diff --git "a/checkpoint-4000/trainer_state.json" "b/checkpoint-4000/trainer_state.json"
--- "a/checkpoint-4000/trainer_state.json"
+++ "b/checkpoint-4000/trainer_state.json"
@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 7.648183556405353,
+  "epoch": 6.4412238325281805,
   "eval_steps": 500,
   "global_step": 4000,
   "is_hyper_param_search": false,
@@ -9,7010 +9,7010 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0076481835564053535,
-      "grad_norm": 0.294921875,
-      "learning_rate": 2.8571428571428567e-05,
-      "loss": 1.3014,
+      "epoch": 0.00644122383252818,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.727272727272727e-05,
+      "loss": 1.3628,
       "step": 4
     },
     {
-      "epoch": 0.015296367112810707,
-      "grad_norm": 0.26953125,
-      "learning_rate": 5.7142857142857135e-05,
-      "loss": 1.3157,
+      "epoch": 0.01288244766505636,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.454545454545454e-05,
+      "loss": 1.3272,
       "step": 8
     },
     {
-      "epoch": 0.022944550669216062,
-      "grad_norm": 0.201171875,
-      "learning_rate": 8.57142857142857e-05,
-      "loss": 1.2369,
+      "epoch": 0.01932367149758454,
+      "grad_norm": 0.375,
+      "learning_rate": 8.18181818181818e-05,
+      "loss": 1.2626,
       "step": 12
     },
     {
-      "epoch": 0.030592734225621414,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00011428571428571427,
-      "loss": 1.207,
+      "epoch": 0.02576489533011272,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00010909090909090908,
+      "loss": 1.2028,
       "step": 16
     },
     {
-      "epoch": 0.03824091778202677,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00014285714285714284,
-      "loss": 1.1666,
+      "epoch": 0.0322061191626409,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013636363636363634,
+      "loss": 1.1822,
       "step": 20
     },
     {
-      "epoch": 0.045889101338432124,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0001714285714285714,
-      "loss": 1.178,
+      "epoch": 0.03864734299516908,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001636363636363636,
+      "loss": 1.2029,
       "step": 24
     },
     {
-      "epoch": 0.05353728489483748,
-      "grad_norm": 0.1552734375,
-      "learning_rate": 0.00019999999999999998,
-      "loss": 1.1533,
+      "epoch": 0.04508856682769726,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001909090909090909,
+      "loss": 1.1609,
       "step": 28
     },
     {
-      "epoch": 0.06118546845124283,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00022857142857142854,
-      "loss": 1.1508,
+      "epoch": 0.05152979066022544,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00021818181818181816,
+      "loss": 1.1137,
       "step": 32
     },
     {
-      "epoch": 0.06883365200764818,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0002571428571428571,
-      "loss": 1.1415,
+      "epoch": 0.057971014492753624,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00024545454545454545,
+      "loss": 1.085,
       "step": 36
     },
     {
-      "epoch": 0.07648183556405354,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0002857142857142857,
-      "loss": 1.1255,
+      "epoch": 0.0644122383252818,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002727272727272727,
+      "loss": 1.1052,
       "step": 40
     },
     {
-      "epoch": 0.0841300191204589,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0002999998274159216,
-      "loss": 1.0581,
+      "epoch": 0.07085346215780998,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0003,
+      "loss": 1.0712,
       "step": 44
     },
     {
-      "epoch": 0.09177820267686425,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00029999844674567734,
-      "loss": 1.0987,
+      "epoch": 0.07729468599033816,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00029999936035650057,
+      "loss": 1.0588,
       "step": 48
     },
     {
-      "epoch": 0.0994263862332696,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0002999956854178972,
-      "loss": 1.089,
+      "epoch": 0.08373590982286634,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002999974414314574,
+      "loss": 1.0531,
       "step": 52
     },
     {
-      "epoch": 0.10707456978967496,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00029999154345799773,
-      "loss": 1.0934,
+      "epoch": 0.09017713365539452,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00029999424324123633,
+      "loss": 0.9953,
       "step": 56
     },
     {
-      "epoch": 0.1147227533460803,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0002999860209041035,
-      "loss": 1.0712,
+      "epoch": 0.0966183574879227,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002999897658131134,
+      "loss": 0.9887,
       "step": 60
     },
     {
-      "epoch": 0.12237093690248566,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00029997911780704675,
-      "loss": 1.0253,
+      "epoch": 0.10305958132045089,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002999840091852746,
+      "loss": 0.9945,
       "step": 64
     },
     {
-      "epoch": 0.13001912045889102,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00029997083423036696,
-      "loss": 1.0526,
+      "epoch": 0.10950080515297907,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029997697340681585,
+      "loss": 0.9306,
       "step": 68
     },
     {
-      "epoch": 0.13766730401529637,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00029996117025031,
-      "loss": 1.0746,
+      "epoch": 0.11594202898550725,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029996865853774236,
+      "loss": 0.9458,
       "step": 72
     },
     {
-      "epoch": 0.14531548757170173,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00029995012595582796,
-      "loss": 1.0502,
+      "epoch": 0.12238325281803543,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029995906464896807,
+      "loss": 0.9487,
       "step": 76
     },
     {
-      "epoch": 0.15296367112810708,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0002999377014485777,
-      "loss": 1.0461,
+      "epoch": 0.1288244766505636,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002999481918223153,
+      "loss": 0.9144,
       "step": 80
     },
     {
-      "epoch": 0.16061185468451242,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00029992389684292025,
-      "loss": 1.0223,
+      "epoch": 0.13526570048309178,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002999360401505139,
+      "loss": 0.9289,
       "step": 84
     },
     {
-      "epoch": 0.1682600382409178,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00029990871226591995,
-      "loss": 1.0218,
+      "epoch": 0.14170692431561996,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029992260973720023,
+      "loss": 0.882,
       "step": 88
     },
     {
-      "epoch": 0.17590822179732313,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00029989214785734286,
-      "loss": 1.0192,
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029990790069691665,
+      "loss": 0.9031,
       "step": 92
     },
     {
-      "epoch": 0.1835564053537285,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00029987420376965577,
-      "loss": 1.0243,
+      "epoch": 0.15458937198067632,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029989191315511055,
+      "loss": 0.9127,
       "step": 96
     },
     {
-      "epoch": 0.19120458891013384,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00029985488016802457,
-      "loss": 1.0202,
+      "epoch": 0.1610305958132045,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002998746472481328,
+      "loss": 0.8803,
       "step": 100
     },
     {
-      "epoch": 0.1988527724665392,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00029983417723031307,
-      "loss": 0.9904,
+      "epoch": 0.16747181964573268,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002998561031232371,
+      "loss": 0.8764,
       "step": 104
     },
     {
-      "epoch": 0.20650095602294455,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.0002998120951470809,
-      "loss": 1.0228,
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029983628093857855,
+      "loss": 0.9189,
       "step": 108
     },
     {
-      "epoch": 0.21414913957934992,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00029978863412158217,
-      "loss": 1.004,
+      "epoch": 0.18035426731078905,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00029981518086321225,
+      "loss": 0.8169,
       "step": 112
     },
     {
-      "epoch": 0.22179732313575526,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002997637943697635,
-      "loss": 0.9903,
+      "epoch": 0.18679549114331723,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00029979280307709176,
+      "loss": 0.8672,
       "step": 116
     },
     {
-      "epoch": 0.2294455066921606,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00029973757612026164,
-      "loss": 0.969,
+      "epoch": 0.1932367149758454,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002997691477710679,
+      "loss": 0.8387,
       "step": 120
     },
     {
-      "epoch": 0.23709369024856597,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029970997961440213,
-      "loss": 0.9869,
+      "epoch": 0.1996779388083736,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002997442151468869,
+      "loss": 0.8039,
       "step": 124
     },
     {
-      "epoch": 0.2447418738049713,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002996810051061963,
-      "loss": 0.9758,
+      "epoch": 0.20611916264090177,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029971800541718854,
+      "loss": 0.8294,
       "step": 128
     },
     {
-      "epoch": 0.25239005736137665,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00029965065286233943,
-      "loss": 0.9615,
+      "epoch": 0.21256038647342995,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002996905188055046,
+      "loss": 0.8228,
       "step": 132
     },
     {
-      "epoch": 0.26003824091778205,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00029961892316220817,
-      "loss": 0.9541,
+      "epoch": 0.21900161030595813,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029966175554625696,
+      "loss": 0.8341,
       "step": 136
     },
     {
-      "epoch": 0.2676864244741874,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.0002995858162978577,
-      "loss": 0.9455,
+      "epoch": 0.22544283413848631,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00029963171588475525,
+      "loss": 0.8095,
       "step": 140
     },
     {
-      "epoch": 0.27533460803059273,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.0002995513325740197,
-      "loss": 0.9566,
+      "epoch": 0.2318840579710145,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002996004000771952,
+      "loss": 0.8285,
       "step": 144
     },
     {
-      "epoch": 0.2829827915869981,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029951547230809865,
-      "loss": 0.9405,
+      "epoch": 0.23832528180354268,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029956780839065616,
+      "loss": 0.8123,
       "step": 148
     },
     {
-      "epoch": 0.29063097514340347,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00029947823583016973,
-      "loss": 0.9119,
+      "epoch": 0.24476650563607086,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00029953394110309887,
+      "loss": 0.7612,
       "step": 152
     },
     {
-      "epoch": 0.2982791586998088,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029943962348297535,
-      "loss": 0.9507,
+      "epoch": 0.25120772946859904,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002994987985033633,
+      "loss": 0.7723,
       "step": 156
     },
     {
-      "epoch": 0.30592734225621415,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00029939963562192196,
-      "loss": 0.9507,
+      "epoch": 0.2576489533011272,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002994623808911659,
+      "loss": 0.8202,
       "step": 160
     },
     {
-      "epoch": 0.3135755258126195,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.000299358272615077,
-      "loss": 0.9815,
+      "epoch": 0.2640901771336554,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00029942468857709715,
+      "loss": 0.7324,
       "step": 164
     },
     {
-      "epoch": 0.32122370936902483,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002993155348431654,
-      "loss": 0.9364,
+      "epoch": 0.27053140096618356,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000299385721882619,
+      "loss": 0.7818,
       "step": 168
     },
     {
-      "epoch": 0.32887189292543023,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0002992714226995661,
-      "loss": 0.9177,
+      "epoch": 0.27697262479871176,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000299345481140062,
+      "loss": 0.7693,
       "step": 172
     },
     {
-      "epoch": 0.3365200764818356,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00029922593659030837,
-      "loss": 0.9224,
+      "epoch": 0.2834138486312399,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00029930396669262255,
+      "loss": 0.7481,
       "step": 176
     },
     {
-      "epoch": 0.3441682600382409,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00029917907693406817,
-      "loss": 0.9359,
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00029926117889435993,
+      "loss": 0.7478,
       "step": 180
     },
     {
-      "epoch": 0.35181644359464626,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00029913084416216415,
-      "loss": 0.9349,
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00029921711811019334,
+      "loss": 0.7581,
       "step": 184
     },
     {
-      "epoch": 0.35946462715105165,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00029908123871855396,
-      "loss": 0.9033,
+      "epoch": 0.3027375201288245,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00029917178471589864,
+      "loss": 0.7131,
       "step": 188
     },
     {
-      "epoch": 0.367112810707457,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.0002990302610598297,
-      "loss": 0.9007,
+      "epoch": 0.30917874396135264,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002991251790981053,
+      "loss": 0.7121,
       "step": 192
     },
     {
-      "epoch": 0.37476099426386233,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00029897791165521434,
-      "loss": 0.9243,
+      "epoch": 0.31561996779388085,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002990773016542932,
+      "loss": 0.7385,
       "step": 196
     },
     {
-      "epoch": 0.3824091778202677,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002989241909865567,
-      "loss": 0.9095,
+      "epoch": 0.322061191626409,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00029902815279278874,
+      "loss": 0.743,
       "step": 200
     },
     {
-      "epoch": 0.390057361376673,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002988690995483276,
-      "loss": 0.8825,
+      "epoch": 0.3285024154589372,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00029897773293276214,
+      "loss": 0.6984,
       "step": 204
     },
     {
-      "epoch": 0.3977055449330784,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029881263784761503,
-      "loss": 0.8929,
+      "epoch": 0.33494363929146537,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000298926042504223,
+      "loss": 0.7278,
       "step": 208
     },
     {
-      "epoch": 0.40535372848948376,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00029875480640411957,
-      "loss": 0.9097,
+      "epoch": 0.3413848631239936,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00029887308194801745,
+      "loss": 0.7043,
       "step": 212
     },
     {
-      "epoch": 0.4130019120458891,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029869560575014945,
-      "loss": 0.8563,
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029881885171582364,
+      "loss": 0.7455,
       "step": 216
     },
     {
-      "epoch": 0.42065009560229444,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00029863503643061585,
-      "loss": 0.8839,
+      "epoch": 0.35426731078904994,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002987633522701486,
+      "loss": 0.7314,
       "step": 220
     },
     {
-      "epoch": 0.42829827915869984,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002985730990030278,
-      "loss": 0.8635,
+      "epoch": 0.3607085346215781,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00029870658408432375,
+      "loss": 0.7344,
       "step": 224
     },
     {
-      "epoch": 0.4359464627151052,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00029850979403748705,
-      "loss": 0.859,
+      "epoch": 0.3671497584541063,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002986485476425011,
+      "loss": 0.7324,
       "step": 228
     },
     {
-      "epoch": 0.4435946462715105,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00029844512211668286,
-      "loss": 0.8256,
+      "epoch": 0.37359098228663445,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002985892434396491,
+      "loss": 0.7197,
       "step": 232
     },
     {
-      "epoch": 0.45124282982791586,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00029837908383588646,
-      "loss": 0.8282,
+      "epoch": 0.38003220611916266,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00029852867198154837,
+      "loss": 0.6616,
       "step": 236
     },
     {
-      "epoch": 0.4588910133843212,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.0002983116798029459,
-      "loss": 0.8579,
+      "epoch": 0.3864734299516908,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002984668337847874,
+      "loss": 0.6325,
       "step": 240
     },
     {
-      "epoch": 0.4665391969407266,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002982429106382801,
-      "loss": 0.8805,
+      "epoch": 0.392914653784219,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002984037293767583,
+      "loss": 0.6445,
       "step": 244
     },
     {
-      "epoch": 0.47418738049713194,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029817277697487347,
-      "loss": 0.823,
+      "epoch": 0.3993558776167472,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00029833935929565194,
+      "loss": 0.6846,
       "step": 248
     },
     {
-      "epoch": 0.4818355640535373,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002981012794582698,
-      "loss": 0.8546,
+      "epoch": 0.4057971014492754,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00029827372409045377,
+      "loss": 0.6976,
       "step": 252
     },
     {
-      "epoch": 0.4894837476099426,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002980284187465665,
-      "loss": 0.8533,
+      "epoch": 0.41223832528180354,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002982068243209389,
+      "loss": 0.7165,
       "step": 256
     },
     {
-      "epoch": 0.497131931166348,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00029795419551040833,
-      "loss": 0.8111,
+      "epoch": 0.41867954911433175,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00029813866055766736,
+      "loss": 0.6647,
       "step": 260
     },
     {
-      "epoch": 0.5047801147227533,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002978786104329816,
-      "loss": 0.8823,
+      "epoch": 0.4251207729468599,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00029806923338197925,
+      "loss": 0.6809,
       "step": 264
     },
     {
-      "epoch": 0.5124282982791587,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002978016642100076,
-      "loss": 0.8839,
+      "epoch": 0.43156199677938806,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029799854338598974,
+      "loss": 0.7285,
       "step": 268
     },
     {
-      "epoch": 0.5200764818355641,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00029772335754973614,
-      "loss": 0.8512,
+      "epoch": 0.43800322061191627,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0002979265911725842,
+      "loss": 0.6978,
       "step": 272
     },
     {
-      "epoch": 0.5277246653919694,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029764369117293925,
-      "loss": 0.8557,
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029785337735541276,
+      "loss": 0.6598,
       "step": 276
     },
     {
-      "epoch": 0.5353728489483748,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002975626658129044,
-      "loss": 0.8163,
+      "epoch": 0.45088566827697263,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002977789025588854,
+      "loss": 0.6534,
       "step": 280
     },
     {
-      "epoch": 0.5430210325047801,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0002974802822154278,
-      "loss": 0.9078,
+      "epoch": 0.4573268921095008,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002977031674181663,
+      "loss": 0.7261,
       "step": 284
     },
     {
-      "epoch": 0.5506692160611855,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00029739654113880755,
-      "loss": 0.8652,
+      "epoch": 0.463768115942029,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00029762617257916873,
+      "loss": 0.6762,
       "step": 288
     },
     {
-      "epoch": 0.5583173996175909,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00029731144335383663,
-      "loss": 0.8551,
+      "epoch": 0.47020933977455714,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002975479186985493,
+      "loss": 0.6625,
       "step": 292
     },
     {
-      "epoch": 0.5659655831739961,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.0002972249896437958,
-      "loss": 0.8536,
+      "epoch": 0.47665056360708535,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002974684064437025,
+      "loss": 0.6617,
       "step": 296
     },
     {
-      "epoch": 0.5736137667304015,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0002971371808044464,
-      "loss": 0.825,
+      "epoch": 0.4830917874396135,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00029738763649275496,
+      "loss": 0.6886,
       "step": 300
     },
     {
-      "epoch": 0.5812619502868069,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.000297048017644023,
-      "loss": 0.8082,
+      "epoch": 0.4895330112721417,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002973056095345596,
+      "loss": 0.6623,
       "step": 304
     },
     {
-      "epoch": 0.5889101338432122,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002969575009832261,
-      "loss": 0.8304,
+      "epoch": 0.49597423510466987,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00029722232626869,
+      "loss": 0.6568,
       "step": 308
     },
     {
-      "epoch": 0.5965583173996176,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00029686563165521435,
-      "loss": 0.8101,
+      "epoch": 0.5024154589371981,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002971377874054341,
+      "loss": 0.6281,
       "step": 312
     },
     {
-      "epoch": 0.6042065009560229,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.00029677241050559707,
-      "loss": 0.8535,
+      "epoch": 0.5088566827697263,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002970519936657884,
+      "loss": 0.6618,
       "step": 316
     },
     {
-      "epoch": 0.6118546845124283,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029667783839242625,
-      "loss": 0.85,
+      "epoch": 0.5152979066022544,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00029696494578145157,
+      "loss": 0.6797,
       "step": 320
     },
     {
-      "epoch": 0.6195028680688337,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002965819161861891,
-      "loss": 0.8101,
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002968766444948185,
+      "loss": 0.6756,
       "step": 324
     },
     {
-      "epoch": 0.627151051625239,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002964846447697994,
-      "loss": 0.8521,
+      "epoch": 0.5281803542673108,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002967870905589739,
+      "loss": 0.698,
       "step": 328
     },
     {
-      "epoch": 0.6347992351816444,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00029638602503858995,
-      "loss": 0.8506,
+      "epoch": 0.534621578099839,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002966962847376855,
+      "loss": 0.6431,
       "step": 332
     },
     {
-      "epoch": 0.6424474187380497,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029628605790030384,
-      "loss": 0.8044,
+      "epoch": 0.5410628019323671,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00029660422780539814,
+      "loss": 0.6713,
       "step": 336
     },
     {
-      "epoch": 0.6500956022944551,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002961847442750866,
-      "loss": 0.8311,
+      "epoch": 0.5475040257648953,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00029651092054722665,
+      "loss": 0.615,
       "step": 340
     },
     {
-      "epoch": 0.6577437858508605,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00029608208509547735,
-      "loss": 0.8705,
+      "epoch": 0.5539452495974235,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002964163637589495,
+      "loss": 0.7173,
       "step": 344
     },
     {
-      "epoch": 0.6653919694072657,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029597808130640027,
-      "loss": 0.8272,
+      "epoch": 0.5603864734299517,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002963205582470017,
+      "loss": 0.6808,
       "step": 348
     },
     {
-      "epoch": 0.6730401529636711,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0002958727338651562,
-      "loss": 0.8035,
+      "epoch": 0.5668276972624798,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00029622350482846844,
+      "loss": 0.6684,
       "step": 352
     },
     {
-      "epoch": 0.6806883365200764,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029576604374141315,
-      "loss": 0.8655,
+      "epoch": 0.573268921095008,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00029612520433107734,
+      "loss": 0.644,
       "step": 356
     },
     {
-      "epoch": 0.6883365200764818,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00029565801191719837,
-      "loss": 0.8585,
+      "epoch": 0.5797101449275363,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002960256575931922,
+      "loss": 0.6599,
       "step": 360
     },
     {
-      "epoch": 0.6959847036328872,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002955486393868884,
-      "loss": 0.8279,
+      "epoch": 0.5861513687600645,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002959248654638053,
+      "loss": 0.7006,
       "step": 364
     },
     {
-      "epoch": 0.7036328871892925,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.0002954379271572004,
-      "loss": 0.8229,
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029582282880253035,
+      "loss": 0.625,
       "step": 368
     },
     {
-      "epoch": 0.7112810707456979,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002953258762471828,
-      "loss": 0.8549,
+      "epoch": 0.5990338164251208,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002957195484795952,
+      "loss": 0.7234,
       "step": 372
     },
     {
-      "epoch": 0.7189292543021033,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002952124876882058,
-      "loss": 0.837,
+      "epoch": 0.605475040257649,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002956150253758344,
+      "loss": 0.6556,
       "step": 376
     },
     {
-      "epoch": 0.7265774378585086,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00029509776252395194,
-      "loss": 0.8084,
+      "epoch": 0.6119162640901772,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00029550926038268146,
+      "loss": 0.6402,
       "step": 380
     },
     {
-      "epoch": 0.734225621414914,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002949817018104066,
-      "loss": 0.8454,
+      "epoch": 0.6183574879227053,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002954022544021617,
+      "loss": 0.6446,
       "step": 384
     },
     {
-      "epoch": 0.7418738049713193,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002948643066158482,
-      "loss": 0.7603,
+      "epoch": 0.6247987117552335,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00029529400834688415,
+      "loss": 0.6379,
       "step": 388
     },
     {
-      "epoch": 0.7495219885277247,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00029474557802083834,
-      "loss": 0.8814,
+      "epoch": 0.6312399355877617,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00029518452314003394,
+      "loss": 0.644,
       "step": 392
     },
     {
-      "epoch": 0.7571701720841301,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002946255171182119,
-      "loss": 0.8678,
+      "epoch": 0.6376811594202898,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002950737997153645,
+      "loss": 0.6413,
       "step": 396
     },
     {
-      "epoch": 0.7648183556405354,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029450412501306675,
-      "loss": 0.8397,
+      "epoch": 0.644122383252818,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029496183901718927,
+      "loss": 0.6249,
       "step": 400
     },
     {
-      "epoch": 0.7724665391969407,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029438140282275413,
-      "loss": 0.797,
+      "epoch": 0.6505636070853462,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00029484864200037415,
+      "loss": 0.5929,
       "step": 404
     },
     {
-      "epoch": 0.780114722753346,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0002942573516768678,
-      "loss": 0.7764,
+      "epoch": 0.6570048309178744,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002947342096303289,
+      "loss": 0.6447,
       "step": 408
     },
     {
-      "epoch": 0.7877629063097514,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00029413197271723385,
-      "loss": 0.7856,
+      "epoch": 0.6634460547504025,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002946185428829991,
+      "loss": 0.641,
       "step": 412
     },
     {
-      "epoch": 0.7954110898661568,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002940052670979004,
-      "loss": 0.7878,
+      "epoch": 0.6698872785829307,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002945016427448579,
+      "loss": 0.6878,
       "step": 416
     },
     {
-      "epoch": 0.8030592734225621,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002938772359851265,
-      "loss": 0.7971,
+      "epoch": 0.6763285024154589,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002943835102128975,
+      "loss": 0.6646,
       "step": 420
     },
     {
-      "epoch": 0.8107074569789675,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029374788055737194,
-      "loss": 0.825,
+      "epoch": 0.6827697262479872,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002942641462946206,
+      "loss": 0.613,
       "step": 424
     },
     {
-      "epoch": 0.8183556405353728,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.000293617202005286,
-      "loss": 0.8105,
+      "epoch": 0.6892109500805152,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00029414355200803197,
+      "loss": 0.6135,
       "step": 428
     },
     {
-      "epoch": 0.8260038240917782,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029348520153169656,
-      "loss": 0.8055,
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002940217283816296,
+      "loss": 0.6145,
       "step": 432
     },
     {
-      "epoch": 0.8336520076481836,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.0002933518803515993,
-      "loss": 0.7616,
+      "epoch": 0.7020933977455717,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002938986764543961,
+      "loss": 0.6199,
       "step": 436
     },
     {
-      "epoch": 0.8413001912045889,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00029321723969214625,
-      "loss": 0.7842,
+      "epoch": 0.7085346215780999,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002937743972757895,
+      "loss": 0.6566,
       "step": 440
     },
     {
-      "epoch": 0.8489483747609943,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0002930812807926343,
-      "loss": 0.803,
+      "epoch": 0.714975845410628,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002936488919057349,
+      "loss": 0.6536,
       "step": 444
     },
     {
-      "epoch": 0.8565965583173997,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0002929440049044945,
-      "loss": 0.8174,
+      "epoch": 0.7214170692431562,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002935221614146148,
+      "loss": 0.6586,
       "step": 448
     },
     {
-      "epoch": 0.864244741873805,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002928054132912797,
-      "loss": 0.836,
+      "epoch": 0.7278582930756844,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002933942068832604,
+      "loss": 0.6234,
       "step": 452
     },
     {
-      "epoch": 0.8718929254302104,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002926655072286536,
-      "loss": 0.7552,
+      "epoch": 0.7342995169082126,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00029326502940294207,
+      "loss": 0.6115,
       "step": 456
     },
     {
-      "epoch": 0.8795411089866156,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00029252428800437854,
-      "loss": 0.7755,
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00029313463007536034,
+      "loss": 0.6205,
       "step": 460
     },
     {
-      "epoch": 0.887189292543021,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029238175691830395,
-      "loss": 0.8101,
+      "epoch": 0.7471819645732689,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002930030100126363,
+      "loss": 0.6185,
       "step": 464
     },
     {
-      "epoch": 0.8948374760994264,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00029223791528235407,
-      "loss": 0.7662,
+      "epoch": 0.7536231884057971,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002928701703373021,
+      "loss": 0.6305,
       "step": 468
     },
     {
-      "epoch": 0.9024856596558317,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00029209276442051634,
-      "loss": 0.7772,
+      "epoch": 0.7600644122383253,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029273611218229165,
+      "loss": 0.6464,
       "step": 472
     },
     {
-      "epoch": 0.9101338432122371,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0002919463056688287,
-      "loss": 0.8135,
+      "epoch": 0.7665056360708534,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002926008366909307,
+      "loss": 0.6488,
       "step": 476
     },
     {
-      "epoch": 0.9177820267686424,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029179854037536773,
-      "loss": 0.7895,
+      "epoch": 0.7729468599033816,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00029246434501692685,
+      "loss": 0.6148,
       "step": 480
     },
     {
-      "epoch": 0.9254302103250478,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002916494699002358,
-      "loss": 0.7644,
+      "epoch": 0.7793880837359098,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00029232663832436047,
+      "loss": 0.5946,
       "step": 484
     },
     {
-      "epoch": 0.9330783938814532,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00029149909561554914,
-      "loss": 0.7804,
+      "epoch": 0.785829307568438,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002921877177876741,
+      "loss": 0.5898,
       "step": 488
     },
     {
-      "epoch": 0.9407265774378585,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002913474189054246,
-      "loss": 0.7779,
+      "epoch": 0.7922705314009661,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002920475845916626,
+      "loss": 0.6435,
       "step": 492
     },
     {
-      "epoch": 0.9483747609942639,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002911944411659672,
-      "loss": 0.781,
+      "epoch": 0.7987117552334944,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029190623993146313,
+      "loss": 0.6605,
       "step": 496
     },
     {
-      "epoch": 0.9560229445506692,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002910401638052574,
-      "loss": 0.8079,
+      "epoch": 0.8051529790660226,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002917636850125449,
+      "loss": 0.6297,
       "step": 500
     },
     {
-      "epoch": 0.9636711281070746,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00029088458824333787,
-      "loss": 0.8332,
+      "epoch": 0.8115942028985508,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00029161992105069905,
+      "loss": 0.6313,
       "step": 504
     },
     {
-      "epoch": 0.97131931166348,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00029072771591220057,
-      "loss": 0.7752,
+      "epoch": 0.8180354267310789,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002914749492720279,
+      "loss": 0.5953,
       "step": 508
     },
     {
-      "epoch": 0.9789674952198852,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029056954825577353,
-      "loss": 0.8038,
+      "epoch": 0.8244766505636071,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029132877091293493,
+      "loss": 0.6615,
       "step": 512
     },
     {
-      "epoch": 0.9866156787762906,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002904100867299077,
-      "loss": 0.7362,
+      "epoch": 0.8309178743961353,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.000291181387220114,
+      "loss": 0.6771,
       "step": 516
     },
     {
-      "epoch": 0.994263862332696,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.0002902493328023633,
-      "loss": 0.8137,
+      "epoch": 0.8373590982286635,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002910327994505387,
+      "loss": 0.5889,
       "step": 520
     },
     {
-      "epoch": 1.0019120458891013,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0002900872879527964,
-      "loss": 0.7389,
+      "epoch": 0.8438003220611916,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002908830088714516,
+      "loss": 0.5781,
       "step": 524
     },
     {
-      "epoch": 1.0095602294455066,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00028992395367274547,
-      "loss": 0.6994,
+      "epoch": 0.8502415458937198,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00029073201676035383,
+      "loss": 0.6182,
       "step": 528
     },
     {
-      "epoch": 1.0172084130019121,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002897593314656174,
-      "loss": 0.7461,
+      "epoch": 0.856682769726248,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029057982440499356,
+      "loss": 0.6226,
       "step": 532
     },
     {
-      "epoch": 1.0248565965583174,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002895934228466738,
-      "loss": 0.6856,
+      "epoch": 0.8631239935587761,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00029042643310335547,
+      "loss": 0.6547,
       "step": 536
     },
     {
-      "epoch": 1.0325047801147227,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.0002894262293430171,
-      "loss": 0.7405,
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00029027184416364956,
+      "loss": 0.6114,
       "step": 540
     },
     {
-      "epoch": 1.0401529636711282,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002892577524935763,
-      "loss": 0.7906,
+      "epoch": 0.8760064412238325,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002901160589043,
+      "loss": 0.6491,
       "step": 544
     },
     {
-      "epoch": 1.0478011472275335,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028908799384909313,
-      "loss": 0.7144,
+      "epoch": 0.8824476650563607,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00028995907865393385,
+      "loss": 0.6375,
       "step": 548
     },
     {
-      "epoch": 1.0554493307839388,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002889169549721073,
-      "loss": 0.7233,
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00028980090475136963,
+      "loss": 0.6083,
       "step": 552
     },
     {
-      "epoch": 1.063097514340344,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00028874463743694265,
-      "loss": 0.7068,
+      "epoch": 0.895330112721417,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002896415385456062,
+      "loss": 0.5879,
       "step": 556
     },
     {
-      "epoch": 1.0707456978967496,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002885710428296922,
-      "loss": 0.7105,
+      "epoch": 0.9017713365539453,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000289480981395811,
+      "loss": 0.6596,
       "step": 560
     },
     {
-      "epoch": 1.0783938814531548,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00028839617274820404,
-      "loss": 0.6563,
+      "epoch": 0.9082125603864735,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00028931923467130855,
+      "loss": 0.5774,
       "step": 564
     },
     {
-      "epoch": 1.0860420650095601,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00028822002880206593,
-      "loss": 0.6956,
+      "epoch": 0.9146537842190016,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00028915629975156867,
+      "loss": 0.6118,
       "step": 568
     },
     {
-      "epoch": 1.0936902485659656,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00028804261261259115,
-      "loss": 0.6669,
+      "epoch": 0.9210950080515298,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002889921780261949,
+      "loss": 0.615,
       "step": 572
     },
     {
-      "epoch": 1.101338432122371,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028786392581280334,
-      "loss": 0.6727,
+      "epoch": 0.927536231884058,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028882687089491234,
+      "loss": 0.6225,
       "step": 576
     },
     {
-      "epoch": 1.1089866156787762,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00028768397004742135,
-      "loss": 0.714,
+      "epoch": 0.9339774557165862,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002886603797675563,
+      "loss": 0.5626,
       "step": 580
     },
     {
-      "epoch": 1.1166347992351817,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00028750274697284423,
-      "loss": 0.666,
+      "epoch": 0.9404186795491143,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002884927060640596,
+      "loss": 0.5886,
       "step": 584
     },
     {
-      "epoch": 1.124282982791587,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00028732025825713587,
-      "loss": 0.6883,
+      "epoch": 0.9468599033816425,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0002883238512144409,
+      "loss": 0.6251,
       "step": 588
     },
     {
-      "epoch": 1.1319311663479923,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00028713650558000983,
-      "loss": 0.7002,
+      "epoch": 0.9533011272141707,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002881538166587921,
+      "loss": 0.6326,
       "step": 592
     },
     {
-      "epoch": 1.1395793499043978,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002869514906328138,
-      "loss": 0.7038,
+      "epoch": 0.9597423510466989,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002879826038472667,
+      "loss": 0.5666,
       "step": 596
     },
     {
-      "epoch": 1.147227533460803,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028676521511851395,
-      "loss": 0.7178,
+      "epoch": 0.966183574879227,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00028781021424006677,
+      "loss": 0.5282,
       "step": 600
     },
     {
-      "epoch": 1.1548757170172084,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002865776807516793,
-      "loss": 0.6752,
+      "epoch": 0.9726247987117552,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00028763664930743087,
+      "loss": 0.6628,
       "step": 604
     },
     {
-      "epoch": 1.1625239005736137,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.0002863888892584659,
-      "loss": 0.7204,
+      "epoch": 0.9790660225442834,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028746191052962146,
+      "loss": 0.5669,
       "step": 608
     },
     {
-      "epoch": 1.1701720841300192,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00028619884237660124,
-      "loss": 0.6967,
+      "epoch": 0.9855072463768116,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028728599939691215,
+      "loss": 0.5955,
       "step": 612
     },
     {
-      "epoch": 1.1778202676864244,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002860075418553676,
-      "loss": 0.705,
+      "epoch": 0.9919484702093397,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00028710891740957507,
+      "loss": 0.5995,
       "step": 616
     },
     {
-      "epoch": 1.1854684512428297,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00028581498945558676,
-      "loss": 0.7208,
+      "epoch": 0.998389694041868,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028693066607786823,
+      "loss": 0.5813,
       "step": 620
     },
     {
-      "epoch": 1.1931166347992352,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00028562118694960316,
-      "loss": 0.6872,
+      "epoch": 1.0048309178743962,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002867512469220222,
+      "loss": 0.5306,
       "step": 624
     },
     {
-      "epoch": 1.2007648183556405,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.0002854261361212679,
-      "loss": 0.7445,
+      "epoch": 1.0112721417069244,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00028657066147222773,
+      "loss": 0.4918,
       "step": 628
     },
     {
-      "epoch": 1.2084130019120458,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028522983876592213,
-      "loss": 0.7267,
+      "epoch": 1.0177133655394526,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00028638891126862224,
+      "loss": 0.5198,
       "step": 632
     },
     {
-      "epoch": 1.2160611854684513,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002850322966903808,
-      "loss": 0.6831,
+      "epoch": 1.0241545893719808,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002862059978612769,
+      "loss": 0.5673,
       "step": 636
     },
     {
-      "epoch": 1.2237093690248566,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00028483351171291576,
-      "loss": 0.7185,
+      "epoch": 1.0305958132045088,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00028602192281018327,
+      "loss": 0.5127,
       "step": 640
     },
     {
-      "epoch": 1.231357552581262,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00028463348566323913,
-      "loss": 0.7019,
+      "epoch": 1.037037037037037,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002858366876852403,
+      "loss": 0.5517,
       "step": 644
     },
     {
-      "epoch": 1.2390057361376674,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00028443222038248645,
-      "loss": 0.7699,
+      "epoch": 1.0434782608695652,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002856502940662403,
+      "loss": 0.5209,
       "step": 648
     },
     {
-      "epoch": 1.2466539196940727,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028422971772319977,
-      "loss": 0.6939,
+      "epoch": 1.0499194847020934,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00028546274354285646,
+      "loss": 0.5362,
       "step": 652
     },
     {
-      "epoch": 1.254302103250478,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00028402597954931046,
-      "loss": 0.6505,
+      "epoch": 1.0563607085346216,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00028527403771462826,
+      "loss": 0.5256,
       "step": 656
     },
     {
-      "epoch": 1.2619502868068833,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028382100773612236,
-      "loss": 0.6632,
+      "epoch": 1.0628019323671498,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00028508417819094844,
+      "loss": 0.5257,
       "step": 660
     },
     {
-      "epoch": 1.2695984703632888,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002836148041702941,
-      "loss": 0.7787,
+      "epoch": 1.069243156199678,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002848931665910492,
+      "loss": 0.4971,
       "step": 664
     },
     {
-      "epoch": 1.277246653919694,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00028340737074982207,
-      "loss": 0.7293,
+      "epoch": 1.075684380032206,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002847010045439882,
+      "loss": 0.5214,
       "step": 668
     },
     {
-      "epoch": 1.2848948374760996,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002831987093840229,
-      "loss": 0.7471,
+      "epoch": 1.0821256038647342,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002845076936886349,
+      "loss": 0.5283,
       "step": 672
     },
     {
-      "epoch": 1.2925430210325048,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00028298882199351565,
-      "loss": 0.7445,
+      "epoch": 1.0885668276972624,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002843132356736563,
+      "loss": 0.5024,
       "step": 676
     },
     {
-      "epoch": 1.3001912045889101,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028277771051020433,
-      "loss": 0.6997,
+      "epoch": 1.0950080515297906,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002841176321575032,
+      "loss": 0.5515,
       "step": 680
     },
     {
-      "epoch": 1.3078393881453154,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00028256537687726017,
-      "loss": 0.7389,
+      "epoch": 1.1014492753623188,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002839208848083958,
+      "loss": 0.5493,
       "step": 684
     },
     {
-      "epoch": 1.3154875717017207,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002823518230491036,
-      "loss": 0.7171,
+      "epoch": 1.107890499194847,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002837229953043096,
+      "loss": 0.4908,
       "step": 688
     },
     {
-      "epoch": 1.3231357552581262,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028213705099138636,
-      "loss": 0.7054,
+      "epoch": 1.1143317230273753,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002835239653329611,
+      "loss": 0.5136,
       "step": 692
     },
     {
-      "epoch": 1.3307839388145315,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00028192106268097334,
-      "loss": 0.6543,
+      "epoch": 1.1207729468599035,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002833237965917934,
+      "loss": 0.5379,
       "step": 696
     },
     {
-      "epoch": 1.338432122370937,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002817038601059243,
-      "loss": 0.7012,
+      "epoch": 1.1272141706924317,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002831224907879614,
+      "loss": 0.5059,
       "step": 700
     },
     {
-      "epoch": 1.3460803059273423,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002814854452654758,
-      "loss": 0.7058,
+      "epoch": 1.1336553945249597,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00028292004963831796,
+      "loss": 0.5231,
       "step": 704
     },
     {
-      "epoch": 1.3537284894837476,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00028126582017002266,
-      "loss": 0.7797,
+      "epoch": 1.1400966183574879,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00028271647486939855,
+      "loss": 0.5223,
       "step": 708
     },
     {
-      "epoch": 1.3613766730401529,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002810449868410994,
-      "loss": 0.6898,
+      "epoch": 1.146537842190016,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002825117682174069,
+      "loss": 0.4907,
       "step": 712
     },
     {
-      "epoch": 1.3690248565965584,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00028082294731136164,
-      "loss": 0.6872,
+      "epoch": 1.1529790660225443,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002823059314282,
+      "loss": 0.4996,
       "step": 716
     },
     {
-      "epoch": 1.3766730401529637,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00028059970362456776,
-      "loss": 0.706,
+      "epoch": 1.1594202898550725,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002820989662572734,
+      "loss": 0.5084,
       "step": 720
     },
     {
-      "epoch": 1.384321223709369,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00028037525783555935,
-      "loss": 0.6971,
+      "epoch": 1.1658615136876007,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002818908744697461,
+      "loss": 0.4909,
       "step": 724
     },
     {
-      "epoch": 1.3919694072657744,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028014961201024304,
-      "loss": 0.7212,
+      "epoch": 1.1723027375201287,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028168165784034566,
+      "loss": 0.5245,
       "step": 728
     },
     {
-      "epoch": 1.3996175908221797,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002799227682255711,
-      "loss": 0.702,
+      "epoch": 1.178743961352657,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00028147131815339267,
+      "loss": 0.5307,
       "step": 732
     },
     {
-      "epoch": 1.407265774378585,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00027969472856952224,
-      "loss": 0.6604,
+      "epoch": 1.1851851851851851,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00028125985720278614,
+      "loss": 0.5213,
       "step": 736
     },
     {
-      "epoch": 1.4149139579349903,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00027946549514108277,
-      "loss": 0.7,
+      "epoch": 1.1916264090177133,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002810472767919876,
+      "loss": 0.5257,
       "step": 740
     },
     {
-      "epoch": 1.4225621414913958,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00027923507005022687,
-      "loss": 0.7281,
+      "epoch": 1.1980676328502415,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002808335787340061,
+      "loss": 0.4913,
       "step": 744
     },
     {
-      "epoch": 1.430210325047801,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00027900345541789746,
-      "loss": 0.6261,
+      "epoch": 1.2045088566827697,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00028061876485138264,
+      "loss": 0.5331,
       "step": 748
     },
     {
-      "epoch": 1.4378585086042066,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002787706533759865,
-      "loss": 0.7131,
+      "epoch": 1.210950080515298,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00028040283697617464,
+      "loss": 0.5055,
       "step": 752
     },
     {
-      "epoch": 1.445506692160612,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00027853666606731547,
-      "loss": 0.664,
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002801857969499402,
+      "loss": 0.5318,
       "step": 756
     },
     {
-      "epoch": 1.4531548757170172,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002783014956456157,
-      "loss": 0.6856,
+      "epoch": 1.2238325281803544,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002799676466237225,
+      "loss": 0.4991,
       "step": 760
     },
     {
-      "epoch": 1.4608030592734225,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002780651442755083,
-      "loss": 0.7422,
+      "epoch": 1.2302737520128824,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002797483878580342,
+      "loss": 0.5059,
       "step": 764
     },
     {
-      "epoch": 1.468451242829828,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0002778276141324844,
-      "loss": 0.6738,
+      "epoch": 1.2367149758454106,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00027952802252284104,
+      "loss": 0.5043,
       "step": 768
     },
     {
-      "epoch": 1.4760994263862333,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002775889074028853,
-      "loss": 0.7595,
+      "epoch": 1.2431561996779388,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002793065524975465,
+      "loss": 0.5747,
       "step": 772
     },
     {
-      "epoch": 1.4837476099426385,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.000277349026283882,
-      "loss": 0.7132,
+      "epoch": 1.249597423510467,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002790839796709755,
+      "loss": 0.5082,
       "step": 776
     },
     {
-      "epoch": 1.491395793499044,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002771079729834552,
-      "loss": 0.7316,
+      "epoch": 1.2560386473429952,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00027886030594135805,
+      "loss": 0.5369,
       "step": 780
     },
     {
-      "epoch": 1.4990439770554493,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002768657497203749,
-      "loss": 0.7314,
+      "epoch": 1.2624798711755234,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002786355332163135,
+      "loss": 0.5423,
       "step": 784
     },
     {
-      "epoch": 1.5066921606118546,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027662235872418005,
-      "loss": 0.7036,
+      "epoch": 1.2689210950080514,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000278409663412834,
+      "loss": 0.4882,
       "step": 788
     },
     {
-      "epoch": 1.51434034416826,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027637780223515793,
-      "loss": 0.7191,
+      "epoch": 1.2753623188405796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002781826984572683,
+      "loss": 0.504,
       "step": 792
     },
     {
-      "epoch": 1.5219885277246654,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00027613208250432353,
-      "loss": 0.712,
+      "epoch": 1.2818035426731078,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002779546402853051,
+      "loss": 0.4872,
       "step": 796
     },
     {
-      "epoch": 1.5296367112810707,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.000275885201793399,
-      "loss": 0.6918,
+      "epoch": 1.288244766505636,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00027772549084195675,
+      "loss": 0.5348,
       "step": 800
     },
     {
-      "epoch": 1.5372848948374762,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002756371623747925,
-      "loss": 0.6822,
+      "epoch": 1.2946859903381642,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00027749525208154265,
+      "loss": 0.5718,
       "step": 804
     },
     {
-      "epoch": 1.5449330783938815,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002753879665315778,
-      "loss": 0.707,
+      "epoch": 1.3011272141706924,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002772639259676726,
+      "loss": 0.5393,
       "step": 808
     },
     {
-      "epoch": 1.5525812619502868,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002751376165574726,
-      "loss": 0.6921,
+      "epoch": 1.3075684380032206,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00027703151447322965,
+      "loss": 0.5421,
       "step": 812
     },
     {
-      "epoch": 1.560229445506692,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002748861147568181,
-      "loss": 0.7048,
+      "epoch": 1.3140096618357489,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002767980195803539,
+      "loss": 0.5555,
       "step": 816
     },
     {
-      "epoch": 1.5678776290630974,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00027463346344455724,
-      "loss": 0.7171,
+      "epoch": 1.320450885668277,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002765634432804253,
+      "loss": 0.553,
       "step": 820
     },
     {
-      "epoch": 1.5755258126195029,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002743796649462137,
-      "loss": 0.6839,
+      "epoch": 1.3268921095008053,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00027632778757404655,
+      "loss": 0.5075,
       "step": 824
     },
     {
-      "epoch": 1.5831739961759084,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027412472159787037,
-      "loss": 0.6722,
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002760910544710261,
+      "loss": 0.4933,
       "step": 828
     },
     {
-      "epoch": 1.5908221797323137,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00027386863574614803,
-      "loss": 0.6538,
+      "epoch": 1.3397745571658615,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00027585324599036133,
+      "loss": 0.5039,
       "step": 832
     },
     {
-      "epoch": 1.598470363288719,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002736114097481833,
-      "loss": 0.7013,
+      "epoch": 1.3462157809983897,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00027561436416022073,
+      "loss": 0.5175,
       "step": 836
     },
     {
-      "epoch": 1.6061185468451242,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002733530459716076,
-      "loss": 0.6967,
+      "epoch": 1.3526570048309179,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00027537441101792715,
+      "loss": 0.5375,
       "step": 840
     },
     {
-      "epoch": 1.6137667304015295,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00027309354679452483,
-      "loss": 0.7052,
+      "epoch": 1.359098228663446,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002751333886099402,
+      "loss": 0.5235,
       "step": 844
     },
     {
-      "epoch": 1.621414913957935,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002728329146054897,
-      "loss": 0.6677,
+      "epoch": 1.3655394524959743,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002748912989918387,
+      "loss": 0.4882,
       "step": 848
     },
     {
-      "epoch": 1.6290630975143403,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00027257115180348557,
-      "loss": 0.7128,
+      "epoch": 1.3719806763285023,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002746481442283034,
+      "loss": 0.5032,
       "step": 852
     },
     {
-      "epoch": 1.6367112810707458,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002723082607979028,
-      "loss": 0.7085,
+      "epoch": 1.3784219001610305,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002744039263930991,
+      "loss": 0.5052,
       "step": 856
     },
     {
-      "epoch": 1.644359464627151,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00027204424400851596,
-      "loss": 0.6719,
+      "epoch": 1.3848631239935587,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002741586475690571,
+      "loss": 0.5538,
       "step": 860
     },
     {
-      "epoch": 1.6520076481835564,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00027177910386546206,
-      "loss": 0.6873,
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002739123098480576,
+      "loss": 0.5457,
       "step": 864
     },
     {
-      "epoch": 1.6596558317399617,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00027151284280921794,
-      "loss": 0.6919,
+      "epoch": 1.3977455716586151,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00027366491533101147,
+      "loss": 0.5111,
       "step": 868
     },
     {
-      "epoch": 1.667304015296367,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002712454632905779,
-      "loss": 0.6967,
+      "epoch": 1.4041867954911433,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002734164661278426,
+      "loss": 0.4902,
       "step": 872
     },
     {
-      "epoch": 1.6749521988527725,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00027097696777063113,
-      "loss": 0.7067,
+      "epoch": 1.4106280193236715,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00027316696435747,
+      "loss": 0.5504,
       "step": 876
     },
     {
-      "epoch": 1.682600382409178,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00027070735872073885,
-      "loss": 0.6303,
+      "epoch": 1.4170692431561998,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00027291641214778937,
+      "loss": 0.5234,
       "step": 880
     },
     {
-      "epoch": 1.6902485659655833,
+      "epoch": 1.423510466988728,
       "grad_norm": 0.271484375,
-      "learning_rate": 0.0002704366386225119,
-      "loss": 0.6828,
+      "learning_rate": 0.0002726648116356554,
+      "loss": 0.5052,
       "step": 884
     },
     {
-      "epoch": 1.6978967495219885,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002701648099677878,
-      "loss": 0.6876,
+      "epoch": 1.4299516908212562,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000272412164966863,
+      "loss": 0.5189,
       "step": 888
     },
     {
-      "epoch": 1.7055449330783938,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002698918752586075,
-      "loss": 0.7079,
+      "epoch": 1.4363929146537842,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00027215847429612965,
+      "loss": 0.4982,
       "step": 892
     },
     {
-      "epoch": 1.7131931166347991,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00026961783700719293,
-      "loss": 0.6956,
+      "epoch": 1.4428341384863124,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002719037417870765,
+      "loss": 0.4916,
       "step": 896
     },
     {
-      "epoch": 1.7208413001912046,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002693426977359233,
-      "loss": 0.7143,
+      "epoch": 1.4492753623188406,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00027164796961221015,
+      "loss": 0.5149,
       "step": 900
     },
     {
-      "epoch": 1.72848948374761,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002690664599773122,
-      "loss": 0.6973,
+      "epoch": 1.4557165861513688,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002713911599529039,
+      "loss": 0.5636,
       "step": 904
     },
     {
-      "epoch": 1.7361376673040154,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00026878912627398434,
-      "loss": 0.68,
+      "epoch": 1.462157809983897,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027113331499937967,
+      "loss": 0.5191,
       "step": 908
     },
     {
-      "epoch": 1.7437858508604207,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0002685106991786519,
-      "loss": 0.7011,
+      "epoch": 1.4685990338164252,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00027087443695068873,
+      "loss": 0.4786,
       "step": 912
     },
     {
-      "epoch": 1.751434034416826,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00026823118125409107,
-      "loss": 0.6862,
+      "epoch": 1.4750402576489532,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002706145280146931,
+      "loss": 0.5033,
       "step": 916
     },
     {
-      "epoch": 1.7590822179732313,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002679505750731189,
-      "loss": 0.6929,
+      "epoch": 1.4814814814814814,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027035359040804703,
+      "loss": 0.4753,
       "step": 920
     },
     {
-      "epoch": 1.7667304015296366,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00026766888321856896,
-      "loss": 0.6927,
+      "epoch": 1.4879227053140096,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002700916263561778,
+      "loss": 0.5255,
       "step": 924
     },
     {
-      "epoch": 1.774378585086042,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.000267386108283268,
-      "loss": 0.6618,
+      "epoch": 1.4943639291465378,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002698286380932667,
+      "loss": 0.5472,
       "step": 928
     },
     {
-      "epoch": 1.7820267686424476,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002671022528700118,
-      "loss": 0.7173,
+      "epoch": 1.500805152979066,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002695646278622302,
+      "loss": 0.4944,
       "step": 932
     },
     {
-      "epoch": 1.7896749521988529,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00026681731959154174,
-      "loss": 0.7314,
+      "epoch": 1.5072463768115942,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002692995979147007,
+      "loss": 0.4677,
       "step": 936
     },
     {
-      "epoch": 1.7973231357552581,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00026653131107052,
-      "loss": 0.7013,
+      "epoch": 1.5136876006441224,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026903355051100734,
+      "loss": 0.5152,
       "step": 940
     },
     {
-      "epoch": 1.8049713193116634,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00026624422993950603,
-      "loss": 0.7591,
+      "epoch": 1.5201288244766507,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002687664879201565,
+      "loss": 0.5287,
       "step": 944
     },
     {
-      "epoch": 1.8126195028680687,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002659560788409321,
-      "loss": 0.6398,
+      "epoch": 1.5265700483091789,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00026849841241981313,
+      "loss": 0.5185,
       "step": 948
     },
     {
-      "epoch": 1.8202676864244742,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002656668604270788,
-      "loss": 0.6778,
+      "epoch": 1.533011272141707,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00026822932629628034,
+      "loss": 0.4925,
       "step": 952
     },
     {
-      "epoch": 1.8279158699808795,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00026537657736005094,
-      "loss": 0.6543,
+      "epoch": 1.539452495974235,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002679592318444808,
+      "loss": 0.4938,
       "step": 956
     },
     {
-      "epoch": 1.835564053537285,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.000265085232311753,
-      "loss": 0.7096,
+      "epoch": 1.5458937198067633,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002676881313679366,
+      "loss": 0.4962,
       "step": 960
     },
     {
-      "epoch": 1.8432122370936903,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00026479282796386416,
-      "loss": 0.6939,
+      "epoch": 1.5523349436392915,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002674160271787498,
+      "loss": 0.4962,
       "step": 964
     },
     {
-      "epoch": 1.8508604206500956,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00026449936700781413,
-      "loss": 0.728,
+      "epoch": 1.5587761674718197,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002671429215975828,
+      "loss": 0.5142,
       "step": 968
     },
     {
-      "epoch": 1.8585086042065009,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002642048521447581,
-      "loss": 0.6862,
+      "epoch": 1.5652173913043477,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026686881695363833,
+      "loss": 0.5361,
       "step": 972
     },
     {
-      "epoch": 1.8661567877629062,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00026390928608555195,
-      "loss": 0.6767,
+      "epoch": 1.5716586151368759,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002665937155846399,
+      "loss": 0.519,
       "step": 976
     },
     {
-      "epoch": 1.8738049713193117,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0002636126715507272,
-      "loss": 0.6229,
+      "epoch": 1.578099838969404,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002663176198368114,
+      "loss": 0.5055,
       "step": 980
     },
     {
-      "epoch": 1.8814531548757172,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.000263315011270466,
-      "loss": 0.6941,
+      "epoch": 1.5845410628019323,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002660405320648576,
+      "loss": 0.5256,
       "step": 984
     },
     {
-      "epoch": 1.8891013384321225,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00026301630798457613,
-      "loss": 0.6854,
+      "epoch": 1.5909822866344605,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002657624546319437,
+      "loss": 0.5103,
       "step": 988
     },
     {
-      "epoch": 1.8967495219885278,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00026271656444246577,
-      "loss": 0.7136,
+      "epoch": 1.5974235104669887,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002654833899096753,
+      "loss": 0.5249,
       "step": 992
     },
     {
-      "epoch": 1.904397705544933,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.000262415783403118,
-      "loss": 0.696,
+      "epoch": 1.603864734299517,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00026520334027807827,
+      "loss": 0.4895,
       "step": 996
     },
     {
-      "epoch": 1.9120458891013383,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00026211396763506546,
-      "loss": 0.6688,
+      "epoch": 1.6103059581320451,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002649223081255782,
+      "loss": 0.5061,
       "step": 1000
     },
     {
-      "epoch": 1.9196940726577438,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002618111199163651,
-      "loss": 0.6953,
+      "epoch": 1.6167471819645733,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026464029584898036,
+      "loss": 0.4781,
       "step": 1004
     },
     {
-      "epoch": 1.9273422562141491,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00026150724303457235,
-      "loss": 0.7481,
+      "epoch": 1.6231884057971016,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00026435730585344896,
+      "loss": 0.4885,
       "step": 1008
     },
     {
-      "epoch": 1.9349904397705546,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002612023397867155,
-      "loss": 0.675,
+      "epoch": 1.6296296296296298,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002640733405524869,
+      "loss": 0.5188,
       "step": 1012
     },
     {
-      "epoch": 1.94263862332696,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00026089641297927,
-      "loss": 0.6684,
+      "epoch": 1.636070853462158,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00026378840236791485,
+      "loss": 0.5386,
       "step": 1016
     },
     {
-      "epoch": 1.9502868068833652,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0002605894654281329,
-      "loss": 0.6547,
+      "epoch": 1.642512077294686,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000263502493729851,
+      "loss": 0.5438,
       "step": 1020
     },
     {
-      "epoch": 1.9579349904397705,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002602814999585963,
-      "loss": 0.7232,
+      "epoch": 1.6489533011272142,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00026321561707668995,
+      "loss": 0.5121,
       "step": 1024
     },
     {
-      "epoch": 1.9655831739961758,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002599725194053219,
-      "loss": 0.7069,
+      "epoch": 1.6553945249597424,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002629277748550823,
+      "loss": 0.4868,
       "step": 1028
     },
     {
-      "epoch": 1.9732313575525813,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002596625266123146,
-      "loss": 0.7303,
+      "epoch": 1.6618357487922706,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002626389695199134,
+      "loss": 0.5199,
       "step": 1032
     },
     {
-      "epoch": 1.9808795411089866,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00025935152443289664,
-      "loss": 0.7096,
+      "epoch": 1.6682769726247986,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002623492035342826,
+      "loss": 0.5424,
       "step": 1036
     },
     {
-      "epoch": 1.988527724665392,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00025903951572968094,
-      "loss": 0.7055,
+      "epoch": 1.6747181964573268,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00026205847936948244,
+      "loss": 0.4983,
       "step": 1040
     },
     {
-      "epoch": 1.9961759082217974,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025872650337454504,
-      "loss": 0.7108,
+      "epoch": 1.681159420289855,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00026176679950497706,
+      "loss": 0.5323,
       "step": 1044
     },
     {
-      "epoch": 2.0038240917782026,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00025841249024860453,
-      "loss": 0.5808,
+      "epoch": 1.6876006441223832,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002614741664283816,
+      "loss": 0.5964,
       "step": 1048
     },
     {
-      "epoch": 2.011472275334608,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00025809747924218667,
-      "loss": 0.58,
+      "epoch": 1.6940418679549114,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00026118058263544056,
+      "loss": 0.5227,
       "step": 1052
     },
     {
-      "epoch": 2.019120458891013,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025778147325480357,
-      "loss": 0.6208,
+      "epoch": 1.7004830917874396,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00026088605063000696,
+      "loss": 0.464,
       "step": 1056
     },
     {
-      "epoch": 2.026768642447419,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002574644751951256,
-      "loss": 0.5692,
+      "epoch": 1.7069243156199678,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002605905729240205,
+      "loss": 0.4978,
       "step": 1060
     },
     {
-      "epoch": 2.0344168260038242,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00025714648798095483,
-      "loss": 0.5891,
+      "epoch": 1.713365539452496,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00026029415203748633,
+      "loss": 0.4983,
       "step": 1064
     },
     {
-      "epoch": 2.0420650095602295,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00025682751453919776,
-      "loss": 0.6214,
+      "epoch": 1.7198067632850242,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002599967904984539,
+      "loss": 0.5166,
       "step": 1068
     },
     {
-      "epoch": 2.049713193116635,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002565075578058388,
-      "loss": 0.5889,
+      "epoch": 1.7262479871175525,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00025969849084299466,
+      "loss": 0.5683,
       "step": 1072
     },
     {
-      "epoch": 2.05736137667304,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002561866207259128,
-      "loss": 0.6099,
+      "epoch": 1.7326892109500807,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00025939925561518126,
+      "loss": 0.486,
       "step": 1076
     },
     {
-      "epoch": 2.0650095602294454,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002558647062534785,
-      "loss": 0.5722,
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002590990873670652,
+      "loss": 0.4655,
       "step": 1080
     },
     {
-      "epoch": 2.0726577437858507,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002555418173515908,
-      "loss": 0.609,
+      "epoch": 1.7455716586151369,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025879798865865533,
+      "loss": 0.4689,
       "step": 1084
     },
     {
-      "epoch": 2.0803059273422564,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.0002552179569922737,
-      "loss": 0.6158,
+      "epoch": 1.752012882447665,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002584959620578962,
+      "loss": 0.424,
       "step": 1088
     },
     {
-      "epoch": 2.0879541108986617,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00025489312815649314,
-      "loss": 0.588,
+      "epoch": 1.7584541062801933,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00025819301014064574,
+      "loss": 0.5134,
       "step": 1092
     },
     {
-      "epoch": 2.095602294455067,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00025456733383412926,
-      "loss": 0.6278,
+      "epoch": 1.7648953301127213,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002578891354906537,
+      "loss": 0.4893,
       "step": 1096
     },
     {
-      "epoch": 2.1032504780114722,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.000254240577023949,
-      "loss": 0.5915,
+      "epoch": 1.7713365539452495,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00025758434069953927,
+      "loss": 0.4887,
       "step": 1100
     },
     {
-      "epoch": 2.1108986615678775,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00025391286073357856,
-      "loss": 0.5764,
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002572786283667692,
+      "loss": 0.5153,
       "step": 1104
     },
     {
-      "epoch": 2.118546845124283,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002535841879794755,
-      "loss": 0.6146,
+      "epoch": 1.7842190016103059,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00025697200109963563,
+      "loss": 0.5056,
       "step": 1108
     },
     {
-      "epoch": 2.126195028680688,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002532545617869014,
-      "loss": 0.5794,
+      "epoch": 1.790660225442834,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002566644615132337,
+      "loss": 0.5319,
       "step": 1112
     },
     {
-      "epoch": 2.133843212237094,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002529239851898935,
-      "loss": 0.6412,
+      "epoch": 1.7971014492753623,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00025635601223043933,
+      "loss": 0.5182,
       "step": 1116
     },
     {
-      "epoch": 2.141491395793499,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00025259246123123706,
-      "loss": 0.6288,
+      "epoch": 1.8035426731078905,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000256046655881887,
+      "loss": 0.5028,
       "step": 1120
     },
     {
-      "epoch": 2.1491395793499044,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002522599929624375,
-      "loss": 0.5644,
+      "epoch": 1.8099838969404187,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.000255736395105947,
+      "loss": 0.5006,
       "step": 1124
     },
     {
-      "epoch": 2.1567877629063097,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00025192658344369193,
-      "loss": 0.6219,
+      "epoch": 1.816425120772947,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002554252325487032,
+      "loss": 0.5234,
       "step": 1128
     },
     {
-      "epoch": 2.164435946462715,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00025159223574386114,
-      "loss": 0.6015,
+      "epoch": 1.8228663446054751,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002551131708639303,
+      "loss": 0.5544,
       "step": 1132
     },
     {
-      "epoch": 2.1720841300191203,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025125695294044156,
-      "loss": 0.612,
+      "epoch": 1.8293075684380034,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00025480021271307156,
+      "loss": 0.4766,
       "step": 1136
     },
     {
-      "epoch": 2.179732313575526,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002509207381195366,
-      "loss": 0.5878,
+      "epoch": 1.8357487922705316,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00025448636076521534,
+      "loss": 0.4615,
       "step": 1140
     },
     {
-      "epoch": 2.1873804971319313,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002505835943758286,
-      "loss": 0.6134,
+      "epoch": 1.8421900161030595,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002541716176970732,
+      "loss": 0.504,
       "step": 1144
     },
     {
-      "epoch": 2.1950286806883366,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00025024552481254993,
-      "loss": 0.5663,
+      "epoch": 1.8486312399355878,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002538559861929566,
+      "loss": 0.5873,
       "step": 1148
     },
     {
-      "epoch": 2.202676864244742,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002499065325414547,
-      "loss": 0.5685,
+      "epoch": 1.855072463768116,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.000253539468944754,
+      "loss": 0.5917,
       "step": 1152
     },
     {
-      "epoch": 2.210325047801147,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00024956662068279027,
-      "loss": 0.5839,
+      "epoch": 1.8615136876006442,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002532220686519081,
+      "loss": 0.4924,
       "step": 1156
     },
     {
-      "epoch": 2.2179732313575524,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00024922579236526807,
-      "loss": 0.611,
+      "epoch": 1.8679549114331722,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00025290378802139273,
+      "loss": 0.4582,
       "step": 1160
     },
     {
-      "epoch": 2.2256214149139577,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024888405072603513,
-      "loss": 0.6218,
+      "epoch": 1.8743961352657004,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002525846297676896,
+      "loss": 0.5639,
       "step": 1164
     },
     {
-      "epoch": 2.2332695984703634,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002485413989106452,
-      "loss": 0.5743,
+      "epoch": 1.8808373590982286,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002522645966127655,
+      "loss": 0.5198,
       "step": 1168
     },
     {
-      "epoch": 2.2409177820267687,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00024819784007302966,
-      "loss": 0.5921,
+      "epoch": 1.8872785829307568,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002519436912860488,
+      "loss": 0.4766,
       "step": 1172
     },
     {
-      "epoch": 2.248565965583174,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00024785337737546863,
-      "loss": 0.5444,
+      "epoch": 1.893719806763285,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002516219165244062,
+      "loss": 0.4583,
       "step": 1176
     },
     {
-      "epoch": 2.2562141491395793,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002475080139885617,
-      "loss": 0.5823,
+      "epoch": 1.9001610305958132,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002512992750721195,
+      "loss": 0.549,
       "step": 1180
     },
     {
-      "epoch": 2.2638623326959846,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00024716175309119875,
-      "loss": 0.5788,
+      "epoch": 1.9066022544283414,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002509757696808622,
+      "loss": 0.4792,
       "step": 1184
     },
     {
-      "epoch": 2.27151051625239,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00024681459787053106,
-      "loss": 0.5666,
+      "epoch": 1.9130434782608696,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002506514031096758,
+      "loss": 0.4834,
       "step": 1188
     },
     {
-      "epoch": 2.2791586998087956,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0002464665515219415,
-      "loss": 0.5988,
+      "epoch": 1.9194847020933978,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00025032617812494664,
+      "loss": 0.4969,
       "step": 1192
     },
     {
-      "epoch": 2.286806883365201,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0002461176172490153,
-      "loss": 0.5851,
+      "epoch": 1.925925925925926,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00025000009750038196,
+      "loss": 0.5553,
       "step": 1196
     },
     {
-      "epoch": 2.294455066921606,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0002457677982635107,
-      "loss": 0.6328,
+      "epoch": 1.9323671497584543,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00024967316401698647,
+      "loss": 0.536,
       "step": 1200
     },
     {
-      "epoch": 2.3021032504780115,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.000245417097785329,
-      "loss": 0.6264,
+      "epoch": 1.9388083735909822,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00024934538046303856,
+      "loss": 0.4848,
       "step": 1204
     },
     {
-      "epoch": 2.3097514340344167,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00024506551904248546,
-      "loss": 0.5853,
+      "epoch": 1.9452495974235104,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002490167496340664,
+      "loss": 0.4984,
       "step": 1208
     },
     {
-      "epoch": 2.317399617590822,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00024471306527107915,
-      "loss": 0.571,
+      "epoch": 1.9516908212560387,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002486872743328244,
+      "loss": 0.4993,
       "step": 1212
     },
     {
-      "epoch": 2.3250478011472273,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002443597397152634,
-      "loss": 0.6076,
+      "epoch": 1.9581320450885669,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000248356957369269,
+      "loss": 0.5265,
       "step": 1216
     },
     {
-      "epoch": 2.332695984703633,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00024400554562721585,
-      "loss": 0.6026,
+      "epoch": 1.9645732689210949,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002480258015605349,
+      "loss": 0.5287,
       "step": 1220
     },
     {
-      "epoch": 2.3403441682600383,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024365048626710843,
-      "loss": 0.6196,
+      "epoch": 1.971014492753623,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002476938097309108,
+      "loss": 0.5616,
       "step": 1224
     },
     {
-      "epoch": 2.3479923518164436,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00024329456490307757,
-      "loss": 0.5704,
+      "epoch": 1.9774557165861513,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002473609847118156,
+      "loss": 0.4542,
       "step": 1228
     },
     {
-      "epoch": 2.355640535372849,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00024293778481119396,
-      "loss": 0.6217,
+      "epoch": 1.9838969404186795,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002470273293417741,
+      "loss": 0.4813,
       "step": 1232
     },
     {
-      "epoch": 2.363288718929254,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002425801492754324,
-      "loss": 0.6238,
+      "epoch": 1.9903381642512077,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00024669284646639287,
+      "loss": 0.5336,
       "step": 1236
     },
     {
-      "epoch": 2.3709369024856595,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00024222166158764161,
-      "loss": 0.5945,
+      "epoch": 1.996779388083736,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024635753893833585,
+      "loss": 0.5528,
       "step": 1240
     },
     {
-      "epoch": 2.378585086042065,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00024186232504751397,
-      "loss": 0.5605,
+      "epoch": 2.003220611916264,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00024602140961730006,
+      "loss": 0.4706,
       "step": 1244
     },
     {
-      "epoch": 2.3862332695984705,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002415021429625551,
-      "loss": 0.6451,
+      "epoch": 2.0096618357487923,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024568446136999134,
+      "loss": 0.4093,
       "step": 1248
     },
     {
-      "epoch": 2.3938814531548758,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024114111864805338,
-      "loss": 0.5581,
+      "epoch": 2.0161030595813205,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00024534669707009974,
+      "loss": 0.3899,
       "step": 1252
     },
     {
-      "epoch": 2.401529636711281,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00024077925542704949,
-      "loss": 0.6212,
+      "epoch": 2.0225442834138487,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002450081195982752,
+      "loss": 0.361,
       "step": 1256
     },
     {
-      "epoch": 2.4091778202676863,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002404165566303057,
-      "loss": 0.6121,
+      "epoch": 2.028985507246377,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00024466873184210273,
+      "loss": 0.3999,
       "step": 1260
     },
     {
-      "epoch": 2.4168260038240916,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00024005302559627561,
-      "loss": 0.578,
+      "epoch": 2.035426731078905,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00024432853669607786,
+      "loss": 0.3753,
       "step": 1264
     },
     {
-      "epoch": 2.424474187380497,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00023968866567107282,
-      "loss": 0.6353,
+      "epoch": 2.0418679549114334,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00024398753706158225,
+      "loss": 0.3951,
       "step": 1268
     },
     {
-      "epoch": 2.4321223709369026,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00023932348020844064,
-      "loss": 0.5811,
+      "epoch": 2.0483091787439616,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00024364573584685848,
+      "loss": 0.3791,
       "step": 1272
     },
     {
-      "epoch": 2.439770554493308,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00023895747256972083,
-      "loss": 0.5834,
+      "epoch": 2.0547504025764893,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00024330313596698553,
+      "loss": 0.4148,
       "step": 1276
     },
     {
-      "epoch": 2.447418738049713,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00023859064612382315,
-      "loss": 0.6234,
+      "epoch": 2.0611916264090175,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024295974034385396,
+      "loss": 0.3767,
       "step": 1280
     },
     {
-      "epoch": 2.4550669216061185,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002382230042471938,
-      "loss": 0.6298,
+      "epoch": 2.0676328502415457,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00024261555190614072,
+      "loss": 0.3743,
       "step": 1284
     },
     {
-      "epoch": 2.462715105162524,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0002378545503237846,
-      "loss": 0.6834,
+      "epoch": 2.074074074074074,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00024227057358928452,
+      "loss": 0.3847,
       "step": 1288
     },
     {
-      "epoch": 2.470363288718929,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00023748528774502194,
-      "loss": 0.6176,
+      "epoch": 2.080515297906602,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024192480833546044,
+      "loss": 0.3627,
       "step": 1292
     },
     {
-      "epoch": 2.478011472275335,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023711521990977554,
-      "loss": 0.6141,
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00024157825909355523,
+      "loss": 0.4324,
       "step": 1296
     },
     {
-      "epoch": 2.48565965583174,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00023674435022432683,
-      "loss": 0.5958,
+      "epoch": 2.0933977455716586,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002412309288191417,
+      "loss": 0.4302,
       "step": 1300
     },
     {
-      "epoch": 2.4933078393881454,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0002363726821023381,
-      "loss": 0.6249,
+      "epoch": 2.099838969404187,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00024088282047445396,
+      "loss": 0.3788,
       "step": 1304
     },
     {
-      "epoch": 2.5009560229445507,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00023600021896482063,
-      "loss": 0.5865,
+      "epoch": 2.106280193236715,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00024053393702836185,
+      "loss": 0.399,
       "step": 1308
     },
     {
-      "epoch": 2.508604206500956,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002356269642401036,
-      "loss": 0.5982,
+      "epoch": 2.112721417069243,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002401842814563457,
+      "loss": 0.387,
       "step": 1312
     },
     {
-      "epoch": 2.5162523900573612,
+      "epoch": 2.1191626409017714,
       "grad_norm": 0.287109375,
-      "learning_rate": 0.0002352529213638022,
-      "loss": 0.5722,
+      "learning_rate": 0.00023983385674047113,
+      "loss": 0.3905,
       "step": 1316
     },
     {
-      "epoch": 2.5239005736137665,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0002348780937787862,
-      "loss": 0.6308,
+      "epoch": 2.1256038647342996,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00023948266586936324,
+      "loss": 0.3715,
       "step": 1320
     },
     {
-      "epoch": 2.5315487571701722,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00023450248493514817,
-      "loss": 0.6104,
+      "epoch": 2.132045088566828,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00023913071183818155,
+      "loss": 0.4474,
       "step": 1324
     },
     {
-      "epoch": 2.5391969407265775,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00023412609829017182,
-      "loss": 0.6053,
+      "epoch": 2.138486312399356,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00023877799764859416,
+      "loss": 0.3759,
       "step": 1328
     },
     {
-      "epoch": 2.546845124282983,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00023374893730830005,
-      "loss": 0.5447,
+      "epoch": 2.1449275362318843,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00023842452630875216,
+      "loss": 0.373,
       "step": 1332
     },
     {
-      "epoch": 2.554493307839388,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00023337100546110313,
-      "loss": 0.5987,
+      "epoch": 2.151368760064412,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002380703008332643,
+      "loss": 0.4218,
       "step": 1336
     },
     {
-      "epoch": 2.5621414913957934,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002329923062272468,
-      "loss": 0.6037,
+      "epoch": 2.1578099838969402,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002377153242431708,
+      "loss": 0.4234,
       "step": 1340
     },
     {
-      "epoch": 2.569789674952199,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0002326128430924602,
-      "loss": 0.6055,
+      "epoch": 2.1642512077294684,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00023735959956591786,
+      "loss": 0.3971,
       "step": 1344
     },
     {
-      "epoch": 2.5774378585086044,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00023223261954950363,
-      "loss": 0.6163,
+      "epoch": 2.1706924315619966,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002370031298353319,
+      "loss": 0.4211,
       "step": 1348
     },
     {
-      "epoch": 2.5850860420650097,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00023185163909813678,
-      "loss": 0.6272,
+      "epoch": 2.177133655394525,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00023664591809159353,
+      "loss": 0.3972,
       "step": 1352
     },
     {
-      "epoch": 2.592734225621415,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023146990524508613,
-      "loss": 0.6303,
+      "epoch": 2.183574879227053,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00023628796738121169,
+      "loss": 0.4185,
       "step": 1356
     },
     {
-      "epoch": 2.6003824091778203,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00023108742150401284,
-      "loss": 0.5862,
+      "epoch": 2.1900161030595813,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023592928075699763,
+      "loss": 0.402,
       "step": 1360
     },
     {
-      "epoch": 2.6080305927342256,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00023070419139548044,
-      "loss": 0.6083,
+      "epoch": 2.1964573268921095,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00023556986127803894,
+      "loss": 0.4056,
       "step": 1364
     },
     {
-      "epoch": 2.615678776290631,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00023032021844692242,
-      "loss": 0.6466,
+      "epoch": 2.2028985507246377,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00023520971200967334,
+      "loss": 0.4506,
       "step": 1368
     },
     {
-      "epoch": 2.623326959847036,
+      "epoch": 2.209339774557166,
       "grad_norm": 0.302734375,
-      "learning_rate": 0.0002299355061926096,
-      "loss": 0.6236,
+      "learning_rate": 0.00023484883602346274,
+      "loss": 0.4093,
       "step": 1372
     },
     {
-      "epoch": 2.6309751434034414,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00022955005817361783,
-      "loss": 0.6203,
+      "epoch": 2.215780998389694,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002344872363971668,
+      "loss": 0.4717,
       "step": 1376
     },
     {
-      "epoch": 2.638623326959847,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022916387793779533,
-      "loss": 0.6101,
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00023412491621471694,
+      "loss": 0.3948,
       "step": 1380
     },
     {
-      "epoch": 2.6462715105162524,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022877696903972984,
-      "loss": 0.5361,
+      "epoch": 2.2286634460547505,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023376187856618972,
+      "loss": 0.3925,
       "step": 1384
     },
     {
-      "epoch": 2.6539196940726577,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022838933504071618,
-      "loss": 0.594,
+      "epoch": 2.2351046698872787,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00023339812654778083,
+      "loss": 0.4324,
       "step": 1388
     },
     {
-      "epoch": 2.661567877629063,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002280009795087233,
-      "loss": 0.6323,
+      "epoch": 2.241545893719807,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002330336632617784,
+      "loss": 0.4557,
       "step": 1392
     },
     {
-      "epoch": 2.6692160611854687,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00022761190601836142,
-      "loss": 0.5863,
+      "epoch": 2.247987117552335,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00023266849181653683,
+      "loss": 0.4301,
       "step": 1396
     },
     {
-      "epoch": 2.676864244741874,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00022722211815084944,
-      "loss": 0.6109,
+      "epoch": 2.2544283413848634,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00023230261532644985,
+      "loss": 0.3799,
       "step": 1400
     },
     {
-      "epoch": 2.6845124282982793,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0002268316194939815,
-      "loss": 0.6062,
+      "epoch": 2.260869565217391,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002319360369119245,
+      "loss": 0.3826,
       "step": 1404
     },
     {
-      "epoch": 2.6921606118546846,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0002264404136420941,
-      "loss": 0.5996,
+      "epoch": 2.2673107890499193,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00023156875969935405,
+      "loss": 0.3862,
       "step": 1408
     },
     {
-      "epoch": 2.69980879541109,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0002260485041960334,
-      "loss": 0.6301,
+      "epoch": 2.2737520128824475,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023120078682109158,
+      "loss": 0.4269,
       "step": 1412
     },
     {
-      "epoch": 2.707456978967495,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022565589476312157,
-      "loss": 0.5807,
+      "epoch": 2.2801932367149758,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00023083212141542328,
+      "loss": 0.4139,
       "step": 1416
     },
     {
-      "epoch": 2.7151051625239004,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00022526258895712377,
-      "loss": 0.6542,
+      "epoch": 2.286634460547504,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00023046276662654143,
+      "loss": 0.3579,
       "step": 1420
     },
     {
-      "epoch": 2.7227533460803057,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022486859039821513,
-      "loss": 0.6001,
+      "epoch": 2.293075684380032,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00023009272560451803,
+      "loss": 0.4,
       "step": 1424
     },
     {
-      "epoch": 2.730401529636711,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00022447390271294697,
-      "loss": 0.5997,
+      "epoch": 2.2995169082125604,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022972200150527745,
+      "loss": 0.3937,
       "step": 1428
     },
     {
-      "epoch": 2.7380497131931167,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022407852953421382,
-      "loss": 0.6401,
+      "epoch": 2.3059581320450886,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00022935059749056992,
+      "loss": 0.4553,
       "step": 1432
     },
     {
-      "epoch": 2.745697896749522,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022368247450121965,
-      "loss": 0.5744,
+      "epoch": 2.312399355877617,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00022897851672794417,
+      "loss": 0.396,
       "step": 1436
     },
     {
-      "epoch": 2.7533460803059273,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00022328574125944476,
-      "loss": 0.5853,
+      "epoch": 2.318840579710145,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00022860576239072084,
+      "loss": 0.5137,
       "step": 1440
     },
     {
-      "epoch": 2.7609942638623326,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00022288833346061182,
-      "loss": 0.5861,
+      "epoch": 2.325281803542673,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022823233765796502,
+      "loss": 0.4085,
       "step": 1444
     },
     {
-      "epoch": 2.768642447418738,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002224902547626526,
-      "loss": 0.6194,
+      "epoch": 2.3317230273752014,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002278582457144595,
+      "loss": 0.3963,
       "step": 1448
     },
     {
-      "epoch": 2.7762906309751436,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00022209150882967398,
-      "loss": 0.604,
+      "epoch": 2.3381642512077296,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00022748348975067733,
+      "loss": 0.4377,
       "step": 1452
     },
     {
-      "epoch": 2.783938814531549,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00022169209933192458,
-      "loss": 0.6312,
+      "epoch": 2.3446054750402574,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00022710807296275472,
+      "loss": 0.4275,
       "step": 1456
     },
     {
-      "epoch": 2.791586998087954,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002212920299457606,
-      "loss": 0.6312,
+      "epoch": 2.3510466988727856,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002267319985524637,
+      "loss": 0.4089,
       "step": 1460
     },
     {
-      "epoch": 2.7992351816443595,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.0002208913043536123,
-      "loss": 0.6089,
+      "epoch": 2.357487922705314,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00022635526972718508,
+      "loss": 0.4386,
       "step": 1464
     },
     {
-      "epoch": 2.8068833652007648,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00022048992624394988,
-      "loss": 0.6349,
+      "epoch": 2.363929146537842,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002259778896998807,
+      "loss": 0.4172,
       "step": 1468
     },
     {
-      "epoch": 2.81453154875717,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00022008789931124976,
-      "loss": 0.591,
+      "epoch": 2.3703703703703702,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00022559986168906637,
+      "loss": 0.4022,
       "step": 1472
     },
     {
-      "epoch": 2.8221797323135753,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002196852272559603,
-      "loss": 0.6085,
+      "epoch": 2.3768115942028984,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00022522118891878418,
+      "loss": 0.4665,
       "step": 1476
     },
     {
-      "epoch": 2.8298279158699806,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00021928191378446795,
-      "loss": 0.6284,
+      "epoch": 2.3832528180354267,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00022484187461857517,
+      "loss": 0.3916,
       "step": 1480
     },
     {
-      "epoch": 2.8374760994263863,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021887796260906304,
-      "loss": 0.5796,
+      "epoch": 2.389694041867955,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00022446192202345156,
+      "loss": 0.3918,
       "step": 1484
     },
     {
-      "epoch": 2.8451242829827916,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021847337744790562,
-      "loss": 0.5739,
+      "epoch": 2.396135265700483,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00022408133437386968,
+      "loss": 0.4198,
       "step": 1488
     },
     {
-      "epoch": 2.852772466539197,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.0002180681620249913,
-      "loss": 0.6,
+      "epoch": 2.4025764895330113,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00022370011491570162,
+      "loss": 0.3635,
       "step": 1492
     },
     {
-      "epoch": 2.860420650095602,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00021766232007011682,
-      "loss": 0.6113,
+      "epoch": 2.4090177133655395,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000223318266900208,
+      "loss": 0.4297,
       "step": 1496
     },
     {
-      "epoch": 2.8680688336520075,
+      "epoch": 2.4154589371980677,
       "grad_norm": 0.29296875,
-      "learning_rate": 0.0002172558553188459,
-      "loss": 0.5949,
+      "learning_rate": 0.00022293579358401023,
+      "loss": 0.3819,
       "step": 1500
     },
     {
-      "epoch": 2.875717017208413,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00021684877151247485,
-      "loss": 0.5929,
+      "epoch": 2.421900161030596,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002225526982290625,
+      "loss": 0.4068,
       "step": 1504
     },
     {
-      "epoch": 2.8833652007648185,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021644107239799786,
-      "loss": 0.6224,
+      "epoch": 2.428341384863124,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00022216898410262428,
+      "loss": 0.3808,
       "step": 1508
     },
     {
-      "epoch": 2.891013384321224,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00021603276172807288,
-      "loss": 0.6439,
+      "epoch": 2.4347826086956523,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022178465447723214,
+      "loss": 0.4037,
       "step": 1512
     },
     {
-      "epoch": 2.898661567877629,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00021562384326098688,
-      "loss": 0.5835,
+      "epoch": 2.4412238325281805,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000221399712630672,
+      "loss": 0.452,
       "step": 1516
     },
     {
-      "epoch": 2.9063097514340344,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0002152143207606211,
-      "loss": 0.6,
+      "epoch": 2.4476650563607087,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002210141618459513,
+      "loss": 0.4127,
       "step": 1520
     },
     {
-      "epoch": 2.9139579349904396,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00021480419799641692,
-      "loss": 0.5959,
+      "epoch": 2.454106280193237,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00022062800541127064,
+      "loss": 0.3894,
       "step": 1524
     },
     {
-      "epoch": 2.921606118546845,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.0002143934787433406,
-      "loss": 0.6111,
+      "epoch": 2.4605475040257647,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00022024124661999613,
+      "loss": 0.4256,
       "step": 1528
     },
     {
-      "epoch": 2.92925430210325,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00021398216678184884,
-      "loss": 0.6072,
+      "epoch": 2.466988727858293,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00021985388877063104,
+      "loss": 0.4556,
       "step": 1532
     },
     {
-      "epoch": 2.936902485659656,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00021357026589785392,
-      "loss": 0.5744,
+      "epoch": 2.473429951690821,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00021946593516678777,
+      "loss": 0.4504,
       "step": 1536
     },
     {
-      "epoch": 2.9445506692160612,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00021315777988268876,
-      "loss": 0.603,
+      "epoch": 2.4798711755233493,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00021907738911715964,
+      "loss": 0.4062,
       "step": 1540
     },
     {
-      "epoch": 2.9521988527724665,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00021274471253307224,
-      "loss": 0.6364,
+      "epoch": 2.4863123993558776,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021868825393549275,
+      "loss": 0.4386,
       "step": 1544
     },
     {
-      "epoch": 2.959847036328872,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00021233106765107407,
-      "loss": 0.5766,
+      "epoch": 2.4927536231884058,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002182985329405576,
+      "loss": 0.3559,
       "step": 1548
     },
     {
-      "epoch": 2.967495219885277,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00021191684904407976,
-      "loss": 0.5715,
+      "epoch": 2.499194847020934,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00021790822945612088,
+      "loss": 0.4244,
       "step": 1552
     },
     {
-      "epoch": 2.975143403441683,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002115020605247558,
-      "loss": 0.6015,
+      "epoch": 2.505636070853462,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0002175173468109171,
+      "loss": 0.4028,
       "step": 1556
     },
     {
-      "epoch": 2.982791586998088,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00021108670591101433,
-      "loss": 0.6017,
+      "epoch": 2.5120772946859904,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00021712588833862014,
+      "loss": 0.3726,
       "step": 1560
     },
     {
-      "epoch": 2.9904397705544934,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021067078902597814,
-      "loss": 0.6201,
+      "epoch": 2.5185185185185186,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00021673385737781492,
+      "loss": 0.481,
       "step": 1564
     },
     {
-      "epoch": 2.9980879541108987,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002102543136979454,
-      "loss": 0.6178,
+      "epoch": 2.524959742351047,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00021634125727196883,
+      "loss": 0.3778,
       "step": 1568
     },
     {
-      "epoch": 3.005736137667304,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00020983728376035448,
-      "loss": 0.464,
+      "epoch": 2.531400966183575,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00021594809136940327,
+      "loss": 0.4438,
       "step": 1572
     },
     {
-      "epoch": 3.0133843212237093,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00020941970305174862,
-      "loss": 0.4966,
+      "epoch": 2.537842190016103,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00021555436302326514,
+      "loss": 0.4399,
       "step": 1576
     },
     {
-      "epoch": 3.0210325047801145,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00020900157541574066,
-      "loss": 0.5019,
+      "epoch": 2.544283413848631,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00021516007559149803,
+      "loss": 0.3979,
       "step": 1580
     },
     {
-      "epoch": 3.0286806883365203,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00020858290470097762,
-      "loss": 0.477,
+      "epoch": 2.550724637681159,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00021476523243681397,
+      "loss": 0.4085,
       "step": 1584
     },
     {
-      "epoch": 3.0363288718929256,
+      "epoch": 2.5571658615136874,
       "grad_norm": 0.291015625,
-      "learning_rate": 0.00020816369476110512,
-      "loss": 0.5126,
+      "learning_rate": 0.0002143698369266643,
+      "loss": 0.3875,
       "step": 1588
     },
     {
-      "epoch": 3.043977055449331,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002077439494547324,
-      "loss": 0.5319,
+      "epoch": 2.5636070853462156,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002139738924332113,
+      "loss": 0.4288,
       "step": 1592
     },
     {
-      "epoch": 3.051625239005736,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00020732367264539612,
-      "loss": 0.4976,
+      "epoch": 2.570048309178744,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002135774023332992,
+      "loss": 0.4155,
       "step": 1596
     },
     {
-      "epoch": 3.0592734225621414,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00020690286820152534,
-      "loss": 0.4638,
+      "epoch": 2.576489533011272,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021318037000842558,
+      "loss": 0.377,
       "step": 1600
     },
     {
-      "epoch": 3.0669216061185467,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0002064815399964057,
-      "loss": 0.5326,
+      "epoch": 2.5829307568438002,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00021278279884471242,
+      "loss": 0.4134,
       "step": 1604
     },
     {
-      "epoch": 3.0745697896749524,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00020605969190814374,
-      "loss": 0.5066,
+      "epoch": 2.5893719806763285,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002123846922328771,
+      "loss": 0.3668,
       "step": 1608
     },
     {
-      "epoch": 3.0822179732313577,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002056373278196313,
-      "loss": 0.4611,
+      "epoch": 2.5958132045088567,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00021198605356820377,
+      "loss": 0.4207,
       "step": 1612
     },
     {
-      "epoch": 3.089866156787763,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002052144516185097,
-      "loss": 0.511,
+      "epoch": 2.602254428341385,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00021158688625051416,
+      "loss": 0.434,
       "step": 1616
     },
     {
-      "epoch": 3.0975143403441683,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020479106719713402,
-      "loss": 0.5338,
+      "epoch": 2.608695652173913,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021118719368413866,
+      "loss": 0.3963,
       "step": 1620
     },
     {
-      "epoch": 3.1051625239005736,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00020436717845253723,
-      "loss": 0.5384,
+      "epoch": 2.6151368760064413,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0002107869792778873,
+      "loss": 0.4366,
       "step": 1624
     },
     {
-      "epoch": 3.112810707456979,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0002039427892863943,
-      "loss": 0.5188,
+      "epoch": 2.6215780998389695,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00021038624644502063,
+      "loss": 0.3604,
       "step": 1628
     },
     {
-      "epoch": 3.120458891013384,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00020351790360498636,
-      "loss": 0.5045,
+      "epoch": 2.6280193236714977,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00020998499860322073,
+      "loss": 0.4029,
       "step": 1632
     },
     {
-      "epoch": 3.12810707456979,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00020309252531916475,
-      "loss": 0.4894,
+      "epoch": 2.634460547504026,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020958323917456186,
+      "loss": 0.429,
       "step": 1636
     },
     {
-      "epoch": 3.135755258126195,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00020266665834431486,
-      "loss": 0.5241,
+      "epoch": 2.640901771336554,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00020918097158548145,
+      "loss": 0.426,
       "step": 1640
     },
     {
-      "epoch": 3.1434034416826004,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00020224030660032023,
-      "loss": 0.4838,
+      "epoch": 2.6473429951690823,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002087781992667509,
+      "loss": 0.4129,
       "step": 1644
     },
     {
-      "epoch": 3.1510516252390057,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00020181347401152652,
-      "loss": 0.5449,
+      "epoch": 2.6537842190016105,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000208374925653446,
+      "loss": 0.3946,
       "step": 1648
     },
     {
-      "epoch": 3.158699808795411,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0002013861645067054,
-      "loss": 0.5428,
+      "epoch": 2.6602254428341388,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00020797115418491816,
+      "loss": 0.3564,
       "step": 1652
     },
     {
-      "epoch": 3.1663479923518163,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00020095838201901798,
-      "loss": 0.5506,
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00020756688830476453,
+      "loss": 0.4553,
       "step": 1656
     },
     {
-      "epoch": 3.173996175908222,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00020053013048597926,
-      "loss": 0.4983,
+      "epoch": 2.6731078904991947,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002071621314607991,
+      "loss": 0.4497,
       "step": 1660
     },
     {
-      "epoch": 3.1816443594646273,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00020010141384942148,
-      "loss": 0.4941,
+      "epoch": 2.679549114331723,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020675688710502293,
+      "loss": 0.3987,
       "step": 1664
     },
     {
-      "epoch": 3.1892925430210326,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001996722360554577,
-      "loss": 0.5406,
+      "epoch": 2.685990338164251,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00020635115869359498,
+      "loss": 0.4695,
       "step": 1668
     },
     {
-      "epoch": 3.196940726577438,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00019924260105444602,
-      "loss": 0.484,
+      "epoch": 2.6924315619967794,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002059449496868024,
+      "loss": 0.4566,
       "step": 1672
     },
     {
-      "epoch": 3.204588910133843,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00019881251280095261,
-      "loss": 0.5374,
+      "epoch": 2.6988727858293076,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020553826354903121,
+      "loss": 0.4199,
       "step": 1676
     },
     {
-      "epoch": 3.2122370936902485,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00019838197525371583,
-      "loss": 0.4977,
+      "epoch": 2.7053140096618358,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00020513110374873676,
+      "loss": 0.3612,
       "step": 1680
     },
     {
-      "epoch": 3.2198852772466537,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.0001979509923756094,
-      "loss": 0.5462,
+      "epoch": 2.711755233494364,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00020472347375841384,
+      "loss": 0.383,
       "step": 1684
     },
     {
-      "epoch": 3.2275334608030595,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0001975195681336061,
-      "loss": 0.4945,
+      "epoch": 2.718196457326892,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002043153770545675,
+      "loss": 0.4051,
       "step": 1688
     },
     {
-      "epoch": 3.2351816443594648,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00019708770649874132,
-      "loss": 0.5042,
+      "epoch": 2.7246376811594204,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00020390681711768312,
+      "loss": 0.4408,
       "step": 1692
     },
     {
-      "epoch": 3.24282982791587,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00019665541144607627,
-      "loss": 0.5369,
+      "epoch": 2.7310789049919486,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00020349779743219682,
+      "loss": 0.4155,
       "step": 1696
     },
     {
-      "epoch": 3.2504780114722753,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00019622268695466166,
-      "loss": 0.5121,
+      "epoch": 2.7375201288244764,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002030883214864657,
+      "loss": 0.4164,
       "step": 1700
     },
     {
-      "epoch": 3.2581261950286806,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.000195789537007501,
-      "loss": 0.4984,
+      "epoch": 2.7439613526570046,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002026783927727381,
+      "loss": 0.4013,
       "step": 1704
     },
     {
-      "epoch": 3.265774378585086,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00019535596559151376,
-      "loss": 0.5391,
+      "epoch": 2.750402576489533,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00020226801478712383,
+      "loss": 0.3839,
       "step": 1708
     },
     {
-      "epoch": 3.2734225621414916,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00019492197669749892,
-      "loss": 0.4778,
+      "epoch": 2.756843800322061,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00020185719102956438,
+      "loss": 0.4691,
       "step": 1712
     },
     {
-      "epoch": 3.281070745697897,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00019448757432009807,
-      "loss": 0.5131,
+      "epoch": 2.763285024154589,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002014459250038031,
+      "loss": 0.3949,
       "step": 1716
     },
     {
-      "epoch": 3.288718929254302,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00019405276245775877,
-      "loss": 0.4901,
+      "epoch": 2.7697262479871174,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00020103422021735507,
+      "loss": 0.3918,
       "step": 1720
     },
     {
-      "epoch": 3.2963671128107075,
+      "epoch": 2.7761674718196456,
       "grad_norm": 0.326171875,
-      "learning_rate": 0.00019361754511269753,
-      "loss": 0.5426,
+      "learning_rate": 0.00020062208018147755,
+      "loss": 0.4027,
       "step": 1724
     },
     {
-      "epoch": 3.3040152963671128,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00019318192629086327,
-      "loss": 0.5105,
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00020020950841113984,
+      "loss": 0.4319,
       "step": 1728
     },
     {
-      "epoch": 3.311663479923518,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00019274591000190028,
-      "loss": 0.5448,
+      "epoch": 2.789049919484702,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019979650842499324,
+      "loss": 0.4255,
       "step": 1732
     },
     {
-      "epoch": 3.3193116634799233,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00019230950025911123,
-      "loss": 0.5079,
+      "epoch": 2.7954911433172303,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019938308374534115,
+      "loss": 0.4403,
       "step": 1736
     },
     {
-      "epoch": 3.3269598470363286,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0001918727010794204,
-      "loss": 0.5185,
+      "epoch": 2.8019323671497585,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019896923789810905,
+      "loss": 0.4311,
       "step": 1740
     },
     {
-      "epoch": 3.3346080305927344,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001914355164833366,
-      "loss": 0.4918,
+      "epoch": 2.8083735909822867,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019855497441281436,
+      "loss": 0.407,
       "step": 1744
     },
     {
-      "epoch": 3.3422562141491396,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00019099795049491621,
-      "loss": 0.5051,
+      "epoch": 2.814814814814815,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019814029682253644,
+      "loss": 0.4184,
       "step": 1748
     },
     {
-      "epoch": 3.349904397705545,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00019056000714172617,
-      "loss": 0.5295,
+      "epoch": 2.821256038647343,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019772520866388605,
+      "loss": 0.3812,
       "step": 1752
     },
     {
-      "epoch": 3.35755258126195,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00019012169045480676,
-      "loss": 0.5455,
+      "epoch": 2.8276972624798713,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019730971347697602,
+      "loss": 0.4228,
       "step": 1756
     },
     {
-      "epoch": 3.3652007648183555,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00018968300446863478,
-      "loss": 0.515,
+      "epoch": 2.8341384863123995,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019689381480539014,
+      "loss": 0.4321,
       "step": 1760
     },
     {
-      "epoch": 3.3728489483747612,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00018924395322108607,
-      "loss": 0.4868,
+      "epoch": 2.8405797101449277,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00019647751619615353,
+      "loss": 0.4321,
       "step": 1764
     },
     {
-      "epoch": 3.3804971319311665,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00018880454075339854,
-      "loss": 0.5087,
+      "epoch": 2.847020933977456,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019606082119970214,
+      "loss": 0.4502,
       "step": 1768
     },
     {
-      "epoch": 3.388145315487572,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00018836477111013495,
-      "loss": 0.5179,
+      "epoch": 2.853462157809984,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019564373336985268,
+      "loss": 0.4298,
       "step": 1772
     },
     {
-      "epoch": 3.395793499043977,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00018792464833914576,
-      "loss": 0.5613,
+      "epoch": 2.8599033816425123,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019522625626377198,
+      "loss": 0.4469,
       "step": 1776
     },
     {
-      "epoch": 3.4034416826003824,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001874841764915317,
-      "loss": 0.5269,
+      "epoch": 2.86634460547504,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019480839344194695,
+      "loss": 0.4033,
       "step": 1780
     },
     {
-      "epoch": 3.4110898661567877,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00018704335962160663,
-      "loss": 0.5024,
+      "epoch": 2.8727858293075683,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019439014846815413,
+      "loss": 0.4381,
       "step": 1784
     },
     {
-      "epoch": 3.418738049713193,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00018660220178686002,
-      "loss": 0.5272,
+      "epoch": 2.8792270531400965,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019397152490942919,
+      "loss": 0.4205,
       "step": 1788
     },
     {
-      "epoch": 3.4263862332695982,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001861607070479199,
-      "loss": 0.4757,
+      "epoch": 2.8856682769726247,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019355252633603668,
+      "loss": 0.4187,
       "step": 1792
     },
     {
-      "epoch": 3.434034416826004,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00018571887946851535,
-      "loss": 0.5167,
+      "epoch": 2.892109500805153,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019313315632143944,
+      "loss": 0.3912,
       "step": 1796
     },
     {
-      "epoch": 3.4416826003824093,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018527672311543887,
-      "loss": 0.5162,
+      "epoch": 2.898550724637681,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019271341844226812,
+      "loss": 0.4236,
       "step": 1800
     },
     {
-      "epoch": 3.4493307839388145,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00018483424205850934,
-      "loss": 0.5711,
+      "epoch": 2.9049919484702094,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001922933162782909,
+      "loss": 0.3677,
       "step": 1804
     },
     {
-      "epoch": 3.45697896749522,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.0001843914403705343,
-      "loss": 0.5373,
+      "epoch": 2.9114331723027376,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019187285341238261,
+      "loss": 0.3979,
       "step": 1808
     },
     {
-      "epoch": 3.464627151051625,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00018394832212727252,
-      "loss": 0.5118,
+      "epoch": 2.917874396135266,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019145203343049453,
+      "loss": 0.3967,
       "step": 1812
     },
     {
-      "epoch": 3.472275334608031,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00018350489140739654,
-      "loss": 0.5283,
+      "epoch": 2.924315619967794,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019103085992162343,
+      "loss": 0.4128,
       "step": 1816
     },
     {
-      "epoch": 3.479923518164436,
+      "epoch": 2.930756843800322,
       "grad_norm": 0.298828125,
-      "learning_rate": 0.00018306115229245506,
-      "loss": 0.4733,
+      "learning_rate": 0.00019060933647778135,
+      "loss": 0.3968,
       "step": 1820
     },
     {
-      "epoch": 3.4875717017208414,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00018261710886683538,
-      "loss": 0.5091,
+      "epoch": 2.9371980676328504,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019018746669396464,
+      "loss": 0.4208,
       "step": 1824
     },
     {
-      "epoch": 3.4952198852772467,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001821727652177258,
-      "loss": 0.5055,
+      "epoch": 2.943639291465378,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018976525416812358,
+      "loss": 0.413,
       "step": 1828
     },
     {
-      "epoch": 3.502868068833652,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00018172812543507813,
-      "loss": 0.518,
+      "epoch": 2.9500805152979064,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018934270250113135,
+      "loss": 0.4122,
       "step": 1832
     },
     {
-      "epoch": 3.5105162523900573,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00018128319361156978,
-      "loss": 0.5309,
+      "epoch": 2.9565217391304346,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00018891981529675376,
+      "loss": 0.3961,
       "step": 1836
     },
     {
-      "epoch": 3.5181644359464626,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001808379738425664,
-      "loss": 0.5381,
+      "epoch": 2.962962962962963,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018849659616161808,
+      "loss": 0.4498,
       "step": 1840
     },
     {
-      "epoch": 3.525812619502868,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00018039247022608393,
-      "loss": 0.5596,
+      "epoch": 2.969404186795491,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018807304870518263,
+      "loss": 0.3935,
       "step": 1844
     },
     {
-      "epoch": 3.5334608030592736,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017994668686275092,
-      "loss": 0.5198,
+      "epoch": 2.975845410628019,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018764917653970567,
+      "loss": 0.4183,
       "step": 1848
     },
     {
-      "epoch": 3.541108986615679,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00017950062785577104,
-      "loss": 0.542,
+      "epoch": 2.9822866344605474,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000187224983280215,
+      "loss": 0.4101,
       "step": 1852
     },
     {
-      "epoch": 3.548757170172084,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00017905429731088497,
-      "loss": 0.5138,
+      "epoch": 2.9887278582930756,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018680047254447665,
+      "loss": 0.4587,
       "step": 1856
     },
     {
-      "epoch": 3.5564053537284894,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001786076993363328,
-      "loss": 0.5144,
+      "epoch": 2.995169082125604,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001863756479529644,
+      "loss": 0.4216,
       "step": 1860
     },
     {
-      "epoch": 3.5640535372848947,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001781608380428161,
-      "loss": 0.5127,
+      "epoch": 3.001610305958132,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018595051312882892,
+      "loss": 0.3842,
       "step": 1864
     },
     {
-      "epoch": 3.5717017208413004,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001777137175434602,
-      "loss": 0.5018,
+      "epoch": 3.0080515297906603,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018552507169786634,
+      "loss": 0.3189,
       "step": 1868
     },
     {
-      "epoch": 3.5793499043977057,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00017726634195377642,
-      "loss": 0.4387,
+      "epoch": 3.0144927536231885,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018509932728848804,
+      "loss": 0.3061,
       "step": 1872
     },
     {
-      "epoch": 3.586998087954111,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00017681871539162382,
-      "loss": 0.5421,
+      "epoch": 3.0209339774557167,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018467328353168934,
+      "loss": 0.3166,
       "step": 1876
     },
     {
-      "epoch": 3.5946462715105163,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00017637084197717163,
-      "loss": 0.5118,
+      "epoch": 3.027375201288245,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018424694406101838,
+      "loss": 0.3081,
       "step": 1880
     },
     {
-      "epoch": 3.6022944550669216,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017592272583286125,
-      "loss": 0.5017,
+      "epoch": 3.033816425120773,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001838203125125455,
+      "loss": 0.2944,
       "step": 1884
     },
     {
-      "epoch": 3.609942638623327,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00017547437108336836,
-      "loss": 0.533,
+      "epoch": 3.0402576489533013,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018339339252483196,
+      "loss": 0.285,
       "step": 1888
     },
     {
-      "epoch": 3.617590822179732,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00017502578185556468,
-      "loss": 0.5954,
+      "epoch": 3.0466988727858295,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018296618773889912,
+      "loss": 0.2926,
       "step": 1892
     },
     {
-      "epoch": 3.6252390057361374,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00017457696227848036,
-      "loss": 0.5098,
+      "epoch": 3.0531400966183573,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000182538701798197,
+      "loss": 0.3019,
       "step": 1896
     },
     {
-      "epoch": 3.632887189292543,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017412791648326566,
-      "loss": 0.4859,
+      "epoch": 3.0595813204508855,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018211093834857379,
+      "loss": 0.2984,
       "step": 1900
     },
     {
-      "epoch": 3.6405353728489485,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001736786486031531,
-      "loss": 0.5466,
+      "epoch": 3.0660225442834137,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018168290103824422,
+      "loss": 0.3185,
       "step": 1904
     },
     {
-      "epoch": 3.6481835564053537,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00017322916277341945,
-      "loss": 0.5258,
+      "epoch": 3.072463768115942,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018125459351775873,
+      "loss": 0.3192,
       "step": 1908
     },
     {
-      "epoch": 3.655831739961759,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00017277946313134758,
-      "loss": 0.5302,
+      "epoch": 3.07890499194847,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018082601943997232,
+      "loss": 0.3459,
       "step": 1912
     },
     {
-      "epoch": 3.6634799235181643,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00017232955381618826,
-      "loss": 0.5487,
+      "epoch": 3.0853462157809983,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018039718246001325,
+      "loss": 0.2837,
       "step": 1916
     },
     {
-      "epoch": 3.67112810707457,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00017187943896912236,
-      "loss": 0.497,
+      "epoch": 3.0917874396135265,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000179968086235252,
+      "loss": 0.3134,
       "step": 1920
     },
     {
-      "epoch": 3.6787762906309753,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001714291227332224,
-      "loss": 0.4907,
+      "epoch": 3.0982286634460547,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017953873442527008,
+      "loss": 0.2907,
       "step": 1924
     },
     {
-      "epoch": 3.6864244741873806,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017097860925341472,
-      "loss": 0.5322,
+      "epoch": 3.104669887278583,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017910913069182872,
+      "loss": 0.3076,
       "step": 1928
     },
     {
-      "epoch": 3.694072657743786,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00017052790267644112,
-      "loss": 0.4859,
+      "epoch": 3.111111111111111,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017867927869883775,
+      "loss": 0.293,
       "step": 1932
     },
     {
-      "epoch": 3.701720841300191,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00017007700715082077,
-      "loss": 0.5101,
+      "epoch": 3.1175523349436394,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017824918211232422,
+      "loss": 0.2775,
       "step": 1936
     },
     {
-      "epoch": 3.7093690248565965,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00016962592682681206,
-      "loss": 0.5091,
+      "epoch": 3.1239935587761676,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017781884460040136,
+      "loss": 0.3037,
       "step": 1940
     },
     {
-      "epoch": 3.7170172084130018,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00016917466585637426,
-      "loss": 0.5399,
+      "epoch": 3.130434782608696,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017738826983323703,
+      "loss": 0.3139,
       "step": 1944
     },
     {
-      "epoch": 3.724665391969407,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.0001687232283931294,
-      "loss": 0.4829,
+      "epoch": 3.136876006441224,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017695746148302252,
+      "loss": 0.3081,
       "step": 1948
     },
     {
-      "epoch": 3.7323135755258128,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00016827161859232418,
-      "loss": 0.5355,
+      "epoch": 3.143317230273752,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017652642322394142,
+      "loss": 0.344,
       "step": 1952
     },
     {
-      "epoch": 3.739961759082218,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00016781984061079138,
-      "loss": 0.5442,
+      "epoch": 3.14975845410628,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017609515873213787,
+      "loss": 0.3006,
       "step": 1956
     },
     {
-      "epoch": 3.7476099426386233,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00016736789860691197,
-      "loss": 0.5238,
+      "epoch": 3.156199677938808,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017566367168568572,
+      "loss": 0.2933,
       "step": 1960
     },
     {
-      "epoch": 3.7552581261950286,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016691579674057657,
-      "loss": 0.5004,
+      "epoch": 3.1626409017713364,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017523196576455663,
+      "loss": 0.2869,
       "step": 1964
     },
     {
-      "epoch": 3.762906309751434,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00016646353917314726,
-      "loss": 0.4718,
+      "epoch": 3.1690821256038646,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017480004465058918,
+      "loss": 0.2935,
       "step": 1968
     },
     {
-      "epoch": 3.7705544933078396,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016601113006741916,
-      "loss": 0.4324,
+      "epoch": 3.175523349436393,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017436791202745706,
+      "loss": 0.3451,
       "step": 1972
     },
     {
-      "epoch": 3.778202676864245,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00016555857358758252,
-      "loss": 0.457,
+      "epoch": 3.181964573268921,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017393557158063803,
+      "loss": 0.3047,
       "step": 1976
     },
     {
-      "epoch": 3.78585086042065,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00016510587389918373,
-      "loss": 0.4992,
+      "epoch": 3.1884057971014492,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017350302699738204,
+      "loss": 0.327,
       "step": 1980
     },
     {
-      "epoch": 3.7934990439770555,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00016465303516908762,
-      "loss": 0.4984,
+      "epoch": 3.1948470209339774,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017307028196668028,
+      "loss": 0.3238,
       "step": 1984
     },
     {
-      "epoch": 3.801147227533461,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001642000615654387,
-      "loss": 0.5278,
+      "epoch": 3.2012882447665056,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001726373401792333,
+      "loss": 0.2957,
       "step": 1988
     },
     {
-      "epoch": 3.808795411089866,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001637469572576229,
-      "loss": 0.5272,
+      "epoch": 3.207729468599034,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017220420532741977,
+      "loss": 0.3124,
       "step": 1992
     },
     {
-      "epoch": 3.8164435946462714,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00016329372641622934,
-      "loss": 0.5019,
+      "epoch": 3.214170692431562,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017177088110526486,
+      "loss": 0.2852,
       "step": 1996
     },
     {
-      "epoch": 3.8240917782026767,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00016284037321301166,
-      "loss": 0.4939,
+      "epoch": 3.2206119162640903,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017133737120840907,
+      "loss": 0.3084,
       "step": 2000
     },
     {
-      "epoch": 3.8317399617590824,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00016238690182084986,
-      "loss": 0.5335,
+      "epoch": 3.2270531400966185,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000170903679334076,
+      "loss": 0.2671,
       "step": 2004
     },
     {
-      "epoch": 3.8393881453154877,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016193331641371176,
-      "loss": 0.5396,
+      "epoch": 3.2334943639291467,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017046980918104164,
+      "loss": 0.2851,
       "step": 2008
     },
     {
-      "epoch": 3.847036328871893,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00016147962116661472,
-      "loss": 0.5078,
+      "epoch": 3.239935587761675,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001700357644496022,
+      "loss": 0.2921,
       "step": 2012
     },
     {
-      "epoch": 3.8546845124282982,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00016102582025558703,
-      "loss": 0.5286,
+      "epoch": 3.246376811594203,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016960154884154298,
+      "loss": 0.2898,
       "step": 2016
     },
     {
-      "epoch": 3.8623326959847035,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00016057191785762964,
-      "loss": 0.504,
+      "epoch": 3.2528180354267313,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016916716606010646,
+      "loss": 0.3277,
       "step": 2020
     },
     {
-      "epoch": 3.8699808795411093,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00016011791815067754,
-      "loss": 0.5131,
+      "epoch": 3.259259259259259,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016873261980996095,
+      "loss": 0.3301,
       "step": 2024
     },
     {
-      "epoch": 3.8776290630975145,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015966382531356144,
-      "loss": 0.5068,
+      "epoch": 3.2657004830917873,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016829791379716896,
+      "loss": 0.3639,
       "step": 2028
     },
     {
-      "epoch": 3.88527724665392,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00015920964352596927,
-      "loss": 0.5257,
+      "epoch": 3.2721417069243155,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016786305172915544,
+      "loss": 0.3492,
       "step": 2032
     },
     {
-      "epoch": 3.892925430210325,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015875537696840775,
-      "loss": 0.5145,
+      "epoch": 3.2785829307568437,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001674280373146763,
+      "loss": 0.3233,
       "step": 2036
     },
     {
-      "epoch": 3.9005736137667304,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001583010298221638,
-      "loss": 0.5219,
+      "epoch": 3.285024154589372,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016699287426378683,
+      "loss": 0.3232,
       "step": 2040
     },
     {
-      "epoch": 3.9082217973231357,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001578466062692661,
-      "loss": 0.5548,
+      "epoch": 3.2914653784219,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001665575662878099,
+      "loss": 0.326,
       "step": 2044
     },
     {
-      "epoch": 3.915869980879541,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015739211049244667,
-      "loss": 0.4981,
+      "epoch": 3.2979066022544283,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016612211709930442,
+      "loss": 0.3169,
       "step": 2048
     },
     {
-      "epoch": 3.9235181644359463,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00015693754667510235,
-      "loss": 0.5184,
+      "epoch": 3.3043478260869565,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016568653041203356,
+      "loss": 0.3036,
       "step": 2052
     },
     {
-      "epoch": 3.9311663479923515,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00015648291900125609,
-      "loss": 0.5207,
+      "epoch": 3.3107890499194848,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016525080994093328,
+      "loss": 0.2987,
       "step": 2056
     },
     {
-      "epoch": 3.9388145315487573,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00015602823165551877,
-      "loss": 0.5194,
+      "epoch": 3.317230273752013,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016481495940208046,
+      "loss": 0.3108,
       "step": 2060
     },
     {
-      "epoch": 3.9464627151051626,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001555734888230505,
-      "loss": 0.4692,
+      "epoch": 3.323671497584541,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001643789825126613,
+      "loss": 0.3119,
       "step": 2064
     },
     {
-      "epoch": 3.954110898661568,
+      "epoch": 3.3301127214170694,
       "grad_norm": 0.330078125,
-      "learning_rate": 0.00015511869468952201,
-      "loss": 0.5247,
+      "learning_rate": 0.0001639428829909396,
+      "loss": 0.342,
       "step": 2068
     },
     {
-      "epoch": 3.961759082217973,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001546638534410763,
-      "loss": 0.4561,
+      "epoch": 3.3365539452495976,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016350666455622497,
+      "loss": 0.3025,
       "step": 2072
     },
     {
-      "epoch": 3.969407265774379,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015420896926429014,
-      "loss": 0.5073,
+      "epoch": 3.342995169082126,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001630703309288412,
+      "loss": 0.3136,
       "step": 2076
     },
     {
-      "epoch": 3.977055449330784,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00015375404634613524,
-      "loss": 0.5323,
+      "epoch": 3.3494363929146536,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00016263388583009463,
+      "loss": 0.2957,
       "step": 2080
     },
     {
-      "epoch": 3.9847036328871894,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015329908887393992,
-      "loss": 0.5223,
+      "epoch": 3.3558776167471818,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001621973329822421,
+      "loss": 0.2948,
       "step": 2084
     },
     {
-      "epoch": 3.9923518164435947,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.0001528441010353508,
-      "loss": 0.5112,
+      "epoch": 3.36231884057971,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016176067610845958,
+      "loss": 0.3298,
       "step": 2088
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.87109375,
-      "learning_rate": 0.00015238908701829378,
-      "loss": 0.5374,
+      "epoch": 3.368760064412238,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016132391893281003,
+      "loss": 0.327,
       "step": 2092
     },
     {
-      "epoch": 4.007648183556405,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001519340510109357,
-      "loss": 0.446,
+      "epoch": 3.3752012882447664,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001608870651802121,
+      "loss": 0.3009,
       "step": 2096
     },
     {
-      "epoch": 4.015296367112811,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015147899720164594,
-      "loss": 0.467,
+      "epoch": 3.3816425120772946,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016045011857640783,
+      "loss": 0.3148,
       "step": 2100
     },
     {
-      "epoch": 4.022944550669216,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015102392977895765,
-      "loss": 0.4112,
+      "epoch": 3.388083735909823,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001600130828479314,
+      "loss": 0.3282,
       "step": 2104
     },
     {
-      "epoch": 4.030592734225621,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00015056885293152932,
-      "loss": 0.4401,
+      "epoch": 3.394524959742351,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001595759617220769,
+      "loss": 0.3203,
       "step": 2108
     },
     {
-      "epoch": 4.038240917782026,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015011377084810624,
-      "loss": 0.406,
+      "epoch": 3.4009661835748792,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015913875892686685,
+      "loss": 0.2977,
       "step": 2112
     },
     {
-      "epoch": 4.045889101338432,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00014965868771748178,
-      "loss": 0.4528,
+      "epoch": 3.4074074074074074,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00015870147819102025,
+      "loss": 0.2806,
       "step": 2116
     },
     {
-      "epoch": 4.053537284894838,
+      "epoch": 3.4138486312399356,
       "grad_norm": 0.337890625,
-      "learning_rate": 0.00014920360772845896,
-      "loss": 0.4345,
+      "learning_rate": 0.00015826412324392085,
+      "loss": 0.3096,
       "step": 2120
     },
     {
-      "epoch": 4.061185468451243,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00014874853506981206,
-      "loss": 0.4349,
+      "epoch": 3.420289855072464,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015782669781558528,
+      "loss": 0.301,
       "step": 2124
     },
     {
-      "epoch": 4.0688336520076485,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00014829347393024764,
-      "loss": 0.4494,
+      "epoch": 3.426731078904992,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00015738920563663136,
+      "loss": 0.3055,
       "step": 2128
     },
     {
-      "epoch": 4.076481835564054,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00014783842849836644,
-      "loss": 0.4159,
+      "epoch": 3.4331723027375203,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00015695165043824605,
+      "loss": 0.3187,
       "step": 2132
     },
     {
-      "epoch": 4.084130019120459,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00014738340296262443,
-      "loss": 0.372,
+      "epoch": 3.4396135265700485,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015651403595215392,
+      "loss": 0.308,
       "step": 2136
     },
     {
-      "epoch": 4.091778202676864,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00014692840151129467,
-      "loss": 0.4621,
+      "epoch": 3.4460547504025767,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00015607636591058506,
+      "loss": 0.3033,
       "step": 2140
     },
     {
-      "epoch": 4.09942638623327,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00014647342833242827,
-      "loss": 0.4095,
+      "epoch": 3.452495974235105,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001556386440462435,
+      "loss": 0.3313,
       "step": 2144
     },
     {
-      "epoch": 4.107074569789675,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00014601848761381633,
-      "loss": 0.4172,
+      "epoch": 3.4589371980676327,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001552008740922751,
+      "loss": 0.2891,
       "step": 2148
     },
     {
-      "epoch": 4.11472275334608,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00014556358354295113,
-      "loss": 0.4197,
+      "epoch": 3.465378421900161,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00015476305978223606,
+      "loss": 0.3416,
       "step": 2152
     },
     {
-      "epoch": 4.1223709369024855,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001451087203069875,
-      "loss": 0.4573,
+      "epoch": 3.471819645732689,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015432520485006055,
+      "loss": 0.2768,
       "step": 2156
     },
     {
-      "epoch": 4.130019120458891,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00014465390209270456,
-      "loss": 0.4117,
+      "epoch": 3.4782608695652173,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00015388731303002954,
+      "loss": 0.3216,
       "step": 2160
     },
     {
-      "epoch": 4.137667304015296,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00014419913308646686,
-      "loss": 0.4148,
+      "epoch": 3.4847020933977455,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001534493880567384,
+      "loss": 0.3112,
       "step": 2164
     },
     {
-      "epoch": 4.145315487571701,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00014374441747418628,
-      "loss": 0.4251,
+      "epoch": 3.4911433172302737,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015301143366506527,
+      "loss": 0.323,
       "step": 2168
     },
     {
-      "epoch": 4.1529636711281075,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00014328975944128292,
-      "loss": 0.4314,
+      "epoch": 3.497584541062802,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015257345359013928,
+      "loss": 0.3406,
       "step": 2172
     },
     {
-      "epoch": 4.160611854684513,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00014283516317264704,
-      "loss": 0.4386,
+      "epoch": 3.50402576489533,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015213545156730847,
+      "loss": 0.2904,
       "step": 2176
     },
     {
-      "epoch": 4.168260038240918,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00014238063285260057,
-      "loss": 0.392,
+      "epoch": 3.5104669887278583,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015169743133210814,
+      "loss": 0.3107,
       "step": 2180
     },
     {
-      "epoch": 4.175908221797323,
+      "epoch": 3.5169082125603865,
       "grad_norm": 0.357421875,
-      "learning_rate": 0.00014192617266485803,
-      "loss": 0.4187,
+      "learning_rate": 0.0001512593966202289,
+      "loss": 0.3377,
       "step": 2184
     },
     {
-      "epoch": 4.183556405353729,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001414717867924888,
-      "loss": 0.4165,
+      "epoch": 3.5233494363929148,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015082135116748483,
+      "loss": 0.3491,
       "step": 2188
     },
     {
-      "epoch": 4.191204588910134,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.0001410174794178779,
-      "loss": 0.4191,
+      "epoch": 3.529790660225443,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015038329870978168,
+      "loss": 0.2865,
       "step": 2192
     },
     {
-      "epoch": 4.198852772466539,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00014056325472268805,
-      "loss": 0.436,
+      "epoch": 3.536231884057971,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014994524298308479,
+      "loss": 0.2913,
       "step": 2196
     },
     {
-      "epoch": 4.2065009560229445,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.0001401091168878209,
-      "loss": 0.4075,
+      "epoch": 3.542673107890499,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001495071877233875,
+      "loss": 0.3163,
       "step": 2200
     },
     {
-      "epoch": 4.21414913957935,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00013965507009337845,
-      "loss": 0.3995,
+      "epoch": 3.549114331723027,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014906913666667913,
+      "loss": 0.2722,
       "step": 2204
     },
     {
-      "epoch": 4.221797323135755,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00013920111851862494,
-      "loss": 0.4474,
+      "epoch": 3.5555555555555554,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014863109354891317,
+      "loss": 0.3163,
       "step": 2208
     },
     {
-      "epoch": 4.22944550669216,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00013874726634194797,
-      "loss": 0.405,
+      "epoch": 3.5619967793880836,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014819306210597536,
+      "loss": 0.3735,
       "step": 2212
     },
     {
-      "epoch": 4.237093690248566,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001382935177408204,
-      "loss": 0.4316,
+      "epoch": 3.5684380032206118,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00014775504607365196,
+      "loss": 0.3303,
       "step": 2216
     },
     {
-      "epoch": 4.244741873804971,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00013783987689176157,
-      "loss": 0.4299,
+      "epoch": 3.57487922705314,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014731704918759765,
+      "loss": 0.2946,
       "step": 2220
     },
     {
-      "epoch": 4.252390057361376,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00013738634797029914,
-      "loss": 0.4347,
+      "epoch": 3.581320450885668,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000146879075183304,
+      "loss": 0.3434,
       "step": 2224
     },
     {
-      "epoch": 4.260038240917782,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00013693293515093052,
-      "loss": 0.4393,
+      "epoch": 3.5877616747181964,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00014644112779606727,
+      "loss": 0.3063,
       "step": 2228
     },
     {
-      "epoch": 4.267686424474188,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00013647964260708436,
-      "loss": 0.44,
+      "epoch": 3.5942028985507246,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014600321076095683,
+      "loss": 0.2962,
       "step": 2232
     },
     {
-      "epoch": 4.275334608030593,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.0001360264745110824,
-      "loss": 0.4194,
+      "epoch": 3.600644122383253,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00014556532781278316,
+      "loss": 0.3006,
       "step": 2236
     },
     {
-      "epoch": 4.282982791586998,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.0001355734350341007,
-      "loss": 0.4259,
+      "epoch": 3.607085346215781,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014512748268606592,
+      "loss": 0.3688,
       "step": 2240
     },
     {
-      "epoch": 4.2906309751434035,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00013512052834613165,
-      "loss": 0.4311,
+      "epoch": 3.6135265700483092,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014468967911500242,
+      "loss": 0.3348,
       "step": 2244
     },
     {
-      "epoch": 4.298279158699809,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00013466775861594523,
-      "loss": 0.4097,
+      "epoch": 3.6199677938808374,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001442519208334353,
+      "loss": 0.3128,
       "step": 2248
     },
     {
-      "epoch": 4.305927342256214,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001342151300110509,
-      "loss": 0.4149,
+      "epoch": 3.6264090177133657,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014381421157482125,
+      "loss": 0.3488,
       "step": 2252
     },
     {
-      "epoch": 4.313575525812619,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001337626466976591,
-      "loss": 0.3983,
+      "epoch": 3.632850241545894,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001433765550721985,
+      "loss": 0.2614,
       "step": 2256
     },
     {
-      "epoch": 4.321223709369025,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.0001333103128406429,
-      "loss": 0.4424,
+      "epoch": 3.639291465378422,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00014293895505815575,
+      "loss": 0.2984,
       "step": 2260
     },
     {
-      "epoch": 4.32887189292543,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00013285813260349982,
-      "loss": 0.4352,
+      "epoch": 3.6457326892109503,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014250141526479953,
+      "loss": 0.3257,
       "step": 2264
     },
     {
-      "epoch": 4.336520076481835,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.0001324061101483132,
-      "loss": 0.4441,
+      "epoch": 3.6521739130434785,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014206393942372314,
+      "loss": 0.3235,
       "step": 2268
     },
     {
-      "epoch": 4.3441682600382405,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00013195424963571424,
-      "loss": 0.421,
+      "epoch": 3.6586151368760067,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001416265312659741,
+      "loss": 0.3435,
       "step": 2272
     },
     {
-      "epoch": 4.351816443594647,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00013150255522484345,
-      "loss": 0.4131,
+      "epoch": 3.6650563607085345,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014118919452202306,
+      "loss": 0.3191,
       "step": 2276
     },
     {
-      "epoch": 4.359464627151052,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00013105103107331255,
-      "loss": 0.3568,
+      "epoch": 3.6714975845410627,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014075193292173126,
+      "loss": 0.2869,
       "step": 2280
     },
     {
-      "epoch": 4.367112810707457,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00013059968133716606,
-      "loss": 0.445,
+      "epoch": 3.677938808373591,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014031475019431934,
+      "loss": 0.3089,
       "step": 2284
     },
     {
-      "epoch": 4.374760994263863,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00013014851017084303,
-      "loss": 0.4267,
+      "epoch": 3.684380032206119,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013987765006833518,
+      "loss": 0.3332,
       "step": 2288
     },
     {
-      "epoch": 4.382409177820268,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00012969752172713905,
-      "loss": 0.4458,
+      "epoch": 3.6908212560386473,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001394406362716221,
+      "loss": 0.3127,
       "step": 2292
     },
     {
-      "epoch": 4.390057361376673,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00012924672015716759,
-      "loss": 0.439,
+      "epoch": 3.6972624798711755,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013900371253128727,
+      "loss": 0.3177,
       "step": 2296
     },
     {
-      "epoch": 4.397705544933078,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012879610961032218,
-      "loss": 0.4792,
+      "epoch": 3.7037037037037037,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001385668825736697,
+      "loss": 0.3324,
       "step": 2300
     },
     {
-      "epoch": 4.405353728489484,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001283456942342383,
-      "loss": 0.4113,
+      "epoch": 3.710144927536232,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001381301501243087,
+      "loss": 0.2785,
       "step": 2304
     },
     {
-      "epoch": 4.413001912045889,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.0001278954781747545,
-      "loss": 0.4548,
+      "epoch": 3.71658615136876,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00013769351890791185,
+      "loss": 0.3274,
       "step": 2308
     },
     {
-      "epoch": 4.420650095602294,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00012744546557587517,
-      "loss": 0.4512,
+      "epoch": 3.7230273752012883,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013725699264832344,
+      "loss": 0.3041,
       "step": 2312
     },
     {
-      "epoch": 4.4282982791587,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00012699566057973168,
-      "loss": 0.4211,
+      "epoch": 3.7294685990338166,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013682057506849256,
+      "loss": 0.3343,
       "step": 2316
     },
     {
-      "epoch": 4.435946462715105,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00012654606732654468,
-      "loss": 0.4256,
+      "epoch": 3.7359098228663448,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013638426989044148,
+      "loss": 0.2785,
       "step": 2320
     },
     {
-      "epoch": 4.44359464627151,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00012609668995458573,
-      "loss": 0.4451,
+      "epoch": 3.7423510466988725,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013594808083523376,
+      "loss": 0.3454,
       "step": 2324
     },
     {
-      "epoch": 4.451242829827915,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001256475326001394,
-      "loss": 0.4703,
+      "epoch": 3.7487922705314007,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013551201162294275,
+      "loss": 0.312,
       "step": 2328
     },
     {
-      "epoch": 4.458891013384322,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00012519859939746504,
-      "loss": 0.4032,
+      "epoch": 3.755233494363929,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013507606597261946,
+      "loss": 0.2885,
       "step": 2332
     },
     {
-      "epoch": 4.466539196940727,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.00012474989447875886,
-      "loss": 0.4324,
+      "epoch": 3.761674718196457,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013464024760226142,
+      "loss": 0.3328,
       "step": 2336
     },
     {
-      "epoch": 4.474187380497132,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.0001243014219741158,
-      "loss": 0.4671,
+      "epoch": 3.7681159420289854,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001342045602287803,
+      "loss": 0.3078,
       "step": 2340
     },
     {
-      "epoch": 4.4818355640535374,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00012385318601149158,
-      "loss": 0.4463,
+      "epoch": 3.7745571658615136,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013376900756797085,
+      "loss": 0.3126,
       "step": 2344
     },
     {
-      "epoch": 4.489483747609943,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00012340519071666467,
-      "loss": 0.4448,
+      "epoch": 3.780998389694042,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013333359333447865,
+      "loss": 0.2941,
       "step": 2348
     },
     {
-      "epoch": 4.497131931166348,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001229574402131982,
-      "loss": 0.4345,
+      "epoch": 3.78743961352657,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001328983212417689,
+      "loss": 0.3251,
       "step": 2352
     },
     {
-      "epoch": 4.504780114722753,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00012250993862240227,
-      "loss": 0.433,
+      "epoch": 3.793880837359098,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001324631950020945,
+      "loss": 0.3367,
       "step": 2356
     },
     {
-      "epoch": 4.512428298279159,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012206269006329593,
-      "loss": 0.4293,
+      "epoch": 3.8003220611916264,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001320282183264643,
+      "loss": 0.3164,
       "step": 2360
     },
     {
-      "epoch": 4.520076481835564,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00012161569865256896,
-      "loss": 0.4413,
+      "epoch": 3.8067632850241546,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00013159339492461176,
+      "loss": 0.3584,
       "step": 2364
     },
     {
-      "epoch": 4.527724665391969,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00012116896850454446,
-      "loss": 0.4446,
+      "epoch": 3.813204508856683,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013115872850496293,
+      "loss": 0.3307,
       "step": 2368
     },
     {
-      "epoch": 4.5353728489483744,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00012072250373114057,
-      "loss": 0.4642,
+      "epoch": 3.819645732689211,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001307242227746053,
+      "loss": 0.3475,
       "step": 2372
     },
     {
-      "epoch": 4.54302103250478,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012027630844183288,
-      "loss": 0.4825,
+      "epoch": 3.8260869565217392,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00013028988143925553,
+      "loss": 0.3058,
       "step": 2376
     },
     {
-      "epoch": 4.550669216061186,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00011983038674361658,
-      "loss": 0.4303,
+      "epoch": 3.8325281803542675,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012985570820322868,
+      "loss": 0.2718,
       "step": 2380
     },
     {
-      "epoch": 4.558317399617591,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00011938474274096844,
-      "loss": 0.4013,
+      "epoch": 3.8389694041867957,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012942170676940576,
+      "loss": 0.3074,
       "step": 2384
     },
     {
-      "epoch": 4.5659655831739965,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00011893938053580933,
-      "loss": 0.4183,
+      "epoch": 3.845410628019324,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012898788083920282,
+      "loss": 0.3177,
       "step": 2388
     },
     {
-      "epoch": 4.573613766730402,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00011849430422746624,
-      "loss": 0.4345,
+      "epoch": 3.851851851851852,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001285542341125389,
+      "loss": 0.3012,
       "step": 2392
     },
     {
-      "epoch": 4.581261950286807,
-      "grad_norm": 0.369140625,
-      "learning_rate": 0.00011804951791263466,
-      "loss": 0.4253,
+      "epoch": 3.8582930756843803,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001281207702878049,
+      "loss": 0.3024,
       "step": 2396
     },
     {
-      "epoch": 4.588910133843212,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00011760502568534081,
-      "loss": 0.473,
+      "epoch": 3.864734299516908,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012768749306183165,
+      "loss": 0.3092,
       "step": 2400
     },
     {
-      "epoch": 4.596558317399618,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011716083163690405,
-      "loss": 0.451,
+      "epoch": 3.8711755233494363,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012725440612985868,
+      "loss": 0.2978,
       "step": 2404
     },
     {
-      "epoch": 4.604206500956023,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00011671693985589913,
-      "loss": 0.4522,
+      "epoch": 3.8776167471819645,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001268215131855025,
+      "loss": 0.3337,
       "step": 2408
     },
     {
-      "epoch": 4.611854684512428,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00011627335442811846,
-      "loss": 0.4193,
+      "epoch": 3.8840579710144927,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012638881792072522,
+      "loss": 0.3278,
       "step": 2412
     },
     {
-      "epoch": 4.6195028680688335,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00011583007943653494,
-      "loss": 0.4616,
+      "epoch": 3.890499194847021,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012595632402580305,
+      "loss": 0.3051,
       "step": 2416
     },
     {
-      "epoch": 4.627151051625239,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00011538711896126369,
-      "loss": 0.4549,
+      "epoch": 3.896940418679549,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012552403518929472,
+      "loss": 0.2764,
       "step": 2420
     },
     {
-      "epoch": 4.634799235181644,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00011494447707952514,
-      "loss": 0.4119,
+      "epoch": 3.9033816425120773,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001250919550980102,
+      "loss": 0.3124,
       "step": 2424
     },
     {
-      "epoch": 4.642447418738049,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.0001145021578656071,
-      "loss": 0.4203,
+      "epoch": 3.9098228663446055,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012466008743697906,
+      "loss": 0.3407,
       "step": 2428
     },
     {
-      "epoch": 4.650095602294455,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00011406016539082747,
-      "loss": 0.4199,
+      "epoch": 3.9162640901771337,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012422843588941925,
+      "loss": 0.3336,
       "step": 2432
     },
     {
-      "epoch": 4.657743785850861,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.00011361850372349667,
-      "loss": 0.4791,
+      "epoch": 3.922705314009662,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012379700413670547,
+      "loss": 0.2992,
       "step": 2436
     },
     {
-      "epoch": 4.665391969407266,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00011317717692888012,
-      "loss": 0.3904,
+      "epoch": 3.92914653784219,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012336579585833798,
+      "loss": 0.3341,
       "step": 2440
     },
     {
-      "epoch": 4.673040152963671,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00011273618906916107,
-      "loss": 0.413,
+      "epoch": 3.9355877616747184,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012293481473191103,
+      "loss": 0.3153,
       "step": 2444
     },
     {
-      "epoch": 4.680688336520077,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00011229554420340289,
-      "loss": 0.4078,
+      "epoch": 3.942028985507246,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012250406443308168,
+      "loss": 0.2993,
       "step": 2448
     },
     {
-      "epoch": 4.688336520076482,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00011185524638751195,
-      "loss": 0.481,
+      "epoch": 3.9484702093397743,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012207354863553825,
+      "loss": 0.3144,
       "step": 2452
     },
     {
-      "epoch": 4.695984703632887,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.0001114152996742003,
-      "loss": 0.4649,
+      "epoch": 3.9549114331723025,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012164327101096923,
+      "loss": 0.3251,
       "step": 2456
     },
     {
-      "epoch": 4.7036328871892925,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00011097570811294803,
-      "loss": 0.4758,
+      "epoch": 3.9613526570048307,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012121323522903167,
+      "loss": 0.2799,
       "step": 2460
     },
     {
-      "epoch": 4.711281070745698,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00011053647574996648,
-      "loss": 0.3909,
+      "epoch": 3.967793880837359,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012078344495732028,
+      "loss": 0.3188,
       "step": 2464
     },
     {
-      "epoch": 4.718929254302103,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.0001100976066281606,
-      "loss": 0.3929,
+      "epoch": 3.974235104669887,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012035390386133558,
+      "loss": 0.3052,
       "step": 2468
     },
     {
-      "epoch": 4.726577437858508,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010965910478709206,
-      "loss": 0.4572,
+      "epoch": 3.9806763285024154,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011992461560445337,
+      "loss": 0.2771,
       "step": 2472
     },
     {
-      "epoch": 4.734225621414914,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00010922097426294166,
-      "loss": 0.422,
+      "epoch": 3.9871175523349436,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011949558384789271,
+      "loss": 0.3164,
       "step": 2476
     },
     {
-      "epoch": 4.741873804971319,
-      "grad_norm": 0.3671875,
-      "learning_rate": 0.00010878321908847259,
-      "loss": 0.4397,
+      "epoch": 3.993558776167472,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011906681225068535,
+      "loss": 0.2902,
       "step": 2480
     },
     {
-      "epoch": 4.749521988527725,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00010834584329299322,
-      "loss": 0.4073,
+      "epoch": 4.0,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00011863830446964417,
+      "loss": 0.3142,
       "step": 2484
     },
     {
-      "epoch": 4.75717017208413,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00010790885090231968,
-      "loss": 0.4209,
+      "epoch": 4.006441223832528,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011821006415933199,
+      "loss": 0.2147,
       "step": 2488
     },
     {
-      "epoch": 4.764818355640536,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00010747224593873933,
-      "loss": 0.4365,
+      "epoch": 4.012882447665056,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011778209497203062,
+      "loss": 0.2092,
       "step": 2492
     },
     {
-      "epoch": 4.772466539196941,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010703603242097322,
-      "loss": 0.4213,
+      "epoch": 4.019323671497585,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011735440055770945,
+      "loss": 0.2548,
       "step": 2496
     },
     {
-      "epoch": 4.780114722753346,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00010660021436413956,
-      "loss": 0.4705,
+      "epoch": 4.025764895330113,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00011692698456399458,
+      "loss": 0.2183,
       "step": 2500
     },
     {
-      "epoch": 4.7877629063097515,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00010616479577971638,
-      "loss": 0.4171,
+      "epoch": 4.032206119162641,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001164998506361374,
+      "loss": 0.2009,
       "step": 2504
     },
     {
-      "epoch": 4.795411089866157,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00010572978067550489,
-      "loss": 0.4357,
+      "epoch": 4.038647342995169,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00011607300241698387,
+      "loss": 0.218,
       "step": 2508
     },
     {
-      "epoch": 4.803059273422562,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00010529517305559244,
-      "loss": 0.4225,
+      "epoch": 4.0450885668276975,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011564644354694312,
+      "loss": 0.2201,
       "step": 2512
     },
     {
-      "epoch": 4.810707456978967,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010486097692031566,
-      "loss": 0.4569,
+      "epoch": 4.051529790660226,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011522017766395665,
+      "loss": 0.2078,
       "step": 2516
     },
     {
-      "epoch": 4.818355640535373,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00010442719626622374,
-      "loss": 0.4801,
+      "epoch": 4.057971014492754,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00011479420840346706,
+      "loss": 0.1932,
       "step": 2520
     },
     {
-      "epoch": 4.826003824091778,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.0001039938350860415,
-      "loss": 0.4476,
+      "epoch": 4.064412238325282,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00011436853939838734,
+      "loss": 0.2217,
       "step": 2524
     },
     {
-      "epoch": 4.833652007648183,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010356089736863282,
-      "loss": 0.4016,
+      "epoch": 4.07085346215781,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001139431742790696,
+      "loss": 0.2448,
       "step": 2528
     },
     {
-      "epoch": 4.8413001912045885,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.0001031283870989638,
-      "loss": 0.4467,
+      "epoch": 4.0772946859903385,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001135181166732743,
+      "loss": 0.2254,
       "step": 2532
     },
     {
-      "epoch": 4.848948374760994,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00010269630825806597,
-      "loss": 0.4236,
+      "epoch": 4.083735909822867,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011309337020613922,
+      "loss": 0.2665,
       "step": 2536
     },
     {
-      "epoch": 4.8565965583174,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00010226466482300006,
-      "loss": 0.426,
+      "epoch": 4.090177133655395,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001126689385001486,
+      "loss": 0.2365,
       "step": 2540
     },
     {
-      "epoch": 4.864244741873805,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00010183346076681882,
-      "loss": 0.4452,
+      "epoch": 4.096618357487923,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011224482517510224,
+      "loss": 0.2341,
       "step": 2544
     },
     {
-      "epoch": 4.871892925430211,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00010140270005853098,
-      "loss": 0.4182,
+      "epoch": 4.1030595813204505,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00011182103384808444,
+      "loss": 0.2015,
       "step": 2548
     },
     {
-      "epoch": 4.879541108986616,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00010097238666306427,
-      "loss": 0.4035,
+      "epoch": 4.109500805152979,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011139756813343359,
+      "loss": 0.2334,
       "step": 2552
     },
     {
-      "epoch": 4.887189292543021,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00010054252454122934,
-      "loss": 0.4187,
+      "epoch": 4.115942028985507,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00011097443164271075,
+      "loss": 0.246,
       "step": 2556
     },
     {
-      "epoch": 4.894837476099426,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.000100113117649683,
-      "loss": 0.4494,
+      "epoch": 4.122383252818035,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011055162798466948,
+      "loss": 0.2322,
       "step": 2560
     },
     {
-      "epoch": 4.902485659655832,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.968416994089189e-05,
-      "loss": 0.4461,
+      "epoch": 4.128824476650563,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011012916076522443,
+      "loss": 0.2178,
       "step": 2564
     },
     {
-      "epoch": 4.910133843212237,
-      "grad_norm": 0.359375,
-      "learning_rate": 9.925568536309619e-05,
-      "loss": 0.4589,
+      "epoch": 4.1352657004830915,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00010970703358742127,
+      "loss": 0.2147,
       "step": 2568
     },
     {
-      "epoch": 4.917782026768642,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.88276678602731e-05,
-      "loss": 0.4329,
+      "epoch": 4.14170692431562,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010928525005140521,
+      "loss": 0.2315,
       "step": 2572
     },
     {
-      "epoch": 4.925430210325048,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.840012137210072e-05,
-      "loss": 0.4048,
+      "epoch": 4.148148148148148,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010886381375439105,
+      "loss": 0.2284,
       "step": 2576
     },
     {
-      "epoch": 4.933078393881453,
-      "grad_norm": 0.40234375,
-      "learning_rate": 9.797304983392164e-05,
-      "loss": 0.4689,
+      "epoch": 4.154589371980676,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001084427282906318,
+      "loss": 0.2568,
       "step": 2580
     },
     {
-      "epoch": 4.940726577437858,
-      "grad_norm": 0.37890625,
-      "learning_rate": 9.75464571767068e-05,
-      "loss": 0.4538,
+      "epoch": 4.161030595813204,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010802199725138869,
+      "loss": 0.2163,
       "step": 2584
     },
     {
-      "epoch": 4.948374760994264,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.712034732701942e-05,
-      "loss": 0.4251,
+      "epoch": 4.1674718196457325,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010760162422489987,
+      "loss": 0.2267,
       "step": 2588
     },
     {
-      "epoch": 4.95602294455067,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.669472420697845e-05,
-      "loss": 0.3808,
+      "epoch": 4.173913043478261,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010718161279635048,
+      "loss": 0.2263,
       "step": 2592
     },
     {
-      "epoch": 4.963671128107075,
+      "epoch": 4.180354267310789,
       "grad_norm": 0.33984375,
-      "learning_rate": 9.626959173422306e-05,
-      "loss": 0.4249,
+      "learning_rate": 0.00010676196654784144,
+      "loss": 0.2395,
       "step": 2596
     },
     {
-      "epoch": 4.97131931166348,
-      "grad_norm": 0.36328125,
-      "learning_rate": 9.5844953821876e-05,
-      "loss": 0.4491,
+      "epoch": 4.186795491143317,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010634268905835949,
+      "loss": 0.2454,
       "step": 2600
     },
     {
-      "epoch": 4.9789674952198855,
-      "grad_norm": 0.33203125,
-      "learning_rate": 9.542081437850801e-05,
-      "loss": 0.3934,
+      "epoch": 4.193236714975845,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00010592378390374612,
+      "loss": 0.2186,
       "step": 2604
     },
     {
-      "epoch": 4.986615678776291,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.49971773081017e-05,
-      "loss": 0.4589,
+      "epoch": 4.199677938808374,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010550525465666751,
+      "loss": 0.2302,
       "step": 2608
     },
     {
-      "epoch": 4.994263862332696,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.457404651001546e-05,
-      "loss": 0.4418,
+      "epoch": 4.206119162640902,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010508710488658385,
+      "loss": 0.2475,
       "step": 2612
     },
     {
-      "epoch": 5.001912045889101,
-      "grad_norm": 0.31640625,
-      "learning_rate": 9.415142587894786e-05,
-      "loss": 0.3326,
+      "epoch": 4.21256038647343,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010466933815971884,
+      "loss": 0.1988,
       "step": 2616
     },
     {
-      "epoch": 5.009560229445507,
-      "grad_norm": 0.314453125,
-      "learning_rate": 9.372931930490147e-05,
-      "loss": 0.3716,
+      "epoch": 4.219001610305958,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010425195803902948,
+      "loss": 0.2137,
       "step": 2620
     },
     {
-      "epoch": 5.017208413001912,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.330773067314747e-05,
-      "loss": 0.4067,
+      "epoch": 4.225442834138486,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010383496808417547,
+      "loss": 0.2564,
       "step": 2624
     },
     {
-      "epoch": 5.024856596558317,
-      "grad_norm": 0.345703125,
-      "learning_rate": 9.28866638641894e-05,
-      "loss": 0.3533,
+      "epoch": 4.231884057971015,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010341837185148903,
+      "loss": 0.2361,
       "step": 2628
     },
     {
-      "epoch": 5.0325047801147225,
-      "grad_norm": 0.3359375,
-      "learning_rate": 9.246612275372786e-05,
-      "loss": 0.3446,
+      "epoch": 4.238325281803543,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010300217289394443,
+      "loss": 0.2324,
       "step": 2632
     },
     {
-      "epoch": 5.040152963671128,
-      "grad_norm": 0.353515625,
-      "learning_rate": 9.204611121262466e-05,
-      "loss": 0.391,
+      "epoch": 4.244766505636071,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00010258637476112782,
+      "loss": 0.2175,
       "step": 2636
     },
     {
-      "epoch": 5.047801147227533,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.16266331068671e-05,
-      "loss": 0.3821,
+      "epoch": 4.251207729468599,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010217098099920676,
+      "loss": 0.2533,
       "step": 2640
     },
     {
-      "epoch": 5.055449330783939,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.120769229753262e-05,
-      "loss": 0.3813,
+      "epoch": 4.2576489533011275,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010175599515090026,
+      "loss": 0.2155,
       "step": 2644
     },
     {
-      "epoch": 5.0630975143403445,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.078929264075293e-05,
-      "loss": 0.3493,
+      "epoch": 4.264090177133656,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010134142075544824,
+      "loss": 0.2299,
       "step": 2648
     },
     {
-      "epoch": 5.07074569789675,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.0371437987679e-05,
-      "loss": 0.3346,
+      "epoch": 4.270531400966184,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010092726134858168,
+      "loss": 0.2776,
       "step": 2652
     },
     {
-      "epoch": 5.078393881453155,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.995413218444502e-05,
-      "loss": 0.3753,
+      "epoch": 4.276972624798712,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010051352046249213,
+      "loss": 0.2079,
       "step": 2656
     },
     {
-      "epoch": 5.08604206500956,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.953737907213346e-05,
-      "loss": 0.3749,
+      "epoch": 4.28341384863124,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010010020162580192,
+      "loss": 0.198,
       "step": 2660
     },
     {
-      "epoch": 5.093690248565966,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.912118248673966e-05,
-      "loss": 0.3673,
+      "epoch": 4.2898550724637685,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.96873083635337e-05,
+      "loss": 0.223,
       "step": 2664
     },
     {
-      "epoch": 5.101338432122371,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.870554625913619e-05,
-      "loss": 0.3591,
+      "epoch": 4.296296296296296,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.927484419708076e-05,
+      "loss": 0.187,
       "step": 2668
     },
     {
-      "epoch": 5.108986615678776,
-      "grad_norm": 0.369140625,
-      "learning_rate": 8.8290474215038e-05,
-      "loss": 0.387,
+      "epoch": 4.302737520128824,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.88628126441768e-05,
+      "loss": 0.2339,
       "step": 2672
     },
     {
-      "epoch": 5.1166347992351815,
-      "grad_norm": 0.359375,
-      "learning_rate": 8.787597017496687e-05,
-      "loss": 0.3774,
+      "epoch": 4.309178743961352,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.84512172188657e-05,
+      "loss": 0.2164,
       "step": 2676
     },
     {
-      "epoch": 5.124282982791587,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.74620379542166e-05,
-      "loss": 0.3711,
+      "epoch": 4.3156199677938805,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.804006143147212e-05,
+      "loss": 0.2328,
       "step": 2680
     },
     {
-      "epoch": 5.131931166347992,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.704868136281742e-05,
-      "loss": 0.3733,
+      "epoch": 4.322061191626409,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.762934878857105e-05,
+      "loss": 0.2577,
       "step": 2684
     },
     {
-      "epoch": 5.139579349904397,
-      "grad_norm": 0.36328125,
-      "learning_rate": 8.663590420550145e-05,
-      "loss": 0.3483,
+      "epoch": 4.328502415458937,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.721908279295812e-05,
+      "loss": 0.2256,
       "step": 2688
     },
     {
-      "epoch": 5.147227533460803,
-      "grad_norm": 0.369140625,
-      "learning_rate": 8.622371028166743e-05,
-      "loss": 0.3773,
+      "epoch": 4.334943639291465,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.680926694361964e-05,
+      "loss": 0.2344,
       "step": 2692
     },
     {
-      "epoch": 5.154875717017209,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.581210338534538e-05,
-      "loss": 0.3921,
+      "epoch": 4.341384863123993,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.639990473570294e-05,
+      "loss": 0.2238,
       "step": 2696
     },
     {
-      "epoch": 5.162523900573614,
-      "grad_norm": 0.365234375,
-      "learning_rate": 8.540108730516248e-05,
-      "loss": 0.355,
+      "epoch": 4.3478260869565215,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.599099966048627e-05,
+      "loss": 0.1847,
       "step": 2700
     },
     {
-      "epoch": 5.170172084130019,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.499066582430748e-05,
-      "loss": 0.3805,
+      "epoch": 4.35426731078905,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.558255520534937e-05,
+      "loss": 0.2451,
       "step": 2704
     },
     {
-      "epoch": 5.177820267686425,
-      "grad_norm": 0.333984375,
-      "learning_rate": 8.45808427204962e-05,
-      "loss": 0.3575,
+      "epoch": 4.360708534621578,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.517457485374336e-05,
+      "loss": 0.2112,
       "step": 2708
     },
     {
-      "epoch": 5.18546845124283,
-      "grad_norm": 0.373046875,
-      "learning_rate": 8.417162176593686e-05,
-      "loss": 0.3948,
+      "epoch": 4.367149758454106,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.476706208516138e-05,
+      "loss": 0.2048,
       "step": 2712
     },
     {
-      "epoch": 5.193116634799235,
-      "grad_norm": 0.345703125,
-      "learning_rate": 8.376300672729504e-05,
-      "loss": 0.3883,
+      "epoch": 4.373590982286634,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.43600203751086e-05,
+      "loss": 0.2036,
       "step": 2716
     },
     {
-      "epoch": 5.2007648183556405,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.335500136565919e-05,
-      "loss": 0.4049,
+      "epoch": 4.3800322061191626,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.395345319507287e-05,
+      "loss": 0.2125,
       "step": 2720
     },
     {
-      "epoch": 5.208413001912046,
-      "grad_norm": 0.376953125,
-      "learning_rate": 8.294760943650605e-05,
-      "loss": 0.3689,
+      "epoch": 4.386473429951691,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.354736401249486e-05,
+      "loss": 0.2199,
       "step": 2724
     },
     {
-      "epoch": 5.216061185468451,
-      "grad_norm": 0.384765625,
-      "learning_rate": 8.254083468966612e-05,
-      "loss": 0.3568,
+      "epoch": 4.392914653784219,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.31417562907387e-05,
+      "loss": 0.2064,
       "step": 2728
     },
     {
-      "epoch": 5.223709369024856,
-      "grad_norm": 0.361328125,
-      "learning_rate": 8.213468086928891e-05,
-      "loss": 0.3961,
+      "epoch": 4.399355877616747,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.273663348906222e-05,
+      "loss": 0.2183,
       "step": 2732
     },
     {
-      "epoch": 5.231357552581262,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.172915171380863e-05,
-      "loss": 0.3587,
+      "epoch": 4.405797101449275,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.233199906258766e-05,
+      "loss": 0.2639,
       "step": 2736
     },
     {
-      "epoch": 5.239005736137667,
-      "grad_norm": 0.357421875,
-      "learning_rate": 8.132425095590999e-05,
-      "loss": 0.407,
+      "epoch": 4.412238325281804,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.192785646227217e-05,
+      "loss": 0.251,
       "step": 2740
     },
     {
-      "epoch": 5.246653919694072,
-      "grad_norm": 0.369140625,
-      "learning_rate": 8.091998232249325e-05,
-      "loss": 0.4085,
+      "epoch": 4.418679549114332,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.152420913487814e-05,
+      "loss": 0.2386,
       "step": 2744
     },
     {
-      "epoch": 5.254302103250478,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.051634953464069e-05,
-      "loss": 0.366,
+      "epoch": 4.42512077294686,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.112106052294418e-05,
+      "loss": 0.217,
       "step": 2748
     },
     {
-      "epoch": 5.261950286806884,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.011335630758169e-05,
-      "loss": 0.3572,
+      "epoch": 4.431561996779388,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.071841406475539e-05,
+      "loss": 0.2102,
       "step": 2752
     },
     {
-      "epoch": 5.269598470363289,
-      "grad_norm": 0.392578125,
-      "learning_rate": 7.971100635065894e-05,
-      "loss": 0.4136,
+      "epoch": 4.438003220611916,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.03162731943144e-05,
+      "loss": 0.2282,
       "step": 2756
     },
     {
-      "epoch": 5.277246653919694,
-      "grad_norm": 0.3671875,
-      "learning_rate": 7.930930336729406e-05,
-      "loss": 0.3311,
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.302734375,
+      "learning_rate": 8.991464134131166e-05,
+      "loss": 0.2395,
       "step": 2760
     },
     {
-      "epoch": 5.2848948374761,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.890825105495376e-05,
-      "loss": 0.351,
+      "epoch": 4.450885668276973,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.951352193109673e-05,
+      "loss": 0.2379,
       "step": 2764
     },
     {
-      "epoch": 5.292543021032505,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.850785310511555e-05,
-      "loss": 0.3403,
+      "epoch": 4.457326892109501,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.911291838464838e-05,
+      "loss": 0.2319,
       "step": 2768
     },
     {
-      "epoch": 5.30019120458891,
+      "epoch": 4.463768115942029,
       "grad_norm": 0.3359375,
-      "learning_rate": 7.810811320323386e-05,
-      "loss": 0.2974,
+      "learning_rate": 8.871283411854619e-05,
+      "loss": 0.2066,
       "step": 2772
     },
     {
-      "epoch": 5.307839388145315,
-      "grad_norm": 0.36328125,
-      "learning_rate": 7.770903502870625e-05,
-      "loss": 0.3635,
+      "epoch": 4.4702093397745575,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.831327254494066e-05,
+      "loss": 0.2086,
       "step": 2776
     },
     {
-      "epoch": 5.315487571701721,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.731062225483933e-05,
-      "loss": 0.3669,
+      "epoch": 4.476650563607086,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.791423707152482e-05,
+      "loss": 0.2454,
       "step": 2780
     },
     {
-      "epoch": 5.323135755258126,
-      "grad_norm": 0.330078125,
-      "learning_rate": 7.6912878548815e-05,
-      "loss": 0.3823,
+      "epoch": 4.483091787439614,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.751573110150443e-05,
+      "loss": 0.2254,
       "step": 2784
     },
     {
-      "epoch": 5.330783938814531,
-      "grad_norm": 0.37890625,
-      "learning_rate": 7.651580757165691e-05,
-      "loss": 0.3644,
+      "epoch": 4.489533011272142,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.711775803356971e-05,
+      "loss": 0.2446,
       "step": 2788
     },
     {
-      "epoch": 5.338432122370937,
-      "grad_norm": 0.361328125,
-      "learning_rate": 7.611941297819643e-05,
-      "loss": 0.3505,
+      "epoch": 4.49597423510467,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.672032126186566e-05,
+      "loss": 0.2154,
       "step": 2792
     },
     {
-      "epoch": 5.346080305927342,
-      "grad_norm": 0.361328125,
-      "learning_rate": 7.572369841703924e-05,
-      "loss": 0.3691,
+      "epoch": 4.5024154589371985,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.632342417596365e-05,
+      "loss": 0.2269,
       "step": 2796
     },
     {
-      "epoch": 5.353728489483748,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.532866753053159e-05,
-      "loss": 0.3229,
+      "epoch": 4.508856682769727,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.592707016083221e-05,
+      "loss": 0.2134,
       "step": 2800
     },
     {
-      "epoch": 5.361376673040153,
-      "grad_norm": 0.39453125,
-      "learning_rate": 7.493432395472711e-05,
-      "loss": 0.3849,
+      "epoch": 4.515297906602254,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.553126259680828e-05,
+      "loss": 0.25,
       "step": 2804
     },
     {
-      "epoch": 5.369024856596559,
-      "grad_norm": 0.375,
-      "learning_rate": 7.454067131935269e-05,
-      "loss": 0.4105,
+      "epoch": 4.521739130434782,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.513600485956835e-05,
+      "loss": 0.2262,
       "step": 2808
     },
     {
-      "epoch": 5.376673040152964,
-      "grad_norm": 0.369140625,
-      "learning_rate": 7.414771324777579e-05,
-      "loss": 0.3686,
+      "epoch": 4.5281803542673105,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.474130032009951e-05,
+      "loss": 0.2062,
       "step": 2812
     },
     {
-      "epoch": 5.384321223709369,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.375545335697085e-05,
-      "loss": 0.3293,
+      "epoch": 4.534621578099839,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.434715234467123e-05,
+      "loss": 0.2293,
       "step": 2816
     },
     {
-      "epoch": 5.3919694072657744,
-      "grad_norm": 0.3828125,
-      "learning_rate": 7.336389525748548e-05,
-      "loss": 0.3892,
+      "epoch": 4.541062801932367,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.395356429480587e-05,
+      "loss": 0.204,
       "step": 2820
     },
     {
-      "epoch": 5.39961759082218,
-      "grad_norm": 0.3984375,
-      "learning_rate": 7.29730425534081e-05,
-      "loss": 0.3562,
+      "epoch": 4.547504025764895,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.356053952725072e-05,
+      "loss": 0.2207,
       "step": 2824
     },
     {
-      "epoch": 5.407265774378585,
-      "grad_norm": 0.376953125,
-      "learning_rate": 7.258289884233417e-05,
-      "loss": 0.363,
+      "epoch": 4.553945249597423,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.316808139394876e-05,
+      "loss": 0.2231,
       "step": 2828
     },
     {
-      "epoch": 5.41491395793499,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.21934677153332e-05,
-      "loss": 0.3517,
+      "epoch": 4.5603864734299515,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.277619324201081e-05,
+      "loss": 0.2321,
       "step": 2832
     },
     {
-      "epoch": 5.422562141491396,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.180475275691573e-05,
-      "loss": 0.39,
+      "epoch": 4.56682769726248,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.238487841368617e-05,
+      "loss": 0.2298,
       "step": 2836
     },
     {
-      "epoch": 5.430210325047801,
-      "grad_norm": 0.376953125,
-      "learning_rate": 7.141675754500049e-05,
-      "loss": 0.3898,
+      "epoch": 4.573268921095008,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.199414024633473e-05,
+      "loss": 0.1997,
       "step": 2840
     },
     {
-      "epoch": 5.437858508604206,
-      "grad_norm": 0.396484375,
-      "learning_rate": 7.102948565088116e-05,
-      "loss": 0.4363,
+      "epoch": 4.579710144927536,
+      "grad_norm": 0.3828125,
+      "learning_rate": 8.160398207239805e-05,
+      "loss": 0.2359,
       "step": 2844
     },
     {
-      "epoch": 5.4455066921606115,
-      "grad_norm": 0.373046875,
-      "learning_rate": 7.064294063919368e-05,
-      "loss": 0.3494,
+      "epoch": 4.586151368760064,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.121440721937157e-05,
+      "loss": 0.216,
       "step": 2848
     },
     {
-      "epoch": 5.453154875717018,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.025712606788362e-05,
-      "loss": 0.3457,
+      "epoch": 4.592592592592593,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.082541900977542e-05,
+      "loss": 0.2374,
       "step": 2852
     },
     {
-      "epoch": 5.460803059273423,
-      "grad_norm": 0.369140625,
-      "learning_rate": 6.987204548817278e-05,
-      "loss": 0.3279,
+      "epoch": 4.599033816425121,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.04370207611267e-05,
+      "loss": 0.2363,
       "step": 2856
     },
     {
-      "epoch": 5.468451242829828,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.948770244452737e-05,
-      "loss": 0.3591,
+      "epoch": 4.605475040257649,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.004921578591091e-05,
+      "loss": 0.214,
       "step": 2860
     },
     {
-      "epoch": 5.4760994263862335,
-      "grad_norm": 0.365234375,
-      "learning_rate": 6.910410047462495e-05,
-      "loss": 0.3488,
+      "epoch": 4.611916264090177,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.966200739155389e-05,
+      "loss": 0.2214,
       "step": 2864
     },
     {
-      "epoch": 5.483747609942639,
-      "grad_norm": 0.3671875,
-      "learning_rate": 6.87212431093215e-05,
-      "loss": 0.3466,
+      "epoch": 4.618357487922705,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.927539888039339e-05,
+      "loss": 0.2431,
       "step": 2868
     },
     {
-      "epoch": 5.491395793499044,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.833913387261973e-05,
-      "loss": 0.4094,
+      "epoch": 4.624798711755234,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.888939354965093e-05,
+      "loss": 0.2104,
       "step": 2872
     },
     {
-      "epoch": 5.499043977055449,
-      "grad_norm": 0.375,
-      "learning_rate": 6.795777628163599e-05,
-      "loss": 0.3803,
+      "epoch": 4.631239935587762,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.850399469140393e-05,
+      "loss": 0.204,
       "step": 2876
     },
     {
-      "epoch": 5.506692160611855,
-      "grad_norm": 0.361328125,
-      "learning_rate": 6.757717384656817e-05,
-      "loss": 0.3832,
+      "epoch": 4.63768115942029,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.811920559255736e-05,
+      "loss": 0.2263,
       "step": 2880
     },
     {
-      "epoch": 5.51434034416826,
-      "grad_norm": 0.37109375,
-      "learning_rate": 6.719733007066331e-05,
-      "loss": 0.3575,
+      "epoch": 4.644122383252818,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.773502953481585e-05,
+      "loss": 0.2161,
       "step": 2884
     },
     {
-      "epoch": 5.521988527724665,
-      "grad_norm": 0.357421875,
-      "learning_rate": 6.68182484501855e-05,
-      "loss": 0.3435,
+      "epoch": 4.650563607085346,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.73514697946556e-05,
+      "loss": 0.2279,
       "step": 2888
     },
     {
-      "epoch": 5.5296367112810705,
-      "grad_norm": 0.359375,
-      "learning_rate": 6.643993247438347e-05,
-      "loss": 0.3674,
+      "epoch": 4.657004830917875,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.696852964329655e-05,
+      "loss": 0.2615,
       "step": 2892
     },
     {
-      "epoch": 5.537284894837476,
-      "grad_norm": 0.361328125,
-      "learning_rate": 6.606238562545859e-05,
-      "loss": 0.3507,
+      "epoch": 4.663446054750403,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.658621234667443e-05,
+      "loss": 0.2407,
       "step": 2896
     },
     {
-      "epoch": 5.544933078393882,
-      "grad_norm": 0.357421875,
-      "learning_rate": 6.568561137853296e-05,
-      "loss": 0.3555,
+      "epoch": 4.669887278582931,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.620452116541291e-05,
+      "loss": 0.2101,
       "step": 2900
     },
     {
-      "epoch": 5.552581261950287,
-      "grad_norm": 0.37890625,
-      "learning_rate": 6.530961320161712e-05,
-      "loss": 0.3786,
+      "epoch": 4.676328502415459,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.582345935479569e-05,
+      "loss": 0.2191,
       "step": 2904
     },
     {
-      "epoch": 5.5602294455066925,
-      "grad_norm": 0.373046875,
-      "learning_rate": 6.493439455557835e-05,
-      "loss": 0.4043,
+      "epoch": 4.6827697262479875,
+      "grad_norm": 0.302734375,
+      "learning_rate": 7.544303016473894e-05,
+      "loss": 0.2159,
       "step": 2908
     },
     {
-      "epoch": 5.567877629063098,
-      "grad_norm": 0.3671875,
-      "learning_rate": 6.455995889410873e-05,
-      "loss": 0.4086,
+      "epoch": 4.689210950080515,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.506323683976344e-05,
+      "loss": 0.2251,
       "step": 2912
     },
     {
-      "epoch": 5.575525812619503,
-      "grad_norm": 0.365234375,
-      "learning_rate": 6.418630966369348e-05,
-      "loss": 0.366,
+      "epoch": 4.695652173913043,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.468408261896701e-05,
+      "loss": 0.1935,
       "step": 2916
     },
     {
-      "epoch": 5.583173996175908,
+      "epoch": 4.702093397745571,
       "grad_norm": 0.341796875,
-      "learning_rate": 6.381345030357899e-05,
-      "loss": 0.3801,
+      "learning_rate": 7.430557073599662e-05,
+      "loss": 0.2123,
       "step": 2920
     },
     {
-      "epoch": 5.590822179732314,
-      "grad_norm": 0.3984375,
-      "learning_rate": 6.344138424574134e-05,
-      "loss": 0.3985,
+      "epoch": 4.708534621578099,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.392770441902116e-05,
+      "loss": 0.2466,
       "step": 2924
     },
     {
-      "epoch": 5.598470363288719,
-      "grad_norm": 0.390625,
-      "learning_rate": 6.307011491485484e-05,
-      "loss": 0.4266,
+      "epoch": 4.714975845410628,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.355048689070389e-05,
+      "loss": 0.2332,
       "step": 2928
     },
     {
-      "epoch": 5.606118546845124,
-      "grad_norm": 0.359375,
-      "learning_rate": 6.269964572826001e-05,
-      "loss": 0.3317,
+      "epoch": 4.721417069243156,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.317392136817453e-05,
+      "loss": 0.2364,
       "step": 2932
     },
     {
-      "epoch": 5.6137667304015295,
-      "grad_norm": 0.384765625,
-      "learning_rate": 6.232998009593275e-05,
-      "loss": 0.3663,
+      "epoch": 4.727858293075684,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.279801106300231e-05,
+      "loss": 0.2662,
       "step": 2936
     },
     {
-      "epoch": 5.621414913957935,
-      "grad_norm": 0.392578125,
-      "learning_rate": 6.196112142045268e-05,
-      "loss": 0.3804,
+      "epoch": 4.734299516908212,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.242275918116832e-05,
+      "loss": 0.2174,
       "step": 2940
     },
     {
-      "epoch": 5.62906309751434,
-      "grad_norm": 0.380859375,
-      "learning_rate": 6.159307309697149e-05,
-      "loss": 0.3782,
+      "epoch": 4.7407407407407405,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.204816892303833e-05,
+      "loss": 0.2135,
       "step": 2944
     },
     {
-      "epoch": 5.636711281070745,
-      "grad_norm": 0.400390625,
-      "learning_rate": 6.122583851318233e-05,
-      "loss": 0.4047,
+      "epoch": 4.747181964573269,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.16742434833352e-05,
+      "loss": 0.231,
       "step": 2948
     },
     {
-      "epoch": 5.644359464627151,
+      "epoch": 4.753623188405797,
       "grad_norm": 0.34765625,
-      "learning_rate": 6.085942104928815e-05,
-      "loss": 0.3875,
+      "learning_rate": 7.1300986051112e-05,
+      "loss": 0.2569,
       "step": 2952
     },
     {
-      "epoch": 5.652007648183556,
-      "grad_norm": 0.388671875,
-      "learning_rate": 6.049382407797076e-05,
-      "loss": 0.3649,
+      "epoch": 4.760064412238325,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.09283998097246e-05,
+      "loss": 0.2072,
       "step": 2956
     },
     {
-      "epoch": 5.659655831739962,
-      "grad_norm": 0.3359375,
-      "learning_rate": 6.012905096435968e-05,
-      "loss": 0.3813,
+      "epoch": 4.766505636070853,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.055648793680466e-05,
+      "loss": 0.2059,
       "step": 2960
     },
     {
-      "epoch": 5.667304015296367,
-      "grad_norm": 0.396484375,
-      "learning_rate": 5.976510506600146e-05,
-      "loss": 0.3955,
+      "epoch": 4.7729468599033815,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.018525360423217e-05,
+      "loss": 0.2429,
       "step": 2964
     },
     {
-      "epoch": 5.674952198852773,
-      "grad_norm": 0.38671875,
-      "learning_rate": 5.9401989732828384e-05,
-      "loss": 0.3574,
+      "epoch": 4.77938808373591,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.981469997810892e-05,
+      "loss": 0.2203,
       "step": 2968
     },
     {
-      "epoch": 5.682600382409178,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.9039708307127816e-05,
-      "loss": 0.333,
+      "epoch": 4.785829307568438,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.944483021873115e-05,
+      "loss": 0.232,
       "step": 2972
     },
     {
-      "epoch": 5.690248565965583,
-      "grad_norm": 0.361328125,
-      "learning_rate": 5.8678264123511626e-05,
-      "loss": 0.3856,
+      "epoch": 4.792270531400966,
+      "grad_norm": 0.318359375,
+      "learning_rate": 6.907564748056273e-05,
+      "loss": 0.2124,
       "step": 2976
     },
     {
-      "epoch": 5.6978967495219885,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.8317660508885e-05,
-      "loss": 0.3931,
+      "epoch": 4.798711755233494,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.870715491220808e-05,
+      "loss": 0.2184,
       "step": 2980
     },
     {
-      "epoch": 5.705544933078394,
-      "grad_norm": 0.375,
-      "learning_rate": 5.795790078241641e-05,
-      "loss": 0.3364,
+      "epoch": 4.805152979066023,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.833935565638559e-05,
+      "loss": 0.238,
       "step": 2984
     },
     {
-      "epoch": 5.713193116634799,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.7598988255506644e-05,
-      "loss": 0.3825,
+      "epoch": 4.811594202898551,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.797225284990064e-05,
+      "loss": 0.2283,
       "step": 2988
     },
     {
-      "epoch": 5.720841300191204,
-      "grad_norm": 0.376953125,
-      "learning_rate": 5.724092623175841e-05,
-      "loss": 0.3654,
+      "epoch": 4.818035426731079,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.760584962361888e-05,
+      "loss": 0.2351,
       "step": 2992
     },
     {
-      "epoch": 5.72848948374761,
-      "grad_norm": 0.373046875,
-      "learning_rate": 5.6883718006946146e-05,
-      "loss": 0.4175,
+      "epoch": 4.824476650563607,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.72401491024396e-05,
+      "loss": 0.2019,
       "step": 2996
     },
     {
-      "epoch": 5.736137667304015,
-      "grad_norm": 0.392578125,
-      "learning_rate": 5.652736686898537e-05,
-      "loss": 0.369,
+      "epoch": 4.830917874396135,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.687515440526882e-05,
+      "loss": 0.242,
       "step": 3000
     },
     {
-      "epoch": 5.743785850860421,
-      "grad_norm": 0.3671875,
-      "learning_rate": 5.61718760979026e-05,
-      "loss": 0.3961,
+      "epoch": 4.837359098228664,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.651086864499305e-05,
+      "loss": 0.2196,
       "step": 3004
     },
     {
-      "epoch": 5.751434034416826,
-      "grad_norm": 0.353515625,
-      "learning_rate": 5.5817248965805096e-05,
-      "loss": 0.3622,
+      "epoch": 4.843800322061192,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.614729492845258e-05,
+      "loss": 0.2146,
       "step": 3008
     },
     {
-      "epoch": 5.759082217973232,
-      "grad_norm": 0.36328125,
-      "learning_rate": 5.546348873685089e-05,
-      "loss": 0.3369,
+      "epoch": 4.85024154589372,
+      "grad_norm": 0.375,
+      "learning_rate": 6.578443635641497e-05,
+      "loss": 0.2232,
       "step": 3012
     },
     {
-      "epoch": 5.766730401529637,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.51105986672185e-05,
-      "loss": 0.3314,
+      "epoch": 4.856682769726248,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.542229602354847e-05,
+      "loss": 0.2319,
       "step": 3016
     },
     {
-      "epoch": 5.774378585086042,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.475858200507708e-05,
-      "loss": 0.3212,
+      "epoch": 4.8631239935587764,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.506087701839593e-05,
+      "loss": 0.2156,
       "step": 3020
     },
     {
-      "epoch": 5.782026768642448,
-      "grad_norm": 0.375,
-      "learning_rate": 5.440744199055663e-05,
-      "loss": 0.3978,
+      "epoch": 4.869565217391305,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.470018242334825e-05,
+      "loss": 0.2372,
       "step": 3024
     },
     {
-      "epoch": 5.789674952198853,
-      "grad_norm": 0.37890625,
-      "learning_rate": 5.4057181855718e-05,
-      "loss": 0.3938,
+      "epoch": 4.876006441223833,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.434021531461818e-05,
+      "loss": 0.2077,
       "step": 3028
     },
     {
-      "epoch": 5.797323135755258,
-      "grad_norm": 0.369140625,
-      "learning_rate": 5.370780482452317e-05,
-      "loss": 0.3613,
+      "epoch": 4.882447665056361,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.398097876221385e-05,
+      "loss": 0.2183,
       "step": 3032
     },
     {
-      "epoch": 5.804971319311663,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.335931411280559e-05,
-      "loss": 0.3717,
+      "epoch": 4.888888888888889,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.362247582991317e-05,
+      "loss": 0.2104,
       "step": 3036
     },
     {
-      "epoch": 5.812619502868069,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.3011712928240787e-05,
-      "loss": 0.3572,
+      "epoch": 4.8953301127214175,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.326470957523686e-05,
+      "loss": 0.2048,
       "step": 3040
     },
     {
-      "epoch": 5.820267686424474,
-      "grad_norm": 0.359375,
-      "learning_rate": 5.2665004470316456e-05,
-      "loss": 0.3993,
+      "epoch": 4.901771336553946,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.29076830494232e-05,
+      "loss": 0.2346,
       "step": 3044
     },
     {
-      "epoch": 5.827915869980879,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.231919193030324e-05,
-      "loss": 0.3708,
+      "epoch": 4.908212560386474,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.255139929740129e-05,
+      "loss": 0.2068,
       "step": 3048
     },
     {
-      "epoch": 5.835564053537285,
-      "grad_norm": 0.37890625,
-      "learning_rate": 5.197427849122549e-05,
-      "loss": 0.3877,
+      "epoch": 4.914653784219001,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.219586135776575e-05,
+      "loss": 0.239,
       "step": 3052
     },
     {
-      "epoch": 5.84321223709369,
-      "grad_norm": 0.34375,
-      "learning_rate": 5.1630267327831494e-05,
-      "loss": 0.3747,
+      "epoch": 4.921095008051529,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.184107226275038e-05,
+      "loss": 0.1814,
       "step": 3056
     },
     {
-      "epoch": 5.850860420650095,
-      "grad_norm": 0.373046875,
-      "learning_rate": 5.128716160656489e-05,
-      "loss": 0.4003,
+      "epoch": 4.927536231884058,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.148703503820224e-05,
+      "loss": 0.2272,
       "step": 3060
     },
     {
-      "epoch": 5.858508604206501,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.0944964485534975e-05,
-      "loss": 0.3654,
+      "epoch": 4.933977455716586,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.113375270355617e-05,
+      "loss": 0.2418,
       "step": 3064
     },
     {
-      "epoch": 5.866156787762907,
-      "grad_norm": 0.39453125,
-      "learning_rate": 5.06036791144879e-05,
-      "loss": 0.3837,
+      "epoch": 4.940418679549114,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.078122827180879e-05,
+      "loss": 0.2723,
       "step": 3068
     },
     {
-      "epoch": 5.873804971319312,
-      "grad_norm": 0.40625,
-      "learning_rate": 5.0263308634777745e-05,
-      "loss": 0.368,
+      "epoch": 4.946859903381642,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.042946474949302e-05,
+      "loss": 0.2407,
       "step": 3072
     },
     {
-      "epoch": 5.881453154875717,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.992385617933734e-05,
-      "loss": 0.3975,
+      "epoch": 4.9533011272141705,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.007846513665207e-05,
+      "loss": 0.2153,
       "step": 3076
     },
     {
-      "epoch": 5.8891013384321225,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.958532487264968e-05,
-      "loss": 0.3837,
+      "epoch": 4.959742351046699,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.972823242681426e-05,
+      "loss": 0.2206,
       "step": 3080
     },
     {
-      "epoch": 5.896749521988528,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.924771783071895e-05,
-      "loss": 0.3744,
+      "epoch": 4.966183574879227,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.937876960696727e-05,
+      "loss": 0.2105,
       "step": 3084
     },
     {
-      "epoch": 5.904397705544933,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.8911038161042136e-05,
-      "loss": 0.3439,
+      "epoch": 4.972624798711755,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.903007965753279e-05,
+      "loss": 0.2526,
       "step": 3088
     },
     {
-      "epoch": 5.912045889101338,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.857528896258012e-05,
-      "loss": 0.4041,
+      "epoch": 4.979066022544283,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.868216555234081e-05,
+      "loss": 0.2168,
       "step": 3092
     },
     {
-      "epoch": 5.919694072657744,
-      "grad_norm": 0.375,
-      "learning_rate": 4.824047332572924e-05,
-      "loss": 0.3753,
+      "epoch": 4.9855072463768115,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.833503025860469e-05,
+      "loss": 0.2174,
       "step": 3096
     },
     {
-      "epoch": 5.927342256214149,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.7906594332293116e-05,
-      "loss": 0.3822,
+      "epoch": 4.99194847020934,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.798867673689553e-05,
+      "loss": 0.2365,
       "step": 3100
     },
     {
-      "epoch": 5.934990439770554,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.75736550554537e-05,
-      "loss": 0.3535,
+      "epoch": 4.998389694041868,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.764310794111711e-05,
+      "loss": 0.1766,
       "step": 3104
     },
     {
-      "epoch": 5.9426386233269595,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.724165855974367e-05,
-      "loss": 0.377,
+      "epoch": 5.004830917874396,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.7298326818480427e-05,
+      "loss": 0.1998,
       "step": 3108
     },
     {
-      "epoch": 5.950286806883366,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.6910607901017715e-05,
-      "loss": 0.359,
+      "epoch": 5.011272141706924,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.695433630947894e-05,
+      "loss": 0.1838,
       "step": 3112
     },
     {
-      "epoch": 5.957934990439771,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.65805061264246e-05,
-      "loss": 0.3415,
+      "epoch": 5.017713365539453,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.661113934786321e-05,
+      "loss": 0.2045,
       "step": 3116
     },
     {
-      "epoch": 5.965583173996176,
-      "grad_norm": 0.390625,
-      "learning_rate": 4.625135627437922e-05,
-      "loss": 0.3968,
+      "epoch": 5.024154589371981,
+      "grad_norm": 0.28125,
+      "learning_rate": 5.626873886061597e-05,
+      "loss": 0.1917,
       "step": 3120
     },
     {
-      "epoch": 5.9732313575525815,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.592316137453439e-05,
-      "loss": 0.3983,
+      "epoch": 5.030595813204509,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.592713776792723e-05,
+      "loss": 0.204,
       "step": 3124
     },
     {
-      "epoch": 5.980879541108987,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.559592444775315e-05,
-      "loss": 0.3466,
+      "epoch": 5.037037037037037,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.5586338983169076e-05,
+      "loss": 0.1471,
       "step": 3128
     },
     {
-      "epoch": 5.988527724665392,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.5269648506080816e-05,
-      "loss": 0.3442,
+      "epoch": 5.043478260869565,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.52463454128714e-05,
+      "loss": 0.1966,
       "step": 3132
     },
     {
-      "epoch": 5.996175908221797,
-      "grad_norm": 0.375,
-      "learning_rate": 4.4944336552717514e-05,
-      "loss": 0.3768,
+      "epoch": 5.049919484702094,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.490715995669641e-05,
+      "loss": 0.1782,
       "step": 3136
     },
     {
-      "epoch": 6.003824091778203,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.461999158199019e-05,
-      "loss": 0.3533,
+      "epoch": 5.056360708534622,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.456878550741453e-05,
+      "loss": 0.1877,
       "step": 3140
     },
     {
-      "epoch": 6.011472275334608,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.429661657932523e-05,
-      "loss": 0.3247,
+      "epoch": 5.06280193236715,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.423122495087915e-05,
+      "loss": 0.1643,
       "step": 3144
     },
     {
-      "epoch": 6.019120458891013,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.397421452122114e-05,
-      "loss": 0.329,
+      "epoch": 5.069243156199678,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.3894481166002674e-05,
+      "loss": 0.1792,
       "step": 3148
     },
     {
-      "epoch": 6.0267686424474185,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.3652788375220787e-05,
-      "loss": 0.3547,
+      "epoch": 5.0756843800322065,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.355855702473125e-05,
+      "loss": 0.1567,
       "step": 3152
     },
     {
-      "epoch": 6.034416826003824,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.333234109988434e-05,
-      "loss": 0.4026,
+      "epoch": 5.082125603864735,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.322345539202086e-05,
+      "loss": 0.2051,
       "step": 3156
     },
     {
-      "epoch": 6.042065009560229,
-      "grad_norm": 0.3828125,
-      "learning_rate": 4.3012875644761955e-05,
-      "loss": 0.3534,
+      "epoch": 5.088566827697263,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.288917912581257e-05,
+      "loss": 0.1754,
       "step": 3160
     },
     {
-      "epoch": 6.049713193116634,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.269439495036678e-05,
-      "loss": 0.3035,
+      "epoch": 5.095008051529791,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.255573107700832e-05,
+      "loss": 0.1824,
       "step": 3164
     },
     {
-      "epoch": 6.0573613766730405,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.2376901948147465e-05,
-      "loss": 0.3553,
+      "epoch": 5.101449275362318,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.222311408944635e-05,
+      "loss": 0.2092,
       "step": 3168
     },
     {
-      "epoch": 6.065009560229446,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.206039956046176e-05,
-      "loss": 0.3602,
+      "epoch": 5.107890499194847,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.189133099987731e-05,
+      "loss": 0.146,
       "step": 3172
     },
     {
-      "epoch": 6.072657743785851,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.174489070054927e-05,
-      "loss": 0.3606,
+      "epoch": 5.114331723027375,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.156038463793981e-05,
+      "loss": 0.1692,
       "step": 3176
     },
     {
-      "epoch": 6.080305927342256,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.143037827250447e-05,
-      "loss": 0.3499,
+      "epoch": 5.120772946859903,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.123027782613636e-05,
+      "loss": 0.1877,
       "step": 3180
     },
     {
-      "epoch": 6.087954110898662,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.1116865171250496e-05,
-      "loss": 0.3447,
+      "epoch": 5.127214170692431,
+      "grad_norm": 0.30078125,
+      "learning_rate": 5.09010133798094e-05,
+      "loss": 0.154,
       "step": 3184
     },
     {
-      "epoch": 6.095602294455067,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.0804354282512016e-05,
-      "loss": 0.3541,
+      "epoch": 5.1336553945249594,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.0572594107116974e-05,
+      "loss": 0.1559,
       "step": 3188
     },
     {
-      "epoch": 6.103250478011472,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.049284848278886e-05,
-      "loss": 0.3638,
+      "epoch": 5.140096618357488,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.0245022809009155e-05,
+      "loss": 0.171,
       "step": 3192
     },
     {
-      "epoch": 6.1108986615678775,
-      "grad_norm": 0.373046875,
-      "learning_rate": 4.01823506393297e-05,
-      "loss": 0.362,
+      "epoch": 5.146537842190016,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.991830227920398e-05,
+      "loss": 0.1774,
       "step": 3196
     },
     {
-      "epoch": 6.118546845124283,
-      "grad_norm": 0.3828125,
-      "learning_rate": 3.987286361010531e-05,
-      "loss": 0.3367,
+      "epoch": 5.152979066022544,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.9592435304163675e-05,
+      "loss": 0.1813,
       "step": 3200
     },
     {
-      "epoch": 6.126195028680688,
-      "grad_norm": 0.361328125,
-      "learning_rate": 3.9564390243782516e-05,
-      "loss": 0.3249,
+      "epoch": 5.159420289855072,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.926742466307069e-05,
+      "loss": 0.1557,
       "step": 3204
     },
     {
-      "epoch": 6.133843212237093,
-      "grad_norm": 0.396484375,
-      "learning_rate": 3.925693337969788e-05,
-      "loss": 0.3328,
+      "epoch": 5.1658615136876005,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.8943273127804345e-05,
+      "loss": 0.1574,
       "step": 3208
     },
     {
-      "epoch": 6.141491395793499,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.895049584783173e-05,
-      "loss": 0.3746,
+      "epoch": 5.172302737520129,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.8619983462916935e-05,
+      "loss": 0.1548,
       "step": 3212
     },
     {
-      "epoch": 6.149139579349905,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.8645080468781676e-05,
-      "loss": 0.3702,
+      "epoch": 5.178743961352657,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.829755842561025e-05,
+      "loss": 0.1888,
       "step": 3216
     },
     {
-      "epoch": 6.15678776290631,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.834069005373724e-05,
-      "loss": 0.3136,
+      "epoch": 5.185185185185185,
+      "grad_norm": 0.2734375,
+      "learning_rate": 4.797600076571194e-05,
+      "loss": 0.2004,
       "step": 3220
     },
     {
-      "epoch": 6.164435946462715,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.8037327404453634e-05,
-      "loss": 0.3542,
+      "epoch": 5.191626409017713,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.7655313225652294e-05,
+      "loss": 0.1587,
       "step": 3224
     },
     {
-      "epoch": 6.172084130019121,
-      "grad_norm": 0.369140625,
-      "learning_rate": 3.77349953132258e-05,
-      "loss": 0.3572,
+      "epoch": 5.1980676328502415,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.7335498540440606e-05,
+      "loss": 0.1669,
       "step": 3228
     },
     {
-      "epoch": 6.179732313575526,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.7433696562863215e-05,
-      "loss": 0.3526,
+      "epoch": 5.20450885668277,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.7016559437642084e-05,
+      "loss": 0.171,
       "step": 3232
     },
     {
-      "epoch": 6.187380497131931,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.7133433926663805e-05,
-      "loss": 0.331,
+      "epoch": 5.210950080515298,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.6698498637354225e-05,
+      "loss": 0.1566,
       "step": 3236
     },
     {
-      "epoch": 6.195028680688337,
-      "grad_norm": 0.376953125,
-      "learning_rate": 3.6834210168388674e-05,
-      "loss": 0.3489,
+      "epoch": 5.217391304347826,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.6381318852184194e-05,
+      "loss": 0.1936,
       "step": 3240
     },
     {
-      "epoch": 6.202676864244742,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.653602804223656e-05,
-      "loss": 0.3329,
+      "epoch": 5.223832528180354,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.606502278722503e-05,
+      "loss": 0.1897,
       "step": 3244
     },
     {
-      "epoch": 6.210325047801147,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.623889029281861e-05,
-      "loss": 0.2646,
+      "epoch": 5.230273752012883,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.574961314003304e-05,
+      "loss": 0.1935,
       "step": 3248
     },
     {
-      "epoch": 6.217973231357552,
-      "grad_norm": 0.373046875,
-      "learning_rate": 3.5942799655132925e-05,
-      "loss": 0.3547,
+      "epoch": 5.236714975845411,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.5435092600604676e-05,
+      "loss": 0.159,
       "step": 3252
     },
     {
-      "epoch": 6.225621414913958,
-      "grad_norm": 0.369140625,
-      "learning_rate": 3.56477588545395e-05,
-      "loss": 0.3547,
+      "epoch": 5.243156199677939,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.5121463851353476e-05,
+      "loss": 0.2065,
       "step": 3256
     },
     {
-      "epoch": 6.233269598470363,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.535377060673524e-05,
-      "loss": 0.3205,
+      "epoch": 5.249597423510467,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.48087295670874e-05,
+      "loss": 0.1993,
       "step": 3260
     },
     {
-      "epoch": 6.240917782026768,
-      "grad_norm": 0.361328125,
-      "learning_rate": 3.506083761772871e-05,
-      "loss": 0.3647,
+      "epoch": 5.256038647342995,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.449689241498569e-05,
+      "loss": 0.1717,
       "step": 3264
     },
     {
-      "epoch": 6.248565965583174,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.476896258381537e-05,
-      "loss": 0.3138,
+      "epoch": 5.262479871175524,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.41859550545765e-05,
+      "loss": 0.1907,
       "step": 3268
     },
     {
-      "epoch": 6.25621414913958,
-      "grad_norm": 0.3828125,
-      "learning_rate": 3.447814819155291e-05,
-      "loss": 0.3071,
+      "epoch": 5.268921095008052,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.387592013771396e-05,
+      "loss": 0.1691,
       "step": 3272
     },
     {
-      "epoch": 6.263862332695985,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.418839711773623e-05,
-      "loss": 0.3353,
+      "epoch": 5.27536231884058,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.356679030855573e-05,
+      "loss": 0.198,
       "step": 3276
     },
     {
-      "epoch": 6.27151051625239,
-      "grad_norm": 0.384765625,
-      "learning_rate": 3.389971202937295e-05,
-      "loss": 0.3497,
+      "epoch": 5.281803542673108,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.32585682035402e-05,
+      "loss": 0.2026,
       "step": 3280
     },
     {
-      "epoch": 6.279158699808796,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.361209558365883e-05,
-      "loss": 0.3125,
+      "epoch": 5.2882447665056365,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.2951256451364264e-05,
+      "loss": 0.1975,
       "step": 3284
     },
     {
-      "epoch": 6.286806883365201,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.332555042795349e-05,
-      "loss": 0.3575,
+      "epoch": 5.294685990338165,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.264485767296081e-05,
+      "loss": 0.1686,
       "step": 3288
     },
     {
-      "epoch": 6.294455066921606,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.304007919975563e-05,
-      "loss": 0.3554,
+      "epoch": 5.301127214170693,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.233937448147635e-05,
+      "loss": 0.1583,
       "step": 3292
     },
     {
-      "epoch": 6.3021032504780115,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.2755684526679196e-05,
-      "loss": 0.3427,
+      "epoch": 5.30756843800322,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.203480948224866e-05,
+      "loss": 0.1777,
       "step": 3296
     },
     {
-      "epoch": 6.309751434034417,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.247236902642905e-05,
-      "loss": 0.3648,
+      "epoch": 5.314009661835748,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.173116527278471e-05,
+      "loss": 0.1616,
       "step": 3300
     },
     {
-      "epoch": 6.317399617590822,
-      "grad_norm": 0.375,
-      "learning_rate": 3.219013530677655e-05,
-      "loss": 0.3269,
+      "epoch": 5.320450885668277,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.142844444273845e-05,
+      "loss": 0.1731,
       "step": 3304
     },
     {
-      "epoch": 6.325047801147227,
+      "epoch": 5.326892109500805,
       "grad_norm": 0.349609375,
-      "learning_rate": 3.190898596553615e-05,
-      "loss": 0.3503,
+      "learning_rate": 4.1126649573888696e-05,
+      "loss": 0.2219,
       "step": 3308
     },
     {
-      "epoch": 6.332695984703633,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.162892359054098e-05,
-      "loss": 0.3239,
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.082578324011716e-05,
+      "loss": 0.1937,
       "step": 3312
     },
     {
-      "epoch": 6.340344168260038,
-      "grad_norm": 0.376953125,
-      "learning_rate": 3.1349950759619255e-05,
-      "loss": 0.3253,
+      "epoch": 5.339774557165861,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.052584800738636e-05,
+      "loss": 0.1891,
       "step": 3316
     },
     {
-      "epoch": 6.347992351816444,
-      "grad_norm": 0.361328125,
-      "learning_rate": 3.107207004057046e-05,
-      "loss": 0.3642,
+      "epoch": 5.3462157809983895,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.0226846433717954e-05,
+      "loss": 0.1811,
       "step": 3320
     },
     {
-      "epoch": 6.355640535372849,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.079528399114189e-05,
-      "loss": 0.3058,
+      "epoch": 5.352657004830918,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.992878106917079e-05,
+      "loss": 0.1768,
       "step": 3324
     },
     {
-      "epoch": 6.363288718929255,
-      "grad_norm": 0.375,
-      "learning_rate": 3.0519595159004853e-05,
-      "loss": 0.3705,
+      "epoch": 5.359098228663446,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.963165445581922e-05,
+      "loss": 0.1985,
       "step": 3328
     },
     {
-      "epoch": 6.37093690248566,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.0245006081731367e-05,
-      "loss": 0.3444,
+      "epoch": 5.365539452495974,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.933546912773119e-05,
+      "loss": 0.154,
       "step": 3332
     },
     {
-      "epoch": 6.378585086042065,
-      "grad_norm": 0.37109375,
-      "learning_rate": 2.9971519286770883e-05,
-      "loss": 0.3181,
+      "epoch": 5.371980676328502,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.904022761094715e-05,
+      "loss": 0.1788,
       "step": 3336
     },
     {
-      "epoch": 6.3862332695984705,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.969913729142668e-05,
-      "loss": 0.3519,
+      "epoch": 5.3784219001610305,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.874593242345785e-05,
+      "loss": 0.1964,
       "step": 3340
     },
     {
-      "epoch": 6.393881453154876,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.9427862602833165e-05,
-      "loss": 0.3261,
+      "epoch": 5.384863123993559,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.845258607518344e-05,
+      "loss": 0.1822,
       "step": 3344
     },
     {
-      "epoch": 6.401529636711281,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.915769771793256e-05,
-      "loss": 0.3554,
+      "epoch": 5.391304347826087,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.816019106795157e-05,
+      "loss": 0.1711,
       "step": 3348
     },
     {
-      "epoch": 6.409177820267686,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.8888645123451694e-05,
-      "loss": 0.3119,
+      "epoch": 5.397745571658615,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.7868749895476624e-05,
+      "loss": 0.1785,
       "step": 3352
     },
     {
-      "epoch": 6.416826003824092,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.862070729587959e-05,
-      "loss": 0.3576,
+      "epoch": 5.404186795491143,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.7578265043337834e-05,
+      "loss": 0.1891,
       "step": 3356
     },
     {
-      "epoch": 6.424474187380497,
-      "grad_norm": 0.3671875,
-      "learning_rate": 2.8353886701444312e-05,
-      "loss": 0.3464,
+      "epoch": 5.4106280193236715,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.72887389889586e-05,
+      "loss": 0.1766,
       "step": 3360
     },
     {
-      "epoch": 6.432122370936902,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.808818579609037e-05,
-      "loss": 0.3362,
+      "epoch": 5.4170692431562,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.700017420158486e-05,
+      "loss": 0.1733,
       "step": 3364
     },
     {
-      "epoch": 6.4397705544933075,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.7823607025456103e-05,
-      "loss": 0.3556,
+      "epoch": 5.423510466988728,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.671257314226471e-05,
+      "loss": 0.1895,
       "step": 3368
     },
     {
-      "epoch": 6.447418738049713,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.7560152824851285e-05,
-      "loss": 0.2955,
+      "epoch": 5.429951690821256,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.642593826382663e-05,
+      "loss": 0.1867,
       "step": 3372
     },
     {
-      "epoch": 6.455066921606119,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.7297825619234515e-05,
-      "loss": 0.3064,
+      "epoch": 5.436392914653784,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.6140272010859166e-05,
+      "loss": 0.1946,
       "step": 3376
     },
     {
-      "epoch": 6.462715105162524,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.7036627823190994e-05,
-      "loss": 0.3553,
+      "epoch": 5.442834138486313,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.585557681968979e-05,
+      "loss": 0.1684,
       "step": 3380
     },
     {
-      "epoch": 6.4703632887189295,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.6776561840910367e-05,
-      "loss": 0.3141,
+      "epoch": 5.449275362318841,
+      "grad_norm": 0.298828125,
+      "learning_rate": 3.5571855118364236e-05,
+      "loss": 0.1886,
       "step": 3384
     },
     {
-      "epoch": 6.478011472275335,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.6517630066164448e-05,
-      "loss": 0.3746,
+      "epoch": 5.455716586151369,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.528910932662577e-05,
+      "loss": 0.199,
       "step": 3388
     },
     {
-      "epoch": 6.48565965583174,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.6259834882285302e-05,
-      "loss": 0.3467,
+      "epoch": 5.462157809983897,
+      "grad_norm": 0.291015625,
+      "learning_rate": 3.5007341855894394e-05,
+      "loss": 0.1877,
       "step": 3392
     },
     {
-      "epoch": 6.493307839388145,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.6003178662143214e-05,
-      "loss": 0.3257,
+      "epoch": 5.468599033816425,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.472655510924656e-05,
+      "loss": 0.1856,
       "step": 3396
     },
     {
-      "epoch": 6.500956022944551,
-      "grad_norm": 0.36328125,
-      "learning_rate": 2.574766376812502e-05,
-      "loss": 0.3231,
+      "epoch": 5.475040257648954,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.4446751481394516e-05,
+      "loss": 0.1818,
       "step": 3400
     },
     {
-      "epoch": 6.508604206500956,
-      "grad_norm": 0.333984375,
-      "learning_rate": 2.5493292552112128e-05,
-      "loss": 0.3102,
+      "epoch": 5.481481481481482,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.4167933358665936e-05,
+      "loss": 0.2009,
       "step": 3404
     },
     {
-      "epoch": 6.516252390057361,
-      "grad_norm": 0.35546875,
-      "learning_rate": 2.5240067355458978e-05,
-      "loss": 0.33,
+      "epoch": 5.48792270531401,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.3890103118983366e-05,
+      "loss": 0.1824,
       "step": 3408
     },
     {
-      "epoch": 6.5239005736137665,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.4987990508971667e-05,
-      "loss": 0.3529,
+      "epoch": 5.494363929146537,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.3613263131844294e-05,
+      "loss": 0.1746,
       "step": 3412
     },
     {
-      "epoch": 6.531548757170172,
-      "grad_norm": 0.375,
-      "learning_rate": 2.4737064332886055e-05,
-      "loss": 0.3285,
+      "epoch": 5.500805152979066,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.333741575830069e-05,
+      "loss": 0.1769,
       "step": 3416
     },
     {
-      "epoch": 6.539196940726577,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.4487291136846894e-05,
-      "loss": 0.3835,
+      "epoch": 5.507246376811594,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.306256335093898e-05,
+      "loss": 0.1822,
       "step": 3420
     },
     {
-      "epoch": 6.546845124282983,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.4238673219886384e-05,
-      "loss": 0.3476,
+      "epoch": 5.513687600644122,
+      "grad_norm": 0.302734375,
+      "learning_rate": 3.278870825385983e-05,
+      "loss": 0.1925,
       "step": 3424
     },
     {
-      "epoch": 6.5544933078393885,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.399121287040275e-05,
-      "loss": 0.3209,
+      "epoch": 5.52012882447665,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.251585280265839e-05,
+      "loss": 0.1923,
       "step": 3428
     },
     {
-      "epoch": 6.562141491395794,
-      "grad_norm": 0.3671875,
-      "learning_rate": 2.3744912366139644e-05,
-      "loss": 0.3498,
+      "epoch": 5.526570048309178,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.224399932440419e-05,
+      "loss": 0.1815,
       "step": 3432
     },
     {
-      "epoch": 6.569789674952199,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.3499773974164825e-05,
-      "loss": 0.348,
+      "epoch": 5.533011272141707,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.1973150137621364e-05,
+      "loss": 0.1738,
       "step": 3436
     },
     {
-      "epoch": 6.577437858508604,
-      "grad_norm": 0.3671875,
-      "learning_rate": 2.325579995084946e-05,
-      "loss": 0.3783,
+      "epoch": 5.539452495974235,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.170330755226893e-05,
+      "loss": 0.191,
       "step": 3440
     },
     {
-      "epoch": 6.58508604206501,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.3012992541847254e-05,
-      "loss": 0.3855,
+      "epoch": 5.545893719806763,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.1434473869720804e-05,
+      "loss": 0.1538,
       "step": 3444
     },
     {
-      "epoch": 6.592734225621415,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.277135398207393e-05,
-      "loss": 0.3163,
+      "epoch": 5.552334943639291,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.116665138274676e-05,
+      "loss": 0.1748,
       "step": 3448
     },
     {
-      "epoch": 6.60038240917782,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.2530886495686506e-05,
-      "loss": 0.3601,
+      "epoch": 5.5587761674718195,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.0899842375492145e-05,
+      "loss": 0.1893,
       "step": 3452
     },
     {
-      "epoch": 6.6080305927342256,
-      "grad_norm": 0.396484375,
-      "learning_rate": 2.229159229606281e-05,
-      "loss": 0.3425,
+      "epoch": 5.565217391304348,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.063404912345897e-05,
+      "loss": 0.1727,
       "step": 3456
     },
     {
-      "epoch": 6.615678776290631,
-      "grad_norm": 0.38671875,
-      "learning_rate": 2.2053473585781377e-05,
-      "loss": 0.324,
+      "epoch": 5.571658615136876,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.036927389348625e-05,
+      "loss": 0.1804,
       "step": 3460
     },
     {
-      "epoch": 6.623326959847036,
-      "grad_norm": 0.37890625,
-      "learning_rate": 2.181653255660072e-05,
-      "loss": 0.3386,
+      "epoch": 5.578099838969404,
+      "grad_norm": 0.283203125,
+      "learning_rate": 3.010551894373075e-05,
+      "loss": 0.1778,
       "step": 3464
     },
     {
-      "epoch": 6.630975143403441,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.1580771389439612e-05,
-      "loss": 0.3434,
+      "epoch": 5.584541062801932,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.9842786523647582e-05,
+      "loss": 0.1679,
       "step": 3468
     },
     {
-      "epoch": 6.638623326959847,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.1346192254356737e-05,
-      "loss": 0.331,
+      "epoch": 5.5909822866344605,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.9581078873971248e-05,
+      "loss": 0.1812,
       "step": 3472
     },
     {
-      "epoch": 6.646271510516252,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.1112797310530716e-05,
-      "loss": 0.3483,
+      "epoch": 5.597423510466989,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.9320398226696367e-05,
+      "loss": 0.188,
       "step": 3476
     },
     {
-      "epoch": 6.653919694072657,
-      "grad_norm": 0.376953125,
-      "learning_rate": 2.08805887062405e-05,
-      "loss": 0.3241,
+      "epoch": 5.603864734299517,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.9060746805058738e-05,
+      "loss": 0.1541,
       "step": 3480
     },
     {
-      "epoch": 6.661567877629063,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.0649568578845205e-05,
-      "loss": 0.3269,
+      "epoch": 5.610305958132045,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8802126823516193e-05,
+      "loss": 0.1671,
       "step": 3484
     },
     {
-      "epoch": 6.669216061185469,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.0419739054764743e-05,
-      "loss": 0.3636,
+      "epoch": 5.616747181964573,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.8544540487729984e-05,
+      "loss": 0.1609,
       "step": 3488
     },
     {
-      "epoch": 6.676864244741874,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.019110224946008e-05,
-      "loss": 0.3549,
+      "epoch": 5.6231884057971016,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.828798999454577e-05,
+      "loss": 0.1488,
       "step": 3492
     },
     {
-      "epoch": 6.684512428298279,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.9963660267413913e-05,
-      "loss": 0.3497,
+      "epoch": 5.62962962962963,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.8032477531974984e-05,
+      "loss": 0.2012,
       "step": 3496
     },
     {
-      "epoch": 6.692160611854685,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.9737415202111144e-05,
-      "loss": 0.3639,
+      "epoch": 5.636070853462158,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.7778005279176053e-05,
+      "loss": 0.208,
       "step": 3500
     },
     {
-      "epoch": 6.69980879541109,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.9512369136019663e-05,
-      "loss": 0.3487,
+      "epoch": 5.642512077294686,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.7524575406435955e-05,
+      "loss": 0.192,
       "step": 3504
     },
     {
-      "epoch": 6.707456978967495,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.9288524140571286e-05,
-      "loss": 0.3196,
+      "epoch": 5.648953301127214,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.7272190075151655e-05,
+      "loss": 0.1582,
       "step": 3508
     },
     {
-      "epoch": 6.7151051625239,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.906588227614254e-05,
-      "loss": 0.3398,
+      "epoch": 5.655394524959743,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.7020851437811608e-05,
+      "loss": 0.1762,
       "step": 3512
     },
     {
-      "epoch": 6.722753346080306,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.8844445592035767e-05,
-      "loss": 0.37,
+      "epoch": 5.661835748792271,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.6770561637977556e-05,
+      "loss": 0.1678,
       "step": 3516
     },
     {
-      "epoch": 6.730401529636711,
-      "grad_norm": 0.37109375,
-      "learning_rate": 1.8624216126460183e-05,
-      "loss": 0.3466,
+      "epoch": 5.668276972624799,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.652132281026598e-05,
+      "loss": 0.1822,
       "step": 3520
     },
     {
-      "epoch": 6.738049713193116,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.8405195906513347e-05,
-      "loss": 0.3638,
+      "epoch": 5.674718196457327,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.6273137080330225e-05,
+      "loss": 0.183,
       "step": 3524
     },
     {
-      "epoch": 6.7456978967495225,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.8187386948162203e-05,
-      "loss": 0.3894,
+      "epoch": 5.681159420289855,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.6026006564842106e-05,
+      "loss": 0.2009,
       "step": 3528
     },
     {
-      "epoch": 6.753346080305928,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.797079125622469e-05,
-      "loss": 0.3109,
+      "epoch": 5.687600644122384,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.577993337147406e-05,
+      "loss": 0.1858,
       "step": 3532
     },
     {
-      "epoch": 6.760994263862333,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.7755410824351363e-05,
-      "loss": 0.3191,
+      "epoch": 5.694041867954912,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.5534919598880887e-05,
+      "loss": 0.203,
       "step": 3536
     },
     {
-      "epoch": 6.768642447418738,
-      "grad_norm": 0.37890625,
-      "learning_rate": 1.7541247635006756e-05,
-      "loss": 0.3548,
+      "epoch": 5.70048309178744,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.5290967336682266e-05,
+      "loss": 0.1588,
       "step": 3540
     },
     {
-      "epoch": 6.776290630975144,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.7328303659451477e-05,
-      "loss": 0.3518,
+      "epoch": 5.706924315619968,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.5048078665444497e-05,
+      "loss": 0.1622,
       "step": 3544
     },
     {
-      "epoch": 6.783938814531549,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.7116580857723872e-05,
-      "loss": 0.367,
+      "epoch": 5.713365539452496,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.4806255656663092e-05,
+      "loss": 0.185,
       "step": 3548
     },
     {
-      "epoch": 6.791586998087954,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.6906081178621917e-05,
-      "loss": 0.3502,
+      "epoch": 5.719806763285024,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.4565500372744845e-05,
+      "loss": 0.1904,
       "step": 3552
     },
     {
-      "epoch": 6.7992351816443595,
-      "grad_norm": 0.376953125,
-      "learning_rate": 1.6696806559685553e-05,
-      "loss": 0.4081,
+      "epoch": 5.726247987117552,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.4325814866990583e-05,
+      "loss": 0.175,
       "step": 3556
     },
     {
-      "epoch": 6.806883365200765,
-      "grad_norm": 0.38671875,
-      "learning_rate": 1.648875892717857e-05,
-      "loss": 0.3557,
+      "epoch": 5.73268921095008,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.4087201183577205e-05,
+      "loss": 0.1699,
       "step": 3560
     },
     {
-      "epoch": 6.81453154875717,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.628194019607099e-05,
-      "loss": 0.3765,
+      "epoch": 5.739130434782608,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.384966135754063e-05,
+      "loss": 0.1823,
       "step": 3564
     },
     {
-      "epoch": 6.822179732313575,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.6076352270021435e-05,
-      "loss": 0.3261,
+      "epoch": 5.745571658615137,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.3613197414758273e-05,
+      "loss": 0.1788,
       "step": 3568
     },
     {
-      "epoch": 6.829827915869981,
-      "grad_norm": 0.40625,
-      "learning_rate": 1.587199704135973e-05,
-      "loss": 0.3734,
+      "epoch": 5.752012882447665,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.3377811371931793e-05,
+      "loss": 0.1794,
       "step": 3572
     },
     {
-      "epoch": 6.837476099426386,
-      "grad_norm": 0.38671875,
-      "learning_rate": 1.5668876391069107e-05,
-      "loss": 0.3523,
+      "epoch": 5.758454106280193,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.3143505236569915e-05,
+      "loss": 0.1684,
       "step": 3576
     },
     {
-      "epoch": 6.845124282982791,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.5466992188769394e-05,
-      "loss": 0.3465,
+      "epoch": 5.764895330112721,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.2910281006971164e-05,
+      "loss": 0.157,
       "step": 3580
     },
     {
-      "epoch": 6.8527724665391965,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.5266346292699522e-05,
-      "loss": 0.3506,
+      "epoch": 5.7713365539452495,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.26781406722071e-05,
+      "loss": 0.1833,
       "step": 3584
     },
     {
-      "epoch": 6.860420650095603,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.5066940549700285e-05,
-      "loss": 0.3112,
+      "epoch": 5.777777777777778,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.2447086212105143e-05,
+      "loss": 0.1945,
       "step": 3588
     },
     {
-      "epoch": 6.868068833652008,
-      "grad_norm": 0.40234375,
-      "learning_rate": 1.4868776795197712e-05,
-      "loss": 0.3271,
+      "epoch": 5.784219001610306,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.2217119597231747e-05,
+      "loss": 0.1801,
       "step": 3592
     },
     {
-      "epoch": 6.875717017208413,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.4671856853185876e-05,
-      "loss": 0.371,
+      "epoch": 5.790660225442834,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.1988242788875532e-05,
+      "loss": 0.1735,
       "step": 3596
     },
     {
-      "epoch": 6.8833652007648185,
-      "grad_norm": 0.384765625,
-      "learning_rate": 1.4476182536210207e-05,
-      "loss": 0.3448,
+      "epoch": 5.797101449275362,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.1760457739030695e-05,
+      "loss": 0.1755,
       "step": 3600
     },
     {
-      "epoch": 6.891013384321224,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.4281755645350873e-05,
-      "loss": 0.3605,
+      "epoch": 5.8035426731078905,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.1533766390380254e-05,
+      "loss": 0.1674,
       "step": 3604
     },
     {
-      "epoch": 6.898661567877629,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.4088577970206044e-05,
-      "loss": 0.3385,
+      "epoch": 5.809983896940419,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.1308170676279547e-05,
+      "loss": 0.138,
       "step": 3608
     },
     {
-      "epoch": 6.906309751434034,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.3896651288875572e-05,
-      "loss": 0.3285,
+      "epoch": 5.816425120772947,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.108367252073961e-05,
+      "loss": 0.1521,
       "step": 3612
     },
     {
-      "epoch": 6.91395793499044,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.3705977367944498e-05,
-      "loss": 0.3316,
+      "epoch": 5.822866344605475,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.0860273838410928e-05,
+      "loss": 0.1771,
       "step": 3616
     },
     {
-      "epoch": 6.921606118546845,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.3516557962466978e-05,
-      "loss": 0.3636,
+      "epoch": 5.829307568438003,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.0637976534567046e-05,
+      "loss": 0.1628,
       "step": 3620
     },
     {
-      "epoch": 6.92925430210325,
-      "grad_norm": 0.380859375,
-      "learning_rate": 1.3328394815949884e-05,
-      "loss": 0.3649,
+      "epoch": 5.835748792270532,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.0416782505088347e-05,
+      "loss": 0.1631,
       "step": 3624
     },
     {
-      "epoch": 6.9369024856596555,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.3141489660336902e-05,
-      "loss": 0.3521,
+      "epoch": 5.84219001610306,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.0196693636445727e-05,
+      "loss": 0.1731,
       "step": 3628
     },
     {
-      "epoch": 6.944550669216062,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.295584421599265e-05,
-      "loss": 0.3177,
+      "epoch": 5.848631239935588,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9977711805684706e-05,
+      "loss": 0.1748,
       "step": 3632
     },
     {
-      "epoch": 6.952198852772467,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.2771460191686656e-05,
-      "loss": 0.3148,
+      "epoch": 5.855072463768116,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.975983888040945e-05,
+      "loss": 0.1872,
       "step": 3636
     },
     {
-      "epoch": 6.959847036328872,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.2588339284577815e-05,
-      "loss": 0.3324,
+      "epoch": 5.861513687600644,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9543076718766538e-05,
+      "loss": 0.1883,
       "step": 3640
     },
     {
-      "epoch": 6.9674952198852775,
-      "grad_norm": 0.37109375,
-      "learning_rate": 1.240648318019859e-05,
-      "loss": 0.3729,
+      "epoch": 5.867954911433173,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.932742716942946e-05,
+      "loss": 0.1543,
       "step": 3644
     },
     {
-      "epoch": 6.975143403441683,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.2225893552439742e-05,
-      "loss": 0.3616,
+      "epoch": 5.874396135265701,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.911289207158254e-05,
+      "loss": 0.1807,
       "step": 3648
     },
     {
-      "epoch": 6.982791586998088,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.2046572063534587e-05,
-      "loss": 0.3261,
+      "epoch": 5.880837359098229,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8899473254905672e-05,
+      "loss": 0.1775,
       "step": 3652
     },
     {
-      "epoch": 6.990439770554493,
-      "grad_norm": 0.384765625,
-      "learning_rate": 1.1868520364044049e-05,
-      "loss": 0.3542,
+      "epoch": 5.887278582930757,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8687172539558208e-05,
+      "loss": 0.1767,
       "step": 3656
     },
     {
-      "epoch": 6.998087954110899,
-      "grad_norm": 0.33203125,
-      "learning_rate": 1.1691740092841228e-05,
-      "loss": 0.3238,
+      "epoch": 5.8937198067632846,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8475991736163835e-05,
+      "loss": 0.1662,
       "step": 3660
     },
     {
-      "epoch": 7.005736137667304,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.151623287709636e-05,
-      "loss": 0.4048,
+      "epoch": 5.900161030595813,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8265932645794827e-05,
+      "loss": 0.1575,
       "step": 3664
     },
     {
-      "epoch": 7.013384321223709,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.1342000332261963e-05,
-      "loss": 0.3609,
+      "epoch": 5.906602254428341,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.805699705995708e-05,
+      "loss": 0.1778,
       "step": 3668
     },
     {
-      "epoch": 7.0210325047801145,
-      "grad_norm": 0.326171875,
-      "learning_rate": 1.1169044062057797e-05,
-      "loss": 0.3145,
+      "epoch": 5.913043478260869,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7849186760574346e-05,
+      "loss": 0.1661,
       "step": 3672
     },
     {
-      "epoch": 7.02868068833652,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.0997365658456164e-05,
-      "loss": 0.3618,
+      "epoch": 5.919484702093397,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7642503519973432e-05,
+      "loss": 0.1603,
       "step": 3676
     },
     {
-      "epoch": 7.036328871892925,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.082696670166736e-05,
-      "loss": 0.3298,
+      "epoch": 5.925925925925926,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7436949100868864e-05,
+      "loss": 0.1603,
       "step": 3680
     },
     {
-      "epoch": 7.04397705544933,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.0657848760124954e-05,
-      "loss": 0.3249,
+      "epoch": 5.932367149758454,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7232525256348013e-05,
+      "loss": 0.1907,
       "step": 3684
     },
     {
-      "epoch": 7.051625239005737,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.0490013390471474e-05,
-      "loss": 0.294,
+      "epoch": 5.938808373590982,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7029233729855883e-05,
+      "loss": 0.1848,
       "step": 3688
     },
     {
-      "epoch": 7.059273422562142,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.0323462137543998e-05,
-      "loss": 0.3264,
+      "epoch": 5.94524959742351,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6827076255180593e-05,
+      "loss": 0.1719,
       "step": 3692
     },
     {
-      "epoch": 7.066921606118547,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.015819653436012e-05,
-      "loss": 0.3164,
+      "epoch": 5.951690821256038,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6626054556438322e-05,
+      "loss": 0.1819,
       "step": 3696
     },
     {
-      "epoch": 7.074569789674952,
-      "grad_norm": 0.34375,
-      "learning_rate": 9.994218102103468e-06,
-      "loss": 0.3317,
+      "epoch": 5.958132045088567,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6426170348058703e-05,
+      "loss": 0.1669,
       "step": 3700
     },
     {
-      "epoch": 7.082217973231358,
-      "grad_norm": 0.37109375,
-      "learning_rate": 9.831528350110119e-06,
-      "loss": 0.3168,
+      "epoch": 5.964573268921095,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6227425334770245e-05,
+      "loss": 0.169,
       "step": 3704
     },
     {
-      "epoch": 7.089866156787763,
-      "grad_norm": 0.39453125,
-      "learning_rate": 9.670128775854513e-06,
-      "loss": 0.3901,
+      "epoch": 5.971014492753623,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6029821211585592e-05,
+      "loss": 0.1723,
       "step": 3708
     },
     {
-      "epoch": 7.097514340344168,
-      "grad_norm": 0.3359375,
-      "learning_rate": 9.51002086493553e-06,
-      "loss": 0.3297,
+      "epoch": 5.977455716586151,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5833359663787392e-05,
+      "loss": 0.2008,
       "step": 3712
     },
     {
-      "epoch": 7.105162523900574,
-      "grad_norm": 0.353515625,
-      "learning_rate": 9.351206091063107e-06,
-      "loss": 0.3484,
+      "epoch": 5.9838969404186795,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.563804236691364e-05,
+      "loss": 0.1523,
       "step": 3716
     },
     {
-      "epoch": 7.112810707456979,
-      "grad_norm": 0.34375,
-      "learning_rate": 9.193685916044469e-06,
-      "loss": 0.3255,
+      "epoch": 5.990338164251208,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.5443870986743562e-05,
+      "loss": 0.1592,
       "step": 3720
     },
     {
-      "epoch": 7.120458891013384,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.03746178977074e-06,
-      "loss": 0.3837,
+      "epoch": 5.996779388083736,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5250847179283243e-05,
+      "loss": 0.2154,
       "step": 3724
     },
     {
-      "epoch": 7.128107074569789,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.882535150203567e-06,
-      "loss": 0.3689,
+      "epoch": 6.003220611916264,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.505897259075171e-05,
+      "loss": 0.1917,
       "step": 3728
     },
     {
-      "epoch": 7.135755258126195,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.728907423361991e-06,
-      "loss": 0.3209,
+      "epoch": 6.009661835748792,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.4868248857566734e-05,
+      "loss": 0.1512,
       "step": 3732
     },
     {
-      "epoch": 7.1434034416826,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.576580023309126e-06,
-      "loss": 0.3297,
+      "epoch": 6.0161030595813205,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4678677606330964e-05,
+      "loss": 0.1889,
       "step": 3736
     },
     {
-      "epoch": 7.151051625239006,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.425554352139313e-06,
-      "loss": 0.3432,
+      "epoch": 6.022544283413849,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.4490260453817898e-05,
+      "loss": 0.1694,
       "step": 3740
     },
     {
-      "epoch": 7.1586998087954115,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.275831799965194e-06,
-      "loss": 0.3514,
+      "epoch": 6.028985507246377,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.4302999006958342e-05,
+      "loss": 0.1365,
       "step": 3744
     },
     {
-      "epoch": 7.166347992351817,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.127413744904804e-06,
-      "loss": 0.3326,
+      "epoch": 6.035426731078905,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.411689486282654e-05,
+      "loss": 0.148,
       "step": 3748
     },
     {
-      "epoch": 7.173996175908222,
-      "grad_norm": 0.373046875,
-      "learning_rate": 7.980301553068985e-06,
-      "loss": 0.3558,
+      "epoch": 6.041867954911433,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.393194960862657e-05,
+      "loss": 0.1744,
       "step": 3752
     },
     {
-      "epoch": 7.181644359464627,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.83449657854886e-06,
-      "loss": 0.3504,
+      "epoch": 6.048309178743962,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.3748164821678759e-05,
+      "loss": 0.1642,
       "step": 3756
     },
     {
-      "epoch": 7.189292543021033,
-      "grad_norm": 0.34375,
-      "learning_rate": 7.690000163403177e-06,
-      "loss": 0.3136,
+      "epoch": 6.05475040257649,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3565542069406433e-05,
+      "loss": 0.1826,
       "step": 3760
     },
     {
-      "epoch": 7.196940726577438,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.546813637646182e-06,
-      "loss": 0.3292,
+      "epoch": 6.061191626409018,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3384082909322375e-05,
+      "loss": 0.1911,
       "step": 3764
     },
     {
-      "epoch": 7.204588910133843,
-      "grad_norm": 0.365234375,
-      "learning_rate": 7.404938319235171e-06,
-      "loss": 0.327,
+      "epoch": 6.067632850241546,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.320378888901546e-05,
+      "loss": 0.134,
       "step": 3768
     },
     {
-      "epoch": 7.2122370936902485,
-      "grad_norm": 0.36328125,
-      "learning_rate": 7.264375514058607e-06,
-      "loss": 0.3526,
+      "epoch": 6.074074074074074,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3024661546137694e-05,
+      "loss": 0.1778,
       "step": 3772
     },
     {
-      "epoch": 7.219885277246654,
-      "grad_norm": 0.333984375,
-      "learning_rate": 7.125126515923752e-06,
-      "loss": 0.3506,
+      "epoch": 6.080515297906603,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.2846702408390975e-05,
+      "loss": 0.1542,
       "step": 3776
     },
     {
-      "epoch": 7.227533460803059,
-      "grad_norm": 0.349609375,
-      "learning_rate": 6.987192606545156e-06,
-      "loss": 0.3355,
+      "epoch": 6.086956521739131,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.2669912993514036e-05,
+      "loss": 0.185,
       "step": 3780
     },
     {
-      "epoch": 7.235181644359464,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.850575055532553e-06,
-      "loss": 0.3484,
+      "epoch": 6.093397745571659,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2494294809269512e-05,
+      "loss": 0.1937,
       "step": 3784
     },
     {
-      "epoch": 7.24282982791587,
-      "grad_norm": 0.36328125,
-      "learning_rate": 6.715275120379271e-06,
-      "loss": 0.366,
+      "epoch": 6.099838969404187,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.2319849353431154e-05,
+      "loss": 0.1642,
       "step": 3788
     },
     {
-      "epoch": 7.250478011472275,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.581294046450753e-06,
-      "loss": 0.322,
+      "epoch": 6.106280193236715,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.2146578113771005e-05,
+      "loss": 0.156,
       "step": 3792
     },
     {
-      "epoch": 7.258126195028681,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.448633066972953e-06,
-      "loss": 0.2989,
+      "epoch": 6.112721417069243,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.1974482568046694e-05,
+      "loss": 0.172,
       "step": 3796
     },
     {
-      "epoch": 7.265774378585086,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.317293403021029e-06,
-      "loss": 0.3377,
+      "epoch": 6.119162640901771,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.1803564183988812e-05,
+      "loss": 0.1655,
       "step": 3800
     },
     {
-      "epoch": 7.273422562141492,
-      "grad_norm": 0.326171875,
-      "learning_rate": 6.187276263508167e-06,
-      "loss": 0.3001,
+      "epoch": 6.125603864734299,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.1633824419288474e-05,
+      "loss": 0.1741,
       "step": 3804
     },
     {
-      "epoch": 7.281070745697897,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.0585828451743925e-06,
-      "loss": 0.3335,
+      "epoch": 6.132045088566827,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.146526472158487e-05,
+      "loss": 0.1805,
       "step": 3808
     },
     {
-      "epoch": 7.288718929254302,
-      "grad_norm": 0.34375,
-      "learning_rate": 5.93121433257554e-06,
-      "loss": 0.3458,
+      "epoch": 6.138486312399356,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.1297886528452882e-05,
+      "loss": 0.1617,
       "step": 3812
     },
     {
-      "epoch": 7.2963671128107075,
-      "grad_norm": 0.3671875,
-      "learning_rate": 5.805171898072369e-06,
-      "loss": 0.2941,
+      "epoch": 6.144927536231884,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.1131691267390757e-05,
+      "loss": 0.1863,
       "step": 3816
     },
     {
-      "epoch": 7.304015296367113,
-      "grad_norm": 0.359375,
-      "learning_rate": 5.680456701819885e-06,
-      "loss": 0.3888,
+      "epoch": 6.151368760064412,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0966680355808122e-05,
+      "loss": 0.2013,
       "step": 3820
     },
     {
-      "epoch": 7.311663479923518,
-      "grad_norm": 0.36328125,
-      "learning_rate": 5.5570698917563994e-06,
-      "loss": 0.3663,
+      "epoch": 6.15780998389694,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.080285520101371e-05,
+      "loss": 0.1683,
       "step": 3824
     },
     {
-      "epoch": 7.319311663479923,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.435012603593219e-06,
-      "loss": 0.292,
+      "epoch": 6.164251207729468,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.0640217200203466e-05,
+      "loss": 0.1729,
       "step": 3828
     },
     {
-      "epoch": 7.326959847036329,
-      "grad_norm": 0.373046875,
-      "learning_rate": 5.3142859608041265e-06,
-      "loss": 0.3628,
+      "epoch": 6.170692431561997,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.047876774044863e-05,
+      "loss": 0.1736,
       "step": 3832
     },
     {
-      "epoch": 7.334608030592734,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.194891074614899e-06,
-      "loss": 0.3377,
+      "epoch": 6.177133655394525,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.0318508198683734e-05,
+      "loss": 0.1757,
       "step": 3836
     },
     {
-      "epoch": 7.342256214149139,
-      "grad_norm": 0.353515625,
-      "learning_rate": 5.076829043993253e-06,
-      "loss": 0.3322,
+      "epoch": 6.183574879227053,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.015943994169523e-05,
+      "loss": 0.1824,
       "step": 3840
     },
     {
-      "epoch": 7.349904397705545,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.960100955638685e-06,
-      "loss": 0.3607,
+      "epoch": 6.190016103059581,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.0001564326109363e-05,
+      "loss": 0.1536,
       "step": 3844
     },
     {
-      "epoch": 7.357552581261951,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.844707883972398e-06,
-      "loss": 0.3303,
+      "epoch": 6.1964573268921095,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.844882698381013e-06,
+      "loss": 0.2085,
       "step": 3848
     },
     {
-      "epoch": 7.365200764818356,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.730650891127457e-06,
-      "loss": 0.3327,
+      "epoch": 6.202898550724638,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.689396394781923e-06,
+      "loss": 0.1665,
       "step": 3852
     },
     {
-      "epoch": 7.372848948374761,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.617931026939115e-06,
-      "loss": 0.3766,
+      "epoch": 6.209339774557166,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.535106741389542e-06,
+      "loss": 0.1714,
       "step": 3856
     },
     {
-      "epoch": 7.3804971319311665,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.506549328934916e-06,
-      "loss": 0.2884,
+      "epoch": 6.215780998389694,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.382015054075465e-06,
+      "loss": 0.1639,
       "step": 3860
     },
     {
-      "epoch": 7.388145315487572,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.39650682232538e-06,
-      "loss": 0.3036,
+      "epoch": 6.222222222222222,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.230122638494408e-06,
+      "loss": 0.1391,
       "step": 3864
     },
     {
-      "epoch": 7.395793499043977,
-      "grad_norm": 0.37890625,
-      "learning_rate": 4.287804519994431e-06,
-      "loss": 0.3398,
+      "epoch": 6.2286634460547505,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.079430790072972e-06,
+      "loss": 0.1514,
       "step": 3868
     },
     {
-      "epoch": 7.403441682600382,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.180443422490115e-06,
-      "loss": 0.3276,
+      "epoch": 6.235104669887279,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.92994079399868e-06,
+      "loss": 0.1324,
       "step": 3872
     },
     {
-      "epoch": 7.411089866156788,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.074424518015384e-06,
-      "loss": 0.3319,
+      "epoch": 6.241545893719807,
+      "grad_norm": 0.306640625,
+      "learning_rate": 8.781653925208887e-06,
+      "loss": 0.1629,
       "step": 3876
     },
     {
-      "epoch": 7.418738049713193,
-      "grad_norm": 0.390625,
-      "learning_rate": 3.969748782418991e-06,
-      "loss": 0.3925,
+      "epoch": 6.247987117552335,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.634571448380056e-06,
+      "loss": 0.1899,
       "step": 3880
     },
     {
-      "epoch": 7.426386233269598,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.8664171791865765e-06,
-      "loss": 0.3556,
+      "epoch": 6.254428341384863,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.488694617916785e-06,
+      "loss": 0.1696,
       "step": 3884
     },
     {
-      "epoch": 7.4340344168260035,
-      "grad_norm": 0.384765625,
-      "learning_rate": 3.764430659431661e-06,
-      "loss": 0.324,
+      "epoch": 6.260869565217392,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.344024677941346e-06,
+      "loss": 0.1652,
       "step": 3888
     },
     {
-      "epoch": 7.441682600382409,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.6637901618870203e-06,
-      "loss": 0.3443,
+      "epoch": 6.26731078904992,
+      "grad_norm": 0.28515625,
+      "learning_rate": 8.200562862282912e-06,
+      "loss": 0.1642,
       "step": 3892
     },
     {
-      "epoch": 7.449330783938814,
-      "grad_norm": 0.3828125,
-      "learning_rate": 3.564496612896006e-06,
-      "loss": 0.3409,
+      "epoch": 6.273752012882448,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.058310394466994e-06,
+      "loss": 0.1458,
       "step": 3896
     },
     {
-      "epoch": 7.45697896749522,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.4665509264039717e-06,
-      "loss": 0.3106,
+      "epoch": 6.280193236714976,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.917268487705175e-06,
+      "loss": 0.1519,
       "step": 3900
     },
     {
-      "epoch": 7.4646271510516256,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.3699540039499263e-06,
-      "loss": 0.3442,
+      "epoch": 6.286634460547504,
+      "grad_norm": 0.287109375,
+      "learning_rate": 7.777438344884645e-06,
+      "loss": 0.1745,
       "step": 3904
     },
     {
-      "epoch": 7.472275334608031,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.274706734658228e-06,
-      "loss": 0.3649,
+      "epoch": 6.293075684380033,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.638821158557962e-06,
+      "loss": 0.1696,
       "step": 3908
     },
     {
-      "epoch": 7.479923518164436,
-      "grad_norm": 0.330078125,
-      "learning_rate": 3.1808099952303045e-06,
-      "loss": 0.3215,
+      "epoch": 6.29951690821256,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.501418110932872e-06,
+      "loss": 0.1634,
       "step": 3912
     },
     {
-      "epoch": 7.487571701720841,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.0882646499367614e-06,
-      "loss": 0.2973,
+      "epoch": 6.305958132045088,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.365230373862274e-06,
+      "loss": 0.1589,
       "step": 3916
     },
     {
-      "epoch": 7.495219885277247,
-      "grad_norm": 0.375,
-      "learning_rate": 2.9970715506092534e-06,
-      "loss": 0.3862,
+      "epoch": 6.312399355877616,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.2302591088341576e-06,
+      "loss": 0.1675,
       "step": 3920
     },
     {
-      "epoch": 7.502868068833652,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.9072315366327424e-06,
-      "loss": 0.3246,
+      "epoch": 6.318840579710145,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.096505466961794e-06,
+      "loss": 0.1718,
       "step": 3924
     },
     {
-      "epoch": 7.510516252390057,
-      "grad_norm": 0.36328125,
-      "learning_rate": 2.8187454349377193e-06,
-      "loss": 0.3823,
+      "epoch": 6.325281803542673,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.963970588973761e-06,
+      "loss": 0.1412,
       "step": 3928
     },
     {
-      "epoch": 7.5181644359464626,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.731614059992676e-06,
-      "loss": 0.3697,
+      "epoch": 6.331723027375201,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.832655605204401e-06,
+      "loss": 0.165,
       "step": 3932
     },
     {
-      "epoch": 7.525812619502868,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.6458382137964805e-06,
-      "loss": 0.3312,
+      "epoch": 6.338164251207729,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.702561635584047e-06,
+      "loss": 0.1496,
       "step": 3936
     },
     {
-      "epoch": 7.533460803059273,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.5614186858711137e-06,
-      "loss": 0.3188,
+      "epoch": 6.344605475040257,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.57368978962956e-06,
+      "loss": 0.1829,
       "step": 3940
     },
     {
-      "epoch": 7.541108986615678,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.47835625325436e-06,
-      "loss": 0.322,
+      "epoch": 6.351046698872786,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.44604116643474e-06,
+      "loss": 0.182,
       "step": 3944
     },
     {
-      "epoch": 7.548757170172085,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.3966516804925784e-06,
-      "loss": 0.3401,
+      "epoch": 6.357487922705314,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.3196168546610634e-06,
+      "loss": 0.1714,
       "step": 3948
     },
     {
-      "epoch": 7.55640535372849,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.3163057196338096e-06,
-      "loss": 0.3388,
+      "epoch": 6.363929146537842,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.194417932528478e-06,
+      "loss": 0.2102,
       "step": 3952
     },
     {
-      "epoch": 7.564053537284895,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.2373191102207646e-06,
-      "loss": 0.3375,
+      "epoch": 6.37037037037037,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.070445467805923e-06,
+      "loss": 0.1761,
       "step": 3956
     },
     {
-      "epoch": 7.5717017208413,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.1596925792839946e-06,
-      "loss": 0.3125,
+      "epoch": 6.3768115942028984,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.947700517802523e-06,
+      "loss": 0.1816,
       "step": 3960
     },
     {
-      "epoch": 7.579349904397706,
-      "grad_norm": 0.37890625,
-      "learning_rate": 2.083426841335284e-06,
-      "loss": 0.3436,
+      "epoch": 6.383252818035427,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.826184129358358e-06,
+      "loss": 0.2008,
       "step": 3964
     },
     {
-      "epoch": 7.586998087954111,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.0085225983610177e-06,
-      "loss": 0.3559,
+      "epoch": 6.389694041867955,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.705897338835724e-06,
+      "loss": 0.1857,
       "step": 3968
     },
     {
-      "epoch": 7.594646271510516,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.9349805398156893e-06,
-      "loss": 0.3167,
+      "epoch": 6.396135265700483,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.58684117211009e-06,
+      "loss": 0.1681,
       "step": 3972
     },
     {
-      "epoch": 7.602294455066922,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.8628013426156386e-06,
-      "loss": 0.3654,
+      "epoch": 6.402576489533011,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.469016644561519e-06,
+      "loss": 0.1409,
       "step": 3976
     },
     {
-      "epoch": 7.609942638623327,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.7919856711327563e-06,
-      "loss": 0.3534,
+      "epoch": 6.4090177133655395,
+      "grad_norm": 0.3046875,
+      "learning_rate": 5.352424761065926e-06,
+      "loss": 0.1647,
       "step": 3980
     },
     {
-      "epoch": 7.617590822179732,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.722534177188406e-06,
-      "loss": 0.3139,
+      "epoch": 6.415458937198068,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.2370665159865045e-06,
+      "loss": 0.1556,
       "step": 3984
     },
     {
-      "epoch": 7.625239005736137,
-      "grad_norm": 0.37109375,
-      "learning_rate": 1.6544475000473957e-06,
-      "loss": 0.3578,
+      "epoch": 6.421900161030596,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.1229428931652775e-06,
+      "loss": 0.1912,
       "step": 3988
     },
     {
-      "epoch": 7.632887189292543,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.5877262664120983e-06,
-      "loss": 0.3272,
+      "epoch": 6.428341384863124,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.010054865914676e-06,
+      "loss": 0.1521,
       "step": 3992
     },
     {
-      "epoch": 7.640535372848948,
-      "grad_norm": 0.373046875,
-      "learning_rate": 1.522371090416724e-06,
-      "loss": 0.3578,
+      "epoch": 6.434782608695652,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.898403397009293e-06,
+      "loss": 0.192,
       "step": 3996
     },
     {
-      "epoch": 7.648183556405353,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.4583825736215749e-06,
-      "loss": 0.3131,
+      "epoch": 6.4412238325281805,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.787989438677625e-06,
+      "loss": 0.1464,
       "step": 4000
     }
   ],
   "logging_steps": 4,
-  "max_steps": 4184,
+  "max_steps": 4347,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 8,
+  "num_train_epochs": 7,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -7026,7 +7026,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.6138957704404664e+18,
+  "total_flos": 2.6355937321911583e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null