diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18076 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 10314,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.656641960144043,
+      "learning_rate": 4.998060888113244e-05,
+      "loss": 9.8019,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3737926483154297,
+      "learning_rate": 4.9961217762264883e-05,
+      "loss": 9.12,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5734024047851562,
+      "learning_rate": 4.9941826643397324e-05,
+      "loss": 8.8002,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0836071968078613,
+      "learning_rate": 4.992243552452977e-05,
+      "loss": 8.5483,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.126896619796753,
+      "learning_rate": 4.9903044405662205e-05,
+      "loss": 8.376,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.795313835144043,
+      "learning_rate": 4.988365328679465e-05,
+      "loss": 8.2286,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.847642183303833,
+      "learning_rate": 4.9864262167927086e-05,
+      "loss": 8.0428,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6489531993865967,
+      "learning_rate": 4.9844871049059533e-05,
+      "loss": 7.8429,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.569993495941162,
+      "learning_rate": 4.9825479930191974e-05,
+      "loss": 7.6876,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3502109050750732,
+      "learning_rate": 4.9806088811324415e-05,
+      "loss": 7.5277,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2252399921417236,
+      "learning_rate": 4.9786697692456855e-05,
+      "loss": 7.3008,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2069296836853027,
+      "learning_rate": 4.97673065735893e-05,
+      "loss": 7.1233,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.133172035217285,
+      "learning_rate": 4.9747915454721736e-05,
+      "loss": 6.9925,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.830748438835144,
+      "learning_rate": 4.9728524335854183e-05,
+      "loss": 6.8808,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7571477890014648,
+      "learning_rate": 4.970913321698662e-05,
+      "loss": 6.7121,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6238043308258057,
+      "learning_rate": 4.9689742098119065e-05,
+      "loss": 6.6879,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.460631012916565,
+      "learning_rate": 4.9670350979251505e-05,
+      "loss": 6.4892,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.399665355682373,
+      "learning_rate": 4.9650959860383946e-05,
+      "loss": 6.4278,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2306325435638428,
+      "learning_rate": 4.9631568741516386e-05,
+      "loss": 6.43,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1850312948226929,
+      "learning_rate": 4.9612177622648833e-05,
+      "loss": 6.3628,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3839452266693115,
+      "learning_rate": 4.959278650378127e-05,
+      "loss": 6.1944,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2195193767547607,
+      "learning_rate": 4.9573395384913715e-05,
+      "loss": 6.2424,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1323491334915161,
+      "learning_rate": 4.9554004266046155e-05,
+      "loss": 6.112,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2749617099761963,
+      "learning_rate": 4.9534613147178596e-05,
+      "loss": 6.13,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2026777267456055,
+      "learning_rate": 4.9515222028311036e-05,
+      "loss": 6.0626,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4500185251235962,
+      "learning_rate": 4.949583090944348e-05,
+      "loss": 6.0918,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2848747968673706,
+      "learning_rate": 4.947643979057592e-05,
+      "loss": 6.1688,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3187015056610107,
+      "learning_rate": 4.9457048671708365e-05,
+      "loss": 6.0728,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.487337589263916,
+      "learning_rate": 4.94376575528408e-05,
+      "loss": 6.0801,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.191052794456482,
+      "learning_rate": 4.9418266433973246e-05,
+      "loss": 6.1361,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5774741172790527,
+      "learning_rate": 4.9398875315105686e-05,
+      "loss": 6.055,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0442208051681519,
+      "learning_rate": 4.937948419623813e-05,
+      "loss": 6.0708,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4346356391906738,
+      "learning_rate": 4.936009307737057e-05,
+      "loss": 6.0169,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2307249307632446,
+      "learning_rate": 4.934070195850301e-05,
+      "loss": 6.0547,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1539560556411743,
+      "learning_rate": 4.932131083963545e-05,
+      "loss": 6.0166,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1614229679107666,
+      "learning_rate": 4.930191972076789e-05,
+      "loss": 6.1866,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0596864223480225,
+      "learning_rate": 4.928252860190033e-05,
+      "loss": 5.9715,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.111221194267273,
+      "learning_rate": 4.926313748303277e-05,
+      "loss": 6.0272,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9217305183410645,
+      "learning_rate": 4.924374636416522e-05,
+      "loss": 6.0135,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.968235969543457,
+      "learning_rate": 4.922435524529765e-05,
+      "loss": 5.9969,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1550750732421875,
+      "learning_rate": 4.92049641264301e-05,
+      "loss": 6.0566,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5655795335769653,
+      "learning_rate": 4.918557300756254e-05,
+      "loss": 5.9619,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2823691368103027,
+      "learning_rate": 4.916618188869498e-05,
+      "loss": 5.876,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.913962721824646,
+      "learning_rate": 4.914679076982742e-05,
+      "loss": 6.0023,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1342990398406982,
+      "learning_rate": 4.912739965095986e-05,
+      "loss": 5.9118,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1372032165527344,
+      "learning_rate": 4.91080085320923e-05,
+      "loss": 5.9543,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1001873016357422,
+      "learning_rate": 4.908861741322475e-05,
+      "loss": 6.0478,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.163060188293457,
+      "learning_rate": 4.906922629435718e-05,
+      "loss": 5.9354,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.022900938987732,
+      "learning_rate": 4.904983517548963e-05,
+      "loss": 5.8227,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2557566165924072,
+      "learning_rate": 4.903044405662207e-05,
+      "loss": 5.898,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.260396122932434,
+      "learning_rate": 4.901105293775451e-05,
+      "loss": 5.9742,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0883902311325073,
+      "learning_rate": 4.899166181888695e-05,
+      "loss": 5.8642,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.07965886592865,
+      "learning_rate": 4.89722707000194e-05,
+      "loss": 5.9836,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.12454092502594,
+      "learning_rate": 4.895287958115183e-05,
+      "loss": 5.9383,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1025103330612183,
+      "learning_rate": 4.893348846228428e-05,
+      "loss": 5.8433,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1046481132507324,
+      "learning_rate": 4.891409734341671e-05,
+      "loss": 5.8649,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.356872320175171,
+      "learning_rate": 4.889470622454916e-05,
+      "loss": 5.9069,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1604045629501343,
+      "learning_rate": 4.88753151056816e-05,
+      "loss": 5.8704,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1360890865325928,
+      "learning_rate": 4.885592398681404e-05,
+      "loss": 5.8723,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8874178528785706,
+      "learning_rate": 4.883653286794648e-05,
+      "loss": 5.8548,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0104761123657227,
+      "learning_rate": 4.881714174907893e-05,
+      "loss": 5.8931,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1234338283538818,
+      "learning_rate": 4.879775063021136e-05,
+      "loss": 5.8915,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2352111339569092,
+      "learning_rate": 4.877835951134381e-05,
+      "loss": 5.9084,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9969356656074524,
+      "learning_rate": 4.875896839247625e-05,
+      "loss": 5.9788,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4352229833602905,
+      "learning_rate": 4.873957727360869e-05,
+      "loss": 5.8121,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3081138134002686,
+      "learning_rate": 4.872018615474113e-05,
+      "loss": 5.8714,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1553027629852295,
+      "learning_rate": 4.870079503587357e-05,
+      "loss": 5.9062,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.328942060470581,
+      "learning_rate": 4.868140391700601e-05,
+      "loss": 5.9659,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.092142939567566,
+      "learning_rate": 4.8662012798138454e-05,
+      "loss": 5.9436,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8468914031982422,
+      "learning_rate": 4.8642621679270894e-05,
+      "loss": 5.9176,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1835949420928955,
+      "learning_rate": 4.8623230560403335e-05,
+      "loss": 5.8968,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.401023507118225,
+      "learning_rate": 4.860383944153578e-05,
+      "loss": 5.8126,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2995976209640503,
+      "learning_rate": 4.8584448322668216e-05,
+      "loss": 5.9769,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0936148166656494,
+      "learning_rate": 4.856505720380066e-05,
+      "loss": 5.8341,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.121054768562317,
+      "learning_rate": 4.85456660849331e-05,
+      "loss": 5.8082,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.011224389076233,
+      "learning_rate": 4.8526274966065544e-05,
+      "loss": 5.8437,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.909989058971405,
+      "learning_rate": 4.8506883847197985e-05,
+      "loss": 5.8217,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9056807160377502,
+      "learning_rate": 4.8487492728330425e-05,
+      "loss": 5.8791,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.990393340587616,
+      "learning_rate": 4.8468101609462866e-05,
+      "loss": 5.9895,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.252443790435791,
+      "learning_rate": 4.844871049059531e-05,
+      "loss": 5.9232,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9842864274978638,
+      "learning_rate": 4.842931937172775e-05,
+      "loss": 5.8068,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.050066351890564,
+      "learning_rate": 4.8409928252860194e-05,
+      "loss": 5.8112,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9735043048858643,
+      "learning_rate": 4.8390537133992635e-05,
+      "loss": 5.8329,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.948833703994751,
+      "learning_rate": 4.8371146015125075e-05,
+      "loss": 5.8564,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0226508378982544,
+      "learning_rate": 4.8351754896257516e-05,
+      "loss": 5.7493,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.09911048412323,
+      "learning_rate": 4.8332363777389956e-05,
+      "loss": 5.7015,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1777935028076172,
+      "learning_rate": 4.83129726585224e-05,
+      "loss": 5.7534,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1616647243499756,
+      "learning_rate": 4.8293581539654844e-05,
+      "loss": 5.8506,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0601903200149536,
+      "learning_rate": 4.827419042078728e-05,
+      "loss": 5.8917,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9961209893226624,
+      "learning_rate": 4.8254799301919725e-05,
+      "loss": 5.797,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0286229848861694,
+      "learning_rate": 4.8235408183052166e-05,
+      "loss": 5.888,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0672661066055298,
+      "learning_rate": 4.8216017064184606e-05,
+      "loss": 5.823,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9099803566932678,
+      "learning_rate": 4.819662594531705e-05,
+      "loss": 5.7411,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1643503904342651,
+      "learning_rate": 4.8177234826449494e-05,
+      "loss": 5.9417,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9987436532974243,
+      "learning_rate": 4.815784370758193e-05,
+      "loss": 5.8176,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1946054697036743,
+      "learning_rate": 4.8138452588714375e-05,
+      "loss": 5.9005,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9317137002944946,
+      "learning_rate": 4.811906146984681e-05,
+      "loss": 5.8017,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0581262111663818,
+      "learning_rate": 4.8099670350979256e-05,
+      "loss": 5.794,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1224080324172974,
+      "learning_rate": 4.80802792321117e-05,
+      "loss": 5.8161,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2604461908340454,
+      "learning_rate": 4.806088811324414e-05,
+      "loss": 5.8109,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1275701522827148,
+      "learning_rate": 4.804149699437658e-05,
+      "loss": 5.8104,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.012634038925171,
+      "learning_rate": 4.802210587550902e-05,
+      "loss": 5.8702,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8785901665687561,
+      "learning_rate": 4.800271475664146e-05,
+      "loss": 5.8498,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8808034062385559,
+      "learning_rate": 4.79833236377739e-05,
+      "loss": 5.7474,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1132124662399292,
+      "learning_rate": 4.796393251890634e-05,
+      "loss": 5.7462,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9468419551849365,
+      "learning_rate": 4.794454140003878e-05,
+      "loss": 5.7767,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9365801215171814,
+      "learning_rate": 4.792515028117123e-05,
+      "loss": 5.8072,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0467591285705566,
+      "learning_rate": 4.790575916230366e-05,
+      "loss": 5.7773,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1843376159667969,
+      "learning_rate": 4.788636804343611e-05,
+      "loss": 5.9357,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.139872431755066,
+      "learning_rate": 4.786697692456855e-05,
+      "loss": 5.7744,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9096530079841614,
+      "learning_rate": 4.784758580570099e-05,
+      "loss": 5.9122,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0682241916656494,
+      "learning_rate": 4.782819468683343e-05,
+      "loss": 5.8587,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.99188232421875,
+      "learning_rate": 4.780880356796588e-05,
+      "loss": 5.8096,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0947283506393433,
+      "learning_rate": 4.778941244909831e-05,
+      "loss": 5.6552,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0802547931671143,
+      "learning_rate": 4.777002133023076e-05,
+      "loss": 5.9245,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.164602518081665,
+      "learning_rate": 4.775063021136319e-05,
+      "loss": 5.8362,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1340216398239136,
+      "learning_rate": 4.773123909249564e-05,
+      "loss": 5.8477,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0476304292678833,
+      "learning_rate": 4.771184797362808e-05,
+      "loss": 5.8336,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9962639808654785,
+      "learning_rate": 4.769245685476052e-05,
+      "loss": 5.7876,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9998891949653625,
+      "learning_rate": 4.767306573589296e-05,
+      "loss": 5.6897,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.054436445236206,
+      "learning_rate": 4.765367461702541e-05,
+      "loss": 5.6461,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9963980913162231,
+      "learning_rate": 4.763428349815784e-05,
+      "loss": 5.7391,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.077802062034607,
+      "learning_rate": 4.761489237929029e-05,
+      "loss": 5.772,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3317667245864868,
+      "learning_rate": 4.759550126042273e-05,
+      "loss": 5.7657,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.028090000152588,
+      "learning_rate": 4.757611014155517e-05,
+      "loss": 5.7417,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9719061255455017,
+      "learning_rate": 4.755671902268761e-05,
+      "loss": 5.8042,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.067749261856079,
+      "learning_rate": 4.753732790382005e-05,
+      "loss": 5.7923,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9830943942070007,
+      "learning_rate": 4.751793678495249e-05,
+      "loss": 5.7911,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0226807594299316,
+      "learning_rate": 4.749854566608494e-05,
+      "loss": 5.7696,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3311954736709595,
+      "learning_rate": 4.7479154547217374e-05,
+      "loss": 5.8689,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9809809923171997,
+      "learning_rate": 4.745976342834982e-05,
+      "loss": 5.8023,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1488884687423706,
+      "learning_rate": 4.744037230948226e-05,
+      "loss": 5.7116,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0571842193603516,
+      "learning_rate": 4.74209811906147e-05,
+      "loss": 5.7105,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9185091257095337,
+      "learning_rate": 4.740159007174714e-05,
+      "loss": 5.8675,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9164991974830627,
+      "learning_rate": 4.738219895287958e-05,
+      "loss": 5.7231,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0493993759155273,
+      "learning_rate": 4.7362807834012024e-05,
+      "loss": 5.8437,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8653802275657654,
+      "learning_rate": 4.7343416715144464e-05,
+      "loss": 5.652,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.975098192691803,
+      "learning_rate": 4.7324025596276905e-05,
+      "loss": 5.7353,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9267313480377197,
+      "learning_rate": 4.7304634477409345e-05,
+      "loss": 5.7886,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9583150148391724,
+      "learning_rate": 4.728524335854179e-05,
+      "loss": 5.6864,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9664693474769592,
+      "learning_rate": 4.7265852239674227e-05,
+      "loss": 5.8314,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8994282484054565,
+      "learning_rate": 4.7246461120806674e-05,
+      "loss": 5.7865,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0095504522323608,
+      "learning_rate": 4.7227070001939114e-05,
+      "loss": 5.7627,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0120139122009277,
+      "learning_rate": 4.7207678883071555e-05,
+      "loss": 5.6769,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9483413100242615,
+      "learning_rate": 4.7188287764203995e-05,
+      "loss": 5.7615,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9208195209503174,
+      "learning_rate": 4.7168896645336436e-05,
+      "loss": 5.7178,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0568493604660034,
+      "learning_rate": 4.7149505526468876e-05,
+      "loss": 5.6305,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9036898016929626,
+      "learning_rate": 4.7130114407601324e-05,
+      "loss": 5.7088,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0219190120697021,
+      "learning_rate": 4.711072328873376e-05,
+      "loss": 5.7359,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1148089170455933,
+      "learning_rate": 4.7091332169866205e-05,
+      "loss": 5.7653,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0402499437332153,
+      "learning_rate": 4.7071941050998645e-05,
+      "loss": 5.7291,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2846848964691162,
+      "learning_rate": 4.7052549932131086e-05,
+      "loss": 5.8432,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0562716722488403,
+      "learning_rate": 4.7033158813263526e-05,
+      "loss": 5.7242,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9270595908164978,
+      "learning_rate": 4.7013767694395974e-05,
+      "loss": 5.735,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9591754674911499,
+      "learning_rate": 4.699437657552841e-05,
+      "loss": 5.6973,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.020099401473999,
+      "learning_rate": 4.6974985456660855e-05,
+      "loss": 5.7211,
+      "step": 624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9728394150733948,
+      "learning_rate": 4.695559433779329e-05,
+      "loss": 5.7558,
+      "step": 628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9725939631462097,
+      "learning_rate": 4.6936203218925736e-05,
+      "loss": 5.8508,
+      "step": 632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0429859161376953,
+      "learning_rate": 4.6916812100058176e-05,
+      "loss": 5.7557,
+      "step": 636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1550756692886353,
+      "learning_rate": 4.689742098119062e-05,
+      "loss": 5.7864,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.277512788772583,
+      "learning_rate": 4.687802986232306e-05,
+      "loss": 5.731,
+      "step": 644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.170222282409668,
+      "learning_rate": 4.6858638743455505e-05,
+      "loss": 5.6716,
+      "step": 648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0293292999267578,
+      "learning_rate": 4.683924762458794e-05,
+      "loss": 5.7762,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0924861431121826,
+      "learning_rate": 4.6819856505720386e-05,
+      "loss": 5.7223,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.949212908744812,
+      "learning_rate": 4.680046538685282e-05,
+      "loss": 5.7425,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0186933279037476,
+      "learning_rate": 4.678107426798527e-05,
+      "loss": 5.8361,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0421589612960815,
+      "learning_rate": 4.676168314911771e-05,
+      "loss": 5.7093,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.061271071434021,
+      "learning_rate": 4.674229203025015e-05,
+      "loss": 5.6382,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0228848457336426,
+      "learning_rate": 4.672290091138259e-05,
+      "loss": 5.6845,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0488477945327759,
+      "learning_rate": 4.670350979251503e-05,
+      "loss": 5.6403,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9394490718841553,
+      "learning_rate": 4.668411867364747e-05,
+      "loss": 5.6733,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.140890121459961,
+      "learning_rate": 4.666472755477991e-05,
+      "loss": 5.7066,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9432203769683838,
+      "learning_rate": 4.664533643591236e-05,
+      "loss": 5.7846,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0256586074829102,
+      "learning_rate": 4.662594531704479e-05,
+      "loss": 5.766,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0383509397506714,
+      "learning_rate": 4.660655419817724e-05,
+      "loss": 5.8138,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9286032319068909,
+      "learning_rate": 4.658716307930968e-05,
+      "loss": 5.7233,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9339187145233154,
+      "learning_rate": 4.656777196044212e-05,
+      "loss": 5.803,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9535828828811646,
+      "learning_rate": 4.654838084157456e-05,
+      "loss": 5.8002,
+      "step": 712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2881035804748535,
+      "learning_rate": 4.6528989722707e-05,
+      "loss": 5.5501,
+      "step": 716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9991912841796875,
+      "learning_rate": 4.650959860383944e-05,
+      "loss": 5.7732,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.071502923965454,
+      "learning_rate": 4.649020748497189e-05,
+      "loss": 5.6494,
+      "step": 724
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9422402381896973,
+      "learning_rate": 4.647081636610432e-05,
+      "loss": 5.7539,
+      "step": 728
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2175371646881104,
+      "learning_rate": 4.645142524723677e-05,
+      "loss": 5.6747,
+      "step": 732
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1459147930145264,
+      "learning_rate": 4.643203412836921e-05,
+      "loss": 5.6895,
+      "step": 736
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9659698009490967,
+      "learning_rate": 4.641264300950165e-05,
+      "loss": 5.6517,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.071323275566101,
+      "learning_rate": 4.639325189063409e-05,
+      "loss": 5.653,
+      "step": 744
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1298364400863647,
+      "learning_rate": 4.637386077176653e-05,
+      "loss": 5.6975,
+      "step": 748
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9889796376228333,
+      "learning_rate": 4.635446965289897e-05,
+      "loss": 5.8007,
+      "step": 752
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9490829110145569,
+      "learning_rate": 4.633507853403142e-05,
+      "loss": 5.6969,
+      "step": 756
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0920511484146118,
+      "learning_rate": 4.6315687415163853e-05,
+      "loss": 5.74,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.039368748664856,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 5.7002,
+      "step": 764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9749051928520203,
+      "learning_rate": 4.627690517742874e-05,
+      "loss": 5.6879,
+      "step": 768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9410939812660217,
+      "learning_rate": 4.625751405856118e-05,
+      "loss": 5.7138,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9284390807151794,
+      "learning_rate": 4.623812293969362e-05,
+      "loss": 5.6772,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.220140814781189,
+      "learning_rate": 4.621873182082606e-05,
+      "loss": 5.6187,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9217903017997742,
+      "learning_rate": 4.6199340701958503e-05,
+      "loss": 5.7051,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9127422571182251,
+      "learning_rate": 4.617994958309095e-05,
+      "loss": 5.6821,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9815553426742554,
+      "learning_rate": 4.6160558464223385e-05,
+      "loss": 5.7365,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8972638249397278,
+      "learning_rate": 4.614116734535583e-05,
+      "loss": 5.6835,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9297725558280945,
+      "learning_rate": 4.612177622648827e-05,
+      "loss": 5.6304,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9928802847862244,
+      "learning_rate": 4.610238510762071e-05,
+      "loss": 5.8121,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9002383351325989,
+      "learning_rate": 4.6082993988753153e-05,
+      "loss": 5.6457,
+      "step": 808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8902143239974976,
+      "learning_rate": 4.6063602869885594e-05,
+      "loss": 5.7276,
+      "step": 812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.069921612739563,
+      "learning_rate": 4.6044211751018035e-05,
+      "loss": 5.6827,
+      "step": 816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9564573764801025,
+      "learning_rate": 4.6024820632150475e-05,
+      "loss": 5.694,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9404289126396179,
+      "learning_rate": 4.6005429513282916e-05,
+      "loss": 5.585,
+      "step": 824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9606718420982361,
+      "learning_rate": 4.5986038394415356e-05,
+      "loss": 5.7258,
+      "step": 828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.893195629119873,
+      "learning_rate": 4.5966647275547803e-05,
+      "loss": 5.6816,
+      "step": 832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9530772566795349,
+      "learning_rate": 4.5947256156680244e-05,
+      "loss": 5.7424,
+      "step": 836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1925373077392578,
+      "learning_rate": 4.5927865037812685e-05,
+      "loss": 5.6958,
+      "step": 840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0818183422088623,
+      "learning_rate": 4.5908473918945125e-05,
+      "loss": 5.6273,
+      "step": 844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 4.5889082800077566e-05,
+      "loss": 5.6699,
+      "step": 848
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.069509744644165,
+      "learning_rate": 4.5869691681210006e-05,
+      "loss": 5.677,
+      "step": 852
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0898882150650024,
+      "learning_rate": 4.5850300562342453e-05,
+      "loss": 5.723,
+      "step": 856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9786805510520935,
+      "learning_rate": 4.583090944347489e-05,
+      "loss": 5.7109,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0218449831008911,
+      "learning_rate": 4.5811518324607335e-05,
+      "loss": 5.7927,
+      "step": 864
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0597094297409058,
+      "learning_rate": 4.579212720573977e-05,
+      "loss": 5.6456,
+      "step": 868
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.015222430229187,
+      "learning_rate": 4.5772736086872216e-05,
+      "loss": 5.7023,
+      "step": 872
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0093826055526733,
+      "learning_rate": 4.5753344968004656e-05,
+      "loss": 5.673,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1618988513946533,
+      "learning_rate": 4.57339538491371e-05,
+      "loss": 5.5492,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9958293437957764,
+      "learning_rate": 4.571456273026954e-05,
+      "loss": 5.5945,
+      "step": 884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9605300426483154,
+      "learning_rate": 4.5695171611401985e-05,
+      "loss": 5.6439,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0881812572479248,
+      "learning_rate": 4.567578049253442e-05,
+      "loss": 5.7346,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1047258377075195,
+      "learning_rate": 4.5656389373666866e-05,
+      "loss": 5.6754,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0597559213638306,
+      "learning_rate": 4.56369982547993e-05,
+      "loss": 5.6655,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0080434083938599,
+      "learning_rate": 4.561760713593175e-05,
+      "loss": 5.637,
+      "step": 904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1175477504730225,
+      "learning_rate": 4.559821601706419e-05,
+      "loss": 5.6114,
+      "step": 908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.309441089630127,
+      "learning_rate": 4.557882489819663e-05,
+      "loss": 5.64,
+      "step": 912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0713026523590088,
+      "learning_rate": 4.555943377932907e-05,
+      "loss": 5.6433,
+      "step": 916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1929994821548462,
+      "learning_rate": 4.5540042660461516e-05,
+      "loss": 5.6503,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0933693647384644,
+      "learning_rate": 4.552065154159395e-05,
+      "loss": 5.68,
+      "step": 924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9823185801506042,
+      "learning_rate": 4.55012604227264e-05,
+      "loss": 5.6941,
+      "step": 928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9883749485015869,
+      "learning_rate": 4.548186930385884e-05,
+      "loss": 5.5973,
+      "step": 932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9289098381996155,
+      "learning_rate": 4.546247818499128e-05,
+      "loss": 5.7444,
+      "step": 936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0262410640716553,
+      "learning_rate": 4.544308706612372e-05,
+      "loss": 5.6235,
+      "step": 940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0129894018173218,
+      "learning_rate": 4.542369594725616e-05,
+      "loss": 5.6469,
+      "step": 944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0500363111495972,
+      "learning_rate": 4.54043048283886e-05,
+      "loss": 5.6281,
+      "step": 948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9239895343780518,
+      "learning_rate": 4.538491370952104e-05,
+      "loss": 5.6151,
+      "step": 952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9726962447166443,
+      "learning_rate": 4.536552259065348e-05,
+      "loss": 5.6146,
+      "step": 956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9822578430175781,
+      "learning_rate": 4.534613147178592e-05,
+      "loss": 5.5852,
+      "step": 960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9551820755004883,
+      "learning_rate": 4.532674035291837e-05,
+      "loss": 5.7069,
+      "step": 964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9644606709480286,
+      "learning_rate": 4.530734923405081e-05,
+      "loss": 5.6178,
+      "step": 968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1014901399612427,
+      "learning_rate": 4.528795811518325e-05,
+      "loss": 5.6131,
+      "step": 972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0390726327896118,
+      "learning_rate": 4.526856699631569e-05,
+      "loss": 5.5601,
+      "step": 976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9854323863983154,
+      "learning_rate": 4.524917587744813e-05,
+      "loss": 5.6175,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9083386659622192,
+      "learning_rate": 4.522978475858057e-05,
+      "loss": 5.6013,
+      "step": 984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.246685266494751,
+      "learning_rate": 4.521039363971301e-05,
+      "loss": 5.7068,
+      "step": 988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0527753829956055,
+      "learning_rate": 4.519100252084545e-05,
+      "loss": 5.6358,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0731735229492188,
+      "learning_rate": 4.51716114019779e-05,
+      "loss": 5.6828,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0552845001220703,
+      "learning_rate": 4.515222028311033e-05,
+      "loss": 5.6828,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.990972101688385,
+      "learning_rate": 4.513282916424278e-05,
+      "loss": 5.7003,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1035810708999634,
+      "learning_rate": 4.511343804537522e-05,
+      "loss": 5.7133,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2226297855377197,
+      "learning_rate": 4.509404692650766e-05,
+      "loss": 5.6895,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1234639883041382,
+      "learning_rate": 4.50746558076401e-05,
+      "loss": 5.7874,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1130845546722412,
+      "learning_rate": 4.505526468877254e-05,
+      "loss": 5.7322,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8997195363044739,
+      "learning_rate": 4.503587356990498e-05,
+      "loss": 5.5914,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9451141953468323,
+      "learning_rate": 4.501648245103743e-05,
+      "loss": 5.68,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0043590068817139,
+      "learning_rate": 4.4997091332169864e-05,
+      "loss": 5.7082,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8845294713973999,
+      "learning_rate": 4.497770021330231e-05,
+      "loss": 5.6198,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0413445234298706,
+      "learning_rate": 4.495830909443475e-05,
+      "loss": 5.6253,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9856407642364502,
+      "learning_rate": 4.493891797556719e-05,
+      "loss": 5.7266,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0194454193115234,
+      "learning_rate": 4.491952685669963e-05,
+      "loss": 5.6567,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9935641884803772,
+      "learning_rate": 4.490013573783208e-05,
+      "loss": 5.639,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0133986473083496,
+      "learning_rate": 4.4880744618964514e-05,
+      "loss": 5.6575,
+      "step": 1056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9149471521377563,
+      "learning_rate": 4.486135350009696e-05,
+      "loss": 5.7549,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1448979377746582,
+      "learning_rate": 4.4841962381229395e-05,
+      "loss": 5.6149,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9014776349067688,
+      "learning_rate": 4.482257126236184e-05,
+      "loss": 5.6667,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9877203702926636,
+      "learning_rate": 4.480318014349428e-05,
+      "loss": 5.6292,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.15706467628479,
+      "learning_rate": 4.4783789024626724e-05,
+      "loss": 5.7073,
+      "step": 1076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0951610803604126,
+      "learning_rate": 4.4764397905759164e-05,
+      "loss": 5.6028,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9684280157089233,
+      "learning_rate": 4.4745006786891605e-05,
+      "loss": 5.6946,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9964583516120911,
+      "learning_rate": 4.4725615668024045e-05,
+      "loss": 5.6245,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9633233547210693,
+      "learning_rate": 4.4706224549156486e-05,
+      "loss": 5.6683,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0245460271835327,
+      "learning_rate": 4.468683343028893e-05,
+      "loss": 5.644,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0242762565612793,
+      "learning_rate": 4.4667442311421374e-05,
+      "loss": 5.6365,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.982917070388794,
+      "learning_rate": 4.4648051192553814e-05,
+      "loss": 5.6458,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.949592113494873,
+      "learning_rate": 4.4628660073686255e-05,
+      "loss": 5.6253,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0238159894943237,
+      "learning_rate": 4.4609268954818695e-05,
+      "loss": 5.6323,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9112494587898254,
+      "learning_rate": 4.4589877835951136e-05,
+      "loss": 5.4746,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9587331414222717,
+      "learning_rate": 4.4570486717083576e-05,
+      "loss": 5.6729,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8947247862815857,
+      "learning_rate": 4.455109559821602e-05,
+      "loss": 5.7247,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9927013516426086,
+      "learning_rate": 4.4531704479348464e-05,
+      "loss": 5.6287,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0964155197143555,
+      "learning_rate": 4.45123133604809e-05,
+      "loss": 5.611,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9409870505332947,
+      "learning_rate": 4.4492922241613345e-05,
+      "loss": 5.6107,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1042007207870483,
+      "learning_rate": 4.447353112274578e-05,
+      "loss": 5.7283,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0457091331481934,
+      "learning_rate": 4.4454140003878226e-05,
+      "loss": 5.6,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.037097692489624,
+      "learning_rate": 4.443474888501067e-05,
+      "loss": 5.7081,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8815637230873108,
+      "learning_rate": 4.441535776614311e-05,
+      "loss": 5.6222,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9820300936698914,
+      "learning_rate": 4.439596664727555e-05,
+      "loss": 5.639,
+      "step": 1156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1689587831497192,
+      "learning_rate": 4.4376575528407995e-05,
+      "loss": 5.6293,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9346803426742554,
+      "learning_rate": 4.435718440954043e-05,
+      "loss": 5.5408,
+      "step": 1164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8906845450401306,
+      "learning_rate": 4.4337793290672876e-05,
+      "loss": 5.5729,
+      "step": 1168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0249162912368774,
+      "learning_rate": 4.431840217180532e-05,
+      "loss": 5.6928,
+      "step": 1172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0330933332443237,
+      "learning_rate": 4.429901105293776e-05,
+      "loss": 5.6583,
+      "step": 1176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2164433002471924,
+      "learning_rate": 4.42796199340702e-05,
+      "loss": 5.6906,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.061458706855774,
+      "learning_rate": 4.426022881520264e-05,
+      "loss": 5.6325,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0236989259719849,
+      "learning_rate": 4.424083769633508e-05,
+      "loss": 5.6806,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.125806212425232,
+      "learning_rate": 4.4221446577467526e-05,
+      "loss": 5.6773,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0632579326629639,
+      "learning_rate": 4.420205545859996e-05,
+      "loss": 5.7194,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0620874166488647,
+      "learning_rate": 4.418266433973241e-05,
+      "loss": 5.5184,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0390667915344238,
+      "learning_rate": 4.416327322086485e-05,
+      "loss": 5.6573,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9894878268241882,
+      "learning_rate": 4.414388210199729e-05,
+      "loss": 5.765,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9994860887527466,
+      "learning_rate": 4.412449098312973e-05,
+      "loss": 5.661,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9990851283073425,
+      "learning_rate": 4.410509986426217e-05,
+      "loss": 5.6168,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0815703868865967,
+      "learning_rate": 4.408570874539461e-05,
+      "loss": 5.6433,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9334216117858887,
+      "learning_rate": 4.406631762652705e-05,
+      "loss": 5.5035,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9055896401405334,
+      "learning_rate": 4.404692650765949e-05,
+      "loss": 5.632,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0294831991195679,
+      "learning_rate": 4.402753538879194e-05,
+      "loss": 5.5739,
+      "step": 1232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.100758671760559,
+      "learning_rate": 4.400814426992438e-05,
+      "loss": 5.6159,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9790968894958496,
+      "learning_rate": 4.398875315105682e-05,
+      "loss": 5.6637,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9665990471839905,
+      "learning_rate": 4.396936203218926e-05,
+      "loss": 5.6361,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0019501447677612,
+      "learning_rate": 4.39499709133217e-05,
+      "loss": 5.5965,
+      "step": 1248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9785341620445251,
+      "learning_rate": 4.393057979445414e-05,
+      "loss": 5.6697,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0229973793029785,
+      "learning_rate": 4.391118867558658e-05,
+      "loss": 5.6391,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9716441035270691,
+      "learning_rate": 4.389179755671902e-05,
+      "loss": 5.6513,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8981690406799316,
+      "learning_rate": 4.387240643785146e-05,
+      "loss": 5.6567,
+      "step": 1264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9904807806015015,
+      "learning_rate": 4.385301531898391e-05,
+      "loss": 5.6761,
+      "step": 1268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9603332877159119,
+      "learning_rate": 4.3833624200116344e-05,
+      "loss": 5.7039,
+      "step": 1272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.044404149055481,
+      "learning_rate": 4.381423308124879e-05,
+      "loss": 5.5329,
+      "step": 1276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1209489107131958,
+      "learning_rate": 4.379484196238123e-05,
+      "loss": 5.6461,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1430855989456177,
+      "learning_rate": 4.377545084351367e-05,
+      "loss": 5.5406,
+      "step": 1284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.004539966583252,
+      "learning_rate": 4.375605972464611e-05,
+      "loss": 5.5827,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0721603631973267,
+      "learning_rate": 4.373666860577856e-05,
+      "loss": 5.5553,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9446938037872314,
+      "learning_rate": 4.3717277486910994e-05,
+      "loss": 5.6481,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8731529116630554,
+      "learning_rate": 4.369788636804344e-05,
+      "loss": 5.6374,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0640937089920044,
+      "learning_rate": 4.3678495249175875e-05,
+      "loss": 5.5943,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0243881940841675,
+      "learning_rate": 4.365910413030832e-05,
+      "loss": 5.5065,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1701568365097046,
+      "learning_rate": 4.363971301144076e-05,
+      "loss": 5.6329,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9212052822113037,
+      "learning_rate": 4.36203218925732e-05,
+      "loss": 5.7735,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9392960667610168,
+      "learning_rate": 4.3600930773705644e-05,
+      "loss": 5.5992,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0432089567184448,
+      "learning_rate": 4.358153965483809e-05,
+      "loss": 5.5948,
+      "step": 1324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9817702770233154,
+      "learning_rate": 4.3562148535970525e-05,
+      "loss": 5.6496,
+      "step": 1328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0641746520996094,
+      "learning_rate": 4.354275741710297e-05,
+      "loss": 5.6238,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0068002939224243,
+      "learning_rate": 4.352336629823541e-05,
+      "loss": 5.6316,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.031501054763794,
+      "learning_rate": 4.350397517936785e-05,
+      "loss": 5.6462,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.980787992477417,
+      "learning_rate": 4.3484584060500294e-05,
+      "loss": 5.6535,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9462695717811584,
+      "learning_rate": 4.3465192941632734e-05,
+      "loss": 5.6083,
+      "step": 1348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8907430768013,
+      "learning_rate": 4.3445801822765175e-05,
+      "loss": 5.615,
+      "step": 1352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2333685159683228,
+      "learning_rate": 4.3426410703897615e-05,
+      "loss": 5.6374,
+      "step": 1356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.165974497795105,
+      "learning_rate": 4.3407019585030056e-05,
+      "loss": 5.5617,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1213206052780151,
+      "learning_rate": 4.33876284661625e-05,
+      "loss": 5.6201,
+      "step": 1364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0781583786010742,
+      "learning_rate": 4.3368237347294944e-05,
+      "loss": 5.639,
+      "step": 1368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0337659120559692,
+      "learning_rate": 4.3348846228427384e-05,
+      "loss": 5.736,
+      "step": 1372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9833229184150696,
+      "learning_rate": 4.3329455109559825e-05,
+      "loss": 5.6153,
+      "step": 1376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0340036153793335,
+      "learning_rate": 4.3310063990692265e-05,
+      "loss": 5.6146,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9522514939308167,
+      "learning_rate": 4.3290672871824706e-05,
+      "loss": 5.646,
+      "step": 1384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.040779948234558,
+      "learning_rate": 4.3271281752957146e-05,
+      "loss": 5.6411,
+      "step": 1388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9653657674789429,
+      "learning_rate": 4.325189063408959e-05,
+      "loss": 5.634,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9610472321510315,
+      "learning_rate": 4.323249951522203e-05,
+      "loss": 5.5112,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0222139358520508,
+      "learning_rate": 4.3213108396354475e-05,
+      "loss": 5.4966,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0731512308120728,
+      "learning_rate": 4.319371727748691e-05,
+      "loss": 5.5193,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.075119972229004,
+      "learning_rate": 4.3174326158619356e-05,
+      "loss": 5.573,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9324813485145569,
+      "learning_rate": 4.3154935039751796e-05,
+      "loss": 5.5821,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8965750336647034,
+      "learning_rate": 4.313554392088424e-05,
+      "loss": 5.4723,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.902160108089447,
+      "learning_rate": 4.311615280201668e-05,
+      "loss": 5.5979,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9165207743644714,
+      "learning_rate": 4.309676168314912e-05,
+      "loss": 5.5698,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0775911808013916,
+      "learning_rate": 4.307737056428156e-05,
+      "loss": 5.4091,
+      "step": 1428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9935070276260376,
+      "learning_rate": 4.3057979445414006e-05,
+      "loss": 5.6084,
+      "step": 1432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.105722188949585,
+      "learning_rate": 4.303858832654644e-05,
+      "loss": 5.6961,
+      "step": 1436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0309725999832153,
+      "learning_rate": 4.301919720767889e-05,
+      "loss": 5.6465,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9571665525436401,
+      "learning_rate": 4.299980608881133e-05,
+      "loss": 5.5844,
+      "step": 1444
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0037786960601807,
+      "learning_rate": 4.298041496994377e-05,
+      "loss": 5.7001,
+      "step": 1448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9573729634284973,
+      "learning_rate": 4.296102385107621e-05,
+      "loss": 5.5545,
+      "step": 1452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0113383531570435,
+      "learning_rate": 4.2941632732208656e-05,
+      "loss": 5.6047,
+      "step": 1456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9878535866737366,
+      "learning_rate": 4.292224161334109e-05,
+      "loss": 5.5582,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9363996386528015,
+      "learning_rate": 4.290285049447354e-05,
+      "loss": 5.5544,
+      "step": 1464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8429636359214783,
+      "learning_rate": 4.288345937560597e-05,
+      "loss": 5.5561,
+      "step": 1468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0051556825637817,
+      "learning_rate": 4.286406825673842e-05,
+      "loss": 5.5236,
+      "step": 1472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9440407156944275,
+      "learning_rate": 4.284467713787086e-05,
+      "loss": 5.6561,
+      "step": 1476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9376720786094666,
+      "learning_rate": 4.28252860190033e-05,
+      "loss": 5.6352,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0096479654312134,
+      "learning_rate": 4.280589490013574e-05,
+      "loss": 5.5187,
+      "step": 1484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.003917932510376,
+      "learning_rate": 4.278650378126818e-05,
+      "loss": 5.5664,
+      "step": 1488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9462191462516785,
+      "learning_rate": 4.276711266240062e-05,
+      "loss": 5.553,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9691956043243408,
+      "learning_rate": 4.274772154353306e-05,
+      "loss": 5.6481,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8949794769287109,
+      "learning_rate": 4.27283304246655e-05,
+      "loss": 5.5607,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9408190250396729,
+      "learning_rate": 4.270893930579795e-05,
+      "loss": 5.5379,
+      "step": 1504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9462267756462097,
+      "learning_rate": 4.268954818693039e-05,
+      "loss": 5.628,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9751488566398621,
+      "learning_rate": 4.267015706806283e-05,
+      "loss": 5.5709,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0487922430038452,
+      "learning_rate": 4.265076594919527e-05,
+      "loss": 5.5999,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0256918668746948,
+      "learning_rate": 4.263137483032771e-05,
+      "loss": 5.5375,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9403053522109985,
+      "learning_rate": 4.261198371146015e-05,
+      "loss": 5.539,
+      "step": 1524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2121844291687012,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 5.5828,
+      "step": 1528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0676014423370361,
+      "learning_rate": 4.257320147372504e-05,
+      "loss": 5.6215,
+      "step": 1532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9083504676818848,
+      "learning_rate": 4.2553810354857473e-05,
+      "loss": 5.7206,
+      "step": 1536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8858234882354736,
+      "learning_rate": 4.253441923598992e-05,
+      "loss": 5.5742,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9309298396110535,
+      "learning_rate": 4.2515028117122355e-05,
+      "loss": 5.5078,
+      "step": 1544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9594176411628723,
+      "learning_rate": 4.24956369982548e-05,
+      "loss": 5.5863,
+      "step": 1548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8728674650192261,
+      "learning_rate": 4.247624587938724e-05,
+      "loss": 5.5323,
+      "step": 1552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9947747588157654,
+      "learning_rate": 4.245685476051968e-05,
+      "loss": 5.5641,
+      "step": 1556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9186658263206482,
+      "learning_rate": 4.2437463641652123e-05,
+      "loss": 5.6258,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9086698889732361,
+      "learning_rate": 4.241807252278457e-05,
+      "loss": 5.6391,
+      "step": 1564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.954474925994873,
+      "learning_rate": 4.2398681403917005e-05,
+      "loss": 5.6077,
+      "step": 1568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9422994256019592,
+      "learning_rate": 4.237929028504945e-05,
+      "loss": 5.4322,
+      "step": 1572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9581114649772644,
+      "learning_rate": 4.235989916618189e-05,
+      "loss": 5.5757,
+      "step": 1576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9948291778564453,
+      "learning_rate": 4.234050804731433e-05,
+      "loss": 5.5409,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0310508012771606,
+      "learning_rate": 4.2321116928446773e-05,
+      "loss": 5.5628,
+      "step": 1584
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.053040862083435,
+      "learning_rate": 4.2301725809579214e-05,
+      "loss": 5.6698,
+      "step": 1588
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.093757152557373,
+      "learning_rate": 4.2282334690711655e-05,
+      "loss": 5.5696,
+      "step": 1592
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0216928720474243,
+      "learning_rate": 4.22629435718441e-05,
+      "loss": 5.6469,
+      "step": 1596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0343660116195679,
+      "learning_rate": 4.2243552452976536e-05,
+      "loss": 5.6364,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9620201587677002,
+      "learning_rate": 4.222416133410898e-05,
+      "loss": 5.6311,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.032458782196045,
+      "learning_rate": 4.2204770215241423e-05,
+      "loss": 5.523,
+      "step": 1608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9948311448097229,
+      "learning_rate": 4.2185379096373864e-05,
+      "loss": 5.5698,
+      "step": 1612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0208816528320312,
+      "learning_rate": 4.2165987977506305e-05,
+      "loss": 5.6042,
+      "step": 1616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1074092388153076,
+      "learning_rate": 4.2146596858638745e-05,
+      "loss": 5.542,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9410656690597534,
+      "learning_rate": 4.2127205739771186e-05,
+      "loss": 5.5188,
+      "step": 1624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0261708498001099,
+      "learning_rate": 4.2107814620903626e-05,
+      "loss": 5.5377,
+      "step": 1628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9042761325836182,
+      "learning_rate": 4.208842350203607e-05,
+      "loss": 5.5023,
+      "step": 1632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0243982076644897,
+      "learning_rate": 4.2069032383168514e-05,
+      "loss": 5.5674,
+      "step": 1636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0028672218322754,
+      "learning_rate": 4.2049641264300955e-05,
+      "loss": 5.5574,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9813360571861267,
+      "learning_rate": 4.2030250145433395e-05,
+      "loss": 5.5664,
+      "step": 1644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9174672365188599,
+      "learning_rate": 4.2010859026565836e-05,
+      "loss": 5.5837,
+      "step": 1648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1121405363082886,
+      "learning_rate": 4.1991467907698276e-05,
+      "loss": 5.5619,
+      "step": 1652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0016354322433472,
+      "learning_rate": 4.197207678883072e-05,
+      "loss": 5.5928,
+      "step": 1656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9536789655685425,
+      "learning_rate": 4.195268566996316e-05,
+      "loss": 5.5211,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0609054565429688,
+      "learning_rate": 4.19332945510956e-05,
+      "loss": 5.5809,
+      "step": 1664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9688783288002014,
+      "learning_rate": 4.191390343222804e-05,
+      "loss": 5.576,
+      "step": 1668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0187081098556519,
+      "learning_rate": 4.1894512313360486e-05,
+      "loss": 5.569,
+      "step": 1672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9904341101646423,
+      "learning_rate": 4.187512119449292e-05,
+      "loss": 5.5725,
+      "step": 1676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1697142124176025,
+      "learning_rate": 4.185573007562537e-05,
+      "loss": 5.566,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9842191934585571,
+      "learning_rate": 4.183633895675781e-05,
+      "loss": 5.5687,
+      "step": 1684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0618923902511597,
+      "learning_rate": 4.181694783789025e-05,
+      "loss": 5.5165,
+      "step": 1688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9923197031021118,
+      "learning_rate": 4.179755671902269e-05,
+      "loss": 5.5625,
+      "step": 1692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0109176635742188,
+      "learning_rate": 4.1778165600155136e-05,
+      "loss": 5.601,
+      "step": 1696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8870441913604736,
+      "learning_rate": 4.175877448128757e-05,
+      "loss": 5.4932,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.004470705986023,
+      "learning_rate": 4.173938336242002e-05,
+      "loss": 5.5546,
+      "step": 1704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9421555399894714,
+      "learning_rate": 4.171999224355245e-05,
+      "loss": 5.4712,
+      "step": 1708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9490504264831543,
+      "learning_rate": 4.17006011246849e-05,
+      "loss": 5.493,
+      "step": 1712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9385312795639038,
+      "learning_rate": 4.168121000581734e-05,
+      "loss": 5.5997,
+      "step": 1716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9013298153877258,
+      "learning_rate": 4.166181888694978e-05,
+      "loss": 5.4883,
+      "step": 1720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9516580104827881,
+      "learning_rate": 4.164242776808222e-05,
+      "loss": 5.5971,
+      "step": 1724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.033234715461731,
+      "learning_rate": 4.162303664921467e-05,
+      "loss": 5.5807,
+      "step": 1728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.042651653289795,
+      "learning_rate": 4.16036455303471e-05,
+      "loss": 5.5076,
+      "step": 1732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9568614959716797,
+      "learning_rate": 4.158425441147955e-05,
+      "loss": 5.5909,
+      "step": 1736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9669156074523926,
+      "learning_rate": 4.156486329261198e-05,
+      "loss": 5.5952,
+      "step": 1740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9101834893226624,
+      "learning_rate": 4.154547217374443e-05,
+      "loss": 5.5554,
+      "step": 1744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9412431716918945,
+      "learning_rate": 4.152608105487687e-05,
+      "loss": 5.5461,
+      "step": 1748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0211936235427856,
+      "learning_rate": 4.150668993600931e-05,
+      "loss": 5.6353,
+      "step": 1752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0581713914871216,
+      "learning_rate": 4.148729881714175e-05,
+      "loss": 5.5807,
+      "step": 1756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9403213858604431,
+      "learning_rate": 4.146790769827419e-05,
+      "loss": 5.5369,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9232079386711121,
+      "learning_rate": 4.144851657940663e-05,
+      "loss": 5.5559,
+      "step": 1764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9828022122383118,
+      "learning_rate": 4.142912546053908e-05,
+      "loss": 5.4692,
+      "step": 1768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9320650696754456,
+      "learning_rate": 4.140973434167152e-05,
+      "loss": 5.5443,
+      "step": 1772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9043223261833191,
+      "learning_rate": 4.139034322280396e-05,
+      "loss": 5.4807,
+      "step": 1776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9783521294593811,
+      "learning_rate": 4.13709521039364e-05,
+      "loss": 5.575,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0388356447219849,
+      "learning_rate": 4.135156098506884e-05,
+      "loss": 5.5553,
+      "step": 1784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9247937202453613,
+      "learning_rate": 4.133216986620128e-05,
+      "loss": 5.5511,
+      "step": 1788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0111491680145264,
+      "learning_rate": 4.131277874733372e-05,
+      "loss": 5.6378,
+      "step": 1792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9572594165802002,
+      "learning_rate": 4.129338762846616e-05,
+      "loss": 5.4827,
+      "step": 1796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9328434467315674,
+      "learning_rate": 4.12739965095986e-05,
+      "loss": 5.5489,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.000696063041687,
+      "learning_rate": 4.125460539073105e-05,
+      "loss": 5.6207,
+      "step": 1804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.891006350517273,
+      "learning_rate": 4.1235214271863484e-05,
+      "loss": 5.525,
+      "step": 1808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0198568105697632,
+      "learning_rate": 4.121582315299593e-05,
+      "loss": 5.5186,
+      "step": 1812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0678110122680664,
+      "learning_rate": 4.119643203412837e-05,
+      "loss": 5.565,
+      "step": 1816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9433650970458984,
+      "learning_rate": 4.117704091526081e-05,
+      "loss": 5.5038,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.982768714427948,
+      "learning_rate": 4.115764979639325e-05,
+      "loss": 5.4981,
+      "step": 1824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9159711003303528,
+      "learning_rate": 4.1138258677525694e-05,
+      "loss": 5.4945,
+      "step": 1828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0521996021270752,
+      "learning_rate": 4.1118867558658134e-05,
+      "loss": 5.5498,
+      "step": 1832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9633331894874573,
+      "learning_rate": 4.109947643979058e-05,
+      "loss": 5.4915,
+      "step": 1836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9132066965103149,
+      "learning_rate": 4.1080085320923015e-05,
+      "loss": 5.5192,
+      "step": 1840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.947010338306427,
+      "learning_rate": 4.106069420205546e-05,
+      "loss": 5.5928,
+      "step": 1844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1576569080352783,
+      "learning_rate": 4.10413030831879e-05,
+      "loss": 5.6224,
+      "step": 1848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.913221538066864,
+      "learning_rate": 4.1021911964320344e-05,
+      "loss": 5.6261,
+      "step": 1852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9803014397621155,
+      "learning_rate": 4.1002520845452784e-05,
+      "loss": 5.4529,
+      "step": 1856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0535297393798828,
+      "learning_rate": 4.0983129726585225e-05,
+      "loss": 5.5471,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9755203723907471,
+      "learning_rate": 4.0963738607717665e-05,
+      "loss": 5.6235,
+      "step": 1864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8834216594696045,
+      "learning_rate": 4.094434748885011e-05,
+      "loss": 5.5566,
+      "step": 1868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9767342209815979,
+      "learning_rate": 4.0924956369982546e-05,
+      "loss": 5.6874,
+      "step": 1872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.018384575843811,
+      "learning_rate": 4.0905565251114994e-05,
+      "loss": 5.5437,
+      "step": 1876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0440962314605713,
+      "learning_rate": 4.0886174132247434e-05,
+      "loss": 5.594,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0711756944656372,
+      "learning_rate": 4.0866783013379875e-05,
+      "loss": 5.55,
+      "step": 1884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9786011576652527,
+      "learning_rate": 4.0847391894512315e-05,
+      "loss": 5.5315,
+      "step": 1888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1673699617385864,
+      "learning_rate": 4.0828000775644756e-05,
+      "loss": 5.596,
+      "step": 1892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0179039239883423,
+      "learning_rate": 4.0808609656777196e-05,
+      "loss": 5.5635,
+      "step": 1896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0204113721847534,
+      "learning_rate": 4.0789218537909644e-05,
+      "loss": 5.5419,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.919965386390686,
+      "learning_rate": 4.076982741904208e-05,
+      "loss": 5.5283,
+      "step": 1904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.911108136177063,
+      "learning_rate": 4.0750436300174525e-05,
+      "loss": 5.4592,
+      "step": 1908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9294359087944031,
+      "learning_rate": 4.0731045181306965e-05,
+      "loss": 5.5676,
+      "step": 1912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.928774893283844,
+      "learning_rate": 4.0711654062439406e-05,
+      "loss": 5.5844,
+      "step": 1916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0778926610946655,
+      "learning_rate": 4.0692262943571846e-05,
+      "loss": 5.5877,
+      "step": 1920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0779755115509033,
+      "learning_rate": 4.067287182470429e-05,
+      "loss": 5.5362,
+      "step": 1924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9737743139266968,
+      "learning_rate": 4.065348070583673e-05,
+      "loss": 5.584,
+      "step": 1928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9039328098297119,
+      "learning_rate": 4.063408958696917e-05,
+      "loss": 5.441,
+      "step": 1932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0125256776809692,
+      "learning_rate": 4.0614698468101615e-05,
+      "loss": 5.5987,
+      "step": 1936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9862051010131836,
+      "learning_rate": 4.059530734923405e-05,
+      "loss": 5.5512,
+      "step": 1940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8931455016136169,
+      "learning_rate": 4.0575916230366496e-05,
+      "loss": 5.5184,
+      "step": 1944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9782811403274536,
+      "learning_rate": 4.055652511149893e-05,
+      "loss": 5.6231,
+      "step": 1948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9657939076423645,
+      "learning_rate": 4.053713399263138e-05,
+      "loss": 5.5536,
+      "step": 1952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9090112447738647,
+      "learning_rate": 4.051774287376382e-05,
+      "loss": 5.51,
+      "step": 1956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9079639911651611,
+      "learning_rate": 4.049835175489626e-05,
+      "loss": 5.4939,
+      "step": 1960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0238642692565918,
+      "learning_rate": 4.04789606360287e-05,
+      "loss": 5.626,
+      "step": 1964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9550356268882751,
+      "learning_rate": 4.0459569517161146e-05,
+      "loss": 5.5831,
+      "step": 1968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0468617677688599,
+      "learning_rate": 4.044017839829358e-05,
+      "loss": 5.5789,
+      "step": 1972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9686053395271301,
+      "learning_rate": 4.042078727942603e-05,
+      "loss": 5.5384,
+      "step": 1976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0532784461975098,
+      "learning_rate": 4.040139616055846e-05,
+      "loss": 5.617,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9046121835708618,
+      "learning_rate": 4.038200504169091e-05,
+      "loss": 5.6226,
+      "step": 1984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9807924628257751,
+      "learning_rate": 4.036261392282335e-05,
+      "loss": 5.5259,
+      "step": 1988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.957099199295044,
+      "learning_rate": 4.034322280395579e-05,
+      "loss": 5.568,
+      "step": 1992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9218006134033203,
+      "learning_rate": 4.032383168508823e-05,
+      "loss": 5.4665,
+      "step": 1996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.190796971321106,
+      "learning_rate": 4.030444056622068e-05,
+      "loss": 5.4891,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9437822699546814,
+      "learning_rate": 4.028504944735311e-05,
+      "loss": 5.4209,
+      "step": 2004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8980192542076111,
+      "learning_rate": 4.026565832848556e-05,
+      "loss": 5.5624,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1388368606567383,
+      "learning_rate": 4.0246267209618e-05,
+      "loss": 5.5448,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9411901235580444,
+      "learning_rate": 4.022687609075044e-05,
+      "loss": 5.552,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9226595163345337,
+      "learning_rate": 4.020748497188288e-05,
+      "loss": 5.4518,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0351731777191162,
+      "learning_rate": 4.018809385301532e-05,
+      "loss": 5.5449,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.89235919713974,
+      "learning_rate": 4.016870273414776e-05,
+      "loss": 5.563,
+      "step": 2028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9486913084983826,
+      "learning_rate": 4.014931161528021e-05,
+      "loss": 5.4925,
+      "step": 2032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9577587842941284,
+      "learning_rate": 4.012992049641264e-05,
+      "loss": 5.4697,
+      "step": 2036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9649032950401306,
+      "learning_rate": 4.011052937754509e-05,
+      "loss": 5.4694,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8921785950660706,
+      "learning_rate": 4.009113825867753e-05,
+      "loss": 5.6328,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9974868893623352,
+      "learning_rate": 4.007174713980997e-05,
+      "loss": 5.572,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9336390495300293,
+      "learning_rate": 4.005235602094241e-05,
+      "loss": 5.5147,
+      "step": 2052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.942483127117157,
+      "learning_rate": 4.003296490207485e-05,
+      "loss": 5.4991,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9286855459213257,
+      "learning_rate": 4.001357378320729e-05,
+      "loss": 5.5299,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9835842251777649,
+      "learning_rate": 3.999418266433973e-05,
+      "loss": 5.6657,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9683178663253784,
+      "learning_rate": 3.997479154547217e-05,
+      "loss": 5.5365,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0575816631317139,
+      "learning_rate": 3.9955400426604614e-05,
+      "loss": 5.6907,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9691389799118042,
+      "learning_rate": 3.993600930773706e-05,
+      "loss": 5.5166,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0539683103561401,
+      "learning_rate": 3.9916618188869495e-05,
+      "loss": 5.6106,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.010002851486206,
+      "learning_rate": 3.989722707000194e-05,
+      "loss": 5.4635,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9025498032569885,
+      "learning_rate": 3.987783595113438e-05,
+      "loss": 5.5538,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9283170700073242,
+      "learning_rate": 3.985844483226682e-05,
+      "loss": 5.5295,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0095292329788208,
+      "learning_rate": 3.9839053713399264e-05,
+      "loss": 5.4738,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.009535312652588,
+      "learning_rate": 3.9819662594531704e-05,
+      "loss": 5.4439,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9733148217201233,
+      "learning_rate": 3.9800271475664145e-05,
+      "loss": 5.5443,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0330760478973389,
+      "learning_rate": 3.978088035679659e-05,
+      "loss": 5.58,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0041800737380981,
+      "learning_rate": 3.9761489237929026e-05,
+      "loss": 5.5228,
+      "step": 2112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.967785120010376,
+      "learning_rate": 3.974209811906147e-05,
+      "loss": 5.6853,
+      "step": 2116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0202077627182007,
+      "learning_rate": 3.9722707000193914e-05,
+      "loss": 5.5428,
+      "step": 2120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9732391834259033,
+      "learning_rate": 3.9703315881326354e-05,
+      "loss": 5.5364,
+      "step": 2124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9813392162322998,
+      "learning_rate": 3.9683924762458795e-05,
+      "loss": 5.5904,
+      "step": 2128
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9656361937522888,
+      "learning_rate": 3.966453364359124e-05,
+      "loss": 5.4623,
+      "step": 2132
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.956072211265564,
+      "learning_rate": 3.9645142524723676e-05,
+      "loss": 5.5391,
+      "step": 2136
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0806626081466675,
+      "learning_rate": 3.962575140585612e-05,
+      "loss": 5.5746,
+      "step": 2140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0280499458312988,
+      "learning_rate": 3.960636028698856e-05,
+      "loss": 5.5504,
+      "step": 2144
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0023061037063599,
+      "learning_rate": 3.9586969168121004e-05,
+      "loss": 5.5687,
+      "step": 2148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9010854363441467,
+      "learning_rate": 3.9567578049253445e-05,
+      "loss": 5.5546,
+      "step": 2152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0272430181503296,
+      "learning_rate": 3.9548186930385885e-05,
+      "loss": 5.4934,
+      "step": 2156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0653831958770752,
+      "learning_rate": 3.9528795811518326e-05,
+      "loss": 5.496,
+      "step": 2160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0314921140670776,
+      "learning_rate": 3.950940469265077e-05,
+      "loss": 5.5062,
+      "step": 2164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9908810257911682,
+      "learning_rate": 3.949001357378321e-05,
+      "loss": 5.6055,
+      "step": 2168
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9573884606361389,
+      "learning_rate": 3.9470622454915654e-05,
+      "loss": 5.5008,
+      "step": 2172
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9359253644943237,
+      "learning_rate": 3.9451231336048095e-05,
+      "loss": 5.6236,
+      "step": 2176
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0594229698181152,
+      "learning_rate": 3.9431840217180535e-05,
+      "loss": 5.4301,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.994006872177124,
+      "learning_rate": 3.9412449098312976e-05,
+      "loss": 5.5638,
+      "step": 2184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9278011322021484,
+      "learning_rate": 3.9393057979445416e-05,
+      "loss": 5.5056,
+      "step": 2188
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0667142868041992,
+      "learning_rate": 3.937366686057786e-05,
+      "loss": 5.6599,
+      "step": 2192
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9456555843353271,
+      "learning_rate": 3.93542757417103e-05,
+      "loss": 5.5152,
+      "step": 2196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9990763068199158,
+      "learning_rate": 3.933488462284274e-05,
+      "loss": 5.4061,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9991575479507446,
+      "learning_rate": 3.931549350397518e-05,
+      "loss": 5.5047,
+      "step": 2204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0041279792785645,
+      "learning_rate": 3.9296102385107626e-05,
+      "loss": 5.5836,
+      "step": 2208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0032880306243896,
+      "learning_rate": 3.927671126624006e-05,
+      "loss": 5.5682,
+      "step": 2212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0610923767089844,
+      "learning_rate": 3.925732014737251e-05,
+      "loss": 5.597,
+      "step": 2216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9000992178916931,
+      "learning_rate": 3.923792902850494e-05,
+      "loss": 5.54,
+      "step": 2220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9666118621826172,
+      "learning_rate": 3.921853790963739e-05,
+      "loss": 5.568,
+      "step": 2224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.963789701461792,
+      "learning_rate": 3.919914679076983e-05,
+      "loss": 5.5229,
+      "step": 2228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.030738353729248,
+      "learning_rate": 3.917975567190227e-05,
+      "loss": 5.6063,
+      "step": 2232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0029246807098389,
+      "learning_rate": 3.916036455303471e-05,
+      "loss": 5.5573,
+      "step": 2236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0004112720489502,
+      "learning_rate": 3.914097343416716e-05,
+      "loss": 5.4992,
+      "step": 2240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9665903449058533,
+      "learning_rate": 3.912158231529959e-05,
+      "loss": 5.475,
+      "step": 2244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.927628219127655,
+      "learning_rate": 3.910219119643204e-05,
+      "loss": 5.4589,
+      "step": 2248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0179831981658936,
+      "learning_rate": 3.908280007756448e-05,
+      "loss": 5.5458,
+      "step": 2252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8914494514465332,
+      "learning_rate": 3.906340895869692e-05,
+      "loss": 5.4927,
+      "step": 2256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.001958966255188,
+      "learning_rate": 3.904401783982936e-05,
+      "loss": 5.4334,
+      "step": 2260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9571743011474609,
+      "learning_rate": 3.90246267209618e-05,
+      "loss": 5.5728,
+      "step": 2264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0079842805862427,
+      "learning_rate": 3.900523560209424e-05,
+      "loss": 5.5844,
+      "step": 2268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9021591544151306,
+      "learning_rate": 3.898584448322669e-05,
+      "loss": 5.5276,
+      "step": 2272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9413024187088013,
+      "learning_rate": 3.896645336435912e-05,
+      "loss": 5.4693,
+      "step": 2276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9716333150863647,
+      "learning_rate": 3.894706224549157e-05,
+      "loss": 5.434,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.020964503288269,
+      "learning_rate": 3.892767112662401e-05,
+      "loss": 5.4197,
+      "step": 2284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9597110748291016,
+      "learning_rate": 3.890828000775645e-05,
+      "loss": 5.4174,
+      "step": 2288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0164399147033691,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 5.4698,
+      "step": 2292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0590660572052002,
+      "learning_rate": 3.886949777002134e-05,
+      "loss": 5.4815,
+      "step": 2296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9020886421203613,
+      "learning_rate": 3.885010665115377e-05,
+      "loss": 5.4197,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9899044632911682,
+      "learning_rate": 3.883071553228622e-05,
+      "loss": 5.5061,
+      "step": 2304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9730533957481384,
+      "learning_rate": 3.881132441341865e-05,
+      "loss": 5.4804,
+      "step": 2308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9274543523788452,
+      "learning_rate": 3.87919332945511e-05,
+      "loss": 5.5486,
+      "step": 2312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9086050391197205,
+      "learning_rate": 3.877254217568354e-05,
+      "loss": 5.3997,
+      "step": 2316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9170548915863037,
+      "learning_rate": 3.875315105681598e-05,
+      "loss": 5.3701,
+      "step": 2320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9133801460266113,
+      "learning_rate": 3.873375993794842e-05,
+      "loss": 5.4241,
+      "step": 2324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9891861081123352,
+      "learning_rate": 3.871436881908086e-05,
+      "loss": 5.5119,
+      "step": 2328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0900509357452393,
+      "learning_rate": 3.86949777002133e-05,
+      "loss": 5.6089,
+      "step": 2332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9711102843284607,
+      "learning_rate": 3.8675586581345743e-05,
+      "loss": 5.5975,
+      "step": 2336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9177804589271545,
+      "learning_rate": 3.8656195462478184e-05,
+      "loss": 5.4848,
+      "step": 2340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9340533018112183,
+      "learning_rate": 3.8636804343610625e-05,
+      "loss": 5.5705,
+      "step": 2344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9871985912322998,
+      "learning_rate": 3.861741322474307e-05,
+      "loss": 5.5552,
+      "step": 2348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9477519392967224,
+      "learning_rate": 3.8598022105875506e-05,
+      "loss": 5.3622,
+      "step": 2352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9251902103424072,
+      "learning_rate": 3.857863098700795e-05,
+      "loss": 5.3964,
+      "step": 2356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9803330898284912,
+      "learning_rate": 3.8559239868140393e-05,
+      "loss": 5.4716,
+      "step": 2360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9429686069488525,
+      "learning_rate": 3.8539848749272834e-05,
+      "loss": 5.4319,
+      "step": 2364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0370640754699707,
+      "learning_rate": 3.8520457630405275e-05,
+      "loss": 5.4297,
+      "step": 2368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9429317712783813,
+      "learning_rate": 3.850106651153772e-05,
+      "loss": 5.4068,
+      "step": 2372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0374614000320435,
+      "learning_rate": 3.8481675392670156e-05,
+      "loss": 5.4975,
+      "step": 2376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9564975500106812,
+      "learning_rate": 3.84622842738026e-05,
+      "loss": 5.5221,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9388477802276611,
+      "learning_rate": 3.844289315493504e-05,
+      "loss": 5.4748,
+      "step": 2384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9292894601821899,
+      "learning_rate": 3.8423502036067484e-05,
+      "loss": 5.6158,
+      "step": 2388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9442563056945801,
+      "learning_rate": 3.8404110917199925e-05,
+      "loss": 5.5371,
+      "step": 2392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0346298217773438,
+      "learning_rate": 3.8384719798332365e-05,
+      "loss": 5.4546,
+      "step": 2396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9618560671806335,
+      "learning_rate": 3.8365328679464806e-05,
+      "loss": 5.5663,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9630410671234131,
+      "learning_rate": 3.834593756059725e-05,
+      "loss": 5.5669,
+      "step": 2404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0021862983703613,
+      "learning_rate": 3.832654644172969e-05,
+      "loss": 5.5155,
+      "step": 2408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.015215277671814,
+      "learning_rate": 3.8307155322862134e-05,
+      "loss": 5.4772,
+      "step": 2412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9358635544776917,
+      "learning_rate": 3.8287764203994574e-05,
+      "loss": 5.4375,
+      "step": 2416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1098734140396118,
+      "learning_rate": 3.8268373085127015e-05,
+      "loss": 5.5493,
+      "step": 2420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9704833030700684,
+      "learning_rate": 3.8248981966259456e-05,
+      "loss": 5.5974,
+      "step": 2424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1319690942764282,
+      "learning_rate": 3.8229590847391896e-05,
+      "loss": 5.5381,
+      "step": 2428
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9214887619018555,
+      "learning_rate": 3.821019972852434e-05,
+      "loss": 5.4689,
+      "step": 2432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9483059048652649,
+      "learning_rate": 3.8190808609656784e-05,
+      "loss": 5.447,
+      "step": 2436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.936059832572937,
+      "learning_rate": 3.817141749078922e-05,
+      "loss": 5.5256,
+      "step": 2440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8800360560417175,
+      "learning_rate": 3.8152026371921665e-05,
+      "loss": 5.4862,
+      "step": 2444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9923036098480225,
+      "learning_rate": 3.8132635253054106e-05,
+      "loss": 5.5005,
+      "step": 2448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9370976090431213,
+      "learning_rate": 3.8113244134186546e-05,
+      "loss": 5.5008,
+      "step": 2452
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9325810670852661,
+      "learning_rate": 3.809385301531899e-05,
+      "loss": 5.5375,
+      "step": 2456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0524775981903076,
+      "learning_rate": 3.807446189645143e-05,
+      "loss": 5.5869,
+      "step": 2460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9688572287559509,
+      "learning_rate": 3.805507077758387e-05,
+      "loss": 5.5657,
+      "step": 2464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9577921628952026,
+      "learning_rate": 3.803567965871631e-05,
+      "loss": 5.4082,
+      "step": 2468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.977051317691803,
+      "learning_rate": 3.801628853984875e-05,
+      "loss": 5.4773,
+      "step": 2472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9854933023452759,
+      "learning_rate": 3.799689742098119e-05,
+      "loss": 5.487,
+      "step": 2476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0686819553375244,
+      "learning_rate": 3.797750630211364e-05,
+      "loss": 5.3842,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9736838936805725,
+      "learning_rate": 3.795811518324607e-05,
+      "loss": 5.5956,
+      "step": 2484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9340422749519348,
+      "learning_rate": 3.793872406437852e-05,
+      "loss": 5.5148,
+      "step": 2488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0714948177337646,
+      "learning_rate": 3.791933294551096e-05,
+      "loss": 5.495,
+      "step": 2492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.037858486175537,
+      "learning_rate": 3.78999418266434e-05,
+      "loss": 5.5312,
+      "step": 2496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1142346858978271,
+      "learning_rate": 3.788055070777584e-05,
+      "loss": 5.6109,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0405195951461792,
+      "learning_rate": 3.786115958890828e-05,
+      "loss": 5.4659,
+      "step": 2504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.106404423713684,
+      "learning_rate": 3.784176847004072e-05,
+      "loss": 5.5187,
+      "step": 2508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9529224634170532,
+      "learning_rate": 3.782237735117317e-05,
+      "loss": 5.441,
+      "step": 2512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0962753295898438,
+      "learning_rate": 3.78029862323056e-05,
+      "loss": 5.5503,
+      "step": 2516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9510455131530762,
+      "learning_rate": 3.778359511343805e-05,
+      "loss": 5.4773,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0108531713485718,
+      "learning_rate": 3.776420399457049e-05,
+      "loss": 5.5567,
+      "step": 2524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9611102938652039,
+      "learning_rate": 3.774481287570293e-05,
+      "loss": 5.4825,
+      "step": 2528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0297412872314453,
+      "learning_rate": 3.772542175683537e-05,
+      "loss": 5.4075,
+      "step": 2532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0067005157470703,
+      "learning_rate": 3.770603063796782e-05,
+      "loss": 5.5345,
+      "step": 2536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0144344568252563,
+      "learning_rate": 3.768663951910025e-05,
+      "loss": 5.5108,
+      "step": 2540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9989475607872009,
+      "learning_rate": 3.76672484002327e-05,
+      "loss": 5.4782,
+      "step": 2544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0596572160720825,
+      "learning_rate": 3.764785728136513e-05,
+      "loss": 5.5088,
+      "step": 2548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.022268533706665,
+      "learning_rate": 3.762846616249758e-05,
+      "loss": 5.4595,
+      "step": 2552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9864400029182434,
+      "learning_rate": 3.760907504363002e-05,
+      "loss": 5.5029,
+      "step": 2556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9575673937797546,
+      "learning_rate": 3.758968392476246e-05,
+      "loss": 5.4772,
+      "step": 2560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9066863059997559,
+      "learning_rate": 3.75702928058949e-05,
+      "loss": 5.5936,
+      "step": 2564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9975427985191345,
+      "learning_rate": 3.755090168702735e-05,
+      "loss": 5.5532,
+      "step": 2568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.935875415802002,
+      "learning_rate": 3.753151056815978e-05,
+      "loss": 5.4235,
+      "step": 2572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0825345516204834,
+      "learning_rate": 3.751211944929223e-05,
+      "loss": 5.3898,
+      "step": 2576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0362260341644287,
+      "learning_rate": 3.7492728330424664e-05,
+      "loss": 5.4427,
+      "step": 2580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8955732583999634,
+      "learning_rate": 3.747333721155711e-05,
+      "loss": 5.5447,
+      "step": 2584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0196340084075928,
+      "learning_rate": 3.745394609268955e-05,
+      "loss": 5.5636,
+      "step": 2588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9525064826011658,
+      "learning_rate": 3.743455497382199e-05,
+      "loss": 5.5,
+      "step": 2592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9297643899917603,
+      "learning_rate": 3.741516385495443e-05,
+      "loss": 5.5382,
+      "step": 2596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9364489912986755,
+      "learning_rate": 3.739577273608687e-05,
+      "loss": 5.5248,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9031673073768616,
+      "learning_rate": 3.7376381617219314e-05,
+      "loss": 5.4689,
+      "step": 2604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.020928978919983,
+      "learning_rate": 3.7356990498351754e-05,
+      "loss": 5.4896,
+      "step": 2608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9407410621643066,
+      "learning_rate": 3.73375993794842e-05,
+      "loss": 5.4259,
+      "step": 2612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0018398761749268,
+      "learning_rate": 3.7318208260616635e-05,
+      "loss": 5.5738,
+      "step": 2616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0489344596862793,
+      "learning_rate": 3.729881714174908e-05,
+      "loss": 5.404,
+      "step": 2620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9759474992752075,
+      "learning_rate": 3.7279426022881516e-05,
+      "loss": 5.4977,
+      "step": 2624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0580724477767944,
+      "learning_rate": 3.7260034904013964e-05,
+      "loss": 5.4477,
+      "step": 2628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0281325578689575,
+      "learning_rate": 3.7240643785146404e-05,
+      "loss": 5.4882,
+      "step": 2632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9635825157165527,
+      "learning_rate": 3.7221252666278845e-05,
+      "loss": 5.4303,
+      "step": 2636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9374428391456604,
+      "learning_rate": 3.7201861547411285e-05,
+      "loss": 5.4308,
+      "step": 2640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0188990831375122,
+      "learning_rate": 3.718247042854373e-05,
+      "loss": 5.4521,
+      "step": 2644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0239014625549316,
+      "learning_rate": 3.7163079309676166e-05,
+      "loss": 5.5686,
+      "step": 2648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.946735680103302,
+      "learning_rate": 3.7143688190808614e-05,
+      "loss": 5.4329,
+      "step": 2652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9370056986808777,
+      "learning_rate": 3.7124297071941054e-05,
+      "loss": 5.5174,
+      "step": 2656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.003072738647461,
+      "learning_rate": 3.7104905953073495e-05,
+      "loss": 5.3857,
+      "step": 2660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9541458487510681,
+      "learning_rate": 3.7085514834205935e-05,
+      "loss": 5.448,
+      "step": 2664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9533443450927734,
+      "learning_rate": 3.7066123715338376e-05,
+      "loss": 5.559,
+      "step": 2668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0223768949508667,
+      "learning_rate": 3.7046732596470816e-05,
+      "loss": 5.503,
+      "step": 2672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2174021005630493,
+      "learning_rate": 3.7027341477603264e-05,
+      "loss": 5.5322,
+      "step": 2676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.117325782775879,
+      "learning_rate": 3.70079503587357e-05,
+      "loss": 5.4969,
+      "step": 2680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9299269318580627,
+      "learning_rate": 3.6988559239868145e-05,
+      "loss": 5.4264,
+      "step": 2684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9750757813453674,
+      "learning_rate": 3.6969168121000585e-05,
+      "loss": 5.4953,
+      "step": 2688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9810564517974854,
+      "learning_rate": 3.6949777002133026e-05,
+      "loss": 5.4762,
+      "step": 2692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0046603679656982,
+      "learning_rate": 3.6930385883265466e-05,
+      "loss": 5.4572,
+      "step": 2696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9024963974952698,
+      "learning_rate": 3.691099476439791e-05,
+      "loss": 5.5038,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9631572961807251,
+      "learning_rate": 3.689160364553035e-05,
+      "loss": 5.4895,
+      "step": 2704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8802670240402222,
+      "learning_rate": 3.6872212526662795e-05,
+      "loss": 5.6245,
+      "step": 2708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9694925546646118,
+      "learning_rate": 3.685282140779523e-05,
+      "loss": 5.4813,
+      "step": 2712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0332534313201904,
+      "learning_rate": 3.6833430288927676e-05,
+      "loss": 5.5064,
+      "step": 2716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9285298585891724,
+      "learning_rate": 3.6814039170060116e-05,
+      "loss": 5.4926,
+      "step": 2720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9079506993293762,
+      "learning_rate": 3.679464805119256e-05,
+      "loss": 5.4837,
+      "step": 2724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0010629892349243,
+      "learning_rate": 3.6775256932325e-05,
+      "loss": 5.4814,
+      "step": 2728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9733301997184753,
+      "learning_rate": 3.675586581345744e-05,
+      "loss": 5.526,
+      "step": 2732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9595903158187866,
+      "learning_rate": 3.673647469458988e-05,
+      "loss": 5.3726,
+      "step": 2736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0144261121749878,
+      "learning_rate": 3.671708357572232e-05,
+      "loss": 5.4286,
+      "step": 2740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0285661220550537,
+      "learning_rate": 3.669769245685476e-05,
+      "loss": 5.5113,
+      "step": 2744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9807763695716858,
+      "learning_rate": 3.66783013379872e-05,
+      "loss": 5.5956,
+      "step": 2748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0480782985687256,
+      "learning_rate": 3.665891021911965e-05,
+      "loss": 5.4205,
+      "step": 2752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8527302145957947,
+      "learning_rate": 3.663951910025208e-05,
+      "loss": 5.5106,
+      "step": 2756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9895337224006653,
+      "learning_rate": 3.662012798138453e-05,
+      "loss": 5.517,
+      "step": 2760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.005570888519287,
+      "learning_rate": 3.660073686251697e-05,
+      "loss": 5.4432,
+      "step": 2764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0083740949630737,
+      "learning_rate": 3.658134574364941e-05,
+      "loss": 5.3863,
+      "step": 2768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9431845545768738,
+      "learning_rate": 3.656195462478185e-05,
+      "loss": 5.3734,
+      "step": 2772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9629083871841431,
+      "learning_rate": 3.65425635059143e-05,
+      "loss": 5.4737,
+      "step": 2776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9649605751037598,
+      "learning_rate": 3.652317238704673e-05,
+      "loss": 5.4022,
+      "step": 2780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9746363162994385,
+      "learning_rate": 3.650378126817918e-05,
+      "loss": 5.4826,
+      "step": 2784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9153027534484863,
+      "learning_rate": 3.648439014931161e-05,
+      "loss": 5.5833,
+      "step": 2788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.940949559211731,
+      "learning_rate": 3.646499903044406e-05,
+      "loss": 5.4852,
+      "step": 2792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9482103586196899,
+      "learning_rate": 3.64456079115765e-05,
+      "loss": 5.4711,
+      "step": 2796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0848538875579834,
+      "learning_rate": 3.642621679270894e-05,
+      "loss": 5.464,
+      "step": 2800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9354459643363953,
+      "learning_rate": 3.640682567384138e-05,
+      "loss": 5.4495,
+      "step": 2804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9546772837638855,
+      "learning_rate": 3.638743455497383e-05,
+      "loss": 5.4562,
+      "step": 2808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9387646317481995,
+      "learning_rate": 3.636804343610626e-05,
+      "loss": 5.4439,
+      "step": 2812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9842014312744141,
+      "learning_rate": 3.634865231723871e-05,
+      "loss": 5.3971,
+      "step": 2816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9438384175300598,
+      "learning_rate": 3.632926119837114e-05,
+      "loss": 5.4875,
+      "step": 2820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.998210072517395,
+      "learning_rate": 3.630987007950359e-05,
+      "loss": 5.4301,
+      "step": 2824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.948137640953064,
+      "learning_rate": 3.629047896063603e-05,
+      "loss": 5.5703,
+      "step": 2828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0335065126419067,
+      "learning_rate": 3.627108784176847e-05,
+      "loss": 5.386,
+      "step": 2832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9774126410484314,
+      "learning_rate": 3.625169672290091e-05,
+      "loss": 5.4288,
+      "step": 2836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.168003797531128,
+      "learning_rate": 3.623230560403336e-05,
+      "loss": 5.4524,
+      "step": 2840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.05010187625885,
+      "learning_rate": 3.621291448516579e-05,
+      "loss": 5.3784,
+      "step": 2844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0008686780929565,
+      "learning_rate": 3.619352336629824e-05,
+      "loss": 5.351,
+      "step": 2848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0319279432296753,
+      "learning_rate": 3.617413224743068e-05,
+      "loss": 5.5112,
+      "step": 2852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9444233775138855,
+      "learning_rate": 3.615474112856312e-05,
+      "loss": 5.4225,
+      "step": 2856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0696698427200317,
+      "learning_rate": 3.613535000969556e-05,
+      "loss": 5.5542,
+      "step": 2860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9493553042411804,
+      "learning_rate": 3.6115958890828e-05,
+      "loss": 5.42,
+      "step": 2864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0606472492218018,
+      "learning_rate": 3.609656777196044e-05,
+      "loss": 5.421,
+      "step": 2868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.045782208442688,
+      "learning_rate": 3.6077176653092884e-05,
+      "loss": 5.5086,
+      "step": 2872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.034601092338562,
+      "learning_rate": 3.6057785534225324e-05,
+      "loss": 5.4582,
+      "step": 2876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0553306341171265,
+      "learning_rate": 3.6038394415357765e-05,
+      "loss": 5.5501,
+      "step": 2880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0070221424102783,
+      "learning_rate": 3.601900329649021e-05,
+      "loss": 5.5016,
+      "step": 2884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0406205654144287,
+      "learning_rate": 3.5999612177622646e-05,
+      "loss": 5.4743,
+      "step": 2888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0793685913085938,
+      "learning_rate": 3.598022105875509e-05,
+      "loss": 5.4889,
+      "step": 2892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1032297611236572,
+      "learning_rate": 3.5960829939887534e-05,
+      "loss": 5.4928,
+      "step": 2896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9864259362220764,
+      "learning_rate": 3.5941438821019974e-05,
+      "loss": 5.4582,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9251708388328552,
+      "learning_rate": 3.5922047702152415e-05,
+      "loss": 5.4329,
+      "step": 2904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9993565082550049,
+      "learning_rate": 3.5902656583284855e-05,
+      "loss": 5.5837,
+      "step": 2908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9946919083595276,
+      "learning_rate": 3.5883265464417296e-05,
+      "loss": 5.4545,
+      "step": 2912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9719089865684509,
+      "learning_rate": 3.586387434554974e-05,
+      "loss": 5.448,
+      "step": 2916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.900641143321991,
+      "learning_rate": 3.584448322668218e-05,
+      "loss": 5.5845,
+      "step": 2920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9279571771621704,
+      "learning_rate": 3.5825092107814624e-05,
+      "loss": 5.4733,
+      "step": 2924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0747668743133545,
+      "learning_rate": 3.5805700988947065e-05,
+      "loss": 5.3805,
+      "step": 2928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8932091593742371,
+      "learning_rate": 3.5786309870079505e-05,
+      "loss": 5.4015,
+      "step": 2932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9807014465332031,
+      "learning_rate": 3.5766918751211946e-05,
+      "loss": 5.5228,
+      "step": 2936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9529114961624146,
+      "learning_rate": 3.5747527632344386e-05,
+      "loss": 5.5328,
+      "step": 2940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9890924692153931,
+      "learning_rate": 3.572813651347683e-05,
+      "loss": 5.4686,
+      "step": 2944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9855780005455017,
+      "learning_rate": 3.5708745394609274e-05,
+      "loss": 5.4921,
+      "step": 2948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9508200287818909,
+      "learning_rate": 3.568935427574171e-05,
+      "loss": 5.5508,
+      "step": 2952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9192949533462524,
+      "learning_rate": 3.5669963156874155e-05,
+      "loss": 5.5484,
+      "step": 2956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9657400846481323,
+      "learning_rate": 3.5650572038006596e-05,
+      "loss": 5.4305,
+      "step": 2960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1960434913635254,
+      "learning_rate": 3.5631180919139036e-05,
+      "loss": 5.4461,
+      "step": 2964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9226086735725403,
+      "learning_rate": 3.561178980027148e-05,
+      "loss": 5.4288,
+      "step": 2968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0317691564559937,
+      "learning_rate": 3.5592398681403924e-05,
+      "loss": 5.5107,
+      "step": 2972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9748375415802002,
+      "learning_rate": 3.557300756253636e-05,
+      "loss": 5.5342,
+      "step": 2976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9513313174247742,
+      "learning_rate": 3.5553616443668805e-05,
+      "loss": 5.4337,
+      "step": 2980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9269315600395203,
+      "learning_rate": 3.553422532480124e-05,
+      "loss": 5.4853,
+      "step": 2984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0136945247650146,
+      "learning_rate": 3.5514834205933686e-05,
+      "loss": 5.4767,
+      "step": 2988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.049842119216919,
+      "learning_rate": 3.549544308706613e-05,
+      "loss": 5.5028,
+      "step": 2992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9470251798629761,
+      "learning_rate": 3.547605196819857e-05,
+      "loss": 5.4391,
+      "step": 2996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1014668941497803,
+      "learning_rate": 3.545666084933101e-05,
+      "loss": 5.4564,
+      "step": 3000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0358186960220337,
+      "learning_rate": 3.543726973046345e-05,
+      "loss": 5.5398,
+      "step": 3004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0642112493515015,
+      "learning_rate": 3.541787861159589e-05,
+      "loss": 5.4548,
+      "step": 3008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9606940150260925,
+      "learning_rate": 3.539848749272833e-05,
+      "loss": 5.5331,
+      "step": 3012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0315513610839844,
+      "learning_rate": 3.537909637386078e-05,
+      "loss": 5.476,
+      "step": 3016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0335862636566162,
+      "learning_rate": 3.535970525499321e-05,
+      "loss": 5.3247,
+      "step": 3020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0279650688171387,
+      "learning_rate": 3.534031413612566e-05,
+      "loss": 5.425,
+      "step": 3024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9862622022628784,
+      "learning_rate": 3.532092301725809e-05,
+      "loss": 5.5486,
+      "step": 3028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0324090719223022,
+      "learning_rate": 3.530153189839054e-05,
+      "loss": 5.546,
+      "step": 3032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0767823457717896,
+      "learning_rate": 3.528214077952298e-05,
+      "loss": 5.4906,
+      "step": 3036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9795736074447632,
+      "learning_rate": 3.526274966065542e-05,
+      "loss": 5.4132,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9786263704299927,
+      "learning_rate": 3.524335854178786e-05,
+      "loss": 5.4225,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9306228756904602,
+      "learning_rate": 3.522396742292031e-05,
+      "loss": 5.5407,
+      "step": 3048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9735816121101379,
+      "learning_rate": 3.520457630405274e-05,
+      "loss": 5.5262,
+      "step": 3052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9719963073730469,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 5.4489,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9754818081855774,
+      "learning_rate": 3.516579406631762e-05,
+      "loss": 5.4656,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0036709308624268,
+      "learning_rate": 3.514640294745007e-05,
+      "loss": 5.4314,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.956697940826416,
+      "learning_rate": 3.512701182858251e-05,
+      "loss": 5.4755,
+      "step": 3068
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1752293109893799,
+      "learning_rate": 3.510762070971495e-05,
+      "loss": 5.4411,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9563004374504089,
+      "learning_rate": 3.508822959084739e-05,
+      "loss": 5.3531,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0313175916671753,
+      "learning_rate": 3.506883847197984e-05,
+      "loss": 5.477,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.997872531414032,
+      "learning_rate": 3.504944735311227e-05,
+      "loss": 5.4479,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9374060034751892,
+      "learning_rate": 3.503005623424472e-05,
+      "loss": 5.4492,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9899947643280029,
+      "learning_rate": 3.501066511537716e-05,
+      "loss": 5.4255,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0138983726501465,
+      "learning_rate": 3.49912739965096e-05,
+      "loss": 5.4234,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9544614553451538,
+      "learning_rate": 3.497188287764204e-05,
+      "loss": 5.4781,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0117629766464233,
+      "learning_rate": 3.495249175877448e-05,
+      "loss": 5.4445,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0312600135803223,
+      "learning_rate": 3.493310063990692e-05,
+      "loss": 5.439,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9983291029930115,
+      "learning_rate": 3.491370952103937e-05,
+      "loss": 5.5003,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0281239748001099,
+      "learning_rate": 3.4894318402171804e-05,
+      "loss": 5.4276,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9426625370979309,
+      "learning_rate": 3.487492728330425e-05,
+      "loss": 5.5267,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0768100023269653,
+      "learning_rate": 3.485553616443669e-05,
+      "loss": 5.5023,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0463875532150269,
+      "learning_rate": 3.483614504556913e-05,
+      "loss": 5.4019,
+      "step": 3128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9380079507827759,
+      "learning_rate": 3.481675392670157e-05,
+      "loss": 5.5364,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9299972653388977,
+      "learning_rate": 3.4797362807834013e-05,
+      "loss": 5.4026,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9217830300331116,
+      "learning_rate": 3.4777971688966454e-05,
+      "loss": 5.6161,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.983069121837616,
+      "learning_rate": 3.4758580570098894e-05,
+      "loss": 5.4112,
+      "step": 3144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0096659660339355,
+      "learning_rate": 3.4739189451231335e-05,
+      "loss": 5.4705,
+      "step": 3148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9388656616210938,
+      "learning_rate": 3.4719798332363776e-05,
+      "loss": 5.4713,
+      "step": 3152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0393484830856323,
+      "learning_rate": 3.470040721349622e-05,
+      "loss": 5.4147,
+      "step": 3156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.001868724822998,
+      "learning_rate": 3.468101609462866e-05,
+      "loss": 5.3701,
+      "step": 3160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9857000708580017,
+      "learning_rate": 3.4661624975761104e-05,
+      "loss": 5.4891,
+      "step": 3164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0586354732513428,
+      "learning_rate": 3.4642233856893544e-05,
+      "loss": 5.4801,
+      "step": 3168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0346976518630981,
+      "learning_rate": 3.4622842738025985e-05,
+      "loss": 5.4555,
+      "step": 3172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9425565004348755,
+      "learning_rate": 3.4603451619158426e-05,
+      "loss": 5.4126,
+      "step": 3176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.984109103679657,
+      "learning_rate": 3.4584060500290866e-05,
+      "loss": 5.3515,
+      "step": 3180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0639657974243164,
+      "learning_rate": 3.456466938142331e-05,
+      "loss": 5.3965,
+      "step": 3184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9687911868095398,
+      "learning_rate": 3.4545278262555754e-05,
+      "loss": 5.4377,
+      "step": 3188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1144814491271973,
+      "learning_rate": 3.452588714368819e-05,
+      "loss": 5.5135,
+      "step": 3192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.94063401222229,
+      "learning_rate": 3.4506496024820635e-05,
+      "loss": 5.4472,
+      "step": 3196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9204466938972473,
+      "learning_rate": 3.4487104905953076e-05,
+      "loss": 5.3214,
+      "step": 3200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0223456621170044,
+      "learning_rate": 3.4467713787085516e-05,
+      "loss": 5.5622,
+      "step": 3204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9197329878807068,
+      "learning_rate": 3.444832266821796e-05,
+      "loss": 5.4799,
+      "step": 3208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9634442329406738,
+      "learning_rate": 3.4428931549350404e-05,
+      "loss": 5.3426,
+      "step": 3212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0423585176467896,
+      "learning_rate": 3.440954043048284e-05,
+      "loss": 5.4419,
+      "step": 3216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0101354122161865,
+      "learning_rate": 3.4390149311615285e-05,
+      "loss": 5.4811,
+      "step": 3220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0531187057495117,
+      "learning_rate": 3.437075819274772e-05,
+      "loss": 5.4645,
+      "step": 3224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0495266914367676,
+      "learning_rate": 3.4351367073880166e-05,
+      "loss": 5.4345,
+      "step": 3228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.029811978340149,
+      "learning_rate": 3.433197595501261e-05,
+      "loss": 5.4607,
+      "step": 3232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9279022812843323,
+      "learning_rate": 3.431258483614505e-05,
+      "loss": 5.4771,
+      "step": 3236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.026243805885315,
+      "learning_rate": 3.429319371727749e-05,
+      "loss": 5.4841,
+      "step": 3240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9428079724311829,
+      "learning_rate": 3.4273802598409935e-05,
+      "loss": 5.5417,
+      "step": 3244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9030874967575073,
+      "learning_rate": 3.425441147954237e-05,
+      "loss": 5.443,
+      "step": 3248
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.981732189655304,
+      "learning_rate": 3.4235020360674816e-05,
+      "loss": 5.5341,
+      "step": 3252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9747270941734314,
+      "learning_rate": 3.421562924180726e-05,
+      "loss": 5.4236,
+      "step": 3256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9781522154808044,
+      "learning_rate": 3.41962381229397e-05,
+      "loss": 5.4404,
+      "step": 3260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9788567423820496,
+      "learning_rate": 3.417684700407214e-05,
+      "loss": 5.3855,
+      "step": 3264
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9978493452072144,
+      "learning_rate": 3.415745588520458e-05,
+      "loss": 5.3939,
+      "step": 3268
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0338048934936523,
+      "learning_rate": 3.413806476633702e-05,
+      "loss": 5.464,
+      "step": 3272
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9052521586418152,
+      "learning_rate": 3.411867364746946e-05,
+      "loss": 5.3242,
+      "step": 3276
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9726389050483704,
+      "learning_rate": 3.40992825286019e-05,
+      "loss": 5.4116,
+      "step": 3280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.987234354019165,
+      "learning_rate": 3.407989140973434e-05,
+      "loss": 5.506,
+      "step": 3284
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0017744302749634,
+      "learning_rate": 3.406050029086679e-05,
+      "loss": 5.5025,
+      "step": 3288
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9768481850624084,
+      "learning_rate": 3.404110917199922e-05,
+      "loss": 5.484,
+      "step": 3292
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9455767273902893,
+      "learning_rate": 3.402171805313167e-05,
+      "loss": 5.4146,
+      "step": 3296
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9832062721252441,
+      "learning_rate": 3.400232693426411e-05,
+      "loss": 5.4591,
+      "step": 3300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8915082216262817,
+      "learning_rate": 3.398293581539655e-05,
+      "loss": 5.3985,
+      "step": 3304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0324420928955078,
+      "learning_rate": 3.396354469652899e-05,
+      "loss": 5.5157,
+      "step": 3308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0704431533813477,
+      "learning_rate": 3.394415357766143e-05,
+      "loss": 5.4155,
+      "step": 3312
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9540812373161316,
+      "learning_rate": 3.392476245879387e-05,
+      "loss": 5.418,
+      "step": 3316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0278005599975586,
+      "learning_rate": 3.390537133992632e-05,
+      "loss": 5.4143,
+      "step": 3320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0574851036071777,
+      "learning_rate": 3.388598022105875e-05,
+      "loss": 5.3889,
+      "step": 3324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1078550815582275,
+      "learning_rate": 3.38665891021912e-05,
+      "loss": 5.4035,
+      "step": 3328
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9876176714897156,
+      "learning_rate": 3.384719798332364e-05,
+      "loss": 5.4488,
+      "step": 3332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9665130376815796,
+      "learning_rate": 3.382780686445608e-05,
+      "loss": 5.3989,
+      "step": 3336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0209985971450806,
+      "learning_rate": 3.380841574558852e-05,
+      "loss": 5.4641,
+      "step": 3340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.920693039894104,
+      "learning_rate": 3.378902462672096e-05,
+      "loss": 5.3932,
+      "step": 3344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0622704029083252,
+      "learning_rate": 3.37696335078534e-05,
+      "loss": 5.3708,
+      "step": 3348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.018336296081543,
+      "learning_rate": 3.375024238898585e-05,
+      "loss": 5.4366,
+      "step": 3352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9588587284088135,
+      "learning_rate": 3.3730851270118284e-05,
+      "loss": 5.4806,
+      "step": 3356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9901473522186279,
+      "learning_rate": 3.371146015125073e-05,
+      "loss": 5.4993,
+      "step": 3360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0226725339889526,
+      "learning_rate": 3.369206903238317e-05,
+      "loss": 5.4429,
+      "step": 3364
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0187616348266602,
+      "learning_rate": 3.367267791351561e-05,
+      "loss": 5.4496,
+      "step": 3368
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9235848188400269,
+      "learning_rate": 3.365328679464805e-05,
+      "loss": 5.4033,
+      "step": 3372
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9983669519424438,
+      "learning_rate": 3.36338956757805e-05,
+      "loss": 5.3515,
+      "step": 3376
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9427633285522461,
+      "learning_rate": 3.3614504556912934e-05,
+      "loss": 5.5194,
+      "step": 3380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9658553004264832,
+      "learning_rate": 3.359511343804538e-05,
+      "loss": 5.489,
+      "step": 3384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9263963103294373,
+      "learning_rate": 3.3575722319177815e-05,
+      "loss": 5.4182,
+      "step": 3388
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0251826047897339,
+      "learning_rate": 3.355633120031026e-05,
+      "loss": 5.5208,
+      "step": 3392
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9037973880767822,
+      "learning_rate": 3.35369400814427e-05,
+      "loss": 5.3744,
+      "step": 3396
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.096735954284668,
+      "learning_rate": 3.351754896257514e-05,
+      "loss": 5.4218,
+      "step": 3400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9776617884635925,
+      "learning_rate": 3.3498157843707584e-05,
+      "loss": 5.4539,
+      "step": 3404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0098875761032104,
+      "learning_rate": 3.3478766724840024e-05,
+      "loss": 5.4252,
+      "step": 3408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0725504159927368,
+      "learning_rate": 3.3459375605972465e-05,
+      "loss": 5.4702,
+      "step": 3412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8836826086044312,
+      "learning_rate": 3.3439984487104905e-05,
+      "loss": 5.4135,
+      "step": 3416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9869784116744995,
+      "learning_rate": 3.3420593368237346e-05,
+      "loss": 5.4617,
+      "step": 3420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0457979440689087,
+      "learning_rate": 3.3401202249369786e-05,
+      "loss": 5.4872,
+      "step": 3424
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0574203729629517,
+      "learning_rate": 3.3381811130502234e-05,
+      "loss": 5.4167,
+      "step": 3428
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.001185655593872,
+      "learning_rate": 3.3362420011634674e-05,
+      "loss": 5.4038,
+      "step": 3432
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0334811210632324,
+      "learning_rate": 3.3343028892767115e-05,
+      "loss": 5.3715,
+      "step": 3436
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0406103134155273,
+      "learning_rate": 3.3323637773899555e-05,
+      "loss": 5.4744,
+      "step": 3440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9958846569061279,
+      "learning_rate": 3.3304246655031996e-05,
+      "loss": 5.4423,
+      "step": 3444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9611808657646179,
+      "learning_rate": 3.3284855536164436e-05,
+      "loss": 5.4374,
+      "step": 3448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0334917306900024,
+      "learning_rate": 3.3265464417296884e-05,
+      "loss": 5.3888,
+      "step": 3452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9894860982894897,
+      "learning_rate": 3.324607329842932e-05,
+      "loss": 5.5356,
+      "step": 3456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9536516070365906,
+      "learning_rate": 3.3226682179561765e-05,
+      "loss": 5.451,
+      "step": 3460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0449178218841553,
+      "learning_rate": 3.32072910606942e-05,
+      "loss": 5.4379,
+      "step": 3464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.958135724067688,
+      "learning_rate": 3.3187899941826646e-05,
+      "loss": 5.394,
+      "step": 3468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.049261212348938,
+      "learning_rate": 3.3168508822959086e-05,
+      "loss": 5.4298,
+      "step": 3472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.018259048461914,
+      "learning_rate": 3.314911770409153e-05,
+      "loss": 5.3572,
+      "step": 3476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.969048798084259,
+      "learning_rate": 3.312972658522397e-05,
+      "loss": 5.4854,
+      "step": 3480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0267409086227417,
+      "learning_rate": 3.3110335466356415e-05,
+      "loss": 5.4474,
+      "step": 3484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9933199286460876,
+      "learning_rate": 3.309094434748885e-05,
+      "loss": 5.4446,
+      "step": 3488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0801371335983276,
+      "learning_rate": 3.3071553228621296e-05,
+      "loss": 5.4434,
+      "step": 3492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.953080415725708,
+      "learning_rate": 3.3052162109753736e-05,
+      "loss": 5.4595,
+      "step": 3496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9589456915855408,
+      "learning_rate": 3.303277099088618e-05,
+      "loss": 5.433,
+      "step": 3500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9284895658493042,
+      "learning_rate": 3.301337987201862e-05,
+      "loss": 5.4761,
+      "step": 3504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9796357154846191,
+      "learning_rate": 3.299398875315106e-05,
+      "loss": 5.3998,
+      "step": 3508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9407968521118164,
+      "learning_rate": 3.29745976342835e-05,
+      "loss": 5.3964,
+      "step": 3512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0362569093704224,
+      "learning_rate": 3.2955206515415946e-05,
+      "loss": 5.5405,
+      "step": 3516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.024808645248413,
+      "learning_rate": 3.293581539654838e-05,
+      "loss": 5.3888,
+      "step": 3520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.185713291168213,
+      "learning_rate": 3.291642427768083e-05,
+      "loss": 5.3961,
+      "step": 3524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9882110357284546,
+      "learning_rate": 3.289703315881327e-05,
+      "loss": 5.4918,
+      "step": 3528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9622325897216797,
+      "learning_rate": 3.287764203994571e-05,
+      "loss": 5.409,
+      "step": 3532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9925034642219543,
+      "learning_rate": 3.285825092107815e-05,
+      "loss": 5.4252,
+      "step": 3536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0400789976119995,
+      "learning_rate": 3.283885980221059e-05,
+      "loss": 5.4415,
+      "step": 3540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0452476739883423,
+      "learning_rate": 3.281946868334303e-05,
+      "loss": 5.4926,
+      "step": 3544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.106879472732544,
+      "learning_rate": 3.280007756447547e-05,
+      "loss": 5.3975,
+      "step": 3548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9826605319976807,
+      "learning_rate": 3.278068644560791e-05,
+      "loss": 5.5062,
+      "step": 3552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0423948764801025,
+      "learning_rate": 3.276129532674035e-05,
+      "loss": 5.4536,
+      "step": 3556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0544408559799194,
+      "learning_rate": 3.27419042078728e-05,
+      "loss": 5.4063,
+      "step": 3560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9723476767539978,
+      "learning_rate": 3.272251308900524e-05,
+      "loss": 5.5508,
+      "step": 3564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0550904273986816,
+      "learning_rate": 3.270312197013768e-05,
+      "loss": 5.3544,
+      "step": 3568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0153075456619263,
+      "learning_rate": 3.268373085127012e-05,
+      "loss": 5.4197,
+      "step": 3572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9704105854034424,
+      "learning_rate": 3.266433973240256e-05,
+      "loss": 5.4494,
+      "step": 3576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.03487229347229,
+      "learning_rate": 3.2644948613535e-05,
+      "loss": 5.4456,
+      "step": 3580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9594029784202576,
+      "learning_rate": 3.262555749466744e-05,
+      "loss": 5.4308,
+      "step": 3584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9959999918937683,
+      "learning_rate": 3.260616637579988e-05,
+      "loss": 5.4625,
+      "step": 3588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0654551982879639,
+      "learning_rate": 3.258677525693233e-05,
+      "loss": 5.4254,
+      "step": 3592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9292247295379639,
+      "learning_rate": 3.256738413806476e-05,
+      "loss": 5.4191,
+      "step": 3596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9685704708099365,
+      "learning_rate": 3.254799301919721e-05,
+      "loss": 5.4356,
+      "step": 3600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0221501588821411,
+      "learning_rate": 3.252860190032965e-05,
+      "loss": 5.3416,
+      "step": 3604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9916023015975952,
+      "learning_rate": 3.250921078146209e-05,
+      "loss": 5.5014,
+      "step": 3608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9550511240959167,
+      "learning_rate": 3.248981966259453e-05,
+      "loss": 5.4269,
+      "step": 3612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0043843984603882,
+      "learning_rate": 3.247042854372698e-05,
+      "loss": 5.4404,
+      "step": 3616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9808074235916138,
+      "learning_rate": 3.245103742485941e-05,
+      "loss": 5.4492,
+      "step": 3620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9921736121177673,
+      "learning_rate": 3.243164630599186e-05,
+      "loss": 5.358,
+      "step": 3624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0501856803894043,
+      "learning_rate": 3.2412255187124294e-05,
+      "loss": 5.3826,
+      "step": 3628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0625114440917969,
+      "learning_rate": 3.239286406825674e-05,
+      "loss": 5.3954,
+      "step": 3632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9541780948638916,
+      "learning_rate": 3.237347294938918e-05,
+      "loss": 5.4487,
+      "step": 3636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0044699907302856,
+      "learning_rate": 3.235408183052162e-05,
+      "loss": 5.4098,
+      "step": 3640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9578327536582947,
+      "learning_rate": 3.233469071165406e-05,
+      "loss": 5.4887,
+      "step": 3644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9381611347198486,
+      "learning_rate": 3.231529959278651e-05,
+      "loss": 5.3075,
+      "step": 3648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9474562406539917,
+      "learning_rate": 3.2295908473918944e-05,
+      "loss": 5.6028,
+      "step": 3652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9133497476577759,
+      "learning_rate": 3.227651735505139e-05,
+      "loss": 5.2818,
+      "step": 3656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0017893314361572,
+      "learning_rate": 3.2257126236183825e-05,
+      "loss": 5.4843,
+      "step": 3660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0744807720184326,
+      "learning_rate": 3.223773511731627e-05,
+      "loss": 5.3414,
+      "step": 3664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9486330151557922,
+      "learning_rate": 3.221834399844871e-05,
+      "loss": 5.4186,
+      "step": 3668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9700675010681152,
+      "learning_rate": 3.2198952879581154e-05,
+      "loss": 5.3885,
+      "step": 3672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0845935344696045,
+      "learning_rate": 3.2179561760713594e-05,
+      "loss": 5.4627,
+      "step": 3676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0409663915634155,
+      "learning_rate": 3.2160170641846035e-05,
+      "loss": 5.4881,
+      "step": 3680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9628760814666748,
+      "learning_rate": 3.2140779522978475e-05,
+      "loss": 5.4565,
+      "step": 3684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.948780357837677,
+      "learning_rate": 3.2121388404110916e-05,
+      "loss": 5.4444,
+      "step": 3688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9025591015815735,
+      "learning_rate": 3.210199728524336e-05,
+      "loss": 5.329,
+      "step": 3692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.059144377708435,
+      "learning_rate": 3.2082606166375804e-05,
+      "loss": 5.4179,
+      "step": 3696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9723076820373535,
+      "learning_rate": 3.2063215047508244e-05,
+      "loss": 5.3901,
+      "step": 3700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0358588695526123,
+      "learning_rate": 3.2043823928640685e-05,
+      "loss": 5.4644,
+      "step": 3704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0152002573013306,
+      "learning_rate": 3.2024432809773125e-05,
+      "loss": 5.5068,
+      "step": 3708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0357673168182373,
+      "learning_rate": 3.2005041690905566e-05,
+      "loss": 5.5428,
+      "step": 3712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0267854928970337,
+      "learning_rate": 3.1985650572038006e-05,
+      "loss": 5.4718,
+      "step": 3716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9849284887313843,
+      "learning_rate": 3.196625945317045e-05,
+      "loss": 5.3864,
+      "step": 3720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0108904838562012,
+      "learning_rate": 3.1946868334302894e-05,
+      "loss": 5.4425,
+      "step": 3724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0127886533737183,
+      "learning_rate": 3.192747721543533e-05,
+      "loss": 5.3571,
+      "step": 3728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9837380051612854,
+      "learning_rate": 3.1908086096567775e-05,
+      "loss": 5.4701,
+      "step": 3732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0696730613708496,
+      "learning_rate": 3.1888694977700216e-05,
+      "loss": 5.5145,
+      "step": 3736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9534905552864075,
+      "learning_rate": 3.1869303858832656e-05,
+      "loss": 5.3795,
+      "step": 3740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.048189401626587,
+      "learning_rate": 3.18499127399651e-05,
+      "loss": 5.3971,
+      "step": 3744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0402449369430542,
+      "learning_rate": 3.183052162109754e-05,
+      "loss": 5.4111,
+      "step": 3748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.022661805152893,
+      "learning_rate": 3.181113050222998e-05,
+      "loss": 5.4513,
+      "step": 3752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9560799598693848,
+      "learning_rate": 3.1791739383362425e-05,
+      "loss": 5.3978,
+      "step": 3756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9077816605567932,
+      "learning_rate": 3.177234826449486e-05,
+      "loss": 5.3508,
+      "step": 3760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.029691219329834,
+      "learning_rate": 3.1752957145627306e-05,
+      "loss": 5.3808,
+      "step": 3764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9867807626724243,
+      "learning_rate": 3.173356602675975e-05,
+      "loss": 5.4433,
+      "step": 3768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0569761991500854,
+      "learning_rate": 3.171417490789219e-05,
+      "loss": 5.4479,
+      "step": 3772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0014142990112305,
+      "learning_rate": 3.169478378902463e-05,
+      "loss": 5.3415,
+      "step": 3776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9502798318862915,
+      "learning_rate": 3.167539267015707e-05,
+      "loss": 5.4549,
+      "step": 3780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9451482892036438,
+      "learning_rate": 3.165600155128951e-05,
+      "loss": 5.43,
+      "step": 3784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.002482295036316,
+      "learning_rate": 3.1636610432421956e-05,
+      "loss": 5.4482,
+      "step": 3788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0348658561706543,
+      "learning_rate": 3.161721931355439e-05,
+      "loss": 5.471,
+      "step": 3792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9318069815635681,
+      "learning_rate": 3.159782819468684e-05,
+      "loss": 5.3646,
+      "step": 3796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1141440868377686,
+      "learning_rate": 3.157843707581928e-05,
+      "loss": 5.4423,
+      "step": 3800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.066623568534851,
+      "learning_rate": 3.155904595695172e-05,
+      "loss": 5.3612,
+      "step": 3804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0264089107513428,
+      "learning_rate": 3.153965483808416e-05,
+      "loss": 5.383,
+      "step": 3808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0358729362487793,
+      "learning_rate": 3.15202637192166e-05,
+      "loss": 5.5152,
+      "step": 3812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9304607510566711,
+      "learning_rate": 3.150087260034904e-05,
+      "loss": 5.3978,
+      "step": 3816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0691871643066406,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 5.5089,
+      "step": 3820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0088552236557007,
+      "learning_rate": 3.146209036261392e-05,
+      "loss": 5.3989,
+      "step": 3824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9631534218788147,
+      "learning_rate": 3.144269924374637e-05,
+      "loss": 5.3578,
+      "step": 3828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.964043378829956,
+      "learning_rate": 3.142330812487881e-05,
+      "loss": 5.4551,
+      "step": 3832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0570262670516968,
+      "learning_rate": 3.140391700601125e-05,
+      "loss": 5.4109,
+      "step": 3836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.975766658782959,
+      "learning_rate": 3.138452588714369e-05,
+      "loss": 5.4265,
+      "step": 3840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9857097268104553,
+      "learning_rate": 3.136513476827613e-05,
+      "loss": 5.4179,
+      "step": 3844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9902443885803223,
+      "learning_rate": 3.134574364940857e-05,
+      "loss": 5.3386,
+      "step": 3848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.974892795085907,
+      "learning_rate": 3.132635253054101e-05,
+      "loss": 5.3901,
+      "step": 3852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9388407468795776,
+      "learning_rate": 3.130696141167346e-05,
+      "loss": 5.5098,
+      "step": 3856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0491032600402832,
+      "learning_rate": 3.128757029280589e-05,
+      "loss": 5.2734,
+      "step": 3860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9660788178443909,
+      "learning_rate": 3.126817917393834e-05,
+      "loss": 5.347,
+      "step": 3864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9738529324531555,
+      "learning_rate": 3.1248788055070774e-05,
+      "loss": 5.4238,
+      "step": 3868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9557338356971741,
+      "learning_rate": 3.122939693620322e-05,
+      "loss": 5.3881,
+      "step": 3872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1574413776397705,
+      "learning_rate": 3.121000581733566e-05,
+      "loss": 5.372,
+      "step": 3876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.046978235244751,
+      "learning_rate": 3.11906146984681e-05,
+      "loss": 5.3993,
+      "step": 3880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0011602640151978,
+      "learning_rate": 3.117122357960054e-05,
+      "loss": 5.3449,
+      "step": 3884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0066341161727905,
+      "learning_rate": 3.115183246073299e-05,
+      "loss": 5.5072,
+      "step": 3888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9333812594413757,
+      "learning_rate": 3.1132441341865424e-05,
+      "loss": 5.4166,
+      "step": 3892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9771477580070496,
+      "learning_rate": 3.111305022299787e-05,
+      "loss": 5.4957,
+      "step": 3896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9672766923904419,
+      "learning_rate": 3.1093659104130305e-05,
+      "loss": 5.5576,
+      "step": 3900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0771820545196533,
+      "learning_rate": 3.107426798526275e-05,
+      "loss": 5.432,
+      "step": 3904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9599045515060425,
+      "learning_rate": 3.105487686639519e-05,
+      "loss": 5.3826,
+      "step": 3908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0193896293640137,
+      "learning_rate": 3.1035485747527633e-05,
+      "loss": 5.3933,
+      "step": 3912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9940204620361328,
+      "learning_rate": 3.1016094628660074e-05,
+      "loss": 5.4366,
+      "step": 3916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9993325471878052,
+      "learning_rate": 3.099670350979252e-05,
+      "loss": 5.3764,
+      "step": 3920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0441081523895264,
+      "learning_rate": 3.0977312390924955e-05,
+      "loss": 5.4612,
+      "step": 3924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.971049964427948,
+      "learning_rate": 3.09579212720574e-05,
+      "loss": 5.365,
+      "step": 3928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9899376034736633,
+      "learning_rate": 3.093853015318984e-05,
+      "loss": 5.4119,
+      "step": 3932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.965894341468811,
+      "learning_rate": 3.0919139034322283e-05,
+      "loss": 5.5259,
+      "step": 3936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.007863998413086,
+      "learning_rate": 3.0899747915454724e-05,
+      "loss": 5.4325,
+      "step": 3940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.032435417175293,
+      "learning_rate": 3.0880356796587164e-05,
+      "loss": 5.5254,
+      "step": 3944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9465643167495728,
+      "learning_rate": 3.0860965677719605e-05,
+      "loss": 5.443,
+      "step": 3948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.112241268157959,
+      "learning_rate": 3.0841574558852046e-05,
+      "loss": 5.4087,
+      "step": 3952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9741985201835632,
+      "learning_rate": 3.0822183439984486e-05,
+      "loss": 5.4688,
+      "step": 3956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0749200582504272,
+      "learning_rate": 3.0802792321116933e-05,
+      "loss": 5.3697,
+      "step": 3960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0521398782730103,
+      "learning_rate": 3.0783401202249374e-05,
+      "loss": 5.4226,
+      "step": 3964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.995364248752594,
+      "learning_rate": 3.0764010083381814e-05,
+      "loss": 5.4333,
+      "step": 3968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.023390769958496,
+      "learning_rate": 3.0744618964514255e-05,
+      "loss": 5.4662,
+      "step": 3972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9972986578941345,
+      "learning_rate": 3.0725227845646696e-05,
+      "loss": 5.4432,
+      "step": 3976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0111867189407349,
+      "learning_rate": 3.0705836726779136e-05,
+      "loss": 5.4875,
+      "step": 3980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.930380642414093,
+      "learning_rate": 3.068644560791158e-05,
+      "loss": 5.335,
+      "step": 3984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0326186418533325,
+      "learning_rate": 3.066705448904402e-05,
+      "loss": 5.3984,
+      "step": 3988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9522334933280945,
+      "learning_rate": 3.064766337017646e-05,
+      "loss": 5.4895,
+      "step": 3992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.048785924911499,
+      "learning_rate": 3.0628272251308905e-05,
+      "loss": 5.3766,
+      "step": 3996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0069239139556885,
+      "learning_rate": 3.060888113244134e-05,
+      "loss": 5.3243,
+      "step": 4000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8954800963401794,
+      "learning_rate": 3.0589490013573786e-05,
+      "loss": 5.3281,
+      "step": 4004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0596482753753662,
+      "learning_rate": 3.057009889470623e-05,
+      "loss": 5.5029,
+      "step": 4008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9929758310317993,
+      "learning_rate": 3.055070777583867e-05,
+      "loss": 5.4649,
+      "step": 4012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9867807626724243,
+      "learning_rate": 3.053131665697111e-05,
+      "loss": 5.3755,
+      "step": 4016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9822360873222351,
+      "learning_rate": 3.0511925538103548e-05,
+      "loss": 5.4723,
+      "step": 4020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9592394232749939,
+      "learning_rate": 3.0492534419235992e-05,
+      "loss": 5.4694,
+      "step": 4024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.993042528629303,
+      "learning_rate": 3.0473143300368433e-05,
+      "loss": 5.4091,
+      "step": 4028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9652045965194702,
+      "learning_rate": 3.0453752181500873e-05,
+      "loss": 5.4072,
+      "step": 4032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9788159728050232,
+      "learning_rate": 3.0434361062633314e-05,
+      "loss": 5.4358,
+      "step": 4036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9906930923461914,
+      "learning_rate": 3.0414969943765758e-05,
+      "loss": 5.4567,
+      "step": 4040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0816450119018555,
+      "learning_rate": 3.0395578824898195e-05,
+      "loss": 5.3435,
+      "step": 4044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0458557605743408,
+      "learning_rate": 3.037618770603064e-05,
+      "loss": 5.5151,
+      "step": 4048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9744831919670105,
+      "learning_rate": 3.0356796587163083e-05,
+      "loss": 5.486,
+      "step": 4052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9657852649688721,
+      "learning_rate": 3.033740546829552e-05,
+      "loss": 5.441,
+      "step": 4056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9370314478874207,
+      "learning_rate": 3.0318014349427964e-05,
+      "loss": 5.4117,
+      "step": 4060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9673814177513123,
+      "learning_rate": 3.02986232305604e-05,
+      "loss": 5.5008,
+      "step": 4064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.004071593284607,
+      "learning_rate": 3.0279232111692845e-05,
+      "loss": 5.4525,
+      "step": 4068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0029621124267578,
+      "learning_rate": 3.025984099282529e-05,
+      "loss": 5.4788,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0012478828430176,
+      "learning_rate": 3.0240449873957726e-05,
+      "loss": 5.443,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9706107974052429,
+      "learning_rate": 3.022105875509017e-05,
+      "loss": 5.35,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9983603358268738,
+      "learning_rate": 3.0201667636222614e-05,
+      "loss": 5.4215,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0420852899551392,
+      "learning_rate": 3.018227651735505e-05,
+      "loss": 5.486,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9456035494804382,
+      "learning_rate": 3.0162885398487495e-05,
+      "loss": 5.4257,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9700554609298706,
+      "learning_rate": 3.014349427961994e-05,
+      "loss": 5.4415,
+      "step": 4096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0411101579666138,
+      "learning_rate": 3.0124103160752376e-05,
+      "loss": 5.4313,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0131868124008179,
+      "learning_rate": 3.010471204188482e-05,
+      "loss": 5.371,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9504354000091553,
+      "learning_rate": 3.0085320923017257e-05,
+      "loss": 5.4643,
+      "step": 4108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9660660624504089,
+      "learning_rate": 3.00659298041497e-05,
+      "loss": 5.4337,
+      "step": 4112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0248061418533325,
+      "learning_rate": 3.0046538685282145e-05,
+      "loss": 5.4113,
+      "step": 4116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0795485973358154,
+      "learning_rate": 3.0027147566414582e-05,
+      "loss": 5.3737,
+      "step": 4120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0381568670272827,
+      "learning_rate": 3.0007756447547026e-05,
+      "loss": 5.4409,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9909124970436096,
+      "learning_rate": 2.998836532867947e-05,
+      "loss": 5.358,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0048110485076904,
+      "learning_rate": 2.9968974209811907e-05,
+      "loss": 5.4109,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9982589483261108,
+      "learning_rate": 2.994958309094435e-05,
+      "loss": 5.4673,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.010875940322876,
+      "learning_rate": 2.9930191972076788e-05,
+      "loss": 5.3333,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0325456857681274,
+      "learning_rate": 2.9910800853209232e-05,
+      "loss": 5.3687,
+      "step": 4144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.057373046875,
+      "learning_rate": 2.9891409734341676e-05,
+      "loss": 5.4218,
+      "step": 4148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.975649356842041,
+      "learning_rate": 2.9872018615474113e-05,
+      "loss": 5.3694,
+      "step": 4152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9869158864021301,
+      "learning_rate": 2.9852627496606557e-05,
+      "loss": 5.3961,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9405871033668518,
+      "learning_rate": 2.9833236377738998e-05,
+      "loss": 5.3882,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9522705078125,
+      "learning_rate": 2.9813845258871438e-05,
+      "loss": 5.4143,
+      "step": 4164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9004665613174438,
+      "learning_rate": 2.979445414000388e-05,
+      "loss": 5.4774,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.01994788646698,
+      "learning_rate": 2.9775063021136323e-05,
+      "loss": 5.4938,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0324288606643677,
+      "learning_rate": 2.975567190226876e-05,
+      "loss": 5.351,
+      "step": 4176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0081384181976318,
+      "learning_rate": 2.9736280783401204e-05,
+      "loss": 5.4233,
+      "step": 4180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9753928780555725,
+      "learning_rate": 2.9716889664533644e-05,
+      "loss": 5.3791,
+      "step": 4184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0055930614471436,
+      "learning_rate": 2.9697498545666085e-05,
+      "loss": 5.4083,
+      "step": 4188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0010572671890259,
+      "learning_rate": 2.967810742679853e-05,
+      "loss": 5.3065,
+      "step": 4192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0279070138931274,
+      "learning_rate": 2.9658716307930966e-05,
+      "loss": 5.359,
+      "step": 4196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9655927419662476,
+      "learning_rate": 2.963932518906341e-05,
+      "loss": 5.4802,
+      "step": 4200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9861381649971008,
+      "learning_rate": 2.9619934070195854e-05,
+      "loss": 5.3633,
+      "step": 4204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0157438516616821,
+      "learning_rate": 2.960054295132829e-05,
+      "loss": 5.3607,
+      "step": 4208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0280277729034424,
+      "learning_rate": 2.9581151832460735e-05,
+      "loss": 5.3693,
+      "step": 4212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.045616865158081,
+      "learning_rate": 2.956176071359318e-05,
+      "loss": 5.4264,
+      "step": 4216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.950824499130249,
+      "learning_rate": 2.9542369594725616e-05,
+      "loss": 5.3266,
+      "step": 4220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9672490358352661,
+      "learning_rate": 2.952297847585806e-05,
+      "loss": 5.4075,
+      "step": 4224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.964094877243042,
+      "learning_rate": 2.9503587356990497e-05,
+      "loss": 5.4298,
+      "step": 4228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9985701441764832,
+      "learning_rate": 2.948419623812294e-05,
+      "loss": 5.4677,
+      "step": 4232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0406873226165771,
+      "learning_rate": 2.9464805119255385e-05,
+      "loss": 5.3258,
+      "step": 4236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0154664516448975,
+      "learning_rate": 2.9445414000387822e-05,
+      "loss": 5.4148,
+      "step": 4240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9826991558074951,
+      "learning_rate": 2.9426022881520266e-05,
+      "loss": 5.3293,
+      "step": 4244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0213288068771362,
+      "learning_rate": 2.940663176265271e-05,
+      "loss": 5.3245,
+      "step": 4248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0005340576171875,
+      "learning_rate": 2.9387240643785147e-05,
+      "loss": 5.4466,
+      "step": 4252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0072083473205566,
+      "learning_rate": 2.936784952491759e-05,
+      "loss": 5.3103,
+      "step": 4256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9992334246635437,
+      "learning_rate": 2.9348458406050028e-05,
+      "loss": 5.4595,
+      "step": 4260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.026963233947754,
+      "learning_rate": 2.9329067287182472e-05,
+      "loss": 5.4082,
+      "step": 4264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0018388032913208,
+      "learning_rate": 2.9309676168314916e-05,
+      "loss": 5.3292,
+      "step": 4268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.106248140335083,
+      "learning_rate": 2.9290285049447353e-05,
+      "loss": 5.3176,
+      "step": 4272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0284066200256348,
+      "learning_rate": 2.9270893930579797e-05,
+      "loss": 5.3839,
+      "step": 4276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0233203172683716,
+      "learning_rate": 2.925150281171224e-05,
+      "loss": 5.4327,
+      "step": 4280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0423204898834229,
+      "learning_rate": 2.9232111692844678e-05,
+      "loss": 5.4134,
+      "step": 4284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0750269889831543,
+      "learning_rate": 2.9212720573977122e-05,
+      "loss": 5.3983,
+      "step": 4288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.035136342048645,
+      "learning_rate": 2.9193329455109562e-05,
+      "loss": 5.3828,
+      "step": 4292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9154065847396851,
+      "learning_rate": 2.9173938336242003e-05,
+      "loss": 5.3261,
+      "step": 4296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0032275915145874,
+      "learning_rate": 2.9154547217374443e-05,
+      "loss": 5.4712,
+      "step": 4300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0405406951904297,
+      "learning_rate": 2.9135156098506884e-05,
+      "loss": 5.3499,
+      "step": 4304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1554255485534668,
+      "learning_rate": 2.9115764979639324e-05,
+      "loss": 5.4097,
+      "step": 4308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.054694414138794,
+      "learning_rate": 2.909637386077177e-05,
+      "loss": 5.3183,
+      "step": 4312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9714403748512268,
+      "learning_rate": 2.907698274190421e-05,
+      "loss": 5.5304,
+      "step": 4316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0622743368148804,
+      "learning_rate": 2.905759162303665e-05,
+      "loss": 5.321,
+      "step": 4320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0464075803756714,
+      "learning_rate": 2.9038200504169093e-05,
+      "loss": 5.3688,
+      "step": 4324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9132739901542664,
+      "learning_rate": 2.901880938530153e-05,
+      "loss": 5.3441,
+      "step": 4328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.013675332069397,
+      "learning_rate": 2.8999418266433974e-05,
+      "loss": 5.4333,
+      "step": 4332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0379986763000488,
+      "learning_rate": 2.898002714756642e-05,
+      "loss": 5.5111,
+      "step": 4336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0619996786117554,
+      "learning_rate": 2.8960636028698856e-05,
+      "loss": 5.3242,
+      "step": 4340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0503803491592407,
+      "learning_rate": 2.89412449098313e-05,
+      "loss": 5.4761,
+      "step": 4344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.006859540939331,
+      "learning_rate": 2.8921853790963737e-05,
+      "loss": 5.4661,
+      "step": 4348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9837526082992554,
+      "learning_rate": 2.890246267209618e-05,
+      "loss": 5.4439,
+      "step": 4352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0037223100662231,
+      "learning_rate": 2.8883071553228624e-05,
+      "loss": 5.3982,
+      "step": 4356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9351251125335693,
+      "learning_rate": 2.886368043436106e-05,
+      "loss": 5.4419,
+      "step": 4360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9317170977592468,
+      "learning_rate": 2.8844289315493506e-05,
+      "loss": 5.4067,
+      "step": 4364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9478697776794434,
+      "learning_rate": 2.882489819662595e-05,
+      "loss": 5.4012,
+      "step": 4368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9834426641464233,
+      "learning_rate": 2.8805507077758387e-05,
+      "loss": 5.4496,
+      "step": 4372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0380594730377197,
+      "learning_rate": 2.878611595889083e-05,
+      "loss": 5.4042,
+      "step": 4376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1123548746109009,
+      "learning_rate": 2.8766724840023268e-05,
+      "loss": 5.4814,
+      "step": 4380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9241830706596375,
+      "learning_rate": 2.874733372115571e-05,
+      "loss": 5.4302,
+      "step": 4384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9564453363418579,
+      "learning_rate": 2.8727942602288156e-05,
+      "loss": 5.3426,
+      "step": 4388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0000247955322266,
+      "learning_rate": 2.8708551483420593e-05,
+      "loss": 5.3799,
+      "step": 4392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9557441473007202,
+      "learning_rate": 2.8689160364553037e-05,
+      "loss": 5.4233,
+      "step": 4396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.920683741569519,
+      "learning_rate": 2.866976924568548e-05,
+      "loss": 5.351,
+      "step": 4400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9885355234146118,
+      "learning_rate": 2.8650378126817918e-05,
+      "loss": 5.4735,
+      "step": 4404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.02256441116333,
+      "learning_rate": 2.863098700795036e-05,
+      "loss": 5.4872,
+      "step": 4408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0346393585205078,
+      "learning_rate": 2.8611595889082806e-05,
+      "loss": 5.4895,
+      "step": 4412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0645418167114258,
+      "learning_rate": 2.8592204770215243e-05,
+      "loss": 5.3678,
+      "step": 4416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9786351323127747,
+      "learning_rate": 2.8572813651347687e-05,
+      "loss": 5.394,
+      "step": 4420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0003682374954224,
+      "learning_rate": 2.8553422532480124e-05,
+      "loss": 5.3394,
+      "step": 4424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0191590785980225,
+      "learning_rate": 2.8534031413612568e-05,
+      "loss": 5.3822,
+      "step": 4428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9041575193405151,
+      "learning_rate": 2.8514640294745008e-05,
+      "loss": 5.3762,
+      "step": 4432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1107348203659058,
+      "learning_rate": 2.849524917587745e-05,
+      "loss": 5.355,
+      "step": 4436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9530418515205383,
+      "learning_rate": 2.847585805700989e-05,
+      "loss": 5.3681,
+      "step": 4440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.988714337348938,
+      "learning_rate": 2.8456466938142333e-05,
+      "loss": 5.3924,
+      "step": 4444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0610371828079224,
+      "learning_rate": 2.8437075819274774e-05,
+      "loss": 5.3354,
+      "step": 4448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.005706548690796,
+      "learning_rate": 2.8417684700407214e-05,
+      "loss": 5.3619,
+      "step": 4452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.053621530532837,
+      "learning_rate": 2.8398293581539658e-05,
+      "loss": 5.344,
+      "step": 4456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0087207555770874,
+      "learning_rate": 2.8378902462672095e-05,
+      "loss": 5.3222,
+      "step": 4460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0023506879806519,
+      "learning_rate": 2.835951134380454e-05,
+      "loss": 5.4015,
+      "step": 4464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9826129078865051,
+      "learning_rate": 2.8340120224936976e-05,
+      "loss": 5.4818,
+      "step": 4468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9970731139183044,
+      "learning_rate": 2.832072910606942e-05,
+      "loss": 5.4439,
+      "step": 4472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0162273645401,
+      "learning_rate": 2.8301337987201864e-05,
+      "loss": 5.4481,
+      "step": 4476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.065722942352295,
+      "learning_rate": 2.82819468683343e-05,
+      "loss": 5.4393,
+      "step": 4480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0227274894714355,
+      "learning_rate": 2.8262555749466745e-05,
+      "loss": 5.4049,
+      "step": 4484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.058347463607788,
+      "learning_rate": 2.824316463059919e-05,
+      "loss": 5.4221,
+      "step": 4488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.981675922870636,
+      "learning_rate": 2.8223773511731626e-05,
+      "loss": 5.4298,
+      "step": 4492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0078340768814087,
+      "learning_rate": 2.820438239286407e-05,
+      "loss": 5.4223,
+      "step": 4496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0829211473464966,
+      "learning_rate": 2.8184991273996508e-05,
+      "loss": 5.4577,
+      "step": 4500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9840219020843506,
+      "learning_rate": 2.816560015512895e-05,
+      "loss": 5.3451,
+      "step": 4504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.059767484664917,
+      "learning_rate": 2.8146209036261395e-05,
+      "loss": 5.4052,
+      "step": 4508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1250829696655273,
+      "learning_rate": 2.8126817917393833e-05,
+      "loss": 5.3476,
+      "step": 4512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9797293543815613,
+      "learning_rate": 2.8107426798526276e-05,
+      "loss": 5.3143,
+      "step": 4516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9784405827522278,
+      "learning_rate": 2.808803567965872e-05,
+      "loss": 5.4072,
+      "step": 4520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0686568021774292,
+      "learning_rate": 2.8068644560791158e-05,
+      "loss": 5.4029,
+      "step": 4524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.958621084690094,
+      "learning_rate": 2.80492534419236e-05,
+      "loss": 5.3838,
+      "step": 4528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.969975471496582,
+      "learning_rate": 2.8029862323056045e-05,
+      "loss": 5.4323,
+      "step": 4532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0803236961364746,
+      "learning_rate": 2.8010471204188483e-05,
+      "loss": 5.4438,
+      "step": 4536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0710035562515259,
+      "learning_rate": 2.7991080085320926e-05,
+      "loss": 5.4503,
+      "step": 4540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9775360226631165,
+      "learning_rate": 2.7971688966453364e-05,
+      "loss": 5.3578,
+      "step": 4544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9491469264030457,
+      "learning_rate": 2.7952297847585808e-05,
+      "loss": 5.3979,
+      "step": 4548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0196774005889893,
+      "learning_rate": 2.793290672871825e-05,
+      "loss": 5.4313,
+      "step": 4552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0691953897476196,
+      "learning_rate": 2.791351560985069e-05,
+      "loss": 5.5029,
+      "step": 4556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0333261489868164,
+      "learning_rate": 2.7894124490983133e-05,
+      "loss": 5.4323,
+      "step": 4560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0331099033355713,
+      "learning_rate": 2.7874733372115573e-05,
+      "loss": 5.4105,
+      "step": 4564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0562598705291748,
+      "learning_rate": 2.7855342253248014e-05,
+      "loss": 5.4442,
+      "step": 4568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0251268148422241,
+      "learning_rate": 2.7835951134380454e-05,
+      "loss": 5.4121,
+      "step": 4572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9381951689720154,
+      "learning_rate": 2.7816560015512898e-05,
+      "loss": 5.303,
+      "step": 4576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0383044481277466,
+      "learning_rate": 2.779716889664534e-05,
+      "loss": 5.3638,
+      "step": 4580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0460588932037354,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 5.3688,
+      "step": 4584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9689496159553528,
+      "learning_rate": 2.775838665891022e-05,
+      "loss": 5.499,
+      "step": 4588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9622393846511841,
+      "learning_rate": 2.773899554004266e-05,
+      "loss": 5.3238,
+      "step": 4592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9752629399299622,
+      "learning_rate": 2.7719604421175104e-05,
+      "loss": 5.3294,
+      "step": 4596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9437540173530579,
+      "learning_rate": 2.770021330230754e-05,
+      "loss": 5.3397,
+      "step": 4600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.041926383972168,
+      "learning_rate": 2.7680822183439985e-05,
+      "loss": 5.4434,
+      "step": 4604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.045983076095581,
+      "learning_rate": 2.766143106457243e-05,
+      "loss": 5.3467,
+      "step": 4608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9513012766838074,
+      "learning_rate": 2.7642039945704866e-05,
+      "loss": 5.4401,
+      "step": 4612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0047541856765747,
+      "learning_rate": 2.762264882683731e-05,
+      "loss": 5.3485,
+      "step": 4616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0038918256759644,
+      "learning_rate": 2.7603257707969747e-05,
+      "loss": 5.4189,
+      "step": 4620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0029683113098145,
+      "learning_rate": 2.758386658910219e-05,
+      "loss": 5.3773,
+      "step": 4624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9756178259849548,
+      "learning_rate": 2.7564475470234635e-05,
+      "loss": 5.4745,
+      "step": 4628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.028535008430481,
+      "learning_rate": 2.7545084351367072e-05,
+      "loss": 5.3767,
+      "step": 4632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0276795625686646,
+      "learning_rate": 2.7525693232499516e-05,
+      "loss": 5.4127,
+      "step": 4636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9759376645088196,
+      "learning_rate": 2.750630211363196e-05,
+      "loss": 5.4136,
+      "step": 4640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9665167331695557,
+      "learning_rate": 2.7486910994764397e-05,
+      "loss": 5.4307,
+      "step": 4644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0241793394088745,
+      "learning_rate": 2.746751987589684e-05,
+      "loss": 5.3753,
+      "step": 4648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.97547847032547,
+      "learning_rate": 2.7448128757029285e-05,
+      "loss": 5.4257,
+      "step": 4652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9529567360877991,
+      "learning_rate": 2.7428737638161722e-05,
+      "loss": 5.4561,
+      "step": 4656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9759612679481506,
+      "learning_rate": 2.7409346519294166e-05,
+      "loss": 5.3565,
+      "step": 4660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0170910358428955,
+      "learning_rate": 2.7389955400426603e-05,
+      "loss": 5.3003,
+      "step": 4664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0059340000152588,
+      "learning_rate": 2.7370564281559047e-05,
+      "loss": 5.4549,
+      "step": 4668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9615738987922668,
+      "learning_rate": 2.735117316269149e-05,
+      "loss": 5.3853,
+      "step": 4672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.07008957862854,
+      "learning_rate": 2.733178204382393e-05,
+      "loss": 5.4426,
+      "step": 4676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0182669162750244,
+      "learning_rate": 2.7312390924956372e-05,
+      "loss": 5.3921,
+      "step": 4680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0533527135849,
+      "learning_rate": 2.7292999806088816e-05,
+      "loss": 5.4001,
+      "step": 4684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0803202390670776,
+      "learning_rate": 2.7273608687221253e-05,
+      "loss": 5.4789,
+      "step": 4688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0540350675582886,
+      "learning_rate": 2.7254217568353697e-05,
+      "loss": 5.4387,
+      "step": 4692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9857541918754578,
+      "learning_rate": 2.7234826449486138e-05,
+      "loss": 5.4231,
+      "step": 4696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9859492778778076,
+      "learning_rate": 2.721543533061858e-05,
+      "loss": 5.4614,
+      "step": 4700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0070449113845825,
+      "learning_rate": 2.719604421175102e-05,
+      "loss": 5.4235,
+      "step": 4704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0052629709243774,
+      "learning_rate": 2.717665309288346e-05,
+      "loss": 5.4255,
+      "step": 4708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0358524322509766,
+      "learning_rate": 2.7157261974015903e-05,
+      "loss": 5.3811,
+      "step": 4712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9567641615867615,
+      "learning_rate": 2.7137870855148344e-05,
+      "loss": 5.3977,
+      "step": 4716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0303080081939697,
+      "learning_rate": 2.7118479736280784e-05,
+      "loss": 5.5686,
+      "step": 4720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9873762726783752,
+      "learning_rate": 2.7099088617413225e-05,
+      "loss": 5.3691,
+      "step": 4724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9921165704727173,
+      "learning_rate": 2.707969749854567e-05,
+      "loss": 5.4681,
+      "step": 4728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9569845795631409,
+      "learning_rate": 2.7060306379678106e-05,
+      "loss": 5.2843,
+      "step": 4732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0275355577468872,
+      "learning_rate": 2.704091526081055e-05,
+      "loss": 5.3745,
+      "step": 4736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.034792184829712,
+      "learning_rate": 2.7021524141942987e-05,
+      "loss": 5.3919,
+      "step": 4740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.05870521068573,
+      "learning_rate": 2.700213302307543e-05,
+      "loss": 5.4458,
+      "step": 4744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1329647302627563,
+      "learning_rate": 2.6982741904207875e-05,
+      "loss": 5.4046,
+      "step": 4748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9993448853492737,
+      "learning_rate": 2.6963350785340312e-05,
+      "loss": 5.3655,
+      "step": 4752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0315309762954712,
+      "learning_rate": 2.6943959666472756e-05,
+      "loss": 5.3993,
+      "step": 4756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0383795499801636,
+      "learning_rate": 2.69245685476052e-05,
+      "loss": 5.3556,
+      "step": 4760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9584831595420837,
+      "learning_rate": 2.6905177428737637e-05,
+      "loss": 5.4116,
+      "step": 4764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0869189500808716,
+      "learning_rate": 2.688578630987008e-05,
+      "loss": 5.3348,
+      "step": 4768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1664848327636719,
+      "learning_rate": 2.6866395191002525e-05,
+      "loss": 5.3155,
+      "step": 4772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0994501113891602,
+      "learning_rate": 2.6847004072134962e-05,
+      "loss": 5.3711,
+      "step": 4776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9989491105079651,
+      "learning_rate": 2.6827612953267406e-05,
+      "loss": 5.3603,
+      "step": 4780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1618521213531494,
+      "learning_rate": 2.6808221834399843e-05,
+      "loss": 5.4217,
+      "step": 4784
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9329230785369873,
+      "learning_rate": 2.6788830715532287e-05,
+      "loss": 5.4326,
+      "step": 4788
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9027351140975952,
+      "learning_rate": 2.676943959666473e-05,
+      "loss": 5.2881,
+      "step": 4792
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9692825078964233,
+      "learning_rate": 2.6750048477797168e-05,
+      "loss": 5.3769,
+      "step": 4796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9520678520202637,
+      "learning_rate": 2.6730657358929612e-05,
+      "loss": 5.377,
+      "step": 4800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9845559000968933,
+      "learning_rate": 2.6711266240062056e-05,
+      "loss": 5.3848,
+      "step": 4804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0046534538269043,
+      "learning_rate": 2.6691875121194493e-05,
+      "loss": 5.3973,
+      "step": 4808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0389518737792969,
+      "learning_rate": 2.6672484002326937e-05,
+      "loss": 5.4557,
+      "step": 4812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0322433710098267,
+      "learning_rate": 2.665309288345938e-05,
+      "loss": 5.3843,
+      "step": 4816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.075420618057251,
+      "learning_rate": 2.6633701764591818e-05,
+      "loss": 5.3968,
+      "step": 4820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.004739761352539,
+      "learning_rate": 2.6614310645724262e-05,
+      "loss": 5.4449,
+      "step": 4824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9824436902999878,
+      "learning_rate": 2.65949195268567e-05,
+      "loss": 5.348,
+      "step": 4828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0312827825546265,
+      "learning_rate": 2.6575528407989143e-05,
+      "loss": 5.438,
+      "step": 4832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9586296677589417,
+      "learning_rate": 2.6556137289121584e-05,
+      "loss": 5.4007,
+      "step": 4836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0318596363067627,
+      "learning_rate": 2.6536746170254024e-05,
+      "loss": 5.3997,
+      "step": 4840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9269830584526062,
+      "learning_rate": 2.6517355051386468e-05,
+      "loss": 5.3597,
+      "step": 4844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0358200073242188,
+      "learning_rate": 2.649796393251891e-05,
+      "loss": 5.3729,
+      "step": 4848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9825155735015869,
+      "learning_rate": 2.647857281365135e-05,
+      "loss": 5.3678,
+      "step": 4852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0111230611801147,
+      "learning_rate": 2.645918169478379e-05,
+      "loss": 5.4858,
+      "step": 4856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.947890043258667,
+      "learning_rate": 2.643979057591623e-05,
+      "loss": 5.3939,
+      "step": 4860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9673914909362793,
+      "learning_rate": 2.642039945704867e-05,
+      "loss": 5.3704,
+      "step": 4864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0731589794158936,
+      "learning_rate": 2.6401008338181115e-05,
+      "loss": 5.4116,
+      "step": 4868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0044986009597778,
+      "learning_rate": 2.6381617219313552e-05,
+      "loss": 5.4318,
+      "step": 4872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9914534687995911,
+      "learning_rate": 2.6362226100445996e-05,
+      "loss": 5.3216,
+      "step": 4876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9887456297874451,
+      "learning_rate": 2.634283498157844e-05,
+      "loss": 5.4072,
+      "step": 4880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0049148797988892,
+      "learning_rate": 2.6323443862710877e-05,
+      "loss": 5.4255,
+      "step": 4884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0179352760314941,
+      "learning_rate": 2.630405274384332e-05,
+      "loss": 5.3108,
+      "step": 4888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9834827184677124,
+      "learning_rate": 2.6284661624975765e-05,
+      "loss": 5.3262,
+      "step": 4892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.032747507095337,
+      "learning_rate": 2.6265270506108202e-05,
+      "loss": 5.3302,
+      "step": 4896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0606615543365479,
+      "learning_rate": 2.6245879387240646e-05,
+      "loss": 5.3351,
+      "step": 4900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0665369033813477,
+      "learning_rate": 2.6226488268373083e-05,
+      "loss": 5.3851,
+      "step": 4904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0293782949447632,
+      "learning_rate": 2.6207097149505527e-05,
+      "loss": 5.3772,
+      "step": 4908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0471354722976685,
+      "learning_rate": 2.618770603063797e-05,
+      "loss": 5.4218,
+      "step": 4912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9955822229385376,
+      "learning_rate": 2.6168314911770408e-05,
+      "loss": 5.3427,
+      "step": 4916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9505192041397095,
+      "learning_rate": 2.6148923792902852e-05,
+      "loss": 5.4442,
+      "step": 4920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0545083284378052,
+      "learning_rate": 2.6129532674035296e-05,
+      "loss": 5.3821,
+      "step": 4924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.077630639076233,
+      "learning_rate": 2.6110141555167733e-05,
+      "loss": 5.4215,
+      "step": 4928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1127928495407104,
+      "learning_rate": 2.6090750436300177e-05,
+      "loss": 5.4017,
+      "step": 4932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0543692111968994,
+      "learning_rate": 2.607135931743262e-05,
+      "loss": 5.3918,
+      "step": 4936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0516538619995117,
+      "learning_rate": 2.6051968198565058e-05,
+      "loss": 5.5141,
+      "step": 4940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.077325463294983,
+      "learning_rate": 2.6032577079697502e-05,
+      "loss": 5.4855,
+      "step": 4944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.034180998802185,
+      "learning_rate": 2.601318596082994e-05,
+      "loss": 5.4097,
+      "step": 4948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0808318853378296,
+      "learning_rate": 2.5993794841962383e-05,
+      "loss": 5.3711,
+      "step": 4952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0749197006225586,
+      "learning_rate": 2.5974403723094827e-05,
+      "loss": 5.3917,
+      "step": 4956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1579383611679077,
+      "learning_rate": 2.5955012604227264e-05,
+      "loss": 5.4251,
+      "step": 4960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0081652402877808,
+      "learning_rate": 2.5935621485359708e-05,
+      "loss": 5.3684,
+      "step": 4964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9971766471862793,
+      "learning_rate": 2.591623036649215e-05,
+      "loss": 5.5078,
+      "step": 4968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9572992920875549,
+      "learning_rate": 2.589683924762459e-05,
+      "loss": 5.4004,
+      "step": 4972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0008381605148315,
+      "learning_rate": 2.5877448128757033e-05,
+      "loss": 5.4194,
+      "step": 4976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0567643642425537,
+      "learning_rate": 2.585805700988947e-05,
+      "loss": 5.4014,
+      "step": 4980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.084086298942566,
+      "learning_rate": 2.5838665891021914e-05,
+      "loss": 5.4578,
+      "step": 4984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.062583088874817,
+      "learning_rate": 2.5819274772154355e-05,
+      "loss": 5.4051,
+      "step": 4988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.018263578414917,
+      "learning_rate": 2.5799883653286795e-05,
+      "loss": 5.4411,
+      "step": 4992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0082392692565918,
+      "learning_rate": 2.5780492534419236e-05,
+      "loss": 5.329,
+      "step": 4996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0163345336914062,
+      "learning_rate": 2.576110141555168e-05,
+      "loss": 5.3487,
+      "step": 5000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9743790030479431,
+      "learning_rate": 2.5741710296684117e-05,
+      "loss": 5.4595,
+      "step": 5004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0049850940704346,
+      "learning_rate": 2.572231917781656e-05,
+      "loss": 5.3147,
+      "step": 5008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9798968434333801,
+      "learning_rate": 2.5702928058949005e-05,
+      "loss": 5.4953,
+      "step": 5012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1075799465179443,
+      "learning_rate": 2.5683536940081442e-05,
+      "loss": 5.4284,
+      "step": 5016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9216572046279907,
+      "learning_rate": 2.5664145821213886e-05,
+      "loss": 5.4338,
+      "step": 5020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0511445999145508,
+      "learning_rate": 2.5644754702346323e-05,
+      "loss": 5.35,
+      "step": 5024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0618550777435303,
+      "learning_rate": 2.5625363583478767e-05,
+      "loss": 5.3539,
+      "step": 5028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0445276498794556,
+      "learning_rate": 2.560597246461121e-05,
+      "loss": 5.3797,
+      "step": 5032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0461180210113525,
+      "learning_rate": 2.5586581345743648e-05,
+      "loss": 5.4262,
+      "step": 5036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1503204107284546,
+      "learning_rate": 2.5567190226876092e-05,
+      "loss": 5.3629,
+      "step": 5040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9792975187301636,
+      "learning_rate": 2.5547799108008536e-05,
+      "loss": 5.3245,
+      "step": 5044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0088895559310913,
+      "learning_rate": 2.5528407989140973e-05,
+      "loss": 5.3447,
+      "step": 5048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9955673217773438,
+      "learning_rate": 2.5509016870273417e-05,
+      "loss": 5.2618,
+      "step": 5052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9601762294769287,
+      "learning_rate": 2.548962575140586e-05,
+      "loss": 5.3711,
+      "step": 5056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9723737835884094,
+      "learning_rate": 2.5470234632538298e-05,
+      "loss": 5.4062,
+      "step": 5060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.961095929145813,
+      "learning_rate": 2.5450843513670742e-05,
+      "loss": 5.4148,
+      "step": 5064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.071572184562683,
+      "learning_rate": 2.543145239480318e-05,
+      "loss": 5.3569,
+      "step": 5068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0116100311279297,
+      "learning_rate": 2.5412061275935623e-05,
+      "loss": 5.4034,
+      "step": 5072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0250638723373413,
+      "learning_rate": 2.5392670157068067e-05,
+      "loss": 5.3628,
+      "step": 5076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9859606027603149,
+      "learning_rate": 2.5373279038200504e-05,
+      "loss": 5.4041,
+      "step": 5080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.01252019405365,
+      "learning_rate": 2.5353887919332948e-05,
+      "loss": 5.4278,
+      "step": 5084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.083108901977539,
+      "learning_rate": 2.5334496800465392e-05,
+      "loss": 5.3863,
+      "step": 5088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.969508945941925,
+      "learning_rate": 2.531510568159783e-05,
+      "loss": 5.3222,
+      "step": 5092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9864810109138489,
+      "learning_rate": 2.5295714562730273e-05,
+      "loss": 5.4214,
+      "step": 5096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9918404221534729,
+      "learning_rate": 2.527632344386271e-05,
+      "loss": 5.3679,
+      "step": 5100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.036550760269165,
+      "learning_rate": 2.5256932324995154e-05,
+      "loss": 5.2626,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9950888752937317,
+      "learning_rate": 2.5237541206127598e-05,
+      "loss": 5.3105,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0199189186096191,
+      "learning_rate": 2.5218150087260035e-05,
+      "loss": 5.3776,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0516811609268188,
+      "learning_rate": 2.519875896839248e-05,
+      "loss": 5.3124,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.078303575515747,
+      "learning_rate": 2.517936784952492e-05,
+      "loss": 5.4144,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.987095296382904,
+      "learning_rate": 2.515997673065736e-05,
+      "loss": 5.349,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9634592533111572,
+      "learning_rate": 2.51405856117898e-05,
+      "loss": 5.2509,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0986416339874268,
+      "learning_rate": 2.5121194492922244e-05,
+      "loss": 5.3225,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0385926961898804,
+      "learning_rate": 2.510180337405468e-05,
+      "loss": 5.3529,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9627379179000854,
+      "learning_rate": 2.5082412255187126e-05,
+      "loss": 5.3675,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.003461480140686,
+      "learning_rate": 2.5063021136319566e-05,
+      "loss": 5.4202,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9771742224693298,
+      "learning_rate": 2.5043630017452007e-05,
+      "loss": 5.3042,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9279887676239014,
+      "learning_rate": 2.502423889858445e-05,
+      "loss": 5.3561,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9281094670295715,
+      "learning_rate": 2.5004847779716888e-05,
+      "loss": 5.3231,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0261962413787842,
+      "learning_rate": 2.498545666084933e-05,
+      "loss": 5.2869,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.956251859664917,
+      "learning_rate": 2.4966065541981772e-05,
+      "loss": 5.3058,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0073554515838623,
+      "learning_rate": 2.4946674423114213e-05,
+      "loss": 5.3455,
+      "step": 5168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9829320907592773,
+      "learning_rate": 2.4927283304246657e-05,
+      "loss": 5.2866,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.067661166191101,
+      "learning_rate": 2.4907892185379097e-05,
+      "loss": 5.3966,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.049310326576233,
+      "learning_rate": 2.4888501066511538e-05,
+      "loss": 5.471,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.052420973777771,
+      "learning_rate": 2.486910994764398e-05,
+      "loss": 5.3907,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0315730571746826,
+      "learning_rate": 2.4849718828776422e-05,
+      "loss": 5.3528,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.165887475013733,
+      "learning_rate": 2.4830327709908863e-05,
+      "loss": 5.399,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9832936525344849,
+      "learning_rate": 2.4810936591041303e-05,
+      "loss": 5.4077,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9647985100746155,
+      "learning_rate": 2.4791545472173747e-05,
+      "loss": 5.3483,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0132298469543457,
+      "learning_rate": 2.4772154353306188e-05,
+      "loss": 5.4671,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9725786447525024,
+      "learning_rate": 2.4752763234438628e-05,
+      "loss": 5.3745,
+      "step": 5208
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0451775789260864,
+      "learning_rate": 2.473337211557107e-05,
+      "loss": 5.4456,
+      "step": 5212
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9887628555297852,
+      "learning_rate": 2.4713980996703513e-05,
+      "loss": 5.333,
+      "step": 5216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.044665813446045,
+      "learning_rate": 2.4694589877835953e-05,
+      "loss": 5.3592,
+      "step": 5220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9709651470184326,
+      "learning_rate": 2.4675198758968394e-05,
+      "loss": 5.3393,
+      "step": 5224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9530077576637268,
+      "learning_rate": 2.4655807640100834e-05,
+      "loss": 5.3602,
+      "step": 5228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9592558145523071,
+      "learning_rate": 2.4636416521233278e-05,
+      "loss": 5.3931,
+      "step": 5232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0525456666946411,
+      "learning_rate": 2.461702540236572e-05,
+      "loss": 5.3367,
+      "step": 5236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0699265003204346,
+      "learning_rate": 2.459763428349816e-05,
+      "loss": 5.3894,
+      "step": 5240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0436670780181885,
+      "learning_rate": 2.4578243164630603e-05,
+      "loss": 5.3643,
+      "step": 5244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0085383653640747,
+      "learning_rate": 2.4558852045763044e-05,
+      "loss": 5.2872,
+      "step": 5248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0748775005340576,
+      "learning_rate": 2.4539460926895484e-05,
+      "loss": 5.4714,
+      "step": 5252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0273828506469727,
+      "learning_rate": 2.4520069808027925e-05,
+      "loss": 5.3271,
+      "step": 5256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9897158145904541,
+      "learning_rate": 2.4500678689160365e-05,
+      "loss": 5.3268,
+      "step": 5260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9869460463523865,
+      "learning_rate": 2.4481287570292806e-05,
+      "loss": 5.3173,
+      "step": 5264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0365619659423828,
+      "learning_rate": 2.4461896451425246e-05,
+      "loss": 5.4168,
+      "step": 5268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.106515884399414,
+      "learning_rate": 2.4442505332557687e-05,
+      "loss": 5.3515,
+      "step": 5272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9866341352462769,
+      "learning_rate": 2.442311421369013e-05,
+      "loss": 5.4644,
+      "step": 5276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1131459474563599,
+      "learning_rate": 2.440372309482257e-05,
+      "loss": 5.2931,
+      "step": 5280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9824268221855164,
+      "learning_rate": 2.4384331975955012e-05,
+      "loss": 5.3094,
+      "step": 5284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0347408056259155,
+      "learning_rate": 2.4364940857087452e-05,
+      "loss": 5.3462,
+      "step": 5288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0148606300354004,
+      "learning_rate": 2.4345549738219896e-05,
+      "loss": 5.3224,
+      "step": 5292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1260877847671509,
+      "learning_rate": 2.4326158619352337e-05,
+      "loss": 5.3337,
+      "step": 5296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.034441590309143,
+      "learning_rate": 2.4306767500484777e-05,
+      "loss": 5.3728,
+      "step": 5300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0142707824707031,
+      "learning_rate": 2.428737638161722e-05,
+      "loss": 5.3848,
+      "step": 5304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9446210861206055,
+      "learning_rate": 2.4267985262749662e-05,
+      "loss": 5.3927,
+      "step": 5308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0099536180496216,
+      "learning_rate": 2.4248594143882102e-05,
+      "loss": 5.3282,
+      "step": 5312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0873996019363403,
+      "learning_rate": 2.4229203025014543e-05,
+      "loss": 5.4533,
+      "step": 5316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9772780537605286,
+      "learning_rate": 2.4209811906146987e-05,
+      "loss": 5.4258,
+      "step": 5320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9593787789344788,
+      "learning_rate": 2.4190420787279427e-05,
+      "loss": 5.369,
+      "step": 5324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0021839141845703,
+      "learning_rate": 2.4171029668411868e-05,
+      "loss": 5.3721,
+      "step": 5328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9823392629623413,
+      "learning_rate": 2.415163854954431e-05,
+      "loss": 5.3277,
+      "step": 5332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1319977045059204,
+      "learning_rate": 2.4132247430676752e-05,
+      "loss": 5.4176,
+      "step": 5336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9715146422386169,
+      "learning_rate": 2.4112856311809193e-05,
+      "loss": 5.3833,
+      "step": 5340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0981075763702393,
+      "learning_rate": 2.4093465192941634e-05,
+      "loss": 5.4515,
+      "step": 5344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9924083948135376,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 5.2979,
+      "step": 5348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9601098895072937,
+      "learning_rate": 2.4054682955206518e-05,
+      "loss": 5.3825,
+      "step": 5352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.134017825126648,
+      "learning_rate": 2.403529183633896e-05,
+      "loss": 5.4259,
+      "step": 5356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0065308809280396,
+      "learning_rate": 2.40159007174714e-05,
+      "loss": 5.2852,
+      "step": 5360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.120582103729248,
+      "learning_rate": 2.3996509598603843e-05,
+      "loss": 5.3872,
+      "step": 5364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0139710903167725,
+      "learning_rate": 2.3977118479736284e-05,
+      "loss": 5.4521,
+      "step": 5368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9832119941711426,
+      "learning_rate": 2.3957727360868724e-05,
+      "loss": 5.4052,
+      "step": 5372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0349886417388916,
+      "learning_rate": 2.3938336242001165e-05,
+      "loss": 5.2876,
+      "step": 5376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0244460105895996,
+      "learning_rate": 2.391894512313361e-05,
+      "loss": 5.4077,
+      "step": 5380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0130488872528076,
+      "learning_rate": 2.389955400426605e-05,
+      "loss": 5.3853,
+      "step": 5384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0110422372817993,
+      "learning_rate": 2.388016288539849e-05,
+      "loss": 5.4025,
+      "step": 5388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1506974697113037,
+      "learning_rate": 2.386077176653093e-05,
+      "loss": 5.4613,
+      "step": 5392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9795562624931335,
+      "learning_rate": 2.384138064766337e-05,
+      "loss": 5.3075,
+      "step": 5396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1148579120635986,
+      "learning_rate": 2.382198952879581e-05,
+      "loss": 5.4263,
+      "step": 5400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.993959367275238,
+      "learning_rate": 2.3802598409928252e-05,
+      "loss": 5.3765,
+      "step": 5404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9957119822502136,
+      "learning_rate": 2.3783207291060692e-05,
+      "loss": 5.3645,
+      "step": 5408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.975545346736908,
+      "learning_rate": 2.3763816172193136e-05,
+      "loss": 5.3522,
+      "step": 5412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9839780330657959,
+      "learning_rate": 2.3744425053325577e-05,
+      "loss": 5.333,
+      "step": 5416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0729457139968872,
+      "learning_rate": 2.3725033934458017e-05,
+      "loss": 5.4589,
+      "step": 5420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0412721633911133,
+      "learning_rate": 2.370564281559046e-05,
+      "loss": 5.3807,
+      "step": 5424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0309202671051025,
+      "learning_rate": 2.3686251696722902e-05,
+      "loss": 5.3716,
+      "step": 5428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.076830506324768,
+      "learning_rate": 2.3666860577855342e-05,
+      "loss": 5.4141,
+      "step": 5432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1014715433120728,
+      "learning_rate": 2.3647469458987783e-05,
+      "loss": 5.3345,
+      "step": 5436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.107503056526184,
+      "learning_rate": 2.3628078340120227e-05,
+      "loss": 5.2971,
+      "step": 5440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0111589431762695,
+      "learning_rate": 2.3608687221252667e-05,
+      "loss": 5.3824,
+      "step": 5444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.014420509338379,
+      "learning_rate": 2.3589296102385108e-05,
+      "loss": 5.4623,
+      "step": 5448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9904616475105286,
+      "learning_rate": 2.356990498351755e-05,
+      "loss": 5.3542,
+      "step": 5452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.061886191368103,
+      "learning_rate": 2.3550513864649992e-05,
+      "loss": 5.4478,
+      "step": 5456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.016271948814392,
+      "learning_rate": 2.3531122745782433e-05,
+      "loss": 5.3555,
+      "step": 5460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9344733953475952,
+      "learning_rate": 2.3511731626914873e-05,
+      "loss": 5.2854,
+      "step": 5464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.050699234008789,
+      "learning_rate": 2.3492340508047314e-05,
+      "loss": 5.3064,
+      "step": 5468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.024604320526123,
+      "learning_rate": 2.3472949389179758e-05,
+      "loss": 5.3682,
+      "step": 5472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0533298254013062,
+      "learning_rate": 2.34535582703122e-05,
+      "loss": 5.3336,
+      "step": 5476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0331076383590698,
+      "learning_rate": 2.343416715144464e-05,
+      "loss": 5.518,
+      "step": 5480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.101364254951477,
+      "learning_rate": 2.3414776032577083e-05,
+      "loss": 5.3366,
+      "step": 5484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9391170740127563,
+      "learning_rate": 2.3395384913709523e-05,
+      "loss": 5.3567,
+      "step": 5488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9690737724304199,
+      "learning_rate": 2.3375993794841964e-05,
+      "loss": 5.3323,
+      "step": 5492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0040618181228638,
+      "learning_rate": 2.3356602675974404e-05,
+      "loss": 5.3209,
+      "step": 5496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9501678347587585,
+      "learning_rate": 2.333721155710685e-05,
+      "loss": 5.3043,
+      "step": 5500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9725291728973389,
+      "learning_rate": 2.331782043823929e-05,
+      "loss": 5.4104,
+      "step": 5504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9561870098114014,
+      "learning_rate": 2.329842931937173e-05,
+      "loss": 5.4013,
+      "step": 5508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0054466724395752,
+      "learning_rate": 2.327903820050417e-05,
+      "loss": 5.3692,
+      "step": 5512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0419256687164307,
+      "learning_rate": 2.3259647081636614e-05,
+      "loss": 5.4082,
+      "step": 5516
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0327070951461792,
+      "learning_rate": 2.3240255962769054e-05,
+      "loss": 5.4721,
+      "step": 5520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9534458518028259,
+      "learning_rate": 2.3220864843901495e-05,
+      "loss": 5.3484,
+      "step": 5524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0182461738586426,
+      "learning_rate": 2.3201473725033936e-05,
+      "loss": 5.3595,
+      "step": 5528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0508521795272827,
+      "learning_rate": 2.3182082606166376e-05,
+      "loss": 5.3651,
+      "step": 5532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9299846887588501,
+      "learning_rate": 2.3162691487298817e-05,
+      "loss": 5.3989,
+      "step": 5536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0000419616699219,
+      "learning_rate": 2.3143300368431257e-05,
+      "loss": 5.2571,
+      "step": 5540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.084428071975708,
+      "learning_rate": 2.31239092495637e-05,
+      "loss": 5.3619,
+      "step": 5544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0404269695281982,
+      "learning_rate": 2.310451813069614e-05,
+      "loss": 5.3425,
+      "step": 5548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.00773286819458,
+      "learning_rate": 2.3085127011828582e-05,
+      "loss": 5.329,
+      "step": 5552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.000942349433899,
+      "learning_rate": 2.3065735892961023e-05,
+      "loss": 5.3361,
+      "step": 5556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.021449089050293,
+      "learning_rate": 2.3046344774093467e-05,
+      "loss": 5.3859,
+      "step": 5560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0686131715774536,
+      "learning_rate": 2.3026953655225907e-05,
+      "loss": 5.4319,
+      "step": 5564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0589152574539185,
+      "learning_rate": 2.3007562536358348e-05,
+      "loss": 5.3367,
+      "step": 5568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0058645009994507,
+      "learning_rate": 2.2988171417490788e-05,
+      "loss": 5.3473,
+      "step": 5572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0159991979599,
+      "learning_rate": 2.2968780298623232e-05,
+      "loss": 5.375,
+      "step": 5576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0634344816207886,
+      "learning_rate": 2.2949389179755673e-05,
+      "loss": 5.401,
+      "step": 5580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.077378273010254,
+      "learning_rate": 2.2929998060888113e-05,
+      "loss": 5.4141,
+      "step": 5584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9465997815132141,
+      "learning_rate": 2.2910606942020554e-05,
+      "loss": 5.357,
+      "step": 5588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.069690465927124,
+      "learning_rate": 2.2891215823152998e-05,
+      "loss": 5.3555,
+      "step": 5592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9830535054206848,
+      "learning_rate": 2.2871824704285438e-05,
+      "loss": 5.3552,
+      "step": 5596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0138301849365234,
+      "learning_rate": 2.285243358541788e-05,
+      "loss": 5.342,
+      "step": 5600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0020854473114014,
+      "learning_rate": 2.2833042466550323e-05,
+      "loss": 5.3489,
+      "step": 5604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0177456140518188,
+      "learning_rate": 2.2813651347682763e-05,
+      "loss": 5.269,
+      "step": 5608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0663460493087769,
+      "learning_rate": 2.2794260228815204e-05,
+      "loss": 5.297,
+      "step": 5612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9828472137451172,
+      "learning_rate": 2.2774869109947644e-05,
+      "loss": 5.3487,
+      "step": 5616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0115931034088135,
+      "learning_rate": 2.2755477991080088e-05,
+      "loss": 5.2995,
+      "step": 5620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.974193274974823,
+      "learning_rate": 2.273608687221253e-05,
+      "loss": 5.3582,
+      "step": 5624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0036460161209106,
+      "learning_rate": 2.271669575334497e-05,
+      "loss": 5.3034,
+      "step": 5628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0930732488632202,
+      "learning_rate": 2.269730463447741e-05,
+      "loss": 5.3337,
+      "step": 5632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9957634806632996,
+      "learning_rate": 2.2677913515609854e-05,
+      "loss": 5.3641,
+      "step": 5636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0515607595443726,
+      "learning_rate": 2.2658522396742294e-05,
+      "loss": 5.5135,
+      "step": 5640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9784524440765381,
+      "learning_rate": 2.2639131277874735e-05,
+      "loss": 5.344,
+      "step": 5644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0054571628570557,
+      "learning_rate": 2.2619740159007175e-05,
+      "loss": 5.3933,
+      "step": 5648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.965859591960907,
+      "learning_rate": 2.260034904013962e-05,
+      "loss": 5.3461,
+      "step": 5652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1478146314620972,
+      "learning_rate": 2.258095792127206e-05,
+      "loss": 5.3515,
+      "step": 5656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9893291592597961,
+      "learning_rate": 2.25615668024045e-05,
+      "loss": 5.2155,
+      "step": 5660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.945925235748291,
+      "learning_rate": 2.254217568353694e-05,
+      "loss": 5.3156,
+      "step": 5664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.020790934562683,
+      "learning_rate": 2.252278456466938e-05,
+      "loss": 5.4443,
+      "step": 5668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9524897933006287,
+      "learning_rate": 2.2503393445801822e-05,
+      "loss": 5.3692,
+      "step": 5672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0599005222320557,
+      "learning_rate": 2.2484002326934266e-05,
+      "loss": 5.3473,
+      "step": 5676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.029077410697937,
+      "learning_rate": 2.2464611208066706e-05,
+      "loss": 5.3266,
+      "step": 5680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0404125452041626,
+      "learning_rate": 2.2445220089199147e-05,
+      "loss": 5.3654,
+      "step": 5684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9789564609527588,
+      "learning_rate": 2.2425828970331587e-05,
+      "loss": 5.3689,
+      "step": 5688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9803569912910461,
+      "learning_rate": 2.2406437851464028e-05,
+      "loss": 5.3231,
+      "step": 5692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.036303997039795,
+      "learning_rate": 2.2387046732596472e-05,
+      "loss": 5.3055,
+      "step": 5696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0285331010818481,
+      "learning_rate": 2.2367655613728912e-05,
+      "loss": 5.3553,
+      "step": 5700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9725428819656372,
+      "learning_rate": 2.2348264494861353e-05,
+      "loss": 5.3338,
+      "step": 5704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0550493001937866,
+      "learning_rate": 2.2328873375993794e-05,
+      "loss": 5.3542,
+      "step": 5708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9990110993385315,
+      "learning_rate": 2.2309482257126237e-05,
+      "loss": 5.3457,
+      "step": 5712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0040626525878906,
+      "learning_rate": 2.2290091138258678e-05,
+      "loss": 5.3244,
+      "step": 5716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0053871870040894,
+      "learning_rate": 2.227070001939112e-05,
+      "loss": 5.3694,
+      "step": 5720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0112636089324951,
+      "learning_rate": 2.2251308900523562e-05,
+      "loss": 5.4549,
+      "step": 5724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0170855522155762,
+      "learning_rate": 2.2231917781656003e-05,
+      "loss": 5.3509,
+      "step": 5728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9416339993476868,
+      "learning_rate": 2.2212526662788444e-05,
+      "loss": 5.2804,
+      "step": 5732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.014902114868164,
+      "learning_rate": 2.2193135543920884e-05,
+      "loss": 5.3263,
+      "step": 5736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0529359579086304,
+      "learning_rate": 2.2173744425053328e-05,
+      "loss": 5.3151,
+      "step": 5740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9608129262924194,
+      "learning_rate": 2.215435330618577e-05,
+      "loss": 5.2865,
+      "step": 5744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0631046295166016,
+      "learning_rate": 2.213496218731821e-05,
+      "loss": 5.3379,
+      "step": 5748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0569404363632202,
+      "learning_rate": 2.211557106845065e-05,
+      "loss": 5.4344,
+      "step": 5752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0010555982589722,
+      "learning_rate": 2.2096179949583094e-05,
+      "loss": 5.3753,
+      "step": 5756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9690674543380737,
+      "learning_rate": 2.2076788830715534e-05,
+      "loss": 5.3373,
+      "step": 5760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.03341805934906,
+      "learning_rate": 2.2057397711847975e-05,
+      "loss": 5.4154,
+      "step": 5764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9742380976676941,
+      "learning_rate": 2.2038006592980415e-05,
+      "loss": 5.4585,
+      "step": 5768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.977780818939209,
+      "learning_rate": 2.201861547411286e-05,
+      "loss": 5.351,
+      "step": 5772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9809128046035767,
+      "learning_rate": 2.19992243552453e-05,
+      "loss": 5.2699,
+      "step": 5776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0754761695861816,
+      "learning_rate": 2.197983323637774e-05,
+      "loss": 5.4224,
+      "step": 5780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1107863187789917,
+      "learning_rate": 2.1960442117510184e-05,
+      "loss": 5.3779,
+      "step": 5784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9794926643371582,
+      "learning_rate": 2.1941050998642625e-05,
+      "loss": 5.4183,
+      "step": 5788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0375285148620605,
+      "learning_rate": 2.1921659879775065e-05,
+      "loss": 5.3357,
+      "step": 5792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.028106451034546,
+      "learning_rate": 2.1902268760907506e-05,
+      "loss": 5.3636,
+      "step": 5796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9496660828590393,
+      "learning_rate": 2.1882877642039946e-05,
+      "loss": 5.4075,
+      "step": 5800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9957082867622375,
+      "learning_rate": 2.1863486523172387e-05,
+      "loss": 5.3508,
+      "step": 5804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0487018823623657,
+      "learning_rate": 2.184409540430483e-05,
+      "loss": 5.3796,
+      "step": 5808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9658469557762146,
+      "learning_rate": 2.182470428543727e-05,
+      "loss": 5.2512,
+      "step": 5812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0086036920547485,
+      "learning_rate": 2.1805313166569712e-05,
+      "loss": 5.3492,
+      "step": 5816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.09529447555542,
+      "learning_rate": 2.1785922047702152e-05,
+      "loss": 5.3167,
+      "step": 5820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0463392734527588,
+      "learning_rate": 2.1766530928834593e-05,
+      "loss": 5.3627,
+      "step": 5824
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0675616264343262,
+      "learning_rate": 2.1747139809967033e-05,
+      "loss": 5.3868,
+      "step": 5828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9881440997123718,
+      "learning_rate": 2.1727748691099477e-05,
+      "loss": 5.3498,
+      "step": 5832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0680946111679077,
+      "learning_rate": 2.1708357572231918e-05,
+      "loss": 5.3376,
+      "step": 5836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1282621622085571,
+      "learning_rate": 2.168896645336436e-05,
+      "loss": 5.3426,
+      "step": 5840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0684411525726318,
+      "learning_rate": 2.1669575334496802e-05,
+      "loss": 5.3769,
+      "step": 5844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0463793277740479,
+      "learning_rate": 2.1650184215629243e-05,
+      "loss": 5.3831,
+      "step": 5848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0346778631210327,
+      "learning_rate": 2.1630793096761683e-05,
+      "loss": 5.3522,
+      "step": 5852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0426431894302368,
+      "learning_rate": 2.1611401977894124e-05,
+      "loss": 5.4098,
+      "step": 5856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0200704336166382,
+      "learning_rate": 2.1592010859026568e-05,
+      "loss": 5.265,
+      "step": 5860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0575344562530518,
+      "learning_rate": 2.157261974015901e-05,
+      "loss": 5.3004,
+      "step": 5864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9970076680183411,
+      "learning_rate": 2.155322862129145e-05,
+      "loss": 5.3595,
+      "step": 5868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0401124954223633,
+      "learning_rate": 2.153383750242389e-05,
+      "loss": 5.5118,
+      "step": 5872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0695228576660156,
+      "learning_rate": 2.1514446383556333e-05,
+      "loss": 5.3078,
+      "step": 5876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0231329202651978,
+      "learning_rate": 2.1495055264688774e-05,
+      "loss": 5.3619,
+      "step": 5880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0032014846801758,
+      "learning_rate": 2.1475664145821214e-05,
+      "loss": 5.2561,
+      "step": 5884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0087443590164185,
+      "learning_rate": 2.1456273026953655e-05,
+      "loss": 5.4882,
+      "step": 5888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.014003038406372,
+      "learning_rate": 2.14368819080861e-05,
+      "loss": 5.4139,
+      "step": 5892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9923560619354248,
+      "learning_rate": 2.141749078921854e-05,
+      "loss": 5.3716,
+      "step": 5896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.022443175315857,
+      "learning_rate": 2.139809967035098e-05,
+      "loss": 5.3654,
+      "step": 5900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0812815427780151,
+      "learning_rate": 2.1378708551483424e-05,
+      "loss": 5.4307,
+      "step": 5904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9952316284179688,
+      "learning_rate": 2.1359317432615864e-05,
+      "loss": 5.3851,
+      "step": 5908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0843716859817505,
+      "learning_rate": 2.1339926313748305e-05,
+      "loss": 5.3507,
+      "step": 5912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9775163531303406,
+      "learning_rate": 2.1320535194880746e-05,
+      "loss": 5.2868,
+      "step": 5916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.051283597946167,
+      "learning_rate": 2.130114407601319e-05,
+      "loss": 5.4404,
+      "step": 5920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0350865125656128,
+      "learning_rate": 2.128175295714563e-05,
+      "loss": 5.3286,
+      "step": 5924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1702210903167725,
+      "learning_rate": 2.126236183827807e-05,
+      "loss": 5.4338,
+      "step": 5928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0268408060073853,
+      "learning_rate": 2.124297071941051e-05,
+      "loss": 5.4108,
+      "step": 5932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.002463459968567,
+      "learning_rate": 2.122357960054295e-05,
+      "loss": 5.2631,
+      "step": 5936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.04947829246521,
+      "learning_rate": 2.1204188481675396e-05,
+      "loss": 5.3011,
+      "step": 5940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0733132362365723,
+      "learning_rate": 2.1184797362807836e-05,
+      "loss": 5.5408,
+      "step": 5944
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1274093389511108,
+      "learning_rate": 2.1165406243940277e-05,
+      "loss": 5.4001,
+      "step": 5948
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0161489248275757,
+      "learning_rate": 2.1146015125072717e-05,
+      "loss": 5.392,
+      "step": 5952
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1048080921173096,
+      "learning_rate": 2.1126624006205158e-05,
+      "loss": 5.3586,
+      "step": 5956
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0635709762573242,
+      "learning_rate": 2.1107232887337598e-05,
+      "loss": 5.3164,
+      "step": 5960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0826531648635864,
+      "learning_rate": 2.1087841768470042e-05,
+      "loss": 5.4166,
+      "step": 5964
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0716171264648438,
+      "learning_rate": 2.1068450649602483e-05,
+      "loss": 5.3028,
+      "step": 5968
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0432461500167847,
+      "learning_rate": 2.1049059530734923e-05,
+      "loss": 5.3777,
+      "step": 5972
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0327306985855103,
+      "learning_rate": 2.1029668411867364e-05,
+      "loss": 5.2718,
+      "step": 5976
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0294878482818604,
+      "learning_rate": 2.1010277292999808e-05,
+      "loss": 5.2555,
+      "step": 5980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0690736770629883,
+      "learning_rate": 2.0990886174132248e-05,
+      "loss": 5.2835,
+      "step": 5984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.060434341430664,
+      "learning_rate": 2.097149505526469e-05,
+      "loss": 5.4269,
+      "step": 5988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1109763383865356,
+      "learning_rate": 2.095210393639713e-05,
+      "loss": 5.4351,
+      "step": 5992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0055971145629883,
+      "learning_rate": 2.0932712817529573e-05,
+      "loss": 5.2801,
+      "step": 5996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.990050733089447,
+      "learning_rate": 2.0913321698662014e-05,
+      "loss": 5.3526,
+      "step": 6000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9774187207221985,
+      "learning_rate": 2.0893930579794454e-05,
+      "loss": 5.3241,
+      "step": 6004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0026575326919556,
+      "learning_rate": 2.0874539460926895e-05,
+      "loss": 5.3575,
+      "step": 6008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9780769944190979,
+      "learning_rate": 2.085514834205934e-05,
+      "loss": 5.2948,
+      "step": 6012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0509179830551147,
+      "learning_rate": 2.083575722319178e-05,
+      "loss": 5.3332,
+      "step": 6016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0327376127243042,
+      "learning_rate": 2.081636610432422e-05,
+      "loss": 5.38,
+      "step": 6020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.105682134628296,
+      "learning_rate": 2.0796974985456664e-05,
+      "loss": 5.3568,
+      "step": 6024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0796016454696655,
+      "learning_rate": 2.0777583866589104e-05,
+      "loss": 5.3871,
+      "step": 6028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.983562171459198,
+      "learning_rate": 2.0758192747721545e-05,
+      "loss": 5.3737,
+      "step": 6032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0870602130889893,
+      "learning_rate": 2.0738801628853985e-05,
+      "loss": 5.2752,
+      "step": 6036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9455908536911011,
+      "learning_rate": 2.071941050998643e-05,
+      "loss": 5.3517,
+      "step": 6040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0257951021194458,
+      "learning_rate": 2.070001939111887e-05,
+      "loss": 5.3043,
+      "step": 6044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0068105459213257,
+      "learning_rate": 2.068062827225131e-05,
+      "loss": 5.2876,
+      "step": 6048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0739387273788452,
+      "learning_rate": 2.066123715338375e-05,
+      "loss": 5.3432,
+      "step": 6052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9713083505630493,
+      "learning_rate": 2.0641846034516195e-05,
+      "loss": 5.368,
+      "step": 6056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9598544239997864,
+      "learning_rate": 2.0622454915648635e-05,
+      "loss": 5.467,
+      "step": 6060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0342966318130493,
+      "learning_rate": 2.0603063796781076e-05,
+      "loss": 5.3495,
+      "step": 6064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9927212595939636,
+      "learning_rate": 2.0583672677913516e-05,
+      "loss": 5.3898,
+      "step": 6068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.985098659992218,
+      "learning_rate": 2.056428155904596e-05,
+      "loss": 5.3315,
+      "step": 6072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1182785034179688,
+      "learning_rate": 2.05448904401784e-05,
+      "loss": 5.3721,
+      "step": 6076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0341837406158447,
+      "learning_rate": 2.052549932131084e-05,
+      "loss": 5.3284,
+      "step": 6080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0580908060073853,
+      "learning_rate": 2.0506108202443282e-05,
+      "loss": 5.3776,
+      "step": 6084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.016257405281067,
+      "learning_rate": 2.0486717083575722e-05,
+      "loss": 5.3578,
+      "step": 6088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0956933498382568,
+      "learning_rate": 2.0467325964708163e-05,
+      "loss": 5.3792,
+      "step": 6092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0442465543746948,
+      "learning_rate": 2.0447934845840604e-05,
+      "loss": 5.2992,
+      "step": 6096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1148533821105957,
+      "learning_rate": 2.0428543726973047e-05,
+      "loss": 5.2835,
+      "step": 6100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.056727409362793,
+      "learning_rate": 2.0409152608105488e-05,
+      "loss": 5.3336,
+      "step": 6104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9661352634429932,
+      "learning_rate": 2.038976148923793e-05,
+      "loss": 5.356,
+      "step": 6108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0282217264175415,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 5.3118,
+      "step": 6112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0714175701141357,
+      "learning_rate": 2.0350979251502813e-05,
+      "loss": 5.3817,
+      "step": 6116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0747861862182617,
+      "learning_rate": 2.0331588132635254e-05,
+      "loss": 5.3961,
+      "step": 6120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9946739673614502,
+      "learning_rate": 2.0312197013767694e-05,
+      "loss": 5.343,
+      "step": 6124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.002004623413086,
+      "learning_rate": 2.0292805894900135e-05,
+      "loss": 5.3654,
+      "step": 6128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0547658205032349,
+      "learning_rate": 2.027341477603258e-05,
+      "loss": 5.3869,
+      "step": 6132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.076196551322937,
+      "learning_rate": 2.025402365716502e-05,
+      "loss": 5.2561,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3180465698242188,
+      "learning_rate": 2.023463253829746e-05,
+      "loss": 5.3457,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0028260946273804,
+      "learning_rate": 2.0215241419429904e-05,
+      "loss": 5.4271,
+      "step": 6144
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0375735759735107,
+      "learning_rate": 2.0195850300562344e-05,
+      "loss": 5.3181,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0055443048477173,
+      "learning_rate": 2.0176459181694785e-05,
+      "loss": 5.3544,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0876275300979614,
+      "learning_rate": 2.0157068062827225e-05,
+      "loss": 5.3487,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.018140196800232,
+      "learning_rate": 2.013767694395967e-05,
+      "loss": 5.4325,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0800230503082275,
+      "learning_rate": 2.011828582509211e-05,
+      "loss": 5.2748,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0820457935333252,
+      "learning_rate": 2.009889470622455e-05,
+      "loss": 5.3661,
+      "step": 6168
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0936886072158813,
+      "learning_rate": 2.007950358735699e-05,
+      "loss": 5.3191,
+      "step": 6172
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0739785432815552,
+      "learning_rate": 2.0060112468489435e-05,
+      "loss": 5.372,
+      "step": 6176
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0785382986068726,
+      "learning_rate": 2.0040721349621875e-05,
+      "loss": 5.3858,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.000064492225647,
+      "learning_rate": 2.0021330230754316e-05,
+      "loss": 5.3544,
+      "step": 6184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.096794605255127,
+      "learning_rate": 2.0001939111886756e-05,
+      "loss": 5.3608,
+      "step": 6188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0260350704193115,
+      "learning_rate": 1.99825479930192e-05,
+      "loss": 5.4049,
+      "step": 6192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0056685209274292,
+      "learning_rate": 1.996315687415164e-05,
+      "loss": 5.3588,
+      "step": 6196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.003697156906128,
+      "learning_rate": 1.994376575528408e-05,
+      "loss": 5.4,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0048468112945557,
+      "learning_rate": 1.9924374636416525e-05,
+      "loss": 5.2654,
+      "step": 6204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9437733292579651,
+      "learning_rate": 1.9904983517548966e-05,
+      "loss": 5.2859,
+      "step": 6208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.961186945438385,
+      "learning_rate": 1.9885592398681406e-05,
+      "loss": 5.3605,
+      "step": 6212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0297905206680298,
+      "learning_rate": 1.9866201279813847e-05,
+      "loss": 5.3202,
+      "step": 6216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3261497020721436,
+      "learning_rate": 1.9846810160946287e-05,
+      "loss": 5.3587,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1081327199935913,
+      "learning_rate": 1.9827419042078728e-05,
+      "loss": 5.427,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0671581029891968,
+      "learning_rate": 1.980802792321117e-05,
+      "loss": 5.453,
+      "step": 6228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0665878057479858,
+      "learning_rate": 1.978863680434361e-05,
+      "loss": 5.3263,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9510138034820557,
+      "learning_rate": 1.9769245685476053e-05,
+      "loss": 5.3364,
+      "step": 6236
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9980200529098511,
+      "learning_rate": 1.9749854566608493e-05,
+      "loss": 5.2064,
+      "step": 6240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0917850732803345,
+      "learning_rate": 1.9730463447740934e-05,
+      "loss": 5.3213,
+      "step": 6244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0644116401672363,
+      "learning_rate": 1.9711072328873374e-05,
+      "loss": 5.2939,
+      "step": 6248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0722259283065796,
+      "learning_rate": 1.969168121000582e-05,
+      "loss": 5.3984,
+      "step": 6252
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0094479322433472,
+      "learning_rate": 1.967229009113826e-05,
+      "loss": 5.3198,
+      "step": 6256
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0126482248306274,
+      "learning_rate": 1.96528989722707e-05,
+      "loss": 5.3356,
+      "step": 6260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.061038613319397,
+      "learning_rate": 1.9633507853403143e-05,
+      "loss": 5.294,
+      "step": 6264
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1081583499908447,
+      "learning_rate": 1.9614116734535584e-05,
+      "loss": 5.3518,
+      "step": 6268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0472965240478516,
+      "learning_rate": 1.9594725615668024e-05,
+      "loss": 5.3526,
+      "step": 6272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9935174584388733,
+      "learning_rate": 1.9575334496800465e-05,
+      "loss": 5.3108,
+      "step": 6276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0617495775222778,
+      "learning_rate": 1.955594337793291e-05,
+      "loss": 5.3518,
+      "step": 6280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0488762855529785,
+      "learning_rate": 1.953655225906535e-05,
+      "loss": 5.2888,
+      "step": 6284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.065202236175537,
+      "learning_rate": 1.951716114019779e-05,
+      "loss": 5.2927,
+      "step": 6288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1323513984680176,
+      "learning_rate": 1.949777002133023e-05,
+      "loss": 5.3774,
+      "step": 6292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0796220302581787,
+      "learning_rate": 1.9478378902462674e-05,
+      "loss": 5.3579,
+      "step": 6296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0631530284881592,
+      "learning_rate": 1.9458987783595115e-05,
+      "loss": 5.2644,
+      "step": 6300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0359313488006592,
+      "learning_rate": 1.9439596664727556e-05,
+      "loss": 5.3949,
+      "step": 6304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0680855512619019,
+      "learning_rate": 1.9420205545859996e-05,
+      "loss": 5.3567,
+      "step": 6308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.013702154159546,
+      "learning_rate": 1.940081442699244e-05,
+      "loss": 5.4166,
+      "step": 6312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9817814230918884,
+      "learning_rate": 1.938142330812488e-05,
+      "loss": 5.4169,
+      "step": 6316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0646679401397705,
+      "learning_rate": 1.936203218925732e-05,
+      "loss": 5.3344,
+      "step": 6320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0351269245147705,
+      "learning_rate": 1.9342641070389765e-05,
+      "loss": 5.358,
+      "step": 6324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9547563195228577,
+      "learning_rate": 1.9323249951522206e-05,
+      "loss": 5.3869,
+      "step": 6328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0959416627883911,
+      "learning_rate": 1.9303858832654646e-05,
+      "loss": 5.3785,
+      "step": 6332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.026036262512207,
+      "learning_rate": 1.9284467713787087e-05,
+      "loss": 5.3627,
+      "step": 6336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0361682176589966,
+      "learning_rate": 1.926507659491953e-05,
+      "loss": 5.3649,
+      "step": 6340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0472002029418945,
+      "learning_rate": 1.924568547605197e-05,
+      "loss": 5.2838,
+      "step": 6344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0799552202224731,
+      "learning_rate": 1.922629435718441e-05,
+      "loss": 5.3879,
+      "step": 6348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0271328687667847,
+      "learning_rate": 1.9206903238316852e-05,
+      "loss": 5.3387,
+      "step": 6352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0310901403427124,
+      "learning_rate": 1.9187512119449293e-05,
+      "loss": 5.4013,
+      "step": 6356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0406147241592407,
+      "learning_rate": 1.9168121000581733e-05,
+      "loss": 5.3939,
+      "step": 6360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.048988938331604,
+      "learning_rate": 1.9148729881714174e-05,
+      "loss": 5.3211,
+      "step": 6364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0325740575790405,
+      "learning_rate": 1.9129338762846614e-05,
+      "loss": 5.4601,
+      "step": 6368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9758936166763306,
+      "learning_rate": 1.9109947643979058e-05,
+      "loss": 5.3075,
+      "step": 6372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.064858078956604,
+      "learning_rate": 1.90905565251115e-05,
+      "loss": 5.366,
+      "step": 6376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9335108399391174,
+      "learning_rate": 1.907116540624394e-05,
+      "loss": 5.2433,
+      "step": 6380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0345276594161987,
+      "learning_rate": 1.9051774287376383e-05,
+      "loss": 5.3438,
+      "step": 6384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0446693897247314,
+      "learning_rate": 1.9032383168508824e-05,
+      "loss": 5.4391,
+      "step": 6388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0137783288955688,
+      "learning_rate": 1.9012992049641264e-05,
+      "loss": 5.3926,
+      "step": 6392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.953395664691925,
+      "learning_rate": 1.8993600930773705e-05,
+      "loss": 5.3703,
+      "step": 6396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0034725666046143,
+      "learning_rate": 1.897420981190615e-05,
+      "loss": 5.3504,
+      "step": 6400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0460563898086548,
+      "learning_rate": 1.895481869303859e-05,
+      "loss": 5.4165,
+      "step": 6404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0266785621643066,
+      "learning_rate": 1.893542757417103e-05,
+      "loss": 5.3853,
+      "step": 6408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1107487678527832,
+      "learning_rate": 1.891603645530347e-05,
+      "loss": 5.4166,
+      "step": 6412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1155457496643066,
+      "learning_rate": 1.8896645336435914e-05,
+      "loss": 5.3838,
+      "step": 6416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0329363346099854,
+      "learning_rate": 1.8877254217568355e-05,
+      "loss": 5.3541,
+      "step": 6420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.028456449508667,
+      "learning_rate": 1.8857863098700795e-05,
+      "loss": 5.3506,
+      "step": 6424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.090142011642456,
+      "learning_rate": 1.8838471979833236e-05,
+      "loss": 5.3296,
+      "step": 6428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.071015477180481,
+      "learning_rate": 1.881908086096568e-05,
+      "loss": 5.3341,
+      "step": 6432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9764314293861389,
+      "learning_rate": 1.879968974209812e-05,
+      "loss": 5.3198,
+      "step": 6436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.046424388885498,
+      "learning_rate": 1.878029862323056e-05,
+      "loss": 5.3087,
+      "step": 6440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0243570804595947,
+      "learning_rate": 1.8760907504363005e-05,
+      "loss": 5.3045,
+      "step": 6444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9760267734527588,
+      "learning_rate": 1.8741516385495445e-05,
+      "loss": 5.2859,
+      "step": 6448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1072561740875244,
+      "learning_rate": 1.8722125266627886e-05,
+      "loss": 5.359,
+      "step": 6452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0756561756134033,
+      "learning_rate": 1.8702734147760326e-05,
+      "loss": 5.409,
+      "step": 6456
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9566421508789062,
+      "learning_rate": 1.868334302889277e-05,
+      "loss": 5.3902,
+      "step": 6460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0135536193847656,
+      "learning_rate": 1.866395191002521e-05,
+      "loss": 5.504,
+      "step": 6464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.139150619506836,
+      "learning_rate": 1.864456079115765e-05,
+      "loss": 5.3502,
+      "step": 6468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0426169633865356,
+      "learning_rate": 1.8625169672290092e-05,
+      "loss": 5.2562,
+      "step": 6472
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0229958295822144,
+      "learning_rate": 1.8605778553422536e-05,
+      "loss": 5.3472,
+      "step": 6476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0061606168746948,
+      "learning_rate": 1.8586387434554976e-05,
+      "loss": 5.2391,
+      "step": 6480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9669675230979919,
+      "learning_rate": 1.8566996315687417e-05,
+      "loss": 5.4266,
+      "step": 6484
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0888794660568237,
+      "learning_rate": 1.8547605196819857e-05,
+      "loss": 5.3798,
+      "step": 6488
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0779885053634644,
+      "learning_rate": 1.8528214077952298e-05,
+      "loss": 5.3159,
+      "step": 6492
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0792943239212036,
+      "learning_rate": 1.850882295908474e-05,
+      "loss": 5.3322,
+      "step": 6496
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0349109172821045,
+      "learning_rate": 1.848943184021718e-05,
+      "loss": 5.3125,
+      "step": 6500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0294796228408813,
+      "learning_rate": 1.8470040721349623e-05,
+      "loss": 5.3059,
+      "step": 6504
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9547713994979858,
+      "learning_rate": 1.8450649602482064e-05,
+      "loss": 5.2556,
+      "step": 6508
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1305149793624878,
+      "learning_rate": 1.8431258483614504e-05,
+      "loss": 5.3476,
+      "step": 6512
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0296196937561035,
+      "learning_rate": 1.8411867364746945e-05,
+      "loss": 5.3404,
+      "step": 6516
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0865687131881714,
+      "learning_rate": 1.839247624587939e-05,
+      "loss": 5.3772,
+      "step": 6520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.065975546836853,
+      "learning_rate": 1.837308512701183e-05,
+      "loss": 5.3577,
+      "step": 6524
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1665056943893433,
+      "learning_rate": 1.835369400814427e-05,
+      "loss": 5.3344,
+      "step": 6528
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0219711065292358,
+      "learning_rate": 1.833430288927671e-05,
+      "loss": 5.3717,
+      "step": 6532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.01799738407135,
+      "learning_rate": 1.8314911770409154e-05,
+      "loss": 5.3329,
+      "step": 6536
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0811940431594849,
+      "learning_rate": 1.8295520651541595e-05,
+      "loss": 5.2708,
+      "step": 6540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0369648933410645,
+      "learning_rate": 1.8276129532674035e-05,
+      "loss": 5.3731,
+      "step": 6544
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0611283779144287,
+      "learning_rate": 1.8256738413806476e-05,
+      "loss": 5.4037,
+      "step": 6548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9396833777427673,
+      "learning_rate": 1.823734729493892e-05,
+      "loss": 5.3158,
+      "step": 6552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9994028210639954,
+      "learning_rate": 1.821795617607136e-05,
+      "loss": 5.3562,
+      "step": 6556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0078858137130737,
+      "learning_rate": 1.81985650572038e-05,
+      "loss": 5.322,
+      "step": 6560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0305253267288208,
+      "learning_rate": 1.8179173938336245e-05,
+      "loss": 5.2791,
+      "step": 6564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0548304319381714,
+      "learning_rate": 1.8159782819468685e-05,
+      "loss": 5.356,
+      "step": 6568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9770938754081726,
+      "learning_rate": 1.8140391700601126e-05,
+      "loss": 5.3234,
+      "step": 6572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.087689757347107,
+      "learning_rate": 1.8121000581733566e-05,
+      "loss": 5.3944,
+      "step": 6576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.14712655544281,
+      "learning_rate": 1.810160946286601e-05,
+      "loss": 5.2835,
+      "step": 6580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1165133714675903,
+      "learning_rate": 1.808221834399845e-05,
+      "loss": 5.2902,
+      "step": 6584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.063067078590393,
+      "learning_rate": 1.806282722513089e-05,
+      "loss": 5.3108,
+      "step": 6588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0933597087860107,
+      "learning_rate": 1.8043436106263332e-05,
+      "loss": 5.431,
+      "step": 6592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0003483295440674,
+      "learning_rate": 1.8024044987395776e-05,
+      "loss": 5.3212,
+      "step": 6596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9593095779418945,
+      "learning_rate": 1.8004653868528216e-05,
+      "loss": 5.3799,
+      "step": 6600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0600883960723877,
+      "learning_rate": 1.7985262749660657e-05,
+      "loss": 5.2845,
+      "step": 6604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.054432988166809,
+      "learning_rate": 1.7965871630793097e-05,
+      "loss": 5.3376,
+      "step": 6608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0431993007659912,
+      "learning_rate": 1.794648051192554e-05,
+      "loss": 5.3923,
+      "step": 6612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1059125661849976,
+      "learning_rate": 1.7927089393057982e-05,
+      "loss": 5.3124,
+      "step": 6616
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0165326595306396,
+      "learning_rate": 1.7907698274190422e-05,
+      "loss": 5.3649,
+      "step": 6620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1015493869781494,
+      "learning_rate": 1.7888307155322863e-05,
+      "loss": 5.3207,
+      "step": 6624
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.062280297279358,
+      "learning_rate": 1.7868916036455303e-05,
+      "loss": 5.3276,
+      "step": 6628
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.101285696029663,
+      "learning_rate": 1.7849524917587744e-05,
+      "loss": 5.4415,
+      "step": 6632
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.039291501045227,
+      "learning_rate": 1.7830133798720184e-05,
+      "loss": 5.3052,
+      "step": 6636
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0250643491744995,
+      "learning_rate": 1.781074267985263e-05,
+      "loss": 5.337,
+      "step": 6640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0628857612609863,
+      "learning_rate": 1.779135156098507e-05,
+      "loss": 5.3236,
+      "step": 6644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0590665340423584,
+      "learning_rate": 1.777196044211751e-05,
+      "loss": 5.2928,
+      "step": 6648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.009734034538269,
+      "learning_rate": 1.775256932324995e-05,
+      "loss": 5.3493,
+      "step": 6652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9915148019790649,
+      "learning_rate": 1.7733178204382394e-05,
+      "loss": 5.3266,
+      "step": 6656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0235103368759155,
+      "learning_rate": 1.7713787085514834e-05,
+      "loss": 5.3359,
+      "step": 6660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.003313422203064,
+      "learning_rate": 1.7694395966647275e-05,
+      "loss": 5.3196,
+      "step": 6664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0449254512786865,
+      "learning_rate": 1.7675004847779716e-05,
+      "loss": 5.3287,
+      "step": 6668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0780065059661865,
+      "learning_rate": 1.765561372891216e-05,
+      "loss": 5.3901,
+      "step": 6672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.134416103363037,
+      "learning_rate": 1.76362226100446e-05,
+      "loss": 5.3008,
+      "step": 6676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.095788836479187,
+      "learning_rate": 1.761683149117704e-05,
+      "loss": 5.3973,
+      "step": 6680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.038216471672058,
+      "learning_rate": 1.7597440372309484e-05,
+      "loss": 5.3031,
+      "step": 6684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.033199429512024,
+      "learning_rate": 1.7578049253441925e-05,
+      "loss": 5.2332,
+      "step": 6688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0989290475845337,
+      "learning_rate": 1.7558658134574366e-05,
+      "loss": 5.3559,
+      "step": 6692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0528738498687744,
+      "learning_rate": 1.7539267015706806e-05,
+      "loss": 5.3435,
+      "step": 6696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0338224172592163,
+      "learning_rate": 1.751987589683925e-05,
+      "loss": 5.4003,
+      "step": 6700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.021791934967041,
+      "learning_rate": 1.750048477797169e-05,
+      "loss": 5.2902,
+      "step": 6704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.069692850112915,
+      "learning_rate": 1.748109365910413e-05,
+      "loss": 5.2897,
+      "step": 6708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0573803186416626,
+      "learning_rate": 1.746170254023657e-05,
+      "loss": 5.386,
+      "step": 6712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0095142126083374,
+      "learning_rate": 1.7442311421369016e-05,
+      "loss": 5.3541,
+      "step": 6716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0076450109481812,
+      "learning_rate": 1.7422920302501456e-05,
+      "loss": 5.3088,
+      "step": 6720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9816309809684753,
+      "learning_rate": 1.7403529183633897e-05,
+      "loss": 5.3906,
+      "step": 6724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1446229219436646,
+      "learning_rate": 1.7384138064766337e-05,
+      "loss": 5.3372,
+      "step": 6728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9934061169624329,
+      "learning_rate": 1.736474694589878e-05,
+      "loss": 5.4092,
+      "step": 6732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9881500005722046,
+      "learning_rate": 1.734535582703122e-05,
+      "loss": 5.2318,
+      "step": 6736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0100693702697754,
+      "learning_rate": 1.7325964708163662e-05,
+      "loss": 5.35,
+      "step": 6740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0123350620269775,
+      "learning_rate": 1.7306573589296106e-05,
+      "loss": 5.2857,
+      "step": 6744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0262787342071533,
+      "learning_rate": 1.7287182470428547e-05,
+      "loss": 5.425,
+      "step": 6748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0568041801452637,
+      "learning_rate": 1.7267791351560987e-05,
+      "loss": 5.3167,
+      "step": 6752
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0791805982589722,
+      "learning_rate": 1.7248400232693428e-05,
+      "loss": 5.3329,
+      "step": 6756
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0418261289596558,
+      "learning_rate": 1.7229009113825868e-05,
+      "loss": 5.3187,
+      "step": 6760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0648219585418701,
+      "learning_rate": 1.720961799495831e-05,
+      "loss": 5.2998,
+      "step": 6764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0518834590911865,
+      "learning_rate": 1.719022687609075e-05,
+      "loss": 5.4489,
+      "step": 6768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1231526136398315,
+      "learning_rate": 1.7170835757223193e-05,
+      "loss": 5.2959,
+      "step": 6772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0702370405197144,
+      "learning_rate": 1.7151444638355634e-05,
+      "loss": 5.321,
+      "step": 6776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0818907022476196,
+      "learning_rate": 1.7132053519488074e-05,
+      "loss": 5.2864,
+      "step": 6780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.990020215511322,
+      "learning_rate": 1.7112662400620515e-05,
+      "loss": 5.2245,
+      "step": 6784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.104169249534607,
+      "learning_rate": 1.7093271281752955e-05,
+      "loss": 5.29,
+      "step": 6788
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0738272666931152,
+      "learning_rate": 1.70738801628854e-05,
+      "loss": 5.2843,
+      "step": 6792
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.009252905845642,
+      "learning_rate": 1.705448904401784e-05,
+      "loss": 5.4593,
+      "step": 6796
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0361759662628174,
+      "learning_rate": 1.703509792515028e-05,
+      "loss": 5.2717,
+      "step": 6800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1224076747894287,
+      "learning_rate": 1.7015706806282724e-05,
+      "loss": 5.3675,
+      "step": 6804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.094973087310791,
+      "learning_rate": 1.6996315687415165e-05,
+      "loss": 5.3096,
+      "step": 6808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0065085887908936,
+      "learning_rate": 1.6976924568547605e-05,
+      "loss": 5.41,
+      "step": 6812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0526219606399536,
+      "learning_rate": 1.6957533449680046e-05,
+      "loss": 5.408,
+      "step": 6816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0697109699249268,
+      "learning_rate": 1.693814233081249e-05,
+      "loss": 5.3375,
+      "step": 6820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0227476358413696,
+      "learning_rate": 1.691875121194493e-05,
+      "loss": 5.3254,
+      "step": 6824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9872763156890869,
+      "learning_rate": 1.689936009307737e-05,
+      "loss": 5.2915,
+      "step": 6828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.028903603553772,
+      "learning_rate": 1.687996897420981e-05,
+      "loss": 5.2651,
+      "step": 6832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.095890998840332,
+      "learning_rate": 1.6860577855342255e-05,
+      "loss": 5.4315,
+      "step": 6836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0813144445419312,
+      "learning_rate": 1.6841186736474696e-05,
+      "loss": 5.3725,
+      "step": 6840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0944348573684692,
+      "learning_rate": 1.6821795617607136e-05,
+      "loss": 5.3591,
+      "step": 6844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.023173451423645,
+      "learning_rate": 1.6802404498739577e-05,
+      "loss": 5.3222,
+      "step": 6848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1330089569091797,
+      "learning_rate": 1.678301337987202e-05,
+      "loss": 5.3623,
+      "step": 6852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1375625133514404,
+      "learning_rate": 1.676362226100446e-05,
+      "loss": 5.3356,
+      "step": 6856
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1181647777557373,
+      "learning_rate": 1.6744231142136902e-05,
+      "loss": 5.3939,
+      "step": 6860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0055108070373535,
+      "learning_rate": 1.6724840023269346e-05,
+      "loss": 5.339,
+      "step": 6864
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.001213788986206,
+      "learning_rate": 1.6705448904401786e-05,
+      "loss": 5.4536,
+      "step": 6868
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0213780403137207,
+      "learning_rate": 1.6686057785534227e-05,
+      "loss": 5.3785,
+      "step": 6872
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0076147317886353,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 5.2837,
+      "step": 6876
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0241953134536743,
+      "learning_rate": 1.664727554779911e-05,
+      "loss": 5.3633,
+      "step": 6880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.022125005722046,
+      "learning_rate": 1.6627884428931552e-05,
+      "loss": 5.3523,
+      "step": 6884
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0576834678649902,
+      "learning_rate": 1.6608493310063992e-05,
+      "loss": 5.315,
+      "step": 6888
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0498878955841064,
+      "learning_rate": 1.6589102191196433e-05,
+      "loss": 5.3896,
+      "step": 6892
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.003936529159546,
+      "learning_rate": 1.6569711072328874e-05,
+      "loss": 5.3293,
+      "step": 6896
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9774566292762756,
+      "learning_rate": 1.6550319953461314e-05,
+      "loss": 5.3388,
+      "step": 6900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0642743110656738,
+      "learning_rate": 1.6530928834593755e-05,
+      "loss": 5.3104,
+      "step": 6904
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9880549907684326,
+      "learning_rate": 1.65115377157262e-05,
+      "loss": 5.3273,
+      "step": 6908
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0308412313461304,
+      "learning_rate": 1.649214659685864e-05,
+      "loss": 5.293,
+      "step": 6912
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0004191398620605,
+      "learning_rate": 1.647275547799108e-05,
+      "loss": 5.3624,
+      "step": 6916
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0223098993301392,
+      "learning_rate": 1.645336435912352e-05,
+      "loss": 5.3298,
+      "step": 6920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0578629970550537,
+      "learning_rate": 1.6433973240255964e-05,
+      "loss": 5.3331,
+      "step": 6924
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0649224519729614,
+      "learning_rate": 1.6414582121388405e-05,
+      "loss": 5.4143,
+      "step": 6928
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0141630172729492,
+      "learning_rate": 1.6395191002520845e-05,
+      "loss": 5.2971,
+      "step": 6932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.06792414188385,
+      "learning_rate": 1.6375799883653286e-05,
+      "loss": 5.359,
+      "step": 6936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0014034509658813,
+      "learning_rate": 1.635640876478573e-05,
+      "loss": 5.3262,
+      "step": 6940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0073959827423096,
+      "learning_rate": 1.633701764591817e-05,
+      "loss": 5.2941,
+      "step": 6944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0175294876098633,
+      "learning_rate": 1.631762652705061e-05,
+      "loss": 5.4023,
+      "step": 6948
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0541809797286987,
+      "learning_rate": 1.629823540818305e-05,
+      "loss": 5.3224,
+      "step": 6952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0851467847824097,
+      "learning_rate": 1.6278844289315495e-05,
+      "loss": 5.3333,
+      "step": 6956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.081163763999939,
+      "learning_rate": 1.6259453170447936e-05,
+      "loss": 5.4189,
+      "step": 6960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0276130437850952,
+      "learning_rate": 1.6240062051580376e-05,
+      "loss": 5.2743,
+      "step": 6964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0390230417251587,
+      "learning_rate": 1.6220670932712817e-05,
+      "loss": 5.3315,
+      "step": 6968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9651821851730347,
+      "learning_rate": 1.620127981384526e-05,
+      "loss": 5.33,
+      "step": 6972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.05519437789917,
+      "learning_rate": 1.61818886949777e-05,
+      "loss": 5.3139,
+      "step": 6976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9778598546981812,
+      "learning_rate": 1.6162497576110142e-05,
+      "loss": 5.303,
+      "step": 6980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0078487396240234,
+      "learning_rate": 1.6143106457242586e-05,
+      "loss": 5.3094,
+      "step": 6984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0488457679748535,
+      "learning_rate": 1.6123715338375026e-05,
+      "loss": 5.2953,
+      "step": 6988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0421743392944336,
+      "learning_rate": 1.6104324219507467e-05,
+      "loss": 5.394,
+      "step": 6992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9976141452789307,
+      "learning_rate": 1.6084933100639907e-05,
+      "loss": 5.3734,
+      "step": 6996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0560002326965332,
+      "learning_rate": 1.606554198177235e-05,
+      "loss": 5.3631,
+      "step": 7000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0117374658584595,
+      "learning_rate": 1.6046150862904792e-05,
+      "loss": 5.3586,
+      "step": 7004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0476561784744263,
+      "learning_rate": 1.6026759744037232e-05,
+      "loss": 5.3215,
+      "step": 7008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0643686056137085,
+      "learning_rate": 1.6007368625169673e-05,
+      "loss": 5.2853,
+      "step": 7012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0695873498916626,
+      "learning_rate": 1.5987977506302117e-05,
+      "loss": 5.3231,
+      "step": 7016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0597788095474243,
+      "learning_rate": 1.5968586387434557e-05,
+      "loss": 5.3185,
+      "step": 7020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0414984226226807,
+      "learning_rate": 1.5949195268566998e-05,
+      "loss": 5.3407,
+      "step": 7024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9906083941459656,
+      "learning_rate": 1.592980414969944e-05,
+      "loss": 5.2572,
+      "step": 7028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9810131192207336,
+      "learning_rate": 1.591041303083188e-05,
+      "loss": 5.2424,
+      "step": 7032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0690699815750122,
+      "learning_rate": 1.589102191196432e-05,
+      "loss": 5.3503,
+      "step": 7036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.037235140800476,
+      "learning_rate": 1.5871630793096763e-05,
+      "loss": 5.3903,
+      "step": 7040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9991822838783264,
+      "learning_rate": 1.5852239674229204e-05,
+      "loss": 5.3439,
+      "step": 7044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0834139585494995,
+      "learning_rate": 1.5832848555361644e-05,
+      "loss": 5.2871,
+      "step": 7048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9821889400482178,
+      "learning_rate": 1.5813457436494085e-05,
+      "loss": 5.248,
+      "step": 7052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0037583112716675,
+      "learning_rate": 1.5794066317626526e-05,
+      "loss": 5.2412,
+      "step": 7056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.059874176979065,
+      "learning_rate": 1.577467519875897e-05,
+      "loss": 5.2965,
+      "step": 7060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0708321332931519,
+      "learning_rate": 1.575528407989141e-05,
+      "loss": 5.1961,
+      "step": 7064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0226424932479858,
+      "learning_rate": 1.573589296102385e-05,
+      "loss": 5.3895,
+      "step": 7068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.088083028793335,
+      "learning_rate": 1.571650184215629e-05,
+      "loss": 5.3202,
+      "step": 7072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1663302183151245,
+      "learning_rate": 1.5697110723288735e-05,
+      "loss": 5.2799,
+      "step": 7076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.049932599067688,
+      "learning_rate": 1.5677719604421175e-05,
+      "loss": 5.3477,
+      "step": 7080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0670647621154785,
+      "learning_rate": 1.5658328485553616e-05,
+      "loss": 5.321,
+      "step": 7084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0027042627334595,
+      "learning_rate": 1.5638937366686057e-05,
+      "loss": 5.2589,
+      "step": 7088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0624514818191528,
+      "learning_rate": 1.56195462478185e-05,
+      "loss": 5.3408,
+      "step": 7092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1033272743225098,
+      "learning_rate": 1.560015512895094e-05,
+      "loss": 5.3141,
+      "step": 7096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9922211766242981,
+      "learning_rate": 1.558076401008338e-05,
+      "loss": 5.3127,
+      "step": 7100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1714577674865723,
+      "learning_rate": 1.5561372891215825e-05,
+      "loss": 5.3356,
+      "step": 7104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.092363715171814,
+      "learning_rate": 1.5541981772348266e-05,
+      "loss": 5.3375,
+      "step": 7108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1520874500274658,
+      "learning_rate": 1.5522590653480707e-05,
+      "loss": 5.38,
+      "step": 7112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0429631471633911,
+      "learning_rate": 1.5503199534613147e-05,
+      "loss": 5.2862,
+      "step": 7116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.133023977279663,
+      "learning_rate": 1.548380841574559e-05,
+      "loss": 5.3376,
+      "step": 7120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0379852056503296,
+      "learning_rate": 1.546441729687803e-05,
+      "loss": 5.3617,
+      "step": 7124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9884594082832336,
+      "learning_rate": 1.5445026178010472e-05,
+      "loss": 5.3478,
+      "step": 7128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0757851600646973,
+      "learning_rate": 1.5425635059142913e-05,
+      "loss": 5.3307,
+      "step": 7132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0310724973678589,
+      "learning_rate": 1.5406243940275357e-05,
+      "loss": 5.2984,
+      "step": 7136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.989493727684021,
+      "learning_rate": 1.5386852821407797e-05,
+      "loss": 5.4285,
+      "step": 7140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.950312614440918,
+      "learning_rate": 1.5367461702540238e-05,
+      "loss": 5.2814,
+      "step": 7144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0180491209030151,
+      "learning_rate": 1.5348070583672678e-05,
+      "loss": 5.3137,
+      "step": 7148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0413585901260376,
+      "learning_rate": 1.5328679464805122e-05,
+      "loss": 5.3889,
+      "step": 7152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0932857990264893,
+      "learning_rate": 1.5309288345937563e-05,
+      "loss": 5.4341,
+      "step": 7156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.06010901927948,
+      "learning_rate": 1.5289897227070003e-05,
+      "loss": 5.3442,
+      "step": 7160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0883803367614746,
+      "learning_rate": 1.5270506108202444e-05,
+      "loss": 5.3341,
+      "step": 7164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0533767938613892,
+      "learning_rate": 1.5251114989334886e-05,
+      "loss": 5.2995,
+      "step": 7168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9598619937896729,
+      "learning_rate": 1.5231723870467326e-05,
+      "loss": 5.3465,
+      "step": 7172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0123393535614014,
+      "learning_rate": 1.5212332751599767e-05,
+      "loss": 5.2892,
+      "step": 7176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0960266590118408,
+      "learning_rate": 1.5192941632732211e-05,
+      "loss": 5.332,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1648544073104858,
+      "learning_rate": 1.5173550513864651e-05,
+      "loss": 5.4285,
+      "step": 7184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.051171064376831,
+      "learning_rate": 1.5154159394997092e-05,
+      "loss": 5.4299,
+      "step": 7188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0466375350952148,
+      "learning_rate": 1.5134768276129533e-05,
+      "loss": 5.3216,
+      "step": 7192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9910897612571716,
+      "learning_rate": 1.5115377157261976e-05,
+      "loss": 5.3598,
+      "step": 7196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1121070384979248,
+      "learning_rate": 1.5095986038394417e-05,
+      "loss": 5.3169,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0514254570007324,
+      "learning_rate": 1.5076594919526858e-05,
+      "loss": 5.3502,
+      "step": 7204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0363171100616455,
+      "learning_rate": 1.5057203800659298e-05,
+      "loss": 5.3192,
+      "step": 7208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0386378765106201,
+      "learning_rate": 1.503781268179174e-05,
+      "loss": 5.3244,
+      "step": 7212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0868582725524902,
+      "learning_rate": 1.5018421562924181e-05,
+      "loss": 5.2998,
+      "step": 7216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.106695532798767,
+      "learning_rate": 1.4999030444056621e-05,
+      "loss": 5.2703,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0727300643920898,
+      "learning_rate": 1.4979639325189065e-05,
+      "loss": 5.2557,
+      "step": 7224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0959160327911377,
+      "learning_rate": 1.4960248206321506e-05,
+      "loss": 5.2666,
+      "step": 7228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0157089233398438,
+      "learning_rate": 1.4940857087453946e-05,
+      "loss": 5.3451,
+      "step": 7232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1046866178512573,
+      "learning_rate": 1.4921465968586387e-05,
+      "loss": 5.2349,
+      "step": 7236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0781642198562622,
+      "learning_rate": 1.4902074849718831e-05,
+      "loss": 5.2609,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0019387006759644,
+      "learning_rate": 1.4882683730851271e-05,
+      "loss": 5.3954,
+      "step": 7244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1083266735076904,
+      "learning_rate": 1.4863292611983712e-05,
+      "loss": 5.2636,
+      "step": 7248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2309002876281738,
+      "learning_rate": 1.4843901493116152e-05,
+      "loss": 5.2955,
+      "step": 7252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.087774634361267,
+      "learning_rate": 1.4824510374248596e-05,
+      "loss": 5.3,
+      "step": 7256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.075287938117981,
+      "learning_rate": 1.4805119255381037e-05,
+      "loss": 5.3727,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0246081352233887,
+      "learning_rate": 1.4785728136513477e-05,
+      "loss": 5.3558,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.124543309211731,
+      "learning_rate": 1.4766337017645918e-05,
+      "loss": 5.3379,
+      "step": 7268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.047892689704895,
+      "learning_rate": 1.474694589877836e-05,
+      "loss": 5.3975,
+      "step": 7272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0381947755813599,
+      "learning_rate": 1.47275547799108e-05,
+      "loss": 5.3515,
+      "step": 7276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0230307579040527,
+      "learning_rate": 1.4708163661043243e-05,
+      "loss": 5.2925,
+      "step": 7280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0596458911895752,
+      "learning_rate": 1.4688772542175685e-05,
+      "loss": 5.3191,
+      "step": 7284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1031346321105957,
+      "learning_rate": 1.4669381423308126e-05,
+      "loss": 5.402,
+      "step": 7288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0289580821990967,
+      "learning_rate": 1.4649990304440566e-05,
+      "loss": 5.3957,
+      "step": 7292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1469511985778809,
+      "learning_rate": 1.4630599185573007e-05,
+      "loss": 5.3244,
+      "step": 7296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0669410228729248,
+      "learning_rate": 1.461120806670545e-05,
+      "loss": 5.431,
+      "step": 7300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.05574631690979,
+      "learning_rate": 1.4591816947837891e-05,
+      "loss": 5.3382,
+      "step": 7304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0296452045440674,
+      "learning_rate": 1.4572425828970332e-05,
+      "loss": 5.337,
+      "step": 7308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0180591344833374,
+      "learning_rate": 1.4553034710102772e-05,
+      "loss": 5.3239,
+      "step": 7312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0508371591567993,
+      "learning_rate": 1.4533643591235216e-05,
+      "loss": 5.2944,
+      "step": 7316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0255225896835327,
+      "learning_rate": 1.4514252472367657e-05,
+      "loss": 5.3574,
+      "step": 7320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0599967241287231,
+      "learning_rate": 1.4494861353500097e-05,
+      "loss": 5.3568,
+      "step": 7324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9832557439804077,
+      "learning_rate": 1.4475470234632538e-05,
+      "loss": 5.2891,
+      "step": 7328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0541869401931763,
+      "learning_rate": 1.4456079115764982e-05,
+      "loss": 5.2502,
+      "step": 7332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.069973111152649,
+      "learning_rate": 1.4436687996897422e-05,
+      "loss": 5.3376,
+      "step": 7336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0768502950668335,
+      "learning_rate": 1.4417296878029863e-05,
+      "loss": 5.2392,
+      "step": 7340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.000628113746643,
+      "learning_rate": 1.4397905759162305e-05,
+      "loss": 5.3293,
+      "step": 7344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9960778951644897,
+      "learning_rate": 1.4378514640294746e-05,
+      "loss": 5.2655,
+      "step": 7348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0327279567718506,
+      "learning_rate": 1.4359123521427186e-05,
+      "loss": 5.3477,
+      "step": 7352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0809035301208496,
+      "learning_rate": 1.4339732402559627e-05,
+      "loss": 5.331,
+      "step": 7356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0690807104110718,
+      "learning_rate": 1.432034128369207e-05,
+      "loss": 5.3986,
+      "step": 7360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0245548486709595,
+      "learning_rate": 1.4300950164824511e-05,
+      "loss": 5.3024,
+      "step": 7364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9999493956565857,
+      "learning_rate": 1.4281559045956952e-05,
+      "loss": 5.2878,
+      "step": 7368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0037769079208374,
+      "learning_rate": 1.4262167927089392e-05,
+      "loss": 5.2216,
+      "step": 7372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.002113699913025,
+      "learning_rate": 1.4242776808221836e-05,
+      "loss": 5.4347,
+      "step": 7376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.04192054271698,
+      "learning_rate": 1.4223385689354277e-05,
+      "loss": 5.3812,
+      "step": 7380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0271295309066772,
+      "learning_rate": 1.4203994570486717e-05,
+      "loss": 5.427,
+      "step": 7384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0071477890014648,
+      "learning_rate": 1.4184603451619158e-05,
+      "loss": 5.2727,
+      "step": 7388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0228255987167358,
+      "learning_rate": 1.4165212332751602e-05,
+      "loss": 5.3244,
+      "step": 7392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0034810304641724,
+      "learning_rate": 1.4145821213884042e-05,
+      "loss": 5.3883,
+      "step": 7396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0305688381195068,
+      "learning_rate": 1.4126430095016483e-05,
+      "loss": 5.3851,
+      "step": 7400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.074646234512329,
+      "learning_rate": 1.4107038976148925e-05,
+      "loss": 5.2406,
+      "step": 7404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9961770176887512,
+      "learning_rate": 1.4087647857281366e-05,
+      "loss": 5.2848,
+      "step": 7408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.034627079963684,
+      "learning_rate": 1.4068256738413808e-05,
+      "loss": 5.2984,
+      "step": 7412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.072096824645996,
+      "learning_rate": 1.4048865619546248e-05,
+      "loss": 5.3182,
+      "step": 7416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0899096727371216,
+      "learning_rate": 1.402947450067869e-05,
+      "loss": 5.3017,
+      "step": 7420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1063376665115356,
+      "learning_rate": 1.4010083381811131e-05,
+      "loss": 5.311,
+      "step": 7424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0748202800750732,
+      "learning_rate": 1.3990692262943572e-05,
+      "loss": 5.318,
+      "step": 7428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0851057767868042,
+      "learning_rate": 1.3971301144076012e-05,
+      "loss": 5.3425,
+      "step": 7432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9942495822906494,
+      "learning_rate": 1.3951910025208456e-05,
+      "loss": 5.3258,
+      "step": 7436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0894228219985962,
+      "learning_rate": 1.3932518906340897e-05,
+      "loss": 5.2769,
+      "step": 7440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9603523015975952,
+      "learning_rate": 1.3913127787473337e-05,
+      "loss": 5.4087,
+      "step": 7444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1049365997314453,
+      "learning_rate": 1.3893736668605778e-05,
+      "loss": 5.339,
+      "step": 7448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.021468162536621,
+      "learning_rate": 1.3874345549738222e-05,
+      "loss": 5.3447,
+      "step": 7452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.088437557220459,
+      "learning_rate": 1.3854954430870662e-05,
+      "loss": 5.3546,
+      "step": 7456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0266107320785522,
+      "learning_rate": 1.3835563312003103e-05,
+      "loss": 5.1867,
+      "step": 7460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.088911533355713,
+      "learning_rate": 1.3816172193135547e-05,
+      "loss": 5.3247,
+      "step": 7464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0675928592681885,
+      "learning_rate": 1.3796781074267987e-05,
+      "loss": 5.2807,
+      "step": 7468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9511438608169556,
+      "learning_rate": 1.3777389955400428e-05,
+      "loss": 5.3774,
+      "step": 7472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0546114444732666,
+      "learning_rate": 1.3757998836532868e-05,
+      "loss": 5.2856,
+      "step": 7476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0445231199264526,
+      "learning_rate": 1.373860771766531e-05,
+      "loss": 5.3805,
+      "step": 7480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0493693351745605,
+      "learning_rate": 1.3719216598797751e-05,
+      "loss": 5.3603,
+      "step": 7484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0047773122787476,
+      "learning_rate": 1.3699825479930192e-05,
+      "loss": 5.3224,
+      "step": 7488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0536508560180664,
+      "learning_rate": 1.3680434361062632e-05,
+      "loss": 5.4235,
+      "step": 7492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0347819328308105,
+      "learning_rate": 1.3661043242195076e-05,
+      "loss": 5.2528,
+      "step": 7496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0392394065856934,
+      "learning_rate": 1.3641652123327517e-05,
+      "loss": 5.4127,
+      "step": 7500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0930792093276978,
+      "learning_rate": 1.3622261004459957e-05,
+      "loss": 5.3221,
+      "step": 7504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9962918758392334,
+      "learning_rate": 1.3602869885592398e-05,
+      "loss": 5.3108,
+      "step": 7508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.030331015586853,
+      "learning_rate": 1.3583478766724842e-05,
+      "loss": 5.2632,
+      "step": 7512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.008636236190796,
+      "learning_rate": 1.3564087647857282e-05,
+      "loss": 5.3089,
+      "step": 7516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.006934642791748,
+      "learning_rate": 1.3544696528989723e-05,
+      "loss": 5.2903,
+      "step": 7520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0194462537765503,
+      "learning_rate": 1.3525305410122167e-05,
+      "loss": 5.3827,
+      "step": 7524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9879323840141296,
+      "learning_rate": 1.3505914291254607e-05,
+      "loss": 5.4307,
+      "step": 7528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0651185512542725,
+      "learning_rate": 1.3486523172387048e-05,
+      "loss": 5.3134,
+      "step": 7532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0858125686645508,
+      "learning_rate": 1.3467132053519488e-05,
+      "loss": 5.3024,
+      "step": 7536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0240780115127563,
+      "learning_rate": 1.344774093465193e-05,
+      "loss": 5.3555,
+      "step": 7540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0629985332489014,
+      "learning_rate": 1.3428349815784373e-05,
+      "loss": 5.3652,
+      "step": 7544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1222716569900513,
+      "learning_rate": 1.3408958696916813e-05,
+      "loss": 5.3845,
+      "step": 7548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0008291006088257,
+      "learning_rate": 1.3389567578049254e-05,
+      "loss": 5.3552,
+      "step": 7552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0860753059387207,
+      "learning_rate": 1.3370176459181696e-05,
+      "loss": 5.3044,
+      "step": 7556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0495448112487793,
+      "learning_rate": 1.3350785340314136e-05,
+      "loss": 5.3259,
+      "step": 7560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0432664155960083,
+      "learning_rate": 1.3331394221446577e-05,
+      "loss": 5.3482,
+      "step": 7564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0440526008605957,
+      "learning_rate": 1.3312003102579018e-05,
+      "loss": 5.2362,
+      "step": 7568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0317118167877197,
+      "learning_rate": 1.3292611983711461e-05,
+      "loss": 5.3579,
+      "step": 7572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9613714218139648,
+      "learning_rate": 1.3273220864843902e-05,
+      "loss": 5.3631,
+      "step": 7576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1608860492706299,
+      "learning_rate": 1.3253829745976343e-05,
+      "loss": 5.3547,
+      "step": 7580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.06599760055542,
+      "learning_rate": 1.3234438627108786e-05,
+      "loss": 5.3728,
+      "step": 7584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0115044116973877,
+      "learning_rate": 1.3215047508241227e-05,
+      "loss": 5.2954,
+      "step": 7588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0504167079925537,
+      "learning_rate": 1.3195656389373668e-05,
+      "loss": 5.3378,
+      "step": 7592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1047917604446411,
+      "learning_rate": 1.3176265270506108e-05,
+      "loss": 5.3153,
+      "step": 7596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0862175226211548,
+      "learning_rate": 1.3156874151638552e-05,
+      "loss": 5.4315,
+      "step": 7600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.05397629737854,
+      "learning_rate": 1.3137483032770993e-05,
+      "loss": 5.3807,
+      "step": 7604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0143108367919922,
+      "learning_rate": 1.3118091913903433e-05,
+      "loss": 5.346,
+      "step": 7608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0487464666366577,
+      "learning_rate": 1.3098700795035874e-05,
+      "loss": 5.2966,
+      "step": 7612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.156467318534851,
+      "learning_rate": 1.3079309676168316e-05,
+      "loss": 5.24,
+      "step": 7616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0586912631988525,
+      "learning_rate": 1.3059918557300756e-05,
+      "loss": 5.3162,
+      "step": 7620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0472930669784546,
+      "learning_rate": 1.3040527438433197e-05,
+      "loss": 5.4309,
+      "step": 7624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0239797830581665,
+      "learning_rate": 1.3021136319565637e-05,
+      "loss": 5.3124,
+      "step": 7628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.084915041923523,
+      "learning_rate": 1.3001745200698081e-05,
+      "loss": 5.382,
+      "step": 7632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.121639370918274,
+      "learning_rate": 1.2982354081830522e-05,
+      "loss": 5.2888,
+      "step": 7636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.06790030002594,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 5.3484,
+      "step": 7640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0664012432098389,
+      "learning_rate": 1.2943571844095406e-05,
+      "loss": 5.3624,
+      "step": 7644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.032593846321106,
+      "learning_rate": 1.2924180725227847e-05,
+      "loss": 5.3109,
+      "step": 7648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0182029008865356,
+      "learning_rate": 1.2904789606360287e-05,
+      "loss": 5.3785,
+      "step": 7652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9787065982818604,
+      "learning_rate": 1.2885398487492728e-05,
+      "loss": 5.2586,
+      "step": 7656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.14923095703125,
+      "learning_rate": 1.2866007368625172e-05,
+      "loss": 5.3378,
+      "step": 7660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0064685344696045,
+      "learning_rate": 1.2846616249757612e-05,
+      "loss": 5.3339,
+      "step": 7664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.015594720840454,
+      "learning_rate": 1.2827225130890053e-05,
+      "loss": 5.2412,
+      "step": 7668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1527953147888184,
+      "learning_rate": 1.2807834012022494e-05,
+      "loss": 5.4026,
+      "step": 7672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.024170994758606,
+      "learning_rate": 1.2788442893154936e-05,
+      "loss": 5.2431,
+      "step": 7676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0834672451019287,
+      "learning_rate": 1.2769051774287378e-05,
+      "loss": 5.3018,
+      "step": 7680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.00787353515625,
+      "learning_rate": 1.2749660655419819e-05,
+      "loss": 5.3129,
+      "step": 7684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0804412364959717,
+      "learning_rate": 1.2730269536552259e-05,
+      "loss": 5.3974,
+      "step": 7688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0361813306808472,
+      "learning_rate": 1.2710878417684701e-05,
+      "loss": 5.291,
+      "step": 7692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9639879465103149,
+      "learning_rate": 1.2691487298817142e-05,
+      "loss": 5.3649,
+      "step": 7696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.035768985748291,
+      "learning_rate": 1.2672096179949582e-05,
+      "loss": 5.3159,
+      "step": 7700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.002968430519104,
+      "learning_rate": 1.2652705061082026e-05,
+      "loss": 5.1975,
+      "step": 7704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0666626691818237,
+      "learning_rate": 1.2633313942214467e-05,
+      "loss": 5.2216,
+      "step": 7708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.031752347946167,
+      "learning_rate": 1.2613922823346907e-05,
+      "loss": 5.4174,
+      "step": 7712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0436227321624756,
+      "learning_rate": 1.2594531704479348e-05,
+      "loss": 5.2295,
+      "step": 7716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0823688507080078,
+      "learning_rate": 1.2575140585611792e-05,
+      "loss": 5.3274,
+      "step": 7720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0667200088500977,
+      "learning_rate": 1.2555749466744232e-05,
+      "loss": 5.3149,
+      "step": 7724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0479573011398315,
+      "learning_rate": 1.2536358347876673e-05,
+      "loss": 5.2653,
+      "step": 7728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1973553895950317,
+      "learning_rate": 1.2516967229009113e-05,
+      "loss": 5.2504,
+      "step": 7732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0791184902191162,
+      "learning_rate": 1.2497576110141556e-05,
+      "loss": 5.2797,
+      "step": 7736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.091112732887268,
+      "learning_rate": 1.2478184991273998e-05,
+      "loss": 5.3129,
+      "step": 7740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0607527494430542,
+      "learning_rate": 1.2458793872406438e-05,
+      "loss": 5.2997,
+      "step": 7744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0152305364608765,
+      "learning_rate": 1.243940275353888e-05,
+      "loss": 5.3655,
+      "step": 7748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.009064793586731,
+      "learning_rate": 1.2420011634671321e-05,
+      "loss": 5.3058,
+      "step": 7752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0338549613952637,
+      "learning_rate": 1.2400620515803762e-05,
+      "loss": 5.2527,
+      "step": 7756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0737717151641846,
+      "learning_rate": 1.2381229396936204e-05,
+      "loss": 5.3859,
+      "step": 7760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0524028539657593,
+      "learning_rate": 1.2361838278068645e-05,
+      "loss": 5.407,
+      "step": 7764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0182230472564697,
+      "learning_rate": 1.2342447159201087e-05,
+      "loss": 5.3243,
+      "step": 7768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.042325496673584,
+      "learning_rate": 1.2323056040333527e-05,
+      "loss": 5.372,
+      "step": 7772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.099138855934143,
+      "learning_rate": 1.230366492146597e-05,
+      "loss": 5.3577,
+      "step": 7776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.007309913635254,
+      "learning_rate": 1.228427380259841e-05,
+      "loss": 5.252,
+      "step": 7780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1510696411132812,
+      "learning_rate": 1.2264882683730852e-05,
+      "loss": 5.3726,
+      "step": 7784
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0110242366790771,
+      "learning_rate": 1.2245491564863293e-05,
+      "loss": 5.3932,
+      "step": 7788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0620808601379395,
+      "learning_rate": 1.2226100445995735e-05,
+      "loss": 5.2691,
+      "step": 7792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.052786946296692,
+      "learning_rate": 1.2206709327128176e-05,
+      "loss": 5.3597,
+      "step": 7796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1259844303131104,
+      "learning_rate": 1.2187318208260618e-05,
+      "loss": 5.3324,
+      "step": 7800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0244724750518799,
+      "learning_rate": 1.2167927089393058e-05,
+      "loss": 5.3643,
+      "step": 7804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1225011348724365,
+      "learning_rate": 1.21485359705255e-05,
+      "loss": 5.3501,
+      "step": 7808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9966182112693787,
+      "learning_rate": 1.2129144851657941e-05,
+      "loss": 5.3966,
+      "step": 7812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.102308988571167,
+      "learning_rate": 1.2109753732790383e-05,
+      "loss": 5.2625,
+      "step": 7816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0900803804397583,
+      "learning_rate": 1.2095210393639714e-05,
+      "loss": 5.365,
+      "step": 7820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1339807510375977,
+      "learning_rate": 1.2075819274772154e-05,
+      "loss": 5.3816,
+      "step": 7824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.077379822731018,
+      "learning_rate": 1.2056428155904597e-05,
+      "loss": 5.1332,
+      "step": 7828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1201417446136475,
+      "learning_rate": 1.2037037037037037e-05,
+      "loss": 5.3037,
+      "step": 7832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1047320365905762,
+      "learning_rate": 1.201764591816948e-05,
+      "loss": 5.4453,
+      "step": 7836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0625344514846802,
+      "learning_rate": 1.1998254799301922e-05,
+      "loss": 5.2676,
+      "step": 7840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0774505138397217,
+      "learning_rate": 1.1978863680434362e-05,
+      "loss": 5.3222,
+      "step": 7844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0557003021240234,
+      "learning_rate": 1.1959472561566804e-05,
+      "loss": 5.3235,
+      "step": 7848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9856312274932861,
+      "learning_rate": 1.1940081442699245e-05,
+      "loss": 5.2947,
+      "step": 7852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0185786485671997,
+      "learning_rate": 1.1920690323831685e-05,
+      "loss": 5.3262,
+      "step": 7856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0777360200881958,
+      "learning_rate": 1.1901299204964126e-05,
+      "loss": 5.254,
+      "step": 7860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0238205194473267,
+      "learning_rate": 1.1881908086096568e-05,
+      "loss": 5.356,
+      "step": 7864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0025471448898315,
+      "learning_rate": 1.1862516967229009e-05,
+      "loss": 5.2541,
+      "step": 7868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.030316710472107,
+      "learning_rate": 1.1843125848361451e-05,
+      "loss": 5.2578,
+      "step": 7872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.091535210609436,
+      "learning_rate": 1.1823734729493891e-05,
+      "loss": 5.3234,
+      "step": 7876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0827471017837524,
+      "learning_rate": 1.1804343610626334e-05,
+      "loss": 5.2876,
+      "step": 7880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0991337299346924,
+      "learning_rate": 1.1784952491758774e-05,
+      "loss": 5.473,
+      "step": 7884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0421675443649292,
+      "learning_rate": 1.1765561372891216e-05,
+      "loss": 5.3059,
+      "step": 7888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1006362438201904,
+      "learning_rate": 1.1746170254023657e-05,
+      "loss": 5.2836,
+      "step": 7892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0210435390472412,
+      "learning_rate": 1.17267791351561e-05,
+      "loss": 5.3632,
+      "step": 7896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9662442207336426,
+      "learning_rate": 1.1707388016288541e-05,
+      "loss": 5.3246,
+      "step": 7900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1299954652786255,
+      "learning_rate": 1.1687996897420982e-05,
+      "loss": 5.2499,
+      "step": 7904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0942116975784302,
+      "learning_rate": 1.1668605778553424e-05,
+      "loss": 5.2834,
+      "step": 7908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0262593030929565,
+      "learning_rate": 1.1649214659685865e-05,
+      "loss": 5.3228,
+      "step": 7912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0624995231628418,
+      "learning_rate": 1.1629823540818307e-05,
+      "loss": 5.2603,
+      "step": 7916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0851109027862549,
+      "learning_rate": 1.1610432421950747e-05,
+      "loss": 5.2343,
+      "step": 7920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0987358093261719,
+      "learning_rate": 1.1591041303083188e-05,
+      "loss": 5.2615,
+      "step": 7924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0852704048156738,
+      "learning_rate": 1.1571650184215629e-05,
+      "loss": 5.2886,
+      "step": 7928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0782544612884521,
+      "learning_rate": 1.155225906534807e-05,
+      "loss": 5.3585,
+      "step": 7932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0928773880004883,
+      "learning_rate": 1.1532867946480511e-05,
+      "loss": 5.2083,
+      "step": 7936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9779551029205322,
+      "learning_rate": 1.1513476827612954e-05,
+      "loss": 5.2969,
+      "step": 7940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0859659910202026,
+      "learning_rate": 1.1494085708745394e-05,
+      "loss": 5.3044,
+      "step": 7944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0626839399337769,
+      "learning_rate": 1.1474694589877836e-05,
+      "loss": 5.3455,
+      "step": 7948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1160836219787598,
+      "learning_rate": 1.1455303471010277e-05,
+      "loss": 5.3082,
+      "step": 7952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.089357614517212,
+      "learning_rate": 1.1435912352142719e-05,
+      "loss": 5.3013,
+      "step": 7956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9616773128509521,
+      "learning_rate": 1.1416521233275161e-05,
+      "loss": 5.2995,
+      "step": 7960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0657833814620972,
+      "learning_rate": 1.1397130114407602e-05,
+      "loss": 5.3208,
+      "step": 7964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0845454931259155,
+      "learning_rate": 1.1377738995540044e-05,
+      "loss": 5.3213,
+      "step": 7968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0332484245300293,
+      "learning_rate": 1.1358347876672485e-05,
+      "loss": 5.2833,
+      "step": 7972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0877047777175903,
+      "learning_rate": 1.1338956757804927e-05,
+      "loss": 5.2889,
+      "step": 7976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0782090425491333,
+      "learning_rate": 1.1319565638937367e-05,
+      "loss": 5.2914,
+      "step": 7980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0773468017578125,
+      "learning_rate": 1.130017452006981e-05,
+      "loss": 5.2853,
+      "step": 7984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1031116247177124,
+      "learning_rate": 1.128078340120225e-05,
+      "loss": 5.2285,
+      "step": 7988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0579017400741577,
+      "learning_rate": 1.126139228233469e-05,
+      "loss": 5.2608,
+      "step": 7992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0039610862731934,
+      "learning_rate": 1.1242001163467133e-05,
+      "loss": 5.2801,
+      "step": 7996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1101170778274536,
+      "learning_rate": 1.1222610044599573e-05,
+      "loss": 5.3003,
+      "step": 8000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1231920719146729,
+      "learning_rate": 1.1203218925732014e-05,
+      "loss": 5.2951,
+      "step": 8004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.104062557220459,
+      "learning_rate": 1.1183827806864456e-05,
+      "loss": 5.2419,
+      "step": 8008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0631533861160278,
+      "learning_rate": 1.1164436687996897e-05,
+      "loss": 5.2927,
+      "step": 8012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0923023223876953,
+      "learning_rate": 1.1145045569129339e-05,
+      "loss": 5.3566,
+      "step": 8016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1646323204040527,
+      "learning_rate": 1.1125654450261781e-05,
+      "loss": 5.229,
+      "step": 8020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0775333642959595,
+      "learning_rate": 1.1106263331394222e-05,
+      "loss": 5.3121,
+      "step": 8024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.101682186126709,
+      "learning_rate": 1.1086872212526664e-05,
+      "loss": 5.3746,
+      "step": 8028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.06061589717865,
+      "learning_rate": 1.1067481093659105e-05,
+      "loss": 5.2878,
+      "step": 8032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0500963926315308,
+      "learning_rate": 1.1048089974791547e-05,
+      "loss": 5.3188,
+      "step": 8036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1280819177627563,
+      "learning_rate": 1.1028698855923987e-05,
+      "loss": 5.2829,
+      "step": 8040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0025454759597778,
+      "learning_rate": 1.100930773705643e-05,
+      "loss": 5.3121,
+      "step": 8044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0445626974105835,
+      "learning_rate": 1.098991661818887e-05,
+      "loss": 5.3264,
+      "step": 8048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0949641466140747,
+      "learning_rate": 1.0970525499321312e-05,
+      "loss": 5.3306,
+      "step": 8052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.044668197631836,
+      "learning_rate": 1.0951134380453753e-05,
+      "loss": 5.3335,
+      "step": 8056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.079334020614624,
+      "learning_rate": 1.0931743261586193e-05,
+      "loss": 5.3383,
+      "step": 8060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0334477424621582,
+      "learning_rate": 1.0912352142718636e-05,
+      "loss": 5.3128,
+      "step": 8064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9899519681930542,
+      "learning_rate": 1.0892961023851076e-05,
+      "loss": 5.3245,
+      "step": 8068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0130146741867065,
+      "learning_rate": 1.0873569904983517e-05,
+      "loss": 5.2891,
+      "step": 8072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0125179290771484,
+      "learning_rate": 1.0854178786115959e-05,
+      "loss": 5.3007,
+      "step": 8076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0482258796691895,
+      "learning_rate": 1.0834787667248401e-05,
+      "loss": 5.2417,
+      "step": 8080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.055640697479248,
+      "learning_rate": 1.0815396548380842e-05,
+      "loss": 5.3779,
+      "step": 8084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0579723119735718,
+      "learning_rate": 1.0796005429513284e-05,
+      "loss": 5.232,
+      "step": 8088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.083598017692566,
+      "learning_rate": 1.0776614310645724e-05,
+      "loss": 5.4277,
+      "step": 8092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9834340214729309,
+      "learning_rate": 1.0757223191778167e-05,
+      "loss": 5.2498,
+      "step": 8096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0400574207305908,
+      "learning_rate": 1.0737832072910607e-05,
+      "loss": 5.3321,
+      "step": 8100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.083775281906128,
+      "learning_rate": 1.071844095404305e-05,
+      "loss": 5.1589,
+      "step": 8104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.035290241241455,
+      "learning_rate": 1.069904983517549e-05,
+      "loss": 5.4106,
+      "step": 8108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0348436832427979,
+      "learning_rate": 1.0679658716307932e-05,
+      "loss": 5.2728,
+      "step": 8112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.058597445487976,
+      "learning_rate": 1.0660267597440373e-05,
+      "loss": 5.2081,
+      "step": 8116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0467309951782227,
+      "learning_rate": 1.0640876478572815e-05,
+      "loss": 5.232,
+      "step": 8120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9697101712226868,
+      "learning_rate": 1.0621485359705256e-05,
+      "loss": 5.2641,
+      "step": 8124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0596665143966675,
+      "learning_rate": 1.0602094240837698e-05,
+      "loss": 5.2865,
+      "step": 8128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0948309898376465,
+      "learning_rate": 1.0582703121970138e-05,
+      "loss": 5.2826,
+      "step": 8132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0270111560821533,
+      "learning_rate": 1.0563312003102579e-05,
+      "loss": 5.2726,
+      "step": 8136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0012414455413818,
+      "learning_rate": 1.0543920884235021e-05,
+      "loss": 5.3184,
+      "step": 8140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.019332766532898,
+      "learning_rate": 1.0524529765367462e-05,
+      "loss": 5.2908,
+      "step": 8144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0300483703613281,
+      "learning_rate": 1.0505138646499904e-05,
+      "loss": 5.3333,
+      "step": 8148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0536975860595703,
+      "learning_rate": 1.0485747527632344e-05,
+      "loss": 5.2944,
+      "step": 8152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9881764650344849,
+      "learning_rate": 1.0466356408764787e-05,
+      "loss": 5.2767,
+      "step": 8156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0163639783859253,
+      "learning_rate": 1.0446965289897227e-05,
+      "loss": 5.2723,
+      "step": 8160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0919207334518433,
+      "learning_rate": 1.042757417102967e-05,
+      "loss": 5.3327,
+      "step": 8164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0306917428970337,
+      "learning_rate": 1.040818305216211e-05,
+      "loss": 5.2565,
+      "step": 8168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0125017166137695,
+      "learning_rate": 1.0388791933294552e-05,
+      "loss": 5.342,
+      "step": 8172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0879104137420654,
+      "learning_rate": 1.0369400814426993e-05,
+      "loss": 5.3122,
+      "step": 8176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9527262449264526,
+      "learning_rate": 1.0350009695559435e-05,
+      "loss": 5.284,
+      "step": 8180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0486680269241333,
+      "learning_rate": 1.0330618576691875e-05,
+      "loss": 5.2428,
+      "step": 8184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0278397798538208,
+      "learning_rate": 1.0311227457824318e-05,
+      "loss": 5.3466,
+      "step": 8188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0635344982147217,
+      "learning_rate": 1.0291836338956758e-05,
+      "loss": 5.2645,
+      "step": 8192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0180613994598389,
+      "learning_rate": 1.02724452200892e-05,
+      "loss": 5.3552,
+      "step": 8196
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0265159606933594,
+      "learning_rate": 1.0253054101221641e-05,
+      "loss": 5.3422,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.029842495918274,
+      "learning_rate": 1.0233662982354082e-05,
+      "loss": 5.4009,
+      "step": 8204
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0848013162612915,
+      "learning_rate": 1.0214271863486524e-05,
+      "loss": 5.3982,
+      "step": 8208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0277658700942993,
+      "learning_rate": 1.0194880744618964e-05,
+      "loss": 5.3567,
+      "step": 8212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0273760557174683,
+      "learning_rate": 1.0175489625751407e-05,
+      "loss": 5.2762,
+      "step": 8216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9951087832450867,
+      "learning_rate": 1.0156098506883847e-05,
+      "loss": 5.3267,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0347890853881836,
+      "learning_rate": 1.013670738801629e-05,
+      "loss": 5.2681,
+      "step": 8224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.076242446899414,
+      "learning_rate": 1.011731626914873e-05,
+      "loss": 5.3621,
+      "step": 8228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.02762770652771,
+      "learning_rate": 1.0097925150281172e-05,
+      "loss": 5.3412,
+      "step": 8232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0394989252090454,
+      "learning_rate": 1.0078534031413613e-05,
+      "loss": 5.2555,
+      "step": 8236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0912150144577026,
+      "learning_rate": 1.0059142912546055e-05,
+      "loss": 5.3784,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0681991577148438,
+      "learning_rate": 1.0039751793678495e-05,
+      "loss": 5.3226,
+      "step": 8244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0572381019592285,
+      "learning_rate": 1.0020360674810938e-05,
+      "loss": 5.2935,
+      "step": 8248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0432649850845337,
+      "learning_rate": 1.0000969555943378e-05,
+      "loss": 5.3087,
+      "step": 8252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0084688663482666,
+      "learning_rate": 9.98157843707582e-06,
+      "loss": 5.2778,
+      "step": 8256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.036868691444397,
+      "learning_rate": 9.962187318208263e-06,
+      "loss": 5.31,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0464235544204712,
+      "learning_rate": 9.942796199340703e-06,
+      "loss": 5.3512,
+      "step": 8264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.001470685005188,
+      "learning_rate": 9.923405080473144e-06,
+      "loss": 5.2698,
+      "step": 8268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0910736322402954,
+      "learning_rate": 9.904013961605584e-06,
+      "loss": 5.363,
+      "step": 8272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.087928056716919,
+      "learning_rate": 9.884622842738026e-06,
+      "loss": 5.3454,
+      "step": 8276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.055014967918396,
+      "learning_rate": 9.865231723870467e-06,
+      "loss": 5.3134,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1186180114746094,
+      "learning_rate": 9.84584060500291e-06,
+      "loss": 5.3016,
+      "step": 8284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0159074068069458,
+      "learning_rate": 9.82644948613535e-06,
+      "loss": 5.3974,
+      "step": 8288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1419733762741089,
+      "learning_rate": 9.807058367267792e-06,
+      "loss": 5.333,
+      "step": 8292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.078598976135254,
+      "learning_rate": 9.787667248400232e-06,
+      "loss": 5.2157,
+      "step": 8296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0978525876998901,
+      "learning_rate": 9.768276129532675e-06,
+      "loss": 5.3084,
+      "step": 8300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0307817459106445,
+      "learning_rate": 9.748885010665115e-06,
+      "loss": 5.2962,
+      "step": 8304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9767160415649414,
+      "learning_rate": 9.729493891797557e-06,
+      "loss": 5.3765,
+      "step": 8308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.048744797706604,
+      "learning_rate": 9.710102772929998e-06,
+      "loss": 5.3491,
+      "step": 8312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1452877521514893,
+      "learning_rate": 9.69071165406244e-06,
+      "loss": 5.3348,
+      "step": 8316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0742149353027344,
+      "learning_rate": 9.671320535194882e-06,
+      "loss": 5.4119,
+      "step": 8320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.016554594039917,
+      "learning_rate": 9.651929416327323e-06,
+      "loss": 5.2832,
+      "step": 8324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1286094188690186,
+      "learning_rate": 9.632538297459765e-06,
+      "loss": 5.3436,
+      "step": 8328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0719375610351562,
+      "learning_rate": 9.613147178592206e-06,
+      "loss": 5.2561,
+      "step": 8332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0239946842193604,
+      "learning_rate": 9.593756059724646e-06,
+      "loss": 5.3411,
+      "step": 8336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.16642165184021,
+      "learning_rate": 9.574364940857087e-06,
+      "loss": 5.2173,
+      "step": 8340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.056943655014038,
+      "learning_rate": 9.554973821989529e-06,
+      "loss": 5.3138,
+      "step": 8344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0310717821121216,
+      "learning_rate": 9.53558270312197e-06,
+      "loss": 5.2147,
+      "step": 8348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0939549207687378,
+      "learning_rate": 9.516191584254412e-06,
+      "loss": 5.3849,
+      "step": 8352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0846009254455566,
+      "learning_rate": 9.496800465386852e-06,
+      "loss": 5.2793,
+      "step": 8356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0984148979187012,
+      "learning_rate": 9.477409346519295e-06,
+      "loss": 5.3142,
+      "step": 8360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.035758376121521,
+      "learning_rate": 9.458018227651735e-06,
+      "loss": 5.264,
+      "step": 8364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0837132930755615,
+      "learning_rate": 9.438627108784177e-06,
+      "loss": 5.432,
+      "step": 8368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0333995819091797,
+      "learning_rate": 9.419235989916618e-06,
+      "loss": 5.2622,
+      "step": 8372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.054474949836731,
+      "learning_rate": 9.39984487104906e-06,
+      "loss": 5.2721,
+      "step": 8376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0750809907913208,
+      "learning_rate": 9.380453752181502e-06,
+      "loss": 5.2433,
+      "step": 8380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.054914116859436,
+      "learning_rate": 9.361062633313943e-06,
+      "loss": 5.3407,
+      "step": 8384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1101247072219849,
+      "learning_rate": 9.341671514446385e-06,
+      "loss": 5.2774,
+      "step": 8388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1054069995880127,
+      "learning_rate": 9.322280395578826e-06,
+      "loss": 5.3978,
+      "step": 8392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0813637971878052,
+      "learning_rate": 9.302889276711268e-06,
+      "loss": 5.3735,
+      "step": 8396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.151734709739685,
+      "learning_rate": 9.283498157843708e-06,
+      "loss": 5.4109,
+      "step": 8400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0774791240692139,
+      "learning_rate": 9.264107038976149e-06,
+      "loss": 5.2963,
+      "step": 8404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0468578338623047,
+      "learning_rate": 9.24471592010859e-06,
+      "loss": 5.334,
+      "step": 8408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0247350931167603,
+      "learning_rate": 9.225324801241032e-06,
+      "loss": 5.3307,
+      "step": 8412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.021700143814087,
+      "learning_rate": 9.205933682373472e-06,
+      "loss": 5.3315,
+      "step": 8416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0852890014648438,
+      "learning_rate": 9.186542563505915e-06,
+      "loss": 5.23,
+      "step": 8420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0435699224472046,
+      "learning_rate": 9.167151444638355e-06,
+      "loss": 5.3353,
+      "step": 8424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0124831199645996,
+      "learning_rate": 9.147760325770797e-06,
+      "loss": 5.3257,
+      "step": 8428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0061957836151123,
+      "learning_rate": 9.128369206903238e-06,
+      "loss": 5.3437,
+      "step": 8432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0675718784332275,
+      "learning_rate": 9.10897808803568e-06,
+      "loss": 5.3543,
+      "step": 8436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0940582752227783,
+      "learning_rate": 9.089586969168122e-06,
+      "loss": 5.3205,
+      "step": 8440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0253922939300537,
+      "learning_rate": 9.070195850300563e-06,
+      "loss": 5.2836,
+      "step": 8444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0632801055908203,
+      "learning_rate": 9.050804731433005e-06,
+      "loss": 5.3634,
+      "step": 8448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.151405692100525,
+      "learning_rate": 9.031413612565446e-06,
+      "loss": 5.266,
+      "step": 8452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1215803623199463,
+      "learning_rate": 9.012022493697888e-06,
+      "loss": 5.3605,
+      "step": 8456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0406687259674072,
+      "learning_rate": 8.992631374830328e-06,
+      "loss": 5.3037,
+      "step": 8460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0305143594741821,
+      "learning_rate": 8.97324025596277e-06,
+      "loss": 5.2419,
+      "step": 8464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0159248113632202,
+      "learning_rate": 8.953849137095211e-06,
+      "loss": 5.2576,
+      "step": 8468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0444971323013306,
+      "learning_rate": 8.934458018227652e-06,
+      "loss": 5.2461,
+      "step": 8472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.062738299369812,
+      "learning_rate": 8.915066899360092e-06,
+      "loss": 5.2934,
+      "step": 8476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0445396900177002,
+      "learning_rate": 8.895675780492534e-06,
+      "loss": 5.3011,
+      "step": 8480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1299471855163574,
+      "learning_rate": 8.876284661624975e-06,
+      "loss": 5.1977,
+      "step": 8484
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0241918563842773,
+      "learning_rate": 8.856893542757417e-06,
+      "loss": 5.2821,
+      "step": 8488
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0978903770446777,
+      "learning_rate": 8.837502423889858e-06,
+      "loss": 5.3411,
+      "step": 8492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0630090236663818,
+      "learning_rate": 8.8181113050223e-06,
+      "loss": 5.2487,
+      "step": 8496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0468335151672363,
+      "learning_rate": 8.798720186154742e-06,
+      "loss": 5.2238,
+      "step": 8500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0701797008514404,
+      "learning_rate": 8.779329067287183e-06,
+      "loss": 5.2461,
+      "step": 8504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0644505023956299,
+      "learning_rate": 8.759937948419625e-06,
+      "loss": 5.2462,
+      "step": 8508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0857114791870117,
+      "learning_rate": 8.740546829552066e-06,
+      "loss": 5.3715,
+      "step": 8512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0911511182785034,
+      "learning_rate": 8.721155710684508e-06,
+      "loss": 5.3341,
+      "step": 8516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.066658854484558,
+      "learning_rate": 8.701764591816948e-06,
+      "loss": 5.2962,
+      "step": 8520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0484204292297363,
+      "learning_rate": 8.68237347294939e-06,
+      "loss": 5.2937,
+      "step": 8524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1011348962783813,
+      "learning_rate": 8.662982354081831e-06,
+      "loss": 5.3372,
+      "step": 8528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1982569694519043,
+      "learning_rate": 8.643591235214273e-06,
+      "loss": 5.1987,
+      "step": 8532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1187928915023804,
+      "learning_rate": 8.624200116346714e-06,
+      "loss": 5.3326,
+      "step": 8536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0191898345947266,
+      "learning_rate": 8.604808997479154e-06,
+      "loss": 5.384,
+      "step": 8540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1834492683410645,
+      "learning_rate": 8.585417878611597e-06,
+      "loss": 5.3774,
+      "step": 8544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0459861755371094,
+      "learning_rate": 8.566026759744037e-06,
+      "loss": 5.3309,
+      "step": 8548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.019656777381897,
+      "learning_rate": 8.546635640876478e-06,
+      "loss": 5.3024,
+      "step": 8552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1104713678359985,
+      "learning_rate": 8.52724452200892e-06,
+      "loss": 5.2309,
+      "step": 8556
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0480828285217285,
+      "learning_rate": 8.507853403141362e-06,
+      "loss": 5.256,
+      "step": 8560
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0698785781860352,
+      "learning_rate": 8.488462284273803e-06,
+      "loss": 5.2322,
+      "step": 8564
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0998084545135498,
+      "learning_rate": 8.469071165406245e-06,
+      "loss": 5.2919,
+      "step": 8568
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0241094827651978,
+      "learning_rate": 8.449680046538685e-06,
+      "loss": 5.3195,
+      "step": 8572
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1492643356323242,
+      "learning_rate": 8.430288927671128e-06,
+      "loss": 5.2311,
+      "step": 8576
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1048632860183716,
+      "learning_rate": 8.410897808803568e-06,
+      "loss": 5.3227,
+      "step": 8580
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0702450275421143,
+      "learning_rate": 8.39150668993601e-06,
+      "loss": 5.3173,
+      "step": 8584
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0409200191497803,
+      "learning_rate": 8.372115571068451e-06,
+      "loss": 5.3185,
+      "step": 8588
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.046713948249817,
+      "learning_rate": 8.352724452200893e-06,
+      "loss": 5.3996,
+      "step": 8592
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.039919376373291,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 5.2483,
+      "step": 8596
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0520331859588623,
+      "learning_rate": 8.313942214465776e-06,
+      "loss": 5.3082,
+      "step": 8600
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0794312953948975,
+      "learning_rate": 8.294551095598217e-06,
+      "loss": 5.382,
+      "step": 8604
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0222445726394653,
+      "learning_rate": 8.275159976730657e-06,
+      "loss": 5.2186,
+      "step": 8608
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.043550729751587,
+      "learning_rate": 8.2557688578631e-06,
+      "loss": 5.2447,
+      "step": 8612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.065027117729187,
+      "learning_rate": 8.23637773899554e-06,
+      "loss": 5.3149,
+      "step": 8616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.089449405670166,
+      "learning_rate": 8.216986620127982e-06,
+      "loss": 5.231,
+      "step": 8620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0330005884170532,
+      "learning_rate": 8.197595501260423e-06,
+      "loss": 5.3099,
+      "step": 8624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.088131070137024,
+      "learning_rate": 8.178204382392865e-06,
+      "loss": 5.3941,
+      "step": 8628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0130773782730103,
+      "learning_rate": 8.158813263525305e-06,
+      "loss": 5.2442,
+      "step": 8632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.035882592201233,
+      "learning_rate": 8.139422144657748e-06,
+      "loss": 5.21,
+      "step": 8636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0550565719604492,
+      "learning_rate": 8.120031025790188e-06,
+      "loss": 5.2888,
+      "step": 8640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1436634063720703,
+      "learning_rate": 8.10063990692263e-06,
+      "loss": 5.3587,
+      "step": 8644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1211497783660889,
+      "learning_rate": 8.081248788055071e-06,
+      "loss": 5.3188,
+      "step": 8648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1205918788909912,
+      "learning_rate": 8.061857669187513e-06,
+      "loss": 5.2688,
+      "step": 8652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0907244682312012,
+      "learning_rate": 8.042466550319954e-06,
+      "loss": 5.2518,
+      "step": 8656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0856692790985107,
+      "learning_rate": 8.023075431452396e-06,
+      "loss": 5.3384,
+      "step": 8660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0303173065185547,
+      "learning_rate": 8.003684312584836e-06,
+      "loss": 5.2906,
+      "step": 8664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0908282995224,
+      "learning_rate": 7.984293193717279e-06,
+      "loss": 5.3331,
+      "step": 8668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.088040828704834,
+      "learning_rate": 7.96490207484972e-06,
+      "loss": 5.3521,
+      "step": 8672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0086363554000854,
+      "learning_rate": 7.94551095598216e-06,
+      "loss": 5.3294,
+      "step": 8676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1544169187545776,
+      "learning_rate": 7.926119837114602e-06,
+      "loss": 5.3184,
+      "step": 8680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0530931949615479,
+      "learning_rate": 7.906728718247042e-06,
+      "loss": 5.2866,
+      "step": 8684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.080121397972107,
+      "learning_rate": 7.887337599379485e-06,
+      "loss": 5.2793,
+      "step": 8688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0175666809082031,
+      "learning_rate": 7.867946480511925e-06,
+      "loss": 5.3282,
+      "step": 8692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0283890962600708,
+      "learning_rate": 7.848555361644367e-06,
+      "loss": 5.3439,
+      "step": 8696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0480095148086548,
+      "learning_rate": 7.829164242776808e-06,
+      "loss": 5.3042,
+      "step": 8700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9947773814201355,
+      "learning_rate": 7.80977312390925e-06,
+      "loss": 5.251,
+      "step": 8704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.033823847770691,
+      "learning_rate": 7.79038200504169e-06,
+      "loss": 5.4016,
+      "step": 8708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.129824161529541,
+      "learning_rate": 7.770990886174133e-06,
+      "loss": 5.2176,
+      "step": 8712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0202304124832153,
+      "learning_rate": 7.751599767306574e-06,
+      "loss": 5.3219,
+      "step": 8716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0748639106750488,
+      "learning_rate": 7.732208648439016e-06,
+      "loss": 5.2945,
+      "step": 8720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0026463270187378,
+      "learning_rate": 7.712817529571456e-06,
+      "loss": 5.3474,
+      "step": 8724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.98891282081604,
+      "learning_rate": 7.693426410703899e-06,
+      "loss": 5.2612,
+      "step": 8728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.079750418663025,
+      "learning_rate": 7.674035291836339e-06,
+      "loss": 5.2398,
+      "step": 8732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.057255744934082,
+      "learning_rate": 7.654644172968781e-06,
+      "loss": 5.4364,
+      "step": 8736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0570470094680786,
+      "learning_rate": 7.635253054101222e-06,
+      "loss": 5.2695,
+      "step": 8740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0664699077606201,
+      "learning_rate": 7.615861935233663e-06,
+      "loss": 5.3378,
+      "step": 8744
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0890289545059204,
+      "learning_rate": 7.5964708163661055e-06,
+      "loss": 5.375,
+      "step": 8748
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0607807636260986,
+      "learning_rate": 7.577079697498546e-06,
+      "loss": 5.3258,
+      "step": 8752
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0911225080490112,
+      "learning_rate": 7.557688578630988e-06,
+      "loss": 5.242,
+      "step": 8756
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1005817651748657,
+      "learning_rate": 7.538297459763429e-06,
+      "loss": 5.2734,
+      "step": 8760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0650907754898071,
+      "learning_rate": 7.51890634089587e-06,
+      "loss": 5.2939,
+      "step": 8764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0343334674835205,
+      "learning_rate": 7.499515222028311e-06,
+      "loss": 5.2601,
+      "step": 8768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.092239260673523,
+      "learning_rate": 7.480124103160753e-06,
+      "loss": 5.266,
+      "step": 8772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.140648603439331,
+      "learning_rate": 7.4607329842931935e-06,
+      "loss": 5.2353,
+      "step": 8776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0743423700332642,
+      "learning_rate": 7.441341865425636e-06,
+      "loss": 5.389,
+      "step": 8780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0870285034179688,
+      "learning_rate": 7.421950746558076e-06,
+      "loss": 5.3212,
+      "step": 8784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0318245887756348,
+      "learning_rate": 7.4025596276905185e-06,
+      "loss": 5.2854,
+      "step": 8788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0597593784332275,
+      "learning_rate": 7.383168508822959e-06,
+      "loss": 5.4204,
+      "step": 8792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0621132850646973,
+      "learning_rate": 7.3637773899554e-06,
+      "loss": 5.2903,
+      "step": 8796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.057024598121643,
+      "learning_rate": 7.344386271087843e-06,
+      "loss": 5.3975,
+      "step": 8800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1537240743637085,
+      "learning_rate": 7.324995152220283e-06,
+      "loss": 5.3118,
+      "step": 8804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.161657691001892,
+      "learning_rate": 7.305604033352725e-06,
+      "loss": 5.3516,
+      "step": 8808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0824769735336304,
+      "learning_rate": 7.286212914485166e-06,
+      "loss": 5.3601,
+      "step": 8812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0471476316452026,
+      "learning_rate": 7.266821795617608e-06,
+      "loss": 5.2858,
+      "step": 8816
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0438990592956543,
+      "learning_rate": 7.247430676750049e-06,
+      "loss": 5.2688,
+      "step": 8820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0383694171905518,
+      "learning_rate": 7.228039557882491e-06,
+      "loss": 5.3404,
+      "step": 8824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.039699673652649,
+      "learning_rate": 7.2086484390149315e-06,
+      "loss": 5.2846,
+      "step": 8828
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1256271600723267,
+      "learning_rate": 7.189257320147373e-06,
+      "loss": 5.327,
+      "step": 8832
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9789720177650452,
+      "learning_rate": 7.169866201279813e-06,
+      "loss": 5.3369,
+      "step": 8836
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0387988090515137,
+      "learning_rate": 7.150475082412256e-06,
+      "loss": 5.3402,
+      "step": 8840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0737075805664062,
+      "learning_rate": 7.131083963544696e-06,
+      "loss": 5.2837,
+      "step": 8844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1215327978134155,
+      "learning_rate": 7.111692844677138e-06,
+      "loss": 5.2237,
+      "step": 8848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0533177852630615,
+      "learning_rate": 7.092301725809579e-06,
+      "loss": 5.3783,
+      "step": 8852
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1658439636230469,
+      "learning_rate": 7.072910606942021e-06,
+      "loss": 5.3175,
+      "step": 8856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0966906547546387,
+      "learning_rate": 7.0535194880744625e-06,
+      "loss": 5.4216,
+      "step": 8860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0952768325805664,
+      "learning_rate": 7.034128369206904e-06,
+      "loss": 5.2748,
+      "step": 8864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.096529483795166,
+      "learning_rate": 7.014737250339345e-06,
+      "loss": 5.1666,
+      "step": 8868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0736936330795288,
+      "learning_rate": 6.995346131471786e-06,
+      "loss": 5.24,
+      "step": 8872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0133376121520996,
+      "learning_rate": 6.975955012604228e-06,
+      "loss": 5.2827,
+      "step": 8876
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0580708980560303,
+      "learning_rate": 6.956563893736669e-06,
+      "loss": 5.3801,
+      "step": 8880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1220327615737915,
+      "learning_rate": 6.937172774869111e-06,
+      "loss": 5.2949,
+      "step": 8884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.136806845664978,
+      "learning_rate": 6.917781656001551e-06,
+      "loss": 5.347,
+      "step": 8888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1147714853286743,
+      "learning_rate": 6.898390537133994e-06,
+      "loss": 5.2648,
+      "step": 8892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0525692701339722,
+      "learning_rate": 6.878999418266434e-06,
+      "loss": 5.2665,
+      "step": 8896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0400636196136475,
+      "learning_rate": 6.8596082993988755e-06,
+      "loss": 5.3571,
+      "step": 8900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0699836015701294,
+      "learning_rate": 6.840217180531316e-06,
+      "loss": 5.3079,
+      "step": 8904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0171644687652588,
+      "learning_rate": 6.820826061663758e-06,
+      "loss": 5.3438,
+      "step": 8908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0426756143569946,
+      "learning_rate": 6.801434942796199e-06,
+      "loss": 5.417,
+      "step": 8912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1138461828231812,
+      "learning_rate": 6.782043823928641e-06,
+      "loss": 5.2769,
+      "step": 8916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0219694375991821,
+      "learning_rate": 6.762652705061083e-06,
+      "loss": 5.2811,
+      "step": 8920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1686511039733887,
+      "learning_rate": 6.743261586193524e-06,
+      "loss": 5.3218,
+      "step": 8924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0407147407531738,
+      "learning_rate": 6.723870467325965e-06,
+      "loss": 5.339,
+      "step": 8928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1056681871414185,
+      "learning_rate": 6.704479348458407e-06,
+      "loss": 5.2758,
+      "step": 8932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0969740152359009,
+      "learning_rate": 6.685088229590848e-06,
+      "loss": 5.2432,
+      "step": 8936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9841113090515137,
+      "learning_rate": 6.6656971107232885e-06,
+      "loss": 5.2596,
+      "step": 8940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1172292232513428,
+      "learning_rate": 6.646305991855731e-06,
+      "loss": 5.2884,
+      "step": 8944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9936596155166626,
+      "learning_rate": 6.626914872988171e-06,
+      "loss": 5.2968,
+      "step": 8948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0389301776885986,
+      "learning_rate": 6.6075237541206135e-06,
+      "loss": 5.2827,
+      "step": 8952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.020494818687439,
+      "learning_rate": 6.588132635253054e-06,
+      "loss": 5.3882,
+      "step": 8956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0391160249710083,
+      "learning_rate": 6.568741516385496e-06,
+      "loss": 5.2172,
+      "step": 8960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0213825702667236,
+      "learning_rate": 6.549350397517937e-06,
+      "loss": 5.3307,
+      "step": 8964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0745649337768555,
+      "learning_rate": 6.529959278650378e-06,
+      "loss": 5.2638,
+      "step": 8968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0567609071731567,
+      "learning_rate": 6.510568159782819e-06,
+      "loss": 5.3164,
+      "step": 8972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0450811386108398,
+      "learning_rate": 6.491177040915261e-06,
+      "loss": 5.3648,
+      "step": 8976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0880790948867798,
+      "learning_rate": 6.471785922047703e-06,
+      "loss": 5.3617,
+      "step": 8980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0606417655944824,
+      "learning_rate": 6.452394803180144e-06,
+      "loss": 5.2201,
+      "step": 8984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0124664306640625,
+      "learning_rate": 6.433003684312586e-06,
+      "loss": 5.2499,
+      "step": 8988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1352604627609253,
+      "learning_rate": 6.4136125654450265e-06,
+      "loss": 5.3782,
+      "step": 8992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1061619520187378,
+      "learning_rate": 6.394221446577468e-06,
+      "loss": 5.3408,
+      "step": 8996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1135718822479248,
+      "learning_rate": 6.374830327709909e-06,
+      "loss": 5.2323,
+      "step": 9000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0155010223388672,
+      "learning_rate": 6.355439208842351e-06,
+      "loss": 5.2942,
+      "step": 9004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.044931173324585,
+      "learning_rate": 6.336048089974791e-06,
+      "loss": 5.3267,
+      "step": 9008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0643398761749268,
+      "learning_rate": 6.316656971107233e-06,
+      "loss": 5.2651,
+      "step": 9012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0203381776809692,
+      "learning_rate": 6.297265852239674e-06,
+      "loss": 5.3493,
+      "step": 9016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0563126802444458,
+      "learning_rate": 6.277874733372116e-06,
+      "loss": 5.362,
+      "step": 9020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0667084455490112,
+      "learning_rate": 6.258483614504557e-06,
+      "loss": 5.2151,
+      "step": 9024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0885251760482788,
+      "learning_rate": 6.239092495636999e-06,
+      "loss": 5.3404,
+      "step": 9028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0210630893707275,
+      "learning_rate": 6.21970137676944e-06,
+      "loss": 5.2937,
+      "step": 9032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1303844451904297,
+      "learning_rate": 6.200310257901881e-06,
+      "loss": 5.3373,
+      "step": 9036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1023499965667725,
+      "learning_rate": 6.180919139034322e-06,
+      "loss": 5.2946,
+      "step": 9040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0469759702682495,
+      "learning_rate": 6.161528020166764e-06,
+      "loss": 5.3265,
+      "step": 9044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9917576313018799,
+      "learning_rate": 6.142136901299205e-06,
+      "loss": 5.2984,
+      "step": 9048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0630229711532593,
+      "learning_rate": 6.122745782431646e-06,
+      "loss": 5.2025,
+      "step": 9052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0385984182357788,
+      "learning_rate": 6.103354663564088e-06,
+      "loss": 5.2934,
+      "step": 9056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0480278730392456,
+      "learning_rate": 6.083963544696529e-06,
+      "loss": 5.2933,
+      "step": 9060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0395824909210205,
+      "learning_rate": 6.0645724258289706e-06,
+      "loss": 5.267,
+      "step": 9064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0996421575546265,
+      "learning_rate": 6.045181306961412e-06,
+      "loss": 5.3497,
+      "step": 9068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0920591354370117,
+      "learning_rate": 6.025790188093853e-06,
+      "loss": 5.2752,
+      "step": 9072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0730946063995361,
+      "learning_rate": 6.006399069226295e-06,
+      "loss": 5.3234,
+      "step": 9076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.077646017074585,
+      "learning_rate": 5.987007950358736e-06,
+      "loss": 5.3009,
+      "step": 9080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0901986360549927,
+      "learning_rate": 5.9676168314911775e-06,
+      "loss": 5.2653,
+      "step": 9084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1307499408721924,
+      "learning_rate": 5.948225712623619e-06,
+      "loss": 5.393,
+      "step": 9088
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1002899408340454,
+      "learning_rate": 5.92883459375606e-06,
+      "loss": 5.2785,
+      "step": 9092
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1348730325698853,
+      "learning_rate": 5.909443474888502e-06,
+      "loss": 5.3749,
+      "step": 9096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2404577732086182,
+      "learning_rate": 5.890052356020943e-06,
+      "loss": 5.2968,
+      "step": 9100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1860145330429077,
+      "learning_rate": 5.8706612371533835e-06,
+      "loss": 5.2951,
+      "step": 9104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0747588872909546,
+      "learning_rate": 5.851270118285825e-06,
+      "loss": 5.2725,
+      "step": 9108
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0642809867858887,
+      "learning_rate": 5.831878999418266e-06,
+      "loss": 5.3347,
+      "step": 9112
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0361077785491943,
+      "learning_rate": 5.812487880550708e-06,
+      "loss": 5.2899,
+      "step": 9116
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.027950406074524,
+      "learning_rate": 5.793096761683149e-06,
+      "loss": 5.3016,
+      "step": 9120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0341978073120117,
+      "learning_rate": 5.7737056428155905e-06,
+      "loss": 5.3066,
+      "step": 9124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0980345010757446,
+      "learning_rate": 5.754314523948033e-06,
+      "loss": 5.2685,
+      "step": 9128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0036580562591553,
+      "learning_rate": 5.734923405080474e-06,
+      "loss": 5.2972,
+      "step": 9132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0951625108718872,
+      "learning_rate": 5.715532286212915e-06,
+      "loss": 5.2363,
+      "step": 9136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.010358214378357,
+      "learning_rate": 5.696141167345356e-06,
+      "loss": 5.2715,
+      "step": 9140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0721516609191895,
+      "learning_rate": 5.676750048477797e-06,
+      "loss": 5.3006,
+      "step": 9144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9922645688056946,
+      "learning_rate": 5.657358929610239e-06,
+      "loss": 5.2985,
+      "step": 9148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0958447456359863,
+      "learning_rate": 5.63796781074268e-06,
+      "loss": 5.2388,
+      "step": 9152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9977266788482666,
+      "learning_rate": 5.6185766918751215e-06,
+      "loss": 5.4273,
+      "step": 9156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1025915145874023,
+      "learning_rate": 5.599185573007563e-06,
+      "loss": 5.3768,
+      "step": 9160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0521866083145142,
+      "learning_rate": 5.579794454140004e-06,
+      "loss": 5.2555,
+      "step": 9164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0546320676803589,
+      "learning_rate": 5.560403335272446e-06,
+      "loss": 5.259,
+      "step": 9168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.084153413772583,
+      "learning_rate": 5.541012216404887e-06,
+      "loss": 5.2733,
+      "step": 9172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1298420429229736,
+      "learning_rate": 5.521621097537328e-06,
+      "loss": 5.2499,
+      "step": 9176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0296047925949097,
+      "learning_rate": 5.502229978669769e-06,
+      "loss": 5.3293,
+      "step": 9180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0231281518936157,
+      "learning_rate": 5.48283885980221e-06,
+      "loss": 5.2614,
+      "step": 9184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1063759326934814,
+      "learning_rate": 5.463447740934653e-06,
+      "loss": 5.3569,
+      "step": 9188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0762827396392822,
+      "learning_rate": 5.444056622067094e-06,
+      "loss": 5.3734,
+      "step": 9192
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0667394399642944,
+      "learning_rate": 5.424665503199535e-06,
+      "loss": 5.3136,
+      "step": 9196
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0663567781448364,
+      "learning_rate": 5.405274384331977e-06,
+      "loss": 5.2178,
+      "step": 9200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1014020442962646,
+      "learning_rate": 5.385883265464417e-06,
+      "loss": 5.3121,
+      "step": 9204
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0944761037826538,
+      "learning_rate": 5.366492146596859e-06,
+      "loss": 5.3321,
+      "step": 9208
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0576825141906738,
+      "learning_rate": 5.3471010277293e-06,
+      "loss": 5.187,
+      "step": 9212
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.102414608001709,
+      "learning_rate": 5.3277099088617414e-06,
+      "loss": 5.2198,
+      "step": 9216
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0515443086624146,
+      "learning_rate": 5.308318789994183e-06,
+      "loss": 5.3509,
+      "step": 9220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0225639343261719,
+      "learning_rate": 5.288927671126624e-06,
+      "loss": 5.3264,
+      "step": 9224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0482409000396729,
+      "learning_rate": 5.269536552259066e-06,
+      "loss": 5.314,
+      "step": 9228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0829366445541382,
+      "learning_rate": 5.250145433391507e-06,
+      "loss": 5.3288,
+      "step": 9232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.018813967704773,
+      "learning_rate": 5.230754314523948e-06,
+      "loss": 5.245,
+      "step": 9236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1042280197143555,
+      "learning_rate": 5.21136319565639e-06,
+      "loss": 5.3871,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0457403659820557,
+      "learning_rate": 5.19197207678883e-06,
+      "loss": 5.3858,
+      "step": 9244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0873547792434692,
+      "learning_rate": 5.1725809579212725e-06,
+      "loss": 5.2907,
+      "step": 9248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0615798234939575,
+      "learning_rate": 5.153189839053714e-06,
+      "loss": 5.2729,
+      "step": 9252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1086784601211548,
+      "learning_rate": 5.133798720186155e-06,
+      "loss": 5.388,
+      "step": 9256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0895764827728271,
+      "learning_rate": 5.114407601318597e-06,
+      "loss": 5.2892,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0369954109191895,
+      "learning_rate": 5.095016482451038e-06,
+      "loss": 5.2989,
+      "step": 9264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.049892544746399,
+      "learning_rate": 5.0756253635834794e-06,
+      "loss": 5.3419,
+      "step": 9268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0920242071151733,
+      "learning_rate": 5.056234244715921e-06,
+      "loss": 5.3116,
+      "step": 9272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0999177694320679,
+      "learning_rate": 5.036843125848361e-06,
+      "loss": 5.2779,
+      "step": 9276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0239474773406982,
+      "learning_rate": 5.017452006980803e-06,
+      "loss": 5.35,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0571128129959106,
+      "learning_rate": 4.998060888113244e-06,
+      "loss": 5.3139,
+      "step": 9284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1082773208618164,
+      "learning_rate": 4.9786697692456855e-06,
+      "loss": 5.2779,
+      "step": 9288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0475021600723267,
+      "learning_rate": 4.959278650378127e-06,
+      "loss": 5.1911,
+      "step": 9292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0756545066833496,
+      "learning_rate": 4.939887531510568e-06,
+      "loss": 5.2851,
+      "step": 9296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9714142680168152,
+      "learning_rate": 4.92049641264301e-06,
+      "loss": 5.2584,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0212279558181763,
+      "learning_rate": 4.901105293775451e-06,
+      "loss": 5.2914,
+      "step": 9304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.038690209388733,
+      "learning_rate": 4.8817141749078924e-06,
+      "loss": 5.3318,
+      "step": 9308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0438801050186157,
+      "learning_rate": 4.862323056040334e-06,
+      "loss": 5.3556,
+      "step": 9312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0903195142745972,
+      "learning_rate": 4.842931937172775e-06,
+      "loss": 5.2413,
+      "step": 9316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.029875636100769,
+      "learning_rate": 4.823540818305217e-06,
+      "loss": 5.2558,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0729329586029053,
+      "learning_rate": 4.804149699437658e-06,
+      "loss": 5.2387,
+      "step": 9324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.012494683265686,
+      "learning_rate": 4.784758580570099e-06,
+      "loss": 5.2488,
+      "step": 9328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0358749628067017,
+      "learning_rate": 4.765367461702541e-06,
+      "loss": 5.2961,
+      "step": 9332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0714225769042969,
+      "learning_rate": 4.745976342834982e-06,
+      "loss": 5.2666,
+      "step": 9336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0506253242492676,
+      "learning_rate": 4.7265852239674235e-06,
+      "loss": 5.2997,
+      "step": 9340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.067970633506775,
+      "learning_rate": 4.707194105099864e-06,
+      "loss": 5.2415,
+      "step": 9344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1353143453598022,
+      "learning_rate": 4.687802986232305e-06,
+      "loss": 5.2687,
+      "step": 9348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0400621891021729,
+      "learning_rate": 4.668411867364747e-06,
+      "loss": 5.269,
+      "step": 9352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1021333932876587,
+      "learning_rate": 4.649020748497188e-06,
+      "loss": 5.3425,
+      "step": 9356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0333983898162842,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 5.3178,
+      "step": 9360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1346157789230347,
+      "learning_rate": 4.610238510762071e-06,
+      "loss": 5.2779,
+      "step": 9364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1103806495666504,
+      "learning_rate": 4.590847391894513e-06,
+      "loss": 5.3102,
+      "step": 9368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.109861135482788,
+      "learning_rate": 4.5714562730269546e-06,
+      "loss": 5.2864,
+      "step": 9372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0895229578018188,
+      "learning_rate": 4.552065154159395e-06,
+      "loss": 5.3538,
+      "step": 9376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0834227800369263,
+      "learning_rate": 4.5326740352918365e-06,
+      "loss": 5.328,
+      "step": 9380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0530760288238525,
+      "learning_rate": 4.513282916424278e-06,
+      "loss": 5.3337,
+      "step": 9384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0888159275054932,
+      "learning_rate": 4.493891797556719e-06,
+      "loss": 5.2769,
+      "step": 9388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9955815076828003,
+      "learning_rate": 4.474500678689161e-06,
+      "loss": 5.3042,
+      "step": 9392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1201573610305786,
+      "learning_rate": 4.455109559821602e-06,
+      "loss": 5.3797,
+      "step": 9396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0425370931625366,
+      "learning_rate": 4.435718440954043e-06,
+      "loss": 5.2661,
+      "step": 9400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.10947847366333,
+      "learning_rate": 4.416327322086485e-06,
+      "loss": 5.3799,
+      "step": 9404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.088013768196106,
+      "learning_rate": 4.396936203218926e-06,
+      "loss": 5.4089,
+      "step": 9408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0372400283813477,
+      "learning_rate": 4.377545084351367e-06,
+      "loss": 5.2819,
+      "step": 9412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.030407428741455,
+      "learning_rate": 4.358153965483808e-06,
+      "loss": 5.3247,
+      "step": 9416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0526556968688965,
+      "learning_rate": 4.3387628466162495e-06,
+      "loss": 5.243,
+      "step": 9420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0113731622695923,
+      "learning_rate": 4.319371727748691e-06,
+      "loss": 5.3229,
+      "step": 9424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1396671533584595,
+      "learning_rate": 4.299980608881133e-06,
+      "loss": 5.2698,
+      "step": 9428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.066570520401001,
+      "learning_rate": 4.2805894900135745e-06,
+      "loss": 5.3542,
+      "step": 9432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0811322927474976,
+      "learning_rate": 4.261198371146016e-06,
+      "loss": 5.2388,
+      "step": 9436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0718379020690918,
+      "learning_rate": 4.241807252278457e-06,
+      "loss": 5.2413,
+      "step": 9440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0602755546569824,
+      "learning_rate": 4.222416133410898e-06,
+      "loss": 5.299,
+      "step": 9444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0844111442565918,
+      "learning_rate": 4.203025014543339e-06,
+      "loss": 5.407,
+      "step": 9448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0383292436599731,
+      "learning_rate": 4.1836338956757805e-06,
+      "loss": 5.3359,
+      "step": 9452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0988202095031738,
+      "learning_rate": 4.164242776808222e-06,
+      "loss": 5.4107,
+      "step": 9456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1423723697662354,
+      "learning_rate": 4.144851657940663e-06,
+      "loss": 5.3404,
+      "step": 9460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1802666187286377,
+      "learning_rate": 4.125460539073105e-06,
+      "loss": 5.247,
+      "step": 9464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9723669290542603,
+      "learning_rate": 4.106069420205546e-06,
+      "loss": 5.2306,
+      "step": 9468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0496985912322998,
+      "learning_rate": 4.0866783013379875e-06,
+      "loss": 5.2929,
+      "step": 9472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0715490579605103,
+      "learning_rate": 4.067287182470429e-06,
+      "loss": 5.2874,
+      "step": 9476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0244431495666504,
+      "learning_rate": 4.04789606360287e-06,
+      "loss": 5.3395,
+      "step": 9480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0559965372085571,
+      "learning_rate": 4.028504944735311e-06,
+      "loss": 5.4143,
+      "step": 9484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0647709369659424,
+      "learning_rate": 4.009113825867752e-06,
+      "loss": 5.3425,
+      "step": 9488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0821641683578491,
+      "learning_rate": 3.989722707000194e-06,
+      "loss": 5.2985,
+      "step": 9492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0766880512237549,
+      "learning_rate": 3.970331588132636e-06,
+      "loss": 5.3269,
+      "step": 9496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0378237962722778,
+      "learning_rate": 3.950940469265077e-06,
+      "loss": 5.2643,
+      "step": 9500
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1061269044876099,
+      "learning_rate": 3.9315493503975185e-06,
+      "loss": 5.3131,
+      "step": 9504
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0113786458969116,
+      "learning_rate": 3.91215823152996e-06,
+      "loss": 5.3312,
+      "step": 9508
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0934914350509644,
+      "learning_rate": 3.8927671126624005e-06,
+      "loss": 5.2506,
+      "step": 9512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.024383544921875,
+      "learning_rate": 3.873375993794842e-06,
+      "loss": 5.3126,
+      "step": 9516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0432902574539185,
+      "learning_rate": 3.853984874927283e-06,
+      "loss": 5.2686,
+      "step": 9520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0603485107421875,
+      "learning_rate": 3.834593756059725e-06,
+      "loss": 5.2765,
+      "step": 9524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0203194618225098,
+      "learning_rate": 3.815202637192166e-06,
+      "loss": 5.2347,
+      "step": 9528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0024125576019287,
+      "learning_rate": 3.7958115183246074e-06,
+      "loss": 5.2144,
+      "step": 9532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.098630666732788,
+      "learning_rate": 3.7764203994570488e-06,
+      "loss": 5.327,
+      "step": 9536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0508458614349365,
+      "learning_rate": 3.75702928058949e-06,
+      "loss": 5.2154,
+      "step": 9540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0391416549682617,
+      "learning_rate": 3.737638161721931e-06,
+      "loss": 5.3172,
+      "step": 9544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0499507188796997,
+      "learning_rate": 3.7182470428543725e-06,
+      "loss": 5.3587,
+      "step": 9548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1225402355194092,
+      "learning_rate": 3.6988559239868147e-06,
+      "loss": 5.2066,
+      "step": 9552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0803992748260498,
+      "learning_rate": 3.6794648051192557e-06,
+      "loss": 5.2294,
+      "step": 9556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0833008289337158,
+      "learning_rate": 3.660073686251697e-06,
+      "loss": 5.3644,
+      "step": 9560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.069279670715332,
+      "learning_rate": 3.6406825673841384e-06,
+      "loss": 5.2461,
+      "step": 9564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.013576626777649,
+      "learning_rate": 3.62129144851658e-06,
+      "loss": 5.3483,
+      "step": 9568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.082595944404602,
+      "learning_rate": 3.601900329649021e-06,
+      "loss": 5.2905,
+      "step": 9572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0759378671646118,
+      "learning_rate": 3.582509210781462e-06,
+      "loss": 5.3518,
+      "step": 9576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.070561170578003,
+      "learning_rate": 3.5631180919139036e-06,
+      "loss": 5.3375,
+      "step": 9580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0099434852600098,
+      "learning_rate": 3.543726973046345e-06,
+      "loss": 5.3133,
+      "step": 9584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1549813747406006,
+      "learning_rate": 3.5243358541787863e-06,
+      "loss": 5.3642,
+      "step": 9588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.024865746498108,
+      "learning_rate": 3.5049447353112273e-06,
+      "loss": 5.3313,
+      "step": 9592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.074829339981079,
+      "learning_rate": 3.4855536164436687e-06,
+      "loss": 5.3294,
+      "step": 9596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0802452564239502,
+      "learning_rate": 3.46616249757611e-06,
+      "loss": 5.3024,
+      "step": 9600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1103779077529907,
+      "learning_rate": 3.4467713787085514e-06,
+      "loss": 5.259,
+      "step": 9604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.122125506401062,
+      "learning_rate": 3.427380259840993e-06,
+      "loss": 5.388,
+      "step": 9608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.106959581375122,
+      "learning_rate": 3.4079891409734346e-06,
+      "loss": 5.344,
+      "step": 9612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.15294349193573,
+      "learning_rate": 3.388598022105876e-06,
+      "loss": 5.3002,
+      "step": 9616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.124644160270691,
+      "learning_rate": 3.3692069032383174e-06,
+      "loss": 5.406,
+      "step": 9620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9895834922790527,
+      "learning_rate": 3.3498157843707584e-06,
+      "loss": 5.3485,
+      "step": 9624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.155667781829834,
+      "learning_rate": 3.3304246655031997e-06,
+      "loss": 5.3161,
+      "step": 9628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0220144987106323,
+      "learning_rate": 3.311033546635641e-06,
+      "loss": 5.3078,
+      "step": 9632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0805678367614746,
+      "learning_rate": 3.2916424277680825e-06,
+      "loss": 5.3262,
+      "step": 9636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0154497623443604,
+      "learning_rate": 3.272251308900524e-06,
+      "loss": 5.2922,
+      "step": 9640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0291320085525513,
+      "learning_rate": 3.252860190032965e-06,
+      "loss": 5.2613,
+      "step": 9644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0393972396850586,
+      "learning_rate": 3.2334690711654062e-06,
+      "loss": 5.3226,
+      "step": 9648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0496784448623657,
+      "learning_rate": 3.2140779522978476e-06,
+      "loss": 5.3214,
+      "step": 9652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1420729160308838,
+      "learning_rate": 3.194686833430289e-06,
+      "loss": 5.3446,
+      "step": 9656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1252871751785278,
+      "learning_rate": 3.17529571456273e-06,
+      "loss": 5.2488,
+      "step": 9660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0620160102844238,
+      "learning_rate": 3.1559045956951714e-06,
+      "loss": 5.3523,
+      "step": 9664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.14591646194458,
+      "learning_rate": 3.1365134768276127e-06,
+      "loss": 5.277,
+      "step": 9668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0520758628845215,
+      "learning_rate": 3.1171223579600545e-06,
+      "loss": 5.2889,
+      "step": 9672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0912026166915894,
+      "learning_rate": 3.0977312390924955e-06,
+      "loss": 5.3545,
+      "step": 9676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0775065422058105,
+      "learning_rate": 3.078340120224937e-06,
+      "loss": 5.3119,
+      "step": 9680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0335484743118286,
+      "learning_rate": 3.0589490013573787e-06,
+      "loss": 5.3732,
+      "step": 9684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0346179008483887,
+      "learning_rate": 3.03955788248982e-06,
+      "loss": 5.2468,
+      "step": 9688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.091991662979126,
+      "learning_rate": 3.020166763622261e-06,
+      "loss": 5.3091,
+      "step": 9692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.07473886013031,
+      "learning_rate": 3.0007756447547024e-06,
+      "loss": 5.2884,
+      "step": 9696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.125009298324585,
+      "learning_rate": 2.981384525887144e-06,
+      "loss": 5.1994,
+      "step": 9700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2514116764068604,
+      "learning_rate": 2.961993407019585e-06,
+      "loss": 5.3651,
+      "step": 9704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1177825927734375,
+      "learning_rate": 2.9426022881520266e-06,
+      "loss": 5.3326,
+      "step": 9708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9657092690467834,
+      "learning_rate": 2.9232111692844675e-06,
+      "loss": 5.2113,
+      "step": 9712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1035900115966797,
+      "learning_rate": 2.9038200504169093e-06,
+      "loss": 5.3596,
+      "step": 9716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0969198942184448,
+      "learning_rate": 2.8844289315493507e-06,
+      "loss": 5.2202,
+      "step": 9720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0725688934326172,
+      "learning_rate": 2.865037812681792e-06,
+      "loss": 5.3028,
+      "step": 9724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0617296695709229,
+      "learning_rate": 2.845646693814233e-06,
+      "loss": 5.2939,
+      "step": 9728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0573393106460571,
+      "learning_rate": 2.8262555749466745e-06,
+      "loss": 5.2684,
+      "step": 9732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0836176872253418,
+      "learning_rate": 2.806864456079116e-06,
+      "loss": 5.2659,
+      "step": 9736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.157784342765808,
+      "learning_rate": 2.7874733372115572e-06,
+      "loss": 5.326,
+      "step": 9740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1517001390457153,
+      "learning_rate": 2.7680822183439986e-06,
+      "loss": 5.3198,
+      "step": 9744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0895476341247559,
+      "learning_rate": 2.74869109947644e-06,
+      "loss": 5.2795,
+      "step": 9748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0055677890777588,
+      "learning_rate": 2.7292999806088814e-06,
+      "loss": 5.2201,
+      "step": 9752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0537750720977783,
+      "learning_rate": 2.7099088617413228e-06,
+      "loss": 5.2557,
+      "step": 9756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1611661911010742,
+      "learning_rate": 2.690517742873764e-06,
+      "loss": 5.2528,
+      "step": 9760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0563558340072632,
+      "learning_rate": 2.671126624006205e-06,
+      "loss": 5.3437,
+      "step": 9764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.08133065700531,
+      "learning_rate": 2.6517355051386465e-06,
+      "loss": 5.278,
+      "step": 9768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0893269777297974,
+      "learning_rate": 2.632344386271088e-06,
+      "loss": 5.4056,
+      "step": 9772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0335135459899902,
+      "learning_rate": 2.6129532674035292e-06,
+      "loss": 5.2513,
+      "step": 9776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0350239276885986,
+      "learning_rate": 2.5935621485359706e-06,
+      "loss": 5.2965,
+      "step": 9780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1059588193893433,
+      "learning_rate": 2.574171029668412e-06,
+      "loss": 5.291,
+      "step": 9784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.109610915184021,
+      "learning_rate": 2.5547799108008534e-06,
+      "loss": 5.3289,
+      "step": 9788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0264822244644165,
+      "learning_rate": 2.5353887919332948e-06,
+      "loss": 5.3212,
+      "step": 9792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0764135122299194,
+      "learning_rate": 2.5159976730657357e-06,
+      "loss": 5.4082,
+      "step": 9796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0444653034210205,
+      "learning_rate": 2.496606554198177e-06,
+      "loss": 5.2801,
+      "step": 9800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.23245108127594,
+      "learning_rate": 2.477215435330619e-06,
+      "loss": 5.4502,
+      "step": 9804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0822020769119263,
+      "learning_rate": 2.4578243164630603e-06,
+      "loss": 5.2585,
+      "step": 9808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0727524757385254,
+      "learning_rate": 2.4384331975955013e-06,
+      "loss": 5.3905,
+      "step": 9812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1105926036834717,
+      "learning_rate": 2.4190420787279427e-06,
+      "loss": 5.3339,
+      "step": 9816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0885437726974487,
+      "learning_rate": 2.399650959860384e-06,
+      "loss": 5.277,
+      "step": 9820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0500329732894897,
+      "learning_rate": 2.3802598409928254e-06,
+      "loss": 5.3009,
+      "step": 9824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0552301406860352,
+      "learning_rate": 2.360868722125267e-06,
+      "loss": 5.3666,
+      "step": 9828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1031043529510498,
+      "learning_rate": 2.3414776032577078e-06,
+      "loss": 5.3025,
+      "step": 9832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0527067184448242,
+      "learning_rate": 2.3220864843901496e-06,
+      "loss": 5.2945,
+      "step": 9836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0818541049957275,
+      "learning_rate": 2.302695365522591e-06,
+      "loss": 5.3044,
+      "step": 9840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.071460485458374,
+      "learning_rate": 2.2833042466550324e-06,
+      "loss": 5.3087,
+      "step": 9844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1150330305099487,
+      "learning_rate": 2.2639131277874733e-06,
+      "loss": 5.2367,
+      "step": 9848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1372606754302979,
+      "learning_rate": 2.2445220089199147e-06,
+      "loss": 5.3709,
+      "step": 9852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1396896839141846,
+      "learning_rate": 2.225130890052356e-06,
+      "loss": 5.2739,
+      "step": 9856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0782867670059204,
+      "learning_rate": 2.2057397711847975e-06,
+      "loss": 5.3314,
+      "step": 9860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0799883604049683,
+      "learning_rate": 2.186348652317239e-06,
+      "loss": 5.2398,
+      "step": 9864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.063509464263916,
+      "learning_rate": 2.1669575334496802e-06,
+      "loss": 5.3209,
+      "step": 9868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.074567437171936,
+      "learning_rate": 2.1475664145821216e-06,
+      "loss": 5.3116,
+      "step": 9872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0963295698165894,
+      "learning_rate": 2.128175295714563e-06,
+      "loss": 5.2356,
+      "step": 9876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0509377717971802,
+      "learning_rate": 2.108784176847004e-06,
+      "loss": 5.3888,
+      "step": 9880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0329670906066895,
+      "learning_rate": 2.0893930579794453e-06,
+      "loss": 5.2065,
+      "step": 9884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0395572185516357,
+      "learning_rate": 2.0700019391118867e-06,
+      "loss": 5.2347,
+      "step": 9888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0882209539413452,
+      "learning_rate": 2.050610820244328e-06,
+      "loss": 5.3839,
+      "step": 9892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0747255086898804,
+      "learning_rate": 2.0312197013767695e-06,
+      "loss": 5.3768,
+      "step": 9896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0426535606384277,
+      "learning_rate": 2.011828582509211e-06,
+      "loss": 5.2553,
+      "step": 9900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9872206449508667,
+      "learning_rate": 1.9924374636416523e-06,
+      "loss": 5.246,
+      "step": 9904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0624208450317383,
+      "learning_rate": 1.9730463447740936e-06,
+      "loss": 5.2859,
+      "step": 9908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0884760618209839,
+      "learning_rate": 1.953655225906535e-06,
+      "loss": 5.3412,
+      "step": 9912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0986016988754272,
+      "learning_rate": 1.934264107038976e-06,
+      "loss": 5.2388,
+      "step": 9916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0865901708602905,
+      "learning_rate": 1.9148729881714174e-06,
+      "loss": 5.2769,
+      "step": 9920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0592422485351562,
+      "learning_rate": 1.8954818693038588e-06,
+      "loss": 5.3446,
+      "step": 9924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0826843976974487,
+      "learning_rate": 1.8760907504363004e-06,
+      "loss": 5.254,
+      "step": 9928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1321392059326172,
+      "learning_rate": 1.8566996315687417e-06,
+      "loss": 5.266,
+      "step": 9932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0519055128097534,
+      "learning_rate": 1.837308512701183e-06,
+      "loss": 5.3199,
+      "step": 9936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0540200471878052,
+      "learning_rate": 1.8179173938336243e-06,
+      "loss": 5.3855,
+      "step": 9940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0404717922210693,
+      "learning_rate": 1.7985262749660657e-06,
+      "loss": 5.3105,
+      "step": 9944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.076065182685852,
+      "learning_rate": 1.7791351560985068e-06,
+      "loss": 5.3344,
+      "step": 9948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0112926959991455,
+      "learning_rate": 1.7597440372309482e-06,
+      "loss": 5.4095,
+      "step": 9952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0115467309951782,
+      "learning_rate": 1.7403529183633898e-06,
+      "loss": 5.2626,
+      "step": 9956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0050586462020874,
+      "learning_rate": 1.720961799495831e-06,
+      "loss": 5.346,
+      "step": 9960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0998096466064453,
+      "learning_rate": 1.7015706806282724e-06,
+      "loss": 5.2961,
+      "step": 9964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0492075681686401,
+      "learning_rate": 1.6821795617607138e-06,
+      "loss": 5.3103,
+      "step": 9968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1068048477172852,
+      "learning_rate": 1.662788442893155e-06,
+      "loss": 5.3157,
+      "step": 9972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0755717754364014,
+      "learning_rate": 1.6433973240255963e-06,
+      "loss": 5.3995,
+      "step": 9976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0472952127456665,
+      "learning_rate": 1.6240062051580375e-06,
+      "loss": 5.2634,
+      "step": 9980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0923290252685547,
+      "learning_rate": 1.6046150862904789e-06,
+      "loss": 5.267,
+      "step": 9984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0469489097595215,
+      "learning_rate": 1.5852239674229205e-06,
+      "loss": 5.2445,
+      "step": 9988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.055059552192688,
+      "learning_rate": 1.5658328485553619e-06,
+      "loss": 5.2376,
+      "step": 9992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9697688817977905,
+      "learning_rate": 1.546441729687803e-06,
+      "loss": 5.2468,
+      "step": 9996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0974935293197632,
+      "learning_rate": 1.5270506108202444e-06,
+      "loss": 5.3911,
+      "step": 10000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0364536046981812,
+      "learning_rate": 1.5076594919526856e-06,
+      "loss": 5.171,
+      "step": 10004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0847764015197754,
+      "learning_rate": 1.4882683730851272e-06,
+      "loss": 5.3467,
+      "step": 10008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0459425449371338,
+      "learning_rate": 1.4688772542175684e-06,
+      "loss": 5.2316,
+      "step": 10012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.084526777267456,
+      "learning_rate": 1.4494861353500097e-06,
+      "loss": 5.279,
+      "step": 10016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.141874074935913,
+      "learning_rate": 1.4300950164824511e-06,
+      "loss": 5.1842,
+      "step": 10020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1359285116195679,
+      "learning_rate": 1.4107038976148925e-06,
+      "loss": 5.2476,
+      "step": 10024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 1.3913127787473339e-06,
+      "loss": 5.4255,
+      "step": 10028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0821614265441895,
+      "learning_rate": 1.371921659879775e-06,
+      "loss": 5.2677,
+      "step": 10032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.108450174331665,
+      "learning_rate": 1.3525305410122164e-06,
+      "loss": 5.3634,
+      "step": 10036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1564322710037231,
+      "learning_rate": 1.3331394221446578e-06,
+      "loss": 5.3358,
+      "step": 10040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0499944686889648,
+      "learning_rate": 1.3137483032770992e-06,
+      "loss": 5.2706,
+      "step": 10044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.115499496459961,
+      "learning_rate": 1.2943571844095404e-06,
+      "loss": 5.2685,
+      "step": 10048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0848426818847656,
+      "learning_rate": 1.274966065541982e-06,
+      "loss": 5.1596,
+      "step": 10052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.014477014541626,
+      "learning_rate": 1.2555749466744232e-06,
+      "loss": 5.3194,
+      "step": 10056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0822529792785645,
+      "learning_rate": 1.2361838278068645e-06,
+      "loss": 5.2521,
+      "step": 10060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.097193717956543,
+      "learning_rate": 1.2167927089393057e-06,
+      "loss": 5.2256,
+      "step": 10064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0768396854400635,
+      "learning_rate": 1.1974015900717473e-06,
+      "loss": 5.3245,
+      "step": 10068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1191695928573608,
+      "learning_rate": 1.1780104712041885e-06,
+      "loss": 5.3045,
+      "step": 10072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0832325220108032,
+      "learning_rate": 1.1586193523366299e-06,
+      "loss": 5.3106,
+      "step": 10076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.083602786064148,
+      "learning_rate": 1.1392282334690712e-06,
+      "loss": 5.3153,
+      "step": 10080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1011203527450562,
+      "learning_rate": 1.1198371146015126e-06,
+      "loss": 5.3236,
+      "step": 10084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1766139268875122,
+      "learning_rate": 1.100445995733954e-06,
+      "loss": 5.2973,
+      "step": 10088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0857359170913696,
+      "learning_rate": 1.0810548768663952e-06,
+      "loss": 5.2986,
+      "step": 10092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1640616655349731,
+      "learning_rate": 1.0616637579988366e-06,
+      "loss": 5.3083,
+      "step": 10096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1388750076293945,
+      "learning_rate": 1.042272639131278e-06,
+      "loss": 5.3472,
+      "step": 10100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1172057390213013,
+      "learning_rate": 1.0228815202637193e-06,
+      "loss": 5.2601,
+      "step": 10104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.078047275543213,
+      "learning_rate": 1.0034904013961605e-06,
+      "loss": 5.2567,
+      "step": 10108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.134878158569336,
+      "learning_rate": 9.840992825286019e-07,
+      "loss": 5.2999,
+      "step": 10112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.049080729484558,
+      "learning_rate": 9.647081636610433e-07,
+      "loss": 5.2823,
+      "step": 10116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0992294549942017,
+      "learning_rate": 9.453170447934847e-07,
+      "loss": 5.264,
+      "step": 10120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9882086515426636,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 5.2944,
+      "step": 10124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0819331407546997,
+      "learning_rate": 9.065348070583674e-07,
+      "loss": 5.3227,
+      "step": 10128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0847752094268799,
+      "learning_rate": 8.871436881908087e-07,
+      "loss": 5.375,
+      "step": 10132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0530517101287842,
+      "learning_rate": 8.6775256932325e-07,
+      "loss": 5.2639,
+      "step": 10136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.010219931602478,
+      "learning_rate": 8.483614504556913e-07,
+      "loss": 5.2476,
+      "step": 10140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9741299748420715,
+      "learning_rate": 8.289703315881327e-07,
+      "loss": 5.3081,
+      "step": 10144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1004730463027954,
+      "learning_rate": 8.09579212720574e-07,
+      "loss": 5.2126,
+      "step": 10148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1317744255065918,
+      "learning_rate": 7.901880938530153e-07,
+      "loss": 5.2943,
+      "step": 10152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1052905321121216,
+      "learning_rate": 7.707969749854567e-07,
+      "loss": 5.4339,
+      "step": 10156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0474963188171387,
+      "learning_rate": 7.51405856117898e-07,
+      "loss": 5.3417,
+      "step": 10160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0810647010803223,
+      "learning_rate": 7.320147372503393e-07,
+      "loss": 5.4219,
+      "step": 10164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0385664701461792,
+      "learning_rate": 7.126236183827807e-07,
+      "loss": 5.3207,
+      "step": 10168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1225430965423584,
+      "learning_rate": 6.93232499515222e-07,
+      "loss": 5.3818,
+      "step": 10172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.021005630493164,
+      "learning_rate": 6.738413806476634e-07,
+      "loss": 5.3095,
+      "step": 10176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0524823665618896,
+      "learning_rate": 6.544502617801048e-07,
+      "loss": 5.2837,
+      "step": 10180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0468004941940308,
+      "learning_rate": 6.350591429125462e-07,
+      "loss": 5.2493,
+      "step": 10184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0307637453079224,
+      "learning_rate": 6.156680240449874e-07,
+      "loss": 5.2559,
+      "step": 10188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0332852602005005,
+      "learning_rate": 5.962769051774288e-07,
+      "loss": 5.4378,
+      "step": 10192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0646268129348755,
+      "learning_rate": 5.768857863098701e-07,
+      "loss": 5.2816,
+      "step": 10196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0905969142913818,
+      "learning_rate": 5.574946674423115e-07,
+      "loss": 5.363,
+      "step": 10200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0439486503601074,
+      "learning_rate": 5.381035485747528e-07,
+      "loss": 5.3199,
+      "step": 10204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0738022327423096,
+      "learning_rate": 5.187124297071941e-07,
+      "loss": 5.251,
+      "step": 10208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.030446171760559,
+      "learning_rate": 4.993213108396354e-07,
+      "loss": 5.3156,
+      "step": 10212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.042098045349121,
+      "learning_rate": 4.799301919720768e-07,
+      "loss": 5.2401,
+      "step": 10216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9965024590492249,
+      "learning_rate": 4.6053907310451814e-07,
+      "loss": 5.2561,
+      "step": 10220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0656026601791382,
+      "learning_rate": 4.411479542369595e-07,
+      "loss": 5.3587,
+      "step": 10224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0533535480499268,
+      "learning_rate": 4.217568353694008e-07,
+      "loss": 5.3381,
+      "step": 10228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0184639692306519,
+      "learning_rate": 4.023657165018422e-07,
+      "loss": 5.3548,
+      "step": 10232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0845178365707397,
+      "learning_rate": 3.829745976342835e-07,
+      "loss": 5.3597,
+      "step": 10236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.068476915359497,
+      "learning_rate": 3.6358347876672485e-07,
+      "loss": 5.2314,
+      "step": 10240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1127116680145264,
+      "learning_rate": 3.441923598991662e-07,
+      "loss": 5.2483,
+      "step": 10244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9709210395812988,
+      "learning_rate": 3.248012410316075e-07,
+      "loss": 5.2665,
+      "step": 10248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0695501565933228,
+      "learning_rate": 3.054101221640489e-07,
+      "loss": 5.2079,
+      "step": 10252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1100637912750244,
+      "learning_rate": 2.860190032964902e-07,
+      "loss": 5.3078,
+      "step": 10256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0800820589065552,
+      "learning_rate": 2.6662788442893156e-07,
+      "loss": 5.4309,
+      "step": 10260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0710211992263794,
+      "learning_rate": 2.472367655613729e-07,
+      "loss": 5.3063,
+      "step": 10264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0967445373535156,
+      "learning_rate": 2.2784564669381422e-07,
+      "loss": 5.2245,
+      "step": 10268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.05315101146698,
+      "learning_rate": 2.084545278262556e-07,
+      "loss": 5.3429,
+      "step": 10272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.97821044921875,
+      "learning_rate": 1.8906340895869693e-07,
+      "loss": 5.2508,
+      "step": 10276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0844597816467285,
+      "learning_rate": 1.6967229009113826e-07,
+      "loss": 5.2811,
+      "step": 10280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0489965677261353,
+      "learning_rate": 1.5028117122357962e-07,
+      "loss": 5.3193,
+      "step": 10284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0502104759216309,
+      "learning_rate": 1.3089005235602095e-07,
+      "loss": 5.3323,
+      "step": 10288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1087231636047363,
+      "learning_rate": 1.114989334884623e-07,
+      "loss": 5.2669,
+      "step": 10292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0919489860534668,
+      "learning_rate": 9.210781462090364e-08,
+      "loss": 5.2765,
+      "step": 10296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.017907738685608,
+      "learning_rate": 7.271669575334497e-08,
+      "loss": 5.3242,
+      "step": 10300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1086193323135376,
+      "learning_rate": 5.332557688578631e-08,
+      "loss": 5.2659,
+      "step": 10304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0139724016189575,
+      "learning_rate": 3.393445801822765e-08,
+      "loss": 5.3324,
+      "step": 10308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9419534802436829,
+      "learning_rate": 1.4543339150668994e-08,
+      "loss": 5.3066,
+      "step": 10312
+    },
+    {
+      "epoch": 1.0,
+      "step": 10314,
+      "total_flos": 8.68702227309527e+16,
+      "train_loss": 5.446075817446203,
+      "train_runtime": 2459.6002,
+      "train_samples_per_second": 67.094,
+      "train_steps_per_second": 4.193
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 10314,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1032,
+  "total_flos": 8.68702227309527e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}