": 2.2848, "step": 384 }, { "epoch": 0.003462650591573617, "grad_norm": 49.866024017333984, "learning_rate": 3.85e-05, "loss": 2.2297, "step": 385 }, { "epoch": 0.0034716444892140684, "grad_norm": 22.154539108276367, "learning_rate": 3.86e-05, "loss": 2.2187, "step": 386 }, { "epoch": 0.003480638386854519, "grad_norm": 212.34315490722656, "learning_rate": 3.8700000000000006e-05, "loss": 2.2701, "step": 387 }, { "epoch": 0.00348963228449497, "grad_norm": 382.63592529296875, "learning_rate": 3.88e-05, "loss": 2.3644, "step": 388 }, { "epoch": 0.0034986261821354212, "grad_norm": 354.2323303222656, "learning_rate": 3.8900000000000004e-05, "loss": 2.4371, "step": 389 }, { "epoch": 0.003507620079775872, "grad_norm": 71.48065948486328, "learning_rate": 3.9000000000000006e-05, "loss": 2.3195, "step": 390 }, { "epoch": 0.003516613977416323, "grad_norm": 576.2838134765625, "learning_rate": 3.91e-05, "loss": 2.6219, "step": 391 }, { "epoch": 0.003525607875056774, "grad_norm": 619.9407958984375, "learning_rate": 3.9200000000000004e-05, "loss": 2.5243, "step": 392 }, { "epoch": 0.003534601772697225, "grad_norm": 179.0673065185547, "learning_rate": 3.9300000000000007e-05, "loss": 2.3726, "step": 393 }, { "epoch": 0.003543595670337676, "grad_norm": 57.12882995605469, "learning_rate": 3.94e-05, "loss": 2.2456, "step": 394 }, { "epoch": 0.003552589567978127, "grad_norm": 14.169979095458984, "learning_rate": 3.9500000000000005e-05, "loss": 2.2046, "step": 395 }, { "epoch": 0.0035615834656185777, "grad_norm": 63.82403564453125, "learning_rate": 3.960000000000001e-05, "loss": 2.2331, "step": 396 }, { "epoch": 0.003570577363259029, "grad_norm": 24.501937866210938, "learning_rate": 3.97e-05, "loss": 2.2535, "step": 397 }, { "epoch": 0.0035795712608994797, "grad_norm": 14.672894477844238, "learning_rate": 3.9800000000000005e-05, "loss": 2.2013, "step": 398 }, { "epoch": 0.0035885651585399305, "grad_norm": 34.786312103271484, "learning_rate": 3.99e-05, "loss": 2.3616, "step": 399 }, { "epoch": 0.0035975590561803817, "grad_norm": 17.01335334777832, "learning_rate": 4e-05, "loss": 2.2084, "step": 400 }, { "epoch": 0.0036065529538208325, "grad_norm": 19.66499900817871, "learning_rate": 4.0100000000000006e-05, "loss": 2.223, "step": 401 }, { "epoch": 0.0036155468514612837, "grad_norm": 47.78734588623047, "learning_rate": 4.02e-05, "loss": 2.1818, "step": 402 }, { "epoch": 0.0036245407491017345, "grad_norm": 18.995853424072266, "learning_rate": 4.0300000000000004e-05, "loss": 2.1758, "step": 403 }, { "epoch": 0.0036335346467421853, "grad_norm": 200.0352325439453, "learning_rate": 4.0400000000000006e-05, "loss": 2.2889, "step": 404 }, { "epoch": 0.0036425285443826365, "grad_norm": 55.98514175415039, "learning_rate": 4.05e-05, "loss": 2.2557, "step": 405 }, { "epoch": 0.0036515224420230873, "grad_norm": 26.741191864013672, "learning_rate": 4.0600000000000004e-05, "loss": 2.3365, "step": 406 }, { "epoch": 0.003660516339663538, "grad_norm": 49.336875915527344, "learning_rate": 4.07e-05, "loss": 2.2778, "step": 407 }, { "epoch": 0.0036695102373039894, "grad_norm": 68.999267578125, "learning_rate": 4.08e-05, "loss": 2.1965, "step": 408 }, { "epoch": 0.00367850413494444, "grad_norm": 38.1204833984375, "learning_rate": 4.09e-05, "loss": 2.3461, "step": 409 }, { "epoch": 0.003687498032584891, "grad_norm": 25.255964279174805, "learning_rate": 4.1e-05, "loss": 2.3932, "step": 410 }, { "epoch": 0.003696491930225342, "grad_norm": 17.3624210357666, "learning_rate": 4.11e-05, "loss": 2.2458, "step": 411 }, { "epoch": 0.003705485827865793, "grad_norm": 19.3513240814209, "learning_rate": 4.12e-05, "loss": 2.2804, "step": 412 }, { "epoch": 0.003714479725506244, "grad_norm": 23.91090965270996, "learning_rate": 4.13e-05, "loss": 2.2758, "step": 413 }, { "epoch": 0.003723473623146695, "grad_norm": 32.457740783691406, "learning_rate": 4.14e-05, "loss": 2.3667, "step": 414 }, { "epoch": 0.0037324675207871458, "grad_norm": 42.071197509765625, "learning_rate": 4.15e-05, "loss": 2.2722, "step": 415 }, { "epoch": 0.003741461418427597, "grad_norm": 21.174034118652344, "learning_rate": 4.16e-05, "loss": 2.3652, "step": 416 }, { "epoch": 0.003750455316068048, "grad_norm": 16.781293869018555, "learning_rate": 4.17e-05, "loss": 2.264, "step": 417 }, { "epoch": 0.0037594492137084986, "grad_norm": 14.720979690551758, "learning_rate": 4.18e-05, "loss": 2.2801, "step": 418 }, { "epoch": 0.00376844311134895, "grad_norm": 45.68321990966797, "learning_rate": 4.19e-05, "loss": 2.2097, "step": 419 }, { "epoch": 0.0037774370089894006, "grad_norm": 13.768199920654297, "learning_rate": 4.2e-05, "loss": 2.3145, "step": 420 }, { "epoch": 0.003786430906629852, "grad_norm": 26.640262603759766, "learning_rate": 4.21e-05, "loss": 2.2069, "step": 421 }, { "epoch": 0.0037954248042703026, "grad_norm": 54.33475875854492, "learning_rate": 4.22e-05, "loss": 2.3673, "step": 422 }, { "epoch": 0.0038044187019107534, "grad_norm": 21.746578216552734, "learning_rate": 4.23e-05, "loss": 2.3421, "step": 423 }, { "epoch": 0.0038134125995512047, "grad_norm": 240.3916778564453, "learning_rate": 4.24e-05, "loss": 2.4255, "step": 424 }, { "epoch": 0.0038224064971916554, "grad_norm": 313.048828125, "learning_rate": 4.25e-05, "loss": 2.2633, "step": 425 }, { "epoch": 0.0038314003948321062, "grad_norm": 118.0916519165039, "learning_rate": 4.26e-05, "loss": 2.2051, "step": 426 }, { "epoch": 0.0038403942924725575, "grad_norm": 32.71201705932617, "learning_rate": 4.27e-05, "loss": 2.3556, "step": 427 }, { "epoch": 0.0038493881901130083, "grad_norm": 90.62095642089844, "learning_rate": 4.2800000000000004e-05, "loss": 2.2523, "step": 428 }, { "epoch": 0.0038583820877534595, "grad_norm": 253.30313110351562, "learning_rate": 4.29e-05, "loss": 2.4426, "step": 429 }, { "epoch": 0.0038673759853939103, "grad_norm": 68.32054901123047, "learning_rate": 4.3e-05, "loss": 2.2475, "step": 430 }, { "epoch": 0.003876369883034361, "grad_norm": 29.561153411865234, "learning_rate": 4.3100000000000004e-05, "loss": 2.2463, "step": 431 }, { "epoch": 0.0038853637806748123, "grad_norm": 47.32217025756836, "learning_rate": 4.32e-05, "loss": 2.2842, "step": 432 }, { "epoch": 0.003894357678315263, "grad_norm": 133.11419677734375, "learning_rate": 4.33e-05, "loss": 2.2752, "step": 433 }, { "epoch": 0.003903351575955714, "grad_norm": 31.708168029785156, "learning_rate": 4.3400000000000005e-05, "loss": 2.2407, "step": 434 }, { "epoch": 0.003912345473596165, "grad_norm": 23.353534698486328, "learning_rate": 4.35e-05, "loss": 2.2194, "step": 435 }, { "epoch": 0.003921339371236616, "grad_norm": 486.08489990234375, "learning_rate": 4.36e-05, "loss": 2.2926, "step": 436 }, { "epoch": 0.003930333268877067, "grad_norm": 794.3959350585938, "learning_rate": 4.3700000000000005e-05, "loss": 2.6735, "step": 437 }, { "epoch": 0.003939327166517518, "grad_norm": 556.8062133789062, "learning_rate": 4.38e-05, "loss": 2.5162, "step": 438 }, { "epoch": 0.003948321064157969, "grad_norm": 327.8409729003906, "learning_rate": 4.39e-05, "loss": 2.3317, "step": 439 }, { "epoch": 0.0039573149617984195, "grad_norm": 24.79322624206543, "learning_rate": 4.4000000000000006e-05, "loss": 2.2953, "step": 440 }, { "epoch": 0.003966308859438871, "grad_norm": 12.116776466369629, "learning_rate": 4.41e-05, "loss": 2.193, "step": 441 }, { "epoch": 0.003975302757079322, "grad_norm": 132.1576385498047, "learning_rate": 4.4200000000000004e-05, "loss": 2.2019, "step": 442 }, { "epoch": 0.003984296654719772, "grad_norm": 191.4755096435547, "learning_rate": 4.43e-05, "loss": 2.3374, "step": 443 }, { "epoch": 0.003993290552360224, "grad_norm": 37.85381317138672, "learning_rate": 4.44e-05, "loss": 2.2976, "step": 444 }, { "epoch": 0.004002284450000675, "grad_norm": 17.471384048461914, "learning_rate": 4.4500000000000004e-05, "loss": 2.2783, "step": 445 }, { "epoch": 0.004011278347641125, "grad_norm": 11.133999824523926, "learning_rate": 4.46e-05, "loss": 2.2805, "step": 446 }, { "epoch": 0.004020272245281576, "grad_norm": 18.621841430664062, "learning_rate": 4.47e-05, "loss": 2.2603, "step": 447 }, { "epoch": 0.004029266142922028, "grad_norm": 16.253196716308594, "learning_rate": 4.4800000000000005e-05, "loss": 2.2795, "step": 448 }, { "epoch": 0.004038260040562478, "grad_norm": 34.07244110107422, "learning_rate": 4.49e-05, "loss": 2.2827, "step": 449 }, { "epoch": 0.004047253938202929, "grad_norm": 17.81850242614746, "learning_rate": 4.5e-05, "loss": 2.222, "step": 450 }, { "epoch": 0.00405624783584338, "grad_norm": 7.954962730407715, "learning_rate": 4.5100000000000005e-05, "loss": 2.1639, "step": 451 }, { "epoch": 0.004065241733483832, "grad_norm": 15.723968505859375, "learning_rate": 4.52e-05, "loss": 2.2212, "step": 452 }, { "epoch": 0.004074235631124282, "grad_norm": 23.19266700744629, "learning_rate": 4.53e-05, "loss": 2.2177, "step": 453 }, { "epoch": 0.004083229528764733, "grad_norm": 9.00516128540039, "learning_rate": 4.5400000000000006e-05, "loss": 2.2727, "step": 454 }, { "epoch": 0.0040922234264051845, "grad_norm": 21.012422561645508, "learning_rate": 4.55e-05, "loss": 2.2464, "step": 455 }, { "epoch": 0.004101217324045635, "grad_norm": 10.173611640930176, "learning_rate": 4.5600000000000004e-05, "loss": 2.1522, "step": 456 }, { "epoch": 0.004110211221686086, "grad_norm": 13.564419746398926, "learning_rate": 4.5700000000000006e-05, "loss": 2.2215, "step": 457 }, { "epoch": 0.004119205119326537, "grad_norm": 11.415655136108398, "learning_rate": 4.58e-05, "loss": 2.2152, "step": 458 }, { "epoch": 0.004128199016966988, "grad_norm": 19.338844299316406, "learning_rate": 4.5900000000000004e-05, "loss": 2.1804, "step": 459 }, { "epoch": 0.004137192914607439, "grad_norm": 25.48905372619629, "learning_rate": 4.600000000000001e-05, "loss": 2.276, "step": 460 }, { "epoch": 0.00414618681224789, "grad_norm": 45.079864501953125, "learning_rate": 4.61e-05, "loss": 2.2027, "step": 461 }, { "epoch": 0.0041551807098883405, "grad_norm": 39.53298568725586, "learning_rate": 4.6200000000000005e-05, "loss": 2.1816, "step": 462 }, { "epoch": 0.004164174607528792, "grad_norm": 18.841827392578125, "learning_rate": 4.630000000000001e-05, "loss": 2.2352, "step": 463 }, { "epoch": 0.004173168505169243, "grad_norm": 5.895425319671631, "learning_rate": 4.64e-05, "loss": 2.2828, "step": 464 }, { "epoch": 0.004182162402809693, "grad_norm": 19.19552993774414, "learning_rate": 4.6500000000000005e-05, "loss": 2.2895, "step": 465 }, { "epoch": 0.0041911563004501445, "grad_norm": 27.967140197753906, "learning_rate": 4.660000000000001e-05, "loss": 2.1892, "step": 466 }, { "epoch": 0.004200150198090596, "grad_norm": 12.54835033416748, "learning_rate": 4.6700000000000003e-05, "loss": 2.1776, "step": 467 }, { "epoch": 0.004209144095731046, "grad_norm": 14.965579986572266, "learning_rate": 4.6800000000000006e-05, "loss": 2.2245, "step": 468 }, { "epoch": 0.004218137993371497, "grad_norm": 10.855680465698242, "learning_rate": 4.69e-05, "loss": 2.1747, "step": 469 }, { "epoch": 0.0042271318910119485, "grad_norm": 15.61082935333252, "learning_rate": 4.7e-05, "loss": 2.1996, "step": 470 }, { "epoch": 0.0042361257886524, "grad_norm": 16.495397567749023, "learning_rate": 4.71e-05, "loss": 2.1717, "step": 471 }, { "epoch": 0.00424511968629285, "grad_norm": 13.10986328125, "learning_rate": 4.72e-05, "loss": 2.2007, "step": 472 }, { "epoch": 0.004254113583933301, "grad_norm": 11.138365745544434, "learning_rate": 4.73e-05, "loss": 2.0885, "step": 473 }, { "epoch": 0.004263107481573753, "grad_norm": 26.061817169189453, "learning_rate": 4.74e-05, "loss": 2.1314, "step": 474 }, { "epoch": 0.004272101379214203, "grad_norm": 10.8087739944458, "learning_rate": 4.75e-05, "loss": 2.1964, "step": 475 }, { "epoch": 0.004281095276854654, "grad_norm": 24.03231430053711, "learning_rate": 4.76e-05, "loss": 2.0867, "step": 476 }, { "epoch": 0.004290089174495105, "grad_norm": 114.8991470336914, "learning_rate": 4.77e-05, "loss": 2.1504, "step": 477 }, { "epoch": 0.004299083072135556, "grad_norm": 40.11175537109375, "learning_rate": 4.78e-05, "loss": 2.3191, "step": 478 }, { "epoch": 0.004308076969776007, "grad_norm": 173.28294372558594, "learning_rate": 4.79e-05, "loss": 2.2462, "step": 479 }, { "epoch": 0.004317070867416458, "grad_norm": 17.962095260620117, "learning_rate": 4.8e-05, "loss": 2.1752, "step": 480 }, { "epoch": 0.004326064765056909, "grad_norm": 25.65735626220703, "learning_rate": 4.8100000000000004e-05, "loss": 2.1044, "step": 481 }, { "epoch": 0.00433505866269736, "grad_norm": 8.496695518493652, "learning_rate": 4.82e-05, "loss": 2.2002, "step": 482 }, { "epoch": 0.004344052560337811, "grad_norm": 74.04930877685547, "learning_rate": 4.83e-05, "loss": 2.1698, "step": 483 }, { "epoch": 0.004353046457978261, "grad_norm": 49.60053253173828, "learning_rate": 4.8400000000000004e-05, "loss": 2.2432, "step": 484 }, { "epoch": 0.004362040355618713, "grad_norm": 25.82587432861328, "learning_rate": 4.85e-05, "loss": 2.1942, "step": 485 }, { "epoch": 0.004371034253259164, "grad_norm": 287.9582214355469, "learning_rate": 4.86e-05, "loss": 2.2702, "step": 486 }, { "epoch": 0.004380028150899614, "grad_norm": 249.64727783203125, "learning_rate": 4.87e-05, "loss": 2.2571, "step": 487 }, { "epoch": 0.004389022048540065, "grad_norm": 65.73006439208984, "learning_rate": 4.88e-05, "loss": 2.2373, "step": 488 }, { "epoch": 0.004398015946180517, "grad_norm": 24.812101364135742, "learning_rate": 4.89e-05, "loss": 2.1121, "step": 489 }, { "epoch": 0.004407009843820968, "grad_norm": 41.44482421875, "learning_rate": 4.9e-05, "loss": 2.1275, "step": 490 }, { "epoch": 0.004416003741461418, "grad_norm": 34.90840148925781, "learning_rate": 4.91e-05, "loss": 2.2112, "step": 491 }, { "epoch": 0.0044249976391018695, "grad_norm": 23.77524185180664, "learning_rate": 4.92e-05, "loss": 2.1756, "step": 492 }, { "epoch": 0.004433991536742321, "grad_norm": 38.70237350463867, "learning_rate": 4.93e-05, "loss": 2.2658, "step": 493 }, { "epoch": 0.004442985434382771, "grad_norm": 7.088588237762451, "learning_rate": 4.94e-05, "loss": 2.1339, "step": 494 }, { "epoch": 0.004451979332023222, "grad_norm": 18.26863670349121, "learning_rate": 4.9500000000000004e-05, "loss": 2.2029, "step": 495 }, { "epoch": 0.0044609732296636735, "grad_norm": 42.24214553833008, "learning_rate": 4.96e-05, "loss": 2.1466, "step": 496 }, { "epoch": 0.004469967127304124, "grad_norm": 15.062081336975098, "learning_rate": 4.97e-05, "loss": 2.1941, "step": 497 }, { "epoch": 0.004478961024944575, "grad_norm": 6.565387725830078, "learning_rate": 4.9800000000000004e-05, "loss": 2.1667, "step": 498 }, { "epoch": 0.004487954922585026, "grad_norm": 11.946210861206055, "learning_rate": 4.99e-05, "loss": 2.1751, "step": 499 }, { "epoch": 0.004496948820225477, "grad_norm": 15.185494422912598, "learning_rate": 5e-05, "loss": 2.1763, "step": 500 }, { "epoch": 0.004505942717865928, "grad_norm": 10.25247573852539, "learning_rate": 4.9999998633018794e-05, "loss": 2.2659, "step": 501 }, { "epoch": 0.004514936615506379, "grad_norm": 9.423380851745605, "learning_rate": 4.9999994532075326e-05, "loss": 2.1479, "step": 502 }, { "epoch": 0.0045239305131468295, "grad_norm": 6.155454158782959, "learning_rate": 4.999998769717004e-05, "loss": 2.1804, "step": 503 }, { "epoch": 0.004532924410787281, "grad_norm": 7.965705871582031, "learning_rate": 4.999997812830369e-05, "loss": 2.2, "step": 504 }, { "epoch": 0.004541918308427732, "grad_norm": 77.55868530273438, "learning_rate": 4.9999965825477314e-05, "loss": 2.217, "step": 505 }, { "epoch": 0.004550912206068182, "grad_norm": 19.80321502685547, "learning_rate": 4.9999950788692264e-05, "loss": 2.1766, "step": 506 }, { "epoch": 0.0045599061037086335, "grad_norm": 17.59615135192871, "learning_rate": 4.9999933017950186e-05, "loss": 2.218, "step": 507 }, { "epoch": 0.004568900001349085, "grad_norm": 797.918212890625, "learning_rate": 4.999991251325301e-05, "loss": 2.4093, "step": 508 }, { "epoch": 0.004577893898989536, "grad_norm": 880.2916259765625, "learning_rate": 4.9999889274603e-05, "loss": 2.5308, "step": 509 }, { "epoch": 0.004586887796629986, "grad_norm": 277.361083984375, "learning_rate": 4.999986330200268e-05, "loss": 2.3883, "step": 510 }, { "epoch": 0.004595881694270438, "grad_norm": 218.80526733398438, "learning_rate": 4.99998345954549e-05, "loss": 2.3332, "step": 511 }, { "epoch": 0.004604875591910889, "grad_norm": 36.54825210571289, "learning_rate": 4.999980315496279e-05, "loss": 2.1921, "step": 512 }, { "epoch": 0.004613869489551339, "grad_norm": 10.002345085144043, "learning_rate": 4.999976898052979e-05, "loss": 2.1759, "step": 513 }, { "epoch": 0.00462286338719179, "grad_norm": 5.7165913581848145, "learning_rate": 4.9999732072159654e-05, "loss": 2.2448, "step": 514 }, { "epoch": 0.004631857284832242, "grad_norm": 7.416532039642334, "learning_rate": 4.999969242985639e-05, "loss": 2.0931, "step": 515 }, { "epoch": 0.004640851182472692, "grad_norm": 21.929731369018555, "learning_rate": 4.999965005362436e-05, "loss": 2.2053, "step": 516 }, { "epoch": 0.004649845080113143, "grad_norm": 10.264759063720703, "learning_rate": 4.999960494346818e-05, "loss": 2.2169, "step": 517 }, { "epoch": 0.0046588389777535944, "grad_norm": 7.747093677520752, "learning_rate": 4.9999557099392795e-05, "loss": 2.2671, "step": 518 }, { "epoch": 0.004667832875394045, "grad_norm": 7.5308380126953125, "learning_rate": 4.999950652140343e-05, "loss": 2.2117, "step": 519 }, { "epoch": 0.004676826773034496, "grad_norm": 7.487093925476074, "learning_rate": 4.999945320950562e-05, "loss": 2.2369, "step": 520 }, { "epoch": 0.004685820670674947, "grad_norm": 4.400563716888428, "learning_rate": 4.999939716370518e-05, "loss": 2.1398, "step": 521 }, { "epoch": 0.004694814568315398, "grad_norm": 11.304590225219727, "learning_rate": 4.999933838400827e-05, "loss": 2.117, "step": 522 }, { "epoch": 0.004703808465955849, "grad_norm": 7.2849273681640625, "learning_rate": 4.99992768704213e-05, "loss": 2.3271, "step": 523 }, { "epoch": 0.0047128023635963, "grad_norm": 15.575254440307617, "learning_rate": 4.9999212622950984e-05, "loss": 2.1764, "step": 524 }, { "epoch": 0.0047217962612367504, "grad_norm": 7.2426228523254395, "learning_rate": 4.999914564160437e-05, "loss": 2.2143, "step": 525 }, { "epoch": 0.004730790158877202, "grad_norm": 8.810681343078613, "learning_rate": 4.999907592638877e-05, "loss": 2.1401, "step": 526 }, { "epoch": 0.004739784056517653, "grad_norm": 13.82747745513916, "learning_rate": 4.999900347731181e-05, "loss": 2.116, "step": 527 }, { "epoch": 0.004748777954158104, "grad_norm": 55.4143180847168, "learning_rate": 4.999892829438142e-05, "loss": 2.1417, "step": 528 }, { "epoch": 0.0047577718517985545, "grad_norm": 68.38306427001953, "learning_rate": 4.999885037760582e-05, "loss": 2.2516, "step": 529 }, { "epoch": 0.004766765749439006, "grad_norm": 20.79741668701172, "learning_rate": 4.999876972699352e-05, "loss": 2.2198, "step": 530 }, { "epoch": 0.004775759647079457, "grad_norm": 9.143597602844238, "learning_rate": 4.999868634255336e-05, "loss": 2.1585, "step": 531 }, { "epoch": 0.004784753544719907, "grad_norm": 6.151281356811523, "learning_rate": 4.999860022429443e-05, "loss": 2.1385, "step": 532 }, { "epoch": 0.0047937474423603585, "grad_norm": 12.984642028808594, "learning_rate": 4.999851137222618e-05, "loss": 2.2133, "step": 533 }, { "epoch": 0.00480274134000081, "grad_norm": 238.07080078125, "learning_rate": 4.9998419786358294e-05, "loss": 2.1878, "step": 534 }, { "epoch": 0.00481173523764126, "grad_norm": 114.31629180908203, "learning_rate": 4.999832546670082e-05, "loss": 2.2205, "step": 535 }, { "epoch": 0.004820729135281711, "grad_norm": 25.34907341003418, "learning_rate": 4.9998228413264056e-05, "loss": 2.1566, "step": 536 }, { "epoch": 0.004829723032922163, "grad_norm": 7.107104778289795, "learning_rate": 4.999812862605861e-05, "loss": 2.1415, "step": 537 }, { "epoch": 0.004838716930562613, "grad_norm": 7.654989242553711, "learning_rate": 4.9998026105095405e-05, "loss": 2.1634, "step": 538 }, { "epoch": 0.004847710828203064, "grad_norm": 7.804510116577148, "learning_rate": 4.999792085038565e-05, "loss": 2.0683, "step": 539 }, { "epoch": 0.004856704725843515, "grad_norm": 8.580718040466309, "learning_rate": 4.999781286194085e-05, "loss": 2.174, "step": 540 }, { "epoch": 0.004865698623483966, "grad_norm": 445.3515930175781, "learning_rate": 4.9997702139772826e-05, "loss": 2.3143, "step": 541 }, { "epoch": 0.004874692521124417, "grad_norm": 479.4335021972656, "learning_rate": 4.9997588683893674e-05, "loss": 2.3901, "step": 542 }, { "epoch": 0.004883686418764868, "grad_norm": 118.88475036621094, "learning_rate": 4.999747249431581e-05, "loss": 2.2949, "step": 543 }, { "epoch": 0.0048926803164053186, "grad_norm": 13.229616165161133, "learning_rate": 4.9997353571051935e-05, "loss": 2.1603, "step": 544 }, { "epoch": 0.00490167421404577, "grad_norm": 249.1796875, "learning_rate": 4.9997231914115064e-05, "loss": 2.3067, "step": 545 }, { "epoch": 0.004910668111686221, "grad_norm": 328.8115539550781, "learning_rate": 4.999710752351848e-05, "loss": 2.3118, "step": 546 }, { "epoch": 0.004919662009326672, "grad_norm": 119.52521514892578, "learning_rate": 4.999698039927581e-05, "loss": 2.1509, "step": 547 }, { "epoch": 0.004928655906967123, "grad_norm": 15.113656044006348, "learning_rate": 4.999685054140094e-05, "loss": 2.1913, "step": 548 }, { "epoch": 0.004937649804607574, "grad_norm": 9.866039276123047, "learning_rate": 4.9996717949908075e-05, "loss": 2.1542, "step": 549 }, { "epoch": 0.004946643702248025, "grad_norm": 178.95773315429688, "learning_rate": 4.9996582624811725e-05, "loss": 2.1589, "step": 550 }, { "epoch": 0.004955637599888475, "grad_norm": 42.39971160888672, "learning_rate": 4.9996444566126676e-05, "loss": 2.1067, "step": 551 }, { "epoch": 0.004964631497528927, "grad_norm": 13.243074417114258, "learning_rate": 4.999630377386803e-05, "loss": 2.0828, "step": 552 }, { "epoch": 0.004973625395169378, "grad_norm": 6.717075824737549, "learning_rate": 4.999616024805119e-05, "loss": 2.0939, "step": 553 }, { "epoch": 0.004982619292809828, "grad_norm": 5.1018829345703125, "learning_rate": 4.9996013988691845e-05, "loss": 2.2054, "step": 554 }, { "epoch": 0.0049916131904502795, "grad_norm": 10.904633522033691, "learning_rate": 4.999586499580599e-05, "loss": 2.1993, "step": 555 }, { "epoch": 0.005000607088090731, "grad_norm": 21.841718673706055, "learning_rate": 4.9995713269409924e-05, "loss": 2.154, "step": 556 }, { "epoch": 0.005009600985731181, "grad_norm": 17.561845779418945, "learning_rate": 4.999555880952023e-05, "loss": 2.1029, "step": 557 }, { "epoch": 0.005018594883371632, "grad_norm": 18.790084838867188, "learning_rate": 4.9995401616153804e-05, "loss": 2.1829, "step": 558 }, { "epoch": 0.0050275887810120835, "grad_norm": 12.541153907775879, "learning_rate": 4.999524168932784e-05, "loss": 2.1549, "step": 559 }, { "epoch": 0.005036582678652534, "grad_norm": 6.247627258300781, "learning_rate": 4.9995079029059824e-05, "loss": 2.116, "step": 560 }, { "epoch": 0.005045576576292985, "grad_norm": 12.468499183654785, "learning_rate": 4.9994913635367546e-05, "loss": 2.017, "step": 561 }, { "epoch": 0.005054570473933436, "grad_norm": 65.92499542236328, "learning_rate": 4.999474550826909e-05, "loss": 2.1548, "step": 562 }, { "epoch": 0.0050635643715738875, "grad_norm": 36.113983154296875, "learning_rate": 4.9994574647782846e-05, "loss": 2.264, "step": 563 }, { "epoch": 0.005072558269214338, "grad_norm": 20.071619033813477, "learning_rate": 4.999440105392749e-05, "loss": 2.1318, "step": 564 }, { "epoch": 0.005081552166854789, "grad_norm": 118.52499389648438, "learning_rate": 4.999422472672202e-05, "loss": 2.2203, "step": 565 }, { "epoch": 0.00509054606449524, "grad_norm": 15.095670700073242, "learning_rate": 4.999404566618571e-05, "loss": 2.1538, "step": 566 }, { "epoch": 0.005099539962135691, "grad_norm": 154.34019470214844, "learning_rate": 4.999386387233815e-05, "loss": 2.1982, "step": 567 }, { "epoch": 0.005108533859776142, "grad_norm": 17.23543357849121, "learning_rate": 4.999367934519921e-05, "loss": 2.1101, "step": 568 }, { "epoch": 0.005117527757416593, "grad_norm": 119.47762298583984, "learning_rate": 4.9993492084789075e-05, "loss": 2.1536, "step": 569 }, { "epoch": 0.0051265216550570435, "grad_norm": 11.38927173614502, "learning_rate": 4.999330209112822e-05, "loss": 2.2522, "step": 570 }, { "epoch": 0.005135515552697495, "grad_norm": 5.554697036743164, "learning_rate": 4.999310936423743e-05, "loss": 2.1634, "step": 571 }, { "epoch": 0.005144509450337946, "grad_norm": 20.293338775634766, "learning_rate": 4.999291390413777e-05, "loss": 2.1109, "step": 572 }, { "epoch": 0.005153503347978396, "grad_norm": 7.7019429206848145, "learning_rate": 4.999271571085063e-05, "loss": 2.1422, "step": 573 }, { "epoch": 0.005162497245618848, "grad_norm": 20.946334838867188, "learning_rate": 4.999251478439767e-05, "loss": 2.1401, "step": 574 }, { "epoch": 0.005171491143259299, "grad_norm": 43.675540924072266, "learning_rate": 4.9992311124800875e-05, "loss": 2.267, "step": 575 }, { "epoch": 0.005180485040899749, "grad_norm": 45.211280822753906, "learning_rate": 4.99921047320825e-05, "loss": 2.1524, "step": 576 }, { "epoch": 0.0051894789385402, "grad_norm": 11.21053409576416, "learning_rate": 4.999189560626514e-05, "loss": 2.0942, "step": 577 }, { "epoch": 0.005198472836180652, "grad_norm": 5.773712158203125, "learning_rate": 4.9991683747371645e-05, "loss": 2.2322, "step": 578 }, { "epoch": 0.005207466733821102, "grad_norm": 68.34687042236328, "learning_rate": 4.9991469155425185e-05, "loss": 2.1402, "step": 579 }, { "epoch": 0.005216460631461553, "grad_norm": 22.906251907348633, "learning_rate": 4.999125183044924e-05, "loss": 2.1033, "step": 580 }, { "epoch": 0.005225454529102004, "grad_norm": 4.81669807434082, "learning_rate": 4.9991031772467554e-05, "loss": 2.1413, "step": 581 }, { "epoch": 0.005234448426742456, "grad_norm": 5.724637985229492, "learning_rate": 4.999080898150422e-05, "loss": 2.1057, "step": 582 }, { "epoch": 0.005243442324382906, "grad_norm": 4.394071102142334, "learning_rate": 4.999058345758359e-05, "loss": 2.1217, "step": 583 }, { "epoch": 0.005252436222023357, "grad_norm": 4.642666339874268, "learning_rate": 4.999035520073032e-05, "loss": 2.1015, "step": 584 }, { "epoch": 0.0052614301196638085, "grad_norm": 93.35617065429688, "learning_rate": 4.999012421096938e-05, "loss": 2.0492, "step": 585 }, { "epoch": 0.005270424017304259, "grad_norm": 121.74952697753906, "learning_rate": 4.998989048832603e-05, "loss": 2.1316, "step": 586 }, { "epoch": 0.00527941791494471, "grad_norm": 8.862252235412598, "learning_rate": 4.998965403282583e-05, "loss": 2.119, "step": 587 }, { "epoch": 0.005288411812585161, "grad_norm": 6.607534885406494, "learning_rate": 4.998941484449463e-05, "loss": 2.131, "step": 588 }, { "epoch": 0.005297405710225612, "grad_norm": 3.972139835357666, "learning_rate": 4.9989172923358594e-05, "loss": 2.0963, "step": 589 }, { "epoch": 0.005306399607866063, "grad_norm": 6.987493991851807, "learning_rate": 4.998892826944418e-05, "loss": 2.1522, "step": 590 }, { "epoch": 0.005315393505506514, "grad_norm": 13.099661827087402, "learning_rate": 4.998868088277814e-05, "loss": 2.1645, "step": 591 }, { "epoch": 0.0053243874031469645, "grad_norm": 6.874795913696289, "learning_rate": 4.998843076338753e-05, "loss": 2.0854, "step": 592 }, { "epoch": 0.005333381300787416, "grad_norm": 7.047823905944824, "learning_rate": 4.99881779112997e-05, "loss": 2.1698, "step": 593 }, { "epoch": 0.005342375198427867, "grad_norm": 6.363389015197754, "learning_rate": 4.9987922326542304e-05, "loss": 2.1075, "step": 594 }, { "epoch": 0.005351369096068317, "grad_norm": 5.937300205230713, "learning_rate": 4.998766400914329e-05, "loss": 2.1136, "step": 595 }, { "epoch": 0.0053603629937087685, "grad_norm": 5.557101726531982, "learning_rate": 4.9987402959130914e-05, "loss": 2.1622, "step": 596 }, { "epoch": 0.00536935689134922, "grad_norm": 7.75043249130249, "learning_rate": 4.998713917653371e-05, "loss": 2.1976, "step": 597 }, { "epoch": 0.00537835078898967, "grad_norm": 11.316961288452148, "learning_rate": 4.9986872661380534e-05, "loss": 2.1447, "step": 598 }, { "epoch": 0.005387344686630121, "grad_norm": 26.577707290649414, "learning_rate": 4.9986603413700536e-05, "loss": 2.1317, "step": 599 }, { "epoch": 0.0053963385842705726, "grad_norm": 4.65086555480957, "learning_rate": 4.9986331433523156e-05, "loss": 2.1428, "step": 600 }, { "epoch": 0.005405332481911024, "grad_norm": 4.814509391784668, "learning_rate": 4.998605672087813e-05, "loss": 2.1471, "step": 601 }, { "epoch": 0.005414326379551474, "grad_norm": 13.799779891967773, "learning_rate": 4.998577927579551e-05, "loss": 2.134, "step": 602 }, { "epoch": 0.005423320277191925, "grad_norm": 3.7790088653564453, "learning_rate": 4.9985499098305636e-05, "loss": 2.0878, "step": 603 }, { "epoch": 0.005432314174832377, "grad_norm": 175.18972778320312, "learning_rate": 4.998521618843914e-05, "loss": 2.112, "step": 604 }, { "epoch": 0.005441308072472827, "grad_norm": 52.25758361816406, "learning_rate": 4.9984930546226975e-05, "loss": 2.1307, "step": 605 }, { "epoch": 0.005450301970113278, "grad_norm": 12.824421882629395, "learning_rate": 4.9984642171700366e-05, "loss": 2.2027, "step": 606 }, { "epoch": 0.005459295867753729, "grad_norm": 46.18326187133789, "learning_rate": 4.998435106489086e-05, "loss": 2.1938, "step": 607 }, { "epoch": 0.00546828976539418, "grad_norm": 108.18930053710938, "learning_rate": 4.998405722583027e-05, "loss": 2.1653, "step": 608 }, { "epoch": 0.005477283663034631, "grad_norm": 7.348074913024902, "learning_rate": 4.9983760654550746e-05, "loss": 2.2004, "step": 609 }, { "epoch": 0.005486277560675082, "grad_norm": 5.328648090362549, "learning_rate": 4.9983461351084735e-05, "loss": 2.0972, "step": 610 }, { "epoch": 0.005495271458315533, "grad_norm": 5.93976354598999, "learning_rate": 4.998315931546495e-05, "loss": 2.0296, "step": 611 }, { "epoch": 0.005504265355955984, "grad_norm": 13.335836410522461, "learning_rate": 4.998285454772441e-05, "loss": 2.1145, "step": 612 }, { "epoch": 0.005513259253596435, "grad_norm": 6.206167697906494, "learning_rate": 4.998254704789647e-05, "loss": 2.1308, "step": 613 }, { "epoch": 0.005522253151236885, "grad_norm": 4.453311443328857, "learning_rate": 4.998223681601473e-05, "loss": 2.1922, "step": 614 }, { "epoch": 0.005531247048877337, "grad_norm": 4.755031108856201, "learning_rate": 4.9981923852113145e-05, "loss": 2.1864, "step": 615 }, { "epoch": 0.005540240946517788, "grad_norm": 9.30239486694336, "learning_rate": 4.998160815622592e-05, "loss": 2.0695, "step": 616 }, { "epoch": 0.005549234844158238, "grad_norm": 6.30414342880249, "learning_rate": 4.99812897283876e-05, "loss": 2.1444, "step": 617 }, { "epoch": 0.0055582287417986894, "grad_norm": 5.58750581741333, "learning_rate": 4.998096856863298e-05, "loss": 2.0972, "step": 618 }, { "epoch": 0.005567222639439141, "grad_norm": 5.299816608428955, "learning_rate": 4.99806446769972e-05, "loss": 2.1112, "step": 619 }, { "epoch": 0.005576216537079592, "grad_norm": 123.18974304199219, "learning_rate": 4.9980318053515686e-05, "loss": 2.0705, "step": 620 }, { "epoch": 0.005585210434720042, "grad_norm": 122.42405700683594, "learning_rate": 4.997998869822413e-05, "loss": 2.1893, "step": 621 }, { "epoch": 0.0055942043323604935, "grad_norm": 32.70368957519531, "learning_rate": 4.9979656611158576e-05, "loss": 2.2274, "step": 622 }, { "epoch": 0.005603198230000945, "grad_norm": 86.51557922363281, "learning_rate": 4.997932179235533e-05, "loss": 2.2865, "step": 623 }, { "epoch": 0.005612192127641395, "grad_norm": 18.89086151123047, "learning_rate": 4.9978984241851013e-05, "loss": 2.2502, "step": 624 }, { "epoch": 0.005621186025281846, "grad_norm": 22.76131248474121, "learning_rate": 4.997864395968252e-05, "loss": 2.1952, "step": 625 }, { "epoch": 0.0056301799229222975, "grad_norm": 48.453453063964844, "learning_rate": 4.9978300945887095e-05, "loss": 2.237, "step": 626 }, { "epoch": 0.005639173820562748, "grad_norm": 150.64227294921875, "learning_rate": 4.997795520050223e-05, "loss": 2.155, "step": 627 }, { "epoch": 0.005648167718203199, "grad_norm": 16.368309020996094, "learning_rate": 4.9977606723565736e-05, "loss": 2.1962, "step": 628 }, { "epoch": 0.00565716161584365, "grad_norm": 26.520477294921875, "learning_rate": 4.997725551511572e-05, "loss": 2.2349, "step": 629 }, { "epoch": 0.005666155513484101, "grad_norm": 37.48980712890625, "learning_rate": 4.997690157519059e-05, "loss": 2.1936, "step": 630 }, { "epoch": 0.005675149411124552, "grad_norm": 105.69886016845703, "learning_rate": 4.997654490382907e-05, "loss": 2.1333, "step": 631 }, { "epoch": 0.005684143308765003, "grad_norm": 27.135730743408203, "learning_rate": 4.9976185501070136e-05, "loss": 2.1193, "step": 632 }, { "epoch": 0.0056931372064054535, "grad_norm": 11.177781105041504, "learning_rate": 4.9975823366953124e-05, "loss": 2.0723, "step": 633 }, { "epoch": 0.005702131104045905, "grad_norm": 4.328058242797852, "learning_rate": 4.9975458501517605e-05, "loss": 2.2055, "step": 634 }, { "epoch": 0.005711125001686356, "grad_norm": 5.190155029296875, "learning_rate": 4.99750909048035e-05, "loss": 2.0535, "step": 635 }, { "epoch": 0.005720118899326806, "grad_norm": 2.9819817543029785, "learning_rate": 4.9974720576850995e-05, "loss": 2.1085, "step": 636 }, { "epoch": 0.0057291127969672576, "grad_norm": 4.863327503204346, "learning_rate": 4.997434751770061e-05, "loss": 2.106, "step": 637 }, { "epoch": 0.005738106694607709, "grad_norm": 3.588758707046509, "learning_rate": 4.997397172739312e-05, "loss": 2.0718, "step": 638 }, { "epoch": 0.00574710059224816, "grad_norm": 8.698785781860352, "learning_rate": 4.997359320596964e-05, "loss": 2.0991, "step": 639 }, { "epoch": 0.00575609448988861, "grad_norm": 16.51296043395996, "learning_rate": 4.997321195347154e-05, "loss": 2.156, "step": 640 }, { "epoch": 0.005765088387529062, "grad_norm": 44.579261779785156, "learning_rate": 4.9972827969940536e-05, "loss": 2.1501, "step": 641 }, { "epoch": 0.005774082285169513, "grad_norm": 4.0765380859375, "learning_rate": 4.9972441255418606e-05, "loss": 2.2285, "step": 642 }, { "epoch": 0.005783076182809963, "grad_norm": 3.2363202571868896, "learning_rate": 4.997205180994805e-05, "loss": 2.1184, "step": 643 }, { "epoch": 0.005792070080450414, "grad_norm": 7.639459609985352, "learning_rate": 4.997165963357145e-05, "loss": 2.1707, "step": 644 }, { "epoch": 0.005801063978090866, "grad_norm": 3.167574644088745, "learning_rate": 4.99712647263317e-05, "loss": 2.054, "step": 645 }, { "epoch": 0.005810057875731316, "grad_norm": 3.169891595840454, "learning_rate": 4.9970867088271976e-05, "loss": 2.098, "step": 646 }, { "epoch": 0.005819051773371767, "grad_norm": 3.608734369277954, "learning_rate": 4.997046671943578e-05, "loss": 2.1069, "step": 647 }, { "epoch": 0.0058280456710122185, "grad_norm": 4.511877059936523, "learning_rate": 4.997006361986688e-05, "loss": 2.0552, "step": 648 }, { "epoch": 0.005837039568652669, "grad_norm": 7.824912071228027, "learning_rate": 4.996965778960937e-05, "loss": 2.1241, "step": 649 }, { "epoch": 0.00584603346629312, "grad_norm": 19.074344635009766, "learning_rate": 4.996924922870762e-05, "loss": 2.1698, "step": 650 }, { "epoch": 0.005855027363933571, "grad_norm": 10.403053283691406, "learning_rate": 4.996883793720632e-05, "loss": 2.1708, "step": 651 }, { "epoch": 0.005864021261574022, "grad_norm": 5.428363800048828, "learning_rate": 4.996842391515044e-05, "loss": 2.1136, "step": 652 }, { "epoch": 0.005873015159214473, "grad_norm": 5.171408176422119, "learning_rate": 4.996800716258526e-05, "loss": 2.0586, "step": 653 }, { "epoch": 0.005882009056854924, "grad_norm": 116.54088592529297, "learning_rate": 4.996758767955636e-05, "loss": 2.1512, "step": 654 }, { "epoch": 0.005891002954495375, "grad_norm": 85.05670166015625, "learning_rate": 4.99671654661096e-05, "loss": 2.3246, "step": 655 }, { "epoch": 0.005899996852135826, "grad_norm": 21.659442901611328, "learning_rate": 4.996674052229117e-05, "loss": 2.0494, "step": 656 }, { "epoch": 0.005908990749776277, "grad_norm": 24.06017303466797, "learning_rate": 4.996631284814754e-05, "loss": 2.0864, "step": 657 }, { "epoch": 0.005917984647416728, "grad_norm": 11.767736434936523, "learning_rate": 4.996588244372545e-05, "loss": 2.0787, "step": 658 }, { "epoch": 0.0059269785450571785, "grad_norm": 3.2828121185302734, "learning_rate": 4.996544930907201e-05, "loss": 2.1498, "step": 659 }, { "epoch": 0.00593597244269763, "grad_norm": 1.9047749042510986, "learning_rate": 4.996501344423456e-05, "loss": 2.0171, "step": 660 }, { "epoch": 0.005944966340338081, "grad_norm": 4.429534912109375, "learning_rate": 4.996457484926079e-05, "loss": 2.1264, "step": 661 }, { "epoch": 0.005953960237978531, "grad_norm": 5.172475337982178, "learning_rate": 4.9964133524198633e-05, "loss": 2.0341, "step": 662 }, { "epoch": 0.0059629541356189825, "grad_norm": 518.8088989257812, "learning_rate": 4.996368946909637e-05, "loss": 2.1847, "step": 663 }, { "epoch": 0.005971948033259434, "grad_norm": 439.8562316894531, "learning_rate": 4.996324268400256e-05, "loss": 2.2054, "step": 664 }, { "epoch": 0.005980941930899884, "grad_norm": 29.56341552734375, "learning_rate": 4.996279316896606e-05, "loss": 2.1238, "step": 665 }, { "epoch": 0.005989935828540335, "grad_norm": 4.9579949378967285, "learning_rate": 4.996234092403603e-05, "loss": 2.1642, "step": 666 }, { "epoch": 0.005998929726180787, "grad_norm": 6.363183498382568, "learning_rate": 4.9961885949261936e-05, "loss": 2.0476, "step": 667 }, { "epoch": 0.006007923623821237, "grad_norm": 2.653773069381714, "learning_rate": 4.996142824469351e-05, "loss": 2.0783, "step": 668 }, { "epoch": 0.006016917521461688, "grad_norm": 5.147334575653076, "learning_rate": 4.996096781038083e-05, "loss": 2.1091, "step": 669 }, { "epoch": 0.006025911419102139, "grad_norm": 4.627228736877441, "learning_rate": 4.996050464637423e-05, "loss": 2.1234, "step": 670 }, { "epoch": 0.00603490531674259, "grad_norm": 2.6882410049438477, "learning_rate": 4.996003875272438e-05, "loss": 2.1365, "step": 671 }, { "epoch": 0.006043899214383041, "grad_norm": 8.325450897216797, "learning_rate": 4.995957012948221e-05, "loss": 2.1131, "step": 672 }, { "epoch": 0.006052893112023492, "grad_norm": 8.386649131774902, "learning_rate": 4.995909877669898e-05, "loss": 2.127, "step": 673 }, { "epoch": 0.0060618870096639434, "grad_norm": 5.3029398918151855, "learning_rate": 4.995862469442624e-05, "loss": 1.9636, "step": 674 }, { "epoch": 0.006070880907304394, "grad_norm": 4.349021911621094, "learning_rate": 4.995814788271582e-05, "loss": 2.063, "step": 675 }, { "epoch": 0.006079874804944845, "grad_norm": 2.573106050491333, "learning_rate": 4.995766834161987e-05, "loss": 2.1037, "step": 676 }, { "epoch": 0.006088868702585296, "grad_norm": 116.49545288085938, "learning_rate": 4.9957186071190834e-05, "loss": 2.1017, "step": 677 }, { "epoch": 0.006097862600225747, "grad_norm": 90.65535736083984, "learning_rate": 4.995670107148146e-05, "loss": 2.0838, "step": 678 }, { "epoch": 0.006106856497866198, "grad_norm": 5.7368292808532715, "learning_rate": 4.995621334254477e-05, "loss": 2.079, "step": 679 }, { "epoch": 0.006115850395506649, "grad_norm": 7.18608283996582, "learning_rate": 4.9955722884434114e-05, "loss": 2.1173, "step": 680 }, { "epoch": 0.006124844293147099, "grad_norm": 5.959943771362305, "learning_rate": 4.9955229697203124e-05, "loss": 2.018, "step": 681 }, { "epoch": 0.006133838190787551, "grad_norm": 7.215688705444336, "learning_rate": 4.995473378090573e-05, "loss": 2.1453, "step": 682 }, { "epoch": 0.006142832088428002, "grad_norm": 3.5249240398406982, "learning_rate": 4.995423513559617e-05, "loss": 2.0339, "step": 683 }, { "epoch": 0.006151825986068452, "grad_norm": 20.786115646362305, "learning_rate": 4.995373376132898e-05, "loss": 2.0342, "step": 684 }, { "epoch": 0.0061608198837089035, "grad_norm": 11.22534465789795, "learning_rate": 4.995322965815898e-05, "loss": 2.1223, "step": 685 }, { "epoch": 0.006169813781349355, "grad_norm": 3.4888298511505127, "learning_rate": 4.9952722826141293e-05, "loss": 2.1477, "step": 686 }, { "epoch": 0.006178807678989805, "grad_norm": 2.956610679626465, "learning_rate": 4.995221326533136e-05, "loss": 2.0732, "step": 687 }, { "epoch": 0.006187801576630256, "grad_norm": 3.0452752113342285, "learning_rate": 4.99517009757849e-05, "loss": 2.033, "step": 688 }, { "epoch": 0.0061967954742707075, "grad_norm": 3.566005229949951, "learning_rate": 4.9951185957557936e-05, "loss": 1.9776, "step": 689 }, { "epoch": 0.006205789371911158, "grad_norm": 6.02570104598999, "learning_rate": 4.995066821070679e-05, "loss": 2.0365, "step": 690 }, { "epoch": 0.006214783269551609, "grad_norm": 5.086747646331787, "learning_rate": 4.9950147735288075e-05, "loss": 2.0209, "step": 691 }, { "epoch": 0.00622377716719206, "grad_norm": 5.322186470031738, "learning_rate": 4.994962453135873e-05, "loss": 2.0767, "step": 692 }, { "epoch": 0.0062327710648325116, "grad_norm": 2.611776113510132, "learning_rate": 4.994909859897594e-05, "loss": 2.0961, "step": 693 }, { "epoch": 0.006241764962472962, "grad_norm": 3.663301706314087, "learning_rate": 4.9948569938197253e-05, "loss": 1.8654, "step": 694 }, { "epoch": 0.006250758860113413, "grad_norm": 3.4063589572906494, "learning_rate": 4.9948038549080456e-05, "loss": 2.0096, "step": 695 }, { "epoch": 0.006259752757753864, "grad_norm": 1.9147545099258423, "learning_rate": 4.994750443168368e-05, "loss": 2.0777, "step": 696 }, { "epoch": 0.006268746655394315, "grad_norm": 2.94858717918396, "learning_rate": 4.994696758606532e-05, "loss": 2.0266, "step": 697 }, { "epoch": 0.006277740553034766, "grad_norm": 4.122600078582764, "learning_rate": 4.9946428012284095e-05, "loss": 1.9921, "step": 698 }, { "epoch": 0.006286734450675217, "grad_norm": 4.549467086791992, "learning_rate": 4.994588571039901e-05, "loss": 2.0673, "step": 699 }, { "epoch": 0.0062957283483156675, "grad_norm": 3.6713383197784424, "learning_rate": 4.994534068046937e-05, "loss": 2.041, "step": 700 }, { "epoch": 0.006304722245956119, "grad_norm": 2.467134952545166, "learning_rate": 4.994479292255477e-05, "loss": 1.9442, "step": 701 }, { "epoch": 0.00631371614359657, "grad_norm": 5.397249221801758, "learning_rate": 4.994424243671513e-05, "loss": 2.0998, "step": 702 }, { "epoch": 0.00632271004123702, "grad_norm": 4.023123741149902, "learning_rate": 4.994368922301064e-05, "loss": 2.065, "step": 703 }, { "epoch": 0.006331703938877472, "grad_norm": 4.139928817749023, "learning_rate": 4.9943133281501795e-05, "loss": 2.094, "step": 704 }, { "epoch": 0.006340697836517923, "grad_norm": 7.147576332092285, "learning_rate": 4.9942574612249394e-05, "loss": 2.0978, "step": 705 }, { "epoch": 0.006349691734158373, "grad_norm": 6.825831890106201, "learning_rate": 4.9942013215314535e-05, "loss": 2.126, "step": 706 }, { "epoch": 0.006358685631798824, "grad_norm": 8.004313468933105, "learning_rate": 4.994144909075861e-05, "loss": 2.0938, "step": 707 }, { "epoch": 0.006367679529439276, "grad_norm": 6.206782817840576, "learning_rate": 4.994088223864332e-05, "loss": 1.9687, "step": 708 }, { "epoch": 0.006376673427079726, "grad_norm": 6.015014171600342, "learning_rate": 4.994031265903063e-05, "loss": 2.0932, "step": 709 }, { "epoch": 0.006385667324720177, "grad_norm": 11.644634246826172, "learning_rate": 4.9939740351982856e-05, "loss": 2.0239, "step": 710 }, { "epoch": 0.0063946612223606284, "grad_norm": 5.23236608505249, "learning_rate": 4.9939165317562575e-05, "loss": 2.0401, "step": 711 }, { "epoch": 0.00640365512000108, "grad_norm": 4.116958141326904, "learning_rate": 4.9938587555832664e-05, "loss": 1.9878, "step": 712 }, { "epoch": 0.00641264901764153, "grad_norm": 5.054750442504883, "learning_rate": 4.993800706685632e-05, "loss": 2.0614, "step": 713 }, { "epoch": 0.006421642915281981, "grad_norm": 147.73779296875, "learning_rate": 4.9937423850697005e-05, "loss": 2.0741, "step": 714 }, { "epoch": 0.0064306368129224325, "grad_norm": 114.04325866699219, "learning_rate": 4.993683790741852e-05, "loss": 2.0862, "step": 715 }, { "epoch": 0.006439630710562883, "grad_norm": 63.89221954345703, "learning_rate": 4.9936249237084934e-05, "loss": 2.1691, "step": 716 }, { "epoch": 0.006448624608203334, "grad_norm": 63.61419677734375, "learning_rate": 4.993565783976062e-05, "loss": 2.1044, "step": 717 }, { "epoch": 0.006457618505843785, "grad_norm": 5.167171001434326, "learning_rate": 4.993506371551025e-05, "loss": 2.1067, "step": 718 }, { "epoch": 0.006466612403484236, "grad_norm": 8.803004264831543, "learning_rate": 4.993446686439881e-05, "loss": 1.9726, "step": 719 }, { "epoch": 0.006475606301124687, "grad_norm": 6.914166450500488, "learning_rate": 4.993386728649156e-05, "loss": 2.1375, "step": 720 }, { "epoch": 0.006484600198765138, "grad_norm": 6.370070934295654, "learning_rate": 4.993326498185406e-05, "loss": 2.0103, "step": 721 }, { "epoch": 0.0064935940964055885, "grad_norm": 2.200678825378418, "learning_rate": 4.993265995055221e-05, "loss": 1.8903, "step": 722 }, { "epoch": 0.00650258799404604, "grad_norm": 1.6513482332229614, "learning_rate": 4.993205219265213e-05, "loss": 1.9637, "step": 723 }, { "epoch": 0.006511581891686491, "grad_norm": 2.7804040908813477, "learning_rate": 4.993144170822032e-05, "loss": 2.0217, "step": 724 }, { "epoch": 0.006520575789326941, "grad_norm": 2.997101306915283, "learning_rate": 4.9930828497323526e-05, "loss": 2.067, "step": 725 }, { "epoch": 0.0065295696869673925, "grad_norm": 2.96132230758667, "learning_rate": 4.9930212560028814e-05, "loss": 2.0453, "step": 726 }, { "epoch": 0.006538563584607844, "grad_norm": 5.870732307434082, "learning_rate": 4.9929593896403534e-05, "loss": 2.0296, "step": 727 }, { "epoch": 0.006547557482248294, "grad_norm": 4.310946941375732, "learning_rate": 4.992897250651535e-05, "loss": 2.1226, "step": 728 }, { "epoch": 0.006556551379888745, "grad_norm": 2.759312868118286, "learning_rate": 4.992834839043221e-05, "loss": 1.9946, "step": 729 }, { "epoch": 0.0065655452775291966, "grad_norm": 7.112692832946777, "learning_rate": 4.9927721548222374e-05, "loss": 2.0332, "step": 730 }, { "epoch": 0.006574539175169648, "grad_norm": 5.675175189971924, "learning_rate": 4.992709197995438e-05, "loss": 2.006, "step": 731 }, { "epoch": 0.006583533072810098, "grad_norm": 16.59166145324707, "learning_rate": 4.992645968569709e-05, "loss": 2.0108, "step": 732 }, { "epoch": 0.006592526970450549, "grad_norm": 3.344719409942627, "learning_rate": 4.992582466551964e-05, "loss": 1.9618, "step": 733 }, { "epoch": 0.006601520868091001, "grad_norm": 30.71370506286621, "learning_rate": 4.992518691949148e-05, "loss": 2.0995, "step": 734 }, { "epoch": 0.006610514765731451, "grad_norm": 5.461675643920898, "learning_rate": 4.992454644768236e-05, "loss": 1.9751, "step": 735 }, { "epoch": 0.006619508663371902, "grad_norm": 12.785653114318848, "learning_rate": 4.992390325016232e-05, "loss": 1.8418, "step": 736 }, { "epoch": 0.006628502561012353, "grad_norm": 4.602219104766846, "learning_rate": 4.9923257327001676e-05, "loss": 2.066, "step": 737 }, { "epoch": 0.006637496458652804, "grad_norm": 3.0194973945617676, "learning_rate": 4.992260867827109e-05, "loss": 1.9127, "step": 738 }, { "epoch": 0.006646490356293255, "grad_norm": 2.6859967708587646, "learning_rate": 4.992195730404149e-05, "loss": 1.9879, "step": 739 }, { "epoch": 0.006655484253933706, "grad_norm": 5.035892486572266, "learning_rate": 4.9921303204384104e-05, "loss": 2.0513, "step": 740 }, { "epoch": 0.006664478151574157, "grad_norm": 3.455760955810547, "learning_rate": 4.9920646379370475e-05, "loss": 1.9694, "step": 741 }, { "epoch": 0.006673472049214608, "grad_norm": 6.330564498901367, "learning_rate": 4.991998682907243e-05, "loss": 1.9597, "step": 742 }, { "epoch": 0.006682465946855059, "grad_norm": 1.7267684936523438, "learning_rate": 4.991932455356208e-05, "loss": 2.0013, "step": 743 }, { "epoch": 0.006691459844495509, "grad_norm": 5.7486162185668945, "learning_rate": 4.9918659552911864e-05, "loss": 1.9907, "step": 744 }, { "epoch": 0.006700453742135961, "grad_norm": 5.849758148193359, "learning_rate": 4.991799182719451e-05, "loss": 2.0595, "step": 745 }, { "epoch": 0.006709447639776412, "grad_norm": 4.720619201660156, "learning_rate": 4.9917321376483026e-05, "loss": 1.9375, "step": 746 }, { "epoch": 0.006718441537416863, "grad_norm": 2.7330431938171387, "learning_rate": 4.991664820085074e-05, "loss": 2.0909, "step": 747 }, { "epoch": 0.0067274354350573135, "grad_norm": 3.8155128955841064, "learning_rate": 4.9915972300371284e-05, "loss": 2.0634, "step": 748 }, { "epoch": 0.006736429332697765, "grad_norm": 2.539130210876465, "learning_rate": 4.9915293675118544e-05, "loss": 1.9258, "step": 749 }, { "epoch": 0.006745423230338216, "grad_norm": 2.686185598373413, "learning_rate": 4.991461232516675e-05, "loss": 1.9462, "step": 750 }, { "epoch": 0.006754417127978666, "grad_norm": 2.5269219875335693, "learning_rate": 4.991392825059041e-05, "loss": 1.9393, "step": 751 }, { "epoch": 0.0067634110256191175, "grad_norm": 2.2180850505828857, "learning_rate": 4.991324145146433e-05, "loss": 1.9955, "step": 752 }, { "epoch": 0.006772404923259569, "grad_norm": 2.9195141792297363, "learning_rate": 4.991255192786363e-05, "loss": 1.9526, "step": 753 }, { "epoch": 0.006781398820900019, "grad_norm": 2.952523946762085, "learning_rate": 4.99118596798637e-05, "loss": 2.007, "step": 754 }, { "epoch": 0.00679039271854047, "grad_norm": 1.8546239137649536, "learning_rate": 4.991116470754025e-05, "loss": 1.9322, "step": 755 }, { "epoch": 0.0067993866161809215, "grad_norm": 1.6663299798965454, "learning_rate": 4.9910467010969286e-05, "loss": 1.9632, "step": 756 }, { "epoch": 0.006808380513821372, "grad_norm": 2.928518772125244, "learning_rate": 4.9909766590227094e-05, "loss": 2.0471, "step": 757 }, { "epoch": 0.006817374411461823, "grad_norm": 3.3369452953338623, "learning_rate": 4.9909063445390285e-05, "loss": 1.8816, "step": 758 }, { "epoch": 0.006826368309102274, "grad_norm": 3.246415615081787, "learning_rate": 4.990835757653575e-05, "loss": 2.025, "step": 759 }, { "epoch": 0.006835362206742725, "grad_norm": 5.6341729164123535, "learning_rate": 4.990764898374067e-05, "loss": 2.0216, "step": 760 }, { "epoch": 0.006844356104383176, "grad_norm": 4.853871822357178, "learning_rate": 4.9906937667082555e-05, "loss": 2.0097, "step": 761 }, { "epoch": 0.006853350002023627, "grad_norm": 8.81838607788086, "learning_rate": 4.990622362663918e-05, "loss": 1.9187, "step": 762 }, { "epoch": 0.0068623438996640775, "grad_norm": 2.1398401260375977, "learning_rate": 4.990550686248864e-05, "loss": 1.9225, "step": 763 }, { "epoch": 0.006871337797304529, "grad_norm": 1.7926567792892456, "learning_rate": 4.9904787374709305e-05, "loss": 1.976, "step": 764 }, { "epoch": 0.00688033169494498, "grad_norm": 2.1338999271392822, "learning_rate": 4.990406516337987e-05, "loss": 1.8915, "step": 765 }, { "epoch": 0.006889325592585431, "grad_norm": 2.8491244316101074, "learning_rate": 4.990334022857932e-05, "loss": 2.0018, "step": 766 }, { "epoch": 0.006898319490225882, "grad_norm": 4.005321502685547, "learning_rate": 4.990261257038691e-05, "loss": 1.9963, "step": 767 }, { "epoch": 0.006907313387866333, "grad_norm": 3.237144947052002, "learning_rate": 4.9901882188882237e-05, "loss": 2.0992, "step": 768 }, { "epoch": 0.006916307285506784, "grad_norm": 3.9469289779663086, "learning_rate": 4.990114908414517e-05, "loss": 2.027, "step": 769 }, { "epoch": 0.006925301183147234, "grad_norm": 12.4454927444458, "learning_rate": 4.9900413256255876e-05, "loss": 1.969, "step": 770 }, { "epoch": 0.006934295080787686, "grad_norm": 4.092862129211426, "learning_rate": 4.989967470529482e-05, "loss": 2.0267, "step": 771 }, { "epoch": 0.006943288978428137, "grad_norm": 8.441617012023926, "learning_rate": 4.9898933431342775e-05, "loss": 2.1058, "step": 772 }, { "epoch": 0.006952282876068587, "grad_norm": 7.716751575469971, "learning_rate": 4.989818943448081e-05, "loss": 1.9575, "step": 773 }, { "epoch": 0.006961276773709038, "grad_norm": 3.444956064224243, "learning_rate": 4.989744271479028e-05, "loss": 1.9229, "step": 774 }, { "epoch": 0.00697027067134949, "grad_norm": 4.643102169036865, "learning_rate": 4.9896693272352846e-05, "loss": 1.9605, "step": 775 }, { "epoch": 0.00697926456898994, "grad_norm": 3.9681947231292725, "learning_rate": 4.989594110725047e-05, "loss": 1.9194, "step": 776 }, { "epoch": 0.006988258466630391, "grad_norm": 4.633117198944092, "learning_rate": 4.9895186219565405e-05, "loss": 1.9232, "step": 777 }, { "epoch": 0.0069972523642708425, "grad_norm": 7.419233322143555, "learning_rate": 4.9894428609380206e-05, "loss": 2.0974, "step": 778 }, { "epoch": 0.007006246261911293, "grad_norm": 6.193356513977051, "learning_rate": 4.989366827677772e-05, "loss": 2.0181, "step": 779 }, { "epoch": 0.007015240159551744, "grad_norm": 2.3698465824127197, "learning_rate": 4.98929052218411e-05, "loss": 2.0276, "step": 780 }, { "epoch": 0.007024234057192195, "grad_norm": 6.084590435028076, "learning_rate": 4.9892139444653784e-05, "loss": 1.8046, "step": 781 }, { "epoch": 0.007033227954832646, "grad_norm": 2.0192339420318604, "learning_rate": 4.989137094529953e-05, "loss": 1.9419, "step": 782 }, { "epoch": 0.007042221852473097, "grad_norm": 3.280902862548828, "learning_rate": 4.989059972386237e-05, "loss": 2.002, "step": 783 }, { "epoch": 0.007051215750113548, "grad_norm": 3.8678269386291504, "learning_rate": 4.988982578042665e-05, "loss": 2.1094, "step": 784 }, { "epoch": 0.007060209647753999, "grad_norm": 3.7394440174102783, "learning_rate": 4.9889049115077005e-05, "loss": 1.996, "step": 785 }, { "epoch": 0.00706920354539445, "grad_norm": 9.715642929077148, "learning_rate": 4.988826972789836e-05, "loss": 1.9838, "step": 786 }, { "epoch": 0.007078197443034901, "grad_norm": 10.669175148010254, "learning_rate": 4.988748761897597e-05, "loss": 1.982, "step": 787 }, { "epoch": 0.007087191340675352, "grad_norm": 6.415482521057129, "learning_rate": 4.988670278839534e-05, "loss": 2.0402, "step": 788 }, { "epoch": 0.0070961852383158025, "grad_norm": 2.265989065170288, "learning_rate": 4.988591523624232e-05, "loss": 2.0449, "step": 789 }, { "epoch": 0.007105179135956254, "grad_norm": 9.280259132385254, "learning_rate": 4.988512496260301e-05, "loss": 2.0306, "step": 790 }, { "epoch": 0.007114173033596705, "grad_norm": 92.1363754272461, "learning_rate": 4.9884331967563856e-05, "loss": 1.9443, "step": 791 }, { "epoch": 0.007123166931237155, "grad_norm": 283.6097717285156, "learning_rate": 4.988353625121158e-05, "loss": 2.1043, "step": 792 }, { "epoch": 0.0071321608288776065, "grad_norm": 239.28402709960938, "learning_rate": 4.988273781363318e-05, "loss": 2.1914, "step": 793 }, { "epoch": 0.007141154726518058, "grad_norm": 4.90164041519165, "learning_rate": 4.988193665491599e-05, "loss": 2.0288, "step": 794 }, { "epoch": 0.007150148624158508, "grad_norm": 3.35483717918396, "learning_rate": 4.988113277514761e-05, "loss": 2.0014, "step": 795 }, { "epoch": 0.007159142521798959, "grad_norm": 5.046658039093018, "learning_rate": 4.9880326174415965e-05, "loss": 1.9064, "step": 796 }, { "epoch": 0.007168136419439411, "grad_norm": 3.212905168533325, "learning_rate": 4.987951685280925e-05, "loss": 2.0173, "step": 797 }, { "epoch": 0.007177130317079861, "grad_norm": 4.597742557525635, "learning_rate": 4.987870481041599e-05, "loss": 2.0612, "step": 798 }, { "epoch": 0.007186124214720312, "grad_norm": 2.937774419784546, "learning_rate": 4.987789004732496e-05, "loss": 1.9424, "step": 799 }, { "epoch": 0.007195118112360763, "grad_norm": 3.295008659362793, "learning_rate": 4.9877072563625285e-05, "loss": 2.0041, "step": 800 }, { "epoch": 0.007204112010001214, "grad_norm": 2.8171606063842773, "learning_rate": 4.987625235940636e-05, "loss": 2.0026, "step": 801 }, { "epoch": 0.007213105907641665, "grad_norm": 3.563558340072632, "learning_rate": 4.9875429434757874e-05, "loss": 1.9951, "step": 802 }, { "epoch": 0.007222099805282116, "grad_norm": 6.484987258911133, "learning_rate": 4.987460378976983e-05, "loss": 1.9376, "step": 803 }, { "epoch": 0.0072310937029225674, "grad_norm": 3.14011287689209, "learning_rate": 4.987377542453251e-05, "loss": 2.0926, "step": 804 }, { "epoch": 0.007240087600563018, "grad_norm": 4.046149253845215, "learning_rate": 4.9872944339136503e-05, "loss": 1.9476, "step": 805 }, { "epoch": 0.007249081498203469, "grad_norm": 1.9229099750518799, "learning_rate": 4.98721105336727e-05, "loss": 1.9644, "step": 806 }, { "epoch": 0.00725807539584392, "grad_norm": 4.156458854675293, "learning_rate": 4.9871274008232286e-05, "loss": 1.9348, "step": 807 }, { "epoch": 0.007267069293484371, "grad_norm": 2.214838981628418, "learning_rate": 4.987043476290674e-05, "loss": 1.9211, "step": 808 }, { "epoch": 0.007276063191124822, "grad_norm": 3.0451478958129883, "learning_rate": 4.986959279778785e-05, "loss": 1.934, "step": 809 }, { "epoch": 0.007285057088765273, "grad_norm": 6.181232929229736, "learning_rate": 4.986874811296767e-05, "loss": 1.921, "step": 810 }, { "epoch": 0.0072940509864057234, "grad_norm": 5.204398155212402, "learning_rate": 4.9867900708538585e-05, "loss": 2.0829, "step": 811 }, { "epoch": 0.007303044884046175, "grad_norm": 3.0582213401794434, "learning_rate": 4.9867050584593265e-05, "loss": 2.005, "step": 812 }, { "epoch": 0.007312038781686626, "grad_norm": 2.129680633544922, "learning_rate": 4.9866197741224694e-05, "loss": 1.9387, "step": 813 }, { "epoch": 0.007321032679327076, "grad_norm": 1.7975072860717773, "learning_rate": 4.986534217852611e-05, "loss": 2.1113, "step": 814 }, { "epoch": 0.0073300265769675275, "grad_norm": 2.243164539337158, "learning_rate": 4.9864483896591094e-05, "loss": 2.0154, "step": 815 }, { "epoch": 0.007339020474607979, "grad_norm": 1.6339006423950195, "learning_rate": 4.986362289551351e-05, "loss": 1.9568, "step": 816 }, { "epoch": 0.007348014372248429, "grad_norm": 3.655214548110962, "learning_rate": 4.98627591753875e-05, "loss": 1.9909, "step": 817 }, { "epoch": 0.00735700826988888, "grad_norm": 2.853417158126831, "learning_rate": 4.986189273630752e-05, "loss": 1.9266, "step": 818 }, { "epoch": 0.0073660021675293315, "grad_norm": 3.3307573795318604, "learning_rate": 4.9861023578368346e-05, "loss": 1.9084, "step": 819 }, { "epoch": 0.007374996065169782, "grad_norm": 9.324799537658691, "learning_rate": 4.9860151701665004e-05, "loss": 1.9924, "step": 820 }, { "epoch": 0.007383989962810233, "grad_norm": 52.28459930419922, "learning_rate": 4.985927710629285e-05, "loss": 1.9948, "step": 821 }, { "epoch": 0.007392983860450684, "grad_norm": 14.636528015136719, "learning_rate": 4.9858399792347525e-05, "loss": 2.0589, "step": 822 }, { "epoch": 0.0074019777580911356, "grad_norm": 5.006237983703613, "learning_rate": 4.9857519759924974e-05, "loss": 1.9374, "step": 823 }, { "epoch": 0.007410971655731586, "grad_norm": 2.8509693145751953, "learning_rate": 4.9856637009121434e-05, "loss": 1.9164, "step": 824 }, { "epoch": 0.007419965553372037, "grad_norm": 2.058683156967163, "learning_rate": 4.985575154003345e-05, "loss": 2.0147, "step": 825 }, { "epoch": 0.007428959451012488, "grad_norm": 9.587484359741211, "learning_rate": 4.985486335275783e-05, "loss": 1.9683, "step": 826 }, { "epoch": 0.007437953348652939, "grad_norm": 34.24576187133789, "learning_rate": 4.985397244739174e-05, "loss": 2.0567, "step": 827 }, { "epoch": 0.00744694724629339, "grad_norm": 5.210579872131348, "learning_rate": 4.985307882403258e-05, "loss": 1.9626, "step": 828 }, { "epoch": 0.007455941143933841, "grad_norm": 2.145352602005005, "learning_rate": 4.9852182482778084e-05, "loss": 1.9584, "step": 829 }, { "epoch": 0.0074649350415742916, "grad_norm": 2.8217432498931885, "learning_rate": 4.985128342372629e-05, "loss": 2.0106, "step": 830 }, { "epoch": 0.007473928939214743, "grad_norm": 5.935822486877441, "learning_rate": 4.985038164697549e-05, "loss": 1.9934, "step": 831 }, { "epoch": 0.007482922836855194, "grad_norm": 3.0496277809143066, "learning_rate": 4.984947715262433e-05, "loss": 2.0064, "step": 832 }, { "epoch": 0.007491916734495644, "grad_norm": 8.335238456726074, "learning_rate": 4.9848569940771696e-05, "loss": 2.1057, "step": 833 }, { "epoch": 0.007500910632136096, "grad_norm": 11.72769546508789, "learning_rate": 4.984766001151681e-05, "loss": 1.9131, "step": 834 }, { "epoch": 0.007509904529776547, "grad_norm": 12.229973793029785, "learning_rate": 4.984674736495919e-05, "loss": 1.9652, "step": 835 }, { "epoch": 0.007518898427416997, "grad_norm": 16.60565948486328, "learning_rate": 4.984583200119863e-05, "loss": 2.1035, "step": 836 }, { "epoch": 0.007527892325057448, "grad_norm": 8.79281234741211, "learning_rate": 4.984491392033524e-05, "loss": 1.8399, "step": 837 }, { "epoch": 0.0075368862226979, "grad_norm": 2.6266536712646484, "learning_rate": 4.9843993122469424e-05, "loss": 2.0328, "step": 838 }, { "epoch": 0.00754588012033835, "grad_norm": 1.687294363975525, "learning_rate": 4.984306960770186e-05, "loss": 2.0665, "step": 839 }, { "epoch": 0.007554874017978801, "grad_norm": 4.115018844604492, "learning_rate": 4.984214337613357e-05, "loss": 2.0124, "step": 840 }, { "epoch": 0.0075638679156192525, "grad_norm": 3.948295831680298, "learning_rate": 4.984121442786582e-05, "loss": 1.8559, "step": 841 }, { "epoch": 0.007572861813259704, "grad_norm": 5.47118616104126, "learning_rate": 4.984028276300021e-05, "loss": 1.8598, "step": 842 }, { "epoch": 0.007581855710900154, "grad_norm": 3.305490732192993, "learning_rate": 4.983934838163863e-05, "loss": 1.9414, "step": 843 }, { "epoch": 0.007590849608540605, "grad_norm": 1.981937050819397, "learning_rate": 4.9838411283883245e-05, "loss": 1.9403, "step": 844 }, { "epoch": 0.0075998435061810565, "grad_norm": 2.802847385406494, "learning_rate": 4.983747146983656e-05, "loss": 1.9169, "step": 845 }, { "epoch": 0.007608837403821507, "grad_norm": 1.624709963798523, "learning_rate": 4.983652893960132e-05, "loss": 1.8823, "step": 846 }, { "epoch": 0.007617831301461958, "grad_norm": 3.2908217906951904, "learning_rate": 4.983558369328063e-05, "loss": 1.9234, "step": 847 }, { "epoch": 0.007626825199102409, "grad_norm": 2.5122551918029785, "learning_rate": 4.983463573097784e-05, "loss": 2.034, "step": 848 }, { "epoch": 0.00763581909674286, "grad_norm": 6.075615406036377, "learning_rate": 4.983368505279663e-05, "loss": 1.9513, "step": 849 }, { "epoch": 0.007644812994383311, "grad_norm": 4.451389312744141, "learning_rate": 4.9832731658840956e-05, "loss": 1.9953, "step": 850 }, { "epoch": 0.007653806892023762, "grad_norm": 2.499955654144287, "learning_rate": 4.983177554921509e-05, "loss": 1.9366, "step": 851 }, { "epoch": 0.0076628007896642125, "grad_norm": 2.8937652111053467, "learning_rate": 4.983081672402358e-05, "loss": 1.9339, "step": 852 }, { "epoch": 0.007671794687304664, "grad_norm": 20.949533462524414, "learning_rate": 4.9829855183371275e-05, "loss": 1.9772, "step": 853 }, { "epoch": 0.007680788584945115, "grad_norm": 4.185976028442383, "learning_rate": 4.982889092736335e-05, "loss": 1.9877, "step": 854 }, { "epoch": 0.007689782482585565, "grad_norm": 2.3809025287628174, "learning_rate": 4.982792395610524e-05, "loss": 1.88, "step": 855 }, { "epoch": 0.0076987763802260165, "grad_norm": 44.777076721191406, "learning_rate": 4.98269542697027e-05, "loss": 2.0112, "step": 856 }, { "epoch": 0.007707770277866468, "grad_norm": 34.3090705871582, "learning_rate": 4.9825981868261756e-05, "loss": 1.9255, "step": 857 }, { "epoch": 0.007716764175506919, "grad_norm": 13.781228065490723, "learning_rate": 4.9825006751888767e-05, "loss": 1.858, "step": 858 }, { "epoch": 0.007725758073147369, "grad_norm": 1.2235125303268433, "learning_rate": 4.982402892069036e-05, "loss": 1.9782, "step": 859 }, { "epoch": 0.007734751970787821, "grad_norm": 1.4614633321762085, "learning_rate": 4.982304837477348e-05, "loss": 1.9112, "step": 860 }, { "epoch": 0.007743745868428272, "grad_norm": 3.8434290885925293, "learning_rate": 4.982206511424534e-05, "loss": 1.9214, "step": 861 }, { "epoch": 0.007752739766068722, "grad_norm": 3.2985260486602783, "learning_rate": 4.982107913921349e-05, "loss": 1.8868, "step": 862 }, { "epoch": 0.007761733663709173, "grad_norm": 1.5997260808944702, "learning_rate": 4.982009044978574e-05, "loss": 1.9178, "step": 863 }, { "epoch": 0.007770727561349625, "grad_norm": 3.1579697132110596, "learning_rate": 4.9819099046070206e-05, "loss": 2.0537, "step": 864 }, { "epoch": 0.007779721458990075, "grad_norm": 1.76636803150177, "learning_rate": 4.981810492817532e-05, "loss": 1.9462, "step": 865 }, { "epoch": 0.007788715356630526, "grad_norm": 2.138521194458008, "learning_rate": 4.98171080962098e-05, "loss": 1.9984, "step": 866 }, { "epoch": 0.007797709254270977, "grad_norm": 2.7925262451171875, "learning_rate": 4.9816108550282646e-05, "loss": 2.0486, "step": 867 }, { "epoch": 0.007806703151911428, "grad_norm": 2.0814309120178223, "learning_rate": 4.981510629050316e-05, "loss": 1.8856, "step": 868 }, { "epoch": 0.00781569704955188, "grad_norm": 2.0061452388763428, "learning_rate": 4.9814101316980977e-05, "loss": 1.9732, "step": 869 }, { "epoch": 0.00782469094719233, "grad_norm": 2.0816330909729004, "learning_rate": 4.981309362982598e-05, "loss": 1.964, "step": 870 }, { "epoch": 0.00783368484483278, "grad_norm": 2.137237548828125, "learning_rate": 4.981208322914836e-05, "loss": 1.9005, "step": 871 }, { "epoch": 0.007842678742473233, "grad_norm": 1.939923644065857, "learning_rate": 4.9811070115058624e-05, "loss": 1.8791, "step": 872 }, { "epoch": 0.007851672640113683, "grad_norm": 5.959467887878418, "learning_rate": 4.981005428766757e-05, "loss": 2.0098, "step": 873 }, { "epoch": 0.007860666537754133, "grad_norm": 3.1744813919067383, "learning_rate": 4.980903574708628e-05, "loss": 1.9179, "step": 874 }, { "epoch": 0.007869660435394586, "grad_norm": 3.2291057109832764, "learning_rate": 4.980801449342613e-05, "loss": 1.9943, "step": 875 }, { "epoch": 0.007878654333035036, "grad_norm": 4.23737907409668, "learning_rate": 4.980699052679882e-05, "loss": 1.963, "step": 876 }, { "epoch": 0.007887648230675486, "grad_norm": 5.72334098815918, "learning_rate": 4.980596384731632e-05, "loss": 1.9203, "step": 877 }, { "epoch": 0.007896642128315938, "grad_norm": 5.560755729675293, "learning_rate": 4.9804934455090915e-05, "loss": 1.9076, "step": 878 }, { "epoch": 0.007905636025956389, "grad_norm": 2.2926976680755615, "learning_rate": 4.9803902350235166e-05, "loss": 1.9764, "step": 879 }, { "epoch": 0.007914629923596839, "grad_norm": 1.8762661218643188, "learning_rate": 4.980286753286195e-05, "loss": 1.8856, "step": 880 }, { "epoch": 0.007923623821237291, "grad_norm": 2.642893075942993, "learning_rate": 4.980183000308442e-05, "loss": 1.7983, "step": 881 }, { "epoch": 0.007932617718877742, "grad_norm": 5.573609352111816, "learning_rate": 4.980078976101607e-05, "loss": 1.9755, "step": 882 }, { "epoch": 0.007941611616518192, "grad_norm": 2.063241720199585, "learning_rate": 4.979974680677063e-05, "loss": 1.9526, "step": 883 }, { "epoch": 0.007950605514158644, "grad_norm": 2.2967402935028076, "learning_rate": 4.979870114046217e-05, "loss": 2.0387, "step": 884 }, { "epoch": 0.007959599411799094, "grad_norm": 2.633618116378784, "learning_rate": 4.9797652762205025e-05, "loss": 1.9579, "step": 885 }, { "epoch": 0.007968593309439545, "grad_norm": 6.970749855041504, "learning_rate": 4.979660167211386e-05, "loss": 1.9995, "step": 886 }, { "epoch": 0.007977587207079997, "grad_norm": 3.016369342803955, "learning_rate": 4.979554787030363e-05, "loss": 1.9284, "step": 887 }, { "epoch": 0.007986581104720447, "grad_norm": 1.7596325874328613, "learning_rate": 4.9794491356889554e-05, "loss": 1.966, "step": 888 }, { "epoch": 0.007995575002360897, "grad_norm": 2.764314651489258, "learning_rate": 4.9793432131987185e-05, "loss": 1.9314, "step": 889 }, { "epoch": 0.00800456890000135, "grad_norm": 2.262730836868286, "learning_rate": 4.979237019571235e-05, "loss": 1.9719, "step": 890 }, { "epoch": 0.0080135627976418, "grad_norm": 6.872565746307373, "learning_rate": 4.979130554818119e-05, "loss": 1.9855, "step": 891 }, { "epoch": 0.00802255669528225, "grad_norm": 2.9007933139801025, "learning_rate": 4.9790238189510124e-05, "loss": 1.9539, "step": 892 }, { "epoch": 0.008031550592922702, "grad_norm": 6.742537498474121, "learning_rate": 4.978916811981589e-05, "loss": 1.908, "step": 893 }, { "epoch": 0.008040544490563153, "grad_norm": 4.260593891143799, "learning_rate": 4.978809533921549e-05, "loss": 2.0017, "step": 894 }, { "epoch": 0.008049538388203603, "grad_norm": 10.132391929626465, "learning_rate": 4.978701984782625e-05, "loss": 2.0573, "step": 895 }, { "epoch": 0.008058532285844055, "grad_norm": 6.60487174987793, "learning_rate": 4.97859416457658e-05, "loss": 1.8291, "step": 896 }, { "epoch": 0.008067526183484506, "grad_norm": 12.799659729003906, "learning_rate": 4.9784860733152026e-05, "loss": 1.9957, "step": 897 }, { "epoch": 0.008076520081124956, "grad_norm": 140.5140838623047, "learning_rate": 4.978377711010315e-05, "loss": 1.9954, "step": 898 }, { "epoch": 0.008085513978765408, "grad_norm": 188.51742553710938, "learning_rate": 4.978269077673767e-05, "loss": 2.0401, "step": 899 }, { "epoch": 0.008094507876405858, "grad_norm": 146.62411499023438, "learning_rate": 4.978160173317438e-05, "loss": 2.0863, "step": 900 }, { "epoch": 0.008103501774046309, "grad_norm": 79.43032836914062, "learning_rate": 4.9780509979532394e-05, "loss": 1.9676, "step": 901 }, { "epoch": 0.00811249567168676, "grad_norm": 46.03955841064453, "learning_rate": 4.977941551593109e-05, "loss": 1.9668, "step": 902 }, { "epoch": 0.008121489569327211, "grad_norm": 2.9399027824401855, "learning_rate": 4.977831834249016e-05, "loss": 1.9262, "step": 903 }, { "epoch": 0.008130483466967663, "grad_norm": 3.323502779006958, "learning_rate": 4.977721845932959e-05, "loss": 1.8704, "step": 904 }, { "epoch": 0.008139477364608114, "grad_norm": 2.397580862045288, "learning_rate": 4.9776115866569654e-05, "loss": 1.7668, "step": 905 }, { "epoch": 0.008148471262248564, "grad_norm": 2.2854068279266357, "learning_rate": 4.977501056433095e-05, "loss": 1.8791, "step": 906 }, { "epoch": 0.008157465159889016, "grad_norm": 3.1162211894989014, "learning_rate": 4.977390255273433e-05, "loss": 2.0284, "step": 907 }, { "epoch": 0.008166459057529466, "grad_norm": 4.6515727043151855, "learning_rate": 4.9772791831900976e-05, "loss": 2.0703, "step": 908 }, { "epoch": 0.008175452955169917, "grad_norm": 5.566706657409668, "learning_rate": 4.977167840195235e-05, "loss": 1.8707, "step": 909 }, { "epoch": 0.008184446852810369, "grad_norm": 3.232285976409912, "learning_rate": 4.9770562263010225e-05, "loss": 1.9405, "step": 910 }, { "epoch": 0.00819344075045082, "grad_norm": 2.6347382068634033, "learning_rate": 4.976944341519665e-05, "loss": 2.0066, "step": 911 }, { "epoch": 0.00820243464809127, "grad_norm": 2.120594024658203, "learning_rate": 4.9768321858633985e-05, "loss": 1.8906, "step": 912 }, { "epoch": 0.008211428545731722, "grad_norm": 1.2899686098098755, "learning_rate": 4.976719759344487e-05, "loss": 1.9043, "step": 913 }, { "epoch": 0.008220422443372172, "grad_norm": 2.818913459777832, "learning_rate": 4.976607061975227e-05, "loss": 1.8631, "step": 914 }, { "epoch": 0.008229416341012622, "grad_norm": 3.584197998046875, "learning_rate": 4.976494093767943e-05, "loss": 1.9723, "step": 915 }, { "epoch": 0.008238410238653075, "grad_norm": 3.11116099357605, "learning_rate": 4.976380854734987e-05, "loss": 1.9534, "step": 916 }, { "epoch": 0.008247404136293525, "grad_norm": 1.9037977457046509, "learning_rate": 4.9762673448887445e-05, "loss": 1.8812, "step": 917 }, { "epoch": 0.008256398033933975, "grad_norm": 1.3624595403671265, "learning_rate": 4.976153564241628e-05, "loss": 1.8846, "step": 918 }, { "epoch": 0.008265391931574427, "grad_norm": 1.8318990468978882, "learning_rate": 4.976039512806081e-05, "loss": 1.8274, "step": 919 }, { "epoch": 0.008274385829214878, "grad_norm": 2.558560848236084, "learning_rate": 4.975925190594575e-05, "loss": 1.9646, "step": 920 }, { "epoch": 0.008283379726855328, "grad_norm": 1.471227765083313, "learning_rate": 4.9758105976196135e-05, "loss": 1.8433, "step": 921 }, { "epoch": 0.00829237362449578, "grad_norm": 2.186021089553833, "learning_rate": 4.975695733893726e-05, "loss": 1.9037, "step": 922 }, { "epoch": 0.00830136752213623, "grad_norm": 2.559056043624878, "learning_rate": 4.9755805994294764e-05, "loss": 1.9224, "step": 923 }, { "epoch": 0.008310361419776681, "grad_norm": 1.8878509998321533, "learning_rate": 4.975465194239454e-05, "loss": 1.989, "step": 924 }, { "epoch": 0.008319355317417133, "grad_norm": 1.8974905014038086, "learning_rate": 4.9753495183362796e-05, "loss": 1.8559, "step": 925 }, { "epoch": 0.008328349215057583, "grad_norm": 2.5387771129608154, "learning_rate": 4.975233571732604e-05, "loss": 1.8846, "step": 926 }, { "epoch": 0.008337343112698034, "grad_norm": 3.0767509937286377, "learning_rate": 4.975117354441106e-05, "loss": 1.9571, "step": 927 }, { "epoch": 0.008346337010338486, "grad_norm": 1.9841474294662476, "learning_rate": 4.975000866474495e-05, "loss": 1.9374, "step": 928 }, { "epoch": 0.008355330907978936, "grad_norm": 555.1681518554688, "learning_rate": 4.974884107845511e-05, "loss": 3.414, "step": 929 }, { "epoch": 0.008364324805619387, "grad_norm": 2504.095703125, "learning_rate": 4.974767078566922e-05, "loss": 2.5588, "step": 930 }, { "epoch": 0.008373318703259839, "grad_norm": 3.425945281982422, "learning_rate": 4.9746497786515256e-05, "loss": 2.0105, "step": 931 }, { "epoch": 0.008382312600900289, "grad_norm": 2.57725191116333, "learning_rate": 4.9745322081121504e-05, "loss": 1.9033, "step": 932 }, { "epoch": 0.00839130649854074, "grad_norm": 3.293555498123169, "learning_rate": 4.9744143669616524e-05, "loss": 1.9284, "step": 933 }, { "epoch": 0.008400300396181191, "grad_norm": 3.24320387840271, "learning_rate": 4.974296255212921e-05, "loss": 1.862, "step": 934 }, { "epoch": 0.008409294293821642, "grad_norm": 2.502941131591797, "learning_rate": 4.9741778728788705e-05, "loss": 1.8517, "step": 935 }, { "epoch": 0.008418288191462092, "grad_norm": 1.9371757507324219, "learning_rate": 4.9740592199724475e-05, "loss": 1.9451, "step": 936 }, { "epoch": 0.008427282089102544, "grad_norm": 1.541487693786621, "learning_rate": 4.9739402965066276e-05, "loss": 1.8815, "step": 937 }, { "epoch": 0.008436275986742995, "grad_norm": 3.783724069595337, "learning_rate": 4.973821102494417e-05, "loss": 1.9794, "step": 938 }, { "epoch": 0.008445269884383445, "grad_norm": 2.684218168258667, "learning_rate": 4.97370163794885e-05, "loss": 1.8163, "step": 939 }, { "epoch": 0.008454263782023897, "grad_norm": 1.3492017984390259, "learning_rate": 4.97358190288299e-05, "loss": 1.8876, "step": 940 }, { "epoch": 0.008463257679664347, "grad_norm": 2.702315092086792, "learning_rate": 4.973461897309933e-05, "loss": 1.8994, "step": 941 }, { "epoch": 0.0084722515773048, "grad_norm": 2.170036792755127, "learning_rate": 4.973341621242802e-05, "loss": 1.9298, "step": 942 }, { "epoch": 0.00848124547494525, "grad_norm": 2.73464035987854, "learning_rate": 4.9732210746947485e-05, "loss": 1.8794, "step": 943 }, { "epoch": 0.0084902393725857, "grad_norm": 1.8489651679992676, "learning_rate": 4.973100257678958e-05, "loss": 1.8754, "step": 944 }, { "epoch": 0.008499233270226152, "grad_norm": 1.766837477684021, "learning_rate": 4.9729791702086414e-05, "loss": 1.9661, "step": 945 }, { "epoch": 0.008508227167866603, "grad_norm": 2.1943469047546387, "learning_rate": 4.9728578122970404e-05, "loss": 1.9636, "step": 946 }, { "epoch": 0.008517221065507053, "grad_norm": 1.7037485837936401, "learning_rate": 4.9727361839574274e-05, "loss": 1.9262, "step": 947 }, { "epoch": 0.008526214963147505, "grad_norm": 1.9452424049377441, "learning_rate": 4.9726142852031024e-05, "loss": 1.9138, "step": 948 }, { "epoch": 0.008535208860787956, "grad_norm": 2.81553053855896, "learning_rate": 4.972492116047397e-05, "loss": 1.9625, "step": 949 }, { "epoch": 0.008544202758428406, "grad_norm": 2.57431960105896, "learning_rate": 4.972369676503672e-05, "loss": 1.9215, "step": 950 }, { "epoch": 0.008553196656068858, "grad_norm": 2.0141494274139404, "learning_rate": 4.972246966585315e-05, "loss": 1.7365, "step": 951 }, { "epoch": 0.008562190553709308, "grad_norm": 38.783119201660156, "learning_rate": 4.972123986305747e-05, "loss": 1.9772, "step": 952 }, { "epoch": 0.008571184451349759, "grad_norm": 4.5533928871154785, "learning_rate": 4.972000735678417e-05, "loss": 1.9083, "step": 953 }, { "epoch": 0.00858017834899021, "grad_norm": 9.014213562011719, "learning_rate": 4.971877214716803e-05, "loss": 1.8316, "step": 954 }, { "epoch": 0.008589172246630661, "grad_norm": 2.1127445697784424, "learning_rate": 4.971753423434413e-05, "loss": 1.8961, "step": 955 }, { "epoch": 0.008598166144271112, "grad_norm": 2.6637797355651855, "learning_rate": 4.971629361844785e-05, "loss": 1.9092, "step": 956 }, { "epoch": 0.008607160041911564, "grad_norm": 1.427209734916687, "learning_rate": 4.9715050299614863e-05, "loss": 1.9069, "step": 957 }, { "epoch": 0.008616153939552014, "grad_norm": 2.1370537281036377, "learning_rate": 4.971380427798113e-05, "loss": 1.7786, "step": 958 }, { "epoch": 0.008625147837192464, "grad_norm": 1.9839975833892822, "learning_rate": 4.971255555368292e-05, "loss": 1.8455, "step": 959 }, { "epoch": 0.008634141734832916, "grad_norm": 2.004221200942993, "learning_rate": 4.971130412685679e-05, "loss": 1.8285, "step": 960 }, { "epoch": 0.008643135632473367, "grad_norm": 1.4012435674667358, "learning_rate": 4.97100499976396e-05, "loss": 1.9219, "step": 961 }, { "epoch": 0.008652129530113817, "grad_norm": 3.5619161128997803, "learning_rate": 4.970879316616848e-05, "loss": 1.8763, "step": 962 }, { "epoch": 0.00866112342775427, "grad_norm": 11.93418025970459, "learning_rate": 4.9707533632580897e-05, "loss": 1.9379, "step": 963 }, { "epoch": 0.00867011732539472, "grad_norm": 2.622504711151123, "learning_rate": 4.970627139701458e-05, "loss": 1.9123, "step": 964 }, { "epoch": 0.00867911122303517, "grad_norm": 362.67572021484375, "learning_rate": 4.970500645960756e-05, "loss": 2.0681, "step": 965 }, { "epoch": 0.008688105120675622, "grad_norm": 142.16639709472656, "learning_rate": 4.970373882049819e-05, "loss": 1.9116, "step": 966 }, { "epoch": 0.008697099018316072, "grad_norm": 4.302046775817871, "learning_rate": 4.970246847982508e-05, "loss": 1.8619, "step": 967 }, { "epoch": 0.008706092915956523, "grad_norm": 1.3895951509475708, "learning_rate": 4.9701195437727156e-05, "loss": 1.9627, "step": 968 }, { "epoch": 0.008715086813596975, "grad_norm": 2.0311763286590576, "learning_rate": 4.969991969434363e-05, "loss": 1.9539, "step": 969 }, { "epoch": 0.008724080711237425, "grad_norm": 2.8236193656921387, "learning_rate": 4.9698641249814036e-05, "loss": 1.9622, "step": 970 }, { "epoch": 0.008733074608877876, "grad_norm": 1.2696911096572876, "learning_rate": 4.9697360104278165e-05, "loss": 1.8872, "step": 971 }, { "epoch": 0.008742068506518328, "grad_norm": 2.578124523162842, "learning_rate": 4.969607625787612e-05, "loss": 1.7937, "step": 972 }, { "epoch": 0.008751062404158778, "grad_norm": 1.6730213165283203, "learning_rate": 4.96947897107483e-05, "loss": 1.8993, "step": 973 }, { "epoch": 0.008760056301799228, "grad_norm": 2.2402238845825195, "learning_rate": 4.9693500463035414e-05, "loss": 1.9205, "step": 974 }, { "epoch": 0.00876905019943968, "grad_norm": 2.1354176998138428, "learning_rate": 4.9692208514878444e-05, "loss": 1.8234, "step": 975 }, { "epoch": 0.00877804409708013, "grad_norm": 438.6123046875, "learning_rate": 4.9690913866418676e-05, "loss": 2.1477, "step": 976 }, { "epoch": 0.008787037994720581, "grad_norm": 388.2720947265625, "learning_rate": 4.968961651779769e-05, "loss": 2.3703, "step": 977 }, { "epoch": 0.008796031892361033, "grad_norm": 346.9206848144531, "learning_rate": 4.9688316469157354e-05, "loss": 2.266, "step": 978 }, { "epoch": 0.008805025790001484, "grad_norm": 63.13435363769531, "learning_rate": 4.968701372063985e-05, "loss": 1.9404, "step": 979 }, { "epoch": 0.008814019687641936, "grad_norm": 3.5552473068237305, "learning_rate": 4.968570827238764e-05, "loss": 1.9421, "step": 980 }, { "epoch": 0.008823013585282386, "grad_norm": 3.3135783672332764, "learning_rate": 4.96844001245435e-05, "loss": 1.8394, "step": 981 }, { "epoch": 0.008832007482922836, "grad_norm": 60.82234573364258, "learning_rate": 4.968308927725046e-05, "loss": 1.9304, "step": 982 }, { "epoch": 0.008841001380563289, "grad_norm": 3.0435380935668945, "learning_rate": 4.9681775730651895e-05, "loss": 1.8503, "step": 983 }, { "epoch": 0.008849995278203739, "grad_norm": 1.5571080446243286, "learning_rate": 4.9680459484891445e-05, "loss": 1.8089, "step": 984 }, { "epoch": 0.00885898917584419, "grad_norm": 3.210838794708252, "learning_rate": 4.967914054011305e-05, "loss": 1.8377, "step": 985 }, { "epoch": 0.008867983073484641, "grad_norm": 2.8126115798950195, "learning_rate": 4.967781889646095e-05, "loss": 1.9654, "step": 986 }, { "epoch": 0.008876976971125092, "grad_norm": 1.3618613481521606, "learning_rate": 4.967649455407968e-05, "loss": 1.8015, "step": 987 }, { "epoch": 0.008885970868765542, "grad_norm": 1.6802117824554443, "learning_rate": 4.9675167513114065e-05, "loss": 1.8316, "step": 988 }, { "epoch": 0.008894964766405994, "grad_norm": 2.3109796047210693, "learning_rate": 4.967383777370923e-05, "loss": 1.954, "step": 989 }, { "epoch": 0.008903958664046445, "grad_norm": 2.644674777984619, "learning_rate": 4.967250533601059e-05, "loss": 2.026, "step": 990 }, { "epoch": 0.008912952561686895, "grad_norm": 2.2049221992492676, "learning_rate": 4.9671170200163864e-05, "loss": 1.829, "step": 991 }, { "epoch": 0.008921946459327347, "grad_norm": 1.7476131916046143, "learning_rate": 4.966983236631505e-05, "loss": 1.915, "step": 992 }, { "epoch": 0.008930940356967797, "grad_norm": 2.012450695037842, "learning_rate": 4.9668491834610466e-05, "loss": 1.8582, "step": 993 }, { "epoch": 0.008939934254608248, "grad_norm": 1.474837303161621, "learning_rate": 4.96671486051967e-05, "loss": 1.9117, "step": 994 }, { "epoch": 0.0089489281522487, "grad_norm": 1.8182295560836792, "learning_rate": 4.966580267822065e-05, "loss": 1.9408, "step": 995 }, { "epoch": 0.00895792204988915, "grad_norm": 2.3502721786499023, "learning_rate": 4.966445405382951e-05, "loss": 1.8961, "step": 996 }, { "epoch": 0.0089669159475296, "grad_norm": 3.871706247329712, "learning_rate": 4.9663102732170754e-05, "loss": 1.8703, "step": 997 }, { "epoch": 0.008975909845170053, "grad_norm": 1.6961944103240967, "learning_rate": 4.9661748713392165e-05, "loss": 1.8936, "step": 998 }, { "epoch": 0.008984903742810503, "grad_norm": 1.272005558013916, "learning_rate": 4.966039199764181e-05, "loss": 1.893, "step": 999 }, { "epoch": 0.008993897640450953, "grad_norm": 2.35463809967041, "learning_rate": 4.965903258506806e-05, "loss": 1.9114, "step": 1000 }, { "epoch": 0.009002891538091405, "grad_norm": 17.072322845458984, "learning_rate": 4.965767047581958e-05, "loss": 1.9652, "step": 1001 }, { "epoch": 0.009011885435731856, "grad_norm": 2.2265431880950928, "learning_rate": 4.9656305670045346e-05, "loss": 1.9169, "step": 1002 }, { "epoch": 0.009020879333372306, "grad_norm": 2.243264675140381, "learning_rate": 4.9654938167894577e-05, "loss": 1.9067, "step": 1003 }, { "epoch": 0.009029873231012758, "grad_norm": 2.680392026901245, "learning_rate": 4.9653567969516844e-05, "loss": 1.8848, "step": 1004 }, { "epoch": 0.009038867128653209, "grad_norm": 2.611449956893921, "learning_rate": 4.965219507506198e-05, "loss": 1.8738, "step": 1005 }, { "epoch": 0.009047861026293659, "grad_norm": 1.7383923530578613, "learning_rate": 4.9650819484680134e-05, "loss": 2.027, "step": 1006 }, { "epoch": 0.009056854923934111, "grad_norm": 2.9273955821990967, "learning_rate": 4.9649441198521726e-05, "loss": 1.9219, "step": 1007 }, { "epoch": 0.009065848821574561, "grad_norm": 1.106629490852356, "learning_rate": 4.964806021673749e-05, "loss": 1.8686, "step": 1008 }, { "epoch": 0.009074842719215012, "grad_norm": 1.6677672863006592, "learning_rate": 4.964667653947844e-05, "loss": 1.942, "step": 1009 }, { "epoch": 0.009083836616855464, "grad_norm": 2.193166732788086, "learning_rate": 4.9645290166895906e-05, "loss": 1.9244, "step": 1010 }, { "epoch": 0.009092830514495914, "grad_norm": 1.939603567123413, "learning_rate": 4.96439010991415e-05, "loss": 1.811, "step": 1011 }, { "epoch": 0.009101824412136365, "grad_norm": 3.3400208950042725, "learning_rate": 4.964250933636711e-05, "loss": 1.788, "step": 1012 }, { "epoch": 0.009110818309776817, "grad_norm": 1.6272523403167725, "learning_rate": 4.9641114878724956e-05, "loss": 1.8575, "step": 1013 }, { "epoch": 0.009119812207417267, "grad_norm": 3.6787490844726562, "learning_rate": 4.963971772636752e-05, "loss": 1.89, "step": 1014 }, { "epoch": 0.00912880610505772, "grad_norm": 1.9048657417297363, "learning_rate": 4.9638317879447606e-05, "loss": 1.9183, "step": 1015 }, { "epoch": 0.00913780000269817, "grad_norm": 1.56377112865448, "learning_rate": 4.963691533811829e-05, "loss": 1.8519, "step": 1016 }, { "epoch": 0.00914679390033862, "grad_norm": 1.8097623586654663, "learning_rate": 4.963551010253294e-05, "loss": 1.8817, "step": 1017 }, { "epoch": 0.009155787797979072, "grad_norm": 1.9876528978347778, "learning_rate": 4.963410217284526e-05, "loss": 1.891, "step": 1018 }, { "epoch": 0.009164781695619522, "grad_norm": 1.4148049354553223, "learning_rate": 4.9632691549209196e-05, "loss": 1.8644, "step": 1019 }, { "epoch": 0.009173775593259973, "grad_norm": 1.417346477508545, "learning_rate": 4.963127823177902e-05, "loss": 1.9671, "step": 1020 }, { "epoch": 0.009182769490900425, "grad_norm": 1.3156847953796387, "learning_rate": 4.962986222070929e-05, "loss": 1.873, "step": 1021 }, { "epoch": 0.009191763388540875, "grad_norm": 1.8055436611175537, "learning_rate": 4.962844351615485e-05, "loss": 1.9456, "step": 1022 }, { "epoch": 0.009200757286181326, "grad_norm": 1.808810830116272, "learning_rate": 4.962702211827087e-05, "loss": 1.9281, "step": 1023 }, { "epoch": 0.009209751183821778, "grad_norm": 3.2672832012176514, "learning_rate": 4.962559802721277e-05, "loss": 1.9276, "step": 1024 }, { "epoch": 0.009218745081462228, "grad_norm": 2.229609489440918, "learning_rate": 4.96241712431363e-05, "loss": 1.9239, "step": 1025 }, { "epoch": 0.009227738979102678, "grad_norm": 2.7016751766204834, "learning_rate": 4.9622741766197477e-05, "loss": 1.8634, "step": 1026 }, { "epoch": 0.00923673287674313, "grad_norm": 1.341886281967163, "learning_rate": 4.9621309596552637e-05, "loss": 1.8144, "step": 1027 }, { "epoch": 0.00924572677438358, "grad_norm": 1.7477009296417236, "learning_rate": 4.961987473435839e-05, "loss": 1.9539, "step": 1028 }, { "epoch": 0.009254720672024031, "grad_norm": 118.58419036865234, "learning_rate": 4.961843717977167e-05, "loss": 1.9389, "step": 1029 }, { "epoch": 0.009263714569664483, "grad_norm": 78.26634216308594, "learning_rate": 4.9616996932949666e-05, "loss": 1.897, "step": 1030 }, { "epoch": 0.009272708467304934, "grad_norm": 33.4898567199707, "learning_rate": 4.961555399404989e-05, "loss": 1.9692, "step": 1031 }, { "epoch": 0.009281702364945384, "grad_norm": 2.159940242767334, "learning_rate": 4.9614108363230135e-05, "loss": 1.8506, "step": 1032 }, { "epoch": 0.009290696262585836, "grad_norm": 44.9131965637207, "learning_rate": 4.96126600406485e-05, "loss": 2.0377, "step": 1033 }, { "epoch": 0.009299690160226286, "grad_norm": 5.286812782287598, "learning_rate": 4.961120902646337e-05, "loss": 1.8353, "step": 1034 }, { "epoch": 0.009308684057866737, "grad_norm": 3.266040086746216, "learning_rate": 4.960975532083342e-05, "loss": 1.9395, "step": 1035 }, { "epoch": 0.009317677955507189, "grad_norm": 7.666036605834961, "learning_rate": 4.960829892391763e-05, "loss": 1.7963, "step": 1036 }, { "epoch": 0.00932667185314764, "grad_norm": 2.2734506130218506, "learning_rate": 4.960683983587526e-05, "loss": 1.8654, "step": 1037 }, { "epoch": 0.00933566575078809, "grad_norm": 1.723211407661438, "learning_rate": 4.960537805686589e-05, "loss": 1.8867, "step": 1038 }, { "epoch": 0.009344659648428542, "grad_norm": 1.9687684774398804, "learning_rate": 4.960391358704937e-05, "loss": 1.8587, "step": 1039 }, { "epoch": 0.009353653546068992, "grad_norm": 2.121542453765869, "learning_rate": 4.960244642658585e-05, "loss": 1.9287, "step": 1040 }, { "epoch": 0.009362647443709442, "grad_norm": 1.856347680091858, "learning_rate": 4.9600976575635774e-05, "loss": 1.9622, "step": 1041 }, { "epoch": 0.009371641341349895, "grad_norm": 1.4679237604141235, "learning_rate": 4.959950403435988e-05, "loss": 1.9125, "step": 1042 }, { "epoch": 0.009380635238990345, "grad_norm": 2.7781500816345215, "learning_rate": 4.9598028802919226e-05, "loss": 1.8641, "step": 1043 }, { "epoch": 0.009389629136630795, "grad_norm": 2.593475580215454, "learning_rate": 4.959655088147511e-05, "loss": 1.8246, "step": 1044 }, { "epoch": 0.009398623034271247, "grad_norm": 2.5981380939483643, "learning_rate": 4.959507027018918e-05, "loss": 1.8914, "step": 1045 }, { "epoch": 0.009407616931911698, "grad_norm": 2.409069061279297, "learning_rate": 4.9593586969223346e-05, "loss": 1.8899, "step": 1046 }, { "epoch": 0.009416610829552148, "grad_norm": 1.8810515403747559, "learning_rate": 4.959210097873981e-05, "loss": 1.8658, "step": 1047 }, { "epoch": 0.0094256047271926, "grad_norm": 1.8913651704788208, "learning_rate": 4.959061229890109e-05, "loss": 1.8399, "step": 1048 }, { "epoch": 0.00943459862483305, "grad_norm": 1.315248727798462, "learning_rate": 4.958912092986997e-05, "loss": 1.9085, "step": 1049 }, { "epoch": 0.009443592522473501, "grad_norm": 1.527173638343811, "learning_rate": 4.958762687180956e-05, "loss": 1.9267, "step": 1050 }, { "epoch": 0.009452586420113953, "grad_norm": 3.5598106384277344, "learning_rate": 4.958613012488324e-05, "loss": 1.9231, "step": 1051 }, { "epoch": 0.009461580317754403, "grad_norm": 4.018421173095703, "learning_rate": 4.95846306892547e-05, "loss": 1.8203, "step": 1052 }, { "epoch": 0.009470574215394855, "grad_norm": 2.0487327575683594, "learning_rate": 4.958312856508791e-05, "loss": 1.8417, "step": 1053 }, { "epoch": 0.009479568113035306, "grad_norm": 5.20163106918335, "learning_rate": 4.9581623752547136e-05, "loss": 1.8954, "step": 1054 }, { "epoch": 0.009488562010675756, "grad_norm": 6.965272426605225, "learning_rate": 4.958011625179695e-05, "loss": 1.9444, "step": 1055 }, { "epoch": 0.009497555908316208, "grad_norm": 28.557777404785156, "learning_rate": 4.9578606063002205e-05, "loss": 2.0284, "step": 1056 }, { "epoch": 0.009506549805956659, "grad_norm": 10.386679649353027, "learning_rate": 4.957709318632805e-05, "loss": 1.8683, "step": 1057 }, { "epoch": 0.009515543703597109, "grad_norm": 9.212545394897461, "learning_rate": 4.957557762193994e-05, "loss": 1.8528, "step": 1058 }, { "epoch": 0.009524537601237561, "grad_norm": 3.3937549591064453, "learning_rate": 4.9574059370003613e-05, "loss": 1.9455, "step": 1059 }, { "epoch": 0.009533531498878011, "grad_norm": 2.7785425186157227, "learning_rate": 4.9572538430685094e-05, "loss": 1.8757, "step": 1060 }, { "epoch": 0.009542525396518462, "grad_norm": 1.7468278408050537, "learning_rate": 4.957101480415072e-05, "loss": 1.8214, "step": 1061 }, { "epoch": 0.009551519294158914, "grad_norm": 3.1171300411224365, "learning_rate": 4.956948849056711e-05, "loss": 1.9355, "step": 1062 }, { "epoch": 0.009560513191799364, "grad_norm": 2.380373954772949, "learning_rate": 4.9567959490101176e-05, "loss": 1.8763, "step": 1063 }, { "epoch": 0.009569507089439815, "grad_norm": 3.901055335998535, "learning_rate": 4.956642780292012e-05, "loss": 1.907, "step": 1064 }, { "epoch": 0.009578500987080267, "grad_norm": 1.4254601001739502, "learning_rate": 4.956489342919147e-05, "loss": 1.8321, "step": 1065 }, { "epoch": 0.009587494884720717, "grad_norm": 9.619344711303711, "learning_rate": 4.9563356369083006e-05, "loss": 1.8327, "step": 1066 }, { "epoch": 0.009596488782361167, "grad_norm": 1.5964345932006836, "learning_rate": 4.9561816622762815e-05, "loss": 1.915, "step": 1067 }, { "epoch": 0.00960548268000162, "grad_norm": 2.5230660438537598, "learning_rate": 4.956027419039928e-05, "loss": 1.9517, "step": 1068 }, { "epoch": 0.00961447657764207, "grad_norm": 6.64260721206665, "learning_rate": 4.955872907216109e-05, "loss": 1.8322, "step": 1069 }, { "epoch": 0.00962347047528252, "grad_norm": 1.897109866142273, "learning_rate": 4.9557181268217227e-05, "loss": 1.8125, "step": 1070 }, { "epoch": 0.009632464372922972, "grad_norm": 1.622819185256958, "learning_rate": 4.9555630778736924e-05, "loss": 1.9067, "step": 1071 }, { "epoch": 0.009641458270563423, "grad_norm": 1.186311960220337, "learning_rate": 4.955407760388976e-05, "loss": 1.8821, "step": 1072 }, { "epoch": 0.009650452168203873, "grad_norm": 2.318347215652466, "learning_rate": 4.955252174384559e-05, "loss": 1.8115, "step": 1073 }, { "epoch": 0.009659446065844325, "grad_norm": 2.215705156326294, "learning_rate": 4.9550963198774556e-05, "loss": 1.9409, "step": 1074 }, { "epoch": 0.009668439963484775, "grad_norm": 2.3281142711639404, "learning_rate": 4.95494019688471e-05, "loss": 1.9058, "step": 1075 }, { "epoch": 0.009677433861125226, "grad_norm": 1.7275751829147339, "learning_rate": 4.954783805423394e-05, "loss": 1.8887, "step": 1076 }, { "epoch": 0.009686427758765678, "grad_norm": 1.9889212846755981, "learning_rate": 4.954627145510614e-05, "loss": 1.9061, "step": 1077 }, { "epoch": 0.009695421656406128, "grad_norm": 5.768620491027832, "learning_rate": 4.954470217163498e-05, "loss": 2.0085, "step": 1078 }, { "epoch": 0.009704415554046579, "grad_norm": 1.5947515964508057, "learning_rate": 4.95431302039921e-05, "loss": 1.9572, "step": 1079 }, { "epoch": 0.00971340945168703, "grad_norm": 1.980930209159851, "learning_rate": 4.95415555523494e-05, "loss": 1.9776, "step": 1080 }, { "epoch": 0.009722403349327481, "grad_norm": 1.6698561906814575, "learning_rate": 4.953997821687909e-05, "loss": 1.8034, "step": 1081 }, { "epoch": 0.009731397246967931, "grad_norm": 1.9077785015106201, "learning_rate": 4.9538398197753646e-05, "loss": 1.8807, "step": 1082 }, { "epoch": 0.009740391144608384, "grad_norm": 1.3882554769515991, "learning_rate": 4.953681549514587e-05, "loss": 1.8531, "step": 1083 }, { "epoch": 0.009749385042248834, "grad_norm": 1.4560352563858032, "learning_rate": 4.9535230109228844e-05, "loss": 1.8884, "step": 1084 }, { "epoch": 0.009758378939889284, "grad_norm": 2.2413580417633057, "learning_rate": 4.953364204017595e-05, "loss": 1.9112, "step": 1085 }, { "epoch": 0.009767372837529736, "grad_norm": 2.730754852294922, "learning_rate": 4.953205128816084e-05, "loss": 1.9373, "step": 1086 }, { "epoch": 0.009776366735170187, "grad_norm": 2.80940580368042, "learning_rate": 4.953045785335748e-05, "loss": 1.8697, "step": 1087 }, { "epoch": 0.009785360632810637, "grad_norm": 1.9984667301177979, "learning_rate": 4.9528861735940144e-05, "loss": 1.8869, "step": 1088 }, { "epoch": 0.00979435453045109, "grad_norm": 2.29573655128479, "learning_rate": 4.952726293608335e-05, "loss": 1.9198, "step": 1089 }, { "epoch": 0.00980334842809154, "grad_norm": 3.68994140625, "learning_rate": 4.952566145396197e-05, "loss": 1.9217, "step": 1090 }, { "epoch": 0.009812342325731992, "grad_norm": 2.9290316104888916, "learning_rate": 4.9524057289751125e-05, "loss": 1.9249, "step": 1091 }, { "epoch": 0.009821336223372442, "grad_norm": 1.826604962348938, "learning_rate": 4.9522450443626244e-05, "loss": 1.8874, "step": 1092 }, { "epoch": 0.009830330121012892, "grad_norm": 2.16923451423645, "learning_rate": 4.952084091576305e-05, "loss": 1.9346, "step": 1093 }, { "epoch": 0.009839324018653344, "grad_norm": 1.5918495655059814, "learning_rate": 4.9519228706337565e-05, "loss": 1.8861, "step": 1094 }, { "epoch": 0.009848317916293795, "grad_norm": 1.5656001567840576, "learning_rate": 4.951761381552609e-05, "loss": 1.7409, "step": 1095 }, { "epoch": 0.009857311813934245, "grad_norm": 3.3073253631591797, "learning_rate": 4.951599624350522e-05, "loss": 1.9138, "step": 1096 }, { "epoch": 0.009866305711574697, "grad_norm": 2.4301917552948, "learning_rate": 4.9514375990451874e-05, "loss": 1.8631, "step": 1097 }, { "epoch": 0.009875299609215148, "grad_norm": 1.85891592502594, "learning_rate": 4.9512753056543226e-05, "loss": 1.8255, "step": 1098 }, { "epoch": 0.009884293506855598, "grad_norm": 2.2511777877807617, "learning_rate": 4.951112744195675e-05, "loss": 1.9533, "step": 1099 }, { "epoch": 0.00989328740449605, "grad_norm": 2.875490188598633, "learning_rate": 4.9509499146870236e-05, "loss": 1.9527, "step": 1100 }, { "epoch": 0.0099022813021365, "grad_norm": 1.7893989086151123, "learning_rate": 4.950786817146174e-05, "loss": 1.8326, "step": 1101 }, { "epoch": 0.00991127519977695, "grad_norm": 1.6562172174453735, "learning_rate": 4.950623451590963e-05, "loss": 1.8855, "step": 1102 }, { "epoch": 0.009920269097417403, "grad_norm": 1.8562512397766113, "learning_rate": 4.950459818039256e-05, "loss": 1.7152, "step": 1103 }, { "epoch": 0.009929262995057853, "grad_norm": 2.159424304962158, "learning_rate": 4.950295916508947e-05, "loss": 1.9001, "step": 1104 }, { "epoch": 0.009938256892698304, "grad_norm": 4.123730659484863, "learning_rate": 4.9501317470179606e-05, "loss": 1.883, "step": 1105 }, { "epoch": 0.009947250790338756, "grad_norm": 139.787841796875, "learning_rate": 4.9499673095842506e-05, "loss": 1.8281, "step": 1106 }, { "epoch": 0.009956244687979206, "grad_norm": 1.6602857112884521, "learning_rate": 4.949802604225799e-05, "loss": 1.9463, "step": 1107 }, { "epoch": 0.009965238585619656, "grad_norm": 8.645430564880371, "learning_rate": 4.949637630960617e-05, "loss": 1.7505, "step": 1108 }, { "epoch": 0.009974232483260109, "grad_norm": 1.6627869606018066, "learning_rate": 4.949472389806748e-05, "loss": 1.8855, "step": 1109 }, { "epoch": 0.009983226380900559, "grad_norm": 1.893190860748291, "learning_rate": 4.94930688078226e-05, "loss": 1.8994, "step": 1110 }, { "epoch": 0.00999222027854101, "grad_norm": 7.066029071807861, "learning_rate": 4.949141103905255e-05, "loss": 1.799, "step": 1111 }, { "epoch": 0.010001214176181461, "grad_norm": 3.522282361984253, "learning_rate": 4.94897505919386e-05, "loss": 2.0333, "step": 1112 }, { "epoch": 0.010010208073821912, "grad_norm": 2.6594889163970947, "learning_rate": 4.948808746666235e-05, "loss": 1.923, "step": 1113 }, { "epoch": 0.010019201971462362, "grad_norm": 1.7435225248336792, "learning_rate": 4.948642166340567e-05, "loss": 1.9911, "step": 1114 }, { "epoch": 0.010028195869102814, "grad_norm": 1.7948875427246094, "learning_rate": 4.948475318235073e-05, "loss": 1.8982, "step": 1115 }, { "epoch": 0.010037189766743265, "grad_norm": 1.8984899520874023, "learning_rate": 4.9483082023680003e-05, "loss": 1.8609, "step": 1116 }, { "epoch": 0.010046183664383715, "grad_norm": 2.4756264686584473, "learning_rate": 4.948140818757623e-05, "loss": 1.8578, "step": 1117 }, { "epoch": 0.010055177562024167, "grad_norm": 1.4138103723526, "learning_rate": 4.947973167422247e-05, "loss": 1.835, "step": 1118 }, { "epoch": 0.010064171459664617, "grad_norm": 1.5873525142669678, "learning_rate": 4.947805248380205e-05, "loss": 1.8284, "step": 1119 }, { "epoch": 0.010073165357305068, "grad_norm": 1.4908703565597534, "learning_rate": 4.9476370616498617e-05, "loss": 1.915, "step": 1120 }, { "epoch": 0.01008215925494552, "grad_norm": 1.1376521587371826, "learning_rate": 4.94746860724961e-05, "loss": 1.8548, "step": 1121 }, { "epoch": 0.01009115315258597, "grad_norm": 1.775139331817627, "learning_rate": 4.9472998851978705e-05, "loss": 1.9232, "step": 1122 }, { "epoch": 0.01010014705022642, "grad_norm": 2.848726749420166, "learning_rate": 4.947130895513094e-05, "loss": 1.8136, "step": 1123 }, { "epoch": 0.010109140947866873, "grad_norm": 2.1533329486846924, "learning_rate": 4.9469616382137635e-05, "loss": 1.9274, "step": 1124 }, { "epoch": 0.010118134845507323, "grad_norm": 2.2761590480804443, "learning_rate": 4.946792113318386e-05, "loss": 1.8654, "step": 1125 }, { "epoch": 0.010127128743147775, "grad_norm": 1.0405824184417725, "learning_rate": 4.946622320845503e-05, "loss": 1.9405, "step": 1126 }, { "epoch": 0.010136122640788225, "grad_norm": 1.4591304063796997, "learning_rate": 4.9464522608136805e-05, "loss": 1.838, "step": 1127 }, { "epoch": 0.010145116538428676, "grad_norm": 1.4084917306900024, "learning_rate": 4.9462819332415175e-05, "loss": 1.8767, "step": 1128 }, { "epoch": 0.010154110436069128, "grad_norm": 2.446413040161133, "learning_rate": 4.94611133814764e-05, "loss": 1.8911, "step": 1129 }, { "epoch": 0.010163104333709578, "grad_norm": 2.2317521572113037, "learning_rate": 4.945940475550703e-05, "loss": 1.9749, "step": 1130 }, { "epoch": 0.010172098231350029, "grad_norm": 1.5114879608154297, "learning_rate": 4.945769345469395e-05, "loss": 1.9047, "step": 1131 }, { "epoch": 0.01018109212899048, "grad_norm": 1.1362504959106445, "learning_rate": 4.945597947922428e-05, "loss": 1.8324, "step": 1132 }, { "epoch": 0.010190086026630931, "grad_norm": 1.402952790260315, "learning_rate": 4.945426282928546e-05, "loss": 1.8686, "step": 1133 }, { "epoch": 0.010199079924271381, "grad_norm": 1.8045059442520142, "learning_rate": 4.945254350506522e-05, "loss": 1.977, "step": 1134 }, { "epoch": 0.010208073821911834, "grad_norm": 1.9946165084838867, "learning_rate": 4.945082150675159e-05, "loss": 1.8939, "step": 1135 }, { "epoch": 0.010217067719552284, "grad_norm": 0.9895839691162109, "learning_rate": 4.9449096834532885e-05, "loss": 1.8736, "step": 1136 }, { "epoch": 0.010226061617192734, "grad_norm": 1.0940806865692139, "learning_rate": 4.944736948859769e-05, "loss": 1.7447, "step": 1137 }, { "epoch": 0.010235055514833186, "grad_norm": 1.5091161727905273, "learning_rate": 4.9445639469134944e-05, "loss": 1.8499, "step": 1138 }, { "epoch": 0.010244049412473637, "grad_norm": 1.067003607749939, "learning_rate": 4.944390677633382e-05, "loss": 1.88, "step": 1139 }, { "epoch": 0.010253043310114087, "grad_norm": 2.034836530685425, "learning_rate": 4.944217141038379e-05, "loss": 1.8558, "step": 1140 }, { "epoch": 0.01026203720775454, "grad_norm": 2.1755740642547607, "learning_rate": 4.9440433371474646e-05, "loss": 1.9008, "step": 1141 }, { "epoch": 0.01027103110539499, "grad_norm": 1.4172083139419556, "learning_rate": 4.943869265979646e-05, "loss": 1.8946, "step": 1142 }, { "epoch": 0.01028002500303544, "grad_norm": 0.8508202433586121, "learning_rate": 4.943694927553958e-05, "loss": 1.7792, "step": 1143 }, { "epoch": 0.010289018900675892, "grad_norm": 2.8203506469726562, "learning_rate": 4.943520321889468e-05, "loss": 1.9006, "step": 1144 }, { "epoch": 0.010298012798316342, "grad_norm": 1.0900795459747314, "learning_rate": 4.9433454490052675e-05, "loss": 1.9254, "step": 1145 }, { "epoch": 0.010307006695956793, "grad_norm": 1.9202090501785278, "learning_rate": 4.943170308920484e-05, "loss": 1.85, "step": 1146 }, { "epoch": 0.010316000593597245, "grad_norm": 1.6025118827819824, "learning_rate": 4.9429949016542675e-05, "loss": 1.8642, "step": 1147 }, { "epoch": 0.010324994491237695, "grad_norm": 1.6239620447158813, "learning_rate": 4.942819227225802e-05, "loss": 1.8318, "step": 1148 }, { "epoch": 0.010333988388878146, "grad_norm": 1.830581545829773, "learning_rate": 4.942643285654299e-05, "loss": 1.8725, "step": 1149 }, { "epoch": 0.010342982286518598, "grad_norm": 3.127344846725464, "learning_rate": 4.9424670769589984e-05, "loss": 1.8795, "step": 1150 }, { "epoch": 0.010351976184159048, "grad_norm": 2.848949432373047, "learning_rate": 4.942290601159171e-05, "loss": 1.8526, "step": 1151 }, { "epoch": 0.010360970081799498, "grad_norm": 2.2335076332092285, "learning_rate": 4.942113858274114e-05, "loss": 1.8177, "step": 1152 }, { "epoch": 0.01036996397943995, "grad_norm": 1.6347110271453857, "learning_rate": 4.941936848323158e-05, "loss": 1.7713, "step": 1153 }, { "epoch": 0.0103789578770804, "grad_norm": 1.8024377822875977, "learning_rate": 4.94175957132566e-05, "loss": 2.0082, "step": 1154 }, { "epoch": 0.010387951774720851, "grad_norm": 1.3315860033035278, "learning_rate": 4.941582027301006e-05, "loss": 1.9111, "step": 1155 }, { "epoch": 0.010396945672361303, "grad_norm": 1.297783374786377, "learning_rate": 4.941404216268612e-05, "loss": 1.8932, "step": 1156 }, { "epoch": 0.010405939570001754, "grad_norm": 1.2000619173049927, "learning_rate": 4.941226138247924e-05, "loss": 1.821, "step": 1157 }, { "epoch": 0.010414933467642204, "grad_norm": 1.4980840682983398, "learning_rate": 4.941047793258416e-05, "loss": 1.9174, "step": 1158 }, { "epoch": 0.010423927365282656, "grad_norm": 1.0269590616226196, "learning_rate": 4.940869181319591e-05, "loss": 1.8612, "step": 1159 }, { "epoch": 0.010432921262923106, "grad_norm": 2.5349740982055664, "learning_rate": 4.940690302450982e-05, "loss": 1.9227, "step": 1160 }, { "epoch": 0.010441915160563557, "grad_norm": 1.889671802520752, "learning_rate": 4.940511156672151e-05, "loss": 1.9047, "step": 1161 }, { "epoch": 0.010450909058204009, "grad_norm": 1.3240622282028198, "learning_rate": 4.9403317440026896e-05, "loss": 1.8657, "step": 1162 }, { "epoch": 0.01045990295584446, "grad_norm": 1.7240233421325684, "learning_rate": 4.940152064462217e-05, "loss": 1.9573, "step": 1163 }, { "epoch": 0.010468896853484911, "grad_norm": 1.8691173791885376, "learning_rate": 4.939972118070384e-05, "loss": 1.8083, "step": 1164 }, { "epoch": 0.010477890751125362, "grad_norm": 1.5035392045974731, "learning_rate": 4.939791904846869e-05, "loss": 1.8409, "step": 1165 }, { "epoch": 0.010486884648765812, "grad_norm": 3.5965306758880615, "learning_rate": 4.939611424811379e-05, "loss": 1.8218, "step": 1166 }, { "epoch": 0.010495878546406264, "grad_norm": 1.7831751108169556, "learning_rate": 4.939430677983651e-05, "loss": 1.8577, "step": 1167 }, { "epoch": 0.010504872444046714, "grad_norm": 1.808497667312622, "learning_rate": 4.939249664383452e-05, "loss": 1.8651, "step": 1168 }, { "epoch": 0.010513866341687165, "grad_norm": 8.165689468383789, "learning_rate": 4.939068384030577e-05, "loss": 1.7653, "step": 1169 }, { "epoch": 0.010522860239327617, "grad_norm": 4.181638240814209, "learning_rate": 4.938886836944851e-05, "loss": 1.8818, "step": 1170 }, { "epoch": 0.010531854136968067, "grad_norm": 1.546927809715271, "learning_rate": 4.938705023146128e-05, "loss": 1.8517, "step": 1171 }, { "epoch": 0.010540848034608518, "grad_norm": 2.2309298515319824, "learning_rate": 4.9385229426542884e-05, "loss": 1.9373, "step": 1172 }, { "epoch": 0.01054984193224897, "grad_norm": 2.8626229763031006, "learning_rate": 4.938340595489247e-05, "loss": 1.831, "step": 1173 }, { "epoch": 0.01055883582988942, "grad_norm": 2.5954203605651855, "learning_rate": 4.9381579816709436e-05, "loss": 1.8583, "step": 1174 }, { "epoch": 0.01056782972752987, "grad_norm": 2.310791254043579, "learning_rate": 4.93797510121935e-05, "loss": 1.7469, "step": 1175 }, { "epoch": 0.010576823625170323, "grad_norm": 3.2758045196533203, "learning_rate": 4.937791954154464e-05, "loss": 1.8918, "step": 1176 }, { "epoch": 0.010585817522810773, "grad_norm": 1.8207217454910278, "learning_rate": 4.9376085404963154e-05, "loss": 1.8152, "step": 1177 }, { "epoch": 0.010594811420451223, "grad_norm": 3.2545933723449707, "learning_rate": 4.937424860264961e-05, "loss": 1.7675, "step": 1178 }, { "epoch": 0.010603805318091675, "grad_norm": 1.5232176780700684, "learning_rate": 4.9372409134804894e-05, "loss": 1.8071, "step": 1179 }, { "epoch": 0.010612799215732126, "grad_norm": 3.1837663650512695, "learning_rate": 4.937056700163015e-05, "loss": 1.8804, "step": 1180 }, { "epoch": 0.010621793113372576, "grad_norm": 2.5293006896972656, "learning_rate": 4.9368722203326844e-05, "loss": 1.8832, "step": 1181 }, { "epoch": 0.010630787011013028, "grad_norm": 1.4530854225158691, "learning_rate": 4.936687474009672e-05, "loss": 1.9887, "step": 1182 }, { "epoch": 0.010639780908653479, "grad_norm": 1.2299060821533203, "learning_rate": 4.9365024612141805e-05, "loss": 1.818, "step": 1183 }, { "epoch": 0.010648774806293929, "grad_norm": 1.4648054838180542, "learning_rate": 4.9363171819664434e-05, "loss": 1.952, "step": 1184 }, { "epoch": 0.010657768703934381, "grad_norm": 3.0476770401000977, "learning_rate": 4.9361316362867215e-05, "loss": 1.9153, "step": 1185 }, { "epoch": 0.010666762601574831, "grad_norm": 1.667232632637024, "learning_rate": 4.935945824195307e-05, "loss": 1.9462, "step": 1186 }, { "epoch": 0.010675756499215282, "grad_norm": 3.4609313011169434, "learning_rate": 4.935759745712519e-05, "loss": 1.867, "step": 1187 }, { "epoch": 0.010684750396855734, "grad_norm": 2.6009020805358887, "learning_rate": 4.935573400858708e-05, "loss": 2.0157, "step": 1188 }, { "epoch": 0.010693744294496184, "grad_norm": 3.5274269580841064, "learning_rate": 4.9353867896542516e-05, "loss": 1.9041, "step": 1189 }, { "epoch": 0.010702738192136635, "grad_norm": 3.8369295597076416, "learning_rate": 4.935199912119558e-05, "loss": 1.8982, "step": 1190 }, { "epoch": 0.010711732089777087, "grad_norm": 1.4186420440673828, "learning_rate": 4.9350127682750623e-05, "loss": 1.9843, "step": 1191 }, { "epoch": 0.010720725987417537, "grad_norm": 1.2024219036102295, "learning_rate": 4.934825358141232e-05, "loss": 1.8085, "step": 1192 }, { "epoch": 0.010729719885057987, "grad_norm": 1.3310720920562744, "learning_rate": 4.934637681738561e-05, "loss": 1.7677, "step": 1193 }, { "epoch": 0.01073871378269844, "grad_norm": 2.024397134780884, "learning_rate": 4.9344497390875734e-05, "loss": 1.8707, "step": 1194 }, { "epoch": 0.01074770768033889, "grad_norm": 2.2088379859924316, "learning_rate": 4.934261530208823e-05, "loss": 1.9011, "step": 1195 }, { "epoch": 0.01075670157797934, "grad_norm": 4.033545017242432, "learning_rate": 4.93407305512289e-05, "loss": 1.8499, "step": 1196 }, { "epoch": 0.010765695475619792, "grad_norm": 1.8474879264831543, "learning_rate": 4.933884313850388e-05, "loss": 1.84, "step": 1197 }, { "epoch": 0.010774689373260243, "grad_norm": 1.8130289316177368, "learning_rate": 4.933695306411957e-05, "loss": 1.7947, "step": 1198 }, { "epoch": 0.010783683270900695, "grad_norm": 3.1804869174957275, "learning_rate": 4.933506032828267e-05, "loss": 1.8327, "step": 1199 }, { "epoch": 0.010792677168541145, "grad_norm": 1.2182505130767822, "learning_rate": 4.933316493120015e-05, "loss": 1.915, "step": 1200 }, { "epoch": 0.010801671066181595, "grad_norm": 2.275679111480713, "learning_rate": 4.93312668730793e-05, "loss": 1.9276, "step": 1201 }, { "epoch": 0.010810664963822048, "grad_norm": 3.0447630882263184, "learning_rate": 4.932936615412769e-05, "loss": 1.8036, "step": 1202 }, { "epoch": 0.010819658861462498, "grad_norm": 1.033819556236267, "learning_rate": 4.9327462774553166e-05, "loss": 1.8736, "step": 1203 }, { "epoch": 0.010828652759102948, "grad_norm": 3.20300555229187, "learning_rate": 4.932555673456389e-05, "loss": 1.8312, "step": 1204 }, { "epoch": 0.0108376466567434, "grad_norm": 2.2383058071136475, "learning_rate": 4.9323648034368316e-05, "loss": 1.8157, "step": 1205 }, { "epoch": 0.01084664055438385, "grad_norm": 2.08383846282959, "learning_rate": 4.932173667417515e-05, "loss": 1.8536, "step": 1206 }, { "epoch": 0.010855634452024301, "grad_norm": 1.684735655784607, "learning_rate": 4.931982265419344e-05, "loss": 1.94, "step": 1207 }, { "epoch": 0.010864628349664753, "grad_norm": 1.423972249031067, "learning_rate": 4.9317905974632475e-05, "loss": 1.7875, "step": 1208 }, { "epoch": 0.010873622247305204, "grad_norm": 1.4680297374725342, "learning_rate": 4.931598663570188e-05, "loss": 1.8181, "step": 1209 }, { "epoch": 0.010882616144945654, "grad_norm": 1.8529033660888672, "learning_rate": 4.931406463761154e-05, "loss": 1.8058, "step": 1210 }, { "epoch": 0.010891610042586106, "grad_norm": 2.7427542209625244, "learning_rate": 4.931213998057166e-05, "loss": 1.9071, "step": 1211 }, { "epoch": 0.010900603940226556, "grad_norm": 2.3561723232269287, "learning_rate": 4.93102126647927e-05, "loss": 1.7741, "step": 1212 }, { "epoch": 0.010909597837867007, "grad_norm": 0.9683165550231934, "learning_rate": 4.930828269048543e-05, "loss": 1.8667, "step": 1213 }, { "epoch": 0.010918591735507459, "grad_norm": 2.084829568862915, "learning_rate": 4.930635005786091e-05, "loss": 1.774, "step": 1214 }, { "epoch": 0.01092758563314791, "grad_norm": 1.2246322631835938, "learning_rate": 4.930441476713049e-05, "loss": 1.7849, "step": 1215 }, { "epoch": 0.01093657953078836, "grad_norm": 1.989516019821167, "learning_rate": 4.930247681850582e-05, "loss": 1.8781, "step": 1216 }, { "epoch": 0.010945573428428812, "grad_norm": 1.048094391822815, "learning_rate": 4.930053621219882e-05, "loss": 1.7403, "step": 1217 }, { "epoch": 0.010954567326069262, "grad_norm": 1.3485957384109497, "learning_rate": 4.929859294842172e-05, "loss": 1.7187, "step": 1218 }, { "epoch": 0.010963561223709712, "grad_norm": 1.6276724338531494, "learning_rate": 4.9296647027387015e-05, "loss": 1.8267, "step": 1219 }, { "epoch": 0.010972555121350164, "grad_norm": 2.4596166610717773, "learning_rate": 4.929469844930753e-05, "loss": 1.8304, "step": 1220 }, { "epoch": 0.010981549018990615, "grad_norm": 1.8605852127075195, "learning_rate": 4.929274721439634e-05, "loss": 1.7972, "step": 1221 }, { "epoch": 0.010990542916631065, "grad_norm": 1.1541024446487427, "learning_rate": 4.929079332286685e-05, "loss": 1.823, "step": 1222 }, { "epoch": 0.010999536814271517, "grad_norm": 3.5439958572387695, "learning_rate": 4.9288836774932723e-05, "loss": 1.7592, "step": 1223 }, { "epoch": 0.011008530711911968, "grad_norm": 1.2599331140518188, "learning_rate": 4.9286877570807915e-05, "loss": 1.8765, "step": 1224 }, { "epoch": 0.011017524609552418, "grad_norm": 1.4835041761398315, "learning_rate": 4.9284915710706695e-05, "loss": 1.8318, "step": 1225 }, { "epoch": 0.01102651850719287, "grad_norm": 2.3044614791870117, "learning_rate": 4.928295119484361e-05, "loss": 1.8984, "step": 1226 }, { "epoch": 0.01103551240483332, "grad_norm": 1.725413203239441, "learning_rate": 4.928098402343348e-05, "loss": 1.8859, "step": 1227 }, { "epoch": 0.01104450630247377, "grad_norm": 1.758030891418457, "learning_rate": 4.927901419669145e-05, "loss": 1.7585, "step": 1228 }, { "epoch": 0.011053500200114223, "grad_norm": 1.3435816764831543, "learning_rate": 4.9277041714832925e-05, "loss": 1.9507, "step": 1229 }, { "epoch": 0.011062494097754673, "grad_norm": 1.8376082181930542, "learning_rate": 4.9275066578073626e-05, "loss": 1.8412, "step": 1230 }, { "epoch": 0.011071487995395124, "grad_norm": 1.6554052829742432, "learning_rate": 4.927308878662954e-05, "loss": 1.8734, "step": 1231 }, { "epoch": 0.011080481893035576, "grad_norm": 1.3696298599243164, "learning_rate": 4.9271108340716955e-05, "loss": 1.8671, "step": 1232 }, { "epoch": 0.011089475790676026, "grad_norm": 1.3664677143096924, "learning_rate": 4.926912524055245e-05, "loss": 1.8803, "step": 1233 }, { "epoch": 0.011098469688316476, "grad_norm": 1.9583055973052979, "learning_rate": 4.926713948635291e-05, "loss": 1.8675, "step": 1234 }, { "epoch": 0.011107463585956929, "grad_norm": 1.3232417106628418, "learning_rate": 4.926515107833547e-05, "loss": 1.8262, "step": 1235 }, { "epoch": 0.011116457483597379, "grad_norm": 2.0523736476898193, "learning_rate": 4.926316001671759e-05, "loss": 1.858, "step": 1236 }, { "epoch": 0.011125451381237831, "grad_norm": 1.9233832359313965, "learning_rate": 4.9261166301717014e-05, "loss": 1.8491, "step": 1237 }, { "epoch": 0.011134445278878281, "grad_norm": 1.6934845447540283, "learning_rate": 4.9259169933551765e-05, "loss": 1.8254, "step": 1238 }, { "epoch": 0.011143439176518732, "grad_norm": 1.158386468887329, "learning_rate": 4.9257170912440165e-05, "loss": 1.9314, "step": 1239 }, { "epoch": 0.011152433074159184, "grad_norm": 1.8263996839523315, "learning_rate": 4.925516923860083e-05, "loss": 1.8837, "step": 1240 }, { "epoch": 0.011161426971799634, "grad_norm": 1.4522002935409546, "learning_rate": 4.925316491225265e-05, "loss": 1.7722, "step": 1241 }, { "epoch": 0.011170420869440085, "grad_norm": 1.4005663394927979, "learning_rate": 4.9251157933614815e-05, "loss": 1.8982, "step": 1242 }, { "epoch": 0.011179414767080537, "grad_norm": 2.027334213256836, "learning_rate": 4.9249148302906816e-05, "loss": 1.7815, "step": 1243 }, { "epoch": 0.011188408664720987, "grad_norm": 3.36435604095459, "learning_rate": 4.924713602034842e-05, "loss": 1.8738, "step": 1244 }, { "epoch": 0.011197402562361437, "grad_norm": 1.5967189073562622, "learning_rate": 4.9245121086159674e-05, "loss": 1.8704, "step": 1245 }, { "epoch": 0.01120639646000189, "grad_norm": 1.671094536781311, "learning_rate": 4.924310350056094e-05, "loss": 1.8076, "step": 1246 }, { "epoch": 0.01121539035764234, "grad_norm": 1.7592309713363647, "learning_rate": 4.9241083263772855e-05, "loss": 1.8805, "step": 1247 }, { "epoch": 0.01122438425528279, "grad_norm": 1.3288494348526, "learning_rate": 4.9239060376016355e-05, "loss": 1.8253, "step": 1248 }, { "epoch": 0.011233378152923242, "grad_norm": 1.9537098407745361, "learning_rate": 4.923703483751265e-05, "loss": 1.9405, "step": 1249 }, { "epoch": 0.011242372050563693, "grad_norm": 1.0627882480621338, "learning_rate": 4.923500664848326e-05, "loss": 1.8506, "step": 1250 }, { "epoch": 0.011251365948204143, "grad_norm": 1.4585059881210327, "learning_rate": 4.923297580914998e-05, "loss": 1.9496, "step": 1251 }, { "epoch": 0.011260359845844595, "grad_norm": 1.8898533582687378, "learning_rate": 4.92309423197349e-05, "loss": 1.8778, "step": 1252 }, { "epoch": 0.011269353743485045, "grad_norm": 1.5207654237747192, "learning_rate": 4.922890618046039e-05, "loss": 1.7528, "step": 1253 }, { "epoch": 0.011278347641125496, "grad_norm": 1.2147852182388306, "learning_rate": 4.9226867391549136e-05, "loss": 1.7918, "step": 1254 }, { "epoch": 0.011287341538765948, "grad_norm": 1.3526878356933594, "learning_rate": 4.9224825953224084e-05, "loss": 1.8394, "step": 1255 }, { "epoch": 0.011296335436406398, "grad_norm": 1.4936753511428833, "learning_rate": 4.922278186570849e-05, "loss": 1.7148, "step": 1256 }, { "epoch": 0.011305329334046849, "grad_norm": 1.0815688371658325, "learning_rate": 4.922073512922589e-05, "loss": 1.8742, "step": 1257 }, { "epoch": 0.0113143232316873, "grad_norm": 2.284883975982666, "learning_rate": 4.921868574400011e-05, "loss": 1.846, "step": 1258 }, { "epoch": 0.011323317129327751, "grad_norm": 0.731486976146698, "learning_rate": 4.9216633710255276e-05, "loss": 1.8443, "step": 1259 }, { "epoch": 0.011332311026968201, "grad_norm": 2.83512282371521, "learning_rate": 4.9214579028215776e-05, "loss": 1.7616, "step": 1260 }, { "epoch": 0.011341304924608653, "grad_norm": 1.4199379682540894, "learning_rate": 4.921252169810633e-05, "loss": 1.7548, "step": 1261 }, { "epoch": 0.011350298822249104, "grad_norm": 1.5982848405838013, "learning_rate": 4.921046172015191e-05, "loss": 1.8004, "step": 1262 }, { "epoch": 0.011359292719889554, "grad_norm": 1.4010401964187622, "learning_rate": 4.920839909457779e-05, "loss": 1.8707, "step": 1263 }, { "epoch": 0.011368286617530006, "grad_norm": 1.732714295387268, "learning_rate": 4.920633382160955e-05, "loss": 1.8823, "step": 1264 }, { "epoch": 0.011377280515170457, "grad_norm": 2.4421274662017822, "learning_rate": 4.920426590147304e-05, "loss": 1.8485, "step": 1265 }, { "epoch": 0.011386274412810907, "grad_norm": 1.7154825925827026, "learning_rate": 4.92021953343944e-05, "loss": 1.855, "step": 1266 }, { "epoch": 0.011395268310451359, "grad_norm": 1.3120369911193848, "learning_rate": 4.920012212060005e-05, "loss": 1.8127, "step": 1267 }, { "epoch": 0.01140426220809181, "grad_norm": 1.077602744102478, "learning_rate": 4.919804626031676e-05, "loss": 1.7972, "step": 1268 }, { "epoch": 0.01141325610573226, "grad_norm": 1.4130492210388184, "learning_rate": 4.919596775377149e-05, "loss": 1.7846, "step": 1269 }, { "epoch": 0.011422250003372712, "grad_norm": 1.2219569683074951, "learning_rate": 4.919388660119156e-05, "loss": 1.7358, "step": 1270 }, { "epoch": 0.011431243901013162, "grad_norm": 1.7677727937698364, "learning_rate": 4.919180280280458e-05, "loss": 1.7333, "step": 1271 }, { "epoch": 0.011440237798653613, "grad_norm": 1.4697858095169067, "learning_rate": 4.918971635883841e-05, "loss": 1.886, "step": 1272 }, { "epoch": 0.011449231696294065, "grad_norm": 3.9753713607788086, "learning_rate": 4.918762726952123e-05, "loss": 1.8704, "step": 1273 }, { "epoch": 0.011458225593934515, "grad_norm": 1.4198626279830933, "learning_rate": 4.9185535535081495e-05, "loss": 1.8454, "step": 1274 }, { "epoch": 0.011467219491574967, "grad_norm": 1.4114727973937988, "learning_rate": 4.918344115574796e-05, "loss": 1.8935, "step": 1275 }, { "epoch": 0.011476213389215418, "grad_norm": 0.9318772554397583, "learning_rate": 4.918134413174966e-05, "loss": 1.7945, "step": 1276 }, { "epoch": 0.011485207286855868, "grad_norm": 2.335042715072632, "learning_rate": 4.917924446331592e-05, "loss": 1.8759, "step": 1277 }, { "epoch": 0.01149420118449632, "grad_norm": 2.0695204734802246, "learning_rate": 4.9177142150676355e-05, "loss": 1.8827, "step": 1278 }, { "epoch": 0.01150319508213677, "grad_norm": 1.187988519668579, "learning_rate": 4.917503719406088e-05, "loss": 1.6878, "step": 1279 }, { "epoch": 0.01151218897977722, "grad_norm": 1.275360345840454, "learning_rate": 4.917292959369968e-05, "loss": 1.98, "step": 1280 }, { "epoch": 0.011521182877417673, "grad_norm": 1.0016882419586182, "learning_rate": 4.9170819349823244e-05, "loss": 1.8712, "step": 1281 }, { "epoch": 0.011530176775058123, "grad_norm": 1.5172919034957886, "learning_rate": 4.9168706462662344e-05, "loss": 1.9505, "step": 1282 }, { "epoch": 0.011539170672698574, "grad_norm": 1.7134175300598145, "learning_rate": 4.9166590932448045e-05, "loss": 1.8041, "step": 1283 }, { "epoch": 0.011548164570339026, "grad_norm": 1.4970893859863281, "learning_rate": 4.9164472759411695e-05, "loss": 1.8797, "step": 1284 }, { "epoch": 0.011557158467979476, "grad_norm": 2.4538016319274902, "learning_rate": 4.9162351943784935e-05, "loss": 1.8818, "step": 1285 }, { "epoch": 0.011566152365619926, "grad_norm": 1.5952671766281128, "learning_rate": 4.916022848579969e-05, "loss": 1.8573, "step": 1286 }, { "epoch": 0.011575146263260378, "grad_norm": 2.1946325302124023, "learning_rate": 4.915810238568819e-05, "loss": 1.8317, "step": 1287 }, { "epoch": 0.011584140160900829, "grad_norm": 2.7667295932769775, "learning_rate": 4.915597364368292e-05, "loss": 1.8622, "step": 1288 }, { "epoch": 0.01159313405854128, "grad_norm": 1.9601869583129883, "learning_rate": 4.9153842260016704e-05, "loss": 1.9494, "step": 1289 }, { "epoch": 0.011602127956181731, "grad_norm": 1.1900784969329834, "learning_rate": 4.9151708234922603e-05, "loss": 1.9281, "step": 1290 }, { "epoch": 0.011611121853822182, "grad_norm": 3.5769097805023193, "learning_rate": 4.914957156863401e-05, "loss": 1.8698, "step": 1291 }, { "epoch": 0.011620115751462632, "grad_norm": 2.737170457839966, "learning_rate": 4.914743226138457e-05, "loss": 1.7249, "step": 1292 }, { "epoch": 0.011629109649103084, "grad_norm": 1.2054256200790405, "learning_rate": 4.914529031340825e-05, "loss": 1.7876, "step": 1293 }, { "epoch": 0.011638103546743534, "grad_norm": 3.631582498550415, "learning_rate": 4.9143145724939284e-05, "loss": 1.8746, "step": 1294 }, { "epoch": 0.011647097444383985, "grad_norm": 2.6612188816070557, "learning_rate": 4.91409984962122e-05, "loss": 1.8648, "step": 1295 }, { "epoch": 0.011656091342024437, "grad_norm": 1.0735119581222534, "learning_rate": 4.913884862746181e-05, "loss": 1.8489, "step": 1296 }, { "epoch": 0.011665085239664887, "grad_norm": 1.3620102405548096, "learning_rate": 4.913669611892323e-05, "loss": 1.853, "step": 1297 }, { "epoch": 0.011674079137305338, "grad_norm": 2.3715150356292725, "learning_rate": 4.913454097083185e-05, "loss": 1.8637, "step": 1298 }, { "epoch": 0.01168307303494579, "grad_norm": 1.4152562618255615, "learning_rate": 4.9132383183423356e-05, "loss": 1.8991, "step": 1299 }, { "epoch": 0.01169206693258624, "grad_norm": 1.9300880432128906, "learning_rate": 4.913022275693372e-05, "loss": 1.8106, "step": 1300 }, { "epoch": 0.01170106083022669, "grad_norm": 1.928333044052124, "learning_rate": 4.9128059691599206e-05, "loss": 1.9228, "step": 1301 }, { "epoch": 0.011710054727867143, "grad_norm": 3.321418285369873, "learning_rate": 4.912589398765635e-05, "loss": 1.7187, "step": 1302 }, { "epoch": 0.011719048625507593, "grad_norm": 1.422747015953064, "learning_rate": 4.912372564534201e-05, "loss": 1.7683, "step": 1303 }, { "epoch": 0.011728042523148043, "grad_norm": 1.040277361869812, "learning_rate": 4.91215546648933e-05, "loss": 1.7418, "step": 1304 }, { "epoch": 0.011737036420788495, "grad_norm": 1.1940758228302002, "learning_rate": 4.9119381046547636e-05, "loss": 1.8063, "step": 1305 }, { "epoch": 0.011746030318428946, "grad_norm": 1.2477617263793945, "learning_rate": 4.911720479054272e-05, "loss": 1.78, "step": 1306 }, { "epoch": 0.011755024216069396, "grad_norm": 1.255989909172058, "learning_rate": 4.911502589711656e-05, "loss": 1.8323, "step": 1307 }, { "epoch": 0.011764018113709848, "grad_norm": 3.495309591293335, "learning_rate": 4.911284436650742e-05, "loss": 1.8506, "step": 1308 }, { "epoch": 0.011773012011350299, "grad_norm": 1.5969996452331543, "learning_rate": 4.9110660198953865e-05, "loss": 1.8757, "step": 1309 }, { "epoch": 0.01178200590899075, "grad_norm": 1.461681842803955, "learning_rate": 4.910847339469477e-05, "loss": 1.877, "step": 1310 }, { "epoch": 0.011790999806631201, "grad_norm": 1.699761986732483, "learning_rate": 4.9106283953969266e-05, "loss": 1.6835, "step": 1311 }, { "epoch": 0.011799993704271651, "grad_norm": 1.2090150117874146, "learning_rate": 4.910409187701679e-05, "loss": 1.8494, "step": 1312 }, { "epoch": 0.011808987601912103, "grad_norm": 1.7570303678512573, "learning_rate": 4.910189716407707e-05, "loss": 1.9767, "step": 1313 }, { "epoch": 0.011817981499552554, "grad_norm": 1.637498378753662, "learning_rate": 4.909969981539011e-05, "loss": 1.7987, "step": 1314 }, { "epoch": 0.011826975397193004, "grad_norm": 2.8890786170959473, "learning_rate": 4.9097499831196216e-05, "loss": 1.8541, "step": 1315 }, { "epoch": 0.011835969294833456, "grad_norm": 1.2725188732147217, "learning_rate": 4.9095297211735966e-05, "loss": 1.8729, "step": 1316 }, { "epoch": 0.011844963192473907, "grad_norm": 1.7973047494888306, "learning_rate": 4.909309195725025e-05, "loss": 1.8872, "step": 1317 }, { "epoch": 0.011853957090114357, "grad_norm": 2.116143226623535, "learning_rate": 4.9090884067980205e-05, "loss": 1.7871, "step": 1318 }, { "epoch": 0.011862950987754809, "grad_norm": 2.6199309825897217, "learning_rate": 4.908867354416731e-05, "loss": 1.8725, "step": 1319 }, { "epoch": 0.01187194488539526, "grad_norm": 1.5852805376052856, "learning_rate": 4.908646038605329e-05, "loss": 1.8592, "step": 1320 }, { "epoch": 0.01188093878303571, "grad_norm": 3.923495292663574, "learning_rate": 4.908424459388018e-05, "loss": 1.8596, "step": 1321 }, { "epoch": 0.011889932680676162, "grad_norm": 1.129482626914978, "learning_rate": 4.9082026167890285e-05, "loss": 1.7707, "step": 1322 }, { "epoch": 0.011898926578316612, "grad_norm": 1.7788457870483398, "learning_rate": 4.9079805108326214e-05, "loss": 1.7891, "step": 1323 }, { "epoch": 0.011907920475957063, "grad_norm": 2.5163111686706543, "learning_rate": 4.907758141543086e-05, "loss": 1.8752, "step": 1324 }, { "epoch": 0.011916914373597515, "grad_norm": 2.898075580596924, "learning_rate": 4.907535508944741e-05, "loss": 1.8152, "step": 1325 }, { "epoch": 0.011925908271237965, "grad_norm": 1.8418488502502441, "learning_rate": 4.907312613061931e-05, "loss": 1.8882, "step": 1326 }, { "epoch": 0.011934902168878415, "grad_norm": 1.5218850374221802, "learning_rate": 4.9070894539190344e-05, "loss": 1.8209, "step": 1327 }, { "epoch": 0.011943896066518868, "grad_norm": 1.2311382293701172, "learning_rate": 4.906866031540453e-05, "loss": 1.8071, "step": 1328 }, { "epoch": 0.011952889964159318, "grad_norm": 1.275814175605774, "learning_rate": 4.906642345950622e-05, "loss": 1.7747, "step": 1329 }, { "epoch": 0.011961883861799768, "grad_norm": 0.9636077284812927, "learning_rate": 4.906418397174002e-05, "loss": 1.8118, "step": 1330 }, { "epoch": 0.01197087775944022, "grad_norm": 1.6456297636032104, "learning_rate": 4.906194185235085e-05, "loss": 1.7927, "step": 1331 }, { "epoch": 0.01197987165708067, "grad_norm": 1.3006504774093628, "learning_rate": 4.905969710158388e-05, "loss": 1.8596, "step": 1332 }, { "epoch": 0.011988865554721121, "grad_norm": 1.5756826400756836, "learning_rate": 4.905744971968461e-05, "loss": 1.8767, "step": 1333 }, { "epoch": 0.011997859452361573, "grad_norm": 1.358364224433899, "learning_rate": 4.9055199706898815e-05, "loss": 1.8781, "step": 1334 }, { "epoch": 0.012006853350002024, "grad_norm": 1.5803292989730835, "learning_rate": 4.905294706347255e-05, "loss": 1.8009, "step": 1335 }, { "epoch": 0.012015847247642474, "grad_norm": 1.3470792770385742, "learning_rate": 4.905069178965215e-05, "loss": 1.872, "step": 1336 }, { "epoch": 0.012024841145282926, "grad_norm": 1.3611164093017578, "learning_rate": 4.904843388568426e-05, "loss": 1.7826, "step": 1337 }, { "epoch": 0.012033835042923376, "grad_norm": 1.393494963645935, "learning_rate": 4.9046173351815796e-05, "loss": 1.7469, "step": 1338 }, { "epoch": 0.012042828940563827, "grad_norm": 1.1285642385482788, "learning_rate": 4.904391018829396e-05, "loss": 1.8459, "step": 1339 }, { "epoch": 0.012051822838204279, "grad_norm": 1.1060659885406494, "learning_rate": 4.904164439536626e-05, "loss": 1.7757, "step": 1340 }, { "epoch": 0.01206081673584473, "grad_norm": 1.2582175731658936, "learning_rate": 4.903937597328048e-05, "loss": 1.8411, "step": 1341 }, { "epoch": 0.01206981063348518, "grad_norm": 1.0940402746200562, "learning_rate": 4.9037104922284685e-05, "loss": 1.8676, "step": 1342 }, { "epoch": 0.012078804531125632, "grad_norm": 1.9910162687301636, "learning_rate": 4.9034831242627226e-05, "loss": 1.8003, "step": 1343 }, { "epoch": 0.012087798428766082, "grad_norm": 1.240732192993164, "learning_rate": 4.903255493455676e-05, "loss": 1.8257, "step": 1344 }, { "epoch": 0.012096792326406532, "grad_norm": 1.409118413925171, "learning_rate": 4.903027599832223e-05, "loss": 1.8675, "step": 1345 }, { "epoch": 0.012105786224046984, "grad_norm": 2.5294911861419678, "learning_rate": 4.9027994434172835e-05, "loss": 1.8919, "step": 1346 }, { "epoch": 0.012114780121687435, "grad_norm": 1.2260996103286743, "learning_rate": 4.9025710242358096e-05, "loss": 1.7979, "step": 1347 }, { "epoch": 0.012123774019327887, "grad_norm": 1.351508617401123, "learning_rate": 4.902342342312781e-05, "loss": 1.7644, "step": 1348 }, { "epoch": 0.012132767916968337, "grad_norm": 2.828436851501465, "learning_rate": 4.9021133976732056e-05, "loss": 1.8753, "step": 1349 }, { "epoch": 0.012141761814608788, "grad_norm": 2.693063497543335, "learning_rate": 4.901884190342121e-05, "loss": 1.9009, "step": 1350 }, { "epoch": 0.01215075571224924, "grad_norm": 2.177013635635376, "learning_rate": 4.901654720344592e-05, "loss": 1.8527, "step": 1351 }, { "epoch": 0.01215974960988969, "grad_norm": 2.032668352127075, "learning_rate": 4.901424987705713e-05, "loss": 1.7702, "step": 1352 }, { "epoch": 0.01216874350753014, "grad_norm": 1.4322946071624756, "learning_rate": 4.901194992450609e-05, "loss": 1.7856, "step": 1353 }, { "epoch": 0.012177737405170592, "grad_norm": 3.160856008529663, "learning_rate": 4.90096473460443e-05, "loss": 1.8508, "step": 1354 }, { "epoch": 0.012186731302811043, "grad_norm": 3.257728099822998, "learning_rate": 4.900734214192358e-05, "loss": 1.9053, "step": 1355 }, { "epoch": 0.012195725200451493, "grad_norm": 1.61920964717865, "learning_rate": 4.900503431239601e-05, "loss": 1.9216, "step": 1356 }, { "epoch": 0.012204719098091945, "grad_norm": 1.5187935829162598, "learning_rate": 4.9002723857713986e-05, "loss": 1.8252, "step": 1357 }, { "epoch": 0.012213712995732396, "grad_norm": 1.8567732572555542, "learning_rate": 4.900041077813017e-05, "loss": 1.7596, "step": 1358 }, { "epoch": 0.012222706893372846, "grad_norm": 1.4208694696426392, "learning_rate": 4.899809507389751e-05, "loss": 1.809, "step": 1359 }, { "epoch": 0.012231700791013298, "grad_norm": 2.4759469032287598, "learning_rate": 4.8995776745269254e-05, "loss": 1.834, "step": 1360 }, { "epoch": 0.012240694688653748, "grad_norm": 2.093702793121338, "learning_rate": 4.8993455792498935e-05, "loss": 1.783, "step": 1361 }, { "epoch": 0.012249688586294199, "grad_norm": 0.9825873970985413, "learning_rate": 4.899113221584036e-05, "loss": 1.8369, "step": 1362 }, { "epoch": 0.012258682483934651, "grad_norm": 1.4569482803344727, "learning_rate": 4.898880601554763e-05, "loss": 1.7732, "step": 1363 }, { "epoch": 0.012267676381575101, "grad_norm": 1.457384467124939, "learning_rate": 4.898647719187515e-05, "loss": 1.7598, "step": 1364 }, { "epoch": 0.012276670279215552, "grad_norm": 1.4181392192840576, "learning_rate": 4.8984145745077584e-05, "loss": 1.9303, "step": 1365 }, { "epoch": 0.012285664176856004, "grad_norm": 1.3104872703552246, "learning_rate": 4.89818116754099e-05, "loss": 1.8466, "step": 1366 }, { "epoch": 0.012294658074496454, "grad_norm": 4.076464653015137, "learning_rate": 4.897947498312735e-05, "loss": 1.7683, "step": 1367 }, { "epoch": 0.012303651972136904, "grad_norm": 1.2978509664535522, "learning_rate": 4.897713566848546e-05, "loss": 1.8823, "step": 1368 }, { "epoch": 0.012312645869777357, "grad_norm": 1.2935636043548584, "learning_rate": 4.8974793731740066e-05, "loss": 1.8201, "step": 1369 }, { "epoch": 0.012321639767417807, "grad_norm": 1.1699429750442505, "learning_rate": 4.8972449173147276e-05, "loss": 1.8521, "step": 1370 }, { "epoch": 0.012330633665058257, "grad_norm": 1.6967957019805908, "learning_rate": 4.8970101992963493e-05, "loss": 1.8677, "step": 1371 }, { "epoch": 0.01233962756269871, "grad_norm": 2.217393159866333, "learning_rate": 4.896775219144538e-05, "loss": 1.844, "step": 1372 }, { "epoch": 0.01234862146033916, "grad_norm": 1.7761520147323608, "learning_rate": 4.896539976884994e-05, "loss": 1.8054, "step": 1373 }, { "epoch": 0.01235761535797961, "grad_norm": 2.134040117263794, "learning_rate": 4.89630447254344e-05, "loss": 1.9001, "step": 1374 }, { "epoch": 0.012366609255620062, "grad_norm": 1.8075480461120605, "learning_rate": 4.8960687061456324e-05, "loss": 1.7467, "step": 1375 }, { "epoch": 0.012375603153260513, "grad_norm": 2.137967824935913, "learning_rate": 4.8958326777173525e-05, "loss": 1.7912, "step": 1376 }, { "epoch": 0.012384597050900963, "grad_norm": 1.3455866575241089, "learning_rate": 4.895596387284414e-05, "loss": 1.8133, "step": 1377 }, { "epoch": 0.012393590948541415, "grad_norm": 3.6329846382141113, "learning_rate": 4.895359834872656e-05, "loss": 1.7828, "step": 1378 }, { "epoch": 0.012402584846181865, "grad_norm": 1.9714648723602295, "learning_rate": 4.895123020507948e-05, "loss": 1.9723, "step": 1379 }, { "epoch": 0.012411578743822316, "grad_norm": 2.3921031951904297, "learning_rate": 4.8948859442161874e-05, "loss": 1.8617, "step": 1380 }, { "epoch": 0.012420572641462768, "grad_norm": 1.8725558519363403, "learning_rate": 4.8946486060233e-05, "loss": 1.896, "step": 1381 }, { "epoch": 0.012429566539103218, "grad_norm": 1.7046856880187988, "learning_rate": 4.8944110059552425e-05, "loss": 1.852, "step": 1382 }, { "epoch": 0.012438560436743669, "grad_norm": 3.738715410232544, "learning_rate": 4.894173144037997e-05, "loss": 1.8181, "step": 1383 }, { "epoch": 0.01244755433438412, "grad_norm": 3.698493719100952, "learning_rate": 4.8939350202975756e-05, "loss": 1.8217, "step": 1384 }, { "epoch": 0.012456548232024571, "grad_norm": 1.3554071187973022, "learning_rate": 4.893696634760019e-05, "loss": 1.8377, "step": 1385 }, { "epoch": 0.012465542129665023, "grad_norm": 1.7178629636764526, "learning_rate": 4.893457987451399e-05, "loss": 1.912, "step": 1386 }, { "epoch": 0.012474536027305473, "grad_norm": 1.7869948148727417, "learning_rate": 4.8932190783978105e-05, "loss": 1.722, "step": 1387 }, { "epoch": 0.012483529924945924, "grad_norm": 1.9533092975616455, "learning_rate": 4.892979907625383e-05, "loss": 1.7923, "step": 1388 }, { "epoch": 0.012492523822586376, "grad_norm": 2.3648841381073, "learning_rate": 4.892740475160269e-05, "loss": 1.8274, "step": 1389 }, { "epoch": 0.012501517720226826, "grad_norm": 2.091163158416748, "learning_rate": 4.8925007810286546e-05, "loss": 1.7412, "step": 1390 }, { "epoch": 0.012510511617867277, "grad_norm": 1.9107555150985718, "learning_rate": 4.8922608252567525e-05, "loss": 1.7679, "step": 1391 }, { "epoch": 0.012519505515507729, "grad_norm": 2.6304662227630615, "learning_rate": 4.892020607870803e-05, "loss": 1.8819, "step": 1392 }, { "epoch": 0.012528499413148179, "grad_norm": 2.2588391304016113, "learning_rate": 4.891780128897077e-05, "loss": 1.8325, "step": 1393 }, { "epoch": 0.01253749331078863, "grad_norm": 1.3935197591781616, "learning_rate": 4.891539388361871e-05, "loss": 1.8378, "step": 1394 }, { "epoch": 0.012546487208429082, "grad_norm": 1.4051916599273682, "learning_rate": 4.891298386291513e-05, "loss": 1.8681, "step": 1395 }, { "epoch": 0.012555481106069532, "grad_norm": 2.046190023422241, "learning_rate": 4.891057122712359e-05, "loss": 1.8297, "step": 1396 }, { "epoch": 0.012564475003709982, "grad_norm": 1.240532636642456, "learning_rate": 4.890815597650793e-05, "loss": 1.8509, "step": 1397 }, { "epoch": 0.012573468901350434, "grad_norm": 1.2061289548873901, "learning_rate": 4.8905738111332275e-05, "loss": 1.7826, "step": 1398 }, { "epoch": 0.012582462798990885, "grad_norm": 1.1452627182006836, "learning_rate": 4.890331763186105e-05, "loss": 1.7851, "step": 1399 }, { "epoch": 0.012591456696631335, "grad_norm": 1.0357736349105835, "learning_rate": 4.8900894538358944e-05, "loss": 1.88, "step": 1400 }, { "epoch": 0.012600450594271787, "grad_norm": 1.6439528465270996, "learning_rate": 4.8898468831090946e-05, "loss": 1.8457, "step": 1401 }, { "epoch": 0.012609444491912238, "grad_norm": 1.5162110328674316, "learning_rate": 4.8896040510322325e-05, "loss": 1.8254, "step": 1402 }, { "epoch": 0.012618438389552688, "grad_norm": 2.9260482788085938, "learning_rate": 4.8893609576318636e-05, "loss": 1.9072, "step": 1403 }, { "epoch": 0.01262743228719314, "grad_norm": 2.1529452800750732, "learning_rate": 4.889117602934574e-05, "loss": 1.8718, "step": 1404 }, { "epoch": 0.01263642618483359, "grad_norm": 1.1687769889831543, "learning_rate": 4.888873986966974e-05, "loss": 1.8048, "step": 1405 }, { "epoch": 0.01264542008247404, "grad_norm": 0.7722979784011841, "learning_rate": 4.8886301097557077e-05, "loss": 1.797, "step": 1406 }, { "epoch": 0.012654413980114493, "grad_norm": 1.5828369855880737, "learning_rate": 4.8883859713274426e-05, "loss": 1.7888, "step": 1407 }, { "epoch": 0.012663407877754943, "grad_norm": 1.9200654029846191, "learning_rate": 4.888141571708879e-05, "loss": 1.8305, "step": 1408 }, { "epoch": 0.012672401775395394, "grad_norm": 2.3463759422302246, "learning_rate": 4.887896910926744e-05, "loss": 1.7895, "step": 1409 }, { "epoch": 0.012681395673035846, "grad_norm": 1.4409575462341309, "learning_rate": 4.8876519890077926e-05, "loss": 1.7971, "step": 1410 }, { "epoch": 0.012690389570676296, "grad_norm": 1.990561604499817, "learning_rate": 4.887406805978809e-05, "loss": 1.703, "step": 1411 }, { "epoch": 0.012699383468316746, "grad_norm": 1.5289450883865356, "learning_rate": 4.887161361866608e-05, "loss": 1.8006, "step": 1412 }, { "epoch": 0.012708377365957198, "grad_norm": 4.412455081939697, "learning_rate": 4.8869156566980276e-05, "loss": 1.9263, "step": 1413 }, { "epoch": 0.012717371263597649, "grad_norm": 1.8012417554855347, "learning_rate": 4.886669690499941e-05, "loss": 1.8319, "step": 1414 }, { "epoch": 0.0127263651612381, "grad_norm": 2.1973876953125, "learning_rate": 4.8864234632992457e-05, "loss": 1.8807, "step": 1415 }, { "epoch": 0.012735359058878551, "grad_norm": 5.839619159698486, "learning_rate": 4.886176975122867e-05, "loss": 1.8023, "step": 1416 }, { "epoch": 0.012744352956519002, "grad_norm": 1.1625697612762451, "learning_rate": 4.885930225997763e-05, "loss": 1.7997, "step": 1417 }, { "epoch": 0.012753346854159452, "grad_norm": 1.6151061058044434, "learning_rate": 4.8856832159509156e-05, "loss": 1.7944, "step": 1418 }, { "epoch": 0.012762340751799904, "grad_norm": 1.1015950441360474, "learning_rate": 4.88543594500934e-05, "loss": 1.7914, "step": 1419 }, { "epoch": 0.012771334649440354, "grad_norm": 2.0917861461639404, "learning_rate": 4.885188413200075e-05, "loss": 1.8464, "step": 1420 }, { "epoch": 0.012780328547080807, "grad_norm": 2.4529812335968018, "learning_rate": 4.884940620550191e-05, "loss": 1.7582, "step": 1421 }, { "epoch": 0.012789322444721257, "grad_norm": 2.761348247528076, "learning_rate": 4.8846925670867865e-05, "loss": 1.8869, "step": 1422 }, { "epoch": 0.012798316342361707, "grad_norm": 2.008761405944824, "learning_rate": 4.884444252836988e-05, "loss": 1.7825, "step": 1423 }, { "epoch": 0.01280731024000216, "grad_norm": 3.652027130126953, "learning_rate": 4.884195677827952e-05, "loss": 1.8542, "step": 1424 }, { "epoch": 0.01281630413764261, "grad_norm": 5.074100971221924, "learning_rate": 4.8839468420868606e-05, "loss": 1.7984, "step": 1425 }, { "epoch": 0.01282529803528306, "grad_norm": 1.1754209995269775, "learning_rate": 4.8836977456409274e-05, "loss": 1.8371, "step": 1426 }, { "epoch": 0.012834291932923512, "grad_norm": 1.4902386665344238, "learning_rate": 4.883448388517391e-05, "loss": 1.7496, "step": 1427 }, { "epoch": 0.012843285830563963, "grad_norm": 1.7821979522705078, "learning_rate": 4.883198770743524e-05, "loss": 1.7378, "step": 1428 }, { "epoch": 0.012852279728204413, "grad_norm": 3.944613218307495, "learning_rate": 4.882948892346622e-05, "loss": 1.887, "step": 1429 }, { "epoch": 0.012861273625844865, "grad_norm": 2.571971893310547, "learning_rate": 4.8826987533540114e-05, "loss": 1.7834, "step": 1430 }, { "epoch": 0.012870267523485315, "grad_norm": 4.4391021728515625, "learning_rate": 4.882448353793048e-05, "loss": 1.9239, "step": 1431 }, { "epoch": 0.012879261421125766, "grad_norm": 1.0481674671173096, "learning_rate": 4.8821976936911134e-05, "loss": 1.8214, "step": 1432 }, { "epoch": 0.012888255318766218, "grad_norm": 1.9484331607818604, "learning_rate": 4.881946773075622e-05, "loss": 1.8401, "step": 1433 }, { "epoch": 0.012897249216406668, "grad_norm": 0.9597821235656738, "learning_rate": 4.8816955919740115e-05, "loss": 1.805, "step": 1434 }, { "epoch": 0.012906243114047119, "grad_norm": 1.5853670835494995, "learning_rate": 4.8814441504137534e-05, "loss": 1.8238, "step": 1435 }, { "epoch": 0.01291523701168757, "grad_norm": 1.5810338258743286, "learning_rate": 4.881192448422342e-05, "loss": 1.8277, "step": 1436 }, { "epoch": 0.012924230909328021, "grad_norm": 2.3278613090515137, "learning_rate": 4.880940486027305e-05, "loss": 1.822, "step": 1437 }, { "epoch": 0.012933224806968471, "grad_norm": 3.090679168701172, "learning_rate": 4.880688263256197e-05, "loss": 1.8113, "step": 1438 }, { "epoch": 0.012942218704608923, "grad_norm": 1.5313953161239624, "learning_rate": 4.8804357801365985e-05, "loss": 1.7847, "step": 1439 }, { "epoch": 0.012951212602249374, "grad_norm": 1.230023741722107, "learning_rate": 4.880183036696123e-05, "loss": 1.7777, "step": 1440 }, { "epoch": 0.012960206499889824, "grad_norm": 1.4640930891036987, "learning_rate": 4.879930032962409e-05, "loss": 1.7954, "step": 1441 }, { "epoch": 0.012969200397530276, "grad_norm": 0.921897828578949, "learning_rate": 4.879676768963124e-05, "loss": 1.896, "step": 1442 }, { "epoch": 0.012978194295170727, "grad_norm": 1.1121348142623901, "learning_rate": 4.8794232447259656e-05, "loss": 1.7267, "step": 1443 }, { "epoch": 0.012987188192811177, "grad_norm": 1.1427109241485596, "learning_rate": 4.879169460278659e-05, "loss": 1.8905, "step": 1444 }, { "epoch": 0.012996182090451629, "grad_norm": 1.4663699865341187, "learning_rate": 4.878915415648957e-05, "loss": 1.7388, "step": 1445 }, { "epoch": 0.01300517598809208, "grad_norm": 1.8921089172363281, "learning_rate": 4.878661110864642e-05, "loss": 1.8948, "step": 1446 }, { "epoch": 0.01301416988573253, "grad_norm": 2.2921788692474365, "learning_rate": 4.8784065459535246e-05, "loss": 1.7811, "step": 1447 }, { "epoch": 0.013023163783372982, "grad_norm": 3.003842353820801, "learning_rate": 4.8781517209434426e-05, "loss": 1.8325, "step": 1448 }, { "epoch": 0.013032157681013432, "grad_norm": 1.8661596775054932, "learning_rate": 4.877896635862263e-05, "loss": 1.8824, "step": 1449 }, { "epoch": 0.013041151578653883, "grad_norm": 2.218923568725586, "learning_rate": 4.877641290737884e-05, "loss": 1.8161, "step": 1450 }, { "epoch": 0.013050145476294335, "grad_norm": 1.1655323505401611, "learning_rate": 4.877385685598228e-05, "loss": 1.8365, "step": 1451 }, { "epoch": 0.013059139373934785, "grad_norm": 1.4436569213867188, "learning_rate": 4.877129820471247e-05, "loss": 1.9182, "step": 1452 }, { "epoch": 0.013068133271575235, "grad_norm": 2.4988341331481934, "learning_rate": 4.8768736953849234e-05, "loss": 1.7064, "step": 1453 }, { "epoch": 0.013077127169215687, "grad_norm": 2.9407799243927, "learning_rate": 4.876617310367265e-05, "loss": 1.9032, "step": 1454 }, { "epoch": 0.013086121066856138, "grad_norm": 1.3968299627304077, "learning_rate": 4.876360665446312e-05, "loss": 1.7687, "step": 1455 }, { "epoch": 0.013095114964496588, "grad_norm": 55.6673583984375, "learning_rate": 4.876103760650129e-05, "loss": 1.9494, "step": 1456 }, { "epoch": 0.01310410886213704, "grad_norm": 2.259777307510376, "learning_rate": 4.8758465960068113e-05, "loss": 1.8443, "step": 1457 }, { "epoch": 0.01311310275977749, "grad_norm": 2.2596142292022705, "learning_rate": 4.8755891715444814e-05, "loss": 1.8621, "step": 1458 }, { "epoch": 0.013122096657417943, "grad_norm": 1.064209222793579, "learning_rate": 4.875331487291292e-05, "loss": 1.8934, "step": 1459 }, { "epoch": 0.013131090555058393, "grad_norm": 1.4731348752975464, "learning_rate": 4.875073543275422e-05, "loss": 1.8221, "step": 1460 }, { "epoch": 0.013140084452698843, "grad_norm": 1.1267800331115723, "learning_rate": 4.87481533952508e-05, "loss": 1.8649, "step": 1461 }, { "epoch": 0.013149078350339296, "grad_norm": 1.0844361782073975, "learning_rate": 4.8745568760685035e-05, "loss": 1.8198, "step": 1462 }, { "epoch": 0.013158072247979746, "grad_norm": 3.615532636642456, "learning_rate": 4.8742981529339565e-05, "loss": 1.8644, "step": 1463 }, { "epoch": 0.013167066145620196, "grad_norm": 1.8813880681991577, "learning_rate": 4.874039170149733e-05, "loss": 1.7669, "step": 1464 }, { "epoch": 0.013176060043260648, "grad_norm": 0.9249187111854553, "learning_rate": 4.8737799277441566e-05, "loss": 1.8214, "step": 1465 }, { "epoch": 0.013185053940901099, "grad_norm": 1.4188538789749146, "learning_rate": 4.8735204257455744e-05, "loss": 1.7746, "step": 1466 }, { "epoch": 0.013194047838541549, "grad_norm": 1.2870889902114868, "learning_rate": 4.873260664182369e-05, "loss": 1.8014, "step": 1467 }, { "epoch": 0.013203041736182001, "grad_norm": 3.4250614643096924, "learning_rate": 4.873000643082944e-05, "loss": 1.7935, "step": 1468 }, { "epoch": 0.013212035633822452, "grad_norm": 1.4405908584594727, "learning_rate": 4.8727403624757365e-05, "loss": 1.7918, "step": 1469 }, { "epoch": 0.013221029531462902, "grad_norm": 1.6677390336990356, "learning_rate": 4.872479822389211e-05, "loss": 1.8362, "step": 1470 }, { "epoch": 0.013230023429103354, "grad_norm": 1.8401039838790894, "learning_rate": 4.8722190228518585e-05, "loss": 1.8025, "step": 1471 }, { "epoch": 0.013239017326743804, "grad_norm": 2.9253580570220947, "learning_rate": 4.871957963892201e-05, "loss": 1.9229, "step": 1472 }, { "epoch": 0.013248011224384255, "grad_norm": 1.1102653741836548, "learning_rate": 4.871696645538786e-05, "loss": 1.7011, "step": 1473 }, { "epoch": 0.013257005122024707, "grad_norm": 1.5358576774597168, "learning_rate": 4.871435067820191e-05, "loss": 1.8703, "step": 1474 }, { "epoch": 0.013265999019665157, "grad_norm": 2.1945979595184326, "learning_rate": 4.871173230765024e-05, "loss": 1.8902, "step": 1475 }, { "epoch": 0.013274992917305608, "grad_norm": 1.4347554445266724, "learning_rate": 4.8709111344019166e-05, "loss": 1.847, "step": 1476 }, { "epoch": 0.01328398681494606, "grad_norm": 1.3465237617492676, "learning_rate": 4.870648778759532e-05, "loss": 1.7316, "step": 1477 }, { "epoch": 0.01329298071258651, "grad_norm": 1.3745577335357666, "learning_rate": 4.870386163866561e-05, "loss": 1.8605, "step": 1478 }, { "epoch": 0.01330197461022696, "grad_norm": 5.627365589141846, "learning_rate": 4.870123289751724e-05, "loss": 1.801, "step": 1479 }, { "epoch": 0.013310968507867412, "grad_norm": 1.938081979751587, "learning_rate": 4.8698601564437675e-05, "loss": 1.744, "step": 1480 }, { "epoch": 0.013319962405507863, "grad_norm": 2.1216671466827393, "learning_rate": 4.869596763971466e-05, "loss": 1.7303, "step": 1481 }, { "epoch": 0.013328956303148313, "grad_norm": 2.3930485248565674, "learning_rate": 4.869333112363625e-05, "loss": 1.9447, "step": 1482 }, { "epoch": 0.013337950200788765, "grad_norm": 1.5260261297225952, "learning_rate": 4.8690692016490786e-05, "loss": 1.7986, "step": 1483 }, { "epoch": 0.013346944098429216, "grad_norm": 1.2044012546539307, "learning_rate": 4.868805031856686e-05, "loss": 1.8229, "step": 1484 }, { "epoch": 0.013355937996069666, "grad_norm": 1.7035765647888184, "learning_rate": 4.868540603015335e-05, "loss": 1.862, "step": 1485 }, { "epoch": 0.013364931893710118, "grad_norm": 2.3510403633117676, "learning_rate": 4.868275915153946e-05, "loss": 1.8362, "step": 1486 }, { "epoch": 0.013373925791350568, "grad_norm": 1.484668254852295, "learning_rate": 4.868010968301463e-05, "loss": 1.7888, "step": 1487 }, { "epoch": 0.013382919688991019, "grad_norm": 2.223858118057251, "learning_rate": 4.867745762486861e-05, "loss": 2.0123, "step": 1488 }, { "epoch": 0.013391913586631471, "grad_norm": 2.4904673099517822, "learning_rate": 4.867480297739142e-05, "loss": 1.8498, "step": 1489 }, { "epoch": 0.013400907484271921, "grad_norm": 1.3459774255752563, "learning_rate": 4.8672145740873374e-05, "loss": 1.8913, "step": 1490 }, { "epoch": 0.013409901381912372, "grad_norm": 1.458339810371399, "learning_rate": 4.866948591560506e-05, "loss": 1.8806, "step": 1491 }, { "epoch": 0.013418895279552824, "grad_norm": 1.5536245107650757, "learning_rate": 4.866682350187735e-05, "loss": 1.9818, "step": 1492 }, { "epoch": 0.013427889177193274, "grad_norm": 2.5166361331939697, "learning_rate": 4.86641584999814e-05, "loss": 1.8216, "step": 1493 }, { "epoch": 0.013436883074833726, "grad_norm": 1.079769492149353, "learning_rate": 4.866149091020867e-05, "loss": 1.852, "step": 1494 }, { "epoch": 0.013445876972474177, "grad_norm": 1.9922525882720947, "learning_rate": 4.865882073285086e-05, "loss": 1.8641, "step": 1495 }, { "epoch": 0.013454870870114627, "grad_norm": 1.4892112016677856, "learning_rate": 4.8656147968199986e-05, "loss": 1.7891, "step": 1496 }, { "epoch": 0.013463864767755079, "grad_norm": 3.1862165927886963, "learning_rate": 4.865347261654834e-05, "loss": 1.7196, "step": 1497 }, { "epoch": 0.01347285866539553, "grad_norm": 3.3255109786987305, "learning_rate": 4.8650794678188485e-05, "loss": 1.9934, "step": 1498 }, { "epoch": 0.01348185256303598, "grad_norm": 2.3318958282470703, "learning_rate": 4.8648114153413284e-05, "loss": 1.7702, "step": 1499 }, { "epoch": 0.013490846460676432, "grad_norm": 2.3319077491760254, "learning_rate": 4.864543104251587e-05, "loss": 1.8135, "step": 1500 }, { "epoch": 0.013499840358316882, "grad_norm": 4.579861640930176, "learning_rate": 4.864274534578967e-05, "loss": 1.8725, "step": 1501 }, { "epoch": 0.013508834255957333, "grad_norm": 1.1121772527694702, "learning_rate": 4.864005706352839e-05, "loss": 1.8221, "step": 1502 }, { "epoch": 0.013517828153597785, "grad_norm": 1.848832130432129, "learning_rate": 4.8637366196026006e-05, "loss": 1.7681, "step": 1503 }, { "epoch": 0.013526822051238235, "grad_norm": 3.0825748443603516, "learning_rate": 4.863467274357679e-05, "loss": 1.8032, "step": 1504 }, { "epoch": 0.013535815948878685, "grad_norm": 1.431194543838501, "learning_rate": 4.863197670647531e-05, "loss": 1.8562, "step": 1505 }, { "epoch": 0.013544809846519137, "grad_norm": 1.6103631258010864, "learning_rate": 4.862927808501638e-05, "loss": 1.8472, "step": 1506 }, { "epoch": 0.013553803744159588, "grad_norm": 4.4154767990112305, "learning_rate": 4.862657687949512e-05, "loss": 1.7728, "step": 1507 }, { "epoch": 0.013562797641800038, "grad_norm": 1.055503487586975, "learning_rate": 4.862387309020694e-05, "loss": 1.8382, "step": 1508 }, { "epoch": 0.01357179153944049, "grad_norm": 1.1550734043121338, "learning_rate": 4.862116671744752e-05, "loss": 1.8316, "step": 1509 }, { "epoch": 0.01358078543708094, "grad_norm": 1.6248600482940674, "learning_rate": 4.861845776151281e-05, "loss": 1.8396, "step": 1510 }, { "epoch": 0.013589779334721391, "grad_norm": 1.861092448234558, "learning_rate": 4.8615746222699086e-05, "loss": 1.7824, "step": 1511 }, { "epoch": 0.013598773232361843, "grad_norm": 1.5808491706848145, "learning_rate": 4.861303210130285e-05, "loss": 1.8286, "step": 1512 }, { "epoch": 0.013607767130002293, "grad_norm": 1.9786161184310913, "learning_rate": 4.861031539762093e-05, "loss": 1.8338, "step": 1513 }, { "epoch": 0.013616761027642744, "grad_norm": 1.8421313762664795, "learning_rate": 4.860759611195041e-05, "loss": 1.835, "step": 1514 }, { "epoch": 0.013625754925283196, "grad_norm": 1.2983936071395874, "learning_rate": 4.860487424458867e-05, "loss": 1.8016, "step": 1515 }, { "epoch": 0.013634748822923646, "grad_norm": 2.103151798248291, "learning_rate": 4.860214979583338e-05, "loss": 1.8487, "step": 1516 }, { "epoch": 0.013643742720564097, "grad_norm": 2.022878408432007, "learning_rate": 4.859942276598247e-05, "loss": 1.746, "step": 1517 }, { "epoch": 0.013652736618204549, "grad_norm": 1.6558146476745605, "learning_rate": 4.859669315533417e-05, "loss": 1.7284, "step": 1518 }, { "epoch": 0.013661730515844999, "grad_norm": 2.133784294128418, "learning_rate": 4.859396096418698e-05, "loss": 1.8083, "step": 1519 }, { "epoch": 0.01367072441348545, "grad_norm": 1.6492441892623901, "learning_rate": 4.8591226192839696e-05, "loss": 1.8149, "step": 1520 }, { "epoch": 0.013679718311125902, "grad_norm": 1.6266766786575317, "learning_rate": 4.858848884159138e-05, "loss": 1.8236, "step": 1521 }, { "epoch": 0.013688712208766352, "grad_norm": 3.012014150619507, "learning_rate": 4.858574891074139e-05, "loss": 1.8124, "step": 1522 }, { "epoch": 0.013697706106406802, "grad_norm": 1.1966755390167236, "learning_rate": 4.858300640058937e-05, "loss": 1.7731, "step": 1523 }, { "epoch": 0.013706700004047254, "grad_norm": 4.502533435821533, "learning_rate": 4.858026131143522e-05, "loss": 1.8868, "step": 1524 }, { "epoch": 0.013715693901687705, "grad_norm": 1.8033208847045898, "learning_rate": 4.857751364357913e-05, "loss": 1.8613, "step": 1525 }, { "epoch": 0.013724687799328155, "grad_norm": 1.476327896118164, "learning_rate": 4.8574763397321614e-05, "loss": 1.8068, "step": 1526 }, { "epoch": 0.013733681696968607, "grad_norm": 173.8759002685547, "learning_rate": 4.857201057296341e-05, "loss": 1.9632, "step": 1527 }, { "epoch": 0.013742675594609058, "grad_norm": 283.15264892578125, "learning_rate": 4.856925517080557e-05, "loss": 2.0847, "step": 1528 }, { "epoch": 0.013751669492249508, "grad_norm": 152.05560302734375, "learning_rate": 4.856649719114943e-05, "loss": 1.8971, "step": 1529 }, { "epoch": 0.01376066338988996, "grad_norm": 4.752237319946289, "learning_rate": 4.856373663429657e-05, "loss": 1.8867, "step": 1530 }, { "epoch": 0.01376965728753041, "grad_norm": 24.231088638305664, "learning_rate": 4.856097350054892e-05, "loss": 1.8306, "step": 1531 }, { "epoch": 0.013778651185170862, "grad_norm": 2.107253074645996, "learning_rate": 4.855820779020862e-05, "loss": 1.8549, "step": 1532 }, { "epoch": 0.013787645082811313, "grad_norm": 4.012585639953613, "learning_rate": 4.8555439503578136e-05, "loss": 1.8685, "step": 1533 }, { "epoch": 0.013796638980451763, "grad_norm": 3.980595350265503, "learning_rate": 4.8552668640960205e-05, "loss": 1.9353, "step": 1534 }, { "epoch": 0.013805632878092215, "grad_norm": 3.184713363647461, "learning_rate": 4.8549895202657844e-05, "loss": 1.8433, "step": 1535 }, { "epoch": 0.013814626775732666, "grad_norm": 2.5665788650512695, "learning_rate": 4.8547119188974345e-05, "loss": 1.8488, "step": 1536 }, { "epoch": 0.013823620673373116, "grad_norm": 1.3410823345184326, "learning_rate": 4.85443406002133e-05, "loss": 1.8526, "step": 1537 }, { "epoch": 0.013832614571013568, "grad_norm": 1.7222508192062378, "learning_rate": 4.8541559436678563e-05, "loss": 1.7139, "step": 1538 }, { "epoch": 0.013841608468654018, "grad_norm": 2.4887657165527344, "learning_rate": 4.8538775698674275e-05, "loss": 1.7552, "step": 1539 }, { "epoch": 0.013850602366294469, "grad_norm": 1.1853159666061401, "learning_rate": 4.853598938650487e-05, "loss": 1.8017, "step": 1540 }, { "epoch": 0.01385959626393492, "grad_norm": 1.2098760604858398, "learning_rate": 4.853320050047505e-05, "loss": 1.7299, "step": 1541 }, { "epoch": 0.013868590161575371, "grad_norm": 2.362501382827759, "learning_rate": 4.8530409040889814e-05, "loss": 1.8119, "step": 1542 }, { "epoch": 0.013877584059215822, "grad_norm": 1.5189224481582642, "learning_rate": 4.8527615008054416e-05, "loss": 1.7861, "step": 1543 }, { "epoch": 0.013886577956856274, "grad_norm": 1.0620697736740112, "learning_rate": 4.8524818402274415e-05, "loss": 1.8308, "step": 1544 }, { "epoch": 0.013895571854496724, "grad_norm": 1.7675049304962158, "learning_rate": 4.852201922385564e-05, "loss": 1.6493, "step": 1545 }, { "epoch": 0.013904565752137174, "grad_norm": 1.4743634462356567, "learning_rate": 4.85192174731042e-05, "loss": 1.8518, "step": 1546 }, { "epoch": 0.013913559649777626, "grad_norm": 1.469042181968689, "learning_rate": 4.851641315032651e-05, "loss": 1.8014, "step": 1547 }, { "epoch": 0.013922553547418077, "grad_norm": 3.847874402999878, "learning_rate": 4.851360625582924e-05, "loss": 1.8243, "step": 1548 }, { "epoch": 0.013931547445058527, "grad_norm": 2.011392116546631, "learning_rate": 4.851079678991933e-05, "loss": 1.7286, "step": 1549 }, { "epoch": 0.01394054134269898, "grad_norm": 1.4734351634979248, "learning_rate": 4.850798475290403e-05, "loss": 1.8559, "step": 1550 }, { "epoch": 0.01394953524033943, "grad_norm": 1.1149499416351318, "learning_rate": 4.850517014509087e-05, "loss": 1.8203, "step": 1551 }, { "epoch": 0.01395852913797988, "grad_norm": 1.0274463891983032, "learning_rate": 4.850235296678763e-05, "loss": 1.7786, "step": 1552 }, { "epoch": 0.013967523035620332, "grad_norm": 1.547009825706482, "learning_rate": 4.8499533218302416e-05, "loss": 1.8153, "step": 1553 }, { "epoch": 0.013976516933260782, "grad_norm": 1.3920140266418457, "learning_rate": 4.849671089994358e-05, "loss": 1.8464, "step": 1554 }, { "epoch": 0.013985510830901233, "grad_norm": 0.957093358039856, "learning_rate": 4.8493886012019765e-05, "loss": 1.8237, "step": 1555 }, { "epoch": 0.013994504728541685, "grad_norm": 2.038073778152466, "learning_rate": 4.84910585548399e-05, "loss": 1.8113, "step": 1556 }, { "epoch": 0.014003498626182135, "grad_norm": 1.61319100856781, "learning_rate": 4.8488228528713186e-05, "loss": 1.7958, "step": 1557 }, { "epoch": 0.014012492523822586, "grad_norm": 1.7915256023406982, "learning_rate": 4.8485395933949114e-05, "loss": 1.85, "step": 1558 }, { "epoch": 0.014021486421463038, "grad_norm": 1.7020790576934814, "learning_rate": 4.848256077085746e-05, "loss": 1.8727, "step": 1559 }, { "epoch": 0.014030480319103488, "grad_norm": 1.3163515329360962, "learning_rate": 4.8479723039748246e-05, "loss": 1.7305, "step": 1560 }, { "epoch": 0.014039474216743938, "grad_norm": 1.229076623916626, "learning_rate": 4.847688274093184e-05, "loss": 1.8, "step": 1561 }, { "epoch": 0.01404846811438439, "grad_norm": 1.7985280752182007, "learning_rate": 4.847403987471883e-05, "loss": 1.6978, "step": 1562 }, { "epoch": 0.014057462012024841, "grad_norm": 1.6685571670532227, "learning_rate": 4.847119444142011e-05, "loss": 1.8324, "step": 1563 }, { "epoch": 0.014066455909665291, "grad_norm": 1.8476347923278809, "learning_rate": 4.846834644134686e-05, "loss": 1.8505, "step": 1564 }, { "epoch": 0.014075449807305743, "grad_norm": 1.6671288013458252, "learning_rate": 4.846549587481052e-05, "loss": 1.7281, "step": 1565 }, { "epoch": 0.014084443704946194, "grad_norm": 2.073594331741333, "learning_rate": 4.8462642742122835e-05, "loss": 1.8758, "step": 1566 }, { "epoch": 0.014093437602586644, "grad_norm": 2.848346471786499, "learning_rate": 4.845978704359581e-05, "loss": 1.8448, "step": 1567 }, { "epoch": 0.014102431500227096, "grad_norm": 1.3396542072296143, "learning_rate": 4.845692877954175e-05, "loss": 1.8423, "step": 1568 }, { "epoch": 0.014111425397867547, "grad_norm": 2.5186805725097656, "learning_rate": 4.845406795027322e-05, "loss": 1.8061, "step": 1569 }, { "epoch": 0.014120419295507999, "grad_norm": 2.990377187728882, "learning_rate": 4.845120455610309e-05, "loss": 1.8717, "step": 1570 }, { "epoch": 0.014129413193148449, "grad_norm": 2.3382620811462402, "learning_rate": 4.8448338597344486e-05, "loss": 1.8101, "step": 1571 }, { "epoch": 0.0141384070907889, "grad_norm": 0.9175812602043152, "learning_rate": 4.8445470074310824e-05, "loss": 1.8001, "step": 1572 }, { "epoch": 0.014147400988429351, "grad_norm": 1.6969122886657715, "learning_rate": 4.8442598987315804e-05, "loss": 1.764, "step": 1573 }, { "epoch": 0.014156394886069802, "grad_norm": 2.458895206451416, "learning_rate": 4.84397253366734e-05, "loss": 1.8294, "step": 1574 }, { "epoch": 0.014165388783710252, "grad_norm": 1.7211194038391113, "learning_rate": 4.8436849122697883e-05, "loss": 1.8236, "step": 1575 }, { "epoch": 0.014174382681350704, "grad_norm": 2.061994791030884, "learning_rate": 4.843397034570378e-05, "loss": 1.7493, "step": 1576 }, { "epoch": 0.014183376578991155, "grad_norm": 2.020867109298706, "learning_rate": 4.843108900600591e-05, "loss": 1.7926, "step": 1577 }, { "epoch": 0.014192370476631605, "grad_norm": 1.1276041269302368, "learning_rate": 4.842820510391938e-05, "loss": 1.783, "step": 1578 }, { "epoch": 0.014201364374272057, "grad_norm": 4.089691162109375, "learning_rate": 4.842531863975957e-05, "loss": 1.8317, "step": 1579 }, { "epoch": 0.014210358271912507, "grad_norm": 2.3015642166137695, "learning_rate": 4.842242961384211e-05, "loss": 1.8355, "step": 1580 }, { "epoch": 0.014219352169552958, "grad_norm": 1.3081742525100708, "learning_rate": 4.841953802648298e-05, "loss": 1.7855, "step": 1581 }, { "epoch": 0.01422834606719341, "grad_norm": 2.9930551052093506, "learning_rate": 4.841664387799838e-05, "loss": 1.7762, "step": 1582 }, { "epoch": 0.01423733996483386, "grad_norm": 1.259583592414856, "learning_rate": 4.841374716870481e-05, "loss": 1.7692, "step": 1583 }, { "epoch": 0.01424633386247431, "grad_norm": 1.6943809986114502, "learning_rate": 4.841084789891905e-05, "loss": 1.8022, "step": 1584 }, { "epoch": 0.014255327760114763, "grad_norm": 1.8553165197372437, "learning_rate": 4.8407946068958165e-05, "loss": 1.8506, "step": 1585 }, { "epoch": 0.014264321657755213, "grad_norm": 1.7261384725570679, "learning_rate": 4.840504167913949e-05, "loss": 1.7534, "step": 1586 }, { "epoch": 0.014273315555395663, "grad_norm": 1.0916295051574707, "learning_rate": 4.8402134729780645e-05, "loss": 1.7983, "step": 1587 }, { "epoch": 0.014282309453036116, "grad_norm": 0.7822823524475098, "learning_rate": 4.839922522119953e-05, "loss": 1.8325, "step": 1588 }, { "epoch": 0.014291303350676566, "grad_norm": 2.950467586517334, "learning_rate": 4.839631315371433e-05, "loss": 1.8389, "step": 1589 }, { "epoch": 0.014300297248317016, "grad_norm": 1.7916351556777954, "learning_rate": 4.83933985276435e-05, "loss": 1.8202, "step": 1590 }, { "epoch": 0.014309291145957468, "grad_norm": 1.571123719215393, "learning_rate": 4.8390481343305765e-05, "loss": 1.8124, "step": 1591 }, { "epoch": 0.014318285043597919, "grad_norm": 1.7090812921524048, "learning_rate": 4.8387561601020174e-05, "loss": 1.8852, "step": 1592 }, { "epoch": 0.014327278941238369, "grad_norm": 1.0493569374084473, "learning_rate": 4.8384639301105994e-05, "loss": 1.8804, "step": 1593 }, { "epoch": 0.014336272838878821, "grad_norm": 1.5576612949371338, "learning_rate": 4.838171444388283e-05, "loss": 1.7811, "step": 1594 }, { "epoch": 0.014345266736519272, "grad_norm": 2.483229398727417, "learning_rate": 4.837878702967052e-05, "loss": 1.8098, "step": 1595 }, { "epoch": 0.014354260634159722, "grad_norm": 2.4935033321380615, "learning_rate": 4.837585705878921e-05, "loss": 1.8948, "step": 1596 }, { "epoch": 0.014363254531800174, "grad_norm": 20.9995174407959, "learning_rate": 4.8372924531559325e-05, "loss": 1.796, "step": 1597 }, { "epoch": 0.014372248429440624, "grad_norm": 1.437670350074768, "learning_rate": 4.836998944830155e-05, "loss": 1.7972, "step": 1598 }, { "epoch": 0.014381242327081075, "grad_norm": 6.437366008758545, "learning_rate": 4.836705180933686e-05, "loss": 1.6567, "step": 1599 }, { "epoch": 0.014390236224721527, "grad_norm": 5.400033950805664, "learning_rate": 4.8364111614986527e-05, "loss": 1.9052, "step": 1600 }, { "epoch": 0.014399230122361977, "grad_norm": 2.0735368728637695, "learning_rate": 4.8361168865572065e-05, "loss": 1.8384, "step": 1601 }, { "epoch": 0.014408224020002428, "grad_norm": 1.9728243350982666, "learning_rate": 4.8358223561415304e-05, "loss": 1.92, "step": 1602 }, { "epoch": 0.01441721791764288, "grad_norm": 1.992037057876587, "learning_rate": 4.8355275702838333e-05, "loss": 1.9085, "step": 1603 }, { "epoch": 0.01442621181528333, "grad_norm": 3.5353219509124756, "learning_rate": 4.8352325290163526e-05, "loss": 1.7601, "step": 1604 }, { "epoch": 0.014435205712923782, "grad_norm": 1.0799016952514648, "learning_rate": 4.834937232371353e-05, "loss": 1.7503, "step": 1605 }, { "epoch": 0.014444199610564232, "grad_norm": 2.4041640758514404, "learning_rate": 4.834641680381129e-05, "loss": 1.7089, "step": 1606 }, { "epoch": 0.014453193508204683, "grad_norm": 1.3034812211990356, "learning_rate": 4.834345873078e-05, "loss": 1.8734, "step": 1607 }, { "epoch": 0.014462187405845135, "grad_norm": 38.97063064575195, "learning_rate": 4.834049810494316e-05, "loss": 1.8043, "step": 1608 }, { "epoch": 0.014471181303485585, "grad_norm": 2.5612423419952393, "learning_rate": 4.833753492662454e-05, "loss": 1.8288, "step": 1609 }, { "epoch": 0.014480175201126036, "grad_norm": 2.002659320831299, "learning_rate": 4.83345691961482e-05, "loss": 1.7253, "step": 1610 }, { "epoch": 0.014489169098766488, "grad_norm": 1.3713325262069702, "learning_rate": 4.8331600913838446e-05, "loss": 1.8035, "step": 1611 }, { "epoch": 0.014498162996406938, "grad_norm": 2.354915142059326, "learning_rate": 4.8328630080019886e-05, "loss": 1.801, "step": 1612 }, { "epoch": 0.014507156894047388, "grad_norm": 2.6289381980895996, "learning_rate": 4.832565669501743e-05, "loss": 1.7912, "step": 1613 }, { "epoch": 0.01451615079168784, "grad_norm": 1.1337374448776245, "learning_rate": 4.832268075915622e-05, "loss": 1.7337, "step": 1614 }, { "epoch": 0.014525144689328291, "grad_norm": 2.500842809677124, "learning_rate": 4.831970227276171e-05, "loss": 1.8231, "step": 1615 }, { "epoch": 0.014534138586968741, "grad_norm": 4.007745742797852, "learning_rate": 4.831672123615961e-05, "loss": 1.7929, "step": 1616 }, { "epoch": 0.014543132484609193, "grad_norm": 1.3288958072662354, "learning_rate": 4.831373764967595e-05, "loss": 1.7801, "step": 1617 }, { "epoch": 0.014552126382249644, "grad_norm": 3.026360034942627, "learning_rate": 4.8310751513636975e-05, "loss": 1.753, "step": 1618 }, { "epoch": 0.014561120279890094, "grad_norm": 0.8103496432304382, "learning_rate": 4.8307762828369275e-05, "loss": 1.8996, "step": 1619 }, { "epoch": 0.014570114177530546, "grad_norm": 2.3469958305358887, "learning_rate": 4.830477159419966e-05, "loss": 1.7914, "step": 1620 }, { "epoch": 0.014579108075170997, "grad_norm": 1.4532898664474487, "learning_rate": 4.8301777811455276e-05, "loss": 1.8344, "step": 1621 }, { "epoch": 0.014588101972811447, "grad_norm": 3.17238450050354, "learning_rate": 4.8298781480463495e-05, "loss": 1.8262, "step": 1622 }, { "epoch": 0.014597095870451899, "grad_norm": 6.063987731933594, "learning_rate": 4.829578260155201e-05, "loss": 1.8374, "step": 1623 }, { "epoch": 0.01460608976809235, "grad_norm": 1.5122120380401611, "learning_rate": 4.829278117504876e-05, "loss": 1.7669, "step": 1624 }, { "epoch": 0.0146150836657328, "grad_norm": 1.6618303060531616, "learning_rate": 4.8289777201281974e-05, "loss": 1.7588, "step": 1625 }, { "epoch": 0.014624077563373252, "grad_norm": 1.3784470558166504, "learning_rate": 4.8286770680580176e-05, "loss": 1.8255, "step": 1626 }, { "epoch": 0.014633071461013702, "grad_norm": 1.1347072124481201, "learning_rate": 4.828376161327215e-05, "loss": 1.8552, "step": 1627 }, { "epoch": 0.014642065358654152, "grad_norm": 2.927276611328125, "learning_rate": 4.8280749999686954e-05, "loss": 1.869, "step": 1628 }, { "epoch": 0.014651059256294605, "grad_norm": 2.266026020050049, "learning_rate": 4.8277735840153945e-05, "loss": 1.8697, "step": 1629 }, { "epoch": 0.014660053153935055, "grad_norm": 1.4562363624572754, "learning_rate": 4.827471913500274e-05, "loss": 1.7509, "step": 1630 }, { "epoch": 0.014669047051575505, "grad_norm": 1.7398287057876587, "learning_rate": 4.827169988456324e-05, "loss": 1.763, "step": 1631 }, { "epoch": 0.014678040949215957, "grad_norm": 1.4238289594650269, "learning_rate": 4.826867808916563e-05, "loss": 1.7767, "step": 1632 }, { "epoch": 0.014687034846856408, "grad_norm": 1.1854379177093506, "learning_rate": 4.826565374914037e-05, "loss": 1.8988, "step": 1633 }, { "epoch": 0.014696028744496858, "grad_norm": 2.195058584213257, "learning_rate": 4.826262686481819e-05, "loss": 1.8103, "step": 1634 }, { "epoch": 0.01470502264213731, "grad_norm": 1.3138450384140015, "learning_rate": 4.8259597436530125e-05, "loss": 1.8406, "step": 1635 }, { "epoch": 0.01471401653977776, "grad_norm": 1.60219407081604, "learning_rate": 4.825656546460745e-05, "loss": 1.8683, "step": 1636 }, { "epoch": 0.014723010437418211, "grad_norm": 3.3320493698120117, "learning_rate": 4.825353094938173e-05, "loss": 1.8531, "step": 1637 }, { "epoch": 0.014732004335058663, "grad_norm": 1.6314222812652588, "learning_rate": 4.825049389118484e-05, "loss": 1.8261, "step": 1638 }, { "epoch": 0.014740998232699113, "grad_norm": 3.2105493545532227, "learning_rate": 4.8247454290348884e-05, "loss": 1.7856, "step": 1639 }, { "epoch": 0.014749992130339564, "grad_norm": 3.377990245819092, "learning_rate": 4.8244412147206284e-05, "loss": 1.9502, "step": 1640 }, { "epoch": 0.014758986027980016, "grad_norm": 1.5134758949279785, "learning_rate": 4.824136746208973e-05, "loss": 1.8135, "step": 1641 }, { "epoch": 0.014767979925620466, "grad_norm": 1.9988751411437988, "learning_rate": 4.8238320235332156e-05, "loss": 1.767, "step": 1642 }, { "epoch": 0.014776973823260918, "grad_norm": 54.28137969970703, "learning_rate": 4.823527046726682e-05, "loss": 1.7752, "step": 1643 }, { "epoch": 0.014785967720901369, "grad_norm": 46.829864501953125, "learning_rate": 4.823221815822725e-05, "loss": 1.8013, "step": 1644 }, { "epoch": 0.014794961618541819, "grad_norm": 3.810288429260254, "learning_rate": 4.822916330854722e-05, "loss": 1.8258, "step": 1645 }, { "epoch": 0.014803955516182271, "grad_norm": 1.446803092956543, "learning_rate": 4.822610591856082e-05, "loss": 1.8708, "step": 1646 }, { "epoch": 0.014812949413822721, "grad_norm": 1.7959927320480347, "learning_rate": 4.822304598860241e-05, "loss": 1.9241, "step": 1647 }, { "epoch": 0.014821943311463172, "grad_norm": 2.549199342727661, "learning_rate": 4.8219983519006587e-05, "loss": 1.8067, "step": 1648 }, { "epoch": 0.014830937209103624, "grad_norm": 1.1751540899276733, "learning_rate": 4.821691851010829e-05, "loss": 1.7568, "step": 1649 }, { "epoch": 0.014839931106744074, "grad_norm": 2.9758663177490234, "learning_rate": 4.821385096224268e-05, "loss": 1.819, "step": 1650 }, { "epoch": 0.014848925004384525, "grad_norm": 3.1565499305725098, "learning_rate": 4.821078087574523e-05, "loss": 1.8491, "step": 1651 }, { "epoch": 0.014857918902024977, "grad_norm": 3.09385347366333, "learning_rate": 4.820770825095169e-05, "loss": 1.8886, "step": 1652 }, { "epoch": 0.014866912799665427, "grad_norm": 2.1547882556915283, "learning_rate": 4.8204633088198064e-05, "loss": 1.7994, "step": 1653 }, { "epoch": 0.014875906697305877, "grad_norm": 1.433874487876892, "learning_rate": 4.8201555387820655e-05, "loss": 1.7334, "step": 1654 }, { "epoch": 0.01488490059494633, "grad_norm": 1.2014200687408447, "learning_rate": 4.819847515015602e-05, "loss": 1.8752, "step": 1655 }, { "epoch": 0.01489389449258678, "grad_norm": 1.4501962661743164, "learning_rate": 4.819539237554103e-05, "loss": 1.7885, "step": 1656 }, { "epoch": 0.01490288839022723, "grad_norm": 2.2591705322265625, "learning_rate": 4.81923070643128e-05, "loss": 1.7783, "step": 1657 }, { "epoch": 0.014911882287867682, "grad_norm": 2.574962615966797, "learning_rate": 4.818921921680874e-05, "loss": 1.882, "step": 1658 }, { "epoch": 0.014920876185508133, "grad_norm": 2.691761016845703, "learning_rate": 4.818612883336654e-05, "loss": 1.7988, "step": 1659 }, { "epoch": 0.014929870083148583, "grad_norm": 2.879997968673706, "learning_rate": 4.8183035914324136e-05, "loss": 1.7822, "step": 1660 }, { "epoch": 0.014938863980789035, "grad_norm": 1.5951367616653442, "learning_rate": 4.817994046001979e-05, "loss": 1.8659, "step": 1661 }, { "epoch": 0.014947857878429486, "grad_norm": 2.2445766925811768, "learning_rate": 4.8176842470792e-05, "loss": 1.9032, "step": 1662 }, { "epoch": 0.014956851776069936, "grad_norm": 2.0278759002685547, "learning_rate": 4.817374194697957e-05, "loss": 1.7057, "step": 1663 }, { "epoch": 0.014965845673710388, "grad_norm": 3.364875078201294, "learning_rate": 4.817063888892155e-05, "loss": 1.8383, "step": 1664 }, { "epoch": 0.014974839571350838, "grad_norm": 2.377976655960083, "learning_rate": 4.81675332969573e-05, "loss": 1.8661, "step": 1665 }, { "epoch": 0.014983833468991289, "grad_norm": 2.344567060470581, "learning_rate": 4.8164425171426456e-05, "loss": 1.8613, "step": 1666 }, { "epoch": 0.01499282736663174, "grad_norm": 1.38979971408844, "learning_rate": 4.816131451266889e-05, "loss": 1.7349, "step": 1667 }, { "epoch": 0.015001821264272191, "grad_norm": 1.548588752746582, "learning_rate": 4.815820132102479e-05, "loss": 1.7784, "step": 1668 }, { "epoch": 0.015010815161912642, "grad_norm": 3.5661778450012207, "learning_rate": 4.815508559683461e-05, "loss": 1.8166, "step": 1669 }, { "epoch": 0.015019809059553094, "grad_norm": 4.3708295822143555, "learning_rate": 4.815196734043909e-05, "loss": 1.9651, "step": 1670 }, { "epoch": 0.015028802957193544, "grad_norm": 1.1792181730270386, "learning_rate": 4.814884655217922e-05, "loss": 1.7625, "step": 1671 }, { "epoch": 0.015037796854833994, "grad_norm": 1.3058805465698242, "learning_rate": 4.81457232323963e-05, "loss": 1.7461, "step": 1672 }, { "epoch": 0.015046790752474446, "grad_norm": 3.0668060779571533, "learning_rate": 4.814259738143189e-05, "loss": 1.8455, "step": 1673 }, { "epoch": 0.015055784650114897, "grad_norm": 1.4026397466659546, "learning_rate": 4.8139468999627825e-05, "loss": 1.8577, "step": 1674 }, { "epoch": 0.015064778547755347, "grad_norm": 2.1205601692199707, "learning_rate": 4.8136338087326216e-05, "loss": 1.726, "step": 1675 }, { "epoch": 0.0150737724453958, "grad_norm": 6.079135417938232, "learning_rate": 4.813320464486946e-05, "loss": 1.7032, "step": 1676 }, { "epoch": 0.01508276634303625, "grad_norm": 1.4485489130020142, "learning_rate": 4.813006867260023e-05, "loss": 1.7788, "step": 1677 }, { "epoch": 0.0150917602406767, "grad_norm": 2.2670578956604004, "learning_rate": 4.812693017086145e-05, "loss": 1.8669, "step": 1678 }, { "epoch": 0.015100754138317152, "grad_norm": 2.9982357025146484, "learning_rate": 4.8123789139996364e-05, "loss": 1.6638, "step": 1679 }, { "epoch": 0.015109748035957602, "grad_norm": 3.1424055099487305, "learning_rate": 4.812064558034847e-05, "loss": 1.7148, "step": 1680 }, { "epoch": 0.015118741933598055, "grad_norm": 4.752673625946045, "learning_rate": 4.811749949226154e-05, "loss": 1.812, "step": 1681 }, { "epoch": 0.015127735831238505, "grad_norm": 1.354101300239563, "learning_rate": 4.811435087607961e-05, "loss": 1.788, "step": 1682 }, { "epoch": 0.015136729728878955, "grad_norm": 2.5098066329956055, "learning_rate": 4.811119973214703e-05, "loss": 1.7605, "step": 1683 }, { "epoch": 0.015145723626519407, "grad_norm": 3.278367519378662, "learning_rate": 4.810804606080839e-05, "loss": 1.8857, "step": 1684 }, { "epoch": 0.015154717524159858, "grad_norm": 4.482705593109131, "learning_rate": 4.810488986240858e-05, "loss": 1.7362, "step": 1685 }, { "epoch": 0.015163711421800308, "grad_norm": 1.4310725927352905, "learning_rate": 4.810173113729274e-05, "loss": 1.8809, "step": 1686 }, { "epoch": 0.01517270531944076, "grad_norm": 2.404642343521118, "learning_rate": 4.809856988580632e-05, "loss": 1.8468, "step": 1687 }, { "epoch": 0.01518169921708121, "grad_norm": 1.851311206817627, "learning_rate": 4.8095406108295036e-05, "loss": 1.8116, "step": 1688 }, { "epoch": 0.015190693114721661, "grad_norm": 1.723460078239441, "learning_rate": 4.809223980510486e-05, "loss": 1.7243, "step": 1689 }, { "epoch": 0.015199687012362113, "grad_norm": 2.492219924926758, "learning_rate": 4.8089070976582054e-05, "loss": 1.7769, "step": 1690 }, { "epoch": 0.015208680910002563, "grad_norm": 1.8249189853668213, "learning_rate": 4.808589962307316e-05, "loss": 1.7779, "step": 1691 }, { "epoch": 0.015217674807643014, "grad_norm": 2.744149923324585, "learning_rate": 4.8082725744924984e-05, "loss": 1.7673, "step": 1692 }, { "epoch": 0.015226668705283466, "grad_norm": 2.3099639415740967, "learning_rate": 4.807954934248464e-05, "loss": 1.6697, "step": 1693 }, { "epoch": 0.015235662602923916, "grad_norm": 1.066101312637329, "learning_rate": 4.807637041609947e-05, "loss": 1.7734, "step": 1694 }, { "epoch": 0.015244656500564367, "grad_norm": 1.2943838834762573, "learning_rate": 4.8073188966117126e-05, "loss": 1.8657, "step": 1695 }, { "epoch": 0.015253650398204819, "grad_norm": 3.412285804748535, "learning_rate": 4.807000499288553e-05, "loss": 1.8188, "step": 1696 }, { "epoch": 0.015262644295845269, "grad_norm": 1.269877314567566, "learning_rate": 4.8066818496752875e-05, "loss": 1.7398, "step": 1697 }, { "epoch": 0.01527163819348572, "grad_norm": 3.1054880619049072, "learning_rate": 4.806362947806763e-05, "loss": 1.6852, "step": 1698 }, { "epoch": 0.015280632091126171, "grad_norm": 1.2002006769180298, "learning_rate": 4.8060437937178536e-05, "loss": 1.7814, "step": 1699 }, { "epoch": 0.015289625988766622, "grad_norm": 1.6226311922073364, "learning_rate": 4.805724387443462e-05, "loss": 1.8048, "step": 1700 }, { "epoch": 0.015298619886407072, "grad_norm": 1.6813570261001587, "learning_rate": 4.8054047290185186e-05, "loss": 1.8213, "step": 1701 }, { "epoch": 0.015307613784047524, "grad_norm": 2.1889772415161133, "learning_rate": 4.805084818477979e-05, "loss": 1.7581, "step": 1702 }, { "epoch": 0.015316607681687975, "grad_norm": 1.956467866897583, "learning_rate": 4.8047646558568304e-05, "loss": 1.853, "step": 1703 }, { "epoch": 0.015325601579328425, "grad_norm": 1.3462538719177246, "learning_rate": 4.804444241190084e-05, "loss": 1.8721, "step": 1704 }, { "epoch": 0.015334595476968877, "grad_norm": 2.049537181854248, "learning_rate": 4.80412357451278e-05, "loss": 1.7067, "step": 1705 }, { "epoch": 0.015343589374609327, "grad_norm": 1.6702535152435303, "learning_rate": 4.803802655859986e-05, "loss": 1.7913, "step": 1706 }, { "epoch": 0.015352583272249778, "grad_norm": 1.7616528272628784, "learning_rate": 4.8034814852667974e-05, "loss": 1.807, "step": 1707 }, { "epoch": 0.01536157716989023, "grad_norm": 1.8447901010513306, "learning_rate": 4.803160062768337e-05, "loss": 1.8542, "step": 1708 }, { "epoch": 0.01537057106753068, "grad_norm": 1.7721887826919556, "learning_rate": 4.802838388399754e-05, "loss": 1.794, "step": 1709 }, { "epoch": 0.01537956496517113, "grad_norm": 0.9180960059165955, "learning_rate": 4.8025164621962284e-05, "loss": 1.7461, "step": 1710 }, { "epoch": 0.015388558862811583, "grad_norm": 1.0209616422653198, "learning_rate": 4.802194284192963e-05, "loss": 1.8931, "step": 1711 }, { "epoch": 0.015397552760452033, "grad_norm": 1.8501290082931519, "learning_rate": 4.801871854425193e-05, "loss": 1.723, "step": 1712 }, { "epoch": 0.015406546658092483, "grad_norm": 1.1178843975067139, "learning_rate": 4.801549172928177e-05, "loss": 1.7421, "step": 1713 }, { "epoch": 0.015415540555732936, "grad_norm": 1.7294328212738037, "learning_rate": 4.801226239737204e-05, "loss": 1.8459, "step": 1714 }, { "epoch": 0.015424534453373386, "grad_norm": 1.0545718669891357, "learning_rate": 4.8009030548875896e-05, "loss": 1.7816, "step": 1715 }, { "epoch": 0.015433528351013838, "grad_norm": 0.8855057954788208, "learning_rate": 4.800579618414676e-05, "loss": 1.7972, "step": 1716 }, { "epoch": 0.015442522248654288, "grad_norm": 1.5310602188110352, "learning_rate": 4.8002559303538344e-05, "loss": 1.8696, "step": 1717 }, { "epoch": 0.015451516146294739, "grad_norm": 1.8585712909698486, "learning_rate": 4.799931990740463e-05, "loss": 1.7839, "step": 1718 }, { "epoch": 0.01546051004393519, "grad_norm": 1.1869429349899292, "learning_rate": 4.7996077996099867e-05, "loss": 1.7719, "step": 1719 }, { "epoch": 0.015469503941575641, "grad_norm": 1.4101265668869019, "learning_rate": 4.799283356997859e-05, "loss": 1.7196, "step": 1720 }, { "epoch": 0.015478497839216092, "grad_norm": 1.8724441528320312, "learning_rate": 4.79895866293956e-05, "loss": 1.8484, "step": 1721 }, { "epoch": 0.015487491736856544, "grad_norm": 1.8255493640899658, "learning_rate": 4.798633717470598e-05, "loss": 1.7924, "step": 1722 }, { "epoch": 0.015496485634496994, "grad_norm": 2.478550434112549, "learning_rate": 4.798308520626509e-05, "loss": 1.7289, "step": 1723 }, { "epoch": 0.015505479532137444, "grad_norm": 2.4164607524871826, "learning_rate": 4.797983072442855e-05, "loss": 1.8004, "step": 1724 }, { "epoch": 0.015514473429777896, "grad_norm": 3.02645206451416, "learning_rate": 4.797657372955228e-05, "loss": 1.824, "step": 1725 }, { "epoch": 0.015523467327418347, "grad_norm": 2.0300581455230713, "learning_rate": 4.797331422199245e-05, "loss": 1.7974, "step": 1726 }, { "epoch": 0.015532461225058797, "grad_norm": 0.9558500647544861, "learning_rate": 4.797005220210551e-05, "loss": 1.6841, "step": 1727 }, { "epoch": 0.01554145512269925, "grad_norm": 1.6371674537658691, "learning_rate": 4.79667876702482e-05, "loss": 1.6478, "step": 1728 }, { "epoch": 0.0155504490203397, "grad_norm": 1.7166789770126343, "learning_rate": 4.7963520626777524e-05, "loss": 1.8473, "step": 1729 }, { "epoch": 0.01555944291798015, "grad_norm": 1.6148964166641235, "learning_rate": 4.796025107205075e-05, "loss": 1.8324, "step": 1730 }, { "epoch": 0.015568436815620602, "grad_norm": 1.5862544775009155, "learning_rate": 4.7956979006425454e-05, "loss": 1.8818, "step": 1731 }, { "epoch": 0.015577430713261052, "grad_norm": 1.4109218120574951, "learning_rate": 4.7953704430259424e-05, "loss": 1.7401, "step": 1732 }, { "epoch": 0.015586424610901503, "grad_norm": 1.547189712524414, "learning_rate": 4.795042734391081e-05, "loss": 1.7993, "step": 1733 }, { "epoch": 0.015595418508541955, "grad_norm": 3.2345640659332275, "learning_rate": 4.794714774773796e-05, "loss": 1.9328, "step": 1734 }, { "epoch": 0.015604412406182405, "grad_norm": 1.837807297706604, "learning_rate": 4.794386564209953e-05, "loss": 1.7145, "step": 1735 }, { "epoch": 0.015613406303822856, "grad_norm": 2.23380184173584, "learning_rate": 4.7940581027354456e-05, "loss": 1.7943, "step": 1736 }, { "epoch": 0.015622400201463308, "grad_norm": 2.284756898880005, "learning_rate": 4.7937293903861926e-05, "loss": 1.846, "step": 1737 }, { "epoch": 0.01563139409910376, "grad_norm": 3.4412975311279297, "learning_rate": 4.7934004271981415e-05, "loss": 1.7191, "step": 1738 }, { "epoch": 0.01564038799674421, "grad_norm": 1.9518203735351562, "learning_rate": 4.793071213207269e-05, "loss": 1.7614, "step": 1739 }, { "epoch": 0.01564938189438466, "grad_norm": 2.871177911758423, "learning_rate": 4.792741748449575e-05, "loss": 1.8493, "step": 1740 }, { "epoch": 0.01565837579202511, "grad_norm": 1.1859278678894043, "learning_rate": 4.792412032961091e-05, "loss": 1.7445, "step": 1741 }, { "epoch": 0.01566736968966556, "grad_norm": 1.7426611185073853, "learning_rate": 4.7920820667778744e-05, "loss": 1.8181, "step": 1742 }, { "epoch": 0.01567636358730601, "grad_norm": 1.4955413341522217, "learning_rate": 4.791751849936008e-05, "loss": 1.8018, "step": 1743 }, { "epoch": 0.015685357484946465, "grad_norm": 1.4823225736618042, "learning_rate": 4.791421382471605e-05, "loss": 1.784, "step": 1744 }, { "epoch": 0.015694351382586916, "grad_norm": 1.5989598035812378, "learning_rate": 4.7910906644208054e-05, "loss": 1.7481, "step": 1745 }, { "epoch": 0.015703345280227366, "grad_norm": 1.5515244007110596, "learning_rate": 4.7907596958197745e-05, "loss": 1.8138, "step": 1746 }, { "epoch": 0.015712339177867816, "grad_norm": 2.348510980606079, "learning_rate": 4.790428476704708e-05, "loss": 1.7824, "step": 1747 }, { "epoch": 0.015721333075508267, "grad_norm": 1.2950592041015625, "learning_rate": 4.790097007111827e-05, "loss": 1.8134, "step": 1748 }, { "epoch": 0.015730326973148717, "grad_norm": 1.6541672945022583, "learning_rate": 4.78976528707738e-05, "loss": 1.8565, "step": 1749 }, { "epoch": 0.01573932087078917, "grad_norm": 1.7645363807678223, "learning_rate": 4.789433316637644e-05, "loss": 1.8155, "step": 1750 }, { "epoch": 0.01574831476842962, "grad_norm": 1.071670651435852, "learning_rate": 4.789101095828922e-05, "loss": 1.7274, "step": 1751 }, { "epoch": 0.015757308666070072, "grad_norm": 2.1450493335723877, "learning_rate": 4.7887686246875465e-05, "loss": 1.8559, "step": 1752 }, { "epoch": 0.015766302563710522, "grad_norm": 1.4066542387008667, "learning_rate": 4.788435903249875e-05, "loss": 1.7345, "step": 1753 }, { "epoch": 0.015775296461350972, "grad_norm": 2.9065680503845215, "learning_rate": 4.788102931552294e-05, "loss": 1.779, "step": 1754 }, { "epoch": 0.015784290358991423, "grad_norm": 1.7117174863815308, "learning_rate": 4.787769709631217e-05, "loss": 1.8603, "step": 1755 }, { "epoch": 0.015793284256631877, "grad_norm": 1.50350022315979, "learning_rate": 4.787436237523083e-05, "loss": 1.8498, "step": 1756 }, { "epoch": 0.015802278154272327, "grad_norm": 1.0733273029327393, "learning_rate": 4.7871025152643616e-05, "loss": 1.8443, "step": 1757 }, { "epoch": 0.015811272051912777, "grad_norm": 1.8059583902359009, "learning_rate": 4.786768542891548e-05, "loss": 1.7372, "step": 1758 }, { "epoch": 0.015820265949553228, "grad_norm": 2.319636344909668, "learning_rate": 4.786434320441164e-05, "loss": 1.9422, "step": 1759 }, { "epoch": 0.015829259847193678, "grad_norm": 1.6729220151901245, "learning_rate": 4.786099847949761e-05, "loss": 1.7799, "step": 1760 }, { "epoch": 0.01583825374483413, "grad_norm": 1.4784845113754272, "learning_rate": 4.785765125453916e-05, "loss": 1.8541, "step": 1761 }, { "epoch": 0.015847247642474582, "grad_norm": 1.8279677629470825, "learning_rate": 4.785430152990233e-05, "loss": 1.8298, "step": 1762 }, { "epoch": 0.015856241540115033, "grad_norm": 1.7301273345947266, "learning_rate": 4.7850949305953444e-05, "loss": 1.8755, "step": 1763 }, { "epoch": 0.015865235437755483, "grad_norm": 1.3777868747711182, "learning_rate": 4.7847594583059096e-05, "loss": 1.8195, "step": 1764 }, { "epoch": 0.015874229335395933, "grad_norm": 2.3328006267547607, "learning_rate": 4.784423736158616e-05, "loss": 1.8556, "step": 1765 }, { "epoch": 0.015883223233036384, "grad_norm": 2.2803874015808105, "learning_rate": 4.784087764190177e-05, "loss": 1.8457, "step": 1766 }, { "epoch": 0.015892217130676834, "grad_norm": 1.4935146570205688, "learning_rate": 4.7837515424373344e-05, "loss": 1.8014, "step": 1767 }, { "epoch": 0.015901211028317288, "grad_norm": 2.0289924144744873, "learning_rate": 4.783415070936857e-05, "loss": 1.6932, "step": 1768 }, { "epoch": 0.01591020492595774, "grad_norm": 1.4899898767471313, "learning_rate": 4.78307834972554e-05, "loss": 1.7443, "step": 1769 }, { "epoch": 0.01591919882359819, "grad_norm": 1.227825403213501, "learning_rate": 4.7827413788402077e-05, "loss": 1.6921, "step": 1770 }, { "epoch": 0.01592819272123864, "grad_norm": 1.2597500085830688, "learning_rate": 4.7824041583177096e-05, "loss": 1.7222, "step": 1771 }, { "epoch": 0.01593718661887909, "grad_norm": 1.9342271089553833, "learning_rate": 4.7820666881949246e-05, "loss": 1.8343, "step": 1772 }, { "epoch": 0.015946180516519543, "grad_norm": 1.7157479524612427, "learning_rate": 4.7817289685087577e-05, "loss": 1.8857, "step": 1773 }, { "epoch": 0.015955174414159994, "grad_norm": 3.4089574813842773, "learning_rate": 4.781390999296141e-05, "loss": 1.8108, "step": 1774 }, { "epoch": 0.015964168311800444, "grad_norm": 2.751850128173828, "learning_rate": 4.781052780594034e-05, "loss": 1.8099, "step": 1775 }, { "epoch": 0.015973162209440894, "grad_norm": 1.3564200401306152, "learning_rate": 4.780714312439426e-05, "loss": 1.7384, "step": 1776 }, { "epoch": 0.015982156107081345, "grad_norm": 5.355855464935303, "learning_rate": 4.780375594869329e-05, "loss": 1.8076, "step": 1777 }, { "epoch": 0.015991150004721795, "grad_norm": 1.351666808128357, "learning_rate": 4.7800366279207854e-05, "loss": 1.7622, "step": 1778 }, { "epoch": 0.01600014390236225, "grad_norm": 3.090740203857422, "learning_rate": 4.779697411630865e-05, "loss": 1.6874, "step": 1779 }, { "epoch": 0.0160091378000027, "grad_norm": 0.8739086389541626, "learning_rate": 4.779357946036661e-05, "loss": 1.7778, "step": 1780 }, { "epoch": 0.01601813169764315, "grad_norm": 3.215498685836792, "learning_rate": 4.779018231175301e-05, "loss": 1.7604, "step": 1781 }, { "epoch": 0.0160271255952836, "grad_norm": 1.5966434478759766, "learning_rate": 4.778678267083933e-05, "loss": 1.7803, "step": 1782 }, { "epoch": 0.01603611949292405, "grad_norm": 1.8294808864593506, "learning_rate": 4.778338053799736e-05, "loss": 1.8296, "step": 1783 }, { "epoch": 0.0160451133905645, "grad_norm": 2.5392637252807617, "learning_rate": 4.777997591359914e-05, "loss": 1.9013, "step": 1784 }, { "epoch": 0.016054107288204954, "grad_norm": 1.1805580854415894, "learning_rate": 4.777656879801701e-05, "loss": 1.8004, "step": 1785 }, { "epoch": 0.016063101185845405, "grad_norm": 1.10187828540802, "learning_rate": 4.7773159191623565e-05, "loss": 1.9051, "step": 1786 }, { "epoch": 0.016072095083485855, "grad_norm": 1.8131848573684692, "learning_rate": 4.776974709479166e-05, "loss": 1.8983, "step": 1787 }, { "epoch": 0.016081088981126306, "grad_norm": 3.434161901473999, "learning_rate": 4.776633250789445e-05, "loss": 1.8227, "step": 1788 }, { "epoch": 0.016090082878766756, "grad_norm": 0.982336163520813, "learning_rate": 4.7762915431305346e-05, "loss": 1.7184, "step": 1789 }, { "epoch": 0.016099076776407206, "grad_norm": 0.9403204917907715, "learning_rate": 4.775949586539803e-05, "loss": 1.7029, "step": 1790 }, { "epoch": 0.01610807067404766, "grad_norm": 3.4730401039123535, "learning_rate": 4.7756073810546476e-05, "loss": 1.7066, "step": 1791 }, { "epoch": 0.01611706457168811, "grad_norm": 2.187244415283203, "learning_rate": 4.775264926712489e-05, "loss": 1.7819, "step": 1792 }, { "epoch": 0.01612605846932856, "grad_norm": 1.1191089153289795, "learning_rate": 4.7749222235507794e-05, "loss": 1.7878, "step": 1793 }, { "epoch": 0.01613505236696901, "grad_norm": 1.8590977191925049, "learning_rate": 4.774579271606996e-05, "loss": 1.8475, "step": 1794 }, { "epoch": 0.01614404626460946, "grad_norm": 1.302479863166809, "learning_rate": 4.774236070918643e-05, "loss": 1.7395, "step": 1795 }, { "epoch": 0.016153040162249912, "grad_norm": 2.1414496898651123, "learning_rate": 4.7738926215232524e-05, "loss": 1.779, "step": 1796 }, { "epoch": 0.016162034059890366, "grad_norm": 1.6297699213027954, "learning_rate": 4.7735489234583844e-05, "loss": 1.6847, "step": 1797 }, { "epoch": 0.016171027957530816, "grad_norm": 4.208197593688965, "learning_rate": 4.773204976761623e-05, "loss": 1.9154, "step": 1798 }, { "epoch": 0.016180021855171266, "grad_norm": 1.7834408283233643, "learning_rate": 4.772860781470584e-05, "loss": 1.8064, "step": 1799 }, { "epoch": 0.016189015752811717, "grad_norm": 3.109675168991089, "learning_rate": 4.7725163376229064e-05, "loss": 1.738, "step": 1800 }, { "epoch": 0.016198009650452167, "grad_norm": 1.2883381843566895, "learning_rate": 4.772171645256259e-05, "loss": 1.696, "step": 1801 }, { "epoch": 0.016207003548092618, "grad_norm": 3.4083049297332764, "learning_rate": 4.771826704408337e-05, "loss": 1.8614, "step": 1802 }, { "epoch": 0.01621599744573307, "grad_norm": 1.6579647064208984, "learning_rate": 4.771481515116862e-05, "loss": 1.7616, "step": 1803 }, { "epoch": 0.01622499134337352, "grad_norm": 1.1972907781600952, "learning_rate": 4.7711360774195835e-05, "loss": 1.7747, "step": 1804 }, { "epoch": 0.016233985241013972, "grad_norm": 2.081352949142456, "learning_rate": 4.770790391354279e-05, "loss": 1.8182, "step": 1805 }, { "epoch": 0.016242979138654422, "grad_norm": 2.1658570766448975, "learning_rate": 4.7704444569587506e-05, "loss": 1.8731, "step": 1806 }, { "epoch": 0.016251973036294873, "grad_norm": 2.597641944885254, "learning_rate": 4.77009827427083e-05, "loss": 1.9573, "step": 1807 }, { "epoch": 0.016260966933935327, "grad_norm": 1.9512230157852173, "learning_rate": 4.769751843328375e-05, "loss": 1.7018, "step": 1808 }, { "epoch": 0.016269960831575777, "grad_norm": 2.6993677616119385, "learning_rate": 4.7694051641692714e-05, "loss": 1.7704, "step": 1809 }, { "epoch": 0.016278954729216227, "grad_norm": 1.8426004648208618, "learning_rate": 4.7690582368314304e-05, "loss": 1.8641, "step": 1810 }, { "epoch": 0.016287948626856678, "grad_norm": 1.9076555967330933, "learning_rate": 4.7687110613527926e-05, "loss": 1.7453, "step": 1811 }, { "epoch": 0.016296942524497128, "grad_norm": 2.035515546798706, "learning_rate": 4.768363637771324e-05, "loss": 1.7672, "step": 1812 }, { "epoch": 0.01630593642213758, "grad_norm": 1.1833724975585938, "learning_rate": 4.768015966125018e-05, "loss": 1.7526, "step": 1813 }, { "epoch": 0.016314930319778032, "grad_norm": 0.8760007619857788, "learning_rate": 4.767668046451896e-05, "loss": 1.8517, "step": 1814 }, { "epoch": 0.016323924217418483, "grad_norm": 2.0089566707611084, "learning_rate": 4.7673198787900063e-05, "loss": 1.754, "step": 1815 }, { "epoch": 0.016332918115058933, "grad_norm": 1.6475714445114136, "learning_rate": 4.7669714631774235e-05, "loss": 1.8944, "step": 1816 }, { "epoch": 0.016341912012699383, "grad_norm": 3.218343734741211, "learning_rate": 4.76662279965225e-05, "loss": 1.8236, "step": 1817 }, { "epoch": 0.016350905910339834, "grad_norm": 1.9155243635177612, "learning_rate": 4.7662738882526146e-05, "loss": 1.9139, "step": 1818 }, { "epoch": 0.016359899807980284, "grad_norm": 3.893763542175293, "learning_rate": 4.765924729016674e-05, "loss": 1.7062, "step": 1819 }, { "epoch": 0.016368893705620738, "grad_norm": 1.716532588005066, "learning_rate": 4.7655753219826114e-05, "loss": 1.7793, "step": 1820 }, { "epoch": 0.016377887603261188, "grad_norm": 2.6965653896331787, "learning_rate": 4.7652256671886386e-05, "loss": 1.7828, "step": 1821 }, { "epoch": 0.01638688150090164, "grad_norm": 1.2937854528427124, "learning_rate": 4.764875764672992e-05, "loss": 1.8004, "step": 1822 }, { "epoch": 0.01639587539854209, "grad_norm": 1.797347903251648, "learning_rate": 4.764525614473937e-05, "loss": 1.7831, "step": 1823 }, { "epoch": 0.01640486929618254, "grad_norm": 1.3847588300704956, "learning_rate": 4.764175216629766e-05, "loss": 1.8209, "step": 1824 }, { "epoch": 0.01641386319382299, "grad_norm": 3.807542562484741, "learning_rate": 4.763824571178798e-05, "loss": 1.767, "step": 1825 }, { "epoch": 0.016422857091463443, "grad_norm": 1.3101102113723755, "learning_rate": 4.763473678159377e-05, "loss": 1.8513, "step": 1826 }, { "epoch": 0.016431850989103894, "grad_norm": 0.9215659499168396, "learning_rate": 4.7631225376098784e-05, "loss": 1.7716, "step": 1827 }, { "epoch": 0.016440844886744344, "grad_norm": 1.2295252084732056, "learning_rate": 4.762771149568701e-05, "loss": 1.8058, "step": 1828 }, { "epoch": 0.016449838784384795, "grad_norm": 0.9341084361076355, "learning_rate": 4.7624195140742734e-05, "loss": 1.767, "step": 1829 }, { "epoch": 0.016458832682025245, "grad_norm": 1.673149585723877, "learning_rate": 4.762067631165049e-05, "loss": 1.7992, "step": 1830 }, { "epoch": 0.016467826579665695, "grad_norm": 1.2018946409225464, "learning_rate": 4.7617155008795094e-05, "loss": 1.7639, "step": 1831 }, { "epoch": 0.01647682047730615, "grad_norm": 1.7744864225387573, "learning_rate": 4.761363123256163e-05, "loss": 1.7383, "step": 1832 }, { "epoch": 0.0164858143749466, "grad_norm": 1.1177303791046143, "learning_rate": 4.761010498333546e-05, "loss": 1.6828, "step": 1833 }, { "epoch": 0.01649480827258705, "grad_norm": 1.2488800287246704, "learning_rate": 4.760657626150219e-05, "loss": 1.7562, "step": 1834 }, { "epoch": 0.0165038021702275, "grad_norm": 0.9924839735031128, "learning_rate": 4.760304506744774e-05, "loss": 1.7631, "step": 1835 }, { "epoch": 0.01651279606786795, "grad_norm": 1.777870774269104, "learning_rate": 4.759951140155826e-05, "loss": 1.7341, "step": 1836 }, { "epoch": 0.0165217899655084, "grad_norm": 2.294938802719116, "learning_rate": 4.759597526422019e-05, "loss": 1.7536, "step": 1837 }, { "epoch": 0.016530783863148855, "grad_norm": 2.3084335327148438, "learning_rate": 4.759243665582024e-05, "loss": 1.7635, "step": 1838 }, { "epoch": 0.016539777760789305, "grad_norm": 1.1973991394042969, "learning_rate": 4.7588895576745385e-05, "loss": 1.8615, "step": 1839 }, { "epoch": 0.016548771658429755, "grad_norm": 2.306457757949829, "learning_rate": 4.7585352027382877e-05, "loss": 1.87, "step": 1840 }, { "epoch": 0.016557765556070206, "grad_norm": 3.060605525970459, "learning_rate": 4.7581806008120216e-05, "loss": 1.9712, "step": 1841 }, { "epoch": 0.016566759453710656, "grad_norm": 1.5962820053100586, "learning_rate": 4.75782575193452e-05, "loss": 1.865, "step": 1842 }, { "epoch": 0.016575753351351107, "grad_norm": 1.5321162939071655, "learning_rate": 4.75747065614459e-05, "loss": 1.7235, "step": 1843 }, { "epoch": 0.01658474724899156, "grad_norm": 2.5975916385650635, "learning_rate": 4.7571153134810634e-05, "loss": 1.7042, "step": 1844 }, { "epoch": 0.01659374114663201, "grad_norm": 0.9650520086288452, "learning_rate": 4.7567597239827974e-05, "loss": 1.6851, "step": 1845 }, { "epoch": 0.01660273504427246, "grad_norm": 1.1925281286239624, "learning_rate": 4.756403887688683e-05, "loss": 1.8438, "step": 1846 }, { "epoch": 0.01661172894191291, "grad_norm": 2.0071001052856445, "learning_rate": 4.7560478046376314e-05, "loss": 1.8507, "step": 1847 }, { "epoch": 0.016620722839553362, "grad_norm": 3.6864712238311768, "learning_rate": 4.755691474868583e-05, "loss": 1.7335, "step": 1848 }, { "epoch": 0.016629716737193816, "grad_norm": 1.8577417135238647, "learning_rate": 4.755334898420507e-05, "loss": 1.8458, "step": 1849 }, { "epoch": 0.016638710634834266, "grad_norm": 1.470793604850769, "learning_rate": 4.754978075332398e-05, "loss": 1.8002, "step": 1850 }, { "epoch": 0.016647704532474716, "grad_norm": 1.2529234886169434, "learning_rate": 4.754621005643276e-05, "loss": 1.8473, "step": 1851 }, { "epoch": 0.016656698430115167, "grad_norm": 2.873105049133301, "learning_rate": 4.754263689392191e-05, "loss": 1.7744, "step": 1852 }, { "epoch": 0.016665692327755617, "grad_norm": 1.300893783569336, "learning_rate": 4.753906126618218e-05, "loss": 1.7305, "step": 1853 }, { "epoch": 0.016674686225396067, "grad_norm": 2.0291919708251953, "learning_rate": 4.753548317360459e-05, "loss": 1.8847, "step": 1854 }, { "epoch": 0.01668368012303652, "grad_norm": 3.254068374633789, "learning_rate": 4.753190261658045e-05, "loss": 1.8614, "step": 1855 }, { "epoch": 0.01669267402067697, "grad_norm": 1.7601491212844849, "learning_rate": 4.752831959550132e-05, "loss": 1.7854, "step": 1856 }, { "epoch": 0.016701667918317422, "grad_norm": 5.800058364868164, "learning_rate": 4.752473411075902e-05, "loss": 1.9272, "step": 1857 }, { "epoch": 0.016710661815957872, "grad_norm": 1.2117910385131836, "learning_rate": 4.752114616274567e-05, "loss": 1.8118, "step": 1858 }, { "epoch": 0.016719655713598323, "grad_norm": 1.4531995058059692, "learning_rate": 4.7517555751853635e-05, "loss": 1.8128, "step": 1859 }, { "epoch": 0.016728649611238773, "grad_norm": 1.9837855100631714, "learning_rate": 4.751396287847556e-05, "loss": 1.795, "step": 1860 }, { "epoch": 0.016737643508879227, "grad_norm": 2.4843268394470215, "learning_rate": 4.751036754300434e-05, "loss": 1.7885, "step": 1861 }, { "epoch": 0.016746637406519677, "grad_norm": 1.8759526014328003, "learning_rate": 4.7506769745833184e-05, "loss": 1.7978, "step": 1862 }, { "epoch": 0.016755631304160128, "grad_norm": 2.5738418102264404, "learning_rate": 4.750316948735553e-05, "loss": 1.7418, "step": 1863 }, { "epoch": 0.016764625201800578, "grad_norm": 1.4375330209732056, "learning_rate": 4.749956676796507e-05, "loss": 1.7478, "step": 1864 }, { "epoch": 0.01677361909944103, "grad_norm": 1.6116092205047607, "learning_rate": 4.7495961588055836e-05, "loss": 1.8503, "step": 1865 }, { "epoch": 0.01678261299708148, "grad_norm": 1.91148841381073, "learning_rate": 4.749235394802207e-05, "loss": 1.7648, "step": 1866 }, { "epoch": 0.016791606894721933, "grad_norm": 3.0706627368927, "learning_rate": 4.7488743848258274e-05, "loss": 1.8029, "step": 1867 }, { "epoch": 0.016800600792362383, "grad_norm": 1.6969808340072632, "learning_rate": 4.7485131289159276e-05, "loss": 1.7393, "step": 1868 }, { "epoch": 0.016809594690002833, "grad_norm": 8.383678436279297, "learning_rate": 4.748151627112012e-05, "loss": 1.8773, "step": 1869 }, { "epoch": 0.016818588587643284, "grad_norm": 218.32174682617188, "learning_rate": 4.747789879453615e-05, "loss": 1.9337, "step": 1870 }, { "epoch": 0.016827582485283734, "grad_norm": 37.34574890136719, "learning_rate": 4.7474278859802965e-05, "loss": 1.7061, "step": 1871 }, { "epoch": 0.016836576382924184, "grad_norm": 2.710949659347534, "learning_rate": 4.7470656467316424e-05, "loss": 1.7965, "step": 1872 }, { "epoch": 0.016845570280564638, "grad_norm": 4.781021595001221, "learning_rate": 4.7467031617472676e-05, "loss": 1.8435, "step": 1873 }, { "epoch": 0.01685456417820509, "grad_norm": 3.376084804534912, "learning_rate": 4.746340431066814e-05, "loss": 1.7413, "step": 1874 }, { "epoch": 0.01686355807584554, "grad_norm": 3.3176639080047607, "learning_rate": 4.7459774547299475e-05, "loss": 1.8391, "step": 1875 }, { "epoch": 0.01687255197348599, "grad_norm": 1.710574746131897, "learning_rate": 4.745614232776363e-05, "loss": 1.8873, "step": 1876 }, { "epoch": 0.01688154587112644, "grad_norm": 2.843493938446045, "learning_rate": 4.745250765245783e-05, "loss": 1.8809, "step": 1877 }, { "epoch": 0.01689053976876689, "grad_norm": 3.4479215145111084, "learning_rate": 4.744887052177954e-05, "loss": 1.6734, "step": 1878 }, { "epoch": 0.016899533666407344, "grad_norm": 1.1058439016342163, "learning_rate": 4.744523093612653e-05, "loss": 1.7888, "step": 1879 }, { "epoch": 0.016908527564047794, "grad_norm": 1.3097167015075684, "learning_rate": 4.7441588895896805e-05, "loss": 1.7603, "step": 1880 }, { "epoch": 0.016917521461688245, "grad_norm": 0.9356676340103149, "learning_rate": 4.7437944401488666e-05, "loss": 1.7326, "step": 1881 }, { "epoch": 0.016926515359328695, "grad_norm": 1.0802165269851685, "learning_rate": 4.743429745330066e-05, "loss": 1.8243, "step": 1882 }, { "epoch": 0.016935509256969145, "grad_norm": 2.2058911323547363, "learning_rate": 4.7430648051731606e-05, "loss": 1.7257, "step": 1883 }, { "epoch": 0.0169445031546096, "grad_norm": 1.9081239700317383, "learning_rate": 4.742699619718061e-05, "loss": 1.7601, "step": 1884 }, { "epoch": 0.01695349705225005, "grad_norm": 0.9012009501457214, "learning_rate": 4.742334189004704e-05, "loss": 1.8419, "step": 1885 }, { "epoch": 0.0169624909498905, "grad_norm": 16.562746047973633, "learning_rate": 4.74196851307305e-05, "loss": 1.827, "step": 1886 }, { "epoch": 0.01697148484753095, "grad_norm": 1.3532160520553589, "learning_rate": 4.7416025919630904e-05, "loss": 1.7658, "step": 1887 }, { "epoch": 0.0169804787451714, "grad_norm": 1.397816777229309, "learning_rate": 4.741236425714841e-05, "loss": 1.6768, "step": 1888 }, { "epoch": 0.01698947264281185, "grad_norm": 51.39292526245117, "learning_rate": 4.740870014368346e-05, "loss": 1.8774, "step": 1889 }, { "epoch": 0.016998466540452305, "grad_norm": 1.873118281364441, "learning_rate": 4.7405033579636756e-05, "loss": 1.7767, "step": 1890 }, { "epoch": 0.017007460438092755, "grad_norm": 1.1507703065872192, "learning_rate": 4.740136456540927e-05, "loss": 1.7712, "step": 1891 }, { "epoch": 0.017016454335733205, "grad_norm": 1.9281471967697144, "learning_rate": 4.739769310140223e-05, "loss": 1.8407, "step": 1892 }, { "epoch": 0.017025448233373656, "grad_norm": 1.067179560661316, "learning_rate": 4.739401918801715e-05, "loss": 1.7805, "step": 1893 }, { "epoch": 0.017034442131014106, "grad_norm": 1.3636798858642578, "learning_rate": 4.7390342825655795e-05, "loss": 1.9002, "step": 1894 }, { "epoch": 0.017043436028654557, "grad_norm": 1.226452112197876, "learning_rate": 4.738666401472022e-05, "loss": 1.7387, "step": 1895 }, { "epoch": 0.01705242992629501, "grad_norm": 1.375666618347168, "learning_rate": 4.7382982755612716e-05, "loss": 1.6875, "step": 1896 }, { "epoch": 0.01706142382393546, "grad_norm": 1.3347879648208618, "learning_rate": 4.7379299048735876e-05, "loss": 1.8056, "step": 1897 }, { "epoch": 0.01707041772157591, "grad_norm": 1.3625867366790771, "learning_rate": 4.7375612894492536e-05, "loss": 1.7298, "step": 1898 }, { "epoch": 0.01707941161921636, "grad_norm": 1.2139872312545776, "learning_rate": 4.737192429328582e-05, "loss": 1.6505, "step": 1899 }, { "epoch": 0.017088405516856812, "grad_norm": 1.1424305438995361, "learning_rate": 4.736823324551909e-05, "loss": 1.7638, "step": 1900 }, { "epoch": 0.017097399414497262, "grad_norm": 2.5335311889648438, "learning_rate": 4.736453975159601e-05, "loss": 1.8454, "step": 1901 }, { "epoch": 0.017106393312137716, "grad_norm": 3.9321887493133545, "learning_rate": 4.736084381192048e-05, "loss": 1.8265, "step": 1902 }, { "epoch": 0.017115387209778166, "grad_norm": 1.0327022075653076, "learning_rate": 4.735714542689669e-05, "loss": 1.7659, "step": 1903 }, { "epoch": 0.017124381107418617, "grad_norm": 1.2502151727676392, "learning_rate": 4.735344459692909e-05, "loss": 1.8544, "step": 1904 }, { "epoch": 0.017133375005059067, "grad_norm": 1.3599454164505005, "learning_rate": 4.73497413224224e-05, "loss": 1.8254, "step": 1905 }, { "epoch": 0.017142368902699517, "grad_norm": 1.3991692066192627, "learning_rate": 4.73460356037816e-05, "loss": 1.8382, "step": 1906 }, { "epoch": 0.017151362800339968, "grad_norm": 1.5937280654907227, "learning_rate": 4.734232744141194e-05, "loss": 1.8199, "step": 1907 }, { "epoch": 0.01716035669798042, "grad_norm": 1.997373104095459, "learning_rate": 4.733861683571894e-05, "loss": 1.7568, "step": 1908 }, { "epoch": 0.017169350595620872, "grad_norm": 1.4168542623519897, "learning_rate": 4.7334903787108397e-05, "loss": 1.8218, "step": 1909 }, { "epoch": 0.017178344493261322, "grad_norm": 1.4413458108901978, "learning_rate": 4.733118829598635e-05, "loss": 1.7704, "step": 1910 }, { "epoch": 0.017187338390901773, "grad_norm": 1.3852629661560059, "learning_rate": 4.732747036275913e-05, "loss": 1.8271, "step": 1911 }, { "epoch": 0.017196332288542223, "grad_norm": 2.0745537281036377, "learning_rate": 4.732374998783332e-05, "loss": 1.7408, "step": 1912 }, { "epoch": 0.017205326186182673, "grad_norm": 1.5210301876068115, "learning_rate": 4.732002717161577e-05, "loss": 1.8701, "step": 1913 }, { "epoch": 0.017214320083823127, "grad_norm": 1.6176573038101196, "learning_rate": 4.7316301914513605e-05, "loss": 1.7061, "step": 1914 }, { "epoch": 0.017223313981463578, "grad_norm": 1.2809315919876099, "learning_rate": 4.7312574216934225e-05, "loss": 1.7675, "step": 1915 }, { "epoch": 0.017232307879104028, "grad_norm": 3.1373283863067627, "learning_rate": 4.730884407928526e-05, "loss": 1.8399, "step": 1916 }, { "epoch": 0.01724130177674448, "grad_norm": 1.1123007535934448, "learning_rate": 4.730511150197466e-05, "loss": 1.746, "step": 1917 }, { "epoch": 0.01725029567438493, "grad_norm": 0.9087127447128296, "learning_rate": 4.7301376485410596e-05, "loss": 1.782, "step": 1918 }, { "epoch": 0.017259289572025382, "grad_norm": 0.7619174718856812, "learning_rate": 4.729763903000153e-05, "loss": 1.8258, "step": 1919 }, { "epoch": 0.017268283469665833, "grad_norm": 2.568309783935547, "learning_rate": 4.7293899136156184e-05, "loss": 1.8378, "step": 1920 }, { "epoch": 0.017277277367306283, "grad_norm": 1.1579744815826416, "learning_rate": 4.729015680428355e-05, "loss": 1.8355, "step": 1921 }, { "epoch": 0.017286271264946734, "grad_norm": 1.0003058910369873, "learning_rate": 4.728641203479287e-05, "loss": 1.7609, "step": 1922 }, { "epoch": 0.017295265162587184, "grad_norm": 2.0153815746307373, "learning_rate": 4.7282664828093686e-05, "loss": 1.7191, "step": 1923 }, { "epoch": 0.017304259060227634, "grad_norm": 1.0943570137023926, "learning_rate": 4.7278915184595774e-05, "loss": 1.7959, "step": 1924 }, { "epoch": 0.017313252957868088, "grad_norm": 1.6851940155029297, "learning_rate": 4.72751631047092e-05, "loss": 1.8288, "step": 1925 }, { "epoch": 0.01732224685550854, "grad_norm": 2.057771921157837, "learning_rate": 4.727140858884427e-05, "loss": 1.964, "step": 1926 }, { "epoch": 0.01733124075314899, "grad_norm": 1.4990499019622803, "learning_rate": 4.726765163741159e-05, "loss": 1.8527, "step": 1927 }, { "epoch": 0.01734023465078944, "grad_norm": 1.3713092803955078, "learning_rate": 4.7263892250822e-05, "loss": 1.6586, "step": 1928 }, { "epoch": 0.01734922854842989, "grad_norm": 0.8301932215690613, "learning_rate": 4.7260130429486627e-05, "loss": 1.7841, "step": 1929 }, { "epoch": 0.01735822244607034, "grad_norm": 1.2117825746536255, "learning_rate": 4.725636617381686e-05, "loss": 1.6991, "step": 1930 }, { "epoch": 0.017367216343710794, "grad_norm": 1.321224331855774, "learning_rate": 4.725259948422435e-05, "loss": 1.7622, "step": 1931 }, { "epoch": 0.017376210241351244, "grad_norm": 1.749241828918457, "learning_rate": 4.7248830361121014e-05, "loss": 1.7775, "step": 1932 }, { "epoch": 0.017385204138991694, "grad_norm": 1.7020156383514404, "learning_rate": 4.724505880491904e-05, "loss": 1.822, "step": 1933 }, { "epoch": 0.017394198036632145, "grad_norm": 0.9136528372764587, "learning_rate": 4.7241284816030885e-05, "loss": 1.6858, "step": 1934 }, { "epoch": 0.017403191934272595, "grad_norm": 2.4850454330444336, "learning_rate": 4.723750839486926e-05, "loss": 1.7604, "step": 1935 }, { "epoch": 0.017412185831913046, "grad_norm": 1.2753173112869263, "learning_rate": 4.7233729541847147e-05, "loss": 1.8753, "step": 1936 }, { "epoch": 0.0174211797295535, "grad_norm": 1.5836960077285767, "learning_rate": 4.72299482573778e-05, "loss": 1.7761, "step": 1937 }, { "epoch": 0.01743017362719395, "grad_norm": 1.4738447666168213, "learning_rate": 4.722616454187474e-05, "loss": 1.8779, "step": 1938 }, { "epoch": 0.0174391675248344, "grad_norm": 1.0590041875839233, "learning_rate": 4.722237839575174e-05, "loss": 1.6653, "step": 1939 }, { "epoch": 0.01744816142247485, "grad_norm": 1.6688333749771118, "learning_rate": 4.721858981942284e-05, "loss": 1.863, "step": 1940 }, { "epoch": 0.0174571553201153, "grad_norm": 1.28078031539917, "learning_rate": 4.7214798813302376e-05, "loss": 1.8086, "step": 1941 }, { "epoch": 0.01746614921775575, "grad_norm": 1.3899922370910645, "learning_rate": 4.7211005377804904e-05, "loss": 1.7307, "step": 1942 }, { "epoch": 0.017475143115396205, "grad_norm": 1.3408582210540771, "learning_rate": 4.720720951334528e-05, "loss": 1.7886, "step": 1943 }, { "epoch": 0.017484137013036655, "grad_norm": 1.6098254919052124, "learning_rate": 4.720341122033862e-05, "loss": 1.7968, "step": 1944 }, { "epoch": 0.017493130910677106, "grad_norm": 4.081319808959961, "learning_rate": 4.719961049920027e-05, "loss": 1.9477, "step": 1945 }, { "epoch": 0.017502124808317556, "grad_norm": 2.731729507446289, "learning_rate": 4.719580735034592e-05, "loss": 1.7313, "step": 1946 }, { "epoch": 0.017511118705958006, "grad_norm": 1.1532678604125977, "learning_rate": 4.719200177419143e-05, "loss": 1.8083, "step": 1947 }, { "epoch": 0.017520112603598457, "grad_norm": 1.0959571599960327, "learning_rate": 4.7188193771153e-05, "loss": 1.6913, "step": 1948 }, { "epoch": 0.01752910650123891, "grad_norm": 1.6646416187286377, "learning_rate": 4.7184383341647055e-05, "loss": 1.7177, "step": 1949 }, { "epoch": 0.01753810039887936, "grad_norm": 1.210663080215454, "learning_rate": 4.71805704860903e-05, "loss": 1.8023, "step": 1950 }, { "epoch": 0.01754709429651981, "grad_norm": 1.1513780355453491, "learning_rate": 4.7176755204899705e-05, "loss": 1.792, "step": 1951 }, { "epoch": 0.01755608819416026, "grad_norm": 1.1846307516098022, "learning_rate": 4.71729374984925e-05, "loss": 1.7087, "step": 1952 }, { "epoch": 0.017565082091800712, "grad_norm": 3.0145461559295654, "learning_rate": 4.71691173672862e-05, "loss": 1.8018, "step": 1953 }, { "epoch": 0.017574075989441162, "grad_norm": 1.1263823509216309, "learning_rate": 4.716529481169854e-05, "loss": 1.922, "step": 1954 }, { "epoch": 0.017583069887081616, "grad_norm": 1.768072485923767, "learning_rate": 4.716146983214757e-05, "loss": 1.7447, "step": 1955 }, { "epoch": 0.017592063784722067, "grad_norm": 1.4011521339416504, "learning_rate": 4.715764242905158e-05, "loss": 1.7768, "step": 1956 }, { "epoch": 0.017601057682362517, "grad_norm": 2.9728055000305176, "learning_rate": 4.715381260282913e-05, "loss": 1.7878, "step": 1957 }, { "epoch": 0.017610051580002967, "grad_norm": 2.5514392852783203, "learning_rate": 4.714998035389904e-05, "loss": 1.9805, "step": 1958 }, { "epoch": 0.017619045477643418, "grad_norm": 1.9937174320220947, "learning_rate": 4.7146145682680396e-05, "loss": 1.7415, "step": 1959 }, { "epoch": 0.01762803937528387, "grad_norm": 1.1720342636108398, "learning_rate": 4.7142308589592556e-05, "loss": 1.6706, "step": 1960 }, { "epoch": 0.017637033272924322, "grad_norm": 1.289574146270752, "learning_rate": 4.713846907505514e-05, "loss": 1.7271, "step": 1961 }, { "epoch": 0.017646027170564772, "grad_norm": 1.125266671180725, "learning_rate": 4.713462713948804e-05, "loss": 1.7164, "step": 1962 }, { "epoch": 0.017655021068205223, "grad_norm": 1.4566586017608643, "learning_rate": 4.713078278331138e-05, "loss": 1.7133, "step": 1963 }, { "epoch": 0.017664014965845673, "grad_norm": 1.4664883613586426, "learning_rate": 4.71269360069456e-05, "loss": 1.8431, "step": 1964 }, { "epoch": 0.017673008863486123, "grad_norm": 1.537471890449524, "learning_rate": 4.7123086810811356e-05, "loss": 1.8017, "step": 1965 }, { "epoch": 0.017682002761126577, "grad_norm": 1.2928153276443481, "learning_rate": 4.711923519532961e-05, "loss": 1.6699, "step": 1966 }, { "epoch": 0.017690996658767028, "grad_norm": 0.9362931251525879, "learning_rate": 4.7115381160921556e-05, "loss": 1.7361, "step": 1967 }, { "epoch": 0.017699990556407478, "grad_norm": 2.310028314590454, "learning_rate": 4.711152470800867e-05, "loss": 1.8399, "step": 1968 }, { "epoch": 0.017708984454047928, "grad_norm": 1.6124167442321777, "learning_rate": 4.710766583701267e-05, "loss": 1.7186, "step": 1969 }, { "epoch": 0.01771797835168838, "grad_norm": 1.1458983421325684, "learning_rate": 4.710380454835559e-05, "loss": 1.8326, "step": 1970 }, { "epoch": 0.01772697224932883, "grad_norm": 1.2335609197616577, "learning_rate": 4.709994084245967e-05, "loss": 1.8258, "step": 1971 }, { "epoch": 0.017735966146969283, "grad_norm": 2.0945217609405518, "learning_rate": 4.709607471974745e-05, "loss": 1.6811, "step": 1972 }, { "epoch": 0.017744960044609733, "grad_norm": 1.5925264358520508, "learning_rate": 4.7092206180641715e-05, "loss": 1.6753, "step": 1973 }, { "epoch": 0.017753953942250184, "grad_norm": 1.837701678276062, "learning_rate": 4.7088335225565535e-05, "loss": 1.6916, "step": 1974 }, { "epoch": 0.017762947839890634, "grad_norm": 2.9025893211364746, "learning_rate": 4.708446185494222e-05, "loss": 1.8601, "step": 1975 }, { "epoch": 0.017771941737531084, "grad_norm": 1.288460373878479, "learning_rate": 4.7080586069195354e-05, "loss": 1.7667, "step": 1976 }, { "epoch": 0.017780935635171535, "grad_norm": 2.161412239074707, "learning_rate": 4.707670786874881e-05, "loss": 1.6357, "step": 1977 }, { "epoch": 0.01778992953281199, "grad_norm": 1.5766348838806152, "learning_rate": 4.7072827254026666e-05, "loss": 1.8227, "step": 1978 }, { "epoch": 0.01779892343045244, "grad_norm": 2.0156702995300293, "learning_rate": 4.706894422545333e-05, "loss": 1.7146, "step": 1979 }, { "epoch": 0.01780791732809289, "grad_norm": 1.1456137895584106, "learning_rate": 4.7065058783453424e-05, "loss": 1.7537, "step": 1980 }, { "epoch": 0.01781691122573334, "grad_norm": 1.1138010025024414, "learning_rate": 4.706117092845187e-05, "loss": 1.658, "step": 1981 }, { "epoch": 0.01782590512337379, "grad_norm": 2.204423189163208, "learning_rate": 4.7057280660873835e-05, "loss": 1.7308, "step": 1982 }, { "epoch": 0.01783489902101424, "grad_norm": 1.3840837478637695, "learning_rate": 4.705338798114475e-05, "loss": 1.7689, "step": 1983 }, { "epoch": 0.017843892918654694, "grad_norm": 1.6691497564315796, "learning_rate": 4.704949288969031e-05, "loss": 1.7995, "step": 1984 }, { "epoch": 0.017852886816295144, "grad_norm": 1.5037338733673096, "learning_rate": 4.704559538693647e-05, "loss": 1.6602, "step": 1985 }, { "epoch": 0.017861880713935595, "grad_norm": 1.5733495950698853, "learning_rate": 4.7041695473309475e-05, "loss": 1.8593, "step": 1986 }, { "epoch": 0.017870874611576045, "grad_norm": 2.037429094314575, "learning_rate": 4.70377931492358e-05, "loss": 1.8064, "step": 1987 }, { "epoch": 0.017879868509216496, "grad_norm": 1.4983861446380615, "learning_rate": 4.703388841514219e-05, "loss": 1.8049, "step": 1988 }, { "epoch": 0.017888862406856946, "grad_norm": 1.423203945159912, "learning_rate": 4.702998127145568e-05, "loss": 1.747, "step": 1989 }, { "epoch": 0.0178978563044974, "grad_norm": 1.5918467044830322, "learning_rate": 4.7026071718603536e-05, "loss": 1.6658, "step": 1990 }, { "epoch": 0.01790685020213785, "grad_norm": 1.807916522026062, "learning_rate": 4.7022159757013314e-05, "loss": 1.6999, "step": 1991 }, { "epoch": 0.0179158440997783, "grad_norm": 1.566667914390564, "learning_rate": 4.70182453871128e-05, "loss": 1.6348, "step": 1992 }, { "epoch": 0.01792483799741875, "grad_norm": 1.9964667558670044, "learning_rate": 4.7014328609330076e-05, "loss": 1.7557, "step": 1993 }, { "epoch": 0.0179338318950592, "grad_norm": 1.7289563417434692, "learning_rate": 4.701040942409348e-05, "loss": 1.7968, "step": 1994 }, { "epoch": 0.017942825792699655, "grad_norm": 1.849856972694397, "learning_rate": 4.700648783183159e-05, "loss": 1.8615, "step": 1995 }, { "epoch": 0.017951819690340105, "grad_norm": 0.9449866414070129, "learning_rate": 4.700256383297329e-05, "loss": 1.7312, "step": 1996 }, { "epoch": 0.017960813587980556, "grad_norm": 1.247435212135315, "learning_rate": 4.699863742794768e-05, "loss": 1.9314, "step": 1997 }, { "epoch": 0.017969807485621006, "grad_norm": 2.6913833618164062, "learning_rate": 4.6994708617184155e-05, "loss": 1.6922, "step": 1998 }, { "epoch": 0.017978801383261456, "grad_norm": 1.8059782981872559, "learning_rate": 4.699077740111237e-05, "loss": 1.7435, "step": 1999 }, { "epoch": 0.017987795280901907, "grad_norm": 1.8075207471847534, "learning_rate": 4.698684378016222e-05, "loss": 1.8178, "step": 2000 }, { "epoch": 0.01799678917854236, "grad_norm": 2.804123640060425, "learning_rate": 4.6982907754763906e-05, "loss": 1.826, "step": 2001 }, { "epoch": 0.01800578307618281, "grad_norm": 2.5402069091796875, "learning_rate": 4.697896932534784e-05, "loss": 1.7519, "step": 2002 }, { "epoch": 0.01801477697382326, "grad_norm": 2.0269501209259033, "learning_rate": 4.6975028492344745e-05, "loss": 1.7593, "step": 2003 }, { "epoch": 0.01802377087146371, "grad_norm": 2.5007810592651367, "learning_rate": 4.697108525618556e-05, "loss": 1.755, "step": 2004 }, { "epoch": 0.018032764769104162, "grad_norm": 1.3249776363372803, "learning_rate": 4.696713961730154e-05, "loss": 1.799, "step": 2005 }, { "epoch": 0.018041758666744612, "grad_norm": 1.4541486501693726, "learning_rate": 4.696319157612414e-05, "loss": 1.8134, "step": 2006 }, { "epoch": 0.018050752564385066, "grad_norm": 1.1957463026046753, "learning_rate": 4.695924113308514e-05, "loss": 1.8537, "step": 2007 }, { "epoch": 0.018059746462025517, "grad_norm": 1.0813088417053223, "learning_rate": 4.695528828861655e-05, "loss": 1.7495, "step": 2008 }, { "epoch": 0.018068740359665967, "grad_norm": 3.56712007522583, "learning_rate": 4.695133304315064e-05, "loss": 1.8962, "step": 2009 }, { "epoch": 0.018077734257306417, "grad_norm": 2.173736095428467, "learning_rate": 4.694737539711994e-05, "loss": 1.8939, "step": 2010 }, { "epoch": 0.018086728154946868, "grad_norm": 1.3032269477844238, "learning_rate": 4.6943415350957276e-05, "loss": 1.7559, "step": 2011 }, { "epoch": 0.018095722052587318, "grad_norm": 1.362119197845459, "learning_rate": 4.6939452905095696e-05, "loss": 1.6948, "step": 2012 }, { "epoch": 0.018104715950227772, "grad_norm": 1.6742527484893799, "learning_rate": 4.6935488059968535e-05, "loss": 1.7548, "step": 2013 }, { "epoch": 0.018113709847868222, "grad_norm": 1.936282753944397, "learning_rate": 4.6931520816009374e-05, "loss": 1.7377, "step": 2014 }, { "epoch": 0.018122703745508673, "grad_norm": 1.1959228515625, "learning_rate": 4.6927551173652075e-05, "loss": 1.7217, "step": 2015 }, { "epoch": 0.018131697643149123, "grad_norm": 3.0763299465179443, "learning_rate": 4.692357913333074e-05, "loss": 1.8213, "step": 2016 }, { "epoch": 0.018140691540789573, "grad_norm": 1.5072171688079834, "learning_rate": 4.691960469547976e-05, "loss": 1.753, "step": 2017 }, { "epoch": 0.018149685438430024, "grad_norm": 1.619361162185669, "learning_rate": 4.691562786053376e-05, "loss": 1.7035, "step": 2018 }, { "epoch": 0.018158679336070477, "grad_norm": 1.6535924673080444, "learning_rate": 4.691164862892765e-05, "loss": 1.7789, "step": 2019 }, { "epoch": 0.018167673233710928, "grad_norm": 1.2890434265136719, "learning_rate": 4.690766700109659e-05, "loss": 1.6908, "step": 2020 }, { "epoch": 0.018176667131351378, "grad_norm": 6.389407157897949, "learning_rate": 4.6903682977476e-05, "loss": 1.7897, "step": 2021 }, { "epoch": 0.01818566102899183, "grad_norm": 1.724674940109253, "learning_rate": 4.6899696558501585e-05, "loss": 1.8001, "step": 2022 }, { "epoch": 0.01819465492663228, "grad_norm": 1.0120362043380737, "learning_rate": 4.689570774460927e-05, "loss": 1.8351, "step": 2023 }, { "epoch": 0.01820364882427273, "grad_norm": 1.6619237661361694, "learning_rate": 4.6891716536235275e-05, "loss": 1.8173, "step": 2024 }, { "epoch": 0.018212642721913183, "grad_norm": 1.4153820276260376, "learning_rate": 4.6887722933816076e-05, "loss": 1.5808, "step": 2025 }, { "epoch": 0.018221636619553633, "grad_norm": 1.5488519668579102, "learning_rate": 4.6883726937788405e-05, "loss": 1.8124, "step": 2026 }, { "epoch": 0.018230630517194084, "grad_norm": 1.9685413837432861, "learning_rate": 4.687972854858925e-05, "loss": 1.7896, "step": 2027 }, { "epoch": 0.018239624414834534, "grad_norm": 2.0616328716278076, "learning_rate": 4.6875727766655887e-05, "loss": 1.8456, "step": 2028 }, { "epoch": 0.018248618312474985, "grad_norm": 1.4026118516921997, "learning_rate": 4.687172459242582e-05, "loss": 1.8497, "step": 2029 }, { "epoch": 0.01825761221011544, "grad_norm": 1.589443564414978, "learning_rate": 4.686771902633684e-05, "loss": 1.6523, "step": 2030 }, { "epoch": 0.01826660610775589, "grad_norm": 1.3182772397994995, "learning_rate": 4.686371106882698e-05, "loss": 1.758, "step": 2031 }, { "epoch": 0.01827560000539634, "grad_norm": 0.8969163298606873, "learning_rate": 4.6859700720334546e-05, "loss": 1.7757, "step": 2032 }, { "epoch": 0.01828459390303679, "grad_norm": 1.2089407444000244, "learning_rate": 4.6855687981298115e-05, "loss": 1.7086, "step": 2033 }, { "epoch": 0.01829358780067724, "grad_norm": 1.4697484970092773, "learning_rate": 4.68516728521565e-05, "loss": 1.7406, "step": 2034 }, { "epoch": 0.01830258169831769, "grad_norm": 1.0593236684799194, "learning_rate": 4.684765533334879e-05, "loss": 1.7829, "step": 2035 }, { "epoch": 0.018311575595958144, "grad_norm": 3.082319736480713, "learning_rate": 4.684363542531436e-05, "loss": 1.6475, "step": 2036 }, { "epoch": 0.018320569493598594, "grad_norm": 1.5255863666534424, "learning_rate": 4.683961312849279e-05, "loss": 1.834, "step": 2037 }, { "epoch": 0.018329563391239045, "grad_norm": 1.6226601600646973, "learning_rate": 4.6835588443323956e-05, "loss": 1.8516, "step": 2038 }, { "epoch": 0.018338557288879495, "grad_norm": 1.6288037300109863, "learning_rate": 4.683156137024801e-05, "loss": 1.8491, "step": 2039 }, { "epoch": 0.018347551186519945, "grad_norm": 3.16646671295166, "learning_rate": 4.682753190970533e-05, "loss": 1.8701, "step": 2040 }, { "epoch": 0.018356545084160396, "grad_norm": 1.5983729362487793, "learning_rate": 4.682350006213658e-05, "loss": 1.8809, "step": 2041 }, { "epoch": 0.01836553898180085, "grad_norm": 1.6236598491668701, "learning_rate": 4.681946582798267e-05, "loss": 1.6511, "step": 2042 }, { "epoch": 0.0183745328794413, "grad_norm": 2.8938143253326416, "learning_rate": 4.6815429207684786e-05, "loss": 1.8414, "step": 2043 }, { "epoch": 0.01838352677708175, "grad_norm": 3.642702102661133, "learning_rate": 4.681139020168436e-05, "loss": 1.7896, "step": 2044 }, { "epoch": 0.0183925206747222, "grad_norm": 2.761230230331421, "learning_rate": 4.68073488104231e-05, "loss": 1.8106, "step": 2045 }, { "epoch": 0.01840151457236265, "grad_norm": 1.8581364154815674, "learning_rate": 4.680330503434296e-05, "loss": 1.8098, "step": 2046 }, { "epoch": 0.0184105084700031, "grad_norm": 1.7268410921096802, "learning_rate": 4.679925887388615e-05, "loss": 1.7846, "step": 2047 }, { "epoch": 0.018419502367643555, "grad_norm": 4.727581024169922, "learning_rate": 4.679521032949518e-05, "loss": 1.9262, "step": 2048 }, { "epoch": 0.018428496265284006, "grad_norm": 1.1492565870285034, "learning_rate": 4.679115940161277e-05, "loss": 1.7905, "step": 2049 }, { "epoch": 0.018437490162924456, "grad_norm": 1.8270750045776367, "learning_rate": 4.678710609068193e-05, "loss": 1.8168, "step": 2050 }, { "epoch": 0.018446484060564906, "grad_norm": 2.3605525493621826, "learning_rate": 4.678305039714593e-05, "loss": 1.7854, "step": 2051 }, { "epoch": 0.018455477958205357, "grad_norm": 2.58406662940979, "learning_rate": 4.677899232144829e-05, "loss": 1.8146, "step": 2052 }, { "epoch": 0.018464471855845807, "grad_norm": 2.2803170680999756, "learning_rate": 4.6774931864032787e-05, "loss": 1.824, "step": 2053 }, { "epoch": 0.01847346575348626, "grad_norm": 1.8729716539382935, "learning_rate": 4.6770869025343475e-05, "loss": 1.9045, "step": 2054 }, { "epoch": 0.01848245965112671, "grad_norm": 1.2087323665618896, "learning_rate": 4.6766803805824655e-05, "loss": 1.7672, "step": 2055 }, { "epoch": 0.01849145354876716, "grad_norm": 1.3885104656219482, "learning_rate": 4.6762736205920896e-05, "loss": 1.7731, "step": 2056 }, { "epoch": 0.018500447446407612, "grad_norm": 1.3012938499450684, "learning_rate": 4.6758666226077034e-05, "loss": 1.7308, "step": 2057 }, { "epoch": 0.018509441344048062, "grad_norm": 1.4353488683700562, "learning_rate": 4.675459386673815e-05, "loss": 1.8682, "step": 2058 }, { "epoch": 0.018518435241688513, "grad_norm": 1.5151594877243042, "learning_rate": 4.675051912834958e-05, "loss": 1.8052, "step": 2059 }, { "epoch": 0.018527429139328967, "grad_norm": 1.8188140392303467, "learning_rate": 4.674644201135694e-05, "loss": 1.8525, "step": 2060 }, { "epoch": 0.018536423036969417, "grad_norm": 1.4011839628219604, "learning_rate": 4.6742362516206094e-05, "loss": 1.8314, "step": 2061 }, { "epoch": 0.018545416934609867, "grad_norm": 1.9763928651809692, "learning_rate": 4.673828064334318e-05, "loss": 1.7157, "step": 2062 }, { "epoch": 0.018554410832250318, "grad_norm": 1.162772297859192, "learning_rate": 4.673419639321457e-05, "loss": 1.8056, "step": 2063 }, { "epoch": 0.018563404729890768, "grad_norm": 2.332754373550415, "learning_rate": 4.673010976626692e-05, "loss": 1.6584, "step": 2064 }, { "epoch": 0.01857239862753122, "grad_norm": 2.508497476577759, "learning_rate": 4.672602076294714e-05, "loss": 1.7466, "step": 2065 }, { "epoch": 0.018581392525171672, "grad_norm": 1.1985446214675903, "learning_rate": 4.67219293837024e-05, "loss": 1.7216, "step": 2066 }, { "epoch": 0.018590386422812123, "grad_norm": 1.433607816696167, "learning_rate": 4.671783562898011e-05, "loss": 1.7058, "step": 2067 }, { "epoch": 0.018599380320452573, "grad_norm": 1.593929409980774, "learning_rate": 4.671373949922797e-05, "loss": 1.8382, "step": 2068 }, { "epoch": 0.018608374218093023, "grad_norm": 2.1381101608276367, "learning_rate": 4.670964099489393e-05, "loss": 1.7539, "step": 2069 }, { "epoch": 0.018617368115733474, "grad_norm": 1.2080533504486084, "learning_rate": 4.670554011642619e-05, "loss": 1.7411, "step": 2070 }, { "epoch": 0.018626362013373927, "grad_norm": 1.3847501277923584, "learning_rate": 4.670143686427322e-05, "loss": 1.7898, "step": 2071 }, { "epoch": 0.018635355911014378, "grad_norm": 3.3527796268463135, "learning_rate": 4.669733123888373e-05, "loss": 1.877, "step": 2072 }, { "epoch": 0.018644349808654828, "grad_norm": 1.000192403793335, "learning_rate": 4.669322324070673e-05, "loss": 1.7051, "step": 2073 }, { "epoch": 0.01865334370629528, "grad_norm": 2.0127580165863037, "learning_rate": 4.6689112870191444e-05, "loss": 1.9304, "step": 2074 }, { "epoch": 0.01866233760393573, "grad_norm": 2.304248094558716, "learning_rate": 4.668500012778738e-05, "loss": 1.8543, "step": 2075 }, { "epoch": 0.01867133150157618, "grad_norm": 3.337897300720215, "learning_rate": 4.668088501394431e-05, "loss": 1.705, "step": 2076 }, { "epoch": 0.018680325399216633, "grad_norm": 2.1452367305755615, "learning_rate": 4.667676752911225e-05, "loss": 1.8019, "step": 2077 }, { "epoch": 0.018689319296857083, "grad_norm": 1.8496088981628418, "learning_rate": 4.6672647673741485e-05, "loss": 1.8255, "step": 2078 }, { "epoch": 0.018698313194497534, "grad_norm": 3.3575830459594727, "learning_rate": 4.666852544828256e-05, "loss": 1.846, "step": 2079 }, { "epoch": 0.018707307092137984, "grad_norm": 1.9009802341461182, "learning_rate": 4.666440085318626e-05, "loss": 1.7955, "step": 2080 }, { "epoch": 0.018716300989778435, "grad_norm": 1.4692115783691406, "learning_rate": 4.666027388890366e-05, "loss": 1.8233, "step": 2081 }, { "epoch": 0.018725294887418885, "grad_norm": 1.9335922002792358, "learning_rate": 4.6656144555886076e-05, "loss": 1.7872, "step": 2082 }, { "epoch": 0.01873428878505934, "grad_norm": 1.5201219320297241, "learning_rate": 4.665201285458508e-05, "loss": 1.7145, "step": 2083 }, { "epoch": 0.01874328268269979, "grad_norm": 1.7875475883483887, "learning_rate": 4.664787878545252e-05, "loss": 1.771, "step": 2084 }, { "epoch": 0.01875227658034024, "grad_norm": 2.3862104415893555, "learning_rate": 4.6643742348940464e-05, "loss": 1.7994, "step": 2085 }, { "epoch": 0.01876127047798069, "grad_norm": 0.9051443934440613, "learning_rate": 4.6639603545501295e-05, "loss": 1.8733, "step": 2086 }, { "epoch": 0.01877026437562114, "grad_norm": 1.100472092628479, "learning_rate": 4.663546237558762e-05, "loss": 1.761, "step": 2087 }, { "epoch": 0.01877925827326159, "grad_norm": 1.9001047611236572, "learning_rate": 4.663131883965231e-05, "loss": 1.7345, "step": 2088 }, { "epoch": 0.018788252170902044, "grad_norm": 2.2512001991271973, "learning_rate": 4.662717293814849e-05, "loss": 1.7768, "step": 2089 }, { "epoch": 0.018797246068542495, "grad_norm": 1.644371747970581, "learning_rate": 4.662302467152955e-05, "loss": 1.8015, "step": 2090 }, { "epoch": 0.018806239966182945, "grad_norm": 2.44160532951355, "learning_rate": 4.661887404024915e-05, "loss": 1.8646, "step": 2091 }, { "epoch": 0.018815233863823395, "grad_norm": 1.4504317045211792, "learning_rate": 4.6614721044761175e-05, "loss": 1.8047, "step": 2092 }, { "epoch": 0.018824227761463846, "grad_norm": 1.0868542194366455, "learning_rate": 4.661056568551982e-05, "loss": 1.756, "step": 2093 }, { "epoch": 0.018833221659104296, "grad_norm": 4.467520713806152, "learning_rate": 4.660640796297948e-05, "loss": 1.6769, "step": 2094 }, { "epoch": 0.01884221555674475, "grad_norm": 1.3139735460281372, "learning_rate": 4.660224787759486e-05, "loss": 1.7826, "step": 2095 }, { "epoch": 0.0188512094543852, "grad_norm": 2.653592824935913, "learning_rate": 4.659808542982088e-05, "loss": 1.7458, "step": 2096 }, { "epoch": 0.01886020335202565, "grad_norm": 1.3718576431274414, "learning_rate": 4.659392062011276e-05, "loss": 1.7285, "step": 2097 }, { "epoch": 0.0188691972496661, "grad_norm": 1.5852669477462769, "learning_rate": 4.658975344892594e-05, "loss": 1.7289, "step": 2098 }, { "epoch": 0.01887819114730655, "grad_norm": 1.94153892993927, "learning_rate": 4.658558391671615e-05, "loss": 1.6902, "step": 2099 }, { "epoch": 0.018887185044947002, "grad_norm": 2.8306474685668945, "learning_rate": 4.6581412023939354e-05, "loss": 1.8901, "step": 2100 }, { "epoch": 0.018896178942587456, "grad_norm": 1.9152741432189941, "learning_rate": 4.657723777105178e-05, "loss": 1.8258, "step": 2101 }, { "epoch": 0.018905172840227906, "grad_norm": 3.2933478355407715, "learning_rate": 4.657306115850992e-05, "loss": 1.8608, "step": 2102 }, { "epoch": 0.018914166737868356, "grad_norm": 1.7155070304870605, "learning_rate": 4.656888218677054e-05, "loss": 1.6688, "step": 2103 }, { "epoch": 0.018923160635508807, "grad_norm": 3.989287853240967, "learning_rate": 4.656470085629062e-05, "loss": 1.8399, "step": 2104 }, { "epoch": 0.018932154533149257, "grad_norm": 1.8837226629257202, "learning_rate": 4.656051716752745e-05, "loss": 1.7293, "step": 2105 }, { "epoch": 0.01894114843078971, "grad_norm": 2.8621439933776855, "learning_rate": 4.655633112093853e-05, "loss": 1.851, "step": 2106 }, { "epoch": 0.01895014232843016, "grad_norm": 2.354987144470215, "learning_rate": 4.655214271698165e-05, "loss": 1.7079, "step": 2107 }, { "epoch": 0.01895913622607061, "grad_norm": 2.054619073867798, "learning_rate": 4.654795195611484e-05, "loss": 1.806, "step": 2108 }, { "epoch": 0.018968130123711062, "grad_norm": 2.372976541519165, "learning_rate": 4.654375883879641e-05, "loss": 1.8489, "step": 2109 }, { "epoch": 0.018977124021351512, "grad_norm": 3.729292154312134, "learning_rate": 4.65395633654849e-05, "loss": 1.6316, "step": 2110 }, { "epoch": 0.018986117918991963, "grad_norm": 1.13978111743927, "learning_rate": 4.653536553663912e-05, "loss": 1.7783, "step": 2111 }, { "epoch": 0.018995111816632416, "grad_norm": 2.03242564201355, "learning_rate": 4.653116535271816e-05, "loss": 1.661, "step": 2112 }, { "epoch": 0.019004105714272867, "grad_norm": 1.55001962184906, "learning_rate": 4.652696281418131e-05, "loss": 1.8081, "step": 2113 }, { "epoch": 0.019013099611913317, "grad_norm": 0.9134846925735474, "learning_rate": 4.652275792148817e-05, "loss": 1.7937, "step": 2114 }, { "epoch": 0.019022093509553768, "grad_norm": 1.2741193771362305, "learning_rate": 4.65185506750986e-05, "loss": 1.7975, "step": 2115 }, { "epoch": 0.019031087407194218, "grad_norm": 1.6057394742965698, "learning_rate": 4.651434107547267e-05, "loss": 1.844, "step": 2116 }, { "epoch": 0.01904008130483467, "grad_norm": 2.311929225921631, "learning_rate": 4.651012912307074e-05, "loss": 1.8192, "step": 2117 }, { "epoch": 0.019049075202475122, "grad_norm": 1.6232632398605347, "learning_rate": 4.650591481835344e-05, "loss": 1.7523, "step": 2118 }, { "epoch": 0.019058069100115572, "grad_norm": 1.6961398124694824, "learning_rate": 4.650169816178162e-05, "loss": 1.7298, "step": 2119 }, { "epoch": 0.019067062997756023, "grad_norm": 2.1314873695373535, "learning_rate": 4.649747915381643e-05, "loss": 1.8235, "step": 2120 }, { "epoch": 0.019076056895396473, "grad_norm": 1.762578010559082, "learning_rate": 4.649325779491923e-05, "loss": 1.7285, "step": 2121 }, { "epoch": 0.019085050793036924, "grad_norm": 2.0781078338623047, "learning_rate": 4.648903408555168e-05, "loss": 1.6974, "step": 2122 }, { "epoch": 0.019094044690677374, "grad_norm": 2.644218683242798, "learning_rate": 4.6484808026175664e-05, "loss": 1.7453, "step": 2123 }, { "epoch": 0.019103038588317828, "grad_norm": 2.133168935775757, "learning_rate": 4.648057961725334e-05, "loss": 1.7943, "step": 2124 }, { "epoch": 0.019112032485958278, "grad_norm": 1.3693124055862427, "learning_rate": 4.6476348859247134e-05, "loss": 1.7382, "step": 2125 }, { "epoch": 0.01912102638359873, "grad_norm": 1.9745917320251465, "learning_rate": 4.64721157526197e-05, "loss": 1.7138, "step": 2126 }, { "epoch": 0.01913002028123918, "grad_norm": 1.3308643102645874, "learning_rate": 4.646788029783398e-05, "loss": 1.8508, "step": 2127 }, { "epoch": 0.01913901417887963, "grad_norm": 1.6118205785751343, "learning_rate": 4.646364249535313e-05, "loss": 1.8088, "step": 2128 }, { "epoch": 0.01914800807652008, "grad_norm": 1.6001852750778198, "learning_rate": 4.645940234564062e-05, "loss": 1.7207, "step": 2129 }, { "epoch": 0.019157001974160533, "grad_norm": 2.048927068710327, "learning_rate": 4.645515984916013e-05, "loss": 1.7913, "step": 2130 }, { "epoch": 0.019165995871800984, "grad_norm": 3.120556354522705, "learning_rate": 4.645091500637562e-05, "loss": 1.8303, "step": 2131 }, { "epoch": 0.019174989769441434, "grad_norm": 2.787069797515869, "learning_rate": 4.644666781775129e-05, "loss": 1.615, "step": 2132 }, { "epoch": 0.019183983667081884, "grad_norm": 1.310383915901184, "learning_rate": 4.644241828375161e-05, "loss": 1.8235, "step": 2133 }, { "epoch": 0.019192977564722335, "grad_norm": 1.5933363437652588, "learning_rate": 4.643816640484131e-05, "loss": 1.8643, "step": 2134 }, { "epoch": 0.019201971462362785, "grad_norm": 1.9600659608840942, "learning_rate": 4.643391218148536e-05, "loss": 1.7306, "step": 2135 }, { "epoch": 0.01921096536000324, "grad_norm": 1.1796625852584839, "learning_rate": 4.6429655614149e-05, "loss": 1.8083, "step": 2136 }, { "epoch": 0.01921995925764369, "grad_norm": 1.1321755647659302, "learning_rate": 4.6425396703297724e-05, "loss": 1.8065, "step": 2137 }, { "epoch": 0.01922895315528414, "grad_norm": 1.5106537342071533, "learning_rate": 4.642113544939728e-05, "loss": 1.7347, "step": 2138 }, { "epoch": 0.01923794705292459, "grad_norm": 1.7763785123825073, "learning_rate": 4.641687185291366e-05, "loss": 1.7587, "step": 2139 }, { "epoch": 0.01924694095056504, "grad_norm": 1.9857326745986938, "learning_rate": 4.6412605914313144e-05, "loss": 1.72, "step": 2140 }, { "epoch": 0.019255934848205494, "grad_norm": 2.5519232749938965, "learning_rate": 4.640833763406223e-05, "loss": 1.7261, "step": 2141 }, { "epoch": 0.019264928745845945, "grad_norm": 1.7469114065170288, "learning_rate": 4.6404067012627704e-05, "loss": 1.5952, "step": 2142 }, { "epoch": 0.019273922643486395, "grad_norm": 3.1920197010040283, "learning_rate": 4.639979405047658e-05, "loss": 1.6747, "step": 2143 }, { "epoch": 0.019282916541126845, "grad_norm": 1.3252753019332886, "learning_rate": 4.639551874807617e-05, "loss": 1.7291, "step": 2144 }, { "epoch": 0.019291910438767296, "grad_norm": 1.7085964679718018, "learning_rate": 4.639124110589399e-05, "loss": 1.8421, "step": 2145 }, { "epoch": 0.019300904336407746, "grad_norm": 3.1263537406921387, "learning_rate": 4.638696112439784e-05, "loss": 1.7097, "step": 2146 }, { "epoch": 0.0193098982340482, "grad_norm": 2.0735437870025635, "learning_rate": 4.638267880405578e-05, "loss": 1.7734, "step": 2147 }, { "epoch": 0.01931889213168865, "grad_norm": 1.6814250946044922, "learning_rate": 4.6378394145336115e-05, "loss": 1.8483, "step": 2148 }, { "epoch": 0.0193278860293291, "grad_norm": 1.7409281730651855, "learning_rate": 4.637410714870741e-05, "loss": 1.7322, "step": 2149 }, { "epoch": 0.01933687992696955, "grad_norm": 1.470043659210205, "learning_rate": 4.6369817814638475e-05, "loss": 1.7402, "step": 2150 }, { "epoch": 0.01934587382461, "grad_norm": 1.6099941730499268, "learning_rate": 4.63655261435984e-05, "loss": 1.7141, "step": 2151 }, { "epoch": 0.01935486772225045, "grad_norm": 3.2834770679473877, "learning_rate": 4.636123213605651e-05, "loss": 1.6715, "step": 2152 }, { "epoch": 0.019363861619890906, "grad_norm": 1.719178318977356, "learning_rate": 4.635693579248238e-05, "loss": 1.8176, "step": 2153 }, { "epoch": 0.019372855517531356, "grad_norm": 1.6343547105789185, "learning_rate": 4.6352637113345865e-05, "loss": 1.802, "step": 2154 }, { "epoch": 0.019381849415171806, "grad_norm": 1.3485571146011353, "learning_rate": 4.634833609911706e-05, "loss": 1.8078, "step": 2155 }, { "epoch": 0.019390843312812257, "grad_norm": 2.1289753913879395, "learning_rate": 4.634403275026632e-05, "loss": 1.7684, "step": 2156 }, { "epoch": 0.019399837210452707, "grad_norm": 1.8365706205368042, "learning_rate": 4.633972706726425e-05, "loss": 1.8043, "step": 2157 }, { "epoch": 0.019408831108093157, "grad_norm": 4.498271465301514, "learning_rate": 4.63354190505817e-05, "loss": 1.6802, "step": 2158 }, { "epoch": 0.01941782500573361, "grad_norm": 3.393364667892456, "learning_rate": 4.633110870068981e-05, "loss": 1.5558, "step": 2159 }, { "epoch": 0.01942681890337406, "grad_norm": 3.582029342651367, "learning_rate": 4.6326796018059936e-05, "loss": 1.7837, "step": 2160 }, { "epoch": 0.019435812801014512, "grad_norm": 2.204375982284546, "learning_rate": 4.632248100316372e-05, "loss": 1.7843, "step": 2161 }, { "epoch": 0.019444806698654962, "grad_norm": 2.106457233428955, "learning_rate": 4.6318163656473043e-05, "loss": 1.7559, "step": 2162 }, { "epoch": 0.019453800596295413, "grad_norm": 2.1652588844299316, "learning_rate": 4.631384397846003e-05, "loss": 1.7909, "step": 2163 }, { "epoch": 0.019462794493935863, "grad_norm": 2.4800820350646973, "learning_rate": 4.630952196959709e-05, "loss": 1.7071, "step": 2164 }, { "epoch": 0.019471788391576317, "grad_norm": 1.2245463132858276, "learning_rate": 4.630519763035687e-05, "loss": 1.8438, "step": 2165 }, { "epoch": 0.019480782289216767, "grad_norm": 1.1622307300567627, "learning_rate": 4.6300870961212264e-05, "loss": 1.7444, "step": 2166 }, { "epoch": 0.019489776186857218, "grad_norm": 0.946546196937561, "learning_rate": 4.629654196263643e-05, "loss": 1.749, "step": 2167 }, { "epoch": 0.019498770084497668, "grad_norm": 1.9194576740264893, "learning_rate": 4.629221063510279e-05, "loss": 1.7895, "step": 2168 }, { "epoch": 0.019507763982138118, "grad_norm": 2.0385642051696777, "learning_rate": 4.6287876979085006e-05, "loss": 1.7408, "step": 2169 }, { "epoch": 0.01951675787977857, "grad_norm": 0.9523259401321411, "learning_rate": 4.6283540995057004e-05, "loss": 1.7747, "step": 2170 }, { "epoch": 0.019525751777419022, "grad_norm": 1.4348504543304443, "learning_rate": 4.627920268349296e-05, "loss": 1.6719, "step": 2171 }, { "epoch": 0.019534745675059473, "grad_norm": 2.7496678829193115, "learning_rate": 4.6274862044867304e-05, "loss": 1.6676, "step": 2172 }, { "epoch": 0.019543739572699923, "grad_norm": 3.4268720149993896, "learning_rate": 4.6270519079654716e-05, "loss": 1.6862, "step": 2173 }, { "epoch": 0.019552733470340374, "grad_norm": 3.2327818870544434, "learning_rate": 4.6266173788330136e-05, "loss": 1.7669, "step": 2174 }, { "epoch": 0.019561727367980824, "grad_norm": 3.3095669746398926, "learning_rate": 4.6261826171368774e-05, "loss": 1.8976, "step": 2175 }, { "epoch": 0.019570721265621274, "grad_norm": 2.899407386779785, "learning_rate": 4.625747622924607e-05, "loss": 1.6485, "step": 2176 }, { "epoch": 0.019579715163261728, "grad_norm": 1.4627596139907837, "learning_rate": 4.625312396243772e-05, "loss": 1.7341, "step": 2177 }, { "epoch": 0.01958870906090218, "grad_norm": 1.1610193252563477, "learning_rate": 4.6248769371419684e-05, "loss": 1.7549, "step": 2178 }, { "epoch": 0.01959770295854263, "grad_norm": 2.516101837158203, "learning_rate": 4.6244412456668185e-05, "loss": 1.7229, "step": 2179 }, { "epoch": 0.01960669685618308, "grad_norm": 1.627119779586792, "learning_rate": 4.6240053218659674e-05, "loss": 1.7607, "step": 2180 }, { "epoch": 0.01961569075382353, "grad_norm": 0.9500035643577576, "learning_rate": 4.623569165787088e-05, "loss": 1.7998, "step": 2181 }, { "epoch": 0.019624684651463983, "grad_norm": 1.4506654739379883, "learning_rate": 4.6231327774778766e-05, "loss": 1.7547, "step": 2182 }, { "epoch": 0.019633678549104434, "grad_norm": 3.65898060798645, "learning_rate": 4.6226961569860575e-05, "loss": 1.99, "step": 2183 }, { "epoch": 0.019642672446744884, "grad_norm": 1.5578161478042603, "learning_rate": 4.622259304359378e-05, "loss": 1.8216, "step": 2184 }, { "epoch": 0.019651666344385334, "grad_norm": 2.4692704677581787, "learning_rate": 4.621822219645612e-05, "loss": 1.8007, "step": 2185 }, { "epoch": 0.019660660242025785, "grad_norm": 2.8048062324523926, "learning_rate": 4.6213849028925574e-05, "loss": 1.6771, "step": 2186 }, { "epoch": 0.019669654139666235, "grad_norm": 1.6007746458053589, "learning_rate": 4.62094735414804e-05, "loss": 1.6454, "step": 2187 }, { "epoch": 0.01967864803730669, "grad_norm": 1.986706256866455, "learning_rate": 4.620509573459909e-05, "loss": 1.7518, "step": 2188 }, { "epoch": 0.01968764193494714, "grad_norm": 1.2139371633529663, "learning_rate": 4.620071560876038e-05, "loss": 1.8063, "step": 2189 }, { "epoch": 0.01969663583258759, "grad_norm": 1.89432954788208, "learning_rate": 4.6196333164443295e-05, "loss": 1.7952, "step": 2190 }, { "epoch": 0.01970562973022804, "grad_norm": 1.839756965637207, "learning_rate": 4.619194840212708e-05, "loss": 1.728, "step": 2191 }, { "epoch": 0.01971462362786849, "grad_norm": 1.0582174062728882, "learning_rate": 4.618756132229125e-05, "loss": 1.7866, "step": 2192 }, { "epoch": 0.01972361752550894, "grad_norm": 2.040465831756592, "learning_rate": 4.618317192541557e-05, "loss": 1.7161, "step": 2193 }, { "epoch": 0.019732611423149395, "grad_norm": 2.4470021724700928, "learning_rate": 4.617878021198006e-05, "loss": 1.7238, "step": 2194 }, { "epoch": 0.019741605320789845, "grad_norm": 1.313166856765747, "learning_rate": 4.617438618246498e-05, "loss": 1.8459, "step": 2195 }, { "epoch": 0.019750599218430295, "grad_norm": 1.6557954549789429, "learning_rate": 4.616998983735087e-05, "loss": 1.8949, "step": 2196 }, { "epoch": 0.019759593116070746, "grad_norm": 0.997327446937561, "learning_rate": 4.616559117711849e-05, "loss": 1.7553, "step": 2197 }, { "epoch": 0.019768587013711196, "grad_norm": 1.4596160650253296, "learning_rate": 4.6161190202248897e-05, "loss": 1.6451, "step": 2198 }, { "epoch": 0.019777580911351646, "grad_norm": 4.417002201080322, "learning_rate": 4.615678691322335e-05, "loss": 1.8097, "step": 2199 }, { "epoch": 0.0197865748089921, "grad_norm": 2.3208796977996826, "learning_rate": 4.6152381310523387e-05, "loss": 1.7553, "step": 2200 }, { "epoch": 0.01979556870663255, "grad_norm": 2.070594072341919, "learning_rate": 4.614797339463082e-05, "loss": 1.8924, "step": 2201 }, { "epoch": 0.019804562604273, "grad_norm": 2.2145535945892334, "learning_rate": 4.614356316602766e-05, "loss": 1.6299, "step": 2202 }, { "epoch": 0.01981355650191345, "grad_norm": 0.8703266978263855, "learning_rate": 4.6139150625196235e-05, "loss": 1.6354, "step": 2203 }, { "epoch": 0.0198225503995539, "grad_norm": 1.852177619934082, "learning_rate": 4.613473577261908e-05, "loss": 1.7249, "step": 2204 }, { "epoch": 0.019831544297194352, "grad_norm": 1.9959479570388794, "learning_rate": 4.6130318608778995e-05, "loss": 1.764, "step": 2205 }, { "epoch": 0.019840538194834806, "grad_norm": 1.1844133138656616, "learning_rate": 4.612589913415904e-05, "loss": 1.7711, "step": 2206 }, { "epoch": 0.019849532092475256, "grad_norm": 2.788403034210205, "learning_rate": 4.6121477349242505e-05, "loss": 1.819, "step": 2207 }, { "epoch": 0.019858525990115707, "grad_norm": 1.5281730890274048, "learning_rate": 4.611705325451297e-05, "loss": 1.7389, "step": 2208 }, { "epoch": 0.019867519887756157, "grad_norm": 2.2549779415130615, "learning_rate": 4.611262685045424e-05, "loss": 1.662, "step": 2209 }, { "epoch": 0.019876513785396607, "grad_norm": 1.302918791770935, "learning_rate": 4.610819813755038e-05, "loss": 1.7962, "step": 2210 }, { "epoch": 0.019885507683037058, "grad_norm": 1.6850008964538574, "learning_rate": 4.6103767116285705e-05, "loss": 1.7465, "step": 2211 }, { "epoch": 0.01989450158067751, "grad_norm": 1.9856219291687012, "learning_rate": 4.6099333787144795e-05, "loss": 1.9358, "step": 2212 }, { "epoch": 0.019903495478317962, "grad_norm": 2.6390187740325928, "learning_rate": 4.609489815061245e-05, "loss": 1.8762, "step": 2213 }, { "epoch": 0.019912489375958412, "grad_norm": 2.3462400436401367, "learning_rate": 4.6090460207173775e-05, "loss": 1.7425, "step": 2214 }, { "epoch": 0.019921483273598863, "grad_norm": 2.059737205505371, "learning_rate": 4.608601995731407e-05, "loss": 1.8411, "step": 2215 }, { "epoch": 0.019930477171239313, "grad_norm": 1.4516576528549194, "learning_rate": 4.608157740151893e-05, "loss": 1.6764, "step": 2216 }, { "epoch": 0.019939471068879767, "grad_norm": 2.2649362087249756, "learning_rate": 4.607713254027418e-05, "loss": 1.7872, "step": 2217 }, { "epoch": 0.019948464966520217, "grad_norm": 2.1351478099823, "learning_rate": 4.607268537406591e-05, "loss": 1.7239, "step": 2218 }, { "epoch": 0.019957458864160667, "grad_norm": 3.15726637840271, "learning_rate": 4.606823590338044e-05, "loss": 1.8314, "step": 2219 }, { "epoch": 0.019966452761801118, "grad_norm": 4.243987083435059, "learning_rate": 4.6063784128704367e-05, "loss": 1.812, "step": 2220 }, { "epoch": 0.019975446659441568, "grad_norm": 1.5010563135147095, "learning_rate": 4.605933005052454e-05, "loss": 1.6968, "step": 2221 }, { "epoch": 0.01998444055708202, "grad_norm": 2.5509402751922607, "learning_rate": 4.6054873669328035e-05, "loss": 1.7451, "step": 2222 }, { "epoch": 0.019993434454722472, "grad_norm": 1.6732513904571533, "learning_rate": 4.60504149856022e-05, "loss": 1.7412, "step": 2223 }, { "epoch": 0.020002428352362923, "grad_norm": 1.4404628276824951, "learning_rate": 4.604595399983463e-05, "loss": 1.7258, "step": 2224 }, { "epoch": 0.020011422250003373, "grad_norm": 1.3979803323745728, "learning_rate": 4.604149071251318e-05, "loss": 1.714, "step": 2225 }, { "epoch": 0.020020416147643823, "grad_norm": 1.3739867210388184, "learning_rate": 4.603702512412593e-05, "loss": 1.7559, "step": 2226 }, { "epoch": 0.020029410045284274, "grad_norm": 1.630112648010254, "learning_rate": 4.6032557235161255e-05, "loss": 1.7324, "step": 2227 }, { "epoch": 0.020038403942924724, "grad_norm": 1.3827682733535767, "learning_rate": 4.602808704610773e-05, "loss": 1.7898, "step": 2228 }, { "epoch": 0.020047397840565178, "grad_norm": 1.1498510837554932, "learning_rate": 4.602361455745423e-05, "loss": 1.7546, "step": 2229 }, { "epoch": 0.02005639173820563, "grad_norm": 2.1381092071533203, "learning_rate": 4.601913976968985e-05, "loss": 1.809, "step": 2230 }, { "epoch": 0.02006538563584608, "grad_norm": 3.69057035446167, "learning_rate": 4.6014662683303946e-05, "loss": 1.7882, "step": 2231 }, { "epoch": 0.02007437953348653, "grad_norm": 1.9872729778289795, "learning_rate": 4.601018329878611e-05, "loss": 1.6813, "step": 2232 }, { "epoch": 0.02008337343112698, "grad_norm": 1.8036212921142578, "learning_rate": 4.600570161662624e-05, "loss": 1.794, "step": 2233 }, { "epoch": 0.02009236732876743, "grad_norm": 1.6398576498031616, "learning_rate": 4.600121763731441e-05, "loss": 1.6571, "step": 2234 }, { "epoch": 0.020101361226407884, "grad_norm": 2.9632186889648438, "learning_rate": 4.5996731361340994e-05, "loss": 1.9164, "step": 2235 }, { "epoch": 0.020110355124048334, "grad_norm": 3.310476779937744, "learning_rate": 4.599224278919661e-05, "loss": 1.7466, "step": 2236 }, { "epoch": 0.020119349021688784, "grad_norm": 1.895015835762024, "learning_rate": 4.5987751921372116e-05, "loss": 1.7677, "step": 2237 }, { "epoch": 0.020128342919329235, "grad_norm": 1.7821534872055054, "learning_rate": 4.598325875835862e-05, "loss": 1.727, "step": 2238 }, { "epoch": 0.020137336816969685, "grad_norm": 2.46008563041687, "learning_rate": 4.59787633006475e-05, "loss": 1.8615, "step": 2239 }, { "epoch": 0.020146330714610135, "grad_norm": 1.7851299047470093, "learning_rate": 4.597426554873037e-05, "loss": 1.7429, "step": 2240 }, { "epoch": 0.02015532461225059, "grad_norm": 1.4020650386810303, "learning_rate": 4.596976550309908e-05, "loss": 1.8065, "step": 2241 }, { "epoch": 0.02016431850989104, "grad_norm": 1.3458341360092163, "learning_rate": 4.596526316424578e-05, "loss": 1.7868, "step": 2242 }, { "epoch": 0.02017331240753149, "grad_norm": 2.1337568759918213, "learning_rate": 4.596075853266281e-05, "loss": 1.7515, "step": 2243 }, { "epoch": 0.02018230630517194, "grad_norm": 1.321604609489441, "learning_rate": 4.59562516088428e-05, "loss": 1.6832, "step": 2244 }, { "epoch": 0.02019130020281239, "grad_norm": 1.237625002861023, "learning_rate": 4.595174239327862e-05, "loss": 1.8606, "step": 2245 }, { "epoch": 0.02020029410045284, "grad_norm": 2.3070228099823, "learning_rate": 4.5947230886463396e-05, "loss": 1.7989, "step": 2246 }, { "epoch": 0.020209287998093295, "grad_norm": 1.9083362817764282, "learning_rate": 4.59427170888905e-05, "loss": 1.6386, "step": 2247 }, { "epoch": 0.020218281895733745, "grad_norm": 2.1421046257019043, "learning_rate": 4.593820100105355e-05, "loss": 1.7779, "step": 2248 }, { "epoch": 0.020227275793374196, "grad_norm": 1.4160685539245605, "learning_rate": 4.5933682623446405e-05, "loss": 1.7794, "step": 2249 }, { "epoch": 0.020236269691014646, "grad_norm": 1.6969027519226074, "learning_rate": 4.592916195656322e-05, "loss": 1.8446, "step": 2250 }, { "epoch": 0.020245263588655096, "grad_norm": 1.2178318500518799, "learning_rate": 4.592463900089834e-05, "loss": 1.6796, "step": 2251 }, { "epoch": 0.02025425748629555, "grad_norm": 1.339027762413025, "learning_rate": 4.5920113756946396e-05, "loss": 1.7301, "step": 2252 }, { "epoch": 0.020263251383936, "grad_norm": 3.8408193588256836, "learning_rate": 4.5915586225202276e-05, "loss": 1.8731, "step": 2253 }, { "epoch": 0.02027224528157645, "grad_norm": 1.3739826679229736, "learning_rate": 4.591105640616108e-05, "loss": 1.7674, "step": 2254 }, { "epoch": 0.0202812391792169, "grad_norm": 1.8437106609344482, "learning_rate": 4.59065243003182e-05, "loss": 1.7974, "step": 2255 }, { "epoch": 0.02029023307685735, "grad_norm": 1.075078010559082, "learning_rate": 4.5901989908169256e-05, "loss": 1.7229, "step": 2256 }, { "epoch": 0.020299226974497802, "grad_norm": 2.184420108795166, "learning_rate": 4.589745323021012e-05, "loss": 1.7521, "step": 2257 }, { "epoch": 0.020308220872138256, "grad_norm": 0.967837929725647, "learning_rate": 4.589291426693692e-05, "loss": 1.6798, "step": 2258 }, { "epoch": 0.020317214769778706, "grad_norm": 0.9452171325683594, "learning_rate": 4.588837301884602e-05, "loss": 1.8082, "step": 2259 }, { "epoch": 0.020326208667419157, "grad_norm": 2.0494179725646973, "learning_rate": 4.588382948643406e-05, "loss": 1.7313, "step": 2260 }, { "epoch": 0.020335202565059607, "grad_norm": 4.940488815307617, "learning_rate": 4.5879283670197906e-05, "loss": 1.9332, "step": 2261 }, { "epoch": 0.020344196462700057, "grad_norm": 1.5312249660491943, "learning_rate": 4.587473557063468e-05, "loss": 1.8594, "step": 2262 }, { "epoch": 0.020353190360340508, "grad_norm": 3.0739502906799316, "learning_rate": 4.5870185188241755e-05, "loss": 1.752, "step": 2263 }, { "epoch": 0.02036218425798096, "grad_norm": 1.1640230417251587, "learning_rate": 4.5865632523516754e-05, "loss": 1.9143, "step": 2264 }, { "epoch": 0.020371178155621412, "grad_norm": 1.1076139211654663, "learning_rate": 4.586107757695755e-05, "loss": 1.6885, "step": 2265 }, { "epoch": 0.020380172053261862, "grad_norm": 2.315984010696411, "learning_rate": 4.585652034906227e-05, "loss": 1.8189, "step": 2266 }, { "epoch": 0.020389165950902313, "grad_norm": 2.4155290126800537, "learning_rate": 4.585196084032928e-05, "loss": 1.6985, "step": 2267 }, { "epoch": 0.020398159848542763, "grad_norm": 2.6599762439727783, "learning_rate": 4.5847399051257204e-05, "loss": 1.7708, "step": 2268 }, { "epoch": 0.020407153746183213, "grad_norm": 2.6259031295776367, "learning_rate": 4.5842834982344916e-05, "loss": 1.7549, "step": 2269 }, { "epoch": 0.020416147643823667, "grad_norm": 1.4755679368972778, "learning_rate": 4.5838268634091524e-05, "loss": 1.74, "step": 2270 }, { "epoch": 0.020425141541464117, "grad_norm": 2.2552127838134766, "learning_rate": 4.5833700006996405e-05, "loss": 1.7179, "step": 2271 }, { "epoch": 0.020434135439104568, "grad_norm": 1.7168699502944946, "learning_rate": 4.582912910155918e-05, "loss": 1.7021, "step": 2272 }, { "epoch": 0.020443129336745018, "grad_norm": 1.9799747467041016, "learning_rate": 4.582455591827971e-05, "loss": 1.7648, "step": 2273 }, { "epoch": 0.02045212323438547, "grad_norm": 0.8853166699409485, "learning_rate": 4.581998045765812e-05, "loss": 1.8468, "step": 2274 }, { "epoch": 0.02046111713202592, "grad_norm": 2.712841749191284, "learning_rate": 4.581540272019476e-05, "loss": 1.7372, "step": 2275 }, { "epoch": 0.020470111029666373, "grad_norm": 1.6172630786895752, "learning_rate": 4.5810822706390254e-05, "loss": 1.6629, "step": 2276 }, { "epoch": 0.020479104927306823, "grad_norm": 1.6308058500289917, "learning_rate": 4.580624041674547e-05, "loss": 1.7234, "step": 2277 }, { "epoch": 0.020488098824947273, "grad_norm": 2.055882692337036, "learning_rate": 4.580165585176152e-05, "loss": 1.8009, "step": 2278 }, { "epoch": 0.020497092722587724, "grad_norm": 2.265346050262451, "learning_rate": 4.5797069011939754e-05, "loss": 1.7405, "step": 2279 }, { "epoch": 0.020506086620228174, "grad_norm": 1.9064559936523438, "learning_rate": 4.579247989778179e-05, "loss": 1.8192, "step": 2280 }, { "epoch": 0.020515080517868624, "grad_norm": 2.034872531890869, "learning_rate": 4.5787888509789486e-05, "loss": 1.7415, "step": 2281 }, { "epoch": 0.02052407441550908, "grad_norm": 2.2891037464141846, "learning_rate": 4.578329484846495e-05, "loss": 1.8283, "step": 2282 }, { "epoch": 0.02053306831314953, "grad_norm": 1.886886715888977, "learning_rate": 4.5778698914310535e-05, "loss": 1.7726, "step": 2283 }, { "epoch": 0.02054206221078998, "grad_norm": 1.4077640771865845, "learning_rate": 4.577410070782885e-05, "loss": 1.8119, "step": 2284 }, { "epoch": 0.02055105610843043, "grad_norm": 2.7739877700805664, "learning_rate": 4.576950022952274e-05, "loss": 1.6496, "step": 2285 }, { "epoch": 0.02056005000607088, "grad_norm": 3.223618268966675, "learning_rate": 4.5764897479895317e-05, "loss": 1.742, "step": 2286 }, { "epoch": 0.020569043903711334, "grad_norm": 3.3934900760650635, "learning_rate": 4.576029245944993e-05, "loss": 1.7863, "step": 2287 }, { "epoch": 0.020578037801351784, "grad_norm": 1.8958532810211182, "learning_rate": 4.575568516869016e-05, "loss": 1.7125, "step": 2288 }, { "epoch": 0.020587031698992234, "grad_norm": 2.368180990219116, "learning_rate": 4.575107560811987e-05, "loss": 1.7309, "step": 2289 }, { "epoch": 0.020596025596632685, "grad_norm": 1.763116478919983, "learning_rate": 4.574646377824315e-05, "loss": 1.8219, "step": 2290 }, { "epoch": 0.020605019494273135, "grad_norm": 1.6222342252731323, "learning_rate": 4.574184967956434e-05, "loss": 1.8357, "step": 2291 }, { "epoch": 0.020614013391913585, "grad_norm": 1.6685752868652344, "learning_rate": 4.573723331258804e-05, "loss": 1.6332, "step": 2292 }, { "epoch": 0.02062300728955404, "grad_norm": 2.48173451423645, "learning_rate": 4.573261467781908e-05, "loss": 1.7689, "step": 2293 }, { "epoch": 0.02063200118719449, "grad_norm": 3.7202978134155273, "learning_rate": 4.572799377576255e-05, "loss": 1.7349, "step": 2294 }, { "epoch": 0.02064099508483494, "grad_norm": 1.170435905456543, "learning_rate": 4.572337060692379e-05, "loss": 1.7168, "step": 2295 }, { "epoch": 0.02064998898247539, "grad_norm": 4.0131516456604, "learning_rate": 4.571874517180837e-05, "loss": 1.7891, "step": 2296 }, { "epoch": 0.02065898288011584, "grad_norm": 2.1273741722106934, "learning_rate": 4.571411747092213e-05, "loss": 1.7133, "step": 2297 }, { "epoch": 0.02066797677775629, "grad_norm": 2.4282209873199463, "learning_rate": 4.570948750477114e-05, "loss": 1.9092, "step": 2298 }, { "epoch": 0.020676970675396745, "grad_norm": 3.3014655113220215, "learning_rate": 4.5704855273861744e-05, "loss": 1.7369, "step": 2299 }, { "epoch": 0.020685964573037195, "grad_norm": 2.334832191467285, "learning_rate": 4.5700220778700504e-05, "loss": 1.6918, "step": 2300 }, { "epoch": 0.020694958470677646, "grad_norm": 2.1993372440338135, "learning_rate": 4.569558401979424e-05, "loss": 1.82, "step": 2301 }, { "epoch": 0.020703952368318096, "grad_norm": 2.0784170627593994, "learning_rate": 4.569094499765002e-05, "loss": 1.6171, "step": 2302 }, { "epoch": 0.020712946265958546, "grad_norm": 11.078575134277344, "learning_rate": 4.568630371277516e-05, "loss": 1.9182, "step": 2303 }, { "epoch": 0.020721940163598997, "grad_norm": 2.154306411743164, "learning_rate": 4.5681660165677236e-05, "loss": 1.6325, "step": 2304 }, { "epoch": 0.02073093406123945, "grad_norm": 3.486635208129883, "learning_rate": 4.567701435686404e-05, "loss": 1.7497, "step": 2305 }, { "epoch": 0.0207399279588799, "grad_norm": 1.8624457120895386, "learning_rate": 4.567236628684366e-05, "loss": 1.7857, "step": 2306 }, { "epoch": 0.02074892185652035, "grad_norm": 2.23759388923645, "learning_rate": 4.566771595612437e-05, "loss": 1.7292, "step": 2307 }, { "epoch": 0.0207579157541608, "grad_norm": 2.55198073387146, "learning_rate": 4.5663063365214734e-05, "loss": 1.8016, "step": 2308 }, { "epoch": 0.020766909651801252, "grad_norm": 2.8357598781585693, "learning_rate": 4.5658408514623566e-05, "loss": 1.7049, "step": 2309 }, { "epoch": 0.020775903549441702, "grad_norm": 3.649618625640869, "learning_rate": 4.565375140485989e-05, "loss": 1.7756, "step": 2310 }, { "epoch": 0.020784897447082156, "grad_norm": 1.487219214439392, "learning_rate": 4.5649092036433016e-05, "loss": 1.81, "step": 2311 }, { "epoch": 0.020793891344722606, "grad_norm": 1.9562350511550903, "learning_rate": 4.564443040985249e-05, "loss": 1.7709, "step": 2312 }, { "epoch": 0.020802885242363057, "grad_norm": 1.5540130138397217, "learning_rate": 4.563976652562809e-05, "loss": 1.7169, "step": 2313 }, { "epoch": 0.020811879140003507, "grad_norm": 1.5260024070739746, "learning_rate": 4.5635100384269845e-05, "loss": 1.7849, "step": 2314 }, { "epoch": 0.020820873037643958, "grad_norm": 1.1957591772079468, "learning_rate": 4.563043198628806e-05, "loss": 1.7006, "step": 2315 }, { "epoch": 0.020829866935284408, "grad_norm": 2.0326545238494873, "learning_rate": 4.562576133219324e-05, "loss": 1.7963, "step": 2316 }, { "epoch": 0.020838860832924862, "grad_norm": 2.109673500061035, "learning_rate": 4.5621088422496175e-05, "loss": 1.7448, "step": 2317 }, { "epoch": 0.020847854730565312, "grad_norm": 1.794297456741333, "learning_rate": 4.561641325770788e-05, "loss": 1.7763, "step": 2318 }, { "epoch": 0.020856848628205762, "grad_norm": 4.233527183532715, "learning_rate": 4.561173583833963e-05, "loss": 1.7554, "step": 2319 }, { "epoch": 0.020865842525846213, "grad_norm": 2.0925405025482178, "learning_rate": 4.560705616490294e-05, "loss": 1.824, "step": 2320 }, { "epoch": 0.020874836423486663, "grad_norm": 2.024864673614502, "learning_rate": 4.5602374237909564e-05, "loss": 1.7818, "step": 2321 }, { "epoch": 0.020883830321127114, "grad_norm": 3.2192177772521973, "learning_rate": 4.5597690057871525e-05, "loss": 1.7253, "step": 2322 }, { "epoch": 0.020892824218767567, "grad_norm": 1.9516295194625854, "learning_rate": 4.559300362530107e-05, "loss": 1.6573, "step": 2323 }, { "epoch": 0.020901818116408018, "grad_norm": 2.554602861404419, "learning_rate": 4.558831494071069e-05, "loss": 1.7531, "step": 2324 }, { "epoch": 0.020910812014048468, "grad_norm": 2.6560280323028564, "learning_rate": 4.5583624004613145e-05, "loss": 1.8745, "step": 2325 }, { "epoch": 0.02091980591168892, "grad_norm": 2.433551549911499, "learning_rate": 4.5578930817521425e-05, "loss": 1.7369, "step": 2326 }, { "epoch": 0.02092879980932937, "grad_norm": 2.8802425861358643, "learning_rate": 4.5574235379948776e-05, "loss": 1.6935, "step": 2327 }, { "epoch": 0.020937793706969823, "grad_norm": 1.8005021810531616, "learning_rate": 4.556953769240868e-05, "loss": 1.7882, "step": 2328 }, { "epoch": 0.020946787604610273, "grad_norm": 1.2202013731002808, "learning_rate": 4.5564837755414864e-05, "loss": 1.7742, "step": 2329 }, { "epoch": 0.020955781502250723, "grad_norm": 1.6693246364593506, "learning_rate": 4.556013556948131e-05, "loss": 1.6809, "step": 2330 }, { "epoch": 0.020964775399891174, "grad_norm": 3.8166472911834717, "learning_rate": 4.555543113512224e-05, "loss": 1.7995, "step": 2331 }, { "epoch": 0.020973769297531624, "grad_norm": 1.9440819025039673, "learning_rate": 4.555072445285213e-05, "loss": 1.7763, "step": 2332 }, { "epoch": 0.020982763195172074, "grad_norm": 4.746578216552734, "learning_rate": 4.5546015523185683e-05, "loss": 1.7582, "step": 2333 }, { "epoch": 0.02099175709281253, "grad_norm": 2.4273557662963867, "learning_rate": 4.554130434663788e-05, "loss": 1.7164, "step": 2334 }, { "epoch": 0.02100075099045298, "grad_norm": 1.877870798110962, "learning_rate": 4.5536590923723906e-05, "loss": 1.8398, "step": 2335 }, { "epoch": 0.02100974488809343, "grad_norm": 2.098482847213745, "learning_rate": 4.5531875254959236e-05, "loss": 1.7609, "step": 2336 }, { "epoch": 0.02101873878573388, "grad_norm": 1.4947023391723633, "learning_rate": 4.552715734085955e-05, "loss": 1.7259, "step": 2337 }, { "epoch": 0.02102773268337433, "grad_norm": 2.0660550594329834, "learning_rate": 4.5522437181940804e-05, "loss": 1.6259, "step": 2338 }, { "epoch": 0.02103672658101478, "grad_norm": 1.267037272453308, "learning_rate": 4.5517714778719175e-05, "loss": 1.7734, "step": 2339 }, { "epoch": 0.021045720478655234, "grad_norm": 2.9227545261383057, "learning_rate": 4.551299013171111e-05, "loss": 1.6779, "step": 2340 }, { "epoch": 0.021054714376295684, "grad_norm": 4.635622501373291, "learning_rate": 4.550826324143329e-05, "loss": 1.8014, "step": 2341 }, { "epoch": 0.021063708273936135, "grad_norm": 3.0436975955963135, "learning_rate": 4.5503534108402635e-05, "loss": 1.6908, "step": 2342 }, { "epoch": 0.021072702171576585, "grad_norm": 2.0508615970611572, "learning_rate": 4.549880273313631e-05, "loss": 1.7534, "step": 2343 }, { "epoch": 0.021081696069217035, "grad_norm": 1.5294864177703857, "learning_rate": 4.549406911615174e-05, "loss": 1.8054, "step": 2344 }, { "epoch": 0.021090689966857486, "grad_norm": 2.613116502761841, "learning_rate": 4.548933325796658e-05, "loss": 1.7134, "step": 2345 }, { "epoch": 0.02109968386449794, "grad_norm": 7.720178127288818, "learning_rate": 4.5484595159098755e-05, "loss": 1.7369, "step": 2346 }, { "epoch": 0.02110867776213839, "grad_norm": 2.882622003555298, "learning_rate": 4.5479854820066385e-05, "loss": 1.7149, "step": 2347 }, { "epoch": 0.02111767165977884, "grad_norm": 1.8212774991989136, "learning_rate": 4.547511224138789e-05, "loss": 1.8624, "step": 2348 }, { "epoch": 0.02112666555741929, "grad_norm": 3.6106045246124268, "learning_rate": 4.5470367423581906e-05, "loss": 1.6332, "step": 2349 }, { "epoch": 0.02113565945505974, "grad_norm": 1.7770973443984985, "learning_rate": 4.546562036716732e-05, "loss": 1.7435, "step": 2350 }, { "epoch": 0.02114465335270019, "grad_norm": 2.5307157039642334, "learning_rate": 4.546087107266325e-05, "loss": 1.7752, "step": 2351 }, { "epoch": 0.021153647250340645, "grad_norm": 1.4969408512115479, "learning_rate": 4.545611954058909e-05, "loss": 1.7298, "step": 2352 }, { "epoch": 0.021162641147981096, "grad_norm": 1.5504119396209717, "learning_rate": 4.545136577146444e-05, "loss": 1.7169, "step": 2353 }, { "epoch": 0.021171635045621546, "grad_norm": 1.8861709833145142, "learning_rate": 4.544660976580919e-05, "loss": 1.6307, "step": 2354 }, { "epoch": 0.021180628943261996, "grad_norm": 4.2253313064575195, "learning_rate": 4.544185152414343e-05, "loss": 1.7855, "step": 2355 }, { "epoch": 0.021189622840902447, "grad_norm": 2.0770530700683594, "learning_rate": 4.543709104698753e-05, "loss": 1.8734, "step": 2356 }, { "epoch": 0.021198616738542897, "grad_norm": 1.7201515436172485, "learning_rate": 4.5432328334862074e-05, "loss": 1.829, "step": 2357 }, { "epoch": 0.02120761063618335, "grad_norm": 5.565993785858154, "learning_rate": 4.542756338828791e-05, "loss": 1.7146, "step": 2358 }, { "epoch": 0.0212166045338238, "grad_norm": 1.38422429561615, "learning_rate": 4.542279620778613e-05, "loss": 1.7264, "step": 2359 }, { "epoch": 0.02122559843146425, "grad_norm": 1.4094674587249756, "learning_rate": 4.541802679387806e-05, "loss": 1.7275, "step": 2360 }, { "epoch": 0.021234592329104702, "grad_norm": 1.4018802642822266, "learning_rate": 4.541325514708528e-05, "loss": 1.7732, "step": 2361 }, { "epoch": 0.021243586226745152, "grad_norm": 2.3935818672180176, "learning_rate": 4.5408481267929605e-05, "loss": 1.6431, "step": 2362 }, { "epoch": 0.021252580124385606, "grad_norm": 2.3367104530334473, "learning_rate": 4.5403705156933095e-05, "loss": 1.8823, "step": 2363 }, { "epoch": 0.021261574022026056, "grad_norm": 1.294884204864502, "learning_rate": 4.539892681461808e-05, "loss": 1.6421, "step": 2364 }, { "epoch": 0.021270567919666507, "grad_norm": 1.848714828491211, "learning_rate": 4.539414624150708e-05, "loss": 1.8001, "step": 2365 }, { "epoch": 0.021279561817306957, "grad_norm": 1.689236044883728, "learning_rate": 4.538936343812293e-05, "loss": 1.791, "step": 2366 }, { "epoch": 0.021288555714947408, "grad_norm": 3.42376971244812, "learning_rate": 4.5384578404988637e-05, "loss": 1.6722, "step": 2367 }, { "epoch": 0.021297549612587858, "grad_norm": 2.636786937713623, "learning_rate": 4.5379791142627505e-05, "loss": 1.7355, "step": 2368 }, { "epoch": 0.02130654351022831, "grad_norm": 2.225850820541382, "learning_rate": 4.537500165156304e-05, "loss": 1.8291, "step": 2369 }, { "epoch": 0.021315537407868762, "grad_norm": 2.43796443939209, "learning_rate": 4.537020993231904e-05, "loss": 1.9183, "step": 2370 }, { "epoch": 0.021324531305509212, "grad_norm": 1.7118300199508667, "learning_rate": 4.536541598541951e-05, "loss": 1.7139, "step": 2371 }, { "epoch": 0.021333525203149663, "grad_norm": 1.9949419498443604, "learning_rate": 4.53606198113887e-05, "loss": 1.6814, "step": 2372 }, { "epoch": 0.021342519100790113, "grad_norm": 1.6315096616744995, "learning_rate": 4.535582141075112e-05, "loss": 1.8251, "step": 2373 }, { "epoch": 0.021351512998430563, "grad_norm": 2.5925848484039307, "learning_rate": 4.535102078403152e-05, "loss": 1.7035, "step": 2374 }, { "epoch": 0.021360506896071017, "grad_norm": 1.1303373575210571, "learning_rate": 4.534621793175487e-05, "loss": 1.7279, "step": 2375 }, { "epoch": 0.021369500793711468, "grad_norm": 2.230166435241699, "learning_rate": 4.534141285444643e-05, "loss": 1.7503, "step": 2376 }, { "epoch": 0.021378494691351918, "grad_norm": 2.021209955215454, "learning_rate": 4.533660555263166e-05, "loss": 1.6679, "step": 2377 }, { "epoch": 0.02138748858899237, "grad_norm": 1.3938347101211548, "learning_rate": 4.533179602683629e-05, "loss": 1.7162, "step": 2378 }, { "epoch": 0.02139648248663282, "grad_norm": 2.0459606647491455, "learning_rate": 4.532698427758627e-05, "loss": 1.8344, "step": 2379 }, { "epoch": 0.02140547638427327, "grad_norm": 2.4584720134735107, "learning_rate": 4.532217030540781e-05, "loss": 1.7216, "step": 2380 }, { "epoch": 0.021414470281913723, "grad_norm": 2.2946348190307617, "learning_rate": 4.531735411082735e-05, "loss": 1.7317, "step": 2381 }, { "epoch": 0.021423464179554173, "grad_norm": 3.456460475921631, "learning_rate": 4.53125356943716e-05, "loss": 1.7226, "step": 2382 }, { "epoch": 0.021432458077194624, "grad_norm": 2.5085268020629883, "learning_rate": 4.5307715056567484e-05, "loss": 1.7289, "step": 2383 }, { "epoch": 0.021441451974835074, "grad_norm": 2.5920472145080566, "learning_rate": 4.530289219794218e-05, "loss": 1.7974, "step": 2384 }, { "epoch": 0.021450445872475524, "grad_norm": 1.7468597888946533, "learning_rate": 4.5298067119023114e-05, "loss": 1.8321, "step": 2385 }, { "epoch": 0.021459439770115975, "grad_norm": 2.3181700706481934, "learning_rate": 4.529323982033795e-05, "loss": 1.6722, "step": 2386 }, { "epoch": 0.02146843366775643, "grad_norm": 1.4935438632965088, "learning_rate": 4.528841030241458e-05, "loss": 1.6694, "step": 2387 }, { "epoch": 0.02147742756539688, "grad_norm": 2.136178731918335, "learning_rate": 4.528357856578117e-05, "loss": 1.8219, "step": 2388 }, { "epoch": 0.02148642146303733, "grad_norm": 2.564387559890747, "learning_rate": 4.5278744610966096e-05, "loss": 1.7622, "step": 2389 }, { "epoch": 0.02149541536067778, "grad_norm": 1.8852046728134155, "learning_rate": 4.5273908438498e-05, "loss": 1.6342, "step": 2390 }, { "epoch": 0.02150440925831823, "grad_norm": 2.6920783519744873, "learning_rate": 4.5269070048905766e-05, "loss": 1.6087, "step": 2391 }, { "epoch": 0.02151340315595868, "grad_norm": 3.356430768966675, "learning_rate": 4.52642294427185e-05, "loss": 1.7844, "step": 2392 }, { "epoch": 0.021522397053599134, "grad_norm": 2.1716184616088867, "learning_rate": 4.525938662046557e-05, "loss": 1.7064, "step": 2393 }, { "epoch": 0.021531390951239585, "grad_norm": 1.9972597360610962, "learning_rate": 4.525454158267658e-05, "loss": 1.7464, "step": 2394 }, { "epoch": 0.021540384848880035, "grad_norm": 9.090784072875977, "learning_rate": 4.524969432988138e-05, "loss": 1.735, "step": 2395 }, { "epoch": 0.021549378746520485, "grad_norm": 2.599846363067627, "learning_rate": 4.524484486261005e-05, "loss": 1.674, "step": 2396 }, { "epoch": 0.021558372644160936, "grad_norm": 2.0945303440093994, "learning_rate": 4.5239993181392913e-05, "loss": 1.8419, "step": 2397 }, { "epoch": 0.02156736654180139, "grad_norm": 1.4711668491363525, "learning_rate": 4.523513928676056e-05, "loss": 1.7538, "step": 2398 }, { "epoch": 0.02157636043944184, "grad_norm": 1.8747934103012085, "learning_rate": 4.5230283179243806e-05, "loss": 1.6886, "step": 2399 }, { "epoch": 0.02158535433708229, "grad_norm": 2.8377668857574463, "learning_rate": 4.522542485937369e-05, "loss": 1.6874, "step": 2400 }, { "epoch": 0.02159434823472274, "grad_norm": 1.665571689605713, "learning_rate": 4.522056432768152e-05, "loss": 1.82, "step": 2401 }, { "epoch": 0.02160334213236319, "grad_norm": 4.107750415802002, "learning_rate": 4.521570158469883e-05, "loss": 1.8884, "step": 2402 }, { "epoch": 0.02161233603000364, "grad_norm": 2.763817548751831, "learning_rate": 4.521083663095742e-05, "loss": 1.6343, "step": 2403 }, { "epoch": 0.021621329927644095, "grad_norm": 1.3035956621170044, "learning_rate": 4.5205969466989304e-05, "loss": 1.7784, "step": 2404 }, { "epoch": 0.021630323825284545, "grad_norm": 1.596833348274231, "learning_rate": 4.520110009332674e-05, "loss": 1.6705, "step": 2405 }, { "epoch": 0.021639317722924996, "grad_norm": 1.5382798910140991, "learning_rate": 4.519622851050224e-05, "loss": 1.6617, "step": 2406 }, { "epoch": 0.021648311620565446, "grad_norm": 2.668247699737549, "learning_rate": 4.5191354719048556e-05, "loss": 1.7493, "step": 2407 }, { "epoch": 0.021657305518205897, "grad_norm": 2.2929046154022217, "learning_rate": 4.5186478719498684e-05, "loss": 1.9305, "step": 2408 }, { "epoch": 0.021666299415846347, "grad_norm": 2.331268072128296, "learning_rate": 4.5181600512385844e-05, "loss": 1.7794, "step": 2409 }, { "epoch": 0.0216752933134868, "grad_norm": 1.8910834789276123, "learning_rate": 4.517672009824351e-05, "loss": 1.6904, "step": 2410 }, { "epoch": 0.02168428721112725, "grad_norm": 2.0468902587890625, "learning_rate": 4.5171837477605404e-05, "loss": 1.7134, "step": 2411 }, { "epoch": 0.0216932811087677, "grad_norm": 2.992893934249878, "learning_rate": 4.5166952651005486e-05, "loss": 1.7458, "step": 2412 }, { "epoch": 0.021702275006408152, "grad_norm": 0.8974065184593201, "learning_rate": 4.516206561897794e-05, "loss": 1.7888, "step": 2413 }, { "epoch": 0.021711268904048602, "grad_norm": 2.16985821723938, "learning_rate": 4.515717638205721e-05, "loss": 1.7559, "step": 2414 }, { "epoch": 0.021720262801689053, "grad_norm": 2.6657633781433105, "learning_rate": 4.515228494077798e-05, "loss": 1.6527, "step": 2415 }, { "epoch": 0.021729256699329506, "grad_norm": 1.9845093488693237, "learning_rate": 4.514739129567517e-05, "loss": 1.7507, "step": 2416 }, { "epoch": 0.021738250596969957, "grad_norm": 2.424546003341675, "learning_rate": 4.514249544728393e-05, "loss": 1.8026, "step": 2417 }, { "epoch": 0.021747244494610407, "grad_norm": 4.5815205574035645, "learning_rate": 4.5137597396139676e-05, "loss": 1.7492, "step": 2418 }, { "epoch": 0.021756238392250857, "grad_norm": 2.550327777862549, "learning_rate": 4.513269714277805e-05, "loss": 1.7699, "step": 2419 }, { "epoch": 0.021765232289891308, "grad_norm": 4.354312896728516, "learning_rate": 4.512779468773494e-05, "loss": 1.6564, "step": 2420 }, { "epoch": 0.021774226187531758, "grad_norm": 1.2560617923736572, "learning_rate": 4.512289003154645e-05, "loss": 1.7723, "step": 2421 }, { "epoch": 0.021783220085172212, "grad_norm": 1.8505702018737793, "learning_rate": 4.5117983174748966e-05, "loss": 1.8287, "step": 2422 }, { "epoch": 0.021792213982812662, "grad_norm": 1.7713243961334229, "learning_rate": 4.51130741178791e-05, "loss": 1.6734, "step": 2423 }, { "epoch": 0.021801207880453113, "grad_norm": 2.115459442138672, "learning_rate": 4.5108162861473665e-05, "loss": 1.7997, "step": 2424 }, { "epoch": 0.021810201778093563, "grad_norm": 2.980518102645874, "learning_rate": 4.510324940606979e-05, "loss": 1.6707, "step": 2425 }, { "epoch": 0.021819195675734013, "grad_norm": 3.4119644165039062, "learning_rate": 4.509833375220478e-05, "loss": 1.8845, "step": 2426 }, { "epoch": 0.021828189573374464, "grad_norm": 2.7175042629241943, "learning_rate": 4.50934159004162e-05, "loss": 1.6964, "step": 2427 }, { "epoch": 0.021837183471014918, "grad_norm": 1.4785937070846558, "learning_rate": 4.508849585124189e-05, "loss": 1.8817, "step": 2428 }, { "epoch": 0.021846177368655368, "grad_norm": 2.72977876663208, "learning_rate": 4.5083573605219856e-05, "loss": 1.7593, "step": 2429 }, { "epoch": 0.02185517126629582, "grad_norm": 2.556875467300415, "learning_rate": 4.5078649162888406e-05, "loss": 1.7746, "step": 2430 }, { "epoch": 0.02186416516393627, "grad_norm": 2.216238498687744, "learning_rate": 4.507372252478608e-05, "loss": 1.8577, "step": 2431 }, { "epoch": 0.02187315906157672, "grad_norm": 1.8527342081069946, "learning_rate": 4.5068793691451635e-05, "loss": 1.728, "step": 2432 }, { "epoch": 0.02188215295921717, "grad_norm": 2.1717731952667236, "learning_rate": 4.506386266342409e-05, "loss": 1.6851, "step": 2433 }, { "epoch": 0.021891146856857623, "grad_norm": 1.9931641817092896, "learning_rate": 4.5058929441242694e-05, "loss": 1.7917, "step": 2434 }, { "epoch": 0.021900140754498074, "grad_norm": 1.3882536888122559, "learning_rate": 4.505399402544692e-05, "loss": 1.6997, "step": 2435 }, { "epoch": 0.021909134652138524, "grad_norm": 1.627825140953064, "learning_rate": 4.504905641657652e-05, "loss": 1.7926, "step": 2436 }, { "epoch": 0.021918128549778974, "grad_norm": 1.489967703819275, "learning_rate": 4.504411661517144e-05, "loss": 1.7866, "step": 2437 }, { "epoch": 0.021927122447419425, "grad_norm": 1.6435235738754272, "learning_rate": 4.503917462177192e-05, "loss": 1.7618, "step": 2438 }, { "epoch": 0.02193611634505988, "grad_norm": 2.0037734508514404, "learning_rate": 4.5034230436918377e-05, "loss": 1.7904, "step": 2439 }, { "epoch": 0.02194511024270033, "grad_norm": 3.1363093852996826, "learning_rate": 4.502928406115152e-05, "loss": 1.7552, "step": 2440 }, { "epoch": 0.02195410414034078, "grad_norm": 2.6401546001434326, "learning_rate": 4.502433549501228e-05, "loss": 1.7894, "step": 2441 }, { "epoch": 0.02196309803798123, "grad_norm": 3.266185760498047, "learning_rate": 4.5019384739041804e-05, "loss": 1.7493, "step": 2442 }, { "epoch": 0.02197209193562168, "grad_norm": 1.8705295324325562, "learning_rate": 4.501443179378152e-05, "loss": 1.7331, "step": 2443 }, { "epoch": 0.02198108583326213, "grad_norm": 3.383681058883667, "learning_rate": 4.500947665977306e-05, "loss": 1.8316, "step": 2444 }, { "epoch": 0.021990079730902584, "grad_norm": 2.5728187561035156, "learning_rate": 4.500451933755833e-05, "loss": 1.6844, "step": 2445 }, { "epoch": 0.021999073628543035, "grad_norm": 32.62267303466797, "learning_rate": 4.499955982767943e-05, "loss": 1.7711, "step": 2446 }, { "epoch": 0.022008067526183485, "grad_norm": 2.911128282546997, "learning_rate": 4.499459813067873e-05, "loss": 1.7697, "step": 2447 }, { "epoch": 0.022017061423823935, "grad_norm": 2.7405428886413574, "learning_rate": 4.4989634247098843e-05, "loss": 1.6779, "step": 2448 }, { "epoch": 0.022026055321464386, "grad_norm": 1.8247190713882446, "learning_rate": 4.498466817748262e-05, "loss": 1.5631, "step": 2449 }, { "epoch": 0.022035049219104836, "grad_norm": 10.378780364990234, "learning_rate": 4.497969992237312e-05, "loss": 1.8428, "step": 2450 }, { "epoch": 0.02204404311674529, "grad_norm": 1.7823221683502197, "learning_rate": 4.497472948231368e-05, "loss": 1.6899, "step": 2451 }, { "epoch": 0.02205303701438574, "grad_norm": 6.158448696136475, "learning_rate": 4.4969756857847854e-05, "loss": 1.7813, "step": 2452 }, { "epoch": 0.02206203091202619, "grad_norm": 3.4771456718444824, "learning_rate": 4.496478204951945e-05, "loss": 1.7516, "step": 2453 }, { "epoch": 0.02207102480966664, "grad_norm": 7.048118591308594, "learning_rate": 4.4959805057872486e-05, "loss": 1.8071, "step": 2454 }, { "epoch": 0.02208001870730709, "grad_norm": 2.4309585094451904, "learning_rate": 4.495482588345126e-05, "loss": 1.671, "step": 2455 }, { "epoch": 0.02208901260494754, "grad_norm": 2.7919580936431885, "learning_rate": 4.4949844526800264e-05, "loss": 1.8067, "step": 2456 }, { "epoch": 0.022098006502587995, "grad_norm": 3.403998613357544, "learning_rate": 4.4944860988464276e-05, "loss": 1.7145, "step": 2457 }, { "epoch": 0.022107000400228446, "grad_norm": 2.629631757736206, "learning_rate": 4.4939875268988273e-05, "loss": 1.7423, "step": 2458 }, { "epoch": 0.022115994297868896, "grad_norm": 4.330638885498047, "learning_rate": 4.493488736891749e-05, "loss": 1.7498, "step": 2459 }, { "epoch": 0.022124988195509347, "grad_norm": 2.9075169563293457, "learning_rate": 4.49298972887974e-05, "loss": 1.7502, "step": 2460 }, { "epoch": 0.022133982093149797, "grad_norm": 1.4923813343048096, "learning_rate": 4.49249050291737e-05, "loss": 1.7178, "step": 2461 }, { "epoch": 0.022142975990790247, "grad_norm": 4.232454776763916, "learning_rate": 4.491991059059235e-05, "loss": 1.6517, "step": 2462 }, { "epoch": 0.0221519698884307, "grad_norm": 7.263999938964844, "learning_rate": 4.491491397359952e-05, "loss": 1.772, "step": 2463 }, { "epoch": 0.02216096378607115, "grad_norm": 5.110044002532959, "learning_rate": 4.490991517874165e-05, "loss": 1.748, "step": 2464 }, { "epoch": 0.022169957683711602, "grad_norm": 1.7948075532913208, "learning_rate": 4.490491420656537e-05, "loss": 1.7066, "step": 2465 }, { "epoch": 0.022178951581352052, "grad_norm": 1.8298461437225342, "learning_rate": 4.489991105761762e-05, "loss": 1.7184, "step": 2466 }, { "epoch": 0.022187945478992502, "grad_norm": 4.800569534301758, "learning_rate": 4.48949057324455e-05, "loss": 1.8346, "step": 2467 }, { "epoch": 0.022196939376632953, "grad_norm": 4.357138633728027, "learning_rate": 4.488989823159641e-05, "loss": 1.6833, "step": 2468 }, { "epoch": 0.022205933274273407, "grad_norm": 2.6826188564300537, "learning_rate": 4.488488855561795e-05, "loss": 1.6421, "step": 2469 }, { "epoch": 0.022214927171913857, "grad_norm": 2.5104122161865234, "learning_rate": 4.487987670505798e-05, "loss": 1.8181, "step": 2470 }, { "epoch": 0.022223921069554307, "grad_norm": 2.237016201019287, "learning_rate": 4.487486268046458e-05, "loss": 1.5667, "step": 2471 }, { "epoch": 0.022232914967194758, "grad_norm": 1.9370285272598267, "learning_rate": 4.486984648238607e-05, "loss": 1.7191, "step": 2472 }, { "epoch": 0.022241908864835208, "grad_norm": 2.5880191326141357, "learning_rate": 4.4864828111371024e-05, "loss": 1.6769, "step": 2473 }, { "epoch": 0.022250902762475662, "grad_norm": 1.7595542669296265, "learning_rate": 4.485980756796825e-05, "loss": 1.646, "step": 2474 }, { "epoch": 0.022259896660116112, "grad_norm": 1.4383643865585327, "learning_rate": 4.4854784852726776e-05, "loss": 1.7808, "step": 2475 }, { "epoch": 0.022268890557756563, "grad_norm": 2.64825701713562, "learning_rate": 4.484975996619589e-05, "loss": 1.6817, "step": 2476 }, { "epoch": 0.022277884455397013, "grad_norm": 2.1559090614318848, "learning_rate": 4.4844732908925084e-05, "loss": 1.7541, "step": 2477 }, { "epoch": 0.022286878353037463, "grad_norm": 1.5271787643432617, "learning_rate": 4.483970368146413e-05, "loss": 1.7494, "step": 2478 }, { "epoch": 0.022295872250677914, "grad_norm": 1.5196506977081299, "learning_rate": 4.483467228436301e-05, "loss": 1.7584, "step": 2479 }, { "epoch": 0.022304866148318368, "grad_norm": 2.0028042793273926, "learning_rate": 4.4829638718171954e-05, "loss": 1.8785, "step": 2480 }, { "epoch": 0.022313860045958818, "grad_norm": 2.368055582046509, "learning_rate": 4.482460298344141e-05, "loss": 1.6668, "step": 2481 }, { "epoch": 0.02232285394359927, "grad_norm": 1.4766929149627686, "learning_rate": 4.4819565080722095e-05, "loss": 1.5933, "step": 2482 }, { "epoch": 0.02233184784123972, "grad_norm": 2.937659978866577, "learning_rate": 4.4814525010564945e-05, "loss": 1.7424, "step": 2483 }, { "epoch": 0.02234084173888017, "grad_norm": 1.420436978340149, "learning_rate": 4.480948277352113e-05, "loss": 1.6814, "step": 2484 }, { "epoch": 0.02234983563652062, "grad_norm": 1.6726047992706299, "learning_rate": 4.480443837014205e-05, "loss": 1.7308, "step": 2485 }, { "epoch": 0.022358829534161073, "grad_norm": 2.072831392288208, "learning_rate": 4.479939180097938e-05, "loss": 1.7107, "step": 2486 }, { "epoch": 0.022367823431801524, "grad_norm": 1.8239449262619019, "learning_rate": 4.479434306658498e-05, "loss": 1.744, "step": 2487 }, { "epoch": 0.022376817329441974, "grad_norm": 1.9663686752319336, "learning_rate": 4.478929216751099e-05, "loss": 1.8421, "step": 2488 }, { "epoch": 0.022385811227082424, "grad_norm": 1.464613914489746, "learning_rate": 4.478423910430975e-05, "loss": 1.7196, "step": 2489 }, { "epoch": 0.022394805124722875, "grad_norm": 4.3503899574279785, "learning_rate": 4.477918387753388e-05, "loss": 1.7342, "step": 2490 }, { "epoch": 0.022403799022363325, "grad_norm": 1.9346001148223877, "learning_rate": 4.47741264877362e-05, "loss": 1.6994, "step": 2491 }, { "epoch": 0.02241279292000378, "grad_norm": 3.3933119773864746, "learning_rate": 4.476906693546977e-05, "loss": 1.7969, "step": 2492 }, { "epoch": 0.02242178681764423, "grad_norm": 2.0753161907196045, "learning_rate": 4.476400522128791e-05, "loss": 1.7382, "step": 2493 }, { "epoch": 0.02243078071528468, "grad_norm": 1.0860116481781006, "learning_rate": 4.4758941345744146e-05, "loss": 1.8311, "step": 2494 }, { "epoch": 0.02243977461292513, "grad_norm": 1.9261865615844727, "learning_rate": 4.4753875309392266e-05, "loss": 1.7586, "step": 2495 }, { "epoch": 0.02244876851056558, "grad_norm": 2.471877098083496, "learning_rate": 4.474880711278628e-05, "loss": 1.7311, "step": 2496 }, { "epoch": 0.02245776240820603, "grad_norm": 3.8031203746795654, "learning_rate": 4.474373675648044e-05, "loss": 1.7095, "step": 2497 }, { "epoch": 0.022466756305846484, "grad_norm": 1.482415795326233, "learning_rate": 4.473866424102923e-05, "loss": 1.6946, "step": 2498 }, { "epoch": 0.022475750203486935, "grad_norm": 1.6045700311660767, "learning_rate": 4.473358956698738e-05, "loss": 1.8206, "step": 2499 }, { "epoch": 0.022484744101127385, "grad_norm": 3.545814275741577, "learning_rate": 4.4728512734909844e-05, "loss": 1.7495, "step": 2500 }, { "epoch": 0.022493737998767836, "grad_norm": 1.9938493967056274, "learning_rate": 4.4723433745351815e-05, "loss": 1.7797, "step": 2501 }, { "epoch": 0.022502731896408286, "grad_norm": 1.8415980339050293, "learning_rate": 4.4718352598868723e-05, "loss": 1.6961, "step": 2502 }, { "epoch": 0.022511725794048736, "grad_norm": 1.303626537322998, "learning_rate": 4.471326929601624e-05, "loss": 1.7442, "step": 2503 }, { "epoch": 0.02252071969168919, "grad_norm": 1.417374610900879, "learning_rate": 4.470818383735027e-05, "loss": 1.6979, "step": 2504 }, { "epoch": 0.02252971358932964, "grad_norm": 1.9532006978988647, "learning_rate": 4.470309622342694e-05, "loss": 1.7538, "step": 2505 }, { "epoch": 0.02253870748697009, "grad_norm": 2.9681174755096436, "learning_rate": 4.469800645480263e-05, "loss": 1.6968, "step": 2506 }, { "epoch": 0.02254770138461054, "grad_norm": 1.8433778285980225, "learning_rate": 4.469291453203395e-05, "loss": 1.8026, "step": 2507 }, { "epoch": 0.02255669528225099, "grad_norm": 2.1546409130096436, "learning_rate": 4.468782045567775e-05, "loss": 1.7224, "step": 2508 }, { "epoch": 0.022565689179891445, "grad_norm": 1.5890923738479614, "learning_rate": 4.46827242262911e-05, "loss": 1.7302, "step": 2509 }, { "epoch": 0.022574683077531896, "grad_norm": 2.3349363803863525, "learning_rate": 4.467762584443131e-05, "loss": 1.7885, "step": 2510 }, { "epoch": 0.022583676975172346, "grad_norm": 2.9211292266845703, "learning_rate": 4.4672525310655965e-05, "loss": 1.8068, "step": 2511 }, { "epoch": 0.022592670872812796, "grad_norm": 1.772200584411621, "learning_rate": 4.466742262552281e-05, "loss": 1.6742, "step": 2512 }, { "epoch": 0.022601664770453247, "grad_norm": 1.908707618713379, "learning_rate": 4.4662317789589904e-05, "loss": 1.7136, "step": 2513 }, { "epoch": 0.022610658668093697, "grad_norm": 3.780785083770752, "learning_rate": 4.465721080341547e-05, "loss": 1.6508, "step": 2514 }, { "epoch": 0.02261965256573415, "grad_norm": 3.334737539291382, "learning_rate": 4.465210166755803e-05, "loss": 1.7946, "step": 2515 }, { "epoch": 0.0226286464633746, "grad_norm": 4.083311080932617, "learning_rate": 4.4646990382576295e-05, "loss": 1.727, "step": 2516 }, { "epoch": 0.02263764036101505, "grad_norm": 3.4165453910827637, "learning_rate": 4.4641876949029224e-05, "loss": 1.702, "step": 2517 }, { "epoch": 0.022646634258655502, "grad_norm": 2.0967862606048584, "learning_rate": 4.463676136747602e-05, "loss": 1.7258, "step": 2518 }, { "epoch": 0.022655628156295952, "grad_norm": 4.003046989440918, "learning_rate": 4.463164363847613e-05, "loss": 1.7679, "step": 2519 }, { "epoch": 0.022664622053936403, "grad_norm": 2.716966152191162, "learning_rate": 4.46265237625892e-05, "loss": 1.7513, "step": 2520 }, { "epoch": 0.022673615951576857, "grad_norm": 3.5525312423706055, "learning_rate": 4.462140174037514e-05, "loss": 1.7409, "step": 2521 }, { "epoch": 0.022682609849217307, "grad_norm": 3.1177711486816406, "learning_rate": 4.461627757239407e-05, "loss": 1.8495, "step": 2522 }, { "epoch": 0.022691603746857757, "grad_norm": 1.046033501625061, "learning_rate": 4.46111512592064e-05, "loss": 1.7216, "step": 2523 }, { "epoch": 0.022700597644498208, "grad_norm": 1.7619438171386719, "learning_rate": 4.460602280137271e-05, "loss": 1.6812, "step": 2524 }, { "epoch": 0.022709591542138658, "grad_norm": 2.5585548877716064, "learning_rate": 4.460089219945382e-05, "loss": 1.7328, "step": 2525 }, { "epoch": 0.02271858543977911, "grad_norm": 1.860069990158081, "learning_rate": 4.459575945401086e-05, "loss": 1.7473, "step": 2526 }, { "epoch": 0.022727579337419562, "grad_norm": 2.7434701919555664, "learning_rate": 4.459062456560509e-05, "loss": 1.7423, "step": 2527 }, { "epoch": 0.022736573235060013, "grad_norm": 3.1500461101531982, "learning_rate": 4.4585487534798076e-05, "loss": 1.7867, "step": 2528 }, { "epoch": 0.022745567132700463, "grad_norm": 10.146349906921387, "learning_rate": 4.4580348362151595e-05, "loss": 1.7696, "step": 2529 }, { "epoch": 0.022754561030340913, "grad_norm": 1.3915654420852661, "learning_rate": 4.457520704822765e-05, "loss": 1.8429, "step": 2530 }, { "epoch": 0.022763554927981364, "grad_norm": 1.9898637533187866, "learning_rate": 4.457006359358849e-05, "loss": 1.7216, "step": 2531 }, { "epoch": 0.022772548825621814, "grad_norm": 1.6907819509506226, "learning_rate": 4.456491799879661e-05, "loss": 1.697, "step": 2532 }, { "epoch": 0.022781542723262268, "grad_norm": 3.413398504257202, "learning_rate": 4.45597702644147e-05, "loss": 1.7696, "step": 2533 }, { "epoch": 0.022790536620902718, "grad_norm": 2.0847651958465576, "learning_rate": 4.455462039100573e-05, "loss": 1.73, "step": 2534 }, { "epoch": 0.02279953051854317, "grad_norm": 7.569029808044434, "learning_rate": 4.454946837913287e-05, "loss": 1.7371, "step": 2535 }, { "epoch": 0.02280852441618362, "grad_norm": 27.83839225769043, "learning_rate": 4.454431422935955e-05, "loss": 1.8859, "step": 2536 }, { "epoch": 0.02281751831382407, "grad_norm": 16.093088150024414, "learning_rate": 4.4539157942249396e-05, "loss": 1.7955, "step": 2537 }, { "epoch": 0.02282651221146452, "grad_norm": 12.691722869873047, "learning_rate": 4.453399951836632e-05, "loss": 1.7344, "step": 2538 }, { "epoch": 0.022835506109104974, "grad_norm": 1.42303466796875, "learning_rate": 4.452883895827442e-05, "loss": 1.7557, "step": 2539 }, { "epoch": 0.022844500006745424, "grad_norm": 4.607066631317139, "learning_rate": 4.452367626253805e-05, "loss": 1.7289, "step": 2540 }, { "epoch": 0.022853493904385874, "grad_norm": 4.663803577423096, "learning_rate": 4.45185114317218e-05, "loss": 1.7589, "step": 2541 }, { "epoch": 0.022862487802026325, "grad_norm": 3.654425621032715, "learning_rate": 4.451334446639048e-05, "loss": 1.7528, "step": 2542 }, { "epoch": 0.022871481699666775, "grad_norm": 4.471340179443359, "learning_rate": 4.4508175367109155e-05, "loss": 1.7021, "step": 2543 }, { "epoch": 0.022880475597307225, "grad_norm": 8.770857810974121, "learning_rate": 4.45030041344431e-05, "loss": 1.7883, "step": 2544 }, { "epoch": 0.02288946949494768, "grad_norm": 2.716670036315918, "learning_rate": 4.449783076895783e-05, "loss": 1.7381, "step": 2545 }, { "epoch": 0.02289846339258813, "grad_norm": 3.0804073810577393, "learning_rate": 4.449265527121911e-05, "loss": 1.7133, "step": 2546 }, { "epoch": 0.02290745729022858, "grad_norm": 3.5980589389801025, "learning_rate": 4.448747764179291e-05, "loss": 1.808, "step": 2547 }, { "epoch": 0.02291645118786903, "grad_norm": 2.1191673278808594, "learning_rate": 4.448229788124545e-05, "loss": 1.6481, "step": 2548 }, { "epoch": 0.02292544508550948, "grad_norm": 4.995469093322754, "learning_rate": 4.447711599014319e-05, "loss": 1.7426, "step": 2549 }, { "epoch": 0.022934438983149934, "grad_norm": 57.55516052246094, "learning_rate": 4.4471931969052816e-05, "loss": 1.8498, "step": 2550 }, { "epoch": 0.022943432880790385, "grad_norm": 18.011457443237305, "learning_rate": 4.446674581854123e-05, "loss": 1.7915, "step": 2551 }, { "epoch": 0.022952426778430835, "grad_norm": 3.3407142162323, "learning_rate": 4.4461557539175594e-05, "loss": 1.8433, "step": 2552 }, { "epoch": 0.022961420676071286, "grad_norm": 3.0757453441619873, "learning_rate": 4.445636713152327e-05, "loss": 1.7396, "step": 2553 }, { "epoch": 0.022970414573711736, "grad_norm": 1.740971326828003, "learning_rate": 4.4451174596151906e-05, "loss": 1.7819, "step": 2554 }, { "epoch": 0.022979408471352186, "grad_norm": 2.2860124111175537, "learning_rate": 4.4445979933629324e-05, "loss": 1.7272, "step": 2555 }, { "epoch": 0.02298840236899264, "grad_norm": 2.6933906078338623, "learning_rate": 4.444078314452362e-05, "loss": 1.7547, "step": 2556 }, { "epoch": 0.02299739626663309, "grad_norm": 2.663923978805542, "learning_rate": 4.443558422940309e-05, "loss": 1.7177, "step": 2557 }, { "epoch": 0.02300639016427354, "grad_norm": 3.1298370361328125, "learning_rate": 4.44303831888363e-05, "loss": 1.7405, "step": 2558 }, { "epoch": 0.02301538406191399, "grad_norm": 7.782910346984863, "learning_rate": 4.442518002339201e-05, "loss": 1.6535, "step": 2559 }, { "epoch": 0.02302437795955444, "grad_norm": 2.03230357170105, "learning_rate": 4.4419974733639244e-05, "loss": 1.7192, "step": 2560 }, { "epoch": 0.023033371857194892, "grad_norm": 2.9448091983795166, "learning_rate": 4.441476732014723e-05, "loss": 1.7249, "step": 2561 }, { "epoch": 0.023042365754835346, "grad_norm": 1.3239023685455322, "learning_rate": 4.4409557783485464e-05, "loss": 1.6441, "step": 2562 }, { "epoch": 0.023051359652475796, "grad_norm": 3.957076072692871, "learning_rate": 4.4404346124223626e-05, "loss": 1.7569, "step": 2563 }, { "epoch": 0.023060353550116246, "grad_norm": 1.8063708543777466, "learning_rate": 4.4399132342931684e-05, "loss": 1.7598, "step": 2564 }, { "epoch": 0.023069347447756697, "grad_norm": 1.5154401063919067, "learning_rate": 4.4393916440179786e-05, "loss": 1.659, "step": 2565 }, { "epoch": 0.023078341345397147, "grad_norm": 1.0538969039916992, "learning_rate": 4.438869841653835e-05, "loss": 1.7038, "step": 2566 }, { "epoch": 0.023087335243037597, "grad_norm": 2.8788247108459473, "learning_rate": 4.438347827257801e-05, "loss": 1.7488, "step": 2567 }, { "epoch": 0.02309632914067805, "grad_norm": 2.7268004417419434, "learning_rate": 4.4378256008869615e-05, "loss": 1.6723, "step": 2568 }, { "epoch": 0.0231053230383185, "grad_norm": 2.401029348373413, "learning_rate": 4.4373031625984285e-05, "loss": 1.6759, "step": 2569 }, { "epoch": 0.023114316935958952, "grad_norm": 4.297641754150391, "learning_rate": 4.436780512449334e-05, "loss": 1.7147, "step": 2570 }, { "epoch": 0.023123310833599402, "grad_norm": 2.161261558532715, "learning_rate": 4.436257650496834e-05, "loss": 1.6654, "step": 2571 }, { "epoch": 0.023132304731239853, "grad_norm": 2.4188125133514404, "learning_rate": 4.43573457679811e-05, "loss": 1.7459, "step": 2572 }, { "epoch": 0.023141298628880303, "grad_norm": 2.7243731021881104, "learning_rate": 4.435211291410363e-05, "loss": 1.7257, "step": 2573 }, { "epoch": 0.023150292526520757, "grad_norm": 4.8212995529174805, "learning_rate": 4.434687794390817e-05, "loss": 1.8274, "step": 2574 }, { "epoch": 0.023159286424161207, "grad_norm": 2.99977707862854, "learning_rate": 4.434164085796724e-05, "loss": 1.717, "step": 2575 }, { "epoch": 0.023168280321801658, "grad_norm": 3.357797861099243, "learning_rate": 4.4336401656853546e-05, "loss": 1.728, "step": 2576 }, { "epoch": 0.023177274219442108, "grad_norm": 2.3460450172424316, "learning_rate": 4.433116034114003e-05, "loss": 1.6498, "step": 2577 }, { "epoch": 0.02318626811708256, "grad_norm": 2.261333703994751, "learning_rate": 4.432591691139989e-05, "loss": 1.7141, "step": 2578 }, { "epoch": 0.02319526201472301, "grad_norm": 3.919997215270996, "learning_rate": 4.432067136820652e-05, "loss": 1.7067, "step": 2579 }, { "epoch": 0.023204255912363463, "grad_norm": 4.087550163269043, "learning_rate": 4.4315423712133595e-05, "loss": 1.7093, "step": 2580 }, { "epoch": 0.023213249810003913, "grad_norm": 4.743485450744629, "learning_rate": 4.431017394375496e-05, "loss": 1.8643, "step": 2581 }, { "epoch": 0.023222243707644363, "grad_norm": 1.4507025480270386, "learning_rate": 4.430492206364474e-05, "loss": 1.8092, "step": 2582 }, { "epoch": 0.023231237605284814, "grad_norm": 1.4695076942443848, "learning_rate": 4.429966807237726e-05, "loss": 1.7249, "step": 2583 }, { "epoch": 0.023240231502925264, "grad_norm": 1.9584555625915527, "learning_rate": 4.4294411970527116e-05, "loss": 1.7525, "step": 2584 }, { "epoch": 0.023249225400565718, "grad_norm": 0.9730052351951599, "learning_rate": 4.4289153758669075e-05, "loss": 1.6573, "step": 2585 }, { "epoch": 0.023258219298206168, "grad_norm": 3.8035545349121094, "learning_rate": 4.4283893437378173e-05, "loss": 1.6959, "step": 2586 }, { "epoch": 0.02326721319584662, "grad_norm": 1.8013588190078735, "learning_rate": 4.427863100722969e-05, "loss": 1.6761, "step": 2587 }, { "epoch": 0.02327620709348707, "grad_norm": 1.3779282569885254, "learning_rate": 4.42733664687991e-05, "loss": 1.6082, "step": 2588 }, { "epoch": 0.02328520099112752, "grad_norm": 2.8634867668151855, "learning_rate": 4.426809982266214e-05, "loss": 1.709, "step": 2589 }, { "epoch": 0.02329419488876797, "grad_norm": 2.1815404891967773, "learning_rate": 4.426283106939474e-05, "loss": 1.7416, "step": 2590 }, { "epoch": 0.023303188786408423, "grad_norm": 2.592116117477417, "learning_rate": 4.425756020957309e-05, "loss": 1.733, "step": 2591 }, { "epoch": 0.023312182684048874, "grad_norm": 1.9898604154586792, "learning_rate": 4.4252287243773617e-05, "loss": 1.7692, "step": 2592 }, { "epoch": 0.023321176581689324, "grad_norm": 2.1133716106414795, "learning_rate": 4.424701217257295e-05, "loss": 1.7795, "step": 2593 }, { "epoch": 0.023330170479329775, "grad_norm": 5.981254577636719, "learning_rate": 4.424173499654798e-05, "loss": 1.6708, "step": 2594 }, { "epoch": 0.023339164376970225, "grad_norm": 4.268572807312012, "learning_rate": 4.42364557162758e-05, "loss": 1.651, "step": 2595 }, { "epoch": 0.023348158274610675, "grad_norm": 4.23325777053833, "learning_rate": 4.4231174332333736e-05, "loss": 1.7085, "step": 2596 }, { "epoch": 0.02335715217225113, "grad_norm": 2.7404489517211914, "learning_rate": 4.422589084529937e-05, "loss": 1.7924, "step": 2597 }, { "epoch": 0.02336614606989158, "grad_norm": 3.655122995376587, "learning_rate": 4.422060525575048e-05, "loss": 1.676, "step": 2598 }, { "epoch": 0.02337513996753203, "grad_norm": 2.0850017070770264, "learning_rate": 4.4215317564265096e-05, "loss": 1.6981, "step": 2599 }, { "epoch": 0.02338413386517248, "grad_norm": 2.9497222900390625, "learning_rate": 4.421002777142148e-05, "loss": 1.7252, "step": 2600 }, { "epoch": 0.02339312776281293, "grad_norm": 2.887794256210327, "learning_rate": 4.42047358777981e-05, "loss": 1.7539, "step": 2601 }, { "epoch": 0.02340212166045338, "grad_norm": 3.9264276027679443, "learning_rate": 4.419944188397369e-05, "loss": 1.6104, "step": 2602 }, { "epoch": 0.023411115558093835, "grad_norm": 8.442761421203613, "learning_rate": 4.419414579052717e-05, "loss": 1.6528, "step": 2603 }, { "epoch": 0.023420109455734285, "grad_norm": 9.16222858428955, "learning_rate": 4.418884759803773e-05, "loss": 1.7297, "step": 2604 }, { "epoch": 0.023429103353374735, "grad_norm": 8.642212867736816, "learning_rate": 4.418354730708476e-05, "loss": 1.6485, "step": 2605 }, { "epoch": 0.023438097251015186, "grad_norm": 2.8616952896118164, "learning_rate": 4.41782449182479e-05, "loss": 1.729, "step": 2606 }, { "epoch": 0.023447091148655636, "grad_norm": 3.018643379211426, "learning_rate": 4.4172940432107015e-05, "loss": 1.6459, "step": 2607 }, { "epoch": 0.023456085046296087, "grad_norm": 1.9177954196929932, "learning_rate": 4.4167633849242196e-05, "loss": 1.6429, "step": 2608 }, { "epoch": 0.02346507894393654, "grad_norm": 11.711039543151855, "learning_rate": 4.4162325170233745e-05, "loss": 1.746, "step": 2609 }, { "epoch": 0.02347407284157699, "grad_norm": 5.09904670715332, "learning_rate": 4.415701439566223e-05, "loss": 1.6583, "step": 2610 }, { "epoch": 0.02348306673921744, "grad_norm": 2.6431360244750977, "learning_rate": 4.415170152610842e-05, "loss": 1.6792, "step": 2611 }, { "epoch": 0.02349206063685789, "grad_norm": 7.673257827758789, "learning_rate": 4.414638656215333e-05, "loss": 1.818, "step": 2612 }, { "epoch": 0.023501054534498342, "grad_norm": 5.714672565460205, "learning_rate": 4.414106950437819e-05, "loss": 1.7783, "step": 2613 }, { "epoch": 0.023510048432138792, "grad_norm": 2.6301169395446777, "learning_rate": 4.413575035336447e-05, "loss": 1.8362, "step": 2614 }, { "epoch": 0.023519042329779246, "grad_norm": 5.371310234069824, "learning_rate": 4.413042910969385e-05, "loss": 1.6229, "step": 2615 }, { "epoch": 0.023528036227419696, "grad_norm": 2.4130959510803223, "learning_rate": 4.412510577394828e-05, "loss": 1.8221, "step": 2616 }, { "epoch": 0.023537030125060147, "grad_norm": 2.375349998474121, "learning_rate": 4.411978034670988e-05, "loss": 1.8032, "step": 2617 }, { "epoch": 0.023546024022700597, "grad_norm": 5.099897384643555, "learning_rate": 4.411445282856106e-05, "loss": 1.5419, "step": 2618 }, { "epoch": 0.023555017920341047, "grad_norm": 1.923096776008606, "learning_rate": 4.410912322008441e-05, "loss": 1.748, "step": 2619 }, { "epoch": 0.0235640118179815, "grad_norm": 3.7113335132598877, "learning_rate": 4.4103791521862784e-05, "loss": 1.7896, "step": 2620 }, { "epoch": 0.02357300571562195, "grad_norm": 2.1036376953125, "learning_rate": 4.409845773447923e-05, "loss": 1.759, "step": 2621 }, { "epoch": 0.023581999613262402, "grad_norm": 2.6264255046844482, "learning_rate": 4.409312185851706e-05, "loss": 1.7335, "step": 2622 }, { "epoch": 0.023590993510902852, "grad_norm": 2.327192783355713, "learning_rate": 4.408778389455978e-05, "loss": 1.7934, "step": 2623 }, { "epoch": 0.023599987408543303, "grad_norm": 2.213585376739502, "learning_rate": 4.408244384319116e-05, "loss": 1.6244, "step": 2624 }, { "epoch": 0.023608981306183753, "grad_norm": 1.868430256843567, "learning_rate": 4.4077101704995166e-05, "loss": 1.7017, "step": 2625 }, { "epoch": 0.023617975203824207, "grad_norm": 2.189706563949585, "learning_rate": 4.407175748055601e-05, "loss": 1.6513, "step": 2626 }, { "epoch": 0.023626969101464657, "grad_norm": 2.4431962966918945, "learning_rate": 4.4066411170458136e-05, "loss": 1.8008, "step": 2627 }, { "epoch": 0.023635962999105108, "grad_norm": 2.3953068256378174, "learning_rate": 4.40610627752862e-05, "loss": 1.8016, "step": 2628 }, { "epoch": 0.023644956896745558, "grad_norm": 1.147969126701355, "learning_rate": 4.4055712295625095e-05, "loss": 1.7896, "step": 2629 }, { "epoch": 0.02365395079438601, "grad_norm": 5.755258083343506, "learning_rate": 4.405035973205994e-05, "loss": 1.7622, "step": 2630 }, { "epoch": 0.02366294469202646, "grad_norm": 2.2921955585479736, "learning_rate": 4.404500508517609e-05, "loss": 1.646, "step": 2631 }, { "epoch": 0.023671938589666913, "grad_norm": 2.543976306915283, "learning_rate": 4.4039648355559124e-05, "loss": 1.8045, "step": 2632 }, { "epoch": 0.023680932487307363, "grad_norm": 4.653142929077148, "learning_rate": 4.4034289543794836e-05, "loss": 1.8551, "step": 2633 }, { "epoch": 0.023689926384947813, "grad_norm": 1.4051008224487305, "learning_rate": 4.402892865046926e-05, "loss": 1.7428, "step": 2634 }, { "epoch": 0.023698920282588264, "grad_norm": 4.3316779136657715, "learning_rate": 4.4023565676168655e-05, "loss": 1.7083, "step": 2635 }, { "epoch": 0.023707914180228714, "grad_norm": 3.0948004722595215, "learning_rate": 4.401820062147951e-05, "loss": 1.7302, "step": 2636 }, { "epoch": 0.023716908077869164, "grad_norm": 1.4380182027816772, "learning_rate": 4.401283348698854e-05, "loss": 1.6311, "step": 2637 }, { "epoch": 0.023725901975509618, "grad_norm": 1.496585726737976, "learning_rate": 4.400746427328268e-05, "loss": 1.7154, "step": 2638 }, { "epoch": 0.02373489587315007, "grad_norm": 3.5473124980926514, "learning_rate": 4.400209298094911e-05, "loss": 1.5646, "step": 2639 }, { "epoch": 0.02374388977079052, "grad_norm": 2.5057926177978516, "learning_rate": 4.399671961057522e-05, "loss": 1.796, "step": 2640 }, { "epoch": 0.02375288366843097, "grad_norm": 1.4700664281845093, "learning_rate": 4.3991344162748626e-05, "loss": 1.7795, "step": 2641 }, { "epoch": 0.02376187756607142, "grad_norm": 3.592540740966797, "learning_rate": 4.3985966638057196e-05, "loss": 1.751, "step": 2642 }, { "epoch": 0.02377087146371187, "grad_norm": 3.8156721591949463, "learning_rate": 4.3980587037089e-05, "loss": 1.6724, "step": 2643 }, { "epoch": 0.023779865361352324, "grad_norm": 3.2661452293395996, "learning_rate": 4.397520536043234e-05, "loss": 1.6104, "step": 2644 }, { "epoch": 0.023788859258992774, "grad_norm": 2.4408576488494873, "learning_rate": 4.396982160867575e-05, "loss": 1.7401, "step": 2645 }, { "epoch": 0.023797853156633225, "grad_norm": 2.4495956897735596, "learning_rate": 4.396443578240799e-05, "loss": 1.8023, "step": 2646 }, { "epoch": 0.023806847054273675, "grad_norm": 2.291236162185669, "learning_rate": 4.395904788221805e-05, "loss": 1.6553, "step": 2647 }, { "epoch": 0.023815840951914125, "grad_norm": 3.266317844390869, "learning_rate": 4.395365790869513e-05, "loss": 1.5597, "step": 2648 }, { "epoch": 0.023824834849554576, "grad_norm": 2.9196999073028564, "learning_rate": 4.3948265862428684e-05, "loss": 1.7363, "step": 2649 }, { "epoch": 0.02383382874719503, "grad_norm": 1.706782579421997, "learning_rate": 4.3942871744008374e-05, "loss": 1.7447, "step": 2650 }, { "epoch": 0.02384282264483548, "grad_norm": 10.057560920715332, "learning_rate": 4.393747555402408e-05, "loss": 1.7959, "step": 2651 }, { "epoch": 0.02385181654247593, "grad_norm": 2.818437337875366, "learning_rate": 4.393207729306594e-05, "loss": 1.7242, "step": 2652 }, { "epoch": 0.02386081044011638, "grad_norm": 2.3812601566314697, "learning_rate": 4.3926676961724287e-05, "loss": 1.7535, "step": 2653 }, { "epoch": 0.02386980433775683, "grad_norm": 2.3439254760742188, "learning_rate": 4.39212745605897e-05, "loss": 1.7075, "step": 2654 }, { "epoch": 0.02387879823539728, "grad_norm": 1.6618317365646362, "learning_rate": 4.391587009025297e-05, "loss": 1.6867, "step": 2655 }, { "epoch": 0.023887792133037735, "grad_norm": 2.8314197063446045, "learning_rate": 4.3910463551305134e-05, "loss": 1.7743, "step": 2656 }, { "epoch": 0.023896786030678185, "grad_norm": 2.3661000728607178, "learning_rate": 4.390505494433743e-05, "loss": 1.697, "step": 2657 }, { "epoch": 0.023905779928318636, "grad_norm": 1.5190093517303467, "learning_rate": 4.389964426994134e-05, "loss": 1.7418, "step": 2658 }, { "epoch": 0.023914773825959086, "grad_norm": 2.8950164318084717, "learning_rate": 4.3894231528708576e-05, "loss": 1.6002, "step": 2659 }, { "epoch": 0.023923767723599536, "grad_norm": 1.363950252532959, "learning_rate": 4.388881672123105e-05, "loss": 1.7394, "step": 2660 }, { "epoch": 0.02393276162123999, "grad_norm": 1.96209716796875, "learning_rate": 4.3883399848100925e-05, "loss": 1.7245, "step": 2661 }, { "epoch": 0.02394175551888044, "grad_norm": 1.6328312158584595, "learning_rate": 4.387798090991059e-05, "loss": 1.6694, "step": 2662 }, { "epoch": 0.02395074941652089, "grad_norm": 4.515115261077881, "learning_rate": 4.3872559907252645e-05, "loss": 1.6633, "step": 2663 }, { "epoch": 0.02395974331416134, "grad_norm": 3.473578691482544, "learning_rate": 4.386713684071991e-05, "loss": 1.692, "step": 2664 }, { "epoch": 0.023968737211801792, "grad_norm": 2.4362151622772217, "learning_rate": 4.386171171090547e-05, "loss": 1.7796, "step": 2665 }, { "epoch": 0.023977731109442242, "grad_norm": 1.331058144569397, "learning_rate": 4.3856284518402594e-05, "loss": 1.7683, "step": 2666 }, { "epoch": 0.023986725007082696, "grad_norm": 1.2272098064422607, "learning_rate": 4.3850855263804794e-05, "loss": 1.6824, "step": 2667 }, { "epoch": 0.023995718904723146, "grad_norm": 3.1904826164245605, "learning_rate": 4.3845423947705796e-05, "loss": 1.5635, "step": 2668 }, { "epoch": 0.024004712802363597, "grad_norm": 2.539644718170166, "learning_rate": 4.383999057069959e-05, "loss": 1.6286, "step": 2669 }, { "epoch": 0.024013706700004047, "grad_norm": 2.0151796340942383, "learning_rate": 4.383455513338032e-05, "loss": 1.6502, "step": 2670 }, { "epoch": 0.024022700597644497, "grad_norm": 2.2896547317504883, "learning_rate": 4.382911763634242e-05, "loss": 1.81, "step": 2671 }, { "epoch": 0.024031694495284948, "grad_norm": 2.697456121444702, "learning_rate": 4.382367808018053e-05, "loss": 1.6416, "step": 2672 }, { "epoch": 0.0240406883929254, "grad_norm": 1.7801804542541504, "learning_rate": 4.38182364654895e-05, "loss": 1.7393, "step": 2673 }, { "epoch": 0.024049682290565852, "grad_norm": 2.3675482273101807, "learning_rate": 4.381279279286443e-05, "loss": 1.7639, "step": 2674 }, { "epoch": 0.024058676188206302, "grad_norm": 3.3090121746063232, "learning_rate": 4.3807347062900624e-05, "loss": 1.7939, "step": 2675 }, { "epoch": 0.024067670085846753, "grad_norm": 1.706540584564209, "learning_rate": 4.3801899276193614e-05, "loss": 1.7505, "step": 2676 }, { "epoch": 0.024076663983487203, "grad_norm": 2.080883502960205, "learning_rate": 4.379644943333917e-05, "loss": 1.6344, "step": 2677 }, { "epoch": 0.024085657881127653, "grad_norm": 3.254476308822632, "learning_rate": 4.3790997534933276e-05, "loss": 1.7908, "step": 2678 }, { "epoch": 0.024094651778768107, "grad_norm": 1.8081237077713013, "learning_rate": 4.3785543581572144e-05, "loss": 1.6967, "step": 2679 }, { "epoch": 0.024103645676408558, "grad_norm": 1.0798168182373047, "learning_rate": 4.3780087573852213e-05, "loss": 1.6959, "step": 2680 }, { "epoch": 0.024112639574049008, "grad_norm": 3.2559027671813965, "learning_rate": 4.3774629512370134e-05, "loss": 1.6632, "step": 2681 }, { "epoch": 0.02412163347168946, "grad_norm": 1.107149362564087, "learning_rate": 4.3769169397722805e-05, "loss": 1.7182, "step": 2682 }, { "epoch": 0.02413062736932991, "grad_norm": 3.5439693927764893, "learning_rate": 4.376370723050732e-05, "loss": 1.6133, "step": 2683 }, { "epoch": 0.02413962126697036, "grad_norm": 2.39306902885437, "learning_rate": 4.375824301132103e-05, "loss": 1.7604, "step": 2684 }, { "epoch": 0.024148615164610813, "grad_norm": 2.9069881439208984, "learning_rate": 4.375277674076149e-05, "loss": 1.5596, "step": 2685 }, { "epoch": 0.024157609062251263, "grad_norm": 4.1222920417785645, "learning_rate": 4.374730841942648e-05, "loss": 1.7166, "step": 2686 }, { "epoch": 0.024166602959891714, "grad_norm": 1.6853325366973877, "learning_rate": 4.3741838047914003e-05, "loss": 1.7295, "step": 2687 }, { "epoch": 0.024175596857532164, "grad_norm": 2.271606683731079, "learning_rate": 4.37363656268223e-05, "loss": 1.689, "step": 2688 }, { "epoch": 0.024184590755172614, "grad_norm": 2.3414177894592285, "learning_rate": 4.373089115674981e-05, "loss": 1.7682, "step": 2689 }, { "epoch": 0.024193584652813065, "grad_norm": 3.0189342498779297, "learning_rate": 4.3725414638295235e-05, "loss": 1.7648, "step": 2690 }, { "epoch": 0.02420257855045352, "grad_norm": 2.1109471321105957, "learning_rate": 4.3719936072057465e-05, "loss": 1.6638, "step": 2691 }, { "epoch": 0.02421157244809397, "grad_norm": 2.627939462661743, "learning_rate": 4.3714455458635625e-05, "loss": 1.7766, "step": 2692 }, { "epoch": 0.02422056634573442, "grad_norm": 1.9264460802078247, "learning_rate": 4.370897279862909e-05, "loss": 1.6816, "step": 2693 }, { "epoch": 0.02422956024337487, "grad_norm": 2.0545530319213867, "learning_rate": 4.3703488092637405e-05, "loss": 1.7829, "step": 2694 }, { "epoch": 0.02423855414101532, "grad_norm": 2.5396499633789062, "learning_rate": 4.369800134126039e-05, "loss": 1.6479, "step": 2695 }, { "epoch": 0.024247548038655774, "grad_norm": 3.6794400215148926, "learning_rate": 4.369251254509805e-05, "loss": 1.8238, "step": 2696 }, { "epoch": 0.024256541936296224, "grad_norm": 2.6968297958374023, "learning_rate": 4.3687021704750644e-05, "loss": 1.7177, "step": 2697 }, { "epoch": 0.024265535833936674, "grad_norm": 2.271775007247925, "learning_rate": 4.368152882081865e-05, "loss": 1.6855, "step": 2698 }, { "epoch": 0.024274529731577125, "grad_norm": 3.9526758193969727, "learning_rate": 4.3676033893902746e-05, "loss": 1.8312, "step": 2699 }, { "epoch": 0.024283523629217575, "grad_norm": 2.451770782470703, "learning_rate": 4.367053692460385e-05, "loss": 1.6901, "step": 2700 }, { "epoch": 0.024292517526858026, "grad_norm": 1.9047315120697021, "learning_rate": 4.366503791352311e-05, "loss": 1.7384, "step": 2701 }, { "epoch": 0.02430151142449848, "grad_norm": 2.2139203548431396, "learning_rate": 4.3659536861261884e-05, "loss": 1.6984, "step": 2702 }, { "epoch": 0.02431050532213893, "grad_norm": 2.7638673782348633, "learning_rate": 4.365403376842177e-05, "loss": 1.6985, "step": 2703 }, { "epoch": 0.02431949921977938, "grad_norm": 1.6364703178405762, "learning_rate": 4.3648528635604556e-05, "loss": 1.6326, "step": 2704 }, { "epoch": 0.02432849311741983, "grad_norm": 1.605384349822998, "learning_rate": 4.3643021463412294e-05, "loss": 1.6475, "step": 2705 }, { "epoch": 0.02433748701506028, "grad_norm": 1.989888072013855, "learning_rate": 4.363751225244723e-05, "loss": 1.7614, "step": 2706 }, { "epoch": 0.02434648091270073, "grad_norm": 1.6122353076934814, "learning_rate": 4.363200100331186e-05, "loss": 1.7884, "step": 2707 }, { "epoch": 0.024355474810341185, "grad_norm": 3.443624258041382, "learning_rate": 4.362648771660886e-05, "loss": 1.7893, "step": 2708 }, { "epoch": 0.024364468707981635, "grad_norm": 3.035010814666748, "learning_rate": 4.362097239294117e-05, "loss": 1.6831, "step": 2709 }, { "epoch": 0.024373462605622086, "grad_norm": 1.5377248525619507, "learning_rate": 4.3615455032911946e-05, "loss": 1.7019, "step": 2710 }, { "epoch": 0.024382456503262536, "grad_norm": 5.212434768676758, "learning_rate": 4.3609935637124535e-05, "loss": 1.7292, "step": 2711 }, { "epoch": 0.024391450400902986, "grad_norm": 3.3096868991851807, "learning_rate": 4.360441420618254e-05, "loss": 1.6224, "step": 2712 }, { "epoch": 0.024400444298543437, "grad_norm": 2.4585206508636475, "learning_rate": 4.359889074068979e-05, "loss": 1.7645, "step": 2713 }, { "epoch": 0.02440943819618389, "grad_norm": 3.2732527256011963, "learning_rate": 4.3593365241250304e-05, "loss": 1.7075, "step": 2714 }, { "epoch": 0.02441843209382434, "grad_norm": 2.6872575283050537, "learning_rate": 4.358783770846836e-05, "loss": 1.6871, "step": 2715 }, { "epoch": 0.02442742599146479, "grad_norm": 3.470177412033081, "learning_rate": 4.358230814294842e-05, "loss": 1.5929, "step": 2716 }, { "epoch": 0.02443641988910524, "grad_norm": 2.641704797744751, "learning_rate": 4.357677654529521e-05, "loss": 1.8209, "step": 2717 }, { "epoch": 0.024445413786745692, "grad_norm": 1.6253741979599, "learning_rate": 4.357124291611363e-05, "loss": 1.7733, "step": 2718 }, { "epoch": 0.024454407684386142, "grad_norm": 5.458434581756592, "learning_rate": 4.356570725600886e-05, "loss": 1.7345, "step": 2719 }, { "epoch": 0.024463401582026596, "grad_norm": 2.020177125930786, "learning_rate": 4.356016956558625e-05, "loss": 1.7674, "step": 2720 }, { "epoch": 0.024472395479667047, "grad_norm": 3.086212396621704, "learning_rate": 4.355462984545141e-05, "loss": 1.8724, "step": 2721 }, { "epoch": 0.024481389377307497, "grad_norm": 3.1671342849731445, "learning_rate": 4.3549088096210136e-05, "loss": 1.6597, "step": 2722 }, { "epoch": 0.024490383274947947, "grad_norm": 2.3960278034210205, "learning_rate": 4.3543544318468485e-05, "loss": 1.7274, "step": 2723 }, { "epoch": 0.024499377172588398, "grad_norm": 1.9079726934432983, "learning_rate": 4.35379985128327e-05, "loss": 1.6882, "step": 2724 }, { "epoch": 0.024508371070228848, "grad_norm": 2.9027903079986572, "learning_rate": 4.3532450679909274e-05, "loss": 1.7112, "step": 2725 }, { "epoch": 0.024517364967869302, "grad_norm": 3.331808090209961, "learning_rate": 4.352690082030491e-05, "loss": 1.7478, "step": 2726 }, { "epoch": 0.024526358865509752, "grad_norm": 2.569451332092285, "learning_rate": 4.352134893462651e-05, "loss": 1.8117, "step": 2727 }, { "epoch": 0.024535352763150203, "grad_norm": 1.740761160850525, "learning_rate": 4.351579502348125e-05, "loss": 1.7744, "step": 2728 }, { "epoch": 0.024544346660790653, "grad_norm": 3.0281341075897217, "learning_rate": 4.3510239087476476e-05, "loss": 1.6768, "step": 2729 }, { "epoch": 0.024553340558431103, "grad_norm": 1.3712553977966309, "learning_rate": 4.3504681127219793e-05, "loss": 1.8812, "step": 2730 }, { "epoch": 0.024562334456071557, "grad_norm": 1.936328649520874, "learning_rate": 4.3499121143319e-05, "loss": 1.6542, "step": 2731 }, { "epoch": 0.024571328353712008, "grad_norm": 2.285932779312134, "learning_rate": 4.349355913638213e-05, "loss": 1.6085, "step": 2732 }, { "epoch": 0.024580322251352458, "grad_norm": 2.2008676528930664, "learning_rate": 4.348799510701744e-05, "loss": 1.7388, "step": 2733 }, { "epoch": 0.024589316148992908, "grad_norm": 1.323438048362732, "learning_rate": 4.3482429055833405e-05, "loss": 1.7561, "step": 2734 }, { "epoch": 0.02459831004663336, "grad_norm": 4.418299198150635, "learning_rate": 4.3476860983438714e-05, "loss": 1.7973, "step": 2735 }, { "epoch": 0.02460730394427381, "grad_norm": 3.4917755126953125, "learning_rate": 4.3471290890442286e-05, "loss": 1.6536, "step": 2736 }, { "epoch": 0.024616297841914263, "grad_norm": 2.9080216884613037, "learning_rate": 4.346571877745326e-05, "loss": 1.905, "step": 2737 }, { "epoch": 0.024625291739554713, "grad_norm": 4.265514850616455, "learning_rate": 4.346014464508098e-05, "loss": 1.6622, "step": 2738 }, { "epoch": 0.024634285637195164, "grad_norm": 1.9536831378936768, "learning_rate": 4.345456849393505e-05, "loss": 1.7136, "step": 2739 }, { "epoch": 0.024643279534835614, "grad_norm": 2.3523261547088623, "learning_rate": 4.3448990324625244e-05, "loss": 1.7869, "step": 2740 }, { "epoch": 0.024652273432476064, "grad_norm": 1.9095805883407593, "learning_rate": 4.34434101377616e-05, "loss": 1.8827, "step": 2741 }, { "epoch": 0.024661267330116515, "grad_norm": 2.462618112564087, "learning_rate": 4.343782793395435e-05, "loss": 1.7248, "step": 2742 }, { "epoch": 0.02467026122775697, "grad_norm": 2.6443352699279785, "learning_rate": 4.3432243713813955e-05, "loss": 1.6825, "step": 2743 }, { "epoch": 0.02467925512539742, "grad_norm": 2.5617287158966064, "learning_rate": 4.3426657477951105e-05, "loss": 1.6399, "step": 2744 }, { "epoch": 0.02468824902303787, "grad_norm": 1.9001339673995972, "learning_rate": 4.342106922697669e-05, "loss": 1.7352, "step": 2745 }, { "epoch": 0.02469724292067832, "grad_norm": 3.5711119174957275, "learning_rate": 4.341547896150185e-05, "loss": 1.6971, "step": 2746 }, { "epoch": 0.02470623681831877, "grad_norm": 2.992297410964966, "learning_rate": 4.340988668213792e-05, "loss": 1.664, "step": 2747 }, { "epoch": 0.02471523071595922, "grad_norm": 2.993309259414673, "learning_rate": 4.340429238949646e-05, "loss": 1.7009, "step": 2748 }, { "epoch": 0.024724224613599674, "grad_norm": 5.6681227684021, "learning_rate": 4.3398696084189245e-05, "loss": 1.7374, "step": 2749 }, { "epoch": 0.024733218511240124, "grad_norm": 6.87704610824585, "learning_rate": 4.3393097766828293e-05, "loss": 1.7542, "step": 2750 }, { "epoch": 0.024742212408880575, "grad_norm": 1.837849497795105, "learning_rate": 4.3387497438025825e-05, "loss": 1.6283, "step": 2751 }, { "epoch": 0.024751206306521025, "grad_norm": 1.1293481588363647, "learning_rate": 4.338189509839427e-05, "loss": 1.6776, "step": 2752 }, { "epoch": 0.024760200204161475, "grad_norm": 1.8185453414916992, "learning_rate": 4.3376290748546325e-05, "loss": 1.7005, "step": 2753 }, { "epoch": 0.024769194101801926, "grad_norm": 2.2429208755493164, "learning_rate": 4.337068438909484e-05, "loss": 1.6415, "step": 2754 }, { "epoch": 0.02477818799944238, "grad_norm": 5.286375522613525, "learning_rate": 4.336507602065293e-05, "loss": 1.7327, "step": 2755 }, { "epoch": 0.02478718189708283, "grad_norm": 2.617262125015259, "learning_rate": 4.335946564383392e-05, "loss": 1.5991, "step": 2756 }, { "epoch": 0.02479617579472328, "grad_norm": 2.3139073848724365, "learning_rate": 4.335385325925135e-05, "loss": 1.7228, "step": 2757 }, { "epoch": 0.02480516969236373, "grad_norm": 2.427985906600952, "learning_rate": 4.3348238867518985e-05, "loss": 1.7015, "step": 2758 }, { "epoch": 0.02481416359000418, "grad_norm": 3.1698403358459473, "learning_rate": 4.33426224692508e-05, "loss": 1.7256, "step": 2759 }, { "epoch": 0.02482315748764463, "grad_norm": 3.854046583175659, "learning_rate": 4.3337004065061e-05, "loss": 1.8241, "step": 2760 }, { "epoch": 0.024832151385285085, "grad_norm": 2.7379086017608643, "learning_rate": 4.3331383655564006e-05, "loss": 1.6849, "step": 2761 }, { "epoch": 0.024841145282925536, "grad_norm": 2.5475165843963623, "learning_rate": 4.332576124137445e-05, "loss": 1.5789, "step": 2762 }, { "epoch": 0.024850139180565986, "grad_norm": 2.3458237648010254, "learning_rate": 4.3320136823107205e-05, "loss": 1.7062, "step": 2763 }, { "epoch": 0.024859133078206436, "grad_norm": 2.4060239791870117, "learning_rate": 4.331451040137734e-05, "loss": 1.8079, "step": 2764 }, { "epoch": 0.024868126975846887, "grad_norm": 2.2996158599853516, "learning_rate": 4.3308881976800146e-05, "loss": 1.7708, "step": 2765 }, { "epoch": 0.024877120873487337, "grad_norm": 2.803224802017212, "learning_rate": 4.330325154999115e-05, "loss": 1.809, "step": 2766 }, { "epoch": 0.02488611477112779, "grad_norm": 4.809998035430908, "learning_rate": 4.329761912156609e-05, "loss": 1.6844, "step": 2767 }, { "epoch": 0.02489510866876824, "grad_norm": 3.1024587154388428, "learning_rate": 4.3291984692140905e-05, "loss": 1.7259, "step": 2768 }, { "epoch": 0.02490410256640869, "grad_norm": 3.199727773666382, "learning_rate": 4.328634826233178e-05, "loss": 1.7236, "step": 2769 }, { "epoch": 0.024913096464049142, "grad_norm": 4.69028902053833, "learning_rate": 4.3280709832755094e-05, "loss": 1.7032, "step": 2770 }, { "epoch": 0.024922090361689592, "grad_norm": 2.6305277347564697, "learning_rate": 4.3275069404027476e-05, "loss": 1.7048, "step": 2771 }, { "epoch": 0.024931084259330046, "grad_norm": 3.320223808288574, "learning_rate": 4.3269426976765734e-05, "loss": 1.7237, "step": 2772 }, { "epoch": 0.024940078156970497, "grad_norm": 3.317105293273926, "learning_rate": 4.3263782551586935e-05, "loss": 1.5996, "step": 2773 }, { "epoch": 0.024949072054610947, "grad_norm": 2.9797630310058594, "learning_rate": 4.325813612910833e-05, "loss": 1.5264, "step": 2774 }, { "epoch": 0.024958065952251397, "grad_norm": 3.486140251159668, "learning_rate": 4.325248770994741e-05, "loss": 1.7109, "step": 2775 }, { "epoch": 0.024967059849891848, "grad_norm": 2.433651924133301, "learning_rate": 4.324683729472188e-05, "loss": 1.557, "step": 2776 }, { "epoch": 0.024976053747532298, "grad_norm": 3.214914560317993, "learning_rate": 4.324118488404965e-05, "loss": 1.8294, "step": 2777 }, { "epoch": 0.024985047645172752, "grad_norm": 3.7471632957458496, "learning_rate": 4.323553047854887e-05, "loss": 1.7279, "step": 2778 }, { "epoch": 0.024994041542813202, "grad_norm": 2.7553870677948, "learning_rate": 4.322987407883788e-05, "loss": 1.4804, "step": 2779 }, { "epoch": 0.025003035440453653, "grad_norm": 6.898322582244873, "learning_rate": 4.3224215685535294e-05, "loss": 1.7945, "step": 2780 }, { "epoch": 0.025012029338094103, "grad_norm": 3.224142551422119, "learning_rate": 4.321855529925987e-05, "loss": 1.6862, "step": 2781 }, { "epoch": 0.025021023235734553, "grad_norm": 2.4588396549224854, "learning_rate": 4.321289292063062e-05, "loss": 1.7307, "step": 2782 }, { "epoch": 0.025030017133375004, "grad_norm": 1.9574995040893555, "learning_rate": 4.32072285502668e-05, "loss": 1.7726, "step": 2783 }, { "epoch": 0.025039011031015457, "grad_norm": 1.9089206457138062, "learning_rate": 4.320156218878783e-05, "loss": 1.7139, "step": 2784 }, { "epoch": 0.025048004928655908, "grad_norm": 3.5203473567962646, "learning_rate": 4.319589383681338e-05, "loss": 1.7749, "step": 2785 }, { "epoch": 0.025056998826296358, "grad_norm": 2.8343000411987305, "learning_rate": 4.3190223494963345e-05, "loss": 1.5501, "step": 2786 }, { "epoch": 0.02506599272393681, "grad_norm": 3.569687604904175, "learning_rate": 4.318455116385781e-05, "loss": 1.7772, "step": 2787 }, { "epoch": 0.02507498662157726, "grad_norm": 2.116938352584839, "learning_rate": 4.31788768441171e-05, "loss": 1.6515, "step": 2788 }, { "epoch": 0.02508398051921771, "grad_norm": 1.9424132108688354, "learning_rate": 4.3173200536361756e-05, "loss": 1.5595, "step": 2789 }, { "epoch": 0.025092974416858163, "grad_norm": 6.397557735443115, "learning_rate": 4.316752224121252e-05, "loss": 1.7091, "step": 2790 }, { "epoch": 0.025101968314498613, "grad_norm": 3.2052409648895264, "learning_rate": 4.316184195929037e-05, "loss": 1.6989, "step": 2791 }, { "epoch": 0.025110962212139064, "grad_norm": 1.3720148801803589, "learning_rate": 4.315615969121649e-05, "loss": 1.6372, "step": 2792 }, { "epoch": 0.025119956109779514, "grad_norm": 4.857582092285156, "learning_rate": 4.3150475437612286e-05, "loss": 1.7294, "step": 2793 }, { "epoch": 0.025128950007419965, "grad_norm": 2.0058066844940186, "learning_rate": 4.3144789199099374e-05, "loss": 1.7422, "step": 2794 }, { "epoch": 0.025137943905060415, "grad_norm": 2.088116407394409, "learning_rate": 4.313910097629959e-05, "loss": 1.6902, "step": 2795 }, { "epoch": 0.02514693780270087, "grad_norm": 2.153123617172241, "learning_rate": 4.3133410769835005e-05, "loss": 1.8027, "step": 2796 }, { "epoch": 0.02515593170034132, "grad_norm": 2.83589243888855, "learning_rate": 4.312771858032787e-05, "loss": 1.7022, "step": 2797 }, { "epoch": 0.02516492559798177, "grad_norm": 2.9059572219848633, "learning_rate": 4.3122024408400693e-05, "loss": 1.7693, "step": 2798 }, { "epoch": 0.02517391949562222, "grad_norm": 3.4941680431365967, "learning_rate": 4.311632825467617e-05, "loss": 1.9235, "step": 2799 }, { "epoch": 0.02518291339326267, "grad_norm": 2.2177064418792725, "learning_rate": 4.311063011977723e-05, "loss": 1.6299, "step": 2800 }, { "epoch": 0.02519190729090312, "grad_norm": 5.249582290649414, "learning_rate": 4.310493000432701e-05, "loss": 1.9723, "step": 2801 }, { "epoch": 0.025200901188543574, "grad_norm": 2.8511924743652344, "learning_rate": 4.309922790894885e-05, "loss": 1.8407, "step": 2802 }, { "epoch": 0.025209895086184025, "grad_norm": 2.8121001720428467, "learning_rate": 4.309352383426635e-05, "loss": 1.7178, "step": 2803 }, { "epoch": 0.025218888983824475, "grad_norm": 2.1476047039031982, "learning_rate": 4.308781778090329e-05, "loss": 1.7517, "step": 2804 }, { "epoch": 0.025227882881464925, "grad_norm": 3.5551021099090576, "learning_rate": 4.308210974948367e-05, "loss": 1.6734, "step": 2805 }, { "epoch": 0.025236876779105376, "grad_norm": 2.1763803958892822, "learning_rate": 4.3076399740631715e-05, "loss": 1.7513, "step": 2806 }, { "epoch": 0.02524587067674583, "grad_norm": 3.0512073040008545, "learning_rate": 4.307068775497185e-05, "loss": 1.7667, "step": 2807 }, { "epoch": 0.02525486457438628, "grad_norm": 1.4659092426300049, "learning_rate": 4.306497379312875e-05, "loss": 1.6603, "step": 2808 }, { "epoch": 0.02526385847202673, "grad_norm": 6.663466930389404, "learning_rate": 4.305925785572728e-05, "loss": 1.7397, "step": 2809 }, { "epoch": 0.02527285236966718, "grad_norm": 2.738049030303955, "learning_rate": 4.305353994339252e-05, "loss": 1.6595, "step": 2810 }, { "epoch": 0.02528184626730763, "grad_norm": 3.4130520820617676, "learning_rate": 4.3047820056749764e-05, "loss": 1.6664, "step": 2811 }, { "epoch": 0.02529084016494808, "grad_norm": 2.26273775100708, "learning_rate": 4.304209819642455e-05, "loss": 1.6309, "step": 2812 }, { "epoch": 0.025299834062588535, "grad_norm": 2.139875650405884, "learning_rate": 4.3036374363042606e-05, "loss": 1.5278, "step": 2813 }, { "epoch": 0.025308827960228986, "grad_norm": 2.195096731185913, "learning_rate": 4.303064855722988e-05, "loss": 1.679, "step": 2814 }, { "epoch": 0.025317821857869436, "grad_norm": 1.9827836751937866, "learning_rate": 4.302492077961253e-05, "loss": 1.6694, "step": 2815 }, { "epoch": 0.025326815755509886, "grad_norm": 1.255738615989685, "learning_rate": 4.3019191030816954e-05, "loss": 1.6541, "step": 2816 }, { "epoch": 0.025335809653150337, "grad_norm": 6.214998245239258, "learning_rate": 4.301345931146973e-05, "loss": 1.7036, "step": 2817 }, { "epoch": 0.025344803550790787, "grad_norm": 3.0498509407043457, "learning_rate": 4.3007725622197674e-05, "loss": 1.637, "step": 2818 }, { "epoch": 0.02535379744843124, "grad_norm": 1.7197359800338745, "learning_rate": 4.300198996362783e-05, "loss": 1.7749, "step": 2819 }, { "epoch": 0.02536279134607169, "grad_norm": 2.155181407928467, "learning_rate": 4.2996252336387414e-05, "loss": 1.7, "step": 2820 }, { "epoch": 0.02537178524371214, "grad_norm": 2.1593902111053467, "learning_rate": 4.299051274110391e-05, "loss": 1.7201, "step": 2821 }, { "epoch": 0.025380779141352592, "grad_norm": 2.8470888137817383, "learning_rate": 4.298477117840498e-05, "loss": 1.5695, "step": 2822 }, { "epoch": 0.025389773038993042, "grad_norm": 2.1279685497283936, "learning_rate": 4.297902764891851e-05, "loss": 1.6252, "step": 2823 }, { "epoch": 0.025398766936633493, "grad_norm": 2.616342306137085, "learning_rate": 4.297328215327261e-05, "loss": 1.7168, "step": 2824 }, { "epoch": 0.025407760834273947, "grad_norm": 2.5890161991119385, "learning_rate": 4.29675346920956e-05, "loss": 1.585, "step": 2825 }, { "epoch": 0.025416754731914397, "grad_norm": 1.4861223697662354, "learning_rate": 4.296178526601601e-05, "loss": 1.6729, "step": 2826 }, { "epoch": 0.025425748629554847, "grad_norm": 4.568517684936523, "learning_rate": 4.295603387566258e-05, "loss": 1.62, "step": 2827 }, { "epoch": 0.025434742527195298, "grad_norm": 3.3383116722106934, "learning_rate": 4.295028052166429e-05, "loss": 1.6989, "step": 2828 }, { "epoch": 0.025443736424835748, "grad_norm": 1.8614529371261597, "learning_rate": 4.2944525204650306e-05, "loss": 1.6787, "step": 2829 }, { "epoch": 0.0254527303224762, "grad_norm": 2.4682905673980713, "learning_rate": 4.293876792525002e-05, "loss": 1.7335, "step": 2830 }, { "epoch": 0.025461724220116652, "grad_norm": 2.2520463466644287, "learning_rate": 4.2933008684093054e-05, "loss": 1.5872, "step": 2831 }, { "epoch": 0.025470718117757103, "grad_norm": 1.3687363862991333, "learning_rate": 4.292724748180921e-05, "loss": 1.6317, "step": 2832 }, { "epoch": 0.025479712015397553, "grad_norm": 6.998400688171387, "learning_rate": 4.2921484319028544e-05, "loss": 1.6309, "step": 2833 }, { "epoch": 0.025488705913038003, "grad_norm": 2.5757572650909424, "learning_rate": 4.2915719196381294e-05, "loss": 1.7085, "step": 2834 }, { "epoch": 0.025497699810678454, "grad_norm": 2.0726234912872314, "learning_rate": 4.2909952114497925e-05, "loss": 1.7827, "step": 2835 }, { "epoch": 0.025506693708318904, "grad_norm": 1.6734914779663086, "learning_rate": 4.290418307400913e-05, "loss": 1.7499, "step": 2836 }, { "epoch": 0.025515687605959358, "grad_norm": 3.4946744441986084, "learning_rate": 4.289841207554578e-05, "loss": 1.7409, "step": 2837 }, { "epoch": 0.025524681503599808, "grad_norm": 2.5932159423828125, "learning_rate": 4.2892639119739e-05, "loss": 1.6819, "step": 2838 }, { "epoch": 0.02553367540124026, "grad_norm": 3.5279123783111572, "learning_rate": 4.288686420722011e-05, "loss": 1.86, "step": 2839 }, { "epoch": 0.02554266929888071, "grad_norm": 3.061887264251709, "learning_rate": 4.288108733862064e-05, "loss": 1.72, "step": 2840 }, { "epoch": 0.02555166319652116, "grad_norm": 2.246880054473877, "learning_rate": 4.287530851457234e-05, "loss": 1.7031, "step": 2841 }, { "epoch": 0.025560657094161613, "grad_norm": 2.1920242309570312, "learning_rate": 4.2869527735707184e-05, "loss": 1.8068, "step": 2842 }, { "epoch": 0.025569650991802063, "grad_norm": 1.857616901397705, "learning_rate": 4.286374500265734e-05, "loss": 1.5058, "step": 2843 }, { "epoch": 0.025578644889442514, "grad_norm": 6.513493061065674, "learning_rate": 4.285796031605519e-05, "loss": 1.6936, "step": 2844 }, { "epoch": 0.025587638787082964, "grad_norm": 2.9064321517944336, "learning_rate": 4.2852173676533356e-05, "loss": 1.765, "step": 2845 }, { "epoch": 0.025596632684723414, "grad_norm": 2.111734390258789, "learning_rate": 4.284638508472465e-05, "loss": 1.6323, "step": 2846 }, { "epoch": 0.025605626582363865, "grad_norm": 2.9215869903564453, "learning_rate": 4.2840594541262105e-05, "loss": 1.807, "step": 2847 }, { "epoch": 0.02561462048000432, "grad_norm": 3.4544436931610107, "learning_rate": 4.283480204677896e-05, "loss": 1.7649, "step": 2848 }, { "epoch": 0.02562361437764477, "grad_norm": 1.9434356689453125, "learning_rate": 4.282900760190868e-05, "loss": 1.6285, "step": 2849 }, { "epoch": 0.02563260827528522, "grad_norm": 1.6531994342803955, "learning_rate": 4.282321120728493e-05, "loss": 1.8504, "step": 2850 }, { "epoch": 0.02564160217292567, "grad_norm": 1.763904333114624, "learning_rate": 4.28174128635416e-05, "loss": 1.6405, "step": 2851 }, { "epoch": 0.02565059607056612, "grad_norm": 2.101077079772949, "learning_rate": 4.281161257131279e-05, "loss": 1.576, "step": 2852 }, { "epoch": 0.02565958996820657, "grad_norm": 2.1988229751586914, "learning_rate": 4.280581033123281e-05, "loss": 1.7509, "step": 2853 }, { "epoch": 0.025668583865847024, "grad_norm": 1.557451844215393, "learning_rate": 4.280000614393618e-05, "loss": 1.6984, "step": 2854 }, { "epoch": 0.025677577763487475, "grad_norm": 4.28115177154541, "learning_rate": 4.279420001005764e-05, "loss": 1.6831, "step": 2855 }, { "epoch": 0.025686571661127925, "grad_norm": 2.81289005279541, "learning_rate": 4.278839193023214e-05, "loss": 1.7563, "step": 2856 }, { "epoch": 0.025695565558768375, "grad_norm": 3.147768020629883, "learning_rate": 4.278258190509484e-05, "loss": 1.6951, "step": 2857 }, { "epoch": 0.025704559456408826, "grad_norm": 2.160245180130005, "learning_rate": 4.277676993528112e-05, "loss": 1.6124, "step": 2858 }, { "epoch": 0.025713553354049276, "grad_norm": 1.4875447750091553, "learning_rate": 4.2770956021426566e-05, "loss": 1.6319, "step": 2859 }, { "epoch": 0.02572254725168973, "grad_norm": 1.9646713733673096, "learning_rate": 4.2765140164166984e-05, "loss": 1.781, "step": 2860 }, { "epoch": 0.02573154114933018, "grad_norm": 3.8731870651245117, "learning_rate": 4.275932236413838e-05, "loss": 1.691, "step": 2861 }, { "epoch": 0.02574053504697063, "grad_norm": 2.474118947982788, "learning_rate": 4.2753502621976986e-05, "loss": 1.6559, "step": 2862 }, { "epoch": 0.02574952894461108, "grad_norm": 3.105426788330078, "learning_rate": 4.2747680938319236e-05, "loss": 1.6917, "step": 2863 }, { "epoch": 0.02575852284225153, "grad_norm": 3.4391956329345703, "learning_rate": 4.274185731380178e-05, "loss": 1.6653, "step": 2864 }, { "epoch": 0.025767516739891982, "grad_norm": 2.4891295433044434, "learning_rate": 4.273603174906149e-05, "loss": 1.7006, "step": 2865 }, { "epoch": 0.025776510637532436, "grad_norm": 2.038447380065918, "learning_rate": 4.273020424473543e-05, "loss": 1.5138, "step": 2866 }, { "epoch": 0.025785504535172886, "grad_norm": 2.127440929412842, "learning_rate": 4.272437480146088e-05, "loss": 1.6458, "step": 2867 }, { "epoch": 0.025794498432813336, "grad_norm": 2.229107141494751, "learning_rate": 4.2718543419875355e-05, "loss": 1.8509, "step": 2868 }, { "epoch": 0.025803492330453787, "grad_norm": 3.9651129245758057, "learning_rate": 4.271271010061657e-05, "loss": 1.6216, "step": 2869 }, { "epoch": 0.025812486228094237, "grad_norm": 2.280771017074585, "learning_rate": 4.270687484432243e-05, "loss": 1.6809, "step": 2870 }, { "epoch": 0.025821480125734687, "grad_norm": 3.568241596221924, "learning_rate": 4.2701037651631084e-05, "loss": 1.7389, "step": 2871 }, { "epoch": 0.02583047402337514, "grad_norm": 5.724792957305908, "learning_rate": 4.2695198523180876e-05, "loss": 1.7017, "step": 2872 }, { "epoch": 0.02583946792101559, "grad_norm": 5.889983654022217, "learning_rate": 4.268935745961036e-05, "loss": 1.6895, "step": 2873 }, { "epoch": 0.025848461818656042, "grad_norm": 3.1376609802246094, "learning_rate": 4.26835144615583e-05, "loss": 1.7368, "step": 2874 }, { "epoch": 0.025857455716296492, "grad_norm": 2.560459613800049, "learning_rate": 4.267766952966369e-05, "loss": 1.6564, "step": 2875 }, { "epoch": 0.025866449613936943, "grad_norm": 2.318441390991211, "learning_rate": 4.267182266456572e-05, "loss": 1.8414, "step": 2876 }, { "epoch": 0.025875443511577396, "grad_norm": 2.823701858520508, "learning_rate": 4.266597386690379e-05, "loss": 1.6363, "step": 2877 }, { "epoch": 0.025884437409217847, "grad_norm": 1.815056562423706, "learning_rate": 4.2660123137317523e-05, "loss": 1.7113, "step": 2878 }, { "epoch": 0.025893431306858297, "grad_norm": 3.78694486618042, "learning_rate": 4.2654270476446736e-05, "loss": 1.9038, "step": 2879 }, { "epoch": 0.025902425204498748, "grad_norm": 1.9953619241714478, "learning_rate": 4.2648415884931476e-05, "loss": 1.6318, "step": 2880 }, { "epoch": 0.025911419102139198, "grad_norm": 3.294437885284424, "learning_rate": 4.264255936341199e-05, "loss": 1.5578, "step": 2881 }, { "epoch": 0.02592041299977965, "grad_norm": 2.077552318572998, "learning_rate": 4.263670091252873e-05, "loss": 1.8045, "step": 2882 }, { "epoch": 0.025929406897420102, "grad_norm": 2.5097010135650635, "learning_rate": 4.263084053292238e-05, "loss": 1.8494, "step": 2883 }, { "epoch": 0.025938400795060552, "grad_norm": 9.004868507385254, "learning_rate": 4.262497822523381e-05, "loss": 1.9141, "step": 2884 }, { "epoch": 0.025947394692701003, "grad_norm": 1.417307734489441, "learning_rate": 4.261911399010413e-05, "loss": 1.6894, "step": 2885 }, { "epoch": 0.025956388590341453, "grad_norm": 3.7236762046813965, "learning_rate": 4.2613247828174635e-05, "loss": 1.6542, "step": 2886 }, { "epoch": 0.025965382487981904, "grad_norm": 2.3737432956695557, "learning_rate": 4.260737974008683e-05, "loss": 1.6568, "step": 2887 }, { "epoch": 0.025974376385622354, "grad_norm": 1.5631675720214844, "learning_rate": 4.260150972648246e-05, "loss": 1.8038, "step": 2888 }, { "epoch": 0.025983370283262808, "grad_norm": 2.029736280441284, "learning_rate": 4.259563778800344e-05, "loss": 1.6871, "step": 2889 }, { "epoch": 0.025992364180903258, "grad_norm": 2.8162574768066406, "learning_rate": 4.258976392529192e-05, "loss": 1.7864, "step": 2890 }, { "epoch": 0.02600135807854371, "grad_norm": 3.2222707271575928, "learning_rate": 4.258388813899028e-05, "loss": 1.5746, "step": 2891 }, { "epoch": 0.02601035197618416, "grad_norm": 2.121157646179199, "learning_rate": 4.257801042974105e-05, "loss": 1.7707, "step": 2892 }, { "epoch": 0.02601934587382461, "grad_norm": 3.041001081466675, "learning_rate": 4.257213079818705e-05, "loss": 1.7781, "step": 2893 }, { "epoch": 0.02602833977146506, "grad_norm": 2.6054534912109375, "learning_rate": 4.256624924497123e-05, "loss": 1.7207, "step": 2894 }, { "epoch": 0.026037333669105513, "grad_norm": 2.0752742290496826, "learning_rate": 4.256036577073681e-05, "loss": 1.6553, "step": 2895 }, { "epoch": 0.026046327566745964, "grad_norm": 1.816106915473938, "learning_rate": 4.255448037612719e-05, "loss": 1.5808, "step": 2896 }, { "epoch": 0.026055321464386414, "grad_norm": 4.710414886474609, "learning_rate": 4.254859306178599e-05, "loss": 1.5881, "step": 2897 }, { "epoch": 0.026064315362026864, "grad_norm": 2.282917022705078, "learning_rate": 4.254270382835703e-05, "loss": 1.6007, "step": 2898 }, { "epoch": 0.026073309259667315, "grad_norm": 2.044408082962036, "learning_rate": 4.2536812676484356e-05, "loss": 1.6319, "step": 2899 }, { "epoch": 0.026082303157307765, "grad_norm": 2.7517340183258057, "learning_rate": 4.2530919606812216e-05, "loss": 1.6675, "step": 2900 }, { "epoch": 0.02609129705494822, "grad_norm": 2.547461986541748, "learning_rate": 4.2525024619985064e-05, "loss": 1.6657, "step": 2901 }, { "epoch": 0.02610029095258867, "grad_norm": 3.10361385345459, "learning_rate": 4.251912771664758e-05, "loss": 1.8708, "step": 2902 }, { "epoch": 0.02610928485022912, "grad_norm": 2.738084554672241, "learning_rate": 4.251322889744461e-05, "loss": 1.583, "step": 2903 }, { "epoch": 0.02611827874786957, "grad_norm": 2.86776065826416, "learning_rate": 4.250732816302127e-05, "loss": 1.728, "step": 2904 }, { "epoch": 0.02612727264551002, "grad_norm": 1.801862120628357, "learning_rate": 4.250142551402284e-05, "loss": 1.7695, "step": 2905 }, { "epoch": 0.02613626654315047, "grad_norm": 1.7925031185150146, "learning_rate": 4.249552095109482e-05, "loss": 1.715, "step": 2906 }, { "epoch": 0.026145260440790925, "grad_norm": 2.917527437210083, "learning_rate": 4.248961447488295e-05, "loss": 1.5379, "step": 2907 }, { "epoch": 0.026154254338431375, "grad_norm": 4.132024765014648, "learning_rate": 4.248370608603313e-05, "loss": 1.7577, "step": 2908 }, { "epoch": 0.026163248236071825, "grad_norm": 2.113306760787964, "learning_rate": 4.247779578519149e-05, "loss": 1.7852, "step": 2909 }, { "epoch": 0.026172242133712276, "grad_norm": 1.9893097877502441, "learning_rate": 4.247188357300439e-05, "loss": 1.6581, "step": 2910 }, { "epoch": 0.026181236031352726, "grad_norm": 2.4478092193603516, "learning_rate": 4.246596945011837e-05, "loss": 1.7363, "step": 2911 }, { "epoch": 0.026190229928993176, "grad_norm": 2.5943756103515625, "learning_rate": 4.246005341718019e-05, "loss": 1.7141, "step": 2912 }, { "epoch": 0.02619922382663363, "grad_norm": 2.2285139560699463, "learning_rate": 4.245413547483682e-05, "loss": 1.7266, "step": 2913 }, { "epoch": 0.02620821772427408, "grad_norm": 4.3628692626953125, "learning_rate": 4.244821562373543e-05, "loss": 1.6649, "step": 2914 }, { "epoch": 0.02621721162191453, "grad_norm": 1.6705400943756104, "learning_rate": 4.244229386452342e-05, "loss": 1.6256, "step": 2915 }, { "epoch": 0.02622620551955498, "grad_norm": 1.8481378555297852, "learning_rate": 4.2436370197848374e-05, "loss": 1.6777, "step": 2916 }, { "epoch": 0.02623519941719543, "grad_norm": 3.035132884979248, "learning_rate": 4.24304446243581e-05, "loss": 1.6971, "step": 2917 }, { "epoch": 0.026244193314835886, "grad_norm": 4.072854518890381, "learning_rate": 4.2424517144700603e-05, "loss": 1.616, "step": 2918 }, { "epoch": 0.026253187212476336, "grad_norm": 3.464848518371582, "learning_rate": 4.2418587759524115e-05, "loss": 1.6014, "step": 2919 }, { "epoch": 0.026262181110116786, "grad_norm": 2.156299114227295, "learning_rate": 4.241265646947706e-05, "loss": 1.748, "step": 2920 }, { "epoch": 0.026271175007757237, "grad_norm": 2.2830607891082764, "learning_rate": 4.240672327520806e-05, "loss": 1.6822, "step": 2921 }, { "epoch": 0.026280168905397687, "grad_norm": 3.136329412460327, "learning_rate": 4.2400788177365985e-05, "loss": 1.6199, "step": 2922 }, { "epoch": 0.026289162803038137, "grad_norm": 2.3878583908081055, "learning_rate": 4.239485117659987e-05, "loss": 1.6465, "step": 2923 }, { "epoch": 0.02629815670067859, "grad_norm": 2.104321002960205, "learning_rate": 4.238891227355899e-05, "loss": 1.7071, "step": 2924 }, { "epoch": 0.02630715059831904, "grad_norm": 1.4780563116073608, "learning_rate": 4.2382971468892806e-05, "loss": 1.8393, "step": 2925 }, { "epoch": 0.026316144495959492, "grad_norm": 2.5206820964813232, "learning_rate": 4.2377028763250995e-05, "loss": 1.6628, "step": 2926 }, { "epoch": 0.026325138393599942, "grad_norm": 1.8609180450439453, "learning_rate": 4.237108415728345e-05, "loss": 1.7323, "step": 2927 }, { "epoch": 0.026334132291240393, "grad_norm": 3.2116687297821045, "learning_rate": 4.236513765164025e-05, "loss": 1.7328, "step": 2928 }, { "epoch": 0.026343126188880843, "grad_norm": 3.40968656539917, "learning_rate": 4.23591892469717e-05, "loss": 1.6279, "step": 2929 }, { "epoch": 0.026352120086521297, "grad_norm": 1.9361984729766846, "learning_rate": 4.235323894392832e-05, "loss": 1.7642, "step": 2930 }, { "epoch": 0.026361113984161747, "grad_norm": 1.9279608726501465, "learning_rate": 4.2347286743160825e-05, "loss": 1.6794, "step": 2931 }, { "epoch": 0.026370107881802198, "grad_norm": 5.222457408905029, "learning_rate": 4.234133264532012e-05, "loss": 1.7592, "step": 2932 }, { "epoch": 0.026379101779442648, "grad_norm": 2.655424118041992, "learning_rate": 4.233537665105737e-05, "loss": 1.6288, "step": 2933 }, { "epoch": 0.026388095677083098, "grad_norm": 2.3905346393585205, "learning_rate": 4.232941876102388e-05, "loss": 1.6585, "step": 2934 }, { "epoch": 0.02639708957472355, "grad_norm": 1.5977022647857666, "learning_rate": 4.23234589758712e-05, "loss": 1.6695, "step": 2935 }, { "epoch": 0.026406083472364002, "grad_norm": 2.3519976139068604, "learning_rate": 4.231749729625111e-05, "loss": 1.625, "step": 2936 }, { "epoch": 0.026415077370004453, "grad_norm": 3.062587022781372, "learning_rate": 4.231153372281554e-05, "loss": 1.6741, "step": 2937 }, { "epoch": 0.026424071267644903, "grad_norm": 2.548214912414551, "learning_rate": 4.230556825621667e-05, "loss": 1.6761, "step": 2938 }, { "epoch": 0.026433065165285353, "grad_norm": 2.345271110534668, "learning_rate": 4.229960089710687e-05, "loss": 1.7241, "step": 2939 }, { "epoch": 0.026442059062925804, "grad_norm": 6.123969078063965, "learning_rate": 4.229363164613874e-05, "loss": 1.8281, "step": 2940 }, { "epoch": 0.026451052960566254, "grad_norm": 3.243075370788574, "learning_rate": 4.228766050396504e-05, "loss": 1.8026, "step": 2941 }, { "epoch": 0.026460046858206708, "grad_norm": 2.48995041847229, "learning_rate": 4.228168747123879e-05, "loss": 1.5331, "step": 2942 }, { "epoch": 0.02646904075584716, "grad_norm": 3.1280791759490967, "learning_rate": 4.227571254861317e-05, "loss": 1.651, "step": 2943 }, { "epoch": 0.02647803465348761, "grad_norm": 2.6898012161254883, "learning_rate": 4.22697357367416e-05, "loss": 1.6803, "step": 2944 }, { "epoch": 0.02648702855112806, "grad_norm": 2.5289742946624756, "learning_rate": 4.2263757036277705e-05, "loss": 1.636, "step": 2945 }, { "epoch": 0.02649602244876851, "grad_norm": 1.4418766498565674, "learning_rate": 4.2257776447875286e-05, "loss": 1.7063, "step": 2946 }, { "epoch": 0.02650501634640896, "grad_norm": 2.4133663177490234, "learning_rate": 4.225179397218839e-05, "loss": 1.7033, "step": 2947 }, { "epoch": 0.026514010244049414, "grad_norm": 2.509361505508423, "learning_rate": 4.2245809609871236e-05, "loss": 1.6612, "step": 2948 }, { "epoch": 0.026523004141689864, "grad_norm": 2.54386043548584, "learning_rate": 4.223982336157828e-05, "loss": 1.5708, "step": 2949 }, { "epoch": 0.026531998039330314, "grad_norm": 2.6849968433380127, "learning_rate": 4.223383522796415e-05, "loss": 1.6321, "step": 2950 }, { "epoch": 0.026540991936970765, "grad_norm": 3.1872901916503906, "learning_rate": 4.2227845209683716e-05, "loss": 1.7279, "step": 2951 }, { "epoch": 0.026549985834611215, "grad_norm": 2.606235980987549, "learning_rate": 4.222185330739203e-05, "loss": 1.7043, "step": 2952 }, { "epoch": 0.02655897973225167, "grad_norm": 3.3207244873046875, "learning_rate": 4.221585952174436e-05, "loss": 1.8861, "step": 2953 }, { "epoch": 0.02656797362989212, "grad_norm": 1.4059269428253174, "learning_rate": 4.220986385339617e-05, "loss": 1.677, "step": 2954 }, { "epoch": 0.02657696752753257, "grad_norm": 3.4107065200805664, "learning_rate": 4.220386630300315e-05, "loss": 1.6774, "step": 2955 }, { "epoch": 0.02658596142517302, "grad_norm": 2.82779598236084, "learning_rate": 4.2197866871221184e-05, "loss": 1.7446, "step": 2956 }, { "epoch": 0.02659495532281347, "grad_norm": 2.0624027252197266, "learning_rate": 4.219186555870634e-05, "loss": 1.7228, "step": 2957 }, { "epoch": 0.02660394922045392, "grad_norm": 3.085062026977539, "learning_rate": 4.2185862366114936e-05, "loss": 1.7607, "step": 2958 }, { "epoch": 0.026612943118094375, "grad_norm": 2.6240010261535645, "learning_rate": 4.217985729410346e-05, "loss": 1.6466, "step": 2959 }, { "epoch": 0.026621937015734825, "grad_norm": 4.042247295379639, "learning_rate": 4.217385034332861e-05, "loss": 1.5581, "step": 2960 }, { "epoch": 0.026630930913375275, "grad_norm": 3.4228081703186035, "learning_rate": 4.216784151444732e-05, "loss": 1.7692, "step": 2961 }, { "epoch": 0.026639924811015726, "grad_norm": 5.954885482788086, "learning_rate": 4.2161830808116685e-05, "loss": 1.805, "step": 2962 }, { "epoch": 0.026648918708656176, "grad_norm": 2.7099056243896484, "learning_rate": 4.215581822499404e-05, "loss": 1.7508, "step": 2963 }, { "epoch": 0.026657912606296626, "grad_norm": 1.7408610582351685, "learning_rate": 4.214980376573691e-05, "loss": 1.8003, "step": 2964 }, { "epoch": 0.02666690650393708, "grad_norm": 1.7443188428878784, "learning_rate": 4.214378743100302e-05, "loss": 1.7997, "step": 2965 }, { "epoch": 0.02667590040157753, "grad_norm": 2.595839262008667, "learning_rate": 4.213776922145031e-05, "loss": 1.6646, "step": 2966 }, { "epoch": 0.02668489429921798, "grad_norm": 2.699599504470825, "learning_rate": 4.2131749137736925e-05, "loss": 1.6302, "step": 2967 }, { "epoch": 0.02669388819685843, "grad_norm": 2.8029770851135254, "learning_rate": 4.212572718052121e-05, "loss": 1.7799, "step": 2968 }, { "epoch": 0.02670288209449888, "grad_norm": 2.0166759490966797, "learning_rate": 4.211970335046172e-05, "loss": 1.6985, "step": 2969 }, { "epoch": 0.026711875992139332, "grad_norm": 2.8542943000793457, "learning_rate": 4.211367764821722e-05, "loss": 1.6936, "step": 2970 }, { "epoch": 0.026720869889779786, "grad_norm": 2.069140911102295, "learning_rate": 4.210765007444666e-05, "loss": 1.6968, "step": 2971 }, { "epoch": 0.026729863787420236, "grad_norm": 1.6795144081115723, "learning_rate": 4.2101620629809205e-05, "loss": 1.6196, "step": 2972 }, { "epoch": 0.026738857685060687, "grad_norm": 2.063993215560913, "learning_rate": 4.209558931496422e-05, "loss": 1.6508, "step": 2973 }, { "epoch": 0.026747851582701137, "grad_norm": 2.37433123588562, "learning_rate": 4.20895561305713e-05, "loss": 1.6218, "step": 2974 }, { "epoch": 0.026756845480341587, "grad_norm": 2.76873517036438, "learning_rate": 4.2083521077290213e-05, "loss": 1.6782, "step": 2975 }, { "epoch": 0.026765839377982038, "grad_norm": 4.737286567687988, "learning_rate": 4.207748415578094e-05, "loss": 1.7314, "step": 2976 }, { "epoch": 0.02677483327562249, "grad_norm": 1.4264144897460938, "learning_rate": 4.2071445366703685e-05, "loss": 1.5756, "step": 2977 }, { "epoch": 0.026783827173262942, "grad_norm": 2.3990767002105713, "learning_rate": 4.2065404710718834e-05, "loss": 1.738, "step": 2978 }, { "epoch": 0.026792821070903392, "grad_norm": 2.1576831340789795, "learning_rate": 4.205936218848697e-05, "loss": 1.6556, "step": 2979 }, { "epoch": 0.026801814968543843, "grad_norm": 1.7269436120986938, "learning_rate": 4.205331780066892e-05, "loss": 1.6211, "step": 2980 }, { "epoch": 0.026810808866184293, "grad_norm": 5.993425369262695, "learning_rate": 4.204727154792566e-05, "loss": 1.6679, "step": 2981 }, { "epoch": 0.026819802763824743, "grad_norm": 2.7868492603302, "learning_rate": 4.204122343091842e-05, "loss": 1.8239, "step": 2982 }, { "epoch": 0.026828796661465197, "grad_norm": 3.3171942234039307, "learning_rate": 4.203517345030861e-05, "loss": 1.6214, "step": 2983 }, { "epoch": 0.026837790559105647, "grad_norm": 2.8388569355010986, "learning_rate": 4.202912160675784e-05, "loss": 1.7651, "step": 2984 }, { "epoch": 0.026846784456746098, "grad_norm": 3.035346746444702, "learning_rate": 4.202306790092792e-05, "loss": 1.6029, "step": 2985 }, { "epoch": 0.026855778354386548, "grad_norm": 1.9770824909210205, "learning_rate": 4.201701233348091e-05, "loss": 1.6852, "step": 2986 }, { "epoch": 0.026864772252027, "grad_norm": 4.480726718902588, "learning_rate": 4.201095490507901e-05, "loss": 1.8206, "step": 2987 }, { "epoch": 0.026873766149667452, "grad_norm": 2.75329852104187, "learning_rate": 4.200489561638466e-05, "loss": 1.6006, "step": 2988 }, { "epoch": 0.026882760047307903, "grad_norm": 1.99849534034729, "learning_rate": 4.199883446806048e-05, "loss": 1.7082, "step": 2989 }, { "epoch": 0.026891753944948353, "grad_norm": 2.488285541534424, "learning_rate": 4.199277146076933e-05, "loss": 1.7458, "step": 2990 }, { "epoch": 0.026900747842588803, "grad_norm": 1.409359097480774, "learning_rate": 4.198670659517424e-05, "loss": 1.7433, "step": 2991 }, { "epoch": 0.026909741740229254, "grad_norm": 1.7848979234695435, "learning_rate": 4.198063987193845e-05, "loss": 1.6649, "step": 2992 }, { "epoch": 0.026918735637869704, "grad_norm": 1.6708652973175049, "learning_rate": 4.1974571291725415e-05, "loss": 1.6353, "step": 2993 }, { "epoch": 0.026927729535510158, "grad_norm": 2.045742988586426, "learning_rate": 4.196850085519879e-05, "loss": 1.7104, "step": 2994 }, { "epoch": 0.02693672343315061, "grad_norm": 2.8486688137054443, "learning_rate": 4.1962428563022414e-05, "loss": 1.6407, "step": 2995 }, { "epoch": 0.02694571733079106, "grad_norm": 2.2256722450256348, "learning_rate": 4.1956354415860364e-05, "loss": 1.6705, "step": 2996 }, { "epoch": 0.02695471122843151, "grad_norm": 2.301121711730957, "learning_rate": 4.1950278414376883e-05, "loss": 1.6867, "step": 2997 }, { "epoch": 0.02696370512607196, "grad_norm": 1.8929758071899414, "learning_rate": 4.194420055923644e-05, "loss": 1.679, "step": 2998 }, { "epoch": 0.02697269902371241, "grad_norm": 3.7899744510650635, "learning_rate": 4.19381208511037e-05, "loss": 1.5678, "step": 2999 }, { "epoch": 0.026981692921352864, "grad_norm": 1.5454990863800049, "learning_rate": 4.193203929064353e-05, "loss": 1.7698, "step": 3000 }, { "epoch": 0.026990686818993314, "grad_norm": 135.7724151611328, "learning_rate": 4.192595587852099e-05, "loss": 6.3536, "step": 3001 }, { "epoch": 0.026999680716633764, "grad_norm": 54.463191986083984, "learning_rate": 4.191987061540138e-05, "loss": 6.0208, "step": 3002 }, { "epoch": 0.027008674614274215, "grad_norm": 36.4810791015625, "learning_rate": 4.191378350195015e-05, "loss": 5.279, "step": 3003 }, { "epoch": 0.027017668511914665, "grad_norm": 60.57801818847656, "learning_rate": 4.190769453883298e-05, "loss": 6.0139, "step": 3004 }, { "epoch": 0.027026662409555115, "grad_norm": 50.58754348754883, "learning_rate": 4.1901603726715766e-05, "loss": 4.8668, "step": 3005 }, { "epoch": 0.02703565630719557, "grad_norm": 84.71070098876953, "learning_rate": 4.1895511066264576e-05, "loss": 4.764, "step": 3006 }, { "epoch": 0.02704465020483602, "grad_norm": 81.87355041503906, "learning_rate": 4.18894165581457e-05, "loss": 4.7897, "step": 3007 }, { "epoch": 0.02705364410247647, "grad_norm": 86.18661499023438, "learning_rate": 4.188332020302561e-05, "loss": 4.5798, "step": 3008 }, { "epoch": 0.02706263800011692, "grad_norm": 29.556934356689453, "learning_rate": 4.187722200157102e-05, "loss": 4.8092, "step": 3009 }, { "epoch": 0.02707163189775737, "grad_norm": 15.70903491973877, "learning_rate": 4.18711219544488e-05, "loss": 3.7685, "step": 3010 }, { "epoch": 0.02708062579539782, "grad_norm": 10.383277893066406, "learning_rate": 4.186502006232605e-05, "loss": 3.0208, "step": 3011 }, { "epoch": 0.027089619693038275, "grad_norm": 14.200499534606934, "learning_rate": 4.1858916325870065e-05, "loss": 3.4479, "step": 3012 }, { "epoch": 0.027098613590678725, "grad_norm": 7.067024230957031, "learning_rate": 4.1852810745748326e-05, "loss": 3.2167, "step": 3013 }, { "epoch": 0.027107607488319176, "grad_norm": 18.172584533691406, "learning_rate": 4.1846703322628556e-05, "loss": 3.4556, "step": 3014 }, { "epoch": 0.027116601385959626, "grad_norm": 10.933493614196777, "learning_rate": 4.184059405717863e-05, "loss": 3.3102, "step": 3015 }, { "epoch": 0.027125595283600076, "grad_norm": 8.070881843566895, "learning_rate": 4.183448295006667e-05, "loss": 3.1792, "step": 3016 }, { "epoch": 0.027134589181240527, "grad_norm": 11.816337585449219, "learning_rate": 4.182837000196095e-05, "loss": 3.0253, "step": 3017 }, { "epoch": 0.02714358307888098, "grad_norm": 18.768280029296875, "learning_rate": 4.182225521352999e-05, "loss": 3.215, "step": 3018 }, { "epoch": 0.02715257697652143, "grad_norm": 6.541604995727539, "learning_rate": 4.18161385854425e-05, "loss": 2.8098, "step": 3019 }, { "epoch": 0.02716157087416188, "grad_norm": 6.671482086181641, "learning_rate": 4.181002011836737e-05, "loss": 3.0591, "step": 3020 }, { "epoch": 0.02717056477180233, "grad_norm": 13.880634307861328, "learning_rate": 4.180389981297371e-05, "loss": 2.6383, "step": 3021 }, { "epoch": 0.027179558669442782, "grad_norm": 18.87247657775879, "learning_rate": 4.179777766993084e-05, "loss": 2.6981, "step": 3022 }, { "epoch": 0.027188552567083232, "grad_norm": 9.017312049865723, "learning_rate": 4.1791653689908254e-05, "loss": 2.702, "step": 3023 }, { "epoch": 0.027197546464723686, "grad_norm": 15.867012023925781, "learning_rate": 4.178552787357567e-05, "loss": 2.3959, "step": 3024 }, { "epoch": 0.027206540362364137, "grad_norm": 15.83353042602539, "learning_rate": 4.177940022160299e-05, "loss": 2.6817, "step": 3025 }, { "epoch": 0.027215534260004587, "grad_norm": 10.416950225830078, "learning_rate": 4.177327073466033e-05, "loss": 2.1476, "step": 3026 }, { "epoch": 0.027224528157645037, "grad_norm": 6.842617034912109, "learning_rate": 4.1767139413418e-05, "loss": 2.2164, "step": 3027 }, { "epoch": 0.027233522055285488, "grad_norm": 4.290499210357666, "learning_rate": 4.176100625854652e-05, "loss": 2.4328, "step": 3028 }, { "epoch": 0.02724251595292594, "grad_norm": 4.508364200592041, "learning_rate": 4.175487127071659e-05, "loss": 2.4621, "step": 3029 }, { "epoch": 0.027251509850566392, "grad_norm": 9.144145965576172, "learning_rate": 4.174873445059913e-05, "loss": 1.9944, "step": 3030 }, { "epoch": 0.027260503748206842, "grad_norm": 5.607731819152832, "learning_rate": 4.174259579886525e-05, "loss": 2.4477, "step": 3031 }, { "epoch": 0.027269497645847292, "grad_norm": 6.461306571960449, "learning_rate": 4.1736455316186265e-05, "loss": 2.386, "step": 3032 }, { "epoch": 0.027278491543487743, "grad_norm": 5.527697563171387, "learning_rate": 4.173031300323369e-05, "loss": 2.0667, "step": 3033 }, { "epoch": 0.027287485441128193, "grad_norm": 3.781141757965088, "learning_rate": 4.172416886067923e-05, "loss": 2.5715, "step": 3034 }, { "epoch": 0.027296479338768647, "grad_norm": 4.1180596351623535, "learning_rate": 4.171802288919482e-05, "loss": 2.7273, "step": 3035 }, { "epoch": 0.027305473236409097, "grad_norm": 3.5386273860931396, "learning_rate": 4.171187508945256e-05, "loss": 2.5204, "step": 3036 }, { "epoch": 0.027314467134049548, "grad_norm": 3.746774911880493, "learning_rate": 4.170572546212477e-05, "loss": 2.4741, "step": 3037 }, { "epoch": 0.027323461031689998, "grad_norm": 2.453432083129883, "learning_rate": 4.169957400788395e-05, "loss": 2.4428, "step": 3038 }, { "epoch": 0.02733245492933045, "grad_norm": 2.2987678050994873, "learning_rate": 4.169342072740283e-05, "loss": 2.0024, "step": 3039 }, { "epoch": 0.0273414488269709, "grad_norm": 2.8574602603912354, "learning_rate": 4.1687265621354314e-05, "loss": 1.8486, "step": 3040 }, { "epoch": 0.027350442724611353, "grad_norm": 2.6134111881256104, "learning_rate": 4.1681108690411526e-05, "loss": 2.1255, "step": 3041 }, { "epoch": 0.027359436622251803, "grad_norm": 3.9120614528656006, "learning_rate": 4.1674949935247767e-05, "loss": 1.9692, "step": 3042 }, { "epoch": 0.027368430519892253, "grad_norm": 4.289495468139648, "learning_rate": 4.166878935653655e-05, "loss": 1.9746, "step": 3043 }, { "epoch": 0.027377424417532704, "grad_norm": 3.1982650756835938, "learning_rate": 4.166262695495159e-05, "loss": 2.3627, "step": 3044 }, { "epoch": 0.027386418315173154, "grad_norm": 2.2008779048919678, "learning_rate": 4.165646273116681e-05, "loss": 2.2751, "step": 3045 }, { "epoch": 0.027395412212813604, "grad_norm": 3.8142881393432617, "learning_rate": 4.165029668585629e-05, "loss": 2.1649, "step": 3046 }, { "epoch": 0.02740440611045406, "grad_norm": 3.549370765686035, "learning_rate": 4.164412881969437e-05, "loss": 2.3398, "step": 3047 }, { "epoch": 0.02741340000809451, "grad_norm": 3.9788920879364014, "learning_rate": 4.1637959133355544e-05, "loss": 2.3667, "step": 3048 }, { "epoch": 0.02742239390573496, "grad_norm": 3.608315944671631, "learning_rate": 4.163178762751452e-05, "loss": 1.9104, "step": 3049 }, { "epoch": 0.02743138780337541, "grad_norm": 2.2390122413635254, "learning_rate": 4.16256143028462e-05, "loss": 1.7725, "step": 3050 }, { "epoch": 0.02744038170101586, "grad_norm": 2.953160285949707, "learning_rate": 4.1619439160025717e-05, "loss": 2.2837, "step": 3051 }, { "epoch": 0.02744937559865631, "grad_norm": 5.7963666915893555, "learning_rate": 4.161326219972833e-05, "loss": 1.7542, "step": 3052 }, { "epoch": 0.027458369496296764, "grad_norm": 4.081873893737793, "learning_rate": 4.1607083422629586e-05, "loss": 2.1032, "step": 3053 }, { "epoch": 0.027467363393937214, "grad_norm": 3.4368042945861816, "learning_rate": 4.160090282940516e-05, "loss": 2.2658, "step": 3054 }, { "epoch": 0.027476357291577665, "grad_norm": 3.2621724605560303, "learning_rate": 4.159472042073096e-05, "loss": 2.3619, "step": 3055 }, { "epoch": 0.027485351189218115, "grad_norm": 3.3290679454803467, "learning_rate": 4.1588536197283076e-05, "loss": 2.2763, "step": 3056 }, { "epoch": 0.027494345086858565, "grad_norm": 5.220732688903809, "learning_rate": 4.1582350159737824e-05, "loss": 1.8041, "step": 3057 }, { "epoch": 0.027503338984499016, "grad_norm": 3.2432384490966797, "learning_rate": 4.157616230877169e-05, "loss": 2.1651, "step": 3058 }, { "epoch": 0.02751233288213947, "grad_norm": 3.3695921897888184, "learning_rate": 4.156997264506136e-05, "loss": 2.662, "step": 3059 }, { "epoch": 0.02752132677977992, "grad_norm": 4.940070152282715, "learning_rate": 4.156378116928375e-05, "loss": 2.1453, "step": 3060 }, { "epoch": 0.02753032067742037, "grad_norm": 2.400968551635742, "learning_rate": 4.1557587882115915e-05, "loss": 2.3067, "step": 3061 }, { "epoch": 0.02753931457506082, "grad_norm": 5.897618770599365, "learning_rate": 4.155139278423518e-05, "loss": 1.9182, "step": 3062 }, { "epoch": 0.02754830847270127, "grad_norm": 6.553978443145752, "learning_rate": 4.1545195876319e-05, "loss": 2.3826, "step": 3063 }, { "epoch": 0.027557302370341725, "grad_norm": 4.258237361907959, "learning_rate": 4.153899715904508e-05, "loss": 2.1553, "step": 3064 }, { "epoch": 0.027566296267982175, "grad_norm": 2.4722869396209717, "learning_rate": 4.1532796633091296e-05, "loss": 2.7671, "step": 3065 }, { "epoch": 0.027575290165622626, "grad_norm": 2.186767578125, "learning_rate": 4.1526594299135737e-05, "loss": 2.1522, "step": 3066 }, { "epoch": 0.027584284063263076, "grad_norm": 3.626065254211426, "learning_rate": 4.152039015785667e-05, "loss": 2.4181, "step": 3067 }, { "epoch": 0.027593277960903526, "grad_norm": 2.640831470489502, "learning_rate": 4.151418420993256e-05, "loss": 2.3971, "step": 3068 }, { "epoch": 0.027602271858543977, "grad_norm": 4.29745626449585, "learning_rate": 4.150797645604211e-05, "loss": 2.5903, "step": 3069 }, { "epoch": 0.02761126575618443, "grad_norm": 4.164549827575684, "learning_rate": 4.150176689686417e-05, "loss": 1.6627, "step": 3070 }, { "epoch": 0.02762025965382488, "grad_norm": 9.698097229003906, "learning_rate": 4.14955555330778e-05, "loss": 2.202, "step": 3071 }, { "epoch": 0.02762925355146533, "grad_norm": 4.5446295738220215, "learning_rate": 4.148934236536229e-05, "loss": 2.3066, "step": 3072 }, { "epoch": 0.02763824744910578, "grad_norm": 2.985602855682373, "learning_rate": 4.148312739439708e-05, "loss": 2.3692, "step": 3073 }, { "epoch": 0.027647241346746232, "grad_norm": 1.844948172569275, "learning_rate": 4.147691062086185e-05, "loss": 2.2292, "step": 3074 }, { "epoch": 0.027656235244386682, "grad_norm": 2.936692476272583, "learning_rate": 4.147069204543645e-05, "loss": 2.2065, "step": 3075 }, { "epoch": 0.027665229142027136, "grad_norm": 3.246051549911499, "learning_rate": 4.146447166880093e-05, "loss": 2.0149, "step": 3076 }, { "epoch": 0.027674223039667586, "grad_norm": 8.493091583251953, "learning_rate": 4.145824949163554e-05, "loss": 2.4045, "step": 3077 }, { "epoch": 0.027683216937308037, "grad_norm": 6.711370468139648, "learning_rate": 4.145202551462072e-05, "loss": 1.7581, "step": 3078 }, { "epoch": 0.027692210834948487, "grad_norm": 2.0546460151672363, "learning_rate": 4.1445799738437136e-05, "loss": 2.158, "step": 3079 }, { "epoch": 0.027701204732588938, "grad_norm": 4.889733791351318, "learning_rate": 4.143957216376561e-05, "loss": 2.2073, "step": 3080 }, { "epoch": 0.027710198630229388, "grad_norm": 3.652509927749634, "learning_rate": 4.14333427912872e-05, "loss": 1.8511, "step": 3081 }, { "epoch": 0.02771919252786984, "grad_norm": 28.932723999023438, "learning_rate": 4.1427111621683123e-05, "loss": 2.1237, "step": 3082 }, { "epoch": 0.027728186425510292, "grad_norm": 2.872149705886841, "learning_rate": 4.142087865563482e-05, "loss": 2.4553, "step": 3083 }, { "epoch": 0.027737180323150742, "grad_norm": 2.5270729064941406, "learning_rate": 4.1414643893823914e-05, "loss": 2.2326, "step": 3084 }, { "epoch": 0.027746174220791193, "grad_norm": 3.750194549560547, "learning_rate": 4.140840733693223e-05, "loss": 2.572, "step": 3085 }, { "epoch": 0.027755168118431643, "grad_norm": 1.9113059043884277, "learning_rate": 4.1402168985641785e-05, "loss": 1.8662, "step": 3086 }, { "epoch": 0.027764162016072094, "grad_norm": 2.114607572555542, "learning_rate": 4.139592884063481e-05, "loss": 2.2566, "step": 3087 }, { "epoch": 0.027773155913712547, "grad_norm": 2.556340217590332, "learning_rate": 4.1389686902593704e-05, "loss": 2.1242, "step": 3088 }, { "epoch": 0.027782149811352998, "grad_norm": 4.419093132019043, "learning_rate": 4.138344317220108e-05, "loss": 2.5831, "step": 3089 }, { "epoch": 0.027791143708993448, "grad_norm": 3.969512462615967, "learning_rate": 4.137719765013974e-05, "loss": 2.0162, "step": 3090 }, { "epoch": 0.0278001376066339, "grad_norm": 4.240414619445801, "learning_rate": 4.137095033709268e-05, "loss": 1.7223, "step": 3091 }, { "epoch": 0.02780913150427435, "grad_norm": 4.162209510803223, "learning_rate": 4.136470123374312e-05, "loss": 2.2151, "step": 3092 }, { "epoch": 0.0278181254019148, "grad_norm": 3.1405136585235596, "learning_rate": 4.135845034077444e-05, "loss": 2.2194, "step": 3093 }, { "epoch": 0.027827119299555253, "grad_norm": 3.116030216217041, "learning_rate": 4.135219765887021e-05, "loss": 2.5085, "step": 3094 }, { "epoch": 0.027836113197195703, "grad_norm": 1.7920465469360352, "learning_rate": 4.134594318871423e-05, "loss": 2.371, "step": 3095 }, { "epoch": 0.027845107094836154, "grad_norm": 3.294434070587158, "learning_rate": 4.133968693099048e-05, "loss": 2.2948, "step": 3096 }, { "epoch": 0.027854100992476604, "grad_norm": 4.317754745483398, "learning_rate": 4.133342888638314e-05, "loss": 2.1576, "step": 3097 }, { "epoch": 0.027863094890117054, "grad_norm": 2.5512516498565674, "learning_rate": 4.132716905557658e-05, "loss": 1.9702, "step": 3098 }, { "epoch": 0.027872088787757508, "grad_norm": 2.659698963165283, "learning_rate": 4.132090743925535e-05, "loss": 2.5757, "step": 3099 }, { "epoch": 0.02788108268539796, "grad_norm": 1.8489782810211182, "learning_rate": 4.131464403810422e-05, "loss": 2.3534, "step": 3100 }, { "epoch": 0.02789007658303841, "grad_norm": 2.7966580390930176, "learning_rate": 4.1308378852808146e-05, "loss": 2.5712, "step": 3101 }, { "epoch": 0.02789907048067886, "grad_norm": 1.7534781694412231, "learning_rate": 4.130211188405229e-05, "loss": 2.1104, "step": 3102 }, { "epoch": 0.02790806437831931, "grad_norm": 1.5903842449188232, "learning_rate": 4.1295843132521973e-05, "loss": 2.1772, "step": 3103 }, { "epoch": 0.02791705827595976, "grad_norm": 6.024909019470215, "learning_rate": 4.128957259890276e-05, "loss": 2.0307, "step": 3104 }, { "epoch": 0.027926052173600214, "grad_norm": 2.0594589710235596, "learning_rate": 4.128330028388038e-05, "loss": 2.2325, "step": 3105 }, { "epoch": 0.027935046071240664, "grad_norm": 3.653573989868164, "learning_rate": 4.127702618814075e-05, "loss": 2.7739, "step": 3106 }, { "epoch": 0.027944039968881115, "grad_norm": 3.4572913646698, "learning_rate": 4.1270750312370024e-05, "loss": 2.1942, "step": 3107 }, { "epoch": 0.027953033866521565, "grad_norm": 2.4217231273651123, "learning_rate": 4.12644726572545e-05, "loss": 1.959, "step": 3108 }, { "epoch": 0.027962027764162015, "grad_norm": 3.619243621826172, "learning_rate": 4.1258193223480704e-05, "loss": 2.2499, "step": 3109 }, { "epoch": 0.027971021661802466, "grad_norm": 4.133096218109131, "learning_rate": 4.1251912011735326e-05, "loss": 2.6729, "step": 3110 }, { "epoch": 0.02798001555944292, "grad_norm": 3.5101542472839355, "learning_rate": 4.1245629022705304e-05, "loss": 2.344, "step": 3111 }, { "epoch": 0.02798900945708337, "grad_norm": 2.8258419036865234, "learning_rate": 4.123934425707771e-05, "loss": 1.9983, "step": 3112 }, { "epoch": 0.02799800335472382, "grad_norm": 1.6894919872283936, "learning_rate": 4.1233057715539836e-05, "loss": 2.6333, "step": 3113 }, { "epoch": 0.02800699725236427, "grad_norm": 7.51069974899292, "learning_rate": 4.1226769398779186e-05, "loss": 1.9825, "step": 3114 }, { "epoch": 0.02801599115000472, "grad_norm": 2.2276382446289062, "learning_rate": 4.122047930748343e-05, "loss": 1.73, "step": 3115 }, { "epoch": 0.02802498504764517, "grad_norm": 2.0561842918395996, "learning_rate": 4.121418744234043e-05, "loss": 2.3657, "step": 3116 }, { "epoch": 0.028033978945285625, "grad_norm": 3.0995891094207764, "learning_rate": 4.120789380403829e-05, "loss": 2.1004, "step": 3117 }, { "epoch": 0.028042972842926076, "grad_norm": 6.902764320373535, "learning_rate": 4.1201598393265244e-05, "loss": 2.1232, "step": 3118 }, { "epoch": 0.028051966740566526, "grad_norm": 3.396804094314575, "learning_rate": 4.119530121070975e-05, "loss": 2.3267, "step": 3119 }, { "epoch": 0.028060960638206976, "grad_norm": 3.929645299911499, "learning_rate": 4.118900225706047e-05, "loss": 2.5841, "step": 3120 }, { "epoch": 0.028069954535847427, "grad_norm": 2.1217966079711914, "learning_rate": 4.118270153300625e-05, "loss": 1.9469, "step": 3121 }, { "epoch": 0.028078948433487877, "grad_norm": 2.858370304107666, "learning_rate": 4.1176399039236116e-05, "loss": 2.2809, "step": 3122 }, { "epoch": 0.02808794233112833, "grad_norm": 3.0742175579071045, "learning_rate": 4.1170094776439295e-05, "loss": 1.8946, "step": 3123 }, { "epoch": 0.02809693622876878, "grad_norm": 4.121082782745361, "learning_rate": 4.116378874530523e-05, "loss": 2.0857, "step": 3124 }, { "epoch": 0.02810593012640923, "grad_norm": 1.9653888940811157, "learning_rate": 4.115748094652352e-05, "loss": 2.3342, "step": 3125 }, { "epoch": 0.028114924024049682, "grad_norm": 2.297305107116699, "learning_rate": 4.1151171380784e-05, "loss": 2.3874, "step": 3126 }, { "epoch": 0.028123917921690132, "grad_norm": 4.029531478881836, "learning_rate": 4.1144860048776653e-05, "loss": 1.6301, "step": 3127 }, { "epoch": 0.028132911819330583, "grad_norm": 3.868661403656006, "learning_rate": 4.1138546951191686e-05, "loss": 2.3426, "step": 3128 }, { "epoch": 0.028141905716971036, "grad_norm": 3.6946258544921875, "learning_rate": 4.1132232088719484e-05, "loss": 2.0033, "step": 3129 }, { "epoch": 0.028150899614611487, "grad_norm": 2.9273486137390137, "learning_rate": 4.112591546205064e-05, "loss": 2.6546, "step": 3130 }, { "epoch": 0.028159893512251937, "grad_norm": 3.0620570182800293, "learning_rate": 4.1119597071875926e-05, "loss": 2.3153, "step": 3131 }, { "epoch": 0.028168887409892387, "grad_norm": 7.146333694458008, "learning_rate": 4.111327691888631e-05, "loss": 2.3122, "step": 3132 }, { "epoch": 0.028177881307532838, "grad_norm": 3.594266891479492, "learning_rate": 4.110695500377295e-05, "loss": 1.6098, "step": 3133 }, { "epoch": 0.028186875205173288, "grad_norm": 2.7578682899475098, "learning_rate": 4.110063132722722e-05, "loss": 2.5206, "step": 3134 }, { "epoch": 0.028195869102813742, "grad_norm": 3.1766514778137207, "learning_rate": 4.1094305889940646e-05, "loss": 2.3234, "step": 3135 }, { "epoch": 0.028204863000454192, "grad_norm": 3.224972724914551, "learning_rate": 4.108797869260498e-05, "loss": 2.3647, "step": 3136 }, { "epoch": 0.028213856898094643, "grad_norm": 4.728072643280029, "learning_rate": 4.1081649735912156e-05, "loss": 2.4751, "step": 3137 }, { "epoch": 0.028222850795735093, "grad_norm": 4.328599452972412, "learning_rate": 4.1075319020554286e-05, "loss": 2.1588, "step": 3138 }, { "epoch": 0.028231844693375543, "grad_norm": 3.9073257446289062, "learning_rate": 4.106898654722371e-05, "loss": 2.4872, "step": 3139 }, { "epoch": 0.028240838591015997, "grad_norm": 3.395214080810547, "learning_rate": 4.106265231661292e-05, "loss": 2.255, "step": 3140 }, { "epoch": 0.028249832488656448, "grad_norm": 2.146472454071045, "learning_rate": 4.1056316329414616e-05, "loss": 2.3922, "step": 3141 }, { "epoch": 0.028258826386296898, "grad_norm": 2.1184942722320557, "learning_rate": 4.1049978586321696e-05, "loss": 1.8372, "step": 3142 }, { "epoch": 0.02826782028393735, "grad_norm": 6.116528034210205, "learning_rate": 4.1043639088027257e-05, "loss": 2.3356, "step": 3143 }, { "epoch": 0.0282768141815778, "grad_norm": 3.0736284255981445, "learning_rate": 4.103729783522456e-05, "loss": 1.9156, "step": 3144 }, { "epoch": 0.02828580807921825, "grad_norm": 1.3594707250595093, "learning_rate": 4.1030954828607095e-05, "loss": 1.8897, "step": 3145 }, { "epoch": 0.028294801976858703, "grad_norm": 5.105855464935303, "learning_rate": 4.102461006886851e-05, "loss": 2.0458, "step": 3146 }, { "epoch": 0.028303795874499153, "grad_norm": 1.7284144163131714, "learning_rate": 4.101826355670265e-05, "loss": 2.2689, "step": 3147 }, { "epoch": 0.028312789772139604, "grad_norm": 2.0737178325653076, "learning_rate": 4.101191529280358e-05, "loss": 2.6112, "step": 3148 }, { "epoch": 0.028321783669780054, "grad_norm": 5.824564456939697, "learning_rate": 4.100556527786552e-05, "loss": 2.4199, "step": 3149 }, { "epoch": 0.028330777567420504, "grad_norm": 3.2998909950256348, "learning_rate": 4.099921351258292e-05, "loss": 1.996, "step": 3150 }, { "epoch": 0.028339771465060955, "grad_norm": 2.3902313709259033, "learning_rate": 4.099285999765037e-05, "loss": 2.5788, "step": 3151 }, { "epoch": 0.02834876536270141, "grad_norm": 5.454022407531738, "learning_rate": 4.0986504733762696e-05, "loss": 2.0259, "step": 3152 }, { "epoch": 0.02835775926034186, "grad_norm": 7.306461334228516, "learning_rate": 4.0980147721614906e-05, "loss": 1.954, "step": 3153 }, { "epoch": 0.02836675315798231, "grad_norm": 2.915823221206665, "learning_rate": 4.0973788961902187e-05, "loss": 2.1068, "step": 3154 }, { "epoch": 0.02837574705562276, "grad_norm": 2.641726016998291, "learning_rate": 4.0967428455319925e-05, "loss": 2.4833, "step": 3155 }, { "epoch": 0.02838474095326321, "grad_norm": 2.861387014389038, "learning_rate": 4.0961066202563684e-05, "loss": 2.182, "step": 3156 }, { "epoch": 0.02839373485090366, "grad_norm": 2.250025749206543, "learning_rate": 4.095470220432925e-05, "loss": 2.6028, "step": 3157 }, { "epoch": 0.028402728748544114, "grad_norm": 2.4638288021087646, "learning_rate": 4.094833646131256e-05, "loss": 2.4293, "step": 3158 }, { "epoch": 0.028411722646184565, "grad_norm": 3.718722105026245, "learning_rate": 4.094196897420979e-05, "loss": 1.9744, "step": 3159 }, { "epoch": 0.028420716543825015, "grad_norm": 25.281524658203125, "learning_rate": 4.093559974371725e-05, "loss": 1.8625, "step": 3160 }, { "epoch": 0.028429710441465465, "grad_norm": 6.270196437835693, "learning_rate": 4.092922877053148e-05, "loss": 2.1938, "step": 3161 }, { "epoch": 0.028438704339105916, "grad_norm": 4.165745735168457, "learning_rate": 4.09228560553492e-05, "loss": 1.886, "step": 3162 }, { "epoch": 0.028447698236746366, "grad_norm": 3.8879573345184326, "learning_rate": 4.091648159886733e-05, "loss": 2.1185, "step": 3163 }, { "epoch": 0.02845669213438682, "grad_norm": 1.82013738155365, "learning_rate": 4.091010540178295e-05, "loss": 2.438, "step": 3164 }, { "epoch": 0.02846568603202727, "grad_norm": 1.5776495933532715, "learning_rate": 4.090372746479337e-05, "loss": 2.493, "step": 3165 }, { "epoch": 0.02847467992966772, "grad_norm": 1.2611743211746216, "learning_rate": 4.0897347788596064e-05, "loss": 2.153, "step": 3166 }, { "epoch": 0.02848367382730817, "grad_norm": 2.1939806938171387, "learning_rate": 4.0890966373888704e-05, "loss": 2.3786, "step": 3167 }, { "epoch": 0.02849266772494862, "grad_norm": 1.9290258884429932, "learning_rate": 4.088458322136916e-05, "loss": 1.6449, "step": 3168 }, { "epoch": 0.02850166162258907, "grad_norm": 3.5956709384918213, "learning_rate": 4.087819833173547e-05, "loss": 2.3184, "step": 3169 }, { "epoch": 0.028510655520229525, "grad_norm": 2.087033987045288, "learning_rate": 4.0871811705685883e-05, "loss": 2.1161, "step": 3170 }, { "epoch": 0.028519649417869976, "grad_norm": 1.456331729888916, "learning_rate": 4.0865423343918836e-05, "loss": 2.3064, "step": 3171 }, { "epoch": 0.028528643315510426, "grad_norm": 2.2013280391693115, "learning_rate": 4.085903324713294e-05, "loss": 2.7429, "step": 3172 }, { "epoch": 0.028537637213150877, "grad_norm": 5.154518127441406, "learning_rate": 4.0852641416027016e-05, "loss": 2.0595, "step": 3173 }, { "epoch": 0.028546631110791327, "grad_norm": 2.222533702850342, "learning_rate": 4.084624785130006e-05, "loss": 2.1323, "step": 3174 }, { "epoch": 0.02855562500843178, "grad_norm": 1.7030072212219238, "learning_rate": 4.0839852553651265e-05, "loss": 2.7251, "step": 3175 }, { "epoch": 0.02856461890607223, "grad_norm": 2.576911687850952, "learning_rate": 4.0833455523780006e-05, "loss": 1.8338, "step": 3176 }, { "epoch": 0.02857361280371268, "grad_norm": 5.397729873657227, "learning_rate": 4.0827056762385855e-05, "loss": 2.0339, "step": 3177 }, { "epoch": 0.028582606701353132, "grad_norm": 3.148986577987671, "learning_rate": 4.082065627016858e-05, "loss": 2.4926, "step": 3178 }, { "epoch": 0.028591600598993582, "grad_norm": 2.2512733936309814, "learning_rate": 4.0814254047828116e-05, "loss": 2.5945, "step": 3179 }, { "epoch": 0.028600594496634033, "grad_norm": 2.6186039447784424, "learning_rate": 4.0807850096064605e-05, "loss": 2.339, "step": 3180 }, { "epoch": 0.028609588394274486, "grad_norm": 3.490780830383301, "learning_rate": 4.080144441557837e-05, "loss": 2.2584, "step": 3181 }, { "epoch": 0.028618582291914937, "grad_norm": 2.951359987258911, "learning_rate": 4.079503700706994e-05, "loss": 1.9852, "step": 3182 }, { "epoch": 0.028627576189555387, "grad_norm": 3.944993495941162, "learning_rate": 4.078862787124e-05, "loss": 2.0303, "step": 3183 }, { "epoch": 0.028636570087195837, "grad_norm": 6.184193134307861, "learning_rate": 4.078221700878946e-05, "loss": 1.9391, "step": 3184 }, { "epoch": 0.028645563984836288, "grad_norm": 5.249210357666016, "learning_rate": 4.0775804420419404e-05, "loss": 2.323, "step": 3185 }, { "epoch": 0.028654557882476738, "grad_norm": 2.322014093399048, "learning_rate": 4.0769390106831084e-05, "loss": 2.0577, "step": 3186 }, { "epoch": 0.028663551780117192, "grad_norm": 4.3748884201049805, "learning_rate": 4.076297406872598e-05, "loss": 2.5096, "step": 3187 }, { "epoch": 0.028672545677757642, "grad_norm": 5.298952102661133, "learning_rate": 4.075655630680572e-05, "loss": 2.4378, "step": 3188 }, { "epoch": 0.028681539575398093, "grad_norm": 3.002617120742798, "learning_rate": 4.0750136821772155e-05, "loss": 2.5241, "step": 3189 }, { "epoch": 0.028690533473038543, "grad_norm": 2.2093453407287598, "learning_rate": 4.0743715614327317e-05, "loss": 2.2691, "step": 3190 }, { "epoch": 0.028699527370678993, "grad_norm": 2.6728267669677734, "learning_rate": 4.07372926851734e-05, "loss": 1.5893, "step": 3191 }, { "epoch": 0.028708521268319444, "grad_norm": 1.891292691230774, "learning_rate": 4.073086803501282e-05, "loss": 2.425, "step": 3192 }, { "epoch": 0.028717515165959898, "grad_norm": 3.9382266998291016, "learning_rate": 4.0724441664548154e-05, "loss": 2.5587, "step": 3193 }, { "epoch": 0.028726509063600348, "grad_norm": 3.898611545562744, "learning_rate": 4.0718013574482195e-05, "loss": 2.3195, "step": 3194 }, { "epoch": 0.0287355029612408, "grad_norm": 4.770773887634277, "learning_rate": 4.07115837655179e-05, "loss": 2.0366, "step": 3195 }, { "epoch": 0.02874449685888125, "grad_norm": 2.057278871536255, "learning_rate": 4.070515223835843e-05, "loss": 2.2337, "step": 3196 }, { "epoch": 0.0287534907565217, "grad_norm": 2.090843677520752, "learning_rate": 4.069871899370713e-05, "loss": 2.2443, "step": 3197 }, { "epoch": 0.02876248465416215, "grad_norm": 2.8579909801483154, "learning_rate": 4.0692284032267516e-05, "loss": 2.0806, "step": 3198 }, { "epoch": 0.028771478551802603, "grad_norm": 2.38395357131958, "learning_rate": 4.068584735474331e-05, "loss": 1.9982, "step": 3199 }, { "epoch": 0.028780472449443054, "grad_norm": 2.35599684715271, "learning_rate": 4.067940896183843e-05, "loss": 2.2506, "step": 3200 }, { "epoch": 0.028789466347083504, "grad_norm": 3.413475751876831, "learning_rate": 4.067296885425695e-05, "loss": 1.95, "step": 3201 }, { "epoch": 0.028798460244723954, "grad_norm": 4.1837382316589355, "learning_rate": 4.0666527032703164e-05, "loss": 2.5609, "step": 3202 }, { "epoch": 0.028807454142364405, "grad_norm": 3.1332669258117676, "learning_rate": 4.066008349788154e-05, "loss": 1.6075, "step": 3203 }, { "epoch": 0.028816448040004855, "grad_norm": 5.490208625793457, "learning_rate": 4.065363825049673e-05, "loss": 2.1677, "step": 3204 }, { "epoch": 0.02882544193764531, "grad_norm": 2.524735450744629, "learning_rate": 4.064719129125356e-05, "loss": 2.0938, "step": 3205 }, { "epoch": 0.02883443583528576, "grad_norm": 4.464412689208984, "learning_rate": 4.064074262085709e-05, "loss": 1.9425, "step": 3206 }, { "epoch": 0.02884342973292621, "grad_norm": 3.7332651615142822, "learning_rate": 4.063429224001252e-05, "loss": 2.2747, "step": 3207 }, { "epoch": 0.02885242363056666, "grad_norm": 2.3687844276428223, "learning_rate": 4.062784014942526e-05, "loss": 2.4281, "step": 3208 }, { "epoch": 0.02886141752820711, "grad_norm": 5.703395366668701, "learning_rate": 4.0621386349800896e-05, "loss": 2.1687, "step": 3209 }, { "epoch": 0.028870411425847564, "grad_norm": 1.5913361310958862, "learning_rate": 4.06149308418452e-05, "loss": 2.1517, "step": 3210 }, { "epoch": 0.028879405323488015, "grad_norm": 5.980081558227539, "learning_rate": 4.0608473626264145e-05, "loss": 1.9842, "step": 3211 }, { "epoch": 0.028888399221128465, "grad_norm": 3.758589744567871, "learning_rate": 4.060201470376389e-05, "loss": 2.0252, "step": 3212 }, { "epoch": 0.028897393118768915, "grad_norm": 10.888694763183594, "learning_rate": 4.0595554075050753e-05, "loss": 1.9601, "step": 3213 }, { "epoch": 0.028906387016409366, "grad_norm": 1.8630120754241943, "learning_rate": 4.058909174083129e-05, "loss": 2.5406, "step": 3214 }, { "epoch": 0.028915380914049816, "grad_norm": 4.058117866516113, "learning_rate": 4.058262770181217e-05, "loss": 2.3633, "step": 3215 }, { "epoch": 0.02892437481169027, "grad_norm": 5.592693328857422, "learning_rate": 4.057616195870032e-05, "loss": 2.5063, "step": 3216 }, { "epoch": 0.02893336870933072, "grad_norm": 3.1954994201660156, "learning_rate": 4.056969451220282e-05, "loss": 2.905, "step": 3217 }, { "epoch": 0.02894236260697117, "grad_norm": 3.0192174911499023, "learning_rate": 4.056322536302693e-05, "loss": 1.8467, "step": 3218 }, { "epoch": 0.02895135650461162, "grad_norm": 1.3952512741088867, "learning_rate": 4.0556754511880115e-05, "loss": 2.158, "step": 3219 }, { "epoch": 0.02896035040225207, "grad_norm": 5.021033763885498, "learning_rate": 4.0550281959470023e-05, "loss": 2.3032, "step": 3220 }, { "epoch": 0.02896934429989252, "grad_norm": 1.853379726409912, "learning_rate": 4.0543807706504464e-05, "loss": 2.2034, "step": 3221 }, { "epoch": 0.028978338197532975, "grad_norm": 8.941792488098145, "learning_rate": 4.0537331753691475e-05, "loss": 2.3309, "step": 3222 }, { "epoch": 0.028987332095173426, "grad_norm": 2.2990055084228516, "learning_rate": 4.053085410173924e-05, "loss": 1.8451, "step": 3223 }, { "epoch": 0.028996325992813876, "grad_norm": 2.0122616291046143, "learning_rate": 4.0524374751356154e-05, "loss": 2.0079, "step": 3224 }, { "epoch": 0.029005319890454326, "grad_norm": 4.179080009460449, "learning_rate": 4.051789370325078e-05, "loss": 2.212, "step": 3225 }, { "epoch": 0.029014313788094777, "grad_norm": 2.0777862071990967, "learning_rate": 4.051141095813189e-05, "loss": 2.1327, "step": 3226 }, { "epoch": 0.029023307685735227, "grad_norm": 1.5539637804031372, "learning_rate": 4.050492651670841e-05, "loss": 2.0647, "step": 3227 }, { "epoch": 0.02903230158337568, "grad_norm": 2.9255666732788086, "learning_rate": 4.049844037968948e-05, "loss": 2.2978, "step": 3228 }, { "epoch": 0.02904129548101613, "grad_norm": 4.005560398101807, "learning_rate": 4.049195254778441e-05, "loss": 2.0323, "step": 3229 }, { "epoch": 0.029050289378656582, "grad_norm": 3.6032602787017822, "learning_rate": 4.048546302170271e-05, "loss": 1.5828, "step": 3230 }, { "epoch": 0.029059283276297032, "grad_norm": 2.553256034851074, "learning_rate": 4.047897180215404e-05, "loss": 2.3962, "step": 3231 }, { "epoch": 0.029068277173937482, "grad_norm": 2.531770944595337, "learning_rate": 4.04724788898483e-05, "loss": 2.5554, "step": 3232 }, { "epoch": 0.029077271071577933, "grad_norm": 1.5501679182052612, "learning_rate": 4.046598428549553e-05, "loss": 1.9135, "step": 3233 }, { "epoch": 0.029086264969218387, "grad_norm": 3.566783905029297, "learning_rate": 4.045948798980597e-05, "loss": 2.2434, "step": 3234 }, { "epoch": 0.029095258866858837, "grad_norm": 2.7577171325683594, "learning_rate": 4.0452990003490047e-05, "loss": 2.3679, "step": 3235 }, { "epoch": 0.029104252764499287, "grad_norm": 3.057166576385498, "learning_rate": 4.044649032725836e-05, "loss": 2.4593, "step": 3236 }, { "epoch": 0.029113246662139738, "grad_norm": 2.762139320373535, "learning_rate": 4.043998896182172e-05, "loss": 2.291, "step": 3237 }, { "epoch": 0.029122240559780188, "grad_norm": 3.3060414791107178, "learning_rate": 4.043348590789111e-05, "loss": 2.6818, "step": 3238 }, { "epoch": 0.02913123445742064, "grad_norm": 2.050816774368286, "learning_rate": 4.042698116617768e-05, "loss": 1.837, "step": 3239 }, { "epoch": 0.029140228355061092, "grad_norm": 4.103585720062256, "learning_rate": 4.042047473739278e-05, "loss": 2.3112, "step": 3240 }, { "epoch": 0.029149222252701543, "grad_norm": 2.104673385620117, "learning_rate": 4.0413966622247943e-05, "loss": 2.1649, "step": 3241 }, { "epoch": 0.029158216150341993, "grad_norm": 2.7185206413269043, "learning_rate": 4.0407456821454906e-05, "loss": 1.9365, "step": 3242 }, { "epoch": 0.029167210047982443, "grad_norm": 3.4596786499023438, "learning_rate": 4.040094533572554e-05, "loss": 1.8971, "step": 3243 }, { "epoch": 0.029176203945622894, "grad_norm": 2.0003161430358887, "learning_rate": 4.039443216577196e-05, "loss": 1.7681, "step": 3244 }, { "epoch": 0.029185197843263344, "grad_norm": 6.0211591720581055, "learning_rate": 4.0387917312306414e-05, "loss": 2.3531, "step": 3245 }, { "epoch": 0.029194191740903798, "grad_norm": 2.3900418281555176, "learning_rate": 4.038140077604137e-05, "loss": 2.1494, "step": 3246 }, { "epoch": 0.02920318563854425, "grad_norm": 2.4880926609039307, "learning_rate": 4.037488255768945e-05, "loss": 1.9524, "step": 3247 }, { "epoch": 0.0292121795361847, "grad_norm": 2.086608409881592, "learning_rate": 4.036836265796351e-05, "loss": 2.2778, "step": 3248 }, { "epoch": 0.02922117343382515, "grad_norm": 2.847870349884033, "learning_rate": 4.036184107757653e-05, "loss": 2.1129, "step": 3249 }, { "epoch": 0.0292301673314656, "grad_norm": 3.948457717895508, "learning_rate": 4.03553178172417e-05, "loss": 2.0142, "step": 3250 }, { "epoch": 0.029239161229106053, "grad_norm": 2.1150753498077393, "learning_rate": 4.0348792877672406e-05, "loss": 2.4087, "step": 3251 }, { "epoch": 0.029248155126746504, "grad_norm": 2.687971830368042, "learning_rate": 4.034226625958219e-05, "loss": 2.4486, "step": 3252 }, { "epoch": 0.029257149024386954, "grad_norm": 2.1269757747650146, "learning_rate": 4.033573796368481e-05, "loss": 1.9155, "step": 3253 }, { "epoch": 0.029266142922027404, "grad_norm": 2.710059642791748, "learning_rate": 4.0329207990694186e-05, "loss": 2.1097, "step": 3254 }, { "epoch": 0.029275136819667855, "grad_norm": 3.3212640285491943, "learning_rate": 4.0322676341324415e-05, "loss": 2.5808, "step": 3255 }, { "epoch": 0.029284130717308305, "grad_norm": 3.190141439437866, "learning_rate": 4.0316143016289795e-05, "loss": 2.1696, "step": 3256 }, { "epoch": 0.02929312461494876, "grad_norm": 3.107213020324707, "learning_rate": 4.0309608016304815e-05, "loss": 2.3517, "step": 3257 }, { "epoch": 0.02930211851258921, "grad_norm": 3.2968926429748535, "learning_rate": 4.030307134208411e-05, "loss": 2.3289, "step": 3258 }, { "epoch": 0.02931111241022966, "grad_norm": 2.112025737762451, "learning_rate": 4.029653299434253e-05, "loss": 2.1726, "step": 3259 }, { "epoch": 0.02932010630787011, "grad_norm": 3.3051886558532715, "learning_rate": 4.028999297379511e-05, "loss": 2.4037, "step": 3260 }, { "epoch": 0.02932910020551056, "grad_norm": 1.1105554103851318, "learning_rate": 4.028345128115704e-05, "loss": 1.9137, "step": 3261 }, { "epoch": 0.02933809410315101, "grad_norm": 2.3117690086364746, "learning_rate": 4.0276907917143714e-05, "loss": 2.4727, "step": 3262 }, { "epoch": 0.029347088000791464, "grad_norm": 1.6814712285995483, "learning_rate": 4.0270362882470715e-05, "loss": 1.7856, "step": 3263 }, { "epoch": 0.029356081898431915, "grad_norm": 3.3382859230041504, "learning_rate": 4.026381617785378e-05, "loss": 2.1227, "step": 3264 }, { "epoch": 0.029365075796072365, "grad_norm": 2.253227710723877, "learning_rate": 4.025726780400886e-05, "loss": 2.3755, "step": 3265 }, { "epoch": 0.029374069693712816, "grad_norm": 5.872115612030029, "learning_rate": 4.025071776165207e-05, "loss": 2.4633, "step": 3266 }, { "epoch": 0.029383063591353266, "grad_norm": 6.51486349105835, "learning_rate": 4.024416605149972e-05, "loss": 2.3157, "step": 3267 }, { "epoch": 0.029392057488993716, "grad_norm": 3.5363011360168457, "learning_rate": 4.023761267426829e-05, "loss": 2.3482, "step": 3268 }, { "epoch": 0.02940105138663417, "grad_norm": 1.5465362071990967, "learning_rate": 4.023105763067444e-05, "loss": 2.257, "step": 3269 }, { "epoch": 0.02941004528427462, "grad_norm": 3.5523605346679688, "learning_rate": 4.0224500921435035e-05, "loss": 2.2594, "step": 3270 }, { "epoch": 0.02941903918191507, "grad_norm": 2.2769815921783447, "learning_rate": 4.0217942547267094e-05, "loss": 2.3038, "step": 3271 }, { "epoch": 0.02942803307955552, "grad_norm": 3.0802955627441406, "learning_rate": 4.021138250888784e-05, "loss": 2.4888, "step": 3272 }, { "epoch": 0.02943702697719597, "grad_norm": 2.3029263019561768, "learning_rate": 4.020482080701466e-05, "loss": 2.495, "step": 3273 }, { "epoch": 0.029446020874836422, "grad_norm": 5.511582851409912, "learning_rate": 4.019825744236514e-05, "loss": 2.0944, "step": 3274 }, { "epoch": 0.029455014772476876, "grad_norm": 2.1373291015625, "learning_rate": 4.019169241565703e-05, "loss": 1.8444, "step": 3275 }, { "epoch": 0.029464008670117326, "grad_norm": 2.1156156063079834, "learning_rate": 4.0185125727608286e-05, "loss": 2.286, "step": 3276 }, { "epoch": 0.029473002567757776, "grad_norm": 3.817440986633301, "learning_rate": 4.017855737893702e-05, "loss": 2.4118, "step": 3277 }, { "epoch": 0.029481996465398227, "grad_norm": 3.3404555320739746, "learning_rate": 4.017198737036154e-05, "loss": 2.3494, "step": 3278 }, { "epoch": 0.029490990363038677, "grad_norm": 1.7974441051483154, "learning_rate": 4.0165415702600326e-05, "loss": 2.085, "step": 3279 }, { "epoch": 0.029499984260679128, "grad_norm": 1.560925006866455, "learning_rate": 4.0158842376372064e-05, "loss": 2.2567, "step": 3280 }, { "epoch": 0.02950897815831958, "grad_norm": 1.4963371753692627, "learning_rate": 4.0152267392395583e-05, "loss": 1.751, "step": 3281 }, { "epoch": 0.02951797205596003, "grad_norm": 2.59458589553833, "learning_rate": 4.014569075138992e-05, "loss": 2.6431, "step": 3282 }, { "epoch": 0.029526965953600482, "grad_norm": 3.4336321353912354, "learning_rate": 4.013911245407429e-05, "loss": 2.0962, "step": 3283 }, { "epoch": 0.029535959851240932, "grad_norm": 3.097058057785034, "learning_rate": 4.013253250116809e-05, "loss": 2.0755, "step": 3284 }, { "epoch": 0.029544953748881383, "grad_norm": 5.463542938232422, "learning_rate": 4.0125950893390876e-05, "loss": 2.6198, "step": 3285 }, { "epoch": 0.029553947646521837, "grad_norm": 3.7046754360198975, "learning_rate": 4.0119367631462416e-05, "loss": 2.6188, "step": 3286 }, { "epoch": 0.029562941544162287, "grad_norm": 3.3695545196533203, "learning_rate": 4.0112782716102644e-05, "loss": 1.8489, "step": 3287 }, { "epoch": 0.029571935441802737, "grad_norm": 2.9100685119628906, "learning_rate": 4.0106196148031674e-05, "loss": 2.4068, "step": 3288 }, { "epoch": 0.029580929339443188, "grad_norm": 1.979823112487793, "learning_rate": 4.009960792796981e-05, "loss": 1.8216, "step": 3289 }, { "epoch": 0.029589923237083638, "grad_norm": 2.7863786220550537, "learning_rate": 4.009301805663752e-05, "loss": 2.3632, "step": 3290 }, { "epoch": 0.02959891713472409, "grad_norm": 2.147399425506592, "learning_rate": 4.0086426534755475e-05, "loss": 2.286, "step": 3291 }, { "epoch": 0.029607911032364542, "grad_norm": 1.7642678022384644, "learning_rate": 4.007983336304449e-05, "loss": 2.2855, "step": 3292 }, { "epoch": 0.029616904930004993, "grad_norm": 2.685920000076294, "learning_rate": 4.007323854222562e-05, "loss": 2.2375, "step": 3293 }, { "epoch": 0.029625898827645443, "grad_norm": 3.8988990783691406, "learning_rate": 4.0066642073020036e-05, "loss": 2.0386, "step": 3294 }, { "epoch": 0.029634892725285893, "grad_norm": 1.778841495513916, "learning_rate": 4.006004395614913e-05, "loss": 2.6029, "step": 3295 }, { "epoch": 0.029643886622926344, "grad_norm": 3.907951831817627, "learning_rate": 4.005344419233446e-05, "loss": 3.2529, "step": 3296 }, { "epoch": 0.029652880520566794, "grad_norm": 3.6310789585113525, "learning_rate": 4.004684278229776e-05, "loss": 1.8507, "step": 3297 }, { "epoch": 0.029661874418207248, "grad_norm": 4.528019428253174, "learning_rate": 4.004023972676097e-05, "loss": 1.9338, "step": 3298 }, { "epoch": 0.029670868315847698, "grad_norm": 2.9773812294006348, "learning_rate": 4.003363502644616e-05, "loss": 1.4874, "step": 3299 }, { "epoch": 0.02967986221348815, "grad_norm": 2.2767298221588135, "learning_rate": 4.002702868207563e-05, "loss": 2.059, "step": 3300 }, { "epoch": 0.0296888561111286, "grad_norm": 3.2745308876037598, "learning_rate": 4.002042069437184e-05, "loss": 2.1171, "step": 3301 }, { "epoch": 0.02969785000876905, "grad_norm": 3.8839833736419678, "learning_rate": 4.001381106405742e-05, "loss": 2.2914, "step": 3302 }, { "epoch": 0.0297068439064095, "grad_norm": 2.911914825439453, "learning_rate": 4.00071997918552e-05, "loss": 2.5146, "step": 3303 }, { "epoch": 0.029715837804049954, "grad_norm": 4.999144554138184, "learning_rate": 4.000058687848817e-05, "loss": 2.3123, "step": 3304 }, { "epoch": 0.029724831701690404, "grad_norm": 2.735229015350342, "learning_rate": 3.999397232467952e-05, "loss": 2.39, "step": 3305 }, { "epoch": 0.029733825599330854, "grad_norm": 1.8078547716140747, "learning_rate": 3.99873561311526e-05, "loss": 2.4691, "step": 3306 }, { "epoch": 0.029742819496971305, "grad_norm": 2.941019058227539, "learning_rate": 3.998073829863093e-05, "loss": 2.1862, "step": 3307 }, { "epoch": 0.029751813394611755, "grad_norm": 3.460770606994629, "learning_rate": 3.9974118827838256e-05, "loss": 1.9813, "step": 3308 }, { "epoch": 0.029760807292252205, "grad_norm": 2.502283811569214, "learning_rate": 3.996749771949846e-05, "loss": 2.377, "step": 3309 }, { "epoch": 0.02976980118989266, "grad_norm": 2.69002103805542, "learning_rate": 3.996087497433562e-05, "loss": 2.3578, "step": 3310 }, { "epoch": 0.02977879508753311, "grad_norm": 3.4034218788146973, "learning_rate": 3.995425059307397e-05, "loss": 2.064, "step": 3311 }, { "epoch": 0.02978778898517356, "grad_norm": 4.322262763977051, "learning_rate": 3.9947624576437975e-05, "loss": 2.2697, "step": 3312 }, { "epoch": 0.02979678288281401, "grad_norm": 3.2600748538970947, "learning_rate": 3.9940996925152226e-05, "loss": 2.1718, "step": 3313 }, { "epoch": 0.02980577678045446, "grad_norm": 2.042640447616577, "learning_rate": 3.9934367639941514e-05, "loss": 2.0404, "step": 3314 }, { "epoch": 0.02981477067809491, "grad_norm": 3.3956491947174072, "learning_rate": 3.9927736721530805e-05, "loss": 2.4141, "step": 3315 }, { "epoch": 0.029823764575735365, "grad_norm": 2.46696138381958, "learning_rate": 3.992110417064526e-05, "loss": 2.0094, "step": 3316 }, { "epoch": 0.029832758473375815, "grad_norm": 2.2790229320526123, "learning_rate": 3.99144699880102e-05, "loss": 1.8526, "step": 3317 }, { "epoch": 0.029841752371016265, "grad_norm": 1.89596688747406, "learning_rate": 3.9907834174351125e-05, "loss": 2.8258, "step": 3318 }, { "epoch": 0.029850746268656716, "grad_norm": 5.993200778961182, "learning_rate": 3.990119673039371e-05, "loss": 2.2965, "step": 3319 }, { "epoch": 0.029859740166297166, "grad_norm": 2.6725239753723145, "learning_rate": 3.9894557656863823e-05, "loss": 2.8972, "step": 3320 }, { "epoch": 0.02986873406393762, "grad_norm": 4.530800819396973, "learning_rate": 3.988791695448751e-05, "loss": 2.3659, "step": 3321 }, { "epoch": 0.02987772796157807, "grad_norm": 2.98868727684021, "learning_rate": 3.988127462399098e-05, "loss": 2.1464, "step": 3322 }, { "epoch": 0.02988672185921852, "grad_norm": 3.6070566177368164, "learning_rate": 3.987463066610063e-05, "loss": 2.1972, "step": 3323 }, { "epoch": 0.02989571575685897, "grad_norm": 3.187333106994629, "learning_rate": 3.986798508154303e-05, "loss": 2.5585, "step": 3324 }, { "epoch": 0.02990470965449942, "grad_norm": 2.423444986343384, "learning_rate": 3.9861337871044954e-05, "loss": 2.4368, "step": 3325 }, { "epoch": 0.029913703552139872, "grad_norm": 2.71205997467041, "learning_rate": 3.9854689035333295e-05, "loss": 1.6773, "step": 3326 }, { "epoch": 0.029922697449780326, "grad_norm": 6.281142711639404, "learning_rate": 3.984803857513518e-05, "loss": 2.3923, "step": 3327 }, { "epoch": 0.029931691347420776, "grad_norm": 2.2859668731689453, "learning_rate": 3.984138649117789e-05, "loss": 1.6474, "step": 3328 }, { "epoch": 0.029940685245061226, "grad_norm": 6.609752655029297, "learning_rate": 3.983473278418889e-05, "loss": 2.1519, "step": 3329 }, { "epoch": 0.029949679142701677, "grad_norm": 4.418726444244385, "learning_rate": 3.9828077454895815e-05, "loss": 1.9958, "step": 3330 }, { "epoch": 0.029958673040342127, "grad_norm": 4.372011661529541, "learning_rate": 3.982142050402649e-05, "loss": 2.495, "step": 3331 }, { "epoch": 0.029967666937982577, "grad_norm": 3.305446147918701, "learning_rate": 3.981476193230889e-05, "loss": 2.5851, "step": 3332 }, { "epoch": 0.02997666083562303, "grad_norm": 4.402859210968018, "learning_rate": 3.980810174047121e-05, "loss": 1.9961, "step": 3333 }, { "epoch": 0.02998565473326348, "grad_norm": 2.893267869949341, "learning_rate": 3.980143992924178e-05, "loss": 2.159, "step": 3334 }, { "epoch": 0.029994648630903932, "grad_norm": 1.5251262187957764, "learning_rate": 3.9794776499349126e-05, "loss": 1.8272, "step": 3335 }, { "epoch": 0.030003642528544382, "grad_norm": 2.934034585952759, "learning_rate": 3.978811145152197e-05, "loss": 2.9505, "step": 3336 }, { "epoch": 0.030012636426184833, "grad_norm": 2.7849252223968506, "learning_rate": 3.978144478648917e-05, "loss": 1.9837, "step": 3337 }, { "epoch": 0.030021630323825283, "grad_norm": 4.492745399475098, "learning_rate": 3.9774776504979805e-05, "loss": 2.266, "step": 3338 }, { "epoch": 0.030030624221465737, "grad_norm": 3.240025281906128, "learning_rate": 3.976810660772308e-05, "loss": 2.521, "step": 3339 }, { "epoch": 0.030039618119106187, "grad_norm": 2.8305504322052, "learning_rate": 3.976143509544843e-05, "loss": 2.183, "step": 3340 }, { "epoch": 0.030048612016746638, "grad_norm": 4.703127384185791, "learning_rate": 3.975476196888543e-05, "loss": 2.3004, "step": 3341 }, { "epoch": 0.030057605914387088, "grad_norm": 7.930714130401611, "learning_rate": 3.9748087228763844e-05, "loss": 2.5213, "step": 3342 }, { "epoch": 0.03006659981202754, "grad_norm": 2.686542272567749, "learning_rate": 3.974141087581361e-05, "loss": 2.293, "step": 3343 }, { "epoch": 0.03007559370966799, "grad_norm": 2.839146614074707, "learning_rate": 3.973473291076485e-05, "loss": 2.2831, "step": 3344 }, { "epoch": 0.030084587607308443, "grad_norm": 2.9863083362579346, "learning_rate": 3.972805333434784e-05, "loss": 1.8813, "step": 3345 }, { "epoch": 0.030093581504948893, "grad_norm": 2.3055953979492188, "learning_rate": 3.972137214729307e-05, "loss": 2.509, "step": 3346 }, { "epoch": 0.030102575402589343, "grad_norm": 2.144429922103882, "learning_rate": 3.971468935033118e-05, "loss": 2.0259, "step": 3347 }, { "epoch": 0.030111569300229794, "grad_norm": 2.1168861389160156, "learning_rate": 3.970800494419298e-05, "loss": 2.0985, "step": 3348 }, { "epoch": 0.030120563197870244, "grad_norm": 3.5302653312683105, "learning_rate": 3.970131892960948e-05, "loss": 2.1029, "step": 3349 }, { "epoch": 0.030129557095510694, "grad_norm": 2.5670557022094727, "learning_rate": 3.969463130731183e-05, "loss": 2.5394, "step": 3350 }, { "epoch": 0.030138550993151148, "grad_norm": 2.6002748012542725, "learning_rate": 3.9687942078031404e-05, "loss": 2.4346, "step": 3351 }, { "epoch": 0.0301475448907916, "grad_norm": 4.314457893371582, "learning_rate": 3.968125124249971e-05, "loss": 2.3692, "step": 3352 }, { "epoch": 0.03015653878843205, "grad_norm": 5.1201677322387695, "learning_rate": 3.967455880144846e-05, "loss": 2.1556, "step": 3353 }, { "epoch": 0.0301655326860725, "grad_norm": 2.8871092796325684, "learning_rate": 3.966786475560952e-05, "loss": 1.9721, "step": 3354 }, { "epoch": 0.03017452658371295, "grad_norm": 1.761716365814209, "learning_rate": 3.966116910571494e-05, "loss": 2.4281, "step": 3355 }, { "epoch": 0.0301835204813534, "grad_norm": 1.8951971530914307, "learning_rate": 3.965447185249695e-05, "loss": 2.6429, "step": 3356 }, { "epoch": 0.030192514378993854, "grad_norm": 3.2166872024536133, "learning_rate": 3.964777299668795e-05, "loss": 1.6466, "step": 3357 }, { "epoch": 0.030201508276634304, "grad_norm": 3.0291852951049805, "learning_rate": 3.964107253902052e-05, "loss": 2.1426, "step": 3358 }, { "epoch": 0.030210502174274755, "grad_norm": 2.273714065551758, "learning_rate": 3.963437048022741e-05, "loss": 1.9312, "step": 3359 }, { "epoch": 0.030219496071915205, "grad_norm": 2.817186117172241, "learning_rate": 3.9627666821041545e-05, "loss": 2.1421, "step": 3360 }, { "epoch": 0.030228489969555655, "grad_norm": 2.3357627391815186, "learning_rate": 3.962096156219603e-05, "loss": 1.947, "step": 3361 }, { "epoch": 0.03023748386719611, "grad_norm": 4.068572044372559, "learning_rate": 3.9614254704424134e-05, "loss": 2.6012, "step": 3362 }, { "epoch": 0.03024647776483656, "grad_norm": 2.0846500396728516, "learning_rate": 3.960754624845932e-05, "loss": 2.4563, "step": 3363 }, { "epoch": 0.03025547166247701, "grad_norm": 2.2644717693328857, "learning_rate": 3.960083619503522e-05, "loss": 2.1892, "step": 3364 }, { "epoch": 0.03026446556011746, "grad_norm": 3.585670232772827, "learning_rate": 3.9594124544885615e-05, "loss": 2.331, "step": 3365 }, { "epoch": 0.03027345945775791, "grad_norm": 4.506518363952637, "learning_rate": 3.95874112987445e-05, "loss": 2.3478, "step": 3366 }, { "epoch": 0.03028245335539836, "grad_norm": 6.630534648895264, "learning_rate": 3.958069645734601e-05, "loss": 1.7625, "step": 3367 }, { "epoch": 0.030291447253038815, "grad_norm": 4.274195671081543, "learning_rate": 3.957398002142448e-05, "loss": 2.1734, "step": 3368 }, { "epoch": 0.030300441150679265, "grad_norm": 2.912423610687256, "learning_rate": 3.9567261991714404e-05, "loss": 2.0906, "step": 3369 }, { "epoch": 0.030309435048319715, "grad_norm": 1.808295726776123, "learning_rate": 3.956054236895046e-05, "loss": 2.2409, "step": 3370 }, { "epoch": 0.030318428945960166, "grad_norm": 3.275714635848999, "learning_rate": 3.955382115386749e-05, "loss": 2.1622, "step": 3371 }, { "epoch": 0.030327422843600616, "grad_norm": 2.9308416843414307, "learning_rate": 3.954709834720053e-05, "loss": 2.4114, "step": 3372 }, { "epoch": 0.030336416741241067, "grad_norm": 4.425062656402588, "learning_rate": 3.954037394968476e-05, "loss": 2.59, "step": 3373 }, { "epoch": 0.03034541063888152, "grad_norm": 2.8823471069335938, "learning_rate": 3.9533647962055555e-05, "loss": 2.4855, "step": 3374 }, { "epoch": 0.03035440453652197, "grad_norm": 1.5312395095825195, "learning_rate": 3.952692038504846e-05, "loss": 2.1235, "step": 3375 }, { "epoch": 0.03036339843416242, "grad_norm": 2.9020469188690186, "learning_rate": 3.9520191219399196e-05, "loss": 2.3148, "step": 3376 }, { "epoch": 0.03037239233180287, "grad_norm": 4.102804660797119, "learning_rate": 3.951346046584364e-05, "loss": 1.9686, "step": 3377 }, { "epoch": 0.030381386229443322, "grad_norm": 3.2844934463500977, "learning_rate": 3.950672812511788e-05, "loss": 2.1558, "step": 3378 }, { "epoch": 0.030390380127083772, "grad_norm": 3.699631690979004, "learning_rate": 3.949999419795815e-05, "loss": 2.0539, "step": 3379 }, { "epoch": 0.030399374024724226, "grad_norm": 2.3833839893341064, "learning_rate": 3.949325868510083e-05, "loss": 2.1831, "step": 3380 }, { "epoch": 0.030408367922364676, "grad_norm": 3.255063772201538, "learning_rate": 3.948652158728255e-05, "loss": 2.2406, "step": 3381 }, { "epoch": 0.030417361820005127, "grad_norm": 2.4463796615600586, "learning_rate": 3.947978290524005e-05, "loss": 1.7855, "step": 3382 }, { "epoch": 0.030426355717645577, "grad_norm": 7.361224174499512, "learning_rate": 3.9473042639710244e-05, "loss": 2.5838, "step": 3383 }, { "epoch": 0.030435349615286027, "grad_norm": 2.137216091156006, "learning_rate": 3.9466300791430274e-05, "loss": 2.6096, "step": 3384 }, { "epoch": 0.030444343512926478, "grad_norm": 4.806118488311768, "learning_rate": 3.945955736113739e-05, "loss": 2.4389, "step": 3385 }, { "epoch": 0.03045333741056693, "grad_norm": 1.2115660905838013, "learning_rate": 3.9452812349569056e-05, "loss": 2.0754, "step": 3386 }, { "epoch": 0.030462331308207382, "grad_norm": 3.167142391204834, "learning_rate": 3.944606575746289e-05, "loss": 1.8421, "step": 3387 }, { "epoch": 0.030471325205847832, "grad_norm": 2.178412914276123, "learning_rate": 3.943931758555669e-05, "loss": 2.4891, "step": 3388 }, { "epoch": 0.030480319103488283, "grad_norm": 4.52821683883667, "learning_rate": 3.9432567834588436e-05, "loss": 1.8387, "step": 3389 }, { "epoch": 0.030489313001128733, "grad_norm": 7.51453971862793, "learning_rate": 3.942581650529625e-05, "loss": 2.4726, "step": 3390 }, { "epoch": 0.030498306898769183, "grad_norm": 1.8719274997711182, "learning_rate": 3.941906359841847e-05, "loss": 2.2018, "step": 3391 }, { "epoch": 0.030507300796409637, "grad_norm": 4.0125322341918945, "learning_rate": 3.941230911469358e-05, "loss": 1.8591, "step": 3392 }, { "epoch": 0.030516294694050088, "grad_norm": 3.2025041580200195, "learning_rate": 3.940555305486022e-05, "loss": 1.926, "step": 3393 }, { "epoch": 0.030525288591690538, "grad_norm": 3.767312526702881, "learning_rate": 3.939879541965724e-05, "loss": 2.6161, "step": 3394 }, { "epoch": 0.03053428248933099, "grad_norm": 2.7026238441467285, "learning_rate": 3.9392036209823644e-05, "loss": 2.468, "step": 3395 }, { "epoch": 0.03054327638697144, "grad_norm": 2.4762394428253174, "learning_rate": 3.9385275426098605e-05, "loss": 2.0348, "step": 3396 }, { "epoch": 0.030552270284611893, "grad_norm": 2.3550844192504883, "learning_rate": 3.9378513069221475e-05, "loss": 2.4539, "step": 3397 }, { "epoch": 0.030561264182252343, "grad_norm": 2.8706889152526855, "learning_rate": 3.937174913993177e-05, "loss": 1.802, "step": 3398 }, { "epoch": 0.030570258079892793, "grad_norm": 3.237880229949951, "learning_rate": 3.936498363896919e-05, "loss": 2.1627, "step": 3399 }, { "epoch": 0.030579251977533244, "grad_norm": 2.168013095855713, "learning_rate": 3.935821656707359e-05, "loss": 2.2797, "step": 3400 }, { "epoch": 0.030588245875173694, "grad_norm": 2.76772141456604, "learning_rate": 3.935144792498502e-05, "loss": 2.2891, "step": 3401 }, { "epoch": 0.030597239772814144, "grad_norm": 14.396310806274414, "learning_rate": 3.934467771344368e-05, "loss": 2.0659, "step": 3402 }, { "epoch": 0.030606233670454598, "grad_norm": 3.376431941986084, "learning_rate": 3.9337905933189956e-05, "loss": 1.7816, "step": 3403 }, { "epoch": 0.03061522756809505, "grad_norm": 2.7424492835998535, "learning_rate": 3.933113258496439e-05, "loss": 2.2886, "step": 3404 }, { "epoch": 0.0306242214657355, "grad_norm": 5.084122657775879, "learning_rate": 3.932435766950772e-05, "loss": 2.0306, "step": 3405 }, { "epoch": 0.03063321536337595, "grad_norm": 3.826967239379883, "learning_rate": 3.931758118756082e-05, "loss": 2.2853, "step": 3406 }, { "epoch": 0.0306422092610164, "grad_norm": 3.463902711868286, "learning_rate": 3.9310803139864775e-05, "loss": 2.3815, "step": 3407 }, { "epoch": 0.03065120315865685, "grad_norm": 2.788301706314087, "learning_rate": 3.930402352716082e-05, "loss": 2.3739, "step": 3408 }, { "epoch": 0.030660197056297304, "grad_norm": 2.788482427597046, "learning_rate": 3.9297242350190345e-05, "loss": 2.5058, "step": 3409 }, { "epoch": 0.030669190953937754, "grad_norm": 2.5618627071380615, "learning_rate": 3.929045960969494e-05, "loss": 2.3603, "step": 3410 }, { "epoch": 0.030678184851578204, "grad_norm": 2.3648533821105957, "learning_rate": 3.928367530641637e-05, "loss": 2.3993, "step": 3411 }, { "epoch": 0.030687178749218655, "grad_norm": 4.632047176361084, "learning_rate": 3.9276889441096534e-05, "loss": 2.0474, "step": 3412 }, { "epoch": 0.030696172646859105, "grad_norm": 2.667696237564087, "learning_rate": 3.927010201447754e-05, "loss": 1.9524, "step": 3413 }, { "epoch": 0.030705166544499556, "grad_norm": 4.857188701629639, "learning_rate": 3.9263313027301643e-05, "loss": 2.1821, "step": 3414 }, { "epoch": 0.03071416044214001, "grad_norm": 2.435000419616699, "learning_rate": 3.925652248031127e-05, "loss": 2.2495, "step": 3415 }, { "epoch": 0.03072315433978046, "grad_norm": 3.0940067768096924, "learning_rate": 3.924973037424905e-05, "loss": 2.0038, "step": 3416 }, { "epoch": 0.03073214823742091, "grad_norm": 4.135527610778809, "learning_rate": 3.924293670985774e-05, "loss": 2.3598, "step": 3417 }, { "epoch": 0.03074114213506136, "grad_norm": 2.4166455268859863, "learning_rate": 3.9236141487880265e-05, "loss": 2.1141, "step": 3418 }, { "epoch": 0.03075013603270181, "grad_norm": 3.100841760635376, "learning_rate": 3.922934470905978e-05, "loss": 2.3393, "step": 3419 }, { "epoch": 0.03075912993034226, "grad_norm": 2.947150707244873, "learning_rate": 3.9222546374139533e-05, "loss": 1.9561, "step": 3420 }, { "epoch": 0.030768123827982715, "grad_norm": 2.8060672283172607, "learning_rate": 3.9215746483863015e-05, "loss": 2.0491, "step": 3421 }, { "epoch": 0.030777117725623165, "grad_norm": 3.410832166671753, "learning_rate": 3.920894503897382e-05, "loss": 2.2992, "step": 3422 }, { "epoch": 0.030786111623263616, "grad_norm": 2.7189183235168457, "learning_rate": 3.920214204021577e-05, "loss": 1.9678, "step": 3423 }, { "epoch": 0.030795105520904066, "grad_norm": 2.5408122539520264, "learning_rate": 3.919533748833282e-05, "loss": 1.9027, "step": 3424 }, { "epoch": 0.030804099418544516, "grad_norm": 2.4887120723724365, "learning_rate": 3.9188531384069096e-05, "loss": 2.4024, "step": 3425 }, { "epoch": 0.030813093316184967, "grad_norm": 3.362849473953247, "learning_rate": 3.9181723728168916e-05, "loss": 1.659, "step": 3426 }, { "epoch": 0.03082208721382542, "grad_norm": 3.779282569885254, "learning_rate": 3.9174914521376746e-05, "loss": 1.971, "step": 3427 }, { "epoch": 0.03083108111146587, "grad_norm": 2.7899739742279053, "learning_rate": 3.9168103764437243e-05, "loss": 1.7633, "step": 3428 }, { "epoch": 0.03084007500910632, "grad_norm": 2.0956947803497314, "learning_rate": 3.9161291458095214e-05, "loss": 2.4606, "step": 3429 }, { "epoch": 0.030849068906746772, "grad_norm": 3.3182578086853027, "learning_rate": 3.915447760309564e-05, "loss": 2.1094, "step": 3430 }, { "epoch": 0.030858062804387222, "grad_norm": 3.026423215866089, "learning_rate": 3.9147662200183675e-05, "loss": 2.4758, "step": 3431 }, { "epoch": 0.030867056702027676, "grad_norm": 2.713522434234619, "learning_rate": 3.914084525010464e-05, "loss": 2.337, "step": 3432 }, { "epoch": 0.030876050599668126, "grad_norm": 1.9852533340454102, "learning_rate": 3.9134026753604034e-05, "loss": 2.3312, "step": 3433 }, { "epoch": 0.030885044497308577, "grad_norm": 2.825181245803833, "learning_rate": 3.912720671142751e-05, "loss": 2.2589, "step": 3434 }, { "epoch": 0.030894038394949027, "grad_norm": 4.129382610321045, "learning_rate": 3.9120385124320894e-05, "loss": 2.0681, "step": 3435 }, { "epoch": 0.030903032292589477, "grad_norm": 5.5702362060546875, "learning_rate": 3.91135619930302e-05, "loss": 2.0795, "step": 3436 }, { "epoch": 0.030912026190229928, "grad_norm": 2.4617063999176025, "learning_rate": 3.9106737318301575e-05, "loss": 2.4848, "step": 3437 }, { "epoch": 0.03092102008787038, "grad_norm": 1.8230080604553223, "learning_rate": 3.909991110088138e-05, "loss": 1.6981, "step": 3438 }, { "epoch": 0.030930013985510832, "grad_norm": 3.931959629058838, "learning_rate": 3.90930833415161e-05, "loss": 2.6906, "step": 3439 }, { "epoch": 0.030939007883151282, "grad_norm": 2.2901222705841064, "learning_rate": 3.9086254040952416e-05, "loss": 2.1478, "step": 3440 }, { "epoch": 0.030948001780791733, "grad_norm": 2.9115920066833496, "learning_rate": 3.907942319993717e-05, "loss": 2.0706, "step": 3441 }, { "epoch": 0.030956995678432183, "grad_norm": 3.176612377166748, "learning_rate": 3.907259081921736e-05, "loss": 2.0645, "step": 3442 }, { "epoch": 0.030965989576072633, "grad_norm": 2.2572555541992188, "learning_rate": 3.906575689954018e-05, "loss": 2.35, "step": 3443 }, { "epoch": 0.030974983473713087, "grad_norm": 2.604419231414795, "learning_rate": 3.9058921441652984e-05, "loss": 2.3909, "step": 3444 }, { "epoch": 0.030983977371353538, "grad_norm": 1.925861120223999, "learning_rate": 3.905208444630327e-05, "loss": 2.0618, "step": 3445 }, { "epoch": 0.030992971268993988, "grad_norm": 4.320574760437012, "learning_rate": 3.904524591423872e-05, "loss": 1.9996, "step": 3446 }, { "epoch": 0.03100196516663444, "grad_norm": 3.2505104541778564, "learning_rate": 3.90384058462072e-05, "loss": 2.2015, "step": 3447 }, { "epoch": 0.03101095906427489, "grad_norm": 2.256129264831543, "learning_rate": 3.903156424295673e-05, "loss": 1.568, "step": 3448 }, { "epoch": 0.03101995296191534, "grad_norm": 3.204224109649658, "learning_rate": 3.9024721105235485e-05, "loss": 2.145, "step": 3449 }, { "epoch": 0.031028946859555793, "grad_norm": 2.9753270149230957, "learning_rate": 3.901787643379182e-05, "loss": 2.5395, "step": 3450 }, { "epoch": 0.031037940757196243, "grad_norm": 2.7557966709136963, "learning_rate": 3.901103022937427e-05, "loss": 2.0691, "step": 3451 }, { "epoch": 0.031046934654836694, "grad_norm": 6.288999557495117, "learning_rate": 3.900418249273152e-05, "loss": 2.2907, "step": 3452 }, { "epoch": 0.031055928552477144, "grad_norm": 2.288966655731201, "learning_rate": 3.899733322461242e-05, "loss": 2.2842, "step": 3453 }, { "epoch": 0.031064922450117594, "grad_norm": 4.064635753631592, "learning_rate": 3.8990482425766006e-05, "loss": 2.3219, "step": 3454 }, { "epoch": 0.031073916347758045, "grad_norm": 4.163564205169678, "learning_rate": 3.898363009694148e-05, "loss": 2.3082, "step": 3455 }, { "epoch": 0.0310829102453985, "grad_norm": 2.0108954906463623, "learning_rate": 3.897677623888818e-05, "loss": 2.1214, "step": 3456 }, { "epoch": 0.03109190414303895, "grad_norm": 5.410860061645508, "learning_rate": 3.8969920852355644e-05, "loss": 2.542, "step": 3457 }, { "epoch": 0.0311008980406794, "grad_norm": 2.4267611503601074, "learning_rate": 3.896306393809357e-05, "loss": 1.9528, "step": 3458 }, { "epoch": 0.03110989193831985, "grad_norm": 2.236835479736328, "learning_rate": 3.895620549685182e-05, "loss": 2.0964, "step": 3459 }, { "epoch": 0.0311188858359603, "grad_norm": 6.5997233390808105, "learning_rate": 3.894934552938041e-05, "loss": 2.2225, "step": 3460 }, { "epoch": 0.03112787973360075, "grad_norm": 2.4210829734802246, "learning_rate": 3.894248403642955e-05, "loss": 2.1529, "step": 3461 }, { "epoch": 0.031136873631241204, "grad_norm": 3.0609254837036133, "learning_rate": 3.89356210187496e-05, "loss": 2.2311, "step": 3462 }, { "epoch": 0.031145867528881654, "grad_norm": 2.943676471710205, "learning_rate": 3.892875647709108e-05, "loss": 2.0448, "step": 3463 }, { "epoch": 0.031154861426522105, "grad_norm": 3.7321176528930664, "learning_rate": 3.8921890412204705e-05, "loss": 2.3851, "step": 3464 }, { "epoch": 0.031163855324162555, "grad_norm": 3.46948504447937, "learning_rate": 3.891502282484132e-05, "loss": 2.4986, "step": 3465 }, { "epoch": 0.031172849221803006, "grad_norm": 2.1093709468841553, "learning_rate": 3.890815371575196e-05, "loss": 1.4791, "step": 3466 }, { "epoch": 0.031181843119443456, "grad_norm": 4.1963887214660645, "learning_rate": 3.890128308568782e-05, "loss": 1.9419, "step": 3467 }, { "epoch": 0.03119083701708391, "grad_norm": 3.144428014755249, "learning_rate": 3.889441093540026e-05, "loss": 2.1147, "step": 3468 }, { "epoch": 0.03119983091472436, "grad_norm": 1.9624632596969604, "learning_rate": 3.888753726564081e-05, "loss": 2.0284, "step": 3469 }, { "epoch": 0.03120882481236481, "grad_norm": 2.9691643714904785, "learning_rate": 3.888066207716117e-05, "loss": 2.1782, "step": 3470 }, { "epoch": 0.03121781871000526, "grad_norm": 2.1838502883911133, "learning_rate": 3.887378537071319e-05, "loss": 2.1739, "step": 3471 }, { "epoch": 0.03122681260764571, "grad_norm": 5.128747463226318, "learning_rate": 3.8866907147048895e-05, "loss": 2.6193, "step": 3472 }, { "epoch": 0.031235806505286165, "grad_norm": 1.8588112592697144, "learning_rate": 3.8860027406920494e-05, "loss": 2.4331, "step": 3473 }, { "epoch": 0.031244800402926615, "grad_norm": 3.0770623683929443, "learning_rate": 3.8853146151080335e-05, "loss": 1.8923, "step": 3474 }, { "epoch": 0.031253794300567066, "grad_norm": 2.2176811695098877, "learning_rate": 3.884626338028094e-05, "loss": 2.406, "step": 3475 }, { "epoch": 0.03126278819820752, "grad_norm": 4.210079193115234, "learning_rate": 3.8839379095274986e-05, "loss": 1.9828, "step": 3476 }, { "epoch": 0.031271782095847966, "grad_norm": 3.7623074054718018, "learning_rate": 3.883249329681535e-05, "loss": 1.7264, "step": 3477 }, { "epoch": 0.03128077599348842, "grad_norm": 10.11509895324707, "learning_rate": 3.882560598565504e-05, "loss": 2.2756, "step": 3478 }, { "epoch": 0.03128976989112887, "grad_norm": 1.971498966217041, "learning_rate": 3.881871716254726e-05, "loss": 2.0913, "step": 3479 }, { "epoch": 0.03129876378876932, "grad_norm": 8.41202163696289, "learning_rate": 3.8811826828245334e-05, "loss": 2.0587, "step": 3480 }, { "epoch": 0.03130775768640977, "grad_norm": 4.420191764831543, "learning_rate": 3.8804934983502794e-05, "loss": 1.5714, "step": 3481 }, { "epoch": 0.03131675158405022, "grad_norm": 1.5411851406097412, "learning_rate": 3.879804162907332e-05, "loss": 2.4808, "step": 3482 }, { "epoch": 0.031325745481690676, "grad_norm": 3.627286434173584, "learning_rate": 3.879114676571076e-05, "loss": 2.4133, "step": 3483 }, { "epoch": 0.03133473937933112, "grad_norm": 3.791470527648926, "learning_rate": 3.878425039416912e-05, "loss": 2.2478, "step": 3484 }, { "epoch": 0.031343733276971576, "grad_norm": 3.572005271911621, "learning_rate": 3.877735251520258e-05, "loss": 1.6297, "step": 3485 }, { "epoch": 0.03135272717461202, "grad_norm": 4.798781871795654, "learning_rate": 3.8770453129565496e-05, "loss": 2.292, "step": 3486 }, { "epoch": 0.03136172107225248, "grad_norm": 5.599769115447998, "learning_rate": 3.876355223801235e-05, "loss": 2.321, "step": 3487 }, { "epoch": 0.03137071496989293, "grad_norm": 1.904028296470642, "learning_rate": 3.8756649841297824e-05, "loss": 2.3071, "step": 3488 }, { "epoch": 0.03137970886753338, "grad_norm": 4.771520614624023, "learning_rate": 3.8749745940176765e-05, "loss": 2.3209, "step": 3489 }, { "epoch": 0.03138870276517383, "grad_norm": 1.6609514951705933, "learning_rate": 3.874284053540416e-05, "loss": 2.476, "step": 3490 }, { "epoch": 0.03139769666281428, "grad_norm": 4.292078971862793, "learning_rate": 3.873593362773517e-05, "loss": 2.3689, "step": 3491 }, { "epoch": 0.03140669056045473, "grad_norm": 4.309085845947266, "learning_rate": 3.872902521792513e-05, "loss": 2.1333, "step": 3492 }, { "epoch": 0.03141568445809518, "grad_norm": 7.990841388702393, "learning_rate": 3.872211530672953e-05, "loss": 1.6088, "step": 3493 }, { "epoch": 0.03142467835573563, "grad_norm": 2.877628803253174, "learning_rate": 3.871520389490404e-05, "loss": 2.3859, "step": 3494 }, { "epoch": 0.03143367225337609, "grad_norm": 4.394637584686279, "learning_rate": 3.870829098320446e-05, "loss": 2.0941, "step": 3495 }, { "epoch": 0.031442666151016534, "grad_norm": 4.008213043212891, "learning_rate": 3.87013765723868e-05, "loss": 2.2239, "step": 3496 }, { "epoch": 0.03145166004865699, "grad_norm": 2.9326043128967285, "learning_rate": 3.869446066320719e-05, "loss": 2.0962, "step": 3497 }, { "epoch": 0.031460653946297434, "grad_norm": 2.937758207321167, "learning_rate": 3.8687543256421955e-05, "loss": 2.4324, "step": 3498 }, { "epoch": 0.03146964784393789, "grad_norm": 4.198010444641113, "learning_rate": 3.868062435278758e-05, "loss": 1.8183, "step": 3499 }, { "epoch": 0.03147864174157834, "grad_norm": 3.929415464401245, "learning_rate": 3.867370395306068e-05, "loss": 2.5114, "step": 3500 }, { "epoch": 0.03148763563921879, "grad_norm": 2.634803295135498, "learning_rate": 3.866678205799807e-05, "loss": 2.1723, "step": 3501 }, { "epoch": 0.03149662953685924, "grad_norm": 2.7829864025115967, "learning_rate": 3.865985866835673e-05, "loss": 2.1344, "step": 3502 }, { "epoch": 0.03150562343449969, "grad_norm": 4.048957347869873, "learning_rate": 3.8652933784893784e-05, "loss": 1.7645, "step": 3503 }, { "epoch": 0.031514617332140143, "grad_norm": 4.2634148597717285, "learning_rate": 3.8646007408366525e-05, "loss": 2.4866, "step": 3504 }, { "epoch": 0.0315236112297806, "grad_norm": 2.4455463886260986, "learning_rate": 3.8639079539532405e-05, "loss": 2.5597, "step": 3505 }, { "epoch": 0.031532605127421044, "grad_norm": 3.591057538986206, "learning_rate": 3.863215017914906e-05, "loss": 1.8197, "step": 3506 }, { "epoch": 0.0315415990250615, "grad_norm": 3.94624662399292, "learning_rate": 3.862521932797427e-05, "loss": 2.5131, "step": 3507 }, { "epoch": 0.031550592922701945, "grad_norm": 2.908414125442505, "learning_rate": 3.861828698676597e-05, "loss": 2.0543, "step": 3508 }, { "epoch": 0.0315595868203424, "grad_norm": 2.9771840572357178, "learning_rate": 3.8611353156282294e-05, "loss": 2.0278, "step": 3509 }, { "epoch": 0.031568580717982846, "grad_norm": 3.1564671993255615, "learning_rate": 3.860441783728149e-05, "loss": 2.2524, "step": 3510 }, { "epoch": 0.0315775746156233, "grad_norm": 2.2181663513183594, "learning_rate": 3.8597481030522005e-05, "loss": 2.3862, "step": 3511 }, { "epoch": 0.03158656851326375, "grad_norm": 2.3167948722839355, "learning_rate": 3.859054273676244e-05, "loss": 2.044, "step": 3512 }, { "epoch": 0.0315955624109042, "grad_norm": 2.8708202838897705, "learning_rate": 3.858360295676155e-05, "loss": 2.1449, "step": 3513 }, { "epoch": 0.031604556308544654, "grad_norm": 3.4565467834472656, "learning_rate": 3.857666169127827e-05, "loss": 2.076, "step": 3514 }, { "epoch": 0.0316135502061851, "grad_norm": 2.5756142139434814, "learning_rate": 3.8569718941071684e-05, "loss": 2.0712, "step": 3515 }, { "epoch": 0.031622544103825555, "grad_norm": 3.5053467750549316, "learning_rate": 3.856277470690103e-05, "loss": 1.5613, "step": 3516 }, { "epoch": 0.03163153800146601, "grad_norm": 3.492436170578003, "learning_rate": 3.855582898952572e-05, "loss": 1.804, "step": 3517 }, { "epoch": 0.031640531899106455, "grad_norm": 2.396042585372925, "learning_rate": 3.8548881789705345e-05, "loss": 1.9452, "step": 3518 }, { "epoch": 0.03164952579674691, "grad_norm": 2.1336233615875244, "learning_rate": 3.8541933108199624e-05, "loss": 1.9892, "step": 3519 }, { "epoch": 0.031658519694387356, "grad_norm": 4.521316051483154, "learning_rate": 3.853498294576845e-05, "loss": 2.1237, "step": 3520 }, { "epoch": 0.03166751359202781, "grad_norm": 3.4189276695251465, "learning_rate": 3.8528031303171895e-05, "loss": 1.9767, "step": 3521 }, { "epoch": 0.03167650748966826, "grad_norm": 2.378126621246338, "learning_rate": 3.8521078181170176e-05, "loss": 1.9506, "step": 3522 }, { "epoch": 0.03168550138730871, "grad_norm": 3.5882856845855713, "learning_rate": 3.8514123580523674e-05, "loss": 2.1895, "step": 3523 }, { "epoch": 0.031694495284949165, "grad_norm": 3.160428047180176, "learning_rate": 3.8507167501992944e-05, "loss": 2.1638, "step": 3524 }, { "epoch": 0.03170348918258961, "grad_norm": 3.025989055633545, "learning_rate": 3.850020994633868e-05, "loss": 2.4785, "step": 3525 }, { "epoch": 0.031712483080230065, "grad_norm": 1.903955101966858, "learning_rate": 3.8493250914321756e-05, "loss": 1.8913, "step": 3526 }, { "epoch": 0.03172147697787051, "grad_norm": 3.727863311767578, "learning_rate": 3.848629040670319e-05, "loss": 2.205, "step": 3527 }, { "epoch": 0.031730470875510966, "grad_norm": 3.433417320251465, "learning_rate": 3.847932842424419e-05, "loss": 1.9223, "step": 3528 }, { "epoch": 0.03173946477315142, "grad_norm": 3.622248888015747, "learning_rate": 3.8472364967706095e-05, "loss": 2.4645, "step": 3529 }, { "epoch": 0.03174845867079187, "grad_norm": 6.204473495483398, "learning_rate": 3.846540003785042e-05, "loss": 1.7768, "step": 3530 }, { "epoch": 0.03175745256843232, "grad_norm": 3.5668773651123047, "learning_rate": 3.8458433635438846e-05, "loss": 2.3077, "step": 3531 }, { "epoch": 0.03176644646607277, "grad_norm": 2.575246810913086, "learning_rate": 3.84514657612332e-05, "loss": 2.1128, "step": 3532 }, { "epoch": 0.03177544036371322, "grad_norm": 2.33121919631958, "learning_rate": 3.844449641599548e-05, "loss": 2.0423, "step": 3533 }, { "epoch": 0.03178443426135367, "grad_norm": 2.5544826984405518, "learning_rate": 3.8437525600487854e-05, "loss": 1.9218, "step": 3534 }, { "epoch": 0.03179342815899412, "grad_norm": 6.555283546447754, "learning_rate": 3.8430553315472626e-05, "loss": 1.9767, "step": 3535 }, { "epoch": 0.031802422056634576, "grad_norm": 2.5270814895629883, "learning_rate": 3.842357956171229e-05, "loss": 2.1676, "step": 3536 }, { "epoch": 0.03181141595427502, "grad_norm": 6.785141468048096, "learning_rate": 3.841660433996946e-05, "loss": 2.0503, "step": 3537 }, { "epoch": 0.03182040985191548, "grad_norm": 2.0753087997436523, "learning_rate": 3.840962765100696e-05, "loss": 2.2342, "step": 3538 }, { "epoch": 0.03182940374955592, "grad_norm": 2.792846202850342, "learning_rate": 3.8402649495587734e-05, "loss": 2.2407, "step": 3539 }, { "epoch": 0.03183839764719638, "grad_norm": 3.5721757411956787, "learning_rate": 3.8395669874474915e-05, "loss": 2.4847, "step": 3540 }, { "epoch": 0.03184739154483683, "grad_norm": 5.158585548400879, "learning_rate": 3.838868878843178e-05, "loss": 2.3833, "step": 3541 }, { "epoch": 0.03185638544247728, "grad_norm": 3.241743564605713, "learning_rate": 3.8381706238221764e-05, "loss": 1.7973, "step": 3542 }, { "epoch": 0.03186537934011773, "grad_norm": 4.5521979331970215, "learning_rate": 3.8374722224608476e-05, "loss": 2.1878, "step": 3543 }, { "epoch": 0.03187437323775818, "grad_norm": 6.24173641204834, "learning_rate": 3.8367736748355676e-05, "loss": 2.2303, "step": 3544 }, { "epoch": 0.03188336713539863, "grad_norm": 2.861536979675293, "learning_rate": 3.8360749810227286e-05, "loss": 2.4946, "step": 3545 }, { "epoch": 0.031892361033039086, "grad_norm": 3.829941749572754, "learning_rate": 3.835376141098738e-05, "loss": 2.4137, "step": 3546 }, { "epoch": 0.03190135493067953, "grad_norm": 1.5474756956100464, "learning_rate": 3.8346771551400204e-05, "loss": 1.7048, "step": 3547 }, { "epoch": 0.03191034882831999, "grad_norm": 8.787674903869629, "learning_rate": 3.8339780232230156e-05, "loss": 2.6529, "step": 3548 }, { "epoch": 0.031919342725960434, "grad_norm": 4.2612409591674805, "learning_rate": 3.833278745424179e-05, "loss": 1.7556, "step": 3549 }, { "epoch": 0.03192833662360089, "grad_norm": 2.5785458087921143, "learning_rate": 3.832579321819985e-05, "loss": 1.5553, "step": 3550 }, { "epoch": 0.031937330521241335, "grad_norm": 4.7794880867004395, "learning_rate": 3.8318797524869185e-05, "loss": 2.4033, "step": 3551 }, { "epoch": 0.03194632441888179, "grad_norm": 2.783947467803955, "learning_rate": 3.8311800375014853e-05, "loss": 2.2016, "step": 3552 }, { "epoch": 0.03195531831652224, "grad_norm": 4.130613803863525, "learning_rate": 3.830480176940205e-05, "loss": 2.0639, "step": 3553 }, { "epoch": 0.03196431221416269, "grad_norm": 3.4136199951171875, "learning_rate": 3.829780170879612e-05, "loss": 1.8322, "step": 3554 }, { "epoch": 0.03197330611180314, "grad_norm": 3.853414535522461, "learning_rate": 3.829080019396259e-05, "loss": 1.8802, "step": 3555 }, { "epoch": 0.03198230000944359, "grad_norm": 3.033212900161743, "learning_rate": 3.828379722566714e-05, "loss": 1.5174, "step": 3556 }, { "epoch": 0.031991293907084044, "grad_norm": 2.575651168823242, "learning_rate": 3.82767928046756e-05, "loss": 1.8913, "step": 3557 }, { "epoch": 0.0320002878047245, "grad_norm": 3.3189454078674316, "learning_rate": 3.826978693175395e-05, "loss": 2.4363, "step": 3558 }, { "epoch": 0.032009281702364945, "grad_norm": 3.4226629734039307, "learning_rate": 3.826277960766835e-05, "loss": 2.0593, "step": 3559 }, { "epoch": 0.0320182756000054, "grad_norm": 3.8278696537017822, "learning_rate": 3.825577083318512e-05, "loss": 2.4017, "step": 3560 }, { "epoch": 0.032027269497645845, "grad_norm": 3.62677264213562, "learning_rate": 3.824876060907072e-05, "loss": 2.211, "step": 3561 }, { "epoch": 0.0320362633952863, "grad_norm": 3.8808176517486572, "learning_rate": 3.8241748936091775e-05, "loss": 2.2397, "step": 3562 }, { "epoch": 0.032045257292926746, "grad_norm": 2.032865524291992, "learning_rate": 3.8234735815015086e-05, "loss": 1.8563, "step": 3563 }, { "epoch": 0.0320542511905672, "grad_norm": 1.7940653562545776, "learning_rate": 3.8227721246607575e-05, "loss": 1.7148, "step": 3564 }, { "epoch": 0.032063245088207654, "grad_norm": 1.9061236381530762, "learning_rate": 3.822070523163636e-05, "loss": 2.1645, "step": 3565 }, { "epoch": 0.0320722389858481, "grad_norm": 5.571143627166748, "learning_rate": 3.82136877708687e-05, "loss": 2.0406, "step": 3566 }, { "epoch": 0.032081232883488554, "grad_norm": 2.4216103553771973, "learning_rate": 3.820666886507201e-05, "loss": 2.5823, "step": 3567 }, { "epoch": 0.032090226781129, "grad_norm": 2.7298367023468018, "learning_rate": 3.8199648515013864e-05, "loss": 2.0343, "step": 3568 }, { "epoch": 0.032099220678769455, "grad_norm": 2.6759443283081055, "learning_rate": 3.819262672146201e-05, "loss": 2.1596, "step": 3569 }, { "epoch": 0.03210821457640991, "grad_norm": 3.2961177825927734, "learning_rate": 3.818560348518433e-05, "loss": 2.0467, "step": 3570 }, { "epoch": 0.032117208474050356, "grad_norm": 3.198172092437744, "learning_rate": 3.8178578806948875e-05, "loss": 2.4879, "step": 3571 }, { "epoch": 0.03212620237169081, "grad_norm": 2.401170253753662, "learning_rate": 3.817155268752385e-05, "loss": 2.207, "step": 3572 }, { "epoch": 0.032135196269331257, "grad_norm": 2.897719383239746, "learning_rate": 3.816452512767764e-05, "loss": 2.3527, "step": 3573 }, { "epoch": 0.03214419016697171, "grad_norm": 2.514460325241089, "learning_rate": 3.815749612817874e-05, "loss": 2.1831, "step": 3574 }, { "epoch": 0.03215318406461216, "grad_norm": 2.8296806812286377, "learning_rate": 3.8150465689795854e-05, "loss": 2.5522, "step": 3575 }, { "epoch": 0.03216217796225261, "grad_norm": 1.4646590948104858, "learning_rate": 3.8143433813297815e-05, "loss": 1.756, "step": 3576 }, { "epoch": 0.032171171859893065, "grad_norm": 1.9657788276672363, "learning_rate": 3.8136400499453604e-05, "loss": 1.953, "step": 3577 }, { "epoch": 0.03218016575753351, "grad_norm": 2.1377246379852295, "learning_rate": 3.81293657490324e-05, "loss": 2.1823, "step": 3578 }, { "epoch": 0.032189159655173966, "grad_norm": 3.540621519088745, "learning_rate": 3.8122329562803485e-05, "loss": 2.3881, "step": 3579 }, { "epoch": 0.03219815355281441, "grad_norm": 2.9319546222686768, "learning_rate": 3.8115291941536345e-05, "loss": 2.1773, "step": 3580 }, { "epoch": 0.032207147450454866, "grad_norm": 3.585906982421875, "learning_rate": 3.810825288600059e-05, "loss": 2.3241, "step": 3581 }, { "epoch": 0.03221614134809532, "grad_norm": 3.2306952476501465, "learning_rate": 3.8101212396966016e-05, "loss": 2.0423, "step": 3582 }, { "epoch": 0.03222513524573577, "grad_norm": 2.6522815227508545, "learning_rate": 3.809417047520254e-05, "loss": 2.1615, "step": 3583 }, { "epoch": 0.03223412914337622, "grad_norm": 3.187256097793579, "learning_rate": 3.808712712148027e-05, "loss": 2.0722, "step": 3584 }, { "epoch": 0.03224312304101667, "grad_norm": 3.494325637817383, "learning_rate": 3.8080082336569455e-05, "loss": 2.3051, "step": 3585 }, { "epoch": 0.03225211693865712, "grad_norm": 2.0459847450256348, "learning_rate": 3.807303612124051e-05, "loss": 2.3233, "step": 3586 }, { "epoch": 0.032261110836297575, "grad_norm": 1.8676209449768066, "learning_rate": 3.806598847626398e-05, "loss": 1.5436, "step": 3587 }, { "epoch": 0.03227010473393802, "grad_norm": 3.6841347217559814, "learning_rate": 3.80589394024106e-05, "loss": 2.1745, "step": 3588 }, { "epoch": 0.032279098631578476, "grad_norm": 3.180654287338257, "learning_rate": 3.8051888900451244e-05, "loss": 2.4693, "step": 3589 }, { "epoch": 0.03228809252921892, "grad_norm": 1.85630202293396, "learning_rate": 3.8044836971156936e-05, "loss": 1.7829, "step": 3590 }, { "epoch": 0.03229708642685938, "grad_norm": 7.125909805297852, "learning_rate": 3.8037783615298865e-05, "loss": 2.2708, "step": 3591 }, { "epoch": 0.032306080324499824, "grad_norm": 2.74249529838562, "learning_rate": 3.803072883364839e-05, "loss": 2.3721, "step": 3592 }, { "epoch": 0.03231507422214028, "grad_norm": 1.9459012746810913, "learning_rate": 3.802367262697699e-05, "loss": 2.3997, "step": 3593 }, { "epoch": 0.03232406811978073, "grad_norm": 2.8532259464263916, "learning_rate": 3.801661499605634e-05, "loss": 2.1885, "step": 3594 }, { "epoch": 0.03233306201742118, "grad_norm": 3.435750722885132, "learning_rate": 3.800955594165825e-05, "loss": 1.8911, "step": 3595 }, { "epoch": 0.03234205591506163, "grad_norm": 3.6897807121276855, "learning_rate": 3.800249546455467e-05, "loss": 1.8469, "step": 3596 }, { "epoch": 0.03235104981270208, "grad_norm": 2.2515180110931396, "learning_rate": 3.7995433565517735e-05, "loss": 2.2665, "step": 3597 }, { "epoch": 0.03236004371034253, "grad_norm": 2.8472747802734375, "learning_rate": 3.798837024531974e-05, "loss": 2.4096, "step": 3598 }, { "epoch": 0.03236903760798299, "grad_norm": 4.182323932647705, "learning_rate": 3.798130550473309e-05, "loss": 1.7639, "step": 3599 }, { "epoch": 0.032378031505623434, "grad_norm": 5.866360187530518, "learning_rate": 3.797423934453038e-05, "loss": 2.2178, "step": 3600 }, { "epoch": 0.03238702540326389, "grad_norm": 3.1232759952545166, "learning_rate": 3.7967171765484375e-05, "loss": 2.1736, "step": 3601 }, { "epoch": 0.032396019300904334, "grad_norm": 3.263235092163086, "learning_rate": 3.7960102768367955e-05, "loss": 2.3317, "step": 3602 }, { "epoch": 0.03240501319854479, "grad_norm": 2.641037702560425, "learning_rate": 3.795303235395418e-05, "loss": 2.133, "step": 3603 }, { "epoch": 0.032414007096185235, "grad_norm": 2.944404363632202, "learning_rate": 3.7945960523016265e-05, "loss": 1.8482, "step": 3604 }, { "epoch": 0.03242300099382569, "grad_norm": 4.8066253662109375, "learning_rate": 3.793888727632757e-05, "loss": 1.9387, "step": 3605 }, { "epoch": 0.03243199489146614, "grad_norm": 3.1100966930389404, "learning_rate": 3.7931812614661616e-05, "loss": 2.6403, "step": 3606 }, { "epoch": 0.03244098878910659, "grad_norm": 2.546287775039673, "learning_rate": 3.792473653879208e-05, "loss": 2.7563, "step": 3607 }, { "epoch": 0.03244998268674704, "grad_norm": 2.576875686645508, "learning_rate": 3.7917659049492785e-05, "loss": 2.1862, "step": 3608 }, { "epoch": 0.03245897658438749, "grad_norm": 3.324324131011963, "learning_rate": 3.791058014753771e-05, "loss": 2.2901, "step": 3609 }, { "epoch": 0.032467970482027944, "grad_norm": 2.5381391048431396, "learning_rate": 3.7903499833701006e-05, "loss": 2.0741, "step": 3610 }, { "epoch": 0.0324769643796684, "grad_norm": 3.3372156620025635, "learning_rate": 3.789641810875697e-05, "loss": 2.283, "step": 3611 }, { "epoch": 0.032485958277308845, "grad_norm": 4.002575397491455, "learning_rate": 3.7889334973480034e-05, "loss": 2.1068, "step": 3612 }, { "epoch": 0.0324949521749493, "grad_norm": 4.788218975067139, "learning_rate": 3.7882250428644795e-05, "loss": 2.6056, "step": 3613 }, { "epoch": 0.032503946072589746, "grad_norm": 1.9388561248779297, "learning_rate": 3.7875164475026025e-05, "loss": 1.9878, "step": 3614 }, { "epoch": 0.0325129399702302, "grad_norm": 3.523447275161743, "learning_rate": 3.786807711339863e-05, "loss": 1.8934, "step": 3615 }, { "epoch": 0.03252193386787065, "grad_norm": 2.511859893798828, "learning_rate": 3.786098834453766e-05, "loss": 1.819, "step": 3616 }, { "epoch": 0.0325309277655111, "grad_norm": 4.583134651184082, "learning_rate": 3.7853898169218346e-05, "loss": 2.2161, "step": 3617 }, { "epoch": 0.032539921663151554, "grad_norm": 5.1994452476501465, "learning_rate": 3.7846806588216055e-05, "loss": 2.1893, "step": 3618 }, { "epoch": 0.032548915560792, "grad_norm": 2.683497667312622, "learning_rate": 3.78397136023063e-05, "loss": 2.1648, "step": 3619 }, { "epoch": 0.032557909458432455, "grad_norm": 2.1133499145507812, "learning_rate": 3.783261921226479e-05, "loss": 1.8911, "step": 3620 }, { "epoch": 0.0325669033560729, "grad_norm": 3.722426176071167, "learning_rate": 3.782552341886733e-05, "loss": 1.8309, "step": 3621 }, { "epoch": 0.032575897253713355, "grad_norm": 2.542320728302002, "learning_rate": 3.781842622288989e-05, "loss": 2.2495, "step": 3622 }, { "epoch": 0.03258489115135381, "grad_norm": 5.489415645599365, "learning_rate": 3.781132762510865e-05, "loss": 2.1178, "step": 3623 }, { "epoch": 0.032593885048994256, "grad_norm": 2.865180015563965, "learning_rate": 3.780422762629988e-05, "loss": 2.677, "step": 3624 }, { "epoch": 0.03260287894663471, "grad_norm": 2.529294729232788, "learning_rate": 3.779712622724003e-05, "loss": 1.7928, "step": 3625 }, { "epoch": 0.03261187284427516, "grad_norm": 5.074844837188721, "learning_rate": 3.779002342870569e-05, "loss": 2.272, "step": 3626 }, { "epoch": 0.03262086674191561, "grad_norm": 3.5045993328094482, "learning_rate": 3.7782919231473626e-05, "loss": 2.2299, "step": 3627 }, { "epoch": 0.032629860639556064, "grad_norm": 3.2225253582000732, "learning_rate": 3.777581363632072e-05, "loss": 2.083, "step": 3628 }, { "epoch": 0.03263885453719651, "grad_norm": 2.674414873123169, "learning_rate": 3.776870664402405e-05, "loss": 1.9905, "step": 3629 }, { "epoch": 0.032647848434836965, "grad_norm": 8.024659156799316, "learning_rate": 3.776159825536082e-05, "loss": 2.4764, "step": 3630 }, { "epoch": 0.03265684233247741, "grad_norm": 2.254582405090332, "learning_rate": 3.7754488471108395e-05, "loss": 2.2806, "step": 3631 }, { "epoch": 0.032665836230117866, "grad_norm": 2.9342548847198486, "learning_rate": 3.7747377292044276e-05, "loss": 1.5559, "step": 3632 }, { "epoch": 0.03267483012775831, "grad_norm": 5.905022621154785, "learning_rate": 3.774026471894615e-05, "loss": 2.2757, "step": 3633 }, { "epoch": 0.03268382402539877, "grad_norm": 3.5302224159240723, "learning_rate": 3.7733150752591826e-05, "loss": 1.9009, "step": 3634 }, { "epoch": 0.03269281792303922, "grad_norm": 1.6774449348449707, "learning_rate": 3.7726035393759285e-05, "loss": 2.077, "step": 3635 }, { "epoch": 0.03270181182067967, "grad_norm": 4.211711406707764, "learning_rate": 3.771891864322664e-05, "loss": 2.1872, "step": 3636 }, { "epoch": 0.03271080571832012, "grad_norm": 2.0395703315734863, "learning_rate": 3.771180050177218e-05, "loss": 2.4258, "step": 3637 }, { "epoch": 0.03271979961596057, "grad_norm": 3.2059576511383057, "learning_rate": 3.770468097017432e-05, "loss": 2.1556, "step": 3638 }, { "epoch": 0.03272879351360102, "grad_norm": 1.988653302192688, "learning_rate": 3.7697560049211666e-05, "loss": 1.832, "step": 3639 }, { "epoch": 0.032737787411241476, "grad_norm": 3.88749098777771, "learning_rate": 3.7690437739662924e-05, "loss": 2.3714, "step": 3640 }, { "epoch": 0.03274678130888192, "grad_norm": 3.184058666229248, "learning_rate": 3.7683314042306996e-05, "loss": 2.5639, "step": 3641 }, { "epoch": 0.032755775206522376, "grad_norm": 2.0092668533325195, "learning_rate": 3.767618895792291e-05, "loss": 1.9798, "step": 3642 }, { "epoch": 0.03276476910416282, "grad_norm": 1.3985826969146729, "learning_rate": 3.766906248728986e-05, "loss": 2.0641, "step": 3643 }, { "epoch": 0.03277376300180328, "grad_norm": 4.978219032287598, "learning_rate": 3.766193463118718e-05, "loss": 2.2991, "step": 3644 }, { "epoch": 0.032782756899443724, "grad_norm": 2.7482504844665527, "learning_rate": 3.7654805390394366e-05, "loss": 2.2874, "step": 3645 }, { "epoch": 0.03279175079708418, "grad_norm": 4.194744110107422, "learning_rate": 3.764767476569106e-05, "loss": 2.5486, "step": 3646 }, { "epoch": 0.03280074469472463, "grad_norm": 4.428000450134277, "learning_rate": 3.764054275785707e-05, "loss": 1.7748, "step": 3647 }, { "epoch": 0.03280973859236508, "grad_norm": 3.4651167392730713, "learning_rate": 3.763340936767231e-05, "loss": 2.2475, "step": 3648 }, { "epoch": 0.03281873249000553, "grad_norm": 3.23488450050354, "learning_rate": 3.762627459591691e-05, "loss": 2.6113, "step": 3649 }, { "epoch": 0.03282772638764598, "grad_norm": 2.4242985248565674, "learning_rate": 3.76191384433711e-05, "loss": 1.9367, "step": 3650 }, { "epoch": 0.03283672028528643, "grad_norm": 6.916306018829346, "learning_rate": 3.761200091081528e-05, "loss": 2.0708, "step": 3651 }, { "epoch": 0.03284571418292689, "grad_norm": 3.7169854640960693, "learning_rate": 3.760486199903e-05, "loss": 2.2996, "step": 3652 }, { "epoch": 0.032854708080567334, "grad_norm": 2.9658379554748535, "learning_rate": 3.759772170879596e-05, "loss": 2.1286, "step": 3653 }, { "epoch": 0.03286370197820779, "grad_norm": 1.941292405128479, "learning_rate": 3.759058004089402e-05, "loss": 1.9222, "step": 3654 }, { "epoch": 0.032872695875848235, "grad_norm": 3.252181053161621, "learning_rate": 3.7583436996105185e-05, "loss": 1.9644, "step": 3655 }, { "epoch": 0.03288168977348869, "grad_norm": 2.739088296890259, "learning_rate": 3.757629257521058e-05, "loss": 1.7514, "step": 3656 }, { "epoch": 0.03289068367112914, "grad_norm": 2.9195210933685303, "learning_rate": 3.7569146778991534e-05, "loss": 2.0576, "step": 3657 }, { "epoch": 0.03289967756876959, "grad_norm": 6.6456074714660645, "learning_rate": 3.7561999608229505e-05, "loss": 2.3358, "step": 3658 }, { "epoch": 0.03290867146641004, "grad_norm": 2.6613645553588867, "learning_rate": 3.7554851063706074e-05, "loss": 2.1524, "step": 3659 }, { "epoch": 0.03291766536405049, "grad_norm": 2.5108654499053955, "learning_rate": 3.7547701146203005e-05, "loss": 2.0349, "step": 3660 }, { "epoch": 0.032926659261690944, "grad_norm": 4.680249214172363, "learning_rate": 3.754054985650222e-05, "loss": 2.2897, "step": 3661 }, { "epoch": 0.03293565315933139, "grad_norm": 3.82123064994812, "learning_rate": 3.753339719538574e-05, "loss": 2.2587, "step": 3662 }, { "epoch": 0.032944647056971844, "grad_norm": 2.8815603256225586, "learning_rate": 3.752624316363579e-05, "loss": 2.0128, "step": 3663 }, { "epoch": 0.0329536409546123, "grad_norm": 2.5879006385803223, "learning_rate": 3.751908776203473e-05, "loss": 2.306, "step": 3664 }, { "epoch": 0.032962634852252745, "grad_norm": 3.3421854972839355, "learning_rate": 3.751193099136505e-05, "loss": 1.8794, "step": 3665 }, { "epoch": 0.0329716287498932, "grad_norm": 3.7255046367645264, "learning_rate": 3.750477285240941e-05, "loss": 2.1195, "step": 3666 }, { "epoch": 0.032980622647533646, "grad_norm": 1.84585702419281, "learning_rate": 3.74976133459506e-05, "loss": 2.4184, "step": 3667 }, { "epoch": 0.0329896165451741, "grad_norm": 2.827592372894287, "learning_rate": 3.7490452472771594e-05, "loss": 1.9807, "step": 3668 }, { "epoch": 0.032998610442814554, "grad_norm": 3.0318124294281006, "learning_rate": 3.748329023365548e-05, "loss": 2.2866, "step": 3669 }, { "epoch": 0.033007604340455, "grad_norm": 2.3541648387908936, "learning_rate": 3.747612662938552e-05, "loss": 2.3768, "step": 3670 }, { "epoch": 0.033016598238095454, "grad_norm": 3.3855743408203125, "learning_rate": 3.74689616607451e-05, "loss": 2.4701, "step": 3671 }, { "epoch": 0.0330255921357359, "grad_norm": 4.738195419311523, "learning_rate": 3.746179532851779e-05, "loss": 1.6529, "step": 3672 }, { "epoch": 0.033034586033376355, "grad_norm": 1.9674732685089111, "learning_rate": 3.7454627633487274e-05, "loss": 1.6271, "step": 3673 }, { "epoch": 0.0330435799310168, "grad_norm": 2.947364091873169, "learning_rate": 3.7447458576437405e-05, "loss": 2.5686, "step": 3674 }, { "epoch": 0.033052573828657256, "grad_norm": 2.341383218765259, "learning_rate": 3.7440288158152187e-05, "loss": 2.2074, "step": 3675 }, { "epoch": 0.03306156772629771, "grad_norm": 2.6924619674682617, "learning_rate": 3.7433116379415754e-05, "loss": 1.699, "step": 3676 }, { "epoch": 0.033070561623938156, "grad_norm": 2.941817045211792, "learning_rate": 3.7425943241012405e-05, "loss": 2.1555, "step": 3677 }, { "epoch": 0.03307955552157861, "grad_norm": 2.940814733505249, "learning_rate": 3.741876874372659e-05, "loss": 2.3618, "step": 3678 }, { "epoch": 0.03308854941921906, "grad_norm": 2.2626335620880127, "learning_rate": 3.741159288834289e-05, "loss": 2.6795, "step": 3679 }, { "epoch": 0.03309754331685951, "grad_norm": 2.7021703720092773, "learning_rate": 3.7404415675646054e-05, "loss": 2.4216, "step": 3680 }, { "epoch": 0.033106537214499965, "grad_norm": 2.706064224243164, "learning_rate": 3.7397237106420965e-05, "loss": 2.2983, "step": 3681 }, { "epoch": 0.03311553111214041, "grad_norm": 2.332305669784546, "learning_rate": 3.7390057181452674e-05, "loss": 2.6016, "step": 3682 }, { "epoch": 0.033124525009780866, "grad_norm": 4.062191486358643, "learning_rate": 3.7382875901526354e-05, "loss": 2.1072, "step": 3683 }, { "epoch": 0.03313351890742131, "grad_norm": 2.48337721824646, "learning_rate": 3.7375693267427346e-05, "loss": 2.5879, "step": 3684 }, { "epoch": 0.033142512805061766, "grad_norm": 10.434313774108887, "learning_rate": 3.7368509279941134e-05, "loss": 2.4072, "step": 3685 }, { "epoch": 0.03315150670270221, "grad_norm": 2.121394634246826, "learning_rate": 3.736132393985333e-05, "loss": 2.4512, "step": 3686 }, { "epoch": 0.03316050060034267, "grad_norm": 3.2384369373321533, "learning_rate": 3.7354137247949735e-05, "loss": 1.7317, "step": 3687 }, { "epoch": 0.03316949449798312, "grad_norm": 2.621616840362549, "learning_rate": 3.7346949205016255e-05, "loss": 2.1395, "step": 3688 }, { "epoch": 0.03317848839562357, "grad_norm": 6.294168949127197, "learning_rate": 3.733975981183898e-05, "loss": 2.3113, "step": 3689 }, { "epoch": 0.03318748229326402, "grad_norm": 2.971346855163574, "learning_rate": 3.733256906920413e-05, "loss": 1.9401, "step": 3690 }, { "epoch": 0.03319647619090447, "grad_norm": 4.003098487854004, "learning_rate": 3.7325376977898055e-05, "loss": 1.9229, "step": 3691 }, { "epoch": 0.03320547008854492, "grad_norm": 2.0209527015686035, "learning_rate": 3.731818353870729e-05, "loss": 2.1276, "step": 3692 }, { "epoch": 0.033214463986185376, "grad_norm": 2.9003236293792725, "learning_rate": 3.7310988752418496e-05, "loss": 1.8771, "step": 3693 }, { "epoch": 0.03322345788382582, "grad_norm": 3.282451629638672, "learning_rate": 3.7303792619818485e-05, "loss": 2.2805, "step": 3694 }, { "epoch": 0.03323245178146628, "grad_norm": 2.714627504348755, "learning_rate": 3.72965951416942e-05, "loss": 2.485, "step": 3695 }, { "epoch": 0.033241445679106724, "grad_norm": 4.425670146942139, "learning_rate": 3.728939631883277e-05, "loss": 2.4304, "step": 3696 }, { "epoch": 0.03325043957674718, "grad_norm": 2.7936434745788574, "learning_rate": 3.728219615202143e-05, "loss": 2.0774, "step": 3697 }, { "epoch": 0.03325943347438763, "grad_norm": 3.993917942047119, "learning_rate": 3.727499464204758e-05, "loss": 2.1595, "step": 3698 }, { "epoch": 0.03326842737202808, "grad_norm": 1.936387062072754, "learning_rate": 3.7267791789698776e-05, "loss": 2.1414, "step": 3699 }, { "epoch": 0.03327742126966853, "grad_norm": 2.43814754486084, "learning_rate": 3.726058759576271e-05, "loss": 2.4035, "step": 3700 }, { "epoch": 0.03328641516730898, "grad_norm": 3.4474198818206787, "learning_rate": 3.725338206102721e-05, "loss": 1.767, "step": 3701 }, { "epoch": 0.03329540906494943, "grad_norm": 4.928455829620361, "learning_rate": 3.724617518628027e-05, "loss": 2.6065, "step": 3702 }, { "epoch": 0.03330440296258988, "grad_norm": 2.974198579788208, "learning_rate": 3.723896697231003e-05, "loss": 2.0146, "step": 3703 }, { "epoch": 0.033313396860230333, "grad_norm": 2.1911609172821045, "learning_rate": 3.723175741990476e-05, "loss": 2.4322, "step": 3704 }, { "epoch": 0.03332239075787079, "grad_norm": 3.333528757095337, "learning_rate": 3.722454652985289e-05, "loss": 2.4274, "step": 3705 }, { "epoch": 0.033331384655511234, "grad_norm": 3.4251928329467773, "learning_rate": 3.7217334302942994e-05, "loss": 2.5013, "step": 3706 }, { "epoch": 0.03334037855315169, "grad_norm": 2.430386543273926, "learning_rate": 3.721012073996378e-05, "loss": 2.5309, "step": 3707 }, { "epoch": 0.033349372450792135, "grad_norm": 2.9592299461364746, "learning_rate": 3.7202905841704124e-05, "loss": 2.2791, "step": 3708 }, { "epoch": 0.03335836634843259, "grad_norm": 6.909228801727295, "learning_rate": 3.719568960895303e-05, "loss": 1.9603, "step": 3709 }, { "epoch": 0.03336736024607304, "grad_norm": 6.584043025970459, "learning_rate": 3.718847204249966e-05, "loss": 1.9682, "step": 3710 }, { "epoch": 0.03337635414371349, "grad_norm": 3.3402557373046875, "learning_rate": 3.718125314313331e-05, "loss": 1.7861, "step": 3711 }, { "epoch": 0.03338534804135394, "grad_norm": 3.1692512035369873, "learning_rate": 3.717403291164343e-05, "loss": 1.8661, "step": 3712 }, { "epoch": 0.03339434193899439, "grad_norm": 3.192655563354492, "learning_rate": 3.716681134881962e-05, "loss": 1.8675, "step": 3713 }, { "epoch": 0.033403335836634844, "grad_norm": 4.599348068237305, "learning_rate": 3.71595884554516e-05, "loss": 2.1068, "step": 3714 }, { "epoch": 0.03341232973427529, "grad_norm": 1.985348105430603, "learning_rate": 3.715236423232928e-05, "loss": 2.1163, "step": 3715 }, { "epoch": 0.033421323631915745, "grad_norm": 2.8290324211120605, "learning_rate": 3.7145138680242675e-05, "loss": 2.2291, "step": 3716 }, { "epoch": 0.0334303175295562, "grad_norm": 2.2228057384490967, "learning_rate": 3.713791179998197e-05, "loss": 2.0558, "step": 3717 }, { "epoch": 0.033439311427196645, "grad_norm": 3.440495014190674, "learning_rate": 3.7130683592337465e-05, "loss": 1.8358, "step": 3718 }, { "epoch": 0.0334483053248371, "grad_norm": 2.2403981685638428, "learning_rate": 3.712345405809965e-05, "loss": 2.2229, "step": 3719 }, { "epoch": 0.033457299222477546, "grad_norm": 3.4343960285186768, "learning_rate": 3.711622319805913e-05, "loss": 1.8858, "step": 3720 }, { "epoch": 0.033466293120118, "grad_norm": 3.764143943786621, "learning_rate": 3.710899101300665e-05, "loss": 2.0621, "step": 3721 }, { "epoch": 0.033475287017758454, "grad_norm": 2.269690752029419, "learning_rate": 3.710175750373312e-05, "loss": 1.7823, "step": 3722 }, { "epoch": 0.0334842809153989, "grad_norm": 2.71130633354187, "learning_rate": 3.7094522671029586e-05, "loss": 1.9636, "step": 3723 }, { "epoch": 0.033493274813039355, "grad_norm": 2.2742295265197754, "learning_rate": 3.708728651568724e-05, "loss": 2.2869, "step": 3724 }, { "epoch": 0.0335022687106798, "grad_norm": 2.09279203414917, "learning_rate": 3.708004903849741e-05, "loss": 1.7631, "step": 3725 }, { "epoch": 0.033511262608320255, "grad_norm": 2.344817876815796, "learning_rate": 3.7072810240251574e-05, "loss": 2.2187, "step": 3726 }, { "epoch": 0.03352025650596071, "grad_norm": 9.60000991821289, "learning_rate": 3.7065570121741366e-05, "loss": 2.2814, "step": 3727 }, { "epoch": 0.033529250403601156, "grad_norm": 2.1673452854156494, "learning_rate": 3.7058328683758545e-05, "loss": 1.6, "step": 3728 }, { "epoch": 0.03353824430124161, "grad_norm": 2.533066749572754, "learning_rate": 3.705108592709503e-05, "loss": 1.9013, "step": 3729 }, { "epoch": 0.03354723819888206, "grad_norm": 2.59710955619812, "learning_rate": 3.704384185254288e-05, "loss": 2.3502, "step": 3730 }, { "epoch": 0.03355623209652251, "grad_norm": 2.1654772758483887, "learning_rate": 3.7036596460894296e-05, "loss": 1.847, "step": 3731 }, { "epoch": 0.03356522599416296, "grad_norm": 2.457502603530884, "learning_rate": 3.7029349752941616e-05, "loss": 2.2298, "step": 3732 }, { "epoch": 0.03357421989180341, "grad_norm": 5.557810306549072, "learning_rate": 3.702210172947733e-05, "loss": 2.176, "step": 3733 }, { "epoch": 0.033583213789443865, "grad_norm": 3.011566638946533, "learning_rate": 3.701485239129408e-05, "loss": 2.1796, "step": 3734 }, { "epoch": 0.03359220768708431, "grad_norm": 2.700047254562378, "learning_rate": 3.700760173918463e-05, "loss": 2.4089, "step": 3735 }, { "epoch": 0.033601201584724766, "grad_norm": 3.1517040729522705, "learning_rate": 3.700034977394192e-05, "loss": 2.178, "step": 3736 }, { "epoch": 0.03361019548236521, "grad_norm": 2.9928436279296875, "learning_rate": 3.699309649635899e-05, "loss": 2.3296, "step": 3737 }, { "epoch": 0.03361918938000567, "grad_norm": 4.997657775878906, "learning_rate": 3.698584190722906e-05, "loss": 1.9493, "step": 3738 }, { "epoch": 0.03362818327764612, "grad_norm": 2.234428644180298, "learning_rate": 3.697858600734548e-05, "loss": 2.0124, "step": 3739 }, { "epoch": 0.03363717717528657, "grad_norm": 2.969073534011841, "learning_rate": 3.697132879750174e-05, "loss": 2.0819, "step": 3740 }, { "epoch": 0.03364617107292702, "grad_norm": 6.490259170532227, "learning_rate": 3.696407027849149e-05, "loss": 2.0347, "step": 3741 }, { "epoch": 0.03365516497056747, "grad_norm": 3.343667507171631, "learning_rate": 3.695681045110849e-05, "loss": 2.0339, "step": 3742 }, { "epoch": 0.03366415886820792, "grad_norm": 3.6742918491363525, "learning_rate": 3.694954931614668e-05, "loss": 2.7793, "step": 3743 }, { "epoch": 0.03367315276584837, "grad_norm": 3.5231597423553467, "learning_rate": 3.694228687440013e-05, "loss": 2.188, "step": 3744 }, { "epoch": 0.03368214666348882, "grad_norm": 3.392496109008789, "learning_rate": 3.693502312666304e-05, "loss": 2.4017, "step": 3745 }, { "epoch": 0.033691140561129276, "grad_norm": 2.1971096992492676, "learning_rate": 3.692775807372976e-05, "loss": 2.085, "step": 3746 }, { "epoch": 0.03370013445876972, "grad_norm": 3.105449914932251, "learning_rate": 3.692049171639479e-05, "loss": 2.1241, "step": 3747 }, { "epoch": 0.03370912835641018, "grad_norm": 3.5273385047912598, "learning_rate": 3.691322405545278e-05, "loss": 2.366, "step": 3748 }, { "epoch": 0.033718122254050624, "grad_norm": 3.117230176925659, "learning_rate": 3.690595509169848e-05, "loss": 2.2652, "step": 3749 }, { "epoch": 0.03372711615169108, "grad_norm": 3.244121551513672, "learning_rate": 3.689868482592684e-05, "loss": 2.2483, "step": 3750 }, { "epoch": 0.03373611004933153, "grad_norm": 2.078463554382324, "learning_rate": 3.6891413258932916e-05, "loss": 2.2042, "step": 3751 }, { "epoch": 0.03374510394697198, "grad_norm": 2.426305055618286, "learning_rate": 3.688414039151192e-05, "loss": 2.1636, "step": 3752 }, { "epoch": 0.03375409784461243, "grad_norm": 2.087386131286621, "learning_rate": 3.68768662244592e-05, "loss": 1.9401, "step": 3753 }, { "epoch": 0.03376309174225288, "grad_norm": 2.529675006866455, "learning_rate": 3.686959075857024e-05, "loss": 2.2346, "step": 3754 }, { "epoch": 0.03377208563989333, "grad_norm": 3.9559578895568848, "learning_rate": 3.6862313994640686e-05, "loss": 1.6469, "step": 3755 }, { "epoch": 0.03378107953753378, "grad_norm": 2.3263845443725586, "learning_rate": 3.68550359334663e-05, "loss": 1.9114, "step": 3756 }, { "epoch": 0.033790073435174234, "grad_norm": 3.5935544967651367, "learning_rate": 3.684775657584302e-05, "loss": 2.5454, "step": 3757 }, { "epoch": 0.03379906733281469, "grad_norm": 2.978955030441284, "learning_rate": 3.6840475922566886e-05, "loss": 2.0409, "step": 3758 }, { "epoch": 0.033808061230455135, "grad_norm": 3.7101871967315674, "learning_rate": 3.68331939744341e-05, "loss": 2.2108, "step": 3759 }, { "epoch": 0.03381705512809559, "grad_norm": 4.3177008628845215, "learning_rate": 3.6825910732241026e-05, "loss": 1.6992, "step": 3760 }, { "epoch": 0.033826049025736035, "grad_norm": 3.244455099105835, "learning_rate": 3.681862619678413e-05, "loss": 2.4985, "step": 3761 }, { "epoch": 0.03383504292337649, "grad_norm": 2.9795994758605957, "learning_rate": 3.681134036886003e-05, "loss": 2.4646, "step": 3762 }, { "epoch": 0.03384403682101694, "grad_norm": 1.7908806800842285, "learning_rate": 3.680405324926553e-05, "loss": 2.2012, "step": 3763 }, { "epoch": 0.03385303071865739, "grad_norm": 2.8807101249694824, "learning_rate": 3.679676483879749e-05, "loss": 1.5361, "step": 3764 }, { "epoch": 0.033862024616297844, "grad_norm": 3.1628971099853516, "learning_rate": 3.678947513825299e-05, "loss": 1.9127, "step": 3765 }, { "epoch": 0.03387101851393829, "grad_norm": 2.5691585540771484, "learning_rate": 3.678218414842922e-05, "loss": 2.0259, "step": 3766 }, { "epoch": 0.033880012411578744, "grad_norm": 3.158815860748291, "learning_rate": 3.677489187012351e-05, "loss": 2.5155, "step": 3767 }, { "epoch": 0.0338890063092192, "grad_norm": 3.130234956741333, "learning_rate": 3.6767598304133324e-05, "loss": 1.9755, "step": 3768 }, { "epoch": 0.033898000206859645, "grad_norm": 1.8863455057144165, "learning_rate": 3.676030345125627e-05, "loss": 1.8456, "step": 3769 }, { "epoch": 0.0339069941045001, "grad_norm": 7.629554271697998, "learning_rate": 3.675300731229012e-05, "loss": 1.9925, "step": 3770 }, { "epoch": 0.033915988002140546, "grad_norm": 2.873331069946289, "learning_rate": 3.6745709888032765e-05, "loss": 2.4921, "step": 3771 }, { "epoch": 0.033924981899781, "grad_norm": 3.6439945697784424, "learning_rate": 3.6738411179282225e-05, "loss": 2.1694, "step": 3772 }, { "epoch": 0.033933975797421446, "grad_norm": 1.7181792259216309, "learning_rate": 3.673111118683669e-05, "loss": 2.1458, "step": 3773 }, { "epoch": 0.0339429696950619, "grad_norm": 4.0575337409973145, "learning_rate": 3.672380991149448e-05, "loss": 2.0758, "step": 3774 }, { "epoch": 0.033951963592702354, "grad_norm": 2.806065559387207, "learning_rate": 3.671650735405404e-05, "loss": 2.3559, "step": 3775 }, { "epoch": 0.0339609574903428, "grad_norm": 2.891570568084717, "learning_rate": 3.6709203515313966e-05, "loss": 1.9812, "step": 3776 }, { "epoch": 0.033969951387983255, "grad_norm": 2.4879541397094727, "learning_rate": 3.670189839607301e-05, "loss": 2.3484, "step": 3777 }, { "epoch": 0.0339789452856237, "grad_norm": 6.784309387207031, "learning_rate": 3.6694591997130026e-05, "loss": 2.0366, "step": 3778 }, { "epoch": 0.033987939183264156, "grad_norm": 5.451055526733398, "learning_rate": 3.6687284319284055e-05, "loss": 2.272, "step": 3779 }, { "epoch": 0.03399693308090461, "grad_norm": 3.346271276473999, "learning_rate": 3.667997536333424e-05, "loss": 2.3799, "step": 3780 }, { "epoch": 0.034005926978545056, "grad_norm": 12.501424789428711, "learning_rate": 3.667266513007988e-05, "loss": 1.9236, "step": 3781 }, { "epoch": 0.03401492087618551, "grad_norm": 2.8113973140716553, "learning_rate": 3.666535362032041e-05, "loss": 1.6758, "step": 3782 }, { "epoch": 0.03402391477382596, "grad_norm": 2.711580991744995, "learning_rate": 3.665804083485541e-05, "loss": 2.4387, "step": 3783 }, { "epoch": 0.03403290867146641, "grad_norm": 5.514557838439941, "learning_rate": 3.665072677448459e-05, "loss": 2.0164, "step": 3784 }, { "epoch": 0.03404190256910686, "grad_norm": 3.560436487197876, "learning_rate": 3.6643411440007804e-05, "loss": 2.2012, "step": 3785 }, { "epoch": 0.03405089646674731, "grad_norm": 3.474595069885254, "learning_rate": 3.663609483222505e-05, "loss": 2.0536, "step": 3786 }, { "epoch": 0.034059890364387765, "grad_norm": 2.5318450927734375, "learning_rate": 3.662877695193646e-05, "loss": 1.8253, "step": 3787 }, { "epoch": 0.03406888426202821, "grad_norm": 5.580223560333252, "learning_rate": 3.6621457799942304e-05, "loss": 2.2689, "step": 3788 }, { "epoch": 0.034077878159668666, "grad_norm": 4.074001789093018, "learning_rate": 3.6614137377043e-05, "loss": 2.5064, "step": 3789 }, { "epoch": 0.03408687205730911, "grad_norm": 3.073441743850708, "learning_rate": 3.66068156840391e-05, "loss": 2.109, "step": 3790 }, { "epoch": 0.03409586595494957, "grad_norm": 2.411109209060669, "learning_rate": 3.6599492721731274e-05, "loss": 2.1947, "step": 3791 }, { "epoch": 0.03410485985259002, "grad_norm": 3.299988269805908, "learning_rate": 3.6592168490920365e-05, "loss": 2.4994, "step": 3792 }, { "epoch": 0.03411385375023047, "grad_norm": 1.9142554998397827, "learning_rate": 3.6584842992407346e-05, "loss": 2.0619, "step": 3793 }, { "epoch": 0.03412284764787092, "grad_norm": 7.04536771774292, "learning_rate": 3.657751622699331e-05, "loss": 1.8979, "step": 3794 }, { "epoch": 0.03413184154551137, "grad_norm": 2.0709383487701416, "learning_rate": 3.65701881954795e-05, "loss": 2.3251, "step": 3795 }, { "epoch": 0.03414083544315182, "grad_norm": 5.413897514343262, "learning_rate": 3.656285889866732e-05, "loss": 2.5721, "step": 3796 }, { "epoch": 0.03414982934079227, "grad_norm": 2.326537847518921, "learning_rate": 3.6555528337358263e-05, "loss": 1.8333, "step": 3797 }, { "epoch": 0.03415882323843272, "grad_norm": 3.3279459476470947, "learning_rate": 3.654819651235401e-05, "loss": 2.1022, "step": 3798 }, { "epoch": 0.03416781713607318, "grad_norm": 5.607339382171631, "learning_rate": 3.6540863424456334e-05, "loss": 2.3895, "step": 3799 }, { "epoch": 0.034176811033713624, "grad_norm": 5.1472883224487305, "learning_rate": 3.65335290744672e-05, "loss": 2.5505, "step": 3800 }, { "epoch": 0.03418580493135408, "grad_norm": 2.7614004611968994, "learning_rate": 3.6526193463188654e-05, "loss": 2.3254, "step": 3801 }, { "epoch": 0.034194798828994524, "grad_norm": 6.46312952041626, "learning_rate": 3.6518856591422934e-05, "loss": 2.5753, "step": 3802 }, { "epoch": 0.03420379272663498, "grad_norm": 2.9432220458984375, "learning_rate": 3.6511518459972365e-05, "loss": 1.959, "step": 3803 }, { "epoch": 0.03421278662427543, "grad_norm": 3.0341222286224365, "learning_rate": 3.650417906963945e-05, "loss": 2.6479, "step": 3804 }, { "epoch": 0.03422178052191588, "grad_norm": 3.7878615856170654, "learning_rate": 3.649683842122681e-05, "loss": 2.3322, "step": 3805 }, { "epoch": 0.03423077441955633, "grad_norm": 1.8143879175186157, "learning_rate": 3.6489496515537204e-05, "loss": 2.1482, "step": 3806 }, { "epoch": 0.03423976831719678, "grad_norm": 1.7759525775909424, "learning_rate": 3.6482153353373525e-05, "loss": 2.3916, "step": 3807 }, { "epoch": 0.03424876221483723, "grad_norm": 4.064060211181641, "learning_rate": 3.647480893553883e-05, "loss": 1.814, "step": 3808 }, { "epoch": 0.03425775611247769, "grad_norm": 6.947946071624756, "learning_rate": 3.646746326283628e-05, "loss": 1.8232, "step": 3809 }, { "epoch": 0.034266750010118134, "grad_norm": 4.8707098960876465, "learning_rate": 3.6460116336069176e-05, "loss": 2.35, "step": 3810 }, { "epoch": 0.03427574390775859, "grad_norm": 3.5949907302856445, "learning_rate": 3.645276815604099e-05, "loss": 1.7577, "step": 3811 }, { "epoch": 0.034284737805399035, "grad_norm": 11.954021453857422, "learning_rate": 3.64454187235553e-05, "loss": 2.1384, "step": 3812 }, { "epoch": 0.03429373170303949, "grad_norm": 4.145819664001465, "learning_rate": 3.6438068039415816e-05, "loss": 1.9824, "step": 3813 }, { "epoch": 0.034302725600679936, "grad_norm": 6.173203945159912, "learning_rate": 3.6430716104426423e-05, "loss": 2.4618, "step": 3814 }, { "epoch": 0.03431171949832039, "grad_norm": 3.5624828338623047, "learning_rate": 3.642336291939109e-05, "loss": 1.7256, "step": 3815 }, { "epoch": 0.03432071339596084, "grad_norm": 3.6224262714385986, "learning_rate": 3.641600848511397e-05, "loss": 2.2138, "step": 3816 }, { "epoch": 0.03432970729360129, "grad_norm": 2.1945388317108154, "learning_rate": 3.640865280239932e-05, "loss": 2.0034, "step": 3817 }, { "epoch": 0.034338701191241744, "grad_norm": 2.216357946395874, "learning_rate": 3.640129587205156e-05, "loss": 2.4626, "step": 3818 }, { "epoch": 0.03434769508888219, "grad_norm": 2.1459851264953613, "learning_rate": 3.639393769487521e-05, "loss": 1.7439, "step": 3819 }, { "epoch": 0.034356688986522645, "grad_norm": 2.583003520965576, "learning_rate": 3.6386578271674984e-05, "loss": 2.1778, "step": 3820 }, { "epoch": 0.0343656828841631, "grad_norm": 2.4784131050109863, "learning_rate": 3.6379217603255664e-05, "loss": 2.394, "step": 3821 }, { "epoch": 0.034374676781803545, "grad_norm": 4.9889326095581055, "learning_rate": 3.637185569042222e-05, "loss": 2.5035, "step": 3822 }, { "epoch": 0.034383670679444, "grad_norm": 3.640698194503784, "learning_rate": 3.636449253397974e-05, "loss": 2.0959, "step": 3823 }, { "epoch": 0.034392664577084446, "grad_norm": 2.0460736751556396, "learning_rate": 3.635712813473343e-05, "loss": 2.4779, "step": 3824 }, { "epoch": 0.0344016584747249, "grad_norm": 2.598148822784424, "learning_rate": 3.634976249348867e-05, "loss": 2.0807, "step": 3825 }, { "epoch": 0.03441065237236535, "grad_norm": 2.5802152156829834, "learning_rate": 3.634239561105095e-05, "loss": 2.0125, "step": 3826 }, { "epoch": 0.0344196462700058, "grad_norm": 1.317373275756836, "learning_rate": 3.633502748822589e-05, "loss": 2.81, "step": 3827 }, { "epoch": 0.034428640167646254, "grad_norm": 1.807120680809021, "learning_rate": 3.632765812581928e-05, "loss": 1.9425, "step": 3828 }, { "epoch": 0.0344376340652867, "grad_norm": 4.807393550872803, "learning_rate": 3.6320287524637e-05, "loss": 2.2002, "step": 3829 }, { "epoch": 0.034446627962927155, "grad_norm": 2.480717182159424, "learning_rate": 3.631291568548509e-05, "loss": 1.9265, "step": 3830 }, { "epoch": 0.0344556218605676, "grad_norm": 4.569138526916504, "learning_rate": 3.630554260916974e-05, "loss": 1.9154, "step": 3831 }, { "epoch": 0.034464615758208056, "grad_norm": 3.439639091491699, "learning_rate": 3.6298168296497246e-05, "loss": 2.0224, "step": 3832 }, { "epoch": 0.03447360965584851, "grad_norm": 3.0423028469085693, "learning_rate": 3.629079274827405e-05, "loss": 1.8521, "step": 3833 }, { "epoch": 0.03448260355348896, "grad_norm": 2.835123062133789, "learning_rate": 3.628341596530674e-05, "loss": 2.4816, "step": 3834 }, { "epoch": 0.03449159745112941, "grad_norm": 2.495570421218872, "learning_rate": 3.627603794840201e-05, "loss": 2.1884, "step": 3835 }, { "epoch": 0.03450059134876986, "grad_norm": 1.7896461486816406, "learning_rate": 3.6268658698366745e-05, "loss": 2.2459, "step": 3836 }, { "epoch": 0.03450958524641031, "grad_norm": 2.746428966522217, "learning_rate": 3.6261278216007894e-05, "loss": 2.0319, "step": 3837 }, { "epoch": 0.034518579144050765, "grad_norm": 1.948797583580017, "learning_rate": 3.625389650213259e-05, "loss": 1.7111, "step": 3838 }, { "epoch": 0.03452757304169121, "grad_norm": 3.0418860912323, "learning_rate": 3.624651355754808e-05, "loss": 2.0592, "step": 3839 }, { "epoch": 0.034536566939331666, "grad_norm": 2.7092819213867188, "learning_rate": 3.623912938306176e-05, "loss": 1.8601, "step": 3840 }, { "epoch": 0.03454556083697211, "grad_norm": 2.441526174545288, "learning_rate": 3.6231743979481157e-05, "loss": 2.3795, "step": 3841 }, { "epoch": 0.034554554734612566, "grad_norm": 3.025935649871826, "learning_rate": 3.6224357347613906e-05, "loss": 1.9835, "step": 3842 }, { "epoch": 0.03456354863225301, "grad_norm": 15.579273223876953, "learning_rate": 3.6216969488267823e-05, "loss": 2.3127, "step": 3843 }, { "epoch": 0.03457254252989347, "grad_norm": 3.2801480293273926, "learning_rate": 3.6209580402250815e-05, "loss": 1.9496, "step": 3844 }, { "epoch": 0.03458153642753392, "grad_norm": 2.852294683456421, "learning_rate": 3.6202190090370944e-05, "loss": 1.866, "step": 3845 }, { "epoch": 0.03459053032517437, "grad_norm": 2.8214070796966553, "learning_rate": 3.619479855343641e-05, "loss": 2.2562, "step": 3846 }, { "epoch": 0.03459952422281482, "grad_norm": 2.9362287521362305, "learning_rate": 3.6187405792255535e-05, "loss": 1.9805, "step": 3847 }, { "epoch": 0.03460851812045527, "grad_norm": 6.36142635345459, "learning_rate": 3.618001180763678e-05, "loss": 2.1122, "step": 3848 }, { "epoch": 0.03461751201809572, "grad_norm": 3.515355348587036, "learning_rate": 3.617261660038875e-05, "loss": 2.4752, "step": 3849 }, { "epoch": 0.034626505915736176, "grad_norm": 2.175198554992676, "learning_rate": 3.616522017132017e-05, "loss": 2.0586, "step": 3850 }, { "epoch": 0.03463549981337662, "grad_norm": 2.2735595703125, "learning_rate": 3.6157822521239886e-05, "loss": 2.2008, "step": 3851 }, { "epoch": 0.03464449371101708, "grad_norm": 2.5579919815063477, "learning_rate": 3.615042365095692e-05, "loss": 2.1818, "step": 3852 }, { "epoch": 0.034653487608657524, "grad_norm": 2.9238128662109375, "learning_rate": 3.614302356128038e-05, "loss": 2.5239, "step": 3853 }, { "epoch": 0.03466248150629798, "grad_norm": 3.1356639862060547, "learning_rate": 3.613562225301953e-05, "loss": 2.2869, "step": 3854 }, { "epoch": 0.034671475403938425, "grad_norm": 4.156601905822754, "learning_rate": 3.6128219726983784e-05, "loss": 2.1665, "step": 3855 }, { "epoch": 0.03468046930157888, "grad_norm": 1.8716399669647217, "learning_rate": 3.612081598398266e-05, "loss": 2.1057, "step": 3856 }, { "epoch": 0.03468946319921933, "grad_norm": 2.5439705848693848, "learning_rate": 3.611341102482581e-05, "loss": 2.0435, "step": 3857 }, { "epoch": 0.03469845709685978, "grad_norm": 1.9671931266784668, "learning_rate": 3.610600485032305e-05, "loss": 2.4499, "step": 3858 }, { "epoch": 0.03470745099450023, "grad_norm": 2.3738818168640137, "learning_rate": 3.609859746128429e-05, "loss": 2.0036, "step": 3859 }, { "epoch": 0.03471644489214068, "grad_norm": 1.9289137125015259, "learning_rate": 3.6091188858519607e-05, "loss": 2.2938, "step": 3860 }, { "epoch": 0.034725438789781134, "grad_norm": 3.410656452178955, "learning_rate": 3.608377904283917e-05, "loss": 2.2735, "step": 3861 }, { "epoch": 0.03473443268742159, "grad_norm": 3.7333545684814453, "learning_rate": 3.607636801505334e-05, "loss": 1.9816, "step": 3862 }, { "epoch": 0.034743426585062034, "grad_norm": 2.7335762977600098, "learning_rate": 3.606895577597255e-05, "loss": 2.1938, "step": 3863 }, { "epoch": 0.03475242048270249, "grad_norm": 2.4549591541290283, "learning_rate": 3.606154232640738e-05, "loss": 2.1209, "step": 3864 }, { "epoch": 0.034761414380342935, "grad_norm": 2.6684980392456055, "learning_rate": 3.6054127667168596e-05, "loss": 2.2777, "step": 3865 }, { "epoch": 0.03477040827798339, "grad_norm": 2.7620296478271484, "learning_rate": 3.604671179906702e-05, "loss": 1.8238, "step": 3866 }, { "epoch": 0.034779402175623836, "grad_norm": 2.209397554397583, "learning_rate": 3.6039294722913655e-05, "loss": 2.6535, "step": 3867 }, { "epoch": 0.03478839607326429, "grad_norm": 5.613131999969482, "learning_rate": 3.603187643951961e-05, "loss": 2.0861, "step": 3868 }, { "epoch": 0.034797389970904744, "grad_norm": 1.999150276184082, "learning_rate": 3.602445694969615e-05, "loss": 2.2884, "step": 3869 }, { "epoch": 0.03480638386854519, "grad_norm": 2.4583771228790283, "learning_rate": 3.601703625425464e-05, "loss": 2.0329, "step": 3870 }, { "epoch": 0.034815377766185644, "grad_norm": 3.088191032409668, "learning_rate": 3.6009614354006626e-05, "loss": 2.0387, "step": 3871 }, { "epoch": 0.03482437166382609, "grad_norm": 2.5702569484710693, "learning_rate": 3.6002191249763725e-05, "loss": 1.9234, "step": 3872 }, { "epoch": 0.034833365561466545, "grad_norm": 5.414108753204346, "learning_rate": 3.5994766942337744e-05, "loss": 2.0518, "step": 3873 }, { "epoch": 0.034842359459107, "grad_norm": 2.028872489929199, "learning_rate": 3.5987341432540565e-05, "loss": 2.4056, "step": 3874 }, { "epoch": 0.034851353356747446, "grad_norm": 2.091247081756592, "learning_rate": 3.597991472118426e-05, "loss": 2.4546, "step": 3875 }, { "epoch": 0.0348603472543879, "grad_norm": 2.84273624420166, "learning_rate": 3.597248680908098e-05, "loss": 2.3472, "step": 3876 }, { "epoch": 0.034869341152028346, "grad_norm": 2.4394264221191406, "learning_rate": 3.5965057697043044e-05, "loss": 2.4035, "step": 3877 }, { "epoch": 0.0348783350496688, "grad_norm": 3.4286117553710938, "learning_rate": 3.5957627385882884e-05, "loss": 2.0645, "step": 3878 }, { "epoch": 0.034887328947309254, "grad_norm": 2.945655584335327, "learning_rate": 3.595019587641307e-05, "loss": 1.8225, "step": 3879 }, { "epoch": 0.0348963228449497, "grad_norm": 2.2977354526519775, "learning_rate": 3.5942763169446295e-05, "loss": 2.5128, "step": 3880 }, { "epoch": 0.034905316742590155, "grad_norm": 1.9970043897628784, "learning_rate": 3.5935329265795396e-05, "loss": 2.5297, "step": 3881 }, { "epoch": 0.0349143106402306, "grad_norm": 1.799513816833496, "learning_rate": 3.592789416627332e-05, "loss": 2.1588, "step": 3882 }, { "epoch": 0.034923304537871055, "grad_norm": 2.815352439880371, "learning_rate": 3.592045787169318e-05, "loss": 2.1022, "step": 3883 }, { "epoch": 0.0349322984355115, "grad_norm": 1.6378451585769653, "learning_rate": 3.591302038286818e-05, "loss": 2.0562, "step": 3884 }, { "epoch": 0.034941292333151956, "grad_norm": 1.8768956661224365, "learning_rate": 3.590558170061168e-05, "loss": 2.2652, "step": 3885 }, { "epoch": 0.03495028623079241, "grad_norm": 2.6672494411468506, "learning_rate": 3.589814182573715e-05, "loss": 2.0443, "step": 3886 }, { "epoch": 0.03495928012843286, "grad_norm": 4.118063926696777, "learning_rate": 3.5890700759058234e-05, "loss": 2.4548, "step": 3887 }, { "epoch": 0.03496827402607331, "grad_norm": 2.5960967540740967, "learning_rate": 3.5883258501388645e-05, "loss": 2.6452, "step": 3888 }, { "epoch": 0.03497726792371376, "grad_norm": 2.805819511413574, "learning_rate": 3.5875815053542275e-05, "loss": 1.9388, "step": 3889 }, { "epoch": 0.03498626182135421, "grad_norm": 4.123222351074219, "learning_rate": 3.586837041633312e-05, "loss": 2.3635, "step": 3890 }, { "epoch": 0.034995255718994665, "grad_norm": 6.508265972137451, "learning_rate": 3.586092459057531e-05, "loss": 1.897, "step": 3891 }, { "epoch": 0.03500424961663511, "grad_norm": 1.632083535194397, "learning_rate": 3.585347757708313e-05, "loss": 2.4871, "step": 3892 }, { "epoch": 0.035013243514275566, "grad_norm": 2.9123482704162598, "learning_rate": 3.5846029376670945e-05, "loss": 2.4213, "step": 3893 }, { "epoch": 0.03502223741191601, "grad_norm": 12.82075023651123, "learning_rate": 3.58385799901533e-05, "loss": 2.6583, "step": 3894 }, { "epoch": 0.03503123130955647, "grad_norm": 4.533183574676514, "learning_rate": 3.5831129418344845e-05, "loss": 2.3615, "step": 3895 }, { "epoch": 0.035040225207196914, "grad_norm": 3.2858657836914062, "learning_rate": 3.582367766206036e-05, "loss": 2.3074, "step": 3896 }, { "epoch": 0.03504921910483737, "grad_norm": 2.838324785232544, "learning_rate": 3.5816224722114754e-05, "loss": 2.2369, "step": 3897 }, { "epoch": 0.03505821300247782, "grad_norm": 2.004145860671997, "learning_rate": 3.580877059932308e-05, "loss": 2.3126, "step": 3898 }, { "epoch": 0.03506720690011827, "grad_norm": 4.2135748863220215, "learning_rate": 3.58013152945005e-05, "loss": 1.6885, "step": 3899 }, { "epoch": 0.03507620079775872, "grad_norm": 7.388514995574951, "learning_rate": 3.579385880846232e-05, "loss": 1.9872, "step": 3900 }, { "epoch": 0.03508519469539917, "grad_norm": 5.729872703552246, "learning_rate": 3.5786401142023975e-05, "loss": 2.083, "step": 3901 }, { "epoch": 0.03509418859303962, "grad_norm": 4.258000373840332, "learning_rate": 3.5778942296001015e-05, "loss": 2.0735, "step": 3902 }, { "epoch": 0.03510318249068008, "grad_norm": 2.780046224594116, "learning_rate": 3.5771482271209134e-05, "loss": 2.0496, "step": 3903 }, { "epoch": 0.03511217638832052, "grad_norm": 4.667655944824219, "learning_rate": 3.5764021068464144e-05, "loss": 2.1002, "step": 3904 }, { "epoch": 0.03512117028596098, "grad_norm": 2.7346973419189453, "learning_rate": 3.5756558688581985e-05, "loss": 2.1668, "step": 3905 }, { "epoch": 0.035130164183601424, "grad_norm": 1.6281448602676392, "learning_rate": 3.5749095132378754e-05, "loss": 2.0978, "step": 3906 }, { "epoch": 0.03513915808124188, "grad_norm": 6.128232955932617, "learning_rate": 3.574163040067063e-05, "loss": 2.4108, "step": 3907 }, { "epoch": 0.035148151978882325, "grad_norm": 2.0067853927612305, "learning_rate": 3.573416449427396e-05, "loss": 1.8789, "step": 3908 }, { "epoch": 0.03515714587652278, "grad_norm": 3.0346102714538574, "learning_rate": 3.57266974140052e-05, "loss": 1.7776, "step": 3909 }, { "epoch": 0.03516613977416323, "grad_norm": 3.956247568130493, "learning_rate": 3.571922916068094e-05, "loss": 2.04, "step": 3910 }, { "epoch": 0.03517513367180368, "grad_norm": 3.5539064407348633, "learning_rate": 3.5711759735117886e-05, "loss": 2.3994, "step": 3911 }, { "epoch": 0.03518412756944413, "grad_norm": 5.894194602966309, "learning_rate": 3.57042891381329e-05, "loss": 2.3476, "step": 3912 }, { "epoch": 0.03519312146708458, "grad_norm": 2.972999095916748, "learning_rate": 3.5696817370542945e-05, "loss": 2.3975, "step": 3913 }, { "epoch": 0.035202115364725034, "grad_norm": 2.0960309505462646, "learning_rate": 3.5689344433165125e-05, "loss": 2.2553, "step": 3914 }, { "epoch": 0.03521110926236549, "grad_norm": 3.8612682819366455, "learning_rate": 3.568187032681667e-05, "loss": 2.2192, "step": 3915 }, { "epoch": 0.035220103160005935, "grad_norm": 4.145002841949463, "learning_rate": 3.567439505231494e-05, "loss": 2.1489, "step": 3916 }, { "epoch": 0.03522909705764639, "grad_norm": 5.623266220092773, "learning_rate": 3.566691861047741e-05, "loss": 2.614, "step": 3917 }, { "epoch": 0.035238090955286835, "grad_norm": 1.9781315326690674, "learning_rate": 3.5659441002121694e-05, "loss": 2.2999, "step": 3918 }, { "epoch": 0.03524708485292729, "grad_norm": 2.345637559890747, "learning_rate": 3.5651962228065545e-05, "loss": 2.406, "step": 3919 }, { "epoch": 0.03525607875056774, "grad_norm": 4.12528133392334, "learning_rate": 3.564448228912682e-05, "loss": 2.5743, "step": 3920 }, { "epoch": 0.03526507264820819, "grad_norm": 3.700047016143799, "learning_rate": 3.563700118612351e-05, "loss": 2.1903, "step": 3921 }, { "epoch": 0.035274066545848644, "grad_norm": 2.174163341522217, "learning_rate": 3.562951891987375e-05, "loss": 2.1809, "step": 3922 }, { "epoch": 0.03528306044348909, "grad_norm": 2.8256731033325195, "learning_rate": 3.5622035491195776e-05, "loss": 2.1517, "step": 3923 }, { "epoch": 0.035292054341129545, "grad_norm": 1.909266471862793, "learning_rate": 3.5614550900907976e-05, "loss": 2.2867, "step": 3924 }, { "epoch": 0.03530104823876999, "grad_norm": 2.9820876121520996, "learning_rate": 3.5607065149828843e-05, "loss": 2.3154, "step": 3925 }, { "epoch": 0.035310042136410445, "grad_norm": 4.151280403137207, "learning_rate": 3.559957823877702e-05, "loss": 1.6979, "step": 3926 }, { "epoch": 0.0353190360340509, "grad_norm": 4.076888561248779, "learning_rate": 3.559209016857125e-05, "loss": 1.9887, "step": 3927 }, { "epoch": 0.035328029931691346, "grad_norm": 2.452928066253662, "learning_rate": 3.558460094003043e-05, "loss": 2.3969, "step": 3928 }, { "epoch": 0.0353370238293318, "grad_norm": 2.1589338779449463, "learning_rate": 3.557711055397356e-05, "loss": 1.7924, "step": 3929 }, { "epoch": 0.03534601772697225, "grad_norm": 2.6424570083618164, "learning_rate": 3.556961901121978e-05, "loss": 2.092, "step": 3930 }, { "epoch": 0.0353550116246127, "grad_norm": 6.212331295013428, "learning_rate": 3.556212631258836e-05, "loss": 1.9947, "step": 3931 }, { "epoch": 0.035364005522253154, "grad_norm": 4.766285419464111, "learning_rate": 3.555463245889869e-05, "loss": 2.4437, "step": 3932 }, { "epoch": 0.0353729994198936, "grad_norm": 5.48333740234375, "learning_rate": 3.554713745097028e-05, "loss": 2.166, "step": 3933 }, { "epoch": 0.035381993317534055, "grad_norm": 3.5461456775665283, "learning_rate": 3.553964128962277e-05, "loss": 2.0092, "step": 3934 }, { "epoch": 0.0353909872151745, "grad_norm": 2.2650158405303955, "learning_rate": 3.553214397567595e-05, "loss": 1.939, "step": 3935 }, { "epoch": 0.035399981112814956, "grad_norm": 5.738270282745361, "learning_rate": 3.5524645509949686e-05, "loss": 2.4924, "step": 3936 }, { "epoch": 0.0354089750104554, "grad_norm": 2.237454891204834, "learning_rate": 3.5517145893264015e-05, "loss": 2.1293, "step": 3937 }, { "epoch": 0.035417968908095857, "grad_norm": 3.354658842086792, "learning_rate": 3.550964512643908e-05, "loss": 2.4302, "step": 3938 }, { "epoch": 0.03542696280573631, "grad_norm": 5.9423675537109375, "learning_rate": 3.5502143210295165e-05, "loss": 1.6485, "step": 3939 }, { "epoch": 0.03543595670337676, "grad_norm": 2.838258743286133, "learning_rate": 3.549464014565265e-05, "loss": 2.6789, "step": 3940 }, { "epoch": 0.03544495060101721, "grad_norm": 5.548557758331299, "learning_rate": 3.548713593333206e-05, "loss": 2.381, "step": 3941 }, { "epoch": 0.03545394449865766, "grad_norm": 8.06951904296875, "learning_rate": 3.547963057415406e-05, "loss": 1.8831, "step": 3942 }, { "epoch": 0.03546293839629811, "grad_norm": 2.311147689819336, "learning_rate": 3.5472124068939416e-05, "loss": 2.0288, "step": 3943 }, { "epoch": 0.035471932293938566, "grad_norm": 2.4802815914154053, "learning_rate": 3.546461641850903e-05, "loss": 2.0306, "step": 3944 }, { "epoch": 0.03548092619157901, "grad_norm": 2.8672399520874023, "learning_rate": 3.545710762368392e-05, "loss": 2.4543, "step": 3945 }, { "epoch": 0.035489920089219466, "grad_norm": 6.853391647338867, "learning_rate": 3.544959768528525e-05, "loss": 1.9886, "step": 3946 }, { "epoch": 0.03549891398685991, "grad_norm": 2.3112761974334717, "learning_rate": 3.544208660413428e-05, "loss": 1.9127, "step": 3947 }, { "epoch": 0.03550790788450037, "grad_norm": 2.6823623180389404, "learning_rate": 3.5434574381052425e-05, "loss": 2.3192, "step": 3948 }, { "epoch": 0.03551690178214082, "grad_norm": 2.6953177452087402, "learning_rate": 3.54270610168612e-05, "loss": 2.2451, "step": 3949 }, { "epoch": 0.03552589567978127, "grad_norm": 2.875086545944214, "learning_rate": 3.5419546512382266e-05, "loss": 1.772, "step": 3950 }, { "epoch": 0.03553488957742172, "grad_norm": 8.146207809448242, "learning_rate": 3.5412030868437376e-05, "loss": 1.9335, "step": 3951 }, { "epoch": 0.03554388347506217, "grad_norm": 1.456997275352478, "learning_rate": 3.5404514085848464e-05, "loss": 1.8374, "step": 3952 }, { "epoch": 0.03555287737270262, "grad_norm": 3.355166435241699, "learning_rate": 3.539699616543752e-05, "loss": 2.3959, "step": 3953 }, { "epoch": 0.03556187127034307, "grad_norm": 3.4386825561523438, "learning_rate": 3.5389477108026714e-05, "loss": 2.0596, "step": 3954 }, { "epoch": 0.03557086516798352, "grad_norm": 1.8050174713134766, "learning_rate": 3.5381956914438305e-05, "loss": 2.428, "step": 3955 }, { "epoch": 0.03557985906562398, "grad_norm": 1.1655333042144775, "learning_rate": 3.53744355854947e-05, "loss": 1.8389, "step": 3956 }, { "epoch": 0.035588852963264424, "grad_norm": 4.0521697998046875, "learning_rate": 3.536691312201842e-05, "loss": 2.4417, "step": 3957 }, { "epoch": 0.03559784686090488, "grad_norm": 3.9751532077789307, "learning_rate": 3.535938952483211e-05, "loss": 2.3228, "step": 3958 }, { "epoch": 0.035606840758545324, "grad_norm": 2.3519551753997803, "learning_rate": 3.535186479475853e-05, "loss": 2.1981, "step": 3959 }, { "epoch": 0.03561583465618578, "grad_norm": 5.27122163772583, "learning_rate": 3.534433893262058e-05, "loss": 1.9446, "step": 3960 }, { "epoch": 0.03562482855382623, "grad_norm": 3.644855260848999, "learning_rate": 3.533681193924128e-05, "loss": 1.5542, "step": 3961 }, { "epoch": 0.03563382245146668, "grad_norm": 2.870882749557495, "learning_rate": 3.532928381544377e-05, "loss": 2.0523, "step": 3962 }, { "epoch": 0.03564281634910713, "grad_norm": 2.728959083557129, "learning_rate": 3.53217545620513e-05, "loss": 2.2927, "step": 3963 }, { "epoch": 0.03565181024674758, "grad_norm": 2.254356861114502, "learning_rate": 3.531422417988728e-05, "loss": 2.2609, "step": 3964 }, { "epoch": 0.035660804144388034, "grad_norm": 1.874548077583313, "learning_rate": 3.530669266977521e-05, "loss": 1.7804, "step": 3965 }, { "epoch": 0.03566979804202848, "grad_norm": 3.2284293174743652, "learning_rate": 3.529916003253872e-05, "loss": 2.0988, "step": 3966 }, { "epoch": 0.035678791939668934, "grad_norm": 3.445888042449951, "learning_rate": 3.5291626269001574e-05, "loss": 2.1322, "step": 3967 }, { "epoch": 0.03568778583730939, "grad_norm": 3.854862689971924, "learning_rate": 3.528409137998766e-05, "loss": 2.3201, "step": 3968 }, { "epoch": 0.035696779734949835, "grad_norm": 3.4192392826080322, "learning_rate": 3.527655536632097e-05, "loss": 2.4309, "step": 3969 }, { "epoch": 0.03570577363259029, "grad_norm": 3.068237066268921, "learning_rate": 3.526901822882564e-05, "loss": 1.9762, "step": 3970 }, { "epoch": 0.035714767530230736, "grad_norm": 4.06576681137085, "learning_rate": 3.5261479968325914e-05, "loss": 1.9462, "step": 3971 }, { "epoch": 0.03572376142787119, "grad_norm": 3.438230037689209, "learning_rate": 3.525394058564616e-05, "loss": 2.3432, "step": 3972 }, { "epoch": 0.03573275532551164, "grad_norm": 1.9968326091766357, "learning_rate": 3.524640008161089e-05, "loss": 2.1459, "step": 3973 }, { "epoch": 0.03574174922315209, "grad_norm": 2.302044153213501, "learning_rate": 3.5238858457044705e-05, "loss": 2.671, "step": 3974 }, { "epoch": 0.035750743120792544, "grad_norm": 4.212534427642822, "learning_rate": 3.523131571277235e-05, "loss": 2.4444, "step": 3975 }, { "epoch": 0.03575973701843299, "grad_norm": 2.492011785507202, "learning_rate": 3.5223771849618695e-05, "loss": 1.9962, "step": 3976 }, { "epoch": 0.035768730916073445, "grad_norm": 2.9679148197174072, "learning_rate": 3.521622686840873e-05, "loss": 2.0991, "step": 3977 }, { "epoch": 0.03577772481371389, "grad_norm": 2.660597085952759, "learning_rate": 3.520868076996755e-05, "loss": 2.307, "step": 3978 }, { "epoch": 0.035786718711354346, "grad_norm": 3.7205746173858643, "learning_rate": 3.5201133555120384e-05, "loss": 2.0296, "step": 3979 }, { "epoch": 0.0357957126089948, "grad_norm": 3.716693162918091, "learning_rate": 3.519358522469259e-05, "loss": 2.4901, "step": 3980 }, { "epoch": 0.035804706506635246, "grad_norm": 3.995272159576416, "learning_rate": 3.518603577950965e-05, "loss": 2.036, "step": 3981 }, { "epoch": 0.0358137004042757, "grad_norm": 4.684288024902344, "learning_rate": 3.517848522039714e-05, "loss": 2.0527, "step": 3982 }, { "epoch": 0.03582269430191615, "grad_norm": 2.8305373191833496, "learning_rate": 3.517093354818079e-05, "loss": 1.7455, "step": 3983 }, { "epoch": 0.0358316881995566, "grad_norm": 2.676189422607422, "learning_rate": 3.516338076368645e-05, "loss": 2.0073, "step": 3984 }, { "epoch": 0.035840682097197055, "grad_norm": 1.386322021484375, "learning_rate": 3.515582686774007e-05, "loss": 2.0838, "step": 3985 }, { "epoch": 0.0358496759948375, "grad_norm": 3.314033031463623, "learning_rate": 3.514827186116772e-05, "loss": 2.1957, "step": 3986 }, { "epoch": 0.035858669892477955, "grad_norm": 4.567304611206055, "learning_rate": 3.514071574479563e-05, "loss": 2.0637, "step": 3987 }, { "epoch": 0.0358676637901184, "grad_norm": 3.1293909549713135, "learning_rate": 3.513315851945011e-05, "loss": 2.0254, "step": 3988 }, { "epoch": 0.035876657687758856, "grad_norm": 2.329288959503174, "learning_rate": 3.51256001859576e-05, "loss": 2.0617, "step": 3989 }, { "epoch": 0.03588565158539931, "grad_norm": 3.91239070892334, "learning_rate": 3.511804074514468e-05, "loss": 2.3205, "step": 3990 }, { "epoch": 0.03589464548303976, "grad_norm": 3.103496551513672, "learning_rate": 3.5110480197838044e-05, "loss": 1.9809, "step": 3991 }, { "epoch": 0.03590363938068021, "grad_norm": 5.7185869216918945, "learning_rate": 3.510291854486448e-05, "loss": 2.408, "step": 3992 }, { "epoch": 0.03591263327832066, "grad_norm": 2.101647138595581, "learning_rate": 3.509535578705095e-05, "loss": 2.1426, "step": 3993 }, { "epoch": 0.03592162717596111, "grad_norm": 2.927654981613159, "learning_rate": 3.508779192522449e-05, "loss": 2.1469, "step": 3994 }, { "epoch": 0.03593062107360156, "grad_norm": 3.417492628097534, "learning_rate": 3.508022696021226e-05, "loss": 2.1746, "step": 3995 }, { "epoch": 0.03593961497124201, "grad_norm": 2.51522159576416, "learning_rate": 3.507266089284157e-05, "loss": 2.5061, "step": 3996 }, { "epoch": 0.035948608868882466, "grad_norm": 3.1327297687530518, "learning_rate": 3.506509372393983e-05, "loss": 2.5058, "step": 3997 }, { "epoch": 0.03595760276652291, "grad_norm": 2.802941083908081, "learning_rate": 3.5057525454334564e-05, "loss": 2.373, "step": 3998 }, { "epoch": 0.03596659666416337, "grad_norm": 2.6871864795684814, "learning_rate": 3.504995608485345e-05, "loss": 2.3063, "step": 3999 }, { "epoch": 0.035975590561803814, "grad_norm": 2.88045334815979, "learning_rate": 3.504238561632424e-05, "loss": 2.3776, "step": 4000 }, { "epoch": 0.03598458445944427, "grad_norm": 2.3918638229370117, "learning_rate": 3.5034814049574836e-05, "loss": 1.8117, "step": 4001 }, { "epoch": 0.03599357835708472, "grad_norm": 1.8376168012619019, "learning_rate": 3.5027241385433264e-05, "loss": 1.5999, "step": 4002 }, { "epoch": 0.03600257225472517, "grad_norm": 2.5054986476898193, "learning_rate": 3.501966762472764e-05, "loss": 1.9793, "step": 4003 }, { "epoch": 0.03601156615236562, "grad_norm": 6.793008804321289, "learning_rate": 3.501209276828623e-05, "loss": 2.2423, "step": 4004 }, { "epoch": 0.03602056005000607, "grad_norm": 3.159755229949951, "learning_rate": 3.500451681693741e-05, "loss": 2.3615, "step": 4005 }, { "epoch": 0.03602955394764652, "grad_norm": 3.5361804962158203, "learning_rate": 3.499693977150967e-05, "loss": 1.8772, "step": 4006 }, { "epoch": 0.03603854784528697, "grad_norm": 2.5805976390838623, "learning_rate": 3.498936163283163e-05, "loss": 1.751, "step": 4007 }, { "epoch": 0.03604754174292742, "grad_norm": 1.9033652544021606, "learning_rate": 3.498178240173202e-05, "loss": 2.5099, "step": 4008 }, { "epoch": 0.03605653564056788, "grad_norm": 4.479966640472412, "learning_rate": 3.49742020790397e-05, "loss": 2.183, "step": 4009 }, { "epoch": 0.036065529538208324, "grad_norm": 3.138979196548462, "learning_rate": 3.496662066558363e-05, "loss": 2.0803, "step": 4010 }, { "epoch": 0.03607452343584878, "grad_norm": 2.87601375579834, "learning_rate": 3.495903816219291e-05, "loss": 1.8994, "step": 4011 }, { "epoch": 0.036083517333489225, "grad_norm": 3.0524110794067383, "learning_rate": 3.495145456969675e-05, "loss": 2.5453, "step": 4012 }, { "epoch": 0.03609251123112968, "grad_norm": 3.7922794818878174, "learning_rate": 3.494386988892448e-05, "loss": 2.3165, "step": 4013 }, { "epoch": 0.03610150512877013, "grad_norm": 3.5387918949127197, "learning_rate": 3.4936284120705545e-05, "loss": 2.3165, "step": 4014 }, { "epoch": 0.03611049902641058, "grad_norm": 2.7174839973449707, "learning_rate": 3.4928697265869515e-05, "loss": 1.8181, "step": 4015 }, { "epoch": 0.03611949292405103, "grad_norm": 2.5107855796813965, "learning_rate": 3.492110932524609e-05, "loss": 2.0667, "step": 4016 }, { "epoch": 0.03612848682169148, "grad_norm": 3.708186149597168, "learning_rate": 3.4913520299665056e-05, "loss": 1.5544, "step": 4017 }, { "epoch": 0.036137480719331934, "grad_norm": 2.326301336288452, "learning_rate": 3.490593018995636e-05, "loss": 1.92, "step": 4018 }, { "epoch": 0.03614647461697238, "grad_norm": 2.4194884300231934, "learning_rate": 3.489833899695002e-05, "loss": 2.3345, "step": 4019 }, { "epoch": 0.036155468514612835, "grad_norm": 2.977360725402832, "learning_rate": 3.489074672147621e-05, "loss": 2.2726, "step": 4020 }, { "epoch": 0.03616446241225329, "grad_norm": 6.784822463989258, "learning_rate": 3.488315336436521e-05, "loss": 1.7762, "step": 4021 }, { "epoch": 0.036173456309893735, "grad_norm": 2.6908490657806396, "learning_rate": 3.487555892644742e-05, "loss": 2.5252, "step": 4022 }, { "epoch": 0.03618245020753419, "grad_norm": 3.4880001544952393, "learning_rate": 3.486796340855335e-05, "loss": 2.4349, "step": 4023 }, { "epoch": 0.036191444105174636, "grad_norm": 2.371725559234619, "learning_rate": 3.486036681151365e-05, "loss": 1.6387, "step": 4024 }, { "epoch": 0.03620043800281509, "grad_norm": 2.801283597946167, "learning_rate": 3.485276913615905e-05, "loss": 2.3395, "step": 4025 }, { "epoch": 0.036209431900455544, "grad_norm": 5.790397644042969, "learning_rate": 3.4845170383320424e-05, "loss": 2.5188, "step": 4026 }, { "epoch": 0.03621842579809599, "grad_norm": 3.232621431350708, "learning_rate": 3.483757055382878e-05, "loss": 2.2906, "step": 4027 }, { "epoch": 0.036227419695736444, "grad_norm": 2.6695809364318848, "learning_rate": 3.4829969648515196e-05, "loss": 2.2123, "step": 4028 }, { "epoch": 0.03623641359337689, "grad_norm": 3.730755090713501, "learning_rate": 3.4822367668210926e-05, "loss": 2.3582, "step": 4029 }, { "epoch": 0.036245407491017345, "grad_norm": 5.371387004852295, "learning_rate": 3.4814764613747286e-05, "loss": 1.8729, "step": 4030 }, { "epoch": 0.0362544013886578, "grad_norm": 3.8718955516815186, "learning_rate": 3.480716048595575e-05, "loss": 2.3089, "step": 4031 }, { "epoch": 0.036263395286298246, "grad_norm": 3.8920185565948486, "learning_rate": 3.479955528566789e-05, "loss": 2.3529, "step": 4032 }, { "epoch": 0.0362723891839387, "grad_norm": 2.209261655807495, "learning_rate": 3.479194901371538e-05, "loss": 1.9301, "step": 4033 }, { "epoch": 0.03628138308157915, "grad_norm": 5.1104207038879395, "learning_rate": 3.4784341670930065e-05, "loss": 2.184, "step": 4034 }, { "epoch": 0.0362903769792196, "grad_norm": 7.364703178405762, "learning_rate": 3.4776733258143854e-05, "loss": 2.4181, "step": 4035 }, { "epoch": 0.03629937087686005, "grad_norm": 5.466034412384033, "learning_rate": 3.476912377618878e-05, "loss": 2.242, "step": 4036 }, { "epoch": 0.0363083647745005, "grad_norm": 2.1104507446289062, "learning_rate": 3.476151322589704e-05, "loss": 2.2011, "step": 4037 }, { "epoch": 0.036317358672140955, "grad_norm": 2.912104368209839, "learning_rate": 3.475390160810087e-05, "loss": 1.8989, "step": 4038 }, { "epoch": 0.0363263525697814, "grad_norm": 3.3401644229888916, "learning_rate": 3.4746288923632695e-05, "loss": 2.5414, "step": 4039 }, { "epoch": 0.036335346467421856, "grad_norm": 7.493204593658447, "learning_rate": 3.473867517332501e-05, "loss": 2.3268, "step": 4040 }, { "epoch": 0.0363443403650623, "grad_norm": 3.7765140533447266, "learning_rate": 3.473106035801045e-05, "loss": 2.1043, "step": 4041 }, { "epoch": 0.036353334262702756, "grad_norm": 2.5455706119537354, "learning_rate": 3.4723444478521764e-05, "loss": 2.2123, "step": 4042 }, { "epoch": 0.03636232816034321, "grad_norm": 4.907833576202393, "learning_rate": 3.47158275356918e-05, "loss": 2.199, "step": 4043 }, { "epoch": 0.03637132205798366, "grad_norm": 2.0221266746520996, "learning_rate": 3.470820953035356e-05, "loss": 1.9354, "step": 4044 }, { "epoch": 0.03638031595562411, "grad_norm": 3.0723609924316406, "learning_rate": 3.470059046334011e-05, "loss": 1.8389, "step": 4045 }, { "epoch": 0.03638930985326456, "grad_norm": 5.070559978485107, "learning_rate": 3.469297033548468e-05, "loss": 2.479, "step": 4046 }, { "epoch": 0.03639830375090501, "grad_norm": 4.5344038009643555, "learning_rate": 3.468534914762058e-05, "loss": 2.3957, "step": 4047 }, { "epoch": 0.03640729764854546, "grad_norm": 2.9133660793304443, "learning_rate": 3.4677726900581265e-05, "loss": 2.5155, "step": 4048 }, { "epoch": 0.03641629154618591, "grad_norm": 4.5918660163879395, "learning_rate": 3.467010359520028e-05, "loss": 2.217, "step": 4049 }, { "epoch": 0.036425285443826366, "grad_norm": 17.822364807128906, "learning_rate": 3.4662479232311306e-05, "loss": 1.9576, "step": 4050 }, { "epoch": 0.03643427934146681, "grad_norm": 2.731257200241089, "learning_rate": 3.465485381274814e-05, "loss": 2.2947, "step": 4051 }, { "epoch": 0.03644327323910727, "grad_norm": 3.2439653873443604, "learning_rate": 3.4647227337344664e-05, "loss": 2.0447, "step": 4052 }, { "epoch": 0.036452267136747714, "grad_norm": 2.992224931716919, "learning_rate": 3.463959980693492e-05, "loss": 2.307, "step": 4053 }, { "epoch": 0.03646126103438817, "grad_norm": 2.7362473011016846, "learning_rate": 3.4631971222353024e-05, "loss": 2.1955, "step": 4054 }, { "epoch": 0.03647025493202862, "grad_norm": 3.2023439407348633, "learning_rate": 3.4624341584433244e-05, "loss": 2.6155, "step": 4055 }, { "epoch": 0.03647924882966907, "grad_norm": 1.9465819597244263, "learning_rate": 3.461671089400994e-05, "loss": 2.6722, "step": 4056 }, { "epoch": 0.03648824272730952, "grad_norm": 1.5223456621170044, "learning_rate": 3.460907915191759e-05, "loss": 2.1778, "step": 4057 }, { "epoch": 0.03649723662494997, "grad_norm": 4.468044281005859, "learning_rate": 3.460144635899079e-05, "loss": 1.9347, "step": 4058 }, { "epoch": 0.03650623052259042, "grad_norm": 12.103440284729004, "learning_rate": 3.459381251606424e-05, "loss": 1.7036, "step": 4059 }, { "epoch": 0.03651522442023088, "grad_norm": 3.9380621910095215, "learning_rate": 3.458617762397279e-05, "loss": 2.2958, "step": 4060 }, { "epoch": 0.036524218317871324, "grad_norm": 4.039039611816406, "learning_rate": 3.457854168355137e-05, "loss": 2.2007, "step": 4061 }, { "epoch": 0.03653321221551178, "grad_norm": 2.4036176204681396, "learning_rate": 3.4570904695635026e-05, "loss": 2.0311, "step": 4062 }, { "epoch": 0.036542206113152224, "grad_norm": 2.6629185676574707, "learning_rate": 3.456326666105894e-05, "loss": 2.3875, "step": 4063 }, { "epoch": 0.03655120001079268, "grad_norm": 1.6093254089355469, "learning_rate": 3.4555627580658384e-05, "loss": 2.2282, "step": 4064 }, { "epoch": 0.036560193908433125, "grad_norm": 2.7007479667663574, "learning_rate": 3.454798745526876e-05, "loss": 2.1355, "step": 4065 }, { "epoch": 0.03656918780607358, "grad_norm": 3.5526280403137207, "learning_rate": 3.454034628572559e-05, "loss": 2.3135, "step": 4066 }, { "epoch": 0.03657818170371403, "grad_norm": 3.3928444385528564, "learning_rate": 3.45327040728645e-05, "loss": 2.4403, "step": 4067 }, { "epoch": 0.03658717560135448, "grad_norm": 2.415529251098633, "learning_rate": 3.4525060817521204e-05, "loss": 2.1126, "step": 4068 }, { "epoch": 0.036596169498994933, "grad_norm": 3.197421073913574, "learning_rate": 3.4517416520531595e-05, "loss": 1.9086, "step": 4069 }, { "epoch": 0.03660516339663538, "grad_norm": 5.188098430633545, "learning_rate": 3.450977118273162e-05, "loss": 2.3842, "step": 4070 }, { "epoch": 0.036614157294275834, "grad_norm": 2.4053807258605957, "learning_rate": 3.450212480495737e-05, "loss": 2.7253, "step": 4071 }, { "epoch": 0.03662315119191629, "grad_norm": 3.329383611679077, "learning_rate": 3.4494477388045035e-05, "loss": 2.049, "step": 4072 }, { "epoch": 0.036632145089556735, "grad_norm": 1.6372860670089722, "learning_rate": 3.448682893283092e-05, "loss": 2.1719, "step": 4073 }, { "epoch": 0.03664113898719719, "grad_norm": 4.356311798095703, "learning_rate": 3.447917944015146e-05, "loss": 1.939, "step": 4074 }, { "epoch": 0.036650132884837636, "grad_norm": 3.308495044708252, "learning_rate": 3.447152891084319e-05, "loss": 1.829, "step": 4075 }, { "epoch": 0.03665912678247809, "grad_norm": 3.91033673286438, "learning_rate": 3.4463877345742767e-05, "loss": 1.7447, "step": 4076 }, { "epoch": 0.036668120680118536, "grad_norm": 3.1973049640655518, "learning_rate": 3.4456224745686935e-05, "loss": 2.2357, "step": 4077 }, { "epoch": 0.03667711457775899, "grad_norm": 2.103635311126709, "learning_rate": 3.444857111151259e-05, "loss": 2.4414, "step": 4078 }, { "epoch": 0.036686108475399444, "grad_norm": 2.8639426231384277, "learning_rate": 3.4440916444056705e-05, "loss": 2.3685, "step": 4079 }, { "epoch": 0.03669510237303989, "grad_norm": 2.402637004852295, "learning_rate": 3.4433260744156396e-05, "loss": 2.1833, "step": 4080 }, { "epoch": 0.036704096270680345, "grad_norm": 2.0387284755706787, "learning_rate": 3.4425604012648874e-05, "loss": 1.6671, "step": 4081 }, { "epoch": 0.03671309016832079, "grad_norm": 3.433450937271118, "learning_rate": 3.441794625037147e-05, "loss": 1.8599, "step": 4082 }, { "epoch": 0.036722084065961245, "grad_norm": 3.858539342880249, "learning_rate": 3.441028745816163e-05, "loss": 2.19, "step": 4083 }, { "epoch": 0.0367310779636017, "grad_norm": 6.199225902557373, "learning_rate": 3.440262763685689e-05, "loss": 2.4431, "step": 4084 }, { "epoch": 0.036740071861242146, "grad_norm": 5.5586771965026855, "learning_rate": 3.439496678729493e-05, "loss": 1.7, "step": 4085 }, { "epoch": 0.0367490657588826, "grad_norm": 3.5227110385894775, "learning_rate": 3.438730491031353e-05, "loss": 2.5, "step": 4086 }, { "epoch": 0.03675805965652305, "grad_norm": 4.793008327484131, "learning_rate": 3.437964200675058e-05, "loss": 2.1033, "step": 4087 }, { "epoch": 0.0367670535541635, "grad_norm": 2.2324182987213135, "learning_rate": 3.4371978077444075e-05, "loss": 2.2263, "step": 4088 }, { "epoch": 0.03677604745180395, "grad_norm": 3.0769262313842773, "learning_rate": 3.436431312323215e-05, "loss": 1.9632, "step": 4089 }, { "epoch": 0.0367850413494444, "grad_norm": 6.096663951873779, "learning_rate": 3.435664714495301e-05, "loss": 2.0028, "step": 4090 }, { "epoch": 0.036794035247084855, "grad_norm": 5.068746089935303, "learning_rate": 3.434898014344501e-05, "loss": 2.3513, "step": 4091 }, { "epoch": 0.0368030291447253, "grad_norm": 3.0394723415374756, "learning_rate": 3.43413121195466e-05, "loss": 2.1402, "step": 4092 }, { "epoch": 0.036812023042365756, "grad_norm": 1.978461503982544, "learning_rate": 3.433364307409634e-05, "loss": 2.5323, "step": 4093 }, { "epoch": 0.0368210169400062, "grad_norm": 2.1528260707855225, "learning_rate": 3.4325973007932907e-05, "loss": 2.5481, "step": 4094 }, { "epoch": 0.03683001083764666, "grad_norm": 1.783646583557129, "learning_rate": 3.4318301921895084e-05, "loss": 2.3843, "step": 4095 }, { "epoch": 0.03683900473528711, "grad_norm": 2.2186732292175293, "learning_rate": 3.431062981682177e-05, "loss": 2.4072, "step": 4096 }, { "epoch": 0.03684799863292756, "grad_norm": 3.2367451190948486, "learning_rate": 3.4302956693551986e-05, "loss": 1.9961, "step": 4097 }, { "epoch": 0.03685699253056801, "grad_norm": 2.6229310035705566, "learning_rate": 3.429528255292485e-05, "loss": 2.3218, "step": 4098 }, { "epoch": 0.03686598642820846, "grad_norm": 6.283295154571533, "learning_rate": 3.428760739577957e-05, "loss": 2.37, "step": 4099 }, { "epoch": 0.03687498032584891, "grad_norm": 3.0178120136260986, "learning_rate": 3.427993122295552e-05, "loss": 1.7527, "step": 4100 }, { "epoch": 0.036883974223489366, "grad_norm": 3.5554511547088623, "learning_rate": 3.4272254035292144e-05, "loss": 2.6033, "step": 4101 }, { "epoch": 0.03689296812112981, "grad_norm": 2.199118137359619, "learning_rate": 3.426457583362901e-05, "loss": 1.9374, "step": 4102 }, { "epoch": 0.03690196201877027, "grad_norm": 2.915051221847534, "learning_rate": 3.4256896618805784e-05, "loss": 2.4182, "step": 4103 }, { "epoch": 0.03691095591641071, "grad_norm": 6.903796195983887, "learning_rate": 3.424921639166226e-05, "loss": 2.1053, "step": 4104 }, { "epoch": 0.03691994981405117, "grad_norm": 3.50443434715271, "learning_rate": 3.424153515303835e-05, "loss": 2.6734, "step": 4105 }, { "epoch": 0.036928943711691614, "grad_norm": 2.8261818885803223, "learning_rate": 3.423385290377404e-05, "loss": 2.0945, "step": 4106 }, { "epoch": 0.03693793760933207, "grad_norm": 2.4238791465759277, "learning_rate": 3.422616964470946e-05, "loss": 1.9376, "step": 4107 }, { "epoch": 0.03694693150697252, "grad_norm": 5.513791084289551, "learning_rate": 3.4218485376684845e-05, "loss": 2.0954, "step": 4108 }, { "epoch": 0.03695592540461297, "grad_norm": 3.1067395210266113, "learning_rate": 3.421080010054052e-05, "loss": 2.0952, "step": 4109 }, { "epoch": 0.03696491930225342, "grad_norm": 2.6548640727996826, "learning_rate": 3.4203113817116957e-05, "loss": 1.9893, "step": 4110 }, { "epoch": 0.03697391319989387, "grad_norm": 4.424537658691406, "learning_rate": 3.419542652725469e-05, "loss": 1.9095, "step": 4111 }, { "epoch": 0.03698290709753432, "grad_norm": 2.8116238117218018, "learning_rate": 3.418773823179441e-05, "loss": 2.7184, "step": 4112 }, { "epoch": 0.03699190099517478, "grad_norm": 3.9188499450683594, "learning_rate": 3.4180048931576884e-05, "loss": 2.4426, "step": 4113 }, { "epoch": 0.037000894892815224, "grad_norm": 2.306633710861206, "learning_rate": 3.417235862744302e-05, "loss": 1.9349, "step": 4114 }, { "epoch": 0.03700988879045568, "grad_norm": 5.844876766204834, "learning_rate": 3.41646673202338e-05, "loss": 2.2182, "step": 4115 }, { "epoch": 0.037018882688096125, "grad_norm": 3.6619365215301514, "learning_rate": 3.4156975010790336e-05, "loss": 2.5464, "step": 4116 }, { "epoch": 0.03702787658573658, "grad_norm": 3.0945823192596436, "learning_rate": 3.414928169995386e-05, "loss": 2.1462, "step": 4117 }, { "epoch": 0.037036870483377025, "grad_norm": 2.580101728439331, "learning_rate": 3.41415873885657e-05, "loss": 2.0282, "step": 4118 }, { "epoch": 0.03704586438101748, "grad_norm": 2.7979483604431152, "learning_rate": 3.4133892077467275e-05, "loss": 2.4057, "step": 4119 }, { "epoch": 0.03705485827865793, "grad_norm": 2.9367053508758545, "learning_rate": 3.412619576750014e-05, "loss": 2.2002, "step": 4120 }, { "epoch": 0.03706385217629838, "grad_norm": 1.9846646785736084, "learning_rate": 3.411849845950598e-05, "loss": 2.6685, "step": 4121 }, { "epoch": 0.037072846073938834, "grad_norm": 3.6218984127044678, "learning_rate": 3.4110800154326515e-05, "loss": 2.0051, "step": 4122 }, { "epoch": 0.03708183997157928, "grad_norm": 3.4673314094543457, "learning_rate": 3.4103100852803656e-05, "loss": 1.8783, "step": 4123 }, { "epoch": 0.037090833869219735, "grad_norm": 3.7446954250335693, "learning_rate": 3.4095400555779366e-05, "loss": 2.2956, "step": 4124 }, { "epoch": 0.03709982776686019, "grad_norm": 3.4104154109954834, "learning_rate": 3.4087699264095745e-05, "loss": 1.6781, "step": 4125 }, { "epoch": 0.037108821664500635, "grad_norm": 5.084015846252441, "learning_rate": 3.4079996978595e-05, "loss": 2.5907, "step": 4126 }, { "epoch": 0.03711781556214109, "grad_norm": 7.101533889770508, "learning_rate": 3.407229370011944e-05, "loss": 2.2006, "step": 4127 }, { "epoch": 0.037126809459781536, "grad_norm": 5.1781158447265625, "learning_rate": 3.406458942951148e-05, "loss": 2.4913, "step": 4128 }, { "epoch": 0.03713580335742199, "grad_norm": 2.1271283626556396, "learning_rate": 3.405688416761364e-05, "loss": 2.0908, "step": 4129 }, { "epoch": 0.03714479725506244, "grad_norm": 2.08803391456604, "learning_rate": 3.4049177915268576e-05, "loss": 2.2071, "step": 4130 }, { "epoch": 0.03715379115270289, "grad_norm": 2.1206555366516113, "learning_rate": 3.404147067331901e-05, "loss": 2.1494, "step": 4131 }, { "epoch": 0.037162785050343344, "grad_norm": 1.8294570446014404, "learning_rate": 3.403376244260781e-05, "loss": 1.9567, "step": 4132 }, { "epoch": 0.03717177894798379, "grad_norm": 2.009303569793701, "learning_rate": 3.402605322397794e-05, "loss": 2.611, "step": 4133 }, { "epoch": 0.037180772845624245, "grad_norm": 2.2809057235717773, "learning_rate": 3.401834301827245e-05, "loss": 2.3139, "step": 4134 }, { "epoch": 0.03718976674326469, "grad_norm": 2.912552833557129, "learning_rate": 3.4010631826334526e-05, "loss": 2.696, "step": 4135 }, { "epoch": 0.037198760640905146, "grad_norm": 3.1116626262664795, "learning_rate": 3.400291964900745e-05, "loss": 2.5299, "step": 4136 }, { "epoch": 0.0372077545385456, "grad_norm": 2.1456894874572754, "learning_rate": 3.3995206487134624e-05, "loss": 2.2505, "step": 4137 }, { "epoch": 0.037216748436186047, "grad_norm": 2.7651071548461914, "learning_rate": 3.3987492341559535e-05, "loss": 2.1287, "step": 4138 }, { "epoch": 0.0372257423338265, "grad_norm": 1.405288577079773, "learning_rate": 3.3979777213125795e-05, "loss": 1.8159, "step": 4139 }, { "epoch": 0.03723473623146695, "grad_norm": 4.490177631378174, "learning_rate": 3.397206110267713e-05, "loss": 2.2297, "step": 4140 }, { "epoch": 0.0372437301291074, "grad_norm": 3.0940866470336914, "learning_rate": 3.396434401105734e-05, "loss": 2.1446, "step": 4141 }, { "epoch": 0.037252724026747855, "grad_norm": 3.345973491668701, "learning_rate": 3.3956625939110376e-05, "loss": 2.3839, "step": 4142 }, { "epoch": 0.0372617179243883, "grad_norm": 5.589799404144287, "learning_rate": 3.394890688768026e-05, "loss": 1.8503, "step": 4143 }, { "epoch": 0.037270711822028756, "grad_norm": 3.037046432495117, "learning_rate": 3.394118685761114e-05, "loss": 1.9068, "step": 4144 }, { "epoch": 0.0372797057196692, "grad_norm": 4.896244049072266, "learning_rate": 3.3933465849747275e-05, "loss": 2.1887, "step": 4145 }, { "epoch": 0.037288699617309656, "grad_norm": 2.54131817817688, "learning_rate": 3.3925743864933014e-05, "loss": 1.919, "step": 4146 }, { "epoch": 0.0372976935149501, "grad_norm": 2.5699729919433594, "learning_rate": 3.391802090401282e-05, "loss": 2.5023, "step": 4147 }, { "epoch": 0.03730668741259056, "grad_norm": 5.856496810913086, "learning_rate": 3.3910296967831266e-05, "loss": 2.2753, "step": 4148 }, { "epoch": 0.03731568131023101, "grad_norm": 2.475019693374634, "learning_rate": 3.390257205723304e-05, "loss": 1.7952, "step": 4149 }, { "epoch": 0.03732467520787146, "grad_norm": 4.478466510772705, "learning_rate": 3.389484617306292e-05, "loss": 2.2211, "step": 4150 }, { "epoch": 0.03733366910551191, "grad_norm": 2.600161552429199, "learning_rate": 3.388711931616578e-05, "loss": 1.7437, "step": 4151 }, { "epoch": 0.03734266300315236, "grad_norm": 2.8699240684509277, "learning_rate": 3.387939148738665e-05, "loss": 2.1566, "step": 4152 }, { "epoch": 0.03735165690079281, "grad_norm": 3.1316568851470947, "learning_rate": 3.387166268757061e-05, "loss": 2.0016, "step": 4153 }, { "epoch": 0.037360650798433266, "grad_norm": 2.6170639991760254, "learning_rate": 3.3863932917562874e-05, "loss": 2.0588, "step": 4154 }, { "epoch": 0.03736964469607371, "grad_norm": 3.225321054458618, "learning_rate": 3.385620217820877e-05, "loss": 1.7683, "step": 4155 }, { "epoch": 0.03737863859371417, "grad_norm": 2.3249475955963135, "learning_rate": 3.38484704703537e-05, "loss": 1.936, "step": 4156 }, { "epoch": 0.037387632491354614, "grad_norm": 3.1367311477661133, "learning_rate": 3.3840737794843206e-05, "loss": 2.1714, "step": 4157 }, { "epoch": 0.03739662638899507, "grad_norm": 2.801952600479126, "learning_rate": 3.383300415252292e-05, "loss": 2.2902, "step": 4158 }, { "epoch": 0.037405620286635514, "grad_norm": 3.258064031600952, "learning_rate": 3.382526954423858e-05, "loss": 2.0333, "step": 4159 }, { "epoch": 0.03741461418427597, "grad_norm": 4.3317484855651855, "learning_rate": 3.381753397083604e-05, "loss": 2.3898, "step": 4160 }, { "epoch": 0.03742360808191642, "grad_norm": 2.3300983905792236, "learning_rate": 3.380979743316122e-05, "loss": 2.3027, "step": 4161 }, { "epoch": 0.03743260197955687, "grad_norm": 2.753230094909668, "learning_rate": 3.3802059932060216e-05, "loss": 2.0436, "step": 4162 }, { "epoch": 0.03744159587719732, "grad_norm": 2.6920852661132812, "learning_rate": 3.379432146837917e-05, "loss": 2.6183, "step": 4163 }, { "epoch": 0.03745058977483777, "grad_norm": 3.7009620666503906, "learning_rate": 3.378658204296434e-05, "loss": 1.8391, "step": 4164 }, { "epoch": 0.037459583672478224, "grad_norm": 3.2558040618896484, "learning_rate": 3.377884165666212e-05, "loss": 2.3496, "step": 4165 }, { "epoch": 0.03746857757011868, "grad_norm": 6.618888854980469, "learning_rate": 3.377110031031896e-05, "loss": 2.5303, "step": 4166 }, { "epoch": 0.037477571467759124, "grad_norm": 2.374894142150879, "learning_rate": 3.3763358004781475e-05, "loss": 1.9878, "step": 4167 }, { "epoch": 0.03748656536539958, "grad_norm": 3.5762155055999756, "learning_rate": 3.375561474089633e-05, "loss": 2.4873, "step": 4168 }, { "epoch": 0.037495559263040025, "grad_norm": 3.0304980278015137, "learning_rate": 3.374787051951032e-05, "loss": 2.679, "step": 4169 }, { "epoch": 0.03750455316068048, "grad_norm": 5.017292499542236, "learning_rate": 3.3740125341470336e-05, "loss": 2.6099, "step": 4170 }, { "epoch": 0.03751354705832093, "grad_norm": 2.8334412574768066, "learning_rate": 3.3732379207623386e-05, "loss": 2.2396, "step": 4171 }, { "epoch": 0.03752254095596138, "grad_norm": 1.907361388206482, "learning_rate": 3.3724632118816584e-05, "loss": 2.0828, "step": 4172 }, { "epoch": 0.03753153485360183, "grad_norm": 4.391378879547119, "learning_rate": 3.371688407589712e-05, "loss": 1.7353, "step": 4173 }, { "epoch": 0.03754052875124228, "grad_norm": 2.8501148223876953, "learning_rate": 3.370913507971233e-05, "loss": 2.2226, "step": 4174 }, { "epoch": 0.037549522648882734, "grad_norm": 5.675631999969482, "learning_rate": 3.3701385131109616e-05, "loss": 2.3989, "step": 4175 }, { "epoch": 0.03755851654652318, "grad_norm": 2.7970597743988037, "learning_rate": 3.36936342309365e-05, "loss": 2.2323, "step": 4176 }, { "epoch": 0.037567510444163635, "grad_norm": 3.817033529281616, "learning_rate": 3.368588238004062e-05, "loss": 2.3872, "step": 4177 }, { "epoch": 0.03757650434180409, "grad_norm": 4.787834644317627, "learning_rate": 3.367812957926971e-05, "loss": 2.1156, "step": 4178 }, { "epoch": 0.037585498239444536, "grad_norm": 4.852682113647461, "learning_rate": 3.367037582947158e-05, "loss": 2.3599, "step": 4179 }, { "epoch": 0.03759449213708499, "grad_norm": 3.1526591777801514, "learning_rate": 3.3662621131494204e-05, "loss": 2.0948, "step": 4180 }, { "epoch": 0.037603486034725436, "grad_norm": 6.8858466148376465, "learning_rate": 3.3654865486185595e-05, "loss": 2.2884, "step": 4181 }, { "epoch": 0.03761247993236589, "grad_norm": 2.37668776512146, "learning_rate": 3.364710889439391e-05, "loss": 2.2146, "step": 4182 }, { "epoch": 0.037621473830006344, "grad_norm": 6.821985244750977, "learning_rate": 3.36393513569674e-05, "loss": 2.0442, "step": 4183 }, { "epoch": 0.03763046772764679, "grad_norm": 2.7474427223205566, "learning_rate": 3.36315928747544e-05, "loss": 2.2302, "step": 4184 }, { "epoch": 0.037639461625287245, "grad_norm": 2.284688949584961, "learning_rate": 3.36238334486034e-05, "loss": 1.9635, "step": 4185 }, { "epoch": 0.03764845552292769, "grad_norm": 3.0201876163482666, "learning_rate": 3.3616073079362926e-05, "loss": 2.1126, "step": 4186 }, { "epoch": 0.037657449420568145, "grad_norm": 3.2934958934783936, "learning_rate": 3.360831176788166e-05, "loss": 2.002, "step": 4187 }, { "epoch": 0.03766644331820859, "grad_norm": 1.9597389698028564, "learning_rate": 3.360054951500837e-05, "loss": 2.1618, "step": 4188 }, { "epoch": 0.037675437215849046, "grad_norm": 3.496727705001831, "learning_rate": 3.35927863215919e-05, "loss": 2.0418, "step": 4189 }, { "epoch": 0.0376844311134895, "grad_norm": 2.855203151702881, "learning_rate": 3.358502218848125e-05, "loss": 2.5248, "step": 4190 }, { "epoch": 0.03769342501112995, "grad_norm": 2.8596277236938477, "learning_rate": 3.357725711652547e-05, "loss": 2.3156, "step": 4191 }, { "epoch": 0.0377024189087704, "grad_norm": 2.8993914127349854, "learning_rate": 3.3569491106573746e-05, "loss": 2.2314, "step": 4192 }, { "epoch": 0.03771141280641085, "grad_norm": 3.0287630558013916, "learning_rate": 3.356172415947537e-05, "loss": 2.6094, "step": 4193 }, { "epoch": 0.0377204067040513, "grad_norm": 2.9539380073547363, "learning_rate": 3.3553956276079714e-05, "loss": 1.8894, "step": 4194 }, { "epoch": 0.037729400601691755, "grad_norm": 2.610567569732666, "learning_rate": 3.3546187457236244e-05, "loss": 2.3754, "step": 4195 }, { "epoch": 0.0377383944993322, "grad_norm": 7.659283638000488, "learning_rate": 3.353841770379458e-05, "loss": 2.5864, "step": 4196 }, { "epoch": 0.037747388396972656, "grad_norm": 5.0745368003845215, "learning_rate": 3.353064701660438e-05, "loss": 2.1987, "step": 4197 }, { "epoch": 0.0377563822946131, "grad_norm": 1.773271083831787, "learning_rate": 3.352287539651545e-05, "loss": 2.741, "step": 4198 }, { "epoch": 0.03776537619225356, "grad_norm": 3.5811431407928467, "learning_rate": 3.351510284437769e-05, "loss": 2.3647, "step": 4199 }, { "epoch": 0.037774370089894004, "grad_norm": 4.919888973236084, "learning_rate": 3.350732936104108e-05, "loss": 2.3209, "step": 4200 }, { "epoch": 0.03778336398753446, "grad_norm": 4.6901140213012695, "learning_rate": 3.3499554947355714e-05, "loss": 2.4404, "step": 4201 }, { "epoch": 0.03779235788517491, "grad_norm": 3.5627589225769043, "learning_rate": 3.3491779604171816e-05, "loss": 2.1924, "step": 4202 }, { "epoch": 0.03780135178281536, "grad_norm": 3.3523812294006348, "learning_rate": 3.348400333233965e-05, "loss": 2.0221, "step": 4203 }, { "epoch": 0.03781034568045581, "grad_norm": 2.914712429046631, "learning_rate": 3.347622613270965e-05, "loss": 1.9679, "step": 4204 }, { "epoch": 0.03781933957809626, "grad_norm": 3.174058198928833, "learning_rate": 3.346844800613229e-05, "loss": 2.2908, "step": 4205 }, { "epoch": 0.03782833347573671, "grad_norm": 4.328569412231445, "learning_rate": 3.3460668953458194e-05, "loss": 2.5909, "step": 4206 }, { "epoch": 0.037837327373377166, "grad_norm": 2.082003593444824, "learning_rate": 3.345288897553806e-05, "loss": 2.1679, "step": 4207 }, { "epoch": 0.03784632127101761, "grad_norm": 3.605571985244751, "learning_rate": 3.344510807322269e-05, "loss": 1.8178, "step": 4208 }, { "epoch": 0.03785531516865807, "grad_norm": 3.9482085704803467, "learning_rate": 3.343732624736301e-05, "loss": 1.9958, "step": 4209 }, { "epoch": 0.037864309066298514, "grad_norm": 2.5546133518218994, "learning_rate": 3.342954349881001e-05, "loss": 2.2334, "step": 4210 }, { "epoch": 0.03787330296393897, "grad_norm": 3.6419260501861572, "learning_rate": 3.342175982841481e-05, "loss": 2.3064, "step": 4211 }, { "epoch": 0.03788229686157942, "grad_norm": 3.2978193759918213, "learning_rate": 3.341397523702862e-05, "loss": 1.9367, "step": 4212 }, { "epoch": 0.03789129075921987, "grad_norm": 3.6504194736480713, "learning_rate": 3.340618972550274e-05, "loss": 2.3255, "step": 4213 }, { "epoch": 0.03790028465686032, "grad_norm": 4.093487739562988, "learning_rate": 3.3398403294688593e-05, "loss": 2.5206, "step": 4214 }, { "epoch": 0.03790927855450077, "grad_norm": 4.331210136413574, "learning_rate": 3.33906159454377e-05, "loss": 1.8045, "step": 4215 }, { "epoch": 0.03791827245214122, "grad_norm": 3.7410449981689453, "learning_rate": 3.338282767860165e-05, "loss": 2.4343, "step": 4216 }, { "epoch": 0.03792726634978167, "grad_norm": 3.2095682621002197, "learning_rate": 3.3375038495032165e-05, "loss": 2.1109, "step": 4217 }, { "epoch": 0.037936260247422124, "grad_norm": 3.8352246284484863, "learning_rate": 3.3367248395581076e-05, "loss": 1.9749, "step": 4218 }, { "epoch": 0.03794525414506258, "grad_norm": 3.8136720657348633, "learning_rate": 3.3359457381100276e-05, "loss": 2.3474, "step": 4219 }, { "epoch": 0.037954248042703025, "grad_norm": 3.4051501750946045, "learning_rate": 3.335166545244178e-05, "loss": 2.1579, "step": 4220 }, { "epoch": 0.03796324194034348, "grad_norm": 3.0376267433166504, "learning_rate": 3.334387261045771e-05, "loss": 2.172, "step": 4221 }, { "epoch": 0.037972235837983925, "grad_norm": 2.2201569080352783, "learning_rate": 3.3336078856000276e-05, "loss": 2.3395, "step": 4222 }, { "epoch": 0.03798122973562438, "grad_norm": 3.5531227588653564, "learning_rate": 3.3328284189921795e-05, "loss": 2.0899, "step": 4223 }, { "epoch": 0.03799022363326483, "grad_norm": 4.542884349822998, "learning_rate": 3.332048861307467e-05, "loss": 2.4072, "step": 4224 }, { "epoch": 0.03799921753090528, "grad_norm": 3.8512489795684814, "learning_rate": 3.3312692126311425e-05, "loss": 2.2371, "step": 4225 }, { "epoch": 0.038008211428545734, "grad_norm": 2.568688154220581, "learning_rate": 3.330489473048467e-05, "loss": 1.9485, "step": 4226 }, { "epoch": 0.03801720532618618, "grad_norm": 7.916805267333984, "learning_rate": 3.32970964264471e-05, "loss": 2.6335, "step": 4227 }, { "epoch": 0.038026199223826634, "grad_norm": 2.569992780685425, "learning_rate": 3.3289297215051553e-05, "loss": 1.9801, "step": 4228 }, { "epoch": 0.03803519312146708, "grad_norm": 4.453756809234619, "learning_rate": 3.328149709715091e-05, "loss": 2.6026, "step": 4229 }, { "epoch": 0.038044187019107535, "grad_norm": 2.6264114379882812, "learning_rate": 3.327369607359821e-05, "loss": 1.8503, "step": 4230 }, { "epoch": 0.03805318091674799, "grad_norm": 4.75861120223999, "learning_rate": 3.326589414524654e-05, "loss": 2.242, "step": 4231 }, { "epoch": 0.038062174814388436, "grad_norm": 4.655515193939209, "learning_rate": 3.325809131294912e-05, "loss": 1.7816, "step": 4232 }, { "epoch": 0.03807116871202889, "grad_norm": 7.702299118041992, "learning_rate": 3.325028757755924e-05, "loss": 1.8865, "step": 4233 }, { "epoch": 0.03808016260966934, "grad_norm": 4.3648176193237305, "learning_rate": 3.324248293993033e-05, "loss": 1.9308, "step": 4234 }, { "epoch": 0.03808915650730979, "grad_norm": 2.720180034637451, "learning_rate": 3.3234677400915865e-05, "loss": 2.4812, "step": 4235 }, { "epoch": 0.038098150404950244, "grad_norm": 3.6674551963806152, "learning_rate": 3.322687096136947e-05, "loss": 2.5692, "step": 4236 }, { "epoch": 0.03810714430259069, "grad_norm": 1.8455675840377808, "learning_rate": 3.321906362214482e-05, "loss": 2.573, "step": 4237 }, { "epoch": 0.038116138200231145, "grad_norm": 5.2171735763549805, "learning_rate": 3.321125538409574e-05, "loss": 2.7018, "step": 4238 }, { "epoch": 0.03812513209787159, "grad_norm": 7.567327499389648, "learning_rate": 3.3203446248076117e-05, "loss": 2.4484, "step": 4239 }, { "epoch": 0.038134125995512046, "grad_norm": 2.2063236236572266, "learning_rate": 3.319563621493994e-05, "loss": 1.9379, "step": 4240 }, { "epoch": 0.03814311989315249, "grad_norm": 3.4237945079803467, "learning_rate": 3.318782528554132e-05, "loss": 1.8623, "step": 4241 }, { "epoch": 0.038152113790792946, "grad_norm": 4.730388164520264, "learning_rate": 3.318001346073443e-05, "loss": 1.9309, "step": 4242 }, { "epoch": 0.0381611076884334, "grad_norm": 1.5881383419036865, "learning_rate": 3.3172200741373563e-05, "loss": 1.9243, "step": 4243 }, { "epoch": 0.03817010158607385, "grad_norm": 2.5153110027313232, "learning_rate": 3.316438712831311e-05, "loss": 2.0842, "step": 4244 }, { "epoch": 0.0381790954837143, "grad_norm": 3.325153112411499, "learning_rate": 3.3156572622407565e-05, "loss": 2.3287, "step": 4245 }, { "epoch": 0.03818808938135475, "grad_norm": 3.4356908798217773, "learning_rate": 3.3148757224511496e-05, "loss": 2.3441, "step": 4246 }, { "epoch": 0.0381970832789952, "grad_norm": 5.24363374710083, "learning_rate": 3.314094093547959e-05, "loss": 2.4466, "step": 4247 }, { "epoch": 0.038206077176635656, "grad_norm": 6.392857551574707, "learning_rate": 3.313312375616662e-05, "loss": 2.5866, "step": 4248 }, { "epoch": 0.0382150710742761, "grad_norm": 3.1582272052764893, "learning_rate": 3.312530568742746e-05, "loss": 2.0558, "step": 4249 }, { "epoch": 0.038224064971916556, "grad_norm": 2.680584669113159, "learning_rate": 3.311748673011709e-05, "loss": 1.9051, "step": 4250 }, { "epoch": 0.038233058869557, "grad_norm": 2.2736330032348633, "learning_rate": 3.310966688509058e-05, "loss": 1.9426, "step": 4251 }, { "epoch": 0.03824205276719746, "grad_norm": 1.9799538850784302, "learning_rate": 3.310184615320309e-05, "loss": 2.1791, "step": 4252 }, { "epoch": 0.03825104666483791, "grad_norm": 2.336535692214966, "learning_rate": 3.309402453530988e-05, "loss": 2.0281, "step": 4253 }, { "epoch": 0.03826004056247836, "grad_norm": 4.512166976928711, "learning_rate": 3.308620203226632e-05, "loss": 2.0092, "step": 4254 }, { "epoch": 0.03826903446011881, "grad_norm": 3.493345022201538, "learning_rate": 3.307837864492786e-05, "loss": 2.8683, "step": 4255 }, { "epoch": 0.03827802835775926, "grad_norm": 3.9420008659362793, "learning_rate": 3.307055437415006e-05, "loss": 2.2641, "step": 4256 }, { "epoch": 0.03828702225539971, "grad_norm": 3.952329158782959, "learning_rate": 3.306272922078857e-05, "loss": 2.416, "step": 4257 }, { "epoch": 0.03829601615304016, "grad_norm": 4.548449993133545, "learning_rate": 3.3054903185699135e-05, "loss": 1.8465, "step": 4258 }, { "epoch": 0.03830501005068061, "grad_norm": 5.318232536315918, "learning_rate": 3.3047076269737586e-05, "loss": 1.691, "step": 4259 }, { "epoch": 0.03831400394832107, "grad_norm": 1.8936216831207275, "learning_rate": 3.3039248473759885e-05, "loss": 2.2928, "step": 4260 }, { "epoch": 0.038322997845961514, "grad_norm": 6.516697883605957, "learning_rate": 3.303141979862206e-05, "loss": 2.2835, "step": 4261 }, { "epoch": 0.03833199174360197, "grad_norm": 4.754917144775391, "learning_rate": 3.302359024518024e-05, "loss": 2.2445, "step": 4262 }, { "epoch": 0.038340985641242414, "grad_norm": 6.107509613037109, "learning_rate": 3.301575981429065e-05, "loss": 2.4953, "step": 4263 }, { "epoch": 0.03834997953888287, "grad_norm": 2.1726794242858887, "learning_rate": 3.3007928506809624e-05, "loss": 2.0467, "step": 4264 }, { "epoch": 0.03835897343652332, "grad_norm": 3.4387717247009277, "learning_rate": 3.300009632359357e-05, "loss": 2.185, "step": 4265 }, { "epoch": 0.03836796733416377, "grad_norm": 3.1552484035491943, "learning_rate": 3.299226326549901e-05, "loss": 2.1557, "step": 4266 }, { "epoch": 0.03837696123180422, "grad_norm": 2.5003039836883545, "learning_rate": 3.298442933338256e-05, "loss": 2.1831, "step": 4267 }, { "epoch": 0.03838595512944467, "grad_norm": 3.402971029281616, "learning_rate": 3.297659452810092e-05, "loss": 2.5321, "step": 4268 }, { "epoch": 0.038394949027085123, "grad_norm": 1.7876968383789062, "learning_rate": 3.296875885051089e-05, "loss": 2.0886, "step": 4269 }, { "epoch": 0.03840394292472557, "grad_norm": 5.312716007232666, "learning_rate": 3.2960922301469385e-05, "loss": 2.3941, "step": 4270 }, { "epoch": 0.038412936822366024, "grad_norm": 2.880190372467041, "learning_rate": 3.295308488183338e-05, "loss": 2.1071, "step": 4271 }, { "epoch": 0.03842193072000648, "grad_norm": 2.3938419818878174, "learning_rate": 3.294524659245996e-05, "loss": 2.4152, "step": 4272 }, { "epoch": 0.038430924617646925, "grad_norm": 2.2519702911376953, "learning_rate": 3.2937407434206334e-05, "loss": 2.511, "step": 4273 }, { "epoch": 0.03843991851528738, "grad_norm": 3.2022182941436768, "learning_rate": 3.292956740792976e-05, "loss": 1.6634, "step": 4274 }, { "epoch": 0.038448912412927826, "grad_norm": 3.0851926803588867, "learning_rate": 3.2921726514487614e-05, "loss": 1.986, "step": 4275 }, { "epoch": 0.03845790631056828, "grad_norm": 3.205518960952759, "learning_rate": 3.291388475473737e-05, "loss": 2.4848, "step": 4276 }, { "epoch": 0.03846690020820873, "grad_norm": 3.708808183670044, "learning_rate": 3.290604212953658e-05, "loss": 2.246, "step": 4277 }, { "epoch": 0.03847589410584918, "grad_norm": 4.568928241729736, "learning_rate": 3.2898198639742915e-05, "loss": 1.9337, "step": 4278 }, { "epoch": 0.038484888003489634, "grad_norm": 6.029495716094971, "learning_rate": 3.2890354286214114e-05, "loss": 1.8727, "step": 4279 }, { "epoch": 0.03849388190113008, "grad_norm": 2.9346089363098145, "learning_rate": 3.2882509069808044e-05, "loss": 1.7012, "step": 4280 }, { "epoch": 0.038502875798770535, "grad_norm": 2.8045032024383545, "learning_rate": 3.287466299138262e-05, "loss": 2.0499, "step": 4281 }, { "epoch": 0.03851186969641099, "grad_norm": 2.0613479614257812, "learning_rate": 3.2866816051795894e-05, "loss": 2.0893, "step": 4282 }, { "epoch": 0.038520863594051435, "grad_norm": 3.532701015472412, "learning_rate": 3.2858968251906003e-05, "loss": 2.016, "step": 4283 }, { "epoch": 0.03852985749169189, "grad_norm": 2.2129287719726562, "learning_rate": 3.2851119592571145e-05, "loss": 2.1464, "step": 4284 }, { "epoch": 0.038538851389332336, "grad_norm": 8.078058242797852, "learning_rate": 3.284327007464966e-05, "loss": 2.2248, "step": 4285 }, { "epoch": 0.03854784528697279, "grad_norm": 2.4483001232147217, "learning_rate": 3.283541969899996e-05, "loss": 3.0032, "step": 4286 }, { "epoch": 0.03855683918461324, "grad_norm": 2.215592622756958, "learning_rate": 3.282756846648052e-05, "loss": 2.3015, "step": 4287 }, { "epoch": 0.03856583308225369, "grad_norm": 1.9844385385513306, "learning_rate": 3.281971637794998e-05, "loss": 1.6304, "step": 4288 }, { "epoch": 0.038574826979894145, "grad_norm": 3.2986278533935547, "learning_rate": 3.281186343426701e-05, "loss": 2.3373, "step": 4289 }, { "epoch": 0.03858382087753459, "grad_norm": 3.95603346824646, "learning_rate": 3.28040096362904e-05, "loss": 1.8765, "step": 4290 }, { "epoch": 0.038592814775175045, "grad_norm": 4.462499618530273, "learning_rate": 3.2796154984879025e-05, "loss": 2.6613, "step": 4291 }, { "epoch": 0.03860180867281549, "grad_norm": 3.491607904434204, "learning_rate": 3.2788299480891866e-05, "loss": 2.211, "step": 4292 }, { "epoch": 0.038610802570455946, "grad_norm": 1.750543236732483, "learning_rate": 3.2780443125187986e-05, "loss": 2.243, "step": 4293 }, { "epoch": 0.0386197964680964, "grad_norm": 3.6609480381011963, "learning_rate": 3.2772585918626534e-05, "loss": 2.1658, "step": 4294 }, { "epoch": 0.03862879036573685, "grad_norm": 10.434111595153809, "learning_rate": 3.276472786206679e-05, "loss": 2.5753, "step": 4295 }, { "epoch": 0.0386377842633773, "grad_norm": 6.286278247833252, "learning_rate": 3.275686895636807e-05, "loss": 1.8842, "step": 4296 }, { "epoch": 0.03864677816101775, "grad_norm": 2.192120313644409, "learning_rate": 3.2749009202389814e-05, "loss": 1.955, "step": 4297 }, { "epoch": 0.0386557720586582, "grad_norm": 4.731518745422363, "learning_rate": 3.2741148600991576e-05, "loss": 2.0122, "step": 4298 }, { "epoch": 0.03866476595629865, "grad_norm": 2.0036702156066895, "learning_rate": 3.2733287153032966e-05, "loss": 1.722, "step": 4299 }, { "epoch": 0.0386737598539391, "grad_norm": 3.270040273666382, "learning_rate": 3.272542485937369e-05, "loss": 2.2067, "step": 4300 }, { "epoch": 0.038682753751579556, "grad_norm": 2.635401964187622, "learning_rate": 3.271756172087357e-05, "loss": 2.0659, "step": 4301 }, { "epoch": 0.03869174764922, "grad_norm": 1.830160140991211, "learning_rate": 3.27096977383925e-05, "loss": 2.3627, "step": 4302 }, { "epoch": 0.03870074154686046, "grad_norm": 2.2782838344573975, "learning_rate": 3.270183291279049e-05, "loss": 1.9288, "step": 4303 }, { "epoch": 0.0387097354445009, "grad_norm": 4.307344913482666, "learning_rate": 3.26939672449276e-05, "loss": 2.0526, "step": 4304 }, { "epoch": 0.03871872934214136, "grad_norm": 2.9579408168792725, "learning_rate": 3.2686100735664024e-05, "loss": 2.1766, "step": 4305 }, { "epoch": 0.03872772323978181, "grad_norm": 2.4457106590270996, "learning_rate": 3.267823338586002e-05, "loss": 1.8889, "step": 4306 }, { "epoch": 0.03873671713742226, "grad_norm": 3.3466873168945312, "learning_rate": 3.2670365196375965e-05, "loss": 1.7078, "step": 4307 }, { "epoch": 0.03874571103506271, "grad_norm": 2.489750623703003, "learning_rate": 3.2662496168072306e-05, "loss": 1.8166, "step": 4308 }, { "epoch": 0.03875470493270316, "grad_norm": 4.285405158996582, "learning_rate": 3.265462630180959e-05, "loss": 2.0166, "step": 4309 }, { "epoch": 0.03876369883034361, "grad_norm": 2.4836697578430176, "learning_rate": 3.264675559844844e-05, "loss": 1.9274, "step": 4310 }, { "epoch": 0.03877269272798406, "grad_norm": 5.535402297973633, "learning_rate": 3.263888405884959e-05, "loss": 2.1136, "step": 4311 }, { "epoch": 0.03878168662562451, "grad_norm": 4.780120372772217, "learning_rate": 3.263101168387388e-05, "loss": 2.178, "step": 4312 }, { "epoch": 0.03879068052326497, "grad_norm": 2.8333160877227783, "learning_rate": 3.26231384743822e-05, "loss": 2.6837, "step": 4313 }, { "epoch": 0.038799674420905414, "grad_norm": 2.244926691055298, "learning_rate": 3.261526443123555e-05, "loss": 2.5685, "step": 4314 }, { "epoch": 0.03880866831854587, "grad_norm": 5.5114545822143555, "learning_rate": 3.260738955529504e-05, "loss": 2.5228, "step": 4315 }, { "epoch": 0.038817662216186315, "grad_norm": 4.720245361328125, "learning_rate": 3.259951384742184e-05, "loss": 2.407, "step": 4316 }, { "epoch": 0.03882665611382677, "grad_norm": 3.124267578125, "learning_rate": 3.259163730847723e-05, "loss": 2.5508, "step": 4317 }, { "epoch": 0.03883565001146722, "grad_norm": 2.701409339904785, "learning_rate": 3.258375993932259e-05, "loss": 1.8516, "step": 4318 }, { "epoch": 0.03884464390910767, "grad_norm": 4.334951877593994, "learning_rate": 3.2575881740819355e-05, "loss": 2.3258, "step": 4319 }, { "epoch": 0.03885363780674812, "grad_norm": 2.385572910308838, "learning_rate": 3.2568002713829084e-05, "loss": 2.4195, "step": 4320 }, { "epoch": 0.03886263170438857, "grad_norm": 2.720125198364258, "learning_rate": 3.256012285921342e-05, "loss": 2.5452, "step": 4321 }, { "epoch": 0.038871625602029024, "grad_norm": 5.401567459106445, "learning_rate": 3.255224217783408e-05, "loss": 1.8175, "step": 4322 }, { "epoch": 0.03888061949966948, "grad_norm": 3.24943208694458, "learning_rate": 3.2544360670552884e-05, "loss": 1.9426, "step": 4323 }, { "epoch": 0.038889613397309925, "grad_norm": 2.7569997310638428, "learning_rate": 3.253647833823176e-05, "loss": 2.3182, "step": 4324 }, { "epoch": 0.03889860729495038, "grad_norm": 4.420102119445801, "learning_rate": 3.252859518173269e-05, "loss": 2.5269, "step": 4325 }, { "epoch": 0.038907601192590825, "grad_norm": 3.918349266052246, "learning_rate": 3.252071120191777e-05, "loss": 2.106, "step": 4326 }, { "epoch": 0.03891659509023128, "grad_norm": 2.3336400985717773, "learning_rate": 3.2512826399649175e-05, "loss": 2.2121, "step": 4327 }, { "epoch": 0.038925588987871726, "grad_norm": 5.694884300231934, "learning_rate": 3.250494077578919e-05, "loss": 1.768, "step": 4328 }, { "epoch": 0.03893458288551218, "grad_norm": 2.8017115592956543, "learning_rate": 3.249705433120016e-05, "loss": 2.4408, "step": 4329 }, { "epoch": 0.038943576783152634, "grad_norm": 3.8953797817230225, "learning_rate": 3.2489167066744547e-05, "loss": 1.5863, "step": 4330 }, { "epoch": 0.03895257068079308, "grad_norm": 2.523782730102539, "learning_rate": 3.248127898328488e-05, "loss": 2.4953, "step": 4331 }, { "epoch": 0.038961564578433534, "grad_norm": 5.3510613441467285, "learning_rate": 3.2473390081683784e-05, "loss": 2.2203, "step": 4332 }, { "epoch": 0.03897055847607398, "grad_norm": 2.482525587081909, "learning_rate": 3.2465500362803994e-05, "loss": 2.568, "step": 4333 }, { "epoch": 0.038979552373714435, "grad_norm": 2.4625487327575684, "learning_rate": 3.2457609827508306e-05, "loss": 1.7405, "step": 4334 }, { "epoch": 0.03898854627135489, "grad_norm": 2.2560412883758545, "learning_rate": 3.244971847665962e-05, "loss": 2.1792, "step": 4335 }, { "epoch": 0.038997540168995336, "grad_norm": 3.143702745437622, "learning_rate": 3.2441826311120924e-05, "loss": 2.3529, "step": 4336 }, { "epoch": 0.03900653406663579, "grad_norm": 2.6585803031921387, "learning_rate": 3.2433933331755285e-05, "loss": 2.1523, "step": 4337 }, { "epoch": 0.039015527964276236, "grad_norm": 3.095194101333618, "learning_rate": 3.2426039539425876e-05, "loss": 2.5771, "step": 4338 }, { "epoch": 0.03902452186191669, "grad_norm": 2.175900459289551, "learning_rate": 3.241814493499595e-05, "loss": 1.6433, "step": 4339 }, { "epoch": 0.03903351575955714, "grad_norm": 7.469650745391846, "learning_rate": 3.241024951932885e-05, "loss": 2.0873, "step": 4340 }, { "epoch": 0.03904250965719759, "grad_norm": 4.276446342468262, "learning_rate": 3.2402353293287995e-05, "loss": 2.2318, "step": 4341 }, { "epoch": 0.039051503554838045, "grad_norm": 1.6434625387191772, "learning_rate": 3.2394456257736916e-05, "loss": 1.947, "step": 4342 }, { "epoch": 0.03906049745247849, "grad_norm": 2.3507871627807617, "learning_rate": 3.238655841353922e-05, "loss": 1.9581, "step": 4343 }, { "epoch": 0.039069491350118946, "grad_norm": 4.28886604309082, "learning_rate": 3.23786597615586e-05, "loss": 1.4966, "step": 4344 }, { "epoch": 0.03907848524775939, "grad_norm": 2.691953659057617, "learning_rate": 3.237076030265884e-05, "loss": 2.4308, "step": 4345 }, { "epoch": 0.039087479145399846, "grad_norm": 2.90053391456604, "learning_rate": 3.236286003770382e-05, "loss": 1.7306, "step": 4346 }, { "epoch": 0.0390964730430403, "grad_norm": 3.197798490524292, "learning_rate": 3.23549589675575e-05, "loss": 1.9488, "step": 4347 }, { "epoch": 0.03910546694068075, "grad_norm": 4.396193027496338, "learning_rate": 3.234705709308391e-05, "loss": 2.3128, "step": 4348 }, { "epoch": 0.0391144608383212, "grad_norm": 1.9443960189819336, "learning_rate": 3.233915441514722e-05, "loss": 2.3206, "step": 4349 }, { "epoch": 0.03912345473596165, "grad_norm": 2.3521363735198975, "learning_rate": 3.2331250934611624e-05, "loss": 2.3881, "step": 4350 }, { "epoch": 0.0391324486336021, "grad_norm": 4.221792221069336, "learning_rate": 3.232334665234145e-05, "loss": 2.3117, "step": 4351 }, { "epoch": 0.03914144253124255, "grad_norm": 6.220218181610107, "learning_rate": 3.23154415692011e-05, "loss": 2.0508, "step": 4352 }, { "epoch": 0.039150436428883, "grad_norm": 2.40482497215271, "learning_rate": 3.230753568605506e-05, "loss": 1.784, "step": 4353 }, { "epoch": 0.039159430326523456, "grad_norm": 5.488173961639404, "learning_rate": 3.22996290037679e-05, "loss": 2.1316, "step": 4354 }, { "epoch": 0.0391684242241639, "grad_norm": 4.096669673919678, "learning_rate": 3.229172152320429e-05, "loss": 1.7975, "step": 4355 }, { "epoch": 0.03917741812180436, "grad_norm": 4.917896747589111, "learning_rate": 3.228381324522898e-05, "loss": 2.4129, "step": 4356 }, { "epoch": 0.039186412019444804, "grad_norm": 3.07080078125, "learning_rate": 3.2275904170706797e-05, "loss": 2.4414, "step": 4357 }, { "epoch": 0.03919540591708526, "grad_norm": 2.382850408554077, "learning_rate": 3.226799430050268e-05, "loss": 2.0, "step": 4358 }, { "epoch": 0.03920439981472571, "grad_norm": 2.975567579269409, "learning_rate": 3.226008363548163e-05, "loss": 2.189, "step": 4359 }, { "epoch": 0.03921339371236616, "grad_norm": 3.426492929458618, "learning_rate": 3.225217217650876e-05, "loss": 2.3918, "step": 4360 }, { "epoch": 0.03922238761000661, "grad_norm": 4.823116302490234, "learning_rate": 3.2244259924449224e-05, "loss": 2.5039, "step": 4361 }, { "epoch": 0.03923138150764706, "grad_norm": 2.697462558746338, "learning_rate": 3.223634688016832e-05, "loss": 2.1698, "step": 4362 }, { "epoch": 0.03924037540528751, "grad_norm": 4.8622002601623535, "learning_rate": 3.222843304453142e-05, "loss": 2.1362, "step": 4363 }, { "epoch": 0.03924936930292797, "grad_norm": 2.611356496810913, "learning_rate": 3.222051841840392e-05, "loss": 1.7669, "step": 4364 }, { "epoch": 0.039258363200568414, "grad_norm": 5.234675407409668, "learning_rate": 3.22126030026514e-05, "loss": 2.8979, "step": 4365 }, { "epoch": 0.03926735709820887, "grad_norm": 3.5468151569366455, "learning_rate": 3.220468679813946e-05, "loss": 2.5014, "step": 4366 }, { "epoch": 0.039276350995849314, "grad_norm": 3.398686647415161, "learning_rate": 3.2196769805733794e-05, "loss": 2.4855, "step": 4367 }, { "epoch": 0.03928534489348977, "grad_norm": 2.1804540157318115, "learning_rate": 3.218885202630021e-05, "loss": 1.8484, "step": 4368 }, { "epoch": 0.039294338791130215, "grad_norm": 3.371901750564575, "learning_rate": 3.218093346070458e-05, "loss": 2.0172, "step": 4369 }, { "epoch": 0.03930333268877067, "grad_norm": 2.337885856628418, "learning_rate": 3.217301410981285e-05, "loss": 2.1971, "step": 4370 }, { "epoch": 0.03931232658641112, "grad_norm": 3.1446309089660645, "learning_rate": 3.216509397449109e-05, "loss": 2.4068, "step": 4371 }, { "epoch": 0.03932132048405157, "grad_norm": 2.8486554622650146, "learning_rate": 3.215717305560543e-05, "loss": 2.3866, "step": 4372 }, { "epoch": 0.03933031438169202, "grad_norm": 3.070387125015259, "learning_rate": 3.214925135402208e-05, "loss": 2.6215, "step": 4373 }, { "epoch": 0.03933930827933247, "grad_norm": 2.6250152587890625, "learning_rate": 3.214132887060736e-05, "loss": 2.228, "step": 4374 }, { "epoch": 0.039348302176972924, "grad_norm": 4.548611164093018, "learning_rate": 3.213340560622763e-05, "loss": 2.0521, "step": 4375 }, { "epoch": 0.03935729607461338, "grad_norm": 5.187291145324707, "learning_rate": 3.21254815617494e-05, "loss": 2.1113, "step": 4376 }, { "epoch": 0.039366289972253825, "grad_norm": 2.735860586166382, "learning_rate": 3.211755673803922e-05, "loss": 2.0938, "step": 4377 }, { "epoch": 0.03937528386989428, "grad_norm": 4.507180213928223, "learning_rate": 3.210963113596374e-05, "loss": 2.4862, "step": 4378 }, { "epoch": 0.039384277767534726, "grad_norm": 2.8748600482940674, "learning_rate": 3.210170475638969e-05, "loss": 2.0377, "step": 4379 }, { "epoch": 0.03939327166517518, "grad_norm": 2.3630306720733643, "learning_rate": 3.2093777600183875e-05, "loss": 2.2321, "step": 4380 }, { "epoch": 0.039402265562815626, "grad_norm": 4.370933532714844, "learning_rate": 3.208584966821321e-05, "loss": 1.9897, "step": 4381 }, { "epoch": 0.03941125946045608, "grad_norm": 3.1442480087280273, "learning_rate": 3.207792096134468e-05, "loss": 2.4438, "step": 4382 }, { "epoch": 0.039420253358096534, "grad_norm": 4.276913166046143, "learning_rate": 3.206999148044535e-05, "loss": 2.4592, "step": 4383 }, { "epoch": 0.03942924725573698, "grad_norm": 5.664295196533203, "learning_rate": 3.2062061226382395e-05, "loss": 2.2333, "step": 4384 }, { "epoch": 0.039438241153377435, "grad_norm": 4.5931854248046875, "learning_rate": 3.205413020002303e-05, "loss": 1.7291, "step": 4385 }, { "epoch": 0.03944723505101788, "grad_norm": 2.871562957763672, "learning_rate": 3.2046198402234594e-05, "loss": 2.1949, "step": 4386 }, { "epoch": 0.039456228948658335, "grad_norm": 4.101726531982422, "learning_rate": 3.20382658338845e-05, "loss": 2.676, "step": 4387 }, { "epoch": 0.03946522284629879, "grad_norm": 2.7857205867767334, "learning_rate": 3.203033249584023e-05, "loss": 1.7857, "step": 4388 }, { "epoch": 0.039474216743939236, "grad_norm": 6.845382213592529, "learning_rate": 3.202239838896937e-05, "loss": 1.9435, "step": 4389 }, { "epoch": 0.03948321064157969, "grad_norm": 3.5767600536346436, "learning_rate": 3.201446351413958e-05, "loss": 2.529, "step": 4390 }, { "epoch": 0.03949220453922014, "grad_norm": 3.667649507522583, "learning_rate": 3.200652787221861e-05, "loss": 2.1024, "step": 4391 }, { "epoch": 0.03950119843686059, "grad_norm": 2.382831335067749, "learning_rate": 3.199859146407429e-05, "loss": 2.2269, "step": 4392 }, { "epoch": 0.039510192334501044, "grad_norm": 1.5013920068740845, "learning_rate": 3.1990654290574516e-05, "loss": 2.542, "step": 4393 }, { "epoch": 0.03951918623214149, "grad_norm": 2.6993794441223145, "learning_rate": 3.198271635258731e-05, "loss": 1.8016, "step": 4394 }, { "epoch": 0.039528180129781945, "grad_norm": 2.7407748699188232, "learning_rate": 3.1974777650980735e-05, "loss": 2.1134, "step": 4395 }, { "epoch": 0.03953717402742239, "grad_norm": 4.554697513580322, "learning_rate": 3.1966838186622965e-05, "loss": 2.246, "step": 4396 }, { "epoch": 0.039546167925062846, "grad_norm": 4.118362903594971, "learning_rate": 3.195889796038225e-05, "loss": 1.8496, "step": 4397 }, { "epoch": 0.03955516182270329, "grad_norm": 4.591533660888672, "learning_rate": 3.1950956973126906e-05, "loss": 2.1944, "step": 4398 }, { "epoch": 0.03956415572034375, "grad_norm": 3.5563111305236816, "learning_rate": 3.194301522572537e-05, "loss": 2.4924, "step": 4399 }, { "epoch": 0.0395731496179842, "grad_norm": 3.0021395683288574, "learning_rate": 3.1935072719046115e-05, "loss": 2.085, "step": 4400 }, { "epoch": 0.03958214351562465, "grad_norm": 3.3939313888549805, "learning_rate": 3.192712945395774e-05, "loss": 2.1009, "step": 4401 }, { "epoch": 0.0395911374132651, "grad_norm": 2.7173895835876465, "learning_rate": 3.19191854313289e-05, "loss": 1.7074, "step": 4402 }, { "epoch": 0.03960013131090555, "grad_norm": 3.3343915939331055, "learning_rate": 3.1911240652028354e-05, "loss": 1.6416, "step": 4403 }, { "epoch": 0.039609125208546, "grad_norm": 5.738031387329102, "learning_rate": 3.190329511692491e-05, "loss": 2.2702, "step": 4404 }, { "epoch": 0.039618119106186456, "grad_norm": 3.629852533340454, "learning_rate": 3.18953488268875e-05, "loss": 2.0151, "step": 4405 }, { "epoch": 0.0396271130038269, "grad_norm": 4.612259864807129, "learning_rate": 3.188740178278511e-05, "loss": 1.9894, "step": 4406 }, { "epoch": 0.039636106901467356, "grad_norm": 3.940159559249878, "learning_rate": 3.187945398548681e-05, "loss": 2.2469, "step": 4407 }, { "epoch": 0.0396451007991078, "grad_norm": 2.5165019035339355, "learning_rate": 3.187150543586176e-05, "loss": 2.4791, "step": 4408 }, { "epoch": 0.03965409469674826, "grad_norm": 2.8162786960601807, "learning_rate": 3.1863556134779216e-05, "loss": 1.598, "step": 4409 }, { "epoch": 0.039663088594388704, "grad_norm": 2.4106125831604004, "learning_rate": 3.185560608310849e-05, "loss": 2.3025, "step": 4410 }, { "epoch": 0.03967208249202916, "grad_norm": 2.642228126525879, "learning_rate": 3.1847655281719e-05, "loss": 2.3933, "step": 4411 }, { "epoch": 0.03968107638966961, "grad_norm": 2.6745035648345947, "learning_rate": 3.1839703731480205e-05, "loss": 2.0419, "step": 4412 }, { "epoch": 0.03969007028731006, "grad_norm": 4.149704456329346, "learning_rate": 3.1831751433261705e-05, "loss": 1.4929, "step": 4413 }, { "epoch": 0.03969906418495051, "grad_norm": 2.5277462005615234, "learning_rate": 3.1823798387933134e-05, "loss": 2.2106, "step": 4414 }, { "epoch": 0.03970805808259096, "grad_norm": 1.8907294273376465, "learning_rate": 3.181584459636423e-05, "loss": 2.3693, "step": 4415 }, { "epoch": 0.03971705198023141, "grad_norm": 2.977020740509033, "learning_rate": 3.180789005942482e-05, "loss": 1.7244, "step": 4416 }, { "epoch": 0.03972604587787187, "grad_norm": 2.4551546573638916, "learning_rate": 3.179993477798477e-05, "loss": 1.8342, "step": 4417 }, { "epoch": 0.039735039775512314, "grad_norm": 2.8113324642181396, "learning_rate": 3.1791978752914085e-05, "loss": 2.0479, "step": 4418 }, { "epoch": 0.03974403367315277, "grad_norm": 2.56562876701355, "learning_rate": 3.178402198508282e-05, "loss": 2.188, "step": 4419 }, { "epoch": 0.039753027570793215, "grad_norm": 2.864905595779419, "learning_rate": 3.1776064475361114e-05, "loss": 2.4693, "step": 4420 }, { "epoch": 0.03976202146843367, "grad_norm": 4.420742988586426, "learning_rate": 3.176810622461917e-05, "loss": 2.3754, "step": 4421 }, { "epoch": 0.039771015366074115, "grad_norm": 3.0094571113586426, "learning_rate": 3.176014723372731e-05, "loss": 2.5497, "step": 4422 }, { "epoch": 0.03978000926371457, "grad_norm": 4.488681793212891, "learning_rate": 3.175218750355592e-05, "loss": 1.9629, "step": 4423 }, { "epoch": 0.03978900316135502, "grad_norm": 3.401949644088745, "learning_rate": 3.174422703497545e-05, "loss": 1.626, "step": 4424 }, { "epoch": 0.03979799705899547, "grad_norm": 3.3223276138305664, "learning_rate": 3.173626582885645e-05, "loss": 2.1393, "step": 4425 }, { "epoch": 0.039806990956635924, "grad_norm": 4.4796833992004395, "learning_rate": 3.172830388606955e-05, "loss": 2.0311, "step": 4426 }, { "epoch": 0.03981598485427637, "grad_norm": 2.1604762077331543, "learning_rate": 3.172034120748546e-05, "loss": 2.4345, "step": 4427 }, { "epoch": 0.039824978751916824, "grad_norm": 5.125059127807617, "learning_rate": 3.171237779397494e-05, "loss": 2.2105, "step": 4428 }, { "epoch": 0.03983397264955728, "grad_norm": 2.1092917919158936, "learning_rate": 3.1704413646408894e-05, "loss": 2.4036, "step": 4429 }, { "epoch": 0.039842966547197725, "grad_norm": 2.964672327041626, "learning_rate": 3.169644876565824e-05, "loss": 2.6079, "step": 4430 }, { "epoch": 0.03985196044483818, "grad_norm": 3.884974718093872, "learning_rate": 3.168848315259402e-05, "loss": 2.0662, "step": 4431 }, { "epoch": 0.039860954342478626, "grad_norm": 2.7391366958618164, "learning_rate": 3.168051680808735e-05, "loss": 1.8761, "step": 4432 }, { "epoch": 0.03986994824011908, "grad_norm": 4.363523483276367, "learning_rate": 3.1672549733009396e-05, "loss": 2.2527, "step": 4433 }, { "epoch": 0.039878942137759534, "grad_norm": 2.182360887527466, "learning_rate": 3.1664581928231426e-05, "loss": 2.1946, "step": 4434 }, { "epoch": 0.03988793603539998, "grad_norm": 3.200847625732422, "learning_rate": 3.165661339462481e-05, "loss": 2.5935, "step": 4435 }, { "epoch": 0.039896929933040434, "grad_norm": 2.851931095123291, "learning_rate": 3.164864413306096e-05, "loss": 2.168, "step": 4436 }, { "epoch": 0.03990592383068088, "grad_norm": 3.9471848011016846, "learning_rate": 3.164067414441138e-05, "loss": 2.7946, "step": 4437 }, { "epoch": 0.039914917728321335, "grad_norm": 2.914346218109131, "learning_rate": 3.163270342954766e-05, "loss": 2.2003, "step": 4438 }, { "epoch": 0.03992391162596178, "grad_norm": 2.5963990688323975, "learning_rate": 3.162473198934147e-05, "loss": 2.421, "step": 4439 }, { "epoch": 0.039932905523602236, "grad_norm": 5.081168174743652, "learning_rate": 3.161675982466454e-05, "loss": 2.5363, "step": 4440 }, { "epoch": 0.03994189942124269, "grad_norm": 3.8369362354278564, "learning_rate": 3.160878693638871e-05, "loss": 2.2349, "step": 4441 }, { "epoch": 0.039950893318883136, "grad_norm": 3.666909694671631, "learning_rate": 3.1600813325385883e-05, "loss": 2.509, "step": 4442 }, { "epoch": 0.03995988721652359, "grad_norm": 3.2705352306365967, "learning_rate": 3.1592838992528024e-05, "loss": 2.6255, "step": 4443 }, { "epoch": 0.03996888111416404, "grad_norm": 2.0913946628570557, "learning_rate": 3.1584863938687206e-05, "loss": 2.2583, "step": 4444 }, { "epoch": 0.03997787501180449, "grad_norm": 2.8258304595947266, "learning_rate": 3.1576888164735575e-05, "loss": 1.9991, "step": 4445 }, { "epoch": 0.039986868909444945, "grad_norm": 6.435807704925537, "learning_rate": 3.156891167154533e-05, "loss": 1.6165, "step": 4446 }, { "epoch": 0.03999586280708539, "grad_norm": 3.3804774284362793, "learning_rate": 3.1560934459988795e-05, "loss": 2.3231, "step": 4447 }, { "epoch": 0.040004856704725845, "grad_norm": 3.593086004257202, "learning_rate": 3.1552956530938324e-05, "loss": 1.8958, "step": 4448 }, { "epoch": 0.04001385060236629, "grad_norm": 2.6224887371063232, "learning_rate": 3.154497788526638e-05, "loss": 1.9272, "step": 4449 }, { "epoch": 0.040022844500006746, "grad_norm": 2.9088077545166016, "learning_rate": 3.1536998523845494e-05, "loss": 2.2838, "step": 4450 }, { "epoch": 0.04003183839764719, "grad_norm": 3.3755316734313965, "learning_rate": 3.152901844754828e-05, "loss": 2.1376, "step": 4451 }, { "epoch": 0.04004083229528765, "grad_norm": 4.887547492980957, "learning_rate": 3.152103765724743e-05, "loss": 2.0521, "step": 4452 }, { "epoch": 0.0400498261929281, "grad_norm": 2.367882013320923, "learning_rate": 3.151305615381569e-05, "loss": 1.7693, "step": 4453 }, { "epoch": 0.04005882009056855, "grad_norm": 3.224438190460205, "learning_rate": 3.150507393812594e-05, "loss": 2.4852, "step": 4454 }, { "epoch": 0.040067813988209, "grad_norm": 3.129493474960327, "learning_rate": 3.149709101105107e-05, "loss": 2.2024, "step": 4455 }, { "epoch": 0.04007680788584945, "grad_norm": 3.362527847290039, "learning_rate": 3.14891073734641e-05, "loss": 2.0337, "step": 4456 }, { "epoch": 0.0400858017834899, "grad_norm": 2.5765397548675537, "learning_rate": 3.148112302623811e-05, "loss": 2.3498, "step": 4457 }, { "epoch": 0.040094795681130356, "grad_norm": 6.152495861053467, "learning_rate": 3.1473137970246246e-05, "loss": 1.6705, "step": 4458 }, { "epoch": 0.0401037895787708, "grad_norm": 3.144834518432617, "learning_rate": 3.1465152206361736e-05, "loss": 2.2484, "step": 4459 }, { "epoch": 0.04011278347641126, "grad_norm": 4.3772478103637695, "learning_rate": 3.145716573545792e-05, "loss": 2.2273, "step": 4460 }, { "epoch": 0.040121777374051704, "grad_norm": 5.1059393882751465, "learning_rate": 3.1449178558408155e-05, "loss": 2.2487, "step": 4461 }, { "epoch": 0.04013077127169216, "grad_norm": 2.6911251544952393, "learning_rate": 3.1441190676085916e-05, "loss": 2.2086, "step": 4462 }, { "epoch": 0.04013976516933261, "grad_norm": 5.261659145355225, "learning_rate": 3.1433202089364756e-05, "loss": 1.8093, "step": 4463 }, { "epoch": 0.04014875906697306, "grad_norm": 7.672554016113281, "learning_rate": 3.1425212799118284e-05, "loss": 2.3062, "step": 4464 }, { "epoch": 0.04015775296461351, "grad_norm": 2.868000030517578, "learning_rate": 3.141722280622021e-05, "loss": 2.7304, "step": 4465 }, { "epoch": 0.04016674686225396, "grad_norm": 2.9154093265533447, "learning_rate": 3.140923211154428e-05, "loss": 2.2463, "step": 4466 }, { "epoch": 0.04017574075989441, "grad_norm": 1.9633097648620605, "learning_rate": 3.140124071596437e-05, "loss": 2.2432, "step": 4467 }, { "epoch": 0.04018473465753486, "grad_norm": 2.3700075149536133, "learning_rate": 3.139324862035441e-05, "loss": 2.1427, "step": 4468 }, { "epoch": 0.04019372855517531, "grad_norm": 2.5848801136016846, "learning_rate": 3.1385255825588387e-05, "loss": 2.3029, "step": 4469 }, { "epoch": 0.04020272245281577, "grad_norm": 4.401237964630127, "learning_rate": 3.137726233254038e-05, "loss": 2.1087, "step": 4470 }, { "epoch": 0.040211716350456214, "grad_norm": 3.1035449504852295, "learning_rate": 3.1369268142084556e-05, "loss": 2.1444, "step": 4471 }, { "epoch": 0.04022071024809667, "grad_norm": 4.548457145690918, "learning_rate": 3.1361273255095146e-05, "loss": 1.9572, "step": 4472 }, { "epoch": 0.040229704145737115, "grad_norm": 3.1576225757598877, "learning_rate": 3.135327767244646e-05, "loss": 2.7233, "step": 4473 }, { "epoch": 0.04023869804337757, "grad_norm": 2.21429705619812, "learning_rate": 3.1345281395012874e-05, "loss": 2.2488, "step": 4474 }, { "epoch": 0.04024769194101802, "grad_norm": 6.351428985595703, "learning_rate": 3.133728442366885e-05, "loss": 2.1201, "step": 4475 }, { "epoch": 0.04025668583865847, "grad_norm": 2.47976016998291, "learning_rate": 3.132928675928894e-05, "loss": 2.3013, "step": 4476 }, { "epoch": 0.04026567973629892, "grad_norm": 3.257352590560913, "learning_rate": 3.132128840274774e-05, "loss": 2.0943, "step": 4477 }, { "epoch": 0.04027467363393937, "grad_norm": 5.758517742156982, "learning_rate": 3.1313289354919945e-05, "loss": 2.0821, "step": 4478 }, { "epoch": 0.040283667531579824, "grad_norm": 4.443996429443359, "learning_rate": 3.130528961668032e-05, "loss": 2.6603, "step": 4479 }, { "epoch": 0.04029266142922027, "grad_norm": 2.6382553577423096, "learning_rate": 3.129728918890371e-05, "loss": 1.995, "step": 4480 }, { "epoch": 0.040301655326860725, "grad_norm": 2.5312957763671875, "learning_rate": 3.1289288072465e-05, "loss": 2.1511, "step": 4481 }, { "epoch": 0.04031064922450118, "grad_norm": 2.3012242317199707, "learning_rate": 3.128128626823922e-05, "loss": 1.7298, "step": 4482 }, { "epoch": 0.040319643122141625, "grad_norm": 4.174227714538574, "learning_rate": 3.127328377710142e-05, "loss": 2.2133, "step": 4483 }, { "epoch": 0.04032863701978208, "grad_norm": 9.364326477050781, "learning_rate": 3.126528059992673e-05, "loss": 2.2749, "step": 4484 }, { "epoch": 0.040337630917422526, "grad_norm": 4.659061908721924, "learning_rate": 3.1257276737590365e-05, "loss": 2.3417, "step": 4485 }, { "epoch": 0.04034662481506298, "grad_norm": 2.5992133617401123, "learning_rate": 3.124927219096764e-05, "loss": 2.3198, "step": 4486 }, { "epoch": 0.040355618712703434, "grad_norm": 3.3685085773468018, "learning_rate": 3.1241266960933904e-05, "loss": 2.274, "step": 4487 }, { "epoch": 0.04036461261034388, "grad_norm": 3.486923933029175, "learning_rate": 3.123326104836458e-05, "loss": 1.7677, "step": 4488 }, { "epoch": 0.040373606507984335, "grad_norm": 8.132150650024414, "learning_rate": 3.122525445413521e-05, "loss": 2.0423, "step": 4489 }, { "epoch": 0.04038260040562478, "grad_norm": 2.5512638092041016, "learning_rate": 3.121724717912138e-05, "loss": 2.2306, "step": 4490 }, { "epoch": 0.040391594303265235, "grad_norm": 2.9294052124023438, "learning_rate": 3.1209239224198724e-05, "loss": 2.3533, "step": 4491 }, { "epoch": 0.04040058820090568, "grad_norm": 2.983555555343628, "learning_rate": 3.120123059024303e-05, "loss": 1.4956, "step": 4492 }, { "epoch": 0.040409582098546136, "grad_norm": 2.9871180057525635, "learning_rate": 3.119322127813006e-05, "loss": 1.673, "step": 4493 }, { "epoch": 0.04041857599618659, "grad_norm": 4.176706314086914, "learning_rate": 3.118521128873573e-05, "loss": 2.3685, "step": 4494 }, { "epoch": 0.04042756989382704, "grad_norm": 5.452051639556885, "learning_rate": 3.117720062293599e-05, "loss": 2.6583, "step": 4495 }, { "epoch": 0.04043656379146749, "grad_norm": 2.488642454147339, "learning_rate": 3.116918928160688e-05, "loss": 2.542, "step": 4496 }, { "epoch": 0.04044555768910794, "grad_norm": 3.087069511413574, "learning_rate": 3.116117726562451e-05, "loss": 2.2111, "step": 4497 }, { "epoch": 0.04045455158674839, "grad_norm": 7.15234899520874, "learning_rate": 3.1153164575865046e-05, "loss": 2.6072, "step": 4498 }, { "epoch": 0.040463545484388845, "grad_norm": 3.638173818588257, "learning_rate": 3.114515121320476e-05, "loss": 1.9829, "step": 4499 }, { "epoch": 0.04047253938202929, "grad_norm": 2.5818729400634766, "learning_rate": 3.1137137178519985e-05, "loss": 2.3155, "step": 4500 }, { "epoch": 0.040481533279669746, "grad_norm": 2.2401468753814697, "learning_rate": 3.11291224726871e-05, "loss": 1.8842, "step": 4501 }, { "epoch": 0.04049052717731019, "grad_norm": 2.4359731674194336, "learning_rate": 3.112110709658261e-05, "loss": 1.8988, "step": 4502 }, { "epoch": 0.040499521074950647, "grad_norm": 4.107000350952148, "learning_rate": 3.111309105108304e-05, "loss": 2.0325, "step": 4503 }, { "epoch": 0.0405085149725911, "grad_norm": 2.9820051193237305, "learning_rate": 3.1105074337065035e-05, "loss": 1.753, "step": 4504 }, { "epoch": 0.04051750887023155, "grad_norm": 2.286076068878174, "learning_rate": 3.1097056955405274e-05, "loss": 2.4881, "step": 4505 }, { "epoch": 0.040526502767872, "grad_norm": 2.4013864994049072, "learning_rate": 3.108903890698053e-05, "loss": 2.2212, "step": 4506 }, { "epoch": 0.04053549666551245, "grad_norm": 3.333650827407837, "learning_rate": 3.108102019266764e-05, "loss": 2.2114, "step": 4507 }, { "epoch": 0.0405444905631529, "grad_norm": 2.9514589309692383, "learning_rate": 3.107300081334353e-05, "loss": 1.917, "step": 4508 }, { "epoch": 0.04055348446079335, "grad_norm": 10.602460861206055, "learning_rate": 3.1064980769885187e-05, "loss": 2.0603, "step": 4509 }, { "epoch": 0.0405624783584338, "grad_norm": 5.373595237731934, "learning_rate": 3.105696006316966e-05, "loss": 2.4813, "step": 4510 }, { "epoch": 0.040571472256074256, "grad_norm": 3.8554179668426514, "learning_rate": 3.1048938694074095e-05, "loss": 2.2675, "step": 4511 }, { "epoch": 0.0405804661537147, "grad_norm": 3.5284087657928467, "learning_rate": 3.1040916663475676e-05, "loss": 2.046, "step": 4512 }, { "epoch": 0.04058946005135516, "grad_norm": 3.689079523086548, "learning_rate": 3.10328939722517e-05, "loss": 2.3533, "step": 4513 }, { "epoch": 0.040598453948995604, "grad_norm": 2.977248430252075, "learning_rate": 3.102487062127951e-05, "loss": 1.9039, "step": 4514 }, { "epoch": 0.04060744784663606, "grad_norm": 13.893518447875977, "learning_rate": 3.101684661143653e-05, "loss": 2.2459, "step": 4515 }, { "epoch": 0.04061644174427651, "grad_norm": 2.655801296234131, "learning_rate": 3.1008821943600244e-05, "loss": 2.4635, "step": 4516 }, { "epoch": 0.04062543564191696, "grad_norm": 3.7471492290496826, "learning_rate": 3.1000796618648225e-05, "loss": 2.1288, "step": 4517 }, { "epoch": 0.04063442953955741, "grad_norm": 3.8757426738739014, "learning_rate": 3.099277063745811e-05, "loss": 1.8932, "step": 4518 }, { "epoch": 0.04064342343719786, "grad_norm": 2.6311488151550293, "learning_rate": 3.0984744000907615e-05, "loss": 2.1857, "step": 4519 }, { "epoch": 0.04065241733483831, "grad_norm": 2.120417833328247, "learning_rate": 3.0976716709874496e-05, "loss": 2.2582, "step": 4520 }, { "epoch": 0.04066141123247876, "grad_norm": 3.7279436588287354, "learning_rate": 3.0968688765236643e-05, "loss": 2.3281, "step": 4521 }, { "epoch": 0.040670405130119214, "grad_norm": 3.610611915588379, "learning_rate": 3.096066016787195e-05, "loss": 1.8932, "step": 4522 }, { "epoch": 0.04067939902775967, "grad_norm": 2.8240807056427, "learning_rate": 3.095263091865843e-05, "loss": 2.4568, "step": 4523 }, { "epoch": 0.040688392925400114, "grad_norm": 3.4517974853515625, "learning_rate": 3.094460101847414e-05, "loss": 2.0357, "step": 4524 }, { "epoch": 0.04069738682304057, "grad_norm": 2.917799234390259, "learning_rate": 3.093657046819722e-05, "loss": 2.4032, "step": 4525 }, { "epoch": 0.040706380720681015, "grad_norm": 4.24191951751709, "learning_rate": 3.0928539268705875e-05, "loss": 2.0906, "step": 4526 }, { "epoch": 0.04071537461832147, "grad_norm": 2.6999173164367676, "learning_rate": 3.092050742087839e-05, "loss": 2.2969, "step": 4527 }, { "epoch": 0.04072436851596192, "grad_norm": 3.671786069869995, "learning_rate": 3.091247492559312e-05, "loss": 2.0587, "step": 4528 }, { "epoch": 0.04073336241360237, "grad_norm": 3.27164626121521, "learning_rate": 3.090444178372848e-05, "loss": 2.085, "step": 4529 }, { "epoch": 0.040742356311242824, "grad_norm": 2.593067169189453, "learning_rate": 3.0896407996162954e-05, "loss": 2.4671, "step": 4530 }, { "epoch": 0.04075135020888327, "grad_norm": 3.188093900680542, "learning_rate": 3.088837356377513e-05, "loss": 2.4326, "step": 4531 }, { "epoch": 0.040760344106523724, "grad_norm": 3.372262954711914, "learning_rate": 3.088033848744362e-05, "loss": 2.5005, "step": 4532 }, { "epoch": 0.04076933800416417, "grad_norm": 3.4084770679473877, "learning_rate": 3.087230276804713e-05, "loss": 2.4582, "step": 4533 }, { "epoch": 0.040778331901804625, "grad_norm": 6.937479019165039, "learning_rate": 3.086426640646444e-05, "loss": 2.0905, "step": 4534 }, { "epoch": 0.04078732579944508, "grad_norm": 4.982515335083008, "learning_rate": 3.08562294035744e-05, "loss": 1.7303, "step": 4535 }, { "epoch": 0.040796319697085526, "grad_norm": 2.857154130935669, "learning_rate": 3.084819176025591e-05, "loss": 1.8803, "step": 4536 }, { "epoch": 0.04080531359472598, "grad_norm": 3.6527657508850098, "learning_rate": 3.0840153477387953e-05, "loss": 2.7055, "step": 4537 }, { "epoch": 0.040814307492366426, "grad_norm": 2.455681324005127, "learning_rate": 3.0832114555849614e-05, "loss": 2.3664, "step": 4538 }, { "epoch": 0.04082330139000688, "grad_norm": 4.747416973114014, "learning_rate": 3.0824074996519975e-05, "loss": 2.4819, "step": 4539 }, { "epoch": 0.040832295287647334, "grad_norm": 2.7658262252807617, "learning_rate": 3.081603480027826e-05, "loss": 1.8566, "step": 4540 }, { "epoch": 0.04084128918528778, "grad_norm": 2.9926388263702393, "learning_rate": 3.080799396800372e-05, "loss": 1.9136, "step": 4541 }, { "epoch": 0.040850283082928235, "grad_norm": 4.580524444580078, "learning_rate": 3.0799952500575694e-05, "loss": 1.5302, "step": 4542 }, { "epoch": 0.04085927698056868, "grad_norm": 6.170639991760254, "learning_rate": 3.079191039887358e-05, "loss": 2.0837, "step": 4543 }, { "epoch": 0.040868270878209136, "grad_norm": 2.2412362098693848, "learning_rate": 3.078386766377686e-05, "loss": 2.2115, "step": 4544 }, { "epoch": 0.04087726477584959, "grad_norm": 2.326286792755127, "learning_rate": 3.077582429616506e-05, "loss": 2.4071, "step": 4545 }, { "epoch": 0.040886258673490036, "grad_norm": 3.490487813949585, "learning_rate": 3.0767780296917803e-05, "loss": 2.0242, "step": 4546 }, { "epoch": 0.04089525257113049, "grad_norm": 3.913335084915161, "learning_rate": 3.075973566691477e-05, "loss": 2.074, "step": 4547 }, { "epoch": 0.04090424646877094, "grad_norm": 2.576481819152832, "learning_rate": 3.07516904070357e-05, "loss": 2.3105, "step": 4548 }, { "epoch": 0.04091324036641139, "grad_norm": 4.732501029968262, "learning_rate": 3.074364451816041e-05, "loss": 2.4196, "step": 4549 }, { "epoch": 0.04092223426405184, "grad_norm": 2.35856556892395, "learning_rate": 3.073559800116879e-05, "loss": 1.6907, "step": 4550 }, { "epoch": 0.04093122816169229, "grad_norm": 6.630499839782715, "learning_rate": 3.0727550856940804e-05, "loss": 2.4379, "step": 4551 }, { "epoch": 0.040940222059332745, "grad_norm": 4.2560272216796875, "learning_rate": 3.0719503086356464e-05, "loss": 1.6236, "step": 4552 }, { "epoch": 0.04094921595697319, "grad_norm": 3.768380880355835, "learning_rate": 3.071145469029586e-05, "loss": 1.8548, "step": 4553 }, { "epoch": 0.040958209854613646, "grad_norm": 3.5521621704101562, "learning_rate": 3.070340566963917e-05, "loss": 2.4573, "step": 4554 }, { "epoch": 0.04096720375225409, "grad_norm": 1.7812074422836304, "learning_rate": 3.06953560252666e-05, "loss": 2.3019, "step": 4555 }, { "epoch": 0.04097619764989455, "grad_norm": 2.545050621032715, "learning_rate": 3.068730575805846e-05, "loss": 2.6382, "step": 4556 }, { "epoch": 0.040985191547535, "grad_norm": 2.1962268352508545, "learning_rate": 3.067925486889512e-05, "loss": 1.9604, "step": 4557 }, { "epoch": 0.04099418544517545, "grad_norm": 1.826528549194336, "learning_rate": 3.0671203358657e-05, "loss": 2.0918, "step": 4558 }, { "epoch": 0.0410031793428159, "grad_norm": 2.184936761856079, "learning_rate": 3.0663151228224605e-05, "loss": 2.2394, "step": 4559 }, { "epoch": 0.04101217324045635, "grad_norm": 4.263561725616455, "learning_rate": 3.065509847847851e-05, "loss": 1.7964, "step": 4560 }, { "epoch": 0.0410211671380968, "grad_norm": 4.149204254150391, "learning_rate": 3.0647045110299346e-05, "loss": 2.2654, "step": 4561 }, { "epoch": 0.04103016103573725, "grad_norm": 2.2871975898742676, "learning_rate": 3.063899112456782e-05, "loss": 2.4051, "step": 4562 }, { "epoch": 0.0410391549333777, "grad_norm": 2.996804714202881, "learning_rate": 3.0630936522164706e-05, "loss": 2.3193, "step": 4563 }, { "epoch": 0.04104814883101816, "grad_norm": 4.705615997314453, "learning_rate": 3.0622881303970834e-05, "loss": 2.2409, "step": 4564 }, { "epoch": 0.041057142728658604, "grad_norm": 3.140028953552246, "learning_rate": 3.061482547086712e-05, "loss": 2.1177, "step": 4565 }, { "epoch": 0.04106613662629906, "grad_norm": 2.8806705474853516, "learning_rate": 3.0606769023734536e-05, "loss": 2.3242, "step": 4566 }, { "epoch": 0.041075130523939504, "grad_norm": 3.4330923557281494, "learning_rate": 3.0598711963454115e-05, "loss": 1.8961, "step": 4567 }, { "epoch": 0.04108412442157996, "grad_norm": 4.710167407989502, "learning_rate": 3.059065429090697e-05, "loss": 1.6409, "step": 4568 }, { "epoch": 0.04109311831922041, "grad_norm": 2.800412654876709, "learning_rate": 3.058259600697428e-05, "loss": 2.2767, "step": 4569 }, { "epoch": 0.04110211221686086, "grad_norm": 3.468817949295044, "learning_rate": 3.05745371125373e-05, "loss": 1.9372, "step": 4570 }, { "epoch": 0.04111110611450131, "grad_norm": 3.7568464279174805, "learning_rate": 3.0566477608477295e-05, "loss": 2.4331, "step": 4571 }, { "epoch": 0.04112010001214176, "grad_norm": 2.5355169773101807, "learning_rate": 3.055841749567569e-05, "loss": 2.0395, "step": 4572 }, { "epoch": 0.04112909390978221, "grad_norm": 5.615525722503662, "learning_rate": 3.0550356775013894e-05, "loss": 2.3574, "step": 4573 }, { "epoch": 0.04113808780742267, "grad_norm": 2.182905435562134, "learning_rate": 3.054229544737343e-05, "loss": 2.5205, "step": 4574 }, { "epoch": 0.041147081705063114, "grad_norm": 3.97705078125, "learning_rate": 3.053423351363586e-05, "loss": 1.7919, "step": 4575 }, { "epoch": 0.04115607560270357, "grad_norm": 3.336205244064331, "learning_rate": 3.052617097468284e-05, "loss": 2.2922, "step": 4576 }, { "epoch": 0.041165069500344015, "grad_norm": 2.891301393508911, "learning_rate": 3.051810783139607e-05, "loss": 1.9826, "step": 4577 }, { "epoch": 0.04117406339798447, "grad_norm": 2.749983549118042, "learning_rate": 3.051004408465733e-05, "loss": 2.3033, "step": 4578 }, { "epoch": 0.041183057295624916, "grad_norm": 4.606414318084717, "learning_rate": 3.050197973534845e-05, "loss": 2.4502, "step": 4579 }, { "epoch": 0.04119205119326537, "grad_norm": 3.7500758171081543, "learning_rate": 3.0493914784351328e-05, "loss": 2.2239, "step": 4580 }, { "epoch": 0.04120104509090582, "grad_norm": 4.545022010803223, "learning_rate": 3.048584923254796e-05, "loss": 2.5089, "step": 4581 }, { "epoch": 0.04121003898854627, "grad_norm": 4.400893211364746, "learning_rate": 3.0477783080820362e-05, "loss": 2.6419, "step": 4582 }, { "epoch": 0.041219032886186724, "grad_norm": 4.0164794921875, "learning_rate": 3.046971633005064e-05, "loss": 1.8725, "step": 4583 }, { "epoch": 0.04122802678382717, "grad_norm": 2.5265371799468994, "learning_rate": 3.046164898112096e-05, "loss": 2.2622, "step": 4584 }, { "epoch": 0.041237020681467625, "grad_norm": 3.893747568130493, "learning_rate": 3.045358103491357e-05, "loss": 1.8848, "step": 4585 }, { "epoch": 0.04124601457910808, "grad_norm": 4.0674309730529785, "learning_rate": 3.0445512492310753e-05, "loss": 2.0346, "step": 4586 }, { "epoch": 0.041255008476748525, "grad_norm": 2.7088096141815186, "learning_rate": 3.0437443354194872e-05, "loss": 2.124, "step": 4587 }, { "epoch": 0.04126400237438898, "grad_norm": 2.813157081604004, "learning_rate": 3.0429373621448365e-05, "loss": 2.0165, "step": 4588 }, { "epoch": 0.041272996272029426, "grad_norm": 4.327122211456299, "learning_rate": 3.042130329495372e-05, "loss": 1.8839, "step": 4589 }, { "epoch": 0.04128199016966988, "grad_norm": 3.794135808944702, "learning_rate": 3.0413232375593497e-05, "loss": 2.1672, "step": 4590 }, { "epoch": 0.04129098406731033, "grad_norm": 2.408588171005249, "learning_rate": 3.0405160864250314e-05, "loss": 1.8898, "step": 4591 }, { "epoch": 0.04129997796495078, "grad_norm": 2.908252239227295, "learning_rate": 3.039708876180688e-05, "loss": 2.5755, "step": 4592 }, { "epoch": 0.041308971862591234, "grad_norm": 2.4161999225616455, "learning_rate": 3.038901606914592e-05, "loss": 2.3031, "step": 4593 }, { "epoch": 0.04131796576023168, "grad_norm": 4.134853839874268, "learning_rate": 3.0380942787150273e-05, "loss": 2.401, "step": 4594 }, { "epoch": 0.041326959657872135, "grad_norm": 4.283020496368408, "learning_rate": 3.037286891670281e-05, "loss": 1.9943, "step": 4595 }, { "epoch": 0.04133595355551258, "grad_norm": 3.8625786304473877, "learning_rate": 3.036479445868648e-05, "loss": 1.7346, "step": 4596 }, { "epoch": 0.041344947453153036, "grad_norm": 2.6459827423095703, "learning_rate": 3.0356719413984296e-05, "loss": 2.2845, "step": 4597 }, { "epoch": 0.04135394135079349, "grad_norm": 2.1875088214874268, "learning_rate": 3.034864378347933e-05, "loss": 1.9255, "step": 4598 }, { "epoch": 0.04136293524843394, "grad_norm": 3.621267795562744, "learning_rate": 3.0340567568054716e-05, "loss": 2.0877, "step": 4599 }, { "epoch": 0.04137192914607439, "grad_norm": 3.2208244800567627, "learning_rate": 3.0332490768593675e-05, "loss": 2.672, "step": 4600 }, { "epoch": 0.04138092304371484, "grad_norm": 2.3920512199401855, "learning_rate": 3.032441338597945e-05, "loss": 2.0539, "step": 4601 }, { "epoch": 0.04138991694135529, "grad_norm": 3.507861614227295, "learning_rate": 3.0316335421095388e-05, "loss": 2.0519, "step": 4602 }, { "epoch": 0.04139891083899574, "grad_norm": 6.4414167404174805, "learning_rate": 3.030825687482487e-05, "loss": 2.1616, "step": 4603 }, { "epoch": 0.04140790473663619, "grad_norm": 1.9339114427566528, "learning_rate": 3.0300177748051373e-05, "loss": 2.1578, "step": 4604 }, { "epoch": 0.041416898634276646, "grad_norm": 4.051889419555664, "learning_rate": 3.0292098041658397e-05, "loss": 1.9082, "step": 4605 }, { "epoch": 0.04142589253191709, "grad_norm": 4.854938983917236, "learning_rate": 3.0284017756529537e-05, "loss": 2.26, "step": 4606 }, { "epoch": 0.041434886429557546, "grad_norm": 2.9465341567993164, "learning_rate": 3.027593689354844e-05, "loss": 2.0914, "step": 4607 }, { "epoch": 0.04144388032719799, "grad_norm": 3.629230260848999, "learning_rate": 3.0267855453598814e-05, "loss": 2.3339, "step": 4608 }, { "epoch": 0.04145287422483845, "grad_norm": 2.8365042209625244, "learning_rate": 3.0259773437564438e-05, "loss": 2.5309, "step": 4609 }, { "epoch": 0.0414618681224789, "grad_norm": 2.9410572052001953, "learning_rate": 3.025169084632915e-05, "loss": 1.9393, "step": 4610 }, { "epoch": 0.04147086202011935, "grad_norm": 2.284728765487671, "learning_rate": 3.024360768077684e-05, "loss": 2.1085, "step": 4611 }, { "epoch": 0.0414798559177598, "grad_norm": 4.251804828643799, "learning_rate": 3.0235523941791484e-05, "loss": 2.159, "step": 4612 }, { "epoch": 0.04148884981540025, "grad_norm": 4.349142551422119, "learning_rate": 3.02274396302571e-05, "loss": 1.7102, "step": 4613 }, { "epoch": 0.0414978437130407, "grad_norm": 3.06392502784729, "learning_rate": 3.0219354747057772e-05, "loss": 2.3244, "step": 4614 }, { "epoch": 0.041506837610681156, "grad_norm": 2.1754441261291504, "learning_rate": 3.021126929307766e-05, "loss": 2.4491, "step": 4615 }, { "epoch": 0.0415158315083216, "grad_norm": 3.444899559020996, "learning_rate": 3.0203183269200975e-05, "loss": 2.1575, "step": 4616 }, { "epoch": 0.04152482540596206, "grad_norm": 9.027824401855469, "learning_rate": 3.019509667631199e-05, "loss": 2.3288, "step": 4617 }, { "epoch": 0.041533819303602504, "grad_norm": 3.2155821323394775, "learning_rate": 3.018700951529504e-05, "loss": 2.2808, "step": 4618 }, { "epoch": 0.04154281320124296, "grad_norm": 3.502128839492798, "learning_rate": 3.017892178703453e-05, "loss": 2.2705, "step": 4619 }, { "epoch": 0.041551807098883405, "grad_norm": 1.7096984386444092, "learning_rate": 3.017083349241492e-05, "loss": 1.6029, "step": 4620 }, { "epoch": 0.04156080099652386, "grad_norm": 2.8882076740264893, "learning_rate": 3.016274463232074e-05, "loss": 2.4646, "step": 4621 }, { "epoch": 0.04156979489416431, "grad_norm": 5.397252559661865, "learning_rate": 3.0154655207636556e-05, "loss": 2.0041, "step": 4622 }, { "epoch": 0.04157878879180476, "grad_norm": 2.732542037963867, "learning_rate": 3.0146565219247036e-05, "loss": 2.5927, "step": 4623 }, { "epoch": 0.04158778268944521, "grad_norm": 2.999408006668091, "learning_rate": 3.0138474668036883e-05, "loss": 2.072, "step": 4624 }, { "epoch": 0.04159677658708566, "grad_norm": 2.824723958969116, "learning_rate": 3.0130383554890856e-05, "loss": 1.9367, "step": 4625 }, { "epoch": 0.041605770484726114, "grad_norm": 3.761340379714966, "learning_rate": 3.012229188069381e-05, "loss": 1.9856, "step": 4626 }, { "epoch": 0.04161476438236657, "grad_norm": 3.7261483669281006, "learning_rate": 3.011419964633062e-05, "loss": 2.0315, "step": 4627 }, { "epoch": 0.041623758280007014, "grad_norm": 4.15814208984375, "learning_rate": 3.0106106852686238e-05, "loss": 1.9268, "step": 4628 }, { "epoch": 0.04163275217764747, "grad_norm": 2.3040733337402344, "learning_rate": 3.00980135006457e-05, "loss": 2.0448, "step": 4629 }, { "epoch": 0.041641746075287915, "grad_norm": 4.512973308563232, "learning_rate": 3.008991959109406e-05, "loss": 2.437, "step": 4630 }, { "epoch": 0.04165073997292837, "grad_norm": 3.0546152591705322, "learning_rate": 3.0081825124916474e-05, "loss": 1.6863, "step": 4631 }, { "epoch": 0.041659733870568816, "grad_norm": 3.272622585296631, "learning_rate": 3.007373010299813e-05, "loss": 1.8992, "step": 4632 }, { "epoch": 0.04166872776820927, "grad_norm": 3.579852819442749, "learning_rate": 3.006563452622429e-05, "loss": 2.1035, "step": 4633 }, { "epoch": 0.041677721665849723, "grad_norm": 1.823140263557434, "learning_rate": 3.005753839548027e-05, "loss": 2.0981, "step": 4634 }, { "epoch": 0.04168671556349017, "grad_norm": 1.834733486175537, "learning_rate": 3.004944171165146e-05, "loss": 2.475, "step": 4635 }, { "epoch": 0.041695709461130624, "grad_norm": 2.0589420795440674, "learning_rate": 3.00413444756233e-05, "loss": 2.1901, "step": 4636 }, { "epoch": 0.04170470335877107, "grad_norm": 4.364973545074463, "learning_rate": 3.003324668828128e-05, "loss": 2.2651, "step": 4637 }, { "epoch": 0.041713697256411525, "grad_norm": 2.4711084365844727, "learning_rate": 3.0025148350510974e-05, "loss": 2.2708, "step": 4638 }, { "epoch": 0.04172269115405198, "grad_norm": 3.224438190460205, "learning_rate": 3.0017049463197993e-05, "loss": 2.3837, "step": 4639 }, { "epoch": 0.041731685051692426, "grad_norm": 2.3661465644836426, "learning_rate": 3.0008950027228033e-05, "loss": 2.0433, "step": 4640 }, { "epoch": 0.04174067894933288, "grad_norm": 2.8729476928710938, "learning_rate": 3.0000850043486823e-05, "loss": 1.9874, "step": 4641 }, { "epoch": 0.041749672846973326, "grad_norm": 3.0814123153686523, "learning_rate": 2.9992749512860173e-05, "loss": 2.3057, "step": 4642 }, { "epoch": 0.04175866674461378, "grad_norm": 1.8564988374710083, "learning_rate": 2.9984648436233947e-05, "loss": 2.5215, "step": 4643 }, { "epoch": 0.04176766064225423, "grad_norm": 2.1685256958007812, "learning_rate": 2.997654681449406e-05, "loss": 1.8758, "step": 4644 }, { "epoch": 0.04177665453989468, "grad_norm": 4.700963497161865, "learning_rate": 2.9968444648526493e-05, "loss": 2.2635, "step": 4645 }, { "epoch": 0.041785648437535135, "grad_norm": 2.5776851177215576, "learning_rate": 2.9960341939217295e-05, "loss": 1.9877, "step": 4646 }, { "epoch": 0.04179464233517558, "grad_norm": 3.0803472995758057, "learning_rate": 2.9952238687452555e-05, "loss": 1.9393, "step": 4647 }, { "epoch": 0.041803636232816035, "grad_norm": 2.981942892074585, "learning_rate": 2.994413489411844e-05, "loss": 2.1755, "step": 4648 }, { "epoch": 0.04181263013045648, "grad_norm": 7.022572040557861, "learning_rate": 2.9936030560101168e-05, "loss": 1.4995, "step": 4649 }, { "epoch": 0.041821624028096936, "grad_norm": 4.759737491607666, "learning_rate": 2.9927925686287006e-05, "loss": 1.9442, "step": 4650 }, { "epoch": 0.04183061792573739, "grad_norm": 2.4352903366088867, "learning_rate": 2.991982027356231e-05, "loss": 2.3616, "step": 4651 }, { "epoch": 0.04183961182337784, "grad_norm": 3.295921802520752, "learning_rate": 2.9911714322813467e-05, "loss": 2.3844, "step": 4652 }, { "epoch": 0.04184860572101829, "grad_norm": 2.785092830657959, "learning_rate": 2.9903607834926927e-05, "loss": 2.3072, "step": 4653 }, { "epoch": 0.04185759961865874, "grad_norm": 2.5290305614471436, "learning_rate": 2.9895500810789202e-05, "loss": 1.7421, "step": 4654 }, { "epoch": 0.04186659351629919, "grad_norm": 3.1894357204437256, "learning_rate": 2.988739325128687e-05, "loss": 1.6238, "step": 4655 }, { "epoch": 0.041875587413939645, "grad_norm": 2.9073965549468994, "learning_rate": 2.987928515730657e-05, "loss": 1.8259, "step": 4656 }, { "epoch": 0.04188458131158009, "grad_norm": 3.5862085819244385, "learning_rate": 2.9871176529734968e-05, "loss": 2.5356, "step": 4657 }, { "epoch": 0.041893575209220546, "grad_norm": 1.9867210388183594, "learning_rate": 2.986306736945883e-05, "loss": 2.4379, "step": 4658 }, { "epoch": 0.04190256910686099, "grad_norm": 3.9151077270507812, "learning_rate": 2.9854957677364957e-05, "loss": 2.186, "step": 4659 }, { "epoch": 0.04191156300450145, "grad_norm": 5.0980706214904785, "learning_rate": 2.984684745434021e-05, "loss": 2.4927, "step": 4660 }, { "epoch": 0.041920556902141894, "grad_norm": 5.189688682556152, "learning_rate": 2.9838736701271514e-05, "loss": 2.2266, "step": 4661 }, { "epoch": 0.04192955079978235, "grad_norm": 5.361749172210693, "learning_rate": 2.9830625419045837e-05, "loss": 2.3416, "step": 4662 }, { "epoch": 0.0419385446974228, "grad_norm": 2.2746431827545166, "learning_rate": 2.9822513608550238e-05, "loss": 2.3301, "step": 4663 }, { "epoch": 0.04194753859506325, "grad_norm": 4.002450942993164, "learning_rate": 2.98144012706718e-05, "loss": 2.8034, "step": 4664 }, { "epoch": 0.0419565324927037, "grad_norm": 3.8837904930114746, "learning_rate": 2.9806288406297676e-05, "loss": 1.8237, "step": 4665 }, { "epoch": 0.04196552639034415, "grad_norm": 1.3276365995407104, "learning_rate": 2.9798175016315078e-05, "loss": 1.7518, "step": 4666 }, { "epoch": 0.0419745202879846, "grad_norm": 2.9217965602874756, "learning_rate": 2.9790061101611276e-05, "loss": 1.8358, "step": 4667 }, { "epoch": 0.04198351418562506, "grad_norm": 2.5761282444000244, "learning_rate": 2.9781946663073596e-05, "loss": 2.477, "step": 4668 }, { "epoch": 0.0419925080832655, "grad_norm": 2.1562230587005615, "learning_rate": 2.9773831701589412e-05, "loss": 2.5594, "step": 4669 }, { "epoch": 0.04200150198090596, "grad_norm": 3.8433148860931396, "learning_rate": 2.9765716218046175e-05, "loss": 2.4918, "step": 4670 }, { "epoch": 0.042010495878546404, "grad_norm": 1.9168798923492432, "learning_rate": 2.9757600213331378e-05, "loss": 2.7903, "step": 4671 }, { "epoch": 0.04201948977618686, "grad_norm": 2.8318514823913574, "learning_rate": 2.9749483688332574e-05, "loss": 2.1933, "step": 4672 }, { "epoch": 0.042028483673827305, "grad_norm": 3.3723320960998535, "learning_rate": 2.9741366643937367e-05, "loss": 2.0005, "step": 4673 }, { "epoch": 0.04203747757146776, "grad_norm": 2.9001760482788086, "learning_rate": 2.9733249081033442e-05, "loss": 1.924, "step": 4674 }, { "epoch": 0.04204647146910821, "grad_norm": 4.469167709350586, "learning_rate": 2.972513100050851e-05, "loss": 2.3103, "step": 4675 }, { "epoch": 0.04205546536674866, "grad_norm": 3.6509358882904053, "learning_rate": 2.9717012403250356e-05, "loss": 2.2515, "step": 4676 }, { "epoch": 0.04206445926438911, "grad_norm": 2.6967644691467285, "learning_rate": 2.9708893290146823e-05, "loss": 1.8414, "step": 4677 }, { "epoch": 0.04207345316202956, "grad_norm": 2.5611534118652344, "learning_rate": 2.970077366208579e-05, "loss": 1.7976, "step": 4678 }, { "epoch": 0.042082447059670014, "grad_norm": 2.6172947883605957, "learning_rate": 2.9692653519955223e-05, "loss": 2.2048, "step": 4679 }, { "epoch": 0.04209144095731047, "grad_norm": 2.9006028175354004, "learning_rate": 2.9684532864643122e-05, "loss": 2.1627, "step": 4680 }, { "epoch": 0.042100434854950915, "grad_norm": 2.094144344329834, "learning_rate": 2.9676411697037547e-05, "loss": 2.0482, "step": 4681 }, { "epoch": 0.04210942875259137, "grad_norm": 3.8280017375946045, "learning_rate": 2.9668290018026617e-05, "loss": 1.8157, "step": 4682 }, { "epoch": 0.042118422650231815, "grad_norm": 4.849945068359375, "learning_rate": 2.966016782849852e-05, "loss": 2.5302, "step": 4683 }, { "epoch": 0.04212741654787227, "grad_norm": 2.507969617843628, "learning_rate": 2.9652045129341467e-05, "loss": 2.2545, "step": 4684 }, { "epoch": 0.04213641044551272, "grad_norm": 1.7411049604415894, "learning_rate": 2.964392192144375e-05, "loss": 2.2525, "step": 4685 }, { "epoch": 0.04214540434315317, "grad_norm": 2.4296963214874268, "learning_rate": 2.963579820569372e-05, "loss": 2.3514, "step": 4686 }, { "epoch": 0.042154398240793624, "grad_norm": 2.8700971603393555, "learning_rate": 2.9627673982979764e-05, "loss": 1.6224, "step": 4687 }, { "epoch": 0.04216339213843407, "grad_norm": 3.9899697303771973, "learning_rate": 2.961954925419035e-05, "loss": 2.0683, "step": 4688 }, { "epoch": 0.042172386036074525, "grad_norm": 2.2432446479797363, "learning_rate": 2.961142402021396e-05, "loss": 2.2663, "step": 4689 }, { "epoch": 0.04218137993371497, "grad_norm": 2.469597101211548, "learning_rate": 2.9603298281939178e-05, "loss": 1.7078, "step": 4690 }, { "epoch": 0.042190373831355425, "grad_norm": 3.2500271797180176, "learning_rate": 2.9595172040254625e-05, "loss": 2.0026, "step": 4691 }, { "epoch": 0.04219936772899588, "grad_norm": 8.856544494628906, "learning_rate": 2.9587045296048955e-05, "loss": 2.2176, "step": 4692 }, { "epoch": 0.042208361626636326, "grad_norm": 2.759798526763916, "learning_rate": 2.9578918050210912e-05, "loss": 2.443, "step": 4693 }, { "epoch": 0.04221735552427678, "grad_norm": 1.6283466815948486, "learning_rate": 2.957079030362928e-05, "loss": 1.5085, "step": 4694 }, { "epoch": 0.04222634942191723, "grad_norm": 1.9393969774246216, "learning_rate": 2.956266205719288e-05, "loss": 2.0155, "step": 4695 }, { "epoch": 0.04223534331955768, "grad_norm": 3.0805041790008545, "learning_rate": 2.9554533311790632e-05, "loss": 2.2241, "step": 4696 }, { "epoch": 0.042244337217198134, "grad_norm": 2.572622060775757, "learning_rate": 2.9546404068311462e-05, "loss": 2.2613, "step": 4697 }, { "epoch": 0.04225333111483858, "grad_norm": 3.9190242290496826, "learning_rate": 2.953827432764438e-05, "loss": 2.3067, "step": 4698 }, { "epoch": 0.042262325012479035, "grad_norm": 3.6484084129333496, "learning_rate": 2.9530144090678435e-05, "loss": 1.9757, "step": 4699 }, { "epoch": 0.04227131891011948, "grad_norm": 4.243458271026611, "learning_rate": 2.952201335830275e-05, "loss": 2.3205, "step": 4700 }, { "epoch": 0.042280312807759936, "grad_norm": 2.916868209838867, "learning_rate": 2.951388213140648e-05, "loss": 2.0359, "step": 4701 }, { "epoch": 0.04228930670540038, "grad_norm": 3.8851306438446045, "learning_rate": 2.9505750410878845e-05, "loss": 1.683, "step": 4702 }, { "epoch": 0.042298300603040836, "grad_norm": 5.658038139343262, "learning_rate": 2.9497618197609124e-05, "loss": 1.6932, "step": 4703 }, { "epoch": 0.04230729450068129, "grad_norm": 2.666339874267578, "learning_rate": 2.948948549248664e-05, "loss": 2.2468, "step": 4704 }, { "epoch": 0.04231628839832174, "grad_norm": 2.48435115814209, "learning_rate": 2.9481352296400766e-05, "loss": 2.0405, "step": 4705 }, { "epoch": 0.04232528229596219, "grad_norm": 2.1106972694396973, "learning_rate": 2.9473218610240943e-05, "loss": 1.7398, "step": 4706 }, { "epoch": 0.04233427619360264, "grad_norm": 4.00709342956543, "learning_rate": 2.9465084434896663e-05, "loss": 2.038, "step": 4707 }, { "epoch": 0.04234327009124309, "grad_norm": 5.440618991851807, "learning_rate": 2.9456949771257458e-05, "loss": 1.7438, "step": 4708 }, { "epoch": 0.042352263988883546, "grad_norm": 1.8420073986053467, "learning_rate": 2.9448814620212926e-05, "loss": 2.1546, "step": 4709 }, { "epoch": 0.04236125788652399, "grad_norm": 2.8794331550598145, "learning_rate": 2.944067898265272e-05, "loss": 2.4101, "step": 4710 }, { "epoch": 0.042370251784164446, "grad_norm": 5.032707691192627, "learning_rate": 2.943254285946654e-05, "loss": 2.046, "step": 4711 }, { "epoch": 0.04237924568180489, "grad_norm": 1.9895223379135132, "learning_rate": 2.9424406251544122e-05, "loss": 2.4914, "step": 4712 }, { "epoch": 0.04238823957944535, "grad_norm": 2.8098812103271484, "learning_rate": 2.9416269159775307e-05, "loss": 2.08, "step": 4713 }, { "epoch": 0.042397233477085794, "grad_norm": 3.575038433074951, "learning_rate": 2.940813158504992e-05, "loss": 2.0527, "step": 4714 }, { "epoch": 0.04240622737472625, "grad_norm": 2.8159661293029785, "learning_rate": 2.9399993528257902e-05, "loss": 2.3937, "step": 4715 }, { "epoch": 0.0424152212723667, "grad_norm": 5.621457576751709, "learning_rate": 2.9391854990289207e-05, "loss": 1.9929, "step": 4716 }, { "epoch": 0.04242421517000715, "grad_norm": 6.6420745849609375, "learning_rate": 2.9383715972033844e-05, "loss": 2.1421, "step": 4717 }, { "epoch": 0.0424332090676476, "grad_norm": 4.183467388153076, "learning_rate": 2.9375576474381905e-05, "loss": 2.0254, "step": 4718 }, { "epoch": 0.04244220296528805, "grad_norm": 3.2745654582977295, "learning_rate": 2.9367436498223493e-05, "loss": 2.3063, "step": 4719 }, { "epoch": 0.0424511968629285, "grad_norm": 7.650079250335693, "learning_rate": 2.9359296044448794e-05, "loss": 1.9355, "step": 4720 }, { "epoch": 0.04246019076056896, "grad_norm": 2.402789354324341, "learning_rate": 2.9351155113948037e-05, "loss": 2.2443, "step": 4721 }, { "epoch": 0.042469184658209404, "grad_norm": 2.846691608428955, "learning_rate": 2.9343013707611493e-05, "loss": 2.0553, "step": 4722 }, { "epoch": 0.04247817855584986, "grad_norm": 4.064048767089844, "learning_rate": 2.9334871826329507e-05, "loss": 2.0402, "step": 4723 }, { "epoch": 0.042487172453490304, "grad_norm": 9.155508041381836, "learning_rate": 2.9326729470992448e-05, "loss": 2.1514, "step": 4724 }, { "epoch": 0.04249616635113076, "grad_norm": 11.149697303771973, "learning_rate": 2.9318586642490763e-05, "loss": 2.5623, "step": 4725 }, { "epoch": 0.04250516024877121, "grad_norm": 1.8705065250396729, "learning_rate": 2.931044334171493e-05, "loss": 2.1263, "step": 4726 }, { "epoch": 0.04251415414641166, "grad_norm": 1.980458378791809, "learning_rate": 2.9302299569555504e-05, "loss": 2.0929, "step": 4727 }, { "epoch": 0.04252314804405211, "grad_norm": 3.6289267539978027, "learning_rate": 2.929415532690306e-05, "loss": 2.258, "step": 4728 }, { "epoch": 0.04253214194169256, "grad_norm": 3.5921785831451416, "learning_rate": 2.9286010614648246e-05, "loss": 1.8593, "step": 4729 }, { "epoch": 0.042541135839333014, "grad_norm": 2.246903657913208, "learning_rate": 2.927786543368175e-05, "loss": 2.1326, "step": 4730 }, { "epoch": 0.04255012973697346, "grad_norm": 2.507826328277588, "learning_rate": 2.926971978489433e-05, "loss": 2.1372, "step": 4731 }, { "epoch": 0.042559123634613914, "grad_norm": 3.453455686569214, "learning_rate": 2.926157366917677e-05, "loss": 2.186, "step": 4732 }, { "epoch": 0.04256811753225437, "grad_norm": 4.487183094024658, "learning_rate": 2.9253427087419916e-05, "loss": 2.4503, "step": 4733 }, { "epoch": 0.042577111429894815, "grad_norm": 4.296563148498535, "learning_rate": 2.9245280040514678e-05, "loss": 1.5416, "step": 4734 }, { "epoch": 0.04258610532753527, "grad_norm": 4.994293689727783, "learning_rate": 2.9237132529351996e-05, "loss": 2.0226, "step": 4735 }, { "epoch": 0.042595099225175716, "grad_norm": 5.011751651763916, "learning_rate": 2.9228984554822868e-05, "loss": 1.7157, "step": 4736 }, { "epoch": 0.04260409312281617, "grad_norm": 3.4111361503601074, "learning_rate": 2.9220836117818344e-05, "loss": 2.2178, "step": 4737 }, { "epoch": 0.04261308702045662, "grad_norm": 3.0871975421905518, "learning_rate": 2.921268721922954e-05, "loss": 1.5758, "step": 4738 }, { "epoch": 0.04262208091809707, "grad_norm": 4.146090507507324, "learning_rate": 2.920453785994759e-05, "loss": 2.5127, "step": 4739 }, { "epoch": 0.042631074815737524, "grad_norm": 4.619262218475342, "learning_rate": 2.9196388040863693e-05, "loss": 1.802, "step": 4740 }, { "epoch": 0.04264006871337797, "grad_norm": 3.617544174194336, "learning_rate": 2.918823776286912e-05, "loss": 2.6367, "step": 4741 }, { "epoch": 0.042649062611018425, "grad_norm": 4.110443115234375, "learning_rate": 2.918008702685516e-05, "loss": 1.5935, "step": 4742 }, { "epoch": 0.04265805650865887, "grad_norm": 3.97080659866333, "learning_rate": 2.917193583371316e-05, "loss": 1.8049, "step": 4743 }, { "epoch": 0.042667050406299326, "grad_norm": 2.2780838012695312, "learning_rate": 2.9163784184334535e-05, "loss": 2.0668, "step": 4744 }, { "epoch": 0.04267604430393978, "grad_norm": 5.620966911315918, "learning_rate": 2.915563207961074e-05, "loss": 2.124, "step": 4745 }, { "epoch": 0.042685038201580226, "grad_norm": 2.6739625930786133, "learning_rate": 2.9147479520433253e-05, "loss": 2.3456, "step": 4746 }, { "epoch": 0.04269403209922068, "grad_norm": 4.014924049377441, "learning_rate": 2.9139326507693654e-05, "loss": 2.5724, "step": 4747 }, { "epoch": 0.04270302599686113, "grad_norm": 5.860874652862549, "learning_rate": 2.9131173042283528e-05, "loss": 2.1287, "step": 4748 }, { "epoch": 0.04271201989450158, "grad_norm": 3.568667411804199, "learning_rate": 2.912301912509452e-05, "loss": 2.005, "step": 4749 }, { "epoch": 0.042721013792142035, "grad_norm": 2.2223074436187744, "learning_rate": 2.9114864757018352e-05, "loss": 2.4175, "step": 4750 }, { "epoch": 0.04273000768978248, "grad_norm": 4.047830581665039, "learning_rate": 2.9106709938946757e-05, "loss": 2.3517, "step": 4751 }, { "epoch": 0.042739001587422935, "grad_norm": 1.9510031938552856, "learning_rate": 2.9098554671771534e-05, "loss": 2.1254, "step": 4752 }, { "epoch": 0.04274799548506338, "grad_norm": 3.4884912967681885, "learning_rate": 2.9090398956384535e-05, "loss": 1.9233, "step": 4753 }, { "epoch": 0.042756989382703836, "grad_norm": 2.2165746688842773, "learning_rate": 2.908224279367766e-05, "loss": 1.9944, "step": 4754 }, { "epoch": 0.04276598328034428, "grad_norm": 3.159351110458374, "learning_rate": 2.9074086184542843e-05, "loss": 2.34, "step": 4755 }, { "epoch": 0.04277497717798474, "grad_norm": 3.0804431438446045, "learning_rate": 2.9065929129872094e-05, "loss": 2.4745, "step": 4756 }, { "epoch": 0.04278397107562519, "grad_norm": 4.14511775970459, "learning_rate": 2.9057771630557444e-05, "loss": 2.1904, "step": 4757 }, { "epoch": 0.04279296497326564, "grad_norm": 2.642125129699707, "learning_rate": 2.904961368749099e-05, "loss": 2.4292, "step": 4758 }, { "epoch": 0.04280195887090609, "grad_norm": 4.150062561035156, "learning_rate": 2.9041455301564874e-05, "loss": 2.0509, "step": 4759 }, { "epoch": 0.04281095276854654, "grad_norm": 3.4167580604553223, "learning_rate": 2.9033296473671278e-05, "loss": 1.5894, "step": 4760 }, { "epoch": 0.04281994666618699, "grad_norm": 3.3076388835906982, "learning_rate": 2.9025137204702448e-05, "loss": 2.2169, "step": 4761 }, { "epoch": 0.042828940563827446, "grad_norm": 5.186434745788574, "learning_rate": 2.9016977495550657e-05, "loss": 2.1816, "step": 4762 }, { "epoch": 0.04283793446146789, "grad_norm": 3.0735435485839844, "learning_rate": 2.900881734710826e-05, "loss": 2.01, "step": 4763 }, { "epoch": 0.04284692835910835, "grad_norm": 8.321554183959961, "learning_rate": 2.900065676026762e-05, "loss": 2.1548, "step": 4764 }, { "epoch": 0.042855922256748794, "grad_norm": 3.9182732105255127, "learning_rate": 2.8992495735921165e-05, "loss": 1.926, "step": 4765 }, { "epoch": 0.04286491615438925, "grad_norm": 3.1436195373535156, "learning_rate": 2.898433427496138e-05, "loss": 1.6696, "step": 4766 }, { "epoch": 0.0428739100520297, "grad_norm": 4.488739967346191, "learning_rate": 2.8976172378280797e-05, "loss": 2.2499, "step": 4767 }, { "epoch": 0.04288290394967015, "grad_norm": 4.233646392822266, "learning_rate": 2.8968010046771975e-05, "loss": 2.1763, "step": 4768 }, { "epoch": 0.0428918978473106, "grad_norm": 2.148001194000244, "learning_rate": 2.8959847281327542e-05, "loss": 2.244, "step": 4769 }, { "epoch": 0.04290089174495105, "grad_norm": 3.019772529602051, "learning_rate": 2.8951684082840163e-05, "loss": 2.4855, "step": 4770 }, { "epoch": 0.0429098856425915, "grad_norm": 2.9217464923858643, "learning_rate": 2.8943520452202555e-05, "loss": 2.1622, "step": 4771 }, { "epoch": 0.04291887954023195, "grad_norm": 2.5013680458068848, "learning_rate": 2.8935356390307477e-05, "loss": 1.9455, "step": 4772 }, { "epoch": 0.0429278734378724, "grad_norm": 3.360123872756958, "learning_rate": 2.8927191898047745e-05, "loss": 2.2764, "step": 4773 }, { "epoch": 0.04293686733551286, "grad_norm": 2.849170684814453, "learning_rate": 2.891902697631621e-05, "loss": 2.2106, "step": 4774 }, { "epoch": 0.042945861233153304, "grad_norm": 3.233994483947754, "learning_rate": 2.8910861626005776e-05, "loss": 1.8598, "step": 4775 }, { "epoch": 0.04295485513079376, "grad_norm": 3.193922281265259, "learning_rate": 2.8902695848009397e-05, "loss": 2.4407, "step": 4776 }, { "epoch": 0.042963849028434205, "grad_norm": 2.5760834217071533, "learning_rate": 2.8894529643220064e-05, "loss": 2.6477, "step": 4777 }, { "epoch": 0.04297284292607466, "grad_norm": 3.278284788131714, "learning_rate": 2.8886363012530826e-05, "loss": 2.3219, "step": 4778 }, { "epoch": 0.04298183682371511, "grad_norm": 2.0642175674438477, "learning_rate": 2.887819595683477e-05, "loss": 1.6783, "step": 4779 }, { "epoch": 0.04299083072135556, "grad_norm": 3.1330041885375977, "learning_rate": 2.8870028477025042e-05, "loss": 2.3101, "step": 4780 }, { "epoch": 0.04299982461899601, "grad_norm": 2.919257879257202, "learning_rate": 2.886186057399481e-05, "loss": 2.2989, "step": 4781 }, { "epoch": 0.04300881851663646, "grad_norm": 3.4763259887695312, "learning_rate": 2.8853692248637315e-05, "loss": 2.4103, "step": 4782 }, { "epoch": 0.043017812414276914, "grad_norm": 2.551605701446533, "learning_rate": 2.884552350184582e-05, "loss": 1.8246, "step": 4783 }, { "epoch": 0.04302680631191736, "grad_norm": 1.9006192684173584, "learning_rate": 2.883735433451367e-05, "loss": 1.5272, "step": 4784 }, { "epoch": 0.043035800209557815, "grad_norm": 3.8372483253479004, "learning_rate": 2.8829184747534214e-05, "loss": 2.2556, "step": 4785 }, { "epoch": 0.04304479410719827, "grad_norm": 3.2144157886505127, "learning_rate": 2.8821014741800873e-05, "loss": 2.1941, "step": 4786 }, { "epoch": 0.043053788004838715, "grad_norm": 3.3725547790527344, "learning_rate": 2.88128443182071e-05, "loss": 1.6919, "step": 4787 }, { "epoch": 0.04306278190247917, "grad_norm": 2.1169590950012207, "learning_rate": 2.880467347764641e-05, "loss": 2.4548, "step": 4788 }, { "epoch": 0.043071775800119616, "grad_norm": 2.614859104156494, "learning_rate": 2.879650222101235e-05, "loss": 1.798, "step": 4789 }, { "epoch": 0.04308076969776007, "grad_norm": 3.285728693008423, "learning_rate": 2.878833054919851e-05, "loss": 1.9271, "step": 4790 }, { "epoch": 0.043089763595400524, "grad_norm": 2.196812868118286, "learning_rate": 2.8780158463098538e-05, "loss": 2.1746, "step": 4791 }, { "epoch": 0.04309875749304097, "grad_norm": 3.577982187271118, "learning_rate": 2.8771985963606128e-05, "loss": 1.3994, "step": 4792 }, { "epoch": 0.043107751390681424, "grad_norm": 1.6286731958389282, "learning_rate": 2.8763813051614996e-05, "loss": 2.4786, "step": 4793 }, { "epoch": 0.04311674528832187, "grad_norm": 1.8249706029891968, "learning_rate": 2.875563972801893e-05, "loss": 2.2949, "step": 4794 }, { "epoch": 0.043125739185962325, "grad_norm": 8.87211799621582, "learning_rate": 2.874746599371175e-05, "loss": 2.3133, "step": 4795 }, { "epoch": 0.04313473308360278, "grad_norm": 3.996387481689453, "learning_rate": 2.8739291849587324e-05, "loss": 2.1586, "step": 4796 }, { "epoch": 0.043143726981243226, "grad_norm": 1.6374413967132568, "learning_rate": 2.8731117296539557e-05, "loss": 1.6262, "step": 4797 }, { "epoch": 0.04315272087888368, "grad_norm": 2.7373743057250977, "learning_rate": 2.8722942335462415e-05, "loss": 2.2597, "step": 4798 }, { "epoch": 0.04316171477652413, "grad_norm": 2.316030263900757, "learning_rate": 2.87147669672499e-05, "loss": 2.2141, "step": 4799 }, { "epoch": 0.04317070867416458, "grad_norm": 6.678574562072754, "learning_rate": 2.870659119279605e-05, "loss": 2.1043, "step": 4800 }, { "epoch": 0.04317970257180503, "grad_norm": 2.726940155029297, "learning_rate": 2.8698415012994967e-05, "loss": 1.8794, "step": 4801 }, { "epoch": 0.04318869646944548, "grad_norm": 3.7419471740722656, "learning_rate": 2.8690238428740772e-05, "loss": 2.1206, "step": 4802 }, { "epoch": 0.043197690367085935, "grad_norm": 4.588747501373291, "learning_rate": 2.868206144092765e-05, "loss": 1.8895, "step": 4803 }, { "epoch": 0.04320668426472638, "grad_norm": 3.88322377204895, "learning_rate": 2.8673884050449827e-05, "loss": 2.2909, "step": 4804 }, { "epoch": 0.043215678162366836, "grad_norm": 2.196803331375122, "learning_rate": 2.8665706258201574e-05, "loss": 2.0767, "step": 4805 }, { "epoch": 0.04322467206000728, "grad_norm": 3.2628333568573, "learning_rate": 2.8657528065077187e-05, "loss": 2.1376, "step": 4806 }, { "epoch": 0.043233665957647736, "grad_norm": 3.2791621685028076, "learning_rate": 2.8649349471971025e-05, "loss": 1.6931, "step": 4807 }, { "epoch": 0.04324265985528819, "grad_norm": 2.3719143867492676, "learning_rate": 2.86411704797775e-05, "loss": 2.3658, "step": 4808 }, { "epoch": 0.04325165375292864, "grad_norm": 2.449040412902832, "learning_rate": 2.8632991089391053e-05, "loss": 1.9908, "step": 4809 }, { "epoch": 0.04326064765056909, "grad_norm": 5.093130588531494, "learning_rate": 2.862481130170615e-05, "loss": 2.4187, "step": 4810 }, { "epoch": 0.04326964154820954, "grad_norm": 3.6478939056396484, "learning_rate": 2.8616631117617336e-05, "loss": 1.7209, "step": 4811 }, { "epoch": 0.04327863544584999, "grad_norm": 3.0439937114715576, "learning_rate": 2.860845053801919e-05, "loss": 2.1629, "step": 4812 }, { "epoch": 0.04328762934349044, "grad_norm": 4.163026332855225, "learning_rate": 2.8600269563806302e-05, "loss": 2.7448, "step": 4813 }, { "epoch": 0.04329662324113089, "grad_norm": 4.233515739440918, "learning_rate": 2.859208819587336e-05, "loss": 2.185, "step": 4814 }, { "epoch": 0.043305617138771346, "grad_norm": 2.311962366104126, "learning_rate": 2.8583906435115047e-05, "loss": 2.3008, "step": 4815 }, { "epoch": 0.04331461103641179, "grad_norm": 5.884310245513916, "learning_rate": 2.8575724282426118e-05, "loss": 2.551, "step": 4816 }, { "epoch": 0.04332360493405225, "grad_norm": 3.0550756454467773, "learning_rate": 2.856754173870136e-05, "loss": 2.2043, "step": 4817 }, { "epoch": 0.043332598831692694, "grad_norm": 3.16469669342041, "learning_rate": 2.8559358804835597e-05, "loss": 2.1844, "step": 4818 }, { "epoch": 0.04334159272933315, "grad_norm": 2.8099453449249268, "learning_rate": 2.85511754817237e-05, "loss": 2.0613, "step": 4819 }, { "epoch": 0.0433505866269736, "grad_norm": 3.5853559970855713, "learning_rate": 2.8542991770260608e-05, "loss": 1.718, "step": 4820 }, { "epoch": 0.04335958052461405, "grad_norm": 2.242112159729004, "learning_rate": 2.853480767134125e-05, "loss": 2.22, "step": 4821 }, { "epoch": 0.0433685744222545, "grad_norm": 2.1495604515075684, "learning_rate": 2.852662318586064e-05, "loss": 2.1922, "step": 4822 }, { "epoch": 0.04337756831989495, "grad_norm": 4.385153770446777, "learning_rate": 2.8518438314713823e-05, "loss": 1.9758, "step": 4823 }, { "epoch": 0.0433865622175354, "grad_norm": 2.472198247909546, "learning_rate": 2.851025305879589e-05, "loss": 1.5633, "step": 4824 }, { "epoch": 0.04339555611517585, "grad_norm": 3.204946756362915, "learning_rate": 2.850206741900195e-05, "loss": 1.8448, "step": 4825 }, { "epoch": 0.043404550012816304, "grad_norm": 2.3729214668273926, "learning_rate": 2.8493881396227186e-05, "loss": 2.2684, "step": 4826 }, { "epoch": 0.04341354391045676, "grad_norm": 2.533449649810791, "learning_rate": 2.8485694991366807e-05, "loss": 2.421, "step": 4827 }, { "epoch": 0.043422537808097204, "grad_norm": 4.9456915855407715, "learning_rate": 2.8477508205316055e-05, "loss": 2.3642, "step": 4828 }, { "epoch": 0.04343153170573766, "grad_norm": 3.4706690311431885, "learning_rate": 2.8469321038970244e-05, "loss": 1.8863, "step": 4829 }, { "epoch": 0.043440525603378105, "grad_norm": 2.683340311050415, "learning_rate": 2.846113349322469e-05, "loss": 2.0937, "step": 4830 }, { "epoch": 0.04344951950101856, "grad_norm": 4.549906253814697, "learning_rate": 2.8452945568974792e-05, "loss": 1.7448, "step": 4831 }, { "epoch": 0.04345851339865901, "grad_norm": 3.226957321166992, "learning_rate": 2.844475726711595e-05, "loss": 2.2335, "step": 4832 }, { "epoch": 0.04346750729629946, "grad_norm": 6.152693271636963, "learning_rate": 2.843656858854364e-05, "loss": 2.0388, "step": 4833 }, { "epoch": 0.043476501193939913, "grad_norm": 3.3068082332611084, "learning_rate": 2.8428379534153355e-05, "loss": 2.1753, "step": 4834 }, { "epoch": 0.04348549509158036, "grad_norm": 2.976961851119995, "learning_rate": 2.8420190104840628e-05, "loss": 2.2082, "step": 4835 }, { "epoch": 0.043494488989220814, "grad_norm": 4.439417362213135, "learning_rate": 2.8412000301501064e-05, "loss": 1.8294, "step": 4836 }, { "epoch": 0.04350348288686127, "grad_norm": 3.1521847248077393, "learning_rate": 2.8403810125030277e-05, "loss": 2.2153, "step": 4837 }, { "epoch": 0.043512476784501715, "grad_norm": 3.8872299194335938, "learning_rate": 2.839561957632393e-05, "loss": 2.0796, "step": 4838 }, { "epoch": 0.04352147068214217, "grad_norm": 1.9696996212005615, "learning_rate": 2.8387428656277724e-05, "loss": 2.0274, "step": 4839 }, { "epoch": 0.043530464579782616, "grad_norm": 2.8557231426239014, "learning_rate": 2.8379237365787426e-05, "loss": 2.1683, "step": 4840 }, { "epoch": 0.04353945847742307, "grad_norm": 2.4449357986450195, "learning_rate": 2.83710457057488e-05, "loss": 2.5292, "step": 4841 }, { "epoch": 0.043548452375063516, "grad_norm": 2.6257379055023193, "learning_rate": 2.8362853677057688e-05, "loss": 2.5779, "step": 4842 }, { "epoch": 0.04355744627270397, "grad_norm": 4.769418239593506, "learning_rate": 2.8354661280609955e-05, "loss": 2.1045, "step": 4843 }, { "epoch": 0.043566440170344424, "grad_norm": 5.471673011779785, "learning_rate": 2.834646851730151e-05, "loss": 2.5172, "step": 4844 }, { "epoch": 0.04357543406798487, "grad_norm": 2.9491236209869385, "learning_rate": 2.8338275388028295e-05, "loss": 2.4963, "step": 4845 }, { "epoch": 0.043584427965625325, "grad_norm": 3.37188458442688, "learning_rate": 2.8330081893686305e-05, "loss": 2.3455, "step": 4846 }, { "epoch": 0.04359342186326577, "grad_norm": 4.480508327484131, "learning_rate": 2.8321888035171567e-05, "loss": 2.077, "step": 4847 }, { "epoch": 0.043602415760906225, "grad_norm": 7.292489051818848, "learning_rate": 2.8313693813380142e-05, "loss": 2.4797, "step": 4848 }, { "epoch": 0.04361140965854668, "grad_norm": 4.422329425811768, "learning_rate": 2.8305499229208148e-05, "loss": 1.6011, "step": 4849 }, { "epoch": 0.043620403556187126, "grad_norm": 1.8012765645980835, "learning_rate": 2.8297304283551728e-05, "loss": 1.4365, "step": 4850 }, { "epoch": 0.04362939745382758, "grad_norm": 5.455534934997559, "learning_rate": 2.8289108977307067e-05, "loss": 2.3007, "step": 4851 }, { "epoch": 0.04363839135146803, "grad_norm": 4.829240322113037, "learning_rate": 2.8280913311370393e-05, "loss": 2.3355, "step": 4852 }, { "epoch": 0.04364738524910848, "grad_norm": 3.628803253173828, "learning_rate": 2.8272717286637973e-05, "loss": 2.1176, "step": 4853 }, { "epoch": 0.04365637914674893, "grad_norm": 2.37267804145813, "learning_rate": 2.8264520904006113e-05, "loss": 2.0671, "step": 4854 }, { "epoch": 0.04366537304438938, "grad_norm": 2.5092878341674805, "learning_rate": 2.825632416437115e-05, "loss": 2.012, "step": 4855 }, { "epoch": 0.043674366942029835, "grad_norm": 3.665663719177246, "learning_rate": 2.8248127068629478e-05, "loss": 1.9748, "step": 4856 }, { "epoch": 0.04368336083967028, "grad_norm": 3.6971778869628906, "learning_rate": 2.8239929617677506e-05, "loss": 2.3537, "step": 4857 }, { "epoch": 0.043692354737310736, "grad_norm": 2.8784544467926025, "learning_rate": 2.823173181241171e-05, "loss": 2.0624, "step": 4858 }, { "epoch": 0.04370134863495118, "grad_norm": 2.2566077709198, "learning_rate": 2.822353365372858e-05, "loss": 2.1585, "step": 4859 }, { "epoch": 0.04371034253259164, "grad_norm": 4.368597030639648, "learning_rate": 2.8215335142524657e-05, "loss": 1.8111, "step": 4860 }, { "epoch": 0.04371933643023209, "grad_norm": 4.796932697296143, "learning_rate": 2.820713627969651e-05, "loss": 2.2668, "step": 4861 }, { "epoch": 0.04372833032787254, "grad_norm": 3.469834804534912, "learning_rate": 2.8198937066140767e-05, "loss": 2.3929, "step": 4862 }, { "epoch": 0.04373732422551299, "grad_norm": 4.307190418243408, "learning_rate": 2.8190737502754077e-05, "loss": 1.9672, "step": 4863 }, { "epoch": 0.04374631812315344, "grad_norm": 2.1231908798217773, "learning_rate": 2.818253759043313e-05, "loss": 1.999, "step": 4864 }, { "epoch": 0.04375531202079389, "grad_norm": 2.04396653175354, "learning_rate": 2.817433733007466e-05, "loss": 2.1255, "step": 4865 }, { "epoch": 0.04376430591843434, "grad_norm": 2.8804259300231934, "learning_rate": 2.8166136722575425e-05, "loss": 2.2028, "step": 4866 }, { "epoch": 0.04377329981607479, "grad_norm": 2.7451181411743164, "learning_rate": 2.815793576883224e-05, "loss": 2.0164, "step": 4867 }, { "epoch": 0.043782293713715247, "grad_norm": 2.673081874847412, "learning_rate": 2.814973446974195e-05, "loss": 1.9604, "step": 4868 }, { "epoch": 0.04379128761135569, "grad_norm": 3.123722791671753, "learning_rate": 2.8141532826201435e-05, "loss": 1.8664, "step": 4869 }, { "epoch": 0.04380028150899615, "grad_norm": 3.0600345134735107, "learning_rate": 2.8133330839107608e-05, "loss": 2.4025, "step": 4870 }, { "epoch": 0.043809275406636594, "grad_norm": 3.0125603675842285, "learning_rate": 2.8125128509357435e-05, "loss": 2.041, "step": 4871 }, { "epoch": 0.04381826930427705, "grad_norm": 3.363447666168213, "learning_rate": 2.811692583784791e-05, "loss": 2.1365, "step": 4872 }, { "epoch": 0.0438272632019175, "grad_norm": 1.9875637292861938, "learning_rate": 2.810872282547605e-05, "loss": 2.4293, "step": 4873 }, { "epoch": 0.04383625709955795, "grad_norm": 2.4021356105804443, "learning_rate": 2.810051947313895e-05, "loss": 2.037, "step": 4874 }, { "epoch": 0.0438452509971984, "grad_norm": 4.595466613769531, "learning_rate": 2.8092315781733696e-05, "loss": 2.7479, "step": 4875 }, { "epoch": 0.04385424489483885, "grad_norm": 4.087653636932373, "learning_rate": 2.8084111752157432e-05, "loss": 2.0722, "step": 4876 }, { "epoch": 0.0438632387924793, "grad_norm": 1.9349615573883057, "learning_rate": 2.8075907385307355e-05, "loss": 2.4148, "step": 4877 }, { "epoch": 0.04387223269011976, "grad_norm": 2.29849910736084, "learning_rate": 2.806770268208066e-05, "loss": 2.2389, "step": 4878 }, { "epoch": 0.043881226587760204, "grad_norm": 5.000744819641113, "learning_rate": 2.8059497643374615e-05, "loss": 2.7411, "step": 4879 }, { "epoch": 0.04389022048540066, "grad_norm": 2.6417675018310547, "learning_rate": 2.8051292270086503e-05, "loss": 2.8074, "step": 4880 }, { "epoch": 0.043899214383041105, "grad_norm": 5.537990570068359, "learning_rate": 2.8043086563113667e-05, "loss": 2.4077, "step": 4881 }, { "epoch": 0.04390820828068156, "grad_norm": 3.1497104167938232, "learning_rate": 2.803488052335346e-05, "loss": 2.2724, "step": 4882 }, { "epoch": 0.043917202178322005, "grad_norm": 6.747270584106445, "learning_rate": 2.8026674151703268e-05, "loss": 2.427, "step": 4883 }, { "epoch": 0.04392619607596246, "grad_norm": 3.7069876194000244, "learning_rate": 2.8018467449060554e-05, "loss": 2.5793, "step": 4884 }, { "epoch": 0.04393518997360291, "grad_norm": 2.932264566421509, "learning_rate": 2.8010260416322774e-05, "loss": 2.4698, "step": 4885 }, { "epoch": 0.04394418387124336, "grad_norm": 2.2557613849639893, "learning_rate": 2.800205305438744e-05, "loss": 2.2086, "step": 4886 }, { "epoch": 0.043953177768883814, "grad_norm": 3.4581384658813477, "learning_rate": 2.7993845364152098e-05, "loss": 2.395, "step": 4887 }, { "epoch": 0.04396217166652426, "grad_norm": 8.201828002929688, "learning_rate": 2.7985637346514328e-05, "loss": 2.2125, "step": 4888 }, { "epoch": 0.043971165564164715, "grad_norm": 6.994072914123535, "learning_rate": 2.7977429002371747e-05, "loss": 2.202, "step": 4889 }, { "epoch": 0.04398015946180517, "grad_norm": 3.8403024673461914, "learning_rate": 2.7969220332622003e-05, "loss": 1.7855, "step": 4890 }, { "epoch": 0.043989153359445615, "grad_norm": 3.351858377456665, "learning_rate": 2.7961011338162795e-05, "loss": 2.1601, "step": 4891 }, { "epoch": 0.04399814725708607, "grad_norm": 4.057693958282471, "learning_rate": 2.7952802019891833e-05, "loss": 2.2925, "step": 4892 }, { "epoch": 0.044007141154726516, "grad_norm": 3.674988269805908, "learning_rate": 2.7944592378706884e-05, "loss": 2.3727, "step": 4893 }, { "epoch": 0.04401613505236697, "grad_norm": 1.9221099615097046, "learning_rate": 2.793638241550574e-05, "loss": 1.7044, "step": 4894 }, { "epoch": 0.04402512895000742, "grad_norm": 2.766446113586426, "learning_rate": 2.792817213118623e-05, "loss": 1.9588, "step": 4895 }, { "epoch": 0.04403412284764787, "grad_norm": 4.458810806274414, "learning_rate": 2.7919961526646215e-05, "loss": 2.2019, "step": 4896 }, { "epoch": 0.044043116745288324, "grad_norm": 2.051345109939575, "learning_rate": 2.7911750602783603e-05, "loss": 2.3656, "step": 4897 }, { "epoch": 0.04405211064292877, "grad_norm": 5.325125694274902, "learning_rate": 2.7903539360496323e-05, "loss": 2.2508, "step": 4898 }, { "epoch": 0.044061104540569225, "grad_norm": 10.051427841186523, "learning_rate": 2.7895327800682336e-05, "loss": 1.7254, "step": 4899 }, { "epoch": 0.04407009843820967, "grad_norm": 5.819668769836426, "learning_rate": 2.788711592423966e-05, "loss": 2.3046, "step": 4900 }, { "epoch": 0.044079092335850126, "grad_norm": 4.151956558227539, "learning_rate": 2.787890373206632e-05, "loss": 2.2546, "step": 4901 }, { "epoch": 0.04408808623349058, "grad_norm": 2.131981134414673, "learning_rate": 2.7870691225060404e-05, "loss": 2.2048, "step": 4902 }, { "epoch": 0.044097080131131026, "grad_norm": 2.444632053375244, "learning_rate": 2.7862478404120008e-05, "loss": 2.1453, "step": 4903 }, { "epoch": 0.04410607402877148, "grad_norm": 3.433208465576172, "learning_rate": 2.7854265270143275e-05, "loss": 2.2751, "step": 4904 }, { "epoch": 0.04411506792641193, "grad_norm": 2.3423242568969727, "learning_rate": 2.7846051824028386e-05, "loss": 2.3844, "step": 4905 }, { "epoch": 0.04412406182405238, "grad_norm": 4.322528839111328, "learning_rate": 2.7837838066673548e-05, "loss": 1.9735, "step": 4906 }, { "epoch": 0.044133055721692835, "grad_norm": 2.7726829051971436, "learning_rate": 2.7829623998977012e-05, "loss": 2.2761, "step": 4907 }, { "epoch": 0.04414204961933328, "grad_norm": 10.344087600708008, "learning_rate": 2.782140962183704e-05, "loss": 2.4031, "step": 4908 }, { "epoch": 0.044151043516973736, "grad_norm": 2.3913183212280273, "learning_rate": 2.781319493615196e-05, "loss": 2.5621, "step": 4909 }, { "epoch": 0.04416003741461418, "grad_norm": 3.6127636432647705, "learning_rate": 2.7804979942820113e-05, "loss": 2.6046, "step": 4910 }, { "epoch": 0.044169031312254636, "grad_norm": 2.892294406890869, "learning_rate": 2.779676464273987e-05, "loss": 2.3806, "step": 4911 }, { "epoch": 0.04417802520989508, "grad_norm": 6.099660873413086, "learning_rate": 2.7788549036809657e-05, "loss": 2.1971, "step": 4912 }, { "epoch": 0.04418701910753554, "grad_norm": 4.377013683319092, "learning_rate": 2.7780333125927917e-05, "loss": 2.0827, "step": 4913 }, { "epoch": 0.04419601300517599, "grad_norm": 3.1136553287506104, "learning_rate": 2.777211691099312e-05, "loss": 1.6505, "step": 4914 }, { "epoch": 0.04420500690281644, "grad_norm": 3.9912233352661133, "learning_rate": 2.776390039290378e-05, "loss": 2.1468, "step": 4915 }, { "epoch": 0.04421400080045689, "grad_norm": 3.0852699279785156, "learning_rate": 2.7755683572558456e-05, "loss": 2.6285, "step": 4916 }, { "epoch": 0.04422299469809734, "grad_norm": 5.474484443664551, "learning_rate": 2.7747466450855718e-05, "loss": 2.0528, "step": 4917 }, { "epoch": 0.04423198859573779, "grad_norm": 2.6658239364624023, "learning_rate": 2.773924902869418e-05, "loss": 2.3327, "step": 4918 }, { "epoch": 0.044240982493378246, "grad_norm": 5.286798477172852, "learning_rate": 2.773103130697249e-05, "loss": 1.9201, "step": 4919 }, { "epoch": 0.04424997639101869, "grad_norm": 3.00640606880188, "learning_rate": 2.7722813286589316e-05, "loss": 2.2662, "step": 4920 }, { "epoch": 0.04425897028865915, "grad_norm": 3.387028694152832, "learning_rate": 2.7714594968443374e-05, "loss": 2.4471, "step": 4921 }, { "epoch": 0.044267964186299594, "grad_norm": 2.558910369873047, "learning_rate": 2.7706376353433415e-05, "loss": 2.105, "step": 4922 }, { "epoch": 0.04427695808394005, "grad_norm": 3.0294251441955566, "learning_rate": 2.7698157442458205e-05, "loss": 1.9925, "step": 4923 }, { "epoch": 0.044285951981580494, "grad_norm": 5.656405448913574, "learning_rate": 2.7689938236416546e-05, "loss": 2.2691, "step": 4924 }, { "epoch": 0.04429494587922095, "grad_norm": 2.693061113357544, "learning_rate": 2.7681718736207298e-05, "loss": 2.1937, "step": 4925 }, { "epoch": 0.0443039397768614, "grad_norm": 5.162328720092773, "learning_rate": 2.767349894272931e-05, "loss": 2.286, "step": 4926 }, { "epoch": 0.04431293367450185, "grad_norm": 7.686915397644043, "learning_rate": 2.76652788568815e-05, "loss": 1.8833, "step": 4927 }, { "epoch": 0.0443219275721423, "grad_norm": 2.725306987762451, "learning_rate": 2.7657058479562802e-05, "loss": 1.9571, "step": 4928 }, { "epoch": 0.04433092146978275, "grad_norm": 2.879467487335205, "learning_rate": 2.7648837811672185e-05, "loss": 2.1416, "step": 4929 }, { "epoch": 0.044339915367423204, "grad_norm": 4.2798171043396, "learning_rate": 2.764061685410865e-05, "loss": 1.9918, "step": 4930 }, { "epoch": 0.04434890926506366, "grad_norm": 3.892133951187134, "learning_rate": 2.7632395607771218e-05, "loss": 1.9346, "step": 4931 }, { "epoch": 0.044357903162704104, "grad_norm": 1.6484057903289795, "learning_rate": 2.7624174073558968e-05, "loss": 2.019, "step": 4932 }, { "epoch": 0.04436689706034456, "grad_norm": 2.1086013317108154, "learning_rate": 2.7615952252370986e-05, "loss": 2.2618, "step": 4933 }, { "epoch": 0.044375890957985005, "grad_norm": 2.34692120552063, "learning_rate": 2.7607730145106392e-05, "loss": 1.8924, "step": 4934 }, { "epoch": 0.04438488485562546, "grad_norm": 2.550607919692993, "learning_rate": 2.7599507752664354e-05, "loss": 2.1684, "step": 4935 }, { "epoch": 0.044393878753265906, "grad_norm": 1.8693647384643555, "learning_rate": 2.7591285075944056e-05, "loss": 1.8907, "step": 4936 }, { "epoch": 0.04440287265090636, "grad_norm": 2.743255853652954, "learning_rate": 2.7583062115844715e-05, "loss": 1.5295, "step": 4937 }, { "epoch": 0.04441186654854681, "grad_norm": 3.8382177352905273, "learning_rate": 2.757483887326559e-05, "loss": 2.5836, "step": 4938 }, { "epoch": 0.04442086044618726, "grad_norm": 3.845322608947754, "learning_rate": 2.7566615349105963e-05, "loss": 2.044, "step": 4939 }, { "epoch": 0.044429854343827714, "grad_norm": 3.428759813308716, "learning_rate": 2.755839154426513e-05, "loss": 2.0629, "step": 4940 }, { "epoch": 0.04443884824146816, "grad_norm": 3.021169900894165, "learning_rate": 2.7550167459642445e-05, "loss": 2.7217, "step": 4941 }, { "epoch": 0.044447842139108615, "grad_norm": 4.300188064575195, "learning_rate": 2.7541943096137286e-05, "loss": 1.9868, "step": 4942 }, { "epoch": 0.04445683603674907, "grad_norm": 4.509927749633789, "learning_rate": 2.7533718454649044e-05, "loss": 2.6207, "step": 4943 }, { "epoch": 0.044465829934389516, "grad_norm": 2.791628360748291, "learning_rate": 2.7525493536077168e-05, "loss": 2.2239, "step": 4944 }, { "epoch": 0.04447482383202997, "grad_norm": 5.94210958480835, "learning_rate": 2.7517268341321112e-05, "loss": 2.0633, "step": 4945 }, { "epoch": 0.044483817729670416, "grad_norm": 2.729296922683716, "learning_rate": 2.7509042871280372e-05, "loss": 1.9585, "step": 4946 }, { "epoch": 0.04449281162731087, "grad_norm": 3.100639820098877, "learning_rate": 2.7500817126854482e-05, "loss": 3.0226, "step": 4947 }, { "epoch": 0.044501805524951324, "grad_norm": 4.243072032928467, "learning_rate": 2.7492591108942993e-05, "loss": 2.0333, "step": 4948 }, { "epoch": 0.04451079942259177, "grad_norm": 5.535202503204346, "learning_rate": 2.748436481844548e-05, "loss": 1.8131, "step": 4949 }, { "epoch": 0.044519793320232225, "grad_norm": 3.410081386566162, "learning_rate": 2.7476138256261575e-05, "loss": 2.7145, "step": 4950 }, { "epoch": 0.04452878721787267, "grad_norm": 2.5904288291931152, "learning_rate": 2.7467911423290897e-05, "loss": 2.2483, "step": 4951 }, { "epoch": 0.044537781115513125, "grad_norm": 3.848156213760376, "learning_rate": 2.7459684320433154e-05, "loss": 2.3474, "step": 4952 }, { "epoch": 0.04454677501315357, "grad_norm": 2.1022415161132812, "learning_rate": 2.7451456948588016e-05, "loss": 2.2607, "step": 4953 }, { "epoch": 0.044555768910794026, "grad_norm": 3.103858709335327, "learning_rate": 2.7443229308655243e-05, "loss": 2.3607, "step": 4954 }, { "epoch": 0.04456476280843448, "grad_norm": 3.100252389907837, "learning_rate": 2.7435001401534586e-05, "loss": 2.3273, "step": 4955 }, { "epoch": 0.04457375670607493, "grad_norm": 3.085465431213379, "learning_rate": 2.7426773228125836e-05, "loss": 2.1286, "step": 4956 }, { "epoch": 0.04458275060371538, "grad_norm": 4.178415775299072, "learning_rate": 2.7418544789328816e-05, "loss": 2.4116, "step": 4957 }, { "epoch": 0.04459174450135583, "grad_norm": 4.681471347808838, "learning_rate": 2.741031608604337e-05, "loss": 2.3347, "step": 4958 }, { "epoch": 0.04460073839899628, "grad_norm": 4.71623420715332, "learning_rate": 2.7402087119169378e-05, "loss": 2.1319, "step": 4959 }, { "epoch": 0.044609732296636735, "grad_norm": 2.670590400695801, "learning_rate": 2.7393857889606756e-05, "loss": 2.1341, "step": 4960 }, { "epoch": 0.04461872619427718, "grad_norm": 4.481751441955566, "learning_rate": 2.738562839825543e-05, "loss": 2.3313, "step": 4961 }, { "epoch": 0.044627720091917636, "grad_norm": 4.4707441329956055, "learning_rate": 2.7377398646015367e-05, "loss": 2.44, "step": 4962 }, { "epoch": 0.04463671398955808, "grad_norm": 3.310023069381714, "learning_rate": 2.736916863378657e-05, "loss": 2.154, "step": 4963 }, { "epoch": 0.04464570788719854, "grad_norm": 5.059321403503418, "learning_rate": 2.7360938362469052e-05, "loss": 2.4668, "step": 4964 }, { "epoch": 0.044654701784838984, "grad_norm": 3.421724557876587, "learning_rate": 2.7352707832962865e-05, "loss": 2.0956, "step": 4965 }, { "epoch": 0.04466369568247944, "grad_norm": 3.0014522075653076, "learning_rate": 2.7344477046168076e-05, "loss": 1.9574, "step": 4966 }, { "epoch": 0.04467268958011989, "grad_norm": 3.8977322578430176, "learning_rate": 2.7336246002984815e-05, "loss": 1.7836, "step": 4967 }, { "epoch": 0.04468168347776034, "grad_norm": 4.485860824584961, "learning_rate": 2.73280147043132e-05, "loss": 2.2761, "step": 4968 }, { "epoch": 0.04469067737540079, "grad_norm": 2.950026512145996, "learning_rate": 2.731978315105339e-05, "loss": 2.1263, "step": 4969 }, { "epoch": 0.04469967127304124, "grad_norm": 2.9019899368286133, "learning_rate": 2.731155134410559e-05, "loss": 2.6064, "step": 4970 }, { "epoch": 0.04470866517068169, "grad_norm": 3.187234401702881, "learning_rate": 2.730331928437001e-05, "loss": 1.8808, "step": 4971 }, { "epoch": 0.044717659068322146, "grad_norm": 3.752746105194092, "learning_rate": 2.729508697274689e-05, "loss": 1.7266, "step": 4972 }, { "epoch": 0.04472665296596259, "grad_norm": 7.6520514488220215, "learning_rate": 2.7286854410136513e-05, "loss": 2.5311, "step": 4973 }, { "epoch": 0.04473564686360305, "grad_norm": 4.129462718963623, "learning_rate": 2.7278621597439176e-05, "loss": 2.2846, "step": 4974 }, { "epoch": 0.044744640761243494, "grad_norm": 4.335081100463867, "learning_rate": 2.727038853555521e-05, "loss": 2.7237, "step": 4975 }, { "epoch": 0.04475363465888395, "grad_norm": 11.533097267150879, "learning_rate": 2.7262155225384962e-05, "loss": 3.1735, "step": 4976 }, { "epoch": 0.044762628556524395, "grad_norm": 3.1894588470458984, "learning_rate": 2.7253921667828823e-05, "loss": 2.1216, "step": 4977 }, { "epoch": 0.04477162245416485, "grad_norm": 4.0704216957092285, "learning_rate": 2.7245687863787195e-05, "loss": 2.2986, "step": 4978 }, { "epoch": 0.0447806163518053, "grad_norm": 3.639423370361328, "learning_rate": 2.7237453814160523e-05, "loss": 2.0811, "step": 4979 }, { "epoch": 0.04478961024944575, "grad_norm": 3.06514573097229, "learning_rate": 2.722921951984927e-05, "loss": 1.8104, "step": 4980 }, { "epoch": 0.0447986041470862, "grad_norm": 5.315256118774414, "learning_rate": 2.7220984981753912e-05, "loss": 2.2377, "step": 4981 }, { "epoch": 0.04480759804472665, "grad_norm": 3.2847325801849365, "learning_rate": 2.721275020077498e-05, "loss": 2.2635, "step": 4982 }, { "epoch": 0.044816591942367104, "grad_norm": 3.5668818950653076, "learning_rate": 2.7204515177813017e-05, "loss": 2.3219, "step": 4983 }, { "epoch": 0.04482558584000756, "grad_norm": 12.546243667602539, "learning_rate": 2.7196279913768584e-05, "loss": 2.2946, "step": 4984 }, { "epoch": 0.044834579737648005, "grad_norm": 6.471747875213623, "learning_rate": 2.7188044409542278e-05, "loss": 2.1387, "step": 4985 }, { "epoch": 0.04484357363528846, "grad_norm": 3.3148844242095947, "learning_rate": 2.7179808666034734e-05, "loss": 2.2863, "step": 4986 }, { "epoch": 0.044852567532928905, "grad_norm": 4.856374740600586, "learning_rate": 2.7171572684146583e-05, "loss": 2.6054, "step": 4987 }, { "epoch": 0.04486156143056936, "grad_norm": 2.375476598739624, "learning_rate": 2.716333646477851e-05, "loss": 2.4676, "step": 4988 }, { "epoch": 0.04487055532820981, "grad_norm": 4.274041652679443, "learning_rate": 2.715510000883122e-05, "loss": 2.4234, "step": 4989 }, { "epoch": 0.04487954922585026, "grad_norm": 2.608119487762451, "learning_rate": 2.7146863317205427e-05, "loss": 2.1144, "step": 4990 }, { "epoch": 0.044888543123490714, "grad_norm": 3.785956621170044, "learning_rate": 2.7138626390801892e-05, "loss": 2.5763, "step": 4991 }, { "epoch": 0.04489753702113116, "grad_norm": 3.1501755714416504, "learning_rate": 2.7130389230521392e-05, "loss": 2.324, "step": 4992 }, { "epoch": 0.044906530918771614, "grad_norm": 6.83990478515625, "learning_rate": 2.7122151837264725e-05, "loss": 1.6707, "step": 4993 }, { "epoch": 0.04491552481641206, "grad_norm": 3.217237710952759, "learning_rate": 2.7113914211932722e-05, "loss": 2.2632, "step": 4994 }, { "epoch": 0.044924518714052515, "grad_norm": 2.2386372089385986, "learning_rate": 2.7105676355426248e-05, "loss": 1.7645, "step": 4995 }, { "epoch": 0.04493351261169297, "grad_norm": 4.097300052642822, "learning_rate": 2.709743826864617e-05, "loss": 2.3151, "step": 4996 }, { "epoch": 0.044942506509333416, "grad_norm": 2.0368943214416504, "learning_rate": 2.708919995249339e-05, "loss": 2.3691, "step": 4997 }, { "epoch": 0.04495150040697387, "grad_norm": 4.16341495513916, "learning_rate": 2.7080961407868848e-05, "loss": 2.2911, "step": 4998 }, { "epoch": 0.04496049430461432, "grad_norm": 3.9103589057922363, "learning_rate": 2.7072722635673493e-05, "loss": 2.2137, "step": 4999 }, { "epoch": 0.04496948820225477, "grad_norm": 2.2765302658081055, "learning_rate": 2.7064483636808313e-05, "loss": 1.9887, "step": 5000 }, { "epoch": 0.044978482099895224, "grad_norm": 2.8304083347320557, "learning_rate": 2.7056244412174297e-05, "loss": 2.4739, "step": 5001 }, { "epoch": 0.04498747599753567, "grad_norm": 3.0284693241119385, "learning_rate": 2.704800496267249e-05, "loss": 2.2025, "step": 5002 }, { "epoch": 0.044996469895176125, "grad_norm": 3.2253036499023438, "learning_rate": 2.7039765289203946e-05, "loss": 2.1497, "step": 5003 }, { "epoch": 0.04500546379281657, "grad_norm": 1.615756869316101, "learning_rate": 2.703152539266972e-05, "loss": 2.3793, "step": 5004 }, { "epoch": 0.045014457690457026, "grad_norm": 4.387635231018066, "learning_rate": 2.7023285273970945e-05, "loss": 2.0654, "step": 5005 }, { "epoch": 0.04502345158809747, "grad_norm": 3.6833205223083496, "learning_rate": 2.701504493400873e-05, "loss": 2.5575, "step": 5006 }, { "epoch": 0.045032445485737926, "grad_norm": 3.371063709259033, "learning_rate": 2.700680437368423e-05, "loss": 1.8493, "step": 5007 }, { "epoch": 0.04504143938337838, "grad_norm": 5.274844169616699, "learning_rate": 2.6998563593898623e-05, "loss": 2.4784, "step": 5008 }, { "epoch": 0.04505043328101883, "grad_norm": 1.4459459781646729, "learning_rate": 2.699032259555311e-05, "loss": 1.8046, "step": 5009 }, { "epoch": 0.04505942717865928, "grad_norm": 5.93959903717041, "learning_rate": 2.6982081379548896e-05, "loss": 2.1197, "step": 5010 }, { "epoch": 0.04506842107629973, "grad_norm": 4.304078102111816, "learning_rate": 2.6973839946787254e-05, "loss": 2.4507, "step": 5011 }, { "epoch": 0.04507741497394018, "grad_norm": 2.394113302230835, "learning_rate": 2.696559829816944e-05, "loss": 1.881, "step": 5012 }, { "epoch": 0.045086408871580635, "grad_norm": 6.959613800048828, "learning_rate": 2.6957356434596752e-05, "loss": 2.4189, "step": 5013 }, { "epoch": 0.04509540276922108, "grad_norm": 3.5029120445251465, "learning_rate": 2.6949114356970513e-05, "loss": 2.086, "step": 5014 }, { "epoch": 0.045104396666861536, "grad_norm": 3.65659761428833, "learning_rate": 2.6940872066192052e-05, "loss": 2.0248, "step": 5015 }, { "epoch": 0.04511339056450198, "grad_norm": 2.484330177307129, "learning_rate": 2.6932629563162743e-05, "loss": 2.5471, "step": 5016 }, { "epoch": 0.04512238446214244, "grad_norm": 4.650343418121338, "learning_rate": 2.6924386848783962e-05, "loss": 2.6997, "step": 5017 }, { "epoch": 0.04513137835978289, "grad_norm": 4.77316951751709, "learning_rate": 2.6916143923957138e-05, "loss": 1.835, "step": 5018 }, { "epoch": 0.04514037225742334, "grad_norm": 3.179671287536621, "learning_rate": 2.6907900789583697e-05, "loss": 2.2398, "step": 5019 }, { "epoch": 0.04514936615506379, "grad_norm": 3.997448444366455, "learning_rate": 2.689965744656508e-05, "loss": 2.0334, "step": 5020 }, { "epoch": 0.04515836005270424, "grad_norm": 3.606933832168579, "learning_rate": 2.689141389580279e-05, "loss": 1.9331, "step": 5021 }, { "epoch": 0.04516735395034469, "grad_norm": 4.449479579925537, "learning_rate": 2.6883170138198323e-05, "loss": 1.8624, "step": 5022 }, { "epoch": 0.04517634784798514, "grad_norm": 3.6294403076171875, "learning_rate": 2.6874926174653196e-05, "loss": 2.2802, "step": 5023 }, { "epoch": 0.04518534174562559, "grad_norm": 2.3155062198638916, "learning_rate": 2.686668200606896e-05, "loss": 2.0849, "step": 5024 }, { "epoch": 0.04519433564326605, "grad_norm": 4.160567760467529, "learning_rate": 2.6858437633347194e-05, "loss": 2.5738, "step": 5025 }, { "epoch": 0.045203329540906494, "grad_norm": 2.783904790878296, "learning_rate": 2.685019305738947e-05, "loss": 1.8203, "step": 5026 }, { "epoch": 0.04521232343854695, "grad_norm": 2.0047266483306885, "learning_rate": 2.684194827909743e-05, "loss": 2.0415, "step": 5027 }, { "epoch": 0.045221317336187394, "grad_norm": 3.0099573135375977, "learning_rate": 2.6833703299372687e-05, "loss": 1.8498, "step": 5028 }, { "epoch": 0.04523031123382785, "grad_norm": 2.912386178970337, "learning_rate": 2.682545811911691e-05, "loss": 1.9568, "step": 5029 }, { "epoch": 0.0452393051314683, "grad_norm": 4.430185317993164, "learning_rate": 2.681721273923178e-05, "loss": 2.4085, "step": 5030 }, { "epoch": 0.04524829902910875, "grad_norm": 3.95168137550354, "learning_rate": 2.6808967160618992e-05, "loss": 2.1371, "step": 5031 }, { "epoch": 0.0452572929267492, "grad_norm": 1.9624909162521362, "learning_rate": 2.6800721384180277e-05, "loss": 2.0607, "step": 5032 }, { "epoch": 0.04526628682438965, "grad_norm": 2.6591684818267822, "learning_rate": 2.679247541081738e-05, "loss": 1.7073, "step": 5033 }, { "epoch": 0.0452752807220301, "grad_norm": 2.5074243545532227, "learning_rate": 2.6784229241432067e-05, "loss": 2.1225, "step": 5034 }, { "epoch": 0.04528427461967055, "grad_norm": 5.541238784790039, "learning_rate": 2.6775982876926126e-05, "loss": 1.9594, "step": 5035 }, { "epoch": 0.045293268517311004, "grad_norm": 2.2623791694641113, "learning_rate": 2.6767736318201364e-05, "loss": 1.9725, "step": 5036 }, { "epoch": 0.04530226241495146, "grad_norm": 4.550655841827393, "learning_rate": 2.6759489566159624e-05, "loss": 2.0444, "step": 5037 }, { "epoch": 0.045311256312591905, "grad_norm": 3.350583553314209, "learning_rate": 2.675124262170275e-05, "loss": 2.1345, "step": 5038 }, { "epoch": 0.04532025021023236, "grad_norm": 2.8841166496276855, "learning_rate": 2.6742995485732606e-05, "loss": 1.6203, "step": 5039 }, { "epoch": 0.045329244107872806, "grad_norm": 2.764591693878174, "learning_rate": 2.6734748159151102e-05, "loss": 2.1213, "step": 5040 }, { "epoch": 0.04533823800551326, "grad_norm": 3.5615580081939697, "learning_rate": 2.6726500642860154e-05, "loss": 2.0992, "step": 5041 }, { "epoch": 0.04534723190315371, "grad_norm": 3.6352360248565674, "learning_rate": 2.671825293776168e-05, "loss": 2.1118, "step": 5042 }, { "epoch": 0.04535622580079416, "grad_norm": 3.089155435562134, "learning_rate": 2.671000504475765e-05, "loss": 2.6768, "step": 5043 }, { "epoch": 0.045365219698434614, "grad_norm": 7.02030086517334, "learning_rate": 2.6701756964750048e-05, "loss": 2.149, "step": 5044 }, { "epoch": 0.04537421359607506, "grad_norm": 3.2538561820983887, "learning_rate": 2.6693508698640852e-05, "loss": 1.9677, "step": 5045 }, { "epoch": 0.045383207493715515, "grad_norm": 3.364259958267212, "learning_rate": 2.66852602473321e-05, "loss": 2.3606, "step": 5046 }, { "epoch": 0.04539220139135596, "grad_norm": 6.755740165710449, "learning_rate": 2.667701161172581e-05, "loss": 1.9894, "step": 5047 }, { "epoch": 0.045401195288996415, "grad_norm": 2.507852554321289, "learning_rate": 2.666876279272407e-05, "loss": 2.3749, "step": 5048 }, { "epoch": 0.04541018918663687, "grad_norm": 3.2537856101989746, "learning_rate": 2.6660513791228918e-05, "loss": 2.047, "step": 5049 }, { "epoch": 0.045419183084277316, "grad_norm": 3.947352647781372, "learning_rate": 2.6652264608142484e-05, "loss": 2.3709, "step": 5050 }, { "epoch": 0.04542817698191777, "grad_norm": 4.551741123199463, "learning_rate": 2.6644015244366883e-05, "loss": 2.3215, "step": 5051 }, { "epoch": 0.04543717087955822, "grad_norm": 6.136465549468994, "learning_rate": 2.663576570080424e-05, "loss": 2.4313, "step": 5052 }, { "epoch": 0.04544616477719867, "grad_norm": 2.6884520053863525, "learning_rate": 2.6627515978356722e-05, "loss": 1.901, "step": 5053 }, { "epoch": 0.045455158674839125, "grad_norm": 3.1607017517089844, "learning_rate": 2.6619266077926504e-05, "loss": 1.9762, "step": 5054 }, { "epoch": 0.04546415257247957, "grad_norm": 2.5834574699401855, "learning_rate": 2.661101600041577e-05, "loss": 2.2253, "step": 5055 }, { "epoch": 0.045473146470120025, "grad_norm": 3.034700632095337, "learning_rate": 2.6602765746726767e-05, "loss": 2.1138, "step": 5056 }, { "epoch": 0.04548214036776047, "grad_norm": 5.624624252319336, "learning_rate": 2.6594515317761705e-05, "loss": 2.2816, "step": 5057 }, { "epoch": 0.045491134265400926, "grad_norm": 2.0225327014923096, "learning_rate": 2.6586264714422836e-05, "loss": 2.3029, "step": 5058 }, { "epoch": 0.04550012816304138, "grad_norm": 2.2266435623168945, "learning_rate": 2.6578013937612456e-05, "loss": 1.743, "step": 5059 }, { "epoch": 0.04550912206068183, "grad_norm": 3.446704864501953, "learning_rate": 2.656976298823284e-05, "loss": 1.8363, "step": 5060 }, { "epoch": 0.04551811595832228, "grad_norm": 2.1071715354919434, "learning_rate": 2.65615118671863e-05, "loss": 1.7932, "step": 5061 }, { "epoch": 0.04552710985596273, "grad_norm": 3.800159454345703, "learning_rate": 2.6553260575375172e-05, "loss": 2.1765, "step": 5062 }, { "epoch": 0.04553610375360318, "grad_norm": 1.7621726989746094, "learning_rate": 2.6545009113701802e-05, "loss": 2.4305, "step": 5063 }, { "epoch": 0.04554509765124363, "grad_norm": 1.9368414878845215, "learning_rate": 2.6536757483068557e-05, "loss": 2.6104, "step": 5064 }, { "epoch": 0.04555409154888408, "grad_norm": 2.252262592315674, "learning_rate": 2.652850568437783e-05, "loss": 2.2126, "step": 5065 }, { "epoch": 0.045563085446524536, "grad_norm": 1.6737862825393677, "learning_rate": 2.6520253718532017e-05, "loss": 1.9369, "step": 5066 }, { "epoch": 0.04557207934416498, "grad_norm": 5.041394233703613, "learning_rate": 2.651200158643354e-05, "loss": 1.5322, "step": 5067 }, { "epoch": 0.045581073241805437, "grad_norm": 2.47914457321167, "learning_rate": 2.650374928898485e-05, "loss": 2.7942, "step": 5068 }, { "epoch": 0.04559006713944588, "grad_norm": 2.6177151203155518, "learning_rate": 2.649549682708839e-05, "loss": 2.0909, "step": 5069 }, { "epoch": 0.04559906103708634, "grad_norm": 2.8211236000061035, "learning_rate": 2.6487244201646645e-05, "loss": 1.9011, "step": 5070 }, { "epoch": 0.04560805493472679, "grad_norm": 2.833315849304199, "learning_rate": 2.6478991413562116e-05, "loss": 2.0108, "step": 5071 }, { "epoch": 0.04561704883236724, "grad_norm": 3.645510196685791, "learning_rate": 2.6470738463737306e-05, "loss": 2.1759, "step": 5072 }, { "epoch": 0.04562604273000769, "grad_norm": 3.542841911315918, "learning_rate": 2.6462485353074752e-05, "loss": 1.9348, "step": 5073 }, { "epoch": 0.04563503662764814, "grad_norm": 3.052518129348755, "learning_rate": 2.6454232082476986e-05, "loss": 2.1464, "step": 5074 }, { "epoch": 0.04564403052528859, "grad_norm": 2.8282887935638428, "learning_rate": 2.6445978652846602e-05, "loss": 1.9498, "step": 5075 }, { "epoch": 0.04565302442292904, "grad_norm": 2.1463308334350586, "learning_rate": 2.6437725065086162e-05, "loss": 2.3036, "step": 5076 }, { "epoch": 0.04566201832056949, "grad_norm": 2.6716153621673584, "learning_rate": 2.6429471320098266e-05, "loss": 2.2616, "step": 5077 }, { "epoch": 0.04567101221820995, "grad_norm": 3.1602587699890137, "learning_rate": 2.6421217418785544e-05, "loss": 1.9879, "step": 5078 }, { "epoch": 0.045680006115850394, "grad_norm": 6.827674865722656, "learning_rate": 2.6412963362050618e-05, "loss": 2.387, "step": 5079 }, { "epoch": 0.04568900001349085, "grad_norm": 2.822599172592163, "learning_rate": 2.6404709150796137e-05, "loss": 2.1554, "step": 5080 }, { "epoch": 0.045697993911131295, "grad_norm": 2.9847495555877686, "learning_rate": 2.6396454785924783e-05, "loss": 1.9544, "step": 5081 }, { "epoch": 0.04570698780877175, "grad_norm": 4.060389995574951, "learning_rate": 2.638820026833923e-05, "loss": 1.7998, "step": 5082 }, { "epoch": 0.0457159817064122, "grad_norm": 5.038862228393555, "learning_rate": 2.637994559894218e-05, "loss": 2.1868, "step": 5083 }, { "epoch": 0.04572497560405265, "grad_norm": 5.029531002044678, "learning_rate": 2.637169077863636e-05, "loss": 2.6822, "step": 5084 }, { "epoch": 0.0457339695016931, "grad_norm": 2.9052393436431885, "learning_rate": 2.63634358083245e-05, "loss": 2.0674, "step": 5085 }, { "epoch": 0.04574296339933355, "grad_norm": 3.330446720123291, "learning_rate": 2.6355180688909348e-05, "loss": 1.7352, "step": 5086 }, { "epoch": 0.045751957296974004, "grad_norm": 3.5729780197143555, "learning_rate": 2.6346925421293668e-05, "loss": 2.3043, "step": 5087 }, { "epoch": 0.04576095119461445, "grad_norm": 3.161336898803711, "learning_rate": 2.633867000638025e-05, "loss": 1.9123, "step": 5088 }, { "epoch": 0.045769945092254904, "grad_norm": 4.466612339019775, "learning_rate": 2.63304144450719e-05, "loss": 2.0756, "step": 5089 }, { "epoch": 0.04577893898989536, "grad_norm": 3.52188777923584, "learning_rate": 2.632215873827142e-05, "loss": 2.1587, "step": 5090 }, { "epoch": 0.045787932887535805, "grad_norm": 3.133718252182007, "learning_rate": 2.6313902886881647e-05, "loss": 2.0414, "step": 5091 }, { "epoch": 0.04579692678517626, "grad_norm": 3.7112300395965576, "learning_rate": 2.6305646891805435e-05, "loss": 2.1122, "step": 5092 }, { "epoch": 0.045805920682816706, "grad_norm": 3.124650001525879, "learning_rate": 2.629739075394564e-05, "loss": 2.408, "step": 5093 }, { "epoch": 0.04581491458045716, "grad_norm": 2.503105878829956, "learning_rate": 2.628913447420514e-05, "loss": 1.6755, "step": 5094 }, { "epoch": 0.045823908478097614, "grad_norm": 3.1595520973205566, "learning_rate": 2.6280878053486828e-05, "loss": 2.2651, "step": 5095 }, { "epoch": 0.04583290237573806, "grad_norm": 2.652385950088501, "learning_rate": 2.627262149269362e-05, "loss": 2.2622, "step": 5096 }, { "epoch": 0.045841896273378514, "grad_norm": 2.774479389190674, "learning_rate": 2.6264364792728446e-05, "loss": 2.2973, "step": 5097 }, { "epoch": 0.04585089017101896, "grad_norm": 4.0915141105651855, "learning_rate": 2.6256107954494242e-05, "loss": 2.0856, "step": 5098 }, { "epoch": 0.045859884068659415, "grad_norm": 3.619215250015259, "learning_rate": 2.624785097889395e-05, "loss": 2.0201, "step": 5099 }, { "epoch": 0.04586887796629987, "grad_norm": 2.6916935443878174, "learning_rate": 2.623959386683056e-05, "loss": 2.1112, "step": 5100 }, { "epoch": 0.045877871863940316, "grad_norm": 3.8351221084594727, "learning_rate": 2.6231336619207044e-05, "loss": 2.1245, "step": 5101 }, { "epoch": 0.04588686576158077, "grad_norm": 3.1218068599700928, "learning_rate": 2.6223079236926407e-05, "loss": 1.9606, "step": 5102 }, { "epoch": 0.045895859659221216, "grad_norm": 3.2530674934387207, "learning_rate": 2.6214821720891665e-05, "loss": 1.9026, "step": 5103 }, { "epoch": 0.04590485355686167, "grad_norm": 2.834376335144043, "learning_rate": 2.6206564072005846e-05, "loss": 2.1464, "step": 5104 }, { "epoch": 0.04591384745450212, "grad_norm": 3.7240493297576904, "learning_rate": 2.6198306291171993e-05, "loss": 2.1176, "step": 5105 }, { "epoch": 0.04592284135214257, "grad_norm": 3.5295441150665283, "learning_rate": 2.6190048379293157e-05, "loss": 2.5283, "step": 5106 }, { "epoch": 0.045931835249783025, "grad_norm": 2.921454906463623, "learning_rate": 2.6181790337272433e-05, "loss": 2.1836, "step": 5107 }, { "epoch": 0.04594082914742347, "grad_norm": 3.8743956089019775, "learning_rate": 2.6173532166012895e-05, "loss": 2.1266, "step": 5108 }, { "epoch": 0.045949823045063926, "grad_norm": 2.879021644592285, "learning_rate": 2.6165273866417633e-05, "loss": 1.9797, "step": 5109 }, { "epoch": 0.04595881694270437, "grad_norm": 3.333691358566284, "learning_rate": 2.6157015439389774e-05, "loss": 2.2043, "step": 5110 }, { "epoch": 0.045967810840344826, "grad_norm": 3.893834114074707, "learning_rate": 2.6148756885832453e-05, "loss": 2.0979, "step": 5111 }, { "epoch": 0.04597680473798528, "grad_norm": 3.3295509815216064, "learning_rate": 2.61404982066488e-05, "loss": 2.1826, "step": 5112 }, { "epoch": 0.04598579863562573, "grad_norm": 4.920767307281494, "learning_rate": 2.6132239402741988e-05, "loss": 2.5452, "step": 5113 }, { "epoch": 0.04599479253326618, "grad_norm": 2.2075579166412354, "learning_rate": 2.612398047501517e-05, "loss": 2.2858, "step": 5114 }, { "epoch": 0.04600378643090663, "grad_norm": 2.9409914016723633, "learning_rate": 2.6115721424371532e-05, "loss": 1.905, "step": 5115 }, { "epoch": 0.04601278032854708, "grad_norm": 2.6833040714263916, "learning_rate": 2.610746225171428e-05, "loss": 2.1632, "step": 5116 }, { "epoch": 0.04602177422618753, "grad_norm": 3.3277456760406494, "learning_rate": 2.6099202957946624e-05, "loss": 2.1989, "step": 5117 }, { "epoch": 0.04603076812382798, "grad_norm": 2.896390676498413, "learning_rate": 2.6090943543971784e-05, "loss": 2.5939, "step": 5118 }, { "epoch": 0.046039762021468436, "grad_norm": 2.5297305583953857, "learning_rate": 2.6082684010692998e-05, "loss": 2.7076, "step": 5119 }, { "epoch": 0.04604875591910888, "grad_norm": 3.8000524044036865, "learning_rate": 2.6074424359013517e-05, "loss": 2.1208, "step": 5120 }, { "epoch": 0.04605774981674934, "grad_norm": 3.7173140048980713, "learning_rate": 2.6066164589836607e-05, "loss": 2.0446, "step": 5121 }, { "epoch": 0.046066743714389784, "grad_norm": 2.0361337661743164, "learning_rate": 2.6057904704065533e-05, "loss": 2.0206, "step": 5122 }, { "epoch": 0.04607573761203024, "grad_norm": 2.7258429527282715, "learning_rate": 2.6049644702603593e-05, "loss": 2.1498, "step": 5123 }, { "epoch": 0.04608473150967069, "grad_norm": 3.7687466144561768, "learning_rate": 2.604138458635409e-05, "loss": 2.1759, "step": 5124 }, { "epoch": 0.04609372540731114, "grad_norm": 4.768552780151367, "learning_rate": 2.6033124356220328e-05, "loss": 2.5302, "step": 5125 }, { "epoch": 0.04610271930495159, "grad_norm": 2.650029182434082, "learning_rate": 2.602486401310565e-05, "loss": 2.2968, "step": 5126 }, { "epoch": 0.04611171320259204, "grad_norm": 3.046252965927124, "learning_rate": 2.6016603557913378e-05, "loss": 2.3807, "step": 5127 }, { "epoch": 0.04612070710023249, "grad_norm": 5.049255847930908, "learning_rate": 2.600834299154687e-05, "loss": 2.1131, "step": 5128 }, { "epoch": 0.04612970099787295, "grad_norm": 7.545327663421631, "learning_rate": 2.6000082314909485e-05, "loss": 2.3501, "step": 5129 }, { "epoch": 0.046138694895513394, "grad_norm": 2.0837740898132324, "learning_rate": 2.599182152890461e-05, "loss": 1.592, "step": 5130 }, { "epoch": 0.04614768879315385, "grad_norm": 2.6050901412963867, "learning_rate": 2.5983560634435616e-05, "loss": 1.8423, "step": 5131 }, { "epoch": 0.046156682690794294, "grad_norm": 2.3437817096710205, "learning_rate": 2.5975299632405914e-05, "loss": 1.9432, "step": 5132 }, { "epoch": 0.04616567658843475, "grad_norm": 2.0068624019622803, "learning_rate": 2.596703852371891e-05, "loss": 3.4761, "step": 5133 }, { "epoch": 0.046174670486075195, "grad_norm": 2.6003262996673584, "learning_rate": 2.595877730927802e-05, "loss": 2.1604, "step": 5134 }, { "epoch": 0.04618366438371565, "grad_norm": 2.8706538677215576, "learning_rate": 2.5950515989986697e-05, "loss": 2.0547, "step": 5135 }, { "epoch": 0.0461926582813561, "grad_norm": 3.139249801635742, "learning_rate": 2.594225456674837e-05, "loss": 2.1435, "step": 5136 }, { "epoch": 0.04620165217899655, "grad_norm": 2.87064790725708, "learning_rate": 2.593399304046649e-05, "loss": 2.0943, "step": 5137 }, { "epoch": 0.046210646076637, "grad_norm": 3.426612138748169, "learning_rate": 2.592573141204455e-05, "loss": 2.1694, "step": 5138 }, { "epoch": 0.04621963997427745, "grad_norm": 3.4315083026885986, "learning_rate": 2.5917469682386008e-05, "loss": 2.0987, "step": 5139 }, { "epoch": 0.046228633871917904, "grad_norm": 3.496582269668579, "learning_rate": 2.5909207852394363e-05, "loss": 2.4367, "step": 5140 }, { "epoch": 0.04623762776955836, "grad_norm": 4.408329486846924, "learning_rate": 2.5900945922973115e-05, "loss": 2.4391, "step": 5141 }, { "epoch": 0.046246621667198805, "grad_norm": 3.1843793392181396, "learning_rate": 2.5892683895025767e-05, "loss": 2.2676, "step": 5142 }, { "epoch": 0.04625561556483926, "grad_norm": 2.474181652069092, "learning_rate": 2.588442176945586e-05, "loss": 2.0628, "step": 5143 }, { "epoch": 0.046264609462479706, "grad_norm": 2.6159379482269287, "learning_rate": 2.5876159547166905e-05, "loss": 1.8906, "step": 5144 }, { "epoch": 0.04627360336012016, "grad_norm": 3.677233934402466, "learning_rate": 2.5867897229062478e-05, "loss": 1.8685, "step": 5145 }, { "epoch": 0.046282597257760606, "grad_norm": 6.876453876495361, "learning_rate": 2.5859634816046102e-05, "loss": 2.3168, "step": 5146 }, { "epoch": 0.04629159115540106, "grad_norm": 3.2355425357818604, "learning_rate": 2.5851372309021354e-05, "loss": 2.2813, "step": 5147 }, { "epoch": 0.046300585053041514, "grad_norm": 4.065851211547852, "learning_rate": 2.5843109708891822e-05, "loss": 1.7703, "step": 5148 }, { "epoch": 0.04630957895068196, "grad_norm": 3.5560128688812256, "learning_rate": 2.583484701656107e-05, "loss": 2.5068, "step": 5149 }, { "epoch": 0.046318572848322415, "grad_norm": 3.497321128845215, "learning_rate": 2.5826584232932706e-05, "loss": 2.6109, "step": 5150 }, { "epoch": 0.04632756674596286, "grad_norm": 4.665919780731201, "learning_rate": 2.5818321358910337e-05, "loss": 2.2116, "step": 5151 }, { "epoch": 0.046336560643603315, "grad_norm": 2.6690828800201416, "learning_rate": 2.581005839539757e-05, "loss": 1.9697, "step": 5152 }, { "epoch": 0.04634555454124377, "grad_norm": 5.027416706085205, "learning_rate": 2.580179534329803e-05, "loss": 2.2849, "step": 5153 }, { "epoch": 0.046354548438884216, "grad_norm": 3.9465904235839844, "learning_rate": 2.5793532203515363e-05, "loss": 1.9721, "step": 5154 }, { "epoch": 0.04636354233652467, "grad_norm": 3.2700536251068115, "learning_rate": 2.578526897695321e-05, "loss": 2.6763, "step": 5155 }, { "epoch": 0.04637253623416512, "grad_norm": 3.09586763381958, "learning_rate": 2.5777005664515224e-05, "loss": 2.6119, "step": 5156 }, { "epoch": 0.04638153013180557, "grad_norm": 2.3922524452209473, "learning_rate": 2.5768742267105055e-05, "loss": 2.2316, "step": 5157 }, { "epoch": 0.04639052402944602, "grad_norm": 3.0257084369659424, "learning_rate": 2.5760478785626397e-05, "loss": 2.0325, "step": 5158 }, { "epoch": 0.04639951792708647, "grad_norm": 3.3396401405334473, "learning_rate": 2.5752215220982922e-05, "loss": 1.8666, "step": 5159 }, { "epoch": 0.046408511824726925, "grad_norm": 2.2505831718444824, "learning_rate": 2.5743951574078314e-05, "loss": 1.6694, "step": 5160 }, { "epoch": 0.04641750572236737, "grad_norm": 2.3353114128112793, "learning_rate": 2.5735687845816285e-05, "loss": 2.2228, "step": 5161 }, { "epoch": 0.046426499620007826, "grad_norm": 2.3149213790893555, "learning_rate": 2.5727424037100544e-05, "loss": 2.094, "step": 5162 }, { "epoch": 0.04643549351764827, "grad_norm": 4.914117813110352, "learning_rate": 2.57191601488348e-05, "loss": 2.2414, "step": 5163 }, { "epoch": 0.04644448741528873, "grad_norm": 1.9890544414520264, "learning_rate": 2.571089618192278e-05, "loss": 2.307, "step": 5164 }, { "epoch": 0.04645348131292918, "grad_norm": 4.990524768829346, "learning_rate": 2.5702632137268223e-05, "loss": 2.1824, "step": 5165 }, { "epoch": 0.04646247521056963, "grad_norm": 2.516310691833496, "learning_rate": 2.5694368015774872e-05, "loss": 1.6445, "step": 5166 }, { "epoch": 0.04647146910821008, "grad_norm": 2.6626689434051514, "learning_rate": 2.568610381834648e-05, "loss": 2.4893, "step": 5167 }, { "epoch": 0.04648046300585053, "grad_norm": 5.6065592765808105, "learning_rate": 2.5677839545886805e-05, "loss": 2.1191, "step": 5168 }, { "epoch": 0.04648945690349098, "grad_norm": 3.0741448402404785, "learning_rate": 2.5669575199299616e-05, "loss": 2.1511, "step": 5169 }, { "epoch": 0.046498450801131436, "grad_norm": 3.0187113285064697, "learning_rate": 2.5661310779488695e-05, "loss": 1.9788, "step": 5170 }, { "epoch": 0.04650744469877188, "grad_norm": 3.1977550983428955, "learning_rate": 2.5653046287357824e-05, "loss": 2.362, "step": 5171 }, { "epoch": 0.046516438596412336, "grad_norm": 2.6257553100585938, "learning_rate": 2.5644781723810785e-05, "loss": 2.5313, "step": 5172 }, { "epoch": 0.04652543249405278, "grad_norm": 2.7663722038269043, "learning_rate": 2.5636517089751388e-05, "loss": 2.0478, "step": 5173 }, { "epoch": 0.04653442639169324, "grad_norm": 4.382857322692871, "learning_rate": 2.562825238608344e-05, "loss": 2.3629, "step": 5174 }, { "epoch": 0.046543420289333684, "grad_norm": 3.3931589126586914, "learning_rate": 2.5619987613710756e-05, "loss": 1.8107, "step": 5175 }, { "epoch": 0.04655241418697414, "grad_norm": 3.393751621246338, "learning_rate": 2.5611722773537156e-05, "loss": 2.2052, "step": 5176 }, { "epoch": 0.04656140808461459, "grad_norm": 3.782433032989502, "learning_rate": 2.5603457866466474e-05, "loss": 2.4152, "step": 5177 }, { "epoch": 0.04657040198225504, "grad_norm": 2.874239206314087, "learning_rate": 2.5595192893402553e-05, "loss": 1.9289, "step": 5178 }, { "epoch": 0.04657939587989549, "grad_norm": 2.7123780250549316, "learning_rate": 2.5586927855249222e-05, "loss": 2.5269, "step": 5179 }, { "epoch": 0.04658838977753594, "grad_norm": 2.939436197280884, "learning_rate": 2.5578662752910347e-05, "loss": 2.1479, "step": 5180 }, { "epoch": 0.04659738367517639, "grad_norm": 3.6181390285491943, "learning_rate": 2.557039758728979e-05, "loss": 2.1625, "step": 5181 }, { "epoch": 0.04660637757281685, "grad_norm": 2.7362334728240967, "learning_rate": 2.55621323592914e-05, "loss": 2.033, "step": 5182 }, { "epoch": 0.046615371470457294, "grad_norm": 8.72860336303711, "learning_rate": 2.5553867069819065e-05, "loss": 2.3304, "step": 5183 }, { "epoch": 0.04662436536809775, "grad_norm": 2.8409061431884766, "learning_rate": 2.5545601719776663e-05, "loss": 2.0288, "step": 5184 }, { "epoch": 0.046633359265738195, "grad_norm": 4.4990010261535645, "learning_rate": 2.553733631006807e-05, "loss": 1.4703, "step": 5185 }, { "epoch": 0.04664235316337865, "grad_norm": 2.0235061645507812, "learning_rate": 2.55290708415972e-05, "loss": 2.1236, "step": 5186 }, { "epoch": 0.046651347061019095, "grad_norm": 5.405665874481201, "learning_rate": 2.5520805315267925e-05, "loss": 1.9617, "step": 5187 }, { "epoch": 0.04666034095865955, "grad_norm": 4.217113971710205, "learning_rate": 2.5512539731984165e-05, "loss": 2.1595, "step": 5188 }, { "epoch": 0.0466693348563, "grad_norm": 4.803822994232178, "learning_rate": 2.5504274092649834e-05, "loss": 1.9642, "step": 5189 }, { "epoch": 0.04667832875394045, "grad_norm": 5.099948406219482, "learning_rate": 2.5496008398168843e-05, "loss": 2.1393, "step": 5190 }, { "epoch": 0.046687322651580904, "grad_norm": 4.123612880706787, "learning_rate": 2.5487742649445123e-05, "loss": 2.2914, "step": 5191 }, { "epoch": 0.04669631654922135, "grad_norm": 5.615482330322266, "learning_rate": 2.5479476847382596e-05, "loss": 2.5954, "step": 5192 }, { "epoch": 0.046705310446861804, "grad_norm": 2.0489306449890137, "learning_rate": 2.547121099288521e-05, "loss": 1.9997, "step": 5193 }, { "epoch": 0.04671430434450226, "grad_norm": 3.5331311225891113, "learning_rate": 2.5462945086856892e-05, "loss": 2.405, "step": 5194 }, { "epoch": 0.046723298242142705, "grad_norm": 5.760573863983154, "learning_rate": 2.5454679130201593e-05, "loss": 2.0993, "step": 5195 }, { "epoch": 0.04673229213978316, "grad_norm": 3.7073330879211426, "learning_rate": 2.544641312382327e-05, "loss": 2.4489, "step": 5196 }, { "epoch": 0.046741286037423606, "grad_norm": 2.803060531616211, "learning_rate": 2.5438147068625882e-05, "loss": 1.6476, "step": 5197 }, { "epoch": 0.04675027993506406, "grad_norm": 3.1334733963012695, "learning_rate": 2.5429880965513386e-05, "loss": 2.0036, "step": 5198 }, { "epoch": 0.04675927383270451, "grad_norm": 6.183271884918213, "learning_rate": 2.5421614815389754e-05, "loss": 2.3146, "step": 5199 }, { "epoch": 0.04676826773034496, "grad_norm": 3.638279438018799, "learning_rate": 2.5413348619158967e-05, "loss": 2.2006, "step": 5200 }, { "epoch": 0.046777261627985414, "grad_norm": 4.137958526611328, "learning_rate": 2.5405082377724987e-05, "loss": 2.6255, "step": 5201 }, { "epoch": 0.04678625552562586, "grad_norm": 3.394043445587158, "learning_rate": 2.539681609199181e-05, "loss": 1.5709, "step": 5202 }, { "epoch": 0.046795249423266315, "grad_norm": 3.8730828762054443, "learning_rate": 2.5388549762863422e-05, "loss": 2.344, "step": 5203 }, { "epoch": 0.04680424332090676, "grad_norm": 3.7443766593933105, "learning_rate": 2.5380283391243815e-05, "loss": 1.5639, "step": 5204 }, { "epoch": 0.046813237218547216, "grad_norm": 1.6988364458084106, "learning_rate": 2.5372016978036995e-05, "loss": 2.4322, "step": 5205 }, { "epoch": 0.04682223111618767, "grad_norm": 2.818317174911499, "learning_rate": 2.536375052414695e-05, "loss": 2.2968, "step": 5206 }, { "epoch": 0.046831225013828116, "grad_norm": 2.972278118133545, "learning_rate": 2.53554840304777e-05, "loss": 2.0498, "step": 5207 }, { "epoch": 0.04684021891146857, "grad_norm": 1.8356956243515015, "learning_rate": 2.5347217497933244e-05, "loss": 2.1876, "step": 5208 }, { "epoch": 0.04684921280910902, "grad_norm": 4.384932518005371, "learning_rate": 2.533895092741761e-05, "loss": 1.9289, "step": 5209 }, { "epoch": 0.04685820670674947, "grad_norm": 2.1846702098846436, "learning_rate": 2.5330684319834814e-05, "loss": 1.8672, "step": 5210 }, { "epoch": 0.046867200604389925, "grad_norm": 3.417445182800293, "learning_rate": 2.532241767608887e-05, "loss": 2.1579, "step": 5211 }, { "epoch": 0.04687619450203037, "grad_norm": 4.276543617248535, "learning_rate": 2.531415099708382e-05, "loss": 1.8191, "step": 5212 }, { "epoch": 0.046885188399670825, "grad_norm": 2.479581117630005, "learning_rate": 2.5305884283723685e-05, "loss": 2.188, "step": 5213 }, { "epoch": 0.04689418229731127, "grad_norm": 3.4722561836242676, "learning_rate": 2.529761753691251e-05, "loss": 1.9465, "step": 5214 }, { "epoch": 0.046903176194951726, "grad_norm": 2.7781426906585693, "learning_rate": 2.528935075755432e-05, "loss": 2.8415, "step": 5215 }, { "epoch": 0.04691217009259217, "grad_norm": 2.9297571182250977, "learning_rate": 2.5281083946553174e-05, "loss": 2.119, "step": 5216 }, { "epoch": 0.04692116399023263, "grad_norm": 3.7519445419311523, "learning_rate": 2.5272817104813107e-05, "loss": 2.6621, "step": 5217 }, { "epoch": 0.04693015788787308, "grad_norm": 9.750064849853516, "learning_rate": 2.526455023323817e-05, "loss": 2.0759, "step": 5218 }, { "epoch": 0.04693915178551353, "grad_norm": 3.2174830436706543, "learning_rate": 2.5256283332732422e-05, "loss": 2.2412, "step": 5219 }, { "epoch": 0.04694814568315398, "grad_norm": 3.9507997035980225, "learning_rate": 2.5248016404199908e-05, "loss": 1.7498, "step": 5220 }, { "epoch": 0.04695713958079443, "grad_norm": 3.5641682147979736, "learning_rate": 2.5239749448544704e-05, "loss": 2.1611, "step": 5221 }, { "epoch": 0.04696613347843488, "grad_norm": 3.2430219650268555, "learning_rate": 2.5231482466670852e-05, "loss": 2.2179, "step": 5222 }, { "epoch": 0.046975127376075336, "grad_norm": 4.94522762298584, "learning_rate": 2.5223215459482425e-05, "loss": 1.9823, "step": 5223 }, { "epoch": 0.04698412127371578, "grad_norm": 4.749909400939941, "learning_rate": 2.5214948427883494e-05, "loss": 2.309, "step": 5224 }, { "epoch": 0.04699311517135624, "grad_norm": 3.255906343460083, "learning_rate": 2.5206681372778124e-05, "loss": 2.4021, "step": 5225 }, { "epoch": 0.047002109068996684, "grad_norm": 4.633800506591797, "learning_rate": 2.5198414295070393e-05, "loss": 1.8647, "step": 5226 }, { "epoch": 0.04701110296663714, "grad_norm": 2.057663917541504, "learning_rate": 2.519014719566437e-05, "loss": 2.1633, "step": 5227 }, { "epoch": 0.047020096864277584, "grad_norm": 4.374258518218994, "learning_rate": 2.5181880075464136e-05, "loss": 2.7557, "step": 5228 }, { "epoch": 0.04702909076191804, "grad_norm": 6.9889044761657715, "learning_rate": 2.5173612935373776e-05, "loss": 2.3907, "step": 5229 }, { "epoch": 0.04703808465955849, "grad_norm": 4.202875137329102, "learning_rate": 2.5165345776297355e-05, "loss": 2.0023, "step": 5230 }, { "epoch": 0.04704707855719894, "grad_norm": 2.744802474975586, "learning_rate": 2.5157078599138977e-05, "loss": 1.9862, "step": 5231 }, { "epoch": 0.04705607245483939, "grad_norm": 7.273871421813965, "learning_rate": 2.514881140480272e-05, "loss": 2.2635, "step": 5232 }, { "epoch": 0.04706506635247984, "grad_norm": 2.189887762069702, "learning_rate": 2.5140544194192665e-05, "loss": 2.1464, "step": 5233 }, { "epoch": 0.04707406025012029, "grad_norm": 4.938686370849609, "learning_rate": 2.5132276968212914e-05, "loss": 2.4847, "step": 5234 }, { "epoch": 0.04708305414776075, "grad_norm": 2.8663625717163086, "learning_rate": 2.512400972776755e-05, "loss": 2.1985, "step": 5235 }, { "epoch": 0.047092048045401194, "grad_norm": 2.718350410461426, "learning_rate": 2.5115742473760668e-05, "loss": 1.8304, "step": 5236 }, { "epoch": 0.04710104194304165, "grad_norm": 4.220222473144531, "learning_rate": 2.5107475207096365e-05, "loss": 2.2845, "step": 5237 }, { "epoch": 0.047110035840682095, "grad_norm": 5.64247989654541, "learning_rate": 2.5099207928678732e-05, "loss": 2.5082, "step": 5238 }, { "epoch": 0.04711902973832255, "grad_norm": 3.0997562408447266, "learning_rate": 2.5090940639411873e-05, "loss": 1.7766, "step": 5239 }, { "epoch": 0.047128023635963, "grad_norm": 10.116399765014648, "learning_rate": 2.508267334019988e-05, "loss": 1.944, "step": 5240 }, { "epoch": 0.04713701753360345, "grad_norm": 3.5021185874938965, "learning_rate": 2.5074406031946855e-05, "loss": 1.8698, "step": 5241 }, { "epoch": 0.0471460114312439, "grad_norm": 3.470512866973877, "learning_rate": 2.5066138715556904e-05, "loss": 1.9464, "step": 5242 }, { "epoch": 0.04715500532888435, "grad_norm": 1.860978603363037, "learning_rate": 2.505787139193411e-05, "loss": 2.0127, "step": 5243 }, { "epoch": 0.047163999226524804, "grad_norm": 4.200709342956543, "learning_rate": 2.5049604061982597e-05, "loss": 1.7976, "step": 5244 }, { "epoch": 0.04717299312416525, "grad_norm": 2.7611818313598633, "learning_rate": 2.5041336726606457e-05, "loss": 2.0532, "step": 5245 }, { "epoch": 0.047181987021805705, "grad_norm": 2.1783173084259033, "learning_rate": 2.5033069386709796e-05, "loss": 1.9965, "step": 5246 }, { "epoch": 0.04719098091944616, "grad_norm": 3.6174590587615967, "learning_rate": 2.5024802043196715e-05, "loss": 1.9129, "step": 5247 }, { "epoch": 0.047199974817086605, "grad_norm": 2.7052974700927734, "learning_rate": 2.5016534696971323e-05, "loss": 2.5129, "step": 5248 }, { "epoch": 0.04720896871472706, "grad_norm": 3.174306869506836, "learning_rate": 2.5008267348937708e-05, "loss": 2.4451, "step": 5249 }, { "epoch": 0.047217962612367506, "grad_norm": 6.680493354797363, "learning_rate": 2.5e-05, "loss": 3.3158, "step": 5250 }, { "epoch": 0.04722695651000796, "grad_norm": 3.5173180103302, "learning_rate": 2.4991732651062294e-05, "loss": 2.3361, "step": 5251 }, { "epoch": 0.047235950407648414, "grad_norm": 4.895158767700195, "learning_rate": 2.4983465303028693e-05, "loss": 2.3037, "step": 5252 }, { "epoch": 0.04724494430528886, "grad_norm": 3.4949095249176025, "learning_rate": 2.4975197956803297e-05, "loss": 2.3885, "step": 5253 }, { "epoch": 0.047253938202929315, "grad_norm": 5.8189778327941895, "learning_rate": 2.4966930613290206e-05, "loss": 2.2376, "step": 5254 }, { "epoch": 0.04726293210056976, "grad_norm": 6.757737159729004, "learning_rate": 2.4958663273393546e-05, "loss": 2.2345, "step": 5255 }, { "epoch": 0.047271925998210215, "grad_norm": 6.926812171936035, "learning_rate": 2.4950395938017405e-05, "loss": 2.2378, "step": 5256 }, { "epoch": 0.04728091989585066, "grad_norm": 2.2063333988189697, "learning_rate": 2.4942128608065894e-05, "loss": 2.3359, "step": 5257 }, { "epoch": 0.047289913793491116, "grad_norm": 3.0184218883514404, "learning_rate": 2.4933861284443112e-05, "loss": 2.1118, "step": 5258 }, { "epoch": 0.04729890769113157, "grad_norm": 1.4949053525924683, "learning_rate": 2.4925593968053147e-05, "loss": 1.8599, "step": 5259 }, { "epoch": 0.04730790158877202, "grad_norm": 2.842205286026001, "learning_rate": 2.4917326659800123e-05, "loss": 2.3191, "step": 5260 }, { "epoch": 0.04731689548641247, "grad_norm": 3.126429796218872, "learning_rate": 2.4909059360588132e-05, "loss": 2.0255, "step": 5261 }, { "epoch": 0.04732588938405292, "grad_norm": 3.0582282543182373, "learning_rate": 2.490079207132127e-05, "loss": 1.5787, "step": 5262 }, { "epoch": 0.04733488328169337, "grad_norm": 3.2342066764831543, "learning_rate": 2.4892524792903637e-05, "loss": 1.79, "step": 5263 }, { "epoch": 0.047343877179333825, "grad_norm": 4.979565620422363, "learning_rate": 2.4884257526239338e-05, "loss": 2.3486, "step": 5264 }, { "epoch": 0.04735287107697427, "grad_norm": 3.8575165271759033, "learning_rate": 2.487599027223246e-05, "loss": 2.2954, "step": 5265 }, { "epoch": 0.047361864974614726, "grad_norm": 6.427733421325684, "learning_rate": 2.4867723031787092e-05, "loss": 1.9714, "step": 5266 }, { "epoch": 0.04737085887225517, "grad_norm": 3.1090545654296875, "learning_rate": 2.4859455805807337e-05, "loss": 1.8274, "step": 5267 }, { "epoch": 0.047379852769895626, "grad_norm": 2.638554811477661, "learning_rate": 2.4851188595197286e-05, "loss": 2.1369, "step": 5268 }, { "epoch": 0.04738884666753607, "grad_norm": 3.2440290451049805, "learning_rate": 2.484292140086103e-05, "loss": 2.2289, "step": 5269 }, { "epoch": 0.04739784056517653, "grad_norm": 2.2045023441314697, "learning_rate": 2.483465422370265e-05, "loss": 2.0658, "step": 5270 }, { "epoch": 0.04740683446281698, "grad_norm": 6.011885166168213, "learning_rate": 2.4826387064626237e-05, "loss": 1.8774, "step": 5271 }, { "epoch": 0.04741582836045743, "grad_norm": 2.9150023460388184, "learning_rate": 2.4818119924535873e-05, "loss": 1.5985, "step": 5272 }, { "epoch": 0.04742482225809788, "grad_norm": 1.7910702228546143, "learning_rate": 2.4809852804335632e-05, "loss": 1.9223, "step": 5273 }, { "epoch": 0.04743381615573833, "grad_norm": 2.3063132762908936, "learning_rate": 2.480158570492961e-05, "loss": 2.2434, "step": 5274 }, { "epoch": 0.04744281005337878, "grad_norm": 1.683037281036377, "learning_rate": 2.4793318627221878e-05, "loss": 2.4168, "step": 5275 }, { "epoch": 0.047451803951019236, "grad_norm": 3.2253754138946533, "learning_rate": 2.4785051572116512e-05, "loss": 1.8739, "step": 5276 }, { "epoch": 0.04746079784865968, "grad_norm": 2.750903367996216, "learning_rate": 2.477678454051758e-05, "loss": 2.0951, "step": 5277 }, { "epoch": 0.04746979174630014, "grad_norm": 2.6179862022399902, "learning_rate": 2.4768517533329157e-05, "loss": 1.9072, "step": 5278 }, { "epoch": 0.047478785643940584, "grad_norm": 2.7825348377227783, "learning_rate": 2.47602505514553e-05, "loss": 1.8224, "step": 5279 }, { "epoch": 0.04748777954158104, "grad_norm": 6.092041492462158, "learning_rate": 2.475198359580009e-05, "loss": 2.2393, "step": 5280 }, { "epoch": 0.04749677343922149, "grad_norm": 4.4628682136535645, "learning_rate": 2.4743716667267584e-05, "loss": 2.2674, "step": 5281 }, { "epoch": 0.04750576733686194, "grad_norm": 2.4769320487976074, "learning_rate": 2.4735449766761835e-05, "loss": 2.4601, "step": 5282 }, { "epoch": 0.04751476123450239, "grad_norm": 2.328380584716797, "learning_rate": 2.4727182895186903e-05, "loss": 2.3771, "step": 5283 }, { "epoch": 0.04752375513214284, "grad_norm": 2.3515284061431885, "learning_rate": 2.4718916053446832e-05, "loss": 2.028, "step": 5284 }, { "epoch": 0.04753274902978329, "grad_norm": 4.798981189727783, "learning_rate": 2.4710649242445686e-05, "loss": 2.071, "step": 5285 }, { "epoch": 0.04754174292742374, "grad_norm": 2.9954745769500732, "learning_rate": 2.4702382463087498e-05, "loss": 1.7393, "step": 5286 }, { "epoch": 0.047550736825064194, "grad_norm": 2.326524019241333, "learning_rate": 2.469411571627632e-05, "loss": 2.0105, "step": 5287 }, { "epoch": 0.04755973072270465, "grad_norm": 2.9254119396209717, "learning_rate": 2.4685849002916183e-05, "loss": 2.5317, "step": 5288 }, { "epoch": 0.047568724620345094, "grad_norm": 3.6786272525787354, "learning_rate": 2.4677582323911135e-05, "loss": 2.5047, "step": 5289 }, { "epoch": 0.04757771851798555, "grad_norm": 3.9505207538604736, "learning_rate": 2.46693156801652e-05, "loss": 2.4171, "step": 5290 }, { "epoch": 0.047586712415625995, "grad_norm": 1.9789010286331177, "learning_rate": 2.46610490725824e-05, "loss": 2.3359, "step": 5291 }, { "epoch": 0.04759570631326645, "grad_norm": 5.911932468414307, "learning_rate": 2.4652782502066755e-05, "loss": 2.2536, "step": 5292 }, { "epoch": 0.0476047002109069, "grad_norm": 1.6799869537353516, "learning_rate": 2.4644515969522306e-05, "loss": 1.9276, "step": 5293 }, { "epoch": 0.04761369410854735, "grad_norm": 3.419605255126953, "learning_rate": 2.4636249475853054e-05, "loss": 2.1169, "step": 5294 }, { "epoch": 0.047622688006187804, "grad_norm": 2.0485999584198, "learning_rate": 2.4627983021963014e-05, "loss": 2.3538, "step": 5295 }, { "epoch": 0.04763168190382825, "grad_norm": 2.850724935531616, "learning_rate": 2.4619716608756187e-05, "loss": 1.9195, "step": 5296 }, { "epoch": 0.047640675801468704, "grad_norm": 4.879880905151367, "learning_rate": 2.4611450237136584e-05, "loss": 1.867, "step": 5297 }, { "epoch": 0.04764966969910915, "grad_norm": 3.289019823074341, "learning_rate": 2.460318390800819e-05, "loss": 2.2771, "step": 5298 }, { "epoch": 0.047658663596749605, "grad_norm": 2.5304980278015137, "learning_rate": 2.4594917622275015e-05, "loss": 1.661, "step": 5299 }, { "epoch": 0.04766765749439006, "grad_norm": 2.8631742000579834, "learning_rate": 2.458665138084104e-05, "loss": 2.2638, "step": 5300 }, { "epoch": 0.047676651392030506, "grad_norm": 1.8742833137512207, "learning_rate": 2.457838518461025e-05, "loss": 2.0134, "step": 5301 }, { "epoch": 0.04768564528967096, "grad_norm": 5.009515285491943, "learning_rate": 2.457011903448662e-05, "loss": 1.9433, "step": 5302 }, { "epoch": 0.047694639187311406, "grad_norm": 14.233297348022461, "learning_rate": 2.4561852931374127e-05, "loss": 2.2043, "step": 5303 }, { "epoch": 0.04770363308495186, "grad_norm": 2.9027493000030518, "learning_rate": 2.455358687617674e-05, "loss": 1.8692, "step": 5304 }, { "epoch": 0.047712626982592314, "grad_norm": 3.113527536392212, "learning_rate": 2.4545320869798406e-05, "loss": 1.9993, "step": 5305 }, { "epoch": 0.04772162088023276, "grad_norm": 2.850205421447754, "learning_rate": 2.4537054913143113e-05, "loss": 2.0994, "step": 5306 }, { "epoch": 0.047730614777873215, "grad_norm": 2.7645747661590576, "learning_rate": 2.45287890071148e-05, "loss": 2.3802, "step": 5307 }, { "epoch": 0.04773960867551366, "grad_norm": 3.2508862018585205, "learning_rate": 2.452052315261741e-05, "loss": 1.9966, "step": 5308 }, { "epoch": 0.047748602573154116, "grad_norm": 4.195789813995361, "learning_rate": 2.4512257350554883e-05, "loss": 2.3305, "step": 5309 }, { "epoch": 0.04775759647079456, "grad_norm": 6.1123809814453125, "learning_rate": 2.4503991601831163e-05, "loss": 2.037, "step": 5310 }, { "epoch": 0.047766590368435016, "grad_norm": 4.3020195960998535, "learning_rate": 2.449572590735017e-05, "loss": 2.1952, "step": 5311 }, { "epoch": 0.04777558426607547, "grad_norm": 2.5504512786865234, "learning_rate": 2.448746026801584e-05, "loss": 2.1532, "step": 5312 }, { "epoch": 0.04778457816371592, "grad_norm": 2.9805588722229004, "learning_rate": 2.4479194684732078e-05, "loss": 2.1907, "step": 5313 }, { "epoch": 0.04779357206135637, "grad_norm": 3.5373525619506836, "learning_rate": 2.447092915840281e-05, "loss": 2.1366, "step": 5314 }, { "epoch": 0.04780256595899682, "grad_norm": 3.7605185508728027, "learning_rate": 2.4462663689931935e-05, "loss": 2.123, "step": 5315 }, { "epoch": 0.04781155985663727, "grad_norm": 5.844115257263184, "learning_rate": 2.4454398280223346e-05, "loss": 2.5254, "step": 5316 }, { "epoch": 0.047820553754277725, "grad_norm": 1.7230080366134644, "learning_rate": 2.444613293018093e-05, "loss": 1.8986, "step": 5317 }, { "epoch": 0.04782954765191817, "grad_norm": 2.14532470703125, "learning_rate": 2.44378676407086e-05, "loss": 1.7131, "step": 5318 }, { "epoch": 0.047838541549558626, "grad_norm": 1.8961410522460938, "learning_rate": 2.4429602412710217e-05, "loss": 2.2916, "step": 5319 }, { "epoch": 0.04784753544719907, "grad_norm": 3.9445865154266357, "learning_rate": 2.4421337247089655e-05, "loss": 1.8038, "step": 5320 }, { "epoch": 0.04785652934483953, "grad_norm": 9.60340404510498, "learning_rate": 2.4413072144750784e-05, "loss": 2.3388, "step": 5321 }, { "epoch": 0.04786552324247998, "grad_norm": 1.927873134613037, "learning_rate": 2.440480710659746e-05, "loss": 2.3024, "step": 5322 }, { "epoch": 0.04787451714012043, "grad_norm": 2.6220614910125732, "learning_rate": 2.4396542133533535e-05, "loss": 1.9913, "step": 5323 }, { "epoch": 0.04788351103776088, "grad_norm": 2.5516958236694336, "learning_rate": 2.4388277226462846e-05, "loss": 1.9497, "step": 5324 }, { "epoch": 0.04789250493540133, "grad_norm": 1.9934635162353516, "learning_rate": 2.438001238628925e-05, "loss": 2.0405, "step": 5325 }, { "epoch": 0.04790149883304178, "grad_norm": 1.8788278102874756, "learning_rate": 2.4371747613916566e-05, "loss": 2.258, "step": 5326 }, { "epoch": 0.04791049273068223, "grad_norm": 2.258190155029297, "learning_rate": 2.436348291024862e-05, "loss": 1.9449, "step": 5327 }, { "epoch": 0.04791948662832268, "grad_norm": 3.244778871536255, "learning_rate": 2.4355218276189224e-05, "loss": 1.8686, "step": 5328 }, { "epoch": 0.04792848052596314, "grad_norm": 3.4548490047454834, "learning_rate": 2.434695371264219e-05, "loss": 2.3787, "step": 5329 }, { "epoch": 0.047937474423603584, "grad_norm": 2.720233201980591, "learning_rate": 2.4338689220511304e-05, "loss": 2.2649, "step": 5330 }, { "epoch": 0.04794646832124404, "grad_norm": 3.9217426776885986, "learning_rate": 2.4330424800700383e-05, "loss": 2.1771, "step": 5331 }, { "epoch": 0.047955462218884484, "grad_norm": 1.9925175905227661, "learning_rate": 2.4322160454113197e-05, "loss": 2.2931, "step": 5332 }, { "epoch": 0.04796445611652494, "grad_norm": 3.504194498062134, "learning_rate": 2.4313896181653524e-05, "loss": 2.0515, "step": 5333 }, { "epoch": 0.04797345001416539, "grad_norm": 2.531358242034912, "learning_rate": 2.430563198422513e-05, "loss": 1.8313, "step": 5334 }, { "epoch": 0.04798244391180584, "grad_norm": 2.0105717182159424, "learning_rate": 2.4297367862731783e-05, "loss": 2.4621, "step": 5335 }, { "epoch": 0.04799143780944629, "grad_norm": 4.867953777313232, "learning_rate": 2.4289103818077223e-05, "loss": 2.4881, "step": 5336 }, { "epoch": 0.04800043170708674, "grad_norm": 6.652677536010742, "learning_rate": 2.428083985116521e-05, "loss": 2.1671, "step": 5337 }, { "epoch": 0.04800942560472719, "grad_norm": 2.7060623168945312, "learning_rate": 2.4272575962899462e-05, "loss": 2.096, "step": 5338 }, { "epoch": 0.04801841950236764, "grad_norm": 4.699619293212891, "learning_rate": 2.426431215418372e-05, "loss": 2.2551, "step": 5339 }, { "epoch": 0.048027413400008094, "grad_norm": 13.786131858825684, "learning_rate": 2.425604842592169e-05, "loss": 2.9586, "step": 5340 }, { "epoch": 0.04803640729764855, "grad_norm": 3.159823179244995, "learning_rate": 2.424778477901709e-05, "loss": 2.3953, "step": 5341 }, { "epoch": 0.048045401195288995, "grad_norm": 2.784008741378784, "learning_rate": 2.4239521214373615e-05, "loss": 1.8158, "step": 5342 }, { "epoch": 0.04805439509292945, "grad_norm": 3.545138120651245, "learning_rate": 2.4231257732894944e-05, "loss": 2.4711, "step": 5343 }, { "epoch": 0.048063388990569895, "grad_norm": 2.5569870471954346, "learning_rate": 2.4222994335484782e-05, "loss": 2.254, "step": 5344 }, { "epoch": 0.04807238288821035, "grad_norm": 2.1440048217773438, "learning_rate": 2.4214731023046793e-05, "loss": 1.781, "step": 5345 }, { "epoch": 0.0480813767858508, "grad_norm": 2.8576953411102295, "learning_rate": 2.420646779648464e-05, "loss": 1.8625, "step": 5346 }, { "epoch": 0.04809037068349125, "grad_norm": 2.4118130207061768, "learning_rate": 2.4198204656701973e-05, "loss": 2.1927, "step": 5347 }, { "epoch": 0.048099364581131704, "grad_norm": 2.8289201259613037, "learning_rate": 2.4189941604602444e-05, "loss": 2.5173, "step": 5348 }, { "epoch": 0.04810835847877215, "grad_norm": 3.0263233184814453, "learning_rate": 2.4181678641089665e-05, "loss": 2.014, "step": 5349 }, { "epoch": 0.048117352376412605, "grad_norm": 3.7633090019226074, "learning_rate": 2.4173415767067297e-05, "loss": 2.1524, "step": 5350 }, { "epoch": 0.04812634627405306, "grad_norm": 2.8044726848602295, "learning_rate": 2.4165152983438932e-05, "loss": 2.2298, "step": 5351 }, { "epoch": 0.048135340171693505, "grad_norm": 2.3959364891052246, "learning_rate": 2.4156890291108187e-05, "loss": 1.9877, "step": 5352 }, { "epoch": 0.04814433406933396, "grad_norm": 3.6083433628082275, "learning_rate": 2.4148627690978648e-05, "loss": 1.9409, "step": 5353 }, { "epoch": 0.048153327966974406, "grad_norm": 2.33804988861084, "learning_rate": 2.4140365183953907e-05, "loss": 1.794, "step": 5354 }, { "epoch": 0.04816232186461486, "grad_norm": 4.5791401863098145, "learning_rate": 2.413210277093753e-05, "loss": 2.2127, "step": 5355 }, { "epoch": 0.04817131576225531, "grad_norm": 3.424569845199585, "learning_rate": 2.412384045283309e-05, "loss": 2.3878, "step": 5356 }, { "epoch": 0.04818030965989576, "grad_norm": 2.497850179672241, "learning_rate": 2.4115578230544147e-05, "loss": 2.2665, "step": 5357 }, { "epoch": 0.048189303557536214, "grad_norm": 2.0800631046295166, "learning_rate": 2.410731610497424e-05, "loss": 1.7503, "step": 5358 }, { "epoch": 0.04819829745517666, "grad_norm": 3.109602212905884, "learning_rate": 2.4099054077026895e-05, "loss": 1.9069, "step": 5359 }, { "epoch": 0.048207291352817115, "grad_norm": 5.146254539489746, "learning_rate": 2.4090792147605647e-05, "loss": 1.807, "step": 5360 }, { "epoch": 0.04821628525045756, "grad_norm": 2.965226173400879, "learning_rate": 2.4082530317614e-05, "loss": 1.9998, "step": 5361 }, { "epoch": 0.048225279148098016, "grad_norm": 3.2227718830108643, "learning_rate": 2.4074268587955455e-05, "loss": 2.3817, "step": 5362 }, { "epoch": 0.04823427304573847, "grad_norm": 2.9907608032226562, "learning_rate": 2.4066006959533508e-05, "loss": 2.3322, "step": 5363 }, { "epoch": 0.04824326694337892, "grad_norm": 4.429880619049072, "learning_rate": 2.4057745433251635e-05, "loss": 1.9186, "step": 5364 }, { "epoch": 0.04825226084101937, "grad_norm": 8.894227027893066, "learning_rate": 2.404948401001331e-05, "loss": 2.0002, "step": 5365 }, { "epoch": 0.04826125473865982, "grad_norm": 3.358358383178711, "learning_rate": 2.404122269072198e-05, "loss": 1.85, "step": 5366 }, { "epoch": 0.04827024863630027, "grad_norm": 2.403371810913086, "learning_rate": 2.4032961476281102e-05, "loss": 1.9982, "step": 5367 }, { "epoch": 0.04827924253394072, "grad_norm": 2.983062744140625, "learning_rate": 2.4024700367594085e-05, "loss": 2.5672, "step": 5368 }, { "epoch": 0.04828823643158117, "grad_norm": 2.356732130050659, "learning_rate": 2.4016439365564386e-05, "loss": 2.585, "step": 5369 }, { "epoch": 0.048297230329221626, "grad_norm": 2.3386690616607666, "learning_rate": 2.4008178471095397e-05, "loss": 2.4624, "step": 5370 }, { "epoch": 0.04830622422686207, "grad_norm": 3.289454460144043, "learning_rate": 2.3999917685090517e-05, "loss": 2.3192, "step": 5371 }, { "epoch": 0.048315218124502526, "grad_norm": 3.0392720699310303, "learning_rate": 2.399165700845314e-05, "loss": 1.7423, "step": 5372 }, { "epoch": 0.04832421202214297, "grad_norm": 3.3981945514678955, "learning_rate": 2.398339644208663e-05, "loss": 2.233, "step": 5373 }, { "epoch": 0.04833320591978343, "grad_norm": 2.0103089809417725, "learning_rate": 2.3975135986894363e-05, "loss": 2.1323, "step": 5374 }, { "epoch": 0.04834219981742388, "grad_norm": 3.4900002479553223, "learning_rate": 2.3966875643779667e-05, "loss": 1.43, "step": 5375 }, { "epoch": 0.04835119371506433, "grad_norm": 2.9265780448913574, "learning_rate": 2.3958615413645913e-05, "loss": 2.1022, "step": 5376 }, { "epoch": 0.04836018761270478, "grad_norm": 2.2034285068511963, "learning_rate": 2.395035529739641e-05, "loss": 2.1167, "step": 5377 }, { "epoch": 0.04836918151034523, "grad_norm": 5.890878677368164, "learning_rate": 2.3942095295934476e-05, "loss": 1.8231, "step": 5378 }, { "epoch": 0.04837817540798568, "grad_norm": 2.6438307762145996, "learning_rate": 2.3933835410163406e-05, "loss": 2.3971, "step": 5379 }, { "epoch": 0.04838716930562613, "grad_norm": 2.4781956672668457, "learning_rate": 2.392557564098649e-05, "loss": 1.9917, "step": 5380 }, { "epoch": 0.04839616320326658, "grad_norm": 3.1769213676452637, "learning_rate": 2.3917315989307005e-05, "loss": 2.6427, "step": 5381 }, { "epoch": 0.04840515710090704, "grad_norm": 3.944859266281128, "learning_rate": 2.390905645602822e-05, "loss": 1.4327, "step": 5382 }, { "epoch": 0.048414150998547484, "grad_norm": 2.295292377471924, "learning_rate": 2.3900797042053382e-05, "loss": 2.39, "step": 5383 }, { "epoch": 0.04842314489618794, "grad_norm": 2.808159828186035, "learning_rate": 2.389253774828572e-05, "loss": 1.8653, "step": 5384 }, { "epoch": 0.048432138793828385, "grad_norm": 2.333935499191284, "learning_rate": 2.388427857562847e-05, "loss": 2.0756, "step": 5385 }, { "epoch": 0.04844113269146884, "grad_norm": 3.895603656768799, "learning_rate": 2.3876019524984843e-05, "loss": 2.2174, "step": 5386 }, { "epoch": 0.04845012658910929, "grad_norm": 3.374799966812134, "learning_rate": 2.3867760597258014e-05, "loss": 2.5072, "step": 5387 }, { "epoch": 0.04845912048674974, "grad_norm": 3.0886595249176025, "learning_rate": 2.3859501793351198e-05, "loss": 2.164, "step": 5388 }, { "epoch": 0.04846811438439019, "grad_norm": 2.8621745109558105, "learning_rate": 2.385124311416755e-05, "loss": 2.0581, "step": 5389 }, { "epoch": 0.04847710828203064, "grad_norm": 1.9633325338363647, "learning_rate": 2.3842984560610228e-05, "loss": 1.9721, "step": 5390 }, { "epoch": 0.048486102179671094, "grad_norm": 3.3031275272369385, "learning_rate": 2.3834726133582373e-05, "loss": 2.221, "step": 5391 }, { "epoch": 0.04849509607731155, "grad_norm": 6.158144950866699, "learning_rate": 2.382646783398712e-05, "loss": 2.0921, "step": 5392 }, { "epoch": 0.048504089974951994, "grad_norm": 1.8536171913146973, "learning_rate": 2.3818209662727577e-05, "loss": 2.0809, "step": 5393 }, { "epoch": 0.04851308387259245, "grad_norm": 3.095855474472046, "learning_rate": 2.3809951620706838e-05, "loss": 2.3671, "step": 5394 }, { "epoch": 0.048522077770232895, "grad_norm": 2.613511085510254, "learning_rate": 2.3801693708828013e-05, "loss": 1.6997, "step": 5395 }, { "epoch": 0.04853107166787335, "grad_norm": 3.0460751056671143, "learning_rate": 2.3793435927994163e-05, "loss": 2.1881, "step": 5396 }, { "epoch": 0.048540065565513796, "grad_norm": 4.098634243011475, "learning_rate": 2.3785178279108344e-05, "loss": 1.7689, "step": 5397 }, { "epoch": 0.04854905946315425, "grad_norm": 3.723113775253296, "learning_rate": 2.3776920763073602e-05, "loss": 1.6491, "step": 5398 }, { "epoch": 0.0485580533607947, "grad_norm": 4.636134624481201, "learning_rate": 2.3768663380792966e-05, "loss": 2.1667, "step": 5399 }, { "epoch": 0.04856704725843515, "grad_norm": 4.21294641494751, "learning_rate": 2.3760406133169443e-05, "loss": 2.0556, "step": 5400 }, { "epoch": 0.048576041156075604, "grad_norm": 1.774176836013794, "learning_rate": 2.375214902110605e-05, "loss": 2.184, "step": 5401 }, { "epoch": 0.04858503505371605, "grad_norm": 1.7265851497650146, "learning_rate": 2.3743892045505764e-05, "loss": 1.9366, "step": 5402 }, { "epoch": 0.048594028951356505, "grad_norm": 2.9765262603759766, "learning_rate": 2.3735635207271557e-05, "loss": 2.4802, "step": 5403 }, { "epoch": 0.04860302284899696, "grad_norm": 2.2710113525390625, "learning_rate": 2.372737850730638e-05, "loss": 2.2691, "step": 5404 }, { "epoch": 0.048612016746637406, "grad_norm": 1.5800433158874512, "learning_rate": 2.3719121946513174e-05, "loss": 1.897, "step": 5405 }, { "epoch": 0.04862101064427786, "grad_norm": 1.8049923181533813, "learning_rate": 2.3710865525794866e-05, "loss": 1.7613, "step": 5406 }, { "epoch": 0.048630004541918306, "grad_norm": 2.428060531616211, "learning_rate": 2.370260924605437e-05, "loss": 2.1922, "step": 5407 }, { "epoch": 0.04863899843955876, "grad_norm": 2.4325525760650635, "learning_rate": 2.369435310819457e-05, "loss": 2.109, "step": 5408 }, { "epoch": 0.04864799233719921, "grad_norm": 3.4525303840637207, "learning_rate": 2.368609711311836e-05, "loss": 2.4386, "step": 5409 }, { "epoch": 0.04865698623483966, "grad_norm": 2.445183277130127, "learning_rate": 2.367784126172859e-05, "loss": 2.5194, "step": 5410 }, { "epoch": 0.048665980132480115, "grad_norm": 1.8727835416793823, "learning_rate": 2.3669585554928113e-05, "loss": 2.2596, "step": 5411 }, { "epoch": 0.04867497403012056, "grad_norm": 1.8058757781982422, "learning_rate": 2.3661329993619757e-05, "loss": 2.2723, "step": 5412 }, { "epoch": 0.048683967927761015, "grad_norm": 3.282318115234375, "learning_rate": 2.3653074578706334e-05, "loss": 2.3518, "step": 5413 }, { "epoch": 0.04869296182540146, "grad_norm": 7.992863655090332, "learning_rate": 2.364481931109066e-05, "loss": 2.0644, "step": 5414 }, { "epoch": 0.048701955723041916, "grad_norm": 3.006014823913574, "learning_rate": 2.3636564191675507e-05, "loss": 1.716, "step": 5415 }, { "epoch": 0.04871094962068237, "grad_norm": 3.694652557373047, "learning_rate": 2.3628309221363645e-05, "loss": 2.0491, "step": 5416 }, { "epoch": 0.04871994351832282, "grad_norm": 7.009035587310791, "learning_rate": 2.3620054401057824e-05, "loss": 2.3212, "step": 5417 }, { "epoch": 0.04872893741596327, "grad_norm": 2.9396142959594727, "learning_rate": 2.361179973166078e-05, "loss": 2.3574, "step": 5418 }, { "epoch": 0.04873793131360372, "grad_norm": 3.475750684738159, "learning_rate": 2.360354521407522e-05, "loss": 2.3372, "step": 5419 }, { "epoch": 0.04874692521124417, "grad_norm": 4.415957450866699, "learning_rate": 2.3595290849203862e-05, "loss": 1.894, "step": 5420 }, { "epoch": 0.04875591910888462, "grad_norm": 3.0996010303497314, "learning_rate": 2.3587036637949388e-05, "loss": 2.3966, "step": 5421 }, { "epoch": 0.04876491300652507, "grad_norm": 2.775540351867676, "learning_rate": 2.3578782581214465e-05, "loss": 2.6899, "step": 5422 }, { "epoch": 0.048773906904165526, "grad_norm": 3.382181406021118, "learning_rate": 2.3570528679901736e-05, "loss": 1.9777, "step": 5423 }, { "epoch": 0.04878290080180597, "grad_norm": 3.229346990585327, "learning_rate": 2.3562274934913847e-05, "loss": 2.0714, "step": 5424 }, { "epoch": 0.04879189469944643, "grad_norm": 1.799443006515503, "learning_rate": 2.3554021347153403e-05, "loss": 2.0035, "step": 5425 }, { "epoch": 0.048800888597086874, "grad_norm": 2.766671657562256, "learning_rate": 2.354576791752301e-05, "loss": 2.4733, "step": 5426 }, { "epoch": 0.04880988249472733, "grad_norm": 4.457692623138428, "learning_rate": 2.3537514646925257e-05, "loss": 2.2045, "step": 5427 }, { "epoch": 0.04881887639236778, "grad_norm": 3.1665737628936768, "learning_rate": 2.3529261536262703e-05, "loss": 1.8398, "step": 5428 }, { "epoch": 0.04882787029000823, "grad_norm": 4.0099005699157715, "learning_rate": 2.352100858643789e-05, "loss": 2.1203, "step": 5429 }, { "epoch": 0.04883686418764868, "grad_norm": 2.3747658729553223, "learning_rate": 2.351275579835336e-05, "loss": 1.9948, "step": 5430 }, { "epoch": 0.04884585808528913, "grad_norm": 3.1279828548431396, "learning_rate": 2.3504503172911622e-05, "loss": 2.7378, "step": 5431 }, { "epoch": 0.04885485198292958, "grad_norm": 2.403646230697632, "learning_rate": 2.349625071101516e-05, "loss": 1.9595, "step": 5432 }, { "epoch": 0.048863845880570037, "grad_norm": 2.256256341934204, "learning_rate": 2.3487998413566463e-05, "loss": 1.7165, "step": 5433 }, { "epoch": 0.04887283977821048, "grad_norm": 3.6985127925872803, "learning_rate": 2.347974628146799e-05, "loss": 2.3598, "step": 5434 }, { "epoch": 0.04888183367585094, "grad_norm": 1.4005361795425415, "learning_rate": 2.3471494315622177e-05, "loss": 2.0822, "step": 5435 }, { "epoch": 0.048890827573491384, "grad_norm": 4.587813854217529, "learning_rate": 2.3463242516931445e-05, "loss": 2.3492, "step": 5436 }, { "epoch": 0.04889982147113184, "grad_norm": 4.426910877227783, "learning_rate": 2.3454990886298207e-05, "loss": 1.7709, "step": 5437 }, { "epoch": 0.048908815368772285, "grad_norm": 4.768750190734863, "learning_rate": 2.3446739424624827e-05, "loss": 2.4654, "step": 5438 }, { "epoch": 0.04891780926641274, "grad_norm": 4.827560901641846, "learning_rate": 2.3438488132813702e-05, "loss": 2.0916, "step": 5439 }, { "epoch": 0.04892680316405319, "grad_norm": 4.710875988006592, "learning_rate": 2.3430237011767167e-05, "loss": 2.1007, "step": 5440 }, { "epoch": 0.04893579706169364, "grad_norm": 2.6409919261932373, "learning_rate": 2.3421986062387553e-05, "loss": 2.049, "step": 5441 }, { "epoch": 0.04894479095933409, "grad_norm": 3.3410146236419678, "learning_rate": 2.3413735285577167e-05, "loss": 2.1834, "step": 5442 }, { "epoch": 0.04895378485697454, "grad_norm": 3.3189871311187744, "learning_rate": 2.3405484682238308e-05, "loss": 1.8763, "step": 5443 }, { "epoch": 0.048962778754614994, "grad_norm": 4.917455673217773, "learning_rate": 2.3397234253273246e-05, "loss": 2.4555, "step": 5444 }, { "epoch": 0.04897177265225545, "grad_norm": 3.4471747875213623, "learning_rate": 2.3388983999584224e-05, "loss": 1.9164, "step": 5445 }, { "epoch": 0.048980766549895895, "grad_norm": 4.46656608581543, "learning_rate": 2.3380733922073502e-05, "loss": 2.058, "step": 5446 }, { "epoch": 0.04898976044753635, "grad_norm": 1.9630366563796997, "learning_rate": 2.3372484021643284e-05, "loss": 1.9026, "step": 5447 }, { "epoch": 0.048998754345176795, "grad_norm": 3.081136465072632, "learning_rate": 2.3364234299195766e-05, "loss": 1.9585, "step": 5448 }, { "epoch": 0.04900774824281725, "grad_norm": 4.311786651611328, "learning_rate": 2.335598475563313e-05, "loss": 2.4354, "step": 5449 }, { "epoch": 0.049016742140457696, "grad_norm": 4.343419075012207, "learning_rate": 2.334773539185752e-05, "loss": 2.0467, "step": 5450 }, { "epoch": 0.04902573603809815, "grad_norm": 2.879625082015991, "learning_rate": 2.333948620877108e-05, "loss": 2.1983, "step": 5451 }, { "epoch": 0.049034729935738604, "grad_norm": 2.7528014183044434, "learning_rate": 2.333123720727594e-05, "loss": 2.0662, "step": 5452 }, { "epoch": 0.04904372383337905, "grad_norm": 4.434741973876953, "learning_rate": 2.332298838827419e-05, "loss": 2.1103, "step": 5453 }, { "epoch": 0.049052717731019504, "grad_norm": 2.832693338394165, "learning_rate": 2.3314739752667906e-05, "loss": 3.072, "step": 5454 }, { "epoch": 0.04906171162865995, "grad_norm": 2.71280574798584, "learning_rate": 2.330649130135915e-05, "loss": 2.0718, "step": 5455 }, { "epoch": 0.049070705526300405, "grad_norm": 2.358755111694336, "learning_rate": 2.329824303524996e-05, "loss": 2.1803, "step": 5456 }, { "epoch": 0.04907969942394086, "grad_norm": 3.542771339416504, "learning_rate": 2.328999495524235e-05, "loss": 1.9077, "step": 5457 }, { "epoch": 0.049088693321581306, "grad_norm": 3.012808084487915, "learning_rate": 2.328174706223832e-05, "loss": 2.182, "step": 5458 }, { "epoch": 0.04909768721922176, "grad_norm": 5.5403151512146, "learning_rate": 2.3273499357139855e-05, "loss": 2.2283, "step": 5459 }, { "epoch": 0.04910668111686221, "grad_norm": 2.4065191745758057, "learning_rate": 2.32652518408489e-05, "loss": 2.4979, "step": 5460 }, { "epoch": 0.04911567501450266, "grad_norm": 3.415018320083618, "learning_rate": 2.32570045142674e-05, "loss": 2.1514, "step": 5461 }, { "epoch": 0.049124668912143114, "grad_norm": 4.950087070465088, "learning_rate": 2.3248757378297263e-05, "loss": 2.3632, "step": 5462 }, { "epoch": 0.04913366280978356, "grad_norm": 3.0695641040802, "learning_rate": 2.324051043384039e-05, "loss": 2.2611, "step": 5463 }, { "epoch": 0.049142656707424015, "grad_norm": 2.546445846557617, "learning_rate": 2.3232263681798632e-05, "loss": 2.3013, "step": 5464 }, { "epoch": 0.04915165060506446, "grad_norm": 3.3812448978424072, "learning_rate": 2.3224017123073877e-05, "loss": 1.9251, "step": 5465 }, { "epoch": 0.049160644502704916, "grad_norm": 2.1788830757141113, "learning_rate": 2.321577075856794e-05, "loss": 2.3168, "step": 5466 }, { "epoch": 0.04916963840034536, "grad_norm": 3.114288091659546, "learning_rate": 2.3207524589182627e-05, "loss": 2.1298, "step": 5467 }, { "epoch": 0.049178632297985816, "grad_norm": 2.275458335876465, "learning_rate": 2.3199278615819732e-05, "loss": 2.4491, "step": 5468 }, { "epoch": 0.04918762619562627, "grad_norm": 1.9607385396957397, "learning_rate": 2.3191032839381017e-05, "loss": 2.3724, "step": 5469 }, { "epoch": 0.04919662009326672, "grad_norm": 3.2227425575256348, "learning_rate": 2.3182787260768223e-05, "loss": 1.7968, "step": 5470 }, { "epoch": 0.04920561399090717, "grad_norm": 4.814120769500732, "learning_rate": 2.3174541880883093e-05, "loss": 2.3189, "step": 5471 }, { "epoch": 0.04921460788854762, "grad_norm": 2.561630964279175, "learning_rate": 2.3166296700627315e-05, "loss": 2.6418, "step": 5472 }, { "epoch": 0.04922360178618807, "grad_norm": 3.4128801822662354, "learning_rate": 2.3158051720902576e-05, "loss": 2.808, "step": 5473 }, { "epoch": 0.049232595683828526, "grad_norm": 3.3458471298217773, "learning_rate": 2.3149806942610532e-05, "loss": 2.3149, "step": 5474 }, { "epoch": 0.04924158958146897, "grad_norm": 5.735185623168945, "learning_rate": 2.3141562366652812e-05, "loss": 1.8824, "step": 5475 }, { "epoch": 0.049250583479109426, "grad_norm": 5.062166213989258, "learning_rate": 2.3133317993931043e-05, "loss": 2.0615, "step": 5476 }, { "epoch": 0.04925957737674987, "grad_norm": 2.5643675327301025, "learning_rate": 2.3125073825346806e-05, "loss": 1.7581, "step": 5477 }, { "epoch": 0.04926857127439033, "grad_norm": 2.5298280715942383, "learning_rate": 2.3116829861801686e-05, "loss": 1.7152, "step": 5478 }, { "epoch": 0.049277565172030774, "grad_norm": 2.80416202545166, "learning_rate": 2.3108586104197212e-05, "loss": 2.1383, "step": 5479 }, { "epoch": 0.04928655906967123, "grad_norm": 3.3138179779052734, "learning_rate": 2.3100342553434924e-05, "loss": 2.6339, "step": 5480 }, { "epoch": 0.04929555296731168, "grad_norm": 4.682737827301025, "learning_rate": 2.309209921041632e-05, "loss": 1.6084, "step": 5481 }, { "epoch": 0.04930454686495213, "grad_norm": 2.6634533405303955, "learning_rate": 2.308385607604287e-05, "loss": 2.2519, "step": 5482 }, { "epoch": 0.04931354076259258, "grad_norm": 3.9833953380584717, "learning_rate": 2.3075613151216037e-05, "loss": 1.6186, "step": 5483 }, { "epoch": 0.04932253466023303, "grad_norm": 3.0456702709198, "learning_rate": 2.3067370436837263e-05, "loss": 1.9612, "step": 5484 }, { "epoch": 0.04933152855787348, "grad_norm": 2.569431781768799, "learning_rate": 2.3059127933807954e-05, "loss": 1.9621, "step": 5485 }, { "epoch": 0.04934052245551394, "grad_norm": 3.0283961296081543, "learning_rate": 2.3050885643029496e-05, "loss": 2.1438, "step": 5486 }, { "epoch": 0.049349516353154384, "grad_norm": 3.1129918098449707, "learning_rate": 2.3042643565403254e-05, "loss": 1.7259, "step": 5487 }, { "epoch": 0.04935851025079484, "grad_norm": 2.7981653213500977, "learning_rate": 2.3034401701830565e-05, "loss": 1.7874, "step": 5488 }, { "epoch": 0.049367504148435284, "grad_norm": 3.7079830169677734, "learning_rate": 2.3026160053212745e-05, "loss": 2.089, "step": 5489 }, { "epoch": 0.04937649804607574, "grad_norm": 2.388298988342285, "learning_rate": 2.30179186204511e-05, "loss": 2.3287, "step": 5490 }, { "epoch": 0.049385491943716185, "grad_norm": 2.5397965908050537, "learning_rate": 2.30096774044469e-05, "loss": 2.3077, "step": 5491 }, { "epoch": 0.04939448584135664, "grad_norm": 3.3680670261383057, "learning_rate": 2.3001436406101383e-05, "loss": 2.0484, "step": 5492 }, { "epoch": 0.04940347973899709, "grad_norm": 2.922257423400879, "learning_rate": 2.2993195626315773e-05, "loss": 2.4967, "step": 5493 }, { "epoch": 0.04941247363663754, "grad_norm": 4.1911845207214355, "learning_rate": 2.2984955065991278e-05, "loss": 2.5361, "step": 5494 }, { "epoch": 0.049421467534277994, "grad_norm": 5.345545768737793, "learning_rate": 2.2976714726029068e-05, "loss": 1.8836, "step": 5495 }, { "epoch": 0.04943046143191844, "grad_norm": 3.6806275844573975, "learning_rate": 2.2968474607330278e-05, "loss": 1.8931, "step": 5496 }, { "epoch": 0.049439455329558894, "grad_norm": 3.3941030502319336, "learning_rate": 2.2960234710796063e-05, "loss": 2.3155, "step": 5497 }, { "epoch": 0.04944844922719935, "grad_norm": 2.0948293209075928, "learning_rate": 2.2951995037327515e-05, "loss": 2.0306, "step": 5498 }, { "epoch": 0.049457443124839795, "grad_norm": 4.0290374755859375, "learning_rate": 2.294375558782571e-05, "loss": 1.9052, "step": 5499 }, { "epoch": 0.04946643702248025, "grad_norm": 6.398050308227539, "learning_rate": 2.2935516363191693e-05, "loss": 1.7656, "step": 5500 }, { "epoch": 0.049475430920120696, "grad_norm": 3.170560121536255, "learning_rate": 2.2927277364326513e-05, "loss": 2.3323, "step": 5501 }, { "epoch": 0.04948442481776115, "grad_norm": 2.547053337097168, "learning_rate": 2.2919038592131158e-05, "loss": 2.2305, "step": 5502 }, { "epoch": 0.0494934187154016, "grad_norm": 2.8269097805023193, "learning_rate": 2.2910800047506616e-05, "loss": 1.846, "step": 5503 }, { "epoch": 0.04950241261304205, "grad_norm": 2.262402057647705, "learning_rate": 2.2902561731353837e-05, "loss": 2.5262, "step": 5504 }, { "epoch": 0.049511406510682504, "grad_norm": 2.268751859664917, "learning_rate": 2.2894323644573758e-05, "loss": 2.0285, "step": 5505 }, { "epoch": 0.04952040040832295, "grad_norm": 4.315337181091309, "learning_rate": 2.288608578806728e-05, "loss": 2.4349, "step": 5506 }, { "epoch": 0.049529394305963405, "grad_norm": 2.712533473968506, "learning_rate": 2.287784816273528e-05, "loss": 2.2132, "step": 5507 }, { "epoch": 0.04953838820360385, "grad_norm": 5.998863697052002, "learning_rate": 2.2869610769478607e-05, "loss": 1.7513, "step": 5508 }, { "epoch": 0.049547382101244306, "grad_norm": 3.932112216949463, "learning_rate": 2.2861373609198107e-05, "loss": 2.2457, "step": 5509 }, { "epoch": 0.04955637599888476, "grad_norm": 3.044017791748047, "learning_rate": 2.2853136682794575e-05, "loss": 1.5878, "step": 5510 }, { "epoch": 0.049565369896525206, "grad_norm": 2.1376829147338867, "learning_rate": 2.2844899991168786e-05, "loss": 1.9677, "step": 5511 }, { "epoch": 0.04957436379416566, "grad_norm": 2.549593210220337, "learning_rate": 2.283666353522149e-05, "loss": 1.746, "step": 5512 }, { "epoch": 0.04958335769180611, "grad_norm": 5.035614967346191, "learning_rate": 2.2828427315853423e-05, "loss": 2.0681, "step": 5513 }, { "epoch": 0.04959235158944656, "grad_norm": 3.6032819747924805, "learning_rate": 2.2820191333965278e-05, "loss": 2.2554, "step": 5514 }, { "epoch": 0.049601345487087015, "grad_norm": 2.4953274726867676, "learning_rate": 2.281195559045772e-05, "loss": 2.1139, "step": 5515 }, { "epoch": 0.04961033938472746, "grad_norm": 2.676811695098877, "learning_rate": 2.280372008623142e-05, "loss": 2.6198, "step": 5516 }, { "epoch": 0.049619333282367915, "grad_norm": 3.082592248916626, "learning_rate": 2.2795484822186993e-05, "loss": 2.0104, "step": 5517 }, { "epoch": 0.04962832718000836, "grad_norm": 2.6314191818237305, "learning_rate": 2.2787249799225024e-05, "loss": 2.3498, "step": 5518 }, { "epoch": 0.049637321077648816, "grad_norm": 2.5036075115203857, "learning_rate": 2.2779015018246097e-05, "loss": 2.2268, "step": 5519 }, { "epoch": 0.04964631497528926, "grad_norm": 3.3655269145965576, "learning_rate": 2.2770780480150744e-05, "loss": 2.2942, "step": 5520 }, { "epoch": 0.04965530887292972, "grad_norm": 2.5793747901916504, "learning_rate": 2.276254618583948e-05, "loss": 2.3337, "step": 5521 }, { "epoch": 0.04966430277057017, "grad_norm": 3.793659210205078, "learning_rate": 2.2754312136212804e-05, "loss": 2.0209, "step": 5522 }, { "epoch": 0.04967329666821062, "grad_norm": 5.761969566345215, "learning_rate": 2.2746078332171182e-05, "loss": 1.962, "step": 5523 }, { "epoch": 0.04968229056585107, "grad_norm": 3.0050179958343506, "learning_rate": 2.2737844774615044e-05, "loss": 2.1868, "step": 5524 }, { "epoch": 0.04969128446349152, "grad_norm": 4.234293460845947, "learning_rate": 2.2729611464444794e-05, "loss": 2.3459, "step": 5525 }, { "epoch": 0.04970027836113197, "grad_norm": 3.5732157230377197, "learning_rate": 2.2721378402560826e-05, "loss": 2.0192, "step": 5526 }, { "epoch": 0.049709272258772426, "grad_norm": 2.2368719577789307, "learning_rate": 2.271314558986349e-05, "loss": 2.262, "step": 5527 }, { "epoch": 0.04971826615641287, "grad_norm": 4.238489627838135, "learning_rate": 2.2704913027253115e-05, "loss": 2.3465, "step": 5528 }, { "epoch": 0.04972726005405333, "grad_norm": 3.089731454849243, "learning_rate": 2.2696680715629996e-05, "loss": 2.062, "step": 5529 }, { "epoch": 0.049736253951693774, "grad_norm": 4.611295223236084, "learning_rate": 2.2688448655894415e-05, "loss": 2.0901, "step": 5530 }, { "epoch": 0.04974524784933423, "grad_norm": 2.7298150062561035, "learning_rate": 2.2680216848946615e-05, "loss": 2.2348, "step": 5531 }, { "epoch": 0.049754241746974674, "grad_norm": 3.918917417526245, "learning_rate": 2.2671985295686813e-05, "loss": 1.841, "step": 5532 }, { "epoch": 0.04976323564461513, "grad_norm": 4.220156669616699, "learning_rate": 2.2663753997015197e-05, "loss": 1.5877, "step": 5533 }, { "epoch": 0.04977222954225558, "grad_norm": 2.2786037921905518, "learning_rate": 2.265552295383192e-05, "loss": 2.0627, "step": 5534 }, { "epoch": 0.04978122343989603, "grad_norm": 3.4631121158599854, "learning_rate": 2.2647292167037144e-05, "loss": 1.927, "step": 5535 }, { "epoch": 0.04979021733753648, "grad_norm": 5.837767601013184, "learning_rate": 2.2639061637530954e-05, "loss": 2.1659, "step": 5536 }, { "epoch": 0.04979921123517693, "grad_norm": 3.6021344661712646, "learning_rate": 2.2630831366213433e-05, "loss": 2.0431, "step": 5537 }, { "epoch": 0.04980820513281738, "grad_norm": 3.34403395652771, "learning_rate": 2.2622601353984635e-05, "loss": 2.5066, "step": 5538 }, { "epoch": 0.04981719903045784, "grad_norm": 4.115644454956055, "learning_rate": 2.261437160174458e-05, "loss": 2.2969, "step": 5539 }, { "epoch": 0.049826192928098284, "grad_norm": 3.870316743850708, "learning_rate": 2.2606142110393247e-05, "loss": 2.2959, "step": 5540 }, { "epoch": 0.04983518682573874, "grad_norm": 3.7166330814361572, "learning_rate": 2.2597912880830624e-05, "loss": 2.0038, "step": 5541 }, { "epoch": 0.049844180723379185, "grad_norm": 2.126807689666748, "learning_rate": 2.2589683913956636e-05, "loss": 2.2124, "step": 5542 }, { "epoch": 0.04985317462101964, "grad_norm": 4.133523941040039, "learning_rate": 2.2581455210671193e-05, "loss": 1.7484, "step": 5543 }, { "epoch": 0.04986216851866009, "grad_norm": 3.765249729156494, "learning_rate": 2.2573226771874173e-05, "loss": 1.9919, "step": 5544 }, { "epoch": 0.04987116241630054, "grad_norm": 3.5767264366149902, "learning_rate": 2.2564998598465423e-05, "loss": 2.1934, "step": 5545 }, { "epoch": 0.04988015631394099, "grad_norm": 4.582919597625732, "learning_rate": 2.255677069134476e-05, "loss": 2.148, "step": 5546 }, { "epoch": 0.04988915021158144, "grad_norm": 2.7810304164886475, "learning_rate": 2.254854305141198e-05, "loss": 2.5468, "step": 5547 }, { "epoch": 0.049898144109221894, "grad_norm": 2.9925599098205566, "learning_rate": 2.2540315679566852e-05, "loss": 2.0853, "step": 5548 }, { "epoch": 0.04990713800686234, "grad_norm": 2.7754697799682617, "learning_rate": 2.2532088576709105e-05, "loss": 2.07, "step": 5549 }, { "epoch": 0.049916131904502795, "grad_norm": 4.991252422332764, "learning_rate": 2.2523861743738434e-05, "loss": 1.7054, "step": 5550 }, { "epoch": 0.04992512580214325, "grad_norm": 2.546196222305298, "learning_rate": 2.2515635181554527e-05, "loss": 2.2937, "step": 5551 }, { "epoch": 0.049934119699783695, "grad_norm": 2.528120994567871, "learning_rate": 2.250740889105702e-05, "loss": 2.1656, "step": 5552 }, { "epoch": 0.04994311359742415, "grad_norm": 2.9049201011657715, "learning_rate": 2.2499182873145523e-05, "loss": 2.0833, "step": 5553 }, { "epoch": 0.049952107495064596, "grad_norm": 2.587519407272339, "learning_rate": 2.2490957128719624e-05, "loss": 2.463, "step": 5554 }, { "epoch": 0.04996110139270505, "grad_norm": 2.3620946407318115, "learning_rate": 2.248273165867889e-05, "loss": 1.892, "step": 5555 }, { "epoch": 0.049970095290345504, "grad_norm": 2.8076703548431396, "learning_rate": 2.2474506463922838e-05, "loss": 2.2772, "step": 5556 }, { "epoch": 0.04997908918798595, "grad_norm": 2.712453603744507, "learning_rate": 2.2466281545350962e-05, "loss": 2.2061, "step": 5557 }, { "epoch": 0.049988083085626404, "grad_norm": 2.88801908493042, "learning_rate": 2.2458056903862726e-05, "loss": 2.3598, "step": 5558 }, { "epoch": 0.04999707698326685, "grad_norm": 3.027956962585449, "learning_rate": 2.2449832540357554e-05, "loss": 2.1002, "step": 5559 }, { "epoch": 0.050006070880907305, "grad_norm": 2.859497308731079, "learning_rate": 2.2441608455734873e-05, "loss": 2.0315, "step": 5560 }, { "epoch": 0.05001506477854775, "grad_norm": 3.2785441875457764, "learning_rate": 2.2433384650894046e-05, "loss": 2.3773, "step": 5561 }, { "epoch": 0.050024058676188206, "grad_norm": 2.2567660808563232, "learning_rate": 2.242516112673441e-05, "loss": 1.9872, "step": 5562 }, { "epoch": 0.05003305257382866, "grad_norm": 5.4915032386779785, "learning_rate": 2.2416937884155288e-05, "loss": 1.8642, "step": 5563 }, { "epoch": 0.05004204647146911, "grad_norm": 3.1124496459960938, "learning_rate": 2.240871492405595e-05, "loss": 2.0672, "step": 5564 }, { "epoch": 0.05005104036910956, "grad_norm": 3.373016357421875, "learning_rate": 2.240049224733566e-05, "loss": 2.404, "step": 5565 }, { "epoch": 0.05006003426675001, "grad_norm": 3.791712999343872, "learning_rate": 2.239226985489361e-05, "loss": 2.1335, "step": 5566 }, { "epoch": 0.05006902816439046, "grad_norm": 4.194141387939453, "learning_rate": 2.2384047747629023e-05, "loss": 1.9454, "step": 5567 }, { "epoch": 0.050078022062030915, "grad_norm": 8.266250610351562, "learning_rate": 2.237582592644104e-05, "loss": 1.6246, "step": 5568 }, { "epoch": 0.05008701595967136, "grad_norm": 3.7372450828552246, "learning_rate": 2.2367604392228788e-05, "loss": 2.0908, "step": 5569 }, { "epoch": 0.050096009857311816, "grad_norm": 8.77468204498291, "learning_rate": 2.2359383145891364e-05, "loss": 2.0242, "step": 5570 }, { "epoch": 0.05010500375495226, "grad_norm": 7.363594055175781, "learning_rate": 2.235116218832782e-05, "loss": 2.2881, "step": 5571 }, { "epoch": 0.050113997652592716, "grad_norm": 2.9095659255981445, "learning_rate": 2.23429415204372e-05, "loss": 1.5408, "step": 5572 }, { "epoch": 0.05012299155023317, "grad_norm": 3.5462043285369873, "learning_rate": 2.2334721143118504e-05, "loss": 2.4406, "step": 5573 }, { "epoch": 0.05013198544787362, "grad_norm": 3.4640626907348633, "learning_rate": 2.23265010572707e-05, "loss": 1.7632, "step": 5574 }, { "epoch": 0.05014097934551407, "grad_norm": 4.244899272918701, "learning_rate": 2.231828126379271e-05, "loss": 2.1283, "step": 5575 }, { "epoch": 0.05014997324315452, "grad_norm": 2.073763370513916, "learning_rate": 2.2310061763583456e-05, "loss": 2.4306, "step": 5576 }, { "epoch": 0.05015896714079497, "grad_norm": 4.532916069030762, "learning_rate": 2.2301842557541808e-05, "loss": 1.9337, "step": 5577 }, { "epoch": 0.05016796103843542, "grad_norm": 3.6205039024353027, "learning_rate": 2.229362364656659e-05, "loss": 2.0986, "step": 5578 }, { "epoch": 0.05017695493607587, "grad_norm": 1.575101613998413, "learning_rate": 2.2285405031556625e-05, "loss": 2.318, "step": 5579 }, { "epoch": 0.050185948833716326, "grad_norm": 4.885843753814697, "learning_rate": 2.2277186713410687e-05, "loss": 1.9709, "step": 5580 }, { "epoch": 0.05019494273135677, "grad_norm": 3.307373285293579, "learning_rate": 2.2268968693027516e-05, "loss": 2.2744, "step": 5581 }, { "epoch": 0.05020393662899723, "grad_norm": 3.2255537509918213, "learning_rate": 2.226075097130583e-05, "loss": 2.1934, "step": 5582 }, { "epoch": 0.050212930526637674, "grad_norm": 3.9901981353759766, "learning_rate": 2.2252533549144288e-05, "loss": 2.1757, "step": 5583 }, { "epoch": 0.05022192442427813, "grad_norm": 1.6420379877090454, "learning_rate": 2.2244316427441553e-05, "loss": 1.9331, "step": 5584 }, { "epoch": 0.05023091832191858, "grad_norm": 3.573782444000244, "learning_rate": 2.223609960709622e-05, "loss": 1.6151, "step": 5585 }, { "epoch": 0.05023991221955903, "grad_norm": 2.595676898956299, "learning_rate": 2.2227883089006887e-05, "loss": 2.1436, "step": 5586 }, { "epoch": 0.05024890611719948, "grad_norm": 2.5311057567596436, "learning_rate": 2.2219666874072093e-05, "loss": 1.9945, "step": 5587 }, { "epoch": 0.05025790001483993, "grad_norm": 4.565000057220459, "learning_rate": 2.221145096319035e-05, "loss": 2.2222, "step": 5588 }, { "epoch": 0.05026689391248038, "grad_norm": 3.2136447429656982, "learning_rate": 2.2203235357260134e-05, "loss": 2.0635, "step": 5589 }, { "epoch": 0.05027588781012083, "grad_norm": 2.5229508876800537, "learning_rate": 2.2195020057179896e-05, "loss": 2.3914, "step": 5590 }, { "epoch": 0.050284881707761284, "grad_norm": 2.5843069553375244, "learning_rate": 2.218680506384804e-05, "loss": 2.7503, "step": 5591 }, { "epoch": 0.05029387560540174, "grad_norm": 5.149191856384277, "learning_rate": 2.217859037816296e-05, "loss": 2.3017, "step": 5592 }, { "epoch": 0.050302869503042184, "grad_norm": 2.2045555114746094, "learning_rate": 2.2170376001022994e-05, "loss": 2.0375, "step": 5593 }, { "epoch": 0.05031186340068264, "grad_norm": 2.6264843940734863, "learning_rate": 2.2162161933326454e-05, "loss": 2.1276, "step": 5594 }, { "epoch": 0.050320857298323085, "grad_norm": 3.2753639221191406, "learning_rate": 2.215394817597162e-05, "loss": 1.7272, "step": 5595 }, { "epoch": 0.05032985119596354, "grad_norm": 3.9096500873565674, "learning_rate": 2.214573472985673e-05, "loss": 2.3534, "step": 5596 }, { "epoch": 0.05033884509360399, "grad_norm": 3.9719579219818115, "learning_rate": 2.213752159588e-05, "loss": 2.3047, "step": 5597 }, { "epoch": 0.05034783899124444, "grad_norm": 3.0361366271972656, "learning_rate": 2.21293087749396e-05, "loss": 1.9776, "step": 5598 }, { "epoch": 0.05035683288888489, "grad_norm": 2.8141369819641113, "learning_rate": 2.2121096267933685e-05, "loss": 2.2879, "step": 5599 }, { "epoch": 0.05036582678652534, "grad_norm": 3.1282389163970947, "learning_rate": 2.2112884075760347e-05, "loss": 2.2676, "step": 5600 }, { "epoch": 0.050374820684165794, "grad_norm": 4.578984260559082, "learning_rate": 2.2104672199317673e-05, "loss": 2.0424, "step": 5601 }, { "epoch": 0.05038381458180624, "grad_norm": 4.3770904541015625, "learning_rate": 2.209646063950369e-05, "loss": 1.9794, "step": 5602 }, { "epoch": 0.050392808479446695, "grad_norm": 4.4820733070373535, "learning_rate": 2.208824939721641e-05, "loss": 2.6297, "step": 5603 }, { "epoch": 0.05040180237708715, "grad_norm": 1.7291291952133179, "learning_rate": 2.2080038473353784e-05, "loss": 2.2026, "step": 5604 }, { "epoch": 0.050410796274727596, "grad_norm": 3.6139140129089355, "learning_rate": 2.2071827868813774e-05, "loss": 2.2438, "step": 5605 }, { "epoch": 0.05041979017236805, "grad_norm": 2.614067792892456, "learning_rate": 2.2063617584494265e-05, "loss": 2.618, "step": 5606 }, { "epoch": 0.050428784070008496, "grad_norm": 3.7276761531829834, "learning_rate": 2.2055407621293122e-05, "loss": 1.8141, "step": 5607 }, { "epoch": 0.05043777796764895, "grad_norm": 3.8056979179382324, "learning_rate": 2.2047197980108176e-05, "loss": 2.6129, "step": 5608 }, { "epoch": 0.050446771865289404, "grad_norm": 2.9379382133483887, "learning_rate": 2.2038988661837214e-05, "loss": 2.7444, "step": 5609 }, { "epoch": 0.05045576576292985, "grad_norm": 2.973733425140381, "learning_rate": 2.2030779667377996e-05, "loss": 2.5003, "step": 5610 }, { "epoch": 0.050464759660570305, "grad_norm": 2.869504451751709, "learning_rate": 2.2022570997628256e-05, "loss": 1.7583, "step": 5611 }, { "epoch": 0.05047375355821075, "grad_norm": 3.4732301235198975, "learning_rate": 2.2014362653485675e-05, "loss": 2.0869, "step": 5612 }, { "epoch": 0.050482747455851205, "grad_norm": 2.8113157749176025, "learning_rate": 2.2006154635847908e-05, "loss": 2.1784, "step": 5613 }, { "epoch": 0.05049174135349166, "grad_norm": 2.3207602500915527, "learning_rate": 2.199794694561257e-05, "loss": 1.9221, "step": 5614 }, { "epoch": 0.050500735251132106, "grad_norm": 2.0728273391723633, "learning_rate": 2.1989739583677238e-05, "loss": 1.4189, "step": 5615 }, { "epoch": 0.05050972914877256, "grad_norm": 3.1661312580108643, "learning_rate": 2.198153255093946e-05, "loss": 2.2388, "step": 5616 }, { "epoch": 0.05051872304641301, "grad_norm": 6.040445804595947, "learning_rate": 2.197332584829673e-05, "loss": 1.6753, "step": 5617 }, { "epoch": 0.05052771694405346, "grad_norm": 3.8409790992736816, "learning_rate": 2.1965119476646547e-05, "loss": 2.5396, "step": 5618 }, { "epoch": 0.05053671084169391, "grad_norm": 3.379208564758301, "learning_rate": 2.195691343688634e-05, "loss": 2.614, "step": 5619 }, { "epoch": 0.05054570473933436, "grad_norm": 6.341452598571777, "learning_rate": 2.19487077299135e-05, "loss": 2.7398, "step": 5620 }, { "epoch": 0.050554698636974815, "grad_norm": 3.431558847427368, "learning_rate": 2.194050235662539e-05, "loss": 1.9115, "step": 5621 }, { "epoch": 0.05056369253461526, "grad_norm": 6.1319355964660645, "learning_rate": 2.1932297317919348e-05, "loss": 2.2726, "step": 5622 }, { "epoch": 0.050572686432255716, "grad_norm": 7.737788677215576, "learning_rate": 2.1924092614692654e-05, "loss": 1.773, "step": 5623 }, { "epoch": 0.05058168032989616, "grad_norm": 2.810791015625, "learning_rate": 2.1915888247842574e-05, "loss": 2.0636, "step": 5624 }, { "epoch": 0.05059067422753662, "grad_norm": 2.708752155303955, "learning_rate": 2.190768421826631e-05, "loss": 2.06, "step": 5625 }, { "epoch": 0.05059966812517707, "grad_norm": 2.7627739906311035, "learning_rate": 2.1899480526861057e-05, "loss": 1.8443, "step": 5626 }, { "epoch": 0.05060866202281752, "grad_norm": 2.0803616046905518, "learning_rate": 2.189127717452395e-05, "loss": 2.0984, "step": 5627 }, { "epoch": 0.05061765592045797, "grad_norm": 3.6739654541015625, "learning_rate": 2.18830741621521e-05, "loss": 1.7434, "step": 5628 }, { "epoch": 0.05062664981809842, "grad_norm": 3.1036200523376465, "learning_rate": 2.187487149064256e-05, "loss": 1.9856, "step": 5629 }, { "epoch": 0.05063564371573887, "grad_norm": 4.229640960693359, "learning_rate": 2.186666916089239e-05, "loss": 2.0895, "step": 5630 }, { "epoch": 0.05064463761337932, "grad_norm": 3.8840668201446533, "learning_rate": 2.1858467173798568e-05, "loss": 1.9636, "step": 5631 }, { "epoch": 0.05065363151101977, "grad_norm": 4.125866413116455, "learning_rate": 2.1850265530258055e-05, "loss": 1.6045, "step": 5632 }, { "epoch": 0.050662625408660227, "grad_norm": 3.4617228507995605, "learning_rate": 2.1842064231167767e-05, "loss": 2.4278, "step": 5633 }, { "epoch": 0.05067161930630067, "grad_norm": 2.6733291149139404, "learning_rate": 2.1833863277424585e-05, "loss": 2.1856, "step": 5634 }, { "epoch": 0.05068061320394113, "grad_norm": 3.0555877685546875, "learning_rate": 2.1825662669925354e-05, "loss": 1.8897, "step": 5635 }, { "epoch": 0.050689607101581574, "grad_norm": 2.6245837211608887, "learning_rate": 2.1817462409566873e-05, "loss": 1.8846, "step": 5636 }, { "epoch": 0.05069860099922203, "grad_norm": 7.898270130157471, "learning_rate": 2.1809262497245925e-05, "loss": 2.4393, "step": 5637 }, { "epoch": 0.05070759489686248, "grad_norm": 3.226712703704834, "learning_rate": 2.1801062933859235e-05, "loss": 2.2462, "step": 5638 }, { "epoch": 0.05071658879450293, "grad_norm": 5.180421352386475, "learning_rate": 2.1792863720303494e-05, "loss": 2.7613, "step": 5639 }, { "epoch": 0.05072558269214338, "grad_norm": 3.1132771968841553, "learning_rate": 2.1784664857475352e-05, "loss": 2.1456, "step": 5640 }, { "epoch": 0.05073457658978383, "grad_norm": 4.475744247436523, "learning_rate": 2.177646634627143e-05, "loss": 1.7053, "step": 5641 }, { "epoch": 0.05074357048742428, "grad_norm": 2.3457350730895996, "learning_rate": 2.1768268187588293e-05, "loss": 1.8743, "step": 5642 }, { "epoch": 0.05075256438506473, "grad_norm": 5.240878582000732, "learning_rate": 2.1760070382322497e-05, "loss": 2.3541, "step": 5643 }, { "epoch": 0.050761558282705184, "grad_norm": 2.8228189945220947, "learning_rate": 2.1751872931370528e-05, "loss": 2.0629, "step": 5644 }, { "epoch": 0.05077055218034564, "grad_norm": 2.026094913482666, "learning_rate": 2.1743675835628856e-05, "loss": 1.6995, "step": 5645 }, { "epoch": 0.050779546077986085, "grad_norm": 6.007753849029541, "learning_rate": 2.1735479095993893e-05, "loss": 2.0609, "step": 5646 }, { "epoch": 0.05078853997562654, "grad_norm": 9.52855110168457, "learning_rate": 2.1727282713362036e-05, "loss": 2.0499, "step": 5647 }, { "epoch": 0.050797533873266985, "grad_norm": 3.7572951316833496, "learning_rate": 2.1719086688629612e-05, "loss": 2.3177, "step": 5648 }, { "epoch": 0.05080652777090744, "grad_norm": 4.077938556671143, "learning_rate": 2.171089102269294e-05, "loss": 1.7724, "step": 5649 }, { "epoch": 0.05081552166854789, "grad_norm": 4.444587230682373, "learning_rate": 2.1702695716448278e-05, "loss": 1.9565, "step": 5650 }, { "epoch": 0.05082451556618834, "grad_norm": 2.8621201515197754, "learning_rate": 2.1694500770791858e-05, "loss": 1.8665, "step": 5651 }, { "epoch": 0.050833509463828794, "grad_norm": 4.1064133644104, "learning_rate": 2.1686306186619863e-05, "loss": 1.9231, "step": 5652 }, { "epoch": 0.05084250336146924, "grad_norm": 3.7369656562805176, "learning_rate": 2.1678111964828445e-05, "loss": 2.1352, "step": 5653 }, { "epoch": 0.050851497259109694, "grad_norm": 2.4633519649505615, "learning_rate": 2.1669918106313704e-05, "loss": 2.0984, "step": 5654 }, { "epoch": 0.05086049115675015, "grad_norm": 4.721185207366943, "learning_rate": 2.1661724611971708e-05, "loss": 2.6612, "step": 5655 }, { "epoch": 0.050869485054390595, "grad_norm": 2.947802782058716, "learning_rate": 2.1653531482698493e-05, "loss": 2.1022, "step": 5656 }, { "epoch": 0.05087847895203105, "grad_norm": 3.6430299282073975, "learning_rate": 2.1645338719390047e-05, "loss": 2.0586, "step": 5657 }, { "epoch": 0.050887472849671496, "grad_norm": 2.482351303100586, "learning_rate": 2.1637146322942315e-05, "loss": 2.191, "step": 5658 }, { "epoch": 0.05089646674731195, "grad_norm": 1.6299551725387573, "learning_rate": 2.1628954294251206e-05, "loss": 2.157, "step": 5659 }, { "epoch": 0.0509054606449524, "grad_norm": 2.68165922164917, "learning_rate": 2.1620762634212586e-05, "loss": 2.632, "step": 5660 }, { "epoch": 0.05091445454259285, "grad_norm": 2.7730045318603516, "learning_rate": 2.161257134372227e-05, "loss": 2.1814, "step": 5661 }, { "epoch": 0.050923448440233304, "grad_norm": 4.3296966552734375, "learning_rate": 2.1604380423676073e-05, "loss": 2.4578, "step": 5662 }, { "epoch": 0.05093244233787375, "grad_norm": 3.9866573810577393, "learning_rate": 2.159618987496973e-05, "loss": 2.1194, "step": 5663 }, { "epoch": 0.050941436235514205, "grad_norm": 2.54345965385437, "learning_rate": 2.1587999698498938e-05, "loss": 2.6535, "step": 5664 }, { "epoch": 0.05095043013315465, "grad_norm": 5.828786373138428, "learning_rate": 2.1579809895159375e-05, "loss": 2.0331, "step": 5665 }, { "epoch": 0.050959424030795106, "grad_norm": 4.044792652130127, "learning_rate": 2.157162046584666e-05, "loss": 1.8381, "step": 5666 }, { "epoch": 0.05096841792843556, "grad_norm": 4.330307960510254, "learning_rate": 2.1563431411456367e-05, "loss": 2.2163, "step": 5667 }, { "epoch": 0.050977411826076006, "grad_norm": 4.453837871551514, "learning_rate": 2.155524273288405e-05, "loss": 2.4631, "step": 5668 }, { "epoch": 0.05098640572371646, "grad_norm": 4.522970199584961, "learning_rate": 2.1547054431025214e-05, "loss": 1.9839, "step": 5669 }, { "epoch": 0.05099539962135691, "grad_norm": 3.1281394958496094, "learning_rate": 2.153886650677531e-05, "loss": 1.8907, "step": 5670 }, { "epoch": 0.05100439351899736, "grad_norm": 3.1135096549987793, "learning_rate": 2.1530678961029762e-05, "loss": 2.1, "step": 5671 }, { "epoch": 0.05101338741663781, "grad_norm": 4.239263534545898, "learning_rate": 2.152249179468395e-05, "loss": 2.3135, "step": 5672 }, { "epoch": 0.05102238131427826, "grad_norm": 4.753326416015625, "learning_rate": 2.1514305008633206e-05, "loss": 2.2117, "step": 5673 }, { "epoch": 0.051031375211918716, "grad_norm": 2.826779365539551, "learning_rate": 2.150611860377282e-05, "loss": 2.2105, "step": 5674 }, { "epoch": 0.05104036910955916, "grad_norm": 2.89357328414917, "learning_rate": 2.1497932580998053e-05, "loss": 2.1411, "step": 5675 }, { "epoch": 0.051049363007199616, "grad_norm": 2.370638847351074, "learning_rate": 2.1489746941204116e-05, "loss": 2.2164, "step": 5676 }, { "epoch": 0.05105835690484006, "grad_norm": 2.875696897506714, "learning_rate": 2.148156168528618e-05, "loss": 2.0154, "step": 5677 }, { "epoch": 0.05106735080248052, "grad_norm": 2.7109649181365967, "learning_rate": 2.1473376814139362e-05, "loss": 2.1201, "step": 5678 }, { "epoch": 0.05107634470012097, "grad_norm": 3.5154542922973633, "learning_rate": 2.146519232865876e-05, "loss": 2.3996, "step": 5679 }, { "epoch": 0.05108533859776142, "grad_norm": 3.5744059085845947, "learning_rate": 2.1457008229739394e-05, "loss": 2.2761, "step": 5680 }, { "epoch": 0.05109433249540187, "grad_norm": 2.3435218334198, "learning_rate": 2.1448824518276295e-05, "loss": 2.0542, "step": 5681 }, { "epoch": 0.05110332639304232, "grad_norm": 2.8864543437957764, "learning_rate": 2.1440641195164405e-05, "loss": 2.4353, "step": 5682 }, { "epoch": 0.05111232029068277, "grad_norm": 3.5247907638549805, "learning_rate": 2.1432458261298646e-05, "loss": 2.2306, "step": 5683 }, { "epoch": 0.051121314188323226, "grad_norm": 2.20542573928833, "learning_rate": 2.1424275717573888e-05, "loss": 2.0654, "step": 5684 }, { "epoch": 0.05113030808596367, "grad_norm": 3.4897255897521973, "learning_rate": 2.141609356488496e-05, "loss": 1.893, "step": 5685 }, { "epoch": 0.05113930198360413, "grad_norm": 4.104620933532715, "learning_rate": 2.140791180412665e-05, "loss": 2.3382, "step": 5686 }, { "epoch": 0.051148295881244574, "grad_norm": 2.121587038040161, "learning_rate": 2.1399730436193697e-05, "loss": 2.2725, "step": 5687 }, { "epoch": 0.05115728977888503, "grad_norm": 3.25046968460083, "learning_rate": 2.1391549461980817e-05, "loss": 2.3551, "step": 5688 }, { "epoch": 0.051166283676525474, "grad_norm": 5.144603729248047, "learning_rate": 2.1383368882382666e-05, "loss": 2.3176, "step": 5689 }, { "epoch": 0.05117527757416593, "grad_norm": 3.8693792819976807, "learning_rate": 2.1375188698293855e-05, "loss": 2.375, "step": 5690 }, { "epoch": 0.05118427147180638, "grad_norm": 1.8157706260681152, "learning_rate": 2.136700891060896e-05, "loss": 2.7003, "step": 5691 }, { "epoch": 0.05119326536944683, "grad_norm": 2.3327972888946533, "learning_rate": 2.1358829520222502e-05, "loss": 1.748, "step": 5692 }, { "epoch": 0.05120225926708728, "grad_norm": 3.3057212829589844, "learning_rate": 2.135065052802897e-05, "loss": 1.8231, "step": 5693 }, { "epoch": 0.05121125316472773, "grad_norm": 1.8098077774047852, "learning_rate": 2.134247193492282e-05, "loss": 1.9193, "step": 5694 }, { "epoch": 0.051220247062368184, "grad_norm": 2.4097743034362793, "learning_rate": 2.1334293741798432e-05, "loss": 2.0552, "step": 5695 }, { "epoch": 0.05122924096000864, "grad_norm": 3.5201780796051025, "learning_rate": 2.1326115949550175e-05, "loss": 1.7509, "step": 5696 }, { "epoch": 0.051238234857649084, "grad_norm": 4.014453411102295, "learning_rate": 2.1317938559072352e-05, "loss": 1.6674, "step": 5697 }, { "epoch": 0.05124722875528954, "grad_norm": 3.571474313735962, "learning_rate": 2.1309761571259237e-05, "loss": 2.4979, "step": 5698 }, { "epoch": 0.051256222652929985, "grad_norm": 4.422110557556152, "learning_rate": 2.1301584987005036e-05, "loss": 2.4553, "step": 5699 }, { "epoch": 0.05126521655057044, "grad_norm": 4.384992599487305, "learning_rate": 2.1293408807203947e-05, "loss": 2.4225, "step": 5700 }, { "epoch": 0.051274210448210886, "grad_norm": 2.3552584648132324, "learning_rate": 2.1285233032750103e-05, "loss": 1.9324, "step": 5701 }, { "epoch": 0.05128320434585134, "grad_norm": 2.543139696121216, "learning_rate": 2.127705766453759e-05, "loss": 2.0098, "step": 5702 }, { "epoch": 0.05129219824349179, "grad_norm": 7.3154168128967285, "learning_rate": 2.126888270346045e-05, "loss": 2.2606, "step": 5703 }, { "epoch": 0.05130119214113224, "grad_norm": 2.7986252307891846, "learning_rate": 2.126070815041269e-05, "loss": 2.2727, "step": 5704 }, { "epoch": 0.051310186038772694, "grad_norm": 3.2764103412628174, "learning_rate": 2.1252534006288264e-05, "loss": 2.4881, "step": 5705 }, { "epoch": 0.05131917993641314, "grad_norm": 3.344090223312378, "learning_rate": 2.1244360271981073e-05, "loss": 2.2967, "step": 5706 }, { "epoch": 0.051328173834053595, "grad_norm": 2.422438621520996, "learning_rate": 2.1236186948385007e-05, "loss": 1.563, "step": 5707 }, { "epoch": 0.05133716773169405, "grad_norm": 4.19499397277832, "learning_rate": 2.1228014036393877e-05, "loss": 1.7627, "step": 5708 }, { "epoch": 0.051346161629334496, "grad_norm": 4.089723110198975, "learning_rate": 2.1219841536901465e-05, "loss": 1.859, "step": 5709 }, { "epoch": 0.05135515552697495, "grad_norm": 3.1726179122924805, "learning_rate": 2.1211669450801493e-05, "loss": 1.8218, "step": 5710 }, { "epoch": 0.051364149424615396, "grad_norm": 4.537609100341797, "learning_rate": 2.120349777898766e-05, "loss": 1.9431, "step": 5711 }, { "epoch": 0.05137314332225585, "grad_norm": 8.080046653747559, "learning_rate": 2.119532652235359e-05, "loss": 2.1163, "step": 5712 }, { "epoch": 0.0513821372198963, "grad_norm": 3.948796033859253, "learning_rate": 2.11871556817929e-05, "loss": 2.1965, "step": 5713 }, { "epoch": 0.05139113111753675, "grad_norm": 2.271238327026367, "learning_rate": 2.1178985258199133e-05, "loss": 1.8804, "step": 5714 }, { "epoch": 0.051400125015177205, "grad_norm": 2.4457314014434814, "learning_rate": 2.117081525246579e-05, "loss": 2.6215, "step": 5715 }, { "epoch": 0.05140911891281765, "grad_norm": 4.196275234222412, "learning_rate": 2.1162645665486332e-05, "loss": 2.2596, "step": 5716 }, { "epoch": 0.051418112810458105, "grad_norm": 4.620303153991699, "learning_rate": 2.115447649815418e-05, "loss": 2.2903, "step": 5717 }, { "epoch": 0.05142710670809855, "grad_norm": 3.8896591663360596, "learning_rate": 2.1146307751362694e-05, "loss": 2.2379, "step": 5718 }, { "epoch": 0.051436100605739006, "grad_norm": 2.3688220977783203, "learning_rate": 2.11381394260052e-05, "loss": 2.2441, "step": 5719 }, { "epoch": 0.05144509450337946, "grad_norm": 2.6962215900421143, "learning_rate": 2.1129971522974967e-05, "loss": 2.0645, "step": 5720 }, { "epoch": 0.05145408840101991, "grad_norm": 3.9879026412963867, "learning_rate": 2.1121804043165232e-05, "loss": 2.2436, "step": 5721 }, { "epoch": 0.05146308229866036, "grad_norm": 2.8724024295806885, "learning_rate": 2.1113636987469183e-05, "loss": 2.2942, "step": 5722 }, { "epoch": 0.05147207619630081, "grad_norm": 2.387303590774536, "learning_rate": 2.1105470356779948e-05, "loss": 2.3891, "step": 5723 }, { "epoch": 0.05148107009394126, "grad_norm": 3.5945045948028564, "learning_rate": 2.1097304151990616e-05, "loss": 2.3703, "step": 5724 }, { "epoch": 0.051490063991581715, "grad_norm": 2.4941048622131348, "learning_rate": 2.1089138373994223e-05, "loss": 2.4304, "step": 5725 }, { "epoch": 0.05149905788922216, "grad_norm": 2.9465620517730713, "learning_rate": 2.1080973023683793e-05, "loss": 1.7779, "step": 5726 }, { "epoch": 0.051508051786862616, "grad_norm": 4.768709182739258, "learning_rate": 2.107280810195226e-05, "loss": 2.2751, "step": 5727 }, { "epoch": 0.05151704568450306, "grad_norm": 8.009357452392578, "learning_rate": 2.106464360969253e-05, "loss": 2.4356, "step": 5728 }, { "epoch": 0.05152603958214352, "grad_norm": 3.229539394378662, "learning_rate": 2.1056479547797454e-05, "loss": 2.3776, "step": 5729 }, { "epoch": 0.051535033479783963, "grad_norm": 2.691683530807495, "learning_rate": 2.1048315917159846e-05, "loss": 1.5572, "step": 5730 }, { "epoch": 0.05154402737742442, "grad_norm": 2.153756618499756, "learning_rate": 2.1040152718672457e-05, "loss": 2.0184, "step": 5731 }, { "epoch": 0.05155302127506487, "grad_norm": 2.7990493774414062, "learning_rate": 2.1031989953228027e-05, "loss": 1.817, "step": 5732 }, { "epoch": 0.05156201517270532, "grad_norm": 3.0611777305603027, "learning_rate": 2.1023827621719206e-05, "loss": 2.2029, "step": 5733 }, { "epoch": 0.05157100907034577, "grad_norm": 3.4464073181152344, "learning_rate": 2.101566572503862e-05, "loss": 2.0954, "step": 5734 }, { "epoch": 0.05158000296798622, "grad_norm": 3.725468397140503, "learning_rate": 2.100750426407884e-05, "loss": 1.8613, "step": 5735 }, { "epoch": 0.05158899686562667, "grad_norm": 4.365272521972656, "learning_rate": 2.0999343239732393e-05, "loss": 1.8421, "step": 5736 }, { "epoch": 0.051597990763267126, "grad_norm": 2.153270721435547, "learning_rate": 2.0991182652891746e-05, "loss": 2.0623, "step": 5737 }, { "epoch": 0.05160698466090757, "grad_norm": 3.673583507537842, "learning_rate": 2.098302250444934e-05, "loss": 1.7434, "step": 5738 }, { "epoch": 0.05161597855854803, "grad_norm": 2.5302672386169434, "learning_rate": 2.0974862795297555e-05, "loss": 1.9941, "step": 5739 }, { "epoch": 0.051624972456188474, "grad_norm": 3.54522442817688, "learning_rate": 2.0966703526328728e-05, "loss": 1.9892, "step": 5740 }, { "epoch": 0.05163396635382893, "grad_norm": 2.0769031047821045, "learning_rate": 2.095854469843513e-05, "loss": 1.7298, "step": 5741 }, { "epoch": 0.051642960251469375, "grad_norm": 5.7586798667907715, "learning_rate": 2.0950386312509015e-05, "loss": 1.81, "step": 5742 }, { "epoch": 0.05165195414910983, "grad_norm": 4.556309700012207, "learning_rate": 2.0942228369442562e-05, "loss": 2.0807, "step": 5743 }, { "epoch": 0.05166094804675028, "grad_norm": 9.846391677856445, "learning_rate": 2.0934070870127912e-05, "loss": 2.0819, "step": 5744 }, { "epoch": 0.05166994194439073, "grad_norm": 3.2079882621765137, "learning_rate": 2.0925913815457153e-05, "loss": 2.0103, "step": 5745 }, { "epoch": 0.05167893584203118, "grad_norm": 3.687513589859009, "learning_rate": 2.0917757206322346e-05, "loss": 2.5841, "step": 5746 }, { "epoch": 0.05168792973967163, "grad_norm": 2.4367525577545166, "learning_rate": 2.090960104361547e-05, "loss": 2.3189, "step": 5747 }, { "epoch": 0.051696923637312084, "grad_norm": 3.3398540019989014, "learning_rate": 2.0901445328228475e-05, "loss": 2.1251, "step": 5748 }, { "epoch": 0.05170591753495254, "grad_norm": 3.4748549461364746, "learning_rate": 2.0893290061053255e-05, "loss": 2.1113, "step": 5749 }, { "epoch": 0.051714911432592985, "grad_norm": 4.0430588722229, "learning_rate": 2.088513524298165e-05, "loss": 2.031, "step": 5750 }, { "epoch": 0.05172390533023344, "grad_norm": 2.5692734718322754, "learning_rate": 2.0876980874905477e-05, "loss": 2.8679, "step": 5751 }, { "epoch": 0.051732899227873885, "grad_norm": 3.496025323867798, "learning_rate": 2.086882695771648e-05, "loss": 1.5239, "step": 5752 }, { "epoch": 0.05174189312551434, "grad_norm": 3.014814615249634, "learning_rate": 2.0860673492306355e-05, "loss": 1.985, "step": 5753 }, { "epoch": 0.05175088702315479, "grad_norm": 4.4749064445495605, "learning_rate": 2.085252047956675e-05, "loss": 1.9626, "step": 5754 }, { "epoch": 0.05175988092079524, "grad_norm": 3.620128631591797, "learning_rate": 2.0844367920389273e-05, "loss": 1.7009, "step": 5755 }, { "epoch": 0.051768874818435694, "grad_norm": 5.559382915496826, "learning_rate": 2.083621581566547e-05, "loss": 2.1961, "step": 5756 }, { "epoch": 0.05177786871607614, "grad_norm": 4.862939834594727, "learning_rate": 2.0828064166286838e-05, "loss": 2.4679, "step": 5757 }, { "epoch": 0.051786862613716594, "grad_norm": 2.2741801738739014, "learning_rate": 2.0819912973144843e-05, "loss": 2.2544, "step": 5758 }, { "epoch": 0.05179585651135704, "grad_norm": 3.7523577213287354, "learning_rate": 2.0811762237130884e-05, "loss": 1.9308, "step": 5759 }, { "epoch": 0.051804850408997495, "grad_norm": 4.9186201095581055, "learning_rate": 2.080361195913631e-05, "loss": 2.0909, "step": 5760 }, { "epoch": 0.05181384430663795, "grad_norm": 3.6369097232818604, "learning_rate": 2.0795462140052423e-05, "loss": 1.8533, "step": 5761 }, { "epoch": 0.051822838204278396, "grad_norm": 2.663358449935913, "learning_rate": 2.0787312780770467e-05, "loss": 1.9432, "step": 5762 }, { "epoch": 0.05183183210191885, "grad_norm": 2.66302490234375, "learning_rate": 2.0779163882181655e-05, "loss": 2.1861, "step": 5763 }, { "epoch": 0.0518408259995593, "grad_norm": 3.144129753112793, "learning_rate": 2.0771015445177135e-05, "loss": 2.6237, "step": 5764 }, { "epoch": 0.05184981989719975, "grad_norm": 4.08966588973999, "learning_rate": 2.0762867470648013e-05, "loss": 2.0718, "step": 5765 }, { "epoch": 0.051858813794840204, "grad_norm": 2.296562671661377, "learning_rate": 2.0754719959485324e-05, "loss": 2.0893, "step": 5766 }, { "epoch": 0.05186780769248065, "grad_norm": 4.296117782592773, "learning_rate": 2.0746572912580086e-05, "loss": 2.1618, "step": 5767 }, { "epoch": 0.051876801590121105, "grad_norm": 2.312896490097046, "learning_rate": 2.073842633082324e-05, "loss": 2.2792, "step": 5768 }, { "epoch": 0.05188579548776155, "grad_norm": 1.9854369163513184, "learning_rate": 2.0730280215105676e-05, "loss": 2.3211, "step": 5769 }, { "epoch": 0.051894789385402006, "grad_norm": 2.8688759803771973, "learning_rate": 2.072213456631825e-05, "loss": 2.2889, "step": 5770 }, { "epoch": 0.05190378328304245, "grad_norm": 3.5189671516418457, "learning_rate": 2.071398938535176e-05, "loss": 2.0285, "step": 5771 }, { "epoch": 0.051912777180682906, "grad_norm": 4.867355823516846, "learning_rate": 2.0705844673096945e-05, "loss": 2.2916, "step": 5772 }, { "epoch": 0.05192177107832336, "grad_norm": 3.417435646057129, "learning_rate": 2.0697700430444505e-05, "loss": 1.8427, "step": 5773 }, { "epoch": 0.05193076497596381, "grad_norm": 3.3530077934265137, "learning_rate": 2.0689556658285074e-05, "loss": 2.0742, "step": 5774 }, { "epoch": 0.05193975887360426, "grad_norm": 3.010962963104248, "learning_rate": 2.068141335750925e-05, "loss": 2.2715, "step": 5775 }, { "epoch": 0.05194875277124471, "grad_norm": 2.51218581199646, "learning_rate": 2.0673270529007555e-05, "loss": 2.2479, "step": 5776 }, { "epoch": 0.05195774666888516, "grad_norm": 2.1503305435180664, "learning_rate": 2.0665128173670502e-05, "loss": 2.2422, "step": 5777 }, { "epoch": 0.051966740566525615, "grad_norm": 2.9737088680267334, "learning_rate": 2.065698629238851e-05, "loss": 2.0511, "step": 5778 }, { "epoch": 0.05197573446416606, "grad_norm": 4.0703630447387695, "learning_rate": 2.0648844886051972e-05, "loss": 2.0614, "step": 5779 }, { "epoch": 0.051984728361806516, "grad_norm": 3.0987277030944824, "learning_rate": 2.0640703955551212e-05, "loss": 1.933, "step": 5780 }, { "epoch": 0.05199372225944696, "grad_norm": 2.4590249061584473, "learning_rate": 2.0632563501776516e-05, "loss": 1.7556, "step": 5781 }, { "epoch": 0.05200271615708742, "grad_norm": 6.969913959503174, "learning_rate": 2.0624423525618098e-05, "loss": 1.8581, "step": 5782 }, { "epoch": 0.052011710054727864, "grad_norm": 3.179901123046875, "learning_rate": 2.0616284027966155e-05, "loss": 2.6058, "step": 5783 }, { "epoch": 0.05202070395236832, "grad_norm": 3.7543888092041016, "learning_rate": 2.06081450097108e-05, "loss": 2.1071, "step": 5784 }, { "epoch": 0.05202969785000877, "grad_norm": 2.4502577781677246, "learning_rate": 2.0600006471742104e-05, "loss": 1.8221, "step": 5785 }, { "epoch": 0.05203869174764922, "grad_norm": 3.6263155937194824, "learning_rate": 2.059186841495008e-05, "loss": 2.2791, "step": 5786 }, { "epoch": 0.05204768564528967, "grad_norm": 5.104333400726318, "learning_rate": 2.0583730840224703e-05, "loss": 1.7573, "step": 5787 }, { "epoch": 0.05205667954293012, "grad_norm": 3.535855531692505, "learning_rate": 2.057559374845588e-05, "loss": 2.0994, "step": 5788 }, { "epoch": 0.05206567344057057, "grad_norm": 3.6937255859375, "learning_rate": 2.056745714053347e-05, "loss": 2.5143, "step": 5789 }, { "epoch": 0.05207466733821103, "grad_norm": 2.566901922225952, "learning_rate": 2.0559321017347285e-05, "loss": 2.1169, "step": 5790 }, { "epoch": 0.052083661235851474, "grad_norm": 2.6976611614227295, "learning_rate": 2.0551185379787076e-05, "loss": 2.2463, "step": 5791 }, { "epoch": 0.05209265513349193, "grad_norm": 2.1550776958465576, "learning_rate": 2.0543050228742548e-05, "loss": 2.1595, "step": 5792 }, { "epoch": 0.052101649031132374, "grad_norm": 4.498174667358398, "learning_rate": 2.0534915565103347e-05, "loss": 1.9818, "step": 5793 }, { "epoch": 0.05211064292877283, "grad_norm": 2.8690185546875, "learning_rate": 2.0526781389759066e-05, "loss": 1.8165, "step": 5794 }, { "epoch": 0.05211963682641328, "grad_norm": 3.6539528369903564, "learning_rate": 2.0518647703599236e-05, "loss": 2.4113, "step": 5795 }, { "epoch": 0.05212863072405373, "grad_norm": 2.8977773189544678, "learning_rate": 2.0510514507513368e-05, "loss": 1.8938, "step": 5796 }, { "epoch": 0.05213762462169418, "grad_norm": 2.2046453952789307, "learning_rate": 2.050238180239088e-05, "loss": 1.9487, "step": 5797 }, { "epoch": 0.05214661851933463, "grad_norm": 2.3905463218688965, "learning_rate": 2.0494249589121157e-05, "loss": 2.121, "step": 5798 }, { "epoch": 0.05215561241697508, "grad_norm": 3.2178456783294678, "learning_rate": 2.0486117868593527e-05, "loss": 2.5538, "step": 5799 }, { "epoch": 0.05216460631461553, "grad_norm": 2.6382107734680176, "learning_rate": 2.047798664169726e-05, "loss": 1.883, "step": 5800 }, { "epoch": 0.052173600212255984, "grad_norm": 4.266504287719727, "learning_rate": 2.0469855909321564e-05, "loss": 2.1381, "step": 5801 }, { "epoch": 0.05218259410989644, "grad_norm": 5.294132709503174, "learning_rate": 2.0461725672355623e-05, "loss": 2.1681, "step": 5802 }, { "epoch": 0.052191588007536885, "grad_norm": 4.123690128326416, "learning_rate": 2.0453595931688544e-05, "loss": 2.4958, "step": 5803 }, { "epoch": 0.05220058190517734, "grad_norm": 3.9471373558044434, "learning_rate": 2.0445466688209373e-05, "loss": 2.3085, "step": 5804 }, { "epoch": 0.052209575802817786, "grad_norm": 2.925266981124878, "learning_rate": 2.043733794280712e-05, "loss": 1.4387, "step": 5805 }, { "epoch": 0.05221856970045824, "grad_norm": 2.0519025325775146, "learning_rate": 2.042920969637073e-05, "loss": 2.3023, "step": 5806 }, { "epoch": 0.05222756359809869, "grad_norm": 6.058498859405518, "learning_rate": 2.0421081949789097e-05, "loss": 1.9724, "step": 5807 }, { "epoch": 0.05223655749573914, "grad_norm": 3.607909917831421, "learning_rate": 2.0412954703951044e-05, "loss": 1.4181, "step": 5808 }, { "epoch": 0.052245551393379594, "grad_norm": 4.994721412658691, "learning_rate": 2.040482795974538e-05, "loss": 2.2017, "step": 5809 }, { "epoch": 0.05225454529102004, "grad_norm": 4.905498504638672, "learning_rate": 2.0396701718060824e-05, "loss": 2.1389, "step": 5810 }, { "epoch": 0.052263539188660495, "grad_norm": 2.672398805618286, "learning_rate": 2.0388575979786044e-05, "loss": 2.1984, "step": 5811 }, { "epoch": 0.05227253308630094, "grad_norm": 2.6052792072296143, "learning_rate": 2.038045074580966e-05, "loss": 1.8796, "step": 5812 }, { "epoch": 0.052281526983941395, "grad_norm": 2.005021572113037, "learning_rate": 2.037232601702024e-05, "loss": 2.1861, "step": 5813 }, { "epoch": 0.05229052088158185, "grad_norm": 2.835474729537964, "learning_rate": 2.036420179430628e-05, "loss": 2.0638, "step": 5814 }, { "epoch": 0.052299514779222296, "grad_norm": 3.09429669380188, "learning_rate": 2.035607807855625e-05, "loss": 1.8839, "step": 5815 }, { "epoch": 0.05230850867686275, "grad_norm": 3.2279179096221924, "learning_rate": 2.0347954870658536e-05, "loss": 2.4396, "step": 5816 }, { "epoch": 0.0523175025745032, "grad_norm": 2.6414377689361572, "learning_rate": 2.0339832171501487e-05, "loss": 1.8421, "step": 5817 }, { "epoch": 0.05232649647214365, "grad_norm": 3.5889673233032227, "learning_rate": 2.033170998197339e-05, "loss": 2.0959, "step": 5818 }, { "epoch": 0.052335490369784105, "grad_norm": 11.178322792053223, "learning_rate": 2.0323588302962462e-05, "loss": 2.0256, "step": 5819 }, { "epoch": 0.05234448426742455, "grad_norm": 4.446950435638428, "learning_rate": 2.031546713535688e-05, "loss": 2.4085, "step": 5820 }, { "epoch": 0.052353478165065005, "grad_norm": 3.2998478412628174, "learning_rate": 2.030734648004478e-05, "loss": 1.8241, "step": 5821 }, { "epoch": 0.05236247206270545, "grad_norm": 3.7894885540008545, "learning_rate": 2.029922633791421e-05, "loss": 2.3058, "step": 5822 }, { "epoch": 0.052371465960345906, "grad_norm": 2.562361240386963, "learning_rate": 2.0291106709853186e-05, "loss": 1.9303, "step": 5823 }, { "epoch": 0.05238045985798635, "grad_norm": 2.511770248413086, "learning_rate": 2.028298759674965e-05, "loss": 2.2475, "step": 5824 }, { "epoch": 0.05238945375562681, "grad_norm": 8.031078338623047, "learning_rate": 2.02748689994915e-05, "loss": 1.9917, "step": 5825 }, { "epoch": 0.05239844765326726, "grad_norm": 2.176940441131592, "learning_rate": 2.0266750918966567e-05, "loss": 1.9969, "step": 5826 }, { "epoch": 0.05240744155090771, "grad_norm": 2.5190494060516357, "learning_rate": 2.0258633356062632e-05, "loss": 1.9621, "step": 5827 }, { "epoch": 0.05241643544854816, "grad_norm": 2.6817941665649414, "learning_rate": 2.0250516311667432e-05, "loss": 1.6762, "step": 5828 }, { "epoch": 0.05242542934618861, "grad_norm": 3.384891986846924, "learning_rate": 2.0242399786668628e-05, "loss": 2.3153, "step": 5829 }, { "epoch": 0.05243442324382906, "grad_norm": 2.993485927581787, "learning_rate": 2.0234283781953834e-05, "loss": 1.9175, "step": 5830 }, { "epoch": 0.052443417141469516, "grad_norm": 3.5577211380004883, "learning_rate": 2.0226168298410597e-05, "loss": 2.1838, "step": 5831 }, { "epoch": 0.05245241103910996, "grad_norm": 2.1746041774749756, "learning_rate": 2.0218053336926417e-05, "loss": 2.3457, "step": 5832 }, { "epoch": 0.052461404936750416, "grad_norm": 3.2770497798919678, "learning_rate": 2.0209938898388726e-05, "loss": 1.6913, "step": 5833 }, { "epoch": 0.05247039883439086, "grad_norm": 3.5659666061401367, "learning_rate": 2.020182498368492e-05, "loss": 1.7382, "step": 5834 }, { "epoch": 0.05247939273203132, "grad_norm": 2.6237494945526123, "learning_rate": 2.0193711593702326e-05, "loss": 2.5135, "step": 5835 }, { "epoch": 0.05248838662967177, "grad_norm": 3.399759531021118, "learning_rate": 2.0185598729328205e-05, "loss": 2.1766, "step": 5836 }, { "epoch": 0.05249738052731222, "grad_norm": 2.518601655960083, "learning_rate": 2.0177486391449764e-05, "loss": 2.4502, "step": 5837 }, { "epoch": 0.05250637442495267, "grad_norm": 3.362213611602783, "learning_rate": 2.0169374580954165e-05, "loss": 2.0184, "step": 5838 }, { "epoch": 0.05251536832259312, "grad_norm": 3.5269792079925537, "learning_rate": 2.0161263298728495e-05, "loss": 1.6506, "step": 5839 }, { "epoch": 0.05252436222023357, "grad_norm": 2.3872640132904053, "learning_rate": 2.0153152545659798e-05, "loss": 1.8288, "step": 5840 }, { "epoch": 0.05253335611787402, "grad_norm": 9.274826049804688, "learning_rate": 2.014504232263505e-05, "loss": 2.3327, "step": 5841 }, { "epoch": 0.05254235001551447, "grad_norm": 4.616566181182861, "learning_rate": 2.0136932630541176e-05, "loss": 2.2587, "step": 5842 }, { "epoch": 0.05255134391315493, "grad_norm": 3.067469835281372, "learning_rate": 2.012882347026504e-05, "loss": 2.6666, "step": 5843 }, { "epoch": 0.052560337810795374, "grad_norm": 4.333625793457031, "learning_rate": 2.0120714842693445e-05, "loss": 2.3334, "step": 5844 }, { "epoch": 0.05256933170843583, "grad_norm": 5.310240745544434, "learning_rate": 2.0112606748713138e-05, "loss": 2.8283, "step": 5845 }, { "epoch": 0.052578325606076275, "grad_norm": 3.594099760055542, "learning_rate": 2.01044991892108e-05, "loss": 1.6793, "step": 5846 }, { "epoch": 0.05258731950371673, "grad_norm": 4.501221656799316, "learning_rate": 2.0096392165073082e-05, "loss": 1.9888, "step": 5847 }, { "epoch": 0.05259631340135718, "grad_norm": 3.690490245819092, "learning_rate": 2.0088285677186542e-05, "loss": 2.472, "step": 5848 }, { "epoch": 0.05260530729899763, "grad_norm": 3.64815616607666, "learning_rate": 2.0080179726437696e-05, "loss": 2.278, "step": 5849 }, { "epoch": 0.05261430119663808, "grad_norm": 3.3224868774414062, "learning_rate": 2.0072074313712997e-05, "loss": 1.9132, "step": 5850 }, { "epoch": 0.05262329509427853, "grad_norm": 2.6847753524780273, "learning_rate": 2.0063969439898845e-05, "loss": 2.3384, "step": 5851 }, { "epoch": 0.052632288991918984, "grad_norm": 4.7252960205078125, "learning_rate": 2.005586510588156e-05, "loss": 2.0137, "step": 5852 }, { "epoch": 0.05264128288955943, "grad_norm": 3.1319360733032227, "learning_rate": 2.0047761312547448e-05, "loss": 2.0031, "step": 5853 }, { "epoch": 0.052650276787199884, "grad_norm": 2.961143970489502, "learning_rate": 2.003965806078271e-05, "loss": 2.3117, "step": 5854 }, { "epoch": 0.05265927068484034, "grad_norm": 3.6009321212768555, "learning_rate": 2.003155535147351e-05, "loss": 2.1423, "step": 5855 }, { "epoch": 0.052668264582480785, "grad_norm": 2.9077165126800537, "learning_rate": 2.002345318550595e-05, "loss": 2.1299, "step": 5856 }, { "epoch": 0.05267725848012124, "grad_norm": 3.178741216659546, "learning_rate": 2.0015351563766062e-05, "loss": 2.4006, "step": 5857 }, { "epoch": 0.052686252377761686, "grad_norm": 2.064624786376953, "learning_rate": 2.000725048713983e-05, "loss": 2.2532, "step": 5858 }, { "epoch": 0.05269524627540214, "grad_norm": 2.4409186840057373, "learning_rate": 1.9999149956513176e-05, "loss": 2.2846, "step": 5859 }, { "epoch": 0.052704240173042594, "grad_norm": 2.364858865737915, "learning_rate": 1.9991049972771972e-05, "loss": 2.2174, "step": 5860 }, { "epoch": 0.05271323407068304, "grad_norm": 4.749326229095459, "learning_rate": 1.9982950536802012e-05, "loss": 2.0402, "step": 5861 }, { "epoch": 0.052722227968323494, "grad_norm": 2.198763370513916, "learning_rate": 1.997485164948903e-05, "loss": 2.0979, "step": 5862 }, { "epoch": 0.05273122186596394, "grad_norm": 4.4120306968688965, "learning_rate": 1.9966753311718727e-05, "loss": 2.2338, "step": 5863 }, { "epoch": 0.052740215763604395, "grad_norm": 4.316563129425049, "learning_rate": 1.995865552437671e-05, "loss": 2.1584, "step": 5864 }, { "epoch": 0.05274920966124485, "grad_norm": 4.131170749664307, "learning_rate": 1.9950558288348542e-05, "loss": 2.3473, "step": 5865 }, { "epoch": 0.052758203558885296, "grad_norm": 3.3873345851898193, "learning_rate": 1.9942461604519728e-05, "loss": 1.9819, "step": 5866 }, { "epoch": 0.05276719745652575, "grad_norm": 2.972454786300659, "learning_rate": 1.9934365473775713e-05, "loss": 2.3795, "step": 5867 }, { "epoch": 0.052776191354166196, "grad_norm": 2.776611328125, "learning_rate": 1.9926269897001875e-05, "loss": 2.2312, "step": 5868 }, { "epoch": 0.05278518525180665, "grad_norm": 2.5899314880371094, "learning_rate": 1.9918174875083535e-05, "loss": 2.2919, "step": 5869 }, { "epoch": 0.0527941791494471, "grad_norm": 2.518409490585327, "learning_rate": 1.9910080408905944e-05, "loss": 2.2038, "step": 5870 }, { "epoch": 0.05280317304708755, "grad_norm": 3.6730000972747803, "learning_rate": 1.9901986499354303e-05, "loss": 2.1212, "step": 5871 }, { "epoch": 0.052812166944728005, "grad_norm": 3.1357250213623047, "learning_rate": 1.989389314731376e-05, "loss": 2.0398, "step": 5872 }, { "epoch": 0.05282116084236845, "grad_norm": 5.96011209487915, "learning_rate": 1.9885800353669387e-05, "loss": 1.962, "step": 5873 }, { "epoch": 0.052830154740008906, "grad_norm": 2.1971845626831055, "learning_rate": 1.98777081193062e-05, "loss": 2.4915, "step": 5874 }, { "epoch": 0.05283914863764935, "grad_norm": 2.264718532562256, "learning_rate": 1.9869616445109147e-05, "loss": 2.4473, "step": 5875 }, { "epoch": 0.052848142535289806, "grad_norm": 4.072366714477539, "learning_rate": 1.986152533196313e-05, "loss": 2.2316, "step": 5876 }, { "epoch": 0.05285713643293026, "grad_norm": 4.138383865356445, "learning_rate": 1.9853434780752973e-05, "loss": 2.3058, "step": 5877 }, { "epoch": 0.05286613033057071, "grad_norm": 2.863352060317993, "learning_rate": 1.9845344792363447e-05, "loss": 2.0543, "step": 5878 }, { "epoch": 0.05287512422821116, "grad_norm": 2.3486502170562744, "learning_rate": 1.983725536767927e-05, "loss": 1.9913, "step": 5879 }, { "epoch": 0.05288411812585161, "grad_norm": 2.6982903480529785, "learning_rate": 1.9829166507585083e-05, "loss": 2.5988, "step": 5880 }, { "epoch": 0.05289311202349206, "grad_norm": 2.5039401054382324, "learning_rate": 1.9821078212965478e-05, "loss": 2.112, "step": 5881 }, { "epoch": 0.05290210592113251, "grad_norm": 4.833558082580566, "learning_rate": 1.981299048470497e-05, "loss": 2.1317, "step": 5882 }, { "epoch": 0.05291109981877296, "grad_norm": 2.5672202110290527, "learning_rate": 1.980490332368802e-05, "loss": 1.9301, "step": 5883 }, { "epoch": 0.052920093716413416, "grad_norm": 2.910810708999634, "learning_rate": 1.9796816730799028e-05, "loss": 2.0501, "step": 5884 }, { "epoch": 0.05292908761405386, "grad_norm": 2.0272045135498047, "learning_rate": 1.9788730706922342e-05, "loss": 2.0614, "step": 5885 }, { "epoch": 0.05293808151169432, "grad_norm": 2.0120091438293457, "learning_rate": 1.978064525294223e-05, "loss": 2.0123, "step": 5886 }, { "epoch": 0.052947075409334764, "grad_norm": 3.2897281646728516, "learning_rate": 1.9772560369742904e-05, "loss": 2.185, "step": 5887 }, { "epoch": 0.05295606930697522, "grad_norm": 2.5656912326812744, "learning_rate": 1.9764476058208522e-05, "loss": 2.3726, "step": 5888 }, { "epoch": 0.05296506320461567, "grad_norm": 3.984572410583496, "learning_rate": 1.9756392319223164e-05, "loss": 2.2098, "step": 5889 }, { "epoch": 0.05297405710225612, "grad_norm": 3.2512612342834473, "learning_rate": 1.974830915367086e-05, "loss": 2.3064, "step": 5890 }, { "epoch": 0.05298305099989657, "grad_norm": 4.661564826965332, "learning_rate": 1.974022656243556e-05, "loss": 1.9023, "step": 5891 }, { "epoch": 0.05299204489753702, "grad_norm": 4.653634548187256, "learning_rate": 1.973214454640119e-05, "loss": 2.1671, "step": 5892 }, { "epoch": 0.05300103879517747, "grad_norm": 3.086078405380249, "learning_rate": 1.972406310645157e-05, "loss": 2.0522, "step": 5893 }, { "epoch": 0.05301003269281792, "grad_norm": 6.728176593780518, "learning_rate": 1.9715982243470472e-05, "loss": 2.094, "step": 5894 }, { "epoch": 0.053019026590458374, "grad_norm": 3.0216317176818848, "learning_rate": 1.9707901958341612e-05, "loss": 2.441, "step": 5895 }, { "epoch": 0.05302802048809883, "grad_norm": 2.9256327152252197, "learning_rate": 1.969982225194864e-05, "loss": 1.85, "step": 5896 }, { "epoch": 0.053037014385739274, "grad_norm": 4.436649322509766, "learning_rate": 1.9691743125175126e-05, "loss": 2.2259, "step": 5897 }, { "epoch": 0.05304600828337973, "grad_norm": 4.530359268188477, "learning_rate": 1.9683664578904614e-05, "loss": 1.8916, "step": 5898 }, { "epoch": 0.053055002181020175, "grad_norm": 3.5933330059051514, "learning_rate": 1.9675586614020553e-05, "loss": 1.9595, "step": 5899 }, { "epoch": 0.05306399607866063, "grad_norm": 3.8951146602630615, "learning_rate": 1.9667509231406334e-05, "loss": 1.9495, "step": 5900 }, { "epoch": 0.05307298997630108, "grad_norm": 2.565790891647339, "learning_rate": 1.9659432431945286e-05, "loss": 1.8899, "step": 5901 }, { "epoch": 0.05308198387394153, "grad_norm": 4.347706317901611, "learning_rate": 1.965135621652068e-05, "loss": 2.151, "step": 5902 }, { "epoch": 0.05309097777158198, "grad_norm": 3.4737772941589355, "learning_rate": 1.9643280586015703e-05, "loss": 2.4463, "step": 5903 }, { "epoch": 0.05309997166922243, "grad_norm": 6.812370777130127, "learning_rate": 1.963520554131352e-05, "loss": 2.0186, "step": 5904 }, { "epoch": 0.053108965566862884, "grad_norm": 3.846179485321045, "learning_rate": 1.962713108329719e-05, "loss": 2.7201, "step": 5905 }, { "epoch": 0.05311795946450334, "grad_norm": 2.5110576152801514, "learning_rate": 1.9619057212849733e-05, "loss": 2.0778, "step": 5906 }, { "epoch": 0.053126953362143785, "grad_norm": 3.846651554107666, "learning_rate": 1.9610983930854087e-05, "loss": 2.167, "step": 5907 }, { "epoch": 0.05313594725978424, "grad_norm": 2.514714479446411, "learning_rate": 1.9602911238193127e-05, "loss": 1.9673, "step": 5908 }, { "epoch": 0.053144941157424685, "grad_norm": 2.883697271347046, "learning_rate": 1.959483913574969e-05, "loss": 2.4502, "step": 5909 }, { "epoch": 0.05315393505506514, "grad_norm": 2.329982042312622, "learning_rate": 1.958676762440651e-05, "loss": 1.99, "step": 5910 }, { "epoch": 0.053162928952705586, "grad_norm": 3.5448689460754395, "learning_rate": 1.957869670504629e-05, "loss": 2.0935, "step": 5911 }, { "epoch": 0.05317192285034604, "grad_norm": 2.8754186630249023, "learning_rate": 1.9570626378551645e-05, "loss": 2.3728, "step": 5912 }, { "epoch": 0.053180916747986494, "grad_norm": 4.714615821838379, "learning_rate": 1.9562556645805137e-05, "loss": 2.7598, "step": 5913 }, { "epoch": 0.05318991064562694, "grad_norm": 3.912590265274048, "learning_rate": 1.955448750768926e-05, "loss": 2.0715, "step": 5914 }, { "epoch": 0.053198904543267395, "grad_norm": 3.073225498199463, "learning_rate": 1.9546418965086442e-05, "loss": 2.2443, "step": 5915 }, { "epoch": 0.05320789844090784, "grad_norm": 3.957714557647705, "learning_rate": 1.9538351018879038e-05, "loss": 2.2491, "step": 5916 }, { "epoch": 0.053216892338548295, "grad_norm": 5.073655605316162, "learning_rate": 1.953028366994936e-05, "loss": 1.7853, "step": 5917 }, { "epoch": 0.05322588623618875, "grad_norm": 2.888134479522705, "learning_rate": 1.9522216919179644e-05, "loss": 2.4395, "step": 5918 }, { "epoch": 0.053234880133829196, "grad_norm": 2.622217893600464, "learning_rate": 1.9514150767452047e-05, "loss": 2.3445, "step": 5919 }, { "epoch": 0.05324387403146965, "grad_norm": 3.7372913360595703, "learning_rate": 1.9506085215648675e-05, "loss": 2.1475, "step": 5920 }, { "epoch": 0.0532528679291101, "grad_norm": 2.2866945266723633, "learning_rate": 1.949802026465156e-05, "loss": 1.4815, "step": 5921 }, { "epoch": 0.05326186182675055, "grad_norm": 2.0815794467926025, "learning_rate": 1.9489955915342673e-05, "loss": 1.8306, "step": 5922 }, { "epoch": 0.053270855724391, "grad_norm": 2.925598621368408, "learning_rate": 1.9481892168603926e-05, "loss": 1.454, "step": 5923 }, { "epoch": 0.05327984962203145, "grad_norm": 3.3912441730499268, "learning_rate": 1.947382902531716e-05, "loss": 1.6579, "step": 5924 }, { "epoch": 0.053288843519671905, "grad_norm": 4.048408031463623, "learning_rate": 1.9465766486364143e-05, "loss": 3.0409, "step": 5925 }, { "epoch": 0.05329783741731235, "grad_norm": 3.735949754714966, "learning_rate": 1.945770455262658e-05, "loss": 1.9267, "step": 5926 }, { "epoch": 0.053306831314952806, "grad_norm": 2.437983751296997, "learning_rate": 1.9449643224986115e-05, "loss": 1.8485, "step": 5927 }, { "epoch": 0.05331582521259325, "grad_norm": 3.8901479244232178, "learning_rate": 1.9441582504324323e-05, "loss": 2.0211, "step": 5928 }, { "epoch": 0.05332481911023371, "grad_norm": 5.159158229827881, "learning_rate": 1.94335223915227e-05, "loss": 2.0369, "step": 5929 }, { "epoch": 0.05333381300787416, "grad_norm": 2.4808599948883057, "learning_rate": 1.9425462887462712e-05, "loss": 2.1936, "step": 5930 }, { "epoch": 0.05334280690551461, "grad_norm": 3.7302889823913574, "learning_rate": 1.941740399302572e-05, "loss": 2.2109, "step": 5931 }, { "epoch": 0.05335180080315506, "grad_norm": 2.901031970977783, "learning_rate": 1.9409345709093036e-05, "loss": 1.9955, "step": 5932 }, { "epoch": 0.05336079470079551, "grad_norm": 3.1095738410949707, "learning_rate": 1.940128803654589e-05, "loss": 1.95, "step": 5933 }, { "epoch": 0.05336978859843596, "grad_norm": 2.7802529335021973, "learning_rate": 1.9393230976265473e-05, "loss": 1.8145, "step": 5934 }, { "epoch": 0.05337878249607641, "grad_norm": 3.673823356628418, "learning_rate": 1.9385174529132884e-05, "loss": 2.5602, "step": 5935 }, { "epoch": 0.05338777639371686, "grad_norm": 3.1698853969573975, "learning_rate": 1.9377118696029172e-05, "loss": 1.9586, "step": 5936 }, { "epoch": 0.053396770291357316, "grad_norm": 2.5201799869537354, "learning_rate": 1.9369063477835297e-05, "loss": 2.066, "step": 5937 }, { "epoch": 0.05340576418899776, "grad_norm": 5.031500339508057, "learning_rate": 1.9361008875432185e-05, "loss": 2.3271, "step": 5938 }, { "epoch": 0.05341475808663822, "grad_norm": 2.804259777069092, "learning_rate": 1.935295488970066e-05, "loss": 1.9951, "step": 5939 }, { "epoch": 0.053423751984278664, "grad_norm": 4.852656841278076, "learning_rate": 1.93449015215215e-05, "loss": 1.9761, "step": 5940 }, { "epoch": 0.05343274588191912, "grad_norm": 3.148024559020996, "learning_rate": 1.9336848771775397e-05, "loss": 2.2564, "step": 5941 }, { "epoch": 0.05344173977955957, "grad_norm": 1.7477471828460693, "learning_rate": 1.9328796641343006e-05, "loss": 2.3971, "step": 5942 }, { "epoch": 0.05345073367720002, "grad_norm": 2.3302133083343506, "learning_rate": 1.9320745131104887e-05, "loss": 1.9589, "step": 5943 }, { "epoch": 0.05345972757484047, "grad_norm": 3.1147549152374268, "learning_rate": 1.9312694241941544e-05, "loss": 2.1395, "step": 5944 }, { "epoch": 0.05346872147248092, "grad_norm": 5.3170905113220215, "learning_rate": 1.930464397473341e-05, "loss": 2.2072, "step": 5945 }, { "epoch": 0.05347771537012137, "grad_norm": 3.291781425476074, "learning_rate": 1.929659433036084e-05, "loss": 2.512, "step": 5946 }, { "epoch": 0.05348670926776183, "grad_norm": 3.2435262203216553, "learning_rate": 1.928854530970415e-05, "loss": 2.1909, "step": 5947 }, { "epoch": 0.053495703165402274, "grad_norm": 2.1753933429718018, "learning_rate": 1.9280496913643542e-05, "loss": 2.4156, "step": 5948 }, { "epoch": 0.05350469706304273, "grad_norm": 6.374592304229736, "learning_rate": 1.9272449143059202e-05, "loss": 2.3915, "step": 5949 }, { "epoch": 0.053513690960683175, "grad_norm": 3.209643840789795, "learning_rate": 1.9264401998831213e-05, "loss": 1.6616, "step": 5950 }, { "epoch": 0.05352268485832363, "grad_norm": 3.308849811553955, "learning_rate": 1.92563554818396e-05, "loss": 1.9976, "step": 5951 }, { "epoch": 0.053531678755964075, "grad_norm": 2.8092844486236572, "learning_rate": 1.9248309592964315e-05, "loss": 2.0349, "step": 5952 }, { "epoch": 0.05354067265360453, "grad_norm": 2.490687370300293, "learning_rate": 1.9240264333085245e-05, "loss": 2.4289, "step": 5953 }, { "epoch": 0.05354966655124498, "grad_norm": 2.5780608654022217, "learning_rate": 1.9232219703082196e-05, "loss": 2.1697, "step": 5954 }, { "epoch": 0.05355866044888543, "grad_norm": 2.6521811485290527, "learning_rate": 1.922417570383494e-05, "loss": 2.0677, "step": 5955 }, { "epoch": 0.053567654346525884, "grad_norm": 6.132172584533691, "learning_rate": 1.9216132336223147e-05, "loss": 2.3205, "step": 5956 }, { "epoch": 0.05357664824416633, "grad_norm": 3.6004436016082764, "learning_rate": 1.9208089601126425e-05, "loss": 1.8626, "step": 5957 }, { "epoch": 0.053585642141806784, "grad_norm": 3.3029534816741943, "learning_rate": 1.9200047499424308e-05, "loss": 2.2715, "step": 5958 }, { "epoch": 0.05359463603944724, "grad_norm": 4.757138729095459, "learning_rate": 1.9192006031996285e-05, "loss": 1.9935, "step": 5959 }, { "epoch": 0.053603629937087685, "grad_norm": 3.1795382499694824, "learning_rate": 1.9183965199721745e-05, "loss": 2.0593, "step": 5960 }, { "epoch": 0.05361262383472814, "grad_norm": 3.4605743885040283, "learning_rate": 1.917592500348003e-05, "loss": 2.3964, "step": 5961 }, { "epoch": 0.053621617732368586, "grad_norm": 3.177278518676758, "learning_rate": 1.9167885444150395e-05, "loss": 2.1966, "step": 5962 }, { "epoch": 0.05363061163000904, "grad_norm": 3.471595525741577, "learning_rate": 1.9159846522612046e-05, "loss": 1.7903, "step": 5963 }, { "epoch": 0.053639605527649487, "grad_norm": 3.5154342651367188, "learning_rate": 1.91518082397441e-05, "loss": 2.6369, "step": 5964 }, { "epoch": 0.05364859942528994, "grad_norm": 3.128739833831787, "learning_rate": 1.9143770596425615e-05, "loss": 2.1948, "step": 5965 }, { "epoch": 0.053657593322930394, "grad_norm": 3.7230496406555176, "learning_rate": 1.9135733593535567e-05, "loss": 2.0493, "step": 5966 }, { "epoch": 0.05366658722057084, "grad_norm": 5.097360610961914, "learning_rate": 1.912769723195287e-05, "loss": 1.9003, "step": 5967 }, { "epoch": 0.053675581118211295, "grad_norm": 5.100334167480469, "learning_rate": 1.9119661512556387e-05, "loss": 2.0346, "step": 5968 }, { "epoch": 0.05368457501585174, "grad_norm": 3.387080192565918, "learning_rate": 1.9111626436224874e-05, "loss": 2.159, "step": 5969 }, { "epoch": 0.053693568913492196, "grad_norm": 3.9910025596618652, "learning_rate": 1.9103592003837045e-05, "loss": 2.3978, "step": 5970 }, { "epoch": 0.05370256281113265, "grad_norm": 3.8852624893188477, "learning_rate": 1.9095558216271532e-05, "loss": 2.4583, "step": 5971 }, { "epoch": 0.053711556708773096, "grad_norm": 3.839998245239258, "learning_rate": 1.908752507440689e-05, "loss": 1.934, "step": 5972 }, { "epoch": 0.05372055060641355, "grad_norm": 2.8122828006744385, "learning_rate": 1.9079492579121607e-05, "loss": 2.3673, "step": 5973 }, { "epoch": 0.053729544504054, "grad_norm": 2.445458173751831, "learning_rate": 1.9071460731294127e-05, "loss": 2.3057, "step": 5974 }, { "epoch": 0.05373853840169445, "grad_norm": 4.73645544052124, "learning_rate": 1.9063429531802786e-05, "loss": 2.1474, "step": 5975 }, { "epoch": 0.053747532299334905, "grad_norm": 2.653430223464966, "learning_rate": 1.9055398981525867e-05, "loss": 2.1866, "step": 5976 }, { "epoch": 0.05375652619697535, "grad_norm": 4.3748393058776855, "learning_rate": 1.904736908134158e-05, "loss": 2.0883, "step": 5977 }, { "epoch": 0.053765520094615805, "grad_norm": 2.375842571258545, "learning_rate": 1.9039339832128057e-05, "loss": 2.189, "step": 5978 }, { "epoch": 0.05377451399225625, "grad_norm": 3.4477379322052, "learning_rate": 1.9031311234763362e-05, "loss": 2.2802, "step": 5979 }, { "epoch": 0.053783507889896706, "grad_norm": 2.676713228225708, "learning_rate": 1.90232832901255e-05, "loss": 1.9924, "step": 5980 }, { "epoch": 0.05379250178753715, "grad_norm": 4.479218006134033, "learning_rate": 1.9015255999092398e-05, "loss": 2.5817, "step": 5981 }, { "epoch": 0.05380149568517761, "grad_norm": 3.1589770317077637, "learning_rate": 1.9007229362541898e-05, "loss": 2.372, "step": 5982 }, { "epoch": 0.05381048958281806, "grad_norm": 2.972790479660034, "learning_rate": 1.899920338135178e-05, "loss": 2.0405, "step": 5983 }, { "epoch": 0.05381948348045851, "grad_norm": 2.973795175552368, "learning_rate": 1.8991178056399765e-05, "loss": 2.0696, "step": 5984 }, { "epoch": 0.05382847737809896, "grad_norm": 1.943598985671997, "learning_rate": 1.8983153388563486e-05, "loss": 2.4542, "step": 5985 }, { "epoch": 0.05383747127573941, "grad_norm": 4.116236209869385, "learning_rate": 1.8975129378720492e-05, "loss": 2.1632, "step": 5986 }, { "epoch": 0.05384646517337986, "grad_norm": 5.5338053703308105, "learning_rate": 1.8967106027748298e-05, "loss": 1.9655, "step": 5987 }, { "epoch": 0.053855459071020316, "grad_norm": 3.1980361938476562, "learning_rate": 1.8959083336524326e-05, "loss": 2.4166, "step": 5988 }, { "epoch": 0.05386445296866076, "grad_norm": 2.1550798416137695, "learning_rate": 1.8951061305925914e-05, "loss": 2.4547, "step": 5989 }, { "epoch": 0.05387344686630122, "grad_norm": 3.6016407012939453, "learning_rate": 1.8943039936830344e-05, "loss": 2.5946, "step": 5990 }, { "epoch": 0.053882440763941664, "grad_norm": 3.183762788772583, "learning_rate": 1.893501923011482e-05, "loss": 2.2472, "step": 5991 }, { "epoch": 0.05389143466158212, "grad_norm": 1.8023124933242798, "learning_rate": 1.8926999186656464e-05, "loss": 2.1573, "step": 5992 }, { "epoch": 0.053900428559222564, "grad_norm": 3.5045125484466553, "learning_rate": 1.8918979807332356e-05, "loss": 2.1342, "step": 5993 }, { "epoch": 0.05390942245686302, "grad_norm": 3.3516807556152344, "learning_rate": 1.8910961093019473e-05, "loss": 2.1709, "step": 5994 }, { "epoch": 0.05391841635450347, "grad_norm": 2.456826686859131, "learning_rate": 1.8902943044594735e-05, "loss": 1.6182, "step": 5995 }, { "epoch": 0.05392741025214392, "grad_norm": 2.9366283416748047, "learning_rate": 1.8894925662934974e-05, "loss": 2.5155, "step": 5996 }, { "epoch": 0.05393640414978437, "grad_norm": 6.0075883865356445, "learning_rate": 1.8886908948916966e-05, "loss": 2.3747, "step": 5997 }, { "epoch": 0.05394539804742482, "grad_norm": 2.925778865814209, "learning_rate": 1.8878892903417404e-05, "loss": 2.3486, "step": 5998 }, { "epoch": 0.05395439194506527, "grad_norm": 3.039971113204956, "learning_rate": 1.8870877527312897e-05, "loss": 1.7173, "step": 5999 }, { "epoch": 0.05396338584270573, "grad_norm": 2.819283962249756, "learning_rate": 1.8862862821480025e-05, "loss": 1.8989, "step": 6000 }, { "epoch": 0.053972379740346174, "grad_norm": 7.304672718048096, "learning_rate": 1.8854848786795243e-05, "loss": 2.4262, "step": 6001 }, { "epoch": 0.05398137363798663, "grad_norm": 9.951163291931152, "learning_rate": 1.8846835424134963e-05, "loss": 2.5135, "step": 6002 }, { "epoch": 0.053990367535627075, "grad_norm": 13.557953834533691, "learning_rate": 1.88388227343755e-05, "loss": 2.0572, "step": 6003 }, { "epoch": 0.05399936143326753, "grad_norm": 6.751053810119629, "learning_rate": 1.883081071839313e-05, "loss": 2.635, "step": 6004 }, { "epoch": 0.054008355330907976, "grad_norm": 3.524333953857422, "learning_rate": 1.8822799377064014e-05, "loss": 2.2634, "step": 6005 }, { "epoch": 0.05401734922854843, "grad_norm": 4.332152843475342, "learning_rate": 1.8814788711264278e-05, "loss": 2.0781, "step": 6006 }, { "epoch": 0.05402634312618888, "grad_norm": 6.693066120147705, "learning_rate": 1.8806778721869945e-05, "loss": 2.1794, "step": 6007 }, { "epoch": 0.05403533702382933, "grad_norm": 7.680394172668457, "learning_rate": 1.8798769409756985e-05, "loss": 1.8626, "step": 6008 }, { "epoch": 0.054044330921469784, "grad_norm": 11.841386795043945, "learning_rate": 1.8790760775801275e-05, "loss": 2.5473, "step": 6009 }, { "epoch": 0.05405332481911023, "grad_norm": 8.649933815002441, "learning_rate": 1.8782752820878634e-05, "loss": 2.3891, "step": 6010 }, { "epoch": 0.054062318716750685, "grad_norm": 4.398743152618408, "learning_rate": 1.8774745545864787e-05, "loss": 2.4105, "step": 6011 }, { "epoch": 0.05407131261439114, "grad_norm": 16.761505126953125, "learning_rate": 1.876673895163542e-05, "loss": 2.4557, "step": 6012 }, { "epoch": 0.054080306512031585, "grad_norm": 17.150291442871094, "learning_rate": 1.8758733039066106e-05, "loss": 2.6311, "step": 6013 }, { "epoch": 0.05408930040967204, "grad_norm": 10.904603004455566, "learning_rate": 1.8750727809032366e-05, "loss": 2.3287, "step": 6014 }, { "epoch": 0.054098294307312486, "grad_norm": 11.358471870422363, "learning_rate": 1.8742723262409634e-05, "loss": 2.5753, "step": 6015 }, { "epoch": 0.05410728820495294, "grad_norm": 4.431844711303711, "learning_rate": 1.873471940007328e-05, "loss": 2.3799, "step": 6016 }, { "epoch": 0.054116282102593394, "grad_norm": 7.443750858306885, "learning_rate": 1.8726716222898592e-05, "loss": 2.4167, "step": 6017 }, { "epoch": 0.05412527600023384, "grad_norm": 10.389894485473633, "learning_rate": 1.871871373176078e-05, "loss": 1.9794, "step": 6018 }, { "epoch": 0.054134269897874294, "grad_norm": 10.4729585647583, "learning_rate": 1.8710711927534996e-05, "loss": 1.8576, "step": 6019 }, { "epoch": 0.05414326379551474, "grad_norm": 8.599879264831543, "learning_rate": 1.87027108110963e-05, "loss": 2.1795, "step": 6020 }, { "epoch": 0.054152257693155195, "grad_norm": 5.667603492736816, "learning_rate": 1.8694710383319685e-05, "loss": 1.8987, "step": 6021 }, { "epoch": 0.05416125159079564, "grad_norm": 4.937087059020996, "learning_rate": 1.868671064508006e-05, "loss": 2.2674, "step": 6022 }, { "epoch": 0.054170245488436096, "grad_norm": 4.329404830932617, "learning_rate": 1.867871159725227e-05, "loss": 2.4177, "step": 6023 }, { "epoch": 0.05417923938607655, "grad_norm": 9.449060440063477, "learning_rate": 1.8670713240711063e-05, "loss": 1.7837, "step": 6024 }, { "epoch": 0.054188233283717, "grad_norm": 2.3096518516540527, "learning_rate": 1.866271557633115e-05, "loss": 2.2127, "step": 6025 }, { "epoch": 0.05419722718135745, "grad_norm": 8.418834686279297, "learning_rate": 1.8654718604987135e-05, "loss": 2.0982, "step": 6026 }, { "epoch": 0.0542062210789979, "grad_norm": 6.354055881500244, "learning_rate": 1.8646722327553552e-05, "loss": 1.9824, "step": 6027 }, { "epoch": 0.05421521497663835, "grad_norm": 5.896059513092041, "learning_rate": 1.863872674490486e-05, "loss": 2.3269, "step": 6028 }, { "epoch": 0.054224208874278805, "grad_norm": 5.664361000061035, "learning_rate": 1.863073185791545e-05, "loss": 2.3073, "step": 6029 }, { "epoch": 0.05423320277191925, "grad_norm": 3.62215518951416, "learning_rate": 1.8622737667459623e-05, "loss": 2.2355, "step": 6030 }, { "epoch": 0.054242196669559706, "grad_norm": 6.457131862640381, "learning_rate": 1.8614744174411626e-05, "loss": 2.0005, "step": 6031 }, { "epoch": 0.05425119056720015, "grad_norm": 8.192627906799316, "learning_rate": 1.8606751379645593e-05, "loss": 2.0235, "step": 6032 }, { "epoch": 0.054260184464840606, "grad_norm": 5.073063850402832, "learning_rate": 1.8598759284035626e-05, "loss": 2.5786, "step": 6033 }, { "epoch": 0.05426917836248105, "grad_norm": 5.402339935302734, "learning_rate": 1.8590767888455723e-05, "loss": 2.2518, "step": 6034 }, { "epoch": 0.05427817226012151, "grad_norm": 1.5450071096420288, "learning_rate": 1.8582777193779805e-05, "loss": 2.1135, "step": 6035 }, { "epoch": 0.05428716615776196, "grad_norm": 10.909832954406738, "learning_rate": 1.8574787200881725e-05, "loss": 2.8076, "step": 6036 }, { "epoch": 0.05429616005540241, "grad_norm": 8.700735092163086, "learning_rate": 1.8566797910635243e-05, "loss": 2.3504, "step": 6037 }, { "epoch": 0.05430515395304286, "grad_norm": 14.784499168395996, "learning_rate": 1.8558809323914083e-05, "loss": 2.3472, "step": 6038 }, { "epoch": 0.05431414785068331, "grad_norm": 4.354684352874756, "learning_rate": 1.8550821441591855e-05, "loss": 2.7698, "step": 6039 }, { "epoch": 0.05432314174832376, "grad_norm": 6.841946125030518, "learning_rate": 1.8542834264542092e-05, "loss": 2.393, "step": 6040 }, { "epoch": 0.054332135645964216, "grad_norm": 4.52779483795166, "learning_rate": 1.8534847793638266e-05, "loss": 2.2549, "step": 6041 }, { "epoch": 0.05434112954360466, "grad_norm": 5.8514790534973145, "learning_rate": 1.8526862029753767e-05, "loss": 2.2707, "step": 6042 }, { "epoch": 0.05435012344124512, "grad_norm": 4.042272090911865, "learning_rate": 1.851887697376189e-05, "loss": 1.9529, "step": 6043 }, { "epoch": 0.054359117338885564, "grad_norm": 14.711100578308105, "learning_rate": 1.8510892626535896e-05, "loss": 2.1362, "step": 6044 }, { "epoch": 0.05436811123652602, "grad_norm": 5.38095760345459, "learning_rate": 1.850290898894893e-05, "loss": 2.2093, "step": 6045 }, { "epoch": 0.054377105134166465, "grad_norm": 9.83702278137207, "learning_rate": 1.849492606187407e-05, "loss": 2.4001, "step": 6046 }, { "epoch": 0.05438609903180692, "grad_norm": 2.565805673599243, "learning_rate": 1.8486943846184314e-05, "loss": 2.5833, "step": 6047 }, { "epoch": 0.05439509292944737, "grad_norm": 4.7984137535095215, "learning_rate": 1.8478962342752583e-05, "loss": 2.2789, "step": 6048 }, { "epoch": 0.05440408682708782, "grad_norm": 3.1529788970947266, "learning_rate": 1.8470981552451725e-05, "loss": 1.7282, "step": 6049 }, { "epoch": 0.05441308072472827, "grad_norm": 2.83282208442688, "learning_rate": 1.8463001476154508e-05, "loss": 2.2612, "step": 6050 }, { "epoch": 0.05442207462236872, "grad_norm": 3.4133169651031494, "learning_rate": 1.8455022114733624e-05, "loss": 2.3106, "step": 6051 }, { "epoch": 0.054431068520009174, "grad_norm": 7.227272987365723, "learning_rate": 1.844704346906168e-05, "loss": 2.2538, "step": 6052 }, { "epoch": 0.05444006241764963, "grad_norm": 4.1020026206970215, "learning_rate": 1.8439065540011207e-05, "loss": 2.7532, "step": 6053 }, { "epoch": 0.054449056315290074, "grad_norm": 5.968801975250244, "learning_rate": 1.843108832845467e-05, "loss": 2.7913, "step": 6054 }, { "epoch": 0.05445805021293053, "grad_norm": 8.13157844543457, "learning_rate": 1.8423111835264434e-05, "loss": 2.2538, "step": 6055 }, { "epoch": 0.054467044110570975, "grad_norm": 15.464434623718262, "learning_rate": 1.84151360613128e-05, "loss": 2.3016, "step": 6056 }, { "epoch": 0.05447603800821143, "grad_norm": 5.024643421173096, "learning_rate": 1.8407161007471978e-05, "loss": 2.284, "step": 6057 }, { "epoch": 0.05448503190585188, "grad_norm": 5.039304256439209, "learning_rate": 1.8399186674614126e-05, "loss": 2.0873, "step": 6058 }, { "epoch": 0.05449402580349233, "grad_norm": 7.543537616729736, "learning_rate": 1.8391213063611293e-05, "loss": 2.5613, "step": 6059 }, { "epoch": 0.054503019701132784, "grad_norm": 3.511875629425049, "learning_rate": 1.8383240175335464e-05, "loss": 2.5587, "step": 6060 }, { "epoch": 0.05451201359877323, "grad_norm": 5.2346649169921875, "learning_rate": 1.8375268010658543e-05, "loss": 2.3605, "step": 6061 }, { "epoch": 0.054521007496413684, "grad_norm": 4.02094030380249, "learning_rate": 1.8367296570452338e-05, "loss": 2.613, "step": 6062 }, { "epoch": 0.05453000139405413, "grad_norm": 3.9006028175354004, "learning_rate": 1.835932585558862e-05, "loss": 2.5193, "step": 6063 }, { "epoch": 0.054538995291694585, "grad_norm": 3.9359588623046875, "learning_rate": 1.8351355866939047e-05, "loss": 1.9432, "step": 6064 }, { "epoch": 0.05454798918933504, "grad_norm": 4.796392917633057, "learning_rate": 1.8343386605375192e-05, "loss": 2.4722, "step": 6065 }, { "epoch": 0.054556983086975486, "grad_norm": 2.7313084602355957, "learning_rate": 1.8335418071768576e-05, "loss": 2.3084, "step": 6066 }, { "epoch": 0.05456597698461594, "grad_norm": 7.480556488037109, "learning_rate": 1.8327450266990616e-05, "loss": 2.2501, "step": 6067 }, { "epoch": 0.054574970882256386, "grad_norm": 2.724433660507202, "learning_rate": 1.8319483191912668e-05, "loss": 2.0772, "step": 6068 }, { "epoch": 0.05458396477989684, "grad_norm": 3.351475238800049, "learning_rate": 1.8311516847405977e-05, "loss": 2.4668, "step": 6069 }, { "epoch": 0.054592958677537294, "grad_norm": 7.964864253997803, "learning_rate": 1.8303551234341763e-05, "loss": 2.2475, "step": 6070 }, { "epoch": 0.05460195257517774, "grad_norm": 4.249490737915039, "learning_rate": 1.8295586353591115e-05, "loss": 1.7749, "step": 6071 }, { "epoch": 0.054610946472818195, "grad_norm": 3.0154213905334473, "learning_rate": 1.828762220602506e-05, "loss": 2.329, "step": 6072 }, { "epoch": 0.05461994037045864, "grad_norm": 5.099632740020752, "learning_rate": 1.8279658792514558e-05, "loss": 1.9135, "step": 6073 }, { "epoch": 0.054628934268099096, "grad_norm": 5.970974445343018, "learning_rate": 1.8271696113930457e-05, "loss": 2.2942, "step": 6074 }, { "epoch": 0.05463792816573954, "grad_norm": 4.0734782218933105, "learning_rate": 1.826373417114355e-05, "loss": 2.2406, "step": 6075 }, { "epoch": 0.054646922063379996, "grad_norm": 6.2573418617248535, "learning_rate": 1.8255772965024554e-05, "loss": 2.29, "step": 6076 }, { "epoch": 0.05465591596102045, "grad_norm": 4.31320858001709, "learning_rate": 1.824781249644409e-05, "loss": 2.342, "step": 6077 }, { "epoch": 0.0546649098586609, "grad_norm": 3.938991069793701, "learning_rate": 1.823985276627269e-05, "loss": 2.0763, "step": 6078 }, { "epoch": 0.05467390375630135, "grad_norm": 3.370471954345703, "learning_rate": 1.8231893775380833e-05, "loss": 2.2616, "step": 6079 }, { "epoch": 0.0546828976539418, "grad_norm": 3.2661211490631104, "learning_rate": 1.8223935524638898e-05, "loss": 2.0931, "step": 6080 }, { "epoch": 0.05469189155158225, "grad_norm": 2.7920873165130615, "learning_rate": 1.821597801491718e-05, "loss": 2.6897, "step": 6081 }, { "epoch": 0.054700885449222705, "grad_norm": 4.565807342529297, "learning_rate": 1.820802124708591e-05, "loss": 1.7001, "step": 6082 }, { "epoch": 0.05470987934686315, "grad_norm": 2.7950439453125, "learning_rate": 1.820006522201523e-05, "loss": 2.1778, "step": 6083 }, { "epoch": 0.054718873244503606, "grad_norm": 2.228316307067871, "learning_rate": 1.8192109940575193e-05, "loss": 1.916, "step": 6084 }, { "epoch": 0.05472786714214405, "grad_norm": 4.088411808013916, "learning_rate": 1.818415540363577e-05, "loss": 2.0031, "step": 6085 }, { "epoch": 0.05473686103978451, "grad_norm": 13.627901077270508, "learning_rate": 1.817620161206687e-05, "loss": 2.3368, "step": 6086 }, { "epoch": 0.05474585493742496, "grad_norm": 5.980828762054443, "learning_rate": 1.8168248566738308e-05, "loss": 2.5319, "step": 6087 }, { "epoch": 0.05475484883506541, "grad_norm": 4.876039505004883, "learning_rate": 1.8160296268519794e-05, "loss": 2.1051, "step": 6088 }, { "epoch": 0.05476384273270586, "grad_norm": 2.9127509593963623, "learning_rate": 1.815234471828101e-05, "loss": 2.4035, "step": 6089 }, { "epoch": 0.05477283663034631, "grad_norm": 4.533393859863281, "learning_rate": 1.814439391689151e-05, "loss": 2.1396, "step": 6090 }, { "epoch": 0.05478183052798676, "grad_norm": 4.287520408630371, "learning_rate": 1.8136443865220787e-05, "loss": 2.4961, "step": 6091 }, { "epoch": 0.05479082442562721, "grad_norm": 25.063316345214844, "learning_rate": 1.8128494564138242e-05, "loss": 2.5209, "step": 6092 }, { "epoch": 0.05479981832326766, "grad_norm": 4.598368167877197, "learning_rate": 1.81205460145132e-05, "loss": 2.3558, "step": 6093 }, { "epoch": 0.05480881222090812, "grad_norm": 5.728954315185547, "learning_rate": 1.8112598217214892e-05, "loss": 1.992, "step": 6094 }, { "epoch": 0.054817806118548563, "grad_norm": 4.898044586181641, "learning_rate": 1.8104651173112503e-05, "loss": 2.343, "step": 6095 }, { "epoch": 0.05482680001618902, "grad_norm": 3.8059258460998535, "learning_rate": 1.8096704883075087e-05, "loss": 2.4103, "step": 6096 }, { "epoch": 0.054835793913829464, "grad_norm": 3.9590299129486084, "learning_rate": 1.808875934797165e-05, "loss": 1.8176, "step": 6097 }, { "epoch": 0.05484478781146992, "grad_norm": 8.0248384475708, "learning_rate": 1.80808145686711e-05, "loss": 2.5834, "step": 6098 }, { "epoch": 0.05485378170911037, "grad_norm": 4.168548583984375, "learning_rate": 1.8072870546042262e-05, "loss": 2.3753, "step": 6099 }, { "epoch": 0.05486277560675082, "grad_norm": 2.452512741088867, "learning_rate": 1.806492728095389e-05, "loss": 2.2466, "step": 6100 }, { "epoch": 0.05487176950439127, "grad_norm": 2.9871175289154053, "learning_rate": 1.805698477427464e-05, "loss": 2.0381, "step": 6101 }, { "epoch": 0.05488076340203172, "grad_norm": 8.058927536010742, "learning_rate": 1.8049043026873096e-05, "loss": 1.7961, "step": 6102 }, { "epoch": 0.05488975729967217, "grad_norm": 10.137011528015137, "learning_rate": 1.804110203961776e-05, "loss": 2.155, "step": 6103 }, { "epoch": 0.05489875119731262, "grad_norm": 3.1005237102508545, "learning_rate": 1.803316181337704e-05, "loss": 1.9895, "step": 6104 }, { "epoch": 0.054907745094953074, "grad_norm": 5.263707637786865, "learning_rate": 1.802522234901927e-05, "loss": 2.6113, "step": 6105 }, { "epoch": 0.05491673899259353, "grad_norm": 11.342692375183105, "learning_rate": 1.8017283647412704e-05, "loss": 2.2051, "step": 6106 }, { "epoch": 0.054925732890233975, "grad_norm": 2.7123770713806152, "learning_rate": 1.8009345709425486e-05, "loss": 2.0425, "step": 6107 }, { "epoch": 0.05493472678787443, "grad_norm": 4.864616870880127, "learning_rate": 1.8001408535925717e-05, "loss": 2.6053, "step": 6108 }, { "epoch": 0.054943720685514875, "grad_norm": 6.524087429046631, "learning_rate": 1.799347212778139e-05, "loss": 2.6269, "step": 6109 }, { "epoch": 0.05495271458315533, "grad_norm": 2.931467056274414, "learning_rate": 1.798553648586042e-05, "loss": 1.8008, "step": 6110 }, { "epoch": 0.05496170848079578, "grad_norm": 3.9683563709259033, "learning_rate": 1.7977601611030633e-05, "loss": 1.8726, "step": 6111 }, { "epoch": 0.05497070237843623, "grad_norm": 16.47687339782715, "learning_rate": 1.7969667504159775e-05, "loss": 2.5009, "step": 6112 }, { "epoch": 0.054979696276076684, "grad_norm": 2.8210010528564453, "learning_rate": 1.79617341661155e-05, "loss": 2.2781, "step": 6113 }, { "epoch": 0.05498869017371713, "grad_norm": 3.1151371002197266, "learning_rate": 1.7953801597765405e-05, "loss": 1.6103, "step": 6114 }, { "epoch": 0.054997684071357585, "grad_norm": 2.713900566101074, "learning_rate": 1.7945869799976973e-05, "loss": 1.6859, "step": 6115 }, { "epoch": 0.05500667796899803, "grad_norm": 5.083333969116211, "learning_rate": 1.7937938773617614e-05, "loss": 2.8396, "step": 6116 }, { "epoch": 0.055015671866638485, "grad_norm": 7.810945987701416, "learning_rate": 1.793000851955465e-05, "loss": 2.0372, "step": 6117 }, { "epoch": 0.05502466576427894, "grad_norm": 3.8276102542877197, "learning_rate": 1.7922079038655326e-05, "loss": 2.5364, "step": 6118 }, { "epoch": 0.055033659661919386, "grad_norm": 4.416284084320068, "learning_rate": 1.79141503317868e-05, "loss": 2.1981, "step": 6119 }, { "epoch": 0.05504265355955984, "grad_norm": 4.324146270751953, "learning_rate": 1.7906222399816124e-05, "loss": 2.2855, "step": 6120 }, { "epoch": 0.05505164745720029, "grad_norm": 7.011447429656982, "learning_rate": 1.789829524361032e-05, "loss": 2.219, "step": 6121 }, { "epoch": 0.05506064135484074, "grad_norm": 3.332423448562622, "learning_rate": 1.7890368864036266e-05, "loss": 2.6731, "step": 6122 }, { "epoch": 0.055069635252481194, "grad_norm": 7.030768394470215, "learning_rate": 1.7882443261960783e-05, "loss": 2.2633, "step": 6123 }, { "epoch": 0.05507862915012164, "grad_norm": 7.151821136474609, "learning_rate": 1.7874518438250597e-05, "loss": 2.023, "step": 6124 }, { "epoch": 0.055087623047762095, "grad_norm": 3.9767112731933594, "learning_rate": 1.7866594393772373e-05, "loss": 1.8762, "step": 6125 }, { "epoch": 0.05509661694540254, "grad_norm": 5.023403167724609, "learning_rate": 1.785867112939265e-05, "loss": 2.322, "step": 6126 }, { "epoch": 0.055105610843042996, "grad_norm": 3.53586745262146, "learning_rate": 1.7850748645977923e-05, "loss": 2.3012, "step": 6127 }, { "epoch": 0.05511460474068345, "grad_norm": 5.097012996673584, "learning_rate": 1.784282694439457e-05, "loss": 2.1671, "step": 6128 }, { "epoch": 0.0551235986383239, "grad_norm": 3.9334752559661865, "learning_rate": 1.7834906025508906e-05, "loss": 2.5381, "step": 6129 }, { "epoch": 0.05513259253596435, "grad_norm": 2.555924654006958, "learning_rate": 1.7826985890187148e-05, "loss": 2.1858, "step": 6130 }, { "epoch": 0.0551415864336048, "grad_norm": 6.713780879974365, "learning_rate": 1.7819066539295432e-05, "loss": 2.1139, "step": 6131 }, { "epoch": 0.05515058033124525, "grad_norm": 7.448620796203613, "learning_rate": 1.7811147973699787e-05, "loss": 2.2649, "step": 6132 }, { "epoch": 0.0551595742288857, "grad_norm": 4.199689865112305, "learning_rate": 1.78032301942662e-05, "loss": 2.9175, "step": 6133 }, { "epoch": 0.05516856812652615, "grad_norm": 2.927948236465454, "learning_rate": 1.7795313201860546e-05, "loss": 2.3687, "step": 6134 }, { "epoch": 0.055177562024166606, "grad_norm": 5.304510593414307, "learning_rate": 1.7787396997348604e-05, "loss": 2.0223, "step": 6135 }, { "epoch": 0.05518655592180705, "grad_norm": 4.592012405395508, "learning_rate": 1.777948158159608e-05, "loss": 2.4265, "step": 6136 }, { "epoch": 0.055195549819447506, "grad_norm": 13.37314510345459, "learning_rate": 1.77715669554686e-05, "loss": 2.6779, "step": 6137 }, { "epoch": 0.05520454371708795, "grad_norm": 4.681512832641602, "learning_rate": 1.7763653119831684e-05, "loss": 1.9668, "step": 6138 }, { "epoch": 0.05521353761472841, "grad_norm": 7.187087535858154, "learning_rate": 1.7755740075550778e-05, "loss": 2.5967, "step": 6139 }, { "epoch": 0.05522253151236886, "grad_norm": 6.7401909828186035, "learning_rate": 1.7747827823491252e-05, "loss": 2.6364, "step": 6140 }, { "epoch": 0.05523152541000931, "grad_norm": 4.481364727020264, "learning_rate": 1.773991636451838e-05, "loss": 2.1551, "step": 6141 }, { "epoch": 0.05524051930764976, "grad_norm": 7.109669208526611, "learning_rate": 1.7732005699497327e-05, "loss": 2.5235, "step": 6142 }, { "epoch": 0.05524951320529021, "grad_norm": 3.545879364013672, "learning_rate": 1.772409582929321e-05, "loss": 2.0225, "step": 6143 }, { "epoch": 0.05525850710293066, "grad_norm": 4.691553592681885, "learning_rate": 1.7716186754771032e-05, "loss": 1.8731, "step": 6144 }, { "epoch": 0.05526750100057111, "grad_norm": 3.504979133605957, "learning_rate": 1.770827847679571e-05, "loss": 2.5045, "step": 6145 }, { "epoch": 0.05527649489821156, "grad_norm": 4.355865001678467, "learning_rate": 1.77003709962321e-05, "loss": 2.5897, "step": 6146 }, { "epoch": 0.05528548879585202, "grad_norm": 9.28866195678711, "learning_rate": 1.7692464313944944e-05, "loss": 2.2181, "step": 6147 }, { "epoch": 0.055294482693492464, "grad_norm": 5.9617133140563965, "learning_rate": 1.768455843079891e-05, "loss": 2.0338, "step": 6148 }, { "epoch": 0.05530347659113292, "grad_norm": 3.636101722717285, "learning_rate": 1.7676653347658552e-05, "loss": 2.3277, "step": 6149 }, { "epoch": 0.055312470488773365, "grad_norm": 3.0552401542663574, "learning_rate": 1.7668749065388385e-05, "loss": 2.549, "step": 6150 }, { "epoch": 0.05532146438641382, "grad_norm": 2.9611663818359375, "learning_rate": 1.7660845584852792e-05, "loss": 2.4492, "step": 6151 }, { "epoch": 0.05533045828405427, "grad_norm": 6.0576581954956055, "learning_rate": 1.7652942906916093e-05, "loss": 2.1098, "step": 6152 }, { "epoch": 0.05533945218169472, "grad_norm": 3.223811626434326, "learning_rate": 1.764504103244251e-05, "loss": 2.1461, "step": 6153 }, { "epoch": 0.05534844607933517, "grad_norm": 1.0834081172943115, "learning_rate": 1.7637139962296185e-05, "loss": 1.5731, "step": 6154 }, { "epoch": 0.05535743997697562, "grad_norm": 3.819230794906616, "learning_rate": 1.7629239697341165e-05, "loss": 1.5886, "step": 6155 }, { "epoch": 0.055366433874616074, "grad_norm": 3.4256269931793213, "learning_rate": 1.7621340238441407e-05, "loss": 2.6145, "step": 6156 }, { "epoch": 0.05537542777225652, "grad_norm": 3.5132195949554443, "learning_rate": 1.7613441586460793e-05, "loss": 1.8328, "step": 6157 }, { "epoch": 0.055384421669896974, "grad_norm": 3.5232787132263184, "learning_rate": 1.7605543742263087e-05, "loss": 2.4546, "step": 6158 }, { "epoch": 0.05539341556753743, "grad_norm": 2.2059824466705322, "learning_rate": 1.759764670671201e-05, "loss": 1.9124, "step": 6159 }, { "epoch": 0.055402409465177875, "grad_norm": 3.9182024002075195, "learning_rate": 1.758975048067116e-05, "loss": 2.0743, "step": 6160 }, { "epoch": 0.05541140336281833, "grad_norm": 8.947427749633789, "learning_rate": 1.7581855065004058e-05, "loss": 1.7376, "step": 6161 }, { "epoch": 0.055420397260458776, "grad_norm": 3.30497670173645, "learning_rate": 1.7573960460574133e-05, "loss": 2.7692, "step": 6162 }, { "epoch": 0.05542939115809923, "grad_norm": 4.163670539855957, "learning_rate": 1.7566066668244724e-05, "loss": 2.682, "step": 6163 }, { "epoch": 0.05543838505573968, "grad_norm": 4.236639976501465, "learning_rate": 1.7558173688879082e-05, "loss": 2.0745, "step": 6164 }, { "epoch": 0.05544737895338013, "grad_norm": 4.563051700592041, "learning_rate": 1.7550281523340382e-05, "loss": 1.9796, "step": 6165 }, { "epoch": 0.055456372851020584, "grad_norm": 7.269571304321289, "learning_rate": 1.7542390172491696e-05, "loss": 2.3084, "step": 6166 }, { "epoch": 0.05546536674866103, "grad_norm": 3.909863233566284, "learning_rate": 1.7534499637196012e-05, "loss": 2.0709, "step": 6167 }, { "epoch": 0.055474360646301485, "grad_norm": 2.9610226154327393, "learning_rate": 1.752660991831622e-05, "loss": 1.8371, "step": 6168 }, { "epoch": 0.05548335454394194, "grad_norm": 5.670548439025879, "learning_rate": 1.751872101671513e-05, "loss": 2.3514, "step": 6169 }, { "epoch": 0.055492348441582386, "grad_norm": 3.1092069149017334, "learning_rate": 1.7510832933255463e-05, "loss": 2.0798, "step": 6170 }, { "epoch": 0.05550134233922284, "grad_norm": 4.2349934577941895, "learning_rate": 1.7502945668799835e-05, "loss": 2.394, "step": 6171 }, { "epoch": 0.055510336236863286, "grad_norm": 5.95779275894165, "learning_rate": 1.7495059224210813e-05, "loss": 2.261, "step": 6172 }, { "epoch": 0.05551933013450374, "grad_norm": 5.954675197601318, "learning_rate": 1.7487173600350824e-05, "loss": 2.3041, "step": 6173 }, { "epoch": 0.05552832403214419, "grad_norm": 3.041757822036743, "learning_rate": 1.7479288798082233e-05, "loss": 2.0356, "step": 6174 }, { "epoch": 0.05553731792978464, "grad_norm": 3.295276403427124, "learning_rate": 1.7471404818267316e-05, "loss": 1.6031, "step": 6175 }, { "epoch": 0.055546311827425095, "grad_norm": 6.574477672576904, "learning_rate": 1.746352166176825e-05, "loss": 2.4384, "step": 6176 }, { "epoch": 0.05555530572506554, "grad_norm": 6.963632583618164, "learning_rate": 1.7455639329447118e-05, "loss": 1.7689, "step": 6177 }, { "epoch": 0.055564299622705995, "grad_norm": 3.0355803966522217, "learning_rate": 1.744775782216593e-05, "loss": 1.8126, "step": 6178 }, { "epoch": 0.05557329352034644, "grad_norm": 3.6917712688446045, "learning_rate": 1.743987714078659e-05, "loss": 2.5351, "step": 6179 }, { "epoch": 0.055582287417986896, "grad_norm": 6.15846586227417, "learning_rate": 1.7431997286170922e-05, "loss": 2.5926, "step": 6180 }, { "epoch": 0.05559128131562735, "grad_norm": 7.7755126953125, "learning_rate": 1.7424118259180654e-05, "loss": 1.8874, "step": 6181 }, { "epoch": 0.0556002752132678, "grad_norm": 5.07634973526001, "learning_rate": 1.7416240060677426e-05, "loss": 2.0601, "step": 6182 }, { "epoch": 0.05560926911090825, "grad_norm": 3.436856508255005, "learning_rate": 1.7408362691522767e-05, "loss": 2.4151, "step": 6183 }, { "epoch": 0.0556182630085487, "grad_norm": 3.730724811553955, "learning_rate": 1.740048615257816e-05, "loss": 1.9594, "step": 6184 }, { "epoch": 0.05562725690618915, "grad_norm": 4.6671671867370605, "learning_rate": 1.7392610444704966e-05, "loss": 2.2357, "step": 6185 }, { "epoch": 0.0556362508038296, "grad_norm": 2.8591582775115967, "learning_rate": 1.7384735568764456e-05, "loss": 1.6982, "step": 6186 }, { "epoch": 0.05564524470147005, "grad_norm": 4.163513660430908, "learning_rate": 1.737686152561781e-05, "loss": 1.7242, "step": 6187 }, { "epoch": 0.055654238599110506, "grad_norm": 2.225860595703125, "learning_rate": 1.736898831612613e-05, "loss": 2.44, "step": 6188 }, { "epoch": 0.05566323249675095, "grad_norm": 3.7733824253082275, "learning_rate": 1.7361115941150413e-05, "loss": 2.0726, "step": 6189 }, { "epoch": 0.05567222639439141, "grad_norm": 3.7729785442352295, "learning_rate": 1.7353244401551566e-05, "loss": 2.4875, "step": 6190 }, { "epoch": 0.055681220292031854, "grad_norm": 3.5991017818450928, "learning_rate": 1.734537369819042e-05, "loss": 1.8554, "step": 6191 }, { "epoch": 0.05569021418967231, "grad_norm": 4.247943878173828, "learning_rate": 1.73375038319277e-05, "loss": 2.1984, "step": 6192 }, { "epoch": 0.05569920808731276, "grad_norm": 4.689177989959717, "learning_rate": 1.732963480362404e-05, "loss": 2.2774, "step": 6193 }, { "epoch": 0.05570820198495321, "grad_norm": 4.382059574127197, "learning_rate": 1.7321766614139982e-05, "loss": 1.9954, "step": 6194 }, { "epoch": 0.05571719588259366, "grad_norm": 8.268787384033203, "learning_rate": 1.7313899264335985e-05, "loss": 2.5543, "step": 6195 }, { "epoch": 0.05572618978023411, "grad_norm": 2.006613254547119, "learning_rate": 1.7306032755072404e-05, "loss": 1.8727, "step": 6196 }, { "epoch": 0.05573518367787456, "grad_norm": 5.207339763641357, "learning_rate": 1.7298167087209522e-05, "loss": 1.6766, "step": 6197 }, { "epoch": 0.055744177575515017, "grad_norm": 4.194074630737305, "learning_rate": 1.7290302261607503e-05, "loss": 2.3999, "step": 6198 }, { "epoch": 0.05575317147315546, "grad_norm": 3.6086626052856445, "learning_rate": 1.7282438279126433e-05, "loss": 1.3847, "step": 6199 }, { "epoch": 0.05576216537079592, "grad_norm": 2.6929454803466797, "learning_rate": 1.7274575140626318e-05, "loss": 2.6477, "step": 6200 }, { "epoch": 0.055771159268436364, "grad_norm": 11.33925724029541, "learning_rate": 1.726671284696705e-05, "loss": 1.8838, "step": 6201 }, { "epoch": 0.05578015316607682, "grad_norm": 4.014926433563232, "learning_rate": 1.725885139900843e-05, "loss": 1.952, "step": 6202 }, { "epoch": 0.055789147063717265, "grad_norm": 2.6136364936828613, "learning_rate": 1.7250990797610185e-05, "loss": 2.0196, "step": 6203 }, { "epoch": 0.05579814096135772, "grad_norm": 2.3150646686553955, "learning_rate": 1.7243131043631937e-05, "loss": 1.8897, "step": 6204 }, { "epoch": 0.05580713485899817, "grad_norm": 4.17164421081543, "learning_rate": 1.723527213793322e-05, "loss": 2.2883, "step": 6205 }, { "epoch": 0.05581612875663862, "grad_norm": 7.681877613067627, "learning_rate": 1.7227414081373468e-05, "loss": 1.912, "step": 6206 }, { "epoch": 0.05582512265427907, "grad_norm": 3.4915213584899902, "learning_rate": 1.7219556874812026e-05, "loss": 2.2774, "step": 6207 }, { "epoch": 0.05583411655191952, "grad_norm": 2.3436014652252197, "learning_rate": 1.7211700519108143e-05, "loss": 2.0802, "step": 6208 }, { "epoch": 0.055843110449559974, "grad_norm": 5.891698837280273, "learning_rate": 1.7203845015120974e-05, "loss": 2.2399, "step": 6209 }, { "epoch": 0.05585210434720043, "grad_norm": 4.779064178466797, "learning_rate": 1.7195990363709604e-05, "loss": 2.2806, "step": 6210 }, { "epoch": 0.055861098244840875, "grad_norm": 3.13875412940979, "learning_rate": 1.7188136565732995e-05, "loss": 2.1521, "step": 6211 }, { "epoch": 0.05587009214248133, "grad_norm": 6.622696876525879, "learning_rate": 1.7180283622050025e-05, "loss": 1.7614, "step": 6212 }, { "epoch": 0.055879086040121775, "grad_norm": 2.6037750244140625, "learning_rate": 1.7172431533519477e-05, "loss": 2.3778, "step": 6213 }, { "epoch": 0.05588807993776223, "grad_norm": 4.415596008300781, "learning_rate": 1.7164580301000056e-05, "loss": 1.8963, "step": 6214 }, { "epoch": 0.055897073835402676, "grad_norm": 4.94844388961792, "learning_rate": 1.7156729925350336e-05, "loss": 2.1967, "step": 6215 }, { "epoch": 0.05590606773304313, "grad_norm": 2.023352861404419, "learning_rate": 1.714888040742885e-05, "loss": 2.3989, "step": 6216 }, { "epoch": 0.055915061630683584, "grad_norm": 6.926364898681641, "learning_rate": 1.7141031748094006e-05, "loss": 2.1744, "step": 6217 }, { "epoch": 0.05592405552832403, "grad_norm": 4.271123886108398, "learning_rate": 1.7133183948204108e-05, "loss": 2.4879, "step": 6218 }, { "epoch": 0.055933049425964484, "grad_norm": 4.143359184265137, "learning_rate": 1.7125337008617386e-05, "loss": 2.2587, "step": 6219 }, { "epoch": 0.05594204332360493, "grad_norm": 4.079606056213379, "learning_rate": 1.7117490930191965e-05, "loss": 1.9885, "step": 6220 }, { "epoch": 0.055951037221245385, "grad_norm": 4.424222946166992, "learning_rate": 1.7109645713785892e-05, "loss": 1.9039, "step": 6221 }, { "epoch": 0.05596003111888584, "grad_norm": 3.783358097076416, "learning_rate": 1.710180136025709e-05, "loss": 2.4137, "step": 6222 }, { "epoch": 0.055969025016526286, "grad_norm": 5.252933025360107, "learning_rate": 1.709395787046343e-05, "loss": 1.7854, "step": 6223 }, { "epoch": 0.05597801891416674, "grad_norm": 3.84177565574646, "learning_rate": 1.708611524526264e-05, "loss": 2.2233, "step": 6224 }, { "epoch": 0.05598701281180719, "grad_norm": 5.121371746063232, "learning_rate": 1.7078273485512392e-05, "loss": 2.4865, "step": 6225 }, { "epoch": 0.05599600670944764, "grad_norm": 2.104495048522949, "learning_rate": 1.707043259207025e-05, "loss": 1.8808, "step": 6226 }, { "epoch": 0.05600500060708809, "grad_norm": 4.3998703956604, "learning_rate": 1.706259256579367e-05, "loss": 1.9998, "step": 6227 }, { "epoch": 0.05601399450472854, "grad_norm": 5.517999172210693, "learning_rate": 1.7054753407540032e-05, "loss": 2.7527, "step": 6228 }, { "epoch": 0.056022988402368995, "grad_norm": 3.8258559703826904, "learning_rate": 1.7046915118166624e-05, "loss": 2.3202, "step": 6229 }, { "epoch": 0.05603198230000944, "grad_norm": 2.878563404083252, "learning_rate": 1.703907769853062e-05, "loss": 2.4636, "step": 6230 }, { "epoch": 0.056040976197649896, "grad_norm": 6.068768501281738, "learning_rate": 1.703124114948911e-05, "loss": 1.8837, "step": 6231 }, { "epoch": 0.05604997009529034, "grad_norm": 4.296798229217529, "learning_rate": 1.7023405471899088e-05, "loss": 1.5424, "step": 6232 }, { "epoch": 0.056058963992930796, "grad_norm": 4.783193588256836, "learning_rate": 1.701557066661745e-05, "loss": 2.0454, "step": 6233 }, { "epoch": 0.05606795789057125, "grad_norm": 4.460155963897705, "learning_rate": 1.700773673450099e-05, "loss": 2.2497, "step": 6234 }, { "epoch": 0.0560769517882117, "grad_norm": 3.320993185043335, "learning_rate": 1.699990367640643e-05, "loss": 1.874, "step": 6235 }, { "epoch": 0.05608594568585215, "grad_norm": 4.272263050079346, "learning_rate": 1.6992071493190382e-05, "loss": 1.9234, "step": 6236 }, { "epoch": 0.0560949395834926, "grad_norm": 3.5546131134033203, "learning_rate": 1.6984240185709353e-05, "loss": 2.5269, "step": 6237 }, { "epoch": 0.05610393348113305, "grad_norm": 3.1427414417266846, "learning_rate": 1.6976409754819767e-05, "loss": 1.823, "step": 6238 }, { "epoch": 0.056112927378773506, "grad_norm": 2.5580708980560303, "learning_rate": 1.6968580201377947e-05, "loss": 2.5108, "step": 6239 }, { "epoch": 0.05612192127641395, "grad_norm": 5.385694980621338, "learning_rate": 1.696075152624012e-05, "loss": 2.2419, "step": 6240 }, { "epoch": 0.056130915174054406, "grad_norm": 4.878279685974121, "learning_rate": 1.695292373026241e-05, "loss": 2.2178, "step": 6241 }, { "epoch": 0.05613990907169485, "grad_norm": 5.148347854614258, "learning_rate": 1.6945096814300874e-05, "loss": 2.4837, "step": 6242 }, { "epoch": 0.05614890296933531, "grad_norm": 7.982915878295898, "learning_rate": 1.6937270779211435e-05, "loss": 2.3491, "step": 6243 }, { "epoch": 0.056157896866975754, "grad_norm": 5.17811918258667, "learning_rate": 1.6929445625849945e-05, "loss": 2.5674, "step": 6244 }, { "epoch": 0.05616689076461621, "grad_norm": 5.508071422576904, "learning_rate": 1.6921621355072144e-05, "loss": 1.6273, "step": 6245 }, { "epoch": 0.05617588466225666, "grad_norm": 3.7769711017608643, "learning_rate": 1.6913797967733688e-05, "loss": 2.4343, "step": 6246 }, { "epoch": 0.05618487855989711, "grad_norm": 7.392800807952881, "learning_rate": 1.6905975464690127e-05, "loss": 2.2283, "step": 6247 }, { "epoch": 0.05619387245753756, "grad_norm": 2.856973886489868, "learning_rate": 1.689815384679692e-05, "loss": 1.8677, "step": 6248 }, { "epoch": 0.05620286635517801, "grad_norm": 2.9929990768432617, "learning_rate": 1.6890333114909425e-05, "loss": 1.9284, "step": 6249 }, { "epoch": 0.05621186025281846, "grad_norm": 5.688886642456055, "learning_rate": 1.6882513269882917e-05, "loss": 2.1362, "step": 6250 }, { "epoch": 0.05622085415045892, "grad_norm": 8.057849884033203, "learning_rate": 1.687469431257255e-05, "loss": 2.3547, "step": 6251 }, { "epoch": 0.056229848048099364, "grad_norm": 5.067206382751465, "learning_rate": 1.6866876243833393e-05, "loss": 2.4698, "step": 6252 }, { "epoch": 0.05623884194573982, "grad_norm": 6.078319549560547, "learning_rate": 1.6859059064520418e-05, "loss": 2.1904, "step": 6253 }, { "epoch": 0.056247835843380264, "grad_norm": 3.382539987564087, "learning_rate": 1.6851242775488506e-05, "loss": 1.9395, "step": 6254 }, { "epoch": 0.05625682974102072, "grad_norm": 4.492697238922119, "learning_rate": 1.684342737759244e-05, "loss": 2.2667, "step": 6255 }, { "epoch": 0.056265823638661165, "grad_norm": 3.0121421813964844, "learning_rate": 1.683561287168689e-05, "loss": 2.2257, "step": 6256 }, { "epoch": 0.05627481753630162, "grad_norm": 6.397670269012451, "learning_rate": 1.6827799258626442e-05, "loss": 1.6401, "step": 6257 }, { "epoch": 0.05628381143394207, "grad_norm": 5.142786979675293, "learning_rate": 1.6819986539265582e-05, "loss": 2.3146, "step": 6258 }, { "epoch": 0.05629280533158252, "grad_norm": 3.966083288192749, "learning_rate": 1.681217471445869e-05, "loss": 1.5673, "step": 6259 }, { "epoch": 0.056301799229222974, "grad_norm": 2.252007007598877, "learning_rate": 1.6804363785060056e-05, "loss": 2.6472, "step": 6260 }, { "epoch": 0.05631079312686342, "grad_norm": 6.645456790924072, "learning_rate": 1.679655375192389e-05, "loss": 2.0061, "step": 6261 }, { "epoch": 0.056319787024503874, "grad_norm": 5.090757846832275, "learning_rate": 1.678874461590426e-05, "loss": 1.6405, "step": 6262 }, { "epoch": 0.05632878092214433, "grad_norm": 4.82916259765625, "learning_rate": 1.6780936377855183e-05, "loss": 2.6684, "step": 6263 }, { "epoch": 0.056337774819784775, "grad_norm": 3.2360339164733887, "learning_rate": 1.6773129038630547e-05, "loss": 2.495, "step": 6264 }, { "epoch": 0.05634676871742523, "grad_norm": 4.154281139373779, "learning_rate": 1.6765322599084147e-05, "loss": 2.363, "step": 6265 }, { "epoch": 0.056355762615065676, "grad_norm": 0.8012027740478516, "learning_rate": 1.6757517060069676e-05, "loss": 1.3543, "step": 6266 }, { "epoch": 0.05636475651270613, "grad_norm": 5.678323745727539, "learning_rate": 1.674971242244076e-05, "loss": 2.035, "step": 6267 }, { "epoch": 0.056373750410346576, "grad_norm": 5.078115463256836, "learning_rate": 1.674190868705089e-05, "loss": 2.3291, "step": 6268 }, { "epoch": 0.05638274430798703, "grad_norm": 8.602823257446289, "learning_rate": 1.6734105854753466e-05, "loss": 2.6765, "step": 6269 }, { "epoch": 0.056391738205627484, "grad_norm": 4.027263164520264, "learning_rate": 1.6726303926401792e-05, "loss": 1.9845, "step": 6270 }, { "epoch": 0.05640073210326793, "grad_norm": 4.4878997802734375, "learning_rate": 1.6718502902849093e-05, "loss": 2.1789, "step": 6271 }, { "epoch": 0.056409726000908385, "grad_norm": 3.0274648666381836, "learning_rate": 1.6710702784948456e-05, "loss": 1.652, "step": 6272 }, { "epoch": 0.05641871989854883, "grad_norm": 2.8719937801361084, "learning_rate": 1.6702903573552904e-05, "loss": 2.5815, "step": 6273 }, { "epoch": 0.056427713796189286, "grad_norm": 7.672677516937256, "learning_rate": 1.669510526951534e-05, "loss": 2.2343, "step": 6274 }, { "epoch": 0.05643670769382974, "grad_norm": 2.6272809505462646, "learning_rate": 1.668730787368858e-05, "loss": 2.4231, "step": 6275 }, { "epoch": 0.056445701591470186, "grad_norm": 4.369339942932129, "learning_rate": 1.6679511386925337e-05, "loss": 2.2953, "step": 6276 }, { "epoch": 0.05645469548911064, "grad_norm": 3.846674680709839, "learning_rate": 1.6671715810078217e-05, "loss": 2.2339, "step": 6277 }, { "epoch": 0.05646368938675109, "grad_norm": 3.507028579711914, "learning_rate": 1.6663921143999733e-05, "loss": 1.6211, "step": 6278 }, { "epoch": 0.05647268328439154, "grad_norm": 4.3051533699035645, "learning_rate": 1.6656127389542292e-05, "loss": 1.8034, "step": 6279 }, { "epoch": 0.056481677182031995, "grad_norm": 2.1968626976013184, "learning_rate": 1.6648334547558226e-05, "loss": 2.4206, "step": 6280 }, { "epoch": 0.05649067107967244, "grad_norm": 2.873520851135254, "learning_rate": 1.6640542618899733e-05, "loss": 1.8516, "step": 6281 }, { "epoch": 0.056499664977312895, "grad_norm": 3.1706435680389404, "learning_rate": 1.6632751604418934e-05, "loss": 2.3659, "step": 6282 }, { "epoch": 0.05650865887495334, "grad_norm": 5.069432735443115, "learning_rate": 1.662496150496784e-05, "loss": 2.4273, "step": 6283 }, { "epoch": 0.056517652772593796, "grad_norm": 3.045910358428955, "learning_rate": 1.6617172321398362e-05, "loss": 2.5234, "step": 6284 }, { "epoch": 0.05652664667023424, "grad_norm": 3.667670488357544, "learning_rate": 1.6609384054562305e-05, "loss": 2.3274, "step": 6285 }, { "epoch": 0.0565356405678747, "grad_norm": 6.235454082489014, "learning_rate": 1.6601596705311402e-05, "loss": 2.3662, "step": 6286 }, { "epoch": 0.05654463446551515, "grad_norm": 7.377414703369141, "learning_rate": 1.6593810274497263e-05, "loss": 2.5859, "step": 6287 }, { "epoch": 0.0565536283631556, "grad_norm": 6.727782726287842, "learning_rate": 1.6586024762971387e-05, "loss": 1.9879, "step": 6288 }, { "epoch": 0.05656262226079605, "grad_norm": 4.334570407867432, "learning_rate": 1.6578240171585195e-05, "loss": 2.1398, "step": 6289 }, { "epoch": 0.0565716161584365, "grad_norm": 2.211258888244629, "learning_rate": 1.6570456501189998e-05, "loss": 2.4501, "step": 6290 }, { "epoch": 0.05658061005607695, "grad_norm": 2.9014675617218018, "learning_rate": 1.6562673752636996e-05, "loss": 2.6715, "step": 6291 }, { "epoch": 0.056589603953717406, "grad_norm": 3.745110511779785, "learning_rate": 1.6554891926777306e-05, "loss": 2.3761, "step": 6292 }, { "epoch": 0.05659859785135785, "grad_norm": 3.967278003692627, "learning_rate": 1.654711102446195e-05, "loss": 2.3045, "step": 6293 }, { "epoch": 0.05660759174899831, "grad_norm": 3.717013120651245, "learning_rate": 1.6539331046541812e-05, "loss": 2.3903, "step": 6294 }, { "epoch": 0.056616585646638753, "grad_norm": 3.3905861377716064, "learning_rate": 1.6531551993867717e-05, "loss": 2.5826, "step": 6295 }, { "epoch": 0.05662557954427921, "grad_norm": 3.047926425933838, "learning_rate": 1.6523773867290366e-05, "loss": 2.0254, "step": 6296 }, { "epoch": 0.056634573441919654, "grad_norm": 5.599710941314697, "learning_rate": 1.651599666766036e-05, "loss": 2.4185, "step": 6297 }, { "epoch": 0.05664356733956011, "grad_norm": 2.2959868907928467, "learning_rate": 1.650822039582819e-05, "loss": 1.4607, "step": 6298 }, { "epoch": 0.05665256123720056, "grad_norm": 14.126157760620117, "learning_rate": 1.6500445052644282e-05, "loss": 2.547, "step": 6299 }, { "epoch": 0.05666155513484101, "grad_norm": 4.543262004852295, "learning_rate": 1.6492670638958924e-05, "loss": 2.6733, "step": 6300 }, { "epoch": 0.05667054903248146, "grad_norm": 5.701849937438965, "learning_rate": 1.6484897155622315e-05, "loss": 1.9694, "step": 6301 }, { "epoch": 0.05667954293012191, "grad_norm": 3.090893507003784, "learning_rate": 1.6477124603484552e-05, "loss": 2.754, "step": 6302 }, { "epoch": 0.05668853682776236, "grad_norm": 2.155146598815918, "learning_rate": 1.6469352983395627e-05, "loss": 1.9448, "step": 6303 }, { "epoch": 0.05669753072540282, "grad_norm": 5.839915752410889, "learning_rate": 1.6461582296205425e-05, "loss": 2.0091, "step": 6304 }, { "epoch": 0.056706524623043264, "grad_norm": 4.916825771331787, "learning_rate": 1.6453812542763752e-05, "loss": 1.9836, "step": 6305 }, { "epoch": 0.05671551852068372, "grad_norm": 3.997638702392578, "learning_rate": 1.6446043723920295e-05, "loss": 2.33, "step": 6306 }, { "epoch": 0.056724512418324165, "grad_norm": 4.06272554397583, "learning_rate": 1.6438275840524634e-05, "loss": 2.2431, "step": 6307 }, { "epoch": 0.05673350631596462, "grad_norm": 2.0098609924316406, "learning_rate": 1.6430508893426256e-05, "loss": 1.7106, "step": 6308 }, { "epoch": 0.05674250021360507, "grad_norm": 7.50196647644043, "learning_rate": 1.642274288347454e-05, "loss": 2.1264, "step": 6309 }, { "epoch": 0.05675149411124552, "grad_norm": 4.136348724365234, "learning_rate": 1.641497781151877e-05, "loss": 1.7763, "step": 6310 }, { "epoch": 0.05676048800888597, "grad_norm": 6.317657947540283, "learning_rate": 1.64072136784081e-05, "loss": 2.4543, "step": 6311 }, { "epoch": 0.05676948190652642, "grad_norm": 2.451409101486206, "learning_rate": 1.6399450484991642e-05, "loss": 1.8144, "step": 6312 }, { "epoch": 0.056778475804166874, "grad_norm": 9.23572063446045, "learning_rate": 1.6391688232118342e-05, "loss": 2.0748, "step": 6313 }, { "epoch": 0.05678746970180732, "grad_norm": 3.550799608230591, "learning_rate": 1.6383926920637077e-05, "loss": 2.052, "step": 6314 }, { "epoch": 0.056796463599447775, "grad_norm": 4.379768371582031, "learning_rate": 1.6376166551396607e-05, "loss": 1.9253, "step": 6315 }, { "epoch": 0.05680545749708823, "grad_norm": 5.364646911621094, "learning_rate": 1.6368407125245598e-05, "loss": 2.4809, "step": 6316 }, { "epoch": 0.056814451394728675, "grad_norm": 4.201511383056641, "learning_rate": 1.6360648643032613e-05, "loss": 2.1885, "step": 6317 }, { "epoch": 0.05682344529236913, "grad_norm": 4.069974899291992, "learning_rate": 1.63528911056061e-05, "loss": 2.2374, "step": 6318 }, { "epoch": 0.056832439190009576, "grad_norm": 6.479613780975342, "learning_rate": 1.6345134513814408e-05, "loss": 1.6866, "step": 6319 }, { "epoch": 0.05684143308765003, "grad_norm": 3.3960678577423096, "learning_rate": 1.6337378868505805e-05, "loss": 2.3583, "step": 6320 }, { "epoch": 0.056850426985290484, "grad_norm": 2.2945594787597656, "learning_rate": 1.632962417052842e-05, "loss": 1.821, "step": 6321 }, { "epoch": 0.05685942088293093, "grad_norm": 3.793215751647949, "learning_rate": 1.63218704207303e-05, "loss": 1.9365, "step": 6322 }, { "epoch": 0.056868414780571384, "grad_norm": 2.7536258697509766, "learning_rate": 1.6314117619959372e-05, "loss": 2.183, "step": 6323 }, { "epoch": 0.05687740867821183, "grad_norm": 17.60282325744629, "learning_rate": 1.6306365769063498e-05, "loss": 2.7362, "step": 6324 }, { "epoch": 0.056886402575852285, "grad_norm": 2.1511282920837402, "learning_rate": 1.6298614868890387e-05, "loss": 2.3301, "step": 6325 }, { "epoch": 0.05689539647349273, "grad_norm": 3.6518959999084473, "learning_rate": 1.6290864920287675e-05, "loss": 2.1924, "step": 6326 }, { "epoch": 0.056904390371133186, "grad_norm": 4.96486759185791, "learning_rate": 1.628311592410288e-05, "loss": 2.3435, "step": 6327 }, { "epoch": 0.05691338426877364, "grad_norm": 4.54649019241333, "learning_rate": 1.6275367881183425e-05, "loss": 2.6072, "step": 6328 }, { "epoch": 0.05692237816641409, "grad_norm": 3.7861459255218506, "learning_rate": 1.626762079237662e-05, "loss": 1.9864, "step": 6329 }, { "epoch": 0.05693137206405454, "grad_norm": 3.0259132385253906, "learning_rate": 1.6259874658529663e-05, "loss": 1.3443, "step": 6330 }, { "epoch": 0.05694036596169499, "grad_norm": 3.917565107345581, "learning_rate": 1.6252129480489688e-05, "loss": 1.6576, "step": 6331 }, { "epoch": 0.05694935985933544, "grad_norm": 11.926139831542969, "learning_rate": 1.6244385259103674e-05, "loss": 1.8772, "step": 6332 }, { "epoch": 0.056958353756975895, "grad_norm": 3.744107961654663, "learning_rate": 1.623664199521853e-05, "loss": 2.7459, "step": 6333 }, { "epoch": 0.05696734765461634, "grad_norm": 14.886246681213379, "learning_rate": 1.6228899689681036e-05, "loss": 1.851, "step": 6334 }, { "epoch": 0.056976341552256796, "grad_norm": 6.252964973449707, "learning_rate": 1.622115834333789e-05, "loss": 2.0958, "step": 6335 }, { "epoch": 0.05698533544989724, "grad_norm": 6.40433931350708, "learning_rate": 1.621341795703566e-05, "loss": 2.6338, "step": 6336 }, { "epoch": 0.056994329347537696, "grad_norm": 2.8708767890930176, "learning_rate": 1.620567853162084e-05, "loss": 2.0983, "step": 6337 }, { "epoch": 0.05700332324517814, "grad_norm": 4.431507110595703, "learning_rate": 1.619794006793979e-05, "loss": 1.7836, "step": 6338 }, { "epoch": 0.0570123171428186, "grad_norm": 3.0669991970062256, "learning_rate": 1.6190202566838783e-05, "loss": 1.8744, "step": 6339 }, { "epoch": 0.05702131104045905, "grad_norm": 7.9771504402160645, "learning_rate": 1.6182466029163975e-05, "loss": 2.3882, "step": 6340 }, { "epoch": 0.0570303049380995, "grad_norm": 3.5082430839538574, "learning_rate": 1.6174730455761427e-05, "loss": 1.5575, "step": 6341 }, { "epoch": 0.05703929883573995, "grad_norm": 3.5434179306030273, "learning_rate": 1.6166995847477083e-05, "loss": 1.997, "step": 6342 }, { "epoch": 0.0570482927333804, "grad_norm": 3.073286533355713, "learning_rate": 1.61592622051568e-05, "loss": 2.1921, "step": 6343 }, { "epoch": 0.05705728663102085, "grad_norm": 8.204341888427734, "learning_rate": 1.6151529529646302e-05, "loss": 2.1184, "step": 6344 }, { "epoch": 0.057066280528661306, "grad_norm": 7.585928916931152, "learning_rate": 1.614379782179124e-05, "loss": 2.2737, "step": 6345 }, { "epoch": 0.05707527442630175, "grad_norm": 8.783267974853516, "learning_rate": 1.6136067082437132e-05, "loss": 1.946, "step": 6346 }, { "epoch": 0.05708426832394221, "grad_norm": 3.5299999713897705, "learning_rate": 1.6128337312429397e-05, "loss": 1.714, "step": 6347 }, { "epoch": 0.057093262221582654, "grad_norm": 3.494563341140747, "learning_rate": 1.612060851261336e-05, "loss": 2.1994, "step": 6348 }, { "epoch": 0.05710225611922311, "grad_norm": 7.800878524780273, "learning_rate": 1.6112880683834216e-05, "loss": 1.8882, "step": 6349 }, { "epoch": 0.05711125001686356, "grad_norm": 10.881048202514648, "learning_rate": 1.6105153826937085e-05, "loss": 2.7284, "step": 6350 }, { "epoch": 0.05712024391450401, "grad_norm": 4.511980056762695, "learning_rate": 1.6097427942766962e-05, "loss": 2.1339, "step": 6351 }, { "epoch": 0.05712923781214446, "grad_norm": 7.530074119567871, "learning_rate": 1.6089703032168733e-05, "loss": 2.2707, "step": 6352 }, { "epoch": 0.05713823170978491, "grad_norm": 3.705821990966797, "learning_rate": 1.608197909598719e-05, "loss": 1.5633, "step": 6353 }, { "epoch": 0.05714722560742536, "grad_norm": 3.224494218826294, "learning_rate": 1.6074256135066995e-05, "loss": 2.478, "step": 6354 }, { "epoch": 0.05715621950506581, "grad_norm": 6.442583084106445, "learning_rate": 1.6066534150252727e-05, "loss": 1.6993, "step": 6355 }, { "epoch": 0.057165213402706264, "grad_norm": 3.973078489303589, "learning_rate": 1.6058813142388856e-05, "loss": 2.0031, "step": 6356 }, { "epoch": 0.05717420730034672, "grad_norm": 2.669619560241699, "learning_rate": 1.6051093112319744e-05, "loss": 2.2535, "step": 6357 }, { "epoch": 0.057183201197987164, "grad_norm": 3.1727001667022705, "learning_rate": 1.604337406088963e-05, "loss": 2.3672, "step": 6358 }, { "epoch": 0.05719219509562762, "grad_norm": 6.631363391876221, "learning_rate": 1.6035655988942662e-05, "loss": 2.8721, "step": 6359 }, { "epoch": 0.057201188993268065, "grad_norm": 4.516825199127197, "learning_rate": 1.602793889732288e-05, "loss": 1.9502, "step": 6360 }, { "epoch": 0.05721018289090852, "grad_norm": 2.7750909328460693, "learning_rate": 1.6020222786874207e-05, "loss": 2.1416, "step": 6361 }, { "epoch": 0.05721917678854897, "grad_norm": 4.0119123458862305, "learning_rate": 1.6012507658440467e-05, "loss": 1.5765, "step": 6362 }, { "epoch": 0.05722817068618942, "grad_norm": 5.239148139953613, "learning_rate": 1.6004793512865382e-05, "loss": 2.7533, "step": 6363 }, { "epoch": 0.05723716458382987, "grad_norm": 4.162419319152832, "learning_rate": 1.5997080350992556e-05, "loss": 1.6926, "step": 6364 }, { "epoch": 0.05724615848147032, "grad_norm": 4.085282325744629, "learning_rate": 1.598936817366548e-05, "loss": 2.0764, "step": 6365 }, { "epoch": 0.057255152379110774, "grad_norm": 7.938693523406982, "learning_rate": 1.598165698172756e-05, "loss": 2.3949, "step": 6366 }, { "epoch": 0.05726414627675122, "grad_norm": 5.2295660972595215, "learning_rate": 1.597394677602207e-05, "loss": 2.0929, "step": 6367 }, { "epoch": 0.057273140174391675, "grad_norm": 5.317360877990723, "learning_rate": 1.596623755739219e-05, "loss": 2.6912, "step": 6368 }, { "epoch": 0.05728213407203213, "grad_norm": 2.8657267093658447, "learning_rate": 1.5958529326680986e-05, "loss": 2.2557, "step": 6369 }, { "epoch": 0.057291127969672576, "grad_norm": 2.67763614654541, "learning_rate": 1.5950822084731426e-05, "loss": 2.2814, "step": 6370 }, { "epoch": 0.05730012186731303, "grad_norm": 2.8247668743133545, "learning_rate": 1.594311583238636e-05, "loss": 2.3648, "step": 6371 }, { "epoch": 0.057309115764953476, "grad_norm": 2.043422222137451, "learning_rate": 1.593541057048853e-05, "loss": 2.3832, "step": 6372 }, { "epoch": 0.05731810966259393, "grad_norm": 3.060668468475342, "learning_rate": 1.592770629988057e-05, "loss": 2.0064, "step": 6373 }, { "epoch": 0.057327103560234384, "grad_norm": 12.605015754699707, "learning_rate": 1.5920003021404995e-05, "loss": 2.1928, "step": 6374 }, { "epoch": 0.05733609745787483, "grad_norm": 6.339964866638184, "learning_rate": 1.591230073590425e-05, "loss": 2.0084, "step": 6375 }, { "epoch": 0.057345091355515285, "grad_norm": 2.565493106842041, "learning_rate": 1.5904599444220636e-05, "loss": 2.2523, "step": 6376 }, { "epoch": 0.05735408525315573, "grad_norm": 3.3376078605651855, "learning_rate": 1.5896899147196353e-05, "loss": 2.6277, "step": 6377 }, { "epoch": 0.057363079150796185, "grad_norm": 5.272970676422119, "learning_rate": 1.588919984567349e-05, "loss": 2.5014, "step": 6378 }, { "epoch": 0.05737207304843663, "grad_norm": 7.7177276611328125, "learning_rate": 1.5881501540494037e-05, "loss": 1.4669, "step": 6379 }, { "epoch": 0.057381066946077086, "grad_norm": 7.856208324432373, "learning_rate": 1.5873804232499863e-05, "loss": 2.275, "step": 6380 }, { "epoch": 0.05739006084371754, "grad_norm": 9.138211250305176, "learning_rate": 1.586610792253273e-05, "loss": 2.421, "step": 6381 }, { "epoch": 0.05739905474135799, "grad_norm": 3.472724437713623, "learning_rate": 1.5858412611434307e-05, "loss": 1.9402, "step": 6382 }, { "epoch": 0.05740804863899844, "grad_norm": 4.891364097595215, "learning_rate": 1.585071830004614e-05, "loss": 1.8047, "step": 6383 }, { "epoch": 0.05741704253663889, "grad_norm": 4.313510417938232, "learning_rate": 1.5843024989209667e-05, "loss": 2.5881, "step": 6384 }, { "epoch": 0.05742603643427934, "grad_norm": 6.629606246948242, "learning_rate": 1.583533267976621e-05, "loss": 2.4628, "step": 6385 }, { "epoch": 0.057435030331919795, "grad_norm": 6.848801136016846, "learning_rate": 1.5827641372556986e-05, "loss": 2.1893, "step": 6386 }, { "epoch": 0.05744402422956024, "grad_norm": 2.43061900138855, "learning_rate": 1.5819951068423115e-05, "loss": 2.1583, "step": 6387 }, { "epoch": 0.057453018127200696, "grad_norm": 3.8574960231781006, "learning_rate": 1.5812261768205596e-05, "loss": 1.9062, "step": 6388 }, { "epoch": 0.05746201202484114, "grad_norm": 1.9891998767852783, "learning_rate": 1.5804573472745314e-05, "loss": 2.4623, "step": 6389 }, { "epoch": 0.0574710059224816, "grad_norm": 4.046595096588135, "learning_rate": 1.5796886182883053e-05, "loss": 1.7693, "step": 6390 }, { "epoch": 0.05747999982012205, "grad_norm": 4.468495845794678, "learning_rate": 1.5789199899459484e-05, "loss": 1.915, "step": 6391 }, { "epoch": 0.0574889937177625, "grad_norm": 3.4017486572265625, "learning_rate": 1.5781514623315164e-05, "loss": 2.4502, "step": 6392 }, { "epoch": 0.05749798761540295, "grad_norm": 6.849021911621094, "learning_rate": 1.577383035529054e-05, "loss": 1.7312, "step": 6393 }, { "epoch": 0.0575069815130434, "grad_norm": 4.770627975463867, "learning_rate": 1.576614709622596e-05, "loss": 1.9691, "step": 6394 }, { "epoch": 0.05751597541068385, "grad_norm": 5.1666035652160645, "learning_rate": 1.5758464846961657e-05, "loss": 2.2514, "step": 6395 }, { "epoch": 0.0575249693083243, "grad_norm": 3.0952959060668945, "learning_rate": 1.5750783608337736e-05, "loss": 1.9008, "step": 6396 }, { "epoch": 0.05753396320596475, "grad_norm": 6.497433662414551, "learning_rate": 1.5743103381194225e-05, "loss": 2.5434, "step": 6397 }, { "epoch": 0.057542957103605206, "grad_norm": 2.989345073699951, "learning_rate": 1.5735424166371e-05, "loss": 1.312, "step": 6398 }, { "epoch": 0.05755195100124565, "grad_norm": 2.3666608333587646, "learning_rate": 1.5727745964707865e-05, "loss": 2.1703, "step": 6399 }, { "epoch": 0.05756094489888611, "grad_norm": 3.789856433868408, "learning_rate": 1.5720068777044476e-05, "loss": 1.9313, "step": 6400 }, { "epoch": 0.057569938796526554, "grad_norm": 4.973116874694824, "learning_rate": 1.571239260422043e-05, "loss": 2.2163, "step": 6401 }, { "epoch": 0.05757893269416701, "grad_norm": 10.256473541259766, "learning_rate": 1.5704717447075162e-05, "loss": 1.7364, "step": 6402 }, { "epoch": 0.05758792659180746, "grad_norm": 5.157270431518555, "learning_rate": 1.5697043306448016e-05, "loss": 2.2472, "step": 6403 }, { "epoch": 0.05759692048944791, "grad_norm": 3.551457405090332, "learning_rate": 1.568937018317823e-05, "loss": 2.4677, "step": 6404 }, { "epoch": 0.05760591438708836, "grad_norm": 8.290508270263672, "learning_rate": 1.5681698078104926e-05, "loss": 2.059, "step": 6405 }, { "epoch": 0.05761490828472881, "grad_norm": 1.8788982629776, "learning_rate": 1.5674026992067096e-05, "loss": 2.1742, "step": 6406 }, { "epoch": 0.05762390218236926, "grad_norm": 4.804436206817627, "learning_rate": 1.5666356925903664e-05, "loss": 2.3304, "step": 6407 }, { "epoch": 0.05763289608000971, "grad_norm": 5.524438381195068, "learning_rate": 1.5658687880453405e-05, "loss": 2.2458, "step": 6408 }, { "epoch": 0.057641889977650164, "grad_norm": 2.8066065311431885, "learning_rate": 1.5651019856554995e-05, "loss": 1.7562, "step": 6409 }, { "epoch": 0.05765088387529062, "grad_norm": 5.64912223815918, "learning_rate": 1.5643352855046996e-05, "loss": 2.2647, "step": 6410 }, { "epoch": 0.057659877772931065, "grad_norm": 9.270191192626953, "learning_rate": 1.5635686876767858e-05, "loss": 2.2152, "step": 6411 }, { "epoch": 0.05766887167057152, "grad_norm": 6.438769817352295, "learning_rate": 1.562802192255593e-05, "loss": 2.0796, "step": 6412 }, { "epoch": 0.057677865568211965, "grad_norm": 6.587494373321533, "learning_rate": 1.5620357993249427e-05, "loss": 2.1474, "step": 6413 }, { "epoch": 0.05768685946585242, "grad_norm": 4.796075820922852, "learning_rate": 1.5612695089686478e-05, "loss": 2.0161, "step": 6414 }, { "epoch": 0.05769585336349287, "grad_norm": 4.402448654174805, "learning_rate": 1.560503321270507e-05, "loss": 2.287, "step": 6415 }, { "epoch": 0.05770484726113332, "grad_norm": 5.52233362197876, "learning_rate": 1.559737236314312e-05, "loss": 1.976, "step": 6416 }, { "epoch": 0.057713841158773774, "grad_norm": 6.2157769203186035, "learning_rate": 1.5589712541838384e-05, "loss": 2.5715, "step": 6417 }, { "epoch": 0.05772283505641422, "grad_norm": 4.753781318664551, "learning_rate": 1.558205374962854e-05, "loss": 2.1105, "step": 6418 }, { "epoch": 0.057731828954054674, "grad_norm": 3.1772708892822266, "learning_rate": 1.5574395987351125e-05, "loss": 2.4807, "step": 6419 }, { "epoch": 0.05774082285169513, "grad_norm": 5.656604766845703, "learning_rate": 1.5566739255843606e-05, "loss": 1.8713, "step": 6420 }, { "epoch": 0.057749816749335575, "grad_norm": 4.10855770111084, "learning_rate": 1.55590835559433e-05, "loss": 2.3128, "step": 6421 }, { "epoch": 0.05775881064697603, "grad_norm": 3.2940642833709717, "learning_rate": 1.5551428888487422e-05, "loss": 2.078, "step": 6422 }, { "epoch": 0.057767804544616476, "grad_norm": 4.229283332824707, "learning_rate": 1.5543775254313074e-05, "loss": 1.7538, "step": 6423 }, { "epoch": 0.05777679844225693, "grad_norm": 3.8789594173431396, "learning_rate": 1.5536122654257246e-05, "loss": 1.6592, "step": 6424 }, { "epoch": 0.05778579233989738, "grad_norm": 2.969128131866455, "learning_rate": 1.5528471089156804e-05, "loss": 1.5536, "step": 6425 }, { "epoch": 0.05779478623753783, "grad_norm": 4.253090858459473, "learning_rate": 1.5520820559848538e-05, "loss": 2.6857, "step": 6426 }, { "epoch": 0.057803780135178284, "grad_norm": 6.3369340896606445, "learning_rate": 1.5513171067169082e-05, "loss": 2.5131, "step": 6427 }, { "epoch": 0.05781277403281873, "grad_norm": 5.186905860900879, "learning_rate": 1.5505522611954975e-05, "loss": 2.714, "step": 6428 }, { "epoch": 0.057821767930459185, "grad_norm": 4.856904029846191, "learning_rate": 1.5497875195042642e-05, "loss": 1.9809, "step": 6429 }, { "epoch": 0.05783076182809963, "grad_norm": 7.233469009399414, "learning_rate": 1.549022881726839e-05, "loss": 2.7687, "step": 6430 }, { "epoch": 0.057839755725740086, "grad_norm": 4.132078170776367, "learning_rate": 1.5482583479468414e-05, "loss": 2.1031, "step": 6431 }, { "epoch": 0.05784874962338054, "grad_norm": 2.7610394954681396, "learning_rate": 1.547493918247879e-05, "loss": 2.4012, "step": 6432 }, { "epoch": 0.057857743521020986, "grad_norm": 4.067428112030029, "learning_rate": 1.546729592713551e-05, "loss": 2.005, "step": 6433 }, { "epoch": 0.05786673741866144, "grad_norm": 9.365584373474121, "learning_rate": 1.5459653714274412e-05, "loss": 2.7853, "step": 6434 }, { "epoch": 0.05787573131630189, "grad_norm": 5.561624526977539, "learning_rate": 1.5452012544731245e-05, "loss": 2.0064, "step": 6435 }, { "epoch": 0.05788472521394234, "grad_norm": 11.158077239990234, "learning_rate": 1.5444372419341625e-05, "loss": 2.6405, "step": 6436 }, { "epoch": 0.05789371911158279, "grad_norm": 5.8557353019714355, "learning_rate": 1.5436733338941072e-05, "loss": 2.7198, "step": 6437 }, { "epoch": 0.05790271300922324, "grad_norm": 9.194714546203613, "learning_rate": 1.5429095304364976e-05, "loss": 2.3391, "step": 6438 }, { "epoch": 0.057911706906863696, "grad_norm": 9.810328483581543, "learning_rate": 1.542145831644864e-05, "loss": 1.8791, "step": 6439 }, { "epoch": 0.05792070080450414, "grad_norm": 7.351736545562744, "learning_rate": 1.541382237602721e-05, "loss": 2.2691, "step": 6440 }, { "epoch": 0.057929694702144596, "grad_norm": 5.251253604888916, "learning_rate": 1.5406187483935757e-05, "loss": 1.7747, "step": 6441 }, { "epoch": 0.05793868859978504, "grad_norm": 12.92046070098877, "learning_rate": 1.5398553641009222e-05, "loss": 2.2283, "step": 6442 }, { "epoch": 0.0579476824974255, "grad_norm": 6.074610710144043, "learning_rate": 1.5390920848082424e-05, "loss": 2.7698, "step": 6443 }, { "epoch": 0.05795667639506595, "grad_norm": 3.8284146785736084, "learning_rate": 1.538328910599006e-05, "loss": 2.151, "step": 6444 }, { "epoch": 0.0579656702927064, "grad_norm": 4.638676166534424, "learning_rate": 1.537565841556676e-05, "loss": 1.4881, "step": 6445 }, { "epoch": 0.05797466419034685, "grad_norm": 5.135349750518799, "learning_rate": 1.5368028777646975e-05, "loss": 2.3986, "step": 6446 }, { "epoch": 0.0579836580879873, "grad_norm": 2.653184175491333, "learning_rate": 1.536040019306509e-05, "loss": 1.7204, "step": 6447 }, { "epoch": 0.05799265198562775, "grad_norm": 5.294336795806885, "learning_rate": 1.5352772662655342e-05, "loss": 2.1873, "step": 6448 }, { "epoch": 0.0580016458832682, "grad_norm": 2.488544464111328, "learning_rate": 1.5345146187251875e-05, "loss": 2.038, "step": 6449 }, { "epoch": 0.05801063978090865, "grad_norm": 3.485295295715332, "learning_rate": 1.5337520767688703e-05, "loss": 1.8435, "step": 6450 }, { "epoch": 0.05801963367854911, "grad_norm": 3.4115805625915527, "learning_rate": 1.532989640479972e-05, "loss": 2.2959, "step": 6451 }, { "epoch": 0.058028627576189554, "grad_norm": 2.849639892578125, "learning_rate": 1.5322273099418737e-05, "loss": 2.3693, "step": 6452 }, { "epoch": 0.05803762147383001, "grad_norm": 2.238187313079834, "learning_rate": 1.5314650852379423e-05, "loss": 2.8617, "step": 6453 }, { "epoch": 0.058046615371470454, "grad_norm": 17.722190856933594, "learning_rate": 1.5307029664515328e-05, "loss": 2.1618, "step": 6454 }, { "epoch": 0.05805560926911091, "grad_norm": 5.786515235900879, "learning_rate": 1.5299409536659895e-05, "loss": 2.2212, "step": 6455 }, { "epoch": 0.05806460316675136, "grad_norm": 3.477175712585449, "learning_rate": 1.529179046964645e-05, "loss": 2.2671, "step": 6456 }, { "epoch": 0.05807359706439181, "grad_norm": 4.283214092254639, "learning_rate": 1.5284172464308193e-05, "loss": 2.1841, "step": 6457 }, { "epoch": 0.05808259096203226, "grad_norm": 4.2658281326293945, "learning_rate": 1.527655552147824e-05, "loss": 2.546, "step": 6458 }, { "epoch": 0.05809158485967271, "grad_norm": 3.9703824520111084, "learning_rate": 1.5268939641989556e-05, "loss": 2.418, "step": 6459 }, { "epoch": 0.058100578757313164, "grad_norm": 8.798530578613281, "learning_rate": 1.5261324826675e-05, "loss": 2.0945, "step": 6460 }, { "epoch": 0.05810957265495362, "grad_norm": 8.429259300231934, "learning_rate": 1.5253711076367316e-05, "loss": 2.1197, "step": 6461 }, { "epoch": 0.058118566552594064, "grad_norm": 5.402380466461182, "learning_rate": 1.5246098391899138e-05, "loss": 2.0986, "step": 6462 }, { "epoch": 0.05812756045023452, "grad_norm": 4.003873348236084, "learning_rate": 1.5238486774102972e-05, "loss": 2.7333, "step": 6463 }, { "epoch": 0.058136554347874965, "grad_norm": 7.007231712341309, "learning_rate": 1.5230876223811219e-05, "loss": 2.439, "step": 6464 }, { "epoch": 0.05814554824551542, "grad_norm": 5.098355293273926, "learning_rate": 1.5223266741856152e-05, "loss": 2.337, "step": 6465 }, { "epoch": 0.058154542143155866, "grad_norm": 4.6742706298828125, "learning_rate": 1.521565832906994e-05, "loss": 2.1907, "step": 6466 }, { "epoch": 0.05816353604079632, "grad_norm": 2.7857227325439453, "learning_rate": 1.5208050986284623e-05, "loss": 2.3473, "step": 6467 }, { "epoch": 0.05817252993843677, "grad_norm": 6.288919448852539, "learning_rate": 1.5200444714332127e-05, "loss": 2.3483, "step": 6468 }, { "epoch": 0.05818152383607722, "grad_norm": 7.703157424926758, "learning_rate": 1.5192839514044263e-05, "loss": 2.6102, "step": 6469 }, { "epoch": 0.058190517733717674, "grad_norm": 4.280155181884766, "learning_rate": 1.5185235386252717e-05, "loss": 2.1993, "step": 6470 }, { "epoch": 0.05819951163135812, "grad_norm": 4.17839241027832, "learning_rate": 1.517763233178908e-05, "loss": 2.5368, "step": 6471 }, { "epoch": 0.058208505528998575, "grad_norm": 3.584146738052368, "learning_rate": 1.5170030351484805e-05, "loss": 2.3672, "step": 6472 }, { "epoch": 0.05821749942663903, "grad_norm": 4.340555667877197, "learning_rate": 1.5162429446171234e-05, "loss": 1.8499, "step": 6473 }, { "epoch": 0.058226493324279475, "grad_norm": 4.1534295082092285, "learning_rate": 1.5154829616679583e-05, "loss": 2.7662, "step": 6474 }, { "epoch": 0.05823548722191993, "grad_norm": 3.2985551357269287, "learning_rate": 1.5147230863840966e-05, "loss": 1.5853, "step": 6475 }, { "epoch": 0.058244481119560376, "grad_norm": 4.870946884155273, "learning_rate": 1.5139633188486357e-05, "loss": 2.1751, "step": 6476 }, { "epoch": 0.05825347501720083, "grad_norm": 6.374031066894531, "learning_rate": 1.5132036591446653e-05, "loss": 1.7208, "step": 6477 }, { "epoch": 0.05826246891484128, "grad_norm": 4.365296840667725, "learning_rate": 1.5124441073552585e-05, "loss": 2.58, "step": 6478 }, { "epoch": 0.05827146281248173, "grad_norm": 7.988781452178955, "learning_rate": 1.5116846635634796e-05, "loss": 1.9706, "step": 6479 }, { "epoch": 0.058280456710122185, "grad_norm": 4.4913554191589355, "learning_rate": 1.5109253278523799e-05, "loss": 1.726, "step": 6480 }, { "epoch": 0.05828945060776263, "grad_norm": 4.041511535644531, "learning_rate": 1.5101661003049994e-05, "loss": 2.7671, "step": 6481 }, { "epoch": 0.058298444505403085, "grad_norm": 3.009864568710327, "learning_rate": 1.5094069810043654e-05, "loss": 2.4732, "step": 6482 }, { "epoch": 0.05830743840304353, "grad_norm": 6.006809234619141, "learning_rate": 1.5086479700334943e-05, "loss": 2.3122, "step": 6483 }, { "epoch": 0.058316432300683986, "grad_norm": 4.31334924697876, "learning_rate": 1.5078890674753918e-05, "loss": 1.8829, "step": 6484 }, { "epoch": 0.05832542619832444, "grad_norm": 11.871665954589844, "learning_rate": 1.5071302734130489e-05, "loss": 2.2393, "step": 6485 }, { "epoch": 0.05833442009596489, "grad_norm": 7.814999580383301, "learning_rate": 1.5063715879294459e-05, "loss": 2.0697, "step": 6486 }, { "epoch": 0.05834341399360534, "grad_norm": 2.2591583728790283, "learning_rate": 1.5056130111075534e-05, "loss": 1.5031, "step": 6487 }, { "epoch": 0.05835240789124579, "grad_norm": 9.248136520385742, "learning_rate": 1.5048545430303262e-05, "loss": 2.6221, "step": 6488 }, { "epoch": 0.05836140178888624, "grad_norm": 3.102334976196289, "learning_rate": 1.5040961837807099e-05, "loss": 2.766, "step": 6489 }, { "epoch": 0.05837039568652669, "grad_norm": 6.284908771514893, "learning_rate": 1.5033379334416375e-05, "loss": 1.9423, "step": 6490 }, { "epoch": 0.05837938958416714, "grad_norm": 3.847289800643921, "learning_rate": 1.5025797920960305e-05, "loss": 2.4643, "step": 6491 }, { "epoch": 0.058388383481807596, "grad_norm": 5.02713680267334, "learning_rate": 1.5018217598267981e-05, "loss": 2.1751, "step": 6492 }, { "epoch": 0.05839737737944804, "grad_norm": 4.048463344573975, "learning_rate": 1.5010638367168373e-05, "loss": 1.762, "step": 6493 }, { "epoch": 0.0584063712770885, "grad_norm": 3.0955989360809326, "learning_rate": 1.5003060228490334e-05, "loss": 2.4408, "step": 6494 }, { "epoch": 0.05841536517472894, "grad_norm": 2.835211992263794, "learning_rate": 1.499548318306259e-05, "loss": 1.8357, "step": 6495 }, { "epoch": 0.0584243590723694, "grad_norm": 4.522455215454102, "learning_rate": 1.498790723171377e-05, "loss": 2.6007, "step": 6496 }, { "epoch": 0.05843335297000985, "grad_norm": 6.458254814147949, "learning_rate": 1.4980332375272366e-05, "loss": 1.7644, "step": 6497 }, { "epoch": 0.0584423468676503, "grad_norm": 3.580179452896118, "learning_rate": 1.4972758614566746e-05, "loss": 2.0327, "step": 6498 }, { "epoch": 0.05845134076529075, "grad_norm": 5.537064552307129, "learning_rate": 1.496518595042517e-05, "loss": 2.1812, "step": 6499 }, { "epoch": 0.0584603346629312, "grad_norm": 5.69214391708374, "learning_rate": 1.495761438367577e-05, "loss": 2.2545, "step": 6500 }, { "epoch": 0.05846932856057165, "grad_norm": 3.868274450302124, "learning_rate": 1.4950043915146561e-05, "loss": 2.1362, "step": 6501 }, { "epoch": 0.058478322458212106, "grad_norm": 3.4960618019104004, "learning_rate": 1.4942474545665433e-05, "loss": 2.6157, "step": 6502 }, { "epoch": 0.05848731635585255, "grad_norm": 4.483057975769043, "learning_rate": 1.4934906276060175e-05, "loss": 2.1023, "step": 6503 }, { "epoch": 0.05849631025349301, "grad_norm": 4.200279712677002, "learning_rate": 1.4927339107158437e-05, "loss": 1.9544, "step": 6504 }, { "epoch": 0.058505304151133454, "grad_norm": 4.531486511230469, "learning_rate": 1.4919773039787748e-05, "loss": 2.6981, "step": 6505 }, { "epoch": 0.05851429804877391, "grad_norm": 3.602950096130371, "learning_rate": 1.4912208074775524e-05, "loss": 2.4754, "step": 6506 }, { "epoch": 0.058523291946414355, "grad_norm": 2.764507532119751, "learning_rate": 1.4904644212949053e-05, "loss": 2.3214, "step": 6507 }, { "epoch": 0.05853228584405481, "grad_norm": 8.765809059143066, "learning_rate": 1.4897081455135514e-05, "loss": 2.2756, "step": 6508 }, { "epoch": 0.05854127974169526, "grad_norm": 5.222254753112793, "learning_rate": 1.488951980216196e-05, "loss": 1.7831, "step": 6509 }, { "epoch": 0.05855027363933571, "grad_norm": 4.788022518157959, "learning_rate": 1.4881959254855324e-05, "loss": 2.3416, "step": 6510 }, { "epoch": 0.05855926753697616, "grad_norm": 3.8831052780151367, "learning_rate": 1.4874399814042403e-05, "loss": 2.5518, "step": 6511 }, { "epoch": 0.05856826143461661, "grad_norm": 3.9307260513305664, "learning_rate": 1.48668414805499e-05, "loss": 2.2338, "step": 6512 }, { "epoch": 0.058577255332257064, "grad_norm": 2.5497562885284424, "learning_rate": 1.4859284255204381e-05, "loss": 2.8639, "step": 6513 }, { "epoch": 0.05858624922989752, "grad_norm": 3.294403553009033, "learning_rate": 1.485172813883228e-05, "loss": 2.44, "step": 6514 }, { "epoch": 0.058595243127537965, "grad_norm": 4.374127388000488, "learning_rate": 1.4844173132259933e-05, "loss": 2.7472, "step": 6515 }, { "epoch": 0.05860423702517842, "grad_norm": 2.8138504028320312, "learning_rate": 1.483661923631355e-05, "loss": 2.683, "step": 6516 }, { "epoch": 0.058613230922818865, "grad_norm": 3.8097643852233887, "learning_rate": 1.4829066451819207e-05, "loss": 2.2312, "step": 6517 }, { "epoch": 0.05862222482045932, "grad_norm": 4.731684684753418, "learning_rate": 1.4821514779602864e-05, "loss": 2.3168, "step": 6518 }, { "epoch": 0.058631218718099766, "grad_norm": 10.511028289794922, "learning_rate": 1.4813964220490361e-05, "loss": 2.697, "step": 6519 }, { "epoch": 0.05864021261574022, "grad_norm": 5.048136234283447, "learning_rate": 1.4806414775307418e-05, "loss": 2.3622, "step": 6520 }, { "epoch": 0.058649206513380674, "grad_norm": 16.656156539916992, "learning_rate": 1.4798866444879617e-05, "loss": 1.9864, "step": 6521 }, { "epoch": 0.05865820041102112, "grad_norm": 4.585415840148926, "learning_rate": 1.4791319230032458e-05, "loss": 2.4013, "step": 6522 }, { "epoch": 0.058667194308661574, "grad_norm": 6.117453575134277, "learning_rate": 1.4783773131591278e-05, "loss": 2.4546, "step": 6523 }, { "epoch": 0.05867618820630202, "grad_norm": 2.5393216609954834, "learning_rate": 1.4776228150381305e-05, "loss": 1.6819, "step": 6524 }, { "epoch": 0.058685182103942475, "grad_norm": 3.8539505004882812, "learning_rate": 1.4768684287227652e-05, "loss": 1.7792, "step": 6525 }, { "epoch": 0.05869417600158293, "grad_norm": 3.6365346908569336, "learning_rate": 1.4761141542955306e-05, "loss": 2.2172, "step": 6526 }, { "epoch": 0.058703169899223376, "grad_norm": 2.4713990688323975, "learning_rate": 1.4753599918389116e-05, "loss": 2.2741, "step": 6527 }, { "epoch": 0.05871216379686383, "grad_norm": 2.88523530960083, "learning_rate": 1.4746059414353841e-05, "loss": 1.9011, "step": 6528 }, { "epoch": 0.058721157694504277, "grad_norm": 3.8957901000976562, "learning_rate": 1.4738520031674092e-05, "loss": 2.5767, "step": 6529 }, { "epoch": 0.05873015159214473, "grad_norm": 6.907022953033447, "learning_rate": 1.4730981771174369e-05, "loss": 2.3501, "step": 6530 }, { "epoch": 0.058739145489785184, "grad_norm": 3.943615436553955, "learning_rate": 1.4723444633679035e-05, "loss": 1.8595, "step": 6531 }, { "epoch": 0.05874813938742563, "grad_norm": 6.491696357727051, "learning_rate": 1.4715908620012342e-05, "loss": 2.2155, "step": 6532 }, { "epoch": 0.058757133285066085, "grad_norm": 4.661561012268066, "learning_rate": 1.4708373730998426e-05, "loss": 2.0192, "step": 6533 }, { "epoch": 0.05876612718270653, "grad_norm": 7.086644172668457, "learning_rate": 1.470083996746128e-05, "loss": 1.8877, "step": 6534 }, { "epoch": 0.058775121080346986, "grad_norm": 5.3124589920043945, "learning_rate": 1.4693307330224798e-05, "loss": 2.0418, "step": 6535 }, { "epoch": 0.05878411497798743, "grad_norm": 2.093712568283081, "learning_rate": 1.4685775820112721e-05, "loss": 2.121, "step": 6536 }, { "epoch": 0.058793108875627886, "grad_norm": 6.454928398132324, "learning_rate": 1.4678245437948701e-05, "loss": 2.0259, "step": 6537 }, { "epoch": 0.05880210277326834, "grad_norm": 3.4132776260375977, "learning_rate": 1.467071618455624e-05, "loss": 1.7598, "step": 6538 }, { "epoch": 0.05881109667090879, "grad_norm": 3.159369707107544, "learning_rate": 1.4663188060758728e-05, "loss": 2.6303, "step": 6539 }, { "epoch": 0.05882009056854924, "grad_norm": 12.283533096313477, "learning_rate": 1.465566106737942e-05, "loss": 2.2436, "step": 6540 }, { "epoch": 0.05882908446618969, "grad_norm": 3.450676918029785, "learning_rate": 1.4648135205241476e-05, "loss": 1.9686, "step": 6541 }, { "epoch": 0.05883807836383014, "grad_norm": 3.4289867877960205, "learning_rate": 1.4640610475167898e-05, "loss": 2.7845, "step": 6542 }, { "epoch": 0.058847072261470595, "grad_norm": 3.5320725440979004, "learning_rate": 1.4633086877981584e-05, "loss": 2.7439, "step": 6543 }, { "epoch": 0.05885606615911104, "grad_norm": 2.924463987350464, "learning_rate": 1.4625564414505304e-05, "loss": 1.962, "step": 6544 }, { "epoch": 0.058865060056751496, "grad_norm": 4.800168037414551, "learning_rate": 1.4618043085561702e-05, "loss": 2.154, "step": 6545 }, { "epoch": 0.05887405395439194, "grad_norm": 3.192927837371826, "learning_rate": 1.4610522891973289e-05, "loss": 2.9809, "step": 6546 }, { "epoch": 0.0588830478520324, "grad_norm": 2.8440351486206055, "learning_rate": 1.4603003834562485e-05, "loss": 2.1899, "step": 6547 }, { "epoch": 0.058892041749672844, "grad_norm": 8.640470504760742, "learning_rate": 1.4595485914151543e-05, "loss": 3.3231, "step": 6548 }, { "epoch": 0.0589010356473133, "grad_norm": 2.520979404449463, "learning_rate": 1.4587969131562618e-05, "loss": 1.8433, "step": 6549 }, { "epoch": 0.05891002954495375, "grad_norm": 3.260313034057617, "learning_rate": 1.4580453487617745e-05, "loss": 2.5383, "step": 6550 }, { "epoch": 0.0589190234425942, "grad_norm": 4.294682502746582, "learning_rate": 1.45729389831388e-05, "loss": 2.1439, "step": 6551 }, { "epoch": 0.05892801734023465, "grad_norm": 7.047526836395264, "learning_rate": 1.4565425618947588e-05, "loss": 2.4525, "step": 6552 }, { "epoch": 0.0589370112378751, "grad_norm": 3.554302930831909, "learning_rate": 1.4557913395865714e-05, "loss": 2.0641, "step": 6553 }, { "epoch": 0.05894600513551555, "grad_norm": 7.054993629455566, "learning_rate": 1.4550402314714755e-05, "loss": 2.4487, "step": 6554 }, { "epoch": 0.05895499903315601, "grad_norm": 2.3669674396514893, "learning_rate": 1.4542892376316078e-05, "loss": 1.7475, "step": 6555 }, { "epoch": 0.058963992930796454, "grad_norm": 3.2979636192321777, "learning_rate": 1.453538358149098e-05, "loss": 2.3341, "step": 6556 }, { "epoch": 0.05897298682843691, "grad_norm": 6.816451072692871, "learning_rate": 1.4527875931060586e-05, "loss": 2.1674, "step": 6557 }, { "epoch": 0.058981980726077354, "grad_norm": 6.4530415534973145, "learning_rate": 1.452036942584595e-05, "loss": 2.324, "step": 6558 }, { "epoch": 0.05899097462371781, "grad_norm": 3.190864324569702, "learning_rate": 1.4512864066667937e-05, "loss": 2.0622, "step": 6559 }, { "epoch": 0.058999968521358255, "grad_norm": 3.5331239700317383, "learning_rate": 1.4505359854347361e-05, "loss": 2.4037, "step": 6560 }, { "epoch": 0.05900896241899871, "grad_norm": 5.109872817993164, "learning_rate": 1.4497856789704844e-05, "loss": 1.8623, "step": 6561 }, { "epoch": 0.05901795631663916, "grad_norm": 5.239011764526367, "learning_rate": 1.449035487356093e-05, "loss": 2.7325, "step": 6562 }, { "epoch": 0.05902695021427961, "grad_norm": 3.458758592605591, "learning_rate": 1.4482854106735993e-05, "loss": 2.1999, "step": 6563 }, { "epoch": 0.05903594411192006, "grad_norm": 4.038429260253906, "learning_rate": 1.447535449005033e-05, "loss": 2.1476, "step": 6564 }, { "epoch": 0.05904493800956051, "grad_norm": 5.872420310974121, "learning_rate": 1.4467856024324056e-05, "loss": 2.4791, "step": 6565 }, { "epoch": 0.059053931907200964, "grad_norm": 1.734207272529602, "learning_rate": 1.446035871037723e-05, "loss": 1.8766, "step": 6566 }, { "epoch": 0.05906292580484142, "grad_norm": 7.018714904785156, "learning_rate": 1.4452862549029722e-05, "loss": 2.4002, "step": 6567 }, { "epoch": 0.059071919702481865, "grad_norm": 4.358508110046387, "learning_rate": 1.4445367541101318e-05, "loss": 1.7868, "step": 6568 }, { "epoch": 0.05908091360012232, "grad_norm": 14.44352912902832, "learning_rate": 1.4437873687411644e-05, "loss": 2.0256, "step": 6569 }, { "epoch": 0.059089907497762766, "grad_norm": 1.8831040859222412, "learning_rate": 1.4430380988780218e-05, "loss": 1.9832, "step": 6570 }, { "epoch": 0.05909890139540322, "grad_norm": 3.3142237663269043, "learning_rate": 1.4422889446026449e-05, "loss": 2.2494, "step": 6571 }, { "epoch": 0.05910789529304367, "grad_norm": 8.769674301147461, "learning_rate": 1.4415399059969578e-05, "loss": 2.4281, "step": 6572 }, { "epoch": 0.05911688919068412, "grad_norm": 7.0004048347473145, "learning_rate": 1.4407909831428751e-05, "loss": 2.0625, "step": 6573 }, { "epoch": 0.059125883088324574, "grad_norm": 2.6451027393341064, "learning_rate": 1.4400421761222981e-05, "loss": 1.9591, "step": 6574 }, { "epoch": 0.05913487698596502, "grad_norm": 2.645962715148926, "learning_rate": 1.439293485017116e-05, "loss": 1.9665, "step": 6575 }, { "epoch": 0.059143870883605475, "grad_norm": 7.772320747375488, "learning_rate": 1.4385449099092027e-05, "loss": 2.1464, "step": 6576 }, { "epoch": 0.05915286478124592, "grad_norm": 4.249725341796875, "learning_rate": 1.4377964508804231e-05, "loss": 1.936, "step": 6577 }, { "epoch": 0.059161858678886375, "grad_norm": 7.758622646331787, "learning_rate": 1.4370481080126248e-05, "loss": 2.3128, "step": 6578 }, { "epoch": 0.05917085257652683, "grad_norm": 5.244017124176025, "learning_rate": 1.436299881387649e-05, "loss": 2.0379, "step": 6579 }, { "epoch": 0.059179846474167276, "grad_norm": 3.800048351287842, "learning_rate": 1.4355517710873184e-05, "loss": 2.1176, "step": 6580 }, { "epoch": 0.05918884037180773, "grad_norm": 3.5124905109405518, "learning_rate": 1.4348037771934464e-05, "loss": 2.3948, "step": 6581 }, { "epoch": 0.05919783426944818, "grad_norm": 1.555365800857544, "learning_rate": 1.4340558997878307e-05, "loss": 2.2754, "step": 6582 }, { "epoch": 0.05920682816708863, "grad_norm": 4.612177848815918, "learning_rate": 1.4333081389522602e-05, "loss": 1.9345, "step": 6583 }, { "epoch": 0.059215822064729084, "grad_norm": 4.147683143615723, "learning_rate": 1.4325604947685062e-05, "loss": 2.1185, "step": 6584 }, { "epoch": 0.05922481596236953, "grad_norm": 5.832060813903809, "learning_rate": 1.4318129673183333e-05, "loss": 2.5705, "step": 6585 }, { "epoch": 0.059233809860009985, "grad_norm": 4.758073806762695, "learning_rate": 1.4310655566834874e-05, "loss": 2.2271, "step": 6586 }, { "epoch": 0.05924280375765043, "grad_norm": 4.481439113616943, "learning_rate": 1.4303182629457062e-05, "loss": 2.447, "step": 6587 }, { "epoch": 0.059251797655290886, "grad_norm": 3.8208870887756348, "learning_rate": 1.4295710861867103e-05, "loss": 2.4693, "step": 6588 }, { "epoch": 0.05926079155293133, "grad_norm": 3.3445239067077637, "learning_rate": 1.4288240264882125e-05, "loss": 2.2226, "step": 6589 }, { "epoch": 0.05926978545057179, "grad_norm": 6.034364700317383, "learning_rate": 1.4280770839319071e-05, "loss": 2.7087, "step": 6590 }, { "epoch": 0.05927877934821224, "grad_norm": 4.734210968017578, "learning_rate": 1.4273302585994804e-05, "loss": 2.3791, "step": 6591 }, { "epoch": 0.05928777324585269, "grad_norm": 5.784308910369873, "learning_rate": 1.4265835505726041e-05, "loss": 2.3498, "step": 6592 }, { "epoch": 0.05929676714349314, "grad_norm": 3.0403385162353516, "learning_rate": 1.4258369599329377e-05, "loss": 2.5566, "step": 6593 }, { "epoch": 0.05930576104113359, "grad_norm": 3.769368886947632, "learning_rate": 1.4250904867621254e-05, "loss": 2.0587, "step": 6594 }, { "epoch": 0.05931475493877404, "grad_norm": 3.5701446533203125, "learning_rate": 1.4243441311418013e-05, "loss": 2.4455, "step": 6595 }, { "epoch": 0.059323748836414496, "grad_norm": 6.024693965911865, "learning_rate": 1.423597893153587e-05, "loss": 1.6643, "step": 6596 }, { "epoch": 0.05933274273405494, "grad_norm": 6.332446098327637, "learning_rate": 1.4228517728790874e-05, "loss": 1.9774, "step": 6597 }, { "epoch": 0.059341736631695396, "grad_norm": 5.166066646575928, "learning_rate": 1.4221057703998985e-05, "loss": 2.2135, "step": 6598 }, { "epoch": 0.05935073052933584, "grad_norm": 11.621291160583496, "learning_rate": 1.4213598857976024e-05, "loss": 1.8464, "step": 6599 }, { "epoch": 0.0593597244269763, "grad_norm": 8.849942207336426, "learning_rate": 1.4206141191537682e-05, "loss": 1.9833, "step": 6600 }, { "epoch": 0.059368718324616744, "grad_norm": 4.247892379760742, "learning_rate": 1.4198684705499498e-05, "loss": 1.5821, "step": 6601 }, { "epoch": 0.0593777122222572, "grad_norm": 9.980263710021973, "learning_rate": 1.4191229400676931e-05, "loss": 2.9811, "step": 6602 }, { "epoch": 0.05938670611989765, "grad_norm": 5.305501937866211, "learning_rate": 1.4183775277885253e-05, "loss": 2.1157, "step": 6603 }, { "epoch": 0.0593957000175381, "grad_norm": 3.695042133331299, "learning_rate": 1.4176322337939646e-05, "loss": 1.9219, "step": 6604 }, { "epoch": 0.05940469391517855, "grad_norm": 4.0804619789123535, "learning_rate": 1.4168870581655159e-05, "loss": 1.3798, "step": 6605 }, { "epoch": 0.059413687812819, "grad_norm": 3.8071670532226562, "learning_rate": 1.416142000984671e-05, "loss": 2.132, "step": 6606 }, { "epoch": 0.05942268171045945, "grad_norm": 7.3829569816589355, "learning_rate": 1.415397062332906e-05, "loss": 2.135, "step": 6607 }, { "epoch": 0.05943167560809991, "grad_norm": 4.027169227600098, "learning_rate": 1.4146522422916889e-05, "loss": 1.892, "step": 6608 }, { "epoch": 0.059440669505740354, "grad_norm": 2.891374349594116, "learning_rate": 1.4139075409424696e-05, "loss": 2.2203, "step": 6609 }, { "epoch": 0.05944966340338081, "grad_norm": 8.55356502532959, "learning_rate": 1.4131629583666888e-05, "loss": 2.4206, "step": 6610 }, { "epoch": 0.059458657301021255, "grad_norm": 2.446195363998413, "learning_rate": 1.412418494645773e-05, "loss": 1.9225, "step": 6611 }, { "epoch": 0.05946765119866171, "grad_norm": 3.5827271938323975, "learning_rate": 1.4116741498611364e-05, "loss": 1.7781, "step": 6612 }, { "epoch": 0.05947664509630216, "grad_norm": 4.5714826583862305, "learning_rate": 1.4109299240941776e-05, "loss": 1.6567, "step": 6613 }, { "epoch": 0.05948563899394261, "grad_norm": 4.313119411468506, "learning_rate": 1.4101858174262856e-05, "loss": 2.6751, "step": 6614 }, { "epoch": 0.05949463289158306, "grad_norm": 4.121906757354736, "learning_rate": 1.4094418299388331e-05, "loss": 2.3586, "step": 6615 }, { "epoch": 0.05950362678922351, "grad_norm": 5.232072830200195, "learning_rate": 1.4086979617131827e-05, "loss": 2.0204, "step": 6616 }, { "epoch": 0.059512620686863964, "grad_norm": 5.816551208496094, "learning_rate": 1.4079542128306822e-05, "loss": 2.2125, "step": 6617 }, { "epoch": 0.05952161458450441, "grad_norm": 2.6314053535461426, "learning_rate": 1.4072105833726684e-05, "loss": 2.1119, "step": 6618 }, { "epoch": 0.059530608482144864, "grad_norm": 4.881569862365723, "learning_rate": 1.406467073420461e-05, "loss": 1.9357, "step": 6619 }, { "epoch": 0.05953960237978532, "grad_norm": 6.612109184265137, "learning_rate": 1.4057236830553704e-05, "loss": 2.5422, "step": 6620 }, { "epoch": 0.059548596277425765, "grad_norm": 3.529949903488159, "learning_rate": 1.4049804123586935e-05, "loss": 1.8668, "step": 6621 }, { "epoch": 0.05955759017506622, "grad_norm": 2.6240127086639404, "learning_rate": 1.4042372614117116e-05, "loss": 1.5825, "step": 6622 }, { "epoch": 0.059566584072706666, "grad_norm": 6.297996997833252, "learning_rate": 1.4034942302956955e-05, "loss": 2.5857, "step": 6623 }, { "epoch": 0.05957557797034712, "grad_norm": 2.918887138366699, "learning_rate": 1.4027513190919012e-05, "loss": 2.6641, "step": 6624 }, { "epoch": 0.059584571867987574, "grad_norm": 2.6226646900177, "learning_rate": 1.4020085278815745e-05, "loss": 2.2835, "step": 6625 }, { "epoch": 0.05959356576562802, "grad_norm": 2.8378589153289795, "learning_rate": 1.401265856745943e-05, "loss": 2.6117, "step": 6626 }, { "epoch": 0.059602559663268474, "grad_norm": 5.163377285003662, "learning_rate": 1.4005233057662267e-05, "loss": 2.7497, "step": 6627 }, { "epoch": 0.05961155356090892, "grad_norm": 5.408797740936279, "learning_rate": 1.3997808750236277e-05, "loss": 2.1836, "step": 6628 }, { "epoch": 0.059620547458549375, "grad_norm": 18.94924545288086, "learning_rate": 1.3990385645993382e-05, "loss": 2.1206, "step": 6629 }, { "epoch": 0.05962954135618982, "grad_norm": 4.053044319152832, "learning_rate": 1.3982963745745352e-05, "loss": 2.4657, "step": 6630 }, { "epoch": 0.059638535253830276, "grad_norm": 3.9196386337280273, "learning_rate": 1.3975543050303863e-05, "loss": 2.0425, "step": 6631 }, { "epoch": 0.05964752915147073, "grad_norm": 21.68605613708496, "learning_rate": 1.3968123560480396e-05, "loss": 2.5129, "step": 6632 }, { "epoch": 0.059656523049111176, "grad_norm": 3.6155874729156494, "learning_rate": 1.3960705277086361e-05, "loss": 1.8861, "step": 6633 }, { "epoch": 0.05966551694675163, "grad_norm": 8.74853515625, "learning_rate": 1.3953288200932988e-05, "loss": 2.2641, "step": 6634 }, { "epoch": 0.05967451084439208, "grad_norm": 6.225401878356934, "learning_rate": 1.3945872332831412e-05, "loss": 2.0173, "step": 6635 }, { "epoch": 0.05968350474203253, "grad_norm": 7.68615198135376, "learning_rate": 1.3938457673592614e-05, "loss": 2.0978, "step": 6636 }, { "epoch": 0.059692498639672985, "grad_norm": 3.364356517791748, "learning_rate": 1.3931044224027468e-05, "loss": 2.399, "step": 6637 }, { "epoch": 0.05970149253731343, "grad_norm": 4.356878280639648, "learning_rate": 1.3923631984946672e-05, "loss": 2.6888, "step": 6638 }, { "epoch": 0.059710486434953886, "grad_norm": 2.9705333709716797, "learning_rate": 1.3916220957160836e-05, "loss": 2.1556, "step": 6639 }, { "epoch": 0.05971948033259433, "grad_norm": 2.581101894378662, "learning_rate": 1.3908811141480408e-05, "loss": 2.1667, "step": 6640 }, { "epoch": 0.059728474230234786, "grad_norm": 4.693850517272949, "learning_rate": 1.3901402538715713e-05, "loss": 2.0154, "step": 6641 }, { "epoch": 0.05973746812787524, "grad_norm": 5.574334144592285, "learning_rate": 1.3893995149676953e-05, "loss": 1.923, "step": 6642 }, { "epoch": 0.05974646202551569, "grad_norm": 2.297542095184326, "learning_rate": 1.3886588975174195e-05, "loss": 2.3115, "step": 6643 }, { "epoch": 0.05975545592315614, "grad_norm": 10.189970970153809, "learning_rate": 1.387918401601735e-05, "loss": 1.9383, "step": 6644 }, { "epoch": 0.05976444982079659, "grad_norm": 9.98621940612793, "learning_rate": 1.3871780273016215e-05, "loss": 2.1868, "step": 6645 }, { "epoch": 0.05977344371843704, "grad_norm": 5.204892158508301, "learning_rate": 1.3864377746980473e-05, "loss": 2.3644, "step": 6646 }, { "epoch": 0.05978243761607749, "grad_norm": 7.540075778961182, "learning_rate": 1.3856976438719626e-05, "loss": 2.4597, "step": 6647 }, { "epoch": 0.05979143151371794, "grad_norm": 3.979429006576538, "learning_rate": 1.3849576349043087e-05, "loss": 2.5781, "step": 6648 }, { "epoch": 0.059800425411358396, "grad_norm": 4.577551364898682, "learning_rate": 1.384217747876011e-05, "loss": 2.4911, "step": 6649 }, { "epoch": 0.05980941930899884, "grad_norm": 3.466019630432129, "learning_rate": 1.383477982867984e-05, "loss": 1.884, "step": 6650 }, { "epoch": 0.0598184132066393, "grad_norm": 4.159086227416992, "learning_rate": 1.3827383399611248e-05, "loss": 2.2366, "step": 6651 }, { "epoch": 0.059827407104279744, "grad_norm": 5.032107830047607, "learning_rate": 1.3819988192363226e-05, "loss": 2.1907, "step": 6652 }, { "epoch": 0.0598364010019202, "grad_norm": 2.288601875305176, "learning_rate": 1.3812594207744472e-05, "loss": 2.4036, "step": 6653 }, { "epoch": 0.05984539489956065, "grad_norm": 4.959178447723389, "learning_rate": 1.3805201446563595e-05, "loss": 2.2521, "step": 6654 }, { "epoch": 0.0598543887972011, "grad_norm": 10.130427360534668, "learning_rate": 1.3797809909629058e-05, "loss": 2.4338, "step": 6655 }, { "epoch": 0.05986338269484155, "grad_norm": 2.8816375732421875, "learning_rate": 1.3790419597749199e-05, "loss": 2.3066, "step": 6656 }, { "epoch": 0.059872376592482, "grad_norm": 8.22860336303711, "learning_rate": 1.3783030511732187e-05, "loss": 1.8406, "step": 6657 }, { "epoch": 0.05988137049012245, "grad_norm": 3.3090720176696777, "learning_rate": 1.3775642652386103e-05, "loss": 2.6912, "step": 6658 }, { "epoch": 0.0598903643877629, "grad_norm": 5.01676607131958, "learning_rate": 1.3768256020518854e-05, "loss": 2.4654, "step": 6659 }, { "epoch": 0.059899358285403353, "grad_norm": 1.5734846591949463, "learning_rate": 1.3760870616938248e-05, "loss": 2.1896, "step": 6660 }, { "epoch": 0.05990835218304381, "grad_norm": 4.39691162109375, "learning_rate": 1.3753486442451919e-05, "loss": 2.2704, "step": 6661 }, { "epoch": 0.059917346080684254, "grad_norm": 5.487411975860596, "learning_rate": 1.3746103497867418e-05, "loss": 1.6219, "step": 6662 }, { "epoch": 0.05992633997832471, "grad_norm": 9.105507850646973, "learning_rate": 1.3738721783992112e-05, "loss": 1.5526, "step": 6663 }, { "epoch": 0.059935333875965155, "grad_norm": 4.04994010925293, "learning_rate": 1.3731341301633268e-05, "loss": 1.9892, "step": 6664 }, { "epoch": 0.05994432777360561, "grad_norm": 4.412807464599609, "learning_rate": 1.3723962051597988e-05, "loss": 2.415, "step": 6665 }, { "epoch": 0.05995332167124606, "grad_norm": 4.893002033233643, "learning_rate": 1.3716584034693266e-05, "loss": 2.198, "step": 6666 }, { "epoch": 0.05996231556888651, "grad_norm": 4.938801288604736, "learning_rate": 1.3709207251725947e-05, "loss": 2.5041, "step": 6667 }, { "epoch": 0.05997130946652696, "grad_norm": 4.076253414154053, "learning_rate": 1.370183170350276e-05, "loss": 2.0294, "step": 6668 }, { "epoch": 0.05998030336416741, "grad_norm": 2.313560724258423, "learning_rate": 1.3694457390830261e-05, "loss": 2.2658, "step": 6669 }, { "epoch": 0.059989297261807864, "grad_norm": 6.691089630126953, "learning_rate": 1.3687084314514907e-05, "loss": 2.6875, "step": 6670 }, { "epoch": 0.05999829115944831, "grad_norm": 3.2706902027130127, "learning_rate": 1.367971247536301e-05, "loss": 2.1186, "step": 6671 }, { "epoch": 0.060007285057088765, "grad_norm": 3.358339309692383, "learning_rate": 1.3672341874180728e-05, "loss": 2.6929, "step": 6672 }, { "epoch": 0.06001627895472922, "grad_norm": 3.018707513809204, "learning_rate": 1.3664972511774116e-05, "loss": 1.8171, "step": 6673 }, { "epoch": 0.060025272852369665, "grad_norm": 4.037342071533203, "learning_rate": 1.3657604388949048e-05, "loss": 2.4228, "step": 6674 }, { "epoch": 0.06003426675001012, "grad_norm": 5.869731903076172, "learning_rate": 1.3650237506511331e-05, "loss": 2.1898, "step": 6675 }, { "epoch": 0.060043260647650566, "grad_norm": 2.7205984592437744, "learning_rate": 1.3642871865266572e-05, "loss": 1.9145, "step": 6676 }, { "epoch": 0.06005225454529102, "grad_norm": 2.426811933517456, "learning_rate": 1.3635507466020273e-05, "loss": 2.3296, "step": 6677 }, { "epoch": 0.060061248442931474, "grad_norm": 2.4901981353759766, "learning_rate": 1.3628144309577784e-05, "loss": 1.7441, "step": 6678 }, { "epoch": 0.06007024234057192, "grad_norm": 5.540495872497559, "learning_rate": 1.3620782396744346e-05, "loss": 2.0468, "step": 6679 }, { "epoch": 0.060079236238212375, "grad_norm": 5.431072235107422, "learning_rate": 1.3613421728325018e-05, "loss": 1.9034, "step": 6680 }, { "epoch": 0.06008823013585282, "grad_norm": 10.173666000366211, "learning_rate": 1.3606062305124789e-05, "loss": 2.1648, "step": 6681 }, { "epoch": 0.060097224033493275, "grad_norm": 4.310715198516846, "learning_rate": 1.3598704127948447e-05, "loss": 2.5538, "step": 6682 }, { "epoch": 0.06010621793113373, "grad_norm": 1.573772668838501, "learning_rate": 1.359134719760069e-05, "loss": 2.0654, "step": 6683 }, { "epoch": 0.060115211828774176, "grad_norm": 4.562710762023926, "learning_rate": 1.3583991514886037e-05, "loss": 2.0154, "step": 6684 }, { "epoch": 0.06012420572641463, "grad_norm": 4.234879970550537, "learning_rate": 1.3576637080608923e-05, "loss": 2.4172, "step": 6685 }, { "epoch": 0.06013319962405508, "grad_norm": 2.981372833251953, "learning_rate": 1.356928389557358e-05, "loss": 2.5298, "step": 6686 }, { "epoch": 0.06014219352169553, "grad_norm": 7.397278785705566, "learning_rate": 1.3561931960584185e-05, "loss": 2.462, "step": 6687 }, { "epoch": 0.06015118741933598, "grad_norm": 3.984812021255493, "learning_rate": 1.3554581276444705e-05, "loss": 2.4876, "step": 6688 }, { "epoch": 0.06016018131697643, "grad_norm": 4.519969940185547, "learning_rate": 1.3547231843959018e-05, "loss": 2.1906, "step": 6689 }, { "epoch": 0.060169175214616885, "grad_norm": 7.7303996086120605, "learning_rate": 1.353988366393083e-05, "loss": 2.3938, "step": 6690 }, { "epoch": 0.06017816911225733, "grad_norm": 1.5980113744735718, "learning_rate": 1.353253673716373e-05, "loss": 1.5604, "step": 6691 }, { "epoch": 0.060187163009897786, "grad_norm": 2.9618914127349854, "learning_rate": 1.3525191064461185e-05, "loss": 1.7359, "step": 6692 }, { "epoch": 0.06019615690753823, "grad_norm": 8.369169235229492, "learning_rate": 1.351784664662648e-05, "loss": 2.6451, "step": 6693 }, { "epoch": 0.06020515080517869, "grad_norm": 2.807729482650757, "learning_rate": 1.3510503484462805e-05, "loss": 2.4622, "step": 6694 }, { "epoch": 0.06021414470281914, "grad_norm": 4.781005382537842, "learning_rate": 1.3503161578773193e-05, "loss": 2.346, "step": 6695 }, { "epoch": 0.06022313860045959, "grad_norm": 2.262437105178833, "learning_rate": 1.3495820930360559e-05, "loss": 1.89, "step": 6696 }, { "epoch": 0.06023213249810004, "grad_norm": 2.44909405708313, "learning_rate": 1.3488481540027637e-05, "loss": 2.1461, "step": 6697 }, { "epoch": 0.06024112639574049, "grad_norm": 12.25643253326416, "learning_rate": 1.3481143408577079e-05, "loss": 2.5, "step": 6698 }, { "epoch": 0.06025012029338094, "grad_norm": 4.681345462799072, "learning_rate": 1.347380653681134e-05, "loss": 2.3853, "step": 6699 }, { "epoch": 0.06025911419102139, "grad_norm": 9.395800590515137, "learning_rate": 1.346647092553281e-05, "loss": 1.8218, "step": 6700 }, { "epoch": 0.06026810808866184, "grad_norm": 3.6953868865966797, "learning_rate": 1.3459136575543663e-05, "loss": 2.2711, "step": 6701 }, { "epoch": 0.060277101986302296, "grad_norm": 4.897243499755859, "learning_rate": 1.3451803487646003e-05, "loss": 1.6269, "step": 6702 }, { "epoch": 0.06028609588394274, "grad_norm": 4.148277759552002, "learning_rate": 1.3444471662641742e-05, "loss": 1.6526, "step": 6703 }, { "epoch": 0.0602950897815832, "grad_norm": 7.332465171813965, "learning_rate": 1.3437141101332695e-05, "loss": 2.4975, "step": 6704 }, { "epoch": 0.060304083679223644, "grad_norm": 4.996962547302246, "learning_rate": 1.3429811804520492e-05, "loss": 1.8216, "step": 6705 }, { "epoch": 0.0603130775768641, "grad_norm": 3.486788749694824, "learning_rate": 1.3422483773006697e-05, "loss": 2.1926, "step": 6706 }, { "epoch": 0.06032207147450455, "grad_norm": 5.07975959777832, "learning_rate": 1.341515700759266e-05, "loss": 2.0034, "step": 6707 }, { "epoch": 0.060331065372145, "grad_norm": 4.3857421875, "learning_rate": 1.3407831509079644e-05, "loss": 2.3616, "step": 6708 }, { "epoch": 0.06034005926978545, "grad_norm": 6.358922004699707, "learning_rate": 1.3400507278268733e-05, "loss": 2.0997, "step": 6709 }, { "epoch": 0.0603490531674259, "grad_norm": 4.913393974304199, "learning_rate": 1.3393184315960918e-05, "loss": 2.2105, "step": 6710 }, { "epoch": 0.06035804706506635, "grad_norm": 3.6948893070220947, "learning_rate": 1.3385862622957008e-05, "loss": 2.2795, "step": 6711 }, { "epoch": 0.0603670409627068, "grad_norm": 2.1938607692718506, "learning_rate": 1.3378542200057697e-05, "loss": 1.6747, "step": 6712 }, { "epoch": 0.060376034860347254, "grad_norm": 5.412569999694824, "learning_rate": 1.337122304806354e-05, "loss": 2.8069, "step": 6713 }, { "epoch": 0.06038502875798771, "grad_norm": 7.861708164215088, "learning_rate": 1.336390516777496e-05, "loss": 2.2429, "step": 6714 }, { "epoch": 0.060394022655628155, "grad_norm": 3.776423215866089, "learning_rate": 1.33565885599922e-05, "loss": 2.2977, "step": 6715 }, { "epoch": 0.06040301655326861, "grad_norm": 3.1470494270324707, "learning_rate": 1.3349273225515413e-05, "loss": 2.7687, "step": 6716 }, { "epoch": 0.060412010450909055, "grad_norm": 7.225902557373047, "learning_rate": 1.3341959165144599e-05, "loss": 2.8166, "step": 6717 }, { "epoch": 0.06042100434854951, "grad_norm": 6.185116291046143, "learning_rate": 1.3334646379679583e-05, "loss": 2.5204, "step": 6718 }, { "epoch": 0.06042999824618996, "grad_norm": 3.5091652870178223, "learning_rate": 1.3327334869920121e-05, "loss": 2.4476, "step": 6719 }, { "epoch": 0.06043899214383041, "grad_norm": 5.46437931060791, "learning_rate": 1.3320024636665757e-05, "loss": 2.1355, "step": 6720 }, { "epoch": 0.060447986041470864, "grad_norm": 3.542346715927124, "learning_rate": 1.3312715680715948e-05, "loss": 2.5627, "step": 6721 }, { "epoch": 0.06045697993911131, "grad_norm": 3.2476933002471924, "learning_rate": 1.3305408002869971e-05, "loss": 2.0668, "step": 6722 }, { "epoch": 0.060465973836751764, "grad_norm": 3.1255509853363037, "learning_rate": 1.3298101603927002e-05, "loss": 2.411, "step": 6723 }, { "epoch": 0.06047496773439222, "grad_norm": 3.8334414958953857, "learning_rate": 1.3290796484686038e-05, "loss": 2.1751, "step": 6724 }, { "epoch": 0.060483961632032665, "grad_norm": 3.596374034881592, "learning_rate": 1.3283492645945966e-05, "loss": 2.0883, "step": 6725 }, { "epoch": 0.06049295552967312, "grad_norm": 3.6237332820892334, "learning_rate": 1.3276190088505525e-05, "loss": 2.5753, "step": 6726 }, { "epoch": 0.060501949427313566, "grad_norm": 2.0802717208862305, "learning_rate": 1.3268888813163317e-05, "loss": 2.3907, "step": 6727 }, { "epoch": 0.06051094332495402, "grad_norm": 4.618504524230957, "learning_rate": 1.3261588820717779e-05, "loss": 2.3055, "step": 6728 }, { "epoch": 0.060519937222594467, "grad_norm": 2.7986109256744385, "learning_rate": 1.3254290111967255e-05, "loss": 1.895, "step": 6729 }, { "epoch": 0.06052893112023492, "grad_norm": 6.170854568481445, "learning_rate": 1.3246992687709889e-05, "loss": 1.6694, "step": 6730 }, { "epoch": 0.060537925017875374, "grad_norm": 4.635559558868408, "learning_rate": 1.3239696548743736e-05, "loss": 2.6097, "step": 6731 }, { "epoch": 0.06054691891551582, "grad_norm": 4.327731609344482, "learning_rate": 1.3232401695866687e-05, "loss": 2.1029, "step": 6732 }, { "epoch": 0.060555912813156275, "grad_norm": 3.452709436416626, "learning_rate": 1.32251081298765e-05, "loss": 2.1466, "step": 6733 }, { "epoch": 0.06056490671079672, "grad_norm": 4.5186052322387695, "learning_rate": 1.321781585157078e-05, "loss": 2.7571, "step": 6734 }, { "epoch": 0.060573900608437176, "grad_norm": 5.45328950881958, "learning_rate": 1.3210524861747015e-05, "loss": 1.6797, "step": 6735 }, { "epoch": 0.06058289450607763, "grad_norm": 3.9979007244110107, "learning_rate": 1.3203235161202515e-05, "loss": 2.6066, "step": 6736 }, { "epoch": 0.060591888403718076, "grad_norm": 5.838160037994385, "learning_rate": 1.3195946750734484e-05, "loss": 2.4884, "step": 6737 }, { "epoch": 0.06060088230135853, "grad_norm": 5.236480712890625, "learning_rate": 1.3188659631139966e-05, "loss": 1.9697, "step": 6738 }, { "epoch": 0.06060987619899898, "grad_norm": 3.843898296356201, "learning_rate": 1.3181373803215874e-05, "loss": 2.2098, "step": 6739 }, { "epoch": 0.06061887009663943, "grad_norm": 5.220137119293213, "learning_rate": 1.3174089267758983e-05, "loss": 2.6532, "step": 6740 }, { "epoch": 0.06062786399427988, "grad_norm": 7.312572956085205, "learning_rate": 1.3166806025565898e-05, "loss": 2.5925, "step": 6741 }, { "epoch": 0.06063685789192033, "grad_norm": 3.466684579849243, "learning_rate": 1.3159524077433127e-05, "loss": 2.3482, "step": 6742 }, { "epoch": 0.060645851789560785, "grad_norm": 4.0336198806762695, "learning_rate": 1.315224342415699e-05, "loss": 1.8843, "step": 6743 }, { "epoch": 0.06065484568720123, "grad_norm": 7.900094032287598, "learning_rate": 1.31449640665337e-05, "loss": 1.7479, "step": 6744 }, { "epoch": 0.060663839584841686, "grad_norm": 13.948092460632324, "learning_rate": 1.3137686005359318e-05, "loss": 1.9744, "step": 6745 }, { "epoch": 0.06067283348248213, "grad_norm": 3.4355990886688232, "learning_rate": 1.3130409241429767e-05, "loss": 2.3022, "step": 6746 }, { "epoch": 0.06068182738012259, "grad_norm": 2.6463842391967773, "learning_rate": 1.3123133775540809e-05, "loss": 2.0483, "step": 6747 }, { "epoch": 0.06069082127776304, "grad_norm": 3.319127082824707, "learning_rate": 1.311585960848809e-05, "loss": 2.5005, "step": 6748 }, { "epoch": 0.06069981517540349, "grad_norm": 2.6843512058258057, "learning_rate": 1.3108586741067086e-05, "loss": 2.638, "step": 6749 }, { "epoch": 0.06070880907304394, "grad_norm": 11.485901832580566, "learning_rate": 1.3101315174073162e-05, "loss": 2.6434, "step": 6750 }, { "epoch": 0.06071780297068439, "grad_norm": 1.9956883192062378, "learning_rate": 1.309404490830152e-05, "loss": 1.8708, "step": 6751 }, { "epoch": 0.06072679686832484, "grad_norm": 4.049040794372559, "learning_rate": 1.3086775944547236e-05, "loss": 2.4442, "step": 6752 }, { "epoch": 0.060735790765965296, "grad_norm": 4.61032772064209, "learning_rate": 1.3079508283605212e-05, "loss": 2.7814, "step": 6753 }, { "epoch": 0.06074478466360574, "grad_norm": 3.6996560096740723, "learning_rate": 1.307224192627025e-05, "loss": 2.298, "step": 6754 }, { "epoch": 0.0607537785612462, "grad_norm": 2.7113730907440186, "learning_rate": 1.306497687333697e-05, "loss": 2.6949, "step": 6755 }, { "epoch": 0.060762772458886644, "grad_norm": 4.238222599029541, "learning_rate": 1.3057713125599874e-05, "loss": 2.1756, "step": 6756 }, { "epoch": 0.0607717663565271, "grad_norm": 4.368277072906494, "learning_rate": 1.3050450683853319e-05, "loss": 2.5932, "step": 6757 }, { "epoch": 0.060780760254167544, "grad_norm": 2.815119743347168, "learning_rate": 1.3043189548891516e-05, "loss": 2.1742, "step": 6758 }, { "epoch": 0.060789754151808, "grad_norm": 5.741359710693359, "learning_rate": 1.3035929721508522e-05, "loss": 2.6598, "step": 6759 }, { "epoch": 0.06079874804944845, "grad_norm": 11.07270622253418, "learning_rate": 1.3028671202498261e-05, "loss": 1.9235, "step": 6760 }, { "epoch": 0.0608077419470889, "grad_norm": 7.577451229095459, "learning_rate": 1.3021413992654529e-05, "loss": 2.1567, "step": 6761 }, { "epoch": 0.06081673584472935, "grad_norm": 2.839812994003296, "learning_rate": 1.3014158092770948e-05, "loss": 2.4345, "step": 6762 }, { "epoch": 0.0608257297423698, "grad_norm": 2.7583999633789062, "learning_rate": 1.3006903503641016e-05, "loss": 2.0598, "step": 6763 }, { "epoch": 0.06083472364001025, "grad_norm": 3.0735080242156982, "learning_rate": 1.2999650226058083e-05, "loss": 2.263, "step": 6764 }, { "epoch": 0.06084371753765071, "grad_norm": 2.5008440017700195, "learning_rate": 1.2992398260815369e-05, "loss": 2.38, "step": 6765 }, { "epoch": 0.060852711435291154, "grad_norm": 4.227540493011475, "learning_rate": 1.2985147608705922e-05, "loss": 2.6998, "step": 6766 }, { "epoch": 0.06086170533293161, "grad_norm": 4.692346572875977, "learning_rate": 1.2977898270522677e-05, "loss": 0.6931, "step": 6767 }, { "epoch": 0.060870699230572055, "grad_norm": 3.982097625732422, "learning_rate": 1.2970650247058388e-05, "loss": 2.0927, "step": 6768 }, { "epoch": 0.06087969312821251, "grad_norm": 3.213762044906616, "learning_rate": 1.2963403539105707e-05, "loss": 2.2319, "step": 6769 }, { "epoch": 0.060888687025852956, "grad_norm": 5.7913312911987305, "learning_rate": 1.2956158147457115e-05, "loss": 2.2485, "step": 6770 }, { "epoch": 0.06089768092349341, "grad_norm": 3.0217173099517822, "learning_rate": 1.2948914072904972e-05, "loss": 2.2498, "step": 6771 }, { "epoch": 0.06090667482113386, "grad_norm": 3.809335947036743, "learning_rate": 1.2941671316241454e-05, "loss": 2.6986, "step": 6772 }, { "epoch": 0.06091566871877431, "grad_norm": 3.755049467086792, "learning_rate": 1.2934429878258648e-05, "loss": 2.1421, "step": 6773 }, { "epoch": 0.060924662616414764, "grad_norm": 4.149538516998291, "learning_rate": 1.2927189759748437e-05, "loss": 2.7859, "step": 6774 }, { "epoch": 0.06093365651405521, "grad_norm": 3.893468141555786, "learning_rate": 1.2919950961502603e-05, "loss": 2.7112, "step": 6775 }, { "epoch": 0.060942650411695665, "grad_norm": 2.3110148906707764, "learning_rate": 1.2912713484312766e-05, "loss": 2.6491, "step": 6776 }, { "epoch": 0.06095164430933612, "grad_norm": 4.063604831695557, "learning_rate": 1.2905477328970423e-05, "loss": 1.7571, "step": 6777 }, { "epoch": 0.060960638206976565, "grad_norm": 2.94872784614563, "learning_rate": 1.2898242496266884e-05, "loss": 2.0582, "step": 6778 }, { "epoch": 0.06096963210461702, "grad_norm": 6.507809162139893, "learning_rate": 1.2891008986993364e-05, "loss": 1.9366, "step": 6779 }, { "epoch": 0.060978626002257466, "grad_norm": 6.612085342407227, "learning_rate": 1.2883776801940884e-05, "loss": 2.284, "step": 6780 }, { "epoch": 0.06098761989989792, "grad_norm": 3.103663682937622, "learning_rate": 1.2876545941900354e-05, "loss": 2.194, "step": 6781 }, { "epoch": 0.06099661379753837, "grad_norm": 2.5381245613098145, "learning_rate": 1.2869316407662535e-05, "loss": 2.0347, "step": 6782 }, { "epoch": 0.06100560769517882, "grad_norm": 4.335866451263428, "learning_rate": 1.2862088200018046e-05, "loss": 1.5438, "step": 6783 }, { "epoch": 0.061014601592819274, "grad_norm": 4.401049613952637, "learning_rate": 1.285486131975733e-05, "loss": 2.1533, "step": 6784 }, { "epoch": 0.06102359549045972, "grad_norm": 4.454504013061523, "learning_rate": 1.2847635767670722e-05, "loss": 2.1768, "step": 6785 }, { "epoch": 0.061032589388100175, "grad_norm": 3.7012200355529785, "learning_rate": 1.2840411544548403e-05, "loss": 2.5003, "step": 6786 }, { "epoch": 0.06104158328574062, "grad_norm": 3.110416889190674, "learning_rate": 1.2833188651180388e-05, "loss": 2.4148, "step": 6787 }, { "epoch": 0.061050577183381076, "grad_norm": 6.215957164764404, "learning_rate": 1.2825967088356572e-05, "loss": 2.6337, "step": 6788 }, { "epoch": 0.06105957108102153, "grad_norm": 4.865861415863037, "learning_rate": 1.2818746856866687e-05, "loss": 2.6996, "step": 6789 }, { "epoch": 0.06106856497866198, "grad_norm": 2.500495672225952, "learning_rate": 1.2811527957500345e-05, "loss": 2.4164, "step": 6790 }, { "epoch": 0.06107755887630243, "grad_norm": 2.3257172107696533, "learning_rate": 1.280431039104697e-05, "loss": 2.3919, "step": 6791 }, { "epoch": 0.06108655277394288, "grad_norm": 3.669248104095459, "learning_rate": 1.279709415829588e-05, "loss": 2.3029, "step": 6792 }, { "epoch": 0.06109554667158333, "grad_norm": 2.8702919483184814, "learning_rate": 1.2789879260036222e-05, "loss": 2.2508, "step": 6793 }, { "epoch": 0.061104540569223785, "grad_norm": 2.2062275409698486, "learning_rate": 1.2782665697057017e-05, "loss": 2.5698, "step": 6794 }, { "epoch": 0.06111353446686423, "grad_norm": 2.670799493789673, "learning_rate": 1.2775453470147106e-05, "loss": 2.3436, "step": 6795 }, { "epoch": 0.061122528364504686, "grad_norm": 2.5685954093933105, "learning_rate": 1.2768242580095241e-05, "loss": 2.554, "step": 6796 }, { "epoch": 0.06113152226214513, "grad_norm": 5.520801067352295, "learning_rate": 1.276103302768997e-05, "loss": 2.3925, "step": 6797 }, { "epoch": 0.061140516159785586, "grad_norm": 9.609354972839355, "learning_rate": 1.2753824813719733e-05, "loss": 2.4345, "step": 6798 }, { "epoch": 0.06114951005742603, "grad_norm": 7.76343297958374, "learning_rate": 1.2746617938972794e-05, "loss": 2.6462, "step": 6799 }, { "epoch": 0.06115850395506649, "grad_norm": 11.960844039916992, "learning_rate": 1.2739412404237306e-05, "loss": 2.5404, "step": 6800 }, { "epoch": 0.06116749785270694, "grad_norm": 4.497251510620117, "learning_rate": 1.2732208210301222e-05, "loss": 2.1535, "step": 6801 }, { "epoch": 0.06117649175034739, "grad_norm": 2.9479618072509766, "learning_rate": 1.2725005357952425e-05, "loss": 2.4509, "step": 6802 }, { "epoch": 0.06118548564798784, "grad_norm": 2.9346823692321777, "learning_rate": 1.2717803847978577e-05, "loss": 2.0735, "step": 6803 }, { "epoch": 0.06119447954562829, "grad_norm": 8.442707061767578, "learning_rate": 1.2710603681167243e-05, "loss": 2.6425, "step": 6804 }, { "epoch": 0.06120347344326874, "grad_norm": 4.842555522918701, "learning_rate": 1.2703404858305806e-05, "loss": 2.3551, "step": 6805 }, { "epoch": 0.061212467340909196, "grad_norm": 3.95101261138916, "learning_rate": 1.2696207380181524e-05, "loss": 1.9784, "step": 6806 }, { "epoch": 0.06122146123854964, "grad_norm": 5.983380317687988, "learning_rate": 1.2689011247581503e-05, "loss": 2.0191, "step": 6807 }, { "epoch": 0.0612304551361901, "grad_norm": 1.764750599861145, "learning_rate": 1.2681816461292715e-05, "loss": 2.5417, "step": 6808 }, { "epoch": 0.061239449033830544, "grad_norm": 6.161182880401611, "learning_rate": 1.2674623022101948e-05, "loss": 2.4624, "step": 6809 }, { "epoch": 0.061248442931471, "grad_norm": 8.198171615600586, "learning_rate": 1.2667430930795877e-05, "loss": 1.7642, "step": 6810 }, { "epoch": 0.061257436829111445, "grad_norm": 3.263648748397827, "learning_rate": 1.2660240188161026e-05, "loss": 1.5675, "step": 6811 }, { "epoch": 0.0612664307267519, "grad_norm": 3.7238874435424805, "learning_rate": 1.2653050794983746e-05, "loss": 2.0744, "step": 6812 }, { "epoch": 0.06127542462439235, "grad_norm": 3.166088342666626, "learning_rate": 1.264586275205028e-05, "loss": 2.5093, "step": 6813 }, { "epoch": 0.0612844185220328, "grad_norm": 4.307883262634277, "learning_rate": 1.2638676060146665e-05, "loss": 2.2314, "step": 6814 }, { "epoch": 0.06129341241967325, "grad_norm": 4.07002067565918, "learning_rate": 1.2631490720058875e-05, "loss": 2.6152, "step": 6815 }, { "epoch": 0.0613024063173137, "grad_norm": 2.8271334171295166, "learning_rate": 1.2624306732572652e-05, "loss": 2.2709, "step": 6816 }, { "epoch": 0.061311400214954154, "grad_norm": 14.153989791870117, "learning_rate": 1.2617124098473649e-05, "loss": 2.8083, "step": 6817 }, { "epoch": 0.06132039411259461, "grad_norm": 3.469456911087036, "learning_rate": 1.2609942818547327e-05, "loss": 2.5229, "step": 6818 }, { "epoch": 0.061329388010235054, "grad_norm": 4.86793327331543, "learning_rate": 1.260276289357904e-05, "loss": 2.4461, "step": 6819 }, { "epoch": 0.06133838190787551, "grad_norm": 3.6923584938049316, "learning_rate": 1.2595584324353943e-05, "loss": 1.6315, "step": 6820 }, { "epoch": 0.061347375805515955, "grad_norm": 5.476335525512695, "learning_rate": 1.2588407111657119e-05, "loss": 2.75, "step": 6821 }, { "epoch": 0.06135636970315641, "grad_norm": 5.109809398651123, "learning_rate": 1.2581231256273419e-05, "loss": 2.363, "step": 6822 }, { "epoch": 0.061365363600796856, "grad_norm": 22.065095901489258, "learning_rate": 1.2574056758987606e-05, "loss": 2.8226, "step": 6823 }, { "epoch": 0.06137435749843731, "grad_norm": 5.520458698272705, "learning_rate": 1.2566883620584257e-05, "loss": 3.0224, "step": 6824 }, { "epoch": 0.061383351396077764, "grad_norm": 2.903902292251587, "learning_rate": 1.255971184184783e-05, "loss": 1.8579, "step": 6825 }, { "epoch": 0.06139234529371821, "grad_norm": 2.625164747238159, "learning_rate": 1.2552541423562592e-05, "loss": 2.9069, "step": 6826 }, { "epoch": 0.061401339191358664, "grad_norm": 5.4922003746032715, "learning_rate": 1.254537236651273e-05, "loss": 2.5123, "step": 6827 }, { "epoch": 0.06141033308899911, "grad_norm": 3.2225289344787598, "learning_rate": 1.2538204671482212e-05, "loss": 2.1117, "step": 6828 }, { "epoch": 0.061419326986639565, "grad_norm": 3.3194820880889893, "learning_rate": 1.2531038339254905e-05, "loss": 2.2042, "step": 6829 }, { "epoch": 0.06142832088428002, "grad_norm": 3.8960676193237305, "learning_rate": 1.2523873370614489e-05, "loss": 2.4059, "step": 6830 }, { "epoch": 0.061437314781920466, "grad_norm": 5.699217319488525, "learning_rate": 1.2516709766344522e-05, "loss": 2.1832, "step": 6831 }, { "epoch": 0.06144630867956092, "grad_norm": 3.4592666625976562, "learning_rate": 1.2509547527228416e-05, "loss": 2.497, "step": 6832 }, { "epoch": 0.061455302577201366, "grad_norm": 6.503941535949707, "learning_rate": 1.2502386654049405e-05, "loss": 2.1737, "step": 6833 }, { "epoch": 0.06146429647484182, "grad_norm": 4.674671649932861, "learning_rate": 1.24952271475906e-05, "loss": 2.5582, "step": 6834 }, { "epoch": 0.061473290372482274, "grad_norm": 5.856832981109619, "learning_rate": 1.2488069008634954e-05, "loss": 1.75, "step": 6835 }, { "epoch": 0.06148228427012272, "grad_norm": 5.6703782081604, "learning_rate": 1.2480912237965276e-05, "loss": 1.889, "step": 6836 }, { "epoch": 0.061491278167763175, "grad_norm": 3.664297580718994, "learning_rate": 1.2473756836364207e-05, "loss": 1.8434, "step": 6837 }, { "epoch": 0.06150027206540362, "grad_norm": 2.8348278999328613, "learning_rate": 1.2466602804614266e-05, "loss": 2.3695, "step": 6838 }, { "epoch": 0.061509265963044076, "grad_norm": 3.597932815551758, "learning_rate": 1.2459450143497778e-05, "loss": 2.2374, "step": 6839 }, { "epoch": 0.06151825986068452, "grad_norm": 3.1138193607330322, "learning_rate": 1.245229885379699e-05, "loss": 1.8824, "step": 6840 }, { "epoch": 0.061527253758324976, "grad_norm": 16.46555519104004, "learning_rate": 1.2445148936293923e-05, "loss": 2.7075, "step": 6841 }, { "epoch": 0.06153624765596543, "grad_norm": 3.9743237495422363, "learning_rate": 1.2438000391770502e-05, "loss": 2.2891, "step": 6842 }, { "epoch": 0.06154524155360588, "grad_norm": 5.17686653137207, "learning_rate": 1.2430853221008461e-05, "loss": 2.4619, "step": 6843 }, { "epoch": 0.06155423545124633, "grad_norm": 3.9454119205474854, "learning_rate": 1.2423707424789424e-05, "loss": 3.3788, "step": 6844 }, { "epoch": 0.06156322934888678, "grad_norm": 9.06855297088623, "learning_rate": 1.2416563003894826e-05, "loss": 2.4078, "step": 6845 }, { "epoch": 0.06157222324652723, "grad_norm": 3.3579628467559814, "learning_rate": 1.2409419959105981e-05, "loss": 2.0105, "step": 6846 }, { "epoch": 0.061581217144167685, "grad_norm": 4.205850601196289, "learning_rate": 1.2402278291204037e-05, "loss": 1.8424, "step": 6847 }, { "epoch": 0.06159021104180813, "grad_norm": 4.75227689743042, "learning_rate": 1.2395138000970008e-05, "loss": 2.3553, "step": 6848 }, { "epoch": 0.061599204939448586, "grad_norm": 5.097429275512695, "learning_rate": 1.2387999089184729e-05, "loss": 1.9237, "step": 6849 }, { "epoch": 0.06160819883708903, "grad_norm": 4.96871280670166, "learning_rate": 1.2380861556628915e-05, "loss": 2.2038, "step": 6850 }, { "epoch": 0.06161719273472949, "grad_norm": 9.184587478637695, "learning_rate": 1.23737254040831e-05, "loss": 2.1862, "step": 6851 }, { "epoch": 0.061626186632369934, "grad_norm": 4.321457386016846, "learning_rate": 1.236659063232769e-05, "loss": 2.0843, "step": 6852 }, { "epoch": 0.06163518053001039, "grad_norm": 7.642993927001953, "learning_rate": 1.2359457242142938e-05, "loss": 1.6481, "step": 6853 }, { "epoch": 0.06164417442765084, "grad_norm": 3.817300319671631, "learning_rate": 1.2352325234308944e-05, "loss": 2.113, "step": 6854 }, { "epoch": 0.06165316832529129, "grad_norm": 3.6986536979675293, "learning_rate": 1.2345194609605636e-05, "loss": 2.1058, "step": 6855 }, { "epoch": 0.06166216222293174, "grad_norm": 4.488257884979248, "learning_rate": 1.2338065368812823e-05, "loss": 2.1887, "step": 6856 }, { "epoch": 0.06167115612057219, "grad_norm": 8.636685371398926, "learning_rate": 1.2330937512710153e-05, "loss": 1.7468, "step": 6857 }, { "epoch": 0.06168015001821264, "grad_norm": 2.7362310886383057, "learning_rate": 1.2323811042077096e-05, "loss": 1.7618, "step": 6858 }, { "epoch": 0.0616891439158531, "grad_norm": 5.068390846252441, "learning_rate": 1.2316685957693008e-05, "loss": 2.5239, "step": 6859 }, { "epoch": 0.061698137813493543, "grad_norm": 3.472768545150757, "learning_rate": 1.2309562260337073e-05, "loss": 2.3901, "step": 6860 }, { "epoch": 0.061707131711134, "grad_norm": 3.71386456489563, "learning_rate": 1.2302439950788342e-05, "loss": 2.1204, "step": 6861 }, { "epoch": 0.061716125608774444, "grad_norm": 3.880943775177002, "learning_rate": 1.2295319029825674e-05, "loss": 2.3437, "step": 6862 }, { "epoch": 0.0617251195064149, "grad_norm": 8.597739219665527, "learning_rate": 1.2288199498227828e-05, "loss": 2.0037, "step": 6863 }, { "epoch": 0.06173411340405535, "grad_norm": 6.437736511230469, "learning_rate": 1.2281081356773364e-05, "loss": 2.017, "step": 6864 }, { "epoch": 0.0617431073016958, "grad_norm": 3.109346389770508, "learning_rate": 1.2273964606240718e-05, "loss": 2.2985, "step": 6865 }, { "epoch": 0.06175210119933625, "grad_norm": 3.334829330444336, "learning_rate": 1.2266849247408171e-05, "loss": 2.0991, "step": 6866 }, { "epoch": 0.0617610950969767, "grad_norm": 7.091006755828857, "learning_rate": 1.2259735281053854e-05, "loss": 2.431, "step": 6867 }, { "epoch": 0.06177008899461715, "grad_norm": 3.644448757171631, "learning_rate": 1.2252622707955725e-05, "loss": 3.0994, "step": 6868 }, { "epoch": 0.0617790828922576, "grad_norm": 3.2847611904144287, "learning_rate": 1.2245511528891618e-05, "loss": 2.1757, "step": 6869 }, { "epoch": 0.061788076789898054, "grad_norm": 4.575556755065918, "learning_rate": 1.2238401744639185e-05, "loss": 2.5446, "step": 6870 }, { "epoch": 0.06179707068753851, "grad_norm": 3.956575393676758, "learning_rate": 1.223129335597595e-05, "loss": 1.636, "step": 6871 }, { "epoch": 0.061806064585178955, "grad_norm": 2.923652410507202, "learning_rate": 1.222418636367928e-05, "loss": 2.3477, "step": 6872 }, { "epoch": 0.06181505848281941, "grad_norm": 7.396069049835205, "learning_rate": 1.2217080768526387e-05, "loss": 1.9746, "step": 6873 }, { "epoch": 0.061824052380459855, "grad_norm": 4.469505310058594, "learning_rate": 1.2209976571294313e-05, "loss": 2.0419, "step": 6874 }, { "epoch": 0.06183304627810031, "grad_norm": 6.315096378326416, "learning_rate": 1.2202873772759981e-05, "loss": 2.3582, "step": 6875 }, { "epoch": 0.06184204017574076, "grad_norm": 2.8557918071746826, "learning_rate": 1.2195772373700126e-05, "loss": 2.2949, "step": 6876 }, { "epoch": 0.06185103407338121, "grad_norm": 3.839931011199951, "learning_rate": 1.2188672374891352e-05, "loss": 1.7905, "step": 6877 }, { "epoch": 0.061860027971021664, "grad_norm": 2.490962266921997, "learning_rate": 1.2181573777110106e-05, "loss": 1.874, "step": 6878 }, { "epoch": 0.06186902186866211, "grad_norm": 4.53511381149292, "learning_rate": 1.2174476581132687e-05, "loss": 1.6228, "step": 6879 }, { "epoch": 0.061878015766302565, "grad_norm": 4.253078937530518, "learning_rate": 1.216738078773522e-05, "loss": 2.2659, "step": 6880 }, { "epoch": 0.06188700966394301, "grad_norm": 3.3815178871154785, "learning_rate": 1.2160286397693693e-05, "loss": 2.5931, "step": 6881 }, { "epoch": 0.061896003561583465, "grad_norm": 2.92668080329895, "learning_rate": 1.2153193411783953e-05, "loss": 1.1235, "step": 6882 }, { "epoch": 0.06190499745922392, "grad_norm": 4.817694187164307, "learning_rate": 1.2146101830781656e-05, "loss": 2.6302, "step": 6883 }, { "epoch": 0.061913991356864366, "grad_norm": 5.480474472045898, "learning_rate": 1.2139011655462337e-05, "loss": 1.597, "step": 6884 }, { "epoch": 0.06192298525450482, "grad_norm": 2.2779715061187744, "learning_rate": 1.2131922886601368e-05, "loss": 1.8721, "step": 6885 }, { "epoch": 0.06193197915214527, "grad_norm": 3.4158096313476562, "learning_rate": 1.2124835524973977e-05, "loss": 1.8047, "step": 6886 }, { "epoch": 0.06194097304978572, "grad_norm": 6.269527435302734, "learning_rate": 1.2117749571355202e-05, "loss": 2.544, "step": 6887 }, { "epoch": 0.061949966947426174, "grad_norm": 5.129368305206299, "learning_rate": 1.2110665026519976e-05, "loss": 2.3332, "step": 6888 }, { "epoch": 0.06195896084506662, "grad_norm": 5.5978102684021, "learning_rate": 1.2103581891243034e-05, "loss": 1.9434, "step": 6889 }, { "epoch": 0.061967954742707075, "grad_norm": 6.183835029602051, "learning_rate": 1.209650016629899e-05, "loss": 2.3024, "step": 6890 }, { "epoch": 0.06197694864034752, "grad_norm": 3.118533134460449, "learning_rate": 1.2089419852462286e-05, "loss": 2.1253, "step": 6891 }, { "epoch": 0.061985942537987976, "grad_norm": 4.880030155181885, "learning_rate": 1.2082340950507226e-05, "loss": 1.9983, "step": 6892 }, { "epoch": 0.06199493643562842, "grad_norm": 3.984203338623047, "learning_rate": 1.2075263461207929e-05, "loss": 2.4073, "step": 6893 }, { "epoch": 0.06200393033326888, "grad_norm": 4.105113983154297, "learning_rate": 1.2068187385338395e-05, "loss": 2.153, "step": 6894 }, { "epoch": 0.06201292423090933, "grad_norm": 3.300806760787964, "learning_rate": 1.2061112723672438e-05, "loss": 2.4266, "step": 6895 }, { "epoch": 0.06202191812854978, "grad_norm": 7.80509614944458, "learning_rate": 1.205403947698374e-05, "loss": 2.2603, "step": 6896 }, { "epoch": 0.06203091202619023, "grad_norm": 12.751304626464844, "learning_rate": 1.204696764604582e-05, "loss": 3.5143, "step": 6897 }, { "epoch": 0.06203990592383068, "grad_norm": 3.595032215118408, "learning_rate": 1.2039897231632055e-05, "loss": 2.3151, "step": 6898 }, { "epoch": 0.06204889982147113, "grad_norm": 6.637812614440918, "learning_rate": 1.2032828234515633e-05, "loss": 2.3041, "step": 6899 }, { "epoch": 0.062057893719111586, "grad_norm": 2.7603909969329834, "learning_rate": 1.202576065546963e-05, "loss": 2.2157, "step": 6900 }, { "epoch": 0.06206688761675203, "grad_norm": 2.3473222255706787, "learning_rate": 1.2018694495266924e-05, "loss": 1.9355, "step": 6901 }, { "epoch": 0.062075881514392486, "grad_norm": 10.859892845153809, "learning_rate": 1.201162975468027e-05, "loss": 2.2252, "step": 6902 }, { "epoch": 0.06208487541203293, "grad_norm": 5.783349514007568, "learning_rate": 1.2004566434482261e-05, "loss": 1.9701, "step": 6903 }, { "epoch": 0.06209386930967339, "grad_norm": 5.695621013641357, "learning_rate": 1.1997504535445336e-05, "loss": 2.2753, "step": 6904 }, { "epoch": 0.06210286320731384, "grad_norm": 8.731057167053223, "learning_rate": 1.1990444058341757e-05, "loss": 2.6657, "step": 6905 }, { "epoch": 0.06211185710495429, "grad_norm": 3.7495779991149902, "learning_rate": 1.1983385003943658e-05, "loss": 1.9619, "step": 6906 }, { "epoch": 0.06212085100259474, "grad_norm": 2.7802953720092773, "learning_rate": 1.1976327373023013e-05, "loss": 2.3751, "step": 6907 }, { "epoch": 0.06212984490023519, "grad_norm": 13.865463256835938, "learning_rate": 1.1969271166351614e-05, "loss": 2.4749, "step": 6908 }, { "epoch": 0.06213883879787564, "grad_norm": 3.2437427043914795, "learning_rate": 1.1962216384701134e-05, "loss": 2.3493, "step": 6909 }, { "epoch": 0.06214783269551609, "grad_norm": 3.9688196182250977, "learning_rate": 1.1955163028843063e-05, "loss": 2.2253, "step": 6910 }, { "epoch": 0.06215682659315654, "grad_norm": 3.7492215633392334, "learning_rate": 1.1948111099548762e-05, "loss": 2.4474, "step": 6911 }, { "epoch": 0.062165820490797, "grad_norm": 5.531002044677734, "learning_rate": 1.1941060597589399e-05, "loss": 2.3041, "step": 6912 }, { "epoch": 0.062174814388437444, "grad_norm": 2.348362684249878, "learning_rate": 1.1934011523736024e-05, "loss": 2.1832, "step": 6913 }, { "epoch": 0.0621838082860779, "grad_norm": 6.084727764129639, "learning_rate": 1.1926963878759495e-05, "loss": 2.9806, "step": 6914 }, { "epoch": 0.062192802183718345, "grad_norm": 3.7542355060577393, "learning_rate": 1.1919917663430552e-05, "loss": 1.9948, "step": 6915 }, { "epoch": 0.0622017960813588, "grad_norm": 7.046555042266846, "learning_rate": 1.1912872878519726e-05, "loss": 2.3724, "step": 6916 }, { "epoch": 0.06221078997899925, "grad_norm": 4.701962947845459, "learning_rate": 1.1905829524797465e-05, "loss": 2.5888, "step": 6917 }, { "epoch": 0.0622197838766397, "grad_norm": 5.291653156280518, "learning_rate": 1.1898787603033995e-05, "loss": 2.171, "step": 6918 }, { "epoch": 0.06222877777428015, "grad_norm": 5.497361183166504, "learning_rate": 1.1891747113999422e-05, "loss": 1.6739, "step": 6919 }, { "epoch": 0.0622377716719206, "grad_norm": 4.818488597869873, "learning_rate": 1.1884708058463668e-05, "loss": 2.7558, "step": 6920 }, { "epoch": 0.062246765569561054, "grad_norm": 5.702832221984863, "learning_rate": 1.1877670437196528e-05, "loss": 2.3742, "step": 6921 }, { "epoch": 0.0622557594672015, "grad_norm": 3.629714012145996, "learning_rate": 1.1870634250967605e-05, "loss": 1.8217, "step": 6922 }, { "epoch": 0.062264753364841954, "grad_norm": 3.240501642227173, "learning_rate": 1.1863599500546399e-05, "loss": 1.8337, "step": 6923 }, { "epoch": 0.06227374726248241, "grad_norm": 3.276278495788574, "learning_rate": 1.1856566186702189e-05, "loss": 1.8648, "step": 6924 }, { "epoch": 0.062282741160122855, "grad_norm": 11.074618339538574, "learning_rate": 1.1849534310204152e-05, "loss": 2.3533, "step": 6925 }, { "epoch": 0.06229173505776331, "grad_norm": 3.624145984649658, "learning_rate": 1.1842503871821262e-05, "loss": 2.0111, "step": 6926 }, { "epoch": 0.062300728955403756, "grad_norm": 2.3751773834228516, "learning_rate": 1.1835474872322367e-05, "loss": 1.6815, "step": 6927 }, { "epoch": 0.06230972285304421, "grad_norm": 4.548739433288574, "learning_rate": 1.1828447312476145e-05, "loss": 2.0291, "step": 6928 }, { "epoch": 0.06231871675068466, "grad_norm": 5.298611640930176, "learning_rate": 1.1821421193051132e-05, "loss": 2.0078, "step": 6929 }, { "epoch": 0.06232771064832511, "grad_norm": 4.018152713775635, "learning_rate": 1.1814396514815676e-05, "loss": 2.4411, "step": 6930 }, { "epoch": 0.062336704545965564, "grad_norm": 4.693087100982666, "learning_rate": 1.1807373278537992e-05, "loss": 2.736, "step": 6931 }, { "epoch": 0.06234569844360601, "grad_norm": 5.366990566253662, "learning_rate": 1.1800351484986141e-05, "loss": 2.5854, "step": 6932 }, { "epoch": 0.062354692341246465, "grad_norm": 7.2326250076293945, "learning_rate": 1.1793331134927996e-05, "loss": 2.6505, "step": 6933 }, { "epoch": 0.06236368623888691, "grad_norm": 4.9624457359313965, "learning_rate": 1.178631222913131e-05, "loss": 2.6215, "step": 6934 }, { "epoch": 0.062372680136527366, "grad_norm": 2.351590871810913, "learning_rate": 1.1779294768363639e-05, "loss": 1.937, "step": 6935 }, { "epoch": 0.06238167403416782, "grad_norm": 2.071228265762329, "learning_rate": 1.1772278753392429e-05, "loss": 2.4311, "step": 6936 }, { "epoch": 0.062390667931808266, "grad_norm": 4.904214382171631, "learning_rate": 1.1765264184984917e-05, "loss": 2.2354, "step": 6937 }, { "epoch": 0.06239966182944872, "grad_norm": 4.679743766784668, "learning_rate": 1.1758251063908227e-05, "loss": 2.0966, "step": 6938 }, { "epoch": 0.06240865572708917, "grad_norm": 4.317685604095459, "learning_rate": 1.1751239390929283e-05, "loss": 2.398, "step": 6939 }, { "epoch": 0.06241764962472962, "grad_norm": 4.280824661254883, "learning_rate": 1.1744229166814888e-05, "loss": 1.8905, "step": 6940 }, { "epoch": 0.062426643522370075, "grad_norm": 5.97560453414917, "learning_rate": 1.1737220392331644e-05, "loss": 2.581, "step": 6941 }, { "epoch": 0.06243563742001052, "grad_norm": 3.3023674488067627, "learning_rate": 1.1730213068246057e-05, "loss": 2.6459, "step": 6942 }, { "epoch": 0.062444631317650975, "grad_norm": 2.4991300106048584, "learning_rate": 1.1723207195324406e-05, "loss": 2.2589, "step": 6943 }, { "epoch": 0.06245362521529142, "grad_norm": 2.7361059188842773, "learning_rate": 1.1716202774332865e-05, "loss": 1.9483, "step": 6944 }, { "epoch": 0.062462619112931876, "grad_norm": 2.1887457370758057, "learning_rate": 1.170919980603741e-05, "loss": 2.1622, "step": 6945 }, { "epoch": 0.06247161301057233, "grad_norm": 4.661910533905029, "learning_rate": 1.1702198291203889e-05, "loss": 2.3253, "step": 6946 }, { "epoch": 0.06248060690821278, "grad_norm": 4.382431983947754, "learning_rate": 1.1695198230597951e-05, "loss": 1.9828, "step": 6947 }, { "epoch": 0.06248960080585323, "grad_norm": 3.448434591293335, "learning_rate": 1.1688199624985149e-05, "loss": 2.1098, "step": 6948 }, { "epoch": 0.06249859470349368, "grad_norm": 4.460287570953369, "learning_rate": 1.1681202475130814e-05, "loss": 1.7621, "step": 6949 }, { "epoch": 0.06250758860113413, "grad_norm": 7.318570137023926, "learning_rate": 1.1674206781800162e-05, "loss": 2.4453, "step": 6950 }, { "epoch": 0.06251658249877458, "grad_norm": 5.3567609786987305, "learning_rate": 1.166721254575821e-05, "loss": 1.5561, "step": 6951 }, { "epoch": 0.06252557639641504, "grad_norm": 9.230003356933594, "learning_rate": 1.1660219767769851e-05, "loss": 2.3151, "step": 6952 }, { "epoch": 0.06253457029405549, "grad_norm": 5.1951799392700195, "learning_rate": 1.165322844859981e-05, "loss": 2.1394, "step": 6953 }, { "epoch": 0.06254356419169593, "grad_norm": 2.536245346069336, "learning_rate": 1.1646238589012628e-05, "loss": 1.8827, "step": 6954 }, { "epoch": 0.06255255808933638, "grad_norm": 4.52073860168457, "learning_rate": 1.1639250189772718e-05, "loss": 2.4527, "step": 6955 }, { "epoch": 0.06256155198697684, "grad_norm": 5.692814826965332, "learning_rate": 1.1632263251644321e-05, "loss": 2.1627, "step": 6956 }, { "epoch": 0.06257054588461729, "grad_norm": 3.470586061477661, "learning_rate": 1.1625277775391526e-05, "loss": 1.9024, "step": 6957 }, { "epoch": 0.06257953978225773, "grad_norm": 3.3481924533843994, "learning_rate": 1.1618293761778235e-05, "loss": 2.3073, "step": 6958 }, { "epoch": 0.0625885336798982, "grad_norm": 2.1929805278778076, "learning_rate": 1.1611311211568229e-05, "loss": 1.9232, "step": 6959 }, { "epoch": 0.06259752757753864, "grad_norm": 2.910792112350464, "learning_rate": 1.1604330125525079e-05, "loss": 2.1555, "step": 6960 }, { "epoch": 0.06260652147517909, "grad_norm": 5.229135513305664, "learning_rate": 1.1597350504412265e-05, "loss": 2.241, "step": 6961 }, { "epoch": 0.06261551537281954, "grad_norm": 5.640956401824951, "learning_rate": 1.1590372348993043e-05, "loss": 2.3034, "step": 6962 }, { "epoch": 0.06262450927046, "grad_norm": 3.639443874359131, "learning_rate": 1.1583395660030547e-05, "loss": 1.9271, "step": 6963 }, { "epoch": 0.06263350316810044, "grad_norm": 8.765067100524902, "learning_rate": 1.157642043828772e-05, "loss": 1.8466, "step": 6964 }, { "epoch": 0.06264249706574089, "grad_norm": 6.315497398376465, "learning_rate": 1.1569446684527383e-05, "loss": 2.1702, "step": 6965 }, { "epoch": 0.06265149096338135, "grad_norm": 6.732358932495117, "learning_rate": 1.1562474399512143e-05, "loss": 2.6444, "step": 6966 }, { "epoch": 0.0626604848610218, "grad_norm": 4.429722309112549, "learning_rate": 1.155550358400452e-05, "loss": 1.7918, "step": 6967 }, { "epoch": 0.06266947875866224, "grad_norm": 3.241931915283203, "learning_rate": 1.1548534238766803e-05, "loss": 1.836, "step": 6968 }, { "epoch": 0.06267847265630269, "grad_norm": 2.597949266433716, "learning_rate": 1.1541566364561166e-05, "loss": 2.2343, "step": 6969 }, { "epoch": 0.06268746655394315, "grad_norm": 4.202613830566406, "learning_rate": 1.1534599962149587e-05, "loss": 2.0536, "step": 6970 }, { "epoch": 0.0626964604515836, "grad_norm": 3.7761478424072266, "learning_rate": 1.152763503229392e-05, "loss": 2.0913, "step": 6971 }, { "epoch": 0.06270545434922405, "grad_norm": 4.111439228057861, "learning_rate": 1.1520671575755823e-05, "loss": 1.9605, "step": 6972 }, { "epoch": 0.06271444824686451, "grad_norm": 6.562988758087158, "learning_rate": 1.1513709593296816e-05, "loss": 2.4171, "step": 6973 }, { "epoch": 0.06272344214450495, "grad_norm": 2.7872564792633057, "learning_rate": 1.150674908567825e-05, "loss": 2.4405, "step": 6974 }, { "epoch": 0.0627324360421454, "grad_norm": 4.612452030181885, "learning_rate": 1.1499790053661327e-05, "loss": 1.9327, "step": 6975 }, { "epoch": 0.06274142993978586, "grad_norm": 5.336350917816162, "learning_rate": 1.1492832498007059e-05, "loss": 2.1277, "step": 6976 }, { "epoch": 0.06275042383742631, "grad_norm": 7.282933235168457, "learning_rate": 1.1485876419476322e-05, "loss": 2.4978, "step": 6977 }, { "epoch": 0.06275941773506676, "grad_norm": 1.8459839820861816, "learning_rate": 1.1478921818829827e-05, "loss": 2.0889, "step": 6978 }, { "epoch": 0.0627684116327072, "grad_norm": 2.6661620140075684, "learning_rate": 1.1471968696828106e-05, "loss": 2.4224, "step": 6979 }, { "epoch": 0.06277740553034766, "grad_norm": 6.256751537322998, "learning_rate": 1.146501705423155e-05, "loss": 2.156, "step": 6980 }, { "epoch": 0.06278639942798811, "grad_norm": 2.570840835571289, "learning_rate": 1.1458066891800378e-05, "loss": 2.5216, "step": 6981 }, { "epoch": 0.06279539332562856, "grad_norm": 4.001133918762207, "learning_rate": 1.1451118210294663e-05, "loss": 2.3162, "step": 6982 }, { "epoch": 0.06280438722326902, "grad_norm": 4.562010765075684, "learning_rate": 1.1444171010474278e-05, "loss": 2.2127, "step": 6983 }, { "epoch": 0.06281338112090946, "grad_norm": 1.601487159729004, "learning_rate": 1.143722529309898e-05, "loss": 2.0744, "step": 6984 }, { "epoch": 0.06282237501854991, "grad_norm": 6.642026424407959, "learning_rate": 1.1430281058928324e-05, "loss": 1.7326, "step": 6985 }, { "epoch": 0.06283136891619036, "grad_norm": 3.5781891345977783, "learning_rate": 1.1423338308721729e-05, "loss": 2.6392, "step": 6986 }, { "epoch": 0.06284036281383082, "grad_norm": 4.1147027015686035, "learning_rate": 1.1416397043238445e-05, "loss": 2.2454, "step": 6987 }, { "epoch": 0.06284935671147127, "grad_norm": 4.727640628814697, "learning_rate": 1.1409457263237566e-05, "loss": 2.3592, "step": 6988 }, { "epoch": 0.06285835060911171, "grad_norm": 3.3967769145965576, "learning_rate": 1.1402518969478e-05, "loss": 2.3136, "step": 6989 }, { "epoch": 0.06286734450675217, "grad_norm": 3.328126907348633, "learning_rate": 1.1395582162718525e-05, "loss": 2.673, "step": 6990 }, { "epoch": 0.06287633840439262, "grad_norm": 2.2742233276367188, "learning_rate": 1.1388646843717718e-05, "loss": 2.3763, "step": 6991 }, { "epoch": 0.06288533230203307, "grad_norm": 2.824533462524414, "learning_rate": 1.138171301323403e-05, "loss": 2.2299, "step": 6992 }, { "epoch": 0.06289432619967353, "grad_norm": 6.058817386627197, "learning_rate": 1.1374780672025734e-05, "loss": 2.2169, "step": 6993 }, { "epoch": 0.06290332009731398, "grad_norm": 6.503373622894287, "learning_rate": 1.1367849820850946e-05, "loss": 2.5048, "step": 6994 }, { "epoch": 0.06291231399495442, "grad_norm": 3.5539040565490723, "learning_rate": 1.1360920460467598e-05, "loss": 1.8367, "step": 6995 }, { "epoch": 0.06292130789259487, "grad_norm": 7.923608779907227, "learning_rate": 1.1353992591633489e-05, "loss": 2.0944, "step": 6996 }, { "epoch": 0.06293030179023533, "grad_norm": 7.645511627197266, "learning_rate": 1.1347066215106225e-05, "loss": 2.147, "step": 6997 }, { "epoch": 0.06293929568787578, "grad_norm": 4.771149158477783, "learning_rate": 1.1340141331643276e-05, "loss": 1.8572, "step": 6998 }, { "epoch": 0.06294828958551622, "grad_norm": 12.016867637634277, "learning_rate": 1.133321794200193e-05, "loss": 2.7135, "step": 6999 }, { "epoch": 0.06295728348315668, "grad_norm": 3.0244545936584473, "learning_rate": 1.1326296046939333e-05, "loss": 2.5515, "step": 7000 }, { "epoch": 0.06296627738079713, "grad_norm": 2.886813163757324, "learning_rate": 1.1319375647212432e-05, "loss": 2.625, "step": 7001 }, { "epoch": 0.06297527127843758, "grad_norm": 3.5171890258789062, "learning_rate": 1.1312456743578042e-05, "loss": 2.3417, "step": 7002 }, { "epoch": 0.06298426517607802, "grad_norm": 6.112205505371094, "learning_rate": 1.1305539336792814e-05, "loss": 2.4221, "step": 7003 }, { "epoch": 0.06299325907371849, "grad_norm": 3.650510311126709, "learning_rate": 1.1298623427613206e-05, "loss": 2.1171, "step": 7004 }, { "epoch": 0.06300225297135893, "grad_norm": 1.7000983953475952, "learning_rate": 1.129170901679554e-05, "loss": 1.405, "step": 7005 }, { "epoch": 0.06301124686899938, "grad_norm": 3.260896921157837, "learning_rate": 1.1284796105095965e-05, "loss": 1.8576, "step": 7006 }, { "epoch": 0.06302024076663984, "grad_norm": 9.11924934387207, "learning_rate": 1.1277884693270477e-05, "loss": 2.1609, "step": 7007 }, { "epoch": 0.06302923466428029, "grad_norm": 3.030734062194824, "learning_rate": 1.1270974782074878e-05, "loss": 2.1426, "step": 7008 }, { "epoch": 0.06303822856192073, "grad_norm": 2.6220145225524902, "learning_rate": 1.1264066372264844e-05, "loss": 2.5018, "step": 7009 }, { "epoch": 0.0630472224595612, "grad_norm": 3.3502814769744873, "learning_rate": 1.1257159464595855e-05, "loss": 1.9704, "step": 7010 }, { "epoch": 0.06305621635720164, "grad_norm": 4.37639856338501, "learning_rate": 1.1250254059823239e-05, "loss": 2.3937, "step": 7011 }, { "epoch": 0.06306521025484209, "grad_norm": 3.037903308868408, "learning_rate": 1.1243350158702171e-05, "loss": 1.8984, "step": 7012 }, { "epoch": 0.06307420415248254, "grad_norm": 25.18252182006836, "learning_rate": 1.1236447761987656e-05, "loss": 2.145, "step": 7013 }, { "epoch": 0.063083198050123, "grad_norm": 3.4633753299713135, "learning_rate": 1.122954687043451e-05, "loss": 2.0984, "step": 7014 }, { "epoch": 0.06309219194776344, "grad_norm": 3.2990286350250244, "learning_rate": 1.1222647484797422e-05, "loss": 2.4209, "step": 7015 }, { "epoch": 0.06310118584540389, "grad_norm": 1.804598093032837, "learning_rate": 1.1215749605830884e-05, "loss": 1.8568, "step": 7016 }, { "epoch": 0.06311017974304435, "grad_norm": 2.1779251098632812, "learning_rate": 1.1208853234289246e-05, "loss": 1.8538, "step": 7017 }, { "epoch": 0.0631191736406848, "grad_norm": 2.238114595413208, "learning_rate": 1.1201958370926682e-05, "loss": 2.4501, "step": 7018 }, { "epoch": 0.06312816753832524, "grad_norm": 3.7447774410247803, "learning_rate": 1.1195065016497213e-05, "loss": 2.2729, "step": 7019 }, { "epoch": 0.06313716143596569, "grad_norm": 3.06439471244812, "learning_rate": 1.118817317175467e-05, "loss": 2.0943, "step": 7020 }, { "epoch": 0.06314615533360615, "grad_norm": 3.7556698322296143, "learning_rate": 1.1181282837452756e-05, "loss": 2.2257, "step": 7021 }, { "epoch": 0.0631551492312466, "grad_norm": 2.83065128326416, "learning_rate": 1.1174394014344963e-05, "loss": 1.8791, "step": 7022 }, { "epoch": 0.06316414312888705, "grad_norm": 4.19040060043335, "learning_rate": 1.1167506703184651e-05, "loss": 2.715, "step": 7023 }, { "epoch": 0.0631731370265275, "grad_norm": 9.433370590209961, "learning_rate": 1.1160620904725014e-05, "loss": 2.4694, "step": 7024 }, { "epoch": 0.06318213092416795, "grad_norm": 3.9629194736480713, "learning_rate": 1.1153736619719077e-05, "loss": 1.9771, "step": 7025 }, { "epoch": 0.0631911248218084, "grad_norm": 2.948051929473877, "learning_rate": 1.1146853848919672e-05, "loss": 1.5133, "step": 7026 }, { "epoch": 0.06320011871944885, "grad_norm": 5.501442909240723, "learning_rate": 1.1139972593079507e-05, "loss": 1.833, "step": 7027 }, { "epoch": 0.06320911261708931, "grad_norm": 5.929480075836182, "learning_rate": 1.1133092852951107e-05, "loss": 2.291, "step": 7028 }, { "epoch": 0.06321810651472975, "grad_norm": 8.735607147216797, "learning_rate": 1.1126214629286816e-05, "loss": 2.0061, "step": 7029 }, { "epoch": 0.0632271004123702, "grad_norm": 8.20920467376709, "learning_rate": 1.1119337922838832e-05, "loss": 2.2554, "step": 7030 }, { "epoch": 0.06323609431001066, "grad_norm": 9.11524772644043, "learning_rate": 1.1112462734359186e-05, "loss": 1.7024, "step": 7031 }, { "epoch": 0.06324508820765111, "grad_norm": 4.368276596069336, "learning_rate": 1.1105589064599744e-05, "loss": 2.5524, "step": 7032 }, { "epoch": 0.06325408210529156, "grad_norm": 2.8150978088378906, "learning_rate": 1.1098716914312184e-05, "loss": 2.5277, "step": 7033 }, { "epoch": 0.06326307600293202, "grad_norm": 3.612044334411621, "learning_rate": 1.109184628424805e-05, "loss": 1.8486, "step": 7034 }, { "epoch": 0.06327206990057246, "grad_norm": 4.606266021728516, "learning_rate": 1.1084977175158687e-05, "loss": 2.1537, "step": 7035 }, { "epoch": 0.06328106379821291, "grad_norm": 5.294416427612305, "learning_rate": 1.107810958779531e-05, "loss": 2.1117, "step": 7036 }, { "epoch": 0.06329005769585336, "grad_norm": 3.555025577545166, "learning_rate": 1.1071243522908916e-05, "loss": 2.3644, "step": 7037 }, { "epoch": 0.06329905159349382, "grad_norm": 3.0164546966552734, "learning_rate": 1.1064378981250409e-05, "loss": 2.0146, "step": 7038 }, { "epoch": 0.06330804549113427, "grad_norm": 4.2986650466918945, "learning_rate": 1.1057515963570452e-05, "loss": 1.6359, "step": 7039 }, { "epoch": 0.06331703938877471, "grad_norm": 3.493750810623169, "learning_rate": 1.1050654470619601e-05, "loss": 2.4332, "step": 7040 }, { "epoch": 0.06332603328641517, "grad_norm": 6.6165080070495605, "learning_rate": 1.1043794503148194e-05, "loss": 2.2333, "step": 7041 }, { "epoch": 0.06333502718405562, "grad_norm": 3.7883849143981934, "learning_rate": 1.1036936061906444e-05, "loss": 2.1731, "step": 7042 }, { "epoch": 0.06334402108169607, "grad_norm": 4.557745456695557, "learning_rate": 1.1030079147644357e-05, "loss": 2.0213, "step": 7043 }, { "epoch": 0.06335301497933651, "grad_norm": 2.9384922981262207, "learning_rate": 1.102322376111183e-05, "loss": 2.1614, "step": 7044 }, { "epoch": 0.06336200887697697, "grad_norm": 6.312346935272217, "learning_rate": 1.1016369903058529e-05, "loss": 2.1553, "step": 7045 }, { "epoch": 0.06337100277461742, "grad_norm": 4.038745880126953, "learning_rate": 1.1009517574234e-05, "loss": 1.9443, "step": 7046 }, { "epoch": 0.06337999667225787, "grad_norm": 9.912015914916992, "learning_rate": 1.1002666775387584e-05, "loss": 1.7042, "step": 7047 }, { "epoch": 0.06338899056989833, "grad_norm": 2.9773800373077393, "learning_rate": 1.0995817507268488e-05, "loss": 2.1418, "step": 7048 }, { "epoch": 0.06339798446753878, "grad_norm": 2.7136709690093994, "learning_rate": 1.0988969770625732e-05, "loss": 2.4164, "step": 7049 }, { "epoch": 0.06340697836517922, "grad_norm": 3.8381521701812744, "learning_rate": 1.0982123566208185e-05, "loss": 2.4172, "step": 7050 }, { "epoch": 0.06341597226281968, "grad_norm": 12.948952674865723, "learning_rate": 1.0975278894764524e-05, "loss": 2.2992, "step": 7051 }, { "epoch": 0.06342496616046013, "grad_norm": 3.509451150894165, "learning_rate": 1.0968435757043275e-05, "loss": 2.021, "step": 7052 }, { "epoch": 0.06343396005810058, "grad_norm": 6.0399556159973145, "learning_rate": 1.0961594153792803e-05, "loss": 2.4012, "step": 7053 }, { "epoch": 0.06344295395574102, "grad_norm": 5.601586818695068, "learning_rate": 1.095475408576128e-05, "loss": 2.827, "step": 7054 }, { "epoch": 0.06345194785338149, "grad_norm": 4.2143635749816895, "learning_rate": 1.0947915553696742e-05, "loss": 1.9032, "step": 7055 }, { "epoch": 0.06346094175102193, "grad_norm": 7.716060161590576, "learning_rate": 1.0941078558347015e-05, "loss": 2.1018, "step": 7056 }, { "epoch": 0.06346993564866238, "grad_norm": 9.372403144836426, "learning_rate": 1.0934243100459817e-05, "loss": 2.797, "step": 7057 }, { "epoch": 0.06347892954630284, "grad_norm": 7.596139907836914, "learning_rate": 1.0927409180782638e-05, "loss": 2.4829, "step": 7058 }, { "epoch": 0.06348792344394329, "grad_norm": 6.180380821228027, "learning_rate": 1.0920576800062842e-05, "loss": 2.4979, "step": 7059 }, { "epoch": 0.06349691734158373, "grad_norm": 4.885025978088379, "learning_rate": 1.091374595904759e-05, "loss": 2.2436, "step": 7060 }, { "epoch": 0.06350591123922418, "grad_norm": 3.1762142181396484, "learning_rate": 1.0906916658483909e-05, "loss": 2.346, "step": 7061 }, { "epoch": 0.06351490513686464, "grad_norm": 5.778194904327393, "learning_rate": 1.0900088899118616e-05, "loss": 2.1916, "step": 7062 }, { "epoch": 0.06352389903450509, "grad_norm": 6.703037738800049, "learning_rate": 1.0893262681698422e-05, "loss": 2.0299, "step": 7063 }, { "epoch": 0.06353289293214553, "grad_norm": 3.3111934661865234, "learning_rate": 1.08864380069698e-05, "loss": 2.5555, "step": 7064 }, { "epoch": 0.063541886829786, "grad_norm": 7.843693733215332, "learning_rate": 1.0879614875679109e-05, "loss": 2.4527, "step": 7065 }, { "epoch": 0.06355088072742644, "grad_norm": 3.846437454223633, "learning_rate": 1.0872793288572497e-05, "loss": 1.9389, "step": 7066 }, { "epoch": 0.06355987462506689, "grad_norm": 3.4276978969573975, "learning_rate": 1.086597324639598e-05, "loss": 2.2097, "step": 7067 }, { "epoch": 0.06356886852270734, "grad_norm": 3.8156473636627197, "learning_rate": 1.0859154749895358e-05, "loss": 2.0735, "step": 7068 }, { "epoch": 0.0635778624203478, "grad_norm": 3.3250226974487305, "learning_rate": 1.0852337799816335e-05, "loss": 2.1822, "step": 7069 }, { "epoch": 0.06358685631798824, "grad_norm": 4.738344669342041, "learning_rate": 1.0845522396904367e-05, "loss": 2.5773, "step": 7070 }, { "epoch": 0.06359585021562869, "grad_norm": 5.805327892303467, "learning_rate": 1.0838708541904797e-05, "loss": 1.7057, "step": 7071 }, { "epoch": 0.06360484411326915, "grad_norm": 3.938621997833252, "learning_rate": 1.0831896235562764e-05, "loss": 2.1644, "step": 7072 }, { "epoch": 0.0636138380109096, "grad_norm": 7.769799709320068, "learning_rate": 1.0825085478623254e-05, "loss": 2.3467, "step": 7073 }, { "epoch": 0.06362283190855005, "grad_norm": 3.9616498947143555, "learning_rate": 1.0818276271831093e-05, "loss": 1.9511, "step": 7074 }, { "epoch": 0.0636318258061905, "grad_norm": 2.4721076488494873, "learning_rate": 1.0811468615930911e-05, "loss": 2.3427, "step": 7075 }, { "epoch": 0.06364081970383095, "grad_norm": 8.32734489440918, "learning_rate": 1.0804662511667189e-05, "loss": 2.0378, "step": 7076 }, { "epoch": 0.0636498136014714, "grad_norm": 2.9716484546661377, "learning_rate": 1.079785795978423e-05, "loss": 2.2981, "step": 7077 }, { "epoch": 0.06365880749911185, "grad_norm": 5.436967372894287, "learning_rate": 1.0791054961026182e-05, "loss": 2.3829, "step": 7078 }, { "epoch": 0.06366780139675231, "grad_norm": 3.779543876647949, "learning_rate": 1.0784253516136991e-05, "loss": 2.2085, "step": 7079 }, { "epoch": 0.06367679529439275, "grad_norm": 3.2265307903289795, "learning_rate": 1.0777453625860472e-05, "loss": 2.0981, "step": 7080 }, { "epoch": 0.0636857891920332, "grad_norm": 2.2811524868011475, "learning_rate": 1.0770655290940224e-05, "loss": 2.107, "step": 7081 }, { "epoch": 0.06369478308967366, "grad_norm": 5.531928062438965, "learning_rate": 1.0763858512119737e-05, "loss": 2.5624, "step": 7082 }, { "epoch": 0.06370377698731411, "grad_norm": 3.6970741748809814, "learning_rate": 1.0757063290142269e-05, "loss": 1.9173, "step": 7083 }, { "epoch": 0.06371277088495456, "grad_norm": 3.4641225337982178, "learning_rate": 1.0750269625750956e-05, "loss": 2.197, "step": 7084 }, { "epoch": 0.063721764782595, "grad_norm": 5.679719924926758, "learning_rate": 1.0743477519688725e-05, "loss": 2.3246, "step": 7085 }, { "epoch": 0.06373075868023546, "grad_norm": 4.059921741485596, "learning_rate": 1.0736686972698364e-05, "loss": 2.1852, "step": 7086 }, { "epoch": 0.06373975257787591, "grad_norm": 3.0419185161590576, "learning_rate": 1.0729897985522455e-05, "loss": 2.1792, "step": 7087 }, { "epoch": 0.06374874647551636, "grad_norm": 3.4877312183380127, "learning_rate": 1.0723110558903465e-05, "loss": 2.1296, "step": 7088 }, { "epoch": 0.06375774037315682, "grad_norm": 5.323472499847412, "learning_rate": 1.0716324693583632e-05, "loss": 1.9844, "step": 7089 }, { "epoch": 0.06376673427079727, "grad_norm": 5.031909942626953, "learning_rate": 1.0709540390305061e-05, "loss": 2.3706, "step": 7090 }, { "epoch": 0.06377572816843771, "grad_norm": 10.888967514038086, "learning_rate": 1.0702757649809663e-05, "loss": 2.34, "step": 7091 }, { "epoch": 0.06378472206607817, "grad_norm": 8.314924240112305, "learning_rate": 1.06959764728392e-05, "loss": 2.712, "step": 7092 }, { "epoch": 0.06379371596371862, "grad_norm": 5.130012035369873, "learning_rate": 1.0689196860135234e-05, "loss": 2.0822, "step": 7093 }, { "epoch": 0.06380270986135907, "grad_norm": 3.5114431381225586, "learning_rate": 1.0682418812439182e-05, "loss": 2.3158, "step": 7094 }, { "epoch": 0.06381170375899951, "grad_norm": 3.2871150970458984, "learning_rate": 1.0675642330492286e-05, "loss": 2.616, "step": 7095 }, { "epoch": 0.06382069765663997, "grad_norm": 7.296317100524902, "learning_rate": 1.0668867415035616e-05, "loss": 2.363, "step": 7096 }, { "epoch": 0.06382969155428042, "grad_norm": 4.586763381958008, "learning_rate": 1.066209406681005e-05, "loss": 2.1414, "step": 7097 }, { "epoch": 0.06383868545192087, "grad_norm": 2.1246871948242188, "learning_rate": 1.0655322286556318e-05, "loss": 2.3168, "step": 7098 }, { "epoch": 0.06384767934956133, "grad_norm": 3.9542267322540283, "learning_rate": 1.0648552075014986e-05, "loss": 2.1395, "step": 7099 }, { "epoch": 0.06385667324720178, "grad_norm": 2.9436230659484863, "learning_rate": 1.064178343292641e-05, "loss": 1.7945, "step": 7100 }, { "epoch": 0.06386566714484222, "grad_norm": 3.0690290927886963, "learning_rate": 1.0635016361030814e-05, "loss": 2.1131, "step": 7101 }, { "epoch": 0.06387466104248267, "grad_norm": 8.230642318725586, "learning_rate": 1.0628250860068228e-05, "loss": 2.4301, "step": 7102 }, { "epoch": 0.06388365494012313, "grad_norm": 4.70937967300415, "learning_rate": 1.0621486930778532e-05, "loss": 2.2571, "step": 7103 }, { "epoch": 0.06389264883776358, "grad_norm": 3.2230963706970215, "learning_rate": 1.0614724573901398e-05, "loss": 1.9159, "step": 7104 }, { "epoch": 0.06390164273540402, "grad_norm": 2.9169037342071533, "learning_rate": 1.0607963790176365e-05, "loss": 2.4093, "step": 7105 }, { "epoch": 0.06391063663304448, "grad_norm": 4.079482078552246, "learning_rate": 1.0601204580342764e-05, "loss": 1.9612, "step": 7106 }, { "epoch": 0.06391963053068493, "grad_norm": 4.888614654541016, "learning_rate": 1.0594446945139783e-05, "loss": 2.9228, "step": 7107 }, { "epoch": 0.06392862442832538, "grad_norm": 4.21964168548584, "learning_rate": 1.0587690885306425e-05, "loss": 2.2926, "step": 7108 }, { "epoch": 0.06393761832596583, "grad_norm": 5.719522476196289, "learning_rate": 1.0580936401581535e-05, "loss": 1.3059, "step": 7109 }, { "epoch": 0.06394661222360629, "grad_norm": 4.369349479675293, "learning_rate": 1.0574183494703748e-05, "loss": 2.1637, "step": 7110 }, { "epoch": 0.06395560612124673, "grad_norm": 4.5037031173706055, "learning_rate": 1.0567432165411576e-05, "loss": 2.2034, "step": 7111 }, { "epoch": 0.06396460001888718, "grad_norm": 3.223085880279541, "learning_rate": 1.0560682414443315e-05, "loss": 1.9114, "step": 7112 }, { "epoch": 0.06397359391652764, "grad_norm": 4.510221481323242, "learning_rate": 1.0553934242537114e-05, "loss": 2.668, "step": 7113 }, { "epoch": 0.06398258781416809, "grad_norm": 5.292646408081055, "learning_rate": 1.0547187650430948e-05, "loss": 2.534, "step": 7114 }, { "epoch": 0.06399158171180853, "grad_norm": 2.959397792816162, "learning_rate": 1.0540442638862618e-05, "loss": 1.8195, "step": 7115 }, { "epoch": 0.064000575609449, "grad_norm": 26.155811309814453, "learning_rate": 1.0533699208569733e-05, "loss": 2.3828, "step": 7116 }, { "epoch": 0.06400956950708944, "grad_norm": 6.323729038238525, "learning_rate": 1.052695736028976e-05, "loss": 1.8711, "step": 7117 }, { "epoch": 0.06401856340472989, "grad_norm": 2.3535659313201904, "learning_rate": 1.0520217094759962e-05, "loss": 1.8819, "step": 7118 }, { "epoch": 0.06402755730237034, "grad_norm": 3.4773292541503906, "learning_rate": 1.0513478412717454e-05, "loss": 2.1218, "step": 7119 }, { "epoch": 0.0640365512000108, "grad_norm": 5.77235221862793, "learning_rate": 1.0506741314899166e-05, "loss": 1.5964, "step": 7120 }, { "epoch": 0.06404554509765124, "grad_norm": 6.314512252807617, "learning_rate": 1.0500005802041867e-05, "loss": 1.8352, "step": 7121 }, { "epoch": 0.06405453899529169, "grad_norm": 3.3132874965667725, "learning_rate": 1.0493271874882124e-05, "loss": 1.9507, "step": 7122 }, { "epoch": 0.06406353289293215, "grad_norm": 2.5372650623321533, "learning_rate": 1.0486539534156356e-05, "loss": 2.14, "step": 7123 }, { "epoch": 0.0640725267905726, "grad_norm": 7.485377311706543, "learning_rate": 1.0479808780600817e-05, "loss": 2.2881, "step": 7124 }, { "epoch": 0.06408152068821305, "grad_norm": 5.468002796173096, "learning_rate": 1.0473079614951545e-05, "loss": 2.0685, "step": 7125 }, { "epoch": 0.06409051458585349, "grad_norm": 8.282814979553223, "learning_rate": 1.046635203794445e-05, "loss": 2.2231, "step": 7126 }, { "epoch": 0.06409950848349395, "grad_norm": 6.8524169921875, "learning_rate": 1.0459626050315241e-05, "loss": 2.7574, "step": 7127 }, { "epoch": 0.0641085023811344, "grad_norm": 5.8245110511779785, "learning_rate": 1.0452901652799477e-05, "loss": 2.3671, "step": 7128 }, { "epoch": 0.06411749627877485, "grad_norm": 4.305452823638916, "learning_rate": 1.0446178846132509e-05, "loss": 2.3904, "step": 7129 }, { "epoch": 0.06412649017641531, "grad_norm": 3.3830549716949463, "learning_rate": 1.0439457631049549e-05, "loss": 1.9517, "step": 7130 }, { "epoch": 0.06413548407405575, "grad_norm": 3.8222804069519043, "learning_rate": 1.0432738008285603e-05, "loss": 2.4856, "step": 7131 }, { "epoch": 0.0641444779716962, "grad_norm": 4.615159034729004, "learning_rate": 1.0426019978575524e-05, "loss": 2.1884, "step": 7132 }, { "epoch": 0.06415347186933666, "grad_norm": 4.479928493499756, "learning_rate": 1.041930354265399e-05, "loss": 2.1999, "step": 7133 }, { "epoch": 0.06416246576697711, "grad_norm": 3.3477625846862793, "learning_rate": 1.041258870125551e-05, "loss": 2.4178, "step": 7134 }, { "epoch": 0.06417145966461756, "grad_norm": 5.419290065765381, "learning_rate": 1.0405875455114387e-05, "loss": 2.3339, "step": 7135 }, { "epoch": 0.064180453562258, "grad_norm": 3.973151683807373, "learning_rate": 1.039916380496479e-05, "loss": 2.599, "step": 7136 }, { "epoch": 0.06418944745989846, "grad_norm": 7.938523292541504, "learning_rate": 1.039245375154068e-05, "loss": 2.3658, "step": 7137 }, { "epoch": 0.06419844135753891, "grad_norm": 3.9356374740600586, "learning_rate": 1.0385745295575866e-05, "loss": 2.6038, "step": 7138 }, { "epoch": 0.06420743525517936, "grad_norm": 3.063079833984375, "learning_rate": 1.0379038437803974e-05, "loss": 2.514, "step": 7139 }, { "epoch": 0.06421642915281982, "grad_norm": 5.427755355834961, "learning_rate": 1.0372333178958462e-05, "loss": 2.441, "step": 7140 }, { "epoch": 0.06422542305046026, "grad_norm": 2.803304672241211, "learning_rate": 1.0365629519772597e-05, "loss": 1.5946, "step": 7141 }, { "epoch": 0.06423441694810071, "grad_norm": 4.523131370544434, "learning_rate": 1.0358927460979493e-05, "loss": 2.2427, "step": 7142 }, { "epoch": 0.06424341084574116, "grad_norm": 5.33206033706665, "learning_rate": 1.0352227003312056e-05, "loss": 2.1739, "step": 7143 }, { "epoch": 0.06425240474338162, "grad_norm": 9.808869361877441, "learning_rate": 1.0345528147503055e-05, "loss": 1.6201, "step": 7144 }, { "epoch": 0.06426139864102207, "grad_norm": 4.127737998962402, "learning_rate": 1.0338830894285065e-05, "loss": 2.2101, "step": 7145 }, { "epoch": 0.06427039253866251, "grad_norm": 4.533472061157227, "learning_rate": 1.0332135244390492e-05, "loss": 2.4221, "step": 7146 }, { "epoch": 0.06427938643630297, "grad_norm": 6.728224277496338, "learning_rate": 1.0325441198551546e-05, "loss": 2.7374, "step": 7147 }, { "epoch": 0.06428838033394342, "grad_norm": 6.683935165405273, "learning_rate": 1.031874875750029e-05, "loss": 2.1913, "step": 7148 }, { "epoch": 0.06429737423158387, "grad_norm": 5.14469575881958, "learning_rate": 1.0312057921968603e-05, "loss": 1.9108, "step": 7149 }, { "epoch": 0.06430636812922431, "grad_norm": 2.951383590698242, "learning_rate": 1.0305368692688174e-05, "loss": 2.5342, "step": 7150 }, { "epoch": 0.06431536202686478, "grad_norm": 3.2194361686706543, "learning_rate": 1.0298681070390528e-05, "loss": 2.6422, "step": 7151 }, { "epoch": 0.06432435592450522, "grad_norm": 2.302138566970825, "learning_rate": 1.0291995055807019e-05, "loss": 1.6115, "step": 7152 }, { "epoch": 0.06433334982214567, "grad_norm": 3.9099574089050293, "learning_rate": 1.0285310649668828e-05, "loss": 2.2714, "step": 7153 }, { "epoch": 0.06434234371978613, "grad_norm": 6.299722194671631, "learning_rate": 1.0278627852706929e-05, "loss": 2.5452, "step": 7154 }, { "epoch": 0.06435133761742658, "grad_norm": 3.4173130989074707, "learning_rate": 1.0271946665652166e-05, "loss": 2.2604, "step": 7155 }, { "epoch": 0.06436033151506702, "grad_norm": 3.9503064155578613, "learning_rate": 1.0265267089235161e-05, "loss": 2.1208, "step": 7156 }, { "epoch": 0.06436932541270748, "grad_norm": 3.1275362968444824, "learning_rate": 1.0258589124186404e-05, "loss": 2.3201, "step": 7157 }, { "epoch": 0.06437831931034793, "grad_norm": 2.0615227222442627, "learning_rate": 1.0251912771236159e-05, "loss": 1.9738, "step": 7158 }, { "epoch": 0.06438731320798838, "grad_norm": 4.424506187438965, "learning_rate": 1.0245238031114578e-05, "loss": 1.9785, "step": 7159 }, { "epoch": 0.06439630710562882, "grad_norm": 3.0794265270233154, "learning_rate": 1.0238564904551574e-05, "loss": 2.7656, "step": 7160 }, { "epoch": 0.06440530100326929, "grad_norm": 4.173437595367432, "learning_rate": 1.0231893392276926e-05, "loss": 1.9914, "step": 7161 }, { "epoch": 0.06441429490090973, "grad_norm": 5.638442039489746, "learning_rate": 1.0225223495020206e-05, "loss": 2.4488, "step": 7162 }, { "epoch": 0.06442328879855018, "grad_norm": 2.1036577224731445, "learning_rate": 1.0218555213510839e-05, "loss": 2.4167, "step": 7163 }, { "epoch": 0.06443228269619064, "grad_norm": 4.083083152770996, "learning_rate": 1.0211888548478033e-05, "loss": 2.5213, "step": 7164 }, { "epoch": 0.06444127659383109, "grad_norm": 4.7286057472229, "learning_rate": 1.0205223500650876e-05, "loss": 2.1183, "step": 7165 }, { "epoch": 0.06445027049147153, "grad_norm": 2.5036818981170654, "learning_rate": 1.0198560070758229e-05, "loss": 2.3101, "step": 7166 }, { "epoch": 0.06445926438911198, "grad_norm": 2.8425209522247314, "learning_rate": 1.0191898259528807e-05, "loss": 2.5645, "step": 7167 }, { "epoch": 0.06446825828675244, "grad_norm": 2.0829615592956543, "learning_rate": 1.0185238067691117e-05, "loss": 1.7118, "step": 7168 }, { "epoch": 0.06447725218439289, "grad_norm": 12.47818374633789, "learning_rate": 1.017857949597352e-05, "loss": 2.1619, "step": 7169 }, { "epoch": 0.06448624608203334, "grad_norm": 5.078470706939697, "learning_rate": 1.0171922545104184e-05, "loss": 2.3574, "step": 7170 }, { "epoch": 0.0644952399796738, "grad_norm": 3.490365743637085, "learning_rate": 1.0165267215811117e-05, "loss": 2.0726, "step": 7171 }, { "epoch": 0.06450423387731424, "grad_norm": 2.149763822555542, "learning_rate": 1.0158613508822112e-05, "loss": 2.1328, "step": 7172 }, { "epoch": 0.06451322777495469, "grad_norm": 5.529705047607422, "learning_rate": 1.015196142486482e-05, "loss": 2.4566, "step": 7173 }, { "epoch": 0.06452222167259515, "grad_norm": 8.320735931396484, "learning_rate": 1.0145310964666713e-05, "loss": 2.2609, "step": 7174 }, { "epoch": 0.0645312155702356, "grad_norm": 10.050851821899414, "learning_rate": 1.0138662128955053e-05, "loss": 1.3984, "step": 7175 }, { "epoch": 0.06454020946787604, "grad_norm": 4.669005870819092, "learning_rate": 1.0132014918456972e-05, "loss": 2.1704, "step": 7176 }, { "epoch": 0.06454920336551649, "grad_norm": 2.8644611835479736, "learning_rate": 1.0125369333899364e-05, "loss": 2.2952, "step": 7177 }, { "epoch": 0.06455819726315695, "grad_norm": 4.578024864196777, "learning_rate": 1.0118725376009025e-05, "loss": 2.3042, "step": 7178 }, { "epoch": 0.0645671911607974, "grad_norm": 1.149499535560608, "learning_rate": 1.0112083045512494e-05, "loss": 1.9775, "step": 7179 }, { "epoch": 0.06457618505843785, "grad_norm": 9.246822357177734, "learning_rate": 1.0105442343136184e-05, "loss": 2.2583, "step": 7180 }, { "epoch": 0.0645851789560783, "grad_norm": 3.424616813659668, "learning_rate": 1.00988032696063e-05, "loss": 1.6696, "step": 7181 }, { "epoch": 0.06459417285371875, "grad_norm": 3.489677667617798, "learning_rate": 1.0092165825648895e-05, "loss": 1.7928, "step": 7182 }, { "epoch": 0.0646031667513592, "grad_norm": 5.866539478302002, "learning_rate": 1.0085530011989803e-05, "loss": 2.3511, "step": 7183 }, { "epoch": 0.06461216064899965, "grad_norm": 9.52908992767334, "learning_rate": 1.0078895829354742e-05, "loss": 2.0099, "step": 7184 }, { "epoch": 0.06462115454664011, "grad_norm": 6.085850238800049, "learning_rate": 1.0072263278469194e-05, "loss": 2.3897, "step": 7185 }, { "epoch": 0.06463014844428056, "grad_norm": 2.7004239559173584, "learning_rate": 1.0065632360058496e-05, "loss": 1.8689, "step": 7186 }, { "epoch": 0.064639142341921, "grad_norm": 3.420656681060791, "learning_rate": 1.0059003074847781e-05, "loss": 1.9117, "step": 7187 }, { "epoch": 0.06464813623956146, "grad_norm": 3.022197723388672, "learning_rate": 1.0052375423562038e-05, "loss": 2.5743, "step": 7188 }, { "epoch": 0.06465713013720191, "grad_norm": 8.565327644348145, "learning_rate": 1.0045749406926026e-05, "loss": 2.4752, "step": 7189 }, { "epoch": 0.06466612403484236, "grad_norm": 9.560060501098633, "learning_rate": 1.0039125025664392e-05, "loss": 2.3464, "step": 7190 }, { "epoch": 0.0646751179324828, "grad_norm": 10.680913925170898, "learning_rate": 1.0032502280501543e-05, "loss": 2.6302, "step": 7191 }, { "epoch": 0.06468411183012326, "grad_norm": 3.063117027282715, "learning_rate": 1.0025881172161752e-05, "loss": 1.9487, "step": 7192 }, { "epoch": 0.06469310572776371, "grad_norm": 4.655853748321533, "learning_rate": 1.0019261701369073e-05, "loss": 2.5819, "step": 7193 }, { "epoch": 0.06470209962540416, "grad_norm": 5.381685256958008, "learning_rate": 1.001264386884741e-05, "loss": 2.7614, "step": 7194 }, { "epoch": 0.06471109352304462, "grad_norm": 8.383171081542969, "learning_rate": 1.0006027675320493e-05, "loss": 1.6572, "step": 7195 }, { "epoch": 0.06472008742068507, "grad_norm": 6.440797805786133, "learning_rate": 9.999413121511833e-06, "loss": 1.9296, "step": 7196 }, { "epoch": 0.06472908131832551, "grad_norm": 4.912498474121094, "learning_rate": 9.992800208144803e-06, "loss": 2.1432, "step": 7197 }, { "epoch": 0.06473807521596597, "grad_norm": 3.061112880706787, "learning_rate": 9.986188935942579e-06, "loss": 2.0416, "step": 7198 }, { "epoch": 0.06474706911360642, "grad_norm": 3.5476036071777344, "learning_rate": 9.97957930562817e-06, "loss": 1.6387, "step": 7199 }, { "epoch": 0.06475606301124687, "grad_norm": 6.67584753036499, "learning_rate": 9.972971317924374e-06, "loss": 2.3568, "step": 7200 }, { "epoch": 0.06476505690888731, "grad_norm": 6.009579658508301, "learning_rate": 9.966364973553852e-06, "loss": 1.6144, "step": 7201 }, { "epoch": 0.06477405080652777, "grad_norm": 3.3558177947998047, "learning_rate": 9.959760273239035e-06, "loss": 2.3487, "step": 7202 }, { "epoch": 0.06478304470416822, "grad_norm": 3.5657949447631836, "learning_rate": 9.953157217702242e-06, "loss": 1.8097, "step": 7203 }, { "epoch": 0.06479203860180867, "grad_norm": 7.778168201446533, "learning_rate": 9.946555807665542e-06, "loss": 2.0922, "step": 7204 }, { "epoch": 0.06480103249944913, "grad_norm": 5.544534206390381, "learning_rate": 9.939956043850876e-06, "loss": 2.1538, "step": 7205 }, { "epoch": 0.06481002639708958, "grad_norm": 2.8538818359375, "learning_rate": 9.933357926979967e-06, "loss": 1.4956, "step": 7206 }, { "epoch": 0.06481902029473002, "grad_norm": 3.8583972454071045, "learning_rate": 9.926761457774389e-06, "loss": 2.0446, "step": 7207 }, { "epoch": 0.06482801419237047, "grad_norm": 6.88946533203125, "learning_rate": 9.920166636955499e-06, "loss": 2.1, "step": 7208 }, { "epoch": 0.06483700809001093, "grad_norm": 6.1566009521484375, "learning_rate": 9.913573465244533e-06, "loss": 2.6523, "step": 7209 }, { "epoch": 0.06484600198765138, "grad_norm": 4.736795425415039, "learning_rate": 9.90698194336248e-06, "loss": 1.9346, "step": 7210 }, { "epoch": 0.06485499588529182, "grad_norm": 6.476447582244873, "learning_rate": 9.900392072030199e-06, "loss": 2.3267, "step": 7211 }, { "epoch": 0.06486398978293229, "grad_norm": 5.5782270431518555, "learning_rate": 9.89380385196833e-06, "loss": 2.4471, "step": 7212 }, { "epoch": 0.06487298368057273, "grad_norm": 5.067736625671387, "learning_rate": 9.88721728389737e-06, "loss": 2.4558, "step": 7213 }, { "epoch": 0.06488197757821318, "grad_norm": 4.243324279785156, "learning_rate": 9.880632368537595e-06, "loss": 2.4611, "step": 7214 }, { "epoch": 0.06489097147585364, "grad_norm": 11.434361457824707, "learning_rate": 9.874049106609135e-06, "loss": 2.0516, "step": 7215 }, { "epoch": 0.06489996537349409, "grad_norm": 2.9553914070129395, "learning_rate": 9.867467498831922e-06, "loss": 2.2812, "step": 7216 }, { "epoch": 0.06490895927113453, "grad_norm": 4.0777153968811035, "learning_rate": 9.86088754592572e-06, "loss": 2.1051, "step": 7217 }, { "epoch": 0.06491795316877498, "grad_norm": 6.7556328773498535, "learning_rate": 9.854309248610085e-06, "loss": 1.5307, "step": 7218 }, { "epoch": 0.06492694706641544, "grad_norm": 3.4520938396453857, "learning_rate": 9.847732607604424e-06, "loss": 2.1198, "step": 7219 }, { "epoch": 0.06493594096405589, "grad_norm": 3.534175395965576, "learning_rate": 9.841157623627947e-06, "loss": 2.3086, "step": 7220 }, { "epoch": 0.06494493486169634, "grad_norm": 2.5806338787078857, "learning_rate": 9.834584297399677e-06, "loss": 2.6058, "step": 7221 }, { "epoch": 0.0649539287593368, "grad_norm": 5.447778701782227, "learning_rate": 9.828012629638464e-06, "loss": 2.2958, "step": 7222 }, { "epoch": 0.06496292265697724, "grad_norm": 5.010773658752441, "learning_rate": 9.821442621062982e-06, "loss": 1.9291, "step": 7223 }, { "epoch": 0.06497191655461769, "grad_norm": 4.7570672035217285, "learning_rate": 9.814874272391722e-06, "loss": 1.9039, "step": 7224 }, { "epoch": 0.06498091045225814, "grad_norm": 6.585616111755371, "learning_rate": 9.808307584342971e-06, "loss": 2.6171, "step": 7225 }, { "epoch": 0.0649899043498986, "grad_norm": 3.7723278999328613, "learning_rate": 9.801742557634872e-06, "loss": 2.1371, "step": 7226 }, { "epoch": 0.06499889824753904, "grad_norm": 3.0170581340789795, "learning_rate": 9.795179192985349e-06, "loss": 2.5637, "step": 7227 }, { "epoch": 0.06500789214517949, "grad_norm": 4.657615661621094, "learning_rate": 9.788617491112167e-06, "loss": 2.1957, "step": 7228 }, { "epoch": 0.06501688604281995, "grad_norm": 14.598859786987305, "learning_rate": 9.782057452732906e-06, "loss": 2.1272, "step": 7229 }, { "epoch": 0.0650258799404604, "grad_norm": 2.3410513401031494, "learning_rate": 9.775499078564973e-06, "loss": 2.1284, "step": 7230 }, { "epoch": 0.06503487383810085, "grad_norm": 4.800159931182861, "learning_rate": 9.768942369325563e-06, "loss": 2.1452, "step": 7231 }, { "epoch": 0.0650438677357413, "grad_norm": 2.901625156402588, "learning_rate": 9.762387325731723e-06, "loss": 1.7056, "step": 7232 }, { "epoch": 0.06505286163338175, "grad_norm": 5.2026801109313965, "learning_rate": 9.755833948500287e-06, "loss": 2.0594, "step": 7233 }, { "epoch": 0.0650618555310222, "grad_norm": 9.699483871459961, "learning_rate": 9.749282238347932e-06, "loss": 2.4479, "step": 7234 }, { "epoch": 0.06507084942866265, "grad_norm": 8.868749618530273, "learning_rate": 9.742732195991142e-06, "loss": 2.1969, "step": 7235 }, { "epoch": 0.06507984332630311, "grad_norm": 7.506131649017334, "learning_rate": 9.736183822146228e-06, "loss": 2.1342, "step": 7236 }, { "epoch": 0.06508883722394355, "grad_norm": 4.297377109527588, "learning_rate": 9.729637117529295e-06, "loss": 2.5039, "step": 7237 }, { "epoch": 0.065097831121584, "grad_norm": 4.715500831604004, "learning_rate": 9.723092082856294e-06, "loss": 2.2687, "step": 7238 }, { "epoch": 0.06510682501922446, "grad_norm": 4.457601070404053, "learning_rate": 9.716548718842968e-06, "loss": 2.2315, "step": 7239 }, { "epoch": 0.06511581891686491, "grad_norm": 5.020309925079346, "learning_rate": 9.710007026204895e-06, "loss": 2.0023, "step": 7240 }, { "epoch": 0.06512481281450536, "grad_norm": 2.1166460514068604, "learning_rate": 9.703467005657466e-06, "loss": 2.1656, "step": 7241 }, { "epoch": 0.0651338067121458, "grad_norm": 2.7564616203308105, "learning_rate": 9.696928657915896e-06, "loss": 2.21, "step": 7242 }, { "epoch": 0.06514280060978626, "grad_norm": 2.647686719894409, "learning_rate": 9.690391983695191e-06, "loss": 2.5908, "step": 7243 }, { "epoch": 0.06515179450742671, "grad_norm": 2.0661332607269287, "learning_rate": 9.683856983710202e-06, "loss": 2.3414, "step": 7244 }, { "epoch": 0.06516078840506716, "grad_norm": 6.796703815460205, "learning_rate": 9.677323658675594e-06, "loss": 2.1434, "step": 7245 }, { "epoch": 0.06516978230270762, "grad_norm": 5.5292277336120605, "learning_rate": 9.670792009305823e-06, "loss": 2.0367, "step": 7246 }, { "epoch": 0.06517877620034807, "grad_norm": 5.851203441619873, "learning_rate": 9.664262036315194e-06, "loss": 2.0211, "step": 7247 }, { "epoch": 0.06518777009798851, "grad_norm": 6.4217448234558105, "learning_rate": 9.657733740417809e-06, "loss": 2.4966, "step": 7248 }, { "epoch": 0.06519676399562896, "grad_norm": 3.089082956314087, "learning_rate": 9.651207122327605e-06, "loss": 2.4196, "step": 7249 }, { "epoch": 0.06520575789326942, "grad_norm": 3.935697078704834, "learning_rate": 9.644682182758306e-06, "loss": 2.5296, "step": 7250 }, { "epoch": 0.06521475179090987, "grad_norm": 3.34030818939209, "learning_rate": 9.638158922423484e-06, "loss": 1.6338, "step": 7251 }, { "epoch": 0.06522374568855031, "grad_norm": 2.563382863998413, "learning_rate": 9.631637342036497e-06, "loss": 2.0229, "step": 7252 }, { "epoch": 0.06523273958619077, "grad_norm": 7.858672142028809, "learning_rate": 9.625117442310546e-06, "loss": 2.4298, "step": 7253 }, { "epoch": 0.06524173348383122, "grad_norm": 3.7875354290008545, "learning_rate": 9.618599223958635e-06, "loss": 2.7156, "step": 7254 }, { "epoch": 0.06525072738147167, "grad_norm": 5.852808952331543, "learning_rate": 9.612082687693597e-06, "loss": 2.2806, "step": 7255 }, { "epoch": 0.06525972127911213, "grad_norm": 3.129194498062134, "learning_rate": 9.605567834228048e-06, "loss": 2.6349, "step": 7256 }, { "epoch": 0.06526871517675258, "grad_norm": 3.983851194381714, "learning_rate": 9.599054664274471e-06, "loss": 1.6061, "step": 7257 }, { "epoch": 0.06527770907439302, "grad_norm": 2.9352493286132812, "learning_rate": 9.592543178545107e-06, "loss": 2.0593, "step": 7258 }, { "epoch": 0.06528670297203347, "grad_norm": 2.3720643520355225, "learning_rate": 9.586033377752057e-06, "loss": 1.5809, "step": 7259 }, { "epoch": 0.06529569686967393, "grad_norm": 3.1746323108673096, "learning_rate": 9.579525262607226e-06, "loss": 2.6716, "step": 7260 }, { "epoch": 0.06530469076731438, "grad_norm": 4.777337551116943, "learning_rate": 9.57301883382233e-06, "loss": 3.1147, "step": 7261 }, { "epoch": 0.06531368466495482, "grad_norm": 3.749465227127075, "learning_rate": 9.566514092108898e-06, "loss": 2.1117, "step": 7262 }, { "epoch": 0.06532267856259528, "grad_norm": 5.918841361999512, "learning_rate": 9.560011038178287e-06, "loss": 2.3077, "step": 7263 }, { "epoch": 0.06533167246023573, "grad_norm": 3.528856039047241, "learning_rate": 9.553509672741645e-06, "loss": 2.5996, "step": 7264 }, { "epoch": 0.06534066635787618, "grad_norm": 5.109477519989014, "learning_rate": 9.547009996509964e-06, "loss": 2.7773, "step": 7265 }, { "epoch": 0.06534966025551663, "grad_norm": 4.847348213195801, "learning_rate": 9.540512010194035e-06, "loss": 2.423, "step": 7266 }, { "epoch": 0.06535865415315709, "grad_norm": 6.887413024902344, "learning_rate": 9.53401571450448e-06, "loss": 2.4901, "step": 7267 }, { "epoch": 0.06536764805079753, "grad_norm": 2.4436187744140625, "learning_rate": 9.527521110151704e-06, "loss": 1.5637, "step": 7268 }, { "epoch": 0.06537664194843798, "grad_norm": 4.798238277435303, "learning_rate": 9.521028197845957e-06, "loss": 2.6094, "step": 7269 }, { "epoch": 0.06538563584607844, "grad_norm": 4.910182476043701, "learning_rate": 9.514536978297303e-06, "loss": 2.368, "step": 7270 }, { "epoch": 0.06539462974371889, "grad_norm": 9.186450004577637, "learning_rate": 9.508047452215593e-06, "loss": 2.3827, "step": 7271 }, { "epoch": 0.06540362364135933, "grad_norm": 3.1733741760253906, "learning_rate": 9.501559620310523e-06, "loss": 2.2843, "step": 7272 }, { "epoch": 0.0654126175389998, "grad_norm": 6.546411037445068, "learning_rate": 9.495073483291592e-06, "loss": 1.8794, "step": 7273 }, { "epoch": 0.06542161143664024, "grad_norm": 4.1234846115112305, "learning_rate": 9.488589041868121e-06, "loss": 2.6883, "step": 7274 }, { "epoch": 0.06543060533428069, "grad_norm": 4.335313320159912, "learning_rate": 9.482106296749221e-06, "loss": 1.9894, "step": 7275 }, { "epoch": 0.06543959923192114, "grad_norm": 4.577296733856201, "learning_rate": 9.475625248643858e-06, "loss": 2.1007, "step": 7276 }, { "epoch": 0.0654485931295616, "grad_norm": 4.961997985839844, "learning_rate": 9.469145898260767e-06, "loss": 2.1306, "step": 7277 }, { "epoch": 0.06545758702720204, "grad_norm": 3.608680009841919, "learning_rate": 9.462668246308529e-06, "loss": 1.9799, "step": 7278 }, { "epoch": 0.06546658092484249, "grad_norm": 7.838219165802002, "learning_rate": 9.456192293495533e-06, "loss": 2.4079, "step": 7279 }, { "epoch": 0.06547557482248295, "grad_norm": 5.000758647918701, "learning_rate": 9.449718040529987e-06, "loss": 2.1382, "step": 7280 }, { "epoch": 0.0654845687201234, "grad_norm": 3.944507122039795, "learning_rate": 9.443245488119886e-06, "loss": 2.6533, "step": 7281 }, { "epoch": 0.06549356261776385, "grad_norm": 2.798320770263672, "learning_rate": 9.43677463697308e-06, "loss": 2.3328, "step": 7282 }, { "epoch": 0.06550255651540429, "grad_norm": 7.7662882804870605, "learning_rate": 9.430305487797191e-06, "loss": 2.2575, "step": 7283 }, { "epoch": 0.06551155041304475, "grad_norm": 7.706813812255859, "learning_rate": 9.423838041299694e-06, "loss": 2.6262, "step": 7284 }, { "epoch": 0.0655205443106852, "grad_norm": 1.5342929363250732, "learning_rate": 9.417372298187833e-06, "loss": 1.9623, "step": 7285 }, { "epoch": 0.06552953820832565, "grad_norm": 4.0396728515625, "learning_rate": 9.410908259168728e-06, "loss": 2.25, "step": 7286 }, { "epoch": 0.06553853210596611, "grad_norm": 4.659127712249756, "learning_rate": 9.404445924949246e-06, "loss": 2.6192, "step": 7287 }, { "epoch": 0.06554752600360655, "grad_norm": 3.149007797241211, "learning_rate": 9.397985296236122e-06, "loss": 1.9778, "step": 7288 }, { "epoch": 0.065556519901247, "grad_norm": 5.488766670227051, "learning_rate": 9.391526373735859e-06, "loss": 1.7967, "step": 7289 }, { "epoch": 0.06556551379888745, "grad_norm": 2.7550225257873535, "learning_rate": 9.385069158154805e-06, "loss": 2.1032, "step": 7290 }, { "epoch": 0.06557450769652791, "grad_norm": 2.409804344177246, "learning_rate": 9.378613650199108e-06, "loss": 2.0621, "step": 7291 }, { "epoch": 0.06558350159416836, "grad_norm": 8.27624797821045, "learning_rate": 9.372159850574749e-06, "loss": 2.0106, "step": 7292 }, { "epoch": 0.0655924954918088, "grad_norm": 3.7074525356292725, "learning_rate": 9.365707759987483e-06, "loss": 2.0901, "step": 7293 }, { "epoch": 0.06560148938944926, "grad_norm": 2.508488178253174, "learning_rate": 9.359257379142907e-06, "loss": 2.3869, "step": 7294 }, { "epoch": 0.06561048328708971, "grad_norm": 2.088977098464966, "learning_rate": 9.352808708746441e-06, "loss": 2.9055, "step": 7295 }, { "epoch": 0.06561947718473016, "grad_norm": 1.9107598066329956, "learning_rate": 9.346361749503279e-06, "loss": 1.9019, "step": 7296 }, { "epoch": 0.06562847108237062, "grad_norm": 3.168135166168213, "learning_rate": 9.339916502118474e-06, "loss": 2.3195, "step": 7297 }, { "epoch": 0.06563746498001106, "grad_norm": 3.930724620819092, "learning_rate": 9.333472967296833e-06, "loss": 2.3124, "step": 7298 }, { "epoch": 0.06564645887765151, "grad_norm": 10.398477554321289, "learning_rate": 9.327031145743054e-06, "loss": 3.0615, "step": 7299 }, { "epoch": 0.06565545277529196, "grad_norm": 2.9169795513153076, "learning_rate": 9.320591038161574e-06, "loss": 2.7527, "step": 7300 }, { "epoch": 0.06566444667293242, "grad_norm": 4.60397481918335, "learning_rate": 9.314152645256696e-06, "loss": 2.2238, "step": 7301 }, { "epoch": 0.06567344057057287, "grad_norm": 3.9417567253112793, "learning_rate": 9.307715967732491e-06, "loss": 2.569, "step": 7302 }, { "epoch": 0.06568243446821331, "grad_norm": 7.791780471801758, "learning_rate": 9.301281006292884e-06, "loss": 1.9974, "step": 7303 }, { "epoch": 0.06569142836585377, "grad_norm": 4.099035739898682, "learning_rate": 9.294847761641565e-06, "loss": 1.6598, "step": 7304 }, { "epoch": 0.06570042226349422, "grad_norm": 4.481351375579834, "learning_rate": 9.2884162344821e-06, "loss": 2.0699, "step": 7305 }, { "epoch": 0.06570941616113467, "grad_norm": 5.6414947509765625, "learning_rate": 9.281986425517808e-06, "loss": 2.1196, "step": 7306 }, { "epoch": 0.06571841005877511, "grad_norm": 5.538795471191406, "learning_rate": 9.275558335451854e-06, "loss": 2.817, "step": 7307 }, { "epoch": 0.06572740395641558, "grad_norm": 3.794971466064453, "learning_rate": 9.26913196498719e-06, "loss": 2.2685, "step": 7308 }, { "epoch": 0.06573639785405602, "grad_norm": 5.2715582847595215, "learning_rate": 9.262707314826616e-06, "loss": 2.7253, "step": 7309 }, { "epoch": 0.06574539175169647, "grad_norm": 3.984691619873047, "learning_rate": 9.25628438567269e-06, "loss": 2.0748, "step": 7310 }, { "epoch": 0.06575438564933693, "grad_norm": 4.0748066902160645, "learning_rate": 9.24986317822785e-06, "loss": 2.3261, "step": 7311 }, { "epoch": 0.06576337954697738, "grad_norm": 2.476614475250244, "learning_rate": 9.243443693194284e-06, "loss": 2.4099, "step": 7312 }, { "epoch": 0.06577237344461782, "grad_norm": 5.111433982849121, "learning_rate": 9.237025931274035e-06, "loss": 2.3257, "step": 7313 }, { "epoch": 0.06578136734225828, "grad_norm": 5.63408899307251, "learning_rate": 9.23060989316892e-06, "loss": 2.5069, "step": 7314 }, { "epoch": 0.06579036123989873, "grad_norm": 4.46622896194458, "learning_rate": 9.224195579580602e-06, "loss": 2.6033, "step": 7315 }, { "epoch": 0.06579935513753918, "grad_norm": 4.123610973358154, "learning_rate": 9.217782991210544e-06, "loss": 2.4478, "step": 7316 }, { "epoch": 0.06580834903517963, "grad_norm": 4.71773624420166, "learning_rate": 9.211372128759994e-06, "loss": 2.512, "step": 7317 }, { "epoch": 0.06581734293282009, "grad_norm": 4.503554821014404, "learning_rate": 9.204962992930063e-06, "loss": 2.0774, "step": 7318 }, { "epoch": 0.06582633683046053, "grad_norm": 3.2885913848876953, "learning_rate": 9.198555584421628e-06, "loss": 2.2082, "step": 7319 }, { "epoch": 0.06583533072810098, "grad_norm": 6.274062156677246, "learning_rate": 9.192149903935405e-06, "loss": 2.7109, "step": 7320 }, { "epoch": 0.06584432462574144, "grad_norm": 3.3591487407684326, "learning_rate": 9.18574595217189e-06, "loss": 1.7829, "step": 7321 }, { "epoch": 0.06585331852338189, "grad_norm": 3.0910613536834717, "learning_rate": 9.179343729831432e-06, "loss": 1.8265, "step": 7322 }, { "epoch": 0.06586231242102233, "grad_norm": 3.2280290126800537, "learning_rate": 9.17294323761414e-06, "loss": 2.4457, "step": 7323 }, { "epoch": 0.06587130631866278, "grad_norm": 2.3742499351501465, "learning_rate": 9.166544476219998e-06, "loss": 2.06, "step": 7324 }, { "epoch": 0.06588030021630324, "grad_norm": 5.010075092315674, "learning_rate": 9.160147446348739e-06, "loss": 2.0864, "step": 7325 }, { "epoch": 0.06588929411394369, "grad_norm": 4.349125385284424, "learning_rate": 9.153752148699945e-06, "loss": 1.524, "step": 7326 }, { "epoch": 0.06589828801158414, "grad_norm": 6.02391242980957, "learning_rate": 9.147358583972988e-06, "loss": 2.1917, "step": 7327 }, { "epoch": 0.0659072819092246, "grad_norm": 3.8503100872039795, "learning_rate": 9.140966752867066e-06, "loss": 1.7489, "step": 7328 }, { "epoch": 0.06591627580686504, "grad_norm": 7.512612819671631, "learning_rate": 9.134576656081162e-06, "loss": 1.913, "step": 7329 }, { "epoch": 0.06592526970450549, "grad_norm": 5.015100955963135, "learning_rate": 9.128188294314119e-06, "loss": 2.2395, "step": 7330 }, { "epoch": 0.06593426360214594, "grad_norm": 2.730726957321167, "learning_rate": 9.121801668264532e-06, "loss": 2.1125, "step": 7331 }, { "epoch": 0.0659432574997864, "grad_norm": 11.073712348937988, "learning_rate": 9.11541677863085e-06, "loss": 1.9181, "step": 7332 }, { "epoch": 0.06595225139742684, "grad_norm": 4.459782600402832, "learning_rate": 9.109033626111299e-06, "loss": 2.2932, "step": 7333 }, { "epoch": 0.06596124529506729, "grad_norm": 8.421353340148926, "learning_rate": 9.102652211403947e-06, "loss": 2.2784, "step": 7334 }, { "epoch": 0.06597023919270775, "grad_norm": 2.573759078979492, "learning_rate": 9.096272535206641e-06, "loss": 2.164, "step": 7335 }, { "epoch": 0.0659792330903482, "grad_norm": 3.401441812515259, "learning_rate": 9.089894598217056e-06, "loss": 3.0632, "step": 7336 }, { "epoch": 0.06598822698798865, "grad_norm": 4.467602729797363, "learning_rate": 9.083518401132679e-06, "loss": 2.5205, "step": 7337 }, { "epoch": 0.06599722088562911, "grad_norm": 4.615909576416016, "learning_rate": 9.077143944650802e-06, "loss": 2.2127, "step": 7338 }, { "epoch": 0.06600621478326955, "grad_norm": 3.132058620452881, "learning_rate": 9.070771229468527e-06, "loss": 2.4774, "step": 7339 }, { "epoch": 0.06601520868091, "grad_norm": 4.170368671417236, "learning_rate": 9.064400256282757e-06, "loss": 2.0153, "step": 7340 }, { "epoch": 0.06602420257855045, "grad_norm": 3.75313138961792, "learning_rate": 9.058031025790222e-06, "loss": 1.8701, "step": 7341 }, { "epoch": 0.06603319647619091, "grad_norm": 6.438293933868408, "learning_rate": 9.051663538687432e-06, "loss": 2.125, "step": 7342 }, { "epoch": 0.06604219037383136, "grad_norm": 5.24947452545166, "learning_rate": 9.045297795670755e-06, "loss": 2.3185, "step": 7343 }, { "epoch": 0.0660511842714718, "grad_norm": 3.769132137298584, "learning_rate": 9.038933797436313e-06, "loss": 2.2918, "step": 7344 }, { "epoch": 0.06606017816911226, "grad_norm": 4.639226913452148, "learning_rate": 9.032571544680086e-06, "loss": 1.4572, "step": 7345 }, { "epoch": 0.06606917206675271, "grad_norm": 5.874194622039795, "learning_rate": 9.026211038097818e-06, "loss": 1.8595, "step": 7346 }, { "epoch": 0.06607816596439316, "grad_norm": 3.943599224090576, "learning_rate": 9.019852278385102e-06, "loss": 2.0763, "step": 7347 }, { "epoch": 0.0660871598620336, "grad_norm": 2.938953161239624, "learning_rate": 9.013495266237306e-06, "loss": 2.2874, "step": 7348 }, { "epoch": 0.06609615375967406, "grad_norm": 2.8642239570617676, "learning_rate": 9.007140002349634e-06, "loss": 2.225, "step": 7349 }, { "epoch": 0.06610514765731451, "grad_norm": 3.546508312225342, "learning_rate": 9.000786487417085e-06, "loss": 2.0501, "step": 7350 }, { "epoch": 0.06611414155495496, "grad_norm": 3.549710273742676, "learning_rate": 8.994434722134481e-06, "loss": 2.1719, "step": 7351 }, { "epoch": 0.06612313545259542, "grad_norm": 5.0488786697387695, "learning_rate": 8.988084707196423e-06, "loss": 2.0816, "step": 7352 }, { "epoch": 0.06613212935023587, "grad_norm": 5.749508857727051, "learning_rate": 8.981736443297357e-06, "loss": 2.0743, "step": 7353 }, { "epoch": 0.06614112324787631, "grad_norm": 7.85929536819458, "learning_rate": 8.9753899311315e-06, "loss": 2.4653, "step": 7354 }, { "epoch": 0.06615011714551677, "grad_norm": 3.375251293182373, "learning_rate": 8.969045171392909e-06, "loss": 2.2114, "step": 7355 }, { "epoch": 0.06615911104315722, "grad_norm": 4.897301197052002, "learning_rate": 8.962702164775435e-06, "loss": 1.7417, "step": 7356 }, { "epoch": 0.06616810494079767, "grad_norm": 3.386094331741333, "learning_rate": 8.956360911972753e-06, "loss": 2.1086, "step": 7357 }, { "epoch": 0.06617709883843811, "grad_norm": 5.012304306030273, "learning_rate": 8.950021413678306e-06, "loss": 2.37, "step": 7358 }, { "epoch": 0.06618609273607858, "grad_norm": 2.7877514362335205, "learning_rate": 8.94368367058539e-06, "loss": 2.0499, "step": 7359 }, { "epoch": 0.06619508663371902, "grad_norm": 3.0804131031036377, "learning_rate": 8.937347683387095e-06, "loss": 1.6116, "step": 7360 }, { "epoch": 0.06620408053135947, "grad_norm": 8.18575382232666, "learning_rate": 8.931013452776299e-06, "loss": 2.2866, "step": 7361 }, { "epoch": 0.06621307442899993, "grad_norm": 3.7890126705169678, "learning_rate": 8.924680979445713e-06, "loss": 2.6607, "step": 7362 }, { "epoch": 0.06622206832664038, "grad_norm": 3.893126964569092, "learning_rate": 8.918350264087847e-06, "loss": 2.1867, "step": 7363 }, { "epoch": 0.06623106222428082, "grad_norm": 4.359896659851074, "learning_rate": 8.912021307395024e-06, "loss": 2.0553, "step": 7364 }, { "epoch": 0.06624005612192127, "grad_norm": 5.575320720672607, "learning_rate": 8.905694110059353e-06, "loss": 2.5321, "step": 7365 }, { "epoch": 0.06624905001956173, "grad_norm": 7.065465450286865, "learning_rate": 8.89936867277279e-06, "loss": 1.827, "step": 7366 }, { "epoch": 0.06625804391720218, "grad_norm": 1.8376741409301758, "learning_rate": 8.893044996227052e-06, "loss": 2.2424, "step": 7367 }, { "epoch": 0.06626703781484262, "grad_norm": 5.006119728088379, "learning_rate": 8.886723081113693e-06, "loss": 2.3679, "step": 7368 }, { "epoch": 0.06627603171248309, "grad_norm": 3.3850107192993164, "learning_rate": 8.880402928124076e-06, "loss": 2.16, "step": 7369 }, { "epoch": 0.06628502561012353, "grad_norm": 2.163510799407959, "learning_rate": 8.874084537949364e-06, "loss": 2.3835, "step": 7370 }, { "epoch": 0.06629401950776398, "grad_norm": 3.326643705368042, "learning_rate": 8.867767911280516e-06, "loss": 2.386, "step": 7371 }, { "epoch": 0.06630301340540443, "grad_norm": 4.3221659660339355, "learning_rate": 8.861453048808325e-06, "loss": 1.6514, "step": 7372 }, { "epoch": 0.06631200730304489, "grad_norm": 4.516854763031006, "learning_rate": 8.855139951223354e-06, "loss": 2.0672, "step": 7373 }, { "epoch": 0.06632100120068533, "grad_norm": 4.9007768630981445, "learning_rate": 8.848828619216007e-06, "loss": 1.742, "step": 7374 }, { "epoch": 0.06632999509832578, "grad_norm": 3.5007436275482178, "learning_rate": 8.842519053476476e-06, "loss": 2.6321, "step": 7375 }, { "epoch": 0.06633898899596624, "grad_norm": 3.9270267486572266, "learning_rate": 8.83621125469478e-06, "loss": 1.8844, "step": 7376 }, { "epoch": 0.06634798289360669, "grad_norm": 3.7022454738616943, "learning_rate": 8.829905223560708e-06, "loss": 1.8786, "step": 7377 }, { "epoch": 0.06635697679124714, "grad_norm": 3.130645751953125, "learning_rate": 8.8236009607639e-06, "loss": 2.0381, "step": 7378 }, { "epoch": 0.0663659706888876, "grad_norm": 2.257274627685547, "learning_rate": 8.817298466993759e-06, "loss": 2.3515, "step": 7379 }, { "epoch": 0.06637496458652804, "grad_norm": 5.592156887054443, "learning_rate": 8.810997742939531e-06, "loss": 2.6675, "step": 7380 }, { "epoch": 0.06638395848416849, "grad_norm": 3.371018409729004, "learning_rate": 8.804698789290247e-06, "loss": 1.7676, "step": 7381 }, { "epoch": 0.06639295238180894, "grad_norm": 3.2780957221984863, "learning_rate": 8.798401606734763e-06, "loss": 2.4241, "step": 7382 }, { "epoch": 0.0664019462794494, "grad_norm": 8.353578567504883, "learning_rate": 8.792106195961714e-06, "loss": 2.3236, "step": 7383 }, { "epoch": 0.06641094017708984, "grad_norm": 2.475710391998291, "learning_rate": 8.785812557659561e-06, "loss": 2.2128, "step": 7384 }, { "epoch": 0.06641993407473029, "grad_norm": 3.54091739654541, "learning_rate": 8.77952069251658e-06, "loss": 2.1318, "step": 7385 }, { "epoch": 0.06642892797237075, "grad_norm": 8.388984680175781, "learning_rate": 8.77323060122082e-06, "loss": 2.2694, "step": 7386 }, { "epoch": 0.0664379218700112, "grad_norm": 8.148481369018555, "learning_rate": 8.766942284460161e-06, "loss": 2.6522, "step": 7387 }, { "epoch": 0.06644691576765165, "grad_norm": 6.530880451202393, "learning_rate": 8.76065574292229e-06, "loss": 1.6587, "step": 7388 }, { "epoch": 0.06645590966529209, "grad_norm": 8.394967079162598, "learning_rate": 8.754370977294702e-06, "loss": 2.2323, "step": 7389 }, { "epoch": 0.06646490356293255, "grad_norm": 3.0811240673065186, "learning_rate": 8.74808798826467e-06, "loss": 1.913, "step": 7390 }, { "epoch": 0.066473897460573, "grad_norm": 4.332529544830322, "learning_rate": 8.741806776519309e-06, "loss": 2.3178, "step": 7391 }, { "epoch": 0.06648289135821345, "grad_norm": 4.01012659072876, "learning_rate": 8.735527342745506e-06, "loss": 2.2935, "step": 7392 }, { "epoch": 0.06649188525585391, "grad_norm": 3.4929630756378174, "learning_rate": 8.729249687629978e-06, "loss": 2.5777, "step": 7393 }, { "epoch": 0.06650087915349435, "grad_norm": 6.232064723968506, "learning_rate": 8.722973811859244e-06, "loss": 2.4818, "step": 7394 }, { "epoch": 0.0665098730511348, "grad_norm": 4.976611137390137, "learning_rate": 8.71669971611963e-06, "loss": 1.8567, "step": 7395 }, { "epoch": 0.06651886694877526, "grad_norm": 4.145784378051758, "learning_rate": 8.710427401097246e-06, "loss": 1.3968, "step": 7396 }, { "epoch": 0.06652786084641571, "grad_norm": 5.894073009490967, "learning_rate": 8.704156867478036e-06, "loss": 1.9628, "step": 7397 }, { "epoch": 0.06653685474405616, "grad_norm": 4.627059459686279, "learning_rate": 8.697888115947725e-06, "loss": 1.7707, "step": 7398 }, { "epoch": 0.0665458486416966, "grad_norm": 7.531310081481934, "learning_rate": 8.691621147191859e-06, "loss": 2.2984, "step": 7399 }, { "epoch": 0.06655484253933706, "grad_norm": 4.32791805267334, "learning_rate": 8.685355961895784e-06, "loss": 2.0509, "step": 7400 }, { "epoch": 0.06656383643697751, "grad_norm": 4.341854572296143, "learning_rate": 8.679092560744662e-06, "loss": 2.419, "step": 7401 }, { "epoch": 0.06657283033461796, "grad_norm": 8.153016090393066, "learning_rate": 8.67283094442343e-06, "loss": 1.9935, "step": 7402 }, { "epoch": 0.06658182423225842, "grad_norm": 4.513627529144287, "learning_rate": 8.666571113616864e-06, "loss": 2.4332, "step": 7403 }, { "epoch": 0.06659081812989887, "grad_norm": 8.03856372833252, "learning_rate": 8.66031306900952e-06, "loss": 2.485, "step": 7404 }, { "epoch": 0.06659981202753931, "grad_norm": 2.7830097675323486, "learning_rate": 8.654056811285772e-06, "loss": 2.7814, "step": 7405 }, { "epoch": 0.06660880592517976, "grad_norm": 6.2588276863098145, "learning_rate": 8.647802341129793e-06, "loss": 2.2208, "step": 7406 }, { "epoch": 0.06661779982282022, "grad_norm": 2.4733469486236572, "learning_rate": 8.641549659225576e-06, "loss": 2.029, "step": 7407 }, { "epoch": 0.06662679372046067, "grad_norm": 3.705824136734009, "learning_rate": 8.635298766256883e-06, "loss": 2.5164, "step": 7408 }, { "epoch": 0.06663578761810111, "grad_norm": 8.53905963897705, "learning_rate": 8.629049662907316e-06, "loss": 2.462, "step": 7409 }, { "epoch": 0.06664478151574157, "grad_norm": 3.38769268989563, "learning_rate": 8.622802349860268e-06, "loss": 1.796, "step": 7410 }, { "epoch": 0.06665377541338202, "grad_norm": 2.8321187496185303, "learning_rate": 8.616556827798927e-06, "loss": 2.1326, "step": 7411 }, { "epoch": 0.06666276931102247, "grad_norm": 2.8645734786987305, "learning_rate": 8.6103130974063e-06, "loss": 2.2583, "step": 7412 }, { "epoch": 0.06667176320866292, "grad_norm": 9.52087116241455, "learning_rate": 8.60407115936519e-06, "loss": 2.5486, "step": 7413 }, { "epoch": 0.06668075710630338, "grad_norm": 4.633691787719727, "learning_rate": 8.597831014358215e-06, "loss": 2.1588, "step": 7414 }, { "epoch": 0.06668975100394382, "grad_norm": 7.940402984619141, "learning_rate": 8.591592663067771e-06, "loss": 2.553, "step": 7415 }, { "epoch": 0.06669874490158427, "grad_norm": 4.504307746887207, "learning_rate": 8.585356106176094e-06, "loss": 2.2828, "step": 7416 }, { "epoch": 0.06670773879922473, "grad_norm": 2.491464138031006, "learning_rate": 8.579121344365185e-06, "loss": 2.0979, "step": 7417 }, { "epoch": 0.06671673269686518, "grad_norm": 2.891463279724121, "learning_rate": 8.572888378316884e-06, "loss": 2.4876, "step": 7418 }, { "epoch": 0.06672572659450562, "grad_norm": 2.7288999557495117, "learning_rate": 8.566657208712797e-06, "loss": 2.3156, "step": 7419 }, { "epoch": 0.06673472049214609, "grad_norm": 3.3602170944213867, "learning_rate": 8.56042783623439e-06, "loss": 2.3508, "step": 7420 }, { "epoch": 0.06674371438978653, "grad_norm": 4.822701454162598, "learning_rate": 8.554200261562866e-06, "loss": 2.6085, "step": 7421 }, { "epoch": 0.06675270828742698, "grad_norm": 4.504384517669678, "learning_rate": 8.547974485379285e-06, "loss": 2.2874, "step": 7422 }, { "epoch": 0.06676170218506743, "grad_norm": 4.042818069458008, "learning_rate": 8.54175050836447e-06, "loss": 2.166, "step": 7423 }, { "epoch": 0.06677069608270789, "grad_norm": 3.084118127822876, "learning_rate": 8.535528331199086e-06, "loss": 2.3022, "step": 7424 }, { "epoch": 0.06677968998034833, "grad_norm": 5.531579971313477, "learning_rate": 8.52930795456355e-06, "loss": 2.6591, "step": 7425 }, { "epoch": 0.06678868387798878, "grad_norm": 3.695723056793213, "learning_rate": 8.523089379138149e-06, "loss": 1.8957, "step": 7426 }, { "epoch": 0.06679767777562924, "grad_norm": 4.20355749130249, "learning_rate": 8.516872605602916e-06, "loss": 2.3896, "step": 7427 }, { "epoch": 0.06680667167326969, "grad_norm": 4.170754432678223, "learning_rate": 8.510657634637717e-06, "loss": 2.8281, "step": 7428 }, { "epoch": 0.06681566557091013, "grad_norm": 6.356776714324951, "learning_rate": 8.504444466922204e-06, "loss": 2.019, "step": 7429 }, { "epoch": 0.06682465946855058, "grad_norm": 6.695620536804199, "learning_rate": 8.49823310313584e-06, "loss": 2.3687, "step": 7430 }, { "epoch": 0.06683365336619104, "grad_norm": 11.708020210266113, "learning_rate": 8.492023543957895e-06, "loss": 1.772, "step": 7431 }, { "epoch": 0.06684264726383149, "grad_norm": 2.018603563308716, "learning_rate": 8.485815790067442e-06, "loss": 2.0232, "step": 7432 }, { "epoch": 0.06685164116147194, "grad_norm": 5.399678707122803, "learning_rate": 8.47960984214334e-06, "loss": 1.7616, "step": 7433 }, { "epoch": 0.0668606350591124, "grad_norm": 3.0477211475372314, "learning_rate": 8.473405700864268e-06, "loss": 2.3575, "step": 7434 }, { "epoch": 0.06686962895675284, "grad_norm": 1.735905408859253, "learning_rate": 8.467203366908707e-06, "loss": 2.438, "step": 7435 }, { "epoch": 0.06687862285439329, "grad_norm": 5.0823493003845215, "learning_rate": 8.461002840954921e-06, "loss": 2.6747, "step": 7436 }, { "epoch": 0.06688761675203375, "grad_norm": 5.033072471618652, "learning_rate": 8.45480412368101e-06, "loss": 2.7489, "step": 7437 }, { "epoch": 0.0668966106496742, "grad_norm": 4.359978199005127, "learning_rate": 8.448607215764823e-06, "loss": 2.1113, "step": 7438 }, { "epoch": 0.06690560454731465, "grad_norm": 5.070064544677734, "learning_rate": 8.442412117884086e-06, "loss": 1.9212, "step": 7439 }, { "epoch": 0.06691459844495509, "grad_norm": 4.279585838317871, "learning_rate": 8.436218830716258e-06, "loss": 1.6727, "step": 7440 }, { "epoch": 0.06692359234259555, "grad_norm": 3.8094754219055176, "learning_rate": 8.430027354938641e-06, "loss": 1.6481, "step": 7441 }, { "epoch": 0.066932586240236, "grad_norm": 8.461146354675293, "learning_rate": 8.423837691228315e-06, "loss": 2.2717, "step": 7442 }, { "epoch": 0.06694158013787645, "grad_norm": 3.3762645721435547, "learning_rate": 8.417649840262184e-06, "loss": 2.4641, "step": 7443 }, { "epoch": 0.06695057403551691, "grad_norm": 5.487915515899658, "learning_rate": 8.411463802716918e-06, "loss": 2.1424, "step": 7444 }, { "epoch": 0.06695956793315735, "grad_norm": 3.2725961208343506, "learning_rate": 8.405279579269046e-06, "loss": 2.0875, "step": 7445 }, { "epoch": 0.0669685618307978, "grad_norm": 3.004772663116455, "learning_rate": 8.399097170594844e-06, "loss": 2.4513, "step": 7446 }, { "epoch": 0.06697755572843825, "grad_norm": 4.146081447601318, "learning_rate": 8.392916577370421e-06, "loss": 2.1603, "step": 7447 }, { "epoch": 0.06698654962607871, "grad_norm": 2.5208051204681396, "learning_rate": 8.386737800271668e-06, "loss": 2.3421, "step": 7448 }, { "epoch": 0.06699554352371916, "grad_norm": 3.6781654357910156, "learning_rate": 8.380560839974296e-06, "loss": 2.7573, "step": 7449 }, { "epoch": 0.0670045374213596, "grad_norm": 6.361738681793213, "learning_rate": 8.374385697153792e-06, "loss": 2.2828, "step": 7450 }, { "epoch": 0.06701353131900006, "grad_norm": 4.668366432189941, "learning_rate": 8.368212372485485e-06, "loss": 2.3397, "step": 7451 }, { "epoch": 0.06702252521664051, "grad_norm": 4.343341827392578, "learning_rate": 8.362040866644457e-06, "loss": 1.9891, "step": 7452 }, { "epoch": 0.06703151911428096, "grad_norm": 3.184880018234253, "learning_rate": 8.355871180305638e-06, "loss": 2.2012, "step": 7453 }, { "epoch": 0.06704051301192142, "grad_norm": 3.8170387744903564, "learning_rate": 8.34970331414371e-06, "loss": 2.339, "step": 7454 }, { "epoch": 0.06704950690956187, "grad_norm": 2.3668203353881836, "learning_rate": 8.343537268833199e-06, "loss": 2.1509, "step": 7455 }, { "epoch": 0.06705850080720231, "grad_norm": 5.788619518280029, "learning_rate": 8.337373045048416e-06, "loss": 2.2729, "step": 7456 }, { "epoch": 0.06706749470484276, "grad_norm": 3.611476421356201, "learning_rate": 8.331210643463455e-06, "loss": 2.3067, "step": 7457 }, { "epoch": 0.06707648860248322, "grad_norm": 3.6642425060272217, "learning_rate": 8.325050064752238e-06, "loss": 2.9622, "step": 7458 }, { "epoch": 0.06708548250012367, "grad_norm": 1.7639399766921997, "learning_rate": 8.318891309588475e-06, "loss": 1.9105, "step": 7459 }, { "epoch": 0.06709447639776411, "grad_norm": 5.580855846405029, "learning_rate": 8.31273437864569e-06, "loss": 2.3416, "step": 7460 }, { "epoch": 0.06710347029540457, "grad_norm": 2.697946548461914, "learning_rate": 8.306579272597173e-06, "loss": 2.1367, "step": 7461 }, { "epoch": 0.06711246419304502, "grad_norm": 4.651733875274658, "learning_rate": 8.300425992116057e-06, "loss": 2.2292, "step": 7462 }, { "epoch": 0.06712145809068547, "grad_norm": 2.6859753131866455, "learning_rate": 8.294274537875232e-06, "loss": 2.1951, "step": 7463 }, { "epoch": 0.06713045198832591, "grad_norm": 3.8497750759124756, "learning_rate": 8.28812491054744e-06, "loss": 2.2255, "step": 7464 }, { "epoch": 0.06713944588596638, "grad_norm": 1.3815503120422363, "learning_rate": 8.281977110805177e-06, "loss": 1.6444, "step": 7465 }, { "epoch": 0.06714843978360682, "grad_norm": 7.632611274719238, "learning_rate": 8.27583113932077e-06, "loss": 2.0827, "step": 7466 }, { "epoch": 0.06715743368124727, "grad_norm": 6.092419147491455, "learning_rate": 8.269686996766315e-06, "loss": 1.6942, "step": 7467 }, { "epoch": 0.06716642757888773, "grad_norm": 4.733609199523926, "learning_rate": 8.263544683813749e-06, "loss": 2.3895, "step": 7468 }, { "epoch": 0.06717542147652818, "grad_norm": 5.288778305053711, "learning_rate": 8.25740420113476e-06, "loss": 1.828, "step": 7469 }, { "epoch": 0.06718441537416862, "grad_norm": 4.690985679626465, "learning_rate": 8.251265549400877e-06, "loss": 2.2352, "step": 7470 }, { "epoch": 0.06719340927180907, "grad_norm": 12.839493751525879, "learning_rate": 8.245128729283414e-06, "loss": 2.9171, "step": 7471 }, { "epoch": 0.06720240316944953, "grad_norm": 4.123460292816162, "learning_rate": 8.238993741453488e-06, "loss": 2.206, "step": 7472 }, { "epoch": 0.06721139706708998, "grad_norm": 8.941717147827148, "learning_rate": 8.232860586582001e-06, "loss": 1.5562, "step": 7473 }, { "epoch": 0.06722039096473043, "grad_norm": 3.208751916885376, "learning_rate": 8.226729265339677e-06, "loss": 2.6414, "step": 7474 }, { "epoch": 0.06722938486237089, "grad_norm": 7.724056243896484, "learning_rate": 8.220599778397017e-06, "loss": 2.4961, "step": 7475 }, { "epoch": 0.06723837876001133, "grad_norm": 6.077045440673828, "learning_rate": 8.214472126424335e-06, "loss": 2.0424, "step": 7476 }, { "epoch": 0.06724737265765178, "grad_norm": 2.699929714202881, "learning_rate": 8.208346310091747e-06, "loss": 2.1653, "step": 7477 }, { "epoch": 0.06725636655529224, "grad_norm": 3.0804948806762695, "learning_rate": 8.202222330069167e-06, "loss": 2.0463, "step": 7478 }, { "epoch": 0.06726536045293269, "grad_norm": 3.5001533031463623, "learning_rate": 8.19610018702629e-06, "loss": 1.9374, "step": 7479 }, { "epoch": 0.06727435435057313, "grad_norm": 1.5374462604522705, "learning_rate": 8.189979881632634e-06, "loss": 1.3691, "step": 7480 }, { "epoch": 0.06728334824821358, "grad_norm": 3.184242010116577, "learning_rate": 8.183861414557511e-06, "loss": 2.1151, "step": 7481 }, { "epoch": 0.06729234214585404, "grad_norm": 2.2811176776885986, "learning_rate": 8.177744786470012e-06, "loss": 1.9312, "step": 7482 }, { "epoch": 0.06730133604349449, "grad_norm": 2.7895495891571045, "learning_rate": 8.171629998039055e-06, "loss": 2.3792, "step": 7483 }, { "epoch": 0.06731032994113494, "grad_norm": 6.675440311431885, "learning_rate": 8.16551704993334e-06, "loss": 2.2556, "step": 7484 }, { "epoch": 0.0673193238387754, "grad_norm": 8.320762634277344, "learning_rate": 8.159405942821375e-06, "loss": 2.553, "step": 7485 }, { "epoch": 0.06732831773641584, "grad_norm": 2.5822839736938477, "learning_rate": 8.153296677371448e-06, "loss": 1.5784, "step": 7486 }, { "epoch": 0.06733731163405629, "grad_norm": 4.285470962524414, "learning_rate": 8.147189254251678e-06, "loss": 2.2522, "step": 7487 }, { "epoch": 0.06734630553169674, "grad_norm": 4.2670793533325195, "learning_rate": 8.141083674129946e-06, "loss": 1.7169, "step": 7488 }, { "epoch": 0.0673552994293372, "grad_norm": 4.713266372680664, "learning_rate": 8.134979937673953e-06, "loss": 2.3385, "step": 7489 }, { "epoch": 0.06736429332697765, "grad_norm": 5.383018493652344, "learning_rate": 8.1288780455512e-06, "loss": 1.8878, "step": 7490 }, { "epoch": 0.06737328722461809, "grad_norm": 2.3105738162994385, "learning_rate": 8.122777998428987e-06, "loss": 2.3597, "step": 7491 }, { "epoch": 0.06738228112225855, "grad_norm": 4.0411601066589355, "learning_rate": 8.116679796974388e-06, "loss": 2.4574, "step": 7492 }, { "epoch": 0.067391275019899, "grad_norm": 3.1504764556884766, "learning_rate": 8.110583441854313e-06, "loss": 2.1019, "step": 7493 }, { "epoch": 0.06740026891753945, "grad_norm": 4.42189884185791, "learning_rate": 8.104488933735432e-06, "loss": 2.3096, "step": 7494 }, { "epoch": 0.06740926281517991, "grad_norm": 3.4144837856292725, "learning_rate": 8.098396273284236e-06, "loss": 2.1965, "step": 7495 }, { "epoch": 0.06741825671282035, "grad_norm": 2.8465394973754883, "learning_rate": 8.092305461167016e-06, "loss": 1.981, "step": 7496 }, { "epoch": 0.0674272506104608, "grad_norm": 3.265697956085205, "learning_rate": 8.086216498049856e-06, "loss": 2.503, "step": 7497 }, { "epoch": 0.06743624450810125, "grad_norm": 4.96607780456543, "learning_rate": 8.080129384598625e-06, "loss": 2.1962, "step": 7498 }, { "epoch": 0.06744523840574171, "grad_norm": 2.516245126724243, "learning_rate": 8.07404412147901e-06, "loss": 1.1129, "step": 7499 }, { "epoch": 0.06745423230338216, "grad_norm": 2.9471426010131836, "learning_rate": 8.067960709356478e-06, "loss": 1.9455, "step": 7500 }, { "epoch": 0.0674632262010226, "grad_norm": 5.945474624633789, "learning_rate": 8.061879148896304e-06, "loss": 2.4985, "step": 7501 }, { "epoch": 0.06747222009866306, "grad_norm": 3.2527096271514893, "learning_rate": 8.05579944076356e-06, "loss": 2.3059, "step": 7502 }, { "epoch": 0.06748121399630351, "grad_norm": 16.322797775268555, "learning_rate": 8.04972158562312e-06, "loss": 2.3043, "step": 7503 }, { "epoch": 0.06749020789394396, "grad_norm": 5.599320411682129, "learning_rate": 8.04364558413964e-06, "loss": 2.0266, "step": 7504 }, { "epoch": 0.0674992017915844, "grad_norm": 3.3041787147521973, "learning_rate": 8.037571436977582e-06, "loss": 2.1508, "step": 7505 }, { "epoch": 0.06750819568922486, "grad_norm": 5.133216857910156, "learning_rate": 8.031499144801216e-06, "loss": 2.5458, "step": 7506 }, { "epoch": 0.06751718958686531, "grad_norm": 3.6276416778564453, "learning_rate": 8.025428708274584e-06, "loss": 2.3273, "step": 7507 }, { "epoch": 0.06752618348450576, "grad_norm": 4.90585994720459, "learning_rate": 8.019360128061551e-06, "loss": 2.3418, "step": 7508 }, { "epoch": 0.06753517738214622, "grad_norm": 2.804438352584839, "learning_rate": 8.01329340482576e-06, "loss": 2.1447, "step": 7509 }, { "epoch": 0.06754417127978667, "grad_norm": 4.229487895965576, "learning_rate": 8.007228539230676e-06, "loss": 2.3362, "step": 7510 }, { "epoch": 0.06755316517742711, "grad_norm": 10.852993965148926, "learning_rate": 8.00116553193952e-06, "loss": 2.2696, "step": 7511 }, { "epoch": 0.06756215907506756, "grad_norm": 2.809643507003784, "learning_rate": 7.995104383615352e-06, "loss": 2.1804, "step": 7512 }, { "epoch": 0.06757115297270802, "grad_norm": 2.33754563331604, "learning_rate": 7.989045094920995e-06, "loss": 2.4674, "step": 7513 }, { "epoch": 0.06758014687034847, "grad_norm": 3.8256027698516846, "learning_rate": 7.98298766651909e-06, "loss": 2.6753, "step": 7514 }, { "epoch": 0.06758914076798891, "grad_norm": 11.2006196975708, "learning_rate": 7.976932099072068e-06, "loss": 2.3859, "step": 7515 }, { "epoch": 0.06759813466562938, "grad_norm": 6.410745620727539, "learning_rate": 7.970878393242168e-06, "loss": 1.9567, "step": 7516 }, { "epoch": 0.06760712856326982, "grad_norm": 2.734118700027466, "learning_rate": 7.964826549691396e-06, "loss": 1.9532, "step": 7517 }, { "epoch": 0.06761612246091027, "grad_norm": 2.1225662231445312, "learning_rate": 7.958776569081586e-06, "loss": 2.4712, "step": 7518 }, { "epoch": 0.06762511635855073, "grad_norm": 4.162598133087158, "learning_rate": 7.952728452074343e-06, "loss": 1.7871, "step": 7519 }, { "epoch": 0.06763411025619118, "grad_norm": 4.862269401550293, "learning_rate": 7.946682199331088e-06, "loss": 2.0691, "step": 7520 }, { "epoch": 0.06764310415383162, "grad_norm": 5.3703742027282715, "learning_rate": 7.940637811513025e-06, "loss": 1.5023, "step": 7521 }, { "epoch": 0.06765209805147207, "grad_norm": 4.160031318664551, "learning_rate": 7.934595289281172e-06, "loss": 2.4489, "step": 7522 }, { "epoch": 0.06766109194911253, "grad_norm": 4.019504547119141, "learning_rate": 7.928554633296312e-06, "loss": 2.2992, "step": 7523 }, { "epoch": 0.06767008584675298, "grad_norm": 5.766537666320801, "learning_rate": 7.922515844219062e-06, "loss": 1.8568, "step": 7524 }, { "epoch": 0.06767907974439342, "grad_norm": 3.899343252182007, "learning_rate": 7.91647892270979e-06, "loss": 1.9859, "step": 7525 }, { "epoch": 0.06768807364203389, "grad_norm": 9.512128829956055, "learning_rate": 7.910443869428704e-06, "loss": 2.6993, "step": 7526 }, { "epoch": 0.06769706753967433, "grad_norm": 6.688137531280518, "learning_rate": 7.90441068503578e-06, "loss": 1.8257, "step": 7527 }, { "epoch": 0.06770606143731478, "grad_norm": 3.370823621749878, "learning_rate": 7.89837937019081e-06, "loss": 1.9774, "step": 7528 }, { "epoch": 0.06771505533495523, "grad_norm": 2.3199691772460938, "learning_rate": 7.89234992555335e-06, "loss": 2.6548, "step": 7529 }, { "epoch": 0.06772404923259569, "grad_norm": 4.438199043273926, "learning_rate": 7.886322351782783e-06, "loss": 1.9323, "step": 7530 }, { "epoch": 0.06773304313023613, "grad_norm": 3.040576934814453, "learning_rate": 7.88029664953828e-06, "loss": 1.6511, "step": 7531 }, { "epoch": 0.06774203702787658, "grad_norm": 2.4226717948913574, "learning_rate": 7.87427281947879e-06, "loss": 2.1404, "step": 7532 }, { "epoch": 0.06775103092551704, "grad_norm": 6.732399940490723, "learning_rate": 7.868250862263074e-06, "loss": 2.3202, "step": 7533 }, { "epoch": 0.06776002482315749, "grad_norm": 2.7009451389312744, "learning_rate": 7.86223077854969e-06, "loss": 1.7172, "step": 7534 }, { "epoch": 0.06776901872079794, "grad_norm": 3.8401176929473877, "learning_rate": 7.856212568996987e-06, "loss": 2.2139, "step": 7535 }, { "epoch": 0.0677780126184384, "grad_norm": 6.671227931976318, "learning_rate": 7.850196234263093e-06, "loss": 2.4927, "step": 7536 }, { "epoch": 0.06778700651607884, "grad_norm": 3.7384727001190186, "learning_rate": 7.844181775005965e-06, "loss": 2.3971, "step": 7537 }, { "epoch": 0.06779600041371929, "grad_norm": 3.5682384967803955, "learning_rate": 7.838169191883316e-06, "loss": 2.1398, "step": 7538 }, { "epoch": 0.06780499431135974, "grad_norm": 8.978638648986816, "learning_rate": 7.83215848555269e-06, "loss": 2.1948, "step": 7539 }, { "epoch": 0.0678139882090002, "grad_norm": 4.593799114227295, "learning_rate": 7.826149656671386e-06, "loss": 2.5454, "step": 7540 }, { "epoch": 0.06782298210664064, "grad_norm": 3.9006707668304443, "learning_rate": 7.82014270589655e-06, "loss": 2.3582, "step": 7541 }, { "epoch": 0.06783197600428109, "grad_norm": 4.472712993621826, "learning_rate": 7.814137633885068e-06, "loss": 2.0129, "step": 7542 }, { "epoch": 0.06784096990192155, "grad_norm": 3.447493076324463, "learning_rate": 7.808134441293666e-06, "loss": 2.2003, "step": 7543 }, { "epoch": 0.067849963799562, "grad_norm": 3.8934619426727295, "learning_rate": 7.802133128778825e-06, "loss": 2.0404, "step": 7544 }, { "epoch": 0.06785895769720245, "grad_norm": 4.676578044891357, "learning_rate": 7.796133696996858e-06, "loss": 2.024, "step": 7545 }, { "epoch": 0.06786795159484289, "grad_norm": 7.839630126953125, "learning_rate": 7.790136146603828e-06, "loss": 1.9074, "step": 7546 }, { "epoch": 0.06787694549248335, "grad_norm": 4.7907609939575195, "learning_rate": 7.784140478255647e-06, "loss": 1.8746, "step": 7547 }, { "epoch": 0.0678859393901238, "grad_norm": 4.247135639190674, "learning_rate": 7.778146692607974e-06, "loss": 2.0167, "step": 7548 }, { "epoch": 0.06789493328776425, "grad_norm": 5.351812362670898, "learning_rate": 7.772154790316294e-06, "loss": 2.0826, "step": 7549 }, { "epoch": 0.06790392718540471, "grad_norm": 2.7021780014038086, "learning_rate": 7.766164772035856e-06, "loss": 2.4426, "step": 7550 }, { "epoch": 0.06791292108304516, "grad_norm": 8.132122039794922, "learning_rate": 7.76017663842173e-06, "loss": 1.7858, "step": 7551 }, { "epoch": 0.0679219149806856, "grad_norm": 9.742754936218262, "learning_rate": 7.754190390128766e-06, "loss": 1.7805, "step": 7552 }, { "epoch": 0.06793090887832605, "grad_norm": 3.5842082500457764, "learning_rate": 7.74820602781162e-06, "loss": 2.6309, "step": 7553 }, { "epoch": 0.06793990277596651, "grad_norm": 5.736945152282715, "learning_rate": 7.742223552124715e-06, "loss": 1.8398, "step": 7554 }, { "epoch": 0.06794889667360696, "grad_norm": 2.657243013381958, "learning_rate": 7.736242963722299e-06, "loss": 1.6759, "step": 7555 }, { "epoch": 0.0679578905712474, "grad_norm": 3.3490488529205322, "learning_rate": 7.730264263258402e-06, "loss": 1.8859, "step": 7556 }, { "epoch": 0.06796688446888786, "grad_norm": 3.3370039463043213, "learning_rate": 7.724287451386833e-06, "loss": 2.0123, "step": 7557 }, { "epoch": 0.06797587836652831, "grad_norm": 5.003236770629883, "learning_rate": 7.718312528761223e-06, "loss": 2.3278, "step": 7558 }, { "epoch": 0.06798487226416876, "grad_norm": 2.4092764854431152, "learning_rate": 7.712339496034956e-06, "loss": 1.8453, "step": 7559 }, { "epoch": 0.06799386616180922, "grad_norm": 9.4157133102417, "learning_rate": 7.706368353861269e-06, "loss": 1.6682, "step": 7560 }, { "epoch": 0.06800286005944967, "grad_norm": 3.425748348236084, "learning_rate": 7.700399102893128e-06, "loss": 1.6255, "step": 7561 }, { "epoch": 0.06801185395709011, "grad_norm": 2.638784885406494, "learning_rate": 7.69443174378334e-06, "loss": 2.2846, "step": 7562 }, { "epoch": 0.06802084785473056, "grad_norm": 3.2500760555267334, "learning_rate": 7.688466277184469e-06, "loss": 1.7125, "step": 7563 }, { "epoch": 0.06802984175237102, "grad_norm": 3.3887104988098145, "learning_rate": 7.682502703748907e-06, "loss": 1.6686, "step": 7564 }, { "epoch": 0.06803883565001147, "grad_norm": 4.3760552406311035, "learning_rate": 7.676541024128798e-06, "loss": 1.8884, "step": 7565 }, { "epoch": 0.06804782954765191, "grad_norm": 4.689037322998047, "learning_rate": 7.67058123897613e-06, "loss": 2.5746, "step": 7566 }, { "epoch": 0.06805682344529237, "grad_norm": 5.091569900512695, "learning_rate": 7.664623348942637e-06, "loss": 1.9259, "step": 7567 }, { "epoch": 0.06806581734293282, "grad_norm": 2.6724281311035156, "learning_rate": 7.65866735467988e-06, "loss": 1.9656, "step": 7568 }, { "epoch": 0.06807481124057327, "grad_norm": 8.891613006591797, "learning_rate": 7.652713256839181e-06, "loss": 2.3584, "step": 7569 }, { "epoch": 0.06808380513821372, "grad_norm": 7.734479904174805, "learning_rate": 7.646761056071686e-06, "loss": 2.2143, "step": 7570 }, { "epoch": 0.06809279903585418, "grad_norm": 3.6209957599639893, "learning_rate": 7.640810753028296e-06, "loss": 1.7855, "step": 7571 }, { "epoch": 0.06810179293349462, "grad_norm": 3.7166802883148193, "learning_rate": 7.634862348359758e-06, "loss": 1.7151, "step": 7572 }, { "epoch": 0.06811078683113507, "grad_norm": 12.61677074432373, "learning_rate": 7.628915842716561e-06, "loss": 2.2922, "step": 7573 }, { "epoch": 0.06811978072877553, "grad_norm": 5.61973762512207, "learning_rate": 7.622971236749013e-06, "loss": 2.7286, "step": 7574 }, { "epoch": 0.06812877462641598, "grad_norm": 3.4473557472229004, "learning_rate": 7.617028531107201e-06, "loss": 2.0669, "step": 7575 }, { "epoch": 0.06813776852405642, "grad_norm": 4.755101680755615, "learning_rate": 7.611087726441013e-06, "loss": 2.2667, "step": 7576 }, { "epoch": 0.06814676242169689, "grad_norm": 3.5610849857330322, "learning_rate": 7.605148823400135e-06, "loss": 1.9495, "step": 7577 }, { "epoch": 0.06815575631933733, "grad_norm": 2.8892531394958496, "learning_rate": 7.59921182263402e-06, "loss": 1.2752, "step": 7578 }, { "epoch": 0.06816475021697778, "grad_norm": 4.821445465087891, "learning_rate": 7.59327672479194e-06, "loss": 2.0723, "step": 7579 }, { "epoch": 0.06817374411461823, "grad_norm": 2.1360416412353516, "learning_rate": 7.587343530522945e-06, "loss": 1.8253, "step": 7580 }, { "epoch": 0.06818273801225869, "grad_norm": 12.110111236572266, "learning_rate": 7.5814122404758905e-06, "loss": 2.4251, "step": 7581 }, { "epoch": 0.06819173190989913, "grad_norm": 4.472696781158447, "learning_rate": 7.5754828552993955e-06, "loss": 2.5683, "step": 7582 }, { "epoch": 0.06820072580753958, "grad_norm": 2.1646273136138916, "learning_rate": 7.5695553756419094e-06, "loss": 2.585, "step": 7583 }, { "epoch": 0.06820971970518004, "grad_norm": 3.9716978073120117, "learning_rate": 7.5636298021516225e-06, "loss": 2.7125, "step": 7584 }, { "epoch": 0.06821871360282049, "grad_norm": 3.02569580078125, "learning_rate": 7.5577061354765835e-06, "loss": 1.8944, "step": 7585 }, { "epoch": 0.06822770750046094, "grad_norm": 2.701047420501709, "learning_rate": 7.551784376264568e-06, "loss": 2.1912, "step": 7586 }, { "epoch": 0.06823670139810138, "grad_norm": 2.6985483169555664, "learning_rate": 7.545864525163188e-06, "loss": 2.4853, "step": 7587 }, { "epoch": 0.06824569529574184, "grad_norm": 2.1804471015930176, "learning_rate": 7.539946582819812e-06, "loss": 2.3009, "step": 7588 }, { "epoch": 0.06825468919338229, "grad_norm": 2.551740884780884, "learning_rate": 7.534030549881638e-06, "loss": 2.3492, "step": 7589 }, { "epoch": 0.06826368309102274, "grad_norm": 2.0381276607513428, "learning_rate": 7.528116426995604e-06, "loss": 1.9649, "step": 7590 }, { "epoch": 0.0682726769886632, "grad_norm": 4.9966511726379395, "learning_rate": 7.522204214808509e-06, "loss": 1.855, "step": 7591 }, { "epoch": 0.06828167088630364, "grad_norm": 4.533949375152588, "learning_rate": 7.5162939139668735e-06, "loss": 1.7273, "step": 7592 }, { "epoch": 0.06829066478394409, "grad_norm": 2.996643304824829, "learning_rate": 7.510385525117058e-06, "loss": 2.5139, "step": 7593 }, { "epoch": 0.06829965868158454, "grad_norm": 4.050113677978516, "learning_rate": 7.504479048905178e-06, "loss": 1.8681, "step": 7594 }, { "epoch": 0.068308652579225, "grad_norm": 1.6919054985046387, "learning_rate": 7.498574485977172e-06, "loss": 2.0977, "step": 7595 }, { "epoch": 0.06831764647686545, "grad_norm": 3.45828914642334, "learning_rate": 7.49267183697874e-06, "loss": 2.2746, "step": 7596 }, { "epoch": 0.06832664037450589, "grad_norm": 2.759734630584717, "learning_rate": 7.486771102555392e-06, "loss": 2.1294, "step": 7597 }, { "epoch": 0.06833563427214635, "grad_norm": 2.5824849605560303, "learning_rate": 7.480872283352428e-06, "loss": 1.9081, "step": 7598 }, { "epoch": 0.0683446281697868, "grad_norm": 7.351283073425293, "learning_rate": 7.47497538001494e-06, "loss": 2.0622, "step": 7599 }, { "epoch": 0.06835362206742725, "grad_norm": 5.520235061645508, "learning_rate": 7.469080393187786e-06, "loss": 2.8619, "step": 7600 }, { "epoch": 0.06836261596506771, "grad_norm": 3.3355939388275146, "learning_rate": 7.463187323515644e-06, "loss": 2.2958, "step": 7601 }, { "epoch": 0.06837160986270815, "grad_norm": 3.135758399963379, "learning_rate": 7.457296171642977e-06, "loss": 2.229, "step": 7602 }, { "epoch": 0.0683806037603486, "grad_norm": 2.262542963027954, "learning_rate": 7.45140693821402e-06, "loss": 1.8935, "step": 7603 }, { "epoch": 0.06838959765798905, "grad_norm": 3.7110908031463623, "learning_rate": 7.445519623872816e-06, "loss": 1.6799, "step": 7604 }, { "epoch": 0.06839859155562951, "grad_norm": 4.400245189666748, "learning_rate": 7.4396342292631895e-06, "loss": 2.8883, "step": 7605 }, { "epoch": 0.06840758545326996, "grad_norm": 4.015838146209717, "learning_rate": 7.433750755028773e-06, "loss": 2.3152, "step": 7606 }, { "epoch": 0.0684165793509104, "grad_norm": 8.741411209106445, "learning_rate": 7.4278692018129535e-06, "loss": 2.1884, "step": 7607 }, { "epoch": 0.06842557324855086, "grad_norm": 3.8026535511016846, "learning_rate": 7.42198957025895e-06, "loss": 1.7599, "step": 7608 }, { "epoch": 0.06843456714619131, "grad_norm": 4.0698981285095215, "learning_rate": 7.416111861009728e-06, "loss": 1.7088, "step": 7609 }, { "epoch": 0.06844356104383176, "grad_norm": 4.6660637855529785, "learning_rate": 7.410236074708077e-06, "loss": 1.9331, "step": 7610 }, { "epoch": 0.0684525549414722, "grad_norm": 3.4654269218444824, "learning_rate": 7.404362211996563e-06, "loss": 2.2602, "step": 7611 }, { "epoch": 0.06846154883911267, "grad_norm": 3.262725591659546, "learning_rate": 7.3984902735175525e-06, "loss": 2.4266, "step": 7612 }, { "epoch": 0.06847054273675311, "grad_norm": 4.226849555969238, "learning_rate": 7.392620259913172e-06, "loss": 2.6279, "step": 7613 }, { "epoch": 0.06847953663439356, "grad_norm": 5.505962371826172, "learning_rate": 7.3867521718253756e-06, "loss": 2.3444, "step": 7614 }, { "epoch": 0.06848853053203402, "grad_norm": 5.794689655303955, "learning_rate": 7.380886009895874e-06, "loss": 2.5008, "step": 7615 }, { "epoch": 0.06849752442967447, "grad_norm": 5.936093330383301, "learning_rate": 7.375021774766189e-06, "loss": 2.4335, "step": 7616 }, { "epoch": 0.06850651832731491, "grad_norm": 6.221467018127441, "learning_rate": 7.369159467077621e-06, "loss": 2.4476, "step": 7617 }, { "epoch": 0.06851551222495537, "grad_norm": 4.056873321533203, "learning_rate": 7.363299087471276e-06, "loss": 2.4175, "step": 7618 }, { "epoch": 0.06852450612259582, "grad_norm": 4.476462364196777, "learning_rate": 7.3574406365880176e-06, "loss": 1.9798, "step": 7619 }, { "epoch": 0.06853350002023627, "grad_norm": 3.008125066757202, "learning_rate": 7.351584115068535e-06, "loss": 2.8111, "step": 7620 }, { "epoch": 0.06854249391787672, "grad_norm": 3.7606968879699707, "learning_rate": 7.345729523553271e-06, "loss": 2.1475, "step": 7621 }, { "epoch": 0.06855148781551718, "grad_norm": 5.504679203033447, "learning_rate": 7.339876862682482e-06, "loss": 2.7369, "step": 7622 }, { "epoch": 0.06856048171315762, "grad_norm": 3.266446113586426, "learning_rate": 7.334026133096211e-06, "loss": 2.3458, "step": 7623 }, { "epoch": 0.06856947561079807, "grad_norm": 3.5694165229797363, "learning_rate": 7.328177335434288e-06, "loss": 2.3749, "step": 7624 }, { "epoch": 0.06857846950843853, "grad_norm": 2.1182138919830322, "learning_rate": 7.3223304703363135e-06, "loss": 1.6195, "step": 7625 }, { "epoch": 0.06858746340607898, "grad_norm": 4.4752092361450195, "learning_rate": 7.316485538441703e-06, "loss": 2.5408, "step": 7626 }, { "epoch": 0.06859645730371942, "grad_norm": 8.55904769897461, "learning_rate": 7.3106425403896535e-06, "loss": 2.1445, "step": 7627 }, { "epoch": 0.06860545120135987, "grad_norm": 3.181638479232788, "learning_rate": 7.3048014768191315e-06, "loss": 2.2703, "step": 7628 }, { "epoch": 0.06861444509900033, "grad_norm": 6.182497501373291, "learning_rate": 7.298962348368918e-06, "loss": 2.1137, "step": 7629 }, { "epoch": 0.06862343899664078, "grad_norm": 3.1190690994262695, "learning_rate": 7.293125155677566e-06, "loss": 2.203, "step": 7630 }, { "epoch": 0.06863243289428123, "grad_norm": 4.3626508712768555, "learning_rate": 7.287289899383437e-06, "loss": 2.6058, "step": 7631 }, { "epoch": 0.06864142679192169, "grad_norm": 6.239588260650635, "learning_rate": 7.2814565801246436e-06, "loss": 2.218, "step": 7632 }, { "epoch": 0.06865042068956213, "grad_norm": 2.236968994140625, "learning_rate": 7.275625198539127e-06, "loss": 2.4853, "step": 7633 }, { "epoch": 0.06865941458720258, "grad_norm": 6.035521507263184, "learning_rate": 7.269795755264583e-06, "loss": 2.6883, "step": 7634 }, { "epoch": 0.06866840848484304, "grad_norm": 2.3262479305267334, "learning_rate": 7.263968250938516e-06, "loss": 2.1071, "step": 7635 }, { "epoch": 0.06867740238248349, "grad_norm": 4.182136535644531, "learning_rate": 7.2581426861982185e-06, "loss": 1.7398, "step": 7636 }, { "epoch": 0.06868639628012393, "grad_norm": 4.500576972961426, "learning_rate": 7.252319061680771e-06, "loss": 2.3822, "step": 7637 }, { "epoch": 0.06869539017776438, "grad_norm": 5.318515777587891, "learning_rate": 7.246497378023018e-06, "loss": 1.891, "step": 7638 }, { "epoch": 0.06870438407540484, "grad_norm": 7.942811012268066, "learning_rate": 7.240677635861626e-06, "loss": 2.2039, "step": 7639 }, { "epoch": 0.06871337797304529, "grad_norm": 2.146995782852173, "learning_rate": 7.234859835833021e-06, "loss": 2.4994, "step": 7640 }, { "epoch": 0.06872237187068574, "grad_norm": 7.736475944519043, "learning_rate": 7.229043978573435e-06, "loss": 2.239, "step": 7641 }, { "epoch": 0.0687313657683262, "grad_norm": 7.48446798324585, "learning_rate": 7.223230064718881e-06, "loss": 2.3225, "step": 7642 }, { "epoch": 0.06874035966596664, "grad_norm": 3.4796934127807617, "learning_rate": 7.217418094905168e-06, "loss": 2.3057, "step": 7643 }, { "epoch": 0.06874935356360709, "grad_norm": 3.46486496925354, "learning_rate": 7.211608069767867e-06, "loss": 1.942, "step": 7644 }, { "epoch": 0.06875834746124754, "grad_norm": 4.764324188232422, "learning_rate": 7.205799989942372e-06, "loss": 2.1605, "step": 7645 }, { "epoch": 0.068767341358888, "grad_norm": 5.306154251098633, "learning_rate": 7.19999385606383e-06, "loss": 2.2958, "step": 7646 }, { "epoch": 0.06877633525652845, "grad_norm": 4.6619768142700195, "learning_rate": 7.194189668767198e-06, "loss": 1.9159, "step": 7647 }, { "epoch": 0.06878532915416889, "grad_norm": 5.802986145019531, "learning_rate": 7.1883874286872124e-06, "loss": 2.498, "step": 7648 }, { "epoch": 0.06879432305180935, "grad_norm": 5.882802963256836, "learning_rate": 7.1825871364584045e-06, "loss": 2.1657, "step": 7649 }, { "epoch": 0.0688033169494498, "grad_norm": 5.59812593460083, "learning_rate": 7.176788792715075e-06, "loss": 2.391, "step": 7650 }, { "epoch": 0.06881231084709025, "grad_norm": 6.378588676452637, "learning_rate": 7.170992398091325e-06, "loss": 1.7689, "step": 7651 }, { "epoch": 0.0688213047447307, "grad_norm": 3.113678455352783, "learning_rate": 7.165197953221048e-06, "loss": 1.8838, "step": 7652 }, { "epoch": 0.06883029864237115, "grad_norm": 4.887275218963623, "learning_rate": 7.1594054587379e-06, "loss": 2.3407, "step": 7653 }, { "epoch": 0.0688392925400116, "grad_norm": 8.33064079284668, "learning_rate": 7.1536149152753505e-06, "loss": 1.8037, "step": 7654 }, { "epoch": 0.06884828643765205, "grad_norm": 3.8997435569763184, "learning_rate": 7.147826323466638e-06, "loss": 2.3519, "step": 7655 }, { "epoch": 0.06885728033529251, "grad_norm": 3.807644844055176, "learning_rate": 7.14203968394481e-06, "loss": 2.356, "step": 7656 }, { "epoch": 0.06886627423293296, "grad_norm": 19.395570755004883, "learning_rate": 7.1362549973426645e-06, "loss": 2.5311, "step": 7657 }, { "epoch": 0.0688752681305734, "grad_norm": 4.717878341674805, "learning_rate": 7.130472264292823e-06, "loss": 2.5894, "step": 7658 }, { "epoch": 0.06888426202821386, "grad_norm": 4.4578680992126465, "learning_rate": 7.124691485427659e-06, "loss": 2.163, "step": 7659 }, { "epoch": 0.06889325592585431, "grad_norm": 4.087946891784668, "learning_rate": 7.118912661379368e-06, "loss": 2.349, "step": 7660 }, { "epoch": 0.06890224982349476, "grad_norm": 4.008911609649658, "learning_rate": 7.11313579277989e-06, "loss": 2.0496, "step": 7661 }, { "epoch": 0.0689112437211352, "grad_norm": 3.3425800800323486, "learning_rate": 7.107360880261002e-06, "loss": 2.54, "step": 7662 }, { "epoch": 0.06892023761877566, "grad_norm": 6.082515716552734, "learning_rate": 7.10158792445422e-06, "loss": 2.0425, "step": 7663 }, { "epoch": 0.06892923151641611, "grad_norm": 5.165694236755371, "learning_rate": 7.0958169259908805e-06, "loss": 2.1996, "step": 7664 }, { "epoch": 0.06893822541405656, "grad_norm": 6.638978004455566, "learning_rate": 7.090047885502077e-06, "loss": 2.192, "step": 7665 }, { "epoch": 0.06894721931169702, "grad_norm": 4.162412166595459, "learning_rate": 7.084280803618718e-06, "loss": 2.2869, "step": 7666 }, { "epoch": 0.06895621320933747, "grad_norm": 4.14676570892334, "learning_rate": 7.078515680971456e-06, "loss": 2.4708, "step": 7667 }, { "epoch": 0.06896520710697791, "grad_norm": 2.9258360862731934, "learning_rate": 7.072752518190792e-06, "loss": 2.3111, "step": 7668 }, { "epoch": 0.06897420100461836, "grad_norm": 4.7684125900268555, "learning_rate": 7.066991315906951e-06, "loss": 2.2532, "step": 7669 }, { "epoch": 0.06898319490225882, "grad_norm": 6.74441385269165, "learning_rate": 7.061232074749985e-06, "loss": 2.1589, "step": 7670 }, { "epoch": 0.06899218879989927, "grad_norm": 4.398915767669678, "learning_rate": 7.055474795349703e-06, "loss": 2.0278, "step": 7671 }, { "epoch": 0.06900118269753971, "grad_norm": 5.578095436096191, "learning_rate": 7.0497194783357174e-06, "loss": 2.0313, "step": 7672 }, { "epoch": 0.06901017659518018, "grad_norm": 4.228109359741211, "learning_rate": 7.04396612433742e-06, "loss": 2.0016, "step": 7673 }, { "epoch": 0.06901917049282062, "grad_norm": 3.953399896621704, "learning_rate": 7.038214733984e-06, "loss": 2.0927, "step": 7674 }, { "epoch": 0.06902816439046107, "grad_norm": 8.853546142578125, "learning_rate": 7.032465307904404e-06, "loss": 1.5946, "step": 7675 }, { "epoch": 0.06903715828810153, "grad_norm": 5.574638366699219, "learning_rate": 7.0267178467273876e-06, "loss": 1.971, "step": 7676 }, { "epoch": 0.06904615218574198, "grad_norm": 2.3280749320983887, "learning_rate": 7.020972351081493e-06, "loss": 2.4444, "step": 7677 }, { "epoch": 0.06905514608338242, "grad_norm": 6.937954425811768, "learning_rate": 7.015228821595024e-06, "loss": 2.3966, "step": 7678 }, { "epoch": 0.06906413998102287, "grad_norm": 6.55820369720459, "learning_rate": 7.009487258896097e-06, "loss": 2.2509, "step": 7679 }, { "epoch": 0.06907313387866333, "grad_norm": 2.45489764213562, "learning_rate": 7.003747663612581e-06, "loss": 2.1725, "step": 7680 }, { "epoch": 0.06908212777630378, "grad_norm": 8.095876693725586, "learning_rate": 6.99801003637218e-06, "loss": 2.0382, "step": 7681 }, { "epoch": 0.06909112167394423, "grad_norm": 3.713296890258789, "learning_rate": 6.992274377802327e-06, "loss": 2.1451, "step": 7682 }, { "epoch": 0.06910011557158469, "grad_norm": 2.9339165687561035, "learning_rate": 6.986540688530283e-06, "loss": 2.8898, "step": 7683 }, { "epoch": 0.06910910946922513, "grad_norm": 3.1713390350341797, "learning_rate": 6.980808969183056e-06, "loss": 2.0707, "step": 7684 }, { "epoch": 0.06911810336686558, "grad_norm": 3.130601167678833, "learning_rate": 6.9750792203874785e-06, "loss": 1.947, "step": 7685 }, { "epoch": 0.06912709726450603, "grad_norm": 5.099554538726807, "learning_rate": 6.969351442770119e-06, "loss": 2.4447, "step": 7686 }, { "epoch": 0.06913609116214649, "grad_norm": 9.605775833129883, "learning_rate": 6.963625636957396e-06, "loss": 2.4276, "step": 7687 }, { "epoch": 0.06914508505978693, "grad_norm": 3.1973557472229004, "learning_rate": 6.957901803575448e-06, "loss": 2.6288, "step": 7688 }, { "epoch": 0.06915407895742738, "grad_norm": 3.3915598392486572, "learning_rate": 6.95217994325024e-06, "loss": 1.6726, "step": 7689 }, { "epoch": 0.06916307285506784, "grad_norm": 3.8360908031463623, "learning_rate": 6.94646005660749e-06, "loss": 2.5547, "step": 7690 }, { "epoch": 0.06917206675270829, "grad_norm": 2.4326720237731934, "learning_rate": 6.940742144272733e-06, "loss": 2.2267, "step": 7691 }, { "epoch": 0.06918106065034874, "grad_norm": 8.62076473236084, "learning_rate": 6.935026206871248e-06, "loss": 2.5678, "step": 7692 }, { "epoch": 0.06919005454798918, "grad_norm": 3.415393590927124, "learning_rate": 6.929312245028152e-06, "loss": 2.0988, "step": 7693 }, { "epoch": 0.06919904844562964, "grad_norm": 3.687582015991211, "learning_rate": 6.923600259368296e-06, "loss": 2.0151, "step": 7694 }, { "epoch": 0.06920804234327009, "grad_norm": 4.785012722015381, "learning_rate": 6.91789025051634e-06, "loss": 2.8423, "step": 7695 }, { "epoch": 0.06921703624091054, "grad_norm": 3.8037657737731934, "learning_rate": 6.912182219096716e-06, "loss": 1.6533, "step": 7696 }, { "epoch": 0.069226030138551, "grad_norm": 3.9255473613739014, "learning_rate": 6.90647616573365e-06, "loss": 2.7331, "step": 7697 }, { "epoch": 0.06923502403619144, "grad_norm": 4.235201358795166, "learning_rate": 6.9007720910511554e-06, "loss": 1.792, "step": 7698 }, { "epoch": 0.06924401793383189, "grad_norm": 7.928804397583008, "learning_rate": 6.895069995673003e-06, "loss": 2.3207, "step": 7699 }, { "epoch": 0.06925301183147235, "grad_norm": 3.777892589569092, "learning_rate": 6.889369880222776e-06, "loss": 2.3824, "step": 7700 }, { "epoch": 0.0692620057291128, "grad_norm": 3.498120069503784, "learning_rate": 6.883671745323833e-06, "loss": 2.2487, "step": 7701 }, { "epoch": 0.06927099962675325, "grad_norm": 6.045474529266357, "learning_rate": 6.877975591599314e-06, "loss": 2.5631, "step": 7702 }, { "epoch": 0.0692799935243937, "grad_norm": 5.093695163726807, "learning_rate": 6.872281419672133e-06, "loss": 1.7114, "step": 7703 }, { "epoch": 0.06928898742203415, "grad_norm": 3.5999302864074707, "learning_rate": 6.86658923016501e-06, "loss": 2.022, "step": 7704 }, { "epoch": 0.0692979813196746, "grad_norm": 4.547231197357178, "learning_rate": 6.860899023700407e-06, "loss": 2.4348, "step": 7705 }, { "epoch": 0.06930697521731505, "grad_norm": 3.016261100769043, "learning_rate": 6.855210800900633e-06, "loss": 2.5653, "step": 7706 }, { "epoch": 0.06931596911495551, "grad_norm": 3.6654975414276123, "learning_rate": 6.849524562387716e-06, "loss": 2.1029, "step": 7707 }, { "epoch": 0.06932496301259596, "grad_norm": 7.284987449645996, "learning_rate": 6.843840308783514e-06, "loss": 1.8867, "step": 7708 }, { "epoch": 0.0693339569102364, "grad_norm": 6.0854716300964355, "learning_rate": 6.83815804070963e-06, "loss": 2.1475, "step": 7709 }, { "epoch": 0.06934295080787685, "grad_norm": 5.217992305755615, "learning_rate": 6.832477758787484e-06, "loss": 2.7471, "step": 7710 }, { "epoch": 0.06935194470551731, "grad_norm": 4.076650619506836, "learning_rate": 6.826799463638239e-06, "loss": 1.9444, "step": 7711 }, { "epoch": 0.06936093860315776, "grad_norm": 3.1201610565185547, "learning_rate": 6.8211231558829e-06, "loss": 2.0936, "step": 7712 }, { "epoch": 0.0693699325007982, "grad_norm": 3.5972137451171875, "learning_rate": 6.815448836142191e-06, "loss": 1.8895, "step": 7713 }, { "epoch": 0.06937892639843866, "grad_norm": 5.160998821258545, "learning_rate": 6.809776505036667e-06, "loss": 2.358, "step": 7714 }, { "epoch": 0.06938792029607911, "grad_norm": 2.4643797874450684, "learning_rate": 6.8041061631866245e-06, "loss": 2.2553, "step": 7715 }, { "epoch": 0.06939691419371956, "grad_norm": 3.137948751449585, "learning_rate": 6.798437811212185e-06, "loss": 2.4062, "step": 7716 }, { "epoch": 0.06940590809136002, "grad_norm": 4.457751750946045, "learning_rate": 6.792771449733215e-06, "loss": 2.0174, "step": 7717 }, { "epoch": 0.06941490198900047, "grad_norm": 2.9499924182891846, "learning_rate": 6.7871070793693815e-06, "loss": 2.6444, "step": 7718 }, { "epoch": 0.06942389588664091, "grad_norm": 3.116267442703247, "learning_rate": 6.781444700740139e-06, "loss": 2.0754, "step": 7719 }, { "epoch": 0.06943288978428136, "grad_norm": 5.152554035186768, "learning_rate": 6.775784314464717e-06, "loss": 2.3474, "step": 7720 }, { "epoch": 0.06944188368192182, "grad_norm": 4.239101886749268, "learning_rate": 6.7701259211621154e-06, "loss": 1.9775, "step": 7721 }, { "epoch": 0.06945087757956227, "grad_norm": 4.624876022338867, "learning_rate": 6.764469521451136e-06, "loss": 2.1076, "step": 7722 }, { "epoch": 0.06945987147720271, "grad_norm": 6.536465167999268, "learning_rate": 6.758815115950362e-06, "loss": 2.4557, "step": 7723 }, { "epoch": 0.06946886537484318, "grad_norm": 2.594398260116577, "learning_rate": 6.753162705278132e-06, "loss": 2.3114, "step": 7724 }, { "epoch": 0.06947785927248362, "grad_norm": 15.670150756835938, "learning_rate": 6.747512290052596e-06, "loss": 2.1758, "step": 7725 }, { "epoch": 0.06948685317012407, "grad_norm": 5.686553955078125, "learning_rate": 6.741863870891671e-06, "loss": 2.0489, "step": 7726 }, { "epoch": 0.06949584706776452, "grad_norm": 11.775077819824219, "learning_rate": 6.736217448413073e-06, "loss": 2.0752, "step": 7727 }, { "epoch": 0.06950484096540498, "grad_norm": 4.035073280334473, "learning_rate": 6.730573023234269e-06, "loss": 2.3419, "step": 7728 }, { "epoch": 0.06951383486304542, "grad_norm": 2.82887864112854, "learning_rate": 6.724930595972537e-06, "loss": 0.7748, "step": 7729 }, { "epoch": 0.06952282876068587, "grad_norm": 5.257516860961914, "learning_rate": 6.71929016724491e-06, "loss": 2.4014, "step": 7730 }, { "epoch": 0.06953182265832633, "grad_norm": 1.8271185159683228, "learning_rate": 6.713651737668228e-06, "loss": 2.4054, "step": 7731 }, { "epoch": 0.06954081655596678, "grad_norm": 6.561675548553467, "learning_rate": 6.708015307859098e-06, "loss": 2.4483, "step": 7732 }, { "epoch": 0.06954981045360722, "grad_norm": 5.785775184631348, "learning_rate": 6.70238087843392e-06, "loss": 2.1214, "step": 7733 }, { "epoch": 0.06955880435124767, "grad_norm": 3.7114803791046143, "learning_rate": 6.696748450008852e-06, "loss": 2.6964, "step": 7734 }, { "epoch": 0.06956779824888813, "grad_norm": 9.54886245727539, "learning_rate": 6.691118023199861e-06, "loss": 2.0515, "step": 7735 }, { "epoch": 0.06957679214652858, "grad_norm": 13.76387882232666, "learning_rate": 6.685489598622669e-06, "loss": 3.0626, "step": 7736 }, { "epoch": 0.06958578604416903, "grad_norm": 4.30763578414917, "learning_rate": 6.6798631768928e-06, "loss": 2.0045, "step": 7737 }, { "epoch": 0.06959477994180949, "grad_norm": 5.171612739562988, "learning_rate": 6.6742387586255505e-06, "loss": 1.8218, "step": 7738 }, { "epoch": 0.06960377383944993, "grad_norm": 10.77538776397705, "learning_rate": 6.668616344436004e-06, "loss": 2.4166, "step": 7739 }, { "epoch": 0.06961276773709038, "grad_norm": 8.130552291870117, "learning_rate": 6.662995934939007e-06, "loss": 1.9099, "step": 7740 }, { "epoch": 0.06962176163473084, "grad_norm": 2.4439733028411865, "learning_rate": 6.657377530749212e-06, "loss": 2.0192, "step": 7741 }, { "epoch": 0.06963075553237129, "grad_norm": 3.2620110511779785, "learning_rate": 6.6517611324810245e-06, "loss": 1.8354, "step": 7742 }, { "epoch": 0.06963974943001174, "grad_norm": 9.226058959960938, "learning_rate": 6.646146740748655e-06, "loss": 2.5139, "step": 7743 }, { "epoch": 0.06964874332765218, "grad_norm": 4.417866230010986, "learning_rate": 6.640534356166081e-06, "loss": 2.3979, "step": 7744 }, { "epoch": 0.06965773722529264, "grad_norm": 2.4956307411193848, "learning_rate": 6.634923979347074e-06, "loss": 1.5065, "step": 7745 }, { "epoch": 0.06966673112293309, "grad_norm": 2.719775438308716, "learning_rate": 6.6293156109051655e-06, "loss": 2.5476, "step": 7746 }, { "epoch": 0.06967572502057354, "grad_norm": 5.625834941864014, "learning_rate": 6.623709251453677e-06, "loss": 2.034, "step": 7747 }, { "epoch": 0.069684718918214, "grad_norm": 2.8849973678588867, "learning_rate": 6.618104901605729e-06, "loss": 1.8181, "step": 7748 }, { "epoch": 0.06969371281585444, "grad_norm": 2.401087999343872, "learning_rate": 6.612502561974182e-06, "loss": 2.0742, "step": 7749 }, { "epoch": 0.06970270671349489, "grad_norm": 4.101959705352783, "learning_rate": 6.606902233171711e-06, "loss": 2.1766, "step": 7750 }, { "epoch": 0.06971170061113534, "grad_norm": 5.9570112228393555, "learning_rate": 6.601303915810755e-06, "loss": 2.2206, "step": 7751 }, { "epoch": 0.0697206945087758, "grad_norm": 6.155735969543457, "learning_rate": 6.5957076105035514e-06, "loss": 2.4093, "step": 7752 }, { "epoch": 0.06972968840641625, "grad_norm": 3.2626590728759766, "learning_rate": 6.5901133178620844e-06, "loss": 1.5642, "step": 7753 }, { "epoch": 0.06973868230405669, "grad_norm": 4.681911468505859, "learning_rate": 6.584521038498153e-06, "loss": 2.5897, "step": 7754 }, { "epoch": 0.06974767620169715, "grad_norm": 2.279080867767334, "learning_rate": 6.5789307730233065e-06, "loss": 2.3576, "step": 7755 }, { "epoch": 0.0697566700993376, "grad_norm": 6.06007719039917, "learning_rate": 6.573342522048895e-06, "loss": 1.8622, "step": 7756 }, { "epoch": 0.06976566399697805, "grad_norm": 3.2553725242614746, "learning_rate": 6.5677562861860415e-06, "loss": 1.6405, "step": 7757 }, { "epoch": 0.06977465789461851, "grad_norm": 3.97701096534729, "learning_rate": 6.562172066045655e-06, "loss": 2.0862, "step": 7758 }, { "epoch": 0.06978365179225895, "grad_norm": 2.9200124740600586, "learning_rate": 6.556589862238405e-06, "loss": 2.0243, "step": 7759 }, { "epoch": 0.0697926456898994, "grad_norm": 3.5185673236846924, "learning_rate": 6.551009675374764e-06, "loss": 1.8424, "step": 7760 }, { "epoch": 0.06980163958753985, "grad_norm": 4.226288318634033, "learning_rate": 6.545431506064959e-06, "loss": 2.0764, "step": 7761 }, { "epoch": 0.06981063348518031, "grad_norm": 2.729398488998413, "learning_rate": 6.539855354919019e-06, "loss": 2.345, "step": 7762 }, { "epoch": 0.06981962738282076, "grad_norm": 4.739338397979736, "learning_rate": 6.5342812225467455e-06, "loss": 2.3106, "step": 7763 }, { "epoch": 0.0698286212804612, "grad_norm": 4.2003865242004395, "learning_rate": 6.528709109557721e-06, "loss": 2.3792, "step": 7764 }, { "epoch": 0.06983761517810166, "grad_norm": 3.006962537765503, "learning_rate": 6.5231390165612884e-06, "loss": 1.9589, "step": 7765 }, { "epoch": 0.06984660907574211, "grad_norm": 8.544995307922363, "learning_rate": 6.517570944166601e-06, "loss": 2.3114, "step": 7766 }, { "epoch": 0.06985560297338256, "grad_norm": 6.033580780029297, "learning_rate": 6.512004892982559e-06, "loss": 2.4658, "step": 7767 }, { "epoch": 0.069864596871023, "grad_norm": 6.313393592834473, "learning_rate": 6.506440863617869e-06, "loss": 2.6993, "step": 7768 }, { "epoch": 0.06987359076866347, "grad_norm": 2.50944185256958, "learning_rate": 6.500878856680998e-06, "loss": 2.0486, "step": 7769 }, { "epoch": 0.06988258466630391, "grad_norm": 2.8580470085144043, "learning_rate": 6.49531887278021e-06, "loss": 2.4094, "step": 7770 }, { "epoch": 0.06989157856394436, "grad_norm": 7.75752592086792, "learning_rate": 6.489760912523524e-06, "loss": 2.3064, "step": 7771 }, { "epoch": 0.06990057246158482, "grad_norm": 4.5310211181640625, "learning_rate": 6.484204976518751e-06, "loss": 2.5519, "step": 7772 }, { "epoch": 0.06990956635922527, "grad_norm": 3.8980319499969482, "learning_rate": 6.478651065373495e-06, "loss": 2.1073, "step": 7773 }, { "epoch": 0.06991856025686571, "grad_norm": 6.796755313873291, "learning_rate": 6.473099179695103e-06, "loss": 2.4592, "step": 7774 }, { "epoch": 0.06992755415450616, "grad_norm": 5.061224460601807, "learning_rate": 6.46754932009073e-06, "loss": 1.9014, "step": 7775 }, { "epoch": 0.06993654805214662, "grad_norm": 1.879467248916626, "learning_rate": 6.462001487167299e-06, "loss": 1.8484, "step": 7776 }, { "epoch": 0.06994554194978707, "grad_norm": 4.033999919891357, "learning_rate": 6.456455681531523e-06, "loss": 1.9709, "step": 7777 }, { "epoch": 0.06995453584742752, "grad_norm": 9.343409538269043, "learning_rate": 6.450911903789863e-06, "loss": 2.1459, "step": 7778 }, { "epoch": 0.06996352974506798, "grad_norm": 4.434802055358887, "learning_rate": 6.445370154548602e-06, "loss": 2.4442, "step": 7779 }, { "epoch": 0.06997252364270842, "grad_norm": 5.227975845336914, "learning_rate": 6.439830434413754e-06, "loss": 2.1812, "step": 7780 }, { "epoch": 0.06998151754034887, "grad_norm": 4.332169532775879, "learning_rate": 6.434292743991144e-06, "loss": 2.0753, "step": 7781 }, { "epoch": 0.06999051143798933, "grad_norm": 3.1318676471710205, "learning_rate": 6.428757083886366e-06, "loss": 2.021, "step": 7782 }, { "epoch": 0.06999950533562978, "grad_norm": 4.755331039428711, "learning_rate": 6.423223454704802e-06, "loss": 1.581, "step": 7783 }, { "epoch": 0.07000849923327022, "grad_norm": 4.8392653465271, "learning_rate": 6.4176918570515834e-06, "loss": 2.3097, "step": 7784 }, { "epoch": 0.07001749313091067, "grad_norm": 7.973690986633301, "learning_rate": 6.412162291531651e-06, "loss": 2.6832, "step": 7785 }, { "epoch": 0.07002648702855113, "grad_norm": 3.5690975189208984, "learning_rate": 6.406634758749699e-06, "loss": 2.8465, "step": 7786 }, { "epoch": 0.07003548092619158, "grad_norm": 8.82128620147705, "learning_rate": 6.40110925931022e-06, "loss": 2.8936, "step": 7787 }, { "epoch": 0.07004447482383203, "grad_norm": 6.823626518249512, "learning_rate": 6.395585793817454e-06, "loss": 2.6071, "step": 7788 }, { "epoch": 0.07005346872147249, "grad_norm": 3.5309691429138184, "learning_rate": 6.390064362875472e-06, "loss": 1.9331, "step": 7789 }, { "epoch": 0.07006246261911293, "grad_norm": 2.9072141647338867, "learning_rate": 6.384544967088063e-06, "loss": 2.4218, "step": 7790 }, { "epoch": 0.07007145651675338, "grad_norm": 5.021820068359375, "learning_rate": 6.379027607058835e-06, "loss": 1.6852, "step": 7791 }, { "epoch": 0.07008045041439383, "grad_norm": 16.01329803466797, "learning_rate": 6.373512283391145e-06, "loss": 2.249, "step": 7792 }, { "epoch": 0.07008944431203429, "grad_norm": 3.9212849140167236, "learning_rate": 6.367998996688149e-06, "loss": 2.4887, "step": 7793 }, { "epoch": 0.07009843820967473, "grad_norm": 14.718862533569336, "learning_rate": 6.362487747552767e-06, "loss": 2.3097, "step": 7794 }, { "epoch": 0.07010743210731518, "grad_norm": 3.098414182662964, "learning_rate": 6.3569785365877125e-06, "loss": 1.4826, "step": 7795 }, { "epoch": 0.07011642600495564, "grad_norm": 3.5540170669555664, "learning_rate": 6.3514713643954475e-06, "loss": 2.1545, "step": 7796 }, { "epoch": 0.07012541990259609, "grad_norm": 4.299899578094482, "learning_rate": 6.345966231578238e-06, "loss": 1.9835, "step": 7797 }, { "epoch": 0.07013441380023654, "grad_norm": 9.545814514160156, "learning_rate": 6.340463138738123e-06, "loss": 2.134, "step": 7798 }, { "epoch": 0.070143407697877, "grad_norm": 9.936206817626953, "learning_rate": 6.334962086476895e-06, "loss": 2.1636, "step": 7799 }, { "epoch": 0.07015240159551744, "grad_norm": 4.559581756591797, "learning_rate": 6.329463075396161e-06, "loss": 2.0013, "step": 7800 }, { "epoch": 0.07016139549315789, "grad_norm": 6.504311561584473, "learning_rate": 6.323966106097257e-06, "loss": 1.8072, "step": 7801 }, { "epoch": 0.07017038939079834, "grad_norm": 3.8817625045776367, "learning_rate": 6.318471179181357e-06, "loss": 2.0481, "step": 7802 }, { "epoch": 0.0701793832884388, "grad_norm": 5.3645339012146, "learning_rate": 6.312978295249356e-06, "loss": 1.6325, "step": 7803 }, { "epoch": 0.07018837718607925, "grad_norm": 19.421476364135742, "learning_rate": 6.307487454901956e-06, "loss": 2.4894, "step": 7804 }, { "epoch": 0.07019737108371969, "grad_norm": 4.201968669891357, "learning_rate": 6.301998658739619e-06, "loss": 2.0053, "step": 7805 }, { "epoch": 0.07020636498136015, "grad_norm": 4.021116256713867, "learning_rate": 6.296511907362607e-06, "loss": 2.0911, "step": 7806 }, { "epoch": 0.0702153588790006, "grad_norm": 1.6286221742630005, "learning_rate": 6.2910272013709135e-06, "loss": 2.0095, "step": 7807 }, { "epoch": 0.07022435277664105, "grad_norm": 5.294590950012207, "learning_rate": 6.285544541364374e-06, "loss": 1.8673, "step": 7808 }, { "epoch": 0.0702333466742815, "grad_norm": 4.024623870849609, "learning_rate": 6.280063927942536e-06, "loss": 2.3879, "step": 7809 }, { "epoch": 0.07024234057192195, "grad_norm": 4.34070348739624, "learning_rate": 6.274585361704774e-06, "loss": 2.2052, "step": 7810 }, { "epoch": 0.0702513344695624, "grad_norm": 3.3180882930755615, "learning_rate": 6.269108843250193e-06, "loss": 2.4005, "step": 7811 }, { "epoch": 0.07026032836720285, "grad_norm": 3.1882495880126953, "learning_rate": 6.263634373177716e-06, "loss": 2.0714, "step": 7812 }, { "epoch": 0.07026932226484331, "grad_norm": 3.4634206295013428, "learning_rate": 6.258161952086e-06, "loss": 2.1733, "step": 7813 }, { "epoch": 0.07027831616248376, "grad_norm": 2.737055778503418, "learning_rate": 6.2526915805735275e-06, "loss": 1.8185, "step": 7814 }, { "epoch": 0.0702873100601242, "grad_norm": 2.4061782360076904, "learning_rate": 6.247223259238511e-06, "loss": 2.1913, "step": 7815 }, { "epoch": 0.07029630395776465, "grad_norm": 8.239388465881348, "learning_rate": 6.2417569886789735e-06, "loss": 2.0002, "step": 7816 }, { "epoch": 0.07030529785540511, "grad_norm": 2.7826294898986816, "learning_rate": 6.2362927694926795e-06, "loss": 3.0277, "step": 7817 }, { "epoch": 0.07031429175304556, "grad_norm": 5.594570159912109, "learning_rate": 6.230830602277199e-06, "loss": 2.2543, "step": 7818 }, { "epoch": 0.070323285650686, "grad_norm": 3.4899401664733887, "learning_rate": 6.225370487629872e-06, "loss": 2.7272, "step": 7819 }, { "epoch": 0.07033227954832647, "grad_norm": 3.582427501678467, "learning_rate": 6.219912426147795e-06, "loss": 2.6331, "step": 7820 }, { "epoch": 0.07034127344596691, "grad_norm": 9.650754928588867, "learning_rate": 6.214456418427858e-06, "loss": 2.4233, "step": 7821 }, { "epoch": 0.07035026734360736, "grad_norm": 5.112184524536133, "learning_rate": 6.209002465066724e-06, "loss": 2.028, "step": 7822 }, { "epoch": 0.07035926124124782, "grad_norm": 9.00005054473877, "learning_rate": 6.203550566660835e-06, "loss": 2.364, "step": 7823 }, { "epoch": 0.07036825513888827, "grad_norm": 3.53507137298584, "learning_rate": 6.198100723806388e-06, "loss": 2.2533, "step": 7824 }, { "epoch": 0.07037724903652871, "grad_norm": 10.4852876663208, "learning_rate": 6.192652937099388e-06, "loss": 2.3527, "step": 7825 }, { "epoch": 0.07038624293416916, "grad_norm": 4.005497455596924, "learning_rate": 6.18720720713557e-06, "loss": 1.9223, "step": 7826 }, { "epoch": 0.07039523683180962, "grad_norm": 3.378408432006836, "learning_rate": 6.181763534510502e-06, "loss": 2.4263, "step": 7827 }, { "epoch": 0.07040423072945007, "grad_norm": 4.4151225090026855, "learning_rate": 6.176321919819472e-06, "loss": 2.4552, "step": 7828 }, { "epoch": 0.07041322462709051, "grad_norm": 3.1444058418273926, "learning_rate": 6.170882363657585e-06, "loss": 2.2337, "step": 7829 }, { "epoch": 0.07042221852473098, "grad_norm": 5.018438339233398, "learning_rate": 6.165444866619685e-06, "loss": 1.9321, "step": 7830 }, { "epoch": 0.07043121242237142, "grad_norm": 9.11778450012207, "learning_rate": 6.160009429300425e-06, "loss": 2.6245, "step": 7831 }, { "epoch": 0.07044020632001187, "grad_norm": 5.157261848449707, "learning_rate": 6.154576052294195e-06, "loss": 1.8399, "step": 7832 }, { "epoch": 0.07044920021765232, "grad_norm": 2.3831613063812256, "learning_rate": 6.149144736195209e-06, "loss": 2.2165, "step": 7833 }, { "epoch": 0.07045819411529278, "grad_norm": 4.116825103759766, "learning_rate": 6.143715481597404e-06, "loss": 2.1514, "step": 7834 }, { "epoch": 0.07046718801293322, "grad_norm": 3.1611318588256836, "learning_rate": 6.138288289094532e-06, "loss": 1.7138, "step": 7835 }, { "epoch": 0.07047618191057367, "grad_norm": 5.087094306945801, "learning_rate": 6.132863159280086e-06, "loss": 2.0054, "step": 7836 }, { "epoch": 0.07048517580821413, "grad_norm": 4.1445393562316895, "learning_rate": 6.127440092747366e-06, "loss": 2.1984, "step": 7837 }, { "epoch": 0.07049416970585458, "grad_norm": 3.171684980392456, "learning_rate": 6.1220190900894175e-06, "loss": 2.1312, "step": 7838 }, { "epoch": 0.07050316360349503, "grad_norm": 5.1762614250183105, "learning_rate": 6.116600151899079e-06, "loss": 1.7284, "step": 7839 }, { "epoch": 0.07051215750113549, "grad_norm": 8.268352508544922, "learning_rate": 6.111183278768956e-06, "loss": 2.4614, "step": 7840 }, { "epoch": 0.07052115139877593, "grad_norm": 4.588499069213867, "learning_rate": 6.105768471291437e-06, "loss": 2.0767, "step": 7841 }, { "epoch": 0.07053014529641638, "grad_norm": 3.7529349327087402, "learning_rate": 6.100355730058666e-06, "loss": 1.8685, "step": 7842 }, { "epoch": 0.07053913919405683, "grad_norm": 6.90172815322876, "learning_rate": 6.094945055662574e-06, "loss": 2.4159, "step": 7843 }, { "epoch": 0.07054813309169729, "grad_norm": 3.4554452896118164, "learning_rate": 6.089536448694877e-06, "loss": 2.1966, "step": 7844 }, { "epoch": 0.07055712698933773, "grad_norm": 2.970940589904785, "learning_rate": 6.084129909747033e-06, "loss": 2.4391, "step": 7845 }, { "epoch": 0.07056612088697818, "grad_norm": 6.126877307891846, "learning_rate": 6.078725439410304e-06, "loss": 2.3725, "step": 7846 }, { "epoch": 0.07057511478461864, "grad_norm": 5.990933895111084, "learning_rate": 6.073323038275713e-06, "loss": 2.6175, "step": 7847 }, { "epoch": 0.07058410868225909, "grad_norm": 4.942292213439941, "learning_rate": 6.067922706934068e-06, "loss": 2.3275, "step": 7848 }, { "epoch": 0.07059310257989954, "grad_norm": 8.492759704589844, "learning_rate": 6.062524445975923e-06, "loss": 2.5139, "step": 7849 }, { "epoch": 0.07060209647753998, "grad_norm": 6.4334821701049805, "learning_rate": 6.057128255991637e-06, "loss": 2.0661, "step": 7850 }, { "epoch": 0.07061109037518044, "grad_norm": 2.5175697803497314, "learning_rate": 6.051734137571322e-06, "loss": 2.6368, "step": 7851 }, { "epoch": 0.07062008427282089, "grad_norm": 9.864127159118652, "learning_rate": 6.046342091304871e-06, "loss": 2.0449, "step": 7852 }, { "epoch": 0.07062907817046134, "grad_norm": 3.0099565982818604, "learning_rate": 6.040952117781953e-06, "loss": 2.664, "step": 7853 }, { "epoch": 0.0706380720681018, "grad_norm": 3.408562421798706, "learning_rate": 6.035564217592013e-06, "loss": 2.2034, "step": 7854 }, { "epoch": 0.07064706596574225, "grad_norm": 2.9124808311462402, "learning_rate": 6.030178391324251e-06, "loss": 1.9736, "step": 7855 }, { "epoch": 0.07065605986338269, "grad_norm": 2.8910465240478516, "learning_rate": 6.024794639567669e-06, "loss": 2.4683, "step": 7856 }, { "epoch": 0.07066505376102315, "grad_norm": 8.069694519042969, "learning_rate": 6.019412962911006e-06, "loss": 2.2289, "step": 7857 }, { "epoch": 0.0706740476586636, "grad_norm": 5.5907182693481445, "learning_rate": 6.014033361942806e-06, "loss": 2.0564, "step": 7858 }, { "epoch": 0.07068304155630405, "grad_norm": 7.719456672668457, "learning_rate": 6.008655837251373e-06, "loss": 2.1852, "step": 7859 }, { "epoch": 0.0706920354539445, "grad_norm": 11.118590354919434, "learning_rate": 6.003280389424789e-06, "loss": 2.4519, "step": 7860 }, { "epoch": 0.07070102935158495, "grad_norm": 19.088008880615234, "learning_rate": 5.9979070190508965e-06, "loss": 2.1966, "step": 7861 }, { "epoch": 0.0707100232492254, "grad_norm": 3.226022243499756, "learning_rate": 5.99253572671733e-06, "loss": 2.3496, "step": 7862 }, { "epoch": 0.07071901714686585, "grad_norm": 5.797511100769043, "learning_rate": 5.987166513011469e-06, "loss": 1.649, "step": 7863 }, { "epoch": 0.07072801104450631, "grad_norm": 4.078891277313232, "learning_rate": 5.981799378520498e-06, "loss": 2.0259, "step": 7864 }, { "epoch": 0.07073700494214676, "grad_norm": 4.021083354949951, "learning_rate": 5.976434323831348e-06, "loss": 2.0142, "step": 7865 }, { "epoch": 0.0707459988397872, "grad_norm": 3.5849499702453613, "learning_rate": 5.971071349530749e-06, "loss": 2.054, "step": 7866 }, { "epoch": 0.07075499273742765, "grad_norm": 5.222200393676758, "learning_rate": 5.96571045620517e-06, "loss": 2.1512, "step": 7867 }, { "epoch": 0.07076398663506811, "grad_norm": 2.4654481410980225, "learning_rate": 5.960351644440878e-06, "loss": 1.9807, "step": 7868 }, { "epoch": 0.07077298053270856, "grad_norm": 4.459303379058838, "learning_rate": 5.954994914823911e-06, "loss": 2.5105, "step": 7869 }, { "epoch": 0.070781974430349, "grad_norm": 3.7987422943115234, "learning_rate": 5.9496402679400594e-06, "loss": 1.9396, "step": 7870 }, { "epoch": 0.07079096832798946, "grad_norm": 3.3511760234832764, "learning_rate": 5.944287704374906e-06, "loss": 2.8397, "step": 7871 }, { "epoch": 0.07079996222562991, "grad_norm": 4.5898261070251465, "learning_rate": 5.9389372247138e-06, "loss": 2.6246, "step": 7872 }, { "epoch": 0.07080895612327036, "grad_norm": 3.8631091117858887, "learning_rate": 5.933588829541869e-06, "loss": 2.0057, "step": 7873 }, { "epoch": 0.0708179500209108, "grad_norm": 8.16714859008789, "learning_rate": 5.928242519443991e-06, "loss": 2.4419, "step": 7874 }, { "epoch": 0.07082694391855127, "grad_norm": 3.204684257507324, "learning_rate": 5.9228982950048416e-06, "loss": 2.4756, "step": 7875 }, { "epoch": 0.07083593781619171, "grad_norm": 3.2388477325439453, "learning_rate": 5.917556156808848e-06, "loss": 1.8271, "step": 7876 }, { "epoch": 0.07084493171383216, "grad_norm": 14.387150764465332, "learning_rate": 5.912216105440221e-06, "loss": 2.0093, "step": 7877 }, { "epoch": 0.07085392561147262, "grad_norm": 3.3338942527770996, "learning_rate": 5.906878141482944e-06, "loss": 2.5783, "step": 7878 }, { "epoch": 0.07086291950911307, "grad_norm": 6.883129119873047, "learning_rate": 5.901542265520777e-06, "loss": 2.4708, "step": 7879 }, { "epoch": 0.07087191340675351, "grad_norm": 3.0712409019470215, "learning_rate": 5.896208478137222e-06, "loss": 2.7011, "step": 7880 }, { "epoch": 0.07088090730439398, "grad_norm": 1.4858578443527222, "learning_rate": 5.890876779915594e-06, "loss": 1.9425, "step": 7881 }, { "epoch": 0.07088990120203442, "grad_norm": 1.8580116033554077, "learning_rate": 5.885547171438946e-06, "loss": 2.0824, "step": 7882 }, { "epoch": 0.07089889509967487, "grad_norm": 5.117650985717773, "learning_rate": 5.88021965329012e-06, "loss": 1.9549, "step": 7883 }, { "epoch": 0.07090788899731532, "grad_norm": 4.173318862915039, "learning_rate": 5.874894226051728e-06, "loss": 1.9819, "step": 7884 }, { "epoch": 0.07091688289495578, "grad_norm": 3.8730762004852295, "learning_rate": 5.869570890306153e-06, "loss": 2.3138, "step": 7885 }, { "epoch": 0.07092587679259622, "grad_norm": 1.8385857343673706, "learning_rate": 5.864249646635539e-06, "loss": 2.5711, "step": 7886 }, { "epoch": 0.07093487069023667, "grad_norm": 6.498500823974609, "learning_rate": 5.85893049562182e-06, "loss": 1.9375, "step": 7887 }, { "epoch": 0.07094386458787713, "grad_norm": 10.278800964355469, "learning_rate": 5.853613437846675e-06, "loss": 1.9646, "step": 7888 }, { "epoch": 0.07095285848551758, "grad_norm": 5.684635162353516, "learning_rate": 5.848298473891581e-06, "loss": 1.8286, "step": 7889 }, { "epoch": 0.07096185238315803, "grad_norm": 2.4375417232513428, "learning_rate": 5.842985604337769e-06, "loss": 1.8412, "step": 7890 }, { "epoch": 0.07097084628079847, "grad_norm": 4.323903560638428, "learning_rate": 5.837674829766257e-06, "loss": 1.8259, "step": 7891 }, { "epoch": 0.07097984017843893, "grad_norm": 10.09202766418457, "learning_rate": 5.8323661507578095e-06, "loss": 2.4003, "step": 7892 }, { "epoch": 0.07098883407607938, "grad_norm": 6.773062705993652, "learning_rate": 5.827059567892981e-06, "loss": 2.3921, "step": 7893 }, { "epoch": 0.07099782797371983, "grad_norm": 5.097596168518066, "learning_rate": 5.821755081752101e-06, "loss": 2.3176, "step": 7894 }, { "epoch": 0.07100682187136029, "grad_norm": 3.5452017784118652, "learning_rate": 5.816452692915242e-06, "loss": 2.3119, "step": 7895 }, { "epoch": 0.07101581576900073, "grad_norm": 6.7246856689453125, "learning_rate": 5.811152401962275e-06, "loss": 1.9282, "step": 7896 }, { "epoch": 0.07102480966664118, "grad_norm": 2.9834365844726562, "learning_rate": 5.80585420947283e-06, "loss": 2.1943, "step": 7897 }, { "epoch": 0.07103380356428164, "grad_norm": 3.918860912322998, "learning_rate": 5.800558116026319e-06, "loss": 2.3655, "step": 7898 }, { "epoch": 0.07104279746192209, "grad_norm": 2.541978359222412, "learning_rate": 5.795264122201902e-06, "loss": 1.7779, "step": 7899 }, { "epoch": 0.07105179135956254, "grad_norm": 5.978222846984863, "learning_rate": 5.78997222857853e-06, "loss": 2.1838, "step": 7900 }, { "epoch": 0.07106078525720298, "grad_norm": 2.6604435443878174, "learning_rate": 5.784682435734906e-06, "loss": 1.7864, "step": 7901 }, { "epoch": 0.07106977915484344, "grad_norm": 3.38655948638916, "learning_rate": 5.779394744249522e-06, "loss": 2.4695, "step": 7902 }, { "epoch": 0.07107877305248389, "grad_norm": 3.3927507400512695, "learning_rate": 5.774109154700633e-06, "loss": 2.2107, "step": 7903 }, { "epoch": 0.07108776695012434, "grad_norm": 5.7702765464782715, "learning_rate": 5.7688256676662666e-06, "loss": 1.7455, "step": 7904 }, { "epoch": 0.0710967608477648, "grad_norm": 3.867229700088501, "learning_rate": 5.763544283724204e-06, "loss": 2.2319, "step": 7905 }, { "epoch": 0.07110575474540524, "grad_norm": 3.1569161415100098, "learning_rate": 5.7582650034520265e-06, "loss": 2.3699, "step": 7906 }, { "epoch": 0.07111474864304569, "grad_norm": 3.710508346557617, "learning_rate": 5.7529878274270476e-06, "loss": 2.2114, "step": 7907 }, { "epoch": 0.07112374254068614, "grad_norm": 2.86437726020813, "learning_rate": 5.747712756226393e-06, "loss": 2.0379, "step": 7908 }, { "epoch": 0.0711327364383266, "grad_norm": 6.359321594238281, "learning_rate": 5.74243979042691e-06, "loss": 2.2202, "step": 7909 }, { "epoch": 0.07114173033596705, "grad_norm": 6.83196496963501, "learning_rate": 5.737168930605272e-06, "loss": 2.0952, "step": 7910 }, { "epoch": 0.07115072423360749, "grad_norm": 3.7750933170318604, "learning_rate": 5.731900177337873e-06, "loss": 1.9969, "step": 7911 }, { "epoch": 0.07115971813124795, "grad_norm": 2.611738443374634, "learning_rate": 5.726633531200906e-06, "loss": 2.1873, "step": 7912 }, { "epoch": 0.0711687120288884, "grad_norm": 6.1405181884765625, "learning_rate": 5.721368992770315e-06, "loss": 2.3778, "step": 7913 }, { "epoch": 0.07117770592652885, "grad_norm": 5.189527988433838, "learning_rate": 5.716106562621826e-06, "loss": 2.3326, "step": 7914 }, { "epoch": 0.0711866998241693, "grad_norm": 3.379000663757324, "learning_rate": 5.710846241330928e-06, "loss": 1.9504, "step": 7915 }, { "epoch": 0.07119569372180976, "grad_norm": 5.868525981903076, "learning_rate": 5.7055880294728855e-06, "loss": 2.0918, "step": 7916 }, { "epoch": 0.0712046876194502, "grad_norm": 9.243414878845215, "learning_rate": 5.700331927622735e-06, "loss": 1.0905, "step": 7917 }, { "epoch": 0.07121368151709065, "grad_norm": 2.8266212940216064, "learning_rate": 5.695077936355259e-06, "loss": 2.3472, "step": 7918 }, { "epoch": 0.07122267541473111, "grad_norm": 2.7544355392456055, "learning_rate": 5.689826056245043e-06, "loss": 2.2463, "step": 7919 }, { "epoch": 0.07123166931237156, "grad_norm": 6.859325408935547, "learning_rate": 5.684576287866411e-06, "loss": 2.581, "step": 7920 }, { "epoch": 0.071240663210012, "grad_norm": 8.37999153137207, "learning_rate": 5.6793286317934846e-06, "loss": 2.0755, "step": 7921 }, { "epoch": 0.07124965710765246, "grad_norm": 2.9898288249969482, "learning_rate": 5.674083088600113e-06, "loss": 2.4264, "step": 7922 }, { "epoch": 0.07125865100529291, "grad_norm": 4.534254550933838, "learning_rate": 5.668839658859976e-06, "loss": 2.1724, "step": 7923 }, { "epoch": 0.07126764490293336, "grad_norm": 8.61820125579834, "learning_rate": 5.663598343146462e-06, "loss": 2.0911, "step": 7924 }, { "epoch": 0.0712766388005738, "grad_norm": 4.075699806213379, "learning_rate": 5.6583591420327684e-06, "loss": 1.9776, "step": 7925 }, { "epoch": 0.07128563269821427, "grad_norm": 3.886843204498291, "learning_rate": 5.653122056091831e-06, "loss": 2.0306, "step": 7926 }, { "epoch": 0.07129462659585471, "grad_norm": 3.6147677898406982, "learning_rate": 5.647887085896386e-06, "loss": 1.993, "step": 7927 }, { "epoch": 0.07130362049349516, "grad_norm": 6.933601379394531, "learning_rate": 5.642654232018898e-06, "loss": 2.3, "step": 7928 }, { "epoch": 0.07131261439113562, "grad_norm": 4.180204391479492, "learning_rate": 5.637423495031658e-06, "loss": 2.3109, "step": 7929 }, { "epoch": 0.07132160828877607, "grad_norm": 3.6720035076141357, "learning_rate": 5.632194875506663e-06, "loss": 2.3856, "step": 7930 }, { "epoch": 0.07133060218641651, "grad_norm": 2.0873215198516846, "learning_rate": 5.6269683740157255e-06, "loss": 2.0543, "step": 7931 }, { "epoch": 0.07133959608405696, "grad_norm": 2.2216556072235107, "learning_rate": 5.621743991130393e-06, "loss": 2.3848, "step": 7932 }, { "epoch": 0.07134858998169742, "grad_norm": 3.662095546722412, "learning_rate": 5.616521727422008e-06, "loss": 1.903, "step": 7933 }, { "epoch": 0.07135758387933787, "grad_norm": 4.86671257019043, "learning_rate": 5.61130158346165e-06, "loss": 2.6497, "step": 7934 }, { "epoch": 0.07136657777697832, "grad_norm": 9.221704483032227, "learning_rate": 5.606083559820219e-06, "loss": 1.9543, "step": 7935 }, { "epoch": 0.07137557167461878, "grad_norm": 6.754184246063232, "learning_rate": 5.600867657068321e-06, "loss": 2.3805, "step": 7936 }, { "epoch": 0.07138456557225922, "grad_norm": 4.811770439147949, "learning_rate": 5.5956538757763685e-06, "loss": 1.9584, "step": 7937 }, { "epoch": 0.07139355946989967, "grad_norm": 3.4241325855255127, "learning_rate": 5.590442216514543e-06, "loss": 2.5299, "step": 7938 }, { "epoch": 0.07140255336754013, "grad_norm": 9.227836608886719, "learning_rate": 5.58523267985277e-06, "loss": 2.1126, "step": 7939 }, { "epoch": 0.07141154726518058, "grad_norm": 4.6254777908325195, "learning_rate": 5.580025266360764e-06, "loss": 1.7525, "step": 7940 }, { "epoch": 0.07142054116282102, "grad_norm": 4.02954626083374, "learning_rate": 5.574819976607987e-06, "loss": 2.2512, "step": 7941 }, { "epoch": 0.07142953506046147, "grad_norm": 7.417486190795898, "learning_rate": 5.569616811163705e-06, "loss": 2.6446, "step": 7942 }, { "epoch": 0.07143852895810193, "grad_norm": 3.636009454727173, "learning_rate": 5.564415770596909e-06, "loss": 1.7949, "step": 7943 }, { "epoch": 0.07144752285574238, "grad_norm": 5.168453216552734, "learning_rate": 5.559216855476387e-06, "loss": 2.0116, "step": 7944 }, { "epoch": 0.07145651675338283, "grad_norm": 4.284126281738281, "learning_rate": 5.554020066370677e-06, "loss": 2.3151, "step": 7945 }, { "epoch": 0.07146551065102329, "grad_norm": 2.680699586868286, "learning_rate": 5.5488254038481045e-06, "loss": 2.2861, "step": 7946 }, { "epoch": 0.07147450454866373, "grad_norm": 6.850200176239014, "learning_rate": 5.543632868476723e-06, "loss": 2.3307, "step": 7947 }, { "epoch": 0.07148349844630418, "grad_norm": 4.2702412605285645, "learning_rate": 5.538442460824417e-06, "loss": 2.5598, "step": 7948 }, { "epoch": 0.07149249234394463, "grad_norm": 2.42665696144104, "learning_rate": 5.533254181458775e-06, "loss": 1.9125, "step": 7949 }, { "epoch": 0.07150148624158509, "grad_norm": 6.445870876312256, "learning_rate": 5.528068030947192e-06, "loss": 2.1513, "step": 7950 }, { "epoch": 0.07151048013922554, "grad_norm": 5.29352331161499, "learning_rate": 5.52288400985681e-06, "loss": 2.1835, "step": 7951 }, { "epoch": 0.07151947403686598, "grad_norm": 8.301692962646484, "learning_rate": 5.5177021187545556e-06, "loss": 2.1316, "step": 7952 }, { "epoch": 0.07152846793450644, "grad_norm": 2.825758457183838, "learning_rate": 5.512522358207092e-06, "loss": 1.8747, "step": 7953 }, { "epoch": 0.07153746183214689, "grad_norm": 5.4273200035095215, "learning_rate": 5.507344728780897e-06, "loss": 1.4322, "step": 7954 }, { "epoch": 0.07154645572978734, "grad_norm": 5.6128387451171875, "learning_rate": 5.50216923104217e-06, "loss": 2.1514, "step": 7955 }, { "epoch": 0.07155544962742778, "grad_norm": 5.500871658325195, "learning_rate": 5.496995865556909e-06, "loss": 2.7166, "step": 7956 }, { "epoch": 0.07156444352506824, "grad_norm": 3.2146949768066406, "learning_rate": 5.491824632890849e-06, "loss": 1.8287, "step": 7957 }, { "epoch": 0.07157343742270869, "grad_norm": 9.675315856933594, "learning_rate": 5.48665553360952e-06, "loss": 2.381, "step": 7958 }, { "epoch": 0.07158243132034914, "grad_norm": 4.456941604614258, "learning_rate": 5.481488568278209e-06, "loss": 2.1226, "step": 7959 }, { "epoch": 0.0715914252179896, "grad_norm": 2.9066452980041504, "learning_rate": 5.476323737461955e-06, "loss": 1.9359, "step": 7960 }, { "epoch": 0.07160041911563005, "grad_norm": 6.815676212310791, "learning_rate": 5.4711610417255835e-06, "loss": 1.8911, "step": 7961 }, { "epoch": 0.07160941301327049, "grad_norm": 2.325045585632324, "learning_rate": 5.466000481633682e-06, "loss": 2.4348, "step": 7962 }, { "epoch": 0.07161840691091095, "grad_norm": 4.212078094482422, "learning_rate": 5.460842057750604e-06, "loss": 2.1068, "step": 7963 }, { "epoch": 0.0716274008085514, "grad_norm": 1.9486802816390991, "learning_rate": 5.4556857706404534e-06, "loss": 1.9367, "step": 7964 }, { "epoch": 0.07163639470619185, "grad_norm": 5.3402204513549805, "learning_rate": 5.45053162086713e-06, "loss": 2.321, "step": 7965 }, { "epoch": 0.0716453886038323, "grad_norm": 2.6646900177001953, "learning_rate": 5.4453796089942635e-06, "loss": 2.0685, "step": 7966 }, { "epoch": 0.07165438250147275, "grad_norm": 7.375187873840332, "learning_rate": 5.440229735585298e-06, "loss": 2.3783, "step": 7967 }, { "epoch": 0.0716633763991132, "grad_norm": 4.278142929077148, "learning_rate": 5.435082001203392e-06, "loss": 2.355, "step": 7968 }, { "epoch": 0.07167237029675365, "grad_norm": 2.7899699211120605, "learning_rate": 5.429936406411512e-06, "loss": 1.9388, "step": 7969 }, { "epoch": 0.07168136419439411, "grad_norm": 2.8030223846435547, "learning_rate": 5.424792951772353e-06, "loss": 2.146, "step": 7970 }, { "epoch": 0.07169035809203456, "grad_norm": 8.988554954528809, "learning_rate": 5.419651637848417e-06, "loss": 2.0158, "step": 7971 }, { "epoch": 0.071699351989675, "grad_norm": 4.287391662597656, "learning_rate": 5.4145124652019295e-06, "loss": 1.7211, "step": 7972 }, { "epoch": 0.07170834588731545, "grad_norm": 6.1192216873168945, "learning_rate": 5.4093754343949146e-06, "loss": 2.377, "step": 7973 }, { "epoch": 0.07171733978495591, "grad_norm": 4.105865478515625, "learning_rate": 5.404240545989145e-06, "loss": 2.2868, "step": 7974 }, { "epoch": 0.07172633368259636, "grad_norm": 3.481978178024292, "learning_rate": 5.399107800546177e-06, "loss": 2.2522, "step": 7975 }, { "epoch": 0.0717353275802368, "grad_norm": 7.118753433227539, "learning_rate": 5.393977198627303e-06, "loss": 1.8247, "step": 7976 }, { "epoch": 0.07174432147787727, "grad_norm": 7.5583977699279785, "learning_rate": 5.388848740793612e-06, "loss": 2.2478, "step": 7977 }, { "epoch": 0.07175331537551771, "grad_norm": 5.309566497802734, "learning_rate": 5.383722427605931e-06, "loss": 2.1176, "step": 7978 }, { "epoch": 0.07176230927315816, "grad_norm": 3.382395029067993, "learning_rate": 5.378598259624873e-06, "loss": 2.0517, "step": 7979 }, { "epoch": 0.07177130317079862, "grad_norm": 7.700477600097656, "learning_rate": 5.373476237410807e-06, "loss": 1.932, "step": 7980 }, { "epoch": 0.07178029706843907, "grad_norm": 3.424539566040039, "learning_rate": 5.368356361523879e-06, "loss": 2.3306, "step": 7981 }, { "epoch": 0.07178929096607951, "grad_norm": 3.10952091217041, "learning_rate": 5.363238632523979e-06, "loss": 2.5268, "step": 7982 }, { "epoch": 0.07179828486371996, "grad_norm": 4.885530948638916, "learning_rate": 5.3581230509707785e-06, "loss": 2.046, "step": 7983 }, { "epoch": 0.07180727876136042, "grad_norm": 3.435837745666504, "learning_rate": 5.353009617423715e-06, "loss": 2.2824, "step": 7984 }, { "epoch": 0.07181627265900087, "grad_norm": 12.303170204162598, "learning_rate": 5.347898332441975e-06, "loss": 2.3496, "step": 7985 }, { "epoch": 0.07182526655664132, "grad_norm": 6.923490524291992, "learning_rate": 5.342789196584527e-06, "loss": 1.7097, "step": 7986 }, { "epoch": 0.07183426045428178, "grad_norm": 3.115241050720215, "learning_rate": 5.3376822104101e-06, "loss": 2.281, "step": 7987 }, { "epoch": 0.07184325435192222, "grad_norm": 4.904194355010986, "learning_rate": 5.332577374477188e-06, "loss": 2.2815, "step": 7988 }, { "epoch": 0.07185224824956267, "grad_norm": 2.011003255844116, "learning_rate": 5.327474689344039e-06, "loss": 2.2372, "step": 7989 }, { "epoch": 0.07186124214720312, "grad_norm": 3.800586223602295, "learning_rate": 5.322374155568688e-06, "loss": 2.2093, "step": 7990 }, { "epoch": 0.07187023604484358, "grad_norm": 5.2557902336120605, "learning_rate": 5.317275773708907e-06, "loss": 2.0129, "step": 7991 }, { "epoch": 0.07187922994248402, "grad_norm": 3.9482951164245605, "learning_rate": 5.312179544322257e-06, "loss": 1.5262, "step": 7992 }, { "epoch": 0.07188822384012447, "grad_norm": 6.176761627197266, "learning_rate": 5.30708546796605e-06, "loss": 2.4913, "step": 7993 }, { "epoch": 0.07189721773776493, "grad_norm": 3.250336170196533, "learning_rate": 5.301993545197376e-06, "loss": 2.3241, "step": 7994 }, { "epoch": 0.07190621163540538, "grad_norm": 4.726370811462402, "learning_rate": 5.296903776573065e-06, "loss": 2.2769, "step": 7995 }, { "epoch": 0.07191520553304583, "grad_norm": 3.808352470397949, "learning_rate": 5.29181616264974e-06, "loss": 2.4135, "step": 7996 }, { "epoch": 0.07192419943068627, "grad_norm": 3.0233800411224365, "learning_rate": 5.286730703983764e-06, "loss": 1.8093, "step": 7997 }, { "epoch": 0.07193319332832673, "grad_norm": 3.3666021823883057, "learning_rate": 5.281647401131279e-06, "loss": 2.0759, "step": 7998 }, { "epoch": 0.07194218722596718, "grad_norm": 3.5234620571136475, "learning_rate": 5.276566254648188e-06, "loss": 1.9827, "step": 7999 }, { "epoch": 0.07195118112360763, "grad_norm": 4.674970626831055, "learning_rate": 5.271487265090163e-06, "loss": 2.0278, "step": 8000 }, { "epoch": 0.07196017502124809, "grad_norm": 0.7578118443489075, "learning_rate": 5.266410433012623e-06, "loss": 0.8358, "step": 8001 }, { "epoch": 0.07196916891888853, "grad_norm": 0.48037630319595337, "learning_rate": 5.261335758970779e-06, "loss": 0.7895, "step": 8002 }, { "epoch": 0.07197816281652898, "grad_norm": 0.5999154448509216, "learning_rate": 5.256263243519569e-06, "loss": 0.7277, "step": 8003 }, { "epoch": 0.07198715671416944, "grad_norm": 0.26563411951065063, "learning_rate": 5.251192887213727e-06, "loss": 1.0834, "step": 8004 }, { "epoch": 0.07199615061180989, "grad_norm": 0.4117529094219208, "learning_rate": 5.24612469060774e-06, "loss": 0.8679, "step": 8005 }, { "epoch": 0.07200514450945034, "grad_norm": 0.4567953646183014, "learning_rate": 5.241058654255865e-06, "loss": 0.7384, "step": 8006 }, { "epoch": 0.07201413840709078, "grad_norm": 0.6682908535003662, "learning_rate": 5.2359947787121e-06, "loss": 0.6695, "step": 8007 }, { "epoch": 0.07202313230473124, "grad_norm": 0.7225403785705566, "learning_rate": 5.23093306453023e-06, "loss": 0.4933, "step": 8008 }, { "epoch": 0.07203212620237169, "grad_norm": 0.6400661468505859, "learning_rate": 5.2258735122638066e-06, "loss": 0.642, "step": 8009 }, { "epoch": 0.07204112010001214, "grad_norm": 0.8410163521766663, "learning_rate": 5.220816122466119e-06, "loss": 0.5636, "step": 8010 }, { "epoch": 0.0720501139976526, "grad_norm": 1.0886105298995972, "learning_rate": 5.215760895690244e-06, "loss": 0.7051, "step": 8011 }, { "epoch": 0.07205910789529305, "grad_norm": 2.51412034034729, "learning_rate": 5.2107078324890095e-06, "loss": 0.9282, "step": 8012 }, { "epoch": 0.07206810179293349, "grad_norm": 0.7449488639831543, "learning_rate": 5.205656933415021e-06, "loss": 0.51, "step": 8013 }, { "epoch": 0.07207709569057394, "grad_norm": 0.8438796997070312, "learning_rate": 5.200608199020623e-06, "loss": 0.081, "step": 8014 }, { "epoch": 0.0720860895882144, "grad_norm": 0.4469670355319977, "learning_rate": 5.195561629857953e-06, "loss": 0.8379, "step": 8015 }, { "epoch": 0.07209508348585485, "grad_norm": 0.2628368139266968, "learning_rate": 5.19051722647888e-06, "loss": 1.1953, "step": 8016 }, { "epoch": 0.0721040773834953, "grad_norm": 0.7715193033218384, "learning_rate": 5.18547498943506e-06, "loss": 0.3222, "step": 8017 }, { "epoch": 0.07211307128113575, "grad_norm": 0.7517063617706299, "learning_rate": 5.1804349192779044e-06, "loss": 0.5303, "step": 8018 }, { "epoch": 0.0721220651787762, "grad_norm": 0.5178043842315674, "learning_rate": 5.175397016558594e-06, "loss": 0.9911, "step": 8019 }, { "epoch": 0.07213105907641665, "grad_norm": 0.5068122744560242, "learning_rate": 5.170361281828054e-06, "loss": 0.7121, "step": 8020 }, { "epoch": 0.07214005297405711, "grad_norm": 0.60454261302948, "learning_rate": 5.165327715636997e-06, "loss": 0.4646, "step": 8021 }, { "epoch": 0.07214904687169756, "grad_norm": 0.5113588571548462, "learning_rate": 5.160296318535876e-06, "loss": 0.898, "step": 8022 }, { "epoch": 0.072158040769338, "grad_norm": 0.44000378251075745, "learning_rate": 5.155267091074916e-06, "loss": 0.9495, "step": 8023 }, { "epoch": 0.07216703466697845, "grad_norm": 0.6372286081314087, "learning_rate": 5.150240033804116e-06, "loss": 0.4413, "step": 8024 }, { "epoch": 0.07217602856461891, "grad_norm": 0.4463580250740051, "learning_rate": 5.145215147273224e-06, "loss": 1.1148, "step": 8025 }, { "epoch": 0.07218502246225936, "grad_norm": 0.6574262976646423, "learning_rate": 5.1401924320317485e-06, "loss": 0.3039, "step": 8026 }, { "epoch": 0.0721940163598998, "grad_norm": 0.40793102979660034, "learning_rate": 5.135171888628975e-06, "loss": 0.6044, "step": 8027 }, { "epoch": 0.07220301025754026, "grad_norm": 0.3608049154281616, "learning_rate": 5.130153517613934e-06, "loss": 0.9756, "step": 8028 }, { "epoch": 0.07221200415518071, "grad_norm": 0.4757501482963562, "learning_rate": 5.125137319535428e-06, "loss": 0.8955, "step": 8029 }, { "epoch": 0.07222099805282116, "grad_norm": 0.3763056993484497, "learning_rate": 5.120123294942022e-06, "loss": 0.8748, "step": 8030 }, { "epoch": 0.0722299919504616, "grad_norm": 0.4395611882209778, "learning_rate": 5.115111444382053e-06, "loss": 0.4484, "step": 8031 }, { "epoch": 0.07223898584810207, "grad_norm": 0.5055692791938782, "learning_rate": 5.110101768403591e-06, "loss": 0.7636, "step": 8032 }, { "epoch": 0.07224797974574251, "grad_norm": 1.6202367544174194, "learning_rate": 5.1050942675544975e-06, "loss": 0.5972, "step": 8033 }, { "epoch": 0.07225697364338296, "grad_norm": 0.8641490936279297, "learning_rate": 5.100088942382389e-06, "loss": 0.3885, "step": 8034 }, { "epoch": 0.07226596754102342, "grad_norm": 0.5224502086639404, "learning_rate": 5.095085793434629e-06, "loss": 0.5565, "step": 8035 }, { "epoch": 0.07227496143866387, "grad_norm": 0.24345608055591583, "learning_rate": 5.090084821258359e-06, "loss": 1.1603, "step": 8036 }, { "epoch": 0.07228395533630431, "grad_norm": 1.341442346572876, "learning_rate": 5.0850860264004795e-06, "loss": 0.6152, "step": 8037 }, { "epoch": 0.07229294923394476, "grad_norm": 0.342251181602478, "learning_rate": 5.0800894094076565e-06, "loss": 0.6781, "step": 8038 }, { "epoch": 0.07230194313158522, "grad_norm": 0.7088210582733154, "learning_rate": 5.075094970826299e-06, "loss": 0.6274, "step": 8039 }, { "epoch": 0.07231093702922567, "grad_norm": 0.8321745991706848, "learning_rate": 5.070102711202607e-06, "loss": 0.5518, "step": 8040 }, { "epoch": 0.07231993092686612, "grad_norm": 0.39217647910118103, "learning_rate": 5.065112631082511e-06, "loss": 0.3335, "step": 8041 }, { "epoch": 0.07232892482450658, "grad_norm": 0.33581364154815674, "learning_rate": 5.060124731011734e-06, "loss": 1.2974, "step": 8042 }, { "epoch": 0.07233791872214702, "grad_norm": 0.43919017910957336, "learning_rate": 5.055139011535723e-06, "loss": 0.8635, "step": 8043 }, { "epoch": 0.07234691261978747, "grad_norm": 0.5097888112068176, "learning_rate": 5.0501554731997356e-06, "loss": 0.9373, "step": 8044 }, { "epoch": 0.07235590651742793, "grad_norm": 0.4147583544254303, "learning_rate": 5.045174116548745e-06, "loss": 0.76, "step": 8045 }, { "epoch": 0.07236490041506838, "grad_norm": 0.2127276211977005, "learning_rate": 5.040194942127518e-06, "loss": 1.0457, "step": 8046 }, { "epoch": 0.07237389431270883, "grad_norm": 0.3788119852542877, "learning_rate": 5.035217950480559e-06, "loss": 0.7385, "step": 8047 }, { "epoch": 0.07238288821034927, "grad_norm": 0.4513744115829468, "learning_rate": 5.030243142152152e-06, "loss": 0.4449, "step": 8048 }, { "epoch": 0.07239188210798973, "grad_norm": 0.5021140575408936, "learning_rate": 5.025270517686317e-06, "loss": 0.2056, "step": 8049 }, { "epoch": 0.07240087600563018, "grad_norm": 0.4376832842826843, "learning_rate": 5.0203000776268825e-06, "loss": 0.8977, "step": 8050 }, { "epoch": 0.07240986990327063, "grad_norm": 0.882815420627594, "learning_rate": 5.015331822517386e-06, "loss": 0.311, "step": 8051 }, { "epoch": 0.07241886380091109, "grad_norm": 0.42897096276283264, "learning_rate": 5.010365752901158e-06, "loss": 0.5164, "step": 8052 }, { "epoch": 0.07242785769855153, "grad_norm": 0.59879070520401, "learning_rate": 5.005401869321275e-06, "loss": 0.2559, "step": 8053 }, { "epoch": 0.07243685159619198, "grad_norm": 0.654643714427948, "learning_rate": 5.000440172320581e-06, "loss": 0.5275, "step": 8054 }, { "epoch": 0.07244584549383243, "grad_norm": 0.471818208694458, "learning_rate": 4.995480662441679e-06, "loss": 0.0794, "step": 8055 }, { "epoch": 0.07245483939147289, "grad_norm": 0.38840627670288086, "learning_rate": 4.990523340226943e-06, "loss": 0.6978, "step": 8056 }, { "epoch": 0.07246383328911334, "grad_norm": 2.1357851028442383, "learning_rate": 4.985568206218483e-06, "loss": 0.4655, "step": 8057 }, { "epoch": 0.07247282718675378, "grad_norm": 0.822580099105835, "learning_rate": 4.980615260958193e-06, "loss": 0.6017, "step": 8058 }, { "epoch": 0.07248182108439424, "grad_norm": 0.30140575766563416, "learning_rate": 4.975664504987729e-06, "loss": 1.1205, "step": 8059 }, { "epoch": 0.07249081498203469, "grad_norm": 0.5141436457633972, "learning_rate": 4.970715938848478e-06, "loss": 0.6202, "step": 8060 }, { "epoch": 0.07249980887967514, "grad_norm": 0.728233814239502, "learning_rate": 4.965769563081627e-06, "loss": 0.6408, "step": 8061 }, { "epoch": 0.0725088027773156, "grad_norm": 0.4174199402332306, "learning_rate": 4.960825378228082e-06, "loss": 1.178, "step": 8062 }, { "epoch": 0.07251779667495604, "grad_norm": 0.4374246597290039, "learning_rate": 4.955883384828558e-06, "loss": 0.7174, "step": 8063 }, { "epoch": 0.07252679057259649, "grad_norm": 1.0038666725158691, "learning_rate": 4.9509435834234865e-06, "loss": 0.4288, "step": 8064 }, { "epoch": 0.07253578447023694, "grad_norm": 0.42305177450180054, "learning_rate": 4.946005974553086e-06, "loss": 0.5876, "step": 8065 }, { "epoch": 0.0725447783678774, "grad_norm": 0.467666357755661, "learning_rate": 4.941070558757316e-06, "loss": 0.961, "step": 8066 }, { "epoch": 0.07255377226551785, "grad_norm": 0.5998733639717102, "learning_rate": 4.936137336575916e-06, "loss": 0.6538, "step": 8067 }, { "epoch": 0.0725627661631583, "grad_norm": 0.4996836185455322, "learning_rate": 4.931206308548361e-06, "loss": 0.6023, "step": 8068 }, { "epoch": 0.07257176006079875, "grad_norm": 1.9325048923492432, "learning_rate": 4.926277475213923e-06, "loss": 0.6259, "step": 8069 }, { "epoch": 0.0725807539584392, "grad_norm": 0.45471078157424927, "learning_rate": 4.9213508371115935e-06, "loss": 0.9514, "step": 8070 }, { "epoch": 0.07258974785607965, "grad_norm": 1.3642030954360962, "learning_rate": 4.9164263947801545e-06, "loss": 0.4573, "step": 8071 }, { "epoch": 0.0725987417537201, "grad_norm": 0.37249594926834106, "learning_rate": 4.911504148758123e-06, "loss": 0.7658, "step": 8072 }, { "epoch": 0.07260773565136056, "grad_norm": 1.0996147394180298, "learning_rate": 4.906584099583803e-06, "loss": 0.7394, "step": 8073 }, { "epoch": 0.072616729549001, "grad_norm": 0.44682812690734863, "learning_rate": 4.9016662477952226e-06, "loss": 0.5651, "step": 8074 }, { "epoch": 0.07262572344664145, "grad_norm": 0.7324913740158081, "learning_rate": 4.896750593930216e-06, "loss": 0.9363, "step": 8075 }, { "epoch": 0.07263471734428191, "grad_norm": 0.41928598284721375, "learning_rate": 4.891837138526334e-06, "loss": 0.4664, "step": 8076 }, { "epoch": 0.07264371124192236, "grad_norm": 0.619937539100647, "learning_rate": 4.886925882120916e-06, "loss": 1.1172, "step": 8077 }, { "epoch": 0.0726527051395628, "grad_norm": 0.45924994349479675, "learning_rate": 4.882016825251037e-06, "loss": 1.0203, "step": 8078 }, { "epoch": 0.07266169903720326, "grad_norm": 0.5529623031616211, "learning_rate": 4.877109968453552e-06, "loss": 0.7886, "step": 8079 }, { "epoch": 0.07267069293484371, "grad_norm": 0.25325295329093933, "learning_rate": 4.872205312265074e-06, "loss": 1.0798, "step": 8080 }, { "epoch": 0.07267968683248416, "grad_norm": 0.5740371346473694, "learning_rate": 4.867302857221953e-06, "loss": 0.9818, "step": 8081 }, { "epoch": 0.0726886807301246, "grad_norm": 0.3090488314628601, "learning_rate": 4.862402603860325e-06, "loss": 0.7827, "step": 8082 }, { "epoch": 0.07269767462776507, "grad_norm": 0.42359212040901184, "learning_rate": 4.857504552716069e-06, "loss": 0.6503, "step": 8083 }, { "epoch": 0.07270666852540551, "grad_norm": 0.7330541014671326, "learning_rate": 4.852608704324841e-06, "loss": 1.0363, "step": 8084 }, { "epoch": 0.07271566242304596, "grad_norm": 0.5526798963546753, "learning_rate": 4.847715059222024e-06, "loss": 0.5235, "step": 8085 }, { "epoch": 0.07272465632068642, "grad_norm": 0.7018617987632751, "learning_rate": 4.8428236179427985e-06, "loss": 0.3822, "step": 8086 }, { "epoch": 0.07273365021832687, "grad_norm": 0.6731918454170227, "learning_rate": 4.837934381022061e-06, "loss": 0.33, "step": 8087 }, { "epoch": 0.07274264411596731, "grad_norm": 0.7440508008003235, "learning_rate": 4.833047348994521e-06, "loss": 0.6653, "step": 8088 }, { "epoch": 0.07275163801360776, "grad_norm": 0.4571261703968048, "learning_rate": 4.828162522394597e-06, "loss": 0.7222, "step": 8089 }, { "epoch": 0.07276063191124822, "grad_norm": 0.8710260391235352, "learning_rate": 4.823279901756497e-06, "loss": 0.9199, "step": 8090 }, { "epoch": 0.07276962580888867, "grad_norm": 0.35384687781333923, "learning_rate": 4.818399487614164e-06, "loss": 0.9422, "step": 8091 }, { "epoch": 0.07277861970652912, "grad_norm": 0.417900413274765, "learning_rate": 4.813521280501329e-06, "loss": 0.5372, "step": 8092 }, { "epoch": 0.07278761360416958, "grad_norm": 0.5112399458885193, "learning_rate": 4.808645280951441e-06, "loss": 0.4352, "step": 8093 }, { "epoch": 0.07279660750181002, "grad_norm": 1.1385201215744019, "learning_rate": 4.803771489497763e-06, "loss": 0.091, "step": 8094 }, { "epoch": 0.07280560139945047, "grad_norm": 0.6715893745422363, "learning_rate": 4.798899906673263e-06, "loss": 0.3161, "step": 8095 }, { "epoch": 0.07281459529709092, "grad_norm": 0.4770410358905792, "learning_rate": 4.794030533010705e-06, "loss": 0.4797, "step": 8096 }, { "epoch": 0.07282358919473138, "grad_norm": 1.0639759302139282, "learning_rate": 4.789163369042582e-06, "loss": 0.482, "step": 8097 }, { "epoch": 0.07283258309237182, "grad_norm": 0.6519335508346558, "learning_rate": 4.784298415301172e-06, "loss": 0.6525, "step": 8098 }, { "epoch": 0.07284157699001227, "grad_norm": 0.3572263717651367, "learning_rate": 4.779435672318485e-06, "loss": 0.5502, "step": 8099 }, { "epoch": 0.07285057088765273, "grad_norm": 0.29478368163108826, "learning_rate": 4.7745751406263165e-06, "loss": 1.0247, "step": 8100 }, { "epoch": 0.07285956478529318, "grad_norm": 0.47979333996772766, "learning_rate": 4.769716820756198e-06, "loss": 1.0105, "step": 8101 }, { "epoch": 0.07286855868293363, "grad_norm": 0.5612058639526367, "learning_rate": 4.764860713239441e-06, "loss": 0.5636, "step": 8102 }, { "epoch": 0.07287755258057409, "grad_norm": 0.49479371309280396, "learning_rate": 4.760006818607088e-06, "loss": 0.7568, "step": 8103 }, { "epoch": 0.07288654647821453, "grad_norm": 0.4802871644496918, "learning_rate": 4.755155137389955e-06, "loss": 0.6724, "step": 8104 }, { "epoch": 0.07289554037585498, "grad_norm": 0.7169194221496582, "learning_rate": 4.750305670118629e-06, "loss": 0.4928, "step": 8105 }, { "epoch": 0.07290453427349543, "grad_norm": 1.114105224609375, "learning_rate": 4.74545841732342e-06, "loss": 0.6932, "step": 8106 }, { "epoch": 0.07291352817113589, "grad_norm": 1.4845776557922363, "learning_rate": 4.740613379534428e-06, "loss": 0.3283, "step": 8107 }, { "epoch": 0.07292252206877634, "grad_norm": 0.37916865944862366, "learning_rate": 4.7357705572814964e-06, "loss": 1.0492, "step": 8108 }, { "epoch": 0.07293151596641678, "grad_norm": 0.7122876644134521, "learning_rate": 4.730929951094237e-06, "loss": 0.461, "step": 8109 }, { "epoch": 0.07294050986405724, "grad_norm": 0.47221359610557556, "learning_rate": 4.726091561502e-06, "loss": 0.6547, "step": 8110 }, { "epoch": 0.07294950376169769, "grad_norm": 0.8121677041053772, "learning_rate": 4.72125538903391e-06, "loss": 0.395, "step": 8111 }, { "epoch": 0.07295849765933814, "grad_norm": 0.48258158564567566, "learning_rate": 4.71642143421884e-06, "loss": 0.7579, "step": 8112 }, { "epoch": 0.07296749155697858, "grad_norm": 0.4000353515148163, "learning_rate": 4.711589697585423e-06, "loss": 0.9595, "step": 8113 }, { "epoch": 0.07297648545461904, "grad_norm": 0.6484460830688477, "learning_rate": 4.706760179662056e-06, "loss": 0.396, "step": 8114 }, { "epoch": 0.07298547935225949, "grad_norm": 2.7375099658966064, "learning_rate": 4.7019328809768895e-06, "loss": 0.1434, "step": 8115 }, { "epoch": 0.07299447324989994, "grad_norm": 1.0431023836135864, "learning_rate": 4.697107802057821e-06, "loss": 0.3416, "step": 8116 }, { "epoch": 0.0730034671475404, "grad_norm": 0.5979581475257874, "learning_rate": 4.692284943432523e-06, "loss": 0.9237, "step": 8117 }, { "epoch": 0.07301246104518085, "grad_norm": 0.6000280976295471, "learning_rate": 4.687464305628405e-06, "loss": 0.56, "step": 8118 }, { "epoch": 0.07302145494282129, "grad_norm": 0.3349648416042328, "learning_rate": 4.682645889172651e-06, "loss": 0.8771, "step": 8119 }, { "epoch": 0.07303044884046175, "grad_norm": 0.9795320630073547, "learning_rate": 4.677829694592198e-06, "loss": 0.9719, "step": 8120 }, { "epoch": 0.0730394427381022, "grad_norm": 0.6821804642677307, "learning_rate": 4.67301572241374e-06, "loss": 0.3059, "step": 8121 }, { "epoch": 0.07304843663574265, "grad_norm": 0.41533708572387695, "learning_rate": 4.668203973163715e-06, "loss": 0.7632, "step": 8122 }, { "epoch": 0.0730574305333831, "grad_norm": 0.37658122181892395, "learning_rate": 4.663394447368344e-06, "loss": 0.8158, "step": 8123 }, { "epoch": 0.07306642443102355, "grad_norm": 0.47663959860801697, "learning_rate": 4.6585871455535725e-06, "loss": 0.6526, "step": 8124 }, { "epoch": 0.073075418328664, "grad_norm": 0.3814954161643982, "learning_rate": 4.653782068245127e-06, "loss": 0.7518, "step": 8125 }, { "epoch": 0.07308441222630445, "grad_norm": 0.48097115755081177, "learning_rate": 4.648979215968485e-06, "loss": 0.0482, "step": 8126 }, { "epoch": 0.07309340612394491, "grad_norm": 0.8168143033981323, "learning_rate": 4.644178589248885e-06, "loss": 0.3838, "step": 8127 }, { "epoch": 0.07310240002158536, "grad_norm": 0.6025754809379578, "learning_rate": 4.639380188611306e-06, "loss": 0.0695, "step": 8128 }, { "epoch": 0.0731113939192258, "grad_norm": 0.3730815649032593, "learning_rate": 4.634584014580496e-06, "loss": 0.7684, "step": 8129 }, { "epoch": 0.07312038781686625, "grad_norm": 0.7987255454063416, "learning_rate": 4.629790067680964e-06, "loss": 0.99, "step": 8130 }, { "epoch": 0.07312938171450671, "grad_norm": 0.2661483883857727, "learning_rate": 4.62499834843696e-06, "loss": 1.0818, "step": 8131 }, { "epoch": 0.07313837561214716, "grad_norm": 0.5253507494926453, "learning_rate": 4.620208857372502e-06, "loss": 0.9563, "step": 8132 }, { "epoch": 0.0731473695097876, "grad_norm": 0.6872766017913818, "learning_rate": 4.6154215950113625e-06, "loss": 1.1605, "step": 8133 }, { "epoch": 0.07315636340742807, "grad_norm": 0.7948797941207886, "learning_rate": 4.610636561877077e-06, "loss": 0.3968, "step": 8134 }, { "epoch": 0.07316535730506851, "grad_norm": 0.35217493772506714, "learning_rate": 4.605853758492914e-06, "loss": 0.7274, "step": 8135 }, { "epoch": 0.07317435120270896, "grad_norm": 0.6802735924720764, "learning_rate": 4.601073185381929e-06, "loss": 0.2898, "step": 8136 }, { "epoch": 0.0731833451003494, "grad_norm": 0.5236994624137878, "learning_rate": 4.596294843066906e-06, "loss": 0.4256, "step": 8137 }, { "epoch": 0.07319233899798987, "grad_norm": 1.1669862270355225, "learning_rate": 4.591518732070402e-06, "loss": 0.5411, "step": 8138 }, { "epoch": 0.07320133289563031, "grad_norm": 0.7437425255775452, "learning_rate": 4.586744852914724e-06, "loss": 0.6623, "step": 8139 }, { "epoch": 0.07321032679327076, "grad_norm": 0.8411116003990173, "learning_rate": 4.581973206121948e-06, "loss": 0.2803, "step": 8140 }, { "epoch": 0.07321932069091122, "grad_norm": 1.4060131311416626, "learning_rate": 4.577203792213874e-06, "loss": 0.5906, "step": 8141 }, { "epoch": 0.07322831458855167, "grad_norm": 0.7642008662223816, "learning_rate": 4.572436611712097e-06, "loss": 0.4446, "step": 8142 }, { "epoch": 0.07323730848619212, "grad_norm": 0.3760044574737549, "learning_rate": 4.567671665137932e-06, "loss": 1.2364, "step": 8143 }, { "epoch": 0.07324630238383258, "grad_norm": 2.1277215480804443, "learning_rate": 4.562908953012474e-06, "loss": 0.0889, "step": 8144 }, { "epoch": 0.07325529628147302, "grad_norm": 1.0360071659088135, "learning_rate": 4.5581484758565665e-06, "loss": 0.0568, "step": 8145 }, { "epoch": 0.07326429017911347, "grad_norm": 0.46262645721435547, "learning_rate": 4.553390234190816e-06, "loss": 0.7598, "step": 8146 }, { "epoch": 0.07327328407675392, "grad_norm": 0.5302947163581848, "learning_rate": 4.548634228535559e-06, "loss": 0.7329, "step": 8147 }, { "epoch": 0.07328227797439438, "grad_norm": 0.34238699078559875, "learning_rate": 4.543880459410924e-06, "loss": 0.7037, "step": 8148 }, { "epoch": 0.07329127187203482, "grad_norm": 0.2888372242450714, "learning_rate": 4.539128927336758e-06, "loss": 1.119, "step": 8149 }, { "epoch": 0.07330026576967527, "grad_norm": 0.8298923969268799, "learning_rate": 4.534379632832692e-06, "loss": 0.9282, "step": 8150 }, { "epoch": 0.07330925966731573, "grad_norm": 0.43673741817474365, "learning_rate": 4.5296325764180965e-06, "loss": 0.9899, "step": 8151 }, { "epoch": 0.07331825356495618, "grad_norm": 0.35383546352386475, "learning_rate": 4.524887758612115e-06, "loss": 1.1194, "step": 8152 }, { "epoch": 0.07332724746259663, "grad_norm": 0.9937562942504883, "learning_rate": 4.520145179933616e-06, "loss": 0.3671, "step": 8153 }, { "epoch": 0.07333624136023707, "grad_norm": 2.1362924575805664, "learning_rate": 4.51540484090125e-06, "loss": 0.4751, "step": 8154 }, { "epoch": 0.07334523525787753, "grad_norm": 0.45611894130706787, "learning_rate": 4.51066674203342e-06, "loss": 0.7224, "step": 8155 }, { "epoch": 0.07335422915551798, "grad_norm": 0.4940088391304016, "learning_rate": 4.505930883848261e-06, "loss": 0.6519, "step": 8156 }, { "epoch": 0.07336322305315843, "grad_norm": 1.1135053634643555, "learning_rate": 4.501197266863691e-06, "loss": 0.8347, "step": 8157 }, { "epoch": 0.07337221695079889, "grad_norm": 0.9217567443847656, "learning_rate": 4.4964658915973676e-06, "loss": 0.5769, "step": 8158 }, { "epoch": 0.07338121084843933, "grad_norm": 1.3947820663452148, "learning_rate": 4.4917367585667144e-06, "loss": 0.8835, "step": 8159 }, { "epoch": 0.07339020474607978, "grad_norm": 1.3749333620071411, "learning_rate": 4.487009868288888e-06, "loss": 0.052, "step": 8160 }, { "epoch": 0.07339919864372024, "grad_norm": 0.6414727568626404, "learning_rate": 4.482285221280827e-06, "loss": 0.5541, "step": 8161 }, { "epoch": 0.07340819254136069, "grad_norm": 0.30642804503440857, "learning_rate": 4.477562818059203e-06, "loss": 0.9603, "step": 8162 }, { "epoch": 0.07341718643900114, "grad_norm": 0.6016037464141846, "learning_rate": 4.472842659140458e-06, "loss": 0.7501, "step": 8163 }, { "epoch": 0.07342618033664158, "grad_norm": 0.4716484248638153, "learning_rate": 4.468124745040767e-06, "loss": 0.7491, "step": 8164 }, { "epoch": 0.07343517423428204, "grad_norm": 0.43111974000930786, "learning_rate": 4.463409076276095e-06, "loss": 0.854, "step": 8165 }, { "epoch": 0.07344416813192249, "grad_norm": 0.6803651452064514, "learning_rate": 4.458695653362125e-06, "loss": 0.5432, "step": 8166 }, { "epoch": 0.07345316202956294, "grad_norm": 0.6923319101333618, "learning_rate": 4.4539844768143214e-06, "loss": 0.3963, "step": 8167 }, { "epoch": 0.0734621559272034, "grad_norm": 0.5295608639717102, "learning_rate": 4.449275547147877e-06, "loss": 0.968, "step": 8168 }, { "epoch": 0.07347114982484385, "grad_norm": 0.6301356554031372, "learning_rate": 4.444568864877769e-06, "loss": 0.5632, "step": 8169 }, { "epoch": 0.07348014372248429, "grad_norm": 3.0851545333862305, "learning_rate": 4.439864430518692e-06, "loss": 0.2556, "step": 8170 }, { "epoch": 0.07348913762012474, "grad_norm": 0.5738159418106079, "learning_rate": 4.435162244585142e-06, "loss": 0.7016, "step": 8171 }, { "epoch": 0.0734981315177652, "grad_norm": 0.6629881858825684, "learning_rate": 4.430462307591326e-06, "loss": 0.6581, "step": 8172 }, { "epoch": 0.07350712541540565, "grad_norm": 0.5860875248908997, "learning_rate": 4.425764620051231e-06, "loss": 0.6466, "step": 8173 }, { "epoch": 0.0735161193130461, "grad_norm": 0.6537678241729736, "learning_rate": 4.421069182478576e-06, "loss": 0.5291, "step": 8174 }, { "epoch": 0.07352511321068655, "grad_norm": 1.0756129026412964, "learning_rate": 4.416375995386857e-06, "loss": 0.3352, "step": 8175 }, { "epoch": 0.073534107108327, "grad_norm": 0.4112701117992401, "learning_rate": 4.411685059289314e-06, "loss": 0.5932, "step": 8176 }, { "epoch": 0.07354310100596745, "grad_norm": 0.3723878264427185, "learning_rate": 4.406996374698944e-06, "loss": 0.7348, "step": 8177 }, { "epoch": 0.0735520949036079, "grad_norm": 0.7879067659378052, "learning_rate": 4.40230994212848e-06, "loss": 0.3318, "step": 8178 }, { "epoch": 0.07356108880124836, "grad_norm": 0.5007966160774231, "learning_rate": 4.397625762090435e-06, "loss": 0.8203, "step": 8179 }, { "epoch": 0.0735700826988888, "grad_norm": 0.49231088161468506, "learning_rate": 4.392943835097069e-06, "loss": 0.4732, "step": 8180 }, { "epoch": 0.07357907659652925, "grad_norm": 0.49794483184814453, "learning_rate": 4.388264161660374e-06, "loss": 0.445, "step": 8181 }, { "epoch": 0.07358807049416971, "grad_norm": 0.7498834729194641, "learning_rate": 4.383586742292128e-06, "loss": 0.3519, "step": 8182 }, { "epoch": 0.07359706439181016, "grad_norm": 0.46211206912994385, "learning_rate": 4.3789115775038276e-06, "loss": 0.5922, "step": 8183 }, { "epoch": 0.0736060582894506, "grad_norm": 0.4384862184524536, "learning_rate": 4.374238667806765e-06, "loss": 0.5644, "step": 8184 }, { "epoch": 0.07361505218709107, "grad_norm": 0.46071937680244446, "learning_rate": 4.369568013711947e-06, "loss": 0.4858, "step": 8185 }, { "epoch": 0.07362404608473151, "grad_norm": 0.9254358410835266, "learning_rate": 4.364899615730159e-06, "loss": 0.5807, "step": 8186 }, { "epoch": 0.07363303998237196, "grad_norm": 0.36335086822509766, "learning_rate": 4.360233474371919e-06, "loss": 1.5275, "step": 8187 }, { "epoch": 0.0736420338800124, "grad_norm": 0.2659264802932739, "learning_rate": 4.35556959014752e-06, "loss": 1.2376, "step": 8188 }, { "epoch": 0.07365102777765287, "grad_norm": 0.41807830333709717, "learning_rate": 4.3509079635669805e-06, "loss": 0.8233, "step": 8189 }, { "epoch": 0.07366002167529331, "grad_norm": 3.162065267562866, "learning_rate": 4.346248595140112e-06, "loss": 0.1301, "step": 8190 }, { "epoch": 0.07366901557293376, "grad_norm": 0.586391031742096, "learning_rate": 4.34159148537644e-06, "loss": 0.3482, "step": 8191 }, { "epoch": 0.07367800947057422, "grad_norm": 0.35461416840553284, "learning_rate": 4.336936634785271e-06, "loss": 0.542, "step": 8192 }, { "epoch": 0.07368700336821467, "grad_norm": 0.5973930954933167, "learning_rate": 4.332284043875637e-06, "loss": 0.063, "step": 8193 }, { "epoch": 0.07369599726585511, "grad_norm": 0.5463537573814392, "learning_rate": 4.327633713156351e-06, "loss": 0.35, "step": 8194 }, { "epoch": 0.07370499116349556, "grad_norm": 3.466291666030884, "learning_rate": 4.322985643135952e-06, "loss": 0.6412, "step": 8195 }, { "epoch": 0.07371398506113602, "grad_norm": 0.5255039930343628, "learning_rate": 4.318339834322768e-06, "loss": 0.3838, "step": 8196 }, { "epoch": 0.07372297895877647, "grad_norm": 0.6938610672950745, "learning_rate": 4.313696287224839e-06, "loss": 0.4379, "step": 8197 }, { "epoch": 0.07373197285641692, "grad_norm": 0.37518206238746643, "learning_rate": 4.309055002349987e-06, "loss": 0.7181, "step": 8198 }, { "epoch": 0.07374096675405738, "grad_norm": 0.3468077778816223, "learning_rate": 4.304415980205767e-06, "loss": 0.9092, "step": 8199 }, { "epoch": 0.07374996065169782, "grad_norm": 0.5665051341056824, "learning_rate": 4.299779221299499e-06, "loss": 0.3313, "step": 8200 }, { "epoch": 0.07375895454933827, "grad_norm": 0.6083886027336121, "learning_rate": 4.295144726138259e-06, "loss": 0.6382, "step": 8201 }, { "epoch": 0.07376794844697873, "grad_norm": 0.41378945112228394, "learning_rate": 4.290512495228857e-06, "loss": 0.8992, "step": 8202 }, { "epoch": 0.07377694234461918, "grad_norm": 0.41916903853416443, "learning_rate": 4.285882529077872e-06, "loss": 0.7669, "step": 8203 }, { "epoch": 0.07378593624225963, "grad_norm": 0.4553501605987549, "learning_rate": 4.281254828191633e-06, "loss": 0.5502, "step": 8204 }, { "epoch": 0.07379493013990007, "grad_norm": 0.44712069630622864, "learning_rate": 4.27662939307622e-06, "loss": 0.3875, "step": 8205 }, { "epoch": 0.07380392403754053, "grad_norm": 0.311555415391922, "learning_rate": 4.272006224237452e-06, "loss": 0.6553, "step": 8206 }, { "epoch": 0.07381291793518098, "grad_norm": 2.0385987758636475, "learning_rate": 4.267385322180928e-06, "loss": 0.3384, "step": 8207 }, { "epoch": 0.07382191183282143, "grad_norm": 0.5150604844093323, "learning_rate": 4.26276668741196e-06, "loss": 0.6792, "step": 8208 }, { "epoch": 0.07383090573046189, "grad_norm": 0.6252371668815613, "learning_rate": 4.258150320435661e-06, "loss": 0.3679, "step": 8209 }, { "epoch": 0.07383989962810233, "grad_norm": 1.6120209693908691, "learning_rate": 4.253536221756851e-06, "loss": 0.8812, "step": 8210 }, { "epoch": 0.07384889352574278, "grad_norm": 0.4774073362350464, "learning_rate": 4.248924391880138e-06, "loss": 0.4158, "step": 8211 }, { "epoch": 0.07385788742338323, "grad_norm": 0.5380010008811951, "learning_rate": 4.244314831309845e-06, "loss": 0.7735, "step": 8212 }, { "epoch": 0.07386688132102369, "grad_norm": 0.34237727522850037, "learning_rate": 4.239707540550084e-06, "loss": 0.8648, "step": 8213 }, { "epoch": 0.07387587521866414, "grad_norm": 0.40482664108276367, "learning_rate": 4.235102520104681e-06, "loss": 0.5656, "step": 8214 }, { "epoch": 0.07388486911630458, "grad_norm": 0.3524165451526642, "learning_rate": 4.230499770477258e-06, "loss": 1.29, "step": 8215 }, { "epoch": 0.07389386301394504, "grad_norm": 1.4761080741882324, "learning_rate": 4.225899292171151e-06, "loss": 0.4885, "step": 8216 }, { "epoch": 0.07390285691158549, "grad_norm": 0.4479531943798065, "learning_rate": 4.221301085689466e-06, "loss": 0.8237, "step": 8217 }, { "epoch": 0.07391185080922594, "grad_norm": 0.396382600069046, "learning_rate": 4.21670515153505e-06, "loss": 0.6314, "step": 8218 }, { "epoch": 0.07392084470686638, "grad_norm": 1.0786864757537842, "learning_rate": 4.212111490210519e-06, "loss": 0.0836, "step": 8219 }, { "epoch": 0.07392983860450685, "grad_norm": 0.3171500265598297, "learning_rate": 4.207520102218213e-06, "loss": 0.8514, "step": 8220 }, { "epoch": 0.07393883250214729, "grad_norm": 0.45714396238327026, "learning_rate": 4.202930988060247e-06, "loss": 0.6041, "step": 8221 }, { "epoch": 0.07394782639978774, "grad_norm": 0.9490932822227478, "learning_rate": 4.198344148238481e-06, "loss": 0.0975, "step": 8222 }, { "epoch": 0.0739568202974282, "grad_norm": 0.3203533887863159, "learning_rate": 4.193759583254531e-06, "loss": 0.9253, "step": 8223 }, { "epoch": 0.07396581419506865, "grad_norm": 0.40048807859420776, "learning_rate": 4.1891772936097445e-06, "loss": 0.663, "step": 8224 }, { "epoch": 0.0739748080927091, "grad_norm": 0.4479226768016815, "learning_rate": 4.184597279805241e-06, "loss": 0.5735, "step": 8225 }, { "epoch": 0.07398380199034955, "grad_norm": 0.6566391587257385, "learning_rate": 4.180019542341889e-06, "loss": 0.6422, "step": 8226 }, { "epoch": 0.07399279588799, "grad_norm": 0.28084251284599304, "learning_rate": 4.175444081720292e-06, "loss": 0.6906, "step": 8227 }, { "epoch": 0.07400178978563045, "grad_norm": 0.5300658941268921, "learning_rate": 4.170870898440821e-06, "loss": 0.7309, "step": 8228 }, { "epoch": 0.0740107836832709, "grad_norm": 1.814934253692627, "learning_rate": 4.166299993003592e-06, "loss": 0.4057, "step": 8229 }, { "epoch": 0.07401977758091136, "grad_norm": 0.5271083116531372, "learning_rate": 4.161731365908481e-06, "loss": 0.7939, "step": 8230 }, { "epoch": 0.0740287714785518, "grad_norm": 0.3446708619594574, "learning_rate": 4.157165017655088e-06, "loss": 0.6175, "step": 8231 }, { "epoch": 0.07403776537619225, "grad_norm": 0.4736447036266327, "learning_rate": 4.152600948742802e-06, "loss": 0.5841, "step": 8232 }, { "epoch": 0.07404675927383271, "grad_norm": 0.5956741571426392, "learning_rate": 4.1480391596707215e-06, "loss": 0.7627, "step": 8233 }, { "epoch": 0.07405575317147316, "grad_norm": 0.3906766176223755, "learning_rate": 4.1434796509377314e-06, "loss": 0.8918, "step": 8234 }, { "epoch": 0.0740647470691136, "grad_norm": 0.6182167530059814, "learning_rate": 4.138922423042449e-06, "loss": 0.8017, "step": 8235 }, { "epoch": 0.07407374096675405, "grad_norm": 0.36508381366729736, "learning_rate": 4.1343674764832526e-06, "loss": 0.651, "step": 8236 }, { "epoch": 0.07408273486439451, "grad_norm": 0.44036632776260376, "learning_rate": 4.129814811758251e-06, "loss": 0.6393, "step": 8237 }, { "epoch": 0.07409172876203496, "grad_norm": 0.7966340184211731, "learning_rate": 4.12526442936533e-06, "loss": 0.47, "step": 8238 }, { "epoch": 0.0741007226596754, "grad_norm": 1.5516277551651, "learning_rate": 4.120716329802102e-06, "loss": 0.4787, "step": 8239 }, { "epoch": 0.07410971655731587, "grad_norm": 0.35034939646720886, "learning_rate": 4.116170513565942e-06, "loss": 0.7761, "step": 8240 }, { "epoch": 0.07411871045495631, "grad_norm": 0.4194398522377014, "learning_rate": 4.11162698115398e-06, "loss": 0.6307, "step": 8241 }, { "epoch": 0.07412770435259676, "grad_norm": 0.7047479152679443, "learning_rate": 4.107085733063088e-06, "loss": 1.21, "step": 8242 }, { "epoch": 0.07413669825023722, "grad_norm": 0.2609134614467621, "learning_rate": 4.1025467697898835e-06, "loss": 1.1003, "step": 8243 }, { "epoch": 0.07414569214787767, "grad_norm": 0.5125914216041565, "learning_rate": 4.098010091830754e-06, "loss": 0.8469, "step": 8244 }, { "epoch": 0.07415468604551811, "grad_norm": 5.512138843536377, "learning_rate": 4.093475699681806e-06, "loss": 0.1475, "step": 8245 }, { "epoch": 0.07416367994315856, "grad_norm": 0.3426024913787842, "learning_rate": 4.0889435938389234e-06, "loss": 0.8534, "step": 8246 }, { "epoch": 0.07417267384079902, "grad_norm": 1.3212965726852417, "learning_rate": 4.084413774797729e-06, "loss": 1.0429, "step": 8247 }, { "epoch": 0.07418166773843947, "grad_norm": 0.41673141717910767, "learning_rate": 4.079886243053607e-06, "loss": 0.5314, "step": 8248 }, { "epoch": 0.07419066163607992, "grad_norm": 0.5267758369445801, "learning_rate": 4.075360999101666e-06, "loss": 0.7615, "step": 8249 }, { "epoch": 0.07419965553372038, "grad_norm": 0.8013662099838257, "learning_rate": 4.070838043436786e-06, "loss": 0.6759, "step": 8250 }, { "epoch": 0.07420864943136082, "grad_norm": 0.41843676567077637, "learning_rate": 4.066317376553597e-06, "loss": 1.0231, "step": 8251 }, { "epoch": 0.07421764332900127, "grad_norm": 0.34707772731781006, "learning_rate": 4.061798998946459e-06, "loss": 0.7878, "step": 8252 }, { "epoch": 0.07422663722664172, "grad_norm": 0.6859583854675293, "learning_rate": 4.057282911109503e-06, "loss": 0.6068, "step": 8253 }, { "epoch": 0.07423563112428218, "grad_norm": 0.366169810295105, "learning_rate": 4.052769113536603e-06, "loss": 1.0348, "step": 8254 }, { "epoch": 0.07424462502192263, "grad_norm": 0.5037016272544861, "learning_rate": 4.048257606721381e-06, "loss": 0.7246, "step": 8255 }, { "epoch": 0.07425361891956307, "grad_norm": 0.5599848031997681, "learning_rate": 4.043748391157204e-06, "loss": 0.2375, "step": 8256 }, { "epoch": 0.07426261281720353, "grad_norm": 0.8938863277435303, "learning_rate": 4.039241467337201e-06, "loss": 0.8298, "step": 8257 }, { "epoch": 0.07427160671484398, "grad_norm": 0.4791821837425232, "learning_rate": 4.03473683575423e-06, "loss": 1.0281, "step": 8258 }, { "epoch": 0.07428060061248443, "grad_norm": 1.6014764308929443, "learning_rate": 4.03023449690092e-06, "loss": 0.6136, "step": 8259 }, { "epoch": 0.07428959451012487, "grad_norm": 0.48144063353538513, "learning_rate": 4.025734451269636e-06, "loss": 0.5504, "step": 8260 }, { "epoch": 0.07429858840776533, "grad_norm": 0.3939114212989807, "learning_rate": 4.021236699352504e-06, "loss": 0.9035, "step": 8261 }, { "epoch": 0.07430758230540578, "grad_norm": 0.8904026746749878, "learning_rate": 4.016741241641381e-06, "loss": 0.555, "step": 8262 }, { "epoch": 0.07431657620304623, "grad_norm": 3.5269598960876465, "learning_rate": 4.012248078627892e-06, "loss": 0.4349, "step": 8263 }, { "epoch": 0.07432557010068669, "grad_norm": 0.48067694902420044, "learning_rate": 4.007757210803395e-06, "loss": 0.7653, "step": 8264 }, { "epoch": 0.07433456399832714, "grad_norm": 0.28983861207962036, "learning_rate": 4.003268638659005e-06, "loss": 1.1078, "step": 8265 }, { "epoch": 0.07434355789596758, "grad_norm": 0.34098342061042786, "learning_rate": 3.998782362685591e-06, "loss": 0.9645, "step": 8266 }, { "epoch": 0.07435255179360804, "grad_norm": 0.4951697289943695, "learning_rate": 3.994298383373768e-06, "loss": 0.4547, "step": 8267 }, { "epoch": 0.07436154569124849, "grad_norm": 0.6220575571060181, "learning_rate": 3.989816701213886e-06, "loss": 0.4369, "step": 8268 }, { "epoch": 0.07437053958888894, "grad_norm": 0.417466402053833, "learning_rate": 3.985337316696066e-06, "loss": 0.728, "step": 8269 }, { "epoch": 0.07437953348652938, "grad_norm": 0.48346036672592163, "learning_rate": 3.98086023031016e-06, "loss": 0.7374, "step": 8270 }, { "epoch": 0.07438852738416984, "grad_norm": 0.6828868389129639, "learning_rate": 3.976385442545774e-06, "loss": 0.5755, "step": 8271 }, { "epoch": 0.07439752128181029, "grad_norm": 0.559496283531189, "learning_rate": 3.971912953892268e-06, "loss": 0.9328, "step": 8272 }, { "epoch": 0.07440651517945074, "grad_norm": 0.9357774257659912, "learning_rate": 3.967442764838752e-06, "loss": 0.4913, "step": 8273 }, { "epoch": 0.0744155090770912, "grad_norm": 0.2908155620098114, "learning_rate": 3.962974875874067e-06, "loss": 0.9685, "step": 8274 }, { "epoch": 0.07442450297473165, "grad_norm": 0.3315260410308838, "learning_rate": 3.958509287486823e-06, "loss": 0.846, "step": 8275 }, { "epoch": 0.07443349687237209, "grad_norm": 0.33557334542274475, "learning_rate": 3.954046000165374e-06, "loss": 0.8725, "step": 8276 }, { "epoch": 0.07444249077001254, "grad_norm": 0.573880136013031, "learning_rate": 3.9495850143978026e-06, "loss": 0.795, "step": 8277 }, { "epoch": 0.074451484667653, "grad_norm": 0.40424078702926636, "learning_rate": 3.945126330671969e-06, "loss": 0.6994, "step": 8278 }, { "epoch": 0.07446047856529345, "grad_norm": 0.41346871852874756, "learning_rate": 3.940669949475462e-06, "loss": 0.7698, "step": 8279 }, { "epoch": 0.0744694724629339, "grad_norm": 0.6880115270614624, "learning_rate": 3.936215871295634e-06, "loss": 0.8595, "step": 8280 }, { "epoch": 0.07447846636057436, "grad_norm": 0.6278561949729919, "learning_rate": 3.931764096619564e-06, "loss": 0.4191, "step": 8281 }, { "epoch": 0.0744874602582148, "grad_norm": 0.41943228244781494, "learning_rate": 3.9273146259341e-06, "loss": 0.937, "step": 8282 }, { "epoch": 0.07449645415585525, "grad_norm": 0.49884045124053955, "learning_rate": 3.922867459725823e-06, "loss": 0.3594, "step": 8283 }, { "epoch": 0.07450544805349571, "grad_norm": 0.6419369578361511, "learning_rate": 3.918422598481078e-06, "loss": 0.2454, "step": 8284 }, { "epoch": 0.07451444195113616, "grad_norm": 0.25919637084007263, "learning_rate": 3.913980042685928e-06, "loss": 0.9548, "step": 8285 }, { "epoch": 0.0745234358487766, "grad_norm": 0.3566941022872925, "learning_rate": 3.909539792826231e-06, "loss": 0.6947, "step": 8286 }, { "epoch": 0.07453242974641705, "grad_norm": 0.6137447953224182, "learning_rate": 3.905101849387547e-06, "loss": 0.7707, "step": 8287 }, { "epoch": 0.07454142364405751, "grad_norm": 0.6808339953422546, "learning_rate": 3.900666212855214e-06, "loss": 0.8454, "step": 8288 }, { "epoch": 0.07455041754169796, "grad_norm": 0.5459151864051819, "learning_rate": 3.896232883714299e-06, "loss": 0.5705, "step": 8289 }, { "epoch": 0.0745594114393384, "grad_norm": 0.5803229808807373, "learning_rate": 3.891801862449629e-06, "loss": 0.6325, "step": 8290 }, { "epoch": 0.07456840533697887, "grad_norm": 0.7745657563209534, "learning_rate": 3.887373149545761e-06, "loss": 0.3721, "step": 8291 }, { "epoch": 0.07457739923461931, "grad_norm": 0.39207354187965393, "learning_rate": 3.882946745487035e-06, "loss": 0.7156, "step": 8292 }, { "epoch": 0.07458639313225976, "grad_norm": 0.6383237242698669, "learning_rate": 3.878522650757499e-06, "loss": 0.9189, "step": 8293 }, { "epoch": 0.0745953870299002, "grad_norm": 0.5605208873748779, "learning_rate": 3.874100865840974e-06, "loss": 0.6385, "step": 8294 }, { "epoch": 0.07460438092754067, "grad_norm": 0.695473849773407, "learning_rate": 3.869681391221011e-06, "loss": 0.4295, "step": 8295 }, { "epoch": 0.07461337482518111, "grad_norm": 0.9956722259521484, "learning_rate": 3.865264227380924e-06, "loss": 0.8796, "step": 8296 }, { "epoch": 0.07462236872282156, "grad_norm": 0.7760307788848877, "learning_rate": 3.860849374803762e-06, "loss": 0.4367, "step": 8297 }, { "epoch": 0.07463136262046202, "grad_norm": 0.8014058470726013, "learning_rate": 3.856436833972341e-06, "loss": 0.5459, "step": 8298 }, { "epoch": 0.07464035651810247, "grad_norm": 0.38650354743003845, "learning_rate": 3.852026605369188e-06, "loss": 0.8231, "step": 8299 }, { "epoch": 0.07464935041574292, "grad_norm": 0.6365307569503784, "learning_rate": 3.847618689476612e-06, "loss": 0.6451, "step": 8300 }, { "epoch": 0.07465834431338338, "grad_norm": 1.565586805343628, "learning_rate": 3.843213086776662e-06, "loss": 0.7527, "step": 8301 }, { "epoch": 0.07466733821102382, "grad_norm": 0.38970956206321716, "learning_rate": 3.838809797751111e-06, "loss": 0.8768, "step": 8302 }, { "epoch": 0.07467633210866427, "grad_norm": 0.5320625901222229, "learning_rate": 3.834408822881511e-06, "loss": 0.817, "step": 8303 }, { "epoch": 0.07468532600630472, "grad_norm": 0.6596502661705017, "learning_rate": 3.8300101626491305e-06, "loss": 0.7914, "step": 8304 }, { "epoch": 0.07469431990394518, "grad_norm": 1.8198093175888062, "learning_rate": 3.825613817535021e-06, "loss": 0.1958, "step": 8305 }, { "epoch": 0.07470331380158562, "grad_norm": 0.593949019908905, "learning_rate": 3.821219788019942e-06, "loss": 0.7043, "step": 8306 }, { "epoch": 0.07471230769922607, "grad_norm": 0.42245304584503174, "learning_rate": 3.816828074584433e-06, "loss": 0.9782, "step": 8307 }, { "epoch": 0.07472130159686653, "grad_norm": 0.4858727753162384, "learning_rate": 3.812438677708749e-06, "loss": 0.8556, "step": 8308 }, { "epoch": 0.07473029549450698, "grad_norm": 0.7625946998596191, "learning_rate": 3.8080515978729247e-06, "loss": 0.2689, "step": 8309 }, { "epoch": 0.07473928939214743, "grad_norm": 0.674423336982727, "learning_rate": 3.8036668355567045e-06, "loss": 0.4896, "step": 8310 }, { "epoch": 0.07474828328978787, "grad_norm": 1.1422349214553833, "learning_rate": 3.7992843912396214e-06, "loss": 0.3507, "step": 8311 }, { "epoch": 0.07475727718742833, "grad_norm": 0.471788614988327, "learning_rate": 3.7949042654009142e-06, "loss": 0.4037, "step": 8312 }, { "epoch": 0.07476627108506878, "grad_norm": 0.45680323243141174, "learning_rate": 3.7905264585196056e-06, "loss": 0.921, "step": 8313 }, { "epoch": 0.07477526498270923, "grad_norm": 1.014906644821167, "learning_rate": 3.786150971074426e-06, "loss": 0.4727, "step": 8314 }, { "epoch": 0.07478425888034969, "grad_norm": 0.5589510798454285, "learning_rate": 3.78177780354389e-06, "loss": 0.7406, "step": 8315 }, { "epoch": 0.07479325277799014, "grad_norm": 0.5094342231750488, "learning_rate": 3.7774069564062182e-06, "loss": 0.8521, "step": 8316 }, { "epoch": 0.07480224667563058, "grad_norm": 0.46003052592277527, "learning_rate": 3.773038430139428e-06, "loss": 0.6094, "step": 8317 }, { "epoch": 0.07481124057327103, "grad_norm": 0.3422051966190338, "learning_rate": 3.7686722252212344e-06, "loss": 0.323, "step": 8318 }, { "epoch": 0.07482023447091149, "grad_norm": 0.2701261341571808, "learning_rate": 3.7643083421291303e-06, "loss": 1.0026, "step": 8319 }, { "epoch": 0.07482922836855194, "grad_norm": 0.3399820625782013, "learning_rate": 3.7599467813403315e-06, "loss": 0.6926, "step": 8320 }, { "epoch": 0.07483822226619238, "grad_norm": 0.6520413160324097, "learning_rate": 3.7555875433318204e-06, "loss": 0.5243, "step": 8321 }, { "epoch": 0.07484721616383284, "grad_norm": 0.4821566343307495, "learning_rate": 3.7512306285803212e-06, "loss": 0.8622, "step": 8322 }, { "epoch": 0.07485621006147329, "grad_norm": 0.8450035452842712, "learning_rate": 3.7468760375622887e-06, "loss": 1.1323, "step": 8323 }, { "epoch": 0.07486520395911374, "grad_norm": 0.4520803689956665, "learning_rate": 3.742523770753939e-06, "loss": 0.8853, "step": 8324 }, { "epoch": 0.0748741978567542, "grad_norm": 0.5575488805770874, "learning_rate": 3.738173828631228e-06, "loss": 0.8928, "step": 8325 }, { "epoch": 0.07488319175439465, "grad_norm": 0.8523882031440735, "learning_rate": 3.7338262116698663e-06, "loss": 0.4244, "step": 8326 }, { "epoch": 0.07489218565203509, "grad_norm": 0.6087777018547058, "learning_rate": 3.7294809203452907e-06, "loss": 0.6382, "step": 8327 }, { "epoch": 0.07490117954967554, "grad_norm": 0.4438583254814148, "learning_rate": 3.725137955132707e-06, "loss": 0.864, "step": 8328 }, { "epoch": 0.074910173447316, "grad_norm": 0.24856172502040863, "learning_rate": 3.7207973165070407e-06, "loss": 1.0225, "step": 8329 }, { "epoch": 0.07491916734495645, "grad_norm": 0.7694800496101379, "learning_rate": 3.7164590049429987e-06, "loss": 0.5488, "step": 8330 }, { "epoch": 0.0749281612425969, "grad_norm": 1.9150689840316772, "learning_rate": 3.7121230209149926e-06, "loss": 0.3434, "step": 8331 }, { "epoch": 0.07493715514023735, "grad_norm": 0.5523433089256287, "learning_rate": 3.7077893648972157e-06, "loss": 0.6461, "step": 8332 }, { "epoch": 0.0749461490378778, "grad_norm": 0.3769187927246094, "learning_rate": 3.703458037363575e-06, "loss": 0.9691, "step": 8333 }, { "epoch": 0.07495514293551825, "grad_norm": 0.22373609244823456, "learning_rate": 3.699129038787749e-06, "loss": 1.0614, "step": 8334 }, { "epoch": 0.0749641368331587, "grad_norm": 0.6724124550819397, "learning_rate": 3.6948023696431354e-06, "loss": 0.8465, "step": 8335 }, { "epoch": 0.07497313073079916, "grad_norm": 0.3364690840244293, "learning_rate": 3.6904780304029135e-06, "loss": 1.021, "step": 8336 }, { "epoch": 0.0749821246284396, "grad_norm": 0.4710368514060974, "learning_rate": 3.6861560215399716e-06, "loss": 0.8187, "step": 8337 }, { "epoch": 0.07499111852608005, "grad_norm": 2.600294351577759, "learning_rate": 3.681836343526965e-06, "loss": 0.2237, "step": 8338 }, { "epoch": 0.07500011242372051, "grad_norm": 0.41248753666877747, "learning_rate": 3.6775189968362822e-06, "loss": 0.786, "step": 8339 }, { "epoch": 0.07500910632136096, "grad_norm": 0.3249988257884979, "learning_rate": 3.6732039819400683e-06, "loss": 0.3809, "step": 8340 }, { "epoch": 0.0750181002190014, "grad_norm": 2.9204626083374023, "learning_rate": 3.6688912993101954e-06, "loss": 0.4521, "step": 8341 }, { "epoch": 0.07502709411664187, "grad_norm": 0.44494006037712097, "learning_rate": 3.6645809494183003e-06, "loss": 0.8328, "step": 8342 }, { "epoch": 0.07503608801428231, "grad_norm": 0.41467636823654175, "learning_rate": 3.660272932735756e-06, "loss": 0.8403, "step": 8343 }, { "epoch": 0.07504508191192276, "grad_norm": 1.1209052801132202, "learning_rate": 3.6559672497336854e-06, "loss": 0.4746, "step": 8344 }, { "epoch": 0.0750540758095632, "grad_norm": 0.3237215280532837, "learning_rate": 3.65166390088294e-06, "loss": 1.1178, "step": 8345 }, { "epoch": 0.07506306970720367, "grad_norm": 0.4744148254394531, "learning_rate": 3.647362886654135e-06, "loss": 0.7995, "step": 8346 }, { "epoch": 0.07507206360484411, "grad_norm": 1.4028921127319336, "learning_rate": 3.643064207517624e-06, "loss": 0.5033, "step": 8347 }, { "epoch": 0.07508105750248456, "grad_norm": 0.5911120772361755, "learning_rate": 3.6387678639434983e-06, "loss": 0.3418, "step": 8348 }, { "epoch": 0.07509005140012502, "grad_norm": 0.25234100222587585, "learning_rate": 3.6344738564016017e-06, "loss": 1.1396, "step": 8349 }, { "epoch": 0.07509904529776547, "grad_norm": 0.4256589114665985, "learning_rate": 3.630182185361522e-06, "loss": 1.0119, "step": 8350 }, { "epoch": 0.07510803919540592, "grad_norm": 0.631626546382904, "learning_rate": 3.6258928512925977e-06, "loss": 0.956, "step": 8351 }, { "epoch": 0.07511703309304636, "grad_norm": 0.47229301929473877, "learning_rate": 3.621605854663887e-06, "loss": 0.6515, "step": 8352 }, { "epoch": 0.07512602699068682, "grad_norm": 1.5979714393615723, "learning_rate": 3.617321195944223e-06, "loss": 0.3563, "step": 8353 }, { "epoch": 0.07513502088832727, "grad_norm": 0.38226112723350525, "learning_rate": 3.613038875602162e-06, "loss": 0.9871, "step": 8354 }, { "epoch": 0.07514401478596772, "grad_norm": 0.35405251383781433, "learning_rate": 3.6087588941060124e-06, "loss": 0.695, "step": 8355 }, { "epoch": 0.07515300868360818, "grad_norm": 0.6965729594230652, "learning_rate": 3.6044812519238303e-06, "loss": 0.2914, "step": 8356 }, { "epoch": 0.07516200258124862, "grad_norm": 0.512387216091156, "learning_rate": 3.6002059495234165e-06, "loss": 1.105, "step": 8357 }, { "epoch": 0.07517099647888907, "grad_norm": 0.8046634793281555, "learning_rate": 3.5959329873723e-06, "loss": 0.0841, "step": 8358 }, { "epoch": 0.07517999037652952, "grad_norm": 1.2521843910217285, "learning_rate": 3.5916623659377756e-06, "loss": 0.7856, "step": 8359 }, { "epoch": 0.07518898427416998, "grad_norm": 0.9233800172805786, "learning_rate": 3.5873940856868656e-06, "loss": 0.3086, "step": 8360 }, { "epoch": 0.07519797817181043, "grad_norm": 0.42301735281944275, "learning_rate": 3.5831281470863425e-06, "loss": 0.7844, "step": 8361 }, { "epoch": 0.07520697206945087, "grad_norm": 0.7032343149185181, "learning_rate": 3.578864550602726e-06, "loss": 0.908, "step": 8362 }, { "epoch": 0.07521596596709133, "grad_norm": 0.2919432818889618, "learning_rate": 3.5746032967022807e-06, "loss": 0.6217, "step": 8363 }, { "epoch": 0.07522495986473178, "grad_norm": 0.28228890895843506, "learning_rate": 3.5703443858510016e-06, "loss": 1.0367, "step": 8364 }, { "epoch": 0.07523395376237223, "grad_norm": 0.3080548644065857, "learning_rate": 3.5660878185146463e-06, "loss": 0.9487, "step": 8365 }, { "epoch": 0.07524294766001269, "grad_norm": 0.3541378676891327, "learning_rate": 3.561833595158698e-06, "loss": 0.7138, "step": 8366 }, { "epoch": 0.07525194155765313, "grad_norm": 0.5446239113807678, "learning_rate": 3.5575817162483926e-06, "loss": 0.7889, "step": 8367 }, { "epoch": 0.07526093545529358, "grad_norm": 0.3944709002971649, "learning_rate": 3.5533321822487143e-06, "loss": 0.9763, "step": 8368 }, { "epoch": 0.07526992935293403, "grad_norm": 0.4479169547557831, "learning_rate": 3.549084993624391e-06, "loss": 0.9824, "step": 8369 }, { "epoch": 0.07527892325057449, "grad_norm": 0.24142324924468994, "learning_rate": 3.544840150839876e-06, "loss": 1.0999, "step": 8370 }, { "epoch": 0.07528791714821494, "grad_norm": 0.5093030333518982, "learning_rate": 3.5405976543593844e-06, "loss": 0.7355, "step": 8371 }, { "epoch": 0.07529691104585538, "grad_norm": 0.41290146112442017, "learning_rate": 3.536357504646873e-06, "loss": 0.748, "step": 8372 }, { "epoch": 0.07530590494349584, "grad_norm": 0.3243314325809479, "learning_rate": 3.5321197021660313e-06, "loss": 0.6651, "step": 8373 }, { "epoch": 0.07531489884113629, "grad_norm": 0.4850621521472931, "learning_rate": 3.5278842473803032e-06, "loss": 1.0501, "step": 8374 }, { "epoch": 0.07532389273877674, "grad_norm": 1.5989028215408325, "learning_rate": 3.523651140752868e-06, "loss": 0.4846, "step": 8375 }, { "epoch": 0.07533288663641718, "grad_norm": 0.36488667130470276, "learning_rate": 3.519420382746663e-06, "loss": 0.9555, "step": 8376 }, { "epoch": 0.07534188053405765, "grad_norm": 0.5377614498138428, "learning_rate": 3.5151919738243416e-06, "loss": 1.2761, "step": 8377 }, { "epoch": 0.07535087443169809, "grad_norm": 0.3437438905239105, "learning_rate": 3.510965914448333e-06, "loss": 0.8527, "step": 8378 }, { "epoch": 0.07535986832933854, "grad_norm": 0.9372901320457458, "learning_rate": 3.5067422050807764e-06, "loss": 0.4403, "step": 8379 }, { "epoch": 0.075368862226979, "grad_norm": 0.3783667981624603, "learning_rate": 3.502520846183577e-06, "loss": 0.8314, "step": 8380 }, { "epoch": 0.07537785612461945, "grad_norm": 0.6540364623069763, "learning_rate": 3.4983018382183762e-06, "loss": 0.9967, "step": 8381 }, { "epoch": 0.0753868500222599, "grad_norm": 0.7870773077011108, "learning_rate": 3.494085181646567e-06, "loss": 0.0859, "step": 8382 }, { "epoch": 0.07539584391990035, "grad_norm": 0.6137230396270752, "learning_rate": 3.4898708769292604e-06, "loss": 0.2366, "step": 8383 }, { "epoch": 0.0754048378175408, "grad_norm": 0.6002194881439209, "learning_rate": 3.4856589245273437e-06, "loss": 0.582, "step": 8384 }, { "epoch": 0.07541383171518125, "grad_norm": 0.33146342635154724, "learning_rate": 3.4814493249014116e-06, "loss": 0.9644, "step": 8385 }, { "epoch": 0.0754228256128217, "grad_norm": 0.32534193992614746, "learning_rate": 3.4772420785118306e-06, "loss": 0.7875, "step": 8386 }, { "epoch": 0.07543181951046216, "grad_norm": 0.3259071707725525, "learning_rate": 3.473037185818695e-06, "loss": 1.4337, "step": 8387 }, { "epoch": 0.0754408134081026, "grad_norm": 0.3856044411659241, "learning_rate": 3.4688346472818546e-06, "loss": 0.8835, "step": 8388 }, { "epoch": 0.07544980730574305, "grad_norm": 0.35841840505599976, "learning_rate": 3.46463446336088e-06, "loss": 0.985, "step": 8389 }, { "epoch": 0.07545880120338351, "grad_norm": 1.2245829105377197, "learning_rate": 3.46043663451511e-06, "loss": 0.5812, "step": 8390 }, { "epoch": 0.07546779510102396, "grad_norm": 0.6813901662826538, "learning_rate": 3.456241161203597e-06, "loss": 0.6332, "step": 8391 }, { "epoch": 0.0754767889986644, "grad_norm": 0.35374557971954346, "learning_rate": 3.4520480438851618e-06, "loss": 0.7714, "step": 8392 }, { "epoch": 0.07548578289630485, "grad_norm": 0.3871983289718628, "learning_rate": 3.447857283018352e-06, "loss": 0.9437, "step": 8393 }, { "epoch": 0.07549477679394531, "grad_norm": 0.8548184037208557, "learning_rate": 3.4436688790614764e-06, "loss": 0.6282, "step": 8394 }, { "epoch": 0.07550377069158576, "grad_norm": 0.7650351524353027, "learning_rate": 3.439482832472554e-06, "loss": 0.4538, "step": 8395 }, { "epoch": 0.0755127645892262, "grad_norm": 0.22801868617534637, "learning_rate": 3.435299143709375e-06, "loss": 0.9784, "step": 8396 }, { "epoch": 0.07552175848686667, "grad_norm": 0.5556910634040833, "learning_rate": 3.4311178132294674e-06, "loss": 0.3866, "step": 8397 }, { "epoch": 0.07553075238450711, "grad_norm": 0.37567245960235596, "learning_rate": 3.4269388414900792e-06, "loss": 1.0238, "step": 8398 }, { "epoch": 0.07553974628214756, "grad_norm": 0.6669659614562988, "learning_rate": 3.422762228948223e-06, "loss": 0.2877, "step": 8399 }, { "epoch": 0.07554874017978801, "grad_norm": 0.5037093758583069, "learning_rate": 3.418587976060653e-06, "loss": 0.7306, "step": 8400 }, { "epoch": 0.07555773407742847, "grad_norm": 0.6397600769996643, "learning_rate": 3.414416083283856e-06, "loss": 0.4295, "step": 8401 }, { "epoch": 0.07556672797506891, "grad_norm": 0.8186016082763672, "learning_rate": 3.410246551074059e-06, "loss": 0.5367, "step": 8402 }, { "epoch": 0.07557572187270936, "grad_norm": 0.5201413631439209, "learning_rate": 3.4060793798872447e-06, "loss": 0.8824, "step": 8403 }, { "epoch": 0.07558471577034982, "grad_norm": 0.36526691913604736, "learning_rate": 3.4019145701791184e-06, "loss": 0.7445, "step": 8404 }, { "epoch": 0.07559370966799027, "grad_norm": 0.506876528263092, "learning_rate": 3.3977521224051427e-06, "loss": 0.7864, "step": 8405 }, { "epoch": 0.07560270356563072, "grad_norm": 0.6265326738357544, "learning_rate": 3.3935920370205154e-06, "loss": 0.4746, "step": 8406 }, { "epoch": 0.07561169746327118, "grad_norm": 1.0299736261367798, "learning_rate": 3.389434314480186e-06, "loss": 0.4173, "step": 8407 }, { "epoch": 0.07562069136091162, "grad_norm": 0.49308475852012634, "learning_rate": 3.3852789552388214e-06, "loss": 0.6879, "step": 8408 }, { "epoch": 0.07562968525855207, "grad_norm": 0.314328134059906, "learning_rate": 3.3811259597508582e-06, "loss": 1.1785, "step": 8409 }, { "epoch": 0.07563867915619252, "grad_norm": 0.3944673538208008, "learning_rate": 3.3769753284704526e-06, "loss": 0.9941, "step": 8410 }, { "epoch": 0.07564767305383298, "grad_norm": 0.6239811182022095, "learning_rate": 3.3728270618515188e-06, "loss": 0.6776, "step": 8411 }, { "epoch": 0.07565666695147343, "grad_norm": 0.7427331805229187, "learning_rate": 3.368681160347692e-06, "loss": 0.3884, "step": 8412 }, { "epoch": 0.07566566084911387, "grad_norm": 0.76864093542099, "learning_rate": 3.364537624412381e-06, "loss": 0.6714, "step": 8413 }, { "epoch": 0.07567465474675433, "grad_norm": 0.82759028673172, "learning_rate": 3.3603964544987043e-06, "loss": 0.3671, "step": 8414 }, { "epoch": 0.07568364864439478, "grad_norm": 1.1404221057891846, "learning_rate": 3.3562576510595385e-06, "loss": 0.3911, "step": 8415 }, { "epoch": 0.07569264254203523, "grad_norm": 0.4048179090023041, "learning_rate": 3.3521212145474905e-06, "loss": 0.2579, "step": 8416 }, { "epoch": 0.07570163643967567, "grad_norm": 0.5728513598442078, "learning_rate": 3.3479871454149214e-06, "loss": 0.8729, "step": 8417 }, { "epoch": 0.07571063033731613, "grad_norm": 0.5440266132354736, "learning_rate": 3.3438554441139247e-06, "loss": 0.7649, "step": 8418 }, { "epoch": 0.07571962423495658, "grad_norm": 0.3141704499721527, "learning_rate": 3.339726111096342e-06, "loss": 0.9276, "step": 8419 }, { "epoch": 0.07572861813259703, "grad_norm": 0.7255260944366455, "learning_rate": 3.3355991468137394e-06, "loss": 1.2468, "step": 8420 }, { "epoch": 0.07573761203023749, "grad_norm": 0.9209005832672119, "learning_rate": 3.3314745517174423e-06, "loss": 0.6127, "step": 8421 }, { "epoch": 0.07574660592787794, "grad_norm": 1.2298367023468018, "learning_rate": 3.327352326258515e-06, "loss": 1.07, "step": 8422 }, { "epoch": 0.07575559982551838, "grad_norm": 0.42664864659309387, "learning_rate": 3.3232324708877495e-06, "loss": 0.4999, "step": 8423 }, { "epoch": 0.07576459372315884, "grad_norm": 0.39712774753570557, "learning_rate": 3.3191149860556963e-06, "loss": 0.8066, "step": 8424 }, { "epoch": 0.07577358762079929, "grad_norm": 0.4824213981628418, "learning_rate": 3.314999872212618e-06, "loss": 0.6494, "step": 8425 }, { "epoch": 0.07578258151843974, "grad_norm": 0.5014702081680298, "learning_rate": 3.3108871298085625e-06, "loss": 0.5688, "step": 8426 }, { "epoch": 0.07579157541608018, "grad_norm": 0.27159741520881653, "learning_rate": 3.3067767592932787e-06, "loss": 1.0497, "step": 8427 }, { "epoch": 0.07580056931372064, "grad_norm": 0.9491857886314392, "learning_rate": 3.302668761116276e-06, "loss": 0.4684, "step": 8428 }, { "epoch": 0.07580956321136109, "grad_norm": 0.7448484301567078, "learning_rate": 3.2985631357267897e-06, "loss": 0.8608, "step": 8429 }, { "epoch": 0.07581855710900154, "grad_norm": 0.4629436135292053, "learning_rate": 3.2944598835738193e-06, "loss": 0.8206, "step": 8430 }, { "epoch": 0.075827551006642, "grad_norm": 0.5083878040313721, "learning_rate": 3.2903590051060694e-06, "loss": 0.7512, "step": 8431 }, { "epoch": 0.07583654490428245, "grad_norm": 0.43731388449668884, "learning_rate": 3.2862605007720287e-06, "loss": 0.8447, "step": 8432 }, { "epoch": 0.0758455388019229, "grad_norm": 0.4168355464935303, "learning_rate": 3.2821643710198914e-06, "loss": 0.7482, "step": 8433 }, { "epoch": 0.07585453269956334, "grad_norm": 0.3151644766330719, "learning_rate": 3.27807061629761e-06, "loss": 0.7558, "step": 8434 }, { "epoch": 0.0758635265972038, "grad_norm": 0.24805623292922974, "learning_rate": 3.2739792370528628e-06, "loss": 0.9603, "step": 8435 }, { "epoch": 0.07587252049484425, "grad_norm": 0.6583631038665771, "learning_rate": 3.269890233733086e-06, "loss": 0.6954, "step": 8436 }, { "epoch": 0.0758815143924847, "grad_norm": 2.0174312591552734, "learning_rate": 3.2658036067854307e-06, "loss": 0.9228, "step": 8437 }, { "epoch": 0.07589050829012516, "grad_norm": 0.36747920513153076, "learning_rate": 3.2617193566568283e-06, "loss": 0.7821, "step": 8438 }, { "epoch": 0.0758995021877656, "grad_norm": 0.4283698797225952, "learning_rate": 3.25763748379391e-06, "loss": 0.7016, "step": 8439 }, { "epoch": 0.07590849608540605, "grad_norm": 0.7665634751319885, "learning_rate": 3.2535579886430718e-06, "loss": 0.359, "step": 8440 }, { "epoch": 0.0759174899830465, "grad_norm": 0.394812673330307, "learning_rate": 3.2494808716504287e-06, "loss": 0.9393, "step": 8441 }, { "epoch": 0.07592648388068696, "grad_norm": 0.5281763672828674, "learning_rate": 3.245406133261858e-06, "loss": 0.4131, "step": 8442 }, { "epoch": 0.0759354777783274, "grad_norm": 0.3546455502510071, "learning_rate": 3.241333773922972e-06, "loss": 0.5238, "step": 8443 }, { "epoch": 0.07594447167596785, "grad_norm": 0.6938077807426453, "learning_rate": 3.237263794079104e-06, "loss": 0.4094, "step": 8444 }, { "epoch": 0.07595346557360831, "grad_norm": 1.9696524143218994, "learning_rate": 3.2331961941753474e-06, "loss": 0.4192, "step": 8445 }, { "epoch": 0.07596245947124876, "grad_norm": 1.2744743824005127, "learning_rate": 3.22913097465653e-06, "loss": 0.3965, "step": 8446 }, { "epoch": 0.0759714533688892, "grad_norm": 0.38299280405044556, "learning_rate": 3.2250681359672214e-06, "loss": 0.6687, "step": 8447 }, { "epoch": 0.07598044726652967, "grad_norm": 0.33991724252700806, "learning_rate": 3.2210076785517186e-06, "loss": 0.8358, "step": 8448 }, { "epoch": 0.07598944116417011, "grad_norm": 0.3193223178386688, "learning_rate": 3.2169496028540775e-06, "loss": 1.1761, "step": 8449 }, { "epoch": 0.07599843506181056, "grad_norm": 0.680895209312439, "learning_rate": 3.2128939093180655e-06, "loss": 0.7082, "step": 8450 }, { "epoch": 0.076007428959451, "grad_norm": 0.3800429701805115, "learning_rate": 3.2088405983872306e-06, "loss": 0.9812, "step": 8451 }, { "epoch": 0.07601642285709147, "grad_norm": 0.49568238854408264, "learning_rate": 3.204789670504821e-06, "loss": 0.811, "step": 8452 }, { "epoch": 0.07602541675473191, "grad_norm": 0.3592278063297272, "learning_rate": 3.200741126113849e-06, "loss": 0.8009, "step": 8453 }, { "epoch": 0.07603441065237236, "grad_norm": 0.7292088270187378, "learning_rate": 3.196694965657046e-06, "loss": 0.2504, "step": 8454 }, { "epoch": 0.07604340455001282, "grad_norm": 0.3082946538925171, "learning_rate": 3.192651189576909e-06, "loss": 0.9402, "step": 8455 }, { "epoch": 0.07605239844765327, "grad_norm": 0.6820107102394104, "learning_rate": 3.1886097983156367e-06, "loss": 0.5529, "step": 8456 }, { "epoch": 0.07606139234529372, "grad_norm": 0.3896775543689728, "learning_rate": 3.184570792315217e-06, "loss": 0.6314, "step": 8457 }, { "epoch": 0.07607038624293416, "grad_norm": 0.7224217057228088, "learning_rate": 3.1805341720173326e-06, "loss": 0.5001, "step": 8458 }, { "epoch": 0.07607938014057462, "grad_norm": 0.3636378049850464, "learning_rate": 3.176499937863428e-06, "loss": 0.7971, "step": 8459 }, { "epoch": 0.07608837403821507, "grad_norm": 0.540843665599823, "learning_rate": 3.1724680902946753e-06, "loss": 0.5218, "step": 8460 }, { "epoch": 0.07609736793585552, "grad_norm": 0.8104742765426636, "learning_rate": 3.168438629752002e-06, "loss": 0.5553, "step": 8461 }, { "epoch": 0.07610636183349598, "grad_norm": 0.40583574771881104, "learning_rate": 3.16441155667605e-06, "loss": 0.6568, "step": 8462 }, { "epoch": 0.07611535573113642, "grad_norm": 0.36365431547164917, "learning_rate": 3.1603868715072205e-06, "loss": 0.807, "step": 8463 }, { "epoch": 0.07612434962877687, "grad_norm": 0.5765535831451416, "learning_rate": 3.156364574685647e-06, "loss": 0.5929, "step": 8464 }, { "epoch": 0.07613334352641733, "grad_norm": 0.23572102189064026, "learning_rate": 3.152344666651208e-06, "loss": 0.8474, "step": 8465 }, { "epoch": 0.07614233742405778, "grad_norm": 0.4757998585700989, "learning_rate": 3.148327147843505e-06, "loss": 0.7773, "step": 8466 }, { "epoch": 0.07615133132169823, "grad_norm": 0.9029171466827393, "learning_rate": 3.1443120187018895e-06, "loss": 0.5158, "step": 8467 }, { "epoch": 0.07616032521933867, "grad_norm": 0.4869462251663208, "learning_rate": 3.140299279665457e-06, "loss": 0.5295, "step": 8468 }, { "epoch": 0.07616931911697913, "grad_norm": 0.35742852091789246, "learning_rate": 3.136288931173026e-06, "loss": 1.1156, "step": 8469 }, { "epoch": 0.07617831301461958, "grad_norm": 1.8894689083099365, "learning_rate": 3.1322809736631654e-06, "loss": 0.1301, "step": 8470 }, { "epoch": 0.07618730691226003, "grad_norm": 0.8453227281570435, "learning_rate": 3.12827540757418e-06, "loss": 0.2301, "step": 8471 }, { "epoch": 0.07619630080990049, "grad_norm": 1.1694034337997437, "learning_rate": 3.1242722333441193e-06, "loss": 0.5266, "step": 8472 }, { "epoch": 0.07620529470754094, "grad_norm": 0.3007834851741791, "learning_rate": 3.12027145141075e-06, "loss": 0.9935, "step": 8473 }, { "epoch": 0.07621428860518138, "grad_norm": 0.3363257348537445, "learning_rate": 3.1162730622116054e-06, "loss": 1.0303, "step": 8474 }, { "epoch": 0.07622328250282183, "grad_norm": 0.8700676560401917, "learning_rate": 3.11227706618393e-06, "loss": 0.8662, "step": 8475 }, { "epoch": 0.07623227640046229, "grad_norm": 0.3813903033733368, "learning_rate": 3.1082834637647274e-06, "loss": 0.7942, "step": 8476 }, { "epoch": 0.07624127029810274, "grad_norm": 0.3115665018558502, "learning_rate": 3.104292255390734e-06, "loss": 0.7849, "step": 8477 }, { "epoch": 0.07625026419574318, "grad_norm": 0.41017961502075195, "learning_rate": 3.100303441498423e-06, "loss": 0.5345, "step": 8478 }, { "epoch": 0.07625925809338364, "grad_norm": 0.4359790086746216, "learning_rate": 3.096317022523998e-06, "loss": 0.6203, "step": 8479 }, { "epoch": 0.07626825199102409, "grad_norm": 0.6925436854362488, "learning_rate": 3.092332998903416e-06, "loss": 0.9218, "step": 8480 }, { "epoch": 0.07627724588866454, "grad_norm": 0.454117089509964, "learning_rate": 3.088351371072354e-06, "loss": 0.9113, "step": 8481 }, { "epoch": 0.07628623978630499, "grad_norm": 0.706346869468689, "learning_rate": 3.0843721394662407e-06, "loss": 0.4594, "step": 8482 }, { "epoch": 0.07629523368394545, "grad_norm": 0.8411538004875183, "learning_rate": 3.0803953045202448e-06, "loss": 0.3354, "step": 8483 }, { "epoch": 0.07630422758158589, "grad_norm": 0.3635079860687256, "learning_rate": 3.076420866669266e-06, "loss": 0.6288, "step": 8484 }, { "epoch": 0.07631322147922634, "grad_norm": 0.438228577375412, "learning_rate": 3.072448826347932e-06, "loss": 0.9821, "step": 8485 }, { "epoch": 0.0763222153768668, "grad_norm": 0.2975573241710663, "learning_rate": 3.068479183990636e-06, "loss": 1.1433, "step": 8486 }, { "epoch": 0.07633120927450725, "grad_norm": 3.8070790767669678, "learning_rate": 3.0645119400314755e-06, "loss": 0.3609, "step": 8487 }, { "epoch": 0.0763402031721477, "grad_norm": 0.36262503266334534, "learning_rate": 3.060547094904309e-06, "loss": 0.6519, "step": 8488 }, { "epoch": 0.07634919706978816, "grad_norm": 0.9845858812332153, "learning_rate": 3.056584649042729e-06, "loss": 0.0748, "step": 8489 }, { "epoch": 0.0763581909674286, "grad_norm": 0.9578110575675964, "learning_rate": 3.0526246028800637e-06, "loss": 0.3716, "step": 8490 }, { "epoch": 0.07636718486506905, "grad_norm": 0.23909087479114532, "learning_rate": 3.048666956849369e-06, "loss": 1.0601, "step": 8491 }, { "epoch": 0.0763761787627095, "grad_norm": 1.0590877532958984, "learning_rate": 3.044711711383455e-06, "loss": 0.7994, "step": 8492 }, { "epoch": 0.07638517266034996, "grad_norm": 0.6226699948310852, "learning_rate": 3.040758866914861e-06, "loss": 0.4933, "step": 8493 }, { "epoch": 0.0763941665579904, "grad_norm": 0.4473448693752289, "learning_rate": 3.036808423875859e-06, "loss": 0.8654, "step": 8494 }, { "epoch": 0.07640316045563085, "grad_norm": 0.7656102180480957, "learning_rate": 3.0328603826984658e-06, "loss": 0.7701, "step": 8495 }, { "epoch": 0.07641215435327131, "grad_norm": 0.385463684797287, "learning_rate": 3.0289147438144346e-06, "loss": 0.6217, "step": 8496 }, { "epoch": 0.07642114825091176, "grad_norm": 0.4012896418571472, "learning_rate": 3.024971507655261e-06, "loss": 1.004, "step": 8497 }, { "epoch": 0.0764301421485522, "grad_norm": 0.4511213004589081, "learning_rate": 3.021030674652156e-06, "loss": 0.3658, "step": 8498 }, { "epoch": 0.07643913604619265, "grad_norm": 0.3992658257484436, "learning_rate": 3.0170922452360973e-06, "loss": 0.7652, "step": 8499 }, { "epoch": 0.07644812994383311, "grad_norm": 1.079973816871643, "learning_rate": 3.013156219837776e-06, "loss": 0.1009, "step": 8500 }, { "epoch": 0.07645712384147356, "grad_norm": 1.3695076704025269, "learning_rate": 3.0092225988876316e-06, "loss": 0.5877, "step": 8501 }, { "epoch": 0.076466117739114, "grad_norm": 0.5195909738540649, "learning_rate": 3.005291382815842e-06, "loss": 0.6923, "step": 8502 }, { "epoch": 0.07647511163675447, "grad_norm": 0.5130648612976074, "learning_rate": 3.001362572052324e-06, "loss": 0.34, "step": 8503 }, { "epoch": 0.07648410553439491, "grad_norm": 0.7596874237060547, "learning_rate": 2.9974361670267153e-06, "loss": 0.7685, "step": 8504 }, { "epoch": 0.07649309943203536, "grad_norm": 0.46039658784866333, "learning_rate": 2.9935121681684138e-06, "loss": 1.1433, "step": 8505 }, { "epoch": 0.07650209332967582, "grad_norm": 0.262819766998291, "learning_rate": 2.9895905759065285e-06, "loss": 1.0206, "step": 8506 }, { "epoch": 0.07651108722731627, "grad_norm": 0.49264389276504517, "learning_rate": 2.9856713906699254e-06, "loss": 0.8455, "step": 8507 }, { "epoch": 0.07652008112495672, "grad_norm": 0.2990308403968811, "learning_rate": 2.9817546128872027e-06, "loss": 0.9925, "step": 8508 }, { "epoch": 0.07652907502259716, "grad_norm": 0.941319465637207, "learning_rate": 2.977840242986693e-06, "loss": 0.0453, "step": 8509 }, { "epoch": 0.07653806892023762, "grad_norm": 0.8126557469367981, "learning_rate": 2.9739282813964627e-06, "loss": 0.7956, "step": 8510 }, { "epoch": 0.07654706281787807, "grad_norm": 0.5655537843704224, "learning_rate": 2.9700187285443242e-06, "loss": 0.6567, "step": 8511 }, { "epoch": 0.07655605671551852, "grad_norm": 0.5627871751785278, "learning_rate": 2.9661115848578114e-06, "loss": 0.7227, "step": 8512 }, { "epoch": 0.07656505061315898, "grad_norm": 0.5845650434494019, "learning_rate": 2.962206850764207e-06, "loss": 0.6654, "step": 8513 }, { "epoch": 0.07657404451079942, "grad_norm": 0.8762015700340271, "learning_rate": 2.9583045266905277e-06, "loss": 0.8544, "step": 8514 }, { "epoch": 0.07658303840843987, "grad_norm": 1.6364173889160156, "learning_rate": 2.954404613063527e-06, "loss": 0.8026, "step": 8515 }, { "epoch": 0.07659203230608032, "grad_norm": 0.5556043386459351, "learning_rate": 2.9505071103097e-06, "loss": 0.6643, "step": 8516 }, { "epoch": 0.07660102620372078, "grad_norm": 0.7008141279220581, "learning_rate": 2.9466120188552547e-06, "loss": 0.5328, "step": 8517 }, { "epoch": 0.07661002010136123, "grad_norm": 0.4044754207134247, "learning_rate": 2.942719339126171e-06, "loss": 0.8907, "step": 8518 }, { "epoch": 0.07661901399900167, "grad_norm": 0.680374264717102, "learning_rate": 2.9388290715481326e-06, "loss": 0.4861, "step": 8519 }, { "epoch": 0.07662800789664213, "grad_norm": 0.66610187292099, "learning_rate": 2.9349412165465773e-06, "loss": 0.3023, "step": 8520 }, { "epoch": 0.07663700179428258, "grad_norm": 0.8133478164672852, "learning_rate": 2.931055774546676e-06, "loss": 0.6966, "step": 8521 }, { "epoch": 0.07664599569192303, "grad_norm": 0.305696964263916, "learning_rate": 2.927172745973339e-06, "loss": 1.1037, "step": 8522 }, { "epoch": 0.07665498958956349, "grad_norm": 0.4894724488258362, "learning_rate": 2.923292131251201e-06, "loss": 0.6041, "step": 8523 }, { "epoch": 0.07666398348720393, "grad_norm": 0.33512789011001587, "learning_rate": 2.9194139308046486e-06, "loss": 0.891, "step": 8524 }, { "epoch": 0.07667297738484438, "grad_norm": 0.4952883720397949, "learning_rate": 2.9155381450577863e-06, "loss": 0.5468, "step": 8525 }, { "epoch": 0.07668197128248483, "grad_norm": 1.6295322179794312, "learning_rate": 2.9116647744344694e-06, "loss": 0.0644, "step": 8526 }, { "epoch": 0.07669096518012529, "grad_norm": 0.6795964241027832, "learning_rate": 2.9077938193582844e-06, "loss": 0.5609, "step": 8527 }, { "epoch": 0.07669995907776574, "grad_norm": 0.5162917971611023, "learning_rate": 2.9039252802525535e-06, "loss": 0.8111, "step": 8528 }, { "epoch": 0.07670895297540618, "grad_norm": 0.3354063034057617, "learning_rate": 2.9000591575403303e-06, "loss": 0.9687, "step": 8529 }, { "epoch": 0.07671794687304664, "grad_norm": 0.6840280890464783, "learning_rate": 2.896195451644415e-06, "loss": 0.3791, "step": 8530 }, { "epoch": 0.07672694077068709, "grad_norm": 1.1614384651184082, "learning_rate": 2.8923341629873286e-06, "loss": 0.4843, "step": 8531 }, { "epoch": 0.07673593466832754, "grad_norm": 0.31089261174201965, "learning_rate": 2.888475291991344e-06, "loss": 0.9908, "step": 8532 }, { "epoch": 0.07674492856596798, "grad_norm": 0.4618804454803467, "learning_rate": 2.8846188390784485e-06, "loss": 0.7576, "step": 8533 }, { "epoch": 0.07675392246360845, "grad_norm": 0.603674590587616, "learning_rate": 2.8807648046703968e-06, "loss": 0.6638, "step": 8534 }, { "epoch": 0.07676291636124889, "grad_norm": 1.4137300252914429, "learning_rate": 2.876913189188643e-06, "loss": 0.2411, "step": 8535 }, { "epoch": 0.07677191025888934, "grad_norm": 0.585340142250061, "learning_rate": 2.8730639930544034e-06, "loss": 0.5327, "step": 8536 }, { "epoch": 0.0767809041565298, "grad_norm": 0.47633543610572815, "learning_rate": 2.8692172166886215e-06, "loss": 1.0903, "step": 8537 }, { "epoch": 0.07678989805417025, "grad_norm": 8.094635963439941, "learning_rate": 2.8653728605119695e-06, "loss": 0.6915, "step": 8538 }, { "epoch": 0.0767988919518107, "grad_norm": 0.4936780631542206, "learning_rate": 2.861530924944858e-06, "loss": 0.6467, "step": 8539 }, { "epoch": 0.07680788584945114, "grad_norm": 0.3276369869709015, "learning_rate": 2.8576914104074425e-06, "loss": 0.8072, "step": 8540 }, { "epoch": 0.0768168797470916, "grad_norm": 0.36500924825668335, "learning_rate": 2.8538543173196098e-06, "loss": 0.7149, "step": 8541 }, { "epoch": 0.07682587364473205, "grad_norm": 1.1243492364883423, "learning_rate": 2.850019646100968e-06, "loss": 0.3375, "step": 8542 }, { "epoch": 0.0768348675423725, "grad_norm": 0.7783758044242859, "learning_rate": 2.84618739717088e-06, "loss": 0.7309, "step": 8543 }, { "epoch": 0.07684386144001296, "grad_norm": 0.5383002758026123, "learning_rate": 2.842357570948423e-06, "loss": 0.4346, "step": 8544 }, { "epoch": 0.0768528553376534, "grad_norm": 0.358438104391098, "learning_rate": 2.838530167852435e-06, "loss": 0.942, "step": 8545 }, { "epoch": 0.07686184923529385, "grad_norm": 0.3093748986721039, "learning_rate": 2.834705188301459e-06, "loss": 0.8598, "step": 8546 }, { "epoch": 0.07687084313293431, "grad_norm": 0.47818875312805176, "learning_rate": 2.8308826327138103e-06, "loss": 0.6252, "step": 8547 }, { "epoch": 0.07687983703057476, "grad_norm": 0.9381404519081116, "learning_rate": 2.827062501507499e-06, "loss": 0.4734, "step": 8548 }, { "epoch": 0.0768888309282152, "grad_norm": 0.37420475482940674, "learning_rate": 2.823244795100302e-06, "loss": 0.6423, "step": 8549 }, { "epoch": 0.07689782482585565, "grad_norm": 0.6751505136489868, "learning_rate": 2.8194295139097048e-06, "loss": 0.7334, "step": 8550 }, { "epoch": 0.07690681872349611, "grad_norm": 0.30849480628967285, "learning_rate": 2.8156166583529568e-06, "loss": 0.811, "step": 8551 }, { "epoch": 0.07691581262113656, "grad_norm": 1.8022565841674805, "learning_rate": 2.8118062288470025e-06, "loss": 0.6129, "step": 8552 }, { "epoch": 0.076924806518777, "grad_norm": 0.3941583037376404, "learning_rate": 2.8079982258085724e-06, "loss": 0.7258, "step": 8553 }, { "epoch": 0.07693380041641747, "grad_norm": 0.42768505215644836, "learning_rate": 2.804192649654086e-06, "loss": 0.5568, "step": 8554 }, { "epoch": 0.07694279431405791, "grad_norm": 0.66610187292099, "learning_rate": 2.8003895007997274e-06, "loss": 0.608, "step": 8555 }, { "epoch": 0.07695178821169836, "grad_norm": 0.8382064700126648, "learning_rate": 2.7965887796613884e-06, "loss": 0.8806, "step": 8556 }, { "epoch": 0.07696078210933881, "grad_norm": 0.8368523716926575, "learning_rate": 2.79279048665472e-06, "loss": 0.2954, "step": 8557 }, { "epoch": 0.07696977600697927, "grad_norm": 0.5772109627723694, "learning_rate": 2.788994622195096e-06, "loss": 0.1935, "step": 8558 }, { "epoch": 0.07697876990461971, "grad_norm": 0.27692273259162903, "learning_rate": 2.785201186697628e-06, "loss": 0.9405, "step": 8559 }, { "epoch": 0.07698776380226016, "grad_norm": 1.4449456930160522, "learning_rate": 2.781410180577157e-06, "loss": 0.8149, "step": 8560 }, { "epoch": 0.07699675769990062, "grad_norm": 0.41670548915863037, "learning_rate": 2.7776216042482623e-06, "loss": 0.6101, "step": 8561 }, { "epoch": 0.07700575159754107, "grad_norm": 0.6372420191764832, "learning_rate": 2.7738354581252653e-06, "loss": 0.6274, "step": 8562 }, { "epoch": 0.07701474549518152, "grad_norm": 0.30744773149490356, "learning_rate": 2.7700517426221982e-06, "loss": 0.8684, "step": 8563 }, { "epoch": 0.07702373939282198, "grad_norm": 0.6419878005981445, "learning_rate": 2.766270458152856e-06, "loss": 0.4364, "step": 8564 }, { "epoch": 0.07703273329046242, "grad_norm": 0.4066171944141388, "learning_rate": 2.7624916051307405e-06, "loss": 0.5974, "step": 8565 }, { "epoch": 0.07704172718810287, "grad_norm": 2.229247808456421, "learning_rate": 2.7587151839691184e-06, "loss": 0.8944, "step": 8566 }, { "epoch": 0.07705072108574332, "grad_norm": 1.0534765720367432, "learning_rate": 2.7549411950809566e-06, "loss": 0.842, "step": 8567 }, { "epoch": 0.07705971498338378, "grad_norm": 0.6661908626556396, "learning_rate": 2.751169638878989e-06, "loss": 0.6493, "step": 8568 }, { "epoch": 0.07706870888102423, "grad_norm": 0.9339955449104309, "learning_rate": 2.7474005157756526e-06, "loss": 0.8197, "step": 8569 }, { "epoch": 0.07707770277866467, "grad_norm": 1.0145213603973389, "learning_rate": 2.743633826183145e-06, "loss": 0.7462, "step": 8570 }, { "epoch": 0.07708669667630513, "grad_norm": 0.4148733615875244, "learning_rate": 2.7398695705133703e-06, "loss": 0.9503, "step": 8571 }, { "epoch": 0.07709569057394558, "grad_norm": 0.5115062594413757, "learning_rate": 2.736107749178002e-06, "loss": 0.3726, "step": 8572 }, { "epoch": 0.07710468447158603, "grad_norm": 0.640799343585968, "learning_rate": 2.7323483625884134e-06, "loss": 0.5884, "step": 8573 }, { "epoch": 0.07711367836922647, "grad_norm": 0.4067390263080597, "learning_rate": 2.7285914111557315e-06, "loss": 0.8393, "step": 8574 }, { "epoch": 0.07712267226686693, "grad_norm": 0.535120964050293, "learning_rate": 2.7248368952908053e-06, "loss": 0.2985, "step": 8575 }, { "epoch": 0.07713166616450738, "grad_norm": 0.4372742772102356, "learning_rate": 2.721084815404229e-06, "loss": 0.6394, "step": 8576 }, { "epoch": 0.07714066006214783, "grad_norm": 1.7682397365570068, "learning_rate": 2.717335171906313e-06, "loss": 0.1905, "step": 8577 }, { "epoch": 0.07714965395978829, "grad_norm": 0.6072855591773987, "learning_rate": 2.7135879652071318e-06, "loss": 0.5386, "step": 8578 }, { "epoch": 0.07715864785742874, "grad_norm": 0.40039709210395813, "learning_rate": 2.709843195716455e-06, "loss": 1.0332, "step": 8579 }, { "epoch": 0.07716764175506918, "grad_norm": 0.31454160809516907, "learning_rate": 2.706100863843822e-06, "loss": 0.8685, "step": 8580 }, { "epoch": 0.07717663565270963, "grad_norm": 0.44854339957237244, "learning_rate": 2.7023609699984743e-06, "loss": 0.8638, "step": 8581 }, { "epoch": 0.07718562955035009, "grad_norm": 0.6786612272262573, "learning_rate": 2.6986235145894073e-06, "loss": 0.8052, "step": 8582 }, { "epoch": 0.07719462344799054, "grad_norm": 0.9006754755973816, "learning_rate": 2.694888498025347e-06, "loss": 0.5299, "step": 8583 }, { "epoch": 0.07720361734563098, "grad_norm": 0.6229459643363953, "learning_rate": 2.691155920714741e-06, "loss": 0.6245, "step": 8584 }, { "epoch": 0.07721261124327145, "grad_norm": 0.7532371282577515, "learning_rate": 2.6874257830657805e-06, "loss": 0.3693, "step": 8585 }, { "epoch": 0.07722160514091189, "grad_norm": 0.46789681911468506, "learning_rate": 2.6836980854863942e-06, "loss": 0.8253, "step": 8586 }, { "epoch": 0.07723059903855234, "grad_norm": 0.4800349473953247, "learning_rate": 2.6799728283842335e-06, "loss": 0.6348, "step": 8587 }, { "epoch": 0.0772395929361928, "grad_norm": 0.7143253087997437, "learning_rate": 2.676250012166684e-06, "loss": 0.4376, "step": 8588 }, { "epoch": 0.07724858683383325, "grad_norm": 0.3939915895462036, "learning_rate": 2.6725296372408755e-06, "loss": 0.8531, "step": 8589 }, { "epoch": 0.0772575807314737, "grad_norm": 0.8626996278762817, "learning_rate": 2.6688117040136464e-06, "loss": 0.6313, "step": 8590 }, { "epoch": 0.07726657462911414, "grad_norm": 0.5403180122375488, "learning_rate": 2.6650962128916046e-06, "loss": 0.7381, "step": 8591 }, { "epoch": 0.0772755685267546, "grad_norm": 0.5375181436538696, "learning_rate": 2.661383164281056e-06, "loss": 0.562, "step": 8592 }, { "epoch": 0.07728456242439505, "grad_norm": 0.8148677349090576, "learning_rate": 2.6576725585880647e-06, "loss": 0.5275, "step": 8593 }, { "epoch": 0.0772935563220355, "grad_norm": 0.4412413239479065, "learning_rate": 2.6539643962184057e-06, "loss": 0.817, "step": 8594 }, { "epoch": 0.07730255021967596, "grad_norm": 0.4643562436103821, "learning_rate": 2.6502586775776076e-06, "loss": 0.8323, "step": 8595 }, { "epoch": 0.0773115441173164, "grad_norm": 0.6177836060523987, "learning_rate": 2.646555403070916e-06, "loss": 0.6921, "step": 8596 }, { "epoch": 0.07732053801495685, "grad_norm": 0.5770148634910583, "learning_rate": 2.642854573103315e-06, "loss": 0.35, "step": 8597 }, { "epoch": 0.0773295319125973, "grad_norm": 0.2423011064529419, "learning_rate": 2.6391561880795253e-06, "loss": 1.0309, "step": 8598 }, { "epoch": 0.07733852581023776, "grad_norm": 0.24904339015483856, "learning_rate": 2.635460248404001e-06, "loss": 1.0483, "step": 8599 }, { "epoch": 0.0773475197078782, "grad_norm": 0.6540228128433228, "learning_rate": 2.6317667544809134e-06, "loss": 0.4767, "step": 8600 }, { "epoch": 0.07735651360551865, "grad_norm": 0.7922264337539673, "learning_rate": 2.6280757067141894e-06, "loss": 0.6034, "step": 8601 }, { "epoch": 0.07736550750315911, "grad_norm": 0.7704839706420898, "learning_rate": 2.624387105507464e-06, "loss": 0.8906, "step": 8602 }, { "epoch": 0.07737450140079956, "grad_norm": 0.8293822407722473, "learning_rate": 2.620700951264127e-06, "loss": 0.4881, "step": 8603 }, { "epoch": 0.07738349529844, "grad_norm": 0.49525460600852966, "learning_rate": 2.6170172443872846e-06, "loss": 0.6604, "step": 8604 }, { "epoch": 0.07739248919608047, "grad_norm": 0.5284490585327148, "learning_rate": 2.6133359852797886e-06, "loss": 0.5224, "step": 8605 }, { "epoch": 0.07740148309372091, "grad_norm": 0.7532519102096558, "learning_rate": 2.609657174344207e-06, "loss": 0.505, "step": 8606 }, { "epoch": 0.07741047699136136, "grad_norm": 0.3464057445526123, "learning_rate": 2.6059808119828526e-06, "loss": 0.8015, "step": 8607 }, { "epoch": 0.0774194708890018, "grad_norm": 0.29089316725730896, "learning_rate": 2.6023068985977756e-06, "loss": 0.639, "step": 8608 }, { "epoch": 0.07742846478664227, "grad_norm": 2.459209680557251, "learning_rate": 2.5986354345907347e-06, "loss": 0.539, "step": 8609 }, { "epoch": 0.07743745868428271, "grad_norm": 0.6833396553993225, "learning_rate": 2.5949664203632428e-06, "loss": 0.9572, "step": 8610 }, { "epoch": 0.07744645258192316, "grad_norm": 0.8041784763336182, "learning_rate": 2.591299856316537e-06, "loss": 0.2679, "step": 8611 }, { "epoch": 0.07745544647956362, "grad_norm": 0.2953178584575653, "learning_rate": 2.5876357428515937e-06, "loss": 0.872, "step": 8612 }, { "epoch": 0.07746444037720407, "grad_norm": 0.3220933973789215, "learning_rate": 2.5839740803691032e-06, "loss": 1.1992, "step": 8613 }, { "epoch": 0.07747343427484452, "grad_norm": 0.3504926860332489, "learning_rate": 2.5803148692695125e-06, "loss": 0.7873, "step": 8614 }, { "epoch": 0.07748242817248496, "grad_norm": 0.591290295124054, "learning_rate": 2.576658109952973e-06, "loss": 0.4186, "step": 8615 }, { "epoch": 0.07749142207012542, "grad_norm": 0.40589743852615356, "learning_rate": 2.5730038028193902e-06, "loss": 0.7353, "step": 8616 }, { "epoch": 0.07750041596776587, "grad_norm": 0.6012634038925171, "learning_rate": 2.569351948268392e-06, "loss": 0.7107, "step": 8617 }, { "epoch": 0.07750940986540632, "grad_norm": 0.38505303859710693, "learning_rate": 2.5657025466993472e-06, "loss": 0.7154, "step": 8618 }, { "epoch": 0.07751840376304678, "grad_norm": 0.670040488243103, "learning_rate": 2.562055598511337e-06, "loss": 0.3611, "step": 8619 }, { "epoch": 0.07752739766068723, "grad_norm": 0.46881213784217834, "learning_rate": 2.558411104103198e-06, "loss": 0.6005, "step": 8620 }, { "epoch": 0.07753639155832767, "grad_norm": 0.7494550347328186, "learning_rate": 2.5547690638734746e-06, "loss": 0.9977, "step": 8621 }, { "epoch": 0.07754538545596812, "grad_norm": 0.48401662707328796, "learning_rate": 2.55112947822046e-06, "loss": 0.862, "step": 8622 }, { "epoch": 0.07755437935360858, "grad_norm": 0.6119019985198975, "learning_rate": 2.547492347542174e-06, "loss": 0.7764, "step": 8623 }, { "epoch": 0.07756337325124903, "grad_norm": 0.3099628984928131, "learning_rate": 2.5438576722363714e-06, "loss": 0.8655, "step": 8624 }, { "epoch": 0.07757236714888947, "grad_norm": 1.4634655714035034, "learning_rate": 2.5402254527005287e-06, "loss": 0.0662, "step": 8625 }, { "epoch": 0.07758136104652993, "grad_norm": 0.5882646441459656, "learning_rate": 2.5365956893318666e-06, "loss": 0.4355, "step": 8626 }, { "epoch": 0.07759035494417038, "grad_norm": 0.504167377948761, "learning_rate": 2.5329683825273233e-06, "loss": 0.7712, "step": 8627 }, { "epoch": 0.07759934884181083, "grad_norm": 0.42697760462760925, "learning_rate": 2.5293435326835785e-06, "loss": 0.7162, "step": 8628 }, { "epoch": 0.07760834273945129, "grad_norm": 0.4423060715198517, "learning_rate": 2.5257211401970405e-06, "loss": 0.821, "step": 8629 }, { "epoch": 0.07761733663709174, "grad_norm": 0.6991807818412781, "learning_rate": 2.522101205463853e-06, "loss": 0.4198, "step": 8630 }, { "epoch": 0.07762633053473218, "grad_norm": 0.7001716494560242, "learning_rate": 2.518483728879878e-06, "loss": 0.6285, "step": 8631 }, { "epoch": 0.07763532443237263, "grad_norm": 0.3174615502357483, "learning_rate": 2.514868710840723e-06, "loss": 0.9347, "step": 8632 }, { "epoch": 0.07764431833001309, "grad_norm": 0.3296900987625122, "learning_rate": 2.5112561517417254e-06, "loss": 0.7466, "step": 8633 }, { "epoch": 0.07765331222765354, "grad_norm": 0.5334970355033875, "learning_rate": 2.5076460519779383e-06, "loss": 0.9324, "step": 8634 }, { "epoch": 0.07766230612529398, "grad_norm": 0.5461183190345764, "learning_rate": 2.5040384119441594e-06, "loss": 0.4366, "step": 8635 }, { "epoch": 0.07767130002293444, "grad_norm": 0.4510369300842285, "learning_rate": 2.500433232034921e-06, "loss": 0.783, "step": 8636 }, { "epoch": 0.07768029392057489, "grad_norm": 0.6756757497787476, "learning_rate": 2.4968305126444795e-06, "loss": 0.6549, "step": 8637 }, { "epoch": 0.07768928781821534, "grad_norm": 0.838635265827179, "learning_rate": 2.493230254166817e-06, "loss": 0.0838, "step": 8638 }, { "epoch": 0.07769828171585579, "grad_norm": 0.7308254241943359, "learning_rate": 2.489632456995661e-06, "loss": 0.6391, "step": 8639 }, { "epoch": 0.07770727561349625, "grad_norm": 0.9165297150611877, "learning_rate": 2.4860371215244484e-06, "loss": 0.7462, "step": 8640 }, { "epoch": 0.07771626951113669, "grad_norm": 0.4444500505924225, "learning_rate": 2.482444248146368e-06, "loss": 1.0142, "step": 8641 }, { "epoch": 0.07772526340877714, "grad_norm": 0.3520982265472412, "learning_rate": 2.4788538372543304e-06, "loss": 0.6923, "step": 8642 }, { "epoch": 0.0777342573064176, "grad_norm": 0.7470948696136475, "learning_rate": 2.4752658892409826e-06, "loss": 0.6841, "step": 8643 }, { "epoch": 0.07774325120405805, "grad_norm": 0.4667431116104126, "learning_rate": 2.4716804044986856e-06, "loss": 0.5551, "step": 8644 }, { "epoch": 0.0777522451016985, "grad_norm": 1.0233113765716553, "learning_rate": 2.4680973834195516e-06, "loss": 0.3514, "step": 8645 }, { "epoch": 0.07776123899933896, "grad_norm": 0.5452296733856201, "learning_rate": 2.4645168263954104e-06, "loss": 1.1144, "step": 8646 }, { "epoch": 0.0777702328969794, "grad_norm": 0.8707544803619385, "learning_rate": 2.4609387338178245e-06, "loss": 0.6139, "step": 8647 }, { "epoch": 0.07777922679461985, "grad_norm": 0.914620041847229, "learning_rate": 2.457363106078095e-06, "loss": 0.7078, "step": 8648 }, { "epoch": 0.0777882206922603, "grad_norm": 0.4155026078224182, "learning_rate": 2.4537899435672446e-06, "loss": 0.8314, "step": 8649 }, { "epoch": 0.07779721458990076, "grad_norm": 0.468456894159317, "learning_rate": 2.4502192466760276e-06, "loss": 0.7125, "step": 8650 }, { "epoch": 0.0778062084875412, "grad_norm": 0.43692559003829956, "learning_rate": 2.446651015794932e-06, "loss": 0.5623, "step": 8651 }, { "epoch": 0.07781520238518165, "grad_norm": 0.576742947101593, "learning_rate": 2.44308525131417e-06, "loss": 0.7739, "step": 8652 }, { "epoch": 0.07782419628282211, "grad_norm": 0.42819491028785706, "learning_rate": 2.439521953623691e-06, "loss": 0.6772, "step": 8653 }, { "epoch": 0.07783319018046256, "grad_norm": 0.6082640290260315, "learning_rate": 2.435961123113173e-06, "loss": 0.7624, "step": 8654 }, { "epoch": 0.077842184078103, "grad_norm": 0.5218389630317688, "learning_rate": 2.4324027601720257e-06, "loss": 0.563, "step": 8655 }, { "epoch": 0.07785117797574345, "grad_norm": 0.39167171716690063, "learning_rate": 2.428846865189377e-06, "loss": 0.9957, "step": 8656 }, { "epoch": 0.07786017187338391, "grad_norm": 1.5019872188568115, "learning_rate": 2.4252934385540993e-06, "loss": 0.4748, "step": 8657 }, { "epoch": 0.07786916577102436, "grad_norm": 0.5030075311660767, "learning_rate": 2.4217424806547984e-06, "loss": 0.2544, "step": 8658 }, { "epoch": 0.0778781596686648, "grad_norm": 0.5703422427177429, "learning_rate": 2.418193991879786e-06, "loss": 0.5275, "step": 8659 }, { "epoch": 0.07788715356630527, "grad_norm": 0.7638204097747803, "learning_rate": 2.414647972617129e-06, "loss": 0.3339, "step": 8660 }, { "epoch": 0.07789614746394571, "grad_norm": 0.41006818413734436, "learning_rate": 2.411104423254615e-06, "loss": 0.85, "step": 8661 }, { "epoch": 0.07790514136158616, "grad_norm": 0.5689763426780701, "learning_rate": 2.407563344179761e-06, "loss": 0.5701, "step": 8662 }, { "epoch": 0.07791413525922661, "grad_norm": 1.262579083442688, "learning_rate": 2.4040247357798084e-06, "loss": 0.1075, "step": 8663 }, { "epoch": 0.07792312915686707, "grad_norm": 0.7719507813453674, "learning_rate": 2.4004885984417443e-06, "loss": 0.354, "step": 8664 }, { "epoch": 0.07793212305450752, "grad_norm": 0.499087929725647, "learning_rate": 2.3969549325522618e-06, "loss": 0.6026, "step": 8665 }, { "epoch": 0.07794111695214796, "grad_norm": 0.9307745099067688, "learning_rate": 2.3934237384978133e-06, "loss": 0.3773, "step": 8666 }, { "epoch": 0.07795011084978842, "grad_norm": 0.3628460764884949, "learning_rate": 2.389895016664542e-06, "loss": 0.3376, "step": 8667 }, { "epoch": 0.07795910474742887, "grad_norm": 0.4854375422000885, "learning_rate": 2.386368767438371e-06, "loss": 0.6463, "step": 8668 }, { "epoch": 0.07796809864506932, "grad_norm": 0.2926967442035675, "learning_rate": 2.382844991204908e-06, "loss": 0.9887, "step": 8669 }, { "epoch": 0.07797709254270978, "grad_norm": 0.7444355487823486, "learning_rate": 2.379323688349516e-06, "loss": 0.7867, "step": 8670 }, { "epoch": 0.07798608644035022, "grad_norm": 0.36624976992607117, "learning_rate": 2.375804859257269e-06, "loss": 0.877, "step": 8671 }, { "epoch": 0.07799508033799067, "grad_norm": 0.9100755453109741, "learning_rate": 2.3722885043129943e-06, "loss": 0.807, "step": 8672 }, { "epoch": 0.07800407423563112, "grad_norm": 0.6296219825744629, "learning_rate": 2.368774623901221e-06, "loss": 0.3527, "step": 8673 }, { "epoch": 0.07801306813327158, "grad_norm": 0.31407976150512695, "learning_rate": 2.3652632184062355e-06, "loss": 0.9646, "step": 8674 }, { "epoch": 0.07802206203091203, "grad_norm": 0.8758073449134827, "learning_rate": 2.361754288212031e-06, "loss": 0.7853, "step": 8675 }, { "epoch": 0.07803105592855247, "grad_norm": 1.3268977403640747, "learning_rate": 2.3582478337023456e-06, "loss": 0.5265, "step": 8676 }, { "epoch": 0.07804004982619293, "grad_norm": 0.545272707939148, "learning_rate": 2.35474385526063e-06, "loss": 0.0618, "step": 8677 }, { "epoch": 0.07804904372383338, "grad_norm": 0.39945539832115173, "learning_rate": 2.3512423532700805e-06, "loss": 0.7635, "step": 8678 }, { "epoch": 0.07805803762147383, "grad_norm": 0.4914548099040985, "learning_rate": 2.3477433281136164e-06, "loss": 0.3552, "step": 8679 }, { "epoch": 0.07806703151911427, "grad_norm": 0.45388269424438477, "learning_rate": 2.3442467801738863e-06, "loss": 0.8448, "step": 8680 }, { "epoch": 0.07807602541675474, "grad_norm": 0.8297141194343567, "learning_rate": 2.340752709833263e-06, "loss": 0.7858, "step": 8681 }, { "epoch": 0.07808501931439518, "grad_norm": 0.9212554693222046, "learning_rate": 2.3372611174738573e-06, "loss": 0.1823, "step": 8682 }, { "epoch": 0.07809401321203563, "grad_norm": 0.43242746591567993, "learning_rate": 2.3337720034775053e-06, "loss": 0.7743, "step": 8683 }, { "epoch": 0.07810300710967609, "grad_norm": 0.8360437154769897, "learning_rate": 2.3302853682257654e-06, "loss": 0.1192, "step": 8684 }, { "epoch": 0.07811200100731654, "grad_norm": 0.36012890934944153, "learning_rate": 2.326801212099938e-06, "loss": 0.9628, "step": 8685 }, { "epoch": 0.07812099490495698, "grad_norm": 0.2858246862888336, "learning_rate": 2.3233195354810335e-06, "loss": 0.972, "step": 8686 }, { "epoch": 0.07812998880259744, "grad_norm": 0.34948408603668213, "learning_rate": 2.319840338749821e-06, "loss": 0.6546, "step": 8687 }, { "epoch": 0.07813898270023789, "grad_norm": 0.3297126889228821, "learning_rate": 2.316363622286763e-06, "loss": 0.6912, "step": 8688 }, { "epoch": 0.07814797659787834, "grad_norm": 0.4704838991165161, "learning_rate": 2.312889386472078e-06, "loss": 0.4543, "step": 8689 }, { "epoch": 0.07815697049551878, "grad_norm": 0.3174577057361603, "learning_rate": 2.309417631685698e-06, "loss": 0.9014, "step": 8690 }, { "epoch": 0.07816596439315925, "grad_norm": 0.33105960488319397, "learning_rate": 2.3059483583072973e-06, "loss": 0.7166, "step": 8691 }, { "epoch": 0.07817495829079969, "grad_norm": 0.3744484484195709, "learning_rate": 2.3024815667162526e-06, "loss": 0.8991, "step": 8692 }, { "epoch": 0.07818395218844014, "grad_norm": 0.5448487997055054, "learning_rate": 2.2990172572917077e-06, "loss": 0.5461, "step": 8693 }, { "epoch": 0.0781929460860806, "grad_norm": 0.46023714542388916, "learning_rate": 2.2955554304124995e-06, "loss": 0.8513, "step": 8694 }, { "epoch": 0.07820193998372105, "grad_norm": 0.8292648196220398, "learning_rate": 2.2920960864572212e-06, "loss": 0.9018, "step": 8695 }, { "epoch": 0.0782109338813615, "grad_norm": 0.7013826370239258, "learning_rate": 2.288639225804168e-06, "loss": 0.5203, "step": 8696 }, { "epoch": 0.07821992777900194, "grad_norm": 0.39766809344291687, "learning_rate": 2.2851848488313872e-06, "loss": 0.8787, "step": 8697 }, { "epoch": 0.0782289216766424, "grad_norm": 0.3637970983982086, "learning_rate": 2.2817329559166293e-06, "loss": 0.6997, "step": 8698 }, { "epoch": 0.07823791557428285, "grad_norm": 0.4947353005409241, "learning_rate": 2.2782835474374125e-06, "loss": 1.018, "step": 8699 }, { "epoch": 0.0782469094719233, "grad_norm": 0.4504292607307434, "learning_rate": 2.2748366237709374e-06, "loss": 1.0712, "step": 8700 }, { "epoch": 0.07825590336956376, "grad_norm": 0.6927943229675293, "learning_rate": 2.271392185294169e-06, "loss": 0.4776, "step": 8701 }, { "epoch": 0.0782648972672042, "grad_norm": 0.48379412293434143, "learning_rate": 2.2679502323837724e-06, "loss": 0.5046, "step": 8702 }, { "epoch": 0.07827389116484465, "grad_norm": 0.3233814835548401, "learning_rate": 2.264510765416164e-06, "loss": 0.6273, "step": 8703 }, { "epoch": 0.0782828850624851, "grad_norm": 0.7669180631637573, "learning_rate": 2.2610737847674784e-06, "loss": 0.6371, "step": 8704 }, { "epoch": 0.07829187896012556, "grad_norm": 0.7411910891532898, "learning_rate": 2.2576392908135707e-06, "loss": 0.5797, "step": 8705 }, { "epoch": 0.078300872857766, "grad_norm": 0.41996487975120544, "learning_rate": 2.2542072839300403e-06, "loss": 0.8559, "step": 8706 }, { "epoch": 0.07830986675540645, "grad_norm": 0.4107441008090973, "learning_rate": 2.250777764492201e-06, "loss": 0.6542, "step": 8707 }, { "epoch": 0.07831886065304691, "grad_norm": 0.4038969874382019, "learning_rate": 2.2473507328751086e-06, "loss": 0.7372, "step": 8708 }, { "epoch": 0.07832785455068736, "grad_norm": 1.1520330905914307, "learning_rate": 2.2439261894535297e-06, "loss": 0.114, "step": 8709 }, { "epoch": 0.0783368484483278, "grad_norm": 0.4764920771121979, "learning_rate": 2.24050413460197e-06, "loss": 0.7211, "step": 8710 }, { "epoch": 0.07834584234596827, "grad_norm": 0.6253266334533691, "learning_rate": 2.237084568694653e-06, "loss": 0.4765, "step": 8711 }, { "epoch": 0.07835483624360871, "grad_norm": 0.963158130645752, "learning_rate": 2.233667492105551e-06, "loss": 0.499, "step": 8712 }, { "epoch": 0.07836383014124916, "grad_norm": 1.232187271118164, "learning_rate": 2.230252905208341e-06, "loss": 0.3399, "step": 8713 }, { "epoch": 0.07837282403888961, "grad_norm": 2.160285711288452, "learning_rate": 2.226840808376443e-06, "loss": 0.4369, "step": 8714 }, { "epoch": 0.07838181793653007, "grad_norm": 0.7859990000724792, "learning_rate": 2.223431201982992e-06, "loss": 0.2412, "step": 8715 }, { "epoch": 0.07839081183417052, "grad_norm": 0.5531406402587891, "learning_rate": 2.2200240864008624e-06, "loss": 0.8363, "step": 8716 }, { "epoch": 0.07839980573181096, "grad_norm": 0.3831285238265991, "learning_rate": 2.216619462002642e-06, "loss": 1.4231, "step": 8717 }, { "epoch": 0.07840879962945142, "grad_norm": 0.4717353284358978, "learning_rate": 2.213217329160672e-06, "loss": 0.8596, "step": 8718 }, { "epoch": 0.07841779352709187, "grad_norm": 0.3810633718967438, "learning_rate": 2.209817688246993e-06, "loss": 0.6902, "step": 8719 }, { "epoch": 0.07842678742473232, "grad_norm": 0.7209190726280212, "learning_rate": 2.2064205396333886e-06, "loss": 0.2915, "step": 8720 }, { "epoch": 0.07843578132237276, "grad_norm": 0.45256122946739197, "learning_rate": 2.203025883691362e-06, "loss": 0.9536, "step": 8721 }, { "epoch": 0.07844477522001322, "grad_norm": 0.49199455976486206, "learning_rate": 2.199633720792152e-06, "loss": 0.4389, "step": 8722 }, { "epoch": 0.07845376911765367, "grad_norm": 0.7407249212265015, "learning_rate": 2.1962440513067175e-06, "loss": 0.3397, "step": 8723 }, { "epoch": 0.07846276301529412, "grad_norm": 0.4684276282787323, "learning_rate": 2.1928568756057448e-06, "loss": 0.7646, "step": 8724 }, { "epoch": 0.07847175691293458, "grad_norm": 0.6369812488555908, "learning_rate": 2.1894721940596554e-06, "loss": 0.5754, "step": 8725 }, { "epoch": 0.07848075081057503, "grad_norm": 0.36349934339523315, "learning_rate": 2.1860900070385963e-06, "loss": 1.0414, "step": 8726 }, { "epoch": 0.07848974470821547, "grad_norm": 0.31025585532188416, "learning_rate": 2.1827103149124313e-06, "loss": 0.8214, "step": 8727 }, { "epoch": 0.07849873860585593, "grad_norm": 0.9482077956199646, "learning_rate": 2.179333118050758e-06, "loss": 0.8846, "step": 8728 }, { "epoch": 0.07850773250349638, "grad_norm": 0.48153194785118103, "learning_rate": 2.1759584168229123e-06, "loss": 0.507, "step": 8729 }, { "epoch": 0.07851672640113683, "grad_norm": 0.3444720208644867, "learning_rate": 2.1725862115979322e-06, "loss": 0.6138, "step": 8730 }, { "epoch": 0.07852572029877727, "grad_norm": 2.2474148273468018, "learning_rate": 2.1692165027446032e-06, "loss": 0.3169, "step": 8731 }, { "epoch": 0.07853471419641773, "grad_norm": 0.49269065260887146, "learning_rate": 2.1658492906314334e-06, "loss": 0.3777, "step": 8732 }, { "epoch": 0.07854370809405818, "grad_norm": 0.4999443292617798, "learning_rate": 2.1624845756266584e-06, "loss": 0.5854, "step": 8733 }, { "epoch": 0.07855270199169863, "grad_norm": 0.38849255442619324, "learning_rate": 2.1591223580982285e-06, "loss": 0.744, "step": 8734 }, { "epoch": 0.07856169588933909, "grad_norm": 1.6698150634765625, "learning_rate": 2.155762638413841e-06, "loss": 0.6465, "step": 8735 }, { "epoch": 0.07857068978697954, "grad_norm": 0.4513590335845947, "learning_rate": 2.152405416940903e-06, "loss": 1.0093, "step": 8736 }, { "epoch": 0.07857968368461998, "grad_norm": 0.2873278260231018, "learning_rate": 2.149050694046559e-06, "loss": 0.7627, "step": 8737 }, { "epoch": 0.07858867758226043, "grad_norm": 0.47682657837867737, "learning_rate": 2.1456984700976733e-06, "loss": 0.7262, "step": 8738 }, { "epoch": 0.07859767147990089, "grad_norm": 0.3591189682483673, "learning_rate": 2.1423487454608455e-06, "loss": 0.8932, "step": 8739 }, { "epoch": 0.07860666537754134, "grad_norm": 0.3851357102394104, "learning_rate": 2.13900152050239e-06, "loss": 0.6691, "step": 8740 }, { "epoch": 0.07861565927518178, "grad_norm": 0.432567298412323, "learning_rate": 2.1356567955883616e-06, "loss": 0.6833, "step": 8741 }, { "epoch": 0.07862465317282225, "grad_norm": 0.6244356632232666, "learning_rate": 2.1323145710845232e-06, "loss": 1.0966, "step": 8742 }, { "epoch": 0.07863364707046269, "grad_norm": 0.6408330202102661, "learning_rate": 2.1289748473563855e-06, "loss": 0.4839, "step": 8743 }, { "epoch": 0.07864264096810314, "grad_norm": 0.45997896790504456, "learning_rate": 2.125637624769172e-06, "loss": 0.9467, "step": 8744 }, { "epoch": 0.0786516348657436, "grad_norm": 0.8297039866447449, "learning_rate": 2.1223029036878395e-06, "loss": 1.1057, "step": 8745 }, { "epoch": 0.07866062876338405, "grad_norm": 0.8187783360481262, "learning_rate": 2.118970684477062e-06, "loss": 0.3671, "step": 8746 }, { "epoch": 0.0786696226610245, "grad_norm": 0.32262736558914185, "learning_rate": 2.1156409675012535e-06, "loss": 1.0355, "step": 8747 }, { "epoch": 0.07867861655866494, "grad_norm": 0.8365736603736877, "learning_rate": 2.112313753124537e-06, "loss": 1.0341, "step": 8748 }, { "epoch": 0.0786876104563054, "grad_norm": 0.8306492567062378, "learning_rate": 2.1089890417107792e-06, "loss": 0.9595, "step": 8749 }, { "epoch": 0.07869660435394585, "grad_norm": 0.34465694427490234, "learning_rate": 2.1056668336235622e-06, "loss": 0.8037, "step": 8750 }, { "epoch": 0.0787055982515863, "grad_norm": 0.2826203405857086, "learning_rate": 2.1023471292262064e-06, "loss": 1.1904, "step": 8751 }, { "epoch": 0.07871459214922676, "grad_norm": 0.6442234516143799, "learning_rate": 2.099029928881735e-06, "loss": 0.9335, "step": 8752 }, { "epoch": 0.0787235860468672, "grad_norm": 0.4050593972206116, "learning_rate": 2.0957152329529217e-06, "loss": 1.0621, "step": 8753 }, { "epoch": 0.07873257994450765, "grad_norm": 0.24350130558013916, "learning_rate": 2.092403041802257e-06, "loss": 1.1391, "step": 8754 }, { "epoch": 0.0787415738421481, "grad_norm": 0.41263633966445923, "learning_rate": 2.089093355791952e-06, "loss": 0.9312, "step": 8755 }, { "epoch": 0.07875056773978856, "grad_norm": 0.7593871355056763, "learning_rate": 2.085786175283949e-06, "loss": 0.4388, "step": 8756 }, { "epoch": 0.078759561637429, "grad_norm": 0.3983156681060791, "learning_rate": 2.0824815006399213e-06, "loss": 0.9679, "step": 8757 }, { "epoch": 0.07876855553506945, "grad_norm": 0.49631813168525696, "learning_rate": 2.079179332221265e-06, "loss": 0.9573, "step": 8758 }, { "epoch": 0.07877754943270991, "grad_norm": 0.5746217966079712, "learning_rate": 2.075879670389089e-06, "loss": 0.3998, "step": 8759 }, { "epoch": 0.07878654333035036, "grad_norm": 0.4340786635875702, "learning_rate": 2.072582515504254e-06, "loss": 0.8249, "step": 8760 }, { "epoch": 0.0787955372279908, "grad_norm": 0.4528609812259674, "learning_rate": 2.0692878679273164e-06, "loss": 0.9705, "step": 8761 }, { "epoch": 0.07880453112563125, "grad_norm": 0.6873425841331482, "learning_rate": 2.0659957280185856e-06, "loss": 1.0969, "step": 8762 }, { "epoch": 0.07881352502327171, "grad_norm": 1.493423581123352, "learning_rate": 2.0627060961380785e-06, "loss": 0.4118, "step": 8763 }, { "epoch": 0.07882251892091216, "grad_norm": 0.5471205711364746, "learning_rate": 2.059418972645552e-06, "loss": 0.6535, "step": 8764 }, { "epoch": 0.07883151281855261, "grad_norm": 0.3593589663505554, "learning_rate": 2.0561343579004715e-06, "loss": 0.7344, "step": 8765 }, { "epoch": 0.07884050671619307, "grad_norm": 0.5632815361022949, "learning_rate": 2.052852252262047e-06, "loss": 0.8607, "step": 8766 }, { "epoch": 0.07884950061383351, "grad_norm": 0.35703009366989136, "learning_rate": 2.0495726560891936e-06, "loss": 0.7386, "step": 8767 }, { "epoch": 0.07885849451147396, "grad_norm": 0.65346759557724, "learning_rate": 2.046295569740572e-06, "loss": 0.3402, "step": 8768 }, { "epoch": 0.07886748840911442, "grad_norm": 1.750712513923645, "learning_rate": 2.0430209935745543e-06, "loss": 0.672, "step": 8769 }, { "epoch": 0.07887648230675487, "grad_norm": 0.49644869565963745, "learning_rate": 2.0397489279492506e-06, "loss": 0.7135, "step": 8770 }, { "epoch": 0.07888547620439532, "grad_norm": 0.5673873424530029, "learning_rate": 2.0364793732224803e-06, "loss": 0.595, "step": 8771 }, { "epoch": 0.07889447010203576, "grad_norm": 1.0070725679397583, "learning_rate": 2.033212329751805e-06, "loss": 0.8053, "step": 8772 }, { "epoch": 0.07890346399967622, "grad_norm": 1.721251368522644, "learning_rate": 2.0299477978944938e-06, "loss": 0.7701, "step": 8773 }, { "epoch": 0.07891245789731667, "grad_norm": 0.972339391708374, "learning_rate": 2.026685778007556e-06, "loss": 0.269, "step": 8774 }, { "epoch": 0.07892145179495712, "grad_norm": 0.6201350092887878, "learning_rate": 2.023426270447723e-06, "loss": 0.772, "step": 8775 }, { "epoch": 0.07893044569259758, "grad_norm": 0.5158722996711731, "learning_rate": 2.0201692755714514e-06, "loss": 0.683, "step": 8776 }, { "epoch": 0.07893943959023803, "grad_norm": 0.34849029779434204, "learning_rate": 2.016914793734914e-06, "loss": 0.9507, "step": 8777 }, { "epoch": 0.07894843348787847, "grad_norm": 0.7428014278411865, "learning_rate": 2.0136628252940217e-06, "loss": 0.3835, "step": 8778 }, { "epoch": 0.07895742738551892, "grad_norm": 0.5857409238815308, "learning_rate": 2.0104133706044053e-06, "loss": 0.661, "step": 8779 }, { "epoch": 0.07896642128315938, "grad_norm": 1.0130071640014648, "learning_rate": 2.007166430021415e-06, "loss": 0.5301, "step": 8780 }, { "epoch": 0.07897541518079983, "grad_norm": 0.42553481459617615, "learning_rate": 2.0039220039001354e-06, "loss": 0.6104, "step": 8781 }, { "epoch": 0.07898440907844027, "grad_norm": 1.2798820734024048, "learning_rate": 2.0006800925953722e-06, "loss": 0.2276, "step": 8782 }, { "epoch": 0.07899340297608073, "grad_norm": 0.3445480465888977, "learning_rate": 1.9974406964616583e-06, "loss": 0.6462, "step": 8783 }, { "epoch": 0.07900239687372118, "grad_norm": 0.8585885763168335, "learning_rate": 1.9942038158532407e-06, "loss": 0.3694, "step": 8784 }, { "epoch": 0.07901139077136163, "grad_norm": 0.4061851501464844, "learning_rate": 1.990969451124111e-06, "loss": 0.0749, "step": 8785 }, { "epoch": 0.07902038466900209, "grad_norm": 0.40800097584724426, "learning_rate": 1.9877376026279647e-06, "loss": 0.9258, "step": 8786 }, { "epoch": 0.07902937856664254, "grad_norm": 0.45122799277305603, "learning_rate": 1.984508270718238e-06, "loss": 1.1714, "step": 8787 }, { "epoch": 0.07903837246428298, "grad_norm": 0.518623948097229, "learning_rate": 1.981281455748077e-06, "loss": 0.7676, "step": 8788 }, { "epoch": 0.07904736636192343, "grad_norm": 0.5449726581573486, "learning_rate": 1.978057158070373e-06, "loss": 0.6136, "step": 8789 }, { "epoch": 0.07905636025956389, "grad_norm": 0.5764589905738831, "learning_rate": 1.9748353780377234e-06, "loss": 0.3863, "step": 8790 }, { "epoch": 0.07906535415720434, "grad_norm": 0.5108487010002136, "learning_rate": 1.9716161160024644e-06, "loss": 0.788, "step": 8791 }, { "epoch": 0.07907434805484478, "grad_norm": 0.5248610377311707, "learning_rate": 1.96839937231664e-06, "loss": 0.76, "step": 8792 }, { "epoch": 0.07908334195248524, "grad_norm": 0.40345197916030884, "learning_rate": 1.9651851473320355e-06, "loss": 0.6444, "step": 8793 }, { "epoch": 0.07909233585012569, "grad_norm": 0.7364603281021118, "learning_rate": 1.961973441400142e-06, "loss": 0.4697, "step": 8794 }, { "epoch": 0.07910132974776614, "grad_norm": 0.6002919673919678, "learning_rate": 1.958764254872206e-06, "loss": 0.3151, "step": 8795 }, { "epoch": 0.07911032364540659, "grad_norm": 0.26762598752975464, "learning_rate": 1.9555575880991633e-06, "loss": 0.9525, "step": 8796 }, { "epoch": 0.07911931754304705, "grad_norm": 0.3217846155166626, "learning_rate": 1.9523534414317006e-06, "loss": 0.6672, "step": 8797 }, { "epoch": 0.0791283114406875, "grad_norm": 0.477459579706192, "learning_rate": 1.9491518152202096e-06, "loss": 0.6776, "step": 8798 }, { "epoch": 0.07913730533832794, "grad_norm": 0.4429590106010437, "learning_rate": 1.9459527098148215e-06, "loss": 0.6304, "step": 8799 }, { "epoch": 0.0791462992359684, "grad_norm": 0.3071509003639221, "learning_rate": 1.9427561255653816e-06, "loss": 1.0532, "step": 8800 }, { "epoch": 0.07915529313360885, "grad_norm": 0.5679381489753723, "learning_rate": 1.9395620628214706e-06, "loss": 0.5053, "step": 8801 }, { "epoch": 0.0791642870312493, "grad_norm": 0.4065302610397339, "learning_rate": 1.9363705219323765e-06, "loss": 0.778, "step": 8802 }, { "epoch": 0.07917328092888974, "grad_norm": 1.3140183687210083, "learning_rate": 1.9331815032471277e-06, "loss": 0.4865, "step": 8803 }, { "epoch": 0.0791822748265302, "grad_norm": 0.27281296253204346, "learning_rate": 1.929995007114474e-06, "loss": 0.8819, "step": 8804 }, { "epoch": 0.07919126872417065, "grad_norm": 0.47285985946655273, "learning_rate": 1.9268110338828747e-06, "loss": 0.498, "step": 8805 }, { "epoch": 0.0792002626218111, "grad_norm": 0.5879923105239868, "learning_rate": 1.923629583900538e-06, "loss": 0.8759, "step": 8806 }, { "epoch": 0.07920925651945156, "grad_norm": 0.6396835446357727, "learning_rate": 1.920450657515366e-06, "loss": 0.3094, "step": 8807 }, { "epoch": 0.079218250417092, "grad_norm": 0.6151683926582336, "learning_rate": 1.917274255075016e-06, "loss": 0.3456, "step": 8808 }, { "epoch": 0.07922724431473245, "grad_norm": 0.34091541171073914, "learning_rate": 1.9141003769268473e-06, "loss": 0.7945, "step": 8809 }, { "epoch": 0.07923623821237291, "grad_norm": 1.7642089128494263, "learning_rate": 1.9109290234179567e-06, "loss": 0.5584, "step": 8810 }, { "epoch": 0.07924523211001336, "grad_norm": 0.9271110892295837, "learning_rate": 1.90776019489515e-06, "loss": 0.7276, "step": 8811 }, { "epoch": 0.0792542260076538, "grad_norm": 0.41185927391052246, "learning_rate": 1.9045938917049727e-06, "loss": 1.0162, "step": 8812 }, { "epoch": 0.07926321990529425, "grad_norm": 1.7987990379333496, "learning_rate": 1.9014301141936752e-06, "loss": 0.3906, "step": 8813 }, { "epoch": 0.07927221380293471, "grad_norm": 0.4072883725166321, "learning_rate": 1.8982688627072615e-06, "loss": 0.8701, "step": 8814 }, { "epoch": 0.07928120770057516, "grad_norm": 0.46238741278648376, "learning_rate": 1.895110137591427e-06, "loss": 0.9613, "step": 8815 }, { "epoch": 0.0792902015982156, "grad_norm": 0.6010200381278992, "learning_rate": 1.891953939191618e-06, "loss": 0.368, "step": 8816 }, { "epoch": 0.07929919549585607, "grad_norm": 0.689606249332428, "learning_rate": 1.8888002678529747e-06, "loss": 0.6064, "step": 8817 }, { "epoch": 0.07930818939349651, "grad_norm": 0.7319777011871338, "learning_rate": 1.8856491239203938e-06, "loss": 0.3676, "step": 8818 }, { "epoch": 0.07931718329113696, "grad_norm": 1.138349175453186, "learning_rate": 1.8825005077384632e-06, "loss": 0.6738, "step": 8819 }, { "epoch": 0.07932617718877741, "grad_norm": 0.5618024468421936, "learning_rate": 1.87935441965153e-06, "loss": 0.6473, "step": 8820 }, { "epoch": 0.07933517108641787, "grad_norm": 0.5339245796203613, "learning_rate": 1.8762108600036326e-06, "loss": 0.9167, "step": 8821 }, { "epoch": 0.07934416498405832, "grad_norm": 0.571256697177887, "learning_rate": 1.8730698291385518e-06, "loss": 0.4322, "step": 8822 }, { "epoch": 0.07935315888169876, "grad_norm": 0.5212585926055908, "learning_rate": 1.8699313273997793e-06, "loss": 0.7113, "step": 8823 }, { "epoch": 0.07936215277933922, "grad_norm": 0.5106498003005981, "learning_rate": 1.8667953551305433e-06, "loss": 0.5312, "step": 8824 }, { "epoch": 0.07937114667697967, "grad_norm": 0.7827973365783691, "learning_rate": 1.8636619126737892e-06, "loss": 0.5903, "step": 8825 }, { "epoch": 0.07938014057462012, "grad_norm": 0.45265668630599976, "learning_rate": 1.8605310003721814e-06, "loss": 0.6231, "step": 8826 }, { "epoch": 0.07938913447226058, "grad_norm": 0.32163670659065247, "learning_rate": 1.8574026185681105e-06, "loss": 0.9797, "step": 8827 }, { "epoch": 0.07939812836990102, "grad_norm": 0.47827619314193726, "learning_rate": 1.8542767676036966e-06, "loss": 0.2504, "step": 8828 }, { "epoch": 0.07940712226754147, "grad_norm": 1.3873395919799805, "learning_rate": 1.8511534478207804e-06, "loss": 0.6746, "step": 8829 }, { "epoch": 0.07941611616518192, "grad_norm": 0.45037412643432617, "learning_rate": 1.8480326595609166e-06, "loss": 0.7542, "step": 8830 }, { "epoch": 0.07942511006282238, "grad_norm": 0.4254176616668701, "learning_rate": 1.8449144031653958e-06, "loss": 0.728, "step": 8831 }, { "epoch": 0.07943410396046283, "grad_norm": 0.3107133209705353, "learning_rate": 1.841798678975215e-06, "loss": 1.0175, "step": 8832 }, { "epoch": 0.07944309785810327, "grad_norm": 0.8605257272720337, "learning_rate": 1.8386854873311181e-06, "loss": 0.2946, "step": 8833 }, { "epoch": 0.07945209175574373, "grad_norm": 0.7627164721488953, "learning_rate": 1.8355748285735525e-06, "loss": 0.528, "step": 8834 }, { "epoch": 0.07946108565338418, "grad_norm": 0.40323561429977417, "learning_rate": 1.832466703042701e-06, "loss": 1.4077, "step": 8835 }, { "epoch": 0.07947007955102463, "grad_norm": 0.7397362589836121, "learning_rate": 1.8293611110784538e-06, "loss": 0.4678, "step": 8836 }, { "epoch": 0.07947907344866507, "grad_norm": 0.3336433172225952, "learning_rate": 1.8262580530204437e-06, "loss": 0.7106, "step": 8837 }, { "epoch": 0.07948806734630554, "grad_norm": 0.6078126430511475, "learning_rate": 1.8231575292080027e-06, "loss": 0.8084, "step": 8838 }, { "epoch": 0.07949706124394598, "grad_norm": 2.053593158721924, "learning_rate": 1.8200595399802151e-06, "loss": 0.59, "step": 8839 }, { "epoch": 0.07950605514158643, "grad_norm": 0.39165928959846497, "learning_rate": 1.8169640856758651e-06, "loss": 0.5289, "step": 8840 }, { "epoch": 0.07951504903922689, "grad_norm": 0.5994317531585693, "learning_rate": 1.8138711666334685e-06, "loss": 0.5561, "step": 8841 }, { "epoch": 0.07952404293686734, "grad_norm": 0.4362727999687195, "learning_rate": 1.8107807831912598e-06, "loss": 0.4791, "step": 8842 }, { "epoch": 0.07953303683450778, "grad_norm": 0.4467857778072357, "learning_rate": 1.8076929356872025e-06, "loss": 0.9412, "step": 8843 }, { "epoch": 0.07954203073214823, "grad_norm": 0.4837181866168976, "learning_rate": 1.8046076244589732e-06, "loss": 0.7877, "step": 8844 }, { "epoch": 0.07955102462978869, "grad_norm": 1.9215998649597168, "learning_rate": 1.80152484984398e-06, "loss": 0.6357, "step": 8845 }, { "epoch": 0.07956001852742914, "grad_norm": 0.3934732973575592, "learning_rate": 1.7984446121793508e-06, "loss": 0.8804, "step": 8846 }, { "epoch": 0.07956901242506959, "grad_norm": 0.5207604169845581, "learning_rate": 1.7953669118019407e-06, "loss": 0.6466, "step": 8847 }, { "epoch": 0.07957800632271005, "grad_norm": 0.44422581791877747, "learning_rate": 1.7922917490483137e-06, "loss": 0.3849, "step": 8848 }, { "epoch": 0.07958700022035049, "grad_norm": 1.119943380355835, "learning_rate": 1.789219124254768e-06, "loss": 0.6873, "step": 8849 }, { "epoch": 0.07959599411799094, "grad_norm": 0.6467533707618713, "learning_rate": 1.7861490377573258e-06, "loss": 0.5712, "step": 8850 }, { "epoch": 0.0796049880156314, "grad_norm": 0.46528738737106323, "learning_rate": 1.7830814898917192e-06, "loss": 1.2403, "step": 8851 }, { "epoch": 0.07961398191327185, "grad_norm": 0.6962966322898865, "learning_rate": 1.7800164809934157e-06, "loss": 0.4772, "step": 8852 }, { "epoch": 0.0796229758109123, "grad_norm": 0.8949600458145142, "learning_rate": 1.7769540113975969e-06, "loss": 0.6367, "step": 8853 }, { "epoch": 0.07963196970855274, "grad_norm": 0.6146500706672668, "learning_rate": 1.7738940814391786e-06, "loss": 0.748, "step": 8854 }, { "epoch": 0.0796409636061932, "grad_norm": 0.39356163144111633, "learning_rate": 1.770836691452779e-06, "loss": 0.7934, "step": 8855 }, { "epoch": 0.07964995750383365, "grad_norm": 0.3615567684173584, "learning_rate": 1.767781841772756e-06, "loss": 0.7241, "step": 8856 }, { "epoch": 0.0796589514014741, "grad_norm": 0.4050396978855133, "learning_rate": 1.764729532733181e-06, "loss": 0.449, "step": 8857 }, { "epoch": 0.07966794529911456, "grad_norm": 0.5867005586624146, "learning_rate": 1.7616797646678479e-06, "loss": 0.5933, "step": 8858 }, { "epoch": 0.079676939196755, "grad_norm": 0.8024190664291382, "learning_rate": 1.7586325379102792e-06, "loss": 0.6977, "step": 8859 }, { "epoch": 0.07968593309439545, "grad_norm": 1.0164445638656616, "learning_rate": 1.7555878527937164e-06, "loss": 0.3873, "step": 8860 }, { "epoch": 0.0796949269920359, "grad_norm": 0.4737911820411682, "learning_rate": 1.7525457096511156e-06, "loss": 0.8527, "step": 8861 }, { "epoch": 0.07970392088967636, "grad_norm": 0.4374101459980011, "learning_rate": 1.749506108815166e-06, "loss": 0.6065, "step": 8862 }, { "epoch": 0.0797129147873168, "grad_norm": 0.6953625679016113, "learning_rate": 1.7464690506182713e-06, "loss": 0.5631, "step": 8863 }, { "epoch": 0.07972190868495725, "grad_norm": 0.40338340401649475, "learning_rate": 1.7434345353925573e-06, "loss": 0.8562, "step": 8864 }, { "epoch": 0.07973090258259771, "grad_norm": 0.4095150828361511, "learning_rate": 1.7404025634698756e-06, "loss": 0.6995, "step": 8865 }, { "epoch": 0.07973989648023816, "grad_norm": 0.39408713579177856, "learning_rate": 1.7373731351818052e-06, "loss": 0.5579, "step": 8866 }, { "epoch": 0.0797488903778786, "grad_norm": 0.7910534143447876, "learning_rate": 1.7343462508596308e-06, "loss": 0.4064, "step": 8867 }, { "epoch": 0.07975788427551907, "grad_norm": 0.3200486898422241, "learning_rate": 1.7313219108343742e-06, "loss": 0.8597, "step": 8868 }, { "epoch": 0.07976687817315951, "grad_norm": 0.3838171064853668, "learning_rate": 1.728300115436765e-06, "loss": 0.613, "step": 8869 }, { "epoch": 0.07977587207079996, "grad_norm": 0.3630114197731018, "learning_rate": 1.7252808649972668e-06, "loss": 1.0141, "step": 8870 }, { "epoch": 0.07978486596844041, "grad_norm": 0.4091061055660248, "learning_rate": 1.7222641598460598e-06, "loss": 0.8251, "step": 8871 }, { "epoch": 0.07979385986608087, "grad_norm": 0.44848117232322693, "learning_rate": 1.7192500003130525e-06, "loss": 0.8514, "step": 8872 }, { "epoch": 0.07980285376372132, "grad_norm": 0.4283837080001831, "learning_rate": 1.716238386727856e-06, "loss": 1.1019, "step": 8873 }, { "epoch": 0.07981184766136176, "grad_norm": 0.8287393450737, "learning_rate": 1.7132293194198263e-06, "loss": 0.3176, "step": 8874 }, { "epoch": 0.07982084155900222, "grad_norm": 1.9533337354660034, "learning_rate": 1.710222798718028e-06, "loss": 0.3916, "step": 8875 }, { "epoch": 0.07982983545664267, "grad_norm": 0.37852829694747925, "learning_rate": 1.7072188249512456e-06, "loss": 1.0076, "step": 8876 }, { "epoch": 0.07983882935428312, "grad_norm": 0.5142437815666199, "learning_rate": 1.7042173984479937e-06, "loss": 0.7266, "step": 8877 }, { "epoch": 0.07984782325192356, "grad_norm": 0.30901697278022766, "learning_rate": 1.7012185195365016e-06, "loss": 1.1176, "step": 8878 }, { "epoch": 0.07985681714956402, "grad_norm": 0.4121398329734802, "learning_rate": 1.6982221885447264e-06, "loss": 0.6426, "step": 8879 }, { "epoch": 0.07986581104720447, "grad_norm": 1.3558603525161743, "learning_rate": 1.6952284058003366e-06, "loss": 0.6862, "step": 8880 }, { "epoch": 0.07987480494484492, "grad_norm": 0.7362377047538757, "learning_rate": 1.692237171630734e-06, "loss": 0.744, "step": 8881 }, { "epoch": 0.07988379884248538, "grad_norm": 0.5283975005149841, "learning_rate": 1.6892484863630264e-06, "loss": 0.5281, "step": 8882 }, { "epoch": 0.07989279274012583, "grad_norm": 0.32479605078697205, "learning_rate": 1.6862623503240581e-06, "loss": 0.9999, "step": 8883 }, { "epoch": 0.07990178663776627, "grad_norm": 0.7495318651199341, "learning_rate": 1.6832787638403846e-06, "loss": 0.3662, "step": 8884 }, { "epoch": 0.07991078053540672, "grad_norm": 0.3797708749771118, "learning_rate": 1.6802977272382974e-06, "loss": 0.8965, "step": 8885 }, { "epoch": 0.07991977443304718, "grad_norm": 0.9264017343521118, "learning_rate": 1.6773192408437832e-06, "loss": 0.2137, "step": 8886 }, { "epoch": 0.07992876833068763, "grad_norm": 0.6109157204627991, "learning_rate": 1.6743433049825785e-06, "loss": 0.6271, "step": 8887 }, { "epoch": 0.07993776222832807, "grad_norm": 0.5997686982154846, "learning_rate": 1.6713699199801148e-06, "loss": 0.5231, "step": 8888 }, { "epoch": 0.07994675612596853, "grad_norm": 1.1522586345672607, "learning_rate": 1.6683990861615623e-06, "loss": 0.5686, "step": 8889 }, { "epoch": 0.07995575002360898, "grad_norm": 0.5697706341743469, "learning_rate": 1.6654308038518058e-06, "loss": 0.6971, "step": 8890 }, { "epoch": 0.07996474392124943, "grad_norm": 0.29194942116737366, "learning_rate": 1.6624650733754605e-06, "loss": 1.244, "step": 8891 }, { "epoch": 0.07997373781888989, "grad_norm": 0.5206452012062073, "learning_rate": 1.659501895056842e-06, "loss": 0.5236, "step": 8892 }, { "epoch": 0.07998273171653034, "grad_norm": 0.6683102250099182, "learning_rate": 1.6565412692200077e-06, "loss": 0.4194, "step": 8893 }, { "epoch": 0.07999172561417078, "grad_norm": 0.3763081133365631, "learning_rate": 1.653583196188721e-06, "loss": 0.6426, "step": 8894 }, { "epoch": 0.08000071951181123, "grad_norm": 1.1877195835113525, "learning_rate": 1.650627676286473e-06, "loss": 0.3328, "step": 8895 }, { "epoch": 0.08000971340945169, "grad_norm": 0.4719311594963074, "learning_rate": 1.6476747098364804e-06, "loss": 0.6651, "step": 8896 }, { "epoch": 0.08001870730709214, "grad_norm": 0.3890596628189087, "learning_rate": 1.6447242971616732e-06, "loss": 0.7601, "step": 8897 }, { "epoch": 0.08002770120473258, "grad_norm": 0.5586602091789246, "learning_rate": 1.6417764385846996e-06, "loss": 0.8464, "step": 8898 }, { "epoch": 0.08003669510237305, "grad_norm": 0.3656972646713257, "learning_rate": 1.6388311344279373e-06, "loss": 0.737, "step": 8899 }, { "epoch": 0.08004568900001349, "grad_norm": 0.7419609427452087, "learning_rate": 1.6358883850134816e-06, "loss": 0.6325, "step": 8900 }, { "epoch": 0.08005468289765394, "grad_norm": 0.39677900075912476, "learning_rate": 1.6329481906631416e-06, "loss": 0.9179, "step": 8901 }, { "epoch": 0.08006367679529439, "grad_norm": 0.5060518383979797, "learning_rate": 1.6300105516984548e-06, "loss": 0.5187, "step": 8902 }, { "epoch": 0.08007267069293485, "grad_norm": 0.47510793805122375, "learning_rate": 1.627075468440678e-06, "loss": 0.8243, "step": 8903 }, { "epoch": 0.0800816645905753, "grad_norm": 0.3329986035823822, "learning_rate": 1.6241429412107906e-06, "loss": 0.1995, "step": 8904 }, { "epoch": 0.08009065848821574, "grad_norm": 0.496293842792511, "learning_rate": 1.6212129703294832e-06, "loss": 0.7467, "step": 8905 }, { "epoch": 0.0800996523858562, "grad_norm": 0.302059531211853, "learning_rate": 1.61828555611718e-06, "loss": 0.725, "step": 8906 }, { "epoch": 0.08010864628349665, "grad_norm": 0.4000271260738373, "learning_rate": 1.6153606988940085e-06, "loss": 0.8093, "step": 8907 }, { "epoch": 0.0801176401811371, "grad_norm": 0.8542802929878235, "learning_rate": 1.6124383989798403e-06, "loss": 0.3326, "step": 8908 }, { "epoch": 0.08012663407877756, "grad_norm": 0.8674601316452026, "learning_rate": 1.6095186566942338e-06, "loss": 0.5301, "step": 8909 }, { "epoch": 0.080135627976418, "grad_norm": 0.4991949200630188, "learning_rate": 1.6066014723565116e-06, "loss": 0.329, "step": 8910 }, { "epoch": 0.08014462187405845, "grad_norm": 0.5541704297065735, "learning_rate": 1.6036868462856764e-06, "loss": 0.7094, "step": 8911 }, { "epoch": 0.0801536157716989, "grad_norm": 0.711337685585022, "learning_rate": 1.6007747788004734e-06, "loss": 0.539, "step": 8912 }, { "epoch": 0.08016260966933936, "grad_norm": 0.6276490688323975, "learning_rate": 1.597865270219359e-06, "loss": 0.5516, "step": 8913 }, { "epoch": 0.0801716035669798, "grad_norm": 0.6514056921005249, "learning_rate": 1.5949583208605151e-06, "loss": 0.5361, "step": 8914 }, { "epoch": 0.08018059746462025, "grad_norm": 0.6820099353790283, "learning_rate": 1.5920539310418341e-06, "loss": 0.8067, "step": 8915 }, { "epoch": 0.08018959136226071, "grad_norm": 0.34045693278312683, "learning_rate": 1.5891521010809513e-06, "loss": 0.9053, "step": 8916 }, { "epoch": 0.08019858525990116, "grad_norm": 0.6842029690742493, "learning_rate": 1.586252831295193e-06, "loss": 0.8711, "step": 8917 }, { "epoch": 0.0802075791575416, "grad_norm": 0.4009655714035034, "learning_rate": 1.5833561220016253e-06, "loss": 0.9537, "step": 8918 }, { "epoch": 0.08021657305518205, "grad_norm": 0.468894898891449, "learning_rate": 1.5804619735170223e-06, "loss": 1.0208, "step": 8919 }, { "epoch": 0.08022556695282251, "grad_norm": 0.7006081938743591, "learning_rate": 1.5775703861578866e-06, "loss": 0.4677, "step": 8920 }, { "epoch": 0.08023456085046296, "grad_norm": 0.34184712171554565, "learning_rate": 1.5746813602404376e-06, "loss": 0.254, "step": 8921 }, { "epoch": 0.08024355474810341, "grad_norm": 0.346100389957428, "learning_rate": 1.5717948960806223e-06, "loss": 0.8335, "step": 8922 }, { "epoch": 0.08025254864574387, "grad_norm": 0.42610907554626465, "learning_rate": 1.5689109939940883e-06, "loss": 0.6245, "step": 8923 }, { "epoch": 0.08026154254338431, "grad_norm": 0.6334729790687561, "learning_rate": 1.5660296542962194e-06, "loss": 0.4551, "step": 8924 }, { "epoch": 0.08027053644102476, "grad_norm": 0.49555885791778564, "learning_rate": 1.5631508773021165e-06, "loss": 0.7719, "step": 8925 }, { "epoch": 0.08027953033866522, "grad_norm": 0.4542671740055084, "learning_rate": 1.5602746633265969e-06, "loss": 0.6627, "step": 8926 }, { "epoch": 0.08028852423630567, "grad_norm": 0.5209537148475647, "learning_rate": 1.557401012684201e-06, "loss": 0.6203, "step": 8927 }, { "epoch": 0.08029751813394612, "grad_norm": 0.3393052816390991, "learning_rate": 1.554529925689177e-06, "loss": 0.9827, "step": 8928 }, { "epoch": 0.08030651203158656, "grad_norm": 0.8123698234558105, "learning_rate": 1.5516614026555187e-06, "loss": 0.9494, "step": 8929 }, { "epoch": 0.08031550592922702, "grad_norm": 0.5290579199790955, "learning_rate": 1.5487954438969139e-06, "loss": 0.4937, "step": 8930 }, { "epoch": 0.08032449982686747, "grad_norm": 0.5106571912765503, "learning_rate": 1.545932049726781e-06, "loss": 0.6989, "step": 8931 }, { "epoch": 0.08033349372450792, "grad_norm": 0.3503676950931549, "learning_rate": 1.543071220458256e-06, "loss": 0.5798, "step": 8932 }, { "epoch": 0.08034248762214838, "grad_norm": 0.42336031794548035, "learning_rate": 1.5402129564041968e-06, "loss": 0.6798, "step": 8933 }, { "epoch": 0.08035148151978883, "grad_norm": 0.41219305992126465, "learning_rate": 1.53735725787717e-06, "loss": 0.7517, "step": 8934 }, { "epoch": 0.08036047541742927, "grad_norm": 0.34058526158332825, "learning_rate": 1.5345041251894871e-06, "loss": 1.0667, "step": 8935 }, { "epoch": 0.08036946931506972, "grad_norm": 1.3343008756637573, "learning_rate": 1.5316535586531483e-06, "loss": 0.8225, "step": 8936 }, { "epoch": 0.08037846321271018, "grad_norm": 1.3016164302825928, "learning_rate": 1.528805558579896e-06, "loss": 0.6818, "step": 8937 }, { "epoch": 0.08038745711035063, "grad_norm": 0.4113394618034363, "learning_rate": 1.5259601252811755e-06, "loss": 0.8788, "step": 8938 }, { "epoch": 0.08039645100799107, "grad_norm": 1.0209956169128418, "learning_rate": 1.5231172590681657e-06, "loss": 0.5635, "step": 8939 }, { "epoch": 0.08040544490563153, "grad_norm": 0.5170311331748962, "learning_rate": 1.5202769602517515e-06, "loss": 0.9944, "step": 8940 }, { "epoch": 0.08041443880327198, "grad_norm": 0.38405725359916687, "learning_rate": 1.5174392291425505e-06, "loss": 0.6519, "step": 8941 }, { "epoch": 0.08042343270091243, "grad_norm": 0.6891795992851257, "learning_rate": 1.5146040660508897e-06, "loss": 0.5135, "step": 8942 }, { "epoch": 0.08043242659855288, "grad_norm": 0.46838894486427307, "learning_rate": 1.5117714712868215e-06, "loss": 0.7425, "step": 8943 }, { "epoch": 0.08044142049619334, "grad_norm": 0.7981269359588623, "learning_rate": 1.508941445160106e-06, "loss": 0.2916, "step": 8944 }, { "epoch": 0.08045041439383378, "grad_norm": 0.41845452785491943, "learning_rate": 1.5061139879802372e-06, "loss": 1.1801, "step": 8945 }, { "epoch": 0.08045940829147423, "grad_norm": 0.2815006375312805, "learning_rate": 1.5032891000564238e-06, "loss": 1.1957, "step": 8946 }, { "epoch": 0.08046840218911469, "grad_norm": 0.525687038898468, "learning_rate": 1.5004667816975847e-06, "loss": 0.5077, "step": 8947 }, { "epoch": 0.08047739608675514, "grad_norm": 0.5982433557510376, "learning_rate": 1.4976470332123655e-06, "loss": 0.5026, "step": 8948 }, { "epoch": 0.08048638998439558, "grad_norm": 0.42937061190605164, "learning_rate": 1.4948298549091327e-06, "loss": 0.9141, "step": 8949 }, { "epoch": 0.08049538388203605, "grad_norm": 0.4327605664730072, "learning_rate": 1.4920152470959707e-06, "loss": 0.8758, "step": 8950 }, { "epoch": 0.08050437777967649, "grad_norm": 0.37551555037498474, "learning_rate": 1.4892032100806723e-06, "loss": 0.915, "step": 8951 }, { "epoch": 0.08051337167731694, "grad_norm": 0.3815748989582062, "learning_rate": 1.4863937441707693e-06, "loss": 0.8238, "step": 8952 }, { "epoch": 0.08052236557495739, "grad_norm": 0.38011854887008667, "learning_rate": 1.4835868496734851e-06, "loss": 0.8418, "step": 8953 }, { "epoch": 0.08053135947259785, "grad_norm": 0.5881704092025757, "learning_rate": 1.4807825268957937e-06, "loss": 1.1675, "step": 8954 }, { "epoch": 0.0805403533702383, "grad_norm": 0.5374636650085449, "learning_rate": 1.4779807761443636e-06, "loss": 0.0757, "step": 8955 }, { "epoch": 0.08054934726787874, "grad_norm": 0.697904646396637, "learning_rate": 1.4751815977255916e-06, "loss": 0.7004, "step": 8956 }, { "epoch": 0.0805583411655192, "grad_norm": 0.4668477773666382, "learning_rate": 1.4723849919455884e-06, "loss": 0.8021, "step": 8957 }, { "epoch": 0.08056733506315965, "grad_norm": 0.6848198771476746, "learning_rate": 1.4695909591101924e-06, "loss": 0.3828, "step": 8958 }, { "epoch": 0.0805763289608001, "grad_norm": 0.4854821562767029, "learning_rate": 1.4667994995249456e-06, "loss": 0.5723, "step": 8959 }, { "epoch": 0.08058532285844054, "grad_norm": 0.3896887004375458, "learning_rate": 1.4640106134951316e-06, "loss": 0.8487, "step": 8960 }, { "epoch": 0.080594316756081, "grad_norm": 0.8516278862953186, "learning_rate": 1.4612243013257255e-06, "loss": 0.4602, "step": 8961 }, { "epoch": 0.08060331065372145, "grad_norm": 1.0585672855377197, "learning_rate": 1.458440563321445e-06, "loss": 0.6757, "step": 8962 }, { "epoch": 0.0806123045513619, "grad_norm": 0.6669263243675232, "learning_rate": 1.4556593997867074e-06, "loss": 0.7134, "step": 8963 }, { "epoch": 0.08062129844900236, "grad_norm": 0.41506826877593994, "learning_rate": 1.4528808110256613e-06, "loss": 0.6528, "step": 8964 }, { "epoch": 0.0806302923466428, "grad_norm": 0.3318576514720917, "learning_rate": 1.4501047973421639e-06, "loss": 0.774, "step": 8965 }, { "epoch": 0.08063928624428325, "grad_norm": 0.6600197553634644, "learning_rate": 1.4473313590397968e-06, "loss": 1.129, "step": 8966 }, { "epoch": 0.08064828014192371, "grad_norm": 1.0866702795028687, "learning_rate": 1.4445604964218651e-06, "loss": 0.0948, "step": 8967 }, { "epoch": 0.08065727403956416, "grad_norm": 0.5384733080863953, "learning_rate": 1.4417922097913845e-06, "loss": 0.4196, "step": 8968 }, { "epoch": 0.0806662679372046, "grad_norm": 0.5057384967803955, "learning_rate": 1.4390264994510848e-06, "loss": 1.1384, "step": 8969 }, { "epoch": 0.08067526183484505, "grad_norm": 0.756625771522522, "learning_rate": 1.4362633657034248e-06, "loss": 0.4089, "step": 8970 }, { "epoch": 0.08068425573248551, "grad_norm": 0.32819101214408875, "learning_rate": 1.4335028088505786e-06, "loss": 0.7802, "step": 8971 }, { "epoch": 0.08069324963012596, "grad_norm": 0.3619702458381653, "learning_rate": 1.4307448291944303e-06, "loss": 0.9235, "step": 8972 }, { "epoch": 0.0807022435277664, "grad_norm": 0.6760005354881287, "learning_rate": 1.4279894270365912e-06, "loss": 0.3789, "step": 8973 }, { "epoch": 0.08071123742540687, "grad_norm": 0.5379469394683838, "learning_rate": 1.425236602678387e-06, "loss": 0.4793, "step": 8974 }, { "epoch": 0.08072023132304731, "grad_norm": 0.6355634331703186, "learning_rate": 1.4224863564208684e-06, "loss": 0.5167, "step": 8975 }, { "epoch": 0.08072922522068776, "grad_norm": 1.6081823110580444, "learning_rate": 1.4197386885647923e-06, "loss": 0.7424, "step": 8976 }, { "epoch": 0.08073821911832821, "grad_norm": 0.34633103013038635, "learning_rate": 1.4169935994106404e-06, "loss": 0.8082, "step": 8977 }, { "epoch": 0.08074721301596867, "grad_norm": 0.42059943079948425, "learning_rate": 1.4142510892586113e-06, "loss": 0.7365, "step": 8978 }, { "epoch": 0.08075620691360912, "grad_norm": 0.38704922795295715, "learning_rate": 1.411511158408621e-06, "loss": 0.9077, "step": 8979 }, { "epoch": 0.08076520081124956, "grad_norm": 0.48976370692253113, "learning_rate": 1.4087738071603075e-06, "loss": 0.6021, "step": 8980 }, { "epoch": 0.08077419470889002, "grad_norm": 0.35210615396499634, "learning_rate": 1.4060390358130227e-06, "loss": 1.0148, "step": 8981 }, { "epoch": 0.08078318860653047, "grad_norm": 1.241452932357788, "learning_rate": 1.4033068446658331e-06, "loss": 0.3919, "step": 8982 }, { "epoch": 0.08079218250417092, "grad_norm": 0.6980345249176025, "learning_rate": 1.4005772340175334e-06, "loss": 0.4954, "step": 8983 }, { "epoch": 0.08080117640181136, "grad_norm": 0.521925151348114, "learning_rate": 1.397850204166623e-06, "loss": 0.5575, "step": 8984 }, { "epoch": 0.08081017029945183, "grad_norm": 1.8952009677886963, "learning_rate": 1.3951257554113306e-06, "loss": 0.3242, "step": 8985 }, { "epoch": 0.08081916419709227, "grad_norm": 0.5358930826187134, "learning_rate": 1.3924038880495926e-06, "loss": 0.3569, "step": 8986 }, { "epoch": 0.08082815809473272, "grad_norm": 0.8826195597648621, "learning_rate": 1.3896846023790793e-06, "loss": 0.5749, "step": 8987 }, { "epoch": 0.08083715199237318, "grad_norm": 0.46789562702178955, "learning_rate": 1.3869678986971528e-06, "loss": 0.5177, "step": 8988 }, { "epoch": 0.08084614589001363, "grad_norm": 0.7954036593437195, "learning_rate": 1.3842537773009228e-06, "loss": 0.4786, "step": 8989 }, { "epoch": 0.08085513978765407, "grad_norm": 0.5082902908325195, "learning_rate": 1.3815422384871879e-06, "loss": 0.7969, "step": 8990 }, { "epoch": 0.08086413368529453, "grad_norm": 0.47143664956092834, "learning_rate": 1.3788332825524858e-06, "loss": 0.5561, "step": 8991 }, { "epoch": 0.08087312758293498, "grad_norm": 0.6273320317268372, "learning_rate": 1.3761269097930602e-06, "loss": 0.8077, "step": 8992 }, { "epoch": 0.08088212148057543, "grad_norm": 0.266725093126297, "learning_rate": 1.3734231205048825e-06, "loss": 0.9894, "step": 8993 }, { "epoch": 0.08089111537821587, "grad_norm": 0.6403798460960388, "learning_rate": 1.3707219149836248e-06, "loss": 1.1096, "step": 8994 }, { "epoch": 0.08090010927585634, "grad_norm": 1.0970507860183716, "learning_rate": 1.368023293524695e-06, "loss": 0.3425, "step": 8995 }, { "epoch": 0.08090910317349678, "grad_norm": 0.41804662346839905, "learning_rate": 1.3653272564232101e-06, "loss": 0.7733, "step": 8996 }, { "epoch": 0.08091809707113723, "grad_norm": 0.2749716639518738, "learning_rate": 1.3626338039739977e-06, "loss": 0.8589, "step": 8997 }, { "epoch": 0.08092709096877769, "grad_norm": 0.3450678884983063, "learning_rate": 1.3599429364716144e-06, "loss": 0.8002, "step": 8998 }, { "epoch": 0.08093608486641814, "grad_norm": 0.46200481057167053, "learning_rate": 1.3572546542103299e-06, "loss": 0.6954, "step": 8999 }, { "epoch": 0.08094507876405858, "grad_norm": 0.6943014860153198, "learning_rate": 1.3545689574841342e-06, "loss": 0.6085, "step": 9000 }, { "epoch": 0.08095407266169903, "grad_norm": 0.3418033719062805, "learning_rate": 1.3518858465867228e-06, "loss": 0.9259, "step": 9001 }, { "epoch": 0.08096306655933949, "grad_norm": 0.43177810311317444, "learning_rate": 1.3492053218115252e-06, "loss": 0.5236, "step": 9002 }, { "epoch": 0.08097206045697994, "grad_norm": 0.6926440596580505, "learning_rate": 1.3465273834516706e-06, "loss": 0.4703, "step": 9003 }, { "epoch": 0.08098105435462039, "grad_norm": 0.5908222794532776, "learning_rate": 1.3438520318000191e-06, "loss": 0.7735, "step": 9004 }, { "epoch": 0.08099004825226085, "grad_norm": 0.4324475824832916, "learning_rate": 1.3411792671491424e-06, "loss": 0.8659, "step": 9005 }, { "epoch": 0.08099904214990129, "grad_norm": 0.6321555972099304, "learning_rate": 1.3385090897913344e-06, "loss": 0.4034, "step": 9006 }, { "epoch": 0.08100803604754174, "grad_norm": 0.5479207038879395, "learning_rate": 1.3358415000185947e-06, "loss": 0.5653, "step": 9007 }, { "epoch": 0.0810170299451822, "grad_norm": 1.0047391653060913, "learning_rate": 1.333176498122654e-06, "loss": 0.3501, "step": 9008 }, { "epoch": 0.08102602384282265, "grad_norm": 0.47808071970939636, "learning_rate": 1.330514084394946e-06, "loss": 0.7069, "step": 9009 }, { "epoch": 0.0810350177404631, "grad_norm": 0.7337067723274231, "learning_rate": 1.3278542591266291e-06, "loss": 0.779, "step": 9010 }, { "epoch": 0.08104401163810354, "grad_norm": 0.3324020802974701, "learning_rate": 1.3251970226085792e-06, "loss": 0.765, "step": 9011 }, { "epoch": 0.081053005535744, "grad_norm": 0.3632524609565735, "learning_rate": 1.3225423751313942e-06, "loss": 0.8582, "step": 9012 }, { "epoch": 0.08106199943338445, "grad_norm": 2.097200393676758, "learning_rate": 1.3198903169853728e-06, "loss": 0.814, "step": 9013 }, { "epoch": 0.0810709933310249, "grad_norm": 0.5140700936317444, "learning_rate": 1.3172408484605464e-06, "loss": 0.7116, "step": 9014 }, { "epoch": 0.08107998722866536, "grad_norm": 0.3424650728702545, "learning_rate": 1.3145939698466502e-06, "loss": 0.9211, "step": 9015 }, { "epoch": 0.0810889811263058, "grad_norm": 0.2776733934879303, "learning_rate": 1.31194968143315e-06, "loss": 0.8942, "step": 9016 }, { "epoch": 0.08109797502394625, "grad_norm": 0.5030879974365234, "learning_rate": 1.3093079835092141e-06, "loss": 0.7199, "step": 9017 }, { "epoch": 0.0811069689215867, "grad_norm": 0.4418683350086212, "learning_rate": 1.3066688763637453e-06, "loss": 0.722, "step": 9018 }, { "epoch": 0.08111596281922716, "grad_norm": 0.5385442972183228, "learning_rate": 1.3040323602853427e-06, "loss": 0.5483, "step": 9019 }, { "epoch": 0.0811249567168676, "grad_norm": 0.3156927227973938, "learning_rate": 1.3013984355623315e-06, "loss": 0.7812, "step": 9020 }, { "epoch": 0.08113395061450805, "grad_norm": 0.5111982822418213, "learning_rate": 1.2987671024827646e-06, "loss": 0.763, "step": 9021 }, { "epoch": 0.08114294451214851, "grad_norm": 0.29023435711860657, "learning_rate": 1.296138361334387e-06, "loss": 1.033, "step": 9022 }, { "epoch": 0.08115193840978896, "grad_norm": 0.4310154318809509, "learning_rate": 1.2935122124046795e-06, "loss": 0.7502, "step": 9023 }, { "epoch": 0.0811609323074294, "grad_norm": 0.317172110080719, "learning_rate": 1.2908886559808348e-06, "loss": 0.9338, "step": 9024 }, { "epoch": 0.08116992620506985, "grad_norm": 1.0641940832138062, "learning_rate": 1.288267692349765e-06, "loss": 0.6179, "step": 9025 }, { "epoch": 0.08117892010271031, "grad_norm": 6.212742805480957, "learning_rate": 1.2856493217980853e-06, "loss": 0.3608, "step": 9026 }, { "epoch": 0.08118791400035076, "grad_norm": 0.2991781532764435, "learning_rate": 1.2830335446121473e-06, "loss": 0.9358, "step": 9027 }, { "epoch": 0.08119690789799121, "grad_norm": 0.44417738914489746, "learning_rate": 1.2804203610779968e-06, "loss": 0.5396, "step": 9028 }, { "epoch": 0.08120590179563167, "grad_norm": 0.48177120089530945, "learning_rate": 1.277809771481417e-06, "loss": 0.7648, "step": 9029 }, { "epoch": 0.08121489569327212, "grad_norm": 0.609722375869751, "learning_rate": 1.2752017761078927e-06, "loss": 0.4241, "step": 9030 }, { "epoch": 0.08122388959091256, "grad_norm": 0.6102551221847534, "learning_rate": 1.2725963752426379e-06, "loss": 0.46, "step": 9031 }, { "epoch": 0.08123288348855302, "grad_norm": 1.4406774044036865, "learning_rate": 1.2699935691705634e-06, "loss": 0.3674, "step": 9032 }, { "epoch": 0.08124187738619347, "grad_norm": 0.4265791177749634, "learning_rate": 1.2673933581763192e-06, "loss": 0.6685, "step": 9033 }, { "epoch": 0.08125087128383392, "grad_norm": 0.5285916924476624, "learning_rate": 1.2647957425442531e-06, "loss": 1.1511, "step": 9034 }, { "epoch": 0.08125986518147436, "grad_norm": 0.48375558853149414, "learning_rate": 1.2622007225584431e-06, "loss": 0.6743, "step": 9035 }, { "epoch": 0.08126885907911482, "grad_norm": 0.6500702500343323, "learning_rate": 1.2596082985026654e-06, "loss": 0.5229, "step": 9036 }, { "epoch": 0.08127785297675527, "grad_norm": 0.7747530937194824, "learning_rate": 1.2570184706604372e-06, "loss": 0.4668, "step": 9037 }, { "epoch": 0.08128684687439572, "grad_norm": 0.7174253463745117, "learning_rate": 1.2544312393149683e-06, "loss": 0.7023, "step": 9038 }, { "epoch": 0.08129584077203618, "grad_norm": 0.47405490279197693, "learning_rate": 1.2518466047492046e-06, "loss": 0.2888, "step": 9039 }, { "epoch": 0.08130483466967663, "grad_norm": 0.4751994013786316, "learning_rate": 1.2492645672457837e-06, "loss": 0.6666, "step": 9040 }, { "epoch": 0.08131382856731707, "grad_norm": 0.40581145882606506, "learning_rate": 1.2466851270870856e-06, "loss": 0.7912, "step": 9041 }, { "epoch": 0.08132282246495752, "grad_norm": 0.7655107378959656, "learning_rate": 1.2441082845551844e-06, "loss": 0.5117, "step": 9042 }, { "epoch": 0.08133181636259798, "grad_norm": 0.2606586217880249, "learning_rate": 1.2415340399318908e-06, "loss": 0.9478, "step": 9043 }, { "epoch": 0.08134081026023843, "grad_norm": 1.3232685327529907, "learning_rate": 1.2389623934987133e-06, "loss": 0.5385, "step": 9044 }, { "epoch": 0.08134980415787887, "grad_norm": 1.0167046785354614, "learning_rate": 1.2363933455368792e-06, "loss": 0.4813, "step": 9045 }, { "epoch": 0.08135879805551934, "grad_norm": 0.8012648224830627, "learning_rate": 1.2338268963273474e-06, "loss": 0.5767, "step": 9046 }, { "epoch": 0.08136779195315978, "grad_norm": 0.4792911112308502, "learning_rate": 1.2312630461507712e-06, "loss": 0.7148, "step": 9047 }, { "epoch": 0.08137678585080023, "grad_norm": 0.5676775574684143, "learning_rate": 1.2287017952875347e-06, "loss": 0.4428, "step": 9048 }, { "epoch": 0.08138577974844069, "grad_norm": 0.3068009614944458, "learning_rate": 1.2261431440177246e-06, "loss": 0.869, "step": 9049 }, { "epoch": 0.08139477364608114, "grad_norm": 1.0213496685028076, "learning_rate": 1.2235870926211619e-06, "loss": 0.8373, "step": 9050 }, { "epoch": 0.08140376754372158, "grad_norm": 0.38501307368278503, "learning_rate": 1.2210336413773644e-06, "loss": 1.6307, "step": 9051 }, { "epoch": 0.08141276144136203, "grad_norm": 0.31148871779441833, "learning_rate": 1.218482790565581e-06, "loss": 1.0146, "step": 9052 }, { "epoch": 0.08142175533900249, "grad_norm": 0.47292816638946533, "learning_rate": 1.215934540464761e-06, "loss": 0.8863, "step": 9053 }, { "epoch": 0.08143074923664294, "grad_norm": 0.7208238244056702, "learning_rate": 1.213388891353584e-06, "loss": 0.5423, "step": 9054 }, { "epoch": 0.08143974313428338, "grad_norm": 1.240457534790039, "learning_rate": 1.2108458435104274e-06, "loss": 1.3739, "step": 9055 }, { "epoch": 0.08144873703192385, "grad_norm": 0.48776453733444214, "learning_rate": 1.2083053972134107e-06, "loss": 0.7077, "step": 9056 }, { "epoch": 0.08145773092956429, "grad_norm": 0.7152127027511597, "learning_rate": 1.2057675527403417e-06, "loss": 0.6879, "step": 9057 }, { "epoch": 0.08146672482720474, "grad_norm": 1.0166653394699097, "learning_rate": 1.2032323103687654e-06, "loss": 0.3787, "step": 9058 }, { "epoch": 0.08147571872484519, "grad_norm": 0.2881104350090027, "learning_rate": 1.2006996703759182e-06, "loss": 0.9178, "step": 9059 }, { "epoch": 0.08148471262248565, "grad_norm": 0.39668217301368713, "learning_rate": 1.1981696330387787e-06, "loss": 0.855, "step": 9060 }, { "epoch": 0.0814937065201261, "grad_norm": 0.40623337030410767, "learning_rate": 1.1956421986340143e-06, "loss": 0.9088, "step": 9061 }, { "epoch": 0.08150270041776654, "grad_norm": 0.24898876249790192, "learning_rate": 1.1931173674380402e-06, "loss": 0.9276, "step": 9062 }, { "epoch": 0.081511694315407, "grad_norm": 0.518517255783081, "learning_rate": 1.1905951397269494e-06, "loss": 0.7358, "step": 9063 }, { "epoch": 0.08152068821304745, "grad_norm": 0.5329123139381409, "learning_rate": 1.188075515776585e-06, "loss": 0.6861, "step": 9064 }, { "epoch": 0.0815296821106879, "grad_norm": 0.6801884174346924, "learning_rate": 1.1855584958624765e-06, "loss": 0.3182, "step": 9065 }, { "epoch": 0.08153867600832834, "grad_norm": 1.2609310150146484, "learning_rate": 1.1830440802598847e-06, "loss": 0.5077, "step": 9066 }, { "epoch": 0.0815476699059688, "grad_norm": 0.6986671090126038, "learning_rate": 1.1805322692437892e-06, "loss": 0.2917, "step": 9067 }, { "epoch": 0.08155666380360925, "grad_norm": 0.3747463822364807, "learning_rate": 1.1780230630888678e-06, "loss": 0.3188, "step": 9068 }, { "epoch": 0.0815656577012497, "grad_norm": 0.6610244512557983, "learning_rate": 1.1755164620695315e-06, "loss": 0.6353, "step": 9069 }, { "epoch": 0.08157465159889016, "grad_norm": 0.8095352649688721, "learning_rate": 1.1730124664598912e-06, "loss": 0.5127, "step": 9070 }, { "epoch": 0.0815836454965306, "grad_norm": 0.9522950649261475, "learning_rate": 1.1705110765337896e-06, "loss": 0.501, "step": 9071 }, { "epoch": 0.08159263939417105, "grad_norm": 0.5818361043930054, "learning_rate": 1.1680122925647658e-06, "loss": 0.3537, "step": 9072 }, { "epoch": 0.08160163329181151, "grad_norm": 0.26472726464271545, "learning_rate": 1.165516114826093e-06, "loss": 0.9628, "step": 9073 }, { "epoch": 0.08161062718945196, "grad_norm": 1.0525561571121216, "learning_rate": 1.1630225435907333e-06, "loss": 0.3422, "step": 9074 }, { "epoch": 0.0816196210870924, "grad_norm": 1.093684434890747, "learning_rate": 1.1605315791313964e-06, "loss": 0.8826, "step": 9075 }, { "epoch": 0.08162861498473285, "grad_norm": 0.3759225010871887, "learning_rate": 1.158043221720481e-06, "loss": 0.7929, "step": 9076 }, { "epoch": 0.08163760888237331, "grad_norm": 0.40888771414756775, "learning_rate": 1.1555574716301199e-06, "loss": 0.8092, "step": 9077 }, { "epoch": 0.08164660278001376, "grad_norm": 0.6119896769523621, "learning_rate": 1.1530743291321367e-06, "loss": 0.6425, "step": 9078 }, { "epoch": 0.08165559667765421, "grad_norm": 0.4340582489967346, "learning_rate": 1.150593794498095e-06, "loss": 0.5001, "step": 9079 }, { "epoch": 0.08166459057529467, "grad_norm": 0.494864821434021, "learning_rate": 1.1481158679992555e-06, "loss": 0.4591, "step": 9080 }, { "epoch": 0.08167358447293512, "grad_norm": 0.3148362636566162, "learning_rate": 1.145640549906607e-06, "loss": 0.9919, "step": 9081 }, { "epoch": 0.08168257837057556, "grad_norm": 0.565215528011322, "learning_rate": 1.1431678404908413e-06, "loss": 0.6979, "step": 9082 }, { "epoch": 0.08169157226821601, "grad_norm": 0.8675175905227661, "learning_rate": 1.1406977400223756e-06, "loss": 0.6153, "step": 9083 }, { "epoch": 0.08170056616585647, "grad_norm": 0.4543282687664032, "learning_rate": 1.1382302487713298e-06, "loss": 0.8029, "step": 9084 }, { "epoch": 0.08170956006349692, "grad_norm": 0.30439043045043945, "learning_rate": 1.135765367007552e-06, "loss": 0.9681, "step": 9085 }, { "epoch": 0.08171855396113736, "grad_norm": 0.6999118328094482, "learning_rate": 1.133303095000593e-06, "loss": 0.4635, "step": 9086 }, { "epoch": 0.08172754785877782, "grad_norm": 0.850005567073822, "learning_rate": 1.130843433019721e-06, "loss": 0.4063, "step": 9087 }, { "epoch": 0.08173654175641827, "grad_norm": 0.7498278617858887, "learning_rate": 1.1283863813339263e-06, "loss": 0.646, "step": 9088 }, { "epoch": 0.08174553565405872, "grad_norm": 0.40980443358421326, "learning_rate": 1.1259319402119079e-06, "loss": 0.7561, "step": 9089 }, { "epoch": 0.08175452955169918, "grad_norm": 0.9036883115768433, "learning_rate": 1.1234801099220787e-06, "loss": 0.6339, "step": 9090 }, { "epoch": 0.08176352344933963, "grad_norm": 0.5086578130722046, "learning_rate": 1.1210308907325634e-06, "loss": 0.8378, "step": 9091 }, { "epoch": 0.08177251734698007, "grad_norm": 0.6839801073074341, "learning_rate": 1.1185842829112142e-06, "loss": 0.6604, "step": 9092 }, { "epoch": 0.08178151124462052, "grad_norm": 0.35440999269485474, "learning_rate": 1.1161402867255754e-06, "loss": 1.0585, "step": 9093 }, { "epoch": 0.08179050514226098, "grad_norm": 0.3268759548664093, "learning_rate": 1.1136989024429305e-06, "loss": 0.921, "step": 9094 }, { "epoch": 0.08179949903990143, "grad_norm": 0.9507691264152527, "learning_rate": 1.1112601303302605e-06, "loss": 0.641, "step": 9095 }, { "epoch": 0.08180849293754187, "grad_norm": 1.2417938709259033, "learning_rate": 1.1088239706542686e-06, "loss": 0.3892, "step": 9096 }, { "epoch": 0.08181748683518233, "grad_norm": 0.8916202783584595, "learning_rate": 1.1063904236813639e-06, "loss": 0.9631, "step": 9097 }, { "epoch": 0.08182648073282278, "grad_norm": 0.2862071692943573, "learning_rate": 1.1039594896776834e-06, "loss": 1.0159, "step": 9098 }, { "epoch": 0.08183547463046323, "grad_norm": 0.6967105269432068, "learning_rate": 1.101531168909059e-06, "loss": 0.6024, "step": 9099 }, { "epoch": 0.08184446852810368, "grad_norm": 0.649246871471405, "learning_rate": 1.0991054616410589e-06, "loss": 0.6867, "step": 9100 }, { "epoch": 0.08185346242574414, "grad_norm": 1.9768787622451782, "learning_rate": 1.0966823681389515e-06, "loss": 0.6961, "step": 9101 }, { "epoch": 0.08186245632338458, "grad_norm": 0.28240966796875, "learning_rate": 1.0942618886677219e-06, "loss": 1.186, "step": 9102 }, { "epoch": 0.08187145022102503, "grad_norm": 0.26518717408180237, "learning_rate": 1.0918440234920696e-06, "loss": 1.204, "step": 9103 }, { "epoch": 0.08188044411866549, "grad_norm": 1.159718632698059, "learning_rate": 1.0894287728764136e-06, "loss": 0.7923, "step": 9104 }, { "epoch": 0.08188943801630594, "grad_norm": 0.38733723759651184, "learning_rate": 1.087016137084873e-06, "loss": 0.8127, "step": 9105 }, { "epoch": 0.08189843191394638, "grad_norm": 0.3783365488052368, "learning_rate": 1.0846061163812954e-06, "loss": 0.7893, "step": 9106 }, { "epoch": 0.08190742581158683, "grad_norm": 0.6166955232620239, "learning_rate": 1.0821987110292365e-06, "loss": 0.4271, "step": 9107 }, { "epoch": 0.08191641970922729, "grad_norm": 0.4101099669933319, "learning_rate": 1.0797939212919717e-06, "loss": 0.6951, "step": 9108 }, { "epoch": 0.08192541360686774, "grad_norm": 0.6134801506996155, "learning_rate": 1.0773917474324741e-06, "loss": 0.5536, "step": 9109 }, { "epoch": 0.08193440750450819, "grad_norm": 0.7971243858337402, "learning_rate": 1.074992189713453e-06, "loss": 0.1018, "step": 9110 }, { "epoch": 0.08194340140214865, "grad_norm": 0.5306267142295837, "learning_rate": 1.0725952483973123e-06, "loss": 0.5315, "step": 9111 }, { "epoch": 0.0819523952997891, "grad_norm": 0.40439125895500183, "learning_rate": 1.0702009237461814e-06, "loss": 0.7715, "step": 9112 }, { "epoch": 0.08196138919742954, "grad_norm": 0.49622923135757446, "learning_rate": 1.0678092160218979e-06, "loss": 0.6033, "step": 9113 }, { "epoch": 0.08197038309507, "grad_norm": 0.45469704270362854, "learning_rate": 1.0654201254860163e-06, "loss": 0.7027, "step": 9114 }, { "epoch": 0.08197937699271045, "grad_norm": 2.664684295654297, "learning_rate": 1.0630336523998113e-06, "loss": 0.3102, "step": 9115 }, { "epoch": 0.0819883708903509, "grad_norm": 0.3513915240764618, "learning_rate": 1.0606497970242518e-06, "loss": 1.055, "step": 9116 }, { "epoch": 0.08199736478799134, "grad_norm": 0.6054510474205017, "learning_rate": 1.0582685596200403e-06, "loss": 0.4195, "step": 9117 }, { "epoch": 0.0820063586856318, "grad_norm": 0.49695634841918945, "learning_rate": 1.0558899404475824e-06, "loss": 0.6054, "step": 9118 }, { "epoch": 0.08201535258327225, "grad_norm": 0.37259218096733093, "learning_rate": 1.0535139397670008e-06, "loss": 0.5453, "step": 9119 }, { "epoch": 0.0820243464809127, "grad_norm": 0.4102901518344879, "learning_rate": 1.051140557838129e-06, "loss": 0.6915, "step": 9120 }, { "epoch": 0.08203334037855316, "grad_norm": 0.5683236122131348, "learning_rate": 1.0487697949205261e-06, "loss": 0.64, "step": 9121 }, { "epoch": 0.0820423342761936, "grad_norm": 0.3690159022808075, "learning_rate": 1.0464016512734432e-06, "loss": 1.2897, "step": 9122 }, { "epoch": 0.08205132817383405, "grad_norm": 0.7007889151573181, "learning_rate": 1.0440361271558646e-06, "loss": 0.2926, "step": 9123 }, { "epoch": 0.0820603220714745, "grad_norm": 0.37068015336990356, "learning_rate": 1.0416732228264752e-06, "loss": 0.7828, "step": 9124 }, { "epoch": 0.08206931596911496, "grad_norm": 0.5056186318397522, "learning_rate": 1.0393129385436824e-06, "loss": 0.6081, "step": 9125 }, { "epoch": 0.0820783098667554, "grad_norm": 0.6135507822036743, "learning_rate": 1.0369552745656013e-06, "loss": 0.7388, "step": 9126 }, { "epoch": 0.08208730376439585, "grad_norm": 0.8443911671638489, "learning_rate": 1.034600231150068e-06, "loss": 0.4195, "step": 9127 }, { "epoch": 0.08209629766203631, "grad_norm": 0.3158336579799652, "learning_rate": 1.0322478085546177e-06, "loss": 1.0266, "step": 9128 }, { "epoch": 0.08210529155967676, "grad_norm": 0.8675842881202698, "learning_rate": 1.0298980070365145e-06, "loss": 0.4853, "step": 9129 }, { "epoch": 0.08211428545731721, "grad_norm": 0.5056912899017334, "learning_rate": 1.0275508268527246e-06, "loss": 0.792, "step": 9130 }, { "epoch": 0.08212327935495767, "grad_norm": 0.593761146068573, "learning_rate": 1.0252062682599345e-06, "loss": 0.7218, "step": 9131 }, { "epoch": 0.08213227325259811, "grad_norm": 0.6601639986038208, "learning_rate": 1.0228643315145419e-06, "loss": 0.444, "step": 9132 }, { "epoch": 0.08214126715023856, "grad_norm": 0.4214666783809662, "learning_rate": 1.0205250168726588e-06, "loss": 0.6638, "step": 9133 }, { "epoch": 0.08215026104787901, "grad_norm": 0.5968158841133118, "learning_rate": 1.0181883245901052e-06, "loss": 0.4114, "step": 9134 }, { "epoch": 0.08215925494551947, "grad_norm": 0.822738528251648, "learning_rate": 1.0158542549224186e-06, "loss": 0.3139, "step": 9135 }, { "epoch": 0.08216824884315992, "grad_norm": 0.28844454884529114, "learning_rate": 1.013522808124856e-06, "loss": 1.3246, "step": 9136 }, { "epoch": 0.08217724274080036, "grad_norm": 5.742728233337402, "learning_rate": 1.0111939844523716e-06, "loss": 0.448, "step": 9137 }, { "epoch": 0.08218623663844082, "grad_norm": 0.4462404251098633, "learning_rate": 1.008867784159645e-06, "loss": 1.0, "step": 9138 }, { "epoch": 0.08219523053608127, "grad_norm": 0.47626709938049316, "learning_rate": 1.00654420750107e-06, "loss": 0.6591, "step": 9139 }, { "epoch": 0.08220422443372172, "grad_norm": 1.696987509727478, "learning_rate": 1.004223254730749e-06, "loss": 0.3398, "step": 9140 }, { "epoch": 0.08221321833136216, "grad_norm": 1.2664070129394531, "learning_rate": 1.0019049261024926e-06, "loss": 0.557, "step": 9141 }, { "epoch": 0.08222221222900263, "grad_norm": 0.40823426842689514, "learning_rate": 9.99589221869837e-07, "loss": 0.7228, "step": 9142 }, { "epoch": 0.08223120612664307, "grad_norm": 0.607486367225647, "learning_rate": 9.972761422860162e-07, "loss": 0.2373, "step": 9143 }, { "epoch": 0.08224020002428352, "grad_norm": 0.45933428406715393, "learning_rate": 9.949656876039882e-07, "loss": 0.8412, "step": 9144 }, { "epoch": 0.08224919392192398, "grad_norm": 0.45315074920654297, "learning_rate": 9.926578580764234e-07, "loss": 0.6314, "step": 9145 }, { "epoch": 0.08225818781956443, "grad_norm": 0.64530348777771, "learning_rate": 9.903526539557007e-07, "loss": 0.8426, "step": 9146 }, { "epoch": 0.08226718171720487, "grad_norm": 1.384777545928955, "learning_rate": 9.880500754939125e-07, "loss": 0.311, "step": 9147 }, { "epoch": 0.08227617561484533, "grad_norm": 0.8221823573112488, "learning_rate": 9.857501229428718e-07, "loss": 0.7892, "step": 9148 }, { "epoch": 0.08228516951248578, "grad_norm": 0.4651743471622467, "learning_rate": 9.834527965540852e-07, "loss": 0.7573, "step": 9149 }, { "epoch": 0.08229416341012623, "grad_norm": 0.6520513296127319, "learning_rate": 9.811580965787965e-07, "loss": 0.2533, "step": 9150 }, { "epoch": 0.08230315730776668, "grad_norm": 0.5852120518684387, "learning_rate": 9.788660232679436e-07, "loss": 0.9854, "step": 9151 }, { "epoch": 0.08231215120540714, "grad_norm": 0.6054171919822693, "learning_rate": 9.76576576872193e-07, "loss": 0.8845, "step": 9152 }, { "epoch": 0.08232114510304758, "grad_norm": 1.0668832063674927, "learning_rate": 9.742897576419052e-07, "loss": 0.4823, "step": 9153 }, { "epoch": 0.08233013900068803, "grad_norm": 0.8199167251586914, "learning_rate": 9.720055658271693e-07, "loss": 0.5323, "step": 9154 }, { "epoch": 0.08233913289832849, "grad_norm": 0.34646403789520264, "learning_rate": 9.69724001677777e-07, "loss": 0.6727, "step": 9155 }, { "epoch": 0.08234812679596894, "grad_norm": 0.595954179763794, "learning_rate": 9.674450654432398e-07, "loss": 0.7388, "step": 9156 }, { "epoch": 0.08235712069360938, "grad_norm": 2.7496538162231445, "learning_rate": 9.651687573727753e-07, "loss": 0.3889, "step": 9157 }, { "epoch": 0.08236611459124983, "grad_norm": 0.7608706951141357, "learning_rate": 9.628950777153256e-07, "loss": 0.4329, "step": 9158 }, { "epoch": 0.08237510848889029, "grad_norm": 0.3508271276950836, "learning_rate": 9.606240267195253e-07, "loss": 0.9009, "step": 9159 }, { "epoch": 0.08238410238653074, "grad_norm": 1.1989837884902954, "learning_rate": 9.5835560463374e-07, "loss": 0.6912, "step": 9160 }, { "epoch": 0.08239309628417119, "grad_norm": 0.583612859249115, "learning_rate": 9.560898117060429e-07, "loss": 0.4257, "step": 9161 }, { "epoch": 0.08240209018181165, "grad_norm": 0.6672049760818481, "learning_rate": 9.538266481842112e-07, "loss": 0.5605, "step": 9162 }, { "epoch": 0.0824110840794521, "grad_norm": 0.7026159763336182, "learning_rate": 9.51566114315744e-07, "loss": 0.7378, "step": 9163 }, { "epoch": 0.08242007797709254, "grad_norm": 0.31682977080345154, "learning_rate": 9.493082103478517e-07, "loss": 0.9361, "step": 9164 }, { "epoch": 0.08242907187473299, "grad_norm": 0.7866734862327576, "learning_rate": 9.470529365274566e-07, "loss": 0.7837, "step": 9165 }, { "epoch": 0.08243806577237345, "grad_norm": 0.9788363575935364, "learning_rate": 9.448002931011834e-07, "loss": 0.3038, "step": 9166 }, { "epoch": 0.0824470596700139, "grad_norm": 0.3337177634239197, "learning_rate": 9.425502803153907e-07, "loss": 0.9622, "step": 9167 }, { "epoch": 0.08245605356765434, "grad_norm": 0.5700885653495789, "learning_rate": 9.403028984161233e-07, "loss": 0.5529, "step": 9168 }, { "epoch": 0.0824650474652948, "grad_norm": 0.5717293620109558, "learning_rate": 9.380581476491623e-07, "loss": 0.9, "step": 9169 }, { "epoch": 0.08247404136293525, "grad_norm": 0.7907506227493286, "learning_rate": 9.358160282599809e-07, "loss": 1.0782, "step": 9170 }, { "epoch": 0.0824830352605757, "grad_norm": 0.3807474374771118, "learning_rate": 9.335765404937829e-07, "loss": 0.5253, "step": 9171 }, { "epoch": 0.08249202915821616, "grad_norm": 0.300790011882782, "learning_rate": 9.31339684595467e-07, "loss": 0.7037, "step": 9172 }, { "epoch": 0.0825010230558566, "grad_norm": 0.9354957342147827, "learning_rate": 9.291054608096622e-07, "loss": 0.3772, "step": 9173 }, { "epoch": 0.08251001695349705, "grad_norm": 0.3123225271701813, "learning_rate": 9.2687386938069e-07, "loss": 0.9429, "step": 9174 }, { "epoch": 0.0825190108511375, "grad_norm": 0.31991416215896606, "learning_rate": 9.246449105525995e-07, "loss": 0.9036, "step": 9175 }, { "epoch": 0.08252800474877796, "grad_norm": 2.076577663421631, "learning_rate": 9.224185845691402e-07, "loss": 0.4076, "step": 9176 }, { "epoch": 0.0825369986464184, "grad_norm": 0.7459611296653748, "learning_rate": 9.201948916737924e-07, "loss": 0.9854, "step": 9177 }, { "epoch": 0.08254599254405885, "grad_norm": 1.4416708946228027, "learning_rate": 9.179738321097225e-07, "loss": 0.8525, "step": 9178 }, { "epoch": 0.08255498644169931, "grad_norm": 0.3489733934402466, "learning_rate": 9.157554061198309e-07, "loss": 1.5205, "step": 9179 }, { "epoch": 0.08256398033933976, "grad_norm": 0.49355897307395935, "learning_rate": 9.135396139467151e-07, "loss": 0.4202, "step": 9180 }, { "epoch": 0.0825729742369802, "grad_norm": 0.6615161299705505, "learning_rate": 9.113264558326923e-07, "loss": 0.273, "step": 9181 }, { "epoch": 0.08258196813462065, "grad_norm": 0.8895791172981262, "learning_rate": 9.09115932019794e-07, "loss": 0.7315, "step": 9182 }, { "epoch": 0.08259096203226111, "grad_norm": 0.5396736264228821, "learning_rate": 9.069080427497573e-07, "loss": 0.7473, "step": 9183 }, { "epoch": 0.08259995592990156, "grad_norm": 0.39424487948417664, "learning_rate": 9.047027882640336e-07, "loss": 0.7068, "step": 9184 }, { "epoch": 0.08260894982754201, "grad_norm": 0.4814402759075165, "learning_rate": 9.025001688037854e-07, "loss": 0.4064, "step": 9185 }, { "epoch": 0.08261794372518247, "grad_norm": 0.6258647441864014, "learning_rate": 9.003001846098897e-07, "loss": 0.0626, "step": 9186 }, { "epoch": 0.08262693762282292, "grad_norm": 0.4915964603424072, "learning_rate": 8.981028359229316e-07, "loss": 1.0281, "step": 9187 }, { "epoch": 0.08263593152046336, "grad_norm": 0.7442896962165833, "learning_rate": 8.959081229832134e-07, "loss": 0.6478, "step": 9188 }, { "epoch": 0.08264492541810382, "grad_norm": 0.4901355803012848, "learning_rate": 8.937160460307375e-07, "loss": 0.7155, "step": 9189 }, { "epoch": 0.08265391931574427, "grad_norm": 0.7538332939147949, "learning_rate": 8.915266053052373e-07, "loss": 0.4709, "step": 9190 }, { "epoch": 0.08266291321338472, "grad_norm": 0.4684353172779083, "learning_rate": 8.893398010461351e-07, "loss": 0.6947, "step": 9191 }, { "epoch": 0.08267190711102516, "grad_norm": 0.4582403004169464, "learning_rate": 8.871556334925895e-07, "loss": 0.9016, "step": 9192 }, { "epoch": 0.08268090100866562, "grad_norm": 0.4588084816932678, "learning_rate": 8.849741028834457e-07, "loss": 0.7382, "step": 9193 }, { "epoch": 0.08268989490630607, "grad_norm": 0.9931371808052063, "learning_rate": 8.827952094572794e-07, "loss": 0.4132, "step": 9194 }, { "epoch": 0.08269888880394652, "grad_norm": 1.6629945039749146, "learning_rate": 8.806189534523668e-07, "loss": 0.4164, "step": 9195 }, { "epoch": 0.08270788270158698, "grad_norm": 1.5669410228729248, "learning_rate": 8.784453351067062e-07, "loss": 0.0613, "step": 9196 }, { "epoch": 0.08271687659922743, "grad_norm": 0.594636857509613, "learning_rate": 8.762743546579937e-07, "loss": 0.0863, "step": 9197 }, { "epoch": 0.08272587049686787, "grad_norm": 0.532166063785553, "learning_rate": 8.741060123436535e-07, "loss": 0.5057, "step": 9198 }, { "epoch": 0.08273486439450832, "grad_norm": 0.6222426891326904, "learning_rate": 8.71940308400801e-07, "loss": 0.3811, "step": 9199 }, { "epoch": 0.08274385829214878, "grad_norm": 0.9123340249061584, "learning_rate": 8.697772430662859e-07, "loss": 0.4206, "step": 9200 }, { "epoch": 0.08275285218978923, "grad_norm": 1.9408503770828247, "learning_rate": 8.676168165766469e-07, "loss": 0.3309, "step": 9201 }, { "epoch": 0.08276184608742967, "grad_norm": 0.34593838453292847, "learning_rate": 8.65459029168153e-07, "loss": 0.8499, "step": 9202 }, { "epoch": 0.08277083998507014, "grad_norm": 0.47889429330825806, "learning_rate": 8.633038810767741e-07, "loss": 0.5756, "step": 9203 }, { "epoch": 0.08277983388271058, "grad_norm": 0.6535888910293579, "learning_rate": 8.611513725381965e-07, "loss": 0.3104, "step": 9204 }, { "epoch": 0.08278882778035103, "grad_norm": 0.4675612151622772, "learning_rate": 8.590015037878068e-07, "loss": 0.7863, "step": 9205 }, { "epoch": 0.08279782167799148, "grad_norm": 0.5362908840179443, "learning_rate": 8.5685427506072e-07, "loss": 0.9065, "step": 9206 }, { "epoch": 0.08280681557563194, "grad_norm": 0.483422189950943, "learning_rate": 8.547096865917537e-07, "loss": 0.8514, "step": 9207 }, { "epoch": 0.08281580947327238, "grad_norm": 0.7542577385902405, "learning_rate": 8.525677386154312e-07, "loss": 0.5792, "step": 9208 }, { "epoch": 0.08282480337091283, "grad_norm": 0.8954369425773621, "learning_rate": 8.504284313659932e-07, "loss": 0.4953, "step": 9209 }, { "epoch": 0.08283379726855329, "grad_norm": 0.6366555094718933, "learning_rate": 8.48291765077397e-07, "loss": 1.0223, "step": 9210 }, { "epoch": 0.08284279116619374, "grad_norm": 0.925036609172821, "learning_rate": 8.461577399833026e-07, "loss": 0.0579, "step": 9211 }, { "epoch": 0.08285178506383419, "grad_norm": 0.730589747428894, "learning_rate": 8.440263563170819e-07, "loss": 0.66, "step": 9212 }, { "epoch": 0.08286077896147465, "grad_norm": 0.6607157588005066, "learning_rate": 8.418976143118206e-07, "loss": 0.6361, "step": 9213 }, { "epoch": 0.08286977285911509, "grad_norm": 0.8585137724876404, "learning_rate": 8.397715142003132e-07, "loss": 0.4071, "step": 9214 }, { "epoch": 0.08287876675675554, "grad_norm": 0.3327944576740265, "learning_rate": 8.376480562150707e-07, "loss": 0.8977, "step": 9215 }, { "epoch": 0.08288776065439599, "grad_norm": 0.44052422046661377, "learning_rate": 8.355272405883075e-07, "loss": 0.7635, "step": 9216 }, { "epoch": 0.08289675455203645, "grad_norm": 0.3267444372177124, "learning_rate": 8.334090675519601e-07, "loss": 0.8564, "step": 9217 }, { "epoch": 0.0829057484496769, "grad_norm": 0.5866419672966003, "learning_rate": 8.312935373376573e-07, "loss": 0.3844, "step": 9218 }, { "epoch": 0.08291474234731734, "grad_norm": 0.3260918855667114, "learning_rate": 8.291806501767613e-07, "loss": 1.0122, "step": 9219 }, { "epoch": 0.0829237362449578, "grad_norm": 0.7436965703964233, "learning_rate": 8.270704063003232e-07, "loss": 0.4523, "step": 9220 }, { "epoch": 0.08293273014259825, "grad_norm": 0.8632025122642517, "learning_rate": 8.249628059391251e-07, "loss": 0.6004, "step": 9221 }, { "epoch": 0.0829417240402387, "grad_norm": 0.36009249091148376, "learning_rate": 8.228578493236466e-07, "loss": 1.2201, "step": 9222 }, { "epoch": 0.08295071793787914, "grad_norm": 0.23952411115169525, "learning_rate": 8.207555366840869e-07, "loss": 0.8992, "step": 9223 }, { "epoch": 0.0829597118355196, "grad_norm": 0.6291777491569519, "learning_rate": 8.186558682503453e-07, "loss": 0.6626, "step": 9224 }, { "epoch": 0.08296870573316005, "grad_norm": 1.689408779144287, "learning_rate": 8.165588442520439e-07, "loss": 0.0823, "step": 9225 }, { "epoch": 0.0829776996308005, "grad_norm": 0.42676445841789246, "learning_rate": 8.144644649185073e-07, "loss": 0.7921, "step": 9226 }, { "epoch": 0.08298669352844096, "grad_norm": 0.4877423346042633, "learning_rate": 8.12372730478772e-07, "loss": 0.9366, "step": 9227 }, { "epoch": 0.0829956874260814, "grad_norm": 0.571127712726593, "learning_rate": 8.102836411615911e-07, "loss": 0.4994, "step": 9228 }, { "epoch": 0.08300468132372185, "grad_norm": 0.3546229600906372, "learning_rate": 8.081971971954261e-07, "loss": 0.6179, "step": 9229 }, { "epoch": 0.08301367522136231, "grad_norm": 0.4031550884246826, "learning_rate": 8.061133988084391e-07, "loss": 0.5568, "step": 9230 }, { "epoch": 0.08302266911900276, "grad_norm": 0.4966057538986206, "learning_rate": 8.040322462285171e-07, "loss": 0.3714, "step": 9231 }, { "epoch": 0.0830316630166432, "grad_norm": 0.7304659485816956, "learning_rate": 8.019537396832533e-07, "loss": 0.4721, "step": 9232 }, { "epoch": 0.08304065691428365, "grad_norm": 0.24751850962638855, "learning_rate": 7.998778793999434e-07, "loss": 0.7346, "step": 9233 }, { "epoch": 0.08304965081192411, "grad_norm": 0.48092132806777954, "learning_rate": 7.97804665605606e-07, "loss": 0.056, "step": 9234 }, { "epoch": 0.08305864470956456, "grad_norm": 0.48133641481399536, "learning_rate": 7.957340985269596e-07, "loss": 0.7615, "step": 9235 }, { "epoch": 0.08306763860720501, "grad_norm": 0.6764446496963501, "learning_rate": 7.936661783904481e-07, "loss": 0.5737, "step": 9236 }, { "epoch": 0.08307663250484547, "grad_norm": 0.37697815895080566, "learning_rate": 7.916009054222046e-07, "loss": 0.587, "step": 9237 }, { "epoch": 0.08308562640248592, "grad_norm": 0.5532207489013672, "learning_rate": 7.895382798480927e-07, "loss": 0.3309, "step": 9238 }, { "epoch": 0.08309462030012636, "grad_norm": 0.44196996092796326, "learning_rate": 7.874783018936738e-07, "loss": 0.2064, "step": 9239 }, { "epoch": 0.08310361419776681, "grad_norm": 0.5399118661880493, "learning_rate": 7.854209717842231e-07, "loss": 0.6323, "step": 9240 }, { "epoch": 0.08311260809540727, "grad_norm": 0.5019822716712952, "learning_rate": 7.833662897447275e-07, "loss": 0.8816, "step": 9241 }, { "epoch": 0.08312160199304772, "grad_norm": 0.553945779800415, "learning_rate": 7.813142559998904e-07, "loss": 0.6786, "step": 9242 }, { "epoch": 0.08313059589068816, "grad_norm": 0.40094271302223206, "learning_rate": 7.792648707741101e-07, "loss": 0.839, "step": 9243 }, { "epoch": 0.08313958978832862, "grad_norm": 2.0179734230041504, "learning_rate": 7.772181342915103e-07, "loss": 0.3643, "step": 9244 }, { "epoch": 0.08314858368596907, "grad_norm": 0.6071616411209106, "learning_rate": 7.751740467759145e-07, "loss": 0.4299, "step": 9245 }, { "epoch": 0.08315757758360952, "grad_norm": 0.5033546686172485, "learning_rate": 7.731326084508661e-07, "loss": 0.4946, "step": 9246 }, { "epoch": 0.08316657148124997, "grad_norm": 0.7201323509216309, "learning_rate": 7.710938195396089e-07, "loss": 0.4128, "step": 9247 }, { "epoch": 0.08317556537889043, "grad_norm": 0.24914419651031494, "learning_rate": 7.690576802651062e-07, "loss": 0.8998, "step": 9248 }, { "epoch": 0.08318455927653087, "grad_norm": 0.4019322991371155, "learning_rate": 7.670241908500214e-07, "loss": 0.7567, "step": 9249 }, { "epoch": 0.08319355317417132, "grad_norm": 0.8254199028015137, "learning_rate": 7.649933515167407e-07, "loss": 0.4273, "step": 9250 }, { "epoch": 0.08320254707181178, "grad_norm": 0.6849421262741089, "learning_rate": 7.629651624873501e-07, "loss": 0.2107, "step": 9251 }, { "epoch": 0.08321154096945223, "grad_norm": 0.5993214845657349, "learning_rate": 7.609396239836474e-07, "loss": 0.6035, "step": 9252 }, { "epoch": 0.08322053486709267, "grad_norm": 0.45398837327957153, "learning_rate": 7.589167362271444e-07, "loss": 0.4385, "step": 9253 }, { "epoch": 0.08322952876473313, "grad_norm": 0.3683181703090668, "learning_rate": 7.568964994390637e-07, "loss": 0.7245, "step": 9254 }, { "epoch": 0.08323852266237358, "grad_norm": 0.8849016427993774, "learning_rate": 7.548789138403317e-07, "loss": 0.2625, "step": 9255 }, { "epoch": 0.08324751656001403, "grad_norm": 0.477235347032547, "learning_rate": 7.528639796515885e-07, "loss": 0.8394, "step": 9256 }, { "epoch": 0.08325651045765448, "grad_norm": 0.5475624799728394, "learning_rate": 7.508516970931883e-07, "loss": 0.9084, "step": 9257 }, { "epoch": 0.08326550435529494, "grad_norm": 0.4811367392539978, "learning_rate": 7.488420663851859e-07, "loss": 0.6982, "step": 9258 }, { "epoch": 0.08327449825293538, "grad_norm": 0.5816320180892944, "learning_rate": 7.468350877473551e-07, "loss": 0.5876, "step": 9259 }, { "epoch": 0.08328349215057583, "grad_norm": 0.37936341762542725, "learning_rate": 7.448307613991734e-07, "loss": 0.7167, "step": 9260 }, { "epoch": 0.08329248604821629, "grad_norm": 0.4634511470794678, "learning_rate": 7.428290875598348e-07, "loss": 0.4404, "step": 9261 }, { "epoch": 0.08330147994585674, "grad_norm": 0.5554715991020203, "learning_rate": 7.408300664482365e-07, "loss": 0.7671, "step": 9262 }, { "epoch": 0.08331047384349718, "grad_norm": 0.44065383076667786, "learning_rate": 7.388336982829897e-07, "loss": 0.7975, "step": 9263 }, { "epoch": 0.08331946774113763, "grad_norm": 2.013162136077881, "learning_rate": 7.368399832824113e-07, "loss": 0.3659, "step": 9264 }, { "epoch": 0.08332846163877809, "grad_norm": 0.9617385864257812, "learning_rate": 7.348489216645327e-07, "loss": 0.0643, "step": 9265 }, { "epoch": 0.08333745553641854, "grad_norm": 0.5153279304504395, "learning_rate": 7.328605136470961e-07, "loss": 0.7022, "step": 9266 }, { "epoch": 0.08334644943405899, "grad_norm": 0.2819739878177643, "learning_rate": 7.308747594475502e-07, "loss": 0.9387, "step": 9267 }, { "epoch": 0.08335544333169945, "grad_norm": 0.3213467299938202, "learning_rate": 7.288916592830486e-07, "loss": 0.8196, "step": 9268 }, { "epoch": 0.0833644372293399, "grad_norm": 0.3474922776222229, "learning_rate": 7.269112133704681e-07, "loss": 0.9788, "step": 9269 }, { "epoch": 0.08337343112698034, "grad_norm": 0.7740414142608643, "learning_rate": 7.249334219263799e-07, "loss": 0.6814, "step": 9270 }, { "epoch": 0.0833824250246208, "grad_norm": 0.5272058844566345, "learning_rate": 7.229582851670774e-07, "loss": 0.3148, "step": 9271 }, { "epoch": 0.08339141892226125, "grad_norm": 0.43487465381622314, "learning_rate": 7.209858033085549e-07, "loss": 0.7116, "step": 9272 }, { "epoch": 0.0834004128199017, "grad_norm": 2.9976651668548584, "learning_rate": 7.190159765665255e-07, "loss": 0.2897, "step": 9273 }, { "epoch": 0.08340940671754214, "grad_norm": 0.502393901348114, "learning_rate": 7.170488051564006e-07, "loss": 0.8537, "step": 9274 }, { "epoch": 0.0834184006151826, "grad_norm": 2.77489972114563, "learning_rate": 7.150842892933107e-07, "loss": 0.6466, "step": 9275 }, { "epoch": 0.08342739451282305, "grad_norm": 0.3167167007923126, "learning_rate": 7.131224291920896e-07, "loss": 0.7657, "step": 9276 }, { "epoch": 0.0834363884104635, "grad_norm": 0.439037948846817, "learning_rate": 7.111632250672851e-07, "loss": 0.8384, "step": 9277 }, { "epoch": 0.08344538230810396, "grad_norm": 0.7386268973350525, "learning_rate": 7.092066771331507e-07, "loss": 0.2901, "step": 9278 }, { "epoch": 0.0834543762057444, "grad_norm": 0.3042357861995697, "learning_rate": 7.07252785603657e-07, "loss": 0.5461, "step": 9279 }, { "epoch": 0.08346337010338485, "grad_norm": 0.6903495192527771, "learning_rate": 7.053015506924748e-07, "loss": 0.9952, "step": 9280 }, { "epoch": 0.0834723640010253, "grad_norm": 0.5129989981651306, "learning_rate": 7.033529726129862e-07, "loss": 0.9895, "step": 9281 }, { "epoch": 0.08348135789866576, "grad_norm": 0.5056366920471191, "learning_rate": 7.0140705157829e-07, "loss": 0.4516, "step": 9282 }, { "epoch": 0.0834903517963062, "grad_norm": 0.4278598725795746, "learning_rate": 6.994637878011828e-07, "loss": 1.0179, "step": 9283 }, { "epoch": 0.08349934569394665, "grad_norm": 0.546110212802887, "learning_rate": 6.975231814941807e-07, "loss": 0.5187, "step": 9284 }, { "epoch": 0.08350833959158711, "grad_norm": 0.9863356351852417, "learning_rate": 6.955852328695056e-07, "loss": 0.659, "step": 9285 }, { "epoch": 0.08351733348922756, "grad_norm": 0.3863963484764099, "learning_rate": 6.936499421390935e-07, "loss": 0.704, "step": 9286 }, { "epoch": 0.08352632738686801, "grad_norm": 0.8386362195014954, "learning_rate": 6.917173095145724e-07, "loss": 0.6255, "step": 9287 }, { "epoch": 0.08353532128450845, "grad_norm": 0.9859722852706909, "learning_rate": 6.897873352073064e-07, "loss": 0.8112, "step": 9288 }, { "epoch": 0.08354431518214891, "grad_norm": 0.3946400582790375, "learning_rate": 6.878600194283408e-07, "loss": 0.7263, "step": 9289 }, { "epoch": 0.08355330907978936, "grad_norm": 0.5449779033660889, "learning_rate": 6.859353623884567e-07, "loss": 0.7923, "step": 9290 }, { "epoch": 0.08356230297742981, "grad_norm": 0.5063990354537964, "learning_rate": 6.840133642981195e-07, "loss": 0.6989, "step": 9291 }, { "epoch": 0.08357129687507027, "grad_norm": 0.45002415776252747, "learning_rate": 6.820940253675273e-07, "loss": 0.317, "step": 9292 }, { "epoch": 0.08358029077271072, "grad_norm": 0.32204797863960266, "learning_rate": 6.80177345806568e-07, "loss": 1.0421, "step": 9293 }, { "epoch": 0.08358928467035116, "grad_norm": 0.6206782460212708, "learning_rate": 6.782633258248517e-07, "loss": 0.8025, "step": 9294 }, { "epoch": 0.08359827856799162, "grad_norm": 0.4346006214618683, "learning_rate": 6.763519656316914e-07, "loss": 0.749, "step": 9295 }, { "epoch": 0.08360727246563207, "grad_norm": 0.7596007585525513, "learning_rate": 6.744432654361088e-07, "loss": 0.5087, "step": 9296 }, { "epoch": 0.08361626636327252, "grad_norm": 0.35737401247024536, "learning_rate": 6.725372254468343e-07, "loss": 1.0412, "step": 9297 }, { "epoch": 0.08362526026091296, "grad_norm": 0.570528507232666, "learning_rate": 6.706338458723205e-07, "loss": 0.4071, "step": 9298 }, { "epoch": 0.08363425415855343, "grad_norm": 0.7093732357025146, "learning_rate": 6.687331269207037e-07, "loss": 0.0853, "step": 9299 }, { "epoch": 0.08364324805619387, "grad_norm": 0.6779006719589233, "learning_rate": 6.668350687998565e-07, "loss": 0.5174, "step": 9300 }, { "epoch": 0.08365224195383432, "grad_norm": 0.7136320471763611, "learning_rate": 6.649396717173378e-07, "loss": 0.993, "step": 9301 }, { "epoch": 0.08366123585147478, "grad_norm": 0.7975353002548218, "learning_rate": 6.630469358804292e-07, "loss": 0.6788, "step": 9302 }, { "epoch": 0.08367022974911523, "grad_norm": 0.3048367500305176, "learning_rate": 6.611568614961178e-07, "loss": 1.1994, "step": 9303 }, { "epoch": 0.08367922364675567, "grad_norm": 0.7401994466781616, "learning_rate": 6.592694487711021e-07, "loss": 0.3027, "step": 9304 }, { "epoch": 0.08368821754439612, "grad_norm": 0.24120236933231354, "learning_rate": 6.573846979117809e-07, "loss": 1.0108, "step": 9305 }, { "epoch": 0.08369721144203658, "grad_norm": 0.40999406576156616, "learning_rate": 6.555026091242727e-07, "loss": 0.7324, "step": 9306 }, { "epoch": 0.08370620533967703, "grad_norm": 0.4631630778312683, "learning_rate": 6.536231826143963e-07, "loss": 0.436, "step": 9307 }, { "epoch": 0.08371519923731748, "grad_norm": 0.39534488320350647, "learning_rate": 6.517464185876842e-07, "loss": 0.7917, "step": 9308 }, { "epoch": 0.08372419313495794, "grad_norm": 0.4602808356285095, "learning_rate": 6.498723172493809e-07, "loss": 0.7635, "step": 9309 }, { "epoch": 0.08373318703259838, "grad_norm": 0.49407103657722473, "learning_rate": 6.480008788044223e-07, "loss": 0.7021, "step": 9310 }, { "epoch": 0.08374218093023883, "grad_norm": 0.6228275895118713, "learning_rate": 6.461321034574836e-07, "loss": 0.6961, "step": 9311 }, { "epoch": 0.08375117482787929, "grad_norm": 0.931683361530304, "learning_rate": 6.442659914129179e-07, "loss": 0.3263, "step": 9312 }, { "epoch": 0.08376016872551974, "grad_norm": 0.6472002267837524, "learning_rate": 6.424025428748065e-07, "loss": 0.5827, "step": 9313 }, { "epoch": 0.08376916262316018, "grad_norm": 0.6295167803764343, "learning_rate": 6.405417580469308e-07, "loss": 0.8004, "step": 9314 }, { "epoch": 0.08377815652080063, "grad_norm": 0.36828258633613586, "learning_rate": 6.386836371327892e-07, "loss": 1.1109, "step": 9315 }, { "epoch": 0.08378715041844109, "grad_norm": 0.41354212164878845, "learning_rate": 6.368281803355691e-07, "loss": 0.8304, "step": 9316 }, { "epoch": 0.08379614431608154, "grad_norm": 0.3290765583515167, "learning_rate": 6.349753878581971e-07, "loss": 0.8561, "step": 9317 }, { "epoch": 0.08380513821372199, "grad_norm": 2.8689544200897217, "learning_rate": 6.331252599032833e-07, "loss": 0.3977, "step": 9318 }, { "epoch": 0.08381413211136245, "grad_norm": 0.7054927349090576, "learning_rate": 6.312777966731575e-07, "loss": 0.5217, "step": 9319 }, { "epoch": 0.0838231260090029, "grad_norm": 0.639707624912262, "learning_rate": 6.2943299836985e-07, "loss": 0.6913, "step": 9320 }, { "epoch": 0.08383211990664334, "grad_norm": 0.42947015166282654, "learning_rate": 6.275908651951129e-07, "loss": 0.7291, "step": 9321 }, { "epoch": 0.08384111380428379, "grad_norm": 0.5891924500465393, "learning_rate": 6.257513973503909e-07, "loss": 0.429, "step": 9322 }, { "epoch": 0.08385010770192425, "grad_norm": 0.45502224564552307, "learning_rate": 6.239145950368535e-07, "loss": 0.7137, "step": 9323 }, { "epoch": 0.0838591015995647, "grad_norm": 0.6101135015487671, "learning_rate": 6.220804584553652e-07, "loss": 0.7069, "step": 9324 }, { "epoch": 0.08386809549720514, "grad_norm": 0.7361465692520142, "learning_rate": 6.202489878065071e-07, "loss": 0.6655, "step": 9325 }, { "epoch": 0.0838770893948456, "grad_norm": 0.6516445279121399, "learning_rate": 6.184201832905662e-07, "loss": 0.7755, "step": 9326 }, { "epoch": 0.08388608329248605, "grad_norm": 0.3906131684780121, "learning_rate": 6.165940451075352e-07, "loss": 0.7605, "step": 9327 }, { "epoch": 0.0838950771901265, "grad_norm": 0.8033998608589172, "learning_rate": 6.147705734571213e-07, "loss": 0.8954, "step": 9328 }, { "epoch": 0.08390407108776694, "grad_norm": 1.3687061071395874, "learning_rate": 6.129497685387342e-07, "loss": 0.563, "step": 9329 }, { "epoch": 0.0839130649854074, "grad_norm": 0.9622273445129395, "learning_rate": 6.111316305514925e-07, "loss": 0.2924, "step": 9330 }, { "epoch": 0.08392205888304785, "grad_norm": 3.430903911590576, "learning_rate": 6.093161596942287e-07, "loss": 0.1482, "step": 9331 }, { "epoch": 0.0839310527806883, "grad_norm": 0.6213619709014893, "learning_rate": 6.075033561654813e-07, "loss": 0.7599, "step": 9332 }, { "epoch": 0.08394004667832876, "grad_norm": 0.3441617786884308, "learning_rate": 6.056932201634918e-07, "loss": 1.0846, "step": 9333 }, { "epoch": 0.0839490405759692, "grad_norm": 0.30971261858940125, "learning_rate": 6.038857518862184e-07, "loss": 0.7645, "step": 9334 }, { "epoch": 0.08395803447360965, "grad_norm": 0.5741382241249084, "learning_rate": 6.020809515313142e-07, "loss": 0.3878, "step": 9335 }, { "epoch": 0.08396702837125011, "grad_norm": 1.5283623933792114, "learning_rate": 6.0027881929616e-07, "loss": 0.4479, "step": 9336 }, { "epoch": 0.08397602226889056, "grad_norm": 0.9055088758468628, "learning_rate": 5.984793553778262e-07, "loss": 0.2608, "step": 9337 }, { "epoch": 0.083985016166531, "grad_norm": 0.24362947046756744, "learning_rate": 5.96682559973108e-07, "loss": 1.0025, "step": 9338 }, { "epoch": 0.08399401006417145, "grad_norm": 0.5600036978721619, "learning_rate": 5.948884332784899e-07, "loss": 0.5429, "step": 9339 }, { "epoch": 0.08400300396181191, "grad_norm": 0.34521928429603577, "learning_rate": 5.930969754901843e-07, "loss": 1.0266, "step": 9340 }, { "epoch": 0.08401199785945236, "grad_norm": 0.32228124141693115, "learning_rate": 5.913081868040931e-07, "loss": 0.7662, "step": 9341 }, { "epoch": 0.08402099175709281, "grad_norm": 0.6849814653396606, "learning_rate": 5.895220674158458e-07, "loss": 0.9162, "step": 9342 }, { "epoch": 0.08402998565473327, "grad_norm": 0.43803542852401733, "learning_rate": 5.877386175207639e-07, "loss": 0.7919, "step": 9343 }, { "epoch": 0.08403897955237372, "grad_norm": 0.2972281575202942, "learning_rate": 5.85957837313883e-07, "loss": 0.993, "step": 9344 }, { "epoch": 0.08404797345001416, "grad_norm": 0.5078462362289429, "learning_rate": 5.841797269899447e-07, "loss": 0.6486, "step": 9345 }, { "epoch": 0.08405696734765461, "grad_norm": 0.7022960782051086, "learning_rate": 5.82404286743407e-07, "loss": 0.4415, "step": 9346 }, { "epoch": 0.08406596124529507, "grad_norm": 0.5876705050468445, "learning_rate": 5.806315167684234e-07, "loss": 0.8373, "step": 9347 }, { "epoch": 0.08407495514293552, "grad_norm": 0.578029215335846, "learning_rate": 5.788614172588608e-07, "loss": 0.7606, "step": 9348 }, { "epoch": 0.08408394904057596, "grad_norm": 0.2859131693840027, "learning_rate": 5.770939884082977e-07, "loss": 1.0223, "step": 9349 }, { "epoch": 0.08409294293821643, "grad_norm": 0.40305763483047485, "learning_rate": 5.753292304100183e-07, "loss": 0.8588, "step": 9350 }, { "epoch": 0.08410193683585687, "grad_norm": 0.38114723563194275, "learning_rate": 5.735671434570127e-07, "loss": 0.7956, "step": 9351 }, { "epoch": 0.08411093073349732, "grad_norm": 0.5870527029037476, "learning_rate": 5.71807727741977e-07, "loss": 0.6449, "step": 9352 }, { "epoch": 0.08411992463113778, "grad_norm": 0.5688018202781677, "learning_rate": 5.700509834573237e-07, "loss": 0.4129, "step": 9353 }, { "epoch": 0.08412891852877823, "grad_norm": 0.7235795855522156, "learning_rate": 5.68296910795163e-07, "loss": 0.4656, "step": 9354 }, { "epoch": 0.08413791242641867, "grad_norm": 0.4389757513999939, "learning_rate": 5.665455099473221e-07, "loss": 0.9174, "step": 9355 }, { "epoch": 0.08414690632405912, "grad_norm": 0.3186652958393097, "learning_rate": 5.647967811053256e-07, "loss": 0.6529, "step": 9356 }, { "epoch": 0.08415590022169958, "grad_norm": 0.4194771349430084, "learning_rate": 5.630507244604205e-07, "loss": 0.7129, "step": 9357 }, { "epoch": 0.08416489411934003, "grad_norm": 0.7576116919517517, "learning_rate": 5.613073402035429e-07, "loss": 0.4167, "step": 9358 }, { "epoch": 0.08417388801698047, "grad_norm": 1.3286798000335693, "learning_rate": 5.595666285253543e-07, "loss": 0.3951, "step": 9359 }, { "epoch": 0.08418288191462094, "grad_norm": 0.3170199394226074, "learning_rate": 5.578285896162106e-07, "loss": 0.8842, "step": 9360 }, { "epoch": 0.08419187581226138, "grad_norm": 0.558043897151947, "learning_rate": 5.560932236661875e-07, "loss": 0.8748, "step": 9361 }, { "epoch": 0.08420086970990183, "grad_norm": 0.3552372455596924, "learning_rate": 5.543605308650557e-07, "loss": 0.8731, "step": 9362 }, { "epoch": 0.08420986360754228, "grad_norm": 0.45800450444221497, "learning_rate": 5.52630511402305e-07, "loss": 0.353, "step": 9363 }, { "epoch": 0.08421885750518274, "grad_norm": 0.5442957878112793, "learning_rate": 5.509031654671231e-07, "loss": 0.7722, "step": 9364 }, { "epoch": 0.08422785140282318, "grad_norm": 0.6581498980522156, "learning_rate": 5.491784932484173e-07, "loss": 0.5798, "step": 9365 }, { "epoch": 0.08423684530046363, "grad_norm": 1.0239492654800415, "learning_rate": 5.474564949347838e-07, "loss": 0.922, "step": 9366 }, { "epoch": 0.08424583919810409, "grad_norm": 0.4358322024345398, "learning_rate": 5.45737170714547e-07, "loss": 0.9038, "step": 9367 }, { "epoch": 0.08425483309574454, "grad_norm": 0.6147909164428711, "learning_rate": 5.44020520775726e-07, "loss": 0.7045, "step": 9368 }, { "epoch": 0.08426382699338499, "grad_norm": 0.31795939803123474, "learning_rate": 5.423065453060566e-07, "loss": 0.9399, "step": 9369 }, { "epoch": 0.08427282089102545, "grad_norm": 0.6476067900657654, "learning_rate": 5.405952444929668e-07, "loss": 0.3979, "step": 9370 }, { "epoch": 0.08428181478866589, "grad_norm": 1.0853075981140137, "learning_rate": 5.388866185236097e-07, "loss": 0.0752, "step": 9371 }, { "epoch": 0.08429080868630634, "grad_norm": 0.3714310824871063, "learning_rate": 5.371806675848329e-07, "loss": 0.7727, "step": 9372 }, { "epoch": 0.08429980258394679, "grad_norm": 0.3915404975414276, "learning_rate": 5.354773918631983e-07, "loss": 0.6265, "step": 9373 }, { "epoch": 0.08430879648158725, "grad_norm": 1.0528596639633179, "learning_rate": 5.337767915449738e-07, "loss": 0.697, "step": 9374 }, { "epoch": 0.0843177903792277, "grad_norm": 0.6707680225372314, "learning_rate": 5.32078866816138e-07, "loss": 0.8335, "step": 9375 }, { "epoch": 0.08432678427686814, "grad_norm": 0.40169501304626465, "learning_rate": 5.303836178623706e-07, "loss": 0.6898, "step": 9376 }, { "epoch": 0.0843357781745086, "grad_norm": 0.3441356420516968, "learning_rate": 5.286910448690591e-07, "loss": 0.6278, "step": 9377 }, { "epoch": 0.08434477207214905, "grad_norm": 0.5006733536720276, "learning_rate": 5.270011480213056e-07, "loss": 0.4369, "step": 9378 }, { "epoch": 0.0843537659697895, "grad_norm": 0.5817052721977234, "learning_rate": 5.253139275039066e-07, "loss": 0.6324, "step": 9379 }, { "epoch": 0.08436275986742994, "grad_norm": 0.3136277496814728, "learning_rate": 5.236293835013839e-07, "loss": 0.9813, "step": 9380 }, { "epoch": 0.0843717537650704, "grad_norm": 0.3983188569545746, "learning_rate": 5.219475161979487e-07, "loss": 0.841, "step": 9381 }, { "epoch": 0.08438074766271085, "grad_norm": 0.37245863676071167, "learning_rate": 5.202683257775342e-07, "loss": 0.9028, "step": 9382 }, { "epoch": 0.0843897415603513, "grad_norm": 0.5883781909942627, "learning_rate": 5.185918124237715e-07, "loss": 1.0825, "step": 9383 }, { "epoch": 0.08439873545799176, "grad_norm": 0.4169425368309021, "learning_rate": 5.169179763199999e-07, "loss": 0.745, "step": 9384 }, { "epoch": 0.0844077293556322, "grad_norm": 0.3206878900527954, "learning_rate": 5.152468176492675e-07, "loss": 0.8515, "step": 9385 }, { "epoch": 0.08441672325327265, "grad_norm": 0.5192161202430725, "learning_rate": 5.135783365943309e-07, "loss": 0.6889, "step": 9386 }, { "epoch": 0.0844257171509131, "grad_norm": 0.3266715705394745, "learning_rate": 5.119125333376523e-07, "loss": 0.7523, "step": 9387 }, { "epoch": 0.08443471104855356, "grad_norm": 0.4164428412914276, "learning_rate": 5.102494080614029e-07, "loss": 0.9253, "step": 9388 }, { "epoch": 0.084443704946194, "grad_norm": 0.37324950098991394, "learning_rate": 5.085889609474593e-07, "loss": 0.8746, "step": 9389 }, { "epoch": 0.08445269884383445, "grad_norm": 0.7610208988189697, "learning_rate": 5.069311921774039e-07, "loss": 0.3376, "step": 9390 }, { "epoch": 0.08446169274147491, "grad_norm": 0.3239977955818176, "learning_rate": 5.052761019325281e-07, "loss": 1.1307, "step": 9391 }, { "epoch": 0.08447068663911536, "grad_norm": 0.6705030202865601, "learning_rate": 5.036236903938285e-07, "loss": 0.7355, "step": 9392 }, { "epoch": 0.08447968053675581, "grad_norm": 0.3141360878944397, "learning_rate": 5.019739577420135e-07, "loss": 0.9533, "step": 9393 }, { "epoch": 0.08448867443439627, "grad_norm": 0.2749394178390503, "learning_rate": 5.003269041574971e-07, "loss": 0.978, "step": 9394 }, { "epoch": 0.08449766833203672, "grad_norm": 0.4125874936580658, "learning_rate": 4.986825298203935e-07, "loss": 0.735, "step": 9395 }, { "epoch": 0.08450666222967716, "grad_norm": 1.3306057453155518, "learning_rate": 4.97040834910531e-07, "loss": 0.8086, "step": 9396 }, { "epoch": 0.08451565612731761, "grad_norm": 1.4497699737548828, "learning_rate": 4.954018196074439e-07, "loss": 0.6246, "step": 9397 }, { "epoch": 0.08452465002495807, "grad_norm": 0.41583970189094543, "learning_rate": 4.937654840903722e-07, "loss": 0.6055, "step": 9398 }, { "epoch": 0.08453364392259852, "grad_norm": 0.31154927611351013, "learning_rate": 4.921318285382587e-07, "loss": 0.7931, "step": 9399 }, { "epoch": 0.08454263782023896, "grad_norm": 0.3771264851093292, "learning_rate": 4.905008531297661e-07, "loss": 0.4821, "step": 9400 }, { "epoch": 0.08455163171787942, "grad_norm": 1.5759341716766357, "learning_rate": 4.88872558043249e-07, "loss": 0.5031, "step": 9401 }, { "epoch": 0.08456062561551987, "grad_norm": 0.5437878370285034, "learning_rate": 4.872469434567762e-07, "loss": 0.6044, "step": 9402 }, { "epoch": 0.08456961951316032, "grad_norm": 0.37107568979263306, "learning_rate": 4.856240095481273e-07, "loss": 0.866, "step": 9403 }, { "epoch": 0.08457861341080077, "grad_norm": 0.6057437062263489, "learning_rate": 4.840037564947747e-07, "loss": 0.4989, "step": 9404 }, { "epoch": 0.08458760730844123, "grad_norm": 0.2840712070465088, "learning_rate": 4.823861844739152e-07, "loss": 1.0605, "step": 9405 }, { "epoch": 0.08459660120608167, "grad_norm": 0.3101247549057007, "learning_rate": 4.807712936624381e-07, "loss": 0.8385, "step": 9406 }, { "epoch": 0.08460559510372212, "grad_norm": 0.6292953491210938, "learning_rate": 4.79159084236952e-07, "loss": 0.3739, "step": 9407 }, { "epoch": 0.08461458900136258, "grad_norm": 0.9086684584617615, "learning_rate": 4.775495563737575e-07, "loss": 0.7622, "step": 9408 }, { "epoch": 0.08462358289900303, "grad_norm": 0.40186214447021484, "learning_rate": 4.759427102488806e-07, "loss": 0.658, "step": 9409 }, { "epoch": 0.08463257679664347, "grad_norm": 0.7342746257781982, "learning_rate": 4.743385460380334e-07, "loss": 0.4292, "step": 9410 }, { "epoch": 0.08464157069428394, "grad_norm": 0.5948224663734436, "learning_rate": 4.727370639166506e-07, "loss": 0.5642, "step": 9411 }, { "epoch": 0.08465056459192438, "grad_norm": 0.6344476938247681, "learning_rate": 4.711382640598616e-07, "loss": 0.6315, "step": 9412 }, { "epoch": 0.08465955848956483, "grad_norm": 1.0144104957580566, "learning_rate": 4.695421466425182e-07, "loss": 0.6085, "step": 9413 }, { "epoch": 0.08466855238720528, "grad_norm": 0.7302782535552979, "learning_rate": 4.67948711839164e-07, "loss": 0.8458, "step": 9414 }, { "epoch": 0.08467754628484574, "grad_norm": 0.3851030170917511, "learning_rate": 4.663579598240569e-07, "loss": 1.1612, "step": 9415 }, { "epoch": 0.08468654018248618, "grad_norm": 0.7271369099617004, "learning_rate": 4.647698907711523e-07, "loss": 0.3057, "step": 9416 }, { "epoch": 0.08469553408012663, "grad_norm": 0.9959169626235962, "learning_rate": 4.631845048541306e-07, "loss": 0.8346, "step": 9417 }, { "epoch": 0.08470452797776709, "grad_norm": 0.48552486300468445, "learning_rate": 4.6160180224635305e-07, "loss": 0.8017, "step": 9418 }, { "epoch": 0.08471352187540754, "grad_norm": 0.30116283893585205, "learning_rate": 4.6002178312091747e-07, "loss": 1.0912, "step": 9419 }, { "epoch": 0.08472251577304798, "grad_norm": 0.4382910132408142, "learning_rate": 4.5844444765059945e-07, "loss": 0.7138, "step": 9420 }, { "epoch": 0.08473150967068843, "grad_norm": 0.42871609330177307, "learning_rate": 4.568697960079027e-07, "loss": 0.8737, "step": 9421 }, { "epoch": 0.08474050356832889, "grad_norm": 0.558975875377655, "learning_rate": 4.552978283650228e-07, "loss": 0.4004, "step": 9422 }, { "epoch": 0.08474949746596934, "grad_norm": 0.5137666463851929, "learning_rate": 4.537285448938694e-07, "loss": 0.4675, "step": 9423 }, { "epoch": 0.08475849136360979, "grad_norm": 0.4008348286151886, "learning_rate": 4.521619457660553e-07, "loss": 0.8665, "step": 9424 }, { "epoch": 0.08476748526125025, "grad_norm": 0.6981993913650513, "learning_rate": 4.505980311529101e-07, "loss": 0.419, "step": 9425 }, { "epoch": 0.0847764791588907, "grad_norm": 0.6124168038368225, "learning_rate": 4.490368012254498e-07, "loss": 0.3901, "step": 9426 }, { "epoch": 0.08478547305653114, "grad_norm": 0.4586964547634125, "learning_rate": 4.4747825615441274e-07, "loss": 0.8546, "step": 9427 }, { "epoch": 0.08479446695417159, "grad_norm": 0.2960319519042969, "learning_rate": 4.459223961102432e-07, "loss": 1.0469, "step": 9428 }, { "epoch": 0.08480346085181205, "grad_norm": 0.6573370695114136, "learning_rate": 4.443692212630829e-07, "loss": 0.8756, "step": 9429 }, { "epoch": 0.0848124547494525, "grad_norm": 0.8886697888374329, "learning_rate": 4.4281873178278475e-07, "loss": 0.2829, "step": 9430 }, { "epoch": 0.08482144864709294, "grad_norm": 0.6646905541419983, "learning_rate": 4.412709278389049e-07, "loss": 0.6097, "step": 9431 }, { "epoch": 0.0848304425447334, "grad_norm": 0.6865703463554382, "learning_rate": 4.3972580960071897e-07, "loss": 0.8213, "step": 9432 }, { "epoch": 0.08483943644237385, "grad_norm": 0.7103006839752197, "learning_rate": 4.381833772371918e-07, "loss": 0.5769, "step": 9433 }, { "epoch": 0.0848484303400143, "grad_norm": 0.7454495429992676, "learning_rate": 4.3664363091700243e-07, "loss": 0.902, "step": 9434 }, { "epoch": 0.08485742423765476, "grad_norm": 2.541304111480713, "learning_rate": 4.351065708085328e-07, "loss": 0.3909, "step": 9435 }, { "epoch": 0.0848664181352952, "grad_norm": 0.6877978444099426, "learning_rate": 4.3357219707987617e-07, "loss": 0.869, "step": 9436 }, { "epoch": 0.08487541203293565, "grad_norm": 0.9010788798332214, "learning_rate": 4.3204050989882895e-07, "loss": 0.0578, "step": 9437 }, { "epoch": 0.0848844059305761, "grad_norm": 3.2383360862731934, "learning_rate": 4.3051150943289596e-07, "loss": 0.5518, "step": 9438 }, { "epoch": 0.08489339982821656, "grad_norm": 0.4615662693977356, "learning_rate": 4.289851958492824e-07, "loss": 0.9108, "step": 9439 }, { "epoch": 0.084902393725857, "grad_norm": 0.4500855505466461, "learning_rate": 4.2746156931490754e-07, "loss": 0.5291, "step": 9440 }, { "epoch": 0.08491138762349745, "grad_norm": 0.83323073387146, "learning_rate": 4.259406299963908e-07, "loss": 1.0109, "step": 9441 }, { "epoch": 0.08492038152113791, "grad_norm": 0.824015200138092, "learning_rate": 4.2442237806006314e-07, "loss": 0.4331, "step": 9442 }, { "epoch": 0.08492937541877836, "grad_norm": 0.5724748373031616, "learning_rate": 4.2290681367194716e-07, "loss": 0.9386, "step": 9443 }, { "epoch": 0.08493836931641881, "grad_norm": 0.3415127992630005, "learning_rate": 4.213939369977993e-07, "loss": 0.7712, "step": 9444 }, { "epoch": 0.08494736321405925, "grad_norm": 0.4020676016807556, "learning_rate": 4.1988374820305377e-07, "loss": 0.8092, "step": 9445 }, { "epoch": 0.08495635711169972, "grad_norm": 0.6873583197593689, "learning_rate": 4.183762474528674e-07, "loss": 0.5212, "step": 9446 }, { "epoch": 0.08496535100934016, "grad_norm": 0.363081693649292, "learning_rate": 4.1687143491209446e-07, "loss": 0.7671, "step": 9447 }, { "epoch": 0.08497434490698061, "grad_norm": 0.39804816246032715, "learning_rate": 4.153693107453005e-07, "loss": 0.8057, "step": 9448 }, { "epoch": 0.08498333880462107, "grad_norm": 0.34867173433303833, "learning_rate": 4.138698751167597e-07, "loss": 0.847, "step": 9449 }, { "epoch": 0.08499233270226152, "grad_norm": 3.928269863128662, "learning_rate": 4.1237312819044085e-07, "loss": 0.5719, "step": 9450 }, { "epoch": 0.08500132659990196, "grad_norm": 0.41442227363586426, "learning_rate": 4.1087907013003247e-07, "loss": 0.7757, "step": 9451 }, { "epoch": 0.08501032049754242, "grad_norm": 0.24817582964897156, "learning_rate": 4.093877010989178e-07, "loss": 0.8195, "step": 9452 }, { "epoch": 0.08501931439518287, "grad_norm": 0.5635285377502441, "learning_rate": 4.0789902126019686e-07, "loss": 0.4128, "step": 9453 }, { "epoch": 0.08502830829282332, "grad_norm": 1.50586998462677, "learning_rate": 4.064130307766617e-07, "loss": 0.4282, "step": 9454 }, { "epoch": 0.08503730219046376, "grad_norm": 0.9674355387687683, "learning_rate": 4.049297298108212e-07, "loss": 0.7403, "step": 9455 }, { "epoch": 0.08504629608810423, "grad_norm": 0.341532826423645, "learning_rate": 4.034491185248873e-07, "loss": 0.874, "step": 9456 }, { "epoch": 0.08505528998574467, "grad_norm": 0.6228776574134827, "learning_rate": 4.0197119708078047e-07, "loss": 1.2412, "step": 9457 }, { "epoch": 0.08506428388338512, "grad_norm": 0.2866643965244293, "learning_rate": 4.004959656401186e-07, "loss": 0.8222, "step": 9458 }, { "epoch": 0.08507327778102558, "grad_norm": 0.5493074655532837, "learning_rate": 3.9902342436423376e-07, "loss": 0.8067, "step": 9459 }, { "epoch": 0.08508227167866603, "grad_norm": 0.4632689356803894, "learning_rate": 3.9755357341415835e-07, "loss": 0.6791, "step": 9460 }, { "epoch": 0.08509126557630647, "grad_norm": 0.7050111293792725, "learning_rate": 3.9608641295063863e-07, "loss": 0.4473, "step": 9461 }, { "epoch": 0.08510025947394692, "grad_norm": 0.4324091672897339, "learning_rate": 3.9462194313411303e-07, "loss": 0.5265, "step": 9462 }, { "epoch": 0.08510925337158738, "grad_norm": 0.7453445196151733, "learning_rate": 3.9316016412474235e-07, "loss": 0.6044, "step": 9463 }, { "epoch": 0.08511824726922783, "grad_norm": 0.42939096689224243, "learning_rate": 3.9170107608237637e-07, "loss": 0.7694, "step": 9464 }, { "epoch": 0.08512724116686828, "grad_norm": 0.4352775514125824, "learning_rate": 3.902446791665848e-07, "loss": 1.0294, "step": 9465 }, { "epoch": 0.08513623506450874, "grad_norm": 0.5222339034080505, "learning_rate": 3.887909735366346e-07, "loss": 0.7472, "step": 9466 }, { "epoch": 0.08514522896214918, "grad_norm": 0.8223403096199036, "learning_rate": 3.8733995935150426e-07, "loss": 0.6366, "step": 9467 }, { "epoch": 0.08515422285978963, "grad_norm": 0.30492886900901794, "learning_rate": 3.8589163676986674e-07, "loss": 1.0254, "step": 9468 }, { "epoch": 0.08516321675743008, "grad_norm": 0.49484333395957947, "learning_rate": 3.8444600595011484e-07, "loss": 0.3793, "step": 9469 }, { "epoch": 0.08517221065507054, "grad_norm": 0.5120353698730469, "learning_rate": 3.83003067050336e-07, "loss": 0.7563, "step": 9470 }, { "epoch": 0.08518120455271098, "grad_norm": 0.5060017704963684, "learning_rate": 3.8156282022833446e-07, "loss": 0.6803, "step": 9471 }, { "epoch": 0.08519019845035143, "grad_norm": 0.517028272151947, "learning_rate": 3.8012526564160933e-07, "loss": 0.5422, "step": 9472 }, { "epoch": 0.08519919234799189, "grad_norm": 0.9290353059768677, "learning_rate": 3.7869040344736817e-07, "loss": 0.0506, "step": 9473 }, { "epoch": 0.08520818624563234, "grad_norm": 0.8655331134796143, "learning_rate": 3.772582338025299e-07, "loss": 0.4236, "step": 9474 }, { "epoch": 0.08521718014327279, "grad_norm": 0.6613210439682007, "learning_rate": 3.758287568637081e-07, "loss": 1.0348, "step": 9475 }, { "epoch": 0.08522617404091325, "grad_norm": 2.8302078247070312, "learning_rate": 3.7440197278723054e-07, "loss": 0.3329, "step": 9476 }, { "epoch": 0.0852351679385537, "grad_norm": 0.6419405341148376, "learning_rate": 3.729778817291307e-07, "loss": 0.5648, "step": 9477 }, { "epoch": 0.08524416183619414, "grad_norm": 0.45628923177719116, "learning_rate": 3.715564838451452e-07, "loss": 0.6727, "step": 9478 }, { "epoch": 0.08525315573383459, "grad_norm": 2.6806883811950684, "learning_rate": 3.7013777929071345e-07, "loss": 0.2304, "step": 9479 }, { "epoch": 0.08526214963147505, "grad_norm": 0.37615811824798584, "learning_rate": 3.687217682209837e-07, "loss": 0.9191, "step": 9480 }, { "epoch": 0.0852711435291155, "grad_norm": 1.5726003646850586, "learning_rate": 3.6730845079080977e-07, "loss": 0.1631, "step": 9481 }, { "epoch": 0.08528013742675594, "grad_norm": 1.1346330642700195, "learning_rate": 3.6589782715474585e-07, "loss": 0.4159, "step": 9482 }, { "epoch": 0.0852891313243964, "grad_norm": 1.0347267389297485, "learning_rate": 3.644898974670602e-07, "loss": 0.3684, "step": 9483 }, { "epoch": 0.08529812522203685, "grad_norm": 1.3167918920516968, "learning_rate": 3.630846618817213e-07, "loss": 0.7999, "step": 9484 }, { "epoch": 0.0853071191196773, "grad_norm": 0.37958404421806335, "learning_rate": 3.616821205524007e-07, "loss": 0.6881, "step": 9485 }, { "epoch": 0.08531611301731774, "grad_norm": 0.5395617485046387, "learning_rate": 3.6028227363248404e-07, "loss": 0.5247, "step": 9486 }, { "epoch": 0.0853251069149582, "grad_norm": 0.613332211971283, "learning_rate": 3.588851212750488e-07, "loss": 0.7006, "step": 9487 }, { "epoch": 0.08533410081259865, "grad_norm": 0.45806413888931274, "learning_rate": 3.574906636328923e-07, "loss": 0.81, "step": 9488 }, { "epoch": 0.0853430947102391, "grad_norm": 2.9720497131347656, "learning_rate": 3.560989008585036e-07, "loss": 0.3801, "step": 9489 }, { "epoch": 0.08535208860787956, "grad_norm": 0.29600152373313904, "learning_rate": 3.5470983310409157e-07, "loss": 0.8789, "step": 9490 }, { "epoch": 0.08536108250552, "grad_norm": 0.43687519431114197, "learning_rate": 3.533234605215568e-07, "loss": 0.4687, "step": 9491 }, { "epoch": 0.08537007640316045, "grad_norm": 0.4728500247001648, "learning_rate": 3.519397832625143e-07, "loss": 0.6564, "step": 9492 }, { "epoch": 0.08537907030080091, "grad_norm": 0.5035286545753479, "learning_rate": 3.5055880147827623e-07, "loss": 0.5302, "step": 9493 }, { "epoch": 0.08538806419844136, "grad_norm": 0.43329983949661255, "learning_rate": 3.4918051531986904e-07, "loss": 1.0311, "step": 9494 }, { "epoch": 0.08539705809608181, "grad_norm": 1.3442271947860718, "learning_rate": 3.478049249380194e-07, "loss": 0.4773, "step": 9495 }, { "epoch": 0.08540605199372225, "grad_norm": 0.29516568779945374, "learning_rate": 3.4643203048315973e-07, "loss": 0.8912, "step": 9496 }, { "epoch": 0.08541504589136271, "grad_norm": 0.6338436603546143, "learning_rate": 3.4506183210542555e-07, "loss": 0.2747, "step": 9497 }, { "epoch": 0.08542403978900316, "grad_norm": 0.7138974070549011, "learning_rate": 3.436943299546608e-07, "loss": 0.5614, "step": 9498 }, { "epoch": 0.08543303368664361, "grad_norm": 0.48987650871276855, "learning_rate": 3.423295241804153e-07, "loss": 0.8784, "step": 9499 }, { "epoch": 0.08544202758428407, "grad_norm": 0.31609538197517395, "learning_rate": 3.4096741493194197e-07, "loss": 0.9556, "step": 9500 }, { "epoch": 0.08545102148192452, "grad_norm": 0.6503754258155823, "learning_rate": 3.396080023581938e-07, "loss": 0.7195, "step": 9501 }, { "epoch": 0.08546001537956496, "grad_norm": 0.2887336015701294, "learning_rate": 3.382512866078408e-07, "loss": 0.9875, "step": 9502 }, { "epoch": 0.08546900927720541, "grad_norm": 0.5318663716316223, "learning_rate": 3.3689726782925033e-07, "loss": 0.6636, "step": 9503 }, { "epoch": 0.08547800317484587, "grad_norm": 0.443109929561615, "learning_rate": 3.3554594617049293e-07, "loss": 0.79, "step": 9504 }, { "epoch": 0.08548699707248632, "grad_norm": 0.6109262108802795, "learning_rate": 3.341973217793476e-07, "loss": 0.7989, "step": 9505 }, { "epoch": 0.08549599097012676, "grad_norm": 0.6849091053009033, "learning_rate": 3.328513948032991e-07, "loss": 0.887, "step": 9506 }, { "epoch": 0.08550498486776723, "grad_norm": 0.5919128656387329, "learning_rate": 3.315081653895352e-07, "loss": 0.5048, "step": 9507 }, { "epoch": 0.08551397876540767, "grad_norm": 0.7052337527275085, "learning_rate": 3.3016763368494963e-07, "loss": 0.6491, "step": 9508 }, { "epoch": 0.08552297266304812, "grad_norm": 0.7423193454742432, "learning_rate": 3.288297998361417e-07, "loss": 0.4762, "step": 9509 }, { "epoch": 0.08553196656068857, "grad_norm": 0.3527366518974304, "learning_rate": 3.2749466398941384e-07, "loss": 0.9234, "step": 9510 }, { "epoch": 0.08554096045832903, "grad_norm": 0.6378151774406433, "learning_rate": 3.2616222629077707e-07, "loss": 0.3051, "step": 9511 }, { "epoch": 0.08554995435596947, "grad_norm": 0.7942308187484741, "learning_rate": 3.248324868859398e-07, "loss": 0.3797, "step": 9512 }, { "epoch": 0.08555894825360992, "grad_norm": 0.3801249861717224, "learning_rate": 3.235054459203246e-07, "loss": 0.8593, "step": 9513 }, { "epoch": 0.08556794215125038, "grad_norm": 0.7264021635055542, "learning_rate": 3.221811035390515e-07, "loss": 0.6137, "step": 9514 }, { "epoch": 0.08557693604889083, "grad_norm": 0.3323279023170471, "learning_rate": 3.2085945988695464e-07, "loss": 0.545, "step": 9515 }, { "epoch": 0.08558592994653128, "grad_norm": 0.46077826619148254, "learning_rate": 3.195405151085573e-07, "loss": 0.4881, "step": 9516 }, { "epoch": 0.08559492384417174, "grad_norm": 0.8702954649925232, "learning_rate": 3.18224269348108e-07, "loss": 0.2621, "step": 9517 }, { "epoch": 0.08560391774181218, "grad_norm": 0.3404834568500519, "learning_rate": 3.169107227495388e-07, "loss": 0.9123, "step": 9518 }, { "epoch": 0.08561291163945263, "grad_norm": 0.5560157895088196, "learning_rate": 3.1559987545650705e-07, "loss": 0.5661, "step": 9519 }, { "epoch": 0.08562190553709308, "grad_norm": 0.8903465270996094, "learning_rate": 3.142917276123564e-07, "loss": 0.8708, "step": 9520 }, { "epoch": 0.08563089943473354, "grad_norm": 1.1681092977523804, "learning_rate": 3.129862793601529e-07, "loss": 0.4033, "step": 9521 }, { "epoch": 0.08563989333237398, "grad_norm": 0.34574535489082336, "learning_rate": 3.116835308426491e-07, "loss": 0.762, "step": 9522 }, { "epoch": 0.08564888723001443, "grad_norm": 0.641056478023529, "learning_rate": 3.103834822023172e-07, "loss": 0.6653, "step": 9523 }, { "epoch": 0.08565788112765489, "grad_norm": 0.5116367340087891, "learning_rate": 3.0908613358132943e-07, "loss": 0.7032, "step": 9524 }, { "epoch": 0.08566687502529534, "grad_norm": 0.9118650555610657, "learning_rate": 3.077914851215585e-07, "loss": 1.1429, "step": 9525 }, { "epoch": 0.08567586892293579, "grad_norm": 0.9290226101875305, "learning_rate": 3.0649953696458567e-07, "loss": 0.8792, "step": 9526 }, { "epoch": 0.08568486282057623, "grad_norm": 1.470211386680603, "learning_rate": 3.052102892516978e-07, "loss": 0.104, "step": 9527 }, { "epoch": 0.0856938567182167, "grad_norm": 0.9277113676071167, "learning_rate": 3.03923742123885e-07, "loss": 0.6674, "step": 9528 }, { "epoch": 0.08570285061585714, "grad_norm": 1.5121216773986816, "learning_rate": 3.0263989572184016e-07, "loss": 0.2481, "step": 9529 }, { "epoch": 0.08571184451349759, "grad_norm": 0.3463083505630493, "learning_rate": 3.013587501859677e-07, "loss": 1.3125, "step": 9530 }, { "epoch": 0.08572083841113805, "grad_norm": 0.6970354318618774, "learning_rate": 3.000803056563667e-07, "loss": 0.4223, "step": 9531 }, { "epoch": 0.0857298323087785, "grad_norm": 0.5909843444824219, "learning_rate": 2.9880456227284745e-07, "loss": 0.7557, "step": 9532 }, { "epoch": 0.08573882620641894, "grad_norm": 0.34916457533836365, "learning_rate": 2.9753152017492347e-07, "loss": 1.0064, "step": 9533 }, { "epoch": 0.0857478201040594, "grad_norm": 0.3694264590740204, "learning_rate": 2.962611795018139e-07, "loss": 0.8122, "step": 9534 }, { "epoch": 0.08575681400169985, "grad_norm": 0.8682466745376587, "learning_rate": 2.949935403924381e-07, "loss": 0.9949, "step": 9535 }, { "epoch": 0.0857658078993403, "grad_norm": 0.4754517078399658, "learning_rate": 2.9372860298542694e-07, "loss": 0.4442, "step": 9536 }, { "epoch": 0.08577480179698074, "grad_norm": 0.38072898983955383, "learning_rate": 2.924663674191114e-07, "loss": 0.8959, "step": 9537 }, { "epoch": 0.0857837956946212, "grad_norm": 0.5197605490684509, "learning_rate": 2.912068338315255e-07, "loss": 0.5507, "step": 9538 }, { "epoch": 0.08579278959226165, "grad_norm": 1.1133219003677368, "learning_rate": 2.8995000236040906e-07, "loss": 0.8969, "step": 9539 }, { "epoch": 0.0858017834899021, "grad_norm": 0.5801394581794739, "learning_rate": 2.886958731432132e-07, "loss": 0.6488, "step": 9540 }, { "epoch": 0.08581077738754256, "grad_norm": 0.2948589026927948, "learning_rate": 2.8744444631708114e-07, "loss": 1.1237, "step": 9541 }, { "epoch": 0.085819771285183, "grad_norm": 0.25310397148132324, "learning_rate": 2.861957220188727e-07, "loss": 1.1231, "step": 9542 }, { "epoch": 0.08582876518282345, "grad_norm": 0.5607098937034607, "learning_rate": 2.849497003851398e-07, "loss": 0.3622, "step": 9543 }, { "epoch": 0.0858377590804639, "grad_norm": 1.0200397968292236, "learning_rate": 2.8370638155215123e-07, "loss": 0.6854, "step": 9544 }, { "epoch": 0.08584675297810436, "grad_norm": 0.5255709290504456, "learning_rate": 2.824657656558705e-07, "loss": 0.3316, "step": 9545 }, { "epoch": 0.0858557468757448, "grad_norm": 0.5631219744682312, "learning_rate": 2.812278528319723e-07, "loss": 0.3934, "step": 9546 }, { "epoch": 0.08586474077338525, "grad_norm": 0.7006236910820007, "learning_rate": 2.7999264321583174e-07, "loss": 0.2734, "step": 9547 }, { "epoch": 0.08587373467102571, "grad_norm": 0.3136926591396332, "learning_rate": 2.7876013694252967e-07, "loss": 0.7308, "step": 9548 }, { "epoch": 0.08588272856866616, "grad_norm": 0.5754534602165222, "learning_rate": 2.775303341468527e-07, "loss": 0.7224, "step": 9549 }, { "epoch": 0.08589172246630661, "grad_norm": 0.2524596154689789, "learning_rate": 2.763032349632877e-07, "loss": 1.2523, "step": 9550 }, { "epoch": 0.08590071636394706, "grad_norm": 1.3109551668167114, "learning_rate": 2.7507883952603284e-07, "loss": 1.1362, "step": 9551 }, { "epoch": 0.08590971026158752, "grad_norm": 0.5910031795501709, "learning_rate": 2.7385714796897553e-07, "loss": 0.4713, "step": 9552 }, { "epoch": 0.08591870415922796, "grad_norm": 0.3975547254085541, "learning_rate": 2.726381604257311e-07, "loss": 0.5749, "step": 9553 }, { "epoch": 0.08592769805686841, "grad_norm": 0.4197461009025574, "learning_rate": 2.714218770295984e-07, "loss": 0.5798, "step": 9554 }, { "epoch": 0.08593669195450887, "grad_norm": 0.40841981768608093, "learning_rate": 2.7020829791359057e-07, "loss": 1.1427, "step": 9555 }, { "epoch": 0.08594568585214932, "grad_norm": 0.30922120809555054, "learning_rate": 2.689974232104209e-07, "loss": 0.6477, "step": 9556 }, { "epoch": 0.08595467974978976, "grad_norm": 0.8640748858451843, "learning_rate": 2.677892530525139e-07, "loss": 0.7037, "step": 9557 }, { "epoch": 0.08596367364743022, "grad_norm": 0.7744613885879517, "learning_rate": 2.6658378757198623e-07, "loss": 0.9518, "step": 9558 }, { "epoch": 0.08597266754507067, "grad_norm": 0.9431270360946655, "learning_rate": 2.6538102690067404e-07, "loss": 0.8278, "step": 9559 }, { "epoch": 0.08598166144271112, "grad_norm": 0.6028528809547424, "learning_rate": 2.641809711700999e-07, "loss": 0.3643, "step": 9560 }, { "epoch": 0.08599065534035157, "grad_norm": 0.26414352655410767, "learning_rate": 2.6298362051150884e-07, "loss": 1.1345, "step": 9561 }, { "epoch": 0.08599964923799203, "grad_norm": 1.4274510145187378, "learning_rate": 2.61788975055835e-07, "loss": 0.9132, "step": 9562 }, { "epoch": 0.08600864313563247, "grad_norm": 0.5265216827392578, "learning_rate": 2.605970349337267e-07, "loss": 0.6378, "step": 9563 }, { "epoch": 0.08601763703327292, "grad_norm": 0.45578041672706604, "learning_rate": 2.594078002755296e-07, "loss": 0.6169, "step": 9564 }, { "epoch": 0.08602663093091338, "grad_norm": 0.4186432957649231, "learning_rate": 2.582212712113008e-07, "loss": 0.6627, "step": 9565 }, { "epoch": 0.08603562482855383, "grad_norm": 0.5536884069442749, "learning_rate": 2.5703744787079485e-07, "loss": 0.3445, "step": 9566 }, { "epoch": 0.08604461872619427, "grad_norm": 0.3541940450668335, "learning_rate": 2.5585633038347213e-07, "loss": 0.8397, "step": 9567 }, { "epoch": 0.08605361262383472, "grad_norm": 0.84224534034729, "learning_rate": 2.5467791887849865e-07, "loss": 0.7531, "step": 9568 }, { "epoch": 0.08606260652147518, "grad_norm": 0.3458433449268341, "learning_rate": 2.535022134847437e-07, "loss": 0.7343, "step": 9569 }, { "epoch": 0.08607160041911563, "grad_norm": 0.3868323266506195, "learning_rate": 2.5232921433078483e-07, "loss": 0.8099, "step": 9570 }, { "epoch": 0.08608059431675608, "grad_norm": 0.2905242443084717, "learning_rate": 2.511589215448917e-07, "loss": 1.1145, "step": 9571 }, { "epoch": 0.08608958821439654, "grad_norm": 0.8894575238227844, "learning_rate": 2.4999133525504813e-07, "loss": 0.2848, "step": 9572 }, { "epoch": 0.08609858211203698, "grad_norm": 0.3548005521297455, "learning_rate": 2.488264555889436e-07, "loss": 0.3265, "step": 9573 }, { "epoch": 0.08610757600967743, "grad_norm": 0.383859783411026, "learning_rate": 2.47664282673965e-07, "loss": 0.8321, "step": 9574 }, { "epoch": 0.08611656990731789, "grad_norm": 0.6990088224411011, "learning_rate": 2.4650481663720525e-07, "loss": 0.5573, "step": 9575 }, { "epoch": 0.08612556380495834, "grad_norm": 0.5720866322517395, "learning_rate": 2.4534805760546566e-07, "loss": 0.3597, "step": 9576 }, { "epoch": 0.08613455770259879, "grad_norm": 1.1408499479293823, "learning_rate": 2.441940057052394e-07, "loss": 0.0562, "step": 9577 }, { "epoch": 0.08614355160023923, "grad_norm": 0.7245225310325623, "learning_rate": 2.430426610627395e-07, "loss": 0.7019, "step": 9578 }, { "epoch": 0.08615254549787969, "grad_norm": 0.44149789214134216, "learning_rate": 2.4189402380387083e-07, "loss": 0.6705, "step": 9579 }, { "epoch": 0.08616153939552014, "grad_norm": 0.7492411732673645, "learning_rate": 2.4074809405425225e-07, "loss": 0.4508, "step": 9580 }, { "epoch": 0.08617053329316059, "grad_norm": 0.6122934818267822, "learning_rate": 2.396048719391919e-07, "loss": 0.408, "step": 9581 }, { "epoch": 0.08617952719080105, "grad_norm": 0.4261980652809143, "learning_rate": 2.384643575837203e-07, "loss": 0.5692, "step": 9582 }, { "epoch": 0.0861885210884415, "grad_norm": 0.47951793670654297, "learning_rate": 2.3732655111255443e-07, "loss": 0.5885, "step": 9583 }, { "epoch": 0.08619751498608194, "grad_norm": 0.4908307194709778, "learning_rate": 2.3619145265013077e-07, "loss": 0.4238, "step": 9584 }, { "epoch": 0.08620650888372239, "grad_norm": 0.5452819466590881, "learning_rate": 2.3505906232057505e-07, "loss": 0.6468, "step": 9585 }, { "epoch": 0.08621550278136285, "grad_norm": 0.508270800113678, "learning_rate": 2.3392938024772993e-07, "loss": 0.8519, "step": 9586 }, { "epoch": 0.0862244966790033, "grad_norm": 0.5896286368370056, "learning_rate": 2.3280240655512987e-07, "loss": 0.7707, "step": 9587 }, { "epoch": 0.08623349057664374, "grad_norm": 2.059823513031006, "learning_rate": 2.316781413660235e-07, "loss": 0.5626, "step": 9588 }, { "epoch": 0.0862424844742842, "grad_norm": 0.4743749499320984, "learning_rate": 2.3055658480335418e-07, "loss": 0.3406, "step": 9589 }, { "epoch": 0.08625147837192465, "grad_norm": 0.5712162852287292, "learning_rate": 2.2943773698977932e-07, "loss": 0.4051, "step": 9590 }, { "epoch": 0.0862604722695651, "grad_norm": 0.9651012420654297, "learning_rate": 2.283215980476483e-07, "loss": 0.0711, "step": 9591 }, { "epoch": 0.08626946616720556, "grad_norm": 0.5918658375740051, "learning_rate": 2.2720816809902734e-07, "loss": 0.7806, "step": 9592 }, { "epoch": 0.086278460064846, "grad_norm": 0.4161301255226135, "learning_rate": 2.260974472656746e-07, "loss": 0.5922, "step": 9593 }, { "epoch": 0.08628745396248645, "grad_norm": 0.7368460297584534, "learning_rate": 2.2498943566905684e-07, "loss": 0.4057, "step": 9594 }, { "epoch": 0.0862964478601269, "grad_norm": 0.28313764929771423, "learning_rate": 2.2388413343034652e-07, "loss": 0.7872, "step": 9595 }, { "epoch": 0.08630544175776736, "grad_norm": 0.5643503665924072, "learning_rate": 2.2278154067041645e-07, "loss": 1.0567, "step": 9596 }, { "epoch": 0.0863144356554078, "grad_norm": 0.5939779877662659, "learning_rate": 2.2168165750984516e-07, "loss": 0.4505, "step": 9597 }, { "epoch": 0.08632342955304825, "grad_norm": 0.30308717489242554, "learning_rate": 2.205844840689142e-07, "loss": 0.921, "step": 9598 }, { "epoch": 0.08633242345068871, "grad_norm": 0.47134190797805786, "learning_rate": 2.194900204676109e-07, "loss": 0.6529, "step": 9599 }, { "epoch": 0.08634141734832916, "grad_norm": 0.34956473112106323, "learning_rate": 2.1839826682562015e-07, "loss": 0.6509, "step": 9600 }, { "epoch": 0.08635041124596961, "grad_norm": 0.3899216949939728, "learning_rate": 2.1730922326233803e-07, "loss": 0.7593, "step": 9601 }, { "epoch": 0.08635940514361005, "grad_norm": 0.6173056960105896, "learning_rate": 2.1622288989685546e-07, "loss": 0.6987, "step": 9602 }, { "epoch": 0.08636839904125052, "grad_norm": 0.4192180037498474, "learning_rate": 2.1513926684797735e-07, "loss": 0.6858, "step": 9603 }, { "epoch": 0.08637739293889096, "grad_norm": 1.084518551826477, "learning_rate": 2.1405835423420618e-07, "loss": 0.9268, "step": 9604 }, { "epoch": 0.08638638683653141, "grad_norm": 0.4302937388420105, "learning_rate": 2.129801521737501e-07, "loss": 0.4402, "step": 9605 }, { "epoch": 0.08639538073417187, "grad_norm": 0.5119165778160095, "learning_rate": 2.1190466078451488e-07, "loss": 0.5255, "step": 9606 }, { "epoch": 0.08640437463181232, "grad_norm": 0.45246317982673645, "learning_rate": 2.1083188018412025e-07, "loss": 0.9693, "step": 9607 }, { "epoch": 0.08641336852945276, "grad_norm": 0.5962128639221191, "learning_rate": 2.0976181048987798e-07, "loss": 0.7333, "step": 9608 }, { "epoch": 0.08642236242709321, "grad_norm": 0.30574992299079895, "learning_rate": 2.0869445181881386e-07, "loss": 0.7068, "step": 9609 }, { "epoch": 0.08643135632473367, "grad_norm": 0.6532962322235107, "learning_rate": 2.0762980428765122e-07, "loss": 0.5571, "step": 9610 }, { "epoch": 0.08644035022237412, "grad_norm": 0.34875059127807617, "learning_rate": 2.0656786801281913e-07, "loss": 0.9068, "step": 9611 }, { "epoch": 0.08644934412001457, "grad_norm": 1.7678126096725464, "learning_rate": 2.0550864311044692e-07, "loss": 0.348, "step": 9612 }, { "epoch": 0.08645833801765503, "grad_norm": 0.6958022117614746, "learning_rate": 2.0445212969637527e-07, "loss": 0.5013, "step": 9613 }, { "epoch": 0.08646733191529547, "grad_norm": 0.41745316982269287, "learning_rate": 2.0339832788613678e-07, "loss": 0.7158, "step": 9614 }, { "epoch": 0.08647632581293592, "grad_norm": 0.47792160511016846, "learning_rate": 2.0234723779497533e-07, "loss": 0.5549, "step": 9615 }, { "epoch": 0.08648531971057638, "grad_norm": 0.6063859462738037, "learning_rate": 2.012988595378379e-07, "loss": 0.5313, "step": 9616 }, { "epoch": 0.08649431360821683, "grad_norm": 0.6578714847564697, "learning_rate": 2.002531932293744e-07, "loss": 0.5622, "step": 9617 }, { "epoch": 0.08650330750585727, "grad_norm": 0.37047749757766724, "learning_rate": 1.9921023898393233e-07, "loss": 1.47, "step": 9618 }, { "epoch": 0.08651230140349772, "grad_norm": 0.8852244019508362, "learning_rate": 1.9816999691557315e-07, "loss": 0.3134, "step": 9619 }, { "epoch": 0.08652129530113818, "grad_norm": 0.5123200416564941, "learning_rate": 1.9713246713805588e-07, "loss": 0.4438, "step": 9620 }, { "epoch": 0.08653028919877863, "grad_norm": 0.5061403512954712, "learning_rate": 1.960976497648398e-07, "loss": 0.3745, "step": 9621 }, { "epoch": 0.08653928309641908, "grad_norm": 0.2858599126338959, "learning_rate": 1.950655449090899e-07, "loss": 0.9418, "step": 9622 }, { "epoch": 0.08654827699405954, "grad_norm": 0.522983193397522, "learning_rate": 1.940361526836826e-07, "loss": 0.5827, "step": 9623 }, { "epoch": 0.08655727089169998, "grad_norm": 0.4299454391002655, "learning_rate": 1.9300947320118335e-07, "loss": 0.6023, "step": 9624 }, { "epoch": 0.08656626478934043, "grad_norm": 0.30001160502433777, "learning_rate": 1.919855065738746e-07, "loss": 0.6901, "step": 9625 }, { "epoch": 0.08657525868698088, "grad_norm": 0.5898848176002502, "learning_rate": 1.9096425291373065e-07, "loss": 0.252, "step": 9626 }, { "epoch": 0.08658425258462134, "grad_norm": 0.4723604917526245, "learning_rate": 1.899457123324344e-07, "loss": 0.6301, "step": 9627 }, { "epoch": 0.08659324648226178, "grad_norm": 0.4636836647987366, "learning_rate": 1.8892988494137732e-07, "loss": 0.5942, "step": 9628 }, { "epoch": 0.08660224037990223, "grad_norm": 0.45409372448921204, "learning_rate": 1.8791677085164273e-07, "loss": 0.732, "step": 9629 }, { "epoch": 0.08661123427754269, "grad_norm": 0.502044677734375, "learning_rate": 1.8690637017402534e-07, "loss": 0.7756, "step": 9630 }, { "epoch": 0.08662022817518314, "grad_norm": 0.7287412285804749, "learning_rate": 1.858986830190229e-07, "loss": 0.4237, "step": 9631 }, { "epoch": 0.08662922207282359, "grad_norm": 0.5236765742301941, "learning_rate": 1.8489370949683339e-07, "loss": 0.6599, "step": 9632 }, { "epoch": 0.08663821597046405, "grad_norm": 0.4137594401836395, "learning_rate": 1.838914497173577e-07, "loss": 0.5959, "step": 9633 }, { "epoch": 0.0866472098681045, "grad_norm": 0.277041494846344, "learning_rate": 1.8289190379020271e-07, "loss": 0.6986, "step": 9634 }, { "epoch": 0.08665620376574494, "grad_norm": 0.8322303295135498, "learning_rate": 1.8189507182467535e-07, "loss": 0.352, "step": 9635 }, { "epoch": 0.08666519766338539, "grad_norm": 0.6151559352874756, "learning_rate": 1.80900953929794e-07, "loss": 0.453, "step": 9636 }, { "epoch": 0.08667419156102585, "grad_norm": 0.9508256912231445, "learning_rate": 1.7990955021426615e-07, "loss": 0.5715, "step": 9637 }, { "epoch": 0.0866831854586663, "grad_norm": 0.5516343712806702, "learning_rate": 1.7892086078651616e-07, "loss": 0.7356, "step": 9638 }, { "epoch": 0.08669217935630674, "grad_norm": 0.44529786705970764, "learning_rate": 1.7793488575466034e-07, "loss": 0.8475, "step": 9639 }, { "epoch": 0.0867011732539472, "grad_norm": 0.30425897240638733, "learning_rate": 1.7695162522652353e-07, "loss": 0.869, "step": 9640 }, { "epoch": 0.08671016715158765, "grad_norm": 1.2733925580978394, "learning_rate": 1.759710793096392e-07, "loss": 0.6901, "step": 9641 }, { "epoch": 0.0867191610492281, "grad_norm": 0.4212096631526947, "learning_rate": 1.7499324811123542e-07, "loss": 0.3335, "step": 9642 }, { "epoch": 0.08672815494686854, "grad_norm": 0.29077112674713135, "learning_rate": 1.740181317382461e-07, "loss": 0.6875, "step": 9643 }, { "epoch": 0.086737148844509, "grad_norm": 1.5133603811264038, "learning_rate": 1.730457302973082e-07, "loss": 0.3974, "step": 9644 }, { "epoch": 0.08674614274214945, "grad_norm": 0.3910375237464905, "learning_rate": 1.720760438947644e-07, "loss": 0.6776, "step": 9645 }, { "epoch": 0.0867551366397899, "grad_norm": 0.5541115403175354, "learning_rate": 1.711090726366521e-07, "loss": 0.7137, "step": 9646 }, { "epoch": 0.08676413053743036, "grad_norm": 0.4927518665790558, "learning_rate": 1.7014481662872562e-07, "loss": 0.8264, "step": 9647 }, { "epoch": 0.0867731244350708, "grad_norm": 0.4583251178264618, "learning_rate": 1.6918327597642835e-07, "loss": 0.8029, "step": 9648 }, { "epoch": 0.08678211833271125, "grad_norm": 0.4451747536659241, "learning_rate": 1.6822445078491788e-07, "loss": 0.8411, "step": 9649 }, { "epoch": 0.0867911122303517, "grad_norm": 0.41033849120140076, "learning_rate": 1.6726834115904643e-07, "loss": 0.667, "step": 9650 }, { "epoch": 0.08680010612799216, "grad_norm": 1.0858383178710938, "learning_rate": 1.6631494720337482e-07, "loss": 0.0549, "step": 9651 }, { "epoch": 0.08680910002563261, "grad_norm": 0.5200047492980957, "learning_rate": 1.653642690221613e-07, "loss": 0.6769, "step": 9652 }, { "epoch": 0.08681809392327305, "grad_norm": 0.5026964545249939, "learning_rate": 1.644163067193727e-07, "loss": 0.633, "step": 9653 }, { "epoch": 0.08682708782091351, "grad_norm": 0.28616058826446533, "learning_rate": 1.6347106039867887e-07, "loss": 1.0665, "step": 9654 }, { "epoch": 0.08683608171855396, "grad_norm": 0.9941072463989258, "learning_rate": 1.6252853016344716e-07, "loss": 0.3954, "step": 9655 }, { "epoch": 0.08684507561619441, "grad_norm": 0.2852889895439148, "learning_rate": 1.6158871611675341e-07, "loss": 0.6686, "step": 9656 }, { "epoch": 0.08685406951383487, "grad_norm": 0.594208300113678, "learning_rate": 1.6065161836137654e-07, "loss": 0.2623, "step": 9657 }, { "epoch": 0.08686306341147532, "grad_norm": 0.9722852110862732, "learning_rate": 1.5971723699979013e-07, "loss": 0.3277, "step": 9658 }, { "epoch": 0.08687205730911576, "grad_norm": 0.41030192375183105, "learning_rate": 1.587855721341819e-07, "loss": 0.3991, "step": 9659 }, { "epoch": 0.08688105120675621, "grad_norm": 0.417084664106369, "learning_rate": 1.578566238664314e-07, "loss": 0.4736, "step": 9660 }, { "epoch": 0.08689004510439667, "grad_norm": 1.5560256242752075, "learning_rate": 1.5693039229813522e-07, "loss": 0.7423, "step": 9661 }, { "epoch": 0.08689903900203712, "grad_norm": 0.8666066527366638, "learning_rate": 1.5600687753057896e-07, "loss": 0.5901, "step": 9662 }, { "epoch": 0.08690803289967756, "grad_norm": 0.2909574806690216, "learning_rate": 1.5508607966475687e-07, "loss": 0.9819, "step": 9663 }, { "epoch": 0.08691702679731803, "grad_norm": 0.9745597839355469, "learning_rate": 1.541679988013689e-07, "loss": 0.5096, "step": 9664 }, { "epoch": 0.08692602069495847, "grad_norm": 0.6601591110229492, "learning_rate": 1.5325263504081256e-07, "loss": 0.5802, "step": 9665 }, { "epoch": 0.08693501459259892, "grad_norm": 0.35133519768714905, "learning_rate": 1.523399884831911e-07, "loss": 0.652, "step": 9666 }, { "epoch": 0.08694400849023937, "grad_norm": 0.404897004365921, "learning_rate": 1.5143005922831355e-07, "loss": 1.0008, "step": 9667 }, { "epoch": 0.08695300238787983, "grad_norm": 1.6706066131591797, "learning_rate": 1.505228473756809e-07, "loss": 0.4931, "step": 9668 }, { "epoch": 0.08696199628552027, "grad_norm": 0.523152768611908, "learning_rate": 1.4961835302451098e-07, "loss": 0.3497, "step": 9669 }, { "epoch": 0.08697099018316072, "grad_norm": 0.39704251289367676, "learning_rate": 1.4871657627371916e-07, "loss": 0.9681, "step": 9670 }, { "epoch": 0.08697998408080118, "grad_norm": 0.35227182507514954, "learning_rate": 1.4781751722191538e-07, "loss": 0.8047, "step": 9671 }, { "epoch": 0.08698897797844163, "grad_norm": 0.6107291579246521, "learning_rate": 1.4692117596742383e-07, "loss": 0.4469, "step": 9672 }, { "epoch": 0.08699797187608208, "grad_norm": 0.8045852780342102, "learning_rate": 1.4602755260826606e-07, "loss": 0.0809, "step": 9673 }, { "epoch": 0.08700696577372254, "grad_norm": 1.0454318523406982, "learning_rate": 1.4513664724216947e-07, "loss": 0.7328, "step": 9674 }, { "epoch": 0.08701595967136298, "grad_norm": 0.6887835264205933, "learning_rate": 1.4424845996655888e-07, "loss": 0.3991, "step": 9675 }, { "epoch": 0.08702495356900343, "grad_norm": 0.7048807144165039, "learning_rate": 1.4336299087856775e-07, "loss": 0.3486, "step": 9676 }, { "epoch": 0.08703394746664388, "grad_norm": 0.35998663306236267, "learning_rate": 1.424802400750269e-07, "loss": 1.1402, "step": 9677 }, { "epoch": 0.08704294136428434, "grad_norm": 0.4910534918308258, "learning_rate": 1.4160020765247862e-07, "loss": 0.7128, "step": 9678 }, { "epoch": 0.08705193526192478, "grad_norm": 0.5616055130958557, "learning_rate": 1.407228937071514e-07, "loss": 1.0654, "step": 9679 }, { "epoch": 0.08706092915956523, "grad_norm": 0.31396710872650146, "learning_rate": 1.3984829833499636e-07, "loss": 0.7479, "step": 9680 }, { "epoch": 0.08706992305720569, "grad_norm": 1.0095430612564087, "learning_rate": 1.389764216316536e-07, "loss": 0.4945, "step": 9681 }, { "epoch": 0.08707891695484614, "grad_norm": 0.7094261646270752, "learning_rate": 1.3810726369247462e-07, "loss": 0.6489, "step": 9682 }, { "epoch": 0.08708791085248659, "grad_norm": 0.7238727807998657, "learning_rate": 1.3724082461250287e-07, "loss": 0.6844, "step": 9683 }, { "epoch": 0.08709690475012703, "grad_norm": 0.3615632951259613, "learning_rate": 1.3637710448649587e-07, "loss": 0.7974, "step": 9684 }, { "epoch": 0.0871058986477675, "grad_norm": 0.3289174437522888, "learning_rate": 1.3551610340890307e-07, "loss": 0.8959, "step": 9685 }, { "epoch": 0.08711489254540794, "grad_norm": 1.0241656303405762, "learning_rate": 1.3465782147389082e-07, "loss": 0.5834, "step": 9686 }, { "epoch": 0.08712388644304839, "grad_norm": 0.49514299631118774, "learning_rate": 1.3380225877531184e-07, "loss": 0.728, "step": 9687 }, { "epoch": 0.08713288034068885, "grad_norm": 0.4237559139728546, "learning_rate": 1.329494154067329e-07, "loss": 0.5655, "step": 9688 }, { "epoch": 0.0871418742383293, "grad_norm": 0.803676426410675, "learning_rate": 1.320992914614183e-07, "loss": 0.7649, "step": 9689 }, { "epoch": 0.08715086813596974, "grad_norm": 0.31434008479118347, "learning_rate": 1.3125188703233816e-07, "loss": 0.7922, "step": 9690 }, { "epoch": 0.08715986203361019, "grad_norm": 0.7463939189910889, "learning_rate": 1.3040720221215997e-07, "loss": 0.6602, "step": 9691 }, { "epoch": 0.08716885593125065, "grad_norm": 0.9763000011444092, "learning_rate": 1.2956523709325986e-07, "loss": 0.2498, "step": 9692 }, { "epoch": 0.0871778498288911, "grad_norm": 0.44674408435821533, "learning_rate": 1.2872599176771417e-07, "loss": 0.5915, "step": 9693 }, { "epoch": 0.08718684372653154, "grad_norm": 1.9579644203186035, "learning_rate": 1.2788946632729947e-07, "loss": 0.6073, "step": 9694 }, { "epoch": 0.087195837624172, "grad_norm": 0.538741946220398, "learning_rate": 1.2705566086350097e-07, "loss": 0.3339, "step": 9695 }, { "epoch": 0.08720483152181245, "grad_norm": 0.3507225811481476, "learning_rate": 1.2622457546749567e-07, "loss": 0.9908, "step": 9696 }, { "epoch": 0.0872138254194529, "grad_norm": 0.39096277952194214, "learning_rate": 1.2539621023017756e-07, "loss": 1.0227, "step": 9697 }, { "epoch": 0.08722281931709336, "grad_norm": 0.4510818123817444, "learning_rate": 1.2457056524212696e-07, "loss": 0.6453, "step": 9698 }, { "epoch": 0.0872318132147338, "grad_norm": 0.4953216314315796, "learning_rate": 1.2374764059364386e-07, "loss": 0.6411, "step": 9699 }, { "epoch": 0.08724080711237425, "grad_norm": 0.9708173871040344, "learning_rate": 1.229274363747146e-07, "loss": 0.3611, "step": 9700 }, { "epoch": 0.0872498010100147, "grad_norm": 0.3456757068634033, "learning_rate": 1.2210995267503966e-07, "loss": 1.0515, "step": 9701 }, { "epoch": 0.08725879490765516, "grad_norm": 0.44572827219963074, "learning_rate": 1.2129518958401974e-07, "loss": 0.5548, "step": 9702 }, { "epoch": 0.0872677888052956, "grad_norm": 0.586336612701416, "learning_rate": 1.2048314719075027e-07, "loss": 0.7418, "step": 9703 }, { "epoch": 0.08727678270293605, "grad_norm": 1.8082722425460815, "learning_rate": 1.1967382558403794e-07, "loss": 0.3404, "step": 9704 }, { "epoch": 0.08728577660057651, "grad_norm": 0.4260518252849579, "learning_rate": 1.1886722485238978e-07, "loss": 1.0208, "step": 9705 }, { "epoch": 0.08729477049821696, "grad_norm": 1.964622974395752, "learning_rate": 1.1806334508401296e-07, "loss": 0.5852, "step": 9706 }, { "epoch": 0.08730376439585741, "grad_norm": 0.4210047721862793, "learning_rate": 1.172621863668233e-07, "loss": 0.7027, "step": 9707 }, { "epoch": 0.08731275829349786, "grad_norm": 0.5997663736343384, "learning_rate": 1.164637487884257e-07, "loss": 0.57, "step": 9708 }, { "epoch": 0.08732175219113832, "grad_norm": 0.6278613209724426, "learning_rate": 1.1566803243614199e-07, "loss": 0.9684, "step": 9709 }, { "epoch": 0.08733074608877876, "grad_norm": 0.6893112659454346, "learning_rate": 1.1487503739698869e-07, "loss": 0.8053, "step": 9710 }, { "epoch": 0.08733973998641921, "grad_norm": 0.5472863912582397, "learning_rate": 1.1408476375768806e-07, "loss": 0.718, "step": 9711 }, { "epoch": 0.08734873388405967, "grad_norm": 0.2888655662536621, "learning_rate": 1.1329721160466267e-07, "loss": 0.9577, "step": 9712 }, { "epoch": 0.08735772778170012, "grad_norm": 1.9033401012420654, "learning_rate": 1.1251238102403528e-07, "loss": 0.2227, "step": 9713 }, { "epoch": 0.08736672167934056, "grad_norm": 0.436921089887619, "learning_rate": 1.1173027210164e-07, "loss": 0.5634, "step": 9714 }, { "epoch": 0.08737571557698103, "grad_norm": 0.3698715269565582, "learning_rate": 1.109508849230001e-07, "loss": 0.8568, "step": 9715 }, { "epoch": 0.08738470947462147, "grad_norm": 0.4482980966567993, "learning_rate": 1.1017421957335294e-07, "loss": 0.6157, "step": 9716 }, { "epoch": 0.08739370337226192, "grad_norm": 0.5292548537254333, "learning_rate": 1.094002761376306e-07, "loss": 0.9108, "step": 9717 }, { "epoch": 0.08740269726990237, "grad_norm": 0.46584025025367737, "learning_rate": 1.0862905470047091e-07, "loss": 1.049, "step": 9718 }, { "epoch": 0.08741169116754283, "grad_norm": 0.41480761766433716, "learning_rate": 1.0786055534621475e-07, "loss": 1.2516, "step": 9719 }, { "epoch": 0.08742068506518327, "grad_norm": 0.4582068920135498, "learning_rate": 1.0709477815890601e-07, "loss": 0.7466, "step": 9720 }, { "epoch": 0.08742967896282372, "grad_norm": 0.8266103863716125, "learning_rate": 1.0633172322228325e-07, "loss": 0.5408, "step": 9721 }, { "epoch": 0.08743867286046418, "grad_norm": 0.2624119520187378, "learning_rate": 1.0557139061979915e-07, "loss": 1.1432, "step": 9722 }, { "epoch": 0.08744766675810463, "grad_norm": 0.7597643136978149, "learning_rate": 1.0481378043459556e-07, "loss": 0.4495, "step": 9723 }, { "epoch": 0.08745666065574507, "grad_norm": 0.6174414157867432, "learning_rate": 1.0405889274953118e-07, "loss": 0.7587, "step": 9724 }, { "epoch": 0.08746565455338552, "grad_norm": 1.2125091552734375, "learning_rate": 1.0330672764715387e-07, "loss": 0.2396, "step": 9725 }, { "epoch": 0.08747464845102598, "grad_norm": 1.96486496925354, "learning_rate": 1.0255728520972285e-07, "loss": 0.57, "step": 9726 }, { "epoch": 0.08748364234866643, "grad_norm": 2.847245693206787, "learning_rate": 1.0181056551919199e-07, "loss": 0.4598, "step": 9727 }, { "epoch": 0.08749263624630688, "grad_norm": 0.5408509373664856, "learning_rate": 1.0106656865722652e-07, "loss": 0.2436, "step": 9728 }, { "epoch": 0.08750163014394734, "grad_norm": 0.394688218832016, "learning_rate": 1.003252947051836e-07, "loss": 0.9332, "step": 9729 }, { "epoch": 0.08751062404158778, "grad_norm": 0.33946293592453003, "learning_rate": 9.958674374413168e-08, "loss": 0.8403, "step": 9730 }, { "epoch": 0.08751961793922823, "grad_norm": 0.36455532908439636, "learning_rate": 9.885091585483397e-08, "loss": 0.9666, "step": 9731 }, { "epoch": 0.08752861183686868, "grad_norm": 0.803732693195343, "learning_rate": 9.811781111776497e-08, "loss": 0.5168, "step": 9732 }, { "epoch": 0.08753760573450914, "grad_norm": 0.4172307550907135, "learning_rate": 9.738742961309111e-08, "loss": 1.0365, "step": 9733 }, { "epoch": 0.08754659963214959, "grad_norm": 0.3151185214519501, "learning_rate": 9.665977142068738e-08, "loss": 1.0653, "step": 9734 }, { "epoch": 0.08755559352979003, "grad_norm": 0.707878053188324, "learning_rate": 9.59348366201318e-08, "loss": 0.7274, "step": 9735 }, { "epoch": 0.08756458742743049, "grad_norm": 0.4735535681247711, "learning_rate": 9.521262529069707e-08, "loss": 0.7745, "step": 9736 }, { "epoch": 0.08757358132507094, "grad_norm": 0.2894141674041748, "learning_rate": 9.449313751136724e-08, "loss": 1.0846, "step": 9737 }, { "epoch": 0.08758257522271139, "grad_norm": 0.6132954359054565, "learning_rate": 9.377637336082101e-08, "loss": 0.7508, "step": 9738 }, { "epoch": 0.08759156912035185, "grad_norm": 0.8208059668540955, "learning_rate": 9.306233291744848e-08, "loss": 0.4546, "step": 9739 }, { "epoch": 0.0876005630179923, "grad_norm": 0.5649977922439575, "learning_rate": 9.235101625932885e-08, "loss": 0.6825, "step": 9740 }, { "epoch": 0.08760955691563274, "grad_norm": 0.45267587900161743, "learning_rate": 9.164242346425544e-08, "loss": 0.6518, "step": 9741 }, { "epoch": 0.08761855081327319, "grad_norm": 0.36832746863365173, "learning_rate": 9.093655460971628e-08, "loss": 0.8218, "step": 9742 }, { "epoch": 0.08762754471091365, "grad_norm": 0.800489604473114, "learning_rate": 9.023340977290518e-08, "loss": 0.0577, "step": 9743 }, { "epoch": 0.0876365386085541, "grad_norm": 0.278694212436676, "learning_rate": 8.953298903071895e-08, "loss": 0.9184, "step": 9744 }, { "epoch": 0.08764553250619454, "grad_norm": 0.6898355484008789, "learning_rate": 8.883529245975186e-08, "loss": 0.6635, "step": 9745 }, { "epoch": 0.087654526403835, "grad_norm": 0.7882290482521057, "learning_rate": 8.814032013630402e-08, "loss": 0.343, "step": 9746 }, { "epoch": 0.08766352030147545, "grad_norm": 0.6116306185722351, "learning_rate": 8.744807213637573e-08, "loss": 0.4978, "step": 9747 }, { "epoch": 0.0876725141991159, "grad_norm": 0.5513296723365784, "learning_rate": 8.675854853567034e-08, "loss": 0.8434, "step": 9748 }, { "epoch": 0.08768150809675634, "grad_norm": 0.4932820796966553, "learning_rate": 8.607174940959417e-08, "loss": 0.5244, "step": 9749 }, { "epoch": 0.0876905019943968, "grad_norm": 0.3705962598323822, "learning_rate": 8.538767483325383e-08, "loss": 0.9965, "step": 9750 }, { "epoch": 0.08769949589203725, "grad_norm": 0.5658516883850098, "learning_rate": 8.470632488145891e-08, "loss": 0.3603, "step": 9751 }, { "epoch": 0.0877084897896777, "grad_norm": 0.4061141312122345, "learning_rate": 8.402769962872203e-08, "loss": 0.6477, "step": 9752 }, { "epoch": 0.08771748368731816, "grad_norm": 0.669914960861206, "learning_rate": 8.335179914925328e-08, "loss": 0.8305, "step": 9753 }, { "epoch": 0.0877264775849586, "grad_norm": 0.598192572593689, "learning_rate": 8.26786235169713e-08, "loss": 0.5423, "step": 9754 }, { "epoch": 0.08773547148259905, "grad_norm": 0.6846851706504822, "learning_rate": 8.200817280549222e-08, "loss": 0.8971, "step": 9755 }, { "epoch": 0.08774446538023951, "grad_norm": 1.5792680978775024, "learning_rate": 8.134044708813515e-08, "loss": 0.6531, "step": 9756 }, { "epoch": 0.08775345927787996, "grad_norm": 0.7289961576461792, "learning_rate": 8.067544643792224e-08, "loss": 0.4379, "step": 9757 }, { "epoch": 0.08776245317552041, "grad_norm": 0.3618885278701782, "learning_rate": 8.001317092757864e-08, "loss": 0.9146, "step": 9758 }, { "epoch": 0.08777144707316085, "grad_norm": 0.5006494522094727, "learning_rate": 7.935362062952423e-08, "loss": 0.6358, "step": 9759 }, { "epoch": 0.08778044097080132, "grad_norm": 0.31008368730545044, "learning_rate": 7.869679561589293e-08, "loss": 0.7413, "step": 9760 }, { "epoch": 0.08778943486844176, "grad_norm": 0.38548943400382996, "learning_rate": 7.804269595851066e-08, "loss": 1.0021, "step": 9761 }, { "epoch": 0.08779842876608221, "grad_norm": 2.0462870597839355, "learning_rate": 7.739132172891184e-08, "loss": 0.5508, "step": 9762 }, { "epoch": 0.08780742266372267, "grad_norm": 0.3944482207298279, "learning_rate": 7.674267299832561e-08, "loss": 0.1993, "step": 9763 }, { "epoch": 0.08781641656136312, "grad_norm": 0.46170783042907715, "learning_rate": 7.609674983768966e-08, "loss": 0.57, "step": 9764 }, { "epoch": 0.08782541045900356, "grad_norm": 1.3623892068862915, "learning_rate": 7.545355231763918e-08, "loss": 0.7733, "step": 9765 }, { "epoch": 0.08783440435664401, "grad_norm": 0.7594032287597656, "learning_rate": 7.48130805085151e-08, "loss": 0.4401, "step": 9766 }, { "epoch": 0.08784339825428447, "grad_norm": 0.9845412969589233, "learning_rate": 7.417533448035863e-08, "loss": 0.2377, "step": 9767 }, { "epoch": 0.08785239215192492, "grad_norm": 0.7111296057701111, "learning_rate": 7.35403143029112e-08, "loss": 0.731, "step": 9768 }, { "epoch": 0.08786138604956537, "grad_norm": 0.31395936012268066, "learning_rate": 7.290802004562003e-08, "loss": 0.7618, "step": 9769 }, { "epoch": 0.08787037994720583, "grad_norm": 0.4095287024974823, "learning_rate": 7.227845177762982e-08, "loss": 0.7641, "step": 9770 }, { "epoch": 0.08787937384484627, "grad_norm": 1.2330039739608765, "learning_rate": 7.165160956779105e-08, "loss": 0.8165, "step": 9771 }, { "epoch": 0.08788836774248672, "grad_norm": 0.4303819239139557, "learning_rate": 7.102749348465165e-08, "loss": 0.707, "step": 9772 }, { "epoch": 0.08789736164012717, "grad_norm": 0.2446850836277008, "learning_rate": 7.040610359646537e-08, "loss": 1.0389, "step": 9773 }, { "epoch": 0.08790635553776763, "grad_norm": 0.779285728931427, "learning_rate": 6.97874399711862e-08, "loss": 0.4491, "step": 9774 }, { "epoch": 0.08791534943540807, "grad_norm": 0.620631217956543, "learning_rate": 6.91715026764711e-08, "loss": 0.7087, "step": 9775 }, { "epoch": 0.08792434333304852, "grad_norm": 0.48732948303222656, "learning_rate": 6.85582917796801e-08, "loss": 0.7222, "step": 9776 }, { "epoch": 0.08793333723068898, "grad_norm": 0.29294946789741516, "learning_rate": 6.794780734786788e-08, "loss": 0.9805, "step": 9777 }, { "epoch": 0.08794233112832943, "grad_norm": 0.6008402109146118, "learning_rate": 6.73400494477977e-08, "loss": 0.934, "step": 9778 }, { "epoch": 0.08795132502596988, "grad_norm": 0.4933336675167084, "learning_rate": 6.673501814593586e-08, "loss": 1.1243, "step": 9779 }, { "epoch": 0.08796031892361034, "grad_norm": 0.4320085346698761, "learning_rate": 6.613271350844608e-08, "loss": 0.8956, "step": 9780 }, { "epoch": 0.08796931282125078, "grad_norm": 0.4508405923843384, "learning_rate": 6.553313560119234e-08, "loss": 0.8014, "step": 9781 }, { "epoch": 0.08797830671889123, "grad_norm": 0.5674914717674255, "learning_rate": 6.493628448974998e-08, "loss": 0.7941, "step": 9782 }, { "epoch": 0.08798730061653168, "grad_norm": 0.33208805322647095, "learning_rate": 6.434216023938344e-08, "loss": 0.0581, "step": 9783 }, { "epoch": 0.08799629451417214, "grad_norm": 0.39595291018486023, "learning_rate": 6.375076291506854e-08, "loss": 0.7977, "step": 9784 }, { "epoch": 0.08800528841181258, "grad_norm": 0.39892643690109253, "learning_rate": 6.316209258148131e-08, "loss": 0.5133, "step": 9785 }, { "epoch": 0.08801428230945303, "grad_norm": 1.1019198894500732, "learning_rate": 6.257614930299527e-08, "loss": 0.0678, "step": 9786 }, { "epoch": 0.08802327620709349, "grad_norm": 0.5051602721214294, "learning_rate": 6.199293314368692e-08, "loss": 0.5473, "step": 9787 }, { "epoch": 0.08803227010473394, "grad_norm": 0.42327868938446045, "learning_rate": 6.14124441673386e-08, "loss": 0.9184, "step": 9788 }, { "epoch": 0.08804126400237439, "grad_norm": 0.60744309425354, "learning_rate": 6.083468243743007e-08, "loss": 0.7356, "step": 9789 }, { "epoch": 0.08805025790001483, "grad_norm": 29.22797393798828, "learning_rate": 6.025964801714412e-08, "loss": 0.5355, "step": 9790 }, { "epoch": 0.0880592517976553, "grad_norm": 0.512666642665863, "learning_rate": 5.968734096936935e-08, "loss": 0.5202, "step": 9791 }, { "epoch": 0.08806824569529574, "grad_norm": 1.9116036891937256, "learning_rate": 5.9117761356689e-08, "loss": 0.8891, "step": 9792 }, { "epoch": 0.08807723959293619, "grad_norm": 0.3013339340686798, "learning_rate": 5.8550909241392174e-08, "loss": 0.8455, "step": 9793 }, { "epoch": 0.08808623349057665, "grad_norm": 0.35659390687942505, "learning_rate": 5.798678468546814e-08, "loss": 1.0699, "step": 9794 }, { "epoch": 0.0880952273882171, "grad_norm": 0.4437927007675171, "learning_rate": 5.742538775061201e-08, "loss": 0.9481, "step": 9795 }, { "epoch": 0.08810422128585754, "grad_norm": 0.5066931247711182, "learning_rate": 5.686671849821079e-08, "loss": 0.3867, "step": 9796 }, { "epoch": 0.088113215183498, "grad_norm": 0.3668416440486908, "learning_rate": 5.631077698936837e-08, "loss": 0.0477, "step": 9797 }, { "epoch": 0.08812220908113845, "grad_norm": 0.6580769419670105, "learning_rate": 5.5757563284875026e-08, "loss": 0.3835, "step": 9798 }, { "epoch": 0.0881312029787789, "grad_norm": 0.8575102090835571, "learning_rate": 5.520707744522957e-08, "loss": 0.3801, "step": 9799 }, { "epoch": 0.08814019687641934, "grad_norm": 0.2698526382446289, "learning_rate": 5.4659319530636633e-08, "loss": 0.7149, "step": 9800 }, { "epoch": 0.0881491907740598, "grad_norm": 0.7953270673751831, "learning_rate": 5.411428960099552e-08, "loss": 0.4866, "step": 9801 }, { "epoch": 0.08815818467170025, "grad_norm": 0.44964635372161865, "learning_rate": 5.3571987715908545e-08, "loss": 0.8104, "step": 9802 }, { "epoch": 0.0881671785693407, "grad_norm": 0.3580918312072754, "learning_rate": 5.3032413934683835e-08, "loss": 0.701, "step": 9803 }, { "epoch": 0.08817617246698116, "grad_norm": 0.384878545999527, "learning_rate": 5.249556831632696e-08, "loss": 0.8381, "step": 9804 }, { "epoch": 0.0881851663646216, "grad_norm": 0.3573433756828308, "learning_rate": 5.196145091954652e-08, "loss": 1.2377, "step": 9805 }, { "epoch": 0.08819416026226205, "grad_norm": 0.3732014298439026, "learning_rate": 5.1430061802751336e-08, "loss": 0.999, "step": 9806 }, { "epoch": 0.0882031541599025, "grad_norm": 0.2876027822494507, "learning_rate": 5.0901401024058825e-08, "loss": 1.0577, "step": 9807 }, { "epoch": 0.08821214805754296, "grad_norm": 0.6238043308258057, "learning_rate": 5.037546864127551e-08, "loss": 0.8706, "step": 9808 }, { "epoch": 0.08822114195518341, "grad_norm": 0.5116254687309265, "learning_rate": 4.985226471192206e-08, "loss": 0.5514, "step": 9809 }, { "epoch": 0.08823013585282385, "grad_norm": 0.5011522173881531, "learning_rate": 4.9331789293211026e-08, "loss": 0.7358, "step": 9810 }, { "epoch": 0.08823912975046432, "grad_norm": 0.33535805344581604, "learning_rate": 4.8814042442063553e-08, "loss": 0.8637, "step": 9811 }, { "epoch": 0.08824812364810476, "grad_norm": 0.3083898425102234, "learning_rate": 4.829902421509824e-08, "loss": 0.672, "step": 9812 }, { "epoch": 0.08825711754574521, "grad_norm": 0.38575002551078796, "learning_rate": 4.77867346686367e-08, "loss": 0.663, "step": 9813 }, { "epoch": 0.08826611144338567, "grad_norm": 0.511073887348175, "learning_rate": 4.727717385870634e-08, "loss": 0.7089, "step": 9814 }, { "epoch": 0.08827510534102612, "grad_norm": 1.0045013427734375, "learning_rate": 4.677034184102369e-08, "loss": 0.9207, "step": 9815 }, { "epoch": 0.08828409923866656, "grad_norm": 0.5878546833992004, "learning_rate": 4.6266238671024975e-08, "loss": 0.6976, "step": 9816 }, { "epoch": 0.08829309313630701, "grad_norm": 0.4010329246520996, "learning_rate": 4.576486440382999e-08, "loss": 0.5362, "step": 9817 }, { "epoch": 0.08830208703394747, "grad_norm": 0.2902154326438904, "learning_rate": 4.526621909426987e-08, "loss": 0.7596, "step": 9818 }, { "epoch": 0.08831108093158792, "grad_norm": 0.4778302013874054, "learning_rate": 4.477030279687877e-08, "loss": 0.6293, "step": 9819 }, { "epoch": 0.08832007482922836, "grad_norm": 4.4740376472473145, "learning_rate": 4.427711556588832e-08, "loss": 0.4578, "step": 9820 }, { "epoch": 0.08832906872686883, "grad_norm": 1.7984330654144287, "learning_rate": 4.378665745523036e-08, "loss": 0.3711, "step": 9821 }, { "epoch": 0.08833806262450927, "grad_norm": 1.816207766532898, "learning_rate": 4.329892851854533e-08, "loss": 0.4185, "step": 9822 }, { "epoch": 0.08834705652214972, "grad_norm": 1.017106533050537, "learning_rate": 4.281392880916557e-08, "loss": 1.0311, "step": 9823 }, { "epoch": 0.08835605041979017, "grad_norm": 0.34715867042541504, "learning_rate": 4.233165838013198e-08, "loss": 0.5941, "step": 9824 }, { "epoch": 0.08836504431743063, "grad_norm": 0.5635330677032471, "learning_rate": 4.18521172841857e-08, "loss": 0.6802, "step": 9825 }, { "epoch": 0.08837403821507107, "grad_norm": 0.4033609628677368, "learning_rate": 4.137530557376534e-08, "loss": 0.7532, "step": 9826 }, { "epoch": 0.08838303211271152, "grad_norm": 1.0165722370147705, "learning_rate": 4.0901223301020863e-08, "loss": 0.4047, "step": 9827 }, { "epoch": 0.08839202601035198, "grad_norm": 0.5271196961402893, "learning_rate": 4.0429870517791344e-08, "loss": 0.5582, "step": 9828 }, { "epoch": 0.08840101990799243, "grad_norm": 0.5750731229782104, "learning_rate": 3.996124727562445e-08, "loss": 0.7387, "step": 9829 }, { "epoch": 0.08841001380563288, "grad_norm": 0.5653998255729675, "learning_rate": 3.949535362576806e-08, "loss": 0.3687, "step": 9830 }, { "epoch": 0.08841900770327332, "grad_norm": 0.734752893447876, "learning_rate": 3.9032189619173096e-08, "loss": 0.7166, "step": 9831 }, { "epoch": 0.08842800160091378, "grad_norm": 0.47661370038986206, "learning_rate": 3.857175530649071e-08, "loss": 0.5081, "step": 9832 }, { "epoch": 0.08843699549855423, "grad_norm": 0.42802006006240845, "learning_rate": 3.811405073807228e-08, "loss": 0.5784, "step": 9833 }, { "epoch": 0.08844598939619468, "grad_norm": 0.6945856213569641, "learning_rate": 3.7659075963969446e-08, "loss": 0.3627, "step": 9834 }, { "epoch": 0.08845498329383514, "grad_norm": 0.7670103311538696, "learning_rate": 3.72068310339424e-08, "loss": 0.4479, "step": 9835 }, { "epoch": 0.08846397719147558, "grad_norm": 0.41893190145492554, "learning_rate": 3.6757315997443255e-08, "loss": 0.8042, "step": 9836 }, { "epoch": 0.08847297108911603, "grad_norm": 0.6763470768928528, "learning_rate": 3.631053090363268e-08, "loss": 0.4223, "step": 9837 }, { "epoch": 0.08848196498675649, "grad_norm": 0.48960620164871216, "learning_rate": 3.586647580136881e-08, "loss": 0.3721, "step": 9838 }, { "epoch": 0.08849095888439694, "grad_norm": 0.3692554831504822, "learning_rate": 3.5425150739218375e-08, "loss": 0.9435, "step": 9839 }, { "epoch": 0.08849995278203739, "grad_norm": 0.6676127910614014, "learning_rate": 3.4986555765434413e-08, "loss": 0.6206, "step": 9840 }, { "epoch": 0.08850894667967783, "grad_norm": 0.544215738773346, "learning_rate": 3.455069092798968e-08, "loss": 0.739, "step": 9841 }, { "epoch": 0.0885179405773183, "grad_norm": 0.5547282099723816, "learning_rate": 3.411755627454605e-08, "loss": 0.6455, "step": 9842 }, { "epoch": 0.08852693447495874, "grad_norm": 0.41399624943733215, "learning_rate": 3.368715185247118e-08, "loss": 0.6281, "step": 9843 }, { "epoch": 0.08853592837259919, "grad_norm": 0.35915979743003845, "learning_rate": 3.3259477708832974e-08, "loss": 0.697, "step": 9844 }, { "epoch": 0.08854492227023965, "grad_norm": 0.37786808609962463, "learning_rate": 3.283453389039959e-08, "loss": 0.9436, "step": 9845 }, { "epoch": 0.0885539161678801, "grad_norm": 1.1409732103347778, "learning_rate": 3.241232044364495e-08, "loss": 0.7567, "step": 9846 }, { "epoch": 0.08856291006552054, "grad_norm": 0.25653165578842163, "learning_rate": 3.1992837414743236e-08, "loss": 0.9206, "step": 9847 }, { "epoch": 0.08857190396316099, "grad_norm": 0.43035373091697693, "learning_rate": 3.157608484956332e-08, "loss": 0.7954, "step": 9848 }, { "epoch": 0.08858089786080145, "grad_norm": 0.7764844298362732, "learning_rate": 3.1162062793682615e-08, "loss": 0.1514, "step": 9849 }, { "epoch": 0.0885898917584419, "grad_norm": 0.4580516219139099, "learning_rate": 3.075077129238158e-08, "loss": 0.7138, "step": 9850 }, { "epoch": 0.08859888565608234, "grad_norm": 1.738675832748413, "learning_rate": 3.0342210390632566e-08, "loss": 0.831, "step": 9851 }, { "epoch": 0.0886078795537228, "grad_norm": 0.3725658655166626, "learning_rate": 2.993638013311928e-08, "loss": 0.6541, "step": 9852 }, { "epoch": 0.08861687345136325, "grad_norm": 1.4975961446762085, "learning_rate": 2.9533280564222887e-08, "loss": 0.4331, "step": 9853 }, { "epoch": 0.0886258673490037, "grad_norm": 1.044998049736023, "learning_rate": 2.9132911728022017e-08, "loss": 0.6471, "step": 9854 }, { "epoch": 0.08863486124664416, "grad_norm": 1.086905598640442, "learning_rate": 2.8735273668303863e-08, "loss": 0.4992, "step": 9855 }, { "epoch": 0.0886438551442846, "grad_norm": 0.49274322390556335, "learning_rate": 2.834036642855309e-08, "loss": 1.003, "step": 9856 }, { "epoch": 0.08865284904192505, "grad_norm": 0.520036518573761, "learning_rate": 2.7948190051954593e-08, "loss": 0.5898, "step": 9857 }, { "epoch": 0.0886618429395655, "grad_norm": 0.980014979839325, "learning_rate": 2.755874458139629e-08, "loss": 0.2441, "step": 9858 }, { "epoch": 0.08867083683720596, "grad_norm": 0.41624581813812256, "learning_rate": 2.717203005946911e-08, "loss": 0.7831, "step": 9859 }, { "epoch": 0.08867983073484641, "grad_norm": 0.6236540079116821, "learning_rate": 2.6788046528461453e-08, "loss": 0.4766, "step": 9860 }, { "epoch": 0.08868882463248685, "grad_norm": 0.7136914730072021, "learning_rate": 2.6406794030367498e-08, "loss": 0.5812, "step": 9861 }, { "epoch": 0.08869781853012731, "grad_norm": 0.4210243821144104, "learning_rate": 2.6028272606881677e-08, "loss": 0.9788, "step": 9862 }, { "epoch": 0.08870681242776776, "grad_norm": 0.4579426348209381, "learning_rate": 2.5652482299393098e-08, "loss": 0.9104, "step": 9863 }, { "epoch": 0.08871580632540821, "grad_norm": 0.4775637090206146, "learning_rate": 2.5279423149002224e-08, "loss": 0.5798, "step": 9864 }, { "epoch": 0.08872480022304866, "grad_norm": 0.3179444372653961, "learning_rate": 2.490909519650142e-08, "loss": 1.0249, "step": 9865 }, { "epoch": 0.08873379412068912, "grad_norm": 0.55256587266922, "learning_rate": 2.4541498482397173e-08, "loss": 0.7053, "step": 9866 }, { "epoch": 0.08874278801832956, "grad_norm": 0.5255864858627319, "learning_rate": 2.4176633046882335e-08, "loss": 0.7859, "step": 9867 }, { "epoch": 0.08875178191597001, "grad_norm": 0.33353424072265625, "learning_rate": 2.3814498929861097e-08, "loss": 0.6007, "step": 9868 }, { "epoch": 0.08876077581361047, "grad_norm": 0.41669395565986633, "learning_rate": 2.3455096170932332e-08, "loss": 1.1563, "step": 9869 }, { "epoch": 0.08876976971125092, "grad_norm": 0.9402024745941162, "learning_rate": 2.309842480940627e-08, "loss": 0.5139, "step": 9870 }, { "epoch": 0.08877876360889136, "grad_norm": 0.8628674745559692, "learning_rate": 2.274448488428227e-08, "loss": 0.556, "step": 9871 }, { "epoch": 0.08878775750653181, "grad_norm": 0.45098966360092163, "learning_rate": 2.2393276434268253e-08, "loss": 0.5332, "step": 9872 }, { "epoch": 0.08879675140417227, "grad_norm": 0.3501611351966858, "learning_rate": 2.2044799497772383e-08, "loss": 0.9971, "step": 9873 }, { "epoch": 0.08880574530181272, "grad_norm": 0.6656119227409363, "learning_rate": 2.1699054112905848e-08, "loss": 0.5032, "step": 9874 }, { "epoch": 0.08881473919945317, "grad_norm": 0.3768710792064667, "learning_rate": 2.1356040317474512e-08, "loss": 0.8061, "step": 9875 }, { "epoch": 0.08882373309709363, "grad_norm": 0.5580398440361023, "learning_rate": 2.1015758148992814e-08, "loss": 0.838, "step": 9876 }, { "epoch": 0.08883272699473407, "grad_norm": 0.47630321979522705, "learning_rate": 2.067820764466988e-08, "loss": 0.5563, "step": 9877 }, { "epoch": 0.08884172089237452, "grad_norm": 0.3114936053752899, "learning_rate": 2.034338884142617e-08, "loss": 0.9859, "step": 9878 }, { "epoch": 0.08885071479001498, "grad_norm": 0.43413466215133667, "learning_rate": 2.0011301775868518e-08, "loss": 0.9353, "step": 9879 }, { "epoch": 0.08885970868765543, "grad_norm": 0.45219886302948, "learning_rate": 1.9681946484320644e-08, "loss": 0.9226, "step": 9880 }, { "epoch": 0.08886870258529588, "grad_norm": 0.5200753808021545, "learning_rate": 1.9355323002798187e-08, "loss": 0.8586, "step": 9881 }, { "epoch": 0.08887769648293632, "grad_norm": 0.8080654740333557, "learning_rate": 1.903143136701979e-08, "loss": 0.0903, "step": 9882 }, { "epoch": 0.08888669038057678, "grad_norm": 0.4322017431259155, "learning_rate": 1.871027161240435e-08, "loss": 0.5224, "step": 9883 }, { "epoch": 0.08889568427821723, "grad_norm": 0.4593682885169983, "learning_rate": 1.8391843774076543e-08, "loss": 0.4398, "step": 9884 }, { "epoch": 0.08890467817585768, "grad_norm": 0.37409090995788574, "learning_rate": 1.8076147886855744e-08, "loss": 0.9877, "step": 9885 }, { "epoch": 0.08891367207349814, "grad_norm": 0.5666894316673279, "learning_rate": 1.7763183985269883e-08, "loss": 0.7598, "step": 9886 }, { "epoch": 0.08892266597113858, "grad_norm": 0.5166069269180298, "learning_rate": 1.745295210353881e-08, "loss": 0.8779, "step": 9887 }, { "epoch": 0.08893165986877903, "grad_norm": 0.5172075033187866, "learning_rate": 1.714545227559372e-08, "loss": 0.6279, "step": 9888 }, { "epoch": 0.08894065376641948, "grad_norm": 0.4938029944896698, "learning_rate": 1.6840684535060492e-08, "loss": 0.625, "step": 9889 }, { "epoch": 0.08894964766405994, "grad_norm": 1.137963056564331, "learning_rate": 1.6538648915270792e-08, "loss": 0.0493, "step": 9890 }, { "epoch": 0.08895864156170039, "grad_norm": 0.26879727840423584, "learning_rate": 1.6239345449250987e-08, "loss": 0.8161, "step": 9891 }, { "epoch": 0.08896763545934083, "grad_norm": 0.537448525428772, "learning_rate": 1.594277416973322e-08, "loss": 0.4624, "step": 9892 }, { "epoch": 0.0889766293569813, "grad_norm": 0.29955244064331055, "learning_rate": 1.5648935109149886e-08, "loss": 0.8448, "step": 9893 }, { "epoch": 0.08898562325462174, "grad_norm": 0.4542538821697235, "learning_rate": 1.535782829963639e-08, "loss": 0.6954, "step": 9894 }, { "epoch": 0.08899461715226219, "grad_norm": 0.34086111187934875, "learning_rate": 1.50694537730256e-08, "loss": 0.7559, "step": 9895 }, { "epoch": 0.08900361104990265, "grad_norm": 0.37522003054618835, "learning_rate": 1.4783811560856175e-08, "loss": 0.5508, "step": 9896 }, { "epoch": 0.0890126049475431, "grad_norm": 1.5341711044311523, "learning_rate": 1.450090169436702e-08, "loss": 0.1216, "step": 9897 }, { "epoch": 0.08902159884518354, "grad_norm": 0.5621117353439331, "learning_rate": 1.422072420449172e-08, "loss": 0.9432, "step": 9898 }, { "epoch": 0.08903059274282399, "grad_norm": 0.5264595746994019, "learning_rate": 1.3943279121872433e-08, "loss": 0.7668, "step": 9899 }, { "epoch": 0.08903958664046445, "grad_norm": 0.7183582782745361, "learning_rate": 1.3668566476848777e-08, "loss": 0.6952, "step": 9900 }, { "epoch": 0.0890485805381049, "grad_norm": 0.24615027010440826, "learning_rate": 1.3396586299466162e-08, "loss": 0.7828, "step": 9901 }, { "epoch": 0.08905757443574534, "grad_norm": 0.5394777655601501, "learning_rate": 1.3127338619467467e-08, "loss": 0.6878, "step": 9902 }, { "epoch": 0.0890665683333858, "grad_norm": 1.320129632949829, "learning_rate": 1.2860823466293025e-08, "loss": 0.0991, "step": 9903 }, { "epoch": 0.08907556223102625, "grad_norm": 0.467526376247406, "learning_rate": 1.2597040869091748e-08, "loss": 0.7513, "step": 9904 }, { "epoch": 0.0890845561286667, "grad_norm": 0.2369871735572815, "learning_rate": 1.233599085671e-08, "loss": 1.4164, "step": 9905 }, { "epoch": 0.08909355002630714, "grad_norm": 0.5823180675506592, "learning_rate": 1.2077673457697169e-08, "loss": 0.5107, "step": 9906 }, { "epoch": 0.0891025439239476, "grad_norm": 2.244266986846924, "learning_rate": 1.1822088700300105e-08, "loss": 0.1381, "step": 9907 }, { "epoch": 0.08911153782158805, "grad_norm": 0.6298990249633789, "learning_rate": 1.1569236612471445e-08, "loss": 0.8204, "step": 9908 }, { "epoch": 0.0891205317192285, "grad_norm": 0.4807043969631195, "learning_rate": 1.1319117221861298e-08, "loss": 0.4537, "step": 9909 }, { "epoch": 0.08912952561686896, "grad_norm": 0.7313010096549988, "learning_rate": 1.107173055582278e-08, "loss": 0.0886, "step": 9910 }, { "epoch": 0.0891385195145094, "grad_norm": 0.3845495581626892, "learning_rate": 1.082707664140925e-08, "loss": 0.6614, "step": 9911 }, { "epoch": 0.08914751341214985, "grad_norm": 0.4897412061691284, "learning_rate": 1.0585155505374312e-08, "loss": 0.4202, "step": 9912 }, { "epoch": 0.0891565073097903, "grad_norm": 0.8057941198348999, "learning_rate": 1.0345967174177352e-08, "loss": 0.4563, "step": 9913 }, { "epoch": 0.08916550120743076, "grad_norm": 2.3710248470306396, "learning_rate": 1.010951167397245e-08, "loss": 0.1446, "step": 9914 }, { "epoch": 0.08917449510507121, "grad_norm": 0.6293322443962097, "learning_rate": 9.875789030622252e-09, "loss": 0.9566, "step": 9915 }, { "epoch": 0.08918348900271166, "grad_norm": 0.5784268379211426, "learning_rate": 9.644799269681314e-09, "loss": 0.0751, "step": 9916 }, { "epoch": 0.08919248290035212, "grad_norm": 2.0896034240722656, "learning_rate": 9.416542416412766e-09, "loss": 0.5934, "step": 9917 }, { "epoch": 0.08920147679799256, "grad_norm": 0.2970486581325531, "learning_rate": 9.191018495779968e-09, "loss": 0.7276, "step": 9918 }, { "epoch": 0.08921047069563301, "grad_norm": 0.4021197259426117, "learning_rate": 8.968227532443752e-09, "loss": 0.9712, "step": 9919 }, { "epoch": 0.08921946459327347, "grad_norm": 0.8435620069503784, "learning_rate": 8.74816955076796e-09, "loss": 1.1416, "step": 9920 }, { "epoch": 0.08922845849091392, "grad_norm": 0.36873647570610046, "learning_rate": 8.530844574819451e-09, "loss": 0.8638, "step": 9921 }, { "epoch": 0.08923745238855436, "grad_norm": 1.0582702159881592, "learning_rate": 8.316252628362553e-09, "loss": 0.7404, "step": 9922 }, { "epoch": 0.08924644628619481, "grad_norm": 0.3108186721801758, "learning_rate": 8.104393734867377e-09, "loss": 0.9082, "step": 9923 }, { "epoch": 0.08925544018383527, "grad_norm": 1.0166114568710327, "learning_rate": 7.895267917501504e-09, "loss": 0.3057, "step": 9924 }, { "epoch": 0.08926443408147572, "grad_norm": 0.3910999894142151, "learning_rate": 7.688875199132751e-09, "loss": 0.4559, "step": 9925 }, { "epoch": 0.08927342797911617, "grad_norm": 0.8004632592201233, "learning_rate": 7.485215602331953e-09, "loss": 0.2777, "step": 9926 }, { "epoch": 0.08928242187675663, "grad_norm": 0.7414090037345886, "learning_rate": 7.284289149372958e-09, "loss": 0.4679, "step": 9927 }, { "epoch": 0.08929141577439707, "grad_norm": 0.9009103178977966, "learning_rate": 7.0860958622298576e-09, "loss": 0.778, "step": 9928 }, { "epoch": 0.08930040967203752, "grad_norm": 0.6225287914276123, "learning_rate": 6.890635762574205e-09, "loss": 0.4857, "step": 9929 }, { "epoch": 0.08930940356967797, "grad_norm": 0.45233115553855896, "learning_rate": 6.697908871780567e-09, "loss": 0.7492, "step": 9930 }, { "epoch": 0.08931839746731843, "grad_norm": 0.46594980359077454, "learning_rate": 6.507915210929305e-09, "loss": 0.4127, "step": 9931 }, { "epoch": 0.08932739136495887, "grad_norm": 0.9904910326004028, "learning_rate": 6.320654800792691e-09, "loss": 0.3213, "step": 9932 }, { "epoch": 0.08933638526259932, "grad_norm": 0.6239996552467346, "learning_rate": 6.136127661854341e-09, "loss": 0.8353, "step": 9933 }, { "epoch": 0.08934537916023978, "grad_norm": 1.0838868618011475, "learning_rate": 5.954333814287005e-09, "loss": 0.7911, "step": 9934 }, { "epoch": 0.08935437305788023, "grad_norm": 0.4186025857925415, "learning_rate": 5.775273277980331e-09, "loss": 1.1184, "step": 9935 }, { "epoch": 0.08936336695552068, "grad_norm": 0.4466906785964966, "learning_rate": 5.5989460725103245e-09, "loss": 0.909, "step": 9936 }, { "epoch": 0.08937236085316114, "grad_norm": 0.4378398358821869, "learning_rate": 5.425352217158786e-09, "loss": 0.7398, "step": 9937 }, { "epoch": 0.08938135475080158, "grad_norm": 0.38323476910591125, "learning_rate": 5.254491730916078e-09, "loss": 0.6782, "step": 9938 }, { "epoch": 0.08939034864844203, "grad_norm": 0.3756667375564575, "learning_rate": 5.086364632458928e-09, "loss": 1.4975, "step": 9939 }, { "epoch": 0.08939934254608248, "grad_norm": 0.8201072216033936, "learning_rate": 4.920970940180958e-09, "loss": 0.5846, "step": 9940 }, { "epoch": 0.08940833644372294, "grad_norm": 0.3783058524131775, "learning_rate": 4.7583106721649225e-09, "loss": 0.6814, "step": 9941 }, { "epoch": 0.08941733034136339, "grad_norm": 0.3927249610424042, "learning_rate": 4.598383846199372e-09, "loss": 0.5613, "step": 9942 }, { "epoch": 0.08942632423900383, "grad_norm": 0.4652377665042877, "learning_rate": 4.441190479775869e-09, "loss": 0.6118, "step": 9943 }, { "epoch": 0.08943531813664429, "grad_norm": 0.8809239864349365, "learning_rate": 4.2867305900806676e-09, "loss": 0.4195, "step": 9944 }, { "epoch": 0.08944431203428474, "grad_norm": 0.5722887516021729, "learning_rate": 4.1350041940113604e-09, "loss": 0.4196, "step": 9945 }, { "epoch": 0.08945330593192519, "grad_norm": 0.7584978342056274, "learning_rate": 3.98601130815468e-09, "loss": 0.1015, "step": 9946 }, { "epoch": 0.08946229982956563, "grad_norm": 0.33563432097435, "learning_rate": 3.8397519488087006e-09, "loss": 0.7591, "step": 9947 }, { "epoch": 0.0894712937272061, "grad_norm": 0.6556146144866943, "learning_rate": 3.6962261319661852e-09, "loss": 0.5894, "step": 9948 }, { "epoch": 0.08948028762484654, "grad_norm": 0.34443819522857666, "learning_rate": 3.5554338733229107e-09, "loss": 0.513, "step": 9949 }, { "epoch": 0.08948928152248699, "grad_norm": 0.4504081904888153, "learning_rate": 3.417375188274896e-09, "loss": 0.9825, "step": 9950 }, { "epoch": 0.08949827542012745, "grad_norm": 2.814979314804077, "learning_rate": 3.2820500919239494e-09, "loss": 0.4319, "step": 9951 }, { "epoch": 0.0895072693177679, "grad_norm": 0.3852732181549072, "learning_rate": 3.1494585990637924e-09, "loss": 0.661, "step": 9952 }, { "epoch": 0.08951626321540834, "grad_norm": 0.3532427251338959, "learning_rate": 3.019600724193938e-09, "loss": 0.8961, "step": 9953 }, { "epoch": 0.08952525711304879, "grad_norm": 0.2954372763633728, "learning_rate": 2.892476481522466e-09, "loss": 1.2318, "step": 9954 }, { "epoch": 0.08953425101068925, "grad_norm": 0.5497097969055176, "learning_rate": 2.768085884943816e-09, "loss": 0.7384, "step": 9955 }, { "epoch": 0.0895432449083297, "grad_norm": 0.455424040555954, "learning_rate": 2.6464289480665485e-09, "loss": 0.6048, "step": 9956 }, { "epoch": 0.08955223880597014, "grad_norm": 2.336987257003784, "learning_rate": 2.5275056841939094e-09, "loss": 0.0725, "step": 9957 }, { "epoch": 0.0895612327036106, "grad_norm": 0.993019700050354, "learning_rate": 2.411316106326611e-09, "loss": 0.3683, "step": 9958 }, { "epoch": 0.08957022660125105, "grad_norm": 1.1579318046569824, "learning_rate": 2.297860227176707e-09, "loss": 0.0555, "step": 9959 }, { "epoch": 0.0895792204988915, "grad_norm": 0.8735564947128296, "learning_rate": 2.1871380591509392e-09, "loss": 0.7791, "step": 9960 }, { "epoch": 0.08958821439653196, "grad_norm": 0.7393962144851685, "learning_rate": 2.0791496143535148e-09, "loss": 0.7817, "step": 9961 }, { "epoch": 0.0895972082941724, "grad_norm": 0.4750567376613617, "learning_rate": 1.973894904597207e-09, "loss": 0.5293, "step": 9962 }, { "epoch": 0.08960620219181285, "grad_norm": 0.7460435032844543, "learning_rate": 1.8713739413922514e-09, "loss": 0.6451, "step": 9963 }, { "epoch": 0.0896151960894533, "grad_norm": 0.39201346039772034, "learning_rate": 1.7715867359491268e-09, "loss": 0.7698, "step": 9964 }, { "epoch": 0.08962418998709376, "grad_norm": 0.6568219661712646, "learning_rate": 1.6745332991813245e-09, "loss": 0.8527, "step": 9965 }, { "epoch": 0.08963318388473421, "grad_norm": 0.3366415798664093, "learning_rate": 1.5802136417025771e-09, "loss": 0.8992, "step": 9966 }, { "epoch": 0.08964217778237465, "grad_norm": 0.518984854221344, "learning_rate": 1.4886277738240806e-09, "loss": 0.7094, "step": 9967 }, { "epoch": 0.08965117168001512, "grad_norm": 0.3349268138408661, "learning_rate": 1.3997757055683735e-09, "loss": 0.8926, "step": 9968 }, { "epoch": 0.08966016557765556, "grad_norm": 0.5839380621910095, "learning_rate": 1.313657446647132e-09, "loss": 0.6063, "step": 9969 }, { "epoch": 0.08966915947529601, "grad_norm": 0.8191852569580078, "learning_rate": 1.2302730064778222e-09, "loss": 0.5973, "step": 9970 }, { "epoch": 0.08967815337293646, "grad_norm": 2.0838754177093506, "learning_rate": 1.1496223941837026e-09, "loss": 0.3852, "step": 9971 }, { "epoch": 0.08968714727057692, "grad_norm": 0.7282170057296753, "learning_rate": 1.0717056185799434e-09, "loss": 0.6379, "step": 9972 }, { "epoch": 0.08969614116821736, "grad_norm": 0.3476618528366089, "learning_rate": 9.965226881875066e-10, "loss": 0.9359, "step": 9973 }, { "epoch": 0.08970513506585781, "grad_norm": 0.2901281714439392, "learning_rate": 9.24073611233145e-10, "loss": 0.788, "step": 9974 }, { "epoch": 0.08971412896349827, "grad_norm": 0.6104804277420044, "learning_rate": 8.543583956355239e-10, "loss": 0.8617, "step": 9975 }, { "epoch": 0.08972312286113872, "grad_norm": 0.7050907015800476, "learning_rate": 7.873770490191002e-10, "loss": 0.5621, "step": 9976 }, { "epoch": 0.08973211675877917, "grad_norm": 2.4172463417053223, "learning_rate": 7.231295787085701e-10, "loss": 0.6707, "step": 9977 }, { "epoch": 0.08974111065641963, "grad_norm": 1.6493456363677979, "learning_rate": 6.616159917316456e-10, "loss": 0.4547, "step": 9978 }, { "epoch": 0.08975010455406007, "grad_norm": 0.4773891568183899, "learning_rate": 6.028362948135024e-10, "loss": 0.6082, "step": 9979 }, { "epoch": 0.08975909845170052, "grad_norm": 0.8443643450737, "learning_rate": 5.467904943851077e-10, "loss": 0.4426, "step": 9980 }, { "epoch": 0.08976809234934097, "grad_norm": 0.9325498342514038, "learning_rate": 4.934785965721167e-10, "loss": 0.52, "step": 9981 }, { "epoch": 0.08977708624698143, "grad_norm": 0.3352440893650055, "learning_rate": 4.42900607205976e-10, "loss": 0.9133, "step": 9982 }, { "epoch": 0.08978608014462187, "grad_norm": 0.44462355971336365, "learning_rate": 3.950565318183719e-10, "loss": 0.6429, "step": 9983 }, { "epoch": 0.08979507404226232, "grad_norm": 0.31531161069869995, "learning_rate": 3.4994637564123025e-10, "loss": 0.8488, "step": 9984 }, { "epoch": 0.08980406793990278, "grad_norm": 0.3524177670478821, "learning_rate": 3.0757014360671686e-10, "loss": 0.9991, "step": 9985 }, { "epoch": 0.08981306183754323, "grad_norm": 1.0540838241577148, "learning_rate": 2.6792784035001296e-10, "loss": 0.576, "step": 9986 }, { "epoch": 0.08982205573518368, "grad_norm": 0.3851053714752197, "learning_rate": 2.310194702093149e-10, "loss": 0.9785, "step": 9987 }, { "epoch": 0.08983104963282412, "grad_norm": 0.6173021197319031, "learning_rate": 1.9684503721473236e-10, "loss": 0.3139, "step": 9988 }, { "epoch": 0.08984004353046458, "grad_norm": 0.3685445189476013, "learning_rate": 1.654045451049413e-10, "loss": 0.8484, "step": 9989 }, { "epoch": 0.08984903742810503, "grad_norm": 0.4995662569999695, "learning_rate": 1.3669799732163313e-10, "loss": 0.6665, "step": 9990 }, { "epoch": 0.08985803132574548, "grad_norm": 0.22089681029319763, "learning_rate": 1.1072539700396345e-10, "loss": 1.0739, "step": 9991 }, { "epoch": 0.08986702522338594, "grad_norm": 0.49564430117607117, "learning_rate": 8.748674698855208e-11, "loss": 0.6558, "step": 9992 }, { "epoch": 0.08987601912102638, "grad_norm": 0.3039294183254242, "learning_rate": 6.698204982058532e-11, "loss": 0.8015, "step": 9993 }, { "epoch": 0.08988501301866683, "grad_norm": 0.6041123270988464, "learning_rate": 4.9211307737162536e-11, "loss": 0.5805, "step": 9994 }, { "epoch": 0.08989400691630728, "grad_norm": 1.1335822343826294, "learning_rate": 3.417452268950072e-11, "loss": 0.139, "step": 9995 }, { "epoch": 0.08990300081394774, "grad_norm": 0.45549607276916504, "learning_rate": 2.1871696315178824e-11, "loss": 0.9808, "step": 9996 }, { "epoch": 0.08991199471158819, "grad_norm": 0.580482542514801, "learning_rate": 1.2302829963117824e-11, "loss": 0.6297, "step": 9997 }, { "epoch": 0.08992098860922863, "grad_norm": 0.7295059561729431, "learning_rate": 5.4679246769273606e-12, "loss": 0.2179, "step": 9998 }, { "epoch": 0.0899299825068691, "grad_norm": 0.3017881214618683, "learning_rate": 1.3669812060079778e-12, "loss": 0.6938, "step": 9999 }, { "epoch": 0.08993897640450954, "grad_norm": 0.46469032764434814, "learning_rate": 0.0, "loss": 0.6448, "step": 10000 }, { "epoch": 0.08993897640450954, "step": 10000, "total_flos": 1.923911762617303e+18, "train_loss": 0.13473155428953468, "train_runtime": 4573.7781, "train_samples_per_second": 17.491, "train_steps_per_second": 2.186 } ], "logging_steps": 1.0, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.923911762617303e+18, "train_batch_size": 4, "trial_name": null, "trial_params": null }