diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,6430 +1,9641 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 7.0,
+  "epoch": 6.447831184056272,
   "eval_steps": 500,
-  "global_step": 3661,
+  "global_step": 5500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0076481835564053535,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.243243243243243e-05,
-      "loss": 1.3011,
+      "epoch": 0.004689331770222743,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 1.2947,
       "step": 4
     },
     {
-      "epoch": 0.015296367112810707,
-      "grad_norm": 0.26171875,
-      "learning_rate": 6.486486486486486e-05,
-      "loss": 1.3104,
+      "epoch": 0.009378663540445486,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 1.2923,
       "step": 8
     },
     {
-      "epoch": 0.022944550669216062,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 9.72972972972973e-05,
-      "loss": 1.2309,
+      "epoch": 0.01406799531066823,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 1.227,
       "step": 12
     },
     {
-      "epoch": 0.030592734225621414,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00012972972972972972,
-      "loss": 1.2051,
+      "epoch": 0.01875732708089097,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 1.2031,
       "step": 16
     },
     {
-      "epoch": 0.03824091778202677,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.00016216216216216215,
-      "loss": 1.1622,
+      "epoch": 0.023446658851113716,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 1.1748,
       "step": 20
     },
     {
-      "epoch": 0.045889101338432124,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0001945945945945946,
-      "loss": 1.1749,
+      "epoch": 0.02813599062133646,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.1424,
       "step": 24
     },
     {
-      "epoch": 0.05353728489483748,
-      "grad_norm": 0.154296875,
-      "learning_rate": 0.00022702702702702703,
-      "loss": 1.149,
+      "epoch": 0.032825322391559206,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014,
+      "loss": 1.1835,
       "step": 28
     },
     {
-      "epoch": 0.06118546845124283,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00025945945945945944,
-      "loss": 1.1455,
+      "epoch": 0.03751465416178194,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 1.1185,
       "step": 32
     },
     {
-      "epoch": 0.06883365200764818,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0002918918918918919,
-      "loss": 1.1358,
+      "epoch": 0.04220398593200469,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 1.139,
       "step": 36
     },
     {
-      "epoch": 0.07648183556405354,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00029999949274434724,
-      "loss": 1.1201,
+      "epoch": 0.04689331770222743,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 1.1278,
       "step": 40
     },
     {
-      "epoch": 0.0841300191204589,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.0002999972382816974,
-      "loss": 1.0549,
+      "epoch": 0.05158264947245018,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 1.0939,
       "step": 44
     },
     {
-      "epoch": 0.09177820267686425,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002999931802773903,
-      "loss": 1.0946,
+      "epoch": 0.05627198124267292,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.0765,
       "step": 48
     },
     {
-      "epoch": 0.0994263862332696,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00029998731878021884,
-      "loss": 1.0841,
+      "epoch": 0.06096131301289566,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00026,
+      "loss": 1.1053,
       "step": 52
     },
     {
-      "epoch": 0.10707456978967496,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00029997965386066057,
-      "loss": 1.0904,
+      "epoch": 0.06565064478311841,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00028,
+      "loss": 1.0675,
       "step": 56
     },
     {
-      "epoch": 0.1147227533460803,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0002999701856108772,
-      "loss": 1.0673,
+      "epoch": 0.07033997655334115,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003,
+      "loss": 1.0904,
       "step": 60
     },
     {
-      "epoch": 0.12237093690248566,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00029995891414471334,
-      "loss": 1.0211,
+      "epoch": 0.07502930832356389,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00029999966103183746,
+      "loss": 1.0871,
       "step": 64
     },
     {
-      "epoch": 0.13001912045889102,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0002999458395976953,
-      "loss": 1.0497,
+      "epoch": 0.07971864009378664,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002999986441288818,
+      "loss": 1.0505,
       "step": 68
     },
     {
-      "epoch": 0.13766730401529637,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002999309621270293,
-      "loss": 1.0711,
+      "epoch": 0.08440797186400938,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.000299996949295729,
+      "loss": 1.0788,
       "step": 72
     },
     {
-      "epoch": 0.14531548757170173,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00029991428191159935,
-      "loss": 1.047,
+      "epoch": 0.08909730363423213,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002999945765400391,
+      "loss": 1.0503,
       "step": 76
     },
     {
-      "epoch": 0.15296367112810708,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00029989579915196574,
-      "loss": 1.0442,
+      "epoch": 0.09378663540445487,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00029999152587253583,
+      "loss": 1.0564,
       "step": 80
     },
     {
-      "epoch": 0.16061185468451242,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.000299875514070362,
-      "loss": 1.0197,
+      "epoch": 0.0984759671746776,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002999877973070069,
+      "loss": 1.0396,
       "step": 84
     },
     {
-      "epoch": 0.1682600382409178,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00029985342691069255,
-      "loss": 1.0185,
+      "epoch": 0.10316529894490035,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00029998339086030395,
+      "loss": 1.0208,
       "step": 88
     },
     {
-      "epoch": 0.17590822179732313,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0002998295379385297,
-      "loss": 1.0168,
+      "epoch": 0.10785463071512309,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00029997830655234217,
+      "loss": 1.057,
       "step": 92
     },
     {
-      "epoch": 0.1835564053537285,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00029980384744111047,
-      "loss": 1.0211,
+      "epoch": 0.11254396248534584,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002999725444061004,
+      "loss": 1.0283,
       "step": 96
     },
     {
-      "epoch": 0.19120458891013384,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002997763557273331,
-      "loss": 1.0178,
+      "epoch": 0.11723329425556858,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002999661044476212,
+      "loss": 0.988,
       "step": 100
     },
     {
-      "epoch": 0.1988527724665392,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.0002997470631277533,
-      "loss": 0.9871,
+      "epoch": 0.12192262602579132,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00029995898670601053,
+      "loss": 1.049,
       "step": 104
     },
     {
-      "epoch": 0.20650095602294455,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002997159699945804,
-      "loss": 1.0197,
+      "epoch": 0.12661195779601406,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002999511912134374,
+      "loss": 0.9804,
       "step": 108
     },
     {
-      "epoch": 0.21414913957934992,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002996830767016731,
-      "loss": 1.0,
+      "epoch": 0.13130128956623682,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002999427180051341,
+      "loss": 1.0172,
       "step": 112
     },
     {
-      "epoch": 0.22179732313575526,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.0002996483836445347,
-      "loss": 0.9885,
+      "epoch": 0.13599062133645956,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00029993356711939615,
+      "loss": 0.9863,
       "step": 116
     },
     {
-      "epoch": 0.2294455066921606,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00029961189124030885,
-      "loss": 0.9664,
+      "epoch": 0.1406799531066823,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002999237385975815,
+      "loss": 0.9799,
       "step": 120
     },
     {
-      "epoch": 0.23709369024856597,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029957359992777404,
-      "loss": 0.9831,
+      "epoch": 0.14536928487690504,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0002999132324841109,
+      "loss": 1.0098,
       "step": 124
     },
     {
-      "epoch": 0.2447418738049713,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029953351016733854,
-      "loss": 0.9718,
+      "epoch": 0.15005861664712777,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0002999020488264676,
+      "loss": 0.9597,
       "step": 128
     },
     {
-      "epoch": 0.25239005736137665,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.000299491622441035,
-      "loss": 0.9592,
+      "epoch": 0.15474794841735054,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002998901876751969,
+      "loss": 0.9958,
       "step": 132
     },
     {
-      "epoch": 0.26003824091778205,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00029944793725251436,
-      "loss": 0.9514,
+      "epoch": 0.15943728018757328,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029987764908390624,
+      "loss": 0.947,
       "step": 136
     },
     {
-      "epoch": 0.2676864244741874,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00029940245512704005,
-      "loss": 0.9435,
+      "epoch": 0.16412661195779601,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002998644331092647,
+      "loss": 0.948,
       "step": 140
     },
     {
-      "epoch": 0.27533460803059273,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0002993551766114815,
-      "loss": 0.9545,
+      "epoch": 0.16881594372801875,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00029985053981100286,
+      "loss": 0.9685,
       "step": 144
     },
     {
-      "epoch": 0.2829827915869981,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00029930610227430767,
-      "loss": 0.9378,
+      "epoch": 0.1735052754982415,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00029983596925191265,
+      "loss": 0.975,
       "step": 148
     },
     {
-      "epoch": 0.29063097514340347,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002992552327055802,
-      "loss": 0.9085,
+      "epoch": 0.17819460726846426,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0002998207214978466,
+      "loss": 0.978,
       "step": 152
     },
     {
-      "epoch": 0.2982791586998088,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002992025685169461,
-      "loss": 0.9482,
+      "epoch": 0.182883939038687,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0002998047966177182,
+      "loss": 0.9702,
       "step": 156
     },
     {
-      "epoch": 0.30592734225621415,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00029914811034163096,
-      "loss": 0.949,
+      "epoch": 0.18757327080890973,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0002997881946835012,
+      "loss": 0.9399,
       "step": 160
     },
     {
-      "epoch": 0.3135755258126195,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029909185883443063,
-      "loss": 0.9796,
+      "epoch": 0.19226260257913247,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029977091577022916,
+      "loss": 0.9403,
       "step": 164
     },
     {
-      "epoch": 0.32122370936902483,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.0002990338146717039,
-      "loss": 0.9351,
+      "epoch": 0.1969519343493552,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002997529599559956,
+      "loss": 0.9342,
       "step": 168
     },
     {
-      "epoch": 0.32887189292543023,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.0002989739785513639,
-      "loss": 0.9166,
+      "epoch": 0.20164126611957797,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00029973432732195303,
+      "loss": 0.9186,
       "step": 172
     },
     {
-      "epoch": 0.3365200764818356,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0002989123511928703,
-      "loss": 0.9204,
+      "epoch": 0.2063305978898007,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002997150179523131,
+      "loss": 0.9377,
       "step": 176
     },
     {
-      "epoch": 0.3441682600382409,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00029884893333722,
-      "loss": 0.9356,
+      "epoch": 0.21101992966002345,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029969503193434606,
+      "loss": 0.9365,
       "step": 180
     },
     {
-      "epoch": 0.35181644359464626,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002987837257469387,
-      "loss": 0.9342,
+      "epoch": 0.21570926143024619,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00029967436935838,
+      "loss": 0.921,
       "step": 184
     },
     {
-      "epoch": 0.35946462715105165,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00029871672920607153,
-      "loss": 0.9026,
+      "epoch": 0.22039859320046892,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00029965303031780126,
+      "loss": 0.9041,
       "step": 188
     },
     {
-      "epoch": 0.367112810707457,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002986479445201737,
-      "loss": 0.8983,
+      "epoch": 0.2250879249706917,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00029963101490905307,
+      "loss": 0.9319,
       "step": 192
     },
     {
-      "epoch": 0.37476099426386233,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002985773725163008,
-      "loss": 0.922,
+      "epoch": 0.22977725674091443,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002996083232316358,
+      "loss": 0.9135,
       "step": 196
     },
     {
-      "epoch": 0.3824091778202677,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002985050140429986,
-      "loss": 0.9099,
+      "epoch": 0.23446658851113716,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002995849553881061,
+      "loss": 0.8845,
       "step": 200
     },
     {
-      "epoch": 0.390057361376673,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002984308699702935,
-      "loss": 0.8825,
+      "epoch": 0.2391559202813599,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029956091148407684,
+      "loss": 0.8891,
       "step": 204
     },
     {
-      "epoch": 0.3977055449330784,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.0002983549411896812,
-      "loss": 0.893,
+      "epoch": 0.24384525205158264,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00029953619162821616,
+      "loss": 0.8917,
       "step": 208
     },
     {
-      "epoch": 0.40535372848948376,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002982772286141167,
-      "loss": 0.9068,
+      "epoch": 0.2485345838218054,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002995107959322474,
+      "loss": 0.9432,
       "step": 212
     },
     {
-      "epoch": 0.4130019120458891,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.000298197733178003,
-      "loss": 0.8536,
+      "epoch": 0.2532239155920281,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00029948472451094823,
+      "loss": 0.9197,
       "step": 216
     },
     {
-      "epoch": 0.42065009560229444,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00029811645583717987,
-      "loss": 0.8843,
+      "epoch": 0.25791324736225085,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002994579774821505,
+      "loss": 0.9129,
       "step": 220
     },
     {
-      "epoch": 0.42829827915869984,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00029803339756891254,
-      "loss": 0.8627,
+      "epoch": 0.26260257913247365,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002994305549667394,
+      "loss": 0.8779,
       "step": 224
     },
     {
-      "epoch": 0.4359464627151052,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00029794855937187963,
-      "loss": 0.8572,
+      "epoch": 0.2672919109026964,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.000299402457088653,
+      "loss": 0.9172,
       "step": 228
     },
     {
-      "epoch": 0.4435946462715105,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002978619422661613,
-      "loss": 0.8255,
+      "epoch": 0.2719812426729191,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002993736839748818,
+      "loss": 0.8779,
       "step": 232
     },
     {
-      "epoch": 0.45124282982791586,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0002977735472932273,
-      "loss": 0.8274,
+      "epoch": 0.27667057444314186,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002993442357554681,
+      "loss": 0.9188,
       "step": 236
     },
     {
-      "epoch": 0.4588910133843212,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00029768337551592394,
-      "loss": 0.8569,
+      "epoch": 0.2813599062133646,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00029931411256350535,
+      "loss": 0.8902,
       "step": 240
     },
     {
-      "epoch": 0.4665391969407266,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00029759142801846143,
-      "loss": 0.8791,
+      "epoch": 0.28604923798358733,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002992833145351376,
+      "loss": 0.8734,
       "step": 244
     },
     {
-      "epoch": 0.47418738049713194,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00029749770590640123,
-      "loss": 0.8215,
+      "epoch": 0.29073856975381007,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002992518418095588,
+      "loss": 0.9317,
       "step": 248
     },
     {
-      "epoch": 0.4818355640535373,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029740221030664216,
-      "loss": 0.8546,
+      "epoch": 0.2954279015240328,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00029921969452901235,
+      "loss": 0.8715,
       "step": 252
     },
     {
-      "epoch": 0.4894837476099426,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029730494236740744,
-      "loss": 0.8518,
+      "epoch": 0.30011723329425555,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002991868728387903,
+      "loss": 0.8825,
       "step": 256
     },
     {
-      "epoch": 0.497131931166348,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002972059032582304,
-      "loss": 0.8115,
+      "epoch": 0.3048065650644783,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00029915337688723277,
+      "loss": 0.926,
       "step": 260
     },
     {
-      "epoch": 0.5047801147227533,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002971050941699407,
-      "loss": 0.8818,
+      "epoch": 0.3094958968347011,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00029911920682572726,
+      "loss": 0.8774,
       "step": 264
     },
     {
-      "epoch": 0.5124282982791587,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00029700251631464993,
-      "loss": 0.8834,
+      "epoch": 0.3141852286049238,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002990843628087079,
+      "loss": 0.8844,
       "step": 268
     },
     {
-      "epoch": 0.5200764818355641,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.000296898170925737,
-      "loss": 0.8502,
+      "epoch": 0.31887456037514655,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002990488449936549,
+      "loss": 0.8683,
       "step": 272
     },
     {
-      "epoch": 0.5277246653919694,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.0002967920592578335,
-      "loss": 0.854,
+      "epoch": 0.3235638921453693,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00029901265354109367,
+      "loss": 0.796,
       "step": 276
     },
     {
-      "epoch": 0.5353728489483748,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002966841825868082,
-      "loss": 0.8164,
+      "epoch": 0.32825322391559203,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002989757886145942,
+      "loss": 0.8603,
       "step": 280
     },
     {
-      "epoch": 0.5430210325047801,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00029657454220975216,
-      "loss": 0.9058,
+      "epoch": 0.33294255568581477,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002989382503807704,
+      "loss": 0.9065,
       "step": 284
     },
     {
-      "epoch": 0.5506692160611855,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00029646313944496297,
-      "loss": 0.8646,
+      "epoch": 0.3376318874560375,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00029890003900927904,
+      "loss": 0.8713,
       "step": 288
     },
     {
-      "epoch": 0.5583173996175909,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00029634997563192866,
-      "loss": 0.8536,
+      "epoch": 0.34232121922626024,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002988611546728194,
+      "loss": 0.8532,
       "step": 292
     },
     {
-      "epoch": 0.5659655831739961,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002962350521313122,
-      "loss": 0.8532,
+      "epoch": 0.347010550996483,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000298821597547132,
+      "loss": 0.8241,
       "step": 296
     },
     {
-      "epoch": 0.5736137667304015,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002961183703249342,
-      "loss": 0.8228,
+      "epoch": 0.3516998827667057,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002987813678109985,
+      "loss": 0.8575,
       "step": 300
     },
     {
-      "epoch": 0.5812619502868069,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002959999316157573,
-      "loss": 0.8088,
+      "epoch": 0.3563892145369285,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00029874046564624,
+      "loss": 0.8889,
       "step": 304
     },
     {
-      "epoch": 0.5889101338432122,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00029587973742786875,
-      "loss": 0.8291,
+      "epoch": 0.36107854630715125,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002986988912377171,
+      "loss": 0.8393,
       "step": 308
     },
     {
-      "epoch": 0.5965583173996176,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002957577892064632,
-      "loss": 0.8087,
+      "epoch": 0.365767878077374,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00029865664477332843,
+      "loss": 0.8675,
       "step": 312
     },
     {
-      "epoch": 0.6042065009560229,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00029563408841782576,
-      "loss": 0.853,
+      "epoch": 0.3704572098475967,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00029861372644401,
+      "loss": 0.8316,
       "step": 316
     },
     {
-      "epoch": 0.6118546845124283,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00029550863654931385,
-      "loss": 0.847,
+      "epoch": 0.37514654161781946,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0002985701364437345,
+      "loss": 0.8705,
       "step": 320
     },
     {
-      "epoch": 0.6195028680688337,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002953814351093398,
-      "loss": 0.8087,
+      "epoch": 0.3798358733880422,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002985258749695102,
+      "loss": 0.8752,
       "step": 324
     },
     {
-      "epoch": 0.627151051625239,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002952524856273524,
-      "loss": 0.8514,
+      "epoch": 0.38452520515826494,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029848094222138024,
+      "loss": 0.8594,
       "step": 328
     },
     {
-      "epoch": 0.6347992351816444,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00029512178965381854,
-      "loss": 0.8501,
+      "epoch": 0.3892145369284877,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002984353384024215,
+      "loss": 0.8335,
       "step": 332
     },
     {
-      "epoch": 0.6424474187380497,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00029498934876020475,
-      "loss": 0.8029,
+      "epoch": 0.3939038686987104,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002983890637187439,
+      "loss": 0.7962,
       "step": 336
     },
     {
-      "epoch": 0.6500956022944551,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029485516453895826,
-      "loss": 0.8293,
+      "epoch": 0.39859320046893315,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0002983421183794893,
+      "loss": 0.8526,
       "step": 340
     },
     {
-      "epoch": 0.6577437858508605,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002947192386034874,
-      "loss": 0.8695,
+      "epoch": 0.40328253223915594,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029829450259683085,
+      "loss": 0.8273,
       "step": 344
     },
     {
-      "epoch": 0.6653919694072657,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029458157258814316,
-      "loss": 0.8249,
+      "epoch": 0.4079718640093787,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029824621658597165,
+      "loss": 0.8174,
       "step": 348
     },
     {
-      "epoch": 0.6730401529636711,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.00029444216814819834,
-      "loss": 0.8009,
+      "epoch": 0.4126611957796014,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00029819726056514383,
+      "loss": 0.836,
       "step": 352
     },
     {
-      "epoch": 0.6806883365200764,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00029430102695982875,
-      "loss": 0.8642,
+      "epoch": 0.41735052754982416,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00029814763475560796,
+      "loss": 0.814,
       "step": 356
     },
     {
-      "epoch": 0.6883365200764818,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00029415815072009237,
-      "loss": 0.8562,
+      "epoch": 0.4220398593200469,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029809733938165157,
+      "loss": 0.8317,
       "step": 360
     },
     {
-      "epoch": 0.6959847036328872,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029401354114690905,
-      "loss": 0.8274,
+      "epoch": 0.42672919109026963,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002980463746705884,
+      "loss": 0.8571,
       "step": 364
     },
     {
-      "epoch": 0.7036328871892925,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002938671999790402,
-      "loss": 0.8214,
+      "epoch": 0.43141852286049237,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00029799474085275734,
+      "loss": 0.8309,
       "step": 368
     },
     {
-      "epoch": 0.7112810707456979,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00029371912897606736,
-      "loss": 0.8537,
+      "epoch": 0.4361078546307151,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029794243816152127,
+      "loss": 0.8289,
       "step": 372
     },
     {
-      "epoch": 0.7189292543021033,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00029356932991837163,
-      "loss": 0.8378,
+      "epoch": 0.44079718640093785,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002978894668332663,
+      "loss": 0.8364,
       "step": 376
     },
     {
-      "epoch": 0.7265774378585086,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002934178046071116,
-      "loss": 0.8064,
+      "epoch": 0.4454865181711606,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029783582710740013,
+      "loss": 0.8101,
       "step": 380
     },
     {
-      "epoch": 0.734225621414914,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002932645548642024,
-      "loss": 0.8427,
+      "epoch": 0.4501758499413834,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00029778151922635175,
+      "loss": 0.8661,
       "step": 384
     },
     {
-      "epoch": 0.7418738049713193,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002931095825322931,
-      "loss": 0.7602,
+      "epoch": 0.4548651817116061,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0002977265434355696,
+      "loss": 0.8419,
       "step": 388
     },
     {
-      "epoch": 0.7495219885277247,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029295288947474513,
-      "loss": 0.881,
+      "epoch": 0.45955451348182885,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000297670899983521,
+      "loss": 0.8444,
       "step": 392
     },
     {
-      "epoch": 0.7571701720841301,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002927944775756095,
-      "loss": 0.865,
+      "epoch": 0.4642438452520516,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00029761458912169064,
+      "loss": 0.8413,
       "step": 396
     },
     {
-      "epoch": 0.7648183556405354,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002926343487396044,
-      "loss": 0.839,
+      "epoch": 0.46893317702227433,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0002975576111045798,
+      "loss": 0.7815,
       "step": 400
     },
     {
-      "epoch": 0.7724665391969407,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029247250489209217,
-      "loss": 0.7939,
+      "epoch": 0.47362250879249707,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002974999661897049,
+      "loss": 0.8496,
       "step": 404
     },
     {
-      "epoch": 0.780114722753346,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029230894797905595,
-      "loss": 0.7748,
+      "epoch": 0.4783118405627198,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002974416546375965,
+      "loss": 0.842,
       "step": 408
     },
     {
-      "epoch": 0.7877629063097514,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00029214367996707676,
-      "loss": 0.7829,
+      "epoch": 0.48300117233294254,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029738267671179793,
+      "loss": 0.8453,
       "step": 412
     },
     {
-      "epoch": 0.7954110898661568,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029197670284330954,
-      "loss": 0.7867,
+      "epoch": 0.4876905041031653,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029732303267886455,
+      "loss": 0.8066,
       "step": 416
     },
     {
-      "epoch": 0.8030592734225621,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00029180801861545906,
-      "loss": 0.7971,
+      "epoch": 0.492379835873388,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00029726272280836206,
+      "loss": 0.749,
       "step": 420
     },
     {
-      "epoch": 0.8107074569789675,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002916376293117564,
-      "loss": 0.8241,
+      "epoch": 0.4970691676436108,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0002972017473728654,
+      "loss": 0.8594,
       "step": 424
     },
     {
-      "epoch": 0.8183556405353728,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00029146553698093387,
-      "loss": 0.8119,
+      "epoch": 0.5017584994138335,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00029714010664795765,
+      "loss": 0.8312,
       "step": 428
     },
     {
-      "epoch": 0.8260038240917782,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00029129174369220087,
-      "loss": 0.8048,
+      "epoch": 0.5064478311840562,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00029707780091222877,
+      "loss": 0.8336,
       "step": 432
     },
     {
-      "epoch": 0.8336520076481836,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029111625153521877,
-      "loss": 0.76,
+      "epoch": 0.511137162954279,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002970148304472742,
+      "loss": 0.8229,
       "step": 436
     },
     {
-      "epoch": 0.8413001912045889,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00029093906262007583,
-      "loss": 0.7833,
+      "epoch": 0.5158264947245017,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002969511955376937,
+      "loss": 0.7516,
       "step": 440
     },
     {
-      "epoch": 0.8489483747609943,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00029076017907726196,
-      "loss": 0.8027,
+      "epoch": 0.5205158264947245,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00029688689647109013,
+      "loss": 0.8158,
       "step": 444
     },
     {
-      "epoch": 0.8565965583173997,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0002905796030576428,
-      "loss": 0.818,
+      "epoch": 0.5252051582649473,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00029682193353806793,
+      "loss": 0.7859,
       "step": 448
     },
     {
-      "epoch": 0.864244741873805,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00029039733673243416,
-      "loss": 0.8358,
+      "epoch": 0.52989449003517,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00029675630703223196,
+      "loss": 0.8025,
       "step": 452
     },
     {
-      "epoch": 0.8718929254302104,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002902133822931759,
-      "loss": 0.7543,
+      "epoch": 0.5345838218053928,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002966900172501861,
+      "loss": 0.7937,
       "step": 456
     },
     {
-      "epoch": 0.8795411089866156,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029002774195170525,
-      "loss": 0.7765,
+      "epoch": 0.5392731535756154,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029662306449153216,
+      "loss": 0.8076,
       "step": 460
     },
     {
-      "epoch": 0.887189292543021,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002898404179401306,
-      "loss": 0.8094,
+      "epoch": 0.5439624853458382,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00029655544905886816,
+      "loss": 0.8268,
       "step": 464
     },
     {
-      "epoch": 0.8948374760994264,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.0002896514125108045,
-      "loss": 0.7657,
+      "epoch": 0.5486518171160609,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002964871712577871,
+      "loss": 0.7727,
       "step": 468
     },
     {
-      "epoch": 0.9024856596558317,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002894607279362966,
-      "loss": 0.7774,
+      "epoch": 0.5533411488862837,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002964182313968757,
+      "loss": 0.809,
       "step": 472
     },
     {
-      "epoch": 0.9101338432122371,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002892683665093662,
-      "loss": 0.8148,
+      "epoch": 0.5580304806565064,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002963486297877128,
+      "loss": 0.8369,
       "step": 476
     },
     {
-      "epoch": 0.9177820267686424,
+      "epoch": 0.5627198124267292,
       "grad_norm": 0.236328125,
-      "learning_rate": 0.0002890743305429348,
-      "loss": 0.7882,
+      "learning_rate": 0.00029627836674486817,
+      "loss": 0.7811,
       "step": 480
     },
     {
-      "epoch": 0.9254302103250478,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.0002888786223700585,
-      "loss": 0.7656,
+      "epoch": 0.567409144196952,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029620744258590097,
+      "loss": 0.7798,
       "step": 484
     },
     {
-      "epoch": 0.9330783938814532,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00028868124434389944,
-      "loss": 0.7802,
+      "epoch": 0.5720984759671747,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002961358576313583,
+      "loss": 0.809,
       "step": 488
     },
     {
-      "epoch": 0.9407265774378585,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00028848219883769805,
-      "loss": 0.7773,
+      "epoch": 0.5767878077373975,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029606361220477364,
+      "loss": 0.7677,
       "step": 492
     },
     {
-      "epoch": 0.9483747609942639,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.000288281488244744,
-      "loss": 0.7803,
+      "epoch": 0.5814771395076201,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0002959907066326658,
+      "loss": 0.7994,
       "step": 496
     },
     {
-      "epoch": 0.9560229445506692,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.000288079114978348,
-      "loss": 0.8056,
+      "epoch": 0.5861664712778429,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00029591714124453693,
+      "loss": 0.8217,
       "step": 500
     },
     {
-      "epoch": 0.9636711281070746,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002878750814718121,
-      "loss": 0.8309,
+      "epoch": 0.5908558030480656,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029584291637287146,
+      "loss": 0.7913,
       "step": 504
     },
     {
-      "epoch": 0.97131931166348,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00028766939017840114,
-      "loss": 0.7737,
+      "epoch": 0.5955451348182884,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029576803235313413,
+      "loss": 0.7679,
       "step": 508
     },
     {
-      "epoch": 0.9789674952198852,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00028746204357131273,
-      "loss": 0.8039,
+      "epoch": 0.6002344665885111,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029569248952376903,
+      "loss": 0.7839,
       "step": 512
     },
     {
-      "epoch": 0.9866156787762906,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002872530441436477,
-      "loss": 0.7341,
+      "epoch": 0.6049237983587339,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00029561628822619775,
+      "loss": 0.8129,
       "step": 516
     },
     {
-      "epoch": 0.994263862332696,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.0002870423944083801,
-      "loss": 0.8122,
+      "epoch": 0.6096131301289566,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029553942880481765,
+      "loss": 0.8105,
       "step": 520
     },
     {
-      "epoch": 1.0019120458891013,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0002868300968983271,
-      "loss": 0.7403,
+      "epoch": 0.6143024618991794,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002954619116070008,
+      "loss": 0.7806,
       "step": 524
     },
     {
-      "epoch": 1.0095602294455066,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0002866161541661185,
-      "loss": 0.697,
+      "epoch": 0.6189917936694022,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00029538373698309193,
+      "loss": 0.7739,
       "step": 528
     },
     {
-      "epoch": 1.0172084130019121,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002864005687841656,
-      "loss": 0.7442,
+      "epoch": 0.6236811254396248,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00029530490528640723,
+      "loss": 0.7882,
       "step": 532
     },
     {
-      "epoch": 1.0248565965583174,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002861833433446312,
-      "loss": 0.6853,
+      "epoch": 0.6283704572098476,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029522541687323253,
+      "loss": 0.7934,
       "step": 536
     },
     {
-      "epoch": 1.0325047801147227,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00028596448045939735,
-      "loss": 0.7398,
+      "epoch": 0.6330597889800703,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00029514527210282163,
+      "loss": 0.8188,
       "step": 540
     },
     {
-      "epoch": 1.0401529636711282,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002857439827600348,
-      "loss": 0.7912,
+      "epoch": 0.6377491207502931,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029506447133739494,
+      "loss": 0.79,
       "step": 544
     },
     {
-      "epoch": 1.0478011472275335,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002855218528977709,
-      "loss": 0.7138,
+      "epoch": 0.6424384525205158,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002949830149421376,
+      "loss": 0.7961,
       "step": 548
     },
     {
-      "epoch": 1.0554493307839388,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00028529809354345794,
-      "loss": 0.726,
+      "epoch": 0.6471277842907386,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00029490090328519795,
+      "loss": 0.798,
       "step": 552
     },
     {
-      "epoch": 1.063097514340344,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002850727073875409,
-      "loss": 0.7058,
+      "epoch": 0.6518171160609613,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029481813673768576,
+      "loss": 0.7388,
       "step": 556
     },
     {
-      "epoch": 1.0707456978967496,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00028484569714002517,
-      "loss": 0.7102,
+      "epoch": 0.6565064478311841,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0002947347156736708,
+      "loss": 0.8278,
       "step": 560
     },
     {
-      "epoch": 1.0783938814531548,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002846170655304438,
-      "loss": 0.6534,
+      "epoch": 0.6611957796014069,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002946506404701808,
+      "loss": 0.7871,
       "step": 564
     },
     {
-      "epoch": 1.0860420650095601,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002843868153078251,
-      "loss": 0.6918,
+      "epoch": 0.6658851113716295,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002945659115071999,
+      "loss": 0.7859,
       "step": 568
     },
     {
-      "epoch": 1.0936902485659656,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.000284154949240659,
-      "loss": 0.6673,
+      "epoch": 0.6705744431418523,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002944805291676672,
+      "loss": 0.8397,
       "step": 572
     },
     {
-      "epoch": 1.101338432122371,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002839214701168644,
-      "loss": 0.6722,
+      "epoch": 0.675263774912075,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002943944938374746,
+      "loss": 0.776,
       "step": 576
     },
     {
-      "epoch": 1.1089866156787762,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028368638074375516,
-      "loss": 0.7141,
+      "epoch": 0.6799531066822978,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002943078059054652,
+      "loss": 0.8045,
       "step": 580
     },
     {
-      "epoch": 1.1166347992351817,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002834496839480063,
-      "loss": 0.665,
+      "epoch": 0.6846424384525205,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002942204657634317,
+      "loss": 0.8548,
       "step": 584
     },
     {
-      "epoch": 1.124282982791587,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00028321138257562066,
-      "loss": 0.6886,
+      "epoch": 0.6893317702227433,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002941324738061145,
+      "loss": 0.741,
       "step": 588
     },
     {
-      "epoch": 1.1319311663479923,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00028297147949189386,
-      "loss": 0.6997,
+      "epoch": 0.694021101992966,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029404383043119984,
+      "loss": 0.7604,
       "step": 592
     },
     {
-      "epoch": 1.1395793499043978,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028272997758138044,
-      "loss": 0.7051,
+      "epoch": 0.6987104337631888,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00029395453603931816,
+      "loss": 0.7914,
       "step": 596
     },
     {
-      "epoch": 1.147227533460803,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00028248687974785896,
-      "loss": 0.7188,
+      "epoch": 0.7033997655334114,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00029386459103404215,
+      "loss": 0.7633,
       "step": 600
     },
     {
-      "epoch": 1.1548757170172084,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002822421889142969,
-      "loss": 0.6757,
+      "epoch": 0.7080890973036342,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002937739958218852,
+      "loss": 0.7783,
       "step": 604
     },
     {
-      "epoch": 1.1625239005736137,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00028199590802281595,
-      "loss": 0.7203,
+      "epoch": 0.712778429073857,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000293682750812299,
+      "loss": 0.7437,
       "step": 608
     },
     {
-      "epoch": 1.1701720841300192,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00028174804003465616,
-      "loss": 0.6943,
+      "epoch": 0.7174677608440797,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00029359085641767244,
+      "loss": 0.746,
       "step": 612
     },
     {
-      "epoch": 1.1778202676864244,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002814985879301408,
-      "loss": 0.7037,
+      "epoch": 0.7221570926143025,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029349831305332914,
+      "loss": 0.774,
       "step": 616
     },
     {
-      "epoch": 1.1854684512428297,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002812475547086401,
-      "loss": 0.7215,
+      "epoch": 0.7268464243845252,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002934051211375258,
+      "loss": 0.7411,
       "step": 620
     },
     {
-      "epoch": 1.1931166347992352,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00028099494338853554,
-      "loss": 0.6863,
+      "epoch": 0.731535756154748,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002933112810914503,
+      "loss": 0.7725,
       "step": 624
     },
     {
-      "epoch": 1.2007648183556405,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002807407570071832,
-      "loss": 0.7432,
+      "epoch": 0.7362250879249707,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002932167933392198,
+      "loss": 0.812,
       "step": 628
     },
     {
-      "epoch": 1.2084130019120458,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00028048499862087757,
-      "loss": 0.7265,
+      "epoch": 0.7409144196951934,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002931216583078789,
+      "loss": 0.8059,
       "step": 632
     },
     {
-      "epoch": 1.2160611854684513,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00028022767130481466,
-      "loss": 0.6848,
+      "epoch": 0.7456037514654161,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002930258764273975,
+      "loss": 0.7539,
       "step": 636
     },
     {
-      "epoch": 1.2237093690248566,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.0002799687781530549,
-      "loss": 0.717,
+      "epoch": 0.7502930832356389,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000292929448130669,
+      "loss": 0.7813,
       "step": 640
     },
     {
-      "epoch": 1.231357552581262,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00027970832227848627,
-      "loss": 0.7011,
+      "epoch": 0.7549824150058617,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002928323738535085,
+      "loss": 0.817,
       "step": 644
     },
     {
-      "epoch": 1.2390057361376674,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002794463068127866,
-      "loss": 0.7696,
+      "epoch": 0.7596717467760844,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00029273465403465045,
+      "loss": 0.7949,
       "step": 648
     },
     {
-      "epoch": 1.2466539196940727,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00027918273490638574,
-      "loss": 0.6922,
+      "epoch": 0.7643610785463072,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002926362891157469,
+      "loss": 0.7825,
       "step": 652
     },
     {
-      "epoch": 1.254302103250478,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002789176097284283,
-      "loss": 0.6521,
+      "epoch": 0.7690504103165299,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002925372795413656,
+      "loss": 0.7385,
       "step": 656
     },
     {
-      "epoch": 1.2619502868068833,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002786509344667349,
-      "loss": 0.6642,
+      "epoch": 0.7737397420867527,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029243762575898775,
+      "loss": 0.8009,
       "step": 660
     },
     {
-      "epoch": 1.2695984703632888,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002783827123277643,
-      "loss": 0.7773,
+      "epoch": 0.7784290738569754,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002923373282190062,
+      "loss": 0.7594,
       "step": 664
     },
     {
-      "epoch": 1.277246653919694,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00027811294653657444,
-      "loss": 0.7314,
+      "epoch": 0.7831184056271981,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00029223638737472325,
+      "loss": 0.7312,
       "step": 668
     },
     {
-      "epoch": 1.2848948374760996,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.000277841640336784,
-      "loss": 0.7461,
+      "epoch": 0.7878077373974208,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029213480368234853,
+      "loss": 0.7294,
       "step": 672
     },
     {
-      "epoch": 1.2925430210325048,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00027756879699053337,
-      "loss": 0.7426,
+      "epoch": 0.7924970691676436,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029203257760099737,
+      "loss": 0.7024,
       "step": 676
     },
     {
-      "epoch": 1.3001912045889101,
+      "epoch": 0.7971864009378663,
       "grad_norm": 0.26171875,
-      "learning_rate": 0.0002772944197784451,
-      "loss": 0.6986,
+      "learning_rate": 0.0002919297095926883,
+      "loss": 0.7424,
       "step": 680
     },
     {
-      "epoch": 1.3078393881453154,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002770185119995849,
-      "loss": 0.7379,
+      "epoch": 0.8018757327080891,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002918262001223408,
+      "loss": 0.7668,
       "step": 684
     },
     {
-      "epoch": 1.3154875717017207,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002767410769714216,
-      "loss": 0.7146,
+      "epoch": 0.8065650644783119,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000291722049657774,
+      "loss": 0.7923,
       "step": 688
     },
     {
-      "epoch": 1.3231357552581262,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002764621180297875,
-      "loss": 0.7061,
+      "epoch": 0.8112543962485346,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002916172586697036,
+      "loss": 0.8041,
       "step": 692
     },
     {
-      "epoch": 1.3307839388145315,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002761816385288382,
-      "loss": 0.6547,
+      "epoch": 0.8159437280187574,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029151182763174053,
+      "loss": 0.7869,
       "step": 696
     },
     {
-      "epoch": 1.338432122370937,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002758996418410122,
-      "loss": 0.7018,
+      "epoch": 0.82063305978898,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002914057570203882,
+      "loss": 0.7462,
       "step": 700
     },
     {
-      "epoch": 1.3460803059273423,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002756161313569904,
-      "loss": 0.7062,
+      "epoch": 0.8253223915592028,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0002912990473150409,
+      "loss": 0.7332,
       "step": 704
     },
     {
-      "epoch": 1.3537284894837476,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00027533111048565537,
-      "loss": 0.778,
+      "epoch": 0.8300117233294255,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000291191698997981,
+      "loss": 0.7944,
       "step": 708
     },
     {
-      "epoch": 1.3613766730401529,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00027504458265405034,
-      "loss": 0.6916,
+      "epoch": 0.8347010550996483,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002910837125543775,
+      "loss": 0.7531,
       "step": 712
     },
     {
-      "epoch": 1.3690248565965584,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00027475655130733786,
-      "loss": 0.6862,
+      "epoch": 0.839390386869871,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002909750884722832,
+      "loss": 0.7751,
       "step": 716
     },
     {
-      "epoch": 1.3766730401529637,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00027446701990875864,
-      "loss": 0.7037,
+      "epoch": 0.8440797186400938,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00029086582724263286,
+      "loss": 0.7757,
       "step": 720
     },
     {
-      "epoch": 1.384321223709369,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00027417599193958964,
-      "loss": 0.6976,
+      "epoch": 0.8487690504103166,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00029075592935924084,
+      "loss": 0.7761,
       "step": 724
     },
     {
-      "epoch": 1.3919694072657744,
+      "epoch": 0.8534583821805393,
       "grad_norm": 0.24609375,
-      "learning_rate": 0.00027388347089910253,
-      "loss": 0.7209,
+      "learning_rate": 0.00029064539531879893,
+      "loss": 0.8099,
       "step": 728
     },
     {
-      "epoch": 1.3996175908221797,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002735894603045211,
-      "loss": 0.7009,
+      "epoch": 0.8581477139507621,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002905342256208741,
+      "loss": 0.759,
       "step": 732
     },
     {
-      "epoch": 1.407265774378585,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002732939636909796,
-      "loss": 0.6583,
+      "epoch": 0.8628370457209847,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002904224207679061,
+      "loss": 0.7609,
       "step": 736
     },
     {
-      "epoch": 1.4149139579349903,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00027299698461147966,
-      "loss": 0.6999,
+      "epoch": 0.8675263774912075,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002903099812652056,
+      "loss": 0.717,
       "step": 740
     },
     {
-      "epoch": 1.4225621414913958,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002726985266368481,
-      "loss": 0.7269,
+      "epoch": 0.8722157092614302,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029019690762095116,
+      "loss": 0.7318,
       "step": 744
     },
     {
-      "epoch": 1.430210325047801,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002723985933556936,
-      "loss": 0.6256,
+      "epoch": 0.876905041031653,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00029008320034618784,
+      "loss": 0.7132,
       "step": 748
     },
     {
-      "epoch": 1.4378585086042066,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00027209718837436353,
-      "loss": 0.7129,
+      "epoch": 0.8815943728018757,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00028996885995482424,
+      "loss": 0.7291,
       "step": 752
     },
     {
-      "epoch": 1.445506692160612,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.000271794315316901,
-      "loss": 0.6623,
+      "epoch": 0.8862837045720985,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002898538869636303,
+      "loss": 0.7854,
       "step": 756
     },
     {
-      "epoch": 1.4531548757170172,
+      "epoch": 0.8909730363423212,
       "grad_norm": 0.259765625,
-      "learning_rate": 0.00027148997782500085,
-      "loss": 0.6869,
+      "learning_rate": 0.0002897382818922352,
+      "loss": 0.7146,
       "step": 760
     },
     {
-      "epoch": 1.4608030592734225,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002711841795579661,
-      "loss": 0.7426,
+      "epoch": 0.895662368112544,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002896220452631247,
+      "loss": 0.8165,
       "step": 764
     },
     {
-      "epoch": 1.468451242829828,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00027087692419266383,
-      "loss": 0.6731,
+      "epoch": 0.9003516998827668,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002895051776016392,
+      "loss": 0.7719,
       "step": 768
     },
     {
-      "epoch": 1.4760994263862333,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00027056821542348114,
-      "loss": 0.7591,
+      "epoch": 0.9050410316529894,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00028938767943597075,
+      "loss": 0.793,
       "step": 772
     },
     {
-      "epoch": 1.4837476099426385,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002702580569622805,
-      "loss": 0.7129,
+      "epoch": 0.9097303634232122,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0002892695512971613,
+      "loss": 0.7722,
       "step": 776
     },
     {
-      "epoch": 1.491395793499044,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002699464525383552,
-      "loss": 0.7307,
+      "epoch": 0.9144196951934349,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002891507937190998,
+      "loss": 0.664,
       "step": 780
     },
     {
-      "epoch": 1.4990439770554493,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002696334058983848,
-      "loss": 0.7317,
+      "epoch": 0.9191090269636577,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002890314072385201,
+      "loss": 0.7462,
       "step": 784
     },
     {
-      "epoch": 1.5066921606118546,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002693189208063894,
-      "loss": 0.6994,
+      "epoch": 0.9237983587338804,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002889113923949985,
+      "loss": 0.7702,
       "step": 788
     },
     {
-      "epoch": 1.51434034416826,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00026900300104368524,
-      "loss": 0.72,
+      "epoch": 0.9284876905041032,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002887907497309511,
+      "loss": 0.7545,
       "step": 792
     },
     {
-      "epoch": 1.5219885277246654,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002686856504088385,
-      "loss": 0.7112,
+      "epoch": 0.9331770222743259,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002886694797916314,
+      "loss": 0.7855,
       "step": 796
     },
     {
-      "epoch": 1.5296367112810707,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00026836687271762015,
-      "loss": 0.6912,
+      "epoch": 0.9378663540445487,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00028854758312512826,
+      "loss": 0.7477,
       "step": 800
     },
     {
-      "epoch": 1.5372848948374762,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002680466718029596,
-      "loss": 0.6801,
+      "epoch": 0.9425556858147714,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00028842506028236274,
+      "loss": 0.7647,
       "step": 804
     },
     {
-      "epoch": 1.5449330783938815,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00026772505151489897,
-      "loss": 0.7077,
+      "epoch": 0.9472450175849941,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002883019118170861,
+      "loss": 0.7492,
       "step": 808
     },
     {
-      "epoch": 1.5525812619502868,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00026740201572054685,
-      "loss": 0.6926,
+      "epoch": 0.9519343493552169,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002881781382858772,
+      "loss": 0.729,
       "step": 812
     },
     {
-      "epoch": 1.560229445506692,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00026707756830403144,
-      "loss": 0.702,
+      "epoch": 0.9566236811254396,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002880537402481397,
+      "loss": 0.7545,
       "step": 816
     },
     {
-      "epoch": 1.5678776290630974,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00026675171316645403,
-      "loss": 0.7178,
+      "epoch": 0.9613130128956624,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002879287182661001,
+      "loss": 0.7451,
       "step": 820
     },
     {
-      "epoch": 1.5755258126195029,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00026642445422584224,
-      "loss": 0.6843,
+      "epoch": 0.9660023446658851,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028780307290480453,
+      "loss": 0.7464,
       "step": 824
     },
     {
-      "epoch": 1.5831739961759084,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002660957954171028,
-      "loss": 0.6722,
+      "epoch": 0.9706916764361079,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00028767680473211683,
+      "loss": 0.7747,
       "step": 828
     },
     {
-      "epoch": 1.5908221797323137,
+      "epoch": 0.9753810082063306,
       "grad_norm": 0.251953125,
-      "learning_rate": 0.00026576574069197406,
-      "loss": 0.6518,
+      "learning_rate": 0.0002875499143187154,
+      "loss": 0.7581,
       "step": 832
     },
     {
-      "epoch": 1.598470363288719,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00026543429401897875,
-      "loss": 0.6998,
+      "epoch": 0.9800703399765534,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00028742240223809116,
+      "loss": 0.7786,
       "step": 836
     },
     {
-      "epoch": 1.6061185468451242,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002651014593833762,
-      "loss": 0.6966,
+      "epoch": 0.984759671746776,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002872942690665445,
+      "loss": 0.8077,
       "step": 840
     },
     {
-      "epoch": 1.6137667304015295,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00026476724078711416,
-      "loss": 0.7054,
+      "epoch": 0.9894490035169988,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002871655153831831,
+      "loss": 0.7733,
       "step": 844
     },
     {
-      "epoch": 1.621414913957935,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00026443164224878115,
-      "loss": 0.6655,
+      "epoch": 0.9941383352872216,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002870361417699188,
+      "loss": 0.7462,
       "step": 848
     },
     {
-      "epoch": 1.6290630975143403,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002640946678035576,
-      "loss": 0.7098,
+      "epoch": 0.9988276670574443,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002869061488114654,
+      "loss": 0.695,
       "step": 852
     },
     {
-      "epoch": 1.6367112810707458,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002637563215031679,
-      "loss": 0.7099,
+      "epoch": 1.003516998827667,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00028677553709533605,
+      "loss": 0.7101,
       "step": 856
     },
     {
-      "epoch": 1.644359464627151,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00026341660741583127,
-      "loss": 0.6704,
+      "epoch": 1.0082063305978899,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00028664430721184013,
+      "loss": 0.6597,
       "step": 860
     },
     {
-      "epoch": 1.6520076481835564,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00026307552962621293,
-      "loss": 0.686,
+      "epoch": 1.0128956623681125,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002865124597540811,
+      "loss": 0.6458,
       "step": 864
     },
     {
-      "epoch": 1.6596558317399617,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00026273309223537507,
-      "loss": 0.69,
+      "epoch": 1.0175849941383353,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002863799953179534,
+      "loss": 0.706,
       "step": 868
     },
     {
-      "epoch": 1.667304015296367,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002623892993607275,
-      "loss": 0.6957,
+      "epoch": 1.022274325908558,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00028624691450214007,
+      "loss": 0.7264,
       "step": 872
     },
     {
-      "epoch": 1.6749521988527725,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00026204415513597813,
-      "loss": 0.7071,
+      "epoch": 1.0269636576787808,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00028611321790810996,
+      "loss": 0.6963,
       "step": 876
     },
     {
-      "epoch": 1.682600382409178,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002616976637110832,
-      "loss": 0.6313,
+      "epoch": 1.0316529894490034,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002859789061401149,
+      "loss": 0.6838,
       "step": 880
     },
     {
-      "epoch": 1.6902485659655833,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002613498292521977,
-      "loss": 0.6809,
+      "epoch": 1.0363423212192262,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00028584397980518705,
+      "loss": 0.7163,
       "step": 884
     },
     {
-      "epoch": 1.6978967495219885,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00026100065594162475,
-      "loss": 0.6867,
+      "epoch": 1.041031652989449,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028570843951313625,
+      "loss": 0.6668,
       "step": 888
     },
     {
-      "epoch": 1.7055449330783938,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00026065014797776575,
-      "loss": 0.7065,
+      "epoch": 1.0457209847596718,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00028557228587654693,
+      "loss": 0.7022,
       "step": 892
     },
     {
-      "epoch": 1.7131931166347991,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0002602983095750698,
-      "loss": 0.6938,
+      "epoch": 1.0504103165298946,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002854355195107758,
+      "loss": 0.6669,
       "step": 896
     },
     {
-      "epoch": 1.7208413001912046,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002599451449639828,
-      "loss": 0.7138,
+      "epoch": 1.0550996483001172,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00028529814103394886,
+      "loss": 0.7281,
       "step": 900
     },
     {
-      "epoch": 1.72848948374761,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00025959065839089684,
-      "loss": 0.6976,
+      "epoch": 1.05978898007034,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00028516015106695833,
+      "loss": 0.714,
       "step": 904
     },
     {
-      "epoch": 1.7361376673040154,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00025923485411809917,
-      "loss": 0.6792,
+      "epoch": 1.0644783118405627,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0002850215502334605,
+      "loss": 0.6515,
       "step": 908
     },
     {
-      "epoch": 1.7437858508604207,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00025887773642372064,
-      "loss": 0.7016,
+      "epoch": 1.0691676436107855,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002848823391598721,
+      "loss": 0.7476,
       "step": 912
     },
     {
-      "epoch": 1.751434034416826,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00025851930960168464,
-      "loss": 0.6845,
+      "epoch": 1.073856975381008,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00028474251847536826,
+      "loss": 0.6498,
       "step": 916
     },
     {
-      "epoch": 1.7590822179732313,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002581595779616552,
-      "loss": 0.6932,
+      "epoch": 1.078546307151231,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000284602088811879,
+      "loss": 0.7022,
       "step": 920
     },
     {
-      "epoch": 1.7667304015296366,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002577985458289852,
-      "loss": 0.6911,
+      "epoch": 1.0832356389214537,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002844610508040868,
+      "loss": 0.6899,
       "step": 924
     },
     {
-      "epoch": 1.774378585086042,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00025743621754466457,
-      "loss": 0.6611,
+      "epoch": 1.0879249706916765,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00028431940508942365,
+      "loss": 0.6563,
       "step": 928
     },
     {
-      "epoch": 1.7820267686424476,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002570725974652679,
-      "loss": 0.7158,
+      "epoch": 1.0926143024618993,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000284177152308068,
+      "loss": 0.6752,
       "step": 932
     },
     {
-      "epoch": 1.7896749521988529,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002567076899629021,
-      "loss": 0.7176,
+      "epoch": 1.0973036342321218,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000284034293102942,
+      "loss": 0.7024,
       "step": 936
     },
     {
-      "epoch": 1.7973231357552581,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002563414994251538,
-      "loss": 0.7001,
+      "epoch": 1.1019929660023446,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00028389082811970873,
+      "loss": 0.6902,
       "step": 940
     },
     {
-      "epoch": 1.8049713193116634,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002559740302550366,
-      "loss": 0.7583,
+      "epoch": 1.1066822977725674,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00028374675800676893,
+      "loss": 0.6658,
       "step": 944
     },
     {
-      "epoch": 1.8126195028680687,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002556052868709383,
-      "loss": 0.6374,
+      "epoch": 1.1113716295427902,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00028360208341525836,
+      "loss": 0.7058,
       "step": 948
     },
     {
-      "epoch": 1.8202676864244742,
+      "epoch": 1.1160609613130128,
       "grad_norm": 0.279296875,
-      "learning_rate": 0.00025523527370656753,
-      "loss": 0.6771,
+      "learning_rate": 0.0002834568049990447,
+      "loss": 0.7162,
       "step": 952
     },
     {
-      "epoch": 1.8279158699808795,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002548639952109006,
-      "loss": 0.6547,
+      "epoch": 1.1207502930832356,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002833109234147249,
+      "loss": 0.6888,
       "step": 956
     },
     {
-      "epoch": 1.835564053537285,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002544914558481279,
-      "loss": 0.7078,
+      "epoch": 1.1254396248534584,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002831644393216216,
+      "loss": 0.6802,
       "step": 960
     },
     {
-      "epoch": 1.8432122370936903,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00025411766009760027,
-      "loss": 0.6913,
+      "epoch": 1.1301289566236812,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00028301735338178086,
+      "loss": 0.6579,
       "step": 964
     },
     {
-      "epoch": 1.8508604206500956,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00025374261245377525,
-      "loss": 0.725,
+      "epoch": 1.134818288393904,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002828696662599686,
+      "loss": 0.6941,
       "step": 968
     },
     {
-      "epoch": 1.8585086042065009,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002533663174261628,
-      "loss": 0.6871,
+      "epoch": 1.1395076201641265,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00028272137862366795,
+      "loss": 0.6928,
       "step": 972
     },
     {
-      "epoch": 1.8661567877629062,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002529887795392713,
-      "loss": 0.6748,
+      "epoch": 1.1441969519343493,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00028257249114307606,
+      "loss": 0.6596,
       "step": 976
     },
     {
-      "epoch": 1.8738049713193117,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00025261000333255305,
-      "loss": 0.6224,
+      "epoch": 1.1488862837045721,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028242300449110114,
+      "loss": 0.6836,
       "step": 980
     },
     {
-      "epoch": 1.8814531548757172,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002522299933603497,
-      "loss": 0.6914,
+      "epoch": 1.153575615474795,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00028227291934335944,
+      "loss": 0.6556,
       "step": 984
     },
     {
-      "epoch": 1.8891013384321225,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002518487541918374,
-      "loss": 0.684,
+      "epoch": 1.1582649472450175,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00028212223637817213,
+      "loss": 0.6663,
       "step": 988
     },
     {
-      "epoch": 1.8967495219885278,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.000251466290410972,
-      "loss": 0.7131,
+      "epoch": 1.1629542790152403,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00028197095627656215,
+      "loss": 0.6999,
       "step": 992
     },
     {
-      "epoch": 1.904397705544933,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002510826066164341,
-      "loss": 0.6956,
+      "epoch": 1.167643610785463,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002818190797222514,
+      "loss": 0.6913,
       "step": 996
     },
     {
-      "epoch": 1.9120458891013383,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00025069770742157317,
-      "loss": 0.6683,
+      "epoch": 1.1723329425556859,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002816666074016575,
+      "loss": 0.6688,
       "step": 1000
     },
     {
-      "epoch": 1.9196940726577438,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00025031159745435267,
-      "loss": 0.6941,
+      "epoch": 1.1770222743259087,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002815135400038905,
+      "loss": 0.6966,
       "step": 1004
     },
     {
-      "epoch": 1.9273422562141491,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0002499242813572942,
-      "loss": 0.7458,
+      "epoch": 1.1817116060961312,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002813598782207501,
+      "loss": 0.6789,
       "step": 1008
     },
     {
-      "epoch": 1.9349904397705546,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002495357637874215,
-      "loss": 0.6724,
+      "epoch": 1.186400937866354,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028120562274672233,
+      "loss": 0.7121,
       "step": 1012
     },
     {
-      "epoch": 1.94263862332696,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002491460494162048,
-      "loss": 0.6662,
+      "epoch": 1.1910902696365768,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002810507742789765,
+      "loss": 0.6333,
       "step": 1016
     },
     {
-      "epoch": 1.9502868068833652,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024875514292950447,
-      "loss": 0.652,
+      "epoch": 1.1957796014067996,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002808953335173619,
+      "loss": 0.683,
       "step": 1020
     },
     {
-      "epoch": 1.9579349904397705,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00024836304902751445,
-      "loss": 0.7223,
+      "epoch": 1.2004689331770222,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00028073930116440484,
+      "loss": 0.6771,
       "step": 1024
     },
     {
-      "epoch": 1.9655831739961758,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002479697724247062,
-      "loss": 0.7065,
+      "epoch": 1.205158264947245,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002805826779253052,
+      "loss": 0.6461,
       "step": 1028
     },
     {
-      "epoch": 1.9732313575525813,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002475753178497716,
-      "loss": 0.7307,
+      "epoch": 1.2098475967174678,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002804254645079337,
+      "loss": 0.665,
       "step": 1032
     },
     {
-      "epoch": 1.9808795411089866,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024717969004556646,
-      "loss": 0.7086,
+      "epoch": 1.2145369284876906,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002802676616228281,
+      "loss": 0.7006,
       "step": 1036
     },
     {
-      "epoch": 1.988527724665392,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002467828937690532,
-      "loss": 0.7051,
+      "epoch": 1.2192262602579134,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00028010926998319055,
+      "loss": 0.6557,
       "step": 1040
     },
     {
-      "epoch": 1.9961759082217974,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002463849337912437,
-      "loss": 0.7084,
+      "epoch": 1.223915592028136,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000279950290304884,
+      "loss": 0.6676,
       "step": 1044
     },
     {
-      "epoch": 2.0038240917782026,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00024598581489714206,
-      "loss": 0.5785,
+      "epoch": 1.2286049237983587,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000279790723306429,
+      "loss": 0.6762,
       "step": 1048
     },
     {
-      "epoch": 2.011472275334608,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002455855418856869,
-      "loss": 0.5793,
+      "epoch": 1.2332942555685815,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00027963056970900085,
+      "loss": 0.6742,
       "step": 1052
     },
     {
-      "epoch": 2.019120458891013,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.000245184119569694,
-      "loss": 0.6212,
+      "epoch": 1.2379835873388043,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002794698302364257,
+      "loss": 0.6682,
       "step": 1056
     },
     {
-      "epoch": 2.026768642447419,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002447815527757979,
-      "loss": 0.569,
+      "epoch": 1.2426729191090269,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002793085056151778,
+      "loss": 0.6691,
       "step": 1060
     },
     {
-      "epoch": 2.0344168260038242,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002443778463443944,
-      "loss": 0.5868,
+      "epoch": 1.2473622508792497,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00027914659657437586,
+      "loss": 0.7082,
       "step": 1064
     },
     {
-      "epoch": 2.0420650095602295,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.0002439730051295818,
-      "loss": 0.6195,
+      "epoch": 1.2520515826494725,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00027898410384578004,
+      "loss": 0.6849,
       "step": 1068
     },
     {
-      "epoch": 2.049713193116635,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002435670339991031,
-      "loss": 0.5886,
+      "epoch": 1.2567409144196953,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00027882102816378853,
+      "loss": 0.6864,
       "step": 1072
     },
     {
-      "epoch": 2.05736137667304,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00024315993783428718,
-      "loss": 0.6097,
+      "epoch": 1.261430246189918,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002786573702654342,
+      "loss": 0.6555,
       "step": 1076
     },
     {
-      "epoch": 2.0650095602294454,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00024275172152999006,
-      "loss": 0.5727,
+      "epoch": 1.2661195779601406,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002784931308903813,
+      "loss": 0.7219,
       "step": 1080
     },
     {
-      "epoch": 2.0726577437858507,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00024234238999453614,
-      "loss": 0.6085,
+      "epoch": 1.2708089097303634,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000278328310780922,
+      "loss": 0.7314,
       "step": 1084
     },
     {
-      "epoch": 2.0803059273422564,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00024193194814965934,
-      "loss": 0.6144,
+      "epoch": 1.2754982415005862,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002781629106819733,
+      "loss": 0.6988,
       "step": 1088
     },
     {
-      "epoch": 2.0879541108986617,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00024152040093044353,
-      "loss": 0.5868,
+      "epoch": 1.2801875732708088,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002779969313410733,
+      "loss": 0.6559,
       "step": 1092
     },
     {
-      "epoch": 2.095602294455067,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00024110775328526352,
-      "loss": 0.6278,
+      "epoch": 1.2848769050410316,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002778303735083784,
+      "loss": 0.6329,
       "step": 1096
     },
     {
-      "epoch": 2.1032504780114722,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00024069401017572543,
-      "loss": 0.5923,
+      "epoch": 1.2895662368112544,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002776632379366591,
+      "loss": 0.6737,
       "step": 1100
     },
     {
-      "epoch": 2.1108986615678775,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00024027917657660713,
-      "loss": 0.5759,
+      "epoch": 1.2942555685814772,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002774955253812973,
+      "loss": 0.6834,
       "step": 1104
     },
     {
-      "epoch": 2.118546845124283,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00023986325747579824,
-      "loss": 0.6138,
+      "epoch": 1.2989449003517,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00027732723660028256,
+      "loss": 0.6357,
       "step": 1108
     },
     {
-      "epoch": 2.126195028680688,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002394462578742403,
-      "loss": 0.5786,
+      "epoch": 1.3036342321219228,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002771583723542087,
+      "loss": 0.6635,
       "step": 1112
     },
     {
-      "epoch": 2.133843212237094,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002390281827858668,
-      "loss": 0.64,
+      "epoch": 1.3083235638921453,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002769889334062705,
+      "loss": 0.6871,
       "step": 1116
     },
     {
-      "epoch": 2.141491395793499,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002386090372375424,
-      "loss": 0.6307,
+      "epoch": 1.3130128956623681,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00027681892052226005,
+      "loss": 0.7013,
       "step": 1120
     },
     {
-      "epoch": 2.1491395793499044,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00023818882626900294,
-      "loss": 0.5641,
+      "epoch": 1.317702227432591,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002766483344705634,
+      "loss": 0.6578,
       "step": 1124
     },
     {
-      "epoch": 2.1567877629063097,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023776755493279473,
-      "loss": 0.623,
+      "epoch": 1.3223915592028135,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00027647717602215704,
+      "loss": 0.66,
       "step": 1128
     },
     {
-      "epoch": 2.164435946462715,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00023734522829421372,
-      "loss": 0.6022,
+      "epoch": 1.3270808909730363,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00027630544595060464,
+      "loss": 0.6884,
       "step": 1132
     },
     {
-      "epoch": 2.1720841300191203,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00023692185143124464,
-      "loss": 0.6121,
+      "epoch": 1.331770222743259,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002761331450320531,
+      "loss": 0.6845,
       "step": 1136
     },
     {
-      "epoch": 2.179732313575526,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00023649742943449996,
-      "loss": 0.5878,
+      "epoch": 1.3364595545134819,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00027596027404522944,
+      "loss": 0.6924,
       "step": 1140
     },
     {
-      "epoch": 2.1873804971319313,
+      "epoch": 1.3411488862837047,
       "grad_norm": 0.2578125,
-      "learning_rate": 0.00023607196740715858,
-      "loss": 0.6143,
+      "learning_rate": 0.0002757868337714372,
+      "loss": 0.6962,
       "step": 1144
     },
     {
-      "epoch": 2.1950286806883366,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00023564547046490468,
-      "loss": 0.5655,
+      "epoch": 1.3458382180539274,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00027561282499455276,
+      "loss": 0.6869,
       "step": 1148
     },
     {
-      "epoch": 2.202676864244742,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00023521794373586603,
-      "loss": 0.5685,
+      "epoch": 1.35052754982415,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00027543824850102187,
+      "loss": 0.7182,
       "step": 1152
     },
     {
-      "epoch": 2.210325047801147,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00023478939236055228,
-      "loss": 0.5845,
+      "epoch": 1.3552168815943728,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00027526310507985626,
+      "loss": 0.6907,
       "step": 1156
     },
     {
-      "epoch": 2.2179732313575524,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00023435982149179346,
-      "loss": 0.6108,
+      "epoch": 1.3599062133645956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002750873955226298,
+      "loss": 0.7224,
       "step": 1160
     },
     {
-      "epoch": 2.2256214149139577,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002339292362946777,
-      "loss": 0.6221,
+      "epoch": 1.3645955451348182,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00027491112062347515,
+      "loss": 0.7173,
       "step": 1164
     },
     {
-      "epoch": 2.2332695984703634,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002334976419464892,
-      "loss": 0.5739,
+      "epoch": 1.369284876905041,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002747342811790799,
+      "loss": 0.6719,
       "step": 1168
     },
     {
-      "epoch": 2.2409177820267687,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00023306504363664613,
-      "loss": 0.5928,
+      "epoch": 1.3739742086752638,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00027455687798868346,
+      "loss": 0.7309,
       "step": 1172
     },
     {
-      "epoch": 2.248565965583174,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00023263144656663801,
-      "loss": 0.5422,
+      "epoch": 1.3786635404454866,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002743789118540728,
+      "loss": 0.7328,
       "step": 1176
     },
     {
-      "epoch": 2.2562141491395793,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00023219685594996347,
-      "loss": 0.5815,
+      "epoch": 1.3833528722157094,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00027420038357957934,
+      "loss": 0.6742,
       "step": 1180
     },
     {
-      "epoch": 2.2638623326959846,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00023176127701206713,
-      "loss": 0.5786,
+      "epoch": 1.388042203985932,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002740212939720751,
+      "loss": 0.7193,
       "step": 1184
     },
     {
-      "epoch": 2.27151051625239,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00023132471499027717,
-      "loss": 0.5634,
+      "epoch": 1.3927315357561547,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002738416438409691,
+      "loss": 0.7321,
       "step": 1188
     },
     {
-      "epoch": 2.2791586998087956,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.0002308871751337422,
-      "loss": 0.5969,
+      "epoch": 1.3974208675263775,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002736614339982036,
+      "loss": 0.7035,
       "step": 1192
     },
     {
-      "epoch": 2.286806883365201,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00023044866270336822,
-      "loss": 0.587,
+      "epoch": 1.4021101992966003,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00027348066525825066,
+      "loss": 0.6952,
       "step": 1196
     },
     {
-      "epoch": 2.294455066921606,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00023000918297175506,
-      "loss": 0.6312,
+      "epoch": 1.4067995310668229,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002732993384381082,
+      "loss": 0.776,
       "step": 1200
     },
     {
-      "epoch": 2.3021032504780115,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00022956874122313347,
-      "loss": 0.6268,
+      "epoch": 1.4114888628370457,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00027311745435729655,
+      "loss": 0.7111,
       "step": 1204
     },
     {
-      "epoch": 2.3097514340344167,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00022912734275330117,
-      "loss": 0.586,
+      "epoch": 1.4161781946072685,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00027293501383785445,
+      "loss": 0.6733,
       "step": 1208
     },
     {
-      "epoch": 2.317399617590822,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00022868499286955943,
-      "loss": 0.571,
+      "epoch": 1.4208675263774913,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00027275201770433574,
+      "loss": 0.7015,
       "step": 1212
     },
     {
-      "epoch": 2.3250478011472273,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00022824169689064915,
-      "loss": 0.6062,
+      "epoch": 1.425556858147714,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002725684667838051,
+      "loss": 0.6968,
       "step": 1216
     },
     {
-      "epoch": 2.332695984703633,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00022779746014668683,
-      "loss": 0.5991,
+      "epoch": 1.4302461899179366,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00027238436190583486,
+      "loss": 0.7071,
       "step": 1220
     },
     {
-      "epoch": 2.3403441682600383,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00022735228797910066,
-      "loss": 0.6193,
+      "epoch": 1.4349355216881594,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00027219970390250094,
+      "loss": 0.6828,
       "step": 1224
     },
     {
-      "epoch": 2.3479923518164436,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002269061857405662,
-      "loss": 0.5719,
+      "epoch": 1.4396248534583822,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000272014493608379,
+      "loss": 0.648,
       "step": 1228
     },
     {
-      "epoch": 2.355640535372849,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00022645915879494202,
-      "loss": 0.6171,
+      "epoch": 1.444314185228605,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002718287318605411,
+      "loss": 0.6828,
       "step": 1232
     },
     {
-      "epoch": 2.363288718929254,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00022601121251720514,
-      "loss": 0.6213,
+      "epoch": 1.4490035169988276,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002716424194985514,
+      "loss": 0.6893,
       "step": 1236
     },
     {
-      "epoch": 2.3709369024856595,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002255623522933866,
-      "loss": 0.5933,
+      "epoch": 1.4536928487690504,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002714555573644627,
+      "loss": 0.7048,
       "step": 1240
     },
     {
-      "epoch": 2.378585086042065,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022511258352050649,
-      "loss": 0.5598,
+      "epoch": 1.4583821805392732,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002712681463028126,
+      "loss": 0.6076,
       "step": 1244
     },
     {
-      "epoch": 2.3862332695984705,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022466191160650916,
-      "loss": 0.6435,
+      "epoch": 1.463071512309496,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00027108018716061945,
+      "loss": 0.7175,
       "step": 1248
     },
     {
-      "epoch": 2.3938814531548758,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00022421034197019822,
-      "loss": 0.5553,
+      "epoch": 1.4677608440797187,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002708916807873787,
+      "loss": 0.6885,
       "step": 1252
     },
     {
-      "epoch": 2.401529636711281,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00022375788004117128,
-      "loss": 0.6193,
+      "epoch": 1.4724501758499413,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002707026280350594,
+      "loss": 0.6368,
       "step": 1256
     },
     {
-      "epoch": 2.4091778202676863,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00022330453125975474,
-      "loss": 0.6117,
+      "epoch": 1.477139507620164,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00027051302975809947,
+      "loss": 0.6759,
       "step": 1260
     },
     {
-      "epoch": 2.4168260038240916,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0002228503010769384,
-      "loss": 0.5761,
+      "epoch": 1.481828839390387,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00027032288681340285,
+      "loss": 0.6619,
       "step": 1264
     },
     {
-      "epoch": 2.424474187380497,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002223951949543098,
-      "loss": 0.6343,
+      "epoch": 1.4865181711606097,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002701322000603347,
+      "loss": 0.7396,
       "step": 1268
     },
     {
-      "epoch": 2.4321223709369026,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00022193921836398875,
-      "loss": 0.5763,
+      "epoch": 1.4912075029308323,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00026994097036071846,
+      "loss": 0.6688,
       "step": 1272
     },
     {
-      "epoch": 2.439770554493308,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00022148237678856138,
-      "loss": 0.5807,
+      "epoch": 1.495896834701055,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000269749198578831,
+      "loss": 0.6439,
       "step": 1276
     },
     {
-      "epoch": 2.447418738049713,
+      "epoch": 1.5005861664712778,
       "grad_norm": 0.279296875,
-      "learning_rate": 0.0002210246757210142,
-      "loss": 0.62,
+      "learning_rate": 0.00026955688558139945,
+      "loss": 0.7026,
       "step": 1280
     },
     {
-      "epoch": 2.4550669216061185,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00022056612066466817,
-      "loss": 0.6255,
+      "epoch": 1.5052754982415006,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002693640322375969,
+      "loss": 0.6609,
       "step": 1284
     },
     {
-      "epoch": 2.462715105162524,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00022010671713311238,
-      "loss": 0.6786,
+      "epoch": 1.5099648300117234,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002691706394190386,
+      "loss": 0.6639,
       "step": 1288
     },
     {
-      "epoch": 2.470363288718929,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.000219646470650138,
-      "loss": 0.6173,
+      "epoch": 1.5146541617819462,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00026897670799977803,
+      "loss": 0.6592,
       "step": 1292
     },
     {
-      "epoch": 2.478011472275335,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00021918538674967156,
-      "loss": 0.6113,
+      "epoch": 1.5193434935521688,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002687822388563028,
+      "loss": 0.631,
       "step": 1296
     },
     {
-      "epoch": 2.48565965583174,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002187234709757087,
-      "loss": 0.5949,
+      "epoch": 1.5240328253223916,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000268587232867531,
+      "loss": 0.671,
       "step": 1300
     },
     {
-      "epoch": 2.4933078393881454,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00021826072888224716,
-      "loss": 0.6248,
+      "epoch": 1.5287221570926142,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00026839169091480685,
+      "loss": 0.7033,
       "step": 1304
     },
     {
-      "epoch": 2.5009560229445507,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021779716603322034,
-      "loss": 0.5849,
+      "epoch": 1.533411488862837,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026819561388189697,
+      "loss": 0.6508,
       "step": 1308
     },
     {
-      "epoch": 2.508604206500956,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002173327880024303,
-      "loss": 0.5947,
+      "epoch": 1.5381008206330598,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026799900265498625,
+      "loss": 0.6661,
       "step": 1312
     },
     {
-      "epoch": 2.5162523900573612,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00021686760037348065,
-      "loss": 0.5689,
+      "epoch": 1.5427901524032825,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002678018581226741,
+      "loss": 0.6565,
       "step": 1316
     },
     {
-      "epoch": 2.5239005736137665,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00021640160873970954,
-      "loss": 0.6302,
+      "epoch": 1.5474794841735053,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00026760418117597007,
+      "loss": 0.7513,
       "step": 1320
     },
     {
-      "epoch": 2.5315487571701722,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00021593481870412217,
-      "loss": 0.6117,
+      "epoch": 1.5521688159437281,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00026740597270829,
+      "loss": 0.6873,
       "step": 1324
     },
     {
-      "epoch": 2.5391969407265775,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0002154672358793238,
-      "loss": 0.6024,
+      "epoch": 1.556858147713951,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026720723361545206,
+      "loss": 0.6724,
       "step": 1328
     },
     {
-      "epoch": 2.546845124282983,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00021499886588745195,
-      "loss": 0.5417,
+      "epoch": 1.5615474794841735,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002670079647956726,
+      "loss": 0.7035,
       "step": 1332
     },
     {
-      "epoch": 2.554493307839388,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00021452971436010886,
-      "loss": 0.5975,
+      "epoch": 1.5662368112543963,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00026680816714956215,
+      "loss": 0.6473,
       "step": 1336
     },
     {
-      "epoch": 2.5621414913957934,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00021405978693829397,
-      "loss": 0.5997,
+      "epoch": 1.5709261430246189,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002666078415801211,
+      "loss": 0.7342,
       "step": 1340
     },
     {
-      "epoch": 2.569789674952199,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00021358908927233576,
-      "loss": 0.6047,
+      "epoch": 1.5756154747948417,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002664069889927361,
+      "loss": 0.7155,
       "step": 1344
     },
     {
-      "epoch": 2.5774378585086044,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021311762702182414,
-      "loss": 0.6135,
+      "epoch": 1.5803048065650644,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00026620561029517555,
+      "loss": 0.7006,
       "step": 1348
     },
     {
-      "epoch": 2.5850860420650097,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00021264540585554215,
-      "loss": 0.6251,
+      "epoch": 1.5849941383352872,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002660037063975857,
+      "loss": 0.6876,
       "step": 1352
     },
     {
-      "epoch": 2.592734225621415,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00021217243145139802,
-      "loss": 0.6308,
+      "epoch": 1.58968347010551,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002658012782124865,
+      "loss": 0.6475,
       "step": 1356
     },
     {
-      "epoch": 2.6003824091778203,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002116987094963567,
-      "loss": 0.5828,
+      "epoch": 1.5943728018757328,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002655983266547673,
+      "loss": 0.6373,
       "step": 1360
     },
     {
-      "epoch": 2.6080305927342256,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00021122424568637157,
-      "loss": 0.6057,
+      "epoch": 1.5990621336459554,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000265394852641683,
+      "loss": 0.6848,
       "step": 1364
     },
     {
-      "epoch": 2.615678776290631,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00021074904572631606,
-      "loss": 0.6435,
+      "epoch": 1.6037514654161782,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002651908570928498,
+      "loss": 0.7114,
       "step": 1368
     },
     {
-      "epoch": 2.623326959847036,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00021027311532991475,
-      "loss": 0.6201,
+      "epoch": 1.608440797186401,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002649863409302411,
+      "loss": 0.7023,
       "step": 1372
     },
     {
-      "epoch": 2.6309751434034414,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00020979646021967503,
-      "loss": 0.6192,
+      "epoch": 1.6131301289566236,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000264781305078183,
+      "loss": 0.6557,
       "step": 1376
     },
     {
-      "epoch": 2.638623326959847,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020931908612681805,
-      "loss": 0.6072,
+      "epoch": 1.6178194607268463,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00026457575046335055,
+      "loss": 0.6688,
       "step": 1380
     },
     {
-      "epoch": 2.6462715105162524,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020884099879120993,
-      "loss": 0.5332,
+      "epoch": 1.6225087924970691,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00026436967801476334,
+      "loss": 0.6833,
       "step": 1384
     },
     {
-      "epoch": 2.6539196940726577,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00020836220396129265,
-      "loss": 0.5923,
+      "epoch": 1.627198124267292,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002641630886637814,
+      "loss": 0.6415,
       "step": 1388
     },
     {
-      "epoch": 2.661567877629063,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00020788270739401505,
-      "loss": 0.6293,
+      "epoch": 1.6318874560375147,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002639559833441008,
+      "loss": 0.7361,
       "step": 1392
     },
     {
-      "epoch": 2.6692160611854687,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00020740251485476345,
-      "loss": 0.5851,
+      "epoch": 1.6365767878077375,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026374836299174984,
+      "loss": 0.6997,
       "step": 1396
     },
     {
-      "epoch": 2.676864244741874,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00020692163211729253,
-      "loss": 0.6088,
+      "epoch": 1.64126611957796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002635402285450842,
+      "loss": 0.6023,
       "step": 1400
     },
     {
-      "epoch": 2.6845124282982793,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.0002064400649636557,
-      "loss": 0.6033,
+      "epoch": 1.6459554513481829,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026333158094478333,
+      "loss": 0.7078,
       "step": 1404
     },
     {
-      "epoch": 2.6921606118546846,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0002059578191841357,
-      "loss": 0.5948,
+      "epoch": 1.6506447831184057,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002631224211338458,
+      "loss": 0.7311,
       "step": 1408
     },
     {
-      "epoch": 2.69980879541109,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00020547490057717499,
-      "loss": 0.6287,
+      "epoch": 1.6553341148886282,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00026291275005758507,
+      "loss": 0.7061,
       "step": 1412
     },
     {
-      "epoch": 2.707456978967495,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00020499131494930602,
-      "loss": 0.5736,
+      "epoch": 1.660023446658851,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00026270256866362554,
+      "loss": 0.6862,
       "step": 1416
     },
     {
-      "epoch": 2.7151051625239004,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002045070681150813,
-      "loss": 0.6496,
+      "epoch": 1.6647127784290738,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002624918779018979,
+      "loss": 0.6868,
       "step": 1420
     },
     {
-      "epoch": 2.7227533460803057,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00020402216589700362,
-      "loss": 0.5993,
+      "epoch": 1.6694021101992966,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00026228067872463475,
+      "loss": 0.7052,
       "step": 1424
     },
     {
-      "epoch": 2.730401529636711,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00020353661412545598,
-      "loss": 0.596,
+      "epoch": 1.6740914419695194,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002620689720863669,
+      "loss": 0.6386,
       "step": 1428
     },
     {
-      "epoch": 2.7380497131931167,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00020305041863863152,
-      "loss": 0.639,
+      "epoch": 1.6787807737397422,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002618567589439185,
+      "loss": 0.6729,
       "step": 1432
     },
     {
-      "epoch": 2.745697896749522,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00020256358528246334,
-      "loss": 0.5703,
+      "epoch": 1.6834701055099648,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00026164404025640276,
+      "loss": 0.655,
       "step": 1436
     },
     {
-      "epoch": 2.7533460803059273,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00020207611991055407,
-      "loss": 0.5838,
+      "epoch": 1.6881594372801876,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002614308169852179,
+      "loss": 0.7085,
       "step": 1440
     },
     {
-      "epoch": 2.7609942638623326,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0002015880283841057,
-      "loss": 0.5845,
+      "epoch": 1.6928487690504102,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00026121709009404264,
+      "loss": 0.7295,
       "step": 1444
     },
     {
-      "epoch": 2.768642447418738,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020109931657184894,
-      "loss": 0.6169,
+      "epoch": 1.697538100820633,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00026100286054883166,
+      "loss": 0.6506,
       "step": 1448
     },
     {
-      "epoch": 2.7762906309751436,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0002006099903499727,
-      "loss": 0.6026,
+      "epoch": 1.7022274325908557,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002607881293178117,
+      "loss": 0.6513,
       "step": 1452
     },
     {
-      "epoch": 2.783938814531549,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00020012005560205356,
-      "loss": 0.6278,
+      "epoch": 1.7069167643610785,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00026057289737147675,
+      "loss": 0.6906,
       "step": 1456
     },
     {
-      "epoch": 2.791586998087954,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0001996295182189847,
-      "loss": 0.6273,
+      "epoch": 1.7116060961313013,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00026035716568258377,
+      "loss": 0.6819,
       "step": 1460
     },
     {
-      "epoch": 2.7992351816443595,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00019913838409890548,
-      "loss": 0.6084,
+      "epoch": 1.7162954279015241,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002601409352261485,
+      "loss": 0.6822,
       "step": 1464
     },
     {
-      "epoch": 2.8068833652007648,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00019864665914713024,
-      "loss": 0.6295,
+      "epoch": 1.720984759671747,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002599242069794407,
+      "loss": 0.6983,
       "step": 1468
     },
     {
-      "epoch": 2.81453154875717,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0001981543492760774,
-      "loss": 0.5889,
+      "epoch": 1.7256740914419695,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00025970698192198026,
+      "loss": 0.663,
       "step": 1472
     },
     {
-      "epoch": 2.8221797323135753,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00019766146040519836,
-      "loss": 0.6064,
+      "epoch": 1.7303634232121923,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00025948926103553196,
+      "loss": 0.7066,
       "step": 1476
     },
     {
-      "epoch": 2.8298279158699806,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00019716799846090634,
-      "loss": 0.6269,
+      "epoch": 1.7350527549824148,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00025927104530410193,
+      "loss": 0.6464,
       "step": 1480
     },
     {
-      "epoch": 2.8374760994263863,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00019667396937650506,
-      "loss": 0.5742,
+      "epoch": 1.7397420867526376,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002590523357139327,
+      "loss": 0.6341,
       "step": 1484
     },
     {
-      "epoch": 2.8451242829827916,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.0001961793790921174,
-      "loss": 0.5701,
+      "epoch": 1.7444314185228604,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025883313325349866,
+      "loss": 0.6986,
       "step": 1488
     },
     {
-      "epoch": 2.852772466539197,
+      "epoch": 1.7491207502930832,
       "grad_norm": 0.2734375,
-      "learning_rate": 0.00019568423355461402,
-      "loss": 0.5973,
+      "learning_rate": 0.0002586134389135019,
+      "loss": 0.6552,
       "step": 1492
     },
     {
-      "epoch": 2.860420650095602,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00019518853871754204,
-      "loss": 0.609,
+      "epoch": 1.753810082063306,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002583932536868676,
+      "loss": 0.681,
       "step": 1496
     },
     {
-      "epoch": 2.8680688336520075,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00019469230054105295,
-      "loss": 0.5944,
+      "epoch": 1.7584994138335288,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00025817257856873956,
+      "loss": 0.6324,
       "step": 1500
     },
     {
-      "epoch": 2.875717017208413,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0001941955249918315,
-      "loss": 0.5914,
+      "epoch": 1.7631887456037516,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00025795141455647554,
+      "loss": 0.6766,
       "step": 1504
     },
     {
-      "epoch": 2.8833652007648185,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00019369821804302365,
-      "loss": 0.6191,
+      "epoch": 1.7678780773739742,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002577297626496431,
+      "loss": 0.6905,
       "step": 1508
     },
     {
-      "epoch": 2.891013384321224,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00019320038567416484,
-      "loss": 0.6409,
+      "epoch": 1.772567409144197,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00025750762385001464,
+      "loss": 0.6752,
       "step": 1512
     },
     {
-      "epoch": 2.898661567877629,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00019270203387110798,
-      "loss": 0.5779,
+      "epoch": 1.7772567409144195,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002572849991615633,
+      "loss": 0.6785,
       "step": 1516
     },
     {
-      "epoch": 2.9063097514340344,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00019220316862595167,
-      "loss": 0.5956,
+      "epoch": 1.7819460726846423,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00025706188959045826,
+      "loss": 0.6516,
       "step": 1520
     },
     {
-      "epoch": 2.9139579349904396,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00019170379593696802,
-      "loss": 0.5916,
+      "epoch": 1.7866354044548651,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00025683829614505993,
+      "loss": 0.628,
       "step": 1524
     },
     {
-      "epoch": 2.921606118546845,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00019120392180853058,
-      "loss": 0.6069,
+      "epoch": 1.791324736225088,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00025661421983591586,
+      "loss": 0.6955,
       "step": 1528
     },
     {
-      "epoch": 2.92925430210325,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0001907035522510421,
-      "loss": 0.6029,
+      "epoch": 1.7960140679953107,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002563896616757558,
+      "loss": 0.6969,
       "step": 1532
     },
     {
-      "epoch": 2.936902485659656,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00019020269328086226,
-      "loss": 0.5706,
+      "epoch": 1.8007033997655335,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00025616462267948726,
+      "loss": 0.6666,
       "step": 1536
     },
     {
-      "epoch": 2.9445506692160612,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0001897013509202354,
-      "loss": 0.6024,
+      "epoch": 1.8053927315357563,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00025593910386419107,
+      "loss": 0.6589,
       "step": 1540
     },
     {
-      "epoch": 2.9521988527724665,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00018919953119721808,
-      "loss": 0.6326,
+      "epoch": 1.8100820633059789,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002557131062491165,
+      "loss": 0.6676,
       "step": 1544
     },
     {
-      "epoch": 2.959847036328872,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0001886972401456065,
-      "loss": 0.5744,
+      "epoch": 1.8147713950762017,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002554866308556769,
+      "loss": 0.6157,
       "step": 1548
     },
     {
-      "epoch": 2.967495219885277,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00018819448380486413,
-      "loss": 0.5679,
+      "epoch": 1.8194607268464242,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002552596787074448,
+      "loss": 0.6262,
       "step": 1552
     },
     {
-      "epoch": 2.975143403441683,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00018769126822004898,
-      "loss": 0.5992,
+      "epoch": 1.824150058616647,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00025503225083014765,
+      "loss": 0.7328,
       "step": 1556
     },
     {
-      "epoch": 2.982791586998088,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00018718759944174086,
-      "loss": 0.5981,
+      "epoch": 1.8288393903868698,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002548043482516629,
+      "loss": 0.6837,
       "step": 1560
     },
     {
-      "epoch": 2.9904397705544934,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0001866834835259688,
-      "loss": 0.6188,
+      "epoch": 1.8335287221570926,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002545759720020134,
+      "loss": 0.7037,
       "step": 1564
     },
     {
-      "epoch": 2.9980879541108987,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.0001861789265341381,
-      "loss": 0.617,
+      "epoch": 1.8382180539273154,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002543471231133628,
+      "loss": 0.6798,
       "step": 1568
     },
     {
-      "epoch": 3.005736137667304,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00018567393453295742,
-      "loss": 0.4644,
+      "epoch": 1.8429073856975382,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002541178026200112,
+      "loss": 0.7137,
       "step": 1572
     },
     {
-      "epoch": 3.0133843212237093,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018516851359436602,
-      "loss": 0.4965,
+      "epoch": 1.847596717467761,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002538880115583896,
+      "loss": 0.7179,
       "step": 1576
     },
     {
-      "epoch": 3.0210325047801145,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00018466266979546057,
-      "loss": 0.501,
+      "epoch": 1.8522860492379836,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002536577509670562,
+      "loss": 0.6664,
       "step": 1580
     },
     {
-      "epoch": 3.0286806883365203,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0001841564092184221,
-      "loss": 0.4787,
+      "epoch": 1.8569753810082064,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002534270218866911,
+      "loss": 0.5868,
       "step": 1584
     },
     {
-      "epoch": 3.0363288718929256,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018364973795044294,
-      "loss": 0.5116,
+      "epoch": 1.861664712778429,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025319582536009175,
+      "loss": 0.68,
       "step": 1588
     },
     {
-      "epoch": 3.043977055449331,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00018314266208365357,
-      "loss": 0.5309,
+      "epoch": 1.8663540445486517,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00025296416243216836,
+      "loss": 0.6532,
       "step": 1592
     },
     {
-      "epoch": 3.051625239005736,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00018263518771504924,
-      "loss": 0.4979,
+      "epoch": 1.8710433763188745,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000252732034149939,
+      "loss": 0.6645,
       "step": 1596
     },
     {
-      "epoch": 3.0592734225621414,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00018212732094641666,
-      "loss": 0.4647,
+      "epoch": 1.8757327080890973,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025249944156252504,
+      "loss": 0.6842,
       "step": 1600
     },
     {
-      "epoch": 3.0669216061185467,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00018161906788426076,
-      "loss": 0.5367,
+      "epoch": 1.88042203985932,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002522663857211461,
+      "loss": 0.6603,
       "step": 1604
     },
     {
-      "epoch": 3.0745697896749524,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00018111043463973122,
-      "loss": 0.5095,
+      "epoch": 1.885111371629543,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00025203286767911575,
+      "loss": 0.704,
       "step": 1608
     },
     {
-      "epoch": 3.0822179732313577,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00018060142732854894,
-      "loss": 0.4615,
+      "epoch": 1.8898007033997657,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002517988884918364,
+      "loss": 0.7084,
       "step": 1612
     },
     {
-      "epoch": 3.089866156787763,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00018009205207093252,
-      "loss": 0.5105,
+      "epoch": 1.8944900351699883,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00025156444921679464,
+      "loss": 0.7358,
       "step": 1616
     },
     {
-      "epoch": 3.0975143403441683,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00017958231499152463,
-      "loss": 0.5326,
+      "epoch": 1.899179366940211,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002513295509135564,
+      "loss": 0.6261,
       "step": 1620
     },
     {
-      "epoch": 3.1051625239005736,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001790722222193186,
-      "loss": 0.5383,
+      "epoch": 1.9038686987104336,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002510941946437625,
+      "loss": 0.7177,
       "step": 1624
     },
     {
-      "epoch": 3.112810707456979,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00017856177988758438,
-      "loss": 0.5192,
+      "epoch": 1.9085580304806564,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00025085838147112315,
+      "loss": 0.6916,
       "step": 1628
     },
     {
-      "epoch": 3.120458891013384,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017805099413379508,
-      "loss": 0.5029,
+      "epoch": 1.9132473622508792,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000250622112461414,
+      "loss": 0.6928,
       "step": 1632
     },
     {
-      "epoch": 3.12810707456979,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00017753987109955297,
-      "loss": 0.4896,
+      "epoch": 1.917936694021102,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025038538868247043,
+      "loss": 0.6989,
       "step": 1636
     },
     {
-      "epoch": 3.135755258126195,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00017702841693051577,
-      "loss": 0.5254,
+      "epoch": 1.9226260257913248,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002501482112041836,
+      "loss": 0.6828,
       "step": 1640
     },
     {
-      "epoch": 3.1434034416826004,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001765166377763227,
-      "loss": 0.4829,
+      "epoch": 1.9273153575615476,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00024991058109849495,
+      "loss": 0.6445,
       "step": 1644
     },
     {
-      "epoch": 3.1510516252390057,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017600453979052055,
-      "loss": 0.5461,
+      "epoch": 1.9320046893317704,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00024967249943939174,
+      "loss": 0.6642,
       "step": 1648
     },
     {
-      "epoch": 3.158699808795411,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001754921291304897,
-      "loss": 0.5415,
+      "epoch": 1.936694021101993,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000249433967302902,
+      "loss": 0.6969,
       "step": 1652
     },
     {
-      "epoch": 3.1663479923518163,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00017497941195737004,
-      "loss": 0.5501,
+      "epoch": 1.9413833528722158,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00024919498576708975,
+      "loss": 0.6826,
       "step": 1656
     },
     {
-      "epoch": 3.173996175908222,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00017446639443598696,
-      "loss": 0.4964,
+      "epoch": 1.9460726846424383,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00024895555591205004,
+      "loss": 0.6732,
       "step": 1660
     },
     {
-      "epoch": 3.1816443594646273,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017395308273477714,
-      "loss": 0.4938,
+      "epoch": 1.9507620164126611,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024871567881990414,
+      "loss": 0.6914,
       "step": 1664
     },
     {
-      "epoch": 3.1892925430210326,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00017343948302571446,
-      "loss": 0.5409,
+      "epoch": 1.955451348182884,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00024847535557479477,
+      "loss": 0.6865,
       "step": 1668
     },
     {
-      "epoch": 3.196940726577438,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00017292560148423578,
-      "loss": 0.4844,
+      "epoch": 1.9601406799531067,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002482345872628809,
+      "loss": 0.6773,
       "step": 1672
     },
     {
-      "epoch": 3.204588910133843,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017241144428916655,
-      "loss": 0.539,
+      "epoch": 1.9648300117233295,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002479933749723332,
+      "loss": 0.6618,
       "step": 1676
     },
     {
-      "epoch": 3.2122370936902485,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017189701762264687,
-      "loss": 0.4974,
+      "epoch": 1.9695193434935523,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00024775171979332867,
+      "loss": 0.6894,
       "step": 1680
     },
     {
-      "epoch": 3.2198852772466537,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.0001713823276700567,
-      "loss": 0.5443,
+      "epoch": 1.9742086752637749,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002475096228180463,
+      "loss": 0.6323,
       "step": 1684
     },
     {
-      "epoch": 3.2275334608030595,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00017086738061994176,
-      "loss": 0.4936,
+      "epoch": 1.9788980070339977,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00024726708514066157,
+      "loss": 0.6447,
       "step": 1688
     },
     {
-      "epoch": 3.2351816443594648,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00017035218266393918,
-      "loss": 0.5027,
+      "epoch": 1.9835873388042204,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002470241078573418,
+      "loss": 0.6645,
       "step": 1692
     },
     {
-      "epoch": 3.24282982791587,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00016983673999670273,
-      "loss": 0.5352,
+      "epoch": 1.988276670574443,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00024678069206624117,
+      "loss": 0.6562,
       "step": 1696
     },
     {
-      "epoch": 3.2504780114722753,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001693210588158287,
-      "loss": 0.5147,
+      "epoch": 1.9929660023446658,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002465368388674958,
+      "loss": 0.7016,
       "step": 1700
     },
     {
-      "epoch": 3.2581261950286806,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00016880514532178123,
-      "loss": 0.5013,
+      "epoch": 1.9976553341148886,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00024629254936321855,
+      "loss": 0.6702,
       "step": 1704
     },
     {
-      "epoch": 3.265774378585086,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00016828900571781767,
-      "loss": 0.5408,
+      "epoch": 2.0023446658851114,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002460478246574944,
+      "loss": 0.6856,
       "step": 1708
     },
     {
-      "epoch": 3.2734225621414916,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00016777264620991414,
-      "loss": 0.4758,
+      "epoch": 2.007033997655334,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024580266585637496,
+      "loss": 0.5902,
       "step": 1712
     },
     {
-      "epoch": 3.281070745697897,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00016725607300669087,
-      "loss": 0.5154,
+      "epoch": 2.011723329425557,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00024555707406787405,
+      "loss": 0.563,
       "step": 1716
     },
     {
-      "epoch": 3.288718929254302,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0001667392923193375,
-      "loss": 0.4882,
+      "epoch": 2.0164126611957798,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002453110504019623,
+      "loss": 0.6143,
       "step": 1720
     },
     {
-      "epoch": 3.2963671128107075,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00016622231036153836,
-      "loss": 0.5423,
+      "epoch": 2.021101992966002,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002450645959705622,
+      "loss": 0.5521,
       "step": 1724
     },
     {
-      "epoch": 3.3040152963671128,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001657051333493978,
-      "loss": 0.509,
+      "epoch": 2.025791324736225,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002448177118875432,
+      "loss": 0.6465,
       "step": 1728
     },
     {
-      "epoch": 3.311663479923518,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00016518776750136578,
-      "loss": 0.5447,
+      "epoch": 2.0304806565064477,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00024457039926871656,
+      "loss": 0.5643,
       "step": 1732
     },
     {
-      "epoch": 3.3193116634799233,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016467021903816237,
-      "loss": 0.5048,
+      "epoch": 2.0351699882766705,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024432265923183025,
+      "loss": 0.5682,
       "step": 1736
     },
     {
-      "epoch": 3.3269598470363286,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00016415249418270364,
-      "loss": 0.5183,
+      "epoch": 2.0398593200468933,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00024407449289656416,
+      "loss": 0.5971,
       "step": 1740
     },
     {
-      "epoch": 3.3346080305927344,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00016363459916002643,
-      "loss": 0.4915,
+      "epoch": 2.044548651817116,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00024382590138452475,
+      "loss": 0.5766,
       "step": 1744
     },
     {
-      "epoch": 3.3422562141491396,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00016311654019721377,
-      "loss": 0.5016,
+      "epoch": 2.049237983587339,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00024357688581924013,
+      "loss": 0.5996,
       "step": 1748
     },
     {
-      "epoch": 3.349904397705545,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00016259832352331978,
-      "loss": 0.5276,
+      "epoch": 2.0539273153575617,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024332744732615496,
+      "loss": 0.5863,
       "step": 1752
     },
     {
-      "epoch": 3.35755258126195,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001620799553692949,
-      "loss": 0.5436,
+      "epoch": 2.0586166471277845,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024307758703262527,
+      "loss": 0.6162,
       "step": 1756
     },
     {
-      "epoch": 3.3652007648183555,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00016156144196791103,
-      "loss": 0.5152,
+      "epoch": 2.063305978898007,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024282730606791365,
+      "loss": 0.6311,
       "step": 1760
     },
     {
-      "epoch": 3.3728489483747612,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0001610427895536863,
-      "loss": 0.4845,
+      "epoch": 2.0679953106682296,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00024257660556318373,
+      "loss": 0.566,
       "step": 1764
     },
     {
-      "epoch": 3.3804971319311665,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016052400436281046,
-      "loss": 0.51,
+      "epoch": 2.0726846424384524,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00024232548665149533,
+      "loss": 0.5363,
       "step": 1768
     },
     {
-      "epoch": 3.388145315487572,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00016000509263306976,
-      "loss": 0.5163,
+      "epoch": 2.077373974208675,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00024207395046779945,
+      "loss": 0.591,
       "step": 1772
     },
     {
-      "epoch": 3.395793499043977,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.0001594860606037719,
-      "loss": 0.559,
+      "epoch": 2.082063305978898,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024182199814893278,
+      "loss": 0.579,
       "step": 1776
     },
     {
-      "epoch": 3.4034416826003824,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001589669145156709,
-      "loss": 0.523,
+      "epoch": 2.086752637749121,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00024156963083361282,
+      "loss": 0.6016,
       "step": 1780
     },
     {
-      "epoch": 3.4110898661567877,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00015844766061089241,
-      "loss": 0.4994,
+      "epoch": 2.0914419695193436,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002413168496624328,
+      "loss": 0.6139,
       "step": 1784
     },
     {
-      "epoch": 3.418738049713193,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00015792830513285838,
-      "loss": 0.5259,
+      "epoch": 2.0961313012895664,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00024106365577785625,
+      "loss": 0.6217,
       "step": 1788
     },
     {
-      "epoch": 3.4263862332695982,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.000157408854326212,
-      "loss": 0.4734,
+      "epoch": 2.100820633059789,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00024081005032421202,
+      "loss": 0.5849,
       "step": 1792
     },
     {
-      "epoch": 3.434034416826004,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015688931443674276,
-      "loss": 0.5163,
+      "epoch": 2.1055099648300115,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002405560344476892,
+      "loss": 0.5774,
       "step": 1796
     },
     {
-      "epoch": 3.4416826003824093,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.0001563696917113112,
-      "loss": 0.514,
+      "epoch": 2.1101992966002343,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00024030160929633165,
+      "loss": 0.6123,
       "step": 1800
     },
     {
-      "epoch": 3.4493307839388145,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00015584999239777393,
-      "loss": 0.5691,
+      "epoch": 2.114888628370457,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024004677602003306,
+      "loss": 0.5666,
       "step": 1804
     },
     {
-      "epoch": 3.45697896749522,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001553302227449084,
-      "loss": 0.5365,
+      "epoch": 2.11957796014068,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00023979153577053167,
+      "loss": 0.5931,
       "step": 1808
     },
     {
-      "epoch": 3.464627151051625,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001548103890023378,
-      "loss": 0.5111,
+      "epoch": 2.1242672919109027,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023953588970140503,
+      "loss": 0.5862,
       "step": 1812
     },
     {
-      "epoch": 3.472275334608031,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00015429049742045591,
-      "loss": 0.5272,
+      "epoch": 2.1289566236811255,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00023927983896806495,
+      "loss": 0.5796,
       "step": 1816
     },
     {
-      "epoch": 3.479923518164436,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.000153770554250352,
-      "loss": 0.4746,
+      "epoch": 2.1336459554513483,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002390233847277519,
+      "loss": 0.5914,
       "step": 1820
     },
     {
-      "epoch": 3.4875717017208414,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015325056574373564,
-      "loss": 0.5091,
+      "epoch": 2.138335287221571,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00023876652813953028,
+      "loss": 0.5639,
       "step": 1824
     },
     {
-      "epoch": 3.4952198852772467,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00015273053815286153,
-      "loss": 0.5043,
+      "epoch": 2.143024618991794,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00023850927036428286,
+      "loss": 0.5648,
       "step": 1828
     },
     {
-      "epoch": 3.502868068833652,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00015221047773045424,
-      "loss": 0.5157,
+      "epoch": 2.147713950762016,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00023825161256470546,
+      "loss": 0.5877,
       "step": 1832
     },
     {
-      "epoch": 3.5105162523900573,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00015169039072963312,
-      "loss": 0.525,
+      "epoch": 2.152403282532239,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023799355590530205,
+      "loss": 0.6211,
       "step": 1836
     },
     {
-      "epoch": 3.5181644359464626,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015117028340383713,
-      "loss": 0.536,
+      "epoch": 2.157092614302462,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00023773510155237918,
+      "loss": 0.6338,
       "step": 1840
     },
     {
-      "epoch": 3.525812619502868,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00015065016200674963,
-      "loss": 0.556,
+      "epoch": 2.1617819460726846,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002374762506740408,
+      "loss": 0.59,
       "step": 1844
     },
     {
-      "epoch": 3.5334608030592736,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00015013003279222312,
-      "loss": 0.5199,
+      "epoch": 2.1664712778429074,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023721700444018296,
+      "loss": 0.6159,
       "step": 1848
     },
     {
-      "epoch": 3.541108986615679,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001496099020142041,
-      "loss": 0.5381,
+      "epoch": 2.17116060961313,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00023695736402248865,
+      "loss": 0.6008,
       "step": 1852
     },
     {
-      "epoch": 3.548757170172084,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00014908977592665787,
-      "loss": 0.5092,
+      "epoch": 2.175849941383353,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00023669733059442238,
+      "loss": 0.5405,
       "step": 1856
     },
     {
-      "epoch": 3.5564053537284894,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00014856966078349339,
-      "loss": 0.5101,
+      "epoch": 2.1805392731535758,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023643690533122467,
+      "loss": 0.5994,
       "step": 1860
     },
     {
-      "epoch": 3.5640535372848947,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00014804956283848793,
-      "loss": 0.5093,
+      "epoch": 2.1852286049237986,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00023617608940990737,
+      "loss": 0.6262,
       "step": 1864
     },
     {
-      "epoch": 3.5717017208413004,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00014752948834521206,
-      "loss": 0.499,
+      "epoch": 2.189917936694021,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002359148840092476,
+      "loss": 0.5902,
       "step": 1868
     },
     {
-      "epoch": 3.5793499043977057,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00014700944355695432,
-      "loss": 0.4342,
+      "epoch": 2.1946072684642437,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00023565329030978297,
+      "loss": 0.5659,
       "step": 1872
     },
     {
-      "epoch": 3.586998087954111,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00014648943472664612,
-      "loss": 0.541,
+      "epoch": 2.1992966002344665,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023539130949380585,
+      "loss": 0.6104,
       "step": 1876
     },
     {
-      "epoch": 3.5946462715105163,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00014596946810678646,
-      "loss": 0.5089,
+      "epoch": 2.2039859320046893,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00023512894274535843,
+      "loss": 0.5938,
       "step": 1880
     },
     {
-      "epoch": 3.6022944550669216,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00014544954994936689,
-      "loss": 0.4995,
+      "epoch": 2.208675263774912,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000234866191250227,
+      "loss": 0.5961,
       "step": 1884
     },
     {
-      "epoch": 3.609942638623327,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001449296865057962,
-      "loss": 0.5299,
+      "epoch": 2.213364595545135,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023460305619593674,
+      "loss": 0.6316,
       "step": 1888
     },
     {
-      "epoch": 3.617590822179732,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00014440988402682526,
-      "loss": 0.5933,
+      "epoch": 2.2180539273153577,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00023433953877174645,
+      "loss": 0.5741,
       "step": 1892
     },
     {
-      "epoch": 3.6252390057361374,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00014389014876247205,
-      "loss": 0.5045,
+      "epoch": 2.2227432590855805,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000234075640168643,
+      "loss": 0.5559,
       "step": 1896
     },
     {
-      "epoch": 3.632887189292543,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00014337048696194625,
-      "loss": 0.4814,
+      "epoch": 2.2274325908558033,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00023381136157933603,
+      "loss": 0.5655,
       "step": 1900
     },
     {
-      "epoch": 3.6405353728489485,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00014285090487357427,
-      "loss": 0.5416,
+      "epoch": 2.2321219226260256,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002335467041982526,
+      "loss": 0.5871,
       "step": 1904
     },
     {
-      "epoch": 3.6481835564053537,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001423314087447241,
-      "loss": 0.5236,
+      "epoch": 2.2368112543962484,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002332816692215318,
+      "loss": 0.5668,
       "step": 1908
     },
     {
-      "epoch": 3.655831739961759,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00014181200482173015,
-      "loss": 0.5281,
+      "epoch": 2.241500586166471,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00023301625784701905,
+      "loss": 0.6187,
       "step": 1912
     },
     {
-      "epoch": 3.6634799235181643,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00014129269934981802,
-      "loss": 0.5446,
+      "epoch": 2.246189917936694,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002327504712742612,
+      "loss": 0.6177,
       "step": 1916
     },
     {
-      "epoch": 3.67112810707457,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00014077349857302983,
-      "loss": 0.4949,
+      "epoch": 2.2508792497069168,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002324843107045008,
+      "loss": 0.6034,
       "step": 1920
     },
     {
-      "epoch": 3.6787762906309753,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00014025440873414863,
-      "loss": 0.4875,
+      "epoch": 2.2555685814771396,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00023221777734067046,
+      "loss": 0.5928,
       "step": 1924
     },
     {
-      "epoch": 3.6864244741873806,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001397354360746237,
-      "loss": 0.528,
+      "epoch": 2.2602579132473624,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000231950872387388,
+      "loss": 0.5952,
       "step": 1928
     },
     {
-      "epoch": 3.694072657743786,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001392165868344953,
-      "loss": 0.4827,
+      "epoch": 2.264947245017585,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002316835970509504,
+      "loss": 0.5139,
       "step": 1932
     },
     {
-      "epoch": 3.701720841300191,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.0001386978672523198,
-      "loss": 0.5073,
+      "epoch": 2.269636576787808,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00023141595253932886,
+      "loss": 0.6095,
       "step": 1936
     },
     {
-      "epoch": 3.7093690248565965,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001381792835650945,
-      "loss": 0.5036,
+      "epoch": 2.2743259085580303,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00023114794006216278,
+      "loss": 0.6051,
       "step": 1940
     },
     {
-      "epoch": 3.7170172084130018,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00013766084200818272,
-      "loss": 0.5396,
+      "epoch": 2.279015240328253,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002308795608307549,
+      "loss": 0.571,
       "step": 1944
     },
     {
-      "epoch": 3.724665391969407,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001371425488152389,
-      "loss": 0.4815,
+      "epoch": 2.283704572098476,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002306108160580654,
+      "loss": 0.6444,
       "step": 1948
     },
     {
-      "epoch": 3.7323135755258128,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001366244102181335,
-      "loss": 0.5306,
+      "epoch": 2.2883939038686987,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00023034170695870665,
+      "loss": 0.5642,
       "step": 1952
     },
     {
-      "epoch": 3.739961759082218,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00013610643244687826,
-      "loss": 0.5419,
+      "epoch": 2.2930832356389215,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00023007223474893736,
+      "loss": 0.59,
       "step": 1956
     },
     {
-      "epoch": 3.7476099426386233,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013558862172955105,
-      "loss": 0.5204,
+      "epoch": 2.2977725674091443,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022980240064665765,
+      "loss": 0.5728,
       "step": 1960
     },
     {
-      "epoch": 3.7552581261950286,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013507098429222115,
-      "loss": 0.4982,
+      "epoch": 2.302461899179367,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002295322058714031,
+      "loss": 0.6141,
       "step": 1964
     },
     {
-      "epoch": 3.762906309751434,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013455352635887438,
-      "loss": 0.4667,
+      "epoch": 2.30715123094959,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002292616516443394,
+      "loss": 0.6183,
       "step": 1968
     },
     {
-      "epoch": 3.7705544933078396,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00013403625415133824,
-      "loss": 0.4302,
+      "epoch": 2.311840562719812,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00022899073918825673,
+      "loss": 0.5835,
       "step": 1972
     },
     {
-      "epoch": 3.778202676864245,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00013351917388920704,
-      "loss": 0.4545,
+      "epoch": 2.316529894490035,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00022871946972756455,
+      "loss": 0.6463,
       "step": 1976
     },
     {
-      "epoch": 3.78585086042065,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00013300229178976722,
-      "loss": 0.4953,
+      "epoch": 2.321219226260258,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00022844784448828554,
+      "loss": 0.6421,
       "step": 1980
     },
     {
-      "epoch": 3.7934990439770555,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001324856140679225,
-      "loss": 0.4966,
+      "epoch": 2.3259085580304806,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002281758646980505,
+      "loss": 0.6004,
       "step": 1984
     },
     {
-      "epoch": 3.801147227533461,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001319691469361193,
-      "loss": 0.5236,
+      "epoch": 2.3305978898007034,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002279035315860926,
+      "loss": 0.6484,
       "step": 1988
     },
     {
-      "epoch": 3.808795411089866,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00013145289660427173,
-      "loss": 0.5244,
+      "epoch": 2.335287221570926,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022763084638324202,
+      "loss": 0.5479,
       "step": 1992
     },
     {
-      "epoch": 3.8164435946462714,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00013093686927968738,
-      "loss": 0.4982,
+      "epoch": 2.339976553341149,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00022735781032192,
+      "loss": 0.6289,
       "step": 1996
     },
     {
-      "epoch": 3.8240917782026767,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00013042107116699228,
-      "loss": 0.4899,
+      "epoch": 2.3446658851113718,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00022708442463613367,
+      "loss": 0.5875,
       "step": 2000
     },
     {
-      "epoch": 3.8317399617590824,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00012990550846805654,
-      "loss": 0.5296,
+      "epoch": 2.3493552168815945,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00022681069056147032,
+      "loss": 0.6117,
       "step": 2004
     },
     {
-      "epoch": 3.8393881453154877,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001293901873819196,
-      "loss": 0.5331,
+      "epoch": 2.3540445486518173,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00022653660933509166,
+      "loss": 0.5951,
       "step": 2008
     },
     {
-      "epoch": 3.847036328871893,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00012887511410471589,
-      "loss": 0.501,
+      "epoch": 2.3587338804220397,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022626218219572858,
+      "loss": 0.6197,
       "step": 2012
     },
     {
-      "epoch": 3.8546845124282982,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012836029482960018,
-      "loss": 0.5254,
+      "epoch": 2.3634232121922625,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00022598741038367523,
+      "loss": 0.5915,
       "step": 2016
     },
     {
-      "epoch": 3.8623326959847035,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00012784573574667316,
-      "loss": 0.5009,
+      "epoch": 2.3681125439624853,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002257122951407836,
+      "loss": 0.6072,
       "step": 2020
     },
     {
-      "epoch": 3.8699808795411093,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00012733144304290697,
-      "loss": 0.5107,
+      "epoch": 2.372801875732708,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002254368377104577,
+      "loss": 0.536,
       "step": 2024
     },
     {
-      "epoch": 3.8776290630975145,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001268174229020709,
-      "loss": 0.5025,
+      "epoch": 2.377491207502931,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002251610393376483,
+      "loss": 0.6213,
       "step": 2028
     },
     {
-      "epoch": 3.88527724665392,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001263036815046571,
-      "loss": 0.5239,
+      "epoch": 2.3821805392731537,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022488490126884692,
+      "loss": 0.5981,
       "step": 2032
     },
     {
-      "epoch": 3.892925430210325,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00012579022502780596,
-      "loss": 0.5112,
+      "epoch": 2.3868698710433764,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00022460842475208038,
+      "loss": 0.619,
       "step": 2036
     },
     {
-      "epoch": 3.9005736137667304,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012527705964523209,
-      "loss": 0.5182,
+      "epoch": 2.3915592028135992,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00022433161103690521,
+      "loss": 0.5714,
       "step": 2040
     },
     {
-      "epoch": 3.9082217973231357,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012476419152715007,
-      "loss": 0.5505,
+      "epoch": 2.3962485345838216,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00022405446137440185,
+      "loss": 0.6044,
       "step": 2044
     },
     {
-      "epoch": 3.915869980879541,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00012425162684020024,
-      "loss": 0.4957,
+      "epoch": 2.4009378663540444,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002237769770171692,
+      "loss": 0.5437,
       "step": 2048
     },
     {
-      "epoch": 3.9235181644359463,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001237393717473745,
-      "loss": 0.5132,
+      "epoch": 2.405627198124267,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00022349915921931866,
+      "loss": 0.5494,
       "step": 2052
     },
     {
-      "epoch": 3.9311663479923515,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001232274324079422,
-      "loss": 0.516,
+      "epoch": 2.41031652989449,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0002232210092364689,
+      "loss": 0.6109,
       "step": 2056
     },
     {
-      "epoch": 3.9388145315487573,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00012271581497737619,
-      "loss": 0.5156,
+      "epoch": 2.4150058616647128,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00022294252832573958,
+      "loss": 0.5868,
       "step": 2060
     },
     {
-      "epoch": 3.9464627151051626,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00012220452560727875,
-      "loss": 0.4621,
+      "epoch": 2.4196951934349356,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00022266371774574633,
+      "loss": 0.6001,
       "step": 2064
     },
     {
-      "epoch": 3.954110898661568,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00012169357044530758,
-      "loss": 0.5206,
+      "epoch": 2.4243845252051583,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00022238457875659455,
+      "loss": 0.5908,
       "step": 2068
     },
     {
-      "epoch": 3.961759082217973,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001211829556351019,
-      "loss": 0.4511,
+      "epoch": 2.429073856975381,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000222105112619874,
+      "loss": 0.6001,
       "step": 2072
     },
     {
-      "epoch": 3.969407265774379,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012067268731620861,
-      "loss": 0.5047,
+      "epoch": 2.433763188745604,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022182532059865305,
+      "loss": 0.6076,
       "step": 2076
     },
     {
-      "epoch": 3.977055449330784,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00012016277162400848,
-      "loss": 0.5295,
+      "epoch": 2.4384525205158267,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00022154520395747279,
+      "loss": 0.5652,
       "step": 2080
     },
     {
-      "epoch": 3.9847036328871894,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00011965321468964237,
-      "loss": 0.5204,
+      "epoch": 2.443141852286049,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002212647639623415,
+      "loss": 0.5713,
       "step": 2084
     },
     {
-      "epoch": 3.9923518164435947,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00011914402263993745,
-      "loss": 0.5064,
+      "epoch": 2.447831184056272,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000220984001880729,
+      "loss": 0.5742,
       "step": 2088
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.87109375,
-      "learning_rate": 0.00011863520159733357,
-      "loss": 0.53,
+      "epoch": 2.4525205158264947,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00022070291898156064,
+      "loss": 0.6326,
       "step": 2092
     },
     {
-      "epoch": 4.007648183556405,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00011812675767980972,
-      "loss": 0.4532,
+      "epoch": 2.4572098475967175,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00022042151653521182,
+      "loss": 0.5809,
       "step": 2096
     },
     {
-      "epoch": 4.015296367112811,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00011761869700081036,
-      "loss": 0.4748,
+      "epoch": 2.4618991793669402,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002201397958135022,
+      "loss": 0.5758,
       "step": 2100
     },
     {
-      "epoch": 4.022944550669216,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00011711102566917194,
-      "loss": 0.4188,
+      "epoch": 2.466588511137163,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00021985775808968982,
+      "loss": 0.5594,
       "step": 2104
     },
     {
-      "epoch": 4.030592734225621,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00011660374978904947,
-      "loss": 0.4466,
+      "epoch": 2.471277842907386,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002195754046384654,
+      "loss": 0.5766,
       "step": 2108
     },
     {
-      "epoch": 4.038240917782026,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00011609687545984315,
-      "loss": 0.4112,
+      "epoch": 2.4759671746776086,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021929273673594677,
+      "loss": 0.5754,
       "step": 2112
     },
     {
-      "epoch": 4.045889101338432,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00011559040877612497,
-      "loss": 0.4566,
+      "epoch": 2.480656506447831,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002190097556596728,
+      "loss": 0.5989,
       "step": 2116
     },
     {
-      "epoch": 4.053537284894838,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00011508435582756545,
-      "loss": 0.4413,
+      "epoch": 2.4853458382180538,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002187264626885979,
+      "loss": 0.5721,
       "step": 2120
     },
     {
-      "epoch": 4.061185468451243,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011457872269886043,
-      "loss": 0.4435,
+      "epoch": 2.4900351699882766,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00021844285910308593,
+      "loss": 0.6148,
       "step": 2124
     },
     {
-      "epoch": 4.0688336520076485,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00011407351546965796,
-      "loss": 0.4568,
+      "epoch": 2.4947245017584994,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00021815894618490482,
+      "loss": 0.6034,
       "step": 2128
     },
     {
-      "epoch": 4.076481835564054,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011356874021448506,
-      "loss": 0.4247,
+      "epoch": 2.499413833528722,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00021787472521722038,
+      "loss": 0.5831,
       "step": 2132
     },
     {
-      "epoch": 4.084130019120459,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00011306440300267482,
-      "loss": 0.3762,
+      "epoch": 2.504103165298945,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002175901974845907,
+      "loss": 0.5933,
       "step": 2136
     },
     {
-      "epoch": 4.091778202676864,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00011256050989829337,
-      "loss": 0.4713,
+      "epoch": 2.5087924970691677,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021730536427296045,
+      "loss": 0.589,
       "step": 2140
     },
     {
-      "epoch": 4.09942638623327,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00011205706696006698,
-      "loss": 0.4178,
+      "epoch": 2.5134818288393905,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00021702022686965471,
+      "loss": 0.5794,
       "step": 2144
     },
     {
-      "epoch": 4.107074569789675,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00011155408024130921,
-      "loss": 0.4266,
+      "epoch": 2.5181711606096133,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00021673478656337365,
+      "loss": 0.6354,
       "step": 2148
     },
     {
-      "epoch": 4.11472275334608,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00011105155578984795,
-      "loss": 0.4242,
+      "epoch": 2.522860492379836,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00021644904464418618,
+      "loss": 0.6022,
       "step": 2152
     },
     {
-      "epoch": 4.1223709369024855,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00011054949964795307,
-      "loss": 0.464,
+      "epoch": 2.5275498241500585,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002161630024035245,
+      "loss": 0.5818,
       "step": 2156
     },
     {
-      "epoch": 4.130019120458891,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00011004791785226347,
-      "loss": 0.4216,
+      "epoch": 2.5322391559202813,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0002158766611341781,
+      "loss": 0.5831,
       "step": 2160
     },
     {
-      "epoch": 4.137667304015296,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00010954681643371462,
-      "loss": 0.4222,
+      "epoch": 2.536928487690504,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000215590022130288,
+      "loss": 0.5985,
       "step": 2164
     },
     {
-      "epoch": 4.145315487571701,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00010904620141746601,
-      "loss": 0.4321,
+      "epoch": 2.541617819460727,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00021530308668734079,
+      "loss": 0.6242,
       "step": 2168
     },
     {
-      "epoch": 4.1529636711281075,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001085460788228287,
-      "loss": 0.441,
+      "epoch": 2.5463071512309496,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002150158561021629,
+      "loss": 0.5457,
       "step": 2172
     },
     {
-      "epoch": 4.160611854684513,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00010804645466319292,
-      "loss": 0.4468,
+      "epoch": 2.5509964830011724,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00021472833167291458,
+      "loss": 0.604,
       "step": 2176
     },
     {
-      "epoch": 4.168260038240918,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.0001075473349459559,
-      "loss": 0.3948,
+      "epoch": 2.5556858147713952,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00021444051469908426,
+      "loss": 0.6074,
       "step": 2180
     },
     {
-      "epoch": 4.175908221797323,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010704872567244948,
-      "loss": 0.4233,
+      "epoch": 2.5603751465416176,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00021415240648148246,
+      "loss": 0.557,
       "step": 2184
     },
     {
-      "epoch": 4.183556405353729,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010655063283786795,
-      "loss": 0.4227,
+      "epoch": 2.5650644783118404,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00021386400832223605,
+      "loss": 0.573,
       "step": 2188
     },
     {
-      "epoch": 4.191204588910134,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00010605306243119617,
-      "loss": 0.4242,
+      "epoch": 2.569753810082063,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002135753215247822,
+      "loss": 0.5672,
       "step": 2192
     },
     {
-      "epoch": 4.198852772466539,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00010555602043513724,
-      "loss": 0.4428,
+      "epoch": 2.574443141852286,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021328634739386279,
+      "loss": 0.5693,
       "step": 2196
     },
     {
-      "epoch": 4.2065009560229445,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00010505951282604088,
-      "loss": 0.4132,
+      "epoch": 2.5791324736225087,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002129970872355182,
+      "loss": 0.5845,
       "step": 2200
     },
     {
-      "epoch": 4.21414913957935,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00010456354557383139,
-      "loss": 0.4046,
+      "epoch": 2.5838218053927315,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00021270754235708152,
+      "loss": 0.588,
       "step": 2204
     },
     {
-      "epoch": 4.221797323135755,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010406812464193584,
-      "loss": 0.4527,
+      "epoch": 2.5885111371629543,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021241771406717275,
+      "loss": 0.5666,
       "step": 2208
     },
     {
-      "epoch": 4.22944550669216,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010357325598721255,
-      "loss": 0.4107,
+      "epoch": 2.593200468933177,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002121276036756926,
+      "loss": 0.5667,
       "step": 2212
     },
     {
-      "epoch": 4.237093690248566,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00010307894555987927,
-      "loss": 0.4375,
+      "epoch": 2.5978898007034,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000211837212493817,
+      "loss": 0.6236,
       "step": 2216
     },
     {
-      "epoch": 4.244741873804971,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00010258519930344179,
-      "loss": 0.4328,
+      "epoch": 2.6025791324736227,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021154654183399077,
+      "loss": 0.5384,
       "step": 2220
     },
     {
-      "epoch": 4.252390057361376,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001020920231546223,
-      "loss": 0.4388,
+      "epoch": 2.6072684642438455,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00021125559300992197,
+      "loss": 0.6194,
       "step": 2224
     },
     {
-      "epoch": 4.260038240917782,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00010159942304328819,
-      "loss": 0.4443,
+      "epoch": 2.611957796014068,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00021096436733657572,
+      "loss": 0.6018,
       "step": 2228
     },
     {
-      "epoch": 4.267686424474188,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00010110740489238066,
-      "loss": 0.4446,
+      "epoch": 2.6166471277842906,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00021067286613016847,
+      "loss": 0.6199,
       "step": 2232
     },
     {
-      "epoch": 4.275334608030593,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00010061597461784346,
-      "loss": 0.4269,
+      "epoch": 2.6213364595545134,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00021038109070816184,
+      "loss": 0.585,
       "step": 2236
     },
     {
-      "epoch": 4.282982791586998,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00010012513812855191,
-      "loss": 0.4326,
+      "epoch": 2.6260257913247362,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00021008904238925704,
+      "loss": 0.5525,
       "step": 2240
     },
     {
-      "epoch": 4.2906309751434035,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.963490132624169e-05,
-      "loss": 0.4342,
+      "epoch": 2.630715123094959,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00020979672249338835,
+      "loss": 0.5492,
       "step": 2244
     },
     {
-      "epoch": 4.298279158699809,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.914527010543795e-05,
-      "loss": 0.4157,
+      "epoch": 2.635404454865182,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020950413234171767,
+      "loss": 0.6457,
       "step": 2248
     },
     {
-      "epoch": 4.305927342256214,
-      "grad_norm": 0.3359375,
-      "learning_rate": 9.865625035338447e-05,
-      "loss": 0.4237,
+      "epoch": 2.6400937866354046,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00020921127325662826,
+      "loss": 0.5911,
       "step": 2252
     },
     {
-      "epoch": 4.313575525812619,
-      "grad_norm": 0.314453125,
-      "learning_rate": 9.816784794997275e-05,
-      "loss": 0.4033,
+      "epoch": 2.644783118405627,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00020891814656171895,
+      "loss": 0.5998,
       "step": 2256
     },
     {
-      "epoch": 4.321223709369025,
-      "grad_norm": 0.359375,
-      "learning_rate": 9.76800687676715e-05,
-      "loss": 0.4436,
+      "epoch": 2.6494724501758498,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00020862475358179787,
+      "loss": 0.6458,
       "step": 2260
     },
     {
-      "epoch": 4.32887189292543,
-      "grad_norm": 0.3359375,
-      "learning_rate": 9.719291867145583e-05,
-      "loss": 0.4384,
+      "epoch": 2.6541617819460726,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00020833109564287675,
+      "loss": 0.5858,
       "step": 2264
     },
     {
-      "epoch": 4.336520076481835,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.670640351873688e-05,
-      "loss": 0.4512,
+      "epoch": 2.6588511137162953,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00020803717407216486,
+      "loss": 0.5901,
       "step": 2268
     },
     {
-      "epoch": 4.3441682600382405,
-      "grad_norm": 0.34375,
-      "learning_rate": 9.62205291592913e-05,
-      "loss": 0.4259,
+      "epoch": 2.663540445486518,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002077429901980629,
+      "loss": 0.5914,
       "step": 2272
     },
     {
-      "epoch": 4.351816443594647,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.573530143519098e-05,
-      "loss": 0.4178,
+      "epoch": 2.668229777256741,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00020744854535015715,
+      "loss": 0.6325,
       "step": 2276
     },
     {
-      "epoch": 4.359464627151052,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.525072618073277e-05,
-      "loss": 0.3608,
+      "epoch": 2.6729191090269637,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00020715384085921327,
+      "loss": 0.6216,
       "step": 2280
     },
     {
-      "epoch": 4.367112810707457,
-      "grad_norm": 0.32421875,
-      "learning_rate": 9.476680922236831e-05,
-      "loss": 0.4489,
+      "epoch": 2.6776084407971865,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020685887805717046,
+      "loss": 0.6035,
       "step": 2284
     },
     {
-      "epoch": 4.374760994263863,
-      "grad_norm": 0.330078125,
-      "learning_rate": 9.428355637863402e-05,
-      "loss": 0.4346,
+      "epoch": 2.6822977725674093,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00020656365827713543,
+      "loss": 0.5751,
       "step": 2288
     },
     {
-      "epoch": 4.382409177820268,
-      "grad_norm": 0.33203125,
-      "learning_rate": 9.380097346008112e-05,
-      "loss": 0.4542,
+      "epoch": 2.686987104337632,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002062681828533762,
+      "loss": 0.5953,
       "step": 2292
     },
     {
-      "epoch": 4.390057361376673,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.331906626920576e-05,
-      "loss": 0.4395,
+      "epoch": 2.691676436107855,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00020597245312131636,
+      "loss": 0.621,
       "step": 2296
     },
     {
-      "epoch": 4.397705544933078,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.283784060037921e-05,
-      "loss": 0.4858,
+      "epoch": 2.6963657678780772,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00020567647041752862,
+      "loss": 0.6052,
       "step": 2300
     },
     {
-      "epoch": 4.405353728489484,
-      "grad_norm": 0.30859375,
-      "learning_rate": 9.235730223977837e-05,
-      "loss": 0.4148,
+      "epoch": 2.7010550996483,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002053802360797292,
+      "loss": 0.6146,
       "step": 2304
     },
     {
-      "epoch": 4.413001912045889,
-      "grad_norm": 0.36328125,
-      "learning_rate": 9.187745696531584e-05,
-      "loss": 0.4579,
+      "epoch": 2.705744431418523,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00020508375144677167,
+      "loss": 0.6077,
       "step": 2308
     },
     {
-      "epoch": 4.420650095602294,
-      "grad_norm": 0.357421875,
-      "learning_rate": 9.139831054657081e-05,
-      "loss": 0.457,
+      "epoch": 2.7104337631887456,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00020478701785864057,
+      "loss": 0.568,
       "step": 2312
     },
     {
-      "epoch": 4.4282982791587,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.091986874471956e-05,
-      "loss": 0.4257,
+      "epoch": 2.7151230949589684,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002044900366564458,
+      "loss": 0.6129,
       "step": 2316
     },
     {
-      "epoch": 4.435946462715105,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.044213731246614e-05,
-      "loss": 0.4287,
+      "epoch": 2.719812426729191,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00020419280918241632,
+      "loss": 0.6165,
       "step": 2320
     },
     {
-      "epoch": 4.44359464627151,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.99651219939732e-05,
-      "loss": 0.4482,
+      "epoch": 2.7245017584994136,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00020389533677989417,
+      "loss": 0.6291,
       "step": 2324
     },
     {
-      "epoch": 4.451242829827915,
-      "grad_norm": 0.345703125,
-      "learning_rate": 8.948882852479305e-05,
-      "loss": 0.4772,
+      "epoch": 2.7291910902696364,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00020359762079332833,
+      "loss": 0.6276,
       "step": 2328
     },
     {
-      "epoch": 4.458891013384322,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.901326263179851e-05,
-      "loss": 0.4067,
+      "epoch": 2.733880422039859,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002032996625682687,
+      "loss": 0.6048,
       "step": 2332
     },
     {
-      "epoch": 4.466539196940727,
-      "grad_norm": 0.373046875,
-      "learning_rate": 8.85384300331142e-05,
-      "loss": 0.4399,
+      "epoch": 2.738569753810082,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002030014634513599,
+      "loss": 0.6189,
       "step": 2336
     },
     {
-      "epoch": 4.474187380497132,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.80643364380477e-05,
-      "loss": 0.4726,
+      "epoch": 2.7432590855803047,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00020270302479033538,
+      "loss": 0.5516,
       "step": 2340
     },
     {
-      "epoch": 4.4818355640535374,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.759098754702099e-05,
-      "loss": 0.4514,
+      "epoch": 2.7479484173505275,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020240434793401124,
+      "loss": 0.6309,
       "step": 2344
     },
     {
-      "epoch": 4.489483747609943,
-      "grad_norm": 0.36328125,
-      "learning_rate": 8.711838905150179e-05,
-      "loss": 0.4502,
+      "epoch": 2.7526377491207503,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00020210543423228,
+      "loss": 0.5694,
       "step": 2348
     },
     {
-      "epoch": 4.497131931166348,
-      "grad_norm": 0.328125,
-      "learning_rate": 8.664654663393516e-05,
-      "loss": 0.4366,
+      "epoch": 2.757327080890973,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00020180628503610484,
+      "loss": 0.6261,
       "step": 2352
     },
     {
-      "epoch": 4.504780114722753,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.617546596767534e-05,
-      "loss": 0.437,
+      "epoch": 2.762016412661196,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000201506901697513,
+      "loss": 0.6279,
       "step": 2356
     },
     {
-      "epoch": 4.512428298279159,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.570515271691723e-05,
-      "loss": 0.4313,
+      "epoch": 2.7667057444314187,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002012072855695902,
+      "loss": 0.596,
       "step": 2360
     },
     {
-      "epoch": 4.520076481835564,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.523561253662864e-05,
-      "loss": 0.447,
+      "epoch": 2.7713950762016415,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00020090743800647403,
+      "loss": 0.5867,
       "step": 2364
     },
     {
-      "epoch": 4.527724665391969,
-      "grad_norm": 0.357421875,
-      "learning_rate": 8.476685107248197e-05,
-      "loss": 0.4488,
+      "epoch": 2.776084407971864,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002006073603633483,
+      "loss": 0.6346,
       "step": 2368
     },
     {
-      "epoch": 4.5353728489483744,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.429887396078655e-05,
-      "loss": 0.469,
+      "epoch": 2.7807737397420866,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00020030705399643646,
+      "loss": 0.6264,
       "step": 2372
     },
     {
-      "epoch": 4.54302103250478,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.38316868284207e-05,
-      "loss": 0.4886,
+      "epoch": 2.7854630715123094,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020000652026299593,
+      "loss": 0.655,
       "step": 2376
     },
     {
-      "epoch": 4.550669216061186,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.336529529276421e-05,
-      "loss": 0.433,
+      "epoch": 2.7901524032825322,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001997057605213115,
+      "loss": 0.606,
       "step": 2380
     },
     {
-      "epoch": 4.558317399617591,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.289970496163085e-05,
-      "loss": 0.4029,
+      "epoch": 2.794841735052755,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019940477613068964,
+      "loss": 0.5661,
       "step": 2384
     },
     {
-      "epoch": 4.5659655831739965,
-      "grad_norm": 0.322265625,
-      "learning_rate": 8.243492143320058e-05,
-      "loss": 0.4198,
+      "epoch": 2.799531066822978,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019910356845145196,
+      "loss": 0.598,
       "step": 2388
     },
     {
-      "epoch": 4.573613766730402,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.197095029595276e-05,
-      "loss": 0.4377,
+      "epoch": 2.8042203985932006,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001988021388449293,
+      "loss": 0.6477,
       "step": 2392
     },
     {
-      "epoch": 4.581261950286807,
-      "grad_norm": 0.359375,
-      "learning_rate": 8.150779712859854e-05,
-      "loss": 0.4263,
+      "epoch": 2.808909730363423,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019850048867345554,
+      "loss": 0.6005,
       "step": 2396
     },
     {
-      "epoch": 4.588910133843212,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.104546750001402e-05,
-      "loss": 0.4784,
+      "epoch": 2.8135990621336457,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001981986193003614,
+      "loss": 0.584,
       "step": 2400
     },
     {
-      "epoch": 4.596558317399618,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.05839669691732e-05,
-      "loss": 0.4549,
+      "epoch": 2.8182883939038685,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001978965320899683,
+      "loss": 0.6111,
       "step": 2404
     },
     {
-      "epoch": 4.604206500956023,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.01233010850811e-05,
-      "loss": 0.4519,
+      "epoch": 2.8229777256740913,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019759422840758228,
+      "loss": 0.6175,
       "step": 2408
     },
     {
-      "epoch": 4.611854684512428,
-      "grad_norm": 0.37109375,
-      "learning_rate": 7.966347538670712e-05,
-      "loss": 0.4242,
+      "epoch": 2.827667057444314,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019729170961948754,
+      "loss": 0.6083,
       "step": 2412
     },
     {
-      "epoch": 4.6195028680688335,
-      "grad_norm": 0.333984375,
-      "learning_rate": 7.92044954029184e-05,
-      "loss": 0.4647,
+      "epoch": 2.832356389214537,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019698897709294062,
+      "loss": 0.5866,
       "step": 2416
     },
     {
-      "epoch": 4.627151051625239,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.874636665241335e-05,
-      "loss": 0.46,
+      "epoch": 2.8370457209847597,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000196686032196164,
+      "loss": 0.5791,
       "step": 2420
     },
     {
-      "epoch": 4.634799235181644,
-      "grad_norm": 0.34375,
-      "learning_rate": 7.828909464365531e-05,
-      "loss": 0.4169,
+      "epoch": 2.8417350527549825,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019638287629834012,
+      "loss": 0.6533,
       "step": 2424
     },
     {
-      "epoch": 4.642447418738049,
-      "grad_norm": 0.349609375,
-      "learning_rate": 7.783268487480626e-05,
-      "loss": 0.4251,
+      "epoch": 2.8464243845252053,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001960795107696048,
+      "loss": 0.5994,
       "step": 2428
     },
     {
-      "epoch": 4.650095602294455,
-      "grad_norm": 0.337890625,
-      "learning_rate": 7.73771428336608e-05,
-      "loss": 0.4232,
+      "epoch": 2.851113716295428,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019577593698104156,
+      "loss": 0.5925,
       "step": 2432
     },
     {
-      "epoch": 4.657743785850861,
-      "grad_norm": 0.3671875,
-      "learning_rate": 7.692247399758008e-05,
-      "loss": 0.4836,
+      "epoch": 2.855803048065651,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019547215630467504,
+      "loss": 0.5999,
       "step": 2436
     },
     {
-      "epoch": 4.665391969407266,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.6468683833426e-05,
-      "loss": 0.3917,
+      "epoch": 2.8604923798358732,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019516817011346494,
+      "loss": 0.6372,
       "step": 2440
     },
     {
-      "epoch": 4.673040152963671,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.601577779749545e-05,
-      "loss": 0.4153,
+      "epoch": 2.865181711606096,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019486397978129977,
+      "loss": 0.5615,
       "step": 2444
     },
     {
-      "epoch": 4.680688336520077,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.55637613354547e-05,
-      "loss": 0.4103,
+      "epoch": 2.869871043376319,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019455958668299075,
+      "loss": 0.5898,
       "step": 2448
     },
     {
-      "epoch": 4.688336520076482,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.511263988227397e-05,
-      "loss": 0.4858,
+      "epoch": 2.8745603751465416,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001942549921942654,
+      "loss": 0.6083,
       "step": 2452
     },
     {
-      "epoch": 4.695984703632887,
-      "grad_norm": 0.369140625,
-      "learning_rate": 7.466241886216198e-05,
-      "loss": 0.468,
+      "epoch": 2.8792497069167644,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019395019769176156,
+      "loss": 0.6447,
       "step": 2456
     },
     {
-      "epoch": 4.7036328871892925,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.421310368850085e-05,
-      "loss": 0.4817,
+      "epoch": 2.883939038686987,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019364520455302103,
+      "loss": 0.6103,
       "step": 2460
     },
     {
-      "epoch": 4.711281070745698,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.376469976378094e-05,
-      "loss": 0.3923,
+      "epoch": 2.88862837045721,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001933400141564833,
+      "loss": 0.5962,
       "step": 2464
     },
     {
-      "epoch": 4.718929254302103,
-      "grad_norm": 0.337890625,
-      "learning_rate": 7.33172124795359e-05,
-      "loss": 0.3954,
+      "epoch": 2.8933177022274323,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019303462788147933,
+      "loss": 0.636,
       "step": 2468
     },
     {
-      "epoch": 4.726577437858508,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.287064721627782e-05,
-      "loss": 0.4606,
+      "epoch": 2.898007033997655,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019272904710822551,
+      "loss": 0.5804,
       "step": 2472
     },
     {
-      "epoch": 4.734225621414914,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.242500934343262e-05,
-      "loss": 0.4253,
+      "epoch": 2.902696365767878,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019242327321781726,
+      "loss": 0.6356,
       "step": 2476
     },
     {
-      "epoch": 4.741873804971319,
-      "grad_norm": 0.3671875,
-      "learning_rate": 7.19803042192754e-05,
-      "loss": 0.4431,
+      "epoch": 2.9073856975381007,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019211730759222272,
+      "loss": 0.574,
       "step": 2480
     },
     {
-      "epoch": 4.749521988527725,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.153653719086604e-05,
-      "loss": 0.4097,
+      "epoch": 2.9120750293083235,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019181115161427662,
+      "loss": 0.6377,
       "step": 2484
     },
     {
-      "epoch": 4.75717017208413,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.109371359398493e-05,
-      "loss": 0.4208,
+      "epoch": 2.9167643610785463,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019150480666767407,
+      "loss": 0.6016,
       "step": 2488
     },
     {
-      "epoch": 4.764818355640536,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.06518387530688e-05,
-      "loss": 0.4396,
+      "epoch": 2.921453692848769,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001911982741369641,
+      "loss": 0.5675,
       "step": 2492
     },
     {
-      "epoch": 4.772466539196941,
-      "grad_norm": 0.328125,
-      "learning_rate": 7.021091798114667e-05,
-      "loss": 0.4212,
+      "epoch": 2.926143024618992,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001908915554075437,
+      "loss": 0.5886,
       "step": 2496
     },
     {
-      "epoch": 4.780114722753346,
-      "grad_norm": 0.33203125,
-      "learning_rate": 6.977095657977603e-05,
-      "loss": 0.4742,
+      "epoch": 2.9308323563892147,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019058465186565132,
+      "loss": 0.6028,
       "step": 2500
     },
     {
-      "epoch": 4.7877629063097515,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.933195983897905e-05,
-      "loss": 0.4198,
+      "epoch": 2.9355216881594375,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001902775648983606,
+      "loss": 0.5958,
       "step": 2504
     },
     {
-      "epoch": 4.795411089866157,
-      "grad_norm": 0.361328125,
-      "learning_rate": 6.889393303717898e-05,
-      "loss": 0.4401,
+      "epoch": 2.9402110199296603,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018997029589357443,
+      "loss": 0.5895,
       "step": 2508
     },
     {
-      "epoch": 4.803059273422562,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.845688144113663e-05,
-      "loss": 0.4273,
+      "epoch": 2.9449003516998826,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018966284624001814,
+      "loss": 0.6345,
       "step": 2512
     },
     {
-      "epoch": 4.810707456978967,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.802081030588722e-05,
-      "loss": 0.4589,
+      "epoch": 2.9495896834701054,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018935521732723376,
+      "loss": 0.5931,
       "step": 2516
     },
     {
-      "epoch": 4.818355640535373,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.758572487467698e-05,
-      "loss": 0.4835,
+      "epoch": 2.954279015240328,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018904741054557325,
+      "loss": 0.6009,
       "step": 2520
     },
     {
-      "epoch": 4.826003824091778,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.715163037890021e-05,
-      "loss": 0.4493,
+      "epoch": 2.958968347010551,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00018873942728619273,
+      "loss": 0.6011,
       "step": 2524
     },
     {
-      "epoch": 4.833652007648183,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.671853203803641e-05,
-      "loss": 0.4032,
+      "epoch": 2.963657678780774,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018843126894104573,
+      "loss": 0.5769,
       "step": 2528
     },
     {
-      "epoch": 4.8413001912045885,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.628643505958742e-05,
-      "loss": 0.4498,
+      "epoch": 2.9683470105509966,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018812293690287715,
+      "loss": 0.5863,
       "step": 2532
     },
     {
-      "epoch": 4.848948374760994,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.585534463901493e-05,
-      "loss": 0.4255,
+      "epoch": 2.9730363423212194,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018781443256521695,
+      "loss": 0.5778,
       "step": 2536
     },
     {
-      "epoch": 4.8565965583174,
-      "grad_norm": 0.388671875,
-      "learning_rate": 6.542526595967795e-05,
-      "loss": 0.4285,
+      "epoch": 2.9777256740914417,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018750575732237379,
+      "loss": 0.6134,
       "step": 2540
     },
     {
-      "epoch": 4.864244741873805,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.499620419277036e-05,
-      "loss": 0.4507,
+      "epoch": 2.9824150058616645,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018719691256942868,
+      "loss": 0.558,
       "step": 2544
     },
     {
-      "epoch": 4.871892925430211,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.456816449725892e-05,
-      "loss": 0.4196,
+      "epoch": 2.9871043376318873,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018688789970222882,
+      "loss": 0.5734,
       "step": 2548
     },
     {
-      "epoch": 4.879541108986616,
-      "grad_norm": 0.345703125,
-      "learning_rate": 6.414115201982134e-05,
-      "loss": 0.4061,
+      "epoch": 2.99179366940211,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018657872011738124,
+      "loss": 0.5789,
       "step": 2552
     },
     {
-      "epoch": 4.887189292543021,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.371517189478403e-05,
-      "loss": 0.4199,
+      "epoch": 2.996483001172333,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001862693752122463,
+      "loss": 0.6177,
       "step": 2556
     },
     {
-      "epoch": 4.894837476099426,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.329022924406061e-05,
-      "loss": 0.452,
+      "epoch": 3.0011723329425557,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001859598663849318,
+      "loss": 0.5135,
       "step": 2560
     },
     {
-      "epoch": 4.902485659655832,
-      "grad_norm": 0.345703125,
-      "learning_rate": 6.286632917709031e-05,
-      "loss": 0.4516,
+      "epoch": 3.0058616647127785,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018565019503428618,
+      "loss": 0.5306,
       "step": 2564
     },
     {
-      "epoch": 4.910133843212237,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.244347679077651e-05,
-      "loss": 0.4622,
+      "epoch": 3.0105509964830013,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018534036255989247,
+      "loss": 0.5166,
       "step": 2568
     },
     {
-      "epoch": 4.917782026768642,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.202167716942543e-05,
-      "loss": 0.4387,
+      "epoch": 3.015240328253224,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018503037036206194,
+      "loss": 0.5773,
       "step": 2572
     },
     {
-      "epoch": 4.925430210325048,
-      "grad_norm": 0.35546875,
-      "learning_rate": 6.160093538468505e-05,
-      "loss": 0.4064,
+      "epoch": 3.019929660023447,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018472021984182777,
+      "loss": 0.5381,
       "step": 2576
     },
     {
-      "epoch": 4.933078393881453,
-      "grad_norm": 0.3671875,
-      "learning_rate": 6.118125649548405e-05,
-      "loss": 0.4728,
+      "epoch": 3.024618991793669,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018440991240093862,
+      "loss": 0.5158,
       "step": 2580
     },
     {
-      "epoch": 4.940726577437858,
-      "grad_norm": 0.35546875,
-      "learning_rate": 6.076264554797112e-05,
-      "loss": 0.4555,
+      "epoch": 3.029308323563892,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018409944944185237,
+      "loss": 0.5442,
       "step": 2584
     },
     {
-      "epoch": 4.948374760994264,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.0345107575454105e-05,
-      "loss": 0.427,
+      "epoch": 3.033997655334115,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001837888323677299,
+      "loss": 0.5519,
       "step": 2588
     },
     {
-      "epoch": 4.95602294455067,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.992864759833963e-05,
-      "loss": 0.3841,
+      "epoch": 3.0386869871043376,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001834780625824285,
+      "loss": 0.525,
       "step": 2592
     },
     {
-      "epoch": 4.963671128107075,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.9513270624072655e-05,
-      "loss": 0.4289,
+      "epoch": 3.0433763188745604,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001831671414904956,
+      "loss": 0.5157,
       "step": 2596
     },
     {
-      "epoch": 4.97131931166348,
-      "grad_norm": 0.353515625,
-      "learning_rate": 5.9098981647076345e-05,
-      "loss": 0.4538,
+      "epoch": 3.048065650644783,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018285607049716256,
+      "loss": 0.5142,
       "step": 2600
     },
     {
-      "epoch": 4.9789674952198855,
-      "grad_norm": 0.33203125,
-      "learning_rate": 5.8685785648691894e-05,
-      "loss": 0.3954,
+      "epoch": 3.052754982415006,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001825448510083383,
+      "loss": 0.4794,
       "step": 2604
     },
     {
-      "epoch": 4.986615678776291,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.8273687597118765e-05,
-      "loss": 0.4614,
+      "epoch": 3.0574443141852288,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018223348443060274,
+      "loss": 0.4835,
       "step": 2608
     },
     {
-      "epoch": 4.994263862332696,
-      "grad_norm": 0.3359375,
-      "learning_rate": 5.786269244735488e-05,
-      "loss": 0.4429,
+      "epoch": 3.0621336459554516,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018192197217120067,
+      "loss": 0.4776,
       "step": 2612
     },
     {
-      "epoch": 5.001912045889101,
-      "grad_norm": 0.30859375,
-      "learning_rate": 5.7452805141137034e-05,
-      "loss": 0.3411,
+      "epoch": 3.066822977725674,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018161031563803523,
+      "loss": 0.5137,
       "step": 2616
     },
     {
-      "epoch": 5.009560229445507,
-      "grad_norm": 0.318359375,
-      "learning_rate": 5.704403060688158e-05,
-      "loss": 0.3973,
+      "epoch": 3.0715123094958967,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018129851623966168,
+      "loss": 0.5287,
       "step": 2620
     },
     {
-      "epoch": 5.017208413001912,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.663637375962489e-05,
-      "loss": 0.4319,
+      "epoch": 3.0762016412661195,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001809865753852809,
+      "loss": 0.5198,
       "step": 2624
     },
     {
-      "epoch": 5.024856596558317,
-      "grad_norm": 0.3203125,
-      "learning_rate": 5.6229839500964635e-05,
-      "loss": 0.3768,
+      "epoch": 3.0808909730363423,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018067449448473321,
+      "loss": 0.4759,
       "step": 2628
     },
     {
-      "epoch": 5.0325047801147225,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.582443271900063e-05,
-      "loss": 0.3683,
+      "epoch": 3.085580304806565,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018036227494849173,
+      "loss": 0.4228,
       "step": 2632
     },
     {
-      "epoch": 5.040152963671128,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.542015828827609e-05,
-      "loss": 0.4145,
+      "epoch": 3.090269636576788,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018004991818765625,
+      "loss": 0.5385,
       "step": 2636
     },
     {
-      "epoch": 5.047801147227533,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.5017021069719014e-05,
-      "loss": 0.4116,
+      "epoch": 3.0949589683470107,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017973742561394675,
+      "loss": 0.5078,
       "step": 2640
     },
     {
-      "epoch": 5.055449330783939,
-      "grad_norm": 0.3359375,
-      "learning_rate": 5.4615025910583756e-05,
-      "loss": 0.4069,
+      "epoch": 3.0996483001172335,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000179424798639697,
+      "loss": 0.4977,
       "step": 2644
     },
     {
-      "epoch": 5.0630975143403445,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.421417764439276e-05,
-      "loss": 0.375,
+      "epoch": 3.1043376318874563,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017911203867784819,
+      "loss": 0.5108,
       "step": 2648
     },
     {
-      "epoch": 5.07074569789675,
-      "grad_norm": 0.333984375,
-      "learning_rate": 5.3814481090878374e-05,
-      "loss": 0.3576,
+      "epoch": 3.1090269636576786,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001787991471419426,
+      "loss": 0.5151,
       "step": 2652
     },
     {
-      "epoch": 5.078393881453155,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.3415941055924974e-05,
-      "loss": 0.4027,
+      "epoch": 3.1137162954279014,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017848612544611714,
+      "loss": 0.5332,
       "step": 2656
     },
     {
-      "epoch": 5.08604206500956,
-      "grad_norm": 0.328125,
-      "learning_rate": 5.301856233151123e-05,
-      "loss": 0.3981,
+      "epoch": 3.118405627198124,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017817297500509702,
+      "loss": 0.5177,
       "step": 2660
     },
     {
-      "epoch": 5.093690248565966,
-      "grad_norm": 0.318359375,
-      "learning_rate": 5.262234969565226e-05,
-      "loss": 0.3928,
+      "epoch": 3.123094958968347,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001778596972341893,
+      "loss": 0.5166,
       "step": 2664
     },
     {
-      "epoch": 5.101338432122371,
-      "grad_norm": 0.33203125,
-      "learning_rate": 5.222730791234246e-05,
-      "loss": 0.3813,
+      "epoch": 3.12778429073857,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017754629354927655,
+      "loss": 0.4979,
       "step": 2668
     },
     {
-      "epoch": 5.108986615678776,
-      "grad_norm": 0.3671875,
-      "learning_rate": 5.183344173149798e-05,
-      "loss": 0.4151,
+      "epoch": 3.1324736225087926,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017723276536681025,
+      "loss": 0.5008,
       "step": 2672
     },
     {
-      "epoch": 5.1166347992351815,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.14407558888998e-05,
-      "loss": 0.4051,
+      "epoch": 3.1371629542790154,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017691911410380485,
+      "loss": 0.5153,
       "step": 2676
     },
     {
-      "epoch": 5.124282982791587,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.104925510613668e-05,
-      "loss": 0.3973,
+      "epoch": 3.141852286049238,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017660534117783084,
+      "loss": 0.4782,
       "step": 2680
     },
     {
-      "epoch": 5.131931166347992,
-      "grad_norm": 0.3203125,
-      "learning_rate": 5.0658944090548436e-05,
-      "loss": 0.3958,
+      "epoch": 3.1465416178194605,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017629144800700866,
+      "loss": 0.4542,
       "step": 2684
     },
     {
-      "epoch": 5.139579349904397,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.0269827535169306e-05,
-      "loss": 0.3731,
+      "epoch": 3.1512309495896833,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017597743601000218,
+      "loss": 0.5261,
       "step": 2688
     },
     {
-      "epoch": 5.147227533460803,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.988191011867153e-05,
-      "loss": 0.4028,
+      "epoch": 3.155920281359906,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00017566330660601236,
+      "loss": 0.4949,
       "step": 2692
     },
     {
-      "epoch": 5.154875717017209,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.9495196505309196e-05,
-      "loss": 0.4201,
+      "epoch": 3.160609613130129,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001753490612147707,
+      "loss": 0.5243,
       "step": 2696
     },
     {
-      "epoch": 5.162523900573614,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.9109691344861886e-05,
-      "loss": 0.3815,
+      "epoch": 3.1652989449003517,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017503470125653309,
+      "loss": 0.5316,
       "step": 2700
     },
     {
-      "epoch": 5.170172084130019,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.8725399272579075e-05,
-      "loss": 0.4043,
+      "epoch": 3.1699882766705745,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017472022815207295,
+      "loss": 0.5134,
       "step": 2704
     },
     {
-      "epoch": 5.177820267686425,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.8342324909124256e-05,
-      "loss": 0.3828,
+      "epoch": 3.1746776084407973,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001744056433226753,
+      "loss": 0.543,
       "step": 2708
     },
     {
-      "epoch": 5.18546845124283,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.7960472860519365e-05,
-      "loss": 0.4199,
+      "epoch": 3.17936694021102,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017409094819013002,
+      "loss": 0.5009,
       "step": 2712
     },
     {
-      "epoch": 5.193116634799235,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.757984771808947e-05,
-      "loss": 0.4071,
+      "epoch": 3.184056271981243,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00017377614417672554,
+      "loss": 0.5103,
       "step": 2716
     },
     {
-      "epoch": 5.2007648183556405,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.72004540584075e-05,
-      "loss": 0.4288,
+      "epoch": 3.1887456037514657,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001734612327052423,
+      "loss": 0.508,
       "step": 2720
     },
     {
-      "epoch": 5.208413001912046,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.682229644323922e-05,
-      "loss": 0.3935,
+      "epoch": 3.193434935521688,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017314621519894652,
+      "loss": 0.4487,
       "step": 2724
     },
     {
-      "epoch": 5.216061185468451,
-      "grad_norm": 0.37109375,
-      "learning_rate": 4.6445379419488436e-05,
-      "loss": 0.3801,
+      "epoch": 3.198124267291911,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017283109308158362,
+      "loss": 0.5308,
       "step": 2728
     },
     {
-      "epoch": 5.223709369024856,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.606970751914229e-05,
-      "loss": 0.4181,
+      "epoch": 3.2028135990621336,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017251586777737175,
+      "loss": 0.5029,
       "step": 2732
     },
     {
-      "epoch": 5.231357552581262,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.569528525921672e-05,
-      "loss": 0.3816,
+      "epoch": 3.2075029308323564,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017220054071099555,
+      "loss": 0.5274,
       "step": 2736
     },
     {
-      "epoch": 5.239005736137667,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.532211714170229e-05,
-      "loss": 0.4331,
+      "epoch": 3.212192262602579,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001718851133075994,
+      "loss": 0.5138,
       "step": 2740
     },
     {
-      "epoch": 5.246653919694072,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.495020765350988e-05,
-      "loss": 0.4331,
+      "epoch": 3.216881594372802,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017156958699278134,
+      "loss": 0.5345,
       "step": 2744
     },
     {
-      "epoch": 5.254302103250478,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.4579561266416855e-05,
-      "loss": 0.3889,
+      "epoch": 3.2215709261430248,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017125396319258635,
+      "loss": 0.5127,
       "step": 2748
     },
     {
-      "epoch": 5.261950286806884,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.421018243701327e-05,
-      "loss": 0.3804,
+      "epoch": 3.2262602579132476,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001709382433335,
+      "loss": 0.4827,
       "step": 2752
     },
     {
-      "epoch": 5.269598470363289,
-      "grad_norm": 0.55859375,
-      "learning_rate": 4.384207560664825e-05,
-      "loss": 0.4386,
+      "epoch": 3.23094958968347,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017062242884244213,
+      "loss": 0.4796,
       "step": 2756
     },
     {
-      "epoch": 5.277246653919694,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.347524520137667e-05,
-      "loss": 0.3547,
+      "epoch": 3.2356389214536927,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017030652114676003,
+      "loss": 0.4927,
       "step": 2760
     },
     {
-      "epoch": 5.2848948374761,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.310969563190578e-05,
-      "loss": 0.3719,
+      "epoch": 3.2403282532239155,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016999052167422247,
+      "loss": 0.5091,
       "step": 2764
     },
     {
-      "epoch": 5.292543021032505,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.274543129354245e-05,
-      "loss": 0.3611,
+      "epoch": 3.2450175849941383,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016967443185301293,
+      "loss": 0.5312,
       "step": 2768
     },
     {
-      "epoch": 5.30019120458891,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.2382456566139985e-05,
-      "loss": 0.3173,
+      "epoch": 3.249706916764361,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016935825311172322,
+      "loss": 0.5339,
       "step": 2772
     },
     {
-      "epoch": 5.307839388145315,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.202077581404574e-05,
-      "loss": 0.3848,
+      "epoch": 3.254396248534584,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016904198687934697,
+      "loss": 0.5263,
       "step": 2776
     },
     {
-      "epoch": 5.315487571701721,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.166039338604838e-05,
-      "loss": 0.3921,
+      "epoch": 3.2590855803048067,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016872563458527332,
+      "loss": 0.4754,
       "step": 2780
     },
     {
-      "epoch": 5.323135755258126,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.130131361532586e-05,
-      "loss": 0.4049,
+      "epoch": 3.2637749120750295,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001684091976592804,
+      "loss": 0.5075,
       "step": 2784
     },
     {
-      "epoch": 5.330783938814531,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.094354081939317e-05,
-      "loss": 0.3866,
+      "epoch": 3.2684642438452522,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016809267753152871,
+      "loss": 0.5461,
       "step": 2788
     },
     {
-      "epoch": 5.338432122370937,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.058707930005048e-05,
-      "loss": 0.3727,
+      "epoch": 3.273153575615475,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016777607563255498,
+      "loss": 0.5318,
       "step": 2792
     },
     {
-      "epoch": 5.346080305927342,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.023193334333132e-05,
-      "loss": 0.3957,
+      "epoch": 3.2778429073856974,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016745939339326532,
+      "loss": 0.5415,
       "step": 2796
     },
     {
-      "epoch": 5.353728489483748,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.9878107219451206e-05,
-      "loss": 0.344,
+      "epoch": 3.28253223915592,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000167142632244929,
+      "loss": 0.5451,
       "step": 2800
     },
     {
-      "epoch": 5.361376673040153,
-      "grad_norm": 0.375,
-      "learning_rate": 3.9525605182756134e-05,
-      "loss": 0.4048,
+      "epoch": 3.287221570926143,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016682579361917196,
+      "loss": 0.5004,
       "step": 2804
     },
     {
-      "epoch": 5.369024856596559,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.917443147167152e-05,
-      "loss": 0.4343,
+      "epoch": 3.2919109026963658,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016650887894797029,
+      "loss": 0.5172,
       "step": 2808
     },
     {
-      "epoch": 5.376673040152964,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.882459030865124e-05,
-      "loss": 0.3908,
+      "epoch": 3.2966002344665886,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016619188966364383,
+      "loss": 0.5349,
       "step": 2812
     },
     {
-      "epoch": 5.384321223709369,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.8476085900126776e-05,
-      "loss": 0.3491,
+      "epoch": 3.3012895662368114,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001658748271988495,
+      "loss": 0.5499,
       "step": 2816
     },
     {
-      "epoch": 5.3919694072657744,
-      "grad_norm": 0.3515625,
-      "learning_rate": 3.8128922436456766e-05,
-      "loss": 0.4092,
+      "epoch": 3.305978898007034,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016555769298657515,
+      "loss": 0.5422,
       "step": 2820
     },
     {
-      "epoch": 5.39961759082218,
-      "grad_norm": 0.375,
-      "learning_rate": 3.7783104091876524e-05,
-      "loss": 0.3766,
+      "epoch": 3.3106682297772565,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016524048846013265,
+      "loss": 0.5366,
       "step": 2824
     },
     {
-      "epoch": 5.407265774378585,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.743863502444783e-05,
-      "loss": 0.3835,
+      "epoch": 3.3153575615474793,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016492321505315194,
+      "loss": 0.5024,
       "step": 2828
     },
     {
-      "epoch": 5.41491395793499,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.709551937600909e-05,
-      "loss": 0.3733,
+      "epoch": 3.320046893317702,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016460587419957407,
+      "loss": 0.5025,
       "step": 2832
     },
     {
-      "epoch": 5.422562141491396,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.675376127212532e-05,
-      "loss": 0.4127,
+      "epoch": 3.324736225087925,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016428846733364502,
+      "loss": 0.4966,
       "step": 2836
     },
     {
-      "epoch": 5.430210325047801,
-      "grad_norm": 0.361328125,
-      "learning_rate": 3.64133648220387e-05,
-      "loss": 0.4091,
+      "epoch": 3.3294255568581477,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016397099588990902,
+      "loss": 0.5281,
       "step": 2840
     },
     {
-      "epoch": 5.437858508604206,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.607433411861912e-05,
-      "loss": 0.4612,
+      "epoch": 3.3341148886283705,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016365346130320233,
+      "loss": 0.4967,
       "step": 2844
     },
     {
-      "epoch": 5.4455066921606115,
-      "grad_norm": 0.330078125,
-      "learning_rate": 3.5736673238314914e-05,
-      "loss": 0.3668,
+      "epoch": 3.3388042203985933,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016333586500864647,
+      "loss": 0.4664,
       "step": 2848
     },
     {
-      "epoch": 5.453154875717018,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.5400386241103946e-05,
-      "loss": 0.3645,
+      "epoch": 3.343493552168816,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016301820844164176,
+      "loss": 0.5526,
       "step": 2852
     },
     {
-      "epoch": 5.460803059273423,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.506547717044472e-05,
-      "loss": 0.348,
+      "epoch": 3.348182883939039,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016270049303786113,
+      "loss": 0.5298,
       "step": 2856
     },
     {
-      "epoch": 5.468451242829828,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.473195005322776e-05,
-      "loss": 0.3812,
+      "epoch": 3.3528722157092616,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001623827202332433,
+      "loss": 0.5579,
       "step": 2860
     },
     {
-      "epoch": 5.4760994263862335,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.439980889972723e-05,
-      "loss": 0.3705,
+      "epoch": 3.357561547479484,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016206489146398655,
+      "loss": 0.5597,
       "step": 2864
     },
     {
-      "epoch": 5.483747609942639,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.406905770355274e-05,
-      "loss": 0.3687,
+      "epoch": 3.362250879249707,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001617470081665419,
+      "loss": 0.5362,
       "step": 2868
     },
     {
-      "epoch": 5.491395793499044,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.373970044160121e-05,
-      "loss": 0.4348,
+      "epoch": 3.3669402110199296,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001614290717776069,
+      "loss": 0.5342,
       "step": 2872
     },
     {
-      "epoch": 5.499043977055449,
-      "grad_norm": 0.36328125,
-      "learning_rate": 3.341174107400916e-05,
-      "loss": 0.4039,
+      "epoch": 3.3716295427901524,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001611110837341191,
+      "loss": 0.4736,
       "step": 2876
     },
     {
-      "epoch": 5.506692160611855,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.30851835441051e-05,
-      "loss": 0.4072,
+      "epoch": 3.376318874560375,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001607930454732495,
+      "loss": 0.5421,
       "step": 2880
     },
     {
-      "epoch": 5.51434034416826,
-      "grad_norm": 0.353515625,
-      "learning_rate": 3.276003177836203e-05,
-      "loss": 0.3796,
+      "epoch": 3.381008206330598,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000160474958432396,
+      "loss": 0.4857,
       "step": 2884
     },
     {
-      "epoch": 5.521988527724665,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.2436289686350285e-05,
-      "loss": 0.3647,
+      "epoch": 3.3856975381008207,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000160156824049177,
+      "loss": 0.5453,
       "step": 2888
     },
     {
-      "epoch": 5.5296367112810705,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.211396116069055e-05,
-      "loss": 0.3895,
+      "epoch": 3.3903868698710435,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015983864376142482,
+      "loss": 0.5494,
       "step": 2892
     },
     {
-      "epoch": 5.537284894837476,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.179305007700697e-05,
-      "loss": 0.3689,
+      "epoch": 3.395076201641266,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001595204190071794,
+      "loss": 0.5328,
       "step": 2896
     },
     {
-      "epoch": 5.544933078393882,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.147356029388067e-05,
-      "loss": 0.3782,
+      "epoch": 3.3997655334114887,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015920215122468146,
+      "loss": 0.4856,
       "step": 2900
     },
     {
-      "epoch": 5.552581261950287,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.115549565280325e-05,
-      "loss": 0.3985,
+      "epoch": 3.4044548651817115,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015888384185236632,
+      "loss": 0.4919,
       "step": 2904
     },
     {
-      "epoch": 5.5602294455066925,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.083885997813066e-05,
-      "loss": 0.4289,
+      "epoch": 3.4091441969519343,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015856549232885712,
+      "loss": 0.5082,
       "step": 2908
     },
     {
-      "epoch": 5.567877629063098,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.052365707703718e-05,
-      "loss": 0.436,
+      "epoch": 3.413833528722157,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015824710409295868,
+      "loss": 0.4919,
       "step": 2912
     },
     {
-      "epoch": 5.575525812619503,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.0209890739469693e-05,
-      "loss": 0.387,
+      "epoch": 3.41852286049238,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001579286785836506,
+      "loss": 0.5208,
       "step": 2916
     },
     {
-      "epoch": 5.583173996175908,
-      "grad_norm": 0.326171875,
-      "learning_rate": 2.989756473810203e-05,
-      "loss": 0.4034,
+      "epoch": 3.4232121922626026,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001576102172400811,
+      "loss": 0.5251,
       "step": 2920
     },
     {
-      "epoch": 5.590822179732314,
-      "grad_norm": 0.3671875,
-      "learning_rate": 2.9586682828289738e-05,
-      "loss": 0.4206,
+      "epoch": 3.4279015240328254,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001572917215015602,
+      "loss": 0.4927,
       "step": 2924
     },
     {
-      "epoch": 5.598470363288719,
-      "grad_norm": 0.36328125,
-      "learning_rate": 2.9277248748024763e-05,
-      "loss": 0.4529,
+      "epoch": 3.4325908558030482,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015697319280755343,
+      "loss": 0.5404,
       "step": 2928
     },
     {
-      "epoch": 5.606118546845124,
-      "grad_norm": 0.330078125,
-      "learning_rate": 2.8969266217890648e-05,
-      "loss": 0.3527,
+      "epoch": 3.437280187573271,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015665463259767525,
+      "loss": 0.5198,
       "step": 2932
     },
     {
-      "epoch": 5.6137667304015295,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.866273894101776e-05,
-      "loss": 0.389,
+      "epoch": 3.4419695193434934,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015633604231168264,
+      "loss": 0.5321,
       "step": 2936
     },
     {
-      "epoch": 5.621414913957935,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.835767060303865e-05,
-      "loss": 0.4017,
+      "epoch": 3.446658851113716,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015601742338946844,
+      "loss": 0.5492,
       "step": 2940
     },
     {
-      "epoch": 5.62906309751434,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.8054064872043917e-05,
-      "loss": 0.3973,
+      "epoch": 3.451348182883939,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015569877727105493,
+      "loss": 0.5115,
       "step": 2944
     },
     {
-      "epoch": 5.636711281070745,
-      "grad_norm": 0.375,
-      "learning_rate": 2.7751925398537993e-05,
-      "loss": 0.4281,
+      "epoch": 3.4560375146541618,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015538010539658728,
+      "loss": 0.4897,
       "step": 2948
     },
     {
-      "epoch": 5.644359464627151,
-      "grad_norm": 0.330078125,
-      "learning_rate": 2.745125581539523e-05,
-      "loss": 0.4071,
+      "epoch": 3.4607268464243846,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015506140920632707,
+      "loss": 0.5417,
       "step": 2952
     },
     {
-      "epoch": 5.652007648183556,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.7152059737816395e-05,
-      "loss": 0.3866,
+      "epoch": 3.4654161781946073,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001547426901406458,
+      "loss": 0.5028,
       "step": 2956
     },
     {
-      "epoch": 5.659655831739962,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.6854340763284954e-05,
-      "loss": 0.4029,
+      "epoch": 3.47010550996483,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015442394964001842,
+      "loss": 0.5215,
       "step": 2960
     },
     {
-      "epoch": 5.667304015296367,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.6558102471523975e-05,
-      "loss": 0.4207,
+      "epoch": 3.474794841735053,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015410518914501655,
+      "loss": 0.5129,
       "step": 2964
     },
     {
-      "epoch": 5.674952198852773,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.6263348424453012e-05,
-      "loss": 0.3769,
+      "epoch": 3.4794841735052753,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015378641009630242,
+      "loss": 0.5582,
       "step": 2968
     },
     {
-      "epoch": 5.682600382409178,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.597008216614534e-05,
-      "loss": 0.3527,
+      "epoch": 3.484173505275498,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.000153467613934622,
+      "loss": 0.5316,
       "step": 2972
     },
     {
-      "epoch": 5.690248565965583,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.5678307222785315e-05,
-      "loss": 0.4091,
+      "epoch": 3.488862837045721,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015314880210079863,
+      "loss": 0.5291,
       "step": 2976
     },
     {
-      "epoch": 5.6978967495219885,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.5388027102625945e-05,
-      "loss": 0.4175,
+      "epoch": 3.4935521688159437,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015282997603572639,
+      "loss": 0.486,
       "step": 2980
     },
     {
-      "epoch": 5.705544933078394,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.5099245295946764e-05,
-      "loss": 0.3557,
+      "epoch": 3.4982415005861665,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015251113718036378,
+      "loss": 0.5083,
       "step": 2984
     },
     {
-      "epoch": 5.713193116634799,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.4811965275011825e-05,
-      "loss": 0.4021,
+      "epoch": 3.5029308323563892,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001521922869757271,
+      "loss": 0.521,
       "step": 2988
     },
     {
-      "epoch": 5.720841300191204,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.4526190494027953e-05,
-      "loss": 0.3868,
+      "epoch": 3.507620164126612,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001518734268628839,
+      "loss": 0.5537,
       "step": 2992
     },
     {
-      "epoch": 5.72848948374761,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.4241924389103227e-05,
-      "loss": 0.4399,
+      "epoch": 3.512309495896835,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015155455828294657,
+      "loss": 0.4843,
       "step": 2996
     },
     {
-      "epoch": 5.736137667304015,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.395917037820566e-05,
-      "loss": 0.3907,
+      "epoch": 3.5169988276670576,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015123568267706575,
+      "loss": 0.5133,
       "step": 3000
     },
     {
-      "epoch": 5.743785850860421,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.3677931861122084e-05,
-      "loss": 0.4195,
+      "epoch": 3.5216881594372804,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015091680148642371,
+      "loss": 0.4921,
       "step": 3004
     },
     {
-      "epoch": 5.751434034416826,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.339821221941731e-05,
-      "loss": 0.3867,
+      "epoch": 3.5263774912075028,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015059791615222817,
+      "loss": 0.5169,
       "step": 3008
     },
     {
-      "epoch": 5.759082217973232,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.312001481639348e-05,
-      "loss": 0.3583,
+      "epoch": 3.5310668229777256,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015027902811570544,
+      "loss": 0.5067,
       "step": 3012
     },
     {
-      "epoch": 5.766730401529637,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.2843342997049445e-05,
-      "loss": 0.3527,
+      "epoch": 3.5357561547479484,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00014996013881809402,
+      "loss": 0.5611,
       "step": 3016
     },
     {
-      "epoch": 5.774378585086042,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.2568200088040867e-05,
-      "loss": 0.3393,
+      "epoch": 3.540445486518171,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014964124970063829,
+      "loss": 0.5201,
       "step": 3020
     },
     {
-      "epoch": 5.782026768642448,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.2294589397639978e-05,
-      "loss": 0.4225,
+      "epoch": 3.545134818288394,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001493223622045816,
+      "loss": 0.5519,
       "step": 3024
     },
     {
-      "epoch": 5.789674952198853,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.2022514215695842e-05,
-      "loss": 0.4191,
+      "epoch": 3.5498241500586167,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014900347777116007,
+      "loss": 0.5112,
       "step": 3028
     },
     {
-      "epoch": 5.797323135755258,
-      "grad_norm": 0.341796875,
-      "learning_rate": 2.175197781359485e-05,
-      "loss": 0.3792,
+      "epoch": 3.5545134818288395,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014868459784159603,
+      "loss": 0.5378,
       "step": 3032
     },
     {
-      "epoch": 5.804971319311663,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.1482983444221402e-05,
-      "loss": 0.3942,
+      "epoch": 3.559202813599062,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001483657238570913,
+      "loss": 0.4843,
       "step": 3036
     },
     {
-      "epoch": 5.812619502868069,
-      "grad_norm": 0.326171875,
-      "learning_rate": 2.1215534341918707e-05,
-      "loss": 0.3753,
+      "epoch": 3.5638921453692847,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014804685725882104,
+      "loss": 0.5048,
       "step": 3040
     },
     {
-      "epoch": 5.820267686424474,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.0949633722449915e-05,
-      "loss": 0.4234,
+      "epoch": 3.5685814771395075,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014772799948792683,
+      "loss": 0.4974,
       "step": 3044
     },
     {
-      "epoch": 5.827915869980879,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.0685284782959566e-05,
-      "loss": 0.3925,
+      "epoch": 3.5732708089097303,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001474091519855105,
+      "loss": 0.5482,
       "step": 3048
     },
     {
-      "epoch": 5.835564053537285,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.0422490701934996e-05,
-      "loss": 0.412,
+      "epoch": 3.577960140679953,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014709031619262737,
+      "loss": 0.4867,
       "step": 3052
     },
     {
-      "epoch": 5.84321223709369,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.0161254639168183e-05,
-      "loss": 0.3981,
+      "epoch": 3.582649472450176,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014677149355027985,
+      "loss": 0.5304,
       "step": 3056
     },
     {
-      "epoch": 5.850860420650095,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.9901579735717743e-05,
-      "loss": 0.4204,
+      "epoch": 3.5873388042203986,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014645268549941107,
+      "loss": 0.5533,
       "step": 3060
     },
     {
-      "epoch": 5.858508604206501,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.964346911387127e-05,
-      "loss": 0.3878,
+      "epoch": 3.5920281359906214,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014613389348089794,
+      "loss": 0.5024,
       "step": 3064
     },
     {
-      "epoch": 5.866156787762907,
-      "grad_norm": 0.373046875,
-      "learning_rate": 1.9386925877107585e-05,
-      "loss": 0.4047,
+      "epoch": 3.5967174677608442,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001458151189355451,
+      "loss": 0.5124,
       "step": 3068
     },
     {
-      "epoch": 5.873804971319312,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.913195311005959e-05,
-      "loss": 0.3899,
+      "epoch": 3.601406799531067,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014549636330407823,
+      "loss": 0.5064,
       "step": 3072
     },
     {
-      "epoch": 5.881453154875717,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.8878553878477105e-05,
-      "loss": 0.4179,
+      "epoch": 3.60609613130129,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001451776280271374,
+      "loss": 0.5421,
       "step": 3076
     },
     {
-      "epoch": 5.8891013384321225,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.8626731229190016e-05,
-      "loss": 0.403,
+      "epoch": 3.610785463071512,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014485891454527083,
+      "loss": 0.4855,
       "step": 3080
     },
     {
-      "epoch": 5.896749521988528,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.8376488190071666e-05,
-      "loss": 0.3925,
+      "epoch": 3.615474794841735,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001445402242989281,
+      "loss": 0.4824,
       "step": 3084
     },
     {
-      "epoch": 5.904397705544933,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.8127827770002423e-05,
-      "loss": 0.3647,
+      "epoch": 3.6201641266119577,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014422155872845387,
+      "loss": 0.5056,
       "step": 3088
     },
     {
-      "epoch": 5.912045889101338,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.7880752958833543e-05,
-      "loss": 0.4301,
+      "epoch": 3.6248534583821805,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014390291927408123,
+      "loss": 0.4958,
       "step": 3092
     },
     {
-      "epoch": 5.919694072657744,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.7635266727351092e-05,
-      "loss": 0.3988,
+      "epoch": 3.6295427901524033,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014358430737592534,
+      "loss": 0.5354,
       "step": 3096
     },
     {
-      "epoch": 5.927342256214149,
+      "epoch": 3.634232121922626,
       "grad_norm": 0.34375,
-      "learning_rate": 1.73913720272404e-05,
-      "loss": 0.4028,
+      "learning_rate": 0.00014326572447397658,
+      "loss": 0.4956,
       "step": 3100
     },
     {
-      "epoch": 5.934990439770554,
-      "grad_norm": 0.328125,
-      "learning_rate": 1.714907179105049e-05,
-      "loss": 0.3756,
+      "epoch": 3.638921453692849,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014294717200809452,
+      "loss": 0.5054,
       "step": 3104
     },
     {
-      "epoch": 5.9426386233269595,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.6908368932158777e-05,
-      "loss": 0.4023,
+      "epoch": 3.6436107854630713,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001426286514180011,
+      "loss": 0.5112,
       "step": 3108
     },
     {
-      "epoch": 5.950286806883366,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.6669266344736104e-05,
-      "loss": 0.3784,
+      "epoch": 3.648300117233294,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014231016414327407,
+      "loss": 0.5005,
       "step": 3112
     },
     {
-      "epoch": 5.957934990439771,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.6431766903711914e-05,
-      "loss": 0.3622,
+      "epoch": 3.652989449003517,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014199171162334077,
+      "loss": 0.5444,
       "step": 3116
     },
     {
-      "epoch": 5.965583173996176,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.6195873464739702e-05,
-      "loss": 0.416,
+      "epoch": 3.6576787807737396,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014167329529747146,
+      "loss": 0.5815,
       "step": 3120
     },
     {
-      "epoch": 5.9732313575525815,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.5961588864162627e-05,
-      "loss": 0.4191,
+      "epoch": 3.6623681125439624,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001413549166047727,
+      "loss": 0.5558,
       "step": 3124
     },
     {
-      "epoch": 5.980879541108987,
-      "grad_norm": 0.328125,
-      "learning_rate": 1.5728915918979477e-05,
-      "loss": 0.3683,
+      "epoch": 3.6670574443141852,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001410365769841811,
+      "loss": 0.5045,
       "step": 3128
     },
     {
-      "epoch": 5.988527724665392,
+      "epoch": 3.671746776084408,
       "grad_norm": 0.337890625,
-      "learning_rate": 1.5497857426810756e-05,
-      "loss": 0.365,
+      "learning_rate": 0.00014071827787445656,
+      "loss": 0.5228,
       "step": 3132
     },
     {
-      "epoch": 5.996175908221797,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.5268416165865055e-05,
-      "loss": 0.401,
+      "epoch": 3.676436107854631,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014040002071417595,
+      "loss": 0.5277,
       "step": 3136
     },
     {
-      "epoch": 6.003824091778203,
-      "grad_norm": 0.33203125,
-      "learning_rate": 1.5040594894905628e-05,
-      "loss": 0.3805,
+      "epoch": 3.6811254396248536,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014008180694172645,
+      "loss": 0.5202,
       "step": 3140
     },
     {
-      "epoch": 6.011472275334608,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.481439635321729e-05,
-      "loss": 0.3635,
+      "epoch": 3.6858147713950764,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013976363799529936,
+      "loss": 0.5646,
       "step": 3144
     },
     {
-      "epoch": 6.019120458891013,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.458982326057338e-05,
-      "loss": 0.37,
+      "epoch": 3.690504103165299,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001394455153128832,
+      "loss": 0.5456,
       "step": 3148
     },
     {
-      "epoch": 6.0267686424474185,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.436687831720314e-05,
-      "loss": 0.3981,
+      "epoch": 3.6951934349355215,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001391274403322574,
+      "loss": 0.5319,
       "step": 3152
     },
     {
-      "epoch": 6.034416826003824,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.4145564203759219e-05,
-      "loss": 0.4484,
+      "epoch": 3.6998827667057443,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013880941449098596,
+      "loss": 0.5385,
       "step": 3156
     },
     {
-      "epoch": 6.042065009560229,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.3925883581285401e-05,
-      "loss": 0.3988,
+      "epoch": 3.704572098475967,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001384914392264106,
+      "loss": 0.5151,
       "step": 3160
     },
     {
-      "epoch": 6.049713193116634,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.3707839091184702e-05,
-      "loss": 0.3422,
+      "epoch": 3.70926143024619,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013817351597564457,
+      "loss": 0.5354,
       "step": 3164
     },
     {
-      "epoch": 6.0573613766730405,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.349143335518752e-05,
-      "loss": 0.3994,
+      "epoch": 3.7139507620164127,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013785564617556603,
+      "loss": 0.5217,
       "step": 3168
     },
     {
-      "epoch": 6.065009560229446,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.3276668975320165e-05,
-      "loss": 0.4038,
+      "epoch": 3.7186400937866355,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013753783126281145,
+      "loss": 0.5383,
       "step": 3172
     },
     {
-      "epoch": 6.072657743785851,
-      "grad_norm": 0.328125,
-      "learning_rate": 1.3063548533873536e-05,
-      "loss": 0.4065,
+      "epoch": 3.7233294255568583,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001372200726737694,
+      "loss": 0.4918,
       "step": 3176
     },
     {
-      "epoch": 6.080305927342256,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.2852074593372142e-05,
-      "loss": 0.3902,
+      "epoch": 3.7280187573270807,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013690237184457377,
+      "loss": 0.4809,
       "step": 3180
     },
     {
-      "epoch": 6.087954110898662,
-      "grad_norm": 0.328125,
-      "learning_rate": 1.2642249696543178e-05,
-      "loss": 0.388,
+      "epoch": 3.7327080890973034,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013658473021109749,
+      "loss": 0.5185,
       "step": 3184
     },
     {
-      "epoch": 6.095602294455067,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.243407636628605e-05,
-      "loss": 0.3979,
+      "epoch": 3.7373974208675262,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013626714920894587,
+      "loss": 0.5666,
       "step": 3188
     },
     {
-      "epoch": 6.103250478011472,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.2227557105642e-05,
-      "loss": 0.404,
+      "epoch": 3.742086752637749,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00013594963027345022,
+      "loss": 0.5398,
       "step": 3192
     },
     {
-      "epoch": 6.1108986615678775,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.2022694397763993e-05,
-      "loss": 0.4059,
+      "epoch": 3.746776084407972,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001356321748396614,
+      "loss": 0.553,
       "step": 3196
     },
     {
-      "epoch": 6.118546845124283,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.1819490705886914e-05,
-      "loss": 0.3766,
+      "epoch": 3.7514654161781946,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013531478434234312,
+      "loss": 0.5552,
       "step": 3200
     },
     {
-      "epoch": 6.126195028680688,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.16179484732979e-05,
-      "loss": 0.3629,
+      "epoch": 3.7561547479484174,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013499746021596582,
+      "loss": 0.543,
       "step": 3204
     },
     {
-      "epoch": 6.133843212237093,
-      "grad_norm": 0.375,
-      "learning_rate": 1.1418070123306989e-05,
-      "loss": 0.3733,
+      "epoch": 3.76084407971864,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013468020389469974,
+      "loss": 0.5163,
       "step": 3208
     },
     {
-      "epoch": 6.141491395793499,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.1219858059217951e-05,
-      "loss": 0.4169,
+      "epoch": 3.765533411488863,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001343630168124088,
+      "loss": 0.5194,
       "step": 3212
     },
     {
-      "epoch": 6.149139579349905,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.1023314664299455e-05,
-      "loss": 0.4131,
+      "epoch": 3.770222743259086,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00013404590040264397,
+      "loss": 0.525,
       "step": 3216
     },
     {
-      "epoch": 6.15678776290631,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.0828442301756312e-05,
-      "loss": 0.3505,
+      "epoch": 3.7749120750293086,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001337288560986368,
+      "loss": 0.4793,
       "step": 3220
     },
     {
-      "epoch": 6.164435946462715,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.0635243314701163e-05,
-      "loss": 0.3978,
+      "epoch": 3.779601406799531,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000133411885333293,
+      "loss": 0.5353,
       "step": 3224
     },
     {
-      "epoch": 6.172084130019121,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.0443720026126273e-05,
-      "loss": 0.3982,
+      "epoch": 3.7842907385697537,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013309498953918583,
+      "loss": 0.5114,
       "step": 3228
     },
     {
-      "epoch": 6.179732313575526,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.025387473887554e-05,
-      "loss": 0.3935,
+      "epoch": 3.7889800703399765,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001327781701485498,
+      "loss": 0.5858,
       "step": 3232
     },
     {
-      "epoch": 6.187380497131931,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.0065709735616917e-05,
-      "loss": 0.3709,
+      "epoch": 3.7936694021101993,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013246142859327402,
+      "loss": 0.548,
       "step": 3236
     },
     {
-      "epoch": 6.195028680688337,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.87922727881484e-06,
-      "loss": 0.3922,
+      "epoch": 3.798358733880422,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001321447663048959,
+      "loss": 0.4745,
       "step": 3240
     },
     {
-      "epoch": 6.202676864244742,
-      "grad_norm": 0.328125,
-      "learning_rate": 9.694429610703153e-06,
-      "loss": 0.3719,
+      "epoch": 3.803048065650645,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013182818471459457,
+      "loss": 0.4773,
       "step": 3244
     },
     {
-      "epoch": 6.210325047801147,
-      "grad_norm": 0.333984375,
-      "learning_rate": 9.511318953258013e-06,
-      "loss": 0.2995,
+      "epoch": 3.8077373974208673,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013151168525318436,
+      "loss": 0.517,
       "step": 3248
     },
     {
-      "epoch": 6.217973231357552,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.329897508171296e-06,
-      "loss": 0.3932,
+      "epoch": 3.81242672919109,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013119526935110852,
+      "loss": 0.5349,
       "step": 3252
     },
     {
-      "epoch": 6.225621414913958,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.150167456824065e-06,
-      "loss": 0.4016,
+      "epoch": 3.817116060961313,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013087893843843264,
+      "loss": 0.5046,
       "step": 3256
     },
     {
-      "epoch": 6.233269598470363,
+      "epoch": 3.8218053927315356,
       "grad_norm": 0.33984375,
-      "learning_rate": 8.972130960260326e-06,
-      "loss": 0.3622,
+      "learning_rate": 0.00013056269394483814,
+      "loss": 0.5361,
       "step": 3260
     },
     {
-      "epoch": 6.240917782026768,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.795790159161098e-06,
-      "loss": 0.4053,
+      "epoch": 3.8264947245017584,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013024653729961586,
+      "loss": 0.5355,
       "step": 3264
     },
     {
-      "epoch": 6.248565965583174,
+      "epoch": 3.831184056271981,
       "grad_norm": 0.3203125,
-      "learning_rate": 8.621147173818587e-06,
-      "loss": 0.3517,
+      "learning_rate": 0.00012993046993165966,
+      "loss": 0.5392,
       "step": 3268
     },
     {
-      "epoch": 6.25621414913958,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.448204104110818e-06,
-      "loss": 0.3448,
+      "epoch": 3.835873388042204,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00012961449326945985,
+      "loss": 0.5118,
       "step": 3272
     },
     {
-      "epoch": 6.263862332695985,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.276963029476275e-06,
-      "loss": 0.3758,
+      "epoch": 3.840562719812427,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012929860874109683,
+      "loss": 0.5205,
       "step": 3276
     },
     {
-      "epoch": 6.27151051625239,
-      "grad_norm": 0.37109375,
-      "learning_rate": 8.107426008888934e-06,
-      "loss": 0.3911,
+      "epoch": 3.8452520515826496,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00012898281777423465,
+      "loss": 0.5297,
       "step": 3280
     },
     {
-      "epoch": 6.279158699808796,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.93959508083351e-06,
-      "loss": 0.3491,
+      "epoch": 3.8499413833528724,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012866712179611427,
+      "loss": 0.5043,
       "step": 3284
     },
     {
-      "epoch": 6.286806883365201,
-      "grad_norm": 0.34375,
-      "learning_rate": 7.773472263280977e-06,
-      "loss": 0.3954,
+      "epoch": 3.854630715123095,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001283515222335476,
+      "loss": 0.5284,
       "step": 3288
     },
     {
-      "epoch": 6.294455066921606,
-      "grad_norm": 0.34375,
-      "learning_rate": 7.609059553664254e-06,
-      "loss": 0.4018,
+      "epoch": 3.859320046893318,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012803602051291064,
+      "loss": 0.5118,
       "step": 3292
     },
     {
-      "epoch": 6.3021032504780115,
-      "grad_norm": 0.330078125,
-      "learning_rate": 7.446358928854207e-06,
-      "loss": 0.3823,
+      "epoch": 3.8640093786635403,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012772061806013728,
+      "loss": 0.5621,
       "step": 3296
     },
     {
-      "epoch": 6.309751434034417,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.2853723451358705e-06,
-      "loss": 0.4097,
+      "epoch": 3.868698710433763,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012740531630071268,
+      "loss": 0.5587,
       "step": 3300
     },
     {
-      "epoch": 6.317399617590822,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.126101738184964e-06,
-      "loss": 0.3676,
+      "epoch": 3.873388042203986,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012709011665966698,
+      "loss": 0.4888,
       "step": 3304
     },
     {
-      "epoch": 6.325047801147227,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.9685490230445615e-06,
-      "loss": 0.3901,
+      "epoch": 3.8780773739742087,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012677502056156878,
+      "loss": 0.4617,
       "step": 3308
     },
     {
-      "epoch": 6.332695984703633,
-      "grad_norm": 0.32421875,
-      "learning_rate": 6.812716094102128e-06,
-      "loss": 0.3652,
+      "epoch": 3.8827667057444315,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012646002943051863,
+      "loss": 0.5277,
       "step": 3312
     },
     {
-      "epoch": 6.340344168260038,
-      "grad_norm": 0.35546875,
-      "learning_rate": 6.658604825066683e-06,
-      "loss": 0.3631,
+      "epoch": 3.8874560375146543,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001261451446901428,
+      "loss": 0.5163,
       "step": 3316
     },
     {
-      "epoch": 6.347992351816444,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.50621706894629e-06,
-      "loss": 0.4053,
+      "epoch": 3.8921453692848766,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00012583036776358652,
+      "loss": 0.4925,
       "step": 3320
     },
     {
-      "epoch": 6.355640535372849,
-      "grad_norm": 0.33203125,
-      "learning_rate": 6.355554658025791e-06,
-      "loss": 0.3451,
+      "epoch": 3.8968347010550994,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00012551570007350796,
+      "loss": 0.5093,
       "step": 3324
     },
     {
-      "epoch": 6.363288718929255,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.206619403844804e-06,
-      "loss": 0.4109,
+      "epoch": 3.9015240328253222,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001252011430420715,
+      "loss": 0.5028,
       "step": 3328
     },
     {
-      "epoch": 6.37093690248566,
-      "grad_norm": 0.318359375,
-      "learning_rate": 6.059413097175808e-06,
-      "loss": 0.3897,
+      "epoch": 3.906213364595545,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001248866980909414,
+      "loss": 0.5412,
       "step": 3332
     },
     {
-      "epoch": 6.378585086042065,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.913937508002797e-06,
-      "loss": 0.3573,
+      "epoch": 3.910902696365768,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012457236664127535,
+      "loss": 0.5431,
       "step": 3336
     },
     {
-      "epoch": 6.3862332695984705,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.770194385499877e-06,
-      "loss": 0.3928,
+      "epoch": 3.9155920281359906,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012425815011371806,
+      "loss": 0.5178,
       "step": 3340
     },
     {
-      "epoch": 6.393881453154876,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.628185458010248e-06,
-      "loss": 0.3662,
+      "epoch": 3.9202813599062134,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012394404992839485,
+      "loss": 0.5413,
       "step": 3344
     },
     {
-      "epoch": 6.401529636711281,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.487912433025493e-06,
-      "loss": 0.3974,
+      "epoch": 3.924970691676436,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001236300675049052,
+      "loss": 0.5313,
       "step": 3348
     },
     {
-      "epoch": 6.409177820267686,
-      "grad_norm": 0.32421875,
-      "learning_rate": 5.349376997164923e-06,
-      "loss": 0.3477,
+      "epoch": 3.929660023446659,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001233162042623165,
+      "loss": 0.5314,
       "step": 3352
     },
     {
-      "epoch": 6.416826003824092,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.212580816155426e-06,
-      "loss": 0.3992,
+      "epoch": 3.934349355216882,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001230024616191572,
+      "loss": 0.5232,
       "step": 3356
     },
     {
-      "epoch": 6.424474187380497,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.077525534811339e-06,
-      "loss": 0.3881,
+      "epoch": 3.9390386869871046,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012268884099341095,
+      "loss": 0.5187,
       "step": 3360
     },
     {
-      "epoch": 6.432122370936902,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.9442127770147385e-06,
-      "loss": 0.3767,
+      "epoch": 3.943728018757327,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012237534380250985,
+      "loss": 0.5218,
       "step": 3364
     },
     {
-      "epoch": 6.4397705544933075,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.812644145695915e-06,
-      "loss": 0.3939,
+      "epoch": 3.9484173505275497,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012206197146332808,
+      "loss": 0.5651,
       "step": 3368
     },
     {
-      "epoch": 6.447418738049713,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.682821222813998e-06,
-      "loss": 0.3322,
+      "epoch": 3.9531066822977725,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00012174872539217565,
+      "loss": 0.5458,
       "step": 3372
     },
     {
-      "epoch": 6.455066921606119,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.554745569338092e-06,
-      "loss": 0.3414,
+      "epoch": 3.9577960140679953,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012143560700479177,
+      "loss": 0.5499,
       "step": 3376
     },
     {
-      "epoch": 6.462715105162524,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.428418725228372e-06,
-      "loss": 0.3982,
+      "epoch": 3.962485345838218,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00012112261771633866,
+      "loss": 0.5396,
       "step": 3380
     },
     {
-      "epoch": 6.4703632887189295,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.303842209417652e-06,
-      "loss": 0.3522,
+      "epoch": 3.967174677608441,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012080975894139508,
+      "loss": 0.5212,
       "step": 3384
     },
     {
-      "epoch": 6.478011472275335,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.181017519793079e-06,
-      "loss": 0.4167,
+      "epoch": 3.9718640093786637,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012049703209394983,
+      "loss": 0.4536,
       "step": 3388
     },
     {
-      "epoch": 6.48565965583174,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.059946133178132e-06,
-      "loss": 0.3875,
+      "epoch": 3.976553341148886,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012018443858739554,
+      "loss": 0.508,
       "step": 3392
     },
     {
-      "epoch": 6.493307839388145,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.94062950531489e-06,
-      "loss": 0.3662,
+      "epoch": 3.981242672919109,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001198719798345221,
+      "loss": 0.5013,
       "step": 3396
     },
     {
-      "epoch": 6.500956022944551,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.823069070846474e-06,
-      "loss": 0.3625,
+      "epoch": 3.9859320046893316,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00011955965724751048,
+      "loss": 0.5184,
       "step": 3400
     },
     {
-      "epoch": 6.508604206500956,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.707266243299861e-06,
-      "loss": 0.3445,
+      "epoch": 3.9906213364595544,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011924747223792619,
+      "loss": 0.5191,
       "step": 3404
     },
     {
-      "epoch": 6.516252390057361,
-      "grad_norm": 0.353515625,
-      "learning_rate": 3.5932224150688526e-06,
-      "loss": 0.3705,
+      "epoch": 3.995310668229777,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00011893542621671296,
+      "loss": 0.5332,
       "step": 3408
     },
     {
-      "epoch": 6.5239005736137665,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.4809389573973e-06,
-      "loss": 0.396,
+      "epoch": 4.0,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00011862352059418636,
+      "loss": 0.4709,
       "step": 3412
     },
     {
-      "epoch": 6.531548757170172,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.3704172203627035e-06,
-      "loss": 0.3689,
+      "epoch": 4.004689331770223,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00011831175678002737,
+      "loss": 0.4506,
       "step": 3416
     },
     {
-      "epoch": 6.539196940726577,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.2616585328599065e-06,
-      "loss": 0.4253,
+      "epoch": 4.009378663540446,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011800013618327605,
+      "loss": 0.4412,
       "step": 3420
     },
     {
-      "epoch": 6.546845124282983,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.154664202585128e-06,
-      "loss": 0.387,
+      "epoch": 4.014067995310668,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011768866021232528,
+      "loss": 0.4716,
       "step": 3424
     },
     {
-      "epoch": 6.5544933078393885,
-      "grad_norm": 0.353515625,
-      "learning_rate": 3.049435516020271e-06,
-      "loss": 0.3585,
+      "epoch": 4.018757327080891,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011737733027491427,
+      "loss": 0.4189,
       "step": 3428
     },
     {
-      "epoch": 6.562141491395794,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.94597373841744e-06,
-      "loss": 0.3912,
+      "epoch": 4.023446658851114,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011706614777812204,
+      "loss": 0.4458,
       "step": 3432
     },
     {
-      "epoch": 6.569789674952199,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.844280113783698e-06,
-      "loss": 0.3863,
+      "epoch": 4.028135990621337,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011675511412836145,
+      "loss": 0.4492,
       "step": 3436
     },
     {
-      "epoch": 6.577437858508604,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.7443558648661656e-06,
-      "loss": 0.4214,
+      "epoch": 4.0328253223915596,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011644423073137259,
+      "loss": 0.4557,
       "step": 3440
     },
     {
-      "epoch": 6.58508604206501,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.646202193137248e-06,
-      "loss": 0.429,
+      "epoch": 4.037514654161782,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011613349899221641,
+      "loss": 0.4215,
       "step": 3444
     },
     {
-      "epoch": 6.592734225621415,
-      "grad_norm": 0.330078125,
-      "learning_rate": 2.549820278780246e-06,
-      "loss": 0.3523,
+      "epoch": 4.042203985932004,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011582292031526844,
+      "loss": 0.467,
       "step": 3448
     },
     {
-      "epoch": 6.60038240917782,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.455211280675168e-06,
-      "loss": 0.4001,
+      "epoch": 4.046893317702227,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011551249610421252,
+      "loss": 0.4608,
       "step": 3452
     },
     {
-      "epoch": 6.6080305927342256,
-      "grad_norm": 0.376953125,
-      "learning_rate": 2.3623763363847246e-06,
-      "loss": 0.3819,
+      "epoch": 4.05158264947245,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011520222776203428,
+      "loss": 0.487,
       "step": 3456
     },
     {
-      "epoch": 6.615678776290631,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.271316562140757e-06,
-      "loss": 0.3632,
+      "epoch": 4.056271981242673,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011489211669101493,
+      "loss": 0.4697,
       "step": 3460
     },
     {
-      "epoch": 6.623326959847036,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.182033052830695e-06,
-      "loss": 0.3842,
+      "epoch": 4.060961313012895,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011458216429272489,
+      "loss": 0.4833,
       "step": 3464
     },
     {
-      "epoch": 6.630975143403441,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.094526881984521e-06,
-      "loss": 0.3844,
+      "epoch": 4.065650644783118,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011427237196801736,
+      "loss": 0.4375,
       "step": 3468
     },
     {
-      "epoch": 6.638623326959847,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.0087991017617598e-06,
-      "loss": 0.3697,
+      "epoch": 4.070339976553341,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011396274111702217,
+      "loss": 0.4756,
       "step": 3472
     },
     {
-      "epoch": 6.646271510516252,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.924850742938894e-06,
-      "loss": 0.3886,
+      "epoch": 4.075029308323564,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011365327313913932,
+      "loss": 0.4834,
       "step": 3476
     },
     {
-      "epoch": 6.653919694072657,
-      "grad_norm": 0.349609375,
-      "learning_rate": 1.8426828148969008e-06,
-      "loss": 0.363,
+      "epoch": 4.079718640093787,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011334396943303271,
+      "loss": 0.4716,
       "step": 3480
     },
     {
-      "epoch": 6.661567877629063,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.7622963056091843e-06,
-      "loss": 0.3671,
+      "epoch": 4.084407971864009,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011303483139662382,
+      "loss": 0.4264,
       "step": 3484
     },
     {
-      "epoch": 6.669216061185469,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.6836921816296644e-06,
-      "loss": 0.4043,
+      "epoch": 4.089097303634232,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011272586042708535,
+      "loss": 0.4405,
       "step": 3488
     },
     {
-      "epoch": 6.676864244741874,
-      "grad_norm": 0.33203125,
-      "learning_rate": 1.6068713880811546e-06,
-      "loss": 0.398,
+      "epoch": 4.093786635404455,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011241705792083484,
+      "loss": 0.3871,
       "step": 3492
     },
     {
-      "epoch": 6.684512428298279,
-      "grad_norm": 0.330078125,
-      "learning_rate": 1.531834848643987e-06,
-      "loss": 0.3847,
+      "epoch": 4.098475967174678,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011210842527352861,
+      "loss": 0.4416,
       "step": 3496
     },
     {
-      "epoch": 6.692160611854685,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.4585834655449547e-06,
-      "loss": 0.4041,
+      "epoch": 4.103165298944901,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011179996388005524,
+      "loss": 0.4453,
       "step": 3500
     },
     {
-      "epoch": 6.69980879541109,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.3871181195464042e-06,
-      "loss": 0.3924,
+      "epoch": 4.107854630715123,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001114916751345292,
+      "loss": 0.4967,
       "step": 3504
     },
     {
-      "epoch": 6.707456978967495,
-      "grad_norm": 0.32421875,
-      "learning_rate": 1.3174396699356937e-06,
-      "loss": 0.3557,
+      "epoch": 4.112543962485346,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011118356043028476,
+      "loss": 0.465,
       "step": 3508
     },
     {
-      "epoch": 6.7151051625239,
+      "epoch": 4.117233294255569,
       "grad_norm": 0.33984375,
-      "learning_rate": 1.2495489545148008e-06,
-      "loss": 0.3767,
+      "learning_rate": 0.00011087562115986965,
+      "loss": 0.4547,
       "step": 3512
     },
     {
-      "epoch": 6.722753346080306,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.1834467895903476e-06,
-      "loss": 0.4122,
+      "epoch": 4.121922626025792,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011056785871503862,
+      "loss": 0.4575,
       "step": 3516
     },
     {
-      "epoch": 6.730401529636711,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.1191339699636426e-06,
-      "loss": 0.3873,
+      "epoch": 4.126611957796014,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011026027448674725,
+      "loss": 0.441,
       "step": 3520
     },
     {
-      "epoch": 6.738049713193116,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.0566112689213035e-06,
-      "loss": 0.4039,
+      "epoch": 4.131301289566236,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010995286986514571,
+      "loss": 0.4565,
       "step": 3524
     },
     {
-      "epoch": 6.7456978967495225,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.958794382257663e-07,
-      "loss": 0.4338,
+      "epoch": 4.135990621336459,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010964564623957239,
+      "loss": 0.4744,
       "step": 3528
     },
     {
-      "epoch": 6.753346080305928,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.369392081063908e-07,
-      "loss": 0.3474,
+      "epoch": 4.140679953106682,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010933860499854768,
+      "loss": 0.4447,
       "step": 3532
     },
     {
-      "epoch": 6.760994263862333,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.797912872506529e-07,
-      "loss": 0.3553,
+      "epoch": 4.145369284876905,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001090317475297677,
+      "loss": 0.5169,
       "step": 3536
     },
     {
-      "epoch": 6.768642447418738,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.244363627955664e-07,
-      "loss": 0.3955,
+      "epoch": 4.150058616647128,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010872507522009781,
+      "loss": 0.461,
       "step": 3540
     },
     {
-      "epoch": 6.776290630975144,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.708751003194569e-07,
-      "loss": 0.3925,
+      "epoch": 4.15474794841735,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010841858945556677,
+      "loss": 0.452,
       "step": 3544
     },
     {
-      "epoch": 6.783938814531549,
-      "grad_norm": 0.333984375,
-      "learning_rate": 7.191081438339685e-07,
-      "loss": 0.4088,
+      "epoch": 4.159437280187573,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010811229162136009,
+      "loss": 0.4486,
       "step": 3548
     },
     {
-      "epoch": 6.791586998087954,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.691361157763198e-07,
-      "loss": 0.3916,
+      "epoch": 4.164126611957796,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010780618310181395,
+      "loss": 0.4286,
       "step": 3552
     },
     {
-      "epoch": 6.7992351816443595,
-      "grad_norm": 0.349609375,
-      "learning_rate": 6.209596170018094e-07,
-      "loss": 0.4524,
+      "epoch": 4.168815943728019,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010750026528040895,
+      "loss": 0.4743,
       "step": 3556
     },
     {
-      "epoch": 6.806883365200765,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.745792267765559e-07,
-      "loss": 0.398,
+      "epoch": 4.173505275498242,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010719453953976375,
+      "loss": 0.4637,
       "step": 3560
     },
     {
-      "epoch": 6.81453154875717,
-      "grad_norm": 0.32421875,
-      "learning_rate": 5.29995502770636e-07,
-      "loss": 0.4172,
+      "epoch": 4.178194607268464,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00010688900726162899,
+      "loss": 0.466,
       "step": 3564
     },
     {
-      "epoch": 6.822179732313575,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.87208981051257e-07,
-      "loss": 0.3622,
+      "epoch": 4.182883939038687,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010658366982688076,
+      "loss": 0.4639,
       "step": 3568
     },
     {
-      "epoch": 6.829827915869981,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.462201760763784e-07,
-      "loss": 0.4181,
+      "epoch": 4.18757327080891,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010627852861551479,
+      "loss": 0.4853,
       "step": 3572
     },
     {
-      "epoch": 6.837476099426386,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.0702958068853373e-07,
-      "loss": 0.3955,
+      "epoch": 4.192262602579133,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010597358500663966,
+      "loss": 0.4674,
       "step": 3576
     },
     {
-      "epoch": 6.845124282982791,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.696376661088685e-07,
-      "loss": 0.3865,
+      "epoch": 4.1969519343493555,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010566884037847111,
+      "loss": 0.4648,
       "step": 3580
     },
     {
-      "epoch": 6.8527724665391965,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.340448819315111e-07,
-      "loss": 0.3909,
+      "epoch": 4.201641266119578,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001053642961083255,
+      "loss": 0.4708,
       "step": 3584
     },
     {
-      "epoch": 6.860420650095603,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.002516561181112e-07,
-      "loss": 0.3462,
+      "epoch": 4.206330597889801,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010505995357261364,
+      "loss": 0.4416,
       "step": 3588
     },
     {
-      "epoch": 6.868068833652008,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.6825839499277634e-07,
-      "loss": 0.3655,
+      "epoch": 4.211019929660023,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010475581414683466,
+      "loss": 0.4991,
       "step": 3592
     },
     {
-      "epoch": 6.875717017208413,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.3806548323710984e-07,
-      "loss": 0.4164,
+      "epoch": 4.215709261430246,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010445187920556956,
+      "loss": 0.4488,
       "step": 3596
     },
     {
-      "epoch": 6.8833652007648185,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.096732838856141e-07,
-      "loss": 0.3865,
+      "epoch": 4.220398593200469,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010414815012247529,
+      "loss": 0.4471,
       "step": 3600
     },
     {
-      "epoch": 6.891013384321224,
+      "epoch": 4.225087924970691,
       "grad_norm": 0.33984375,
-      "learning_rate": 1.8308213832134434e-07,
-      "loss": 0.4012,
+      "learning_rate": 0.00010384462827027838,
+      "loss": 0.4286,
       "step": 3604
     },
     {
-      "epoch": 6.898661567877629,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.5829236627177833e-07,
-      "loss": 0.3761,
+      "epoch": 4.229777256740914,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010354131502076875,
+      "loss": 0.4893,
       "step": 3608
     },
     {
-      "epoch": 6.906309751434034,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.3530426580496968e-07,
-      "loss": 0.3674,
+      "epoch": 4.234466588511137,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00010323821174479363,
+      "loss": 0.4489,
       "step": 3612
     },
     {
-      "epoch": 6.91395793499044,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.1411811332600051e-07,
-      "loss": 0.3708,
+      "epoch": 4.23915592028136,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.000102935319812251,
+      "loss": 0.4349,
       "step": 3616
     },
     {
-      "epoch": 6.921606118546845,
-      "grad_norm": 0.333984375,
-      "learning_rate": 9.473416357361762e-08,
-      "loss": 0.4026,
+      "epoch": 4.243845252051583,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001026326405920839,
+      "loss": 0.4235,
       "step": 3620
     },
     {
-      "epoch": 6.92925430210325,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.715264961718481e-08,
-      "loss": 0.4073,
+      "epoch": 4.248534583821805,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010233017545227389,
+      "loss": 0.4619,
       "step": 3624
     },
     {
-      "epoch": 6.9369024856596555,
-      "grad_norm": 0.3359375,
-      "learning_rate": 6.137378285386851e-08,
-      "loss": 0.3944,
+      "epoch": 4.253223915592028,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010202792575983502,
+      "loss": 0.4251,
       "step": 3628
     },
     {
-      "epoch": 6.944550669216062,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.739775300612314e-08,
-      "loss": 0.352,
+      "epoch": 4.257913247362251,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010172589288080759,
+      "loss": 0.4545,
       "step": 3632
     },
     {
-      "epoch": 6.952198852772467,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.522472811939292e-08,
-      "loss": 0.3563,
+      "epoch": 4.262602579132474,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010142407818025201,
+      "loss": 0.425,
       "step": 3636
     },
     {
-      "epoch": 6.959847036328872,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.4854854560096837e-08,
-      "loss": 0.3673,
+      "epoch": 4.2672919109026966,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010112248302224263,
+      "loss": 0.4444,
       "step": 3640
     },
     {
-      "epoch": 6.9674952198852775,
+      "epoch": 4.271981242672919,
       "grad_norm": 0.353515625,
-      "learning_rate": 1.6288257013830075e-08,
-      "loss": 0.4149,
+      "learning_rate": 0.00010082110876986147,
+      "loss": 0.4439,
       "step": 3644
     },
     {
-      "epoch": 6.975143403441683,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.52503848396513e-09,
-      "loss": 0.4061,
+      "epoch": 4.276670574443142,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00010051995678519231,
+      "loss": 0.4381,
       "step": 3648
     },
     {
-      "epoch": 6.982791586998088,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.565280290269591e-09,
-      "loss": 0.3616,
+      "epoch": 4.281359906213365,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00010021902842931421,
+      "loss": 0.4599,
       "step": 3652
     },
     {
-      "epoch": 6.990439770554493,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.4090420680734627e-09,
-      "loss": 0.3985,
+      "epoch": 4.286049237983588,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.991832506229558e-05,
+      "loss": 0.4676,
       "step": 3656
     },
     {
-      "epoch": 6.998087954110899,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.6361767436507997e-11,
-      "loss": 0.3654,
+      "epoch": 4.29073856975381,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.961784804318803e-05,
+      "loss": 0.4528,
       "step": 3660
     },
     {
-      "epoch": 7.0,
-      "step": 3661,
-      "total_flos": 2.3920283694378516e+18,
-      "train_loss": 0.561195495056344,
-      "train_runtime": 14692.9794,
-      "train_samples_per_second": 7.96,
-      "train_steps_per_second": 0.249
+      "epoch": 4.295427901524032,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.931759873002012e-05,
+      "loss": 0.4812,
+      "step": 3664
+    },
+    {
+      "epoch": 4.300117233294255,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.901757847979136e-05,
+      "loss": 0.451,
+      "step": 3668
+    },
+    {
+      "epoch": 4.304806565064478,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.871778864846578e-05,
+      "loss": 0.47,
+      "step": 3672
+    },
+    {
+      "epoch": 4.309495896834701,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.841823059096629e-05,
+      "loss": 0.4477,
+      "step": 3676
+    },
+    {
+      "epoch": 4.314185228604924,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.811890566116806e-05,
+      "loss": 0.4822,
+      "step": 3680
+    },
+    {
+      "epoch": 4.318874560375146,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.781981521189283e-05,
+      "loss": 0.4965,
+      "step": 3684
+    },
+    {
+      "epoch": 4.323563892145369,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.75209605949023e-05,
+      "loss": 0.4346,
+      "step": 3688
+    },
+    {
+      "epoch": 4.328253223915592,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.722234316089256e-05,
+      "loss": 0.4458,
+      "step": 3692
+    },
+    {
+      "epoch": 4.332942555685815,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.692396425948768e-05,
+      "loss": 0.4547,
+      "step": 3696
+    },
+    {
+      "epoch": 4.337631887456038,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.662582523923357e-05,
+      "loss": 0.4411,
+      "step": 3700
+    },
+    {
+      "epoch": 4.34232121922626,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.632792744759207e-05,
+      "loss": 0.4091,
+      "step": 3704
+    },
+    {
+      "epoch": 4.347010550996483,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.603027223093474e-05,
+      "loss": 0.4727,
+      "step": 3708
+    },
+    {
+      "epoch": 4.351699882766706,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.573286093453682e-05,
+      "loss": 0.4794,
+      "step": 3712
+    },
+    {
+      "epoch": 4.356389214536929,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.543569490257111e-05,
+      "loss": 0.4868,
+      "step": 3716
+    },
+    {
+      "epoch": 4.3610785463071515,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.513877547810192e-05,
+      "loss": 0.4537,
+      "step": 3720
+    },
+    {
+      "epoch": 4.365767878077374,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.484210400307903e-05,
+      "loss": 0.414,
+      "step": 3724
+    },
+    {
+      "epoch": 4.370457209847597,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.454568181833151e-05,
+      "loss": 0.445,
+      "step": 3728
+    },
+    {
+      "epoch": 4.37514654161782,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.424951026356183e-05,
+      "loss": 0.4561,
+      "step": 3732
+    },
+    {
+      "epoch": 4.379835873388042,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.395359067733974e-05,
+      "loss": 0.4429,
+      "step": 3736
+    },
+    {
+      "epoch": 4.384525205158265,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.365792439709609e-05,
+      "loss": 0.4832,
+      "step": 3740
+    },
+    {
+      "epoch": 4.389214536928487,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.336251275911702e-05,
+      "loss": 0.4795,
+      "step": 3744
+    },
+    {
+      "epoch": 4.39390386869871,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.306735709853765e-05,
+      "loss": 0.4701,
+      "step": 3748
+    },
+    {
+      "epoch": 4.398593200468933,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.277245874933633e-05,
+      "loss": 0.4642,
+      "step": 3752
+    },
+    {
+      "epoch": 4.403282532239156,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.247781904432847e-05,
+      "loss": 0.4756,
+      "step": 3756
+    },
+    {
+      "epoch": 4.407971864009379,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.218343931516034e-05,
+      "loss": 0.4885,
+      "step": 3760
+    },
+    {
+      "epoch": 4.412661195779601,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.188932089230338e-05,
+      "loss": 0.4598,
+      "step": 3764
+    },
+    {
+      "epoch": 4.417350527549824,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.159546510504807e-05,
+      "loss": 0.4661,
+      "step": 3768
+    },
+    {
+      "epoch": 4.422039859320047,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.130187328149779e-05,
+      "loss": 0.4818,
+      "step": 3772
+    },
+    {
+      "epoch": 4.42672919109027,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.100854674856293e-05,
+      "loss": 0.4438,
+      "step": 3776
+    },
+    {
+      "epoch": 4.4314185228604925,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.071548683195495e-05,
+      "loss": 0.4307,
+      "step": 3780
+    },
+    {
+      "epoch": 4.436107854630715,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.042269485618021e-05,
+      "loss": 0.44,
+      "step": 3784
+    },
+    {
+      "epoch": 4.440797186400938,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.013017214453422e-05,
+      "loss": 0.4822,
+      "step": 3788
+    },
+    {
+      "epoch": 4.445486518171161,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.983792001909543e-05,
+      "loss": 0.5057,
+      "step": 3792
+    },
+    {
+      "epoch": 4.450175849941384,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.954593980071941e-05,
+      "loss": 0.4424,
+      "step": 3796
+    },
+    {
+      "epoch": 4.4548651817116065,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.925423280903274e-05,
+      "loss": 0.4472,
+      "step": 3800
+    },
+    {
+      "epoch": 4.459554513481828,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.896280036242722e-05,
+      "loss": 0.4449,
+      "step": 3804
+    },
+    {
+      "epoch": 4.464243845252051,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.86716437780538e-05,
+      "loss": 0.4707,
+      "step": 3808
+    },
+    {
+      "epoch": 4.468933177022274,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.838076437181663e-05,
+      "loss": 0.4591,
+      "step": 3812
+    },
+    {
+      "epoch": 4.473622508792497,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.80901634583672e-05,
+      "loss": 0.4404,
+      "step": 3816
+    },
+    {
+      "epoch": 4.47831184056272,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.779984235109825e-05,
+      "loss": 0.4562,
+      "step": 3820
+    },
+    {
+      "epoch": 4.483001172332942,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.750980236213792e-05,
+      "loss": 0.4657,
+      "step": 3824
+    },
+    {
+      "epoch": 4.487690504103165,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.722004480234381e-05,
+      "loss": 0.4363,
+      "step": 3828
+    },
+    {
+      "epoch": 4.492379835873388,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.693057098129729e-05,
+      "loss": 0.468,
+      "step": 3832
+    },
+    {
+      "epoch": 4.497069167643611,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.664138220729686e-05,
+      "loss": 0.5015,
+      "step": 3836
+    },
+    {
+      "epoch": 4.5017584994138335,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.63524797873532e-05,
+      "loss": 0.4253,
+      "step": 3840
+    },
+    {
+      "epoch": 4.506447831184056,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.606386502718258e-05,
+      "loss": 0.4468,
+      "step": 3844
+    },
+    {
+      "epoch": 4.511137162954279,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.577553923120111e-05,
+      "loss": 0.481,
+      "step": 3848
+    },
+    {
+      "epoch": 4.515826494724502,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.548750370251915e-05,
+      "loss": 0.468,
+      "step": 3852
+    },
+    {
+      "epoch": 4.520515826494725,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.519975974293485e-05,
+      "loss": 0.4859,
+      "step": 3856
+    },
+    {
+      "epoch": 4.5252051582649475,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.49123086529289e-05,
+      "loss": 0.3966,
+      "step": 3860
+    },
+    {
+      "epoch": 4.52989449003517,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.462515173165817e-05,
+      "loss": 0.4594,
+      "step": 3864
+    },
+    {
+      "epoch": 4.534583821805393,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.433829027695e-05,
+      "loss": 0.4901,
+      "step": 3868
+    },
+    {
+      "epoch": 4.539273153575616,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.405172558529643e-05,
+      "loss": 0.4108,
+      "step": 3872
+    },
+    {
+      "epoch": 4.543962485345839,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.376545895184815e-05,
+      "loss": 0.4007,
+      "step": 3876
+    },
+    {
+      "epoch": 4.548651817116061,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.347949167040894e-05,
+      "loss": 0.4615,
+      "step": 3880
+    },
+    {
+      "epoch": 4.553341148886283,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.319382503342938e-05,
+      "loss": 0.4131,
+      "step": 3884
+    },
+    {
+      "epoch": 4.558030480656506,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.290846033200158e-05,
+      "loss": 0.4706,
+      "step": 3888
+    },
+    {
+      "epoch": 4.562719812426729,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.262339885585274e-05,
+      "loss": 0.4386,
+      "step": 3892
+    },
+    {
+      "epoch": 4.567409144196952,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.233864189333967e-05,
+      "loss": 0.4437,
+      "step": 3896
+    },
+    {
+      "epoch": 4.572098475967175,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.20541907314431e-05,
+      "loss": 0.4487,
+      "step": 3900
+    },
+    {
+      "epoch": 4.576787807737397,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.177004665576147e-05,
+      "loss": 0.4597,
+      "step": 3904
+    },
+    {
+      "epoch": 4.58147713950762,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.148621095050537e-05,
+      "loss": 0.4854,
+      "step": 3908
+    },
+    {
+      "epoch": 4.586166471277843,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.120268489849164e-05,
+      "loss": 0.4578,
+      "step": 3912
+    },
+    {
+      "epoch": 4.590855803048066,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.091946978113782e-05,
+      "loss": 0.4828,
+      "step": 3916
+    },
+    {
+      "epoch": 4.5955451348182885,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.063656687845592e-05,
+      "loss": 0.4351,
+      "step": 3920
+    },
+    {
+      "epoch": 4.600234466588511,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.035397746904695e-05,
+      "loss": 0.4401,
+      "step": 3924
+    },
+    {
+      "epoch": 4.604923798358734,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.007170283009517e-05,
+      "loss": 0.4649,
+      "step": 3928
+    },
+    {
+      "epoch": 4.609613130128957,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.978974423736202e-05,
+      "loss": 0.4128,
+      "step": 3932
+    },
+    {
+      "epoch": 4.61430246189918,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.950810296518076e-05,
+      "loss": 0.4591,
+      "step": 3936
+    },
+    {
+      "epoch": 4.6189917936694025,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.922678028645032e-05,
+      "loss": 0.5247,
+      "step": 3940
+    },
+    {
+      "epoch": 4.623681125439624,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.89457774726298e-05,
+      "loss": 0.4786,
+      "step": 3944
+    },
+    {
+      "epoch": 4.628370457209847,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.866509579373261e-05,
+      "loss": 0.4466,
+      "step": 3948
+    },
+    {
+      "epoch": 4.63305978898007,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.838473651832077e-05,
+      "loss": 0.4672,
+      "step": 3952
+    },
+    {
+      "epoch": 4.637749120750293,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.810470091349925e-05,
+      "loss": 0.454,
+      "step": 3956
+    },
+    {
+      "epoch": 4.642438452520516,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.782499024491004e-05,
+      "loss": 0.4429,
+      "step": 3960
+    },
+    {
+      "epoch": 4.647127784290738,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.754560577672674e-05,
+      "loss": 0.4261,
+      "step": 3964
+    },
+    {
+      "epoch": 4.651817116060961,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.726654877164847e-05,
+      "loss": 0.4814,
+      "step": 3968
+    },
+    {
+      "epoch": 4.656506447831184,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.698782049089438e-05,
+      "loss": 0.4318,
+      "step": 3972
+    },
+    {
+      "epoch": 4.661195779601407,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.67094221941981e-05,
+      "loss": 0.4282,
+      "step": 3976
+    },
+    {
+      "epoch": 4.6658851113716295,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.64313551398017e-05,
+      "loss": 0.4523,
+      "step": 3980
+    },
+    {
+      "epoch": 4.670574443141852,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.615362058445022e-05,
+      "loss": 0.4633,
+      "step": 3984
+    },
+    {
+      "epoch": 4.675263774912075,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.587621978338586e-05,
+      "loss": 0.4672,
+      "step": 3988
+    },
+    {
+      "epoch": 4.679953106682298,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.559915399034266e-05,
+      "loss": 0.4357,
+      "step": 3992
+    },
+    {
+      "epoch": 4.684642438452521,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.532242445754029e-05,
+      "loss": 0.4189,
+      "step": 3996
+    },
+    {
+      "epoch": 4.6893317702227435,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.504603243567874e-05,
+      "loss": 0.4741,
+      "step": 4000
+    },
+    {
+      "epoch": 4.694021101992966,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.476997917393269e-05,
+      "loss": 0.4782,
+      "step": 4004
+    },
+    {
+      "epoch": 4.698710433763189,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.449426591994565e-05,
+      "loss": 0.4787,
+      "step": 4008
+    },
+    {
+      "epoch": 4.703399765533412,
+      "grad_norm": 0.36328125,
+      "learning_rate": 7.421889391982454e-05,
+      "loss": 0.4287,
+      "step": 4012
+    },
+    {
+      "epoch": 4.708089097303635,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.394386441813388e-05,
+      "loss": 0.4493,
+      "step": 4016
+    },
+    {
+      "epoch": 4.7127784290738575,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.366917865789027e-05,
+      "loss": 0.4431,
+      "step": 4020
+    },
+    {
+      "epoch": 4.717467760844079,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.339483788055672e-05,
+      "loss": 0.3989,
+      "step": 4024
+    },
+    {
+      "epoch": 4.722157092614302,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.312084332603706e-05,
+      "loss": 0.4511,
+      "step": 4028
+    },
+    {
+      "epoch": 4.726846424384525,
+      "grad_norm": 0.375,
+      "learning_rate": 7.284719623267044e-05,
+      "loss": 0.4183,
+      "step": 4032
+    },
+    {
+      "epoch": 4.731535756154748,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.257389783722548e-05,
+      "loss": 0.4115,
+      "step": 4036
+    },
+    {
+      "epoch": 4.7362250879249705,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.2300949374895e-05,
+      "loss": 0.4635,
+      "step": 4040
+    },
+    {
+      "epoch": 4.740914419695193,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.202835207929014e-05,
+      "loss": 0.4566,
+      "step": 4044
+    },
+    {
+      "epoch": 4.745603751465416,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.175610718243493e-05,
+      "loss": 0.4718,
+      "step": 4048
+    },
+    {
+      "epoch": 4.750293083235639,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.148421591476086e-05,
+      "loss": 0.4123,
+      "step": 4052
+    },
+    {
+      "epoch": 4.754982415005862,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.121267950510082e-05,
+      "loss": 0.4439,
+      "step": 4056
+    },
+    {
+      "epoch": 4.7596717467760845,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.094149918068432e-05,
+      "loss": 0.4509,
+      "step": 4060
+    },
+    {
+      "epoch": 4.764361078546307,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.067067616713117e-05,
+      "loss": 0.4291,
+      "step": 4064
+    },
+    {
+      "epoch": 4.76905041031653,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.040021168844653e-05,
+      "loss": 0.4565,
+      "step": 4068
+    },
+    {
+      "epoch": 4.773739742086753,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.013010696701502e-05,
+      "loss": 0.4885,
+      "step": 4072
+    },
+    {
+      "epoch": 4.778429073856976,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.986036322359522e-05,
+      "loss": 0.3944,
+      "step": 4076
+    },
+    {
+      "epoch": 4.7831184056271985,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.959098167731447e-05,
+      "loss": 0.4708,
+      "step": 4080
+    },
+    {
+      "epoch": 4.78780773739742,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.93219635456629e-05,
+      "loss": 0.4989,
+      "step": 4084
+    },
+    {
+      "epoch": 4.792497069167643,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.905331004448843e-05,
+      "loss": 0.4702,
+      "step": 4088
+    },
+    {
+      "epoch": 4.797186400937866,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.878502238799062e-05,
+      "loss": 0.4528,
+      "step": 4092
+    },
+    {
+      "epoch": 4.801875732708089,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.851710178871596e-05,
+      "loss": 0.4384,
+      "step": 4096
+    },
+    {
+      "epoch": 4.8065650644783116,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.824954945755177e-05,
+      "loss": 0.4377,
+      "step": 4100
+    },
+    {
+      "epoch": 4.811254396248534,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.798236660372095e-05,
+      "loss": 0.4406,
+      "step": 4104
+    },
+    {
+      "epoch": 4.815943728018757,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.77155544347767e-05,
+      "loss": 0.4855,
+      "step": 4108
+    },
+    {
+      "epoch": 4.82063305978898,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.744911415659665e-05,
+      "loss": 0.4849,
+      "step": 4112
+    },
+    {
+      "epoch": 4.825322391559203,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.718304697337785e-05,
+      "loss": 0.4904,
+      "step": 4116
+    },
+    {
+      "epoch": 4.8300117233294255,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.691735408763097e-05,
+      "loss": 0.4598,
+      "step": 4120
+    },
+    {
+      "epoch": 4.834701055099648,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.66520367001751e-05,
+      "loss": 0.4602,
+      "step": 4124
+    },
+    {
+      "epoch": 4.839390386869871,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.638709601013215e-05,
+      "loss": 0.4598,
+      "step": 4128
+    },
+    {
+      "epoch": 4.844079718640094,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.612253321492157e-05,
+      "loss": 0.438,
+      "step": 4132
+    },
+    {
+      "epoch": 4.848769050410317,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.585834951025496e-05,
+      "loss": 0.4723,
+      "step": 4136
+    },
+    {
+      "epoch": 4.8534583821805395,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.559454609013043e-05,
+      "loss": 0.4737,
+      "step": 4140
+    },
+    {
+      "epoch": 4.858147713950762,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.533112414682754e-05,
+      "loss": 0.4772,
+      "step": 4144
+    },
+    {
+      "epoch": 4.862837045720985,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.506808487090163e-05,
+      "loss": 0.4312,
+      "step": 4148
+    },
+    {
+      "epoch": 4.867526377491208,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.48054294511785e-05,
+      "loss": 0.4765,
+      "step": 4152
+    },
+    {
+      "epoch": 4.872215709261431,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.454315907474926e-05,
+      "loss": 0.4566,
+      "step": 4156
+    },
+    {
+      "epoch": 4.8769050410316535,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.428127492696454e-05,
+      "loss": 0.4499,
+      "step": 4160
+    },
+    {
+      "epoch": 4.881594372801875,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.401977819142972e-05,
+      "loss": 0.4875,
+      "step": 4164
+    },
+    {
+      "epoch": 4.886283704572098,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.375867004999882e-05,
+      "loss": 0.4595,
+      "step": 4168
+    },
+    {
+      "epoch": 4.890973036342321,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.349795168276994e-05,
+      "loss": 0.4624,
+      "step": 4172
+    },
+    {
+      "epoch": 4.895662368112544,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.323762426807939e-05,
+      "loss": 0.4611,
+      "step": 4176
+    },
+    {
+      "epoch": 4.9003516998827665,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.297768898249649e-05,
+      "loss": 0.4109,
+      "step": 4180
+    },
+    {
+      "epoch": 4.905041031652989,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.271814700081852e-05,
+      "loss": 0.4552,
+      "step": 4184
+    },
+    {
+      "epoch": 4.909730363423212,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.245899949606498e-05,
+      "loss": 0.4127,
+      "step": 4188
+    },
+    {
+      "epoch": 4.914419695193435,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.220024763947263e-05,
+      "loss": 0.4686,
+      "step": 4192
+    },
+    {
+      "epoch": 4.919109026963658,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.194189260049003e-05,
+      "loss": 0.4711,
+      "step": 4196
+    },
+    {
+      "epoch": 4.9237983587338805,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.168393554677224e-05,
+      "loss": 0.4401,
+      "step": 4200
+    },
+    {
+      "epoch": 4.928487690504103,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.142637764417566e-05,
+      "loss": 0.4706,
+      "step": 4204
+    },
+    {
+      "epoch": 4.933177022274326,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.116922005675262e-05,
+      "loss": 0.4727,
+      "step": 4208
+    },
+    {
+      "epoch": 4.937866354044549,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.0912463946746346e-05,
+      "loss": 0.4737,
+      "step": 4212
+    },
+    {
+      "epoch": 4.942555685814772,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.065611047458538e-05,
+      "loss": 0.4833,
+      "step": 4216
+    },
+    {
+      "epoch": 4.9472450175849945,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.04001607988787e-05,
+      "loss": 0.4372,
+      "step": 4220
+    },
+    {
+      "epoch": 4.951934349355217,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.0144616076410114e-05,
+      "loss": 0.4829,
+      "step": 4224
+    },
+    {
+      "epoch": 4.956623681125439,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.9889477462133234e-05,
+      "loss": 0.428,
+      "step": 4228
+    },
+    {
+      "epoch": 4.961313012895662,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.963474610916643e-05,
+      "loss": 0.4963,
+      "step": 4232
+    },
+    {
+      "epoch": 4.966002344665885,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.938042316878719e-05,
+      "loss": 0.5161,
+      "step": 4236
+    },
+    {
+      "epoch": 4.9706916764361075,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.912650979042729e-05,
+      "loss": 0.4917,
+      "step": 4240
+    },
+    {
+      "epoch": 4.97538100820633,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.8873007121667314e-05,
+      "loss": 0.4658,
+      "step": 4244
+    },
+    {
+      "epoch": 4.980070339976553,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.861991630823185e-05,
+      "loss": 0.4614,
+      "step": 4248
+    },
+    {
+      "epoch": 4.984759671746776,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.8367238493983885e-05,
+      "loss": 0.428,
+      "step": 4252
+    },
+    {
+      "epoch": 4.989449003516999,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.81149748209198e-05,
+      "loss": 0.4652,
+      "step": 4256
+    },
+    {
+      "epoch": 4.9941383352872215,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.7863126429164445e-05,
+      "loss": 0.4367,
+      "step": 4260
+    },
+    {
+      "epoch": 4.998827667057444,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.761169445696552e-05,
+      "loss": 0.4746,
+      "step": 4264
+    },
+    {
+      "epoch": 5.003516998827667,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.7360680040688915e-05,
+      "loss": 0.5014,
+      "step": 4268
+    },
+    {
+      "epoch": 5.00820633059789,
+      "grad_norm": 0.3046875,
+      "learning_rate": 5.711008431481318e-05,
+      "loss": 0.4015,
+      "step": 4272
+    },
+    {
+      "epoch": 5.012895662368113,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.6859908411924634e-05,
+      "loss": 0.3825,
+      "step": 4276
+    },
+    {
+      "epoch": 5.0175849941383355,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.6610153462712144e-05,
+      "loss": 0.4382,
+      "step": 4280
+    },
+    {
+      "epoch": 5.022274325908558,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.6360820595962e-05,
+      "loss": 0.4462,
+      "step": 4284
+    },
+    {
+      "epoch": 5.026963657678781,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.611191093855304e-05,
+      "loss": 0.4038,
+      "step": 4288
+    },
+    {
+      "epoch": 5.031652989449004,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.5863425615451144e-05,
+      "loss": 0.434,
+      "step": 4292
+    },
+    {
+      "epoch": 5.036342321219227,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.5615365749704586e-05,
+      "loss": 0.4023,
+      "step": 4296
+    },
+    {
+      "epoch": 5.041031652989449,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.536773246243861e-05,
+      "loss": 0.3909,
+      "step": 4300
+    },
+    {
+      "epoch": 5.045720984759671,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.512052687285052e-05,
+      "loss": 0.4292,
+      "step": 4304
+    },
+    {
+      "epoch": 5.050410316529894,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.487375009820477e-05,
+      "loss": 0.4089,
+      "step": 4308
+    },
+    {
+      "epoch": 5.055099648300117,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.4627403253827436e-05,
+      "loss": 0.442,
+      "step": 4312
+    },
+    {
+      "epoch": 5.05978898007034,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.438148745310182e-05,
+      "loss": 0.4045,
+      "step": 4316
+    },
+    {
+      "epoch": 5.0644783118405625,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.413600380746286e-05,
+      "loss": 0.3942,
+      "step": 4320
+    },
+    {
+      "epoch": 5.069167643610785,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.3890953426392544e-05,
+      "loss": 0.3925,
+      "step": 4324
+    },
+    {
+      "epoch": 5.073856975381008,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.364633741741448e-05,
+      "loss": 0.4398,
+      "step": 4328
+    },
+    {
+      "epoch": 5.078546307151231,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.340215688608918e-05,
+      "loss": 0.4871,
+      "step": 4332
+    },
+    {
+      "epoch": 5.083235638921454,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.315841293600906e-05,
+      "loss": 0.4335,
+      "step": 4336
+    },
+    {
+      "epoch": 5.0879249706916765,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.2915106668793214e-05,
+      "loss": 0.4057,
+      "step": 4340
+    },
+    {
+      "epoch": 5.092614302461899,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.2672239184082845e-05,
+      "loss": 0.4184,
+      "step": 4344
+    },
+    {
+      "epoch": 5.097303634232122,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.242981157953567e-05,
+      "loss": 0.399,
+      "step": 4348
+    },
+    {
+      "epoch": 5.101992966002345,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.2187824950821725e-05,
+      "loss": 0.4237,
+      "step": 4352
+    },
+    {
+      "epoch": 5.106682297772568,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.194628039161778e-05,
+      "loss": 0.454,
+      "step": 4356
+    },
+    {
+      "epoch": 5.1113716295427905,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.170517899360267e-05,
+      "loss": 0.4397,
+      "step": 4360
+    },
+    {
+      "epoch": 5.116060961313013,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.1464521846452464e-05,
+      "loss": 0.4265,
+      "step": 4364
+    },
+    {
+      "epoch": 5.120750293083236,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.12243100378352e-05,
+      "loss": 0.4019,
+      "step": 4368
+    },
+    {
+      "epoch": 5.125439624853458,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.098454465340638e-05,
+      "loss": 0.3929,
+      "step": 4372
+    },
+    {
+      "epoch": 5.130128956623681,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.074522677680372e-05,
+      "loss": 0.416,
+      "step": 4376
+    },
+    {
+      "epoch": 5.1348182883939035,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.050635748964239e-05,
+      "loss": 0.4315,
+      "step": 4380
+    },
+    {
+      "epoch": 5.139507620164126,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.0267937871510304e-05,
+      "loss": 0.3818,
+      "step": 4384
+    },
+    {
+      "epoch": 5.144196951934349,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.0029968999962726e-05,
+      "loss": 0.3975,
+      "step": 4388
+    },
+    {
+      "epoch": 5.148886283704572,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.97924519505181e-05,
+      "loss": 0.4051,
+      "step": 4392
+    },
+    {
+      "epoch": 5.153575615474795,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.955538779665256e-05,
+      "loss": 0.3967,
+      "step": 4396
+    },
+    {
+      "epoch": 5.1582649472450175,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.9318777609795536e-05,
+      "loss": 0.4136,
+      "step": 4400
+    },
+    {
+      "epoch": 5.16295427901524,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.90826224593246e-05,
+      "loss": 0.4419,
+      "step": 4404
+    },
+    {
+      "epoch": 5.167643610785463,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.884692341256072e-05,
+      "loss": 0.45,
+      "step": 4408
+    },
+    {
+      "epoch": 5.172332942555686,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.8611681534763635e-05,
+      "loss": 0.389,
+      "step": 4412
+    },
+    {
+      "epoch": 5.177022274325909,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.837689788912667e-05,
+      "loss": 0.4161,
+      "step": 4416
+    },
+    {
+      "epoch": 5.1817116060961315,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.814257353677241e-05,
+      "loss": 0.3891,
+      "step": 4420
+    },
+    {
+      "epoch": 5.186400937866354,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.7908709536747224e-05,
+      "loss": 0.4633,
+      "step": 4424
+    },
+    {
+      "epoch": 5.191090269636577,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.7675306946017296e-05,
+      "loss": 0.4279,
+      "step": 4428
+    },
+    {
+      "epoch": 5.1957796014068,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.74423668194632e-05,
+      "loss": 0.4269,
+      "step": 4432
+    },
+    {
+      "epoch": 5.200468933177023,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.720989020987535e-05,
+      "loss": 0.4207,
+      "step": 4436
+    },
+    {
+      "epoch": 5.205158264947245,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.697787816794947e-05,
+      "loss": 0.492,
+      "step": 4440
+    },
+    {
+      "epoch": 5.209847596717467,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.674633174228138e-05,
+      "loss": 0.4622,
+      "step": 4444
+    },
+    {
+      "epoch": 5.21453692848769,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.651525197936275e-05,
+      "loss": 0.4413,
+      "step": 4448
+    },
+    {
+      "epoch": 5.219226260257913,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.6284639923575934e-05,
+      "loss": 0.4275,
+      "step": 4452
+    },
+    {
+      "epoch": 5.223915592028136,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.6054496617189554e-05,
+      "loss": 0.3943,
+      "step": 4456
+    },
+    {
+      "epoch": 5.2286049237983585,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.582482310035365e-05,
+      "loss": 0.4183,
+      "step": 4460
+    },
+    {
+      "epoch": 5.233294255568581,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.559562041109499e-05,
+      "loss": 0.439,
+      "step": 4464
+    },
+    {
+      "epoch": 5.237983587338804,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.53668895853125e-05,
+      "loss": 0.4362,
+      "step": 4468
+    },
+    {
+      "epoch": 5.242672919109027,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.5138631656772346e-05,
+      "loss": 0.4021,
+      "step": 4472
+    },
+    {
+      "epoch": 5.24736225087925,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.4910847657103555e-05,
+      "loss": 0.4298,
+      "step": 4476
+    },
+    {
+      "epoch": 5.2520515826494725,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.468353861579306e-05,
+      "loss": 0.3793,
+      "step": 4480
+    },
+    {
+      "epoch": 5.256740914419695,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.44567055601812e-05,
+      "loss": 0.3934,
+      "step": 4484
+    },
+    {
+      "epoch": 5.261430246189918,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.423034951545718e-05,
+      "loss": 0.3962,
+      "step": 4488
+    },
+    {
+      "epoch": 5.266119577960141,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.4004471504654196e-05,
+      "loss": 0.4206,
+      "step": 4492
+    },
+    {
+      "epoch": 5.270808909730364,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.377907254864496e-05,
+      "loss": 0.3844,
+      "step": 4496
+    },
+    {
+      "epoch": 5.275498241500586,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.355415366613702e-05,
+      "loss": 0.4442,
+      "step": 4500
+    },
+    {
+      "epoch": 5.280187573270809,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.332971587366837e-05,
+      "loss": 0.4124,
+      "step": 4504
+    },
+    {
+      "epoch": 5.284876905041032,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.3105760185602476e-05,
+      "loss": 0.443,
+      "step": 4508
+    },
+    {
+      "epoch": 5.289566236811254,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.2882287614123965e-05,
+      "loss": 0.4309,
+      "step": 4512
+    },
+    {
+      "epoch": 5.294255568581477,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.2659299169234056e-05,
+      "loss": 0.4248,
+      "step": 4516
+    },
+    {
+      "epoch": 5.2989449003516995,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.24367958587458e-05,
+      "loss": 0.4564,
+      "step": 4520
+    },
+    {
+      "epoch": 5.303634232121922,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.221477868827978e-05,
+      "loss": 0.4004,
+      "step": 4524
+    },
+    {
+      "epoch": 5.308323563892145,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.1993248661259324e-05,
+      "loss": 0.4194,
+      "step": 4528
+    },
+    {
+      "epoch": 5.313012895662368,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.1772206778906104e-05,
+      "loss": 0.3892,
+      "step": 4532
+    },
+    {
+      "epoch": 5.317702227432591,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.155165404023561e-05,
+      "loss": 0.415,
+      "step": 4536
+    },
+    {
+      "epoch": 5.3223915592028135,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.1331591442052534e-05,
+      "loss": 0.4516,
+      "step": 4540
+    },
+    {
+      "epoch": 5.327080890973036,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.111201997894651e-05,
+      "loss": 0.3757,
+      "step": 4544
+    },
+    {
+      "epoch": 5.331770222743259,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.089294064328725e-05,
+      "loss": 0.3921,
+      "step": 4548
+    },
+    {
+      "epoch": 5.336459554513482,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.067435442522043e-05,
+      "loss": 0.4161,
+      "step": 4552
+    },
+    {
+      "epoch": 5.341148886283705,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.045626231266294e-05,
+      "loss": 0.455,
+      "step": 4556
+    },
+    {
+      "epoch": 5.3458382180539274,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.023866529129848e-05,
+      "loss": 0.424,
+      "step": 4560
+    },
+    {
+      "epoch": 5.35052754982415,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.002156434457333e-05,
+      "loss": 0.4199,
+      "step": 4564
+    },
+    {
+      "epoch": 5.355216881594373,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.980496045369155e-05,
+      "loss": 0.4172,
+      "step": 4568
+    },
+    {
+      "epoch": 5.359906213364596,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.95888545976108e-05,
+      "loss": 0.4208,
+      "step": 4572
+    },
+    {
+      "epoch": 5.364595545134819,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.937324775303773e-05,
+      "loss": 0.3733,
+      "step": 4576
+    },
+    {
+      "epoch": 5.369284876905041,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.915814089442388e-05,
+      "loss": 0.4418,
+      "step": 4580
+    },
+    {
+      "epoch": 5.373974208675264,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.894353499396086e-05,
+      "loss": 0.4322,
+      "step": 4584
+    },
+    {
+      "epoch": 5.378663540445486,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.872943102157622e-05,
+      "loss": 0.4462,
+      "step": 4588
+    },
+    {
+      "epoch": 5.383352872215709,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.851582994492912e-05,
+      "loss": 0.4307,
+      "step": 4592
+    },
+    {
+      "epoch": 5.388042203985932,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.830273272940564e-05,
+      "loss": 0.4003,
+      "step": 4596
+    },
+    {
+      "epoch": 5.3927315357561545,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.8090140338114843e-05,
+      "loss": 0.4188,
+      "step": 4600
+    },
+    {
+      "epoch": 5.397420867526377,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.787805373188405e-05,
+      "loss": 0.4103,
+      "step": 4604
+    },
+    {
+      "epoch": 5.4021101992966,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.766647386925467e-05,
+      "loss": 0.4376,
+      "step": 4608
+    },
+    {
+      "epoch": 5.406799531066823,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.745540170647788e-05,
+      "loss": 0.4164,
+      "step": 4612
+    },
+    {
+      "epoch": 5.411488862837046,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.724483819751022e-05,
+      "loss": 0.4355,
+      "step": 4616
+    },
+    {
+      "epoch": 5.4161781946072685,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.703478429400945e-05,
+      "loss": 0.4503,
+      "step": 4620
+    },
+    {
+      "epoch": 5.420867526377491,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.6825240945329946e-05,
+      "loss": 0.4262,
+      "step": 4624
+    },
+    {
+      "epoch": 5.425556858147714,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.661620909851878e-05,
+      "loss": 0.4333,
+      "step": 4628
+    },
+    {
+      "epoch": 5.430246189917937,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.640768969831113e-05,
+      "loss": 0.4332,
+      "step": 4632
+    },
+    {
+      "epoch": 5.43493552168816,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.619968368712613e-05,
+      "loss": 0.3988,
+      "step": 4636
+    },
+    {
+      "epoch": 5.439624853458382,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.599219200506277e-05,
+      "loss": 0.3978,
+      "step": 4640
+    },
+    {
+      "epoch": 5.444314185228605,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.5785215589895224e-05,
+      "loss": 0.4265,
+      "step": 4644
+    },
+    {
+      "epoch": 5.449003516998828,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.557875537706914e-05,
+      "loss": 0.4031,
+      "step": 4648
+    },
+    {
+      "epoch": 5.453692848769051,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.5372812299696934e-05,
+      "loss": 0.4482,
+      "step": 4652
+    },
+    {
+      "epoch": 5.458382180539273,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.5167387288554014e-05,
+      "loss": 0.3847,
+      "step": 4656
+    },
+    {
+      "epoch": 5.4630715123094955,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.496248127207415e-05,
+      "loss": 0.4316,
+      "step": 4660
+    },
+    {
+      "epoch": 5.467760844079718,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.475809517634554e-05,
+      "loss": 0.4049,
+      "step": 4664
+    },
+    {
+      "epoch": 5.472450175849941,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.455422992510664e-05,
+      "loss": 0.43,
+      "step": 4668
+    },
+    {
+      "epoch": 5.477139507620164,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.435088643974177e-05,
+      "loss": 0.4267,
+      "step": 4672
+    },
+    {
+      "epoch": 5.481828839390387,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.41480656392773e-05,
+      "loss": 0.4264,
+      "step": 4676
+    },
+    {
+      "epoch": 5.4865181711606095,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.394576844037695e-05,
+      "loss": 0.3992,
+      "step": 4680
+    },
+    {
+      "epoch": 5.491207502930832,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.374399575733835e-05,
+      "loss": 0.4328,
+      "step": 4684
+    },
+    {
+      "epoch": 5.495896834701055,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.3542748502088325e-05,
+      "loss": 0.4447,
+      "step": 4688
+    },
+    {
+      "epoch": 5.500586166471278,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.334202758417896e-05,
+      "loss": 0.4483,
+      "step": 4692
+    },
+    {
+      "epoch": 5.505275498241501,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.314183391078373e-05,
+      "loss": 0.4673,
+      "step": 4696
+    },
+    {
+      "epoch": 5.509964830011723,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.294216838669295e-05,
+      "loss": 0.4609,
+      "step": 4700
+    },
+    {
+      "epoch": 5.514654161781946,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.2743031914310104e-05,
+      "loss": 0.4391,
+      "step": 4704
+    },
+    {
+      "epoch": 5.519343493552169,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.254442539364749e-05,
+      "loss": 0.4274,
+      "step": 4708
+    },
+    {
+      "epoch": 5.524032825322392,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.2346349722322274e-05,
+      "loss": 0.3941,
+      "step": 4712
+    },
+    {
+      "epoch": 5.528722157092615,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.2148805795552406e-05,
+      "loss": 0.3779,
+      "step": 4716
+    },
+    {
+      "epoch": 5.533411488862837,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.195179450615252e-05,
+      "loss": 0.3988,
+      "step": 4720
+    },
+    {
+      "epoch": 5.53810082063306,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.175531674453012e-05,
+      "loss": 0.4613,
+      "step": 4724
+    },
+    {
+      "epoch": 5.542790152403283,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.155937339868117e-05,
+      "loss": 0.4266,
+      "step": 4728
+    },
+    {
+      "epoch": 5.547479484173505,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.136396535418653e-05,
+      "loss": 0.4355,
+      "step": 4732
+    },
+    {
+      "epoch": 5.552168815943728,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.1169093494207547e-05,
+      "loss": 0.4324,
+      "step": 4736
+    },
+    {
+      "epoch": 5.5568581477139505,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.097475869948228e-05,
+      "loss": 0.4061,
+      "step": 4740
+    },
+    {
+      "epoch": 5.561547479484173,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.078096184832158e-05,
+      "loss": 0.4486,
+      "step": 4744
+    },
+    {
+      "epoch": 5.566236811254396,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.058770381660487e-05,
+      "loss": 0.4335,
+      "step": 4748
+    },
+    {
+      "epoch": 5.570926143024619,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.0394985477776522e-05,
+      "loss": 0.4712,
+      "step": 4752
+    },
+    {
+      "epoch": 5.575615474794842,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.0202807702841493e-05,
+      "loss": 0.4249,
+      "step": 4756
+    },
+    {
+      "epoch": 5.5803048065650644,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.0011171360361815e-05,
+      "loss": 0.4202,
+      "step": 4760
+    },
+    {
+      "epoch": 5.584994138335287,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9820077316452417e-05,
+      "loss": 0.4492,
+      "step": 4764
+    },
+    {
+      "epoch": 5.58968347010551,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.962952643477718e-05,
+      "loss": 0.4444,
+      "step": 4768
+    },
+    {
+      "epoch": 5.594372801875733,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.9439519576545302e-05,
+      "loss": 0.4112,
+      "step": 4772
+    },
+    {
+      "epoch": 5.599062133645956,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.925005760050704e-05,
+      "loss": 0.4111,
+      "step": 4776
+    },
+    {
+      "epoch": 5.603751465416178,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.906114136295018e-05,
+      "loss": 0.423,
+      "step": 4780
+    },
+    {
+      "epoch": 5.608440797186401,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.8872771717695858e-05,
+      "loss": 0.4275,
+      "step": 4784
+    },
+    {
+      "epoch": 5.613130128956624,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.8684949516094947e-05,
+      "loss": 0.4399,
+      "step": 4788
+    },
+    {
+      "epoch": 5.617819460726847,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.8497675607024046e-05,
+      "loss": 0.4434,
+      "step": 4792
+    },
+    {
+      "epoch": 5.622508792497069,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.831095083688169e-05,
+      "loss": 0.4714,
+      "step": 4796
+    },
+    {
+      "epoch": 5.6271981242672915,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.812477604958465e-05,
+      "loss": 0.4226,
+      "step": 4800
+    },
+    {
+      "epoch": 5.631887456037514,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.793915208656387e-05,
+      "loss": 0.4441,
+      "step": 4804
+    },
+    {
+      "epoch": 5.636576787807737,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.775407978676093e-05,
+      "loss": 0.441,
+      "step": 4808
+    },
+    {
+      "epoch": 5.64126611957796,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.7569559986624023e-05,
+      "loss": 0.425,
+      "step": 4812
+    },
+    {
+      "epoch": 5.645955451348183,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7385593520104276e-05,
+      "loss": 0.4097,
+      "step": 4816
+    },
+    {
+      "epoch": 5.6506447831184055,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.7202181218652113e-05,
+      "loss": 0.4147,
+      "step": 4820
+    },
+    {
+      "epoch": 5.655334114888628,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.701932391121323e-05,
+      "loss": 0.3897,
+      "step": 4824
+    },
+    {
+      "epoch": 5.660023446658851,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.6837022424225048e-05,
+      "loss": 0.4353,
+      "step": 4828
+    },
+    {
+      "epoch": 5.664712778429074,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.6655277581612838e-05,
+      "loss": 0.4483,
+      "step": 4832
+    },
+    {
+      "epoch": 5.669402110199297,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.647409020478623e-05,
+      "loss": 0.3887,
+      "step": 4836
+    },
+    {
+      "epoch": 5.674091441969519,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.629346111263521e-05,
+      "loss": 0.4746,
+      "step": 4840
+    },
+    {
+      "epoch": 5.678780773739742,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6113391121526573e-05,
+      "loss": 0.4224,
+      "step": 4844
+    },
+    {
+      "epoch": 5.683470105509965,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.593388104530031e-05,
+      "loss": 0.4432,
+      "step": 4848
+    },
+    {
+      "epoch": 5.688159437280188,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.5754931695265674e-05,
+      "loss": 0.4395,
+      "step": 4852
+    },
+    {
+      "epoch": 5.692848769050411,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.5576543880197847e-05,
+      "loss": 0.4717,
+      "step": 4856
+    },
+    {
+      "epoch": 5.697538100820633,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.539871840633399e-05,
+      "loss": 0.4418,
+      "step": 4860
+    },
+    {
+      "epoch": 5.702227432590856,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.522145607736976e-05,
+      "loss": 0.4035,
+      "step": 4864
+    },
+    {
+      "epoch": 5.706916764361079,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.5044757694455642e-05,
+      "loss": 0.4121,
+      "step": 4868
+    },
+    {
+      "epoch": 5.711606096131302,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.4868624056193264e-05,
+      "loss": 0.4117,
+      "step": 4872
+    },
+    {
+      "epoch": 5.716295427901524,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.469305595863199e-05,
+      "loss": 0.4515,
+      "step": 4876
+    },
+    {
+      "epoch": 5.7209847596717465,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.4518054195265024e-05,
+      "loss": 0.4104,
+      "step": 4880
+    },
+    {
+      "epoch": 5.725674091441969,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.4343619557026102e-05,
+      "loss": 0.4226,
+      "step": 4884
+    },
+    {
+      "epoch": 5.730363423212192,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4169752832285723e-05,
+      "loss": 0.4267,
+      "step": 4888
+    },
+    {
+      "epoch": 5.735052754982415,
+      "grad_norm": 0.375,
+      "learning_rate": 2.3996454806847624e-05,
+      "loss": 0.4508,
+      "step": 4892
+    },
+    {
+      "epoch": 5.739742086752638,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3823726263945442e-05,
+      "loss": 0.4046,
+      "step": 4896
+    },
+    {
+      "epoch": 5.74443141852286,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.3651567984238707e-05,
+      "loss": 0.4372,
+      "step": 4900
+    },
+    {
+      "epoch": 5.749120750293083,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.3479980745809885e-05,
+      "loss": 0.4356,
+      "step": 4904
+    },
+    {
+      "epoch": 5.753810082063306,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3308965324160374e-05,
+      "loss": 0.4222,
+      "step": 4908
+    },
+    {
+      "epoch": 5.758499413833529,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.313852249220735e-05,
+      "loss": 0.4257,
+      "step": 4912
+    },
+    {
+      "epoch": 5.763188745603752,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.2968653020280036e-05,
+      "loss": 0.387,
+      "step": 4916
+    },
+    {
+      "epoch": 5.767878077373974,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.2799357676116287e-05,
+      "loss": 0.3745,
+      "step": 4920
+    },
+    {
+      "epoch": 5.772567409144197,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.2630637224859283e-05,
+      "loss": 0.4284,
+      "step": 4924
+    },
+    {
+      "epoch": 5.77725674091442,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.246249242905377e-05,
+      "loss": 0.4014,
+      "step": 4928
+    },
+    {
+      "epoch": 5.781946072684643,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.2294924048642888e-05,
+      "loss": 0.4091,
+      "step": 4932
+    },
+    {
+      "epoch": 5.786635404454865,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.212793284096458e-05,
+      "loss": 0.4428,
+      "step": 4936
+    },
+    {
+      "epoch": 5.7913247362250875,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.196151956074821e-05,
+      "loss": 0.4281,
+      "step": 4940
+    },
+    {
+      "epoch": 5.79601406799531,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.179568496011116e-05,
+      "loss": 0.4555,
+      "step": 4944
+    },
+    {
+      "epoch": 5.800703399765533,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.1630429788555376e-05,
+      "loss": 0.4294,
+      "step": 4948
+    },
+    {
+      "epoch": 5.805392731535756,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.146575479296418e-05,
+      "loss": 0.3904,
+      "step": 4952
+    },
+    {
+      "epoch": 5.810082063305979,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1301660717598575e-05,
+      "loss": 0.417,
+      "step": 4956
+    },
+    {
+      "epoch": 5.814771395076201,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.1138148304094177e-05,
+      "loss": 0.4715,
+      "step": 4960
+    },
+    {
+      "epoch": 5.819460726846424,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.0975218291457645e-05,
+      "loss": 0.403,
+      "step": 4964
+    },
+    {
+      "epoch": 5.824150058616647,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.0812871416063477e-05,
+      "loss": 0.399,
+      "step": 4968
+    },
+    {
+      "epoch": 5.82883939038687,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.0651108411650685e-05,
+      "loss": 0.4323,
+      "step": 4972
+    },
+    {
+      "epoch": 5.833528722157093,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.0489930009319287e-05,
+      "loss": 0.4407,
+      "step": 4976
+    },
+    {
+      "epoch": 5.838218053927315,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.0329336937527312e-05,
+      "loss": 0.4169,
+      "step": 4980
+    },
+    {
+      "epoch": 5.842907385697538,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.0169329922087218e-05,
+      "loss": 0.4336,
+      "step": 4984
+    },
+    {
+      "epoch": 5.847596717467761,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.000990968616287e-05,
+      "loss": 0.4445,
+      "step": 4988
+    },
+    {
+      "epoch": 5.852286049237984,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.985107695026601e-05,
+      "loss": 0.4177,
+      "step": 4992
+    },
+    {
+      "epoch": 5.856975381008207,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9692832432253154e-05,
+      "loss": 0.4066,
+      "step": 4996
+    },
+    {
+      "epoch": 5.861664712778429,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9535176847322416e-05,
+      "loss": 0.3923,
+      "step": 5000
+    },
+    {
+      "epoch": 5.866354044548652,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.937811090801004e-05,
+      "loss": 0.3981,
+      "step": 5004
+    },
+    {
+      "epoch": 5.871043376318875,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9221635324187513e-05,
+      "loss": 0.4235,
+      "step": 5008
+    },
+    {
+      "epoch": 5.875732708089098,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9065750803057907e-05,
+      "loss": 0.4204,
+      "step": 5012
+    },
+    {
+      "epoch": 5.88042203985932,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8910458049153173e-05,
+      "loss": 0.4582,
+      "step": 5016
+    },
+    {
+      "epoch": 5.8851113716295425,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8755757764330632e-05,
+      "loss": 0.4433,
+      "step": 5020
+    },
+    {
+      "epoch": 5.889800703399765,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.860165064776985e-05,
+      "loss": 0.4477,
+      "step": 5024
+    },
+    {
+      "epoch": 5.894490035169988,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8448137395969636e-05,
+      "loss": 0.3817,
+      "step": 5028
+    },
+    {
+      "epoch": 5.899179366940211,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8295218702744662e-05,
+      "loss": 0.4144,
+      "step": 5032
+    },
+    {
+      "epoch": 5.903868698710434,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8142895259222584e-05,
+      "loss": 0.4391,
+      "step": 5036
+    },
+    {
+      "epoch": 5.908558030480656,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7991167753840673e-05,
+      "loss": 0.4012,
+      "step": 5040
+    },
+    {
+      "epoch": 5.913247362250879,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.784003687234281e-05,
+      "loss": 0.4313,
+      "step": 5044
+    },
+    {
+      "epoch": 5.917936694021102,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7689503297776464e-05,
+      "loss": 0.4179,
+      "step": 5048
+    },
+    {
+      "epoch": 5.922626025791325,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.753956771048946e-05,
+      "loss": 0.4241,
+      "step": 5052
+    },
+    {
+      "epoch": 5.927315357561548,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7390230788127024e-05,
+      "loss": 0.4458,
+      "step": 5056
+    },
+    {
+      "epoch": 5.93200468933177,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7241493205628644e-05,
+      "loss": 0.3708,
+      "step": 5060
+    },
+    {
+      "epoch": 5.936694021101993,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.709335563522507e-05,
+      "loss": 0.4046,
+      "step": 5064
+    },
+    {
+      "epoch": 5.941383352872216,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6945818746435248e-05,
+      "loss": 0.4018,
+      "step": 5068
+    },
+    {
+      "epoch": 5.946072684642439,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6798883206063217e-05,
+      "loss": 0.437,
+      "step": 5072
+    },
+    {
+      "epoch": 5.950762016412662,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.665254967819532e-05,
+      "loss": 0.4529,
+      "step": 5076
+    },
+    {
+      "epoch": 5.9554513481828835,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6506818824196965e-05,
+      "loss": 0.4405,
+      "step": 5080
+    },
+    {
+      "epoch": 5.960140679953106,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.636169130270973e-05,
+      "loss": 0.4241,
+      "step": 5084
+    },
+    {
+      "epoch": 5.964830011723329,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6217167769648398e-05,
+      "loss": 0.4551,
+      "step": 5088
+    },
+    {
+      "epoch": 5.969519343493552,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6073248878198032e-05,
+      "loss": 0.427,
+      "step": 5092
+    },
+    {
+      "epoch": 5.974208675263775,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5929935278810883e-05,
+      "loss": 0.3894,
+      "step": 5096
+    },
+    {
+      "epoch": 5.978898007033997,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.578722761920359e-05,
+      "loss": 0.434,
+      "step": 5100
+    },
+    {
+      "epoch": 5.98358733880422,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5645126544354253e-05,
+      "loss": 0.4446,
+      "step": 5104
+    },
+    {
+      "epoch": 5.988276670574443,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.550363269649932e-05,
+      "loss": 0.4205,
+      "step": 5108
+    },
+    {
+      "epoch": 5.992966002344666,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.536274671513098e-05,
+      "loss": 0.4032,
+      "step": 5112
+    },
+    {
+      "epoch": 5.997655334114889,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5222469236994061e-05,
+      "loss": 0.4029,
+      "step": 5116
+    },
+    {
+      "epoch": 6.002344665885111,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5082800896083186e-05,
+      "loss": 0.3855,
+      "step": 5120
+    },
+    {
+      "epoch": 6.007033997655334,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4943742323639995e-05,
+      "loss": 0.4209,
+      "step": 5124
+    },
+    {
+      "epoch": 6.011723329425557,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4805294148150171e-05,
+      "loss": 0.4114,
+      "step": 5128
+    },
+    {
+      "epoch": 6.01641266119578,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4667456995340731e-05,
+      "loss": 0.4657,
+      "step": 5132
+    },
+    {
+      "epoch": 6.021101992966003,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4530231488177058e-05,
+      "loss": 0.4312,
+      "step": 5136
+    },
+    {
+      "epoch": 6.025791324736225,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4393618246860239e-05,
+      "loss": 0.43,
+      "step": 5140
+    },
+    {
+      "epoch": 6.030480656506448,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4257617888824096e-05,
+      "loss": 0.4133,
+      "step": 5144
+    },
+    {
+      "epoch": 6.035169988276671,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4122231028732516e-05,
+      "loss": 0.4065,
+      "step": 5148
+    },
+    {
+      "epoch": 6.039859320046894,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.398745827847667e-05,
+      "loss": 0.4284,
+      "step": 5152
+    },
+    {
+      "epoch": 6.044548651817116,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.385330024717215e-05,
+      "loss": 0.3909,
+      "step": 5156
+    },
+    {
+      "epoch": 6.049237983587338,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3719757541156317e-05,
+      "loss": 0.3842,
+      "step": 5160
+    },
+    {
+      "epoch": 6.053927315357561,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3586830763985479e-05,
+      "loss": 0.3678,
+      "step": 5164
+    },
+    {
+      "epoch": 6.058616647127784,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3454520516432282e-05,
+      "loss": 0.4318,
+      "step": 5168
+    },
+    {
+      "epoch": 6.063305978898007,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3322827396482888e-05,
+      "loss": 0.4121,
+      "step": 5172
+    },
+    {
+      "epoch": 6.06799531066823,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3191751999334237e-05,
+      "loss": 0.4247,
+      "step": 5176
+    },
+    {
+      "epoch": 6.072684642438452,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3061294917391558e-05,
+      "loss": 0.375,
+      "step": 5180
+    },
+    {
+      "epoch": 6.077373974208675,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2931456740265406e-05,
+      "loss": 0.4611,
+      "step": 5184
+    },
+    {
+      "epoch": 6.082063305978898,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2802238054769298e-05,
+      "loss": 0.4174,
+      "step": 5188
+    },
+    {
+      "epoch": 6.086752637749121,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2673639444916805e-05,
+      "loss": 0.4337,
+      "step": 5192
+    },
+    {
+      "epoch": 6.091441969519344,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2545661491919057e-05,
+      "loss": 0.3709,
+      "step": 5196
+    },
+    {
+      "epoch": 6.096131301289566,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2418304774182075e-05,
+      "loss": 0.452,
+      "step": 5200
+    },
+    {
+      "epoch": 6.100820633059789,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2291569867304112e-05,
+      "loss": 0.435,
+      "step": 5204
+    },
+    {
+      "epoch": 6.105509964830012,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2165457344073238e-05,
+      "loss": 0.4482,
+      "step": 5208
+    },
+    {
+      "epoch": 6.110199296600235,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2039967774464448e-05,
+      "loss": 0.3896,
+      "step": 5212
+    },
+    {
+      "epoch": 6.1148886283704575,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1915101725637383e-05,
+      "loss": 0.4166,
+      "step": 5216
+    },
+    {
+      "epoch": 6.11957796014068,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1790859761933563e-05,
+      "loss": 0.3914,
+      "step": 5220
+    },
+    {
+      "epoch": 6.124267291910903,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.166724244487387e-05,
+      "loss": 0.4055,
+      "step": 5224
+    },
+    {
+      "epoch": 6.128956623681125,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1544250333156207e-05,
+      "loss": 0.409,
+      "step": 5228
+    },
+    {
+      "epoch": 6.133645955451348,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.142188398265259e-05,
+      "loss": 0.3903,
+      "step": 5232
+    },
+    {
+      "epoch": 6.138335287221571,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1300143946407064e-05,
+      "loss": 0.397,
+      "step": 5236
+    },
+    {
+      "epoch": 6.143024618991793,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.1179030774632851e-05,
+      "loss": 0.3783,
+      "step": 5240
+    },
+    {
+      "epoch": 6.147713950762016,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1058545014710146e-05,
+      "loss": 0.4203,
+      "step": 5244
+    },
+    {
+      "epoch": 6.152403282532239,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.093868721118339e-05,
+      "loss": 0.433,
+      "step": 5248
+    },
+    {
+      "epoch": 6.157092614302462,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0819457905758978e-05,
+      "loss": 0.4351,
+      "step": 5252
+    },
+    {
+      "epoch": 6.161781946072685,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0700857637302779e-05,
+      "loss": 0.4498,
+      "step": 5256
+    },
+    {
+      "epoch": 6.166471277842907,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.058288694183762e-05,
+      "loss": 0.4151,
+      "step": 5260
+    },
+    {
+      "epoch": 6.17116060961313,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.0465546352541055e-05,
+      "loss": 0.4446,
+      "step": 5264
+    },
+    {
+      "epoch": 6.175849941383353,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.034883639974261e-05,
+      "loss": 0.4044,
+      "step": 5268
+    },
+    {
+      "epoch": 6.180539273153576,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0232757610921833e-05,
+      "loss": 0.3634,
+      "step": 5272
+    },
+    {
+      "epoch": 6.185228604923799,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0117310510705528e-05,
+      "loss": 0.4032,
+      "step": 5276
+    },
+    {
+      "epoch": 6.189917936694021,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0002495620865558e-05,
+      "loss": 0.4418,
+      "step": 5280
+    },
+    {
+      "epoch": 6.194607268464244,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.888313460316549e-06,
+      "loss": 0.4224,
+      "step": 5284
+    },
+    {
+      "epoch": 6.199296600234467,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.77476454511335e-06,
+      "loss": 0.4139,
+      "step": 5288
+    },
+    {
+      "epoch": 6.20398593200469,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.661849388448866e-06,
+      "loss": 0.3725,
+      "step": 5292
+    },
+    {
+      "epoch": 6.2086752637749125,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.549568500651695e-06,
+      "loss": 0.4023,
+      "step": 5296
+    },
+    {
+      "epoch": 6.213364595545134,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.437922389183772e-06,
+      "loss": 0.3973,
+      "step": 5300
+    },
+    {
+      "epoch": 6.218053927315357,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.3269115586381e-06,
+      "loss": 0.4492,
+      "step": 5304
+    },
+    {
+      "epoch": 6.22274325908558,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.216536510736528e-06,
+      "loss": 0.3933,
+      "step": 5308
+    },
+    {
+      "epoch": 6.227432590855803,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.106797744327449e-06,
+      "loss": 0.4262,
+      "step": 5312
+    },
+    {
+      "epoch": 6.232121922626026,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.997695755383444e-06,
+      "loss": 0.386,
+      "step": 5316
+    },
+    {
+      "epoch": 6.236811254396248,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.889231036999245e-06,
+      "loss": 0.4026,
+      "step": 5320
+    },
+    {
+      "epoch": 6.241500586166471,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.781404079389304e-06,
+      "loss": 0.4596,
+      "step": 5324
+    },
+    {
+      "epoch": 6.246189917936694,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.674215369885695e-06,
+      "loss": 0.4081,
+      "step": 5328
+    },
+    {
+      "epoch": 6.250879249706917,
+      "grad_norm": 0.369140625,
+      "learning_rate": 8.567665392935918e-06,
+      "loss": 0.4246,
+      "step": 5332
+    },
+    {
+      "epoch": 6.25556858147714,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.461754630100581e-06,
+      "loss": 0.4221,
+      "step": 5336
+    },
+    {
+      "epoch": 6.260257913247362,
+      "grad_norm": 0.375,
+      "learning_rate": 8.356483560051468e-06,
+      "loss": 0.4704,
+      "step": 5340
+    },
+    {
+      "epoch": 6.264947245017585,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.251852658569014e-06,
+      "loss": 0.3896,
+      "step": 5344
+    },
+    {
+      "epoch": 6.269636576787808,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.147862398540545e-06,
+      "loss": 0.4065,
+      "step": 5348
+    },
+    {
+      "epoch": 6.274325908558031,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.044513249957874e-06,
+      "loss": 0.4074,
+      "step": 5352
+    },
+    {
+      "epoch": 6.2790152403282535,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.94180567991528e-06,
+      "loss": 0.4237,
+      "step": 5356
+    },
+    {
+      "epoch": 6.283704572098476,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.839740152607398e-06,
+      "loss": 0.4043,
+      "step": 5360
+    },
+    {
+      "epoch": 6.288393903868699,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.738317129327049e-06,
+      "loss": 0.4098,
+      "step": 5364
+    },
+    {
+      "epoch": 6.293083235638921,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.63753706846329e-06,
+      "loss": 0.4149,
+      "step": 5368
+    },
+    {
+      "epoch": 6.297772567409144,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.537400425499191e-06,
+      "loss": 0.3909,
+      "step": 5372
+    },
+    {
+      "epoch": 6.302461899179367,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.437907653009878e-06,
+      "loss": 0.3893,
+      "step": 5376
+    },
+    {
+      "epoch": 6.307151230949589,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.339059200660441e-06,
+      "loss": 0.4365,
+      "step": 5380
+    },
+    {
+      "epoch": 6.311840562719812,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.240855515203897e-06,
+      "loss": 0.4586,
+      "step": 5384
+    },
+    {
+      "epoch": 6.316529894490035,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.143297040479262e-06,
+      "loss": 0.4519,
+      "step": 5388
+    },
+    {
+      "epoch": 6.321219226260258,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.046384217409401e-06,
+      "loss": 0.3872,
+      "step": 5392
+    },
+    {
+      "epoch": 6.325908558030481,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.950117483999145e-06,
+      "loss": 0.4004,
+      "step": 5396
+    },
+    {
+      "epoch": 6.330597889800703,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.854497275333282e-06,
+      "loss": 0.3972,
+      "step": 5400
+    },
+    {
+      "epoch": 6.335287221570926,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.759524023574514e-06,
+      "loss": 0.367,
+      "step": 5404
+    },
+    {
+      "epoch": 6.339976553341149,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.6651981579616705e-06,
+      "loss": 0.4688,
+      "step": 5408
+    },
+    {
+      "epoch": 6.344665885111372,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.57152010480762e-06,
+      "loss": 0.4207,
+      "step": 5412
+    },
+    {
+      "epoch": 6.3493552168815945,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.4784902874973734e-06,
+      "loss": 0.413,
+      "step": 5416
+    },
+    {
+      "epoch": 6.354044548651817,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.3861091264861995e-06,
+      "loss": 0.448,
+      "step": 5420
+    },
+    {
+      "epoch": 6.35873388042204,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.2943770392977826e-06,
+      "loss": 0.4018,
+      "step": 5424
+    },
+    {
+      "epoch": 6.363423212192263,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.203294440522183e-06,
+      "loss": 0.416,
+      "step": 5428
+    },
+    {
+      "epoch": 6.368112543962486,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.112861741814063e-06,
+      "loss": 0.4392,
+      "step": 5432
+    },
+    {
+      "epoch": 6.3728018757327085,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.023079351890881e-06,
+      "loss": 0.4195,
+      "step": 5436
+    },
+    {
+      "epoch": 6.377491207502931,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.933947676530881e-06,
+      "loss": 0.4029,
+      "step": 5440
+    },
+    {
+      "epoch": 6.382180539273153,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.845467118571445e-06,
+      "loss": 0.4476,
+      "step": 5444
+    },
+    {
+      "epoch": 6.386869871043376,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.757638077907123e-06,
+      "loss": 0.4326,
+      "step": 5448
+    },
+    {
+      "epoch": 6.391559202813599,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.670460951487854e-06,
+      "loss": 0.4126,
+      "step": 5452
+    },
+    {
+      "epoch": 6.396248534583822,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.583936133317285e-06,
+      "loss": 0.4201,
+      "step": 5456
+    },
+    {
+      "epoch": 6.400937866354044,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.498064014450837e-06,
+      "loss": 0.4105,
+      "step": 5460
+    },
+    {
+      "epoch": 6.405627198124267,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.4128449829940745e-06,
+      "loss": 0.4393,
+      "step": 5464
+    },
+    {
+      "epoch": 6.41031652989449,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.3282794241007895e-06,
+      "loss": 0.3989,
+      "step": 5468
+    },
+    {
+      "epoch": 6.415005861664713,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.244367719971454e-06,
+      "loss": 0.409,
+      "step": 5472
+    },
+    {
+      "epoch": 6.4196951934349356,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.161110249851353e-06,
+      "loss": 0.4121,
+      "step": 5476
+    },
+    {
+      "epoch": 6.424384525205158,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.078507390028852e-06,
+      "loss": 0.4181,
+      "step": 5480
+    },
+    {
+      "epoch": 6.429073856975381,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.996559513833903e-06,
+      "loss": 0.4494,
+      "step": 5484
+    },
+    {
+      "epoch": 6.433763188745604,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.915266991636025e-06,
+      "loss": 0.4472,
+      "step": 5488
+    },
+    {
+      "epoch": 6.438452520515827,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.83463019084297e-06,
+      "loss": 0.3834,
+      "step": 5492
+    },
+    {
+      "epoch": 6.4431418522860495,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.754649475898814e-06,
+      "loss": 0.3881,
+      "step": 5496
+    },
+    {
+      "epoch": 6.447831184056272,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.675325208282471e-06,
+      "loss": 0.4295,
+      "step": 5500
     }
   ],
   "logging_steps": 4,
-  "max_steps": 3661,
+  "max_steps": 5971,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 7,
   "save_steps": 500,
@@ -6440,7 +9651,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.3920283694378516e+18,
+  "total_flos": 3.5958144579529605e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null