diff --git "a/checkpoint-6500/trainer_state.json" "b/checkpoint-6500/trainer_state.json"
--- "a/checkpoint-6500/trainer_state.json"
+++ "b/checkpoint-6500/trainer_state.json"
@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 6.937033084311633,
+  "epoch": 6.653019447287615,
   "eval_steps": 500,
   "global_step": 6500,
   "is_hyper_param_search": false,
@@ -9,11383 +9,11383 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.004268943436499467,
+      "epoch": 0.0040941658137154556,
       "grad_norm": 0.283203125,
-      "learning_rate": 1.8181818181818182e-05,
-      "loss": 1.3096,
+      "learning_rate": 1.7391304347826085e-05,
+      "loss": 1.2477,
       "step": 4
     },
     {
-      "epoch": 0.008537886872998933,
-      "grad_norm": 0.291015625,
-      "learning_rate": 3.6363636363636364e-05,
-      "loss": 1.2874,
+      "epoch": 0.008188331627430911,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.478260869565217e-05,
+      "loss": 1.292,
       "step": 8
     },
     {
-      "epoch": 0.012806830309498399,
-      "grad_norm": 0.244140625,
-      "learning_rate": 5.454545454545454e-05,
-      "loss": 1.2297,
+      "epoch": 0.012282497441146366,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.2173913043478256e-05,
+      "loss": 1.201,
       "step": 12
     },
     {
-      "epoch": 0.017075773745997867,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 7.272727272727273e-05,
-      "loss": 1.1818,
+      "epoch": 0.016376663254861822,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.956521739130434e-05,
+      "loss": 1.218,
       "step": 16
     },
     {
-      "epoch": 0.021344717182497332,
-      "grad_norm": 0.244140625,
-      "learning_rate": 9.09090909090909e-05,
-      "loss": 1.2125,
+      "epoch": 0.02047082906857728,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.695652173913043e-05,
+      "loss": 1.2022,
       "step": 20
     },
     {
-      "epoch": 0.025613660618996798,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00010909090909090908,
-      "loss": 1.1827,
+      "epoch": 0.02456499488229273,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010434782608695651,
+      "loss": 1.1587,
       "step": 24
     },
     {
-      "epoch": 0.029882604055496264,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00012727272727272725,
-      "loss": 1.1747,
+      "epoch": 0.028659160696008188,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012173913043478261,
+      "loss": 1.1458,
       "step": 28
     },
     {
-      "epoch": 0.03415154749199573,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00014545454545454546,
-      "loss": 1.13,
+      "epoch": 0.032753326509723645,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013913043478260868,
+      "loss": 1.1289,
       "step": 32
     },
     {
-      "epoch": 0.0384204909284952,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0001636363636363636,
-      "loss": 1.1684,
+      "epoch": 0.0368474923234391,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00015652173913043477,
+      "loss": 1.1263,
       "step": 36
     },
     {
-      "epoch": 0.042689434364994665,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.0001818181818181818,
-      "loss": 1.0837,
+      "epoch": 0.04094165813715456,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00017391304347826085,
+      "loss": 1.1077,
       "step": 40
     },
     {
-      "epoch": 0.04695837780149413,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00019999999999999998,
-      "loss": 1.1134,
+      "epoch": 0.04503582395087001,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001913043478260869,
+      "loss": 1.1116,
       "step": 44
     },
     {
-      "epoch": 0.051227321237993596,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00021818181818181816,
-      "loss": 1.0833,
+      "epoch": 0.04912998976458546,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00020869565217391303,
+      "loss": 1.0815,
       "step": 48
     },
     {
-      "epoch": 0.05549626467449306,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00023636363636363633,
-      "loss": 1.0918,
+      "epoch": 0.05322415557830092,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002260869565217391,
+      "loss": 1.1132,
       "step": 52
     },
     {
-      "epoch": 0.05976520811099253,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0002545454545454545,
-      "loss": 1.0809,
+      "epoch": 0.057318321392016376,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00024347826086956522,
+      "loss": 1.1032,
       "step": 56
     },
     {
-      "epoch": 0.064034151547492,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0002727272727272727,
-      "loss": 1.088,
+      "epoch": 0.06141248720573183,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002608695652173913,
+      "loss": 1.0802,
       "step": 60
     },
     {
-      "epoch": 0.06830309498399147,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0002909090909090909,
-      "loss": 1.0458,
+      "epoch": 0.06550665301944729,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00027826086956521737,
+      "loss": 1.0692,
       "step": 64
     },
     {
-      "epoch": 0.07257203842049093,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0002999999297687884,
-      "loss": 1.0937,
+      "epoch": 0.06960081883316274,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00029565217391304345,
+      "loss": 1.0476,
       "step": 68
     },
     {
-      "epoch": 0.0768409818569904,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00029999936791949057,
-      "loss": 1.0496,
+      "epoch": 0.0736949846468782,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00029999985464629347,
+      "loss": 1.0535,
       "step": 72
     },
     {
-      "epoch": 0.08110992529348986,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002999982442229994,
-      "loss": 1.0712,
+      "epoch": 0.07778915046059365,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00029999920863038815,
+      "loss": 1.042,
       "step": 76
     },
     {
-      "epoch": 0.08537886872998933,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00029999655868352383,
-      "loss": 1.0374,
+      "epoch": 0.08188331627430911,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002999980458040957,
+      "loss": 1.0121,
       "step": 80
     },
     {
-      "epoch": 0.08964781216648879,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0002999943113073774,
-      "loss": 1.0349,
+      "epoch": 0.08597748208802457,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002999963661714225,
+      "loss": 1.0423,
       "step": 84
     },
     {
-      "epoch": 0.09391675560298826,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00029999150210297795,
-      "loss": 1.0417,
+      "epoch": 0.09007164790174002,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0002999941697381556,
+      "loss": 1.0429,
       "step": 88
     },
     {
-      "epoch": 0.09818569903948772,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0002999881310808479,
-      "loss": 1.0053,
+      "epoch": 0.09416581371545547,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002999914565118627,
+      "loss": 1.0316,
       "step": 92
     },
     {
-      "epoch": 0.10245464247598719,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002999841982536139,
-      "loss": 1.0185,
+      "epoch": 0.09825997952917093,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002999882265018919,
+      "loss": 1.0462,
       "step": 96
     },
     {
-      "epoch": 0.10672358591248667,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.000299979703636007,
-      "loss": 1.0001,
+      "epoch": 0.1023541453428864,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000299984479719372,
+      "loss": 1.0391,
       "step": 100
     },
     {
-      "epoch": 0.11099252934898612,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002999746472448626,
-      "loss": 1.0292,
+      "epoch": 0.10644831115660185,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00029998021617721224,
+      "loss": 1.0167,
       "step": 104
     },
     {
-      "epoch": 0.1152614727854856,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00029996902909912017,
-      "loss": 1.0005,
+      "epoch": 0.1105424769703173,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002999754358901023,
+      "loss": 1.001,
       "step": 108
     },
     {
-      "epoch": 0.11953041622198506,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.0002999628492198234,
-      "loss": 0.9856,
+      "epoch": 0.11463664278403275,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00029997013887451236,
+      "loss": 1.0101,
       "step": 112
     },
     {
-      "epoch": 0.12379935965848453,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00029995610763012,
-      "loss": 0.9663,
+      "epoch": 0.1187308085977482,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0002999643251486927,
+      "loss": 0.9859,
       "step": 116
     },
     {
-      "epoch": 0.128068303094984,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029994880435526184,
-      "loss": 0.9839,
+      "epoch": 0.12282497441146366,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002999579947326742,
+      "loss": 1.0245,
       "step": 120
     },
     {
-      "epoch": 0.13233724653148346,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0002999409394226044,
-      "loss": 1.0058,
+      "epoch": 0.1269191402251791,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002999511476482678,
+      "loss": 0.9762,
       "step": 124
     },
     {
-      "epoch": 0.13660618996798293,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.0002999325128616071,
-      "loss": 1.0166,
+      "epoch": 0.13101330603889458,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00029994378391906453,
+      "loss": 0.9698,
       "step": 128
     },
     {
-      "epoch": 0.14087513340448238,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00029992352470383307,
-      "loss": 0.9636,
+      "epoch": 0.13510747185261002,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002999359035704355,
+      "loss": 0.9787,
       "step": 132
     },
     {
-      "epoch": 0.14514407684098185,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.0002999139749829488,
-      "loss": 0.9827,
+      "epoch": 0.13920163766632548,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00029992750662953196,
+      "loss": 1.0197,
       "step": 136
     },
     {
-      "epoch": 0.14941302027748132,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029990386373472444,
-      "loss": 0.9433,
+      "epoch": 0.14329580348004095,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00029991859312528476,
+      "loss": 0.9958,
       "step": 140
     },
     {
-      "epoch": 0.1536819637139808,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.0002998931909970333,
-      "loss": 0.9327,
+      "epoch": 0.1473899692937564,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00029990916308840476,
+      "loss": 0.9495,
       "step": 144
     },
     {
-      "epoch": 0.15795090715048027,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002998819568098519,
-      "loss": 0.9729,
+      "epoch": 0.15148413510747186,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002998992165513824,
+      "loss": 1.0079,
       "step": 148
     },
     {
-      "epoch": 0.1622198505869797,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002998701612152596,
-      "loss": 0.9278,
+      "epoch": 0.1555783009211873,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00029988875354848766,
+      "loss": 0.9369,
       "step": 152
     },
     {
-      "epoch": 0.1664887940234792,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00029985780425743886,
-      "loss": 0.9414,
+      "epoch": 0.15967246673490276,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00029987777411576996,
+      "loss": 0.9607,
       "step": 156
     },
     {
-      "epoch": 0.17075773745997866,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.0002998448859826747,
-      "loss": 0.9558,
+      "epoch": 0.16376663254861823,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002998662782910581,
+      "loss": 0.9773,
       "step": 160
     },
     {
-      "epoch": 0.17502668089647813,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0002998314064393546,
-      "loss": 0.948,
+      "epoch": 0.16786079836233367,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002998542661139601,
+      "loss": 0.9218,
       "step": 164
     },
     {
-      "epoch": 0.17929562433297758,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002998173656779683,
-      "loss": 0.9017,
+      "epoch": 0.17195496417604914,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002998417376258628,
+      "loss": 0.9753,
       "step": 168
     },
     {
-      "epoch": 0.18356456776947705,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0002998027637511078,
-      "loss": 0.9413,
+      "epoch": 0.17604912998976457,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00029982869286993225,
+      "loss": 0.9777,
       "step": 172
     },
     {
-      "epoch": 0.18783351120597652,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002997876007134671,
-      "loss": 0.9348,
+      "epoch": 0.18014329580348004,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029981513189111314,
+      "loss": 0.9389,
       "step": 176
     },
     {
-      "epoch": 0.192102454642476,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0002997718766218419,
-      "loss": 0.9005,
+      "epoch": 0.1842374616171955,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00029980105473612865,
+      "loss": 0.9107,
       "step": 180
     },
     {
-      "epoch": 0.19637139807897544,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029975559153512917,
-      "loss": 0.9079,
+      "epoch": 0.18833162743091095,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002997864614534805,
+      "loss": 0.9503,
       "step": 184
     },
     {
-      "epoch": 0.2006403415154749,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002997387455143275,
-      "loss": 0.9624,
+      "epoch": 0.19242579324462641,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00029977135209344874,
+      "loss": 0.9258,
       "step": 188
     },
     {
-      "epoch": 0.20490928495197439,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0002997213386225363,
-      "loss": 0.9399,
+      "epoch": 0.19651995905834185,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0002997557267080915,
+      "loss": 0.9276,
       "step": 192
     },
     {
-      "epoch": 0.20917822838847386,
-      "grad_norm": 0.25,
-      "learning_rate": 0.000299703370924956,
-      "loss": 0.9132,
+      "epoch": 0.20061412487205732,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002997395853512448,
+      "loss": 0.9303,
       "step": 196
     },
     {
-      "epoch": 0.21344717182497333,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00029968484248888767,
-      "loss": 0.9243,
+      "epoch": 0.2047082906857728,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00029972292807852233,
+      "loss": 0.9378,
       "step": 200
     },
     {
-      "epoch": 0.21771611526147278,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002996657533837324,
-      "loss": 0.8978,
+      "epoch": 0.20880245649948823,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00029970575494731543,
+      "loss": 0.9271,
       "step": 204
     },
     {
-      "epoch": 0.22198505869797225,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002996461036809917,
-      "loss": 0.9,
+      "epoch": 0.2128966223132037,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00029968806601679283,
+      "loss": 0.9574,
       "step": 208
     },
     {
-      "epoch": 0.22625400213447172,
-      "grad_norm": 0.22265625,
-      "learning_rate": 0.0002996258934542667,
-      "loss": 0.9007,
+      "epoch": 0.21699078812691913,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00029966986134790025,
+      "loss": 0.9354,
       "step": 212
     },
     {
-      "epoch": 0.2305229455709712,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029960512277925816,
-      "loss": 0.9005,
+      "epoch": 0.2210849539406346,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002996511410033605,
+      "loss": 0.9185,
       "step": 216
     },
     {
-      "epoch": 0.23479188900747064,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0002995837917337659,
-      "loss": 0.8934,
+      "epoch": 0.22517911975435004,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.000299631905047673,
+      "loss": 0.9327,
       "step": 220
     },
     {
-      "epoch": 0.2390608324439701,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00029956190039768897,
-      "loss": 0.9122,
+      "epoch": 0.2292732855680655,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00029961215354711376,
+      "loss": 0.9141,
       "step": 224
     },
     {
-      "epoch": 0.24332977588046958,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.000299539448853025,
-      "loss": 0.8902,
+      "epoch": 0.23336745138178097,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000299591886569735,
+      "loss": 0.9369,
       "step": 228
     },
     {
-      "epoch": 0.24759871931696906,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.0002995164371838698,
-      "loss": 0.9001,
+      "epoch": 0.2374616171954964,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0002995711041853649,
+      "loss": 0.9163,
       "step": 232
     },
     {
-      "epoch": 0.2518676627534685,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029949286547641737,
-      "loss": 0.8653,
+      "epoch": 0.24155578300921188,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00029954980646560765,
+      "loss": 0.8721,
       "step": 236
     },
     {
-      "epoch": 0.256136606189968,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00029946873381895943,
-      "loss": 0.9168,
+      "epoch": 0.24564994882292732,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0002995279934838427,
+      "loss": 0.8767,
       "step": 240
     },
     {
-      "epoch": 0.26040554962646745,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00029944404230188503,
-      "loss": 0.9061,
+      "epoch": 0.24974411463664278,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00029950566531522515,
+      "loss": 0.8851,
       "step": 244
     },
     {
-      "epoch": 0.2646744930629669,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00029941879101768037,
-      "loss": 0.896,
+      "epoch": 0.2538382804503582,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00029948282203668477,
+      "loss": 0.9255,
       "step": 248
     },
     {
-      "epoch": 0.2689434364994664,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002993929800609282,
-      "loss": 0.9208,
+      "epoch": 0.2579324462640737,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029945946372692635,
+      "loss": 0.8865,
       "step": 252
     },
     {
-      "epoch": 0.27321237993596587,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00029936660952830773,
-      "loss": 0.8437,
+      "epoch": 0.26202661207778916,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00029943559046642903,
+      "loss": 0.8514,
       "step": 256
     },
     {
-      "epoch": 0.27748132337246534,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00029933967951859404,
-      "loss": 0.8265,
+      "epoch": 0.2661207778915046,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029941120233744625,
+      "loss": 0.9019,
       "step": 260
     },
     {
-      "epoch": 0.28175026680896476,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00029931219013265786,
-      "loss": 0.8543,
+      "epoch": 0.27021494370522003,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00029938629942400546,
+      "loss": 0.9104,
       "step": 264
     },
     {
-      "epoch": 0.28601921024546423,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00029928414147346535,
-      "loss": 0.8803,
+      "epoch": 0.2743091095189355,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029936088181190754,
+      "loss": 0.8915,
       "step": 268
     },
     {
-      "epoch": 0.2902881536819637,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0002992555336460772,
-      "loss": 0.8672,
+      "epoch": 0.27840327533265097,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.000299334949588727,
+      "loss": 0.889,
       "step": 272
     },
     {
-      "epoch": 0.2945570971184632,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002992263667576488,
-      "loss": 0.862,
+      "epoch": 0.28249744114636643,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00029930850284381116,
+      "loss": 0.9146,
       "step": 276
     },
     {
-      "epoch": 0.29882604055496265,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0002991966409174295,
-      "loss": 0.85,
+      "epoch": 0.2865916069600819,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029928154166828025,
+      "loss": 0.8627,
       "step": 280
     },
     {
-      "epoch": 0.3030949839914621,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002991663562367622,
-      "loss": 0.8714,
+      "epoch": 0.2906857727737973,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002992540661550268,
+      "loss": 0.8735,
       "step": 284
     },
     {
-      "epoch": 0.3073639274279616,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.00029913551282908325,
-      "loss": 0.8543,
+      "epoch": 0.2947799385875128,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029922607639871557,
+      "loss": 0.8819,
       "step": 288
     },
     {
-      "epoch": 0.31163287086446106,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00029910411080992164,
-      "loss": 0.8684,
+      "epoch": 0.29887410440122825,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000299197572495783,
+      "loss": 0.8979,
       "step": 292
     },
     {
-      "epoch": 0.31590181430096054,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00029907215029689873,
-      "loss": 0.8473,
+      "epoch": 0.3029682702149437,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029916855454443706,
+      "loss": 0.9047,
       "step": 296
     },
     {
-      "epoch": 0.32017075773745995,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00029903963140972796,
-      "loss": 0.8555,
+      "epoch": 0.3070624360286592,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002991390226446568,
+      "loss": 0.8861,
       "step": 300
     },
     {
-      "epoch": 0.3244397011739594,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0002990065542702141,
-      "loss": 0.9187,
+      "epoch": 0.3111566018423746,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.000299108976898192,
+      "loss": 0.8957,
       "step": 304
     },
     {
-      "epoch": 0.3287086446104589,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00029897291900225293,
-      "loss": 0.8884,
+      "epoch": 0.31525076765609006,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000299078417408563,
+      "loss": 0.8565,
       "step": 308
     },
     {
-      "epoch": 0.3329775880469584,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002989387257318309,
-      "loss": 0.8533,
+      "epoch": 0.3193449334698055,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029904734428105997,
+      "loss": 0.8656,
       "step": 312
     },
     {
-      "epoch": 0.33724653148345785,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.0002989039745870245,
-      "loss": 0.8292,
+      "epoch": 0.323439099283521,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000299015757622743,
+      "loss": 0.8339,
       "step": 316
     },
     {
-      "epoch": 0.3415154749199573,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002988686656979999,
-      "loss": 0.8913,
+      "epoch": 0.32753326509723646,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00029898365754244135,
+      "loss": 0.813,
       "step": 320
     },
     {
-      "epoch": 0.3457844183564568,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00029883279919701224,
-      "loss": 0.8792,
+      "epoch": 0.33162743091095187,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00029895104415075336,
+      "loss": 0.8969,
       "step": 324
     },
     {
-      "epoch": 0.35005336179295626,
+      "epoch": 0.33572159672466734,
       "grad_norm": 0.248046875,
-      "learning_rate": 0.0002987963752184054,
-      "loss": 0.853,
+      "learning_rate": 0.0002989179175600459,
+      "loss": 0.7858,
       "step": 328
     },
     {
-      "epoch": 0.35432230522945574,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002987593938986113,
-      "loss": 0.8576,
+      "epoch": 0.3398157625383828,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002988842778844539,
+      "loss": 0.8451,
       "step": 332
     },
     {
-      "epoch": 0.35859124866595515,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029872185537614963,
-      "loss": 0.8549,
+      "epoch": 0.34390992835209827,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029885012523988034,
+      "loss": 0.8592,
       "step": 336
     },
     {
-      "epoch": 0.3628601921024546,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.000298683759791627,
-      "loss": 0.8209,
+      "epoch": 0.34800409416581374,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0002988154597439954,
+      "loss": 0.8831,
       "step": 340
     },
     {
-      "epoch": 0.3671291355389541,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00029864510728773674,
-      "loss": 0.8402,
+      "epoch": 0.35209825997952915,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002987802815162363,
+      "loss": 0.8825,
       "step": 344
     },
     {
-      "epoch": 0.37139807897545357,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00029860589800925796,
-      "loss": 0.8256,
+      "epoch": 0.3561924257932446,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002987445906778068,
+      "loss": 0.817,
       "step": 348
     },
     {
-      "epoch": 0.37566702241195304,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00029856613210305565,
-      "loss": 0.8581,
+      "epoch": 0.3602865916069601,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00029870838735167684,
+      "loss": 0.8549,
       "step": 352
     },
     {
-      "epoch": 0.3799359658484525,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002985258097180794,
-      "loss": 0.8795,
+      "epoch": 0.36438075742067555,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0002986716716625822,
+      "loss": 0.8458,
       "step": 356
     },
     {
-      "epoch": 0.384204909284952,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00029848493100536325,
-      "loss": 0.8492,
+      "epoch": 0.368474923234391,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002986344437370238,
+      "loss": 0.8335,
       "step": 360
     },
     {
-      "epoch": 0.38847385272145146,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00029844349611802526,
-      "loss": 0.8277,
+      "epoch": 0.3725690890481064,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029859670370326757,
+      "loss": 0.8258,
       "step": 364
     },
     {
-      "epoch": 0.3927427961579509,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002984015052112665,
-      "loss": 0.7891,
+      "epoch": 0.3766632548618219,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0002985584516913437,
+      "loss": 0.8816,
       "step": 368
     },
     {
-      "epoch": 0.39701173959445035,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002983589584423708,
-      "loss": 0.8472,
+      "epoch": 0.38075742067553736,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0002985196878330466,
+      "loss": 0.8361,
       "step": 372
     },
     {
-      "epoch": 0.4012806830309498,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00029831585597070404,
-      "loss": 0.7938,
+      "epoch": 0.38485158648925283,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000298480412261934,
+      "loss": 0.8259,
       "step": 376
     },
     {
-      "epoch": 0.4055496264674493,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002982721979577136,
-      "loss": 0.8323,
+      "epoch": 0.3889457523029683,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002984406251133268,
+      "loss": 0.8277,
       "step": 380
     },
     {
-      "epoch": 0.40981856990394877,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00029822798456692774,
-      "loss": 0.8585,
+      "epoch": 0.3930399181166837,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002984003265243084,
+      "loss": 0.861,
       "step": 384
     },
     {
-      "epoch": 0.41408751334044824,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0002981832159639548,
-      "loss": 0.7854,
+      "epoch": 0.3971340839303992,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00029835951663372446,
+      "loss": 0.8145,
       "step": 388
     },
     {
-      "epoch": 0.4183564567769477,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.0002981378923164832,
-      "loss": 0.8226,
+      "epoch": 0.40122824974411464,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002983181955821822,
+      "loss": 0.8501,
       "step": 392
     },
     {
-      "epoch": 0.4226254002134472,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00029809201379427986,
-      "loss": 0.7998,
+      "epoch": 0.4053224155578301,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029827636351205004,
+      "loss": 0.8362,
       "step": 396
     },
     {
-      "epoch": 0.42689434364994666,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00029804558056919036,
-      "loss": 0.8507,
+      "epoch": 0.4094165813715456,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00029823402056745706,
+      "loss": 0.7834,
       "step": 400
     },
     {
-      "epoch": 0.4311632870864461,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002979985928151379,
-      "loss": 0.8109,
+      "epoch": 0.413510747185261,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002981911668942925,
+      "loss": 0.8486,
       "step": 404
     },
     {
-      "epoch": 0.43543223052294555,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002979510507081229,
-      "loss": 0.797,
+      "epoch": 0.41760491299897645,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00029814780264020535,
+      "loss": 0.8006,
       "step": 408
     },
     {
-      "epoch": 0.439701173959445,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029790295442622204,
-      "loss": 0.7914,
+      "epoch": 0.4216990788126919,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029810392795460365,
+      "loss": 0.8553,
       "step": 412
     },
     {
-      "epoch": 0.4439701173959445,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029785430414958785,
-      "loss": 0.8193,
+      "epoch": 0.4257932446264074,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00029805954298865413,
+      "loss": 0.8271,
       "step": 416
     },
     {
-      "epoch": 0.44823906083244397,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029780510006044794,
-      "loss": 0.8435,
+      "epoch": 0.42988741044012285,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002980146478952818,
+      "loss": 0.8035,
       "step": 420
     },
     {
-      "epoch": 0.45250800426894344,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00029775534234310414,
-      "loss": 0.7973,
+      "epoch": 0.43398157625383826,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000297969242829169,
+      "loss": 0.7901,
       "step": 424
     },
     {
-      "epoch": 0.4567769477054429,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002977050311839322,
-      "loss": 0.8333,
+      "epoch": 0.43807574206755373,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002979233279467554,
+      "loss": 0.7974,
       "step": 428
     },
     {
-      "epoch": 0.4610458911419424,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00029765416677138095,
-      "loss": 0.7915,
+      "epoch": 0.4421699078812692,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000297876903406237,
+      "loss": 0.8294,
       "step": 432
     },
     {
-      "epoch": 0.46531483457844186,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002976027492959712,
-      "loss": 0.8047,
+      "epoch": 0.44626407369498466,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.000297829969367566,
+      "loss": 0.7868,
       "step": 436
     },
     {
-      "epoch": 0.4695837780149413,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002975507789502956,
-      "loss": 0.8397,
+      "epoch": 0.4503582395087001,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002977825259924497,
+      "loss": 0.842,
       "step": 440
     },
     {
-      "epoch": 0.47385272145144075,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00029749825592901755,
-      "loss": 0.8664,
+      "epoch": 0.45445240532241554,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029773457344435067,
+      "loss": 0.8359,
       "step": 444
     },
     {
-      "epoch": 0.4781216648879402,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002974451804288706,
-      "loss": 0.8272,
+      "epoch": 0.458546571136131,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002976861118884856,
+      "loss": 0.8127,
       "step": 448
     },
     {
-      "epoch": 0.4823906083244397,
+      "epoch": 0.4626407369498465,
       "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029739155264865783,
-      "loss": 0.8314,
+      "learning_rate": 0.00029763714149182483,
+      "loss": 0.8251,
       "step": 452
     },
     {
-      "epoch": 0.48665955176093917,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002973373727892508,
-      "loss": 0.8114,
+      "epoch": 0.46673490276356194,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002975876624230921,
+      "loss": 0.8145,
       "step": 456
     },
     {
-      "epoch": 0.49092849519743864,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.000297282641053589,
-      "loss": 0.8129,
+      "epoch": 0.47082906857727735,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002975376748527636,
+      "loss": 0.814,
       "step": 460
     },
     {
-      "epoch": 0.4951974386339381,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002972273576466792,
-      "loss": 0.8096,
+      "epoch": 0.4749232343909928,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029748717895306746,
+      "loss": 0.8419,
       "step": 464
     },
     {
-      "epoch": 0.4994663820704376,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029717152277559445,
-      "loss": 0.7984,
+      "epoch": 0.4790174002047083,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0002974361748979834,
+      "loss": 0.814,
       "step": 468
     },
     {
-      "epoch": 0.503735325506937,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00029711513664947334,
-      "loss": 0.816,
+      "epoch": 0.48311156601842375,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00029738466286324176,
+      "loss": 0.8097,
       "step": 472
     },
     {
-      "epoch": 0.5080042689434365,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002970581994795194,
-      "loss": 0.8384,
+      "epoch": 0.4872057318321392,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029733264302632325,
+      "loss": 0.7909,
       "step": 476
     },
     {
-      "epoch": 0.512273212379936,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002970007114790001,
-      "loss": 0.7474,
+      "epoch": 0.49129989764585463,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002972801155664581,
+      "loss": 0.8078,
       "step": 480
     },
     {
-      "epoch": 0.5165421558164355,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029694267286324604,
-      "loss": 0.8275,
+      "epoch": 0.4953940634595701,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00029722708066462543,
+      "loss": 0.8108,
       "step": 484
     },
     {
-      "epoch": 0.5208110992529349,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00029688408384965056,
-      "loss": 0.8404,
+      "epoch": 0.49948822927328557,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00029717353850355286,
+      "loss": 0.852,
       "step": 488
     },
     {
-      "epoch": 0.5250800426894343,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0002968249446576683,
-      "loss": 0.8155,
+      "epoch": 0.503582395087001,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002971194892677157,
+      "loss": 0.7383,
       "step": 492
     },
     {
-      "epoch": 0.5293489861259338,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002967652555088148,
-      "loss": 0.7902,
+      "epoch": 0.5076765609007164,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002970649331433362,
+      "loss": 0.7806,
       "step": 496
     },
     {
-      "epoch": 0.5336179295624333,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00029670501662666546,
-      "loss": 0.8009,
+      "epoch": 0.5117707267144319,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002970098703183832,
+      "loss": 0.7578,
       "step": 500
     },
     {
-      "epoch": 0.5378868729989328,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.0002966442282368551,
-      "loss": 0.8057,
+      "epoch": 0.5158648925281474,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002969543009825713,
+      "loss": 0.8208,
       "step": 504
     },
     {
-      "epoch": 0.5421558164354322,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002965828905670763,
-      "loss": 0.8033,
+      "epoch": 0.5199590583418628,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002968982253273603,
+      "loss": 0.8233,
       "step": 508
     },
     {
-      "epoch": 0.5464247598719317,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029652100384707956,
-      "loss": 0.7977,
+      "epoch": 0.5240532241555783,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0002968416435459544,
+      "loss": 0.8092,
       "step": 512
     },
     {
-      "epoch": 0.5506937033084311,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00029645856830867165,
-      "loss": 0.7863,
+      "epoch": 0.5281473899692938,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029678455583330156,
+      "loss": 0.8246,
       "step": 516
     },
     {
-      "epoch": 0.5549626467449307,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029639558418571507,
-      "loss": 0.7807,
+      "epoch": 0.5322415557830092,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002967269623860931,
+      "loss": 0.7451,
       "step": 520
     },
     {
-      "epoch": 0.5592315901814301,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00029633205171412724,
-      "loss": 0.7969,
+      "epoch": 0.5363357215967247,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00029666886340276263,
+      "loss": 0.759,
       "step": 524
     },
     {
-      "epoch": 0.5635005336179295,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029626797113187935,
-      "loss": 0.8441,
+      "epoch": 0.5404298874104401,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00029661025908348556,
+      "loss": 0.8068,
       "step": 528
     },
     {
-      "epoch": 0.567769477054429,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00029620334267899584,
-      "loss": 0.7523,
+      "epoch": 0.5445240532241555,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002965511496301784,
+      "loss": 0.7771,
       "step": 532
     },
     {
-      "epoch": 0.5720384204909285,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.000296138166597553,
-      "loss": 0.7924,
+      "epoch": 0.548618219037871,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002964915352464982,
+      "loss": 0.8213,
       "step": 536
     },
     {
-      "epoch": 0.576307363927428,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029607244313167863,
-      "loss": 0.7733,
+      "epoch": 0.5527123848515865,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002964314161378415,
+      "loss": 0.8283,
       "step": 540
     },
     {
-      "epoch": 0.5805763073639274,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002960061725275507,
+      "epoch": 0.5568065506653019,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000296370792511344,
       "loss": 0.7587,
       "step": 544
     },
     {
-      "epoch": 0.5848452508004269,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029593935503339656,
-      "loss": 0.807,
+      "epoch": 0.5609007164790174,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0002963096645758795,
+      "loss": 0.7708,
       "step": 548
     },
     {
-      "epoch": 0.5891141942369263,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00029587199089949214,
-      "loss": 0.8088,
+      "epoch": 0.5649948822927329,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029624803254205953,
+      "loss": 0.8349,
       "step": 552
     },
     {
-      "epoch": 0.5933831376734259,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.0002958040803781609,
-      "loss": 0.74,
+      "epoch": 0.5690890481064483,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002961858966222324,
+      "loss": 0.7805,
       "step": 556
     },
     {
-      "epoch": 0.5976520811099253,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002957356237237726,
-      "loss": 0.7626,
+      "epoch": 0.5731832139201638,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002961232570304824,
+      "loss": 0.757,
       "step": 560
     },
     {
-      "epoch": 0.6019210245464247,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00029566662119274306,
-      "loss": 0.7676,
+      "epoch": 0.5772773797338793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002960601139826294,
+      "loss": 0.7807,
       "step": 564
     },
     {
-      "epoch": 0.6061899679829242,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029559707304353247,
-      "loss": 0.8182,
+      "epoch": 0.5813715455475946,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00029599646769622775,
+      "loss": 0.7528,
       "step": 568
     },
     {
-      "epoch": 0.6104589114194237,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002955269795366448,
-      "loss": 0.762,
+      "epoch": 0.5854657113613101,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00029593231839056554,
+      "loss": 0.817,
       "step": 572
     },
     {
-      "epoch": 0.6147278548559232,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00029545634093462676,
-      "loss": 0.8269,
+      "epoch": 0.5895598771750256,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002958676662866643,
+      "loss": 0.8212,
       "step": 576
     },
     {
-      "epoch": 0.6189967982924226,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002953851575020669,
-      "loss": 0.7654,
+      "epoch": 0.593654042988741,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00029580251160727766,
+      "loss": 0.7561,
       "step": 580
     },
     {
-      "epoch": 0.6232657417289221,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002953134295055943,
-      "loss": 0.7842,
+      "epoch": 0.5977482088024565,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00029573685457689086,
+      "loss": 0.8184,
       "step": 584
     },
     {
-      "epoch": 0.6275346851654215,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002952411572138779,
-      "loss": 0.8006,
+      "epoch": 0.601842374616172,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00029567069542172004,
+      "loss": 0.7591,
       "step": 588
     },
     {
-      "epoch": 0.6318036286019211,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002951683408976255,
-      "loss": 0.8125,
+      "epoch": 0.6059365404298874,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002956040343697114,
+      "loss": 0.7598,
       "step": 592
     },
     {
-      "epoch": 0.6360725720384205,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00029509498082958246,
-      "loss": 0.7812,
+      "epoch": 0.6100307062436029,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002955368716505401,
+      "loss": 0.7583,
       "step": 596
     },
     {
-      "epoch": 0.6403415154749199,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00029502107728453095,
-      "loss": 0.7902,
+      "epoch": 0.6141248720573184,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002954692074956102,
+      "loss": 0.7787,
       "step": 600
     },
     {
-      "epoch": 0.6446104589114194,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002949466305392889,
-      "loss": 0.8181,
+      "epoch": 0.6182190378710338,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00029540104213805307,
+      "loss": 0.8317,
       "step": 604
     },
     {
-      "epoch": 0.6488794023479189,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002948716408727086,
-      "loss": 0.7977,
+      "epoch": 0.6223132036847492,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00029533237581272706,
+      "loss": 0.7833,
       "step": 608
     },
     {
-      "epoch": 0.6531483457844184,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00029479610856567625,
-      "loss": 0.8335,
+      "epoch": 0.6264073694984647,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00029526320875621656,
+      "loss": 0.7263,
       "step": 612
     },
     {
-      "epoch": 0.6574172892209178,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002947200339011105,
-      "loss": 0.7471,
+      "epoch": 0.6305015353121801,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029519354120683116,
+      "loss": 0.8114,
       "step": 616
     },
     {
-      "epoch": 0.6616862326574173,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002946434171639613,
-      "loss": 0.7966,
+      "epoch": 0.6345957011258956,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002951233734046049,
+      "loss": 0.785,
       "step": 620
     },
     {
-      "epoch": 0.6659551760939167,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002945662586412093,
-      "loss": 0.7977,
+      "epoch": 0.638689866939611,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002950527055912955,
+      "loss": 0.8104,
       "step": 624
     },
     {
-      "epoch": 0.6702241195304163,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029448855862186445,
-      "loss": 0.7886,
+      "epoch": 0.6427840327533265,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029498153801038303,
+      "loss": 0.7885,
       "step": 628
     },
     {
-      "epoch": 0.6744930629669157,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029441031739696477,
-      "loss": 0.8127,
+      "epoch": 0.646878198567042,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00029490987090707,
+      "loss": 0.7883,
       "step": 632
     },
     {
-      "epoch": 0.6787620064034151,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002943315352595756,
-      "loss": 0.7775,
+      "epoch": 0.6509723643807575,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002948377045282796,
+      "loss": 0.8154,
       "step": 636
     },
     {
-      "epoch": 0.6830309498399146,
+      "epoch": 0.6550665301944729,
       "grad_norm": 0.25,
-      "learning_rate": 0.0002942522125047884,
-      "loss": 0.7769,
+      "learning_rate": 0.0002947650391226555,
+      "loss": 0.7979,
       "step": 640
     },
     {
-      "epoch": 0.687299893276414,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00029417234942971947,
-      "loss": 0.7934,
+      "epoch": 0.6591606960081884,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00029469187494056046,
+      "loss": 0.7896,
       "step": 644
     },
     {
-      "epoch": 0.6915688367129136,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.0002940919463335091,
-      "loss": 0.7217,
+      "epoch": 0.6632548618219037,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002946182122340759,
+      "loss": 0.7295,
       "step": 648
     },
     {
-      "epoch": 0.695837780149413,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002940110035173201,
-      "loss": 0.7463,
+      "epoch": 0.6673490276356192,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0002945440512570009,
+      "loss": 0.7889,
       "step": 652
     },
     {
-      "epoch": 0.7001067235859125,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00029392952128433727,
-      "loss": 0.7861,
+      "epoch": 0.6714431934493347,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00029446939226485125,
+      "loss": 0.7931,
       "step": 656
     },
     {
-      "epoch": 0.7043756670224119,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002938474999397655,
-      "loss": 0.7669,
+      "epoch": 0.6755373592630501,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029439423551485844,
+      "loss": 0.7944,
       "step": 660
     },
     {
-      "epoch": 0.7086446104589115,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00029376493979082916,
-      "loss": 0.7686,
+      "epoch": 0.6796315250767656,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002943185812659693,
+      "loss": 0.8238,
       "step": 664
     },
     {
-      "epoch": 0.7129135538954109,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.0002936818411467709,
-      "loss": 0.768,
+      "epoch": 0.6837256908904811,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029424242977884436,
+      "loss": 0.827,
       "step": 668
     },
     {
-      "epoch": 0.7171824973319103,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00029359820431885025,
-      "loss": 0.7736,
+      "epoch": 0.6878198567041965,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00029416578131585765,
+      "loss": 0.7974,
       "step": 672
     },
     {
-      "epoch": 0.7214514407684098,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.0002935140296203426,
-      "loss": 0.8166,
+      "epoch": 0.691914022517912,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00029408863614109533,
+      "loss": 0.8207,
       "step": 676
     },
     {
-      "epoch": 0.7257203842049093,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00029342931736653816,
-      "loss": 0.7935,
+      "epoch": 0.6960081883316275,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000294010994520355,
+      "loss": 0.7433,
       "step": 680
     },
     {
-      "epoch": 0.7299893276414088,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002933440678747404,
-      "loss": 0.7525,
+      "epoch": 0.7001023541453428,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00029393285672114477,
+      "loss": 0.8231,
       "step": 684
     },
     {
-      "epoch": 0.7342582710779082,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00029325828146426543,
-      "loss": 0.7382,
+      "epoch": 0.7041965199590583,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002938542230126821,
+      "loss": 0.8487,
       "step": 688
     },
     {
-      "epoch": 0.7385272145144077,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002931719584564402,
-      "loss": 0.7566,
+      "epoch": 0.7082906857727738,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002937750936658933,
+      "loss": 0.7503,
       "step": 692
     },
     {
-      "epoch": 0.7427961579509071,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002930850991746017,
-      "loss": 0.7944,
+      "epoch": 0.7123848515864892,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00029369546895341225,
+      "loss": 0.8059,
       "step": 696
     },
     {
-      "epoch": 0.7470651013874067,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029299770394409553,
-      "loss": 0.8058,
+      "epoch": 0.7164790174002047,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002936153491495796,
+      "loss": 0.7972,
       "step": 700
     },
     {
-      "epoch": 0.7513340448239061,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002929097730922749,
-      "loss": 0.7256,
+      "epoch": 0.7205731832139202,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002935347345304417,
+      "loss": 0.7924,
       "step": 704
     },
     {
-      "epoch": 0.7556029882604055,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.0002928213069484992,
-      "loss": 0.8137,
+      "epoch": 0.7246673490276356,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029345362537374996,
+      "loss": 0.7478,
       "step": 708
     },
     {
-      "epoch": 0.759871931696905,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002927323058441328,
-      "loss": 0.8153,
+      "epoch": 0.7287615148413511,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002933720219589595,
+      "loss": 0.7237,
       "step": 712
     },
     {
-      "epoch": 0.7641408751334045,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.000292642770112544,
-      "loss": 0.8005,
+      "epoch": 0.7328556806550666,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029328992456722835,
+      "loss": 0.7898,
       "step": 716
     },
     {
-      "epoch": 0.768409818569904,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002925527000891035,
-      "loss": 0.7505,
+      "epoch": 0.736949846468782,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00029320733348141666,
+      "loss": 0.7344,
       "step": 720
     },
     {
-      "epoch": 0.7726787620064034,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00029246209611118336,
-      "loss": 0.7822,
+      "epoch": 0.7410440122824974,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00029312424898608546,
+      "loss": 0.8047,
       "step": 724
     },
     {
-      "epoch": 0.7769477054429029,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00029237095851815555,
-      "loss": 0.8299,
+      "epoch": 0.7451381780962129,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002930406713674957,
+      "loss": 0.7926,
       "step": 728
     },
     {
-      "epoch": 0.7812166488794023,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002922792876513909,
-      "loss": 0.7657,
+      "epoch": 0.7492323439099283,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029295660091360764,
+      "loss": 0.7631,
       "step": 732
     },
     {
-      "epoch": 0.7854855923159018,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002921870838542576,
-      "loss": 0.7274,
+      "epoch": 0.7533265097236438,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00029287203791407917,
+      "loss": 0.7931,
       "step": 736
     },
     {
-      "epoch": 0.7897545357524013,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002920943474721201,
-      "loss": 0.735,
+      "epoch": 0.7574206755373593,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00029278698266026545,
+      "loss": 0.8366,
       "step": 740
     },
     {
-      "epoch": 0.7940234791889007,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002920010788523377,
-      "loss": 0.7685,
+      "epoch": 0.7615148413510747,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002927014354452177,
+      "loss": 0.7789,
       "step": 744
     },
     {
-      "epoch": 0.7982924226254002,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.00029190727834426327,
-      "loss": 0.7693,
+      "epoch": 0.7656090071647902,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000292615396563682,
+      "loss": 0.7381,
       "step": 748
     },
     {
-      "epoch": 0.8025613660618997,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.000291812946299242,
-      "loss": 0.7253,
+      "epoch": 0.7697031729785057,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00029252886631209846,
+      "loss": 0.7583,
       "step": 752
     },
     {
-      "epoch": 0.8068303094983992,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00029171808307061,
-      "loss": 0.7483,
+      "epoch": 0.7737973387922211,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002924418449886003,
+      "loss": 0.7299,
       "step": 756
     },
     {
-      "epoch": 0.8110992529348986,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00029162268901369306,
-      "loss": 0.7539,
+      "epoch": 0.7778915046059366,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00029235433289301257,
+      "loss": 0.753,
       "step": 760
     },
     {
-      "epoch": 0.8153681963713981,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002915267644858052,
-      "loss": 0.7549,
+      "epoch": 0.781985670419652,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002922663303268512,
+      "loss": 0.7687,
       "step": 764
     },
     {
-      "epoch": 0.8196371398078975,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.0002914303098462474,
-      "loss": 0.8225,
+      "epoch": 0.7860798362333674,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00029217783759332214,
+      "loss": 0.7792,
       "step": 768
     },
     {
-      "epoch": 0.823906083244397,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00029133332545630645,
-      "loss": 0.7879,
+      "epoch": 0.7901740020470829,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00029208885499732004,
+      "loss": 0.7712,
       "step": 772
     },
     {
-      "epoch": 0.8281750266808965,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0002912358116792531,
-      "loss": 0.7922,
+      "epoch": 0.7942681678607983,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002919993828454274,
+      "loss": 0.7814,
       "step": 776
     },
     {
-      "epoch": 0.8324439701173959,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00029113776888034125,
-      "loss": 0.7629,
+      "epoch": 0.7983623336745138,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002919094214459134,
+      "loss": 0.7868,
       "step": 780
     },
     {
-      "epoch": 0.8367129135538954,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002910391974268063,
-      "loss": 0.7856,
+      "epoch": 0.8024564994882293,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000291818971108733,
+      "loss": 0.759,
       "step": 784
     },
     {
-      "epoch": 0.8409818569903948,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00029094009768786366,
-      "loss": 0.7591,
+      "epoch": 0.8065506653019447,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002917280321455255,
+      "loss": 0.722,
       "step": 788
     },
     {
-      "epoch": 0.8452508004268944,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00029084047003470766,
-      "loss": 0.7624,
+      "epoch": 0.8106448311156602,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00029163660486961404,
+      "loss": 0.7269,
       "step": 792
     },
     {
-      "epoch": 0.8495197438633938,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029074031484051005,
-      "loss": 0.7285,
+      "epoch": 0.8147389969293757,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002915446895960041,
+      "loss": 0.7571,
       "step": 796
     },
     {
-      "epoch": 0.8537886872998933,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00029063963248041844,
-      "loss": 0.7958,
+      "epoch": 0.8188331627430911,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002914522866413823,
+      "loss": 0.7498,
       "step": 800
     },
     {
-      "epoch": 0.8580576307363927,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00029053842333155516,
-      "loss": 0.8189,
+      "epoch": 0.8229273285568065,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00029135939632411576,
+      "loss": 0.7437,
       "step": 804
     },
     {
-      "epoch": 0.8623265741728922,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0002904366877730156,
-      "loss": 0.7676,
+      "epoch": 0.827021494370522,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00029126601896425084,
+      "loss": 0.7524,
       "step": 808
     },
     {
-      "epoch": 0.8665955176093917,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.000290334426185867,
-      "loss": 0.7937,
+      "epoch": 0.8311156601842374,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002911721548835116,
+      "loss": 0.783,
       "step": 812
     },
     {
-      "epoch": 0.8708644610458911,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00029023163895314685,
-      "loss": 0.7136,
+      "epoch": 0.8352098259979529,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002910778044052995,
+      "loss": 0.7625,
       "step": 816
     },
     {
-      "epoch": 0.8751334044823906,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00029012832645986154,
-      "loss": 0.7553,
+      "epoch": 0.8393039918116684,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00029098296785469153,
+      "loss": 0.8289,
       "step": 820
     },
     {
-      "epoch": 0.87940234791889,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00029002448909298497,
-      "loss": 0.7844,
+      "epoch": 0.8433981576253838,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00029088764555843953,
+      "loss": 0.7998,
       "step": 824
     },
     {
-      "epoch": 0.8836712913553896,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00028992012724145694,
-      "loss": 0.7752,
+      "epoch": 0.8474923234390993,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002907918378449689,
+      "loss": 0.7699,
       "step": 828
     },
     {
-      "epoch": 0.887940234791889,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.0002898152412961819,
-      "loss": 0.769,
+      "epoch": 0.8515864892528148,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00029069554504437757,
+      "loss": 0.7366,
       "step": 832
     },
     {
-      "epoch": 0.8922091782283885,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002897098316500273,
-      "loss": 0.7243,
+      "epoch": 0.8556806550665302,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002905987674884347,
+      "loss": 0.7756,
       "step": 836
     },
     {
-      "epoch": 0.8964781216648879,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00028960389869782225,
-      "loss": 0.7912,
+      "epoch": 0.8597748208802457,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00029050150551057977,
+      "loss": 0.8355,
       "step": 840
     },
     {
-      "epoch": 0.9007470651013874,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.000289497442836356,
-      "loss": 0.7629,
+      "epoch": 0.8638689866939611,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00029040375944592114,
+      "loss": 0.7178,
       "step": 844
     },
     {
-      "epoch": 0.9050160085378869,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00028939046446437635,
-      "loss": 0.8218,
+      "epoch": 0.8679631525076765,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00029030552963123517,
+      "loss": 0.7798,
       "step": 848
     },
     {
-      "epoch": 0.9092849519743863,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002892829639825884,
-      "loss": 0.7341,
+      "epoch": 0.872057318321392,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002902068164049649,
+      "loss": 0.7485,
       "step": 852
     },
     {
-      "epoch": 0.9135538954108858,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00028917494179365273,
-      "loss": 0.7561,
+      "epoch": 0.8761514841351075,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0002901076201072189,
+      "loss": 0.7272,
       "step": 856
     },
     {
-      "epoch": 0.9178228388473852,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00028906639830218414,
-      "loss": 0.7344,
+      "epoch": 0.8802456499488229,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00029000794107977016,
+      "loss": 0.7575,
       "step": 860
     },
     {
-      "epoch": 0.9220917822838848,
+      "epoch": 0.8843398157625384,
       "grad_norm": 0.26171875,
-      "learning_rate": 0.00028895733391475,
-      "loss": 0.8057,
+      "learning_rate": 0.0002899077796660549,
+      "loss": 0.7391,
       "step": 864
     },
     {
-      "epoch": 0.9263607257203842,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.000288847749039869,
-      "loss": 0.6914,
+      "epoch": 0.8884339815762539,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0002898071362111713,
+      "loss": 0.7357,
       "step": 868
     },
     {
-      "epoch": 0.9306296691568837,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00028873764408800907,
-      "loss": 0.7737,
+      "epoch": 0.8925281473899693,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00028970601106187844,
+      "loss": 0.7975,
       "step": 872
     },
     {
-      "epoch": 0.9348986125933831,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002886270194715864,
-      "loss": 0.7899,
+      "epoch": 0.8966223132036848,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000289604404566595,
+      "loss": 0.7639,
       "step": 876
     },
     {
-      "epoch": 0.9391675560298826,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.0002885158756049636,
-      "loss": 0.7628,
+      "epoch": 0.9007164790174002,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00028950231707539813,
+      "loss": 0.7482,
       "step": 880
     },
     {
-      "epoch": 0.9434364994663821,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00028840421290444817,
-      "loss": 0.7288,
+      "epoch": 0.9048106448311156,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0002893997489400221,
+      "loss": 0.7782,
       "step": 884
     },
     {
-      "epoch": 0.9477054429028815,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002882920317882911,
-      "loss": 0.7506,
+      "epoch": 0.9089048106448311,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028929670051385745,
+      "loss": 0.7509,
       "step": 888
     },
     {
-      "epoch": 0.951974386339381,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002881793326766852,
-      "loss": 0.743,
+      "epoch": 0.9129989764585466,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002891931721519492,
+      "loss": 0.7274,
       "step": 892
     },
     {
-      "epoch": 0.9562433297758804,
-      "grad_norm": 0.232421875,
-      "learning_rate": 0.00028806611599176325,
-      "loss": 0.7591,
+      "epoch": 0.917093142272262,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002890891642109962,
+      "loss": 0.7512,
       "step": 896
     },
     {
-      "epoch": 0.96051227321238,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00028795238215759703,
-      "loss": 0.7597,
+      "epoch": 0.9211873080859775,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002889846770493496,
+      "loss": 0.7375,
       "step": 900
     },
     {
-      "epoch": 0.9647812166488794,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00028783813160019526,
-      "loss": 0.7295,
+      "epoch": 0.925281473899693,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028887971102701154,
+      "loss": 0.6872,
       "step": 904
     },
     {
-      "epoch": 0.9690501600853789,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00028772336474750193,
-      "loss": 0.7862,
+      "epoch": 0.9293756397134084,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002887742665056342,
+      "loss": 0.7695,
       "step": 908
     },
     {
-      "epoch": 0.9733191035218783,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.0002876080820293953,
-      "loss": 0.7436,
+      "epoch": 0.9334698055271239,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002886683438485183,
+      "loss": 0.7726,
       "step": 912
     },
     {
-      "epoch": 0.9775880469583778,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00028749228387768555,
-      "loss": 0.7865,
+      "epoch": 0.9375639713408394,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000288561943420612,
+      "loss": 0.7426,
       "step": 916
     },
     {
-      "epoch": 0.9818569903948773,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00028737597072611357,
-      "loss": 0.8468,
+      "epoch": 0.9416581371545547,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002884550655885095,
+      "loss": 0.7311,
       "step": 920
     },
     {
-      "epoch": 0.9861259338313767,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002872591430103495,
-      "loss": 0.7507,
+      "epoch": 0.9457523029682702,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00028834771072044994,
+      "loss": 0.7497,
       "step": 924
     },
     {
-      "epoch": 0.9903948772678762,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00028714180116799044,
-      "loss": 0.7358,
+      "epoch": 0.9498464687819856,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00028823987918631596,
+      "loss": 0.7134,
       "step": 928
     },
     {
-      "epoch": 0.9946638207043756,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002870239456385596,
-      "loss": 0.808,
+      "epoch": 0.9539406345957011,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002881315713576326,
+      "loss": 0.7818,
       "step": 932
     },
     {
-      "epoch": 0.9989327641408752,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00028690557686350424,
-      "loss": 0.7362,
+      "epoch": 0.9580348004094166,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0002880227876075659,
+      "loss": 0.7358,
       "step": 936
     },
     {
-      "epoch": 1.0032017075773747,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028678669528619383,
-      "loss": 0.6681,
+      "epoch": 0.962128966223132,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00028791352831092164,
+      "loss": 0.7309,
       "step": 940
     },
     {
-      "epoch": 1.007470651013874,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00028666730135191884,
-      "loss": 0.7536,
+      "epoch": 0.9662231320368475,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002878037938441441,
+      "loss": 0.7227,
       "step": 944
     },
     {
-      "epoch": 1.0117395944503735,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00028654739550788887,
-      "loss": 0.7087,
+      "epoch": 0.970317297850563,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002876935845853146,
+      "loss": 0.7828,
       "step": 948
     },
     {
-      "epoch": 1.016008537886873,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002864269782032308,
-      "loss": 0.7074,
+      "epoch": 0.9744114636642784,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002875829009141505,
+      "loss": 0.7156,
       "step": 952
     },
     {
-      "epoch": 1.0202774813233724,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002863060498889874,
-      "loss": 0.7294,
+      "epoch": 0.9785056294779939,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00028747174321200363,
+      "loss": 0.7602,
       "step": 956
     },
     {
-      "epoch": 1.024546424759872,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002861846110181155,
-      "loss": 0.6662,
+      "epoch": 0.9825997952917093,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000287360111861859,
+      "loss": 0.7051,
       "step": 960
     },
     {
-      "epoch": 1.0288153681963714,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002860626620454842,
-      "loss": 0.6669,
+      "epoch": 0.9866939611054247,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028724800724833354,
+      "loss": 0.6861,
       "step": 964
     },
     {
-      "epoch": 1.033084311632871,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.0002859402034278735,
-      "loss": 0.6592,
+      "epoch": 0.9907881269191402,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00028713542975767486,
+      "loss": 0.7947,
       "step": 968
     },
     {
-      "epoch": 1.0373532550693703,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.000285817235623972,
-      "loss": 0.6988,
+      "epoch": 0.9948822927328557,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002870223797777598,
+      "loss": 0.7499,
       "step": 972
     },
     {
-      "epoch": 1.0416221985058698,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00028569375909437585,
-      "loss": 0.7218,
+      "epoch": 0.9989764585465711,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002869088576980931,
+      "loss": 0.7674,
       "step": 976
     },
     {
-      "epoch": 1.0458911419423693,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002855697743015866,
-      "loss": 0.6944,
+      "epoch": 1.0030706243602865,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002867948639098061,
+      "loss": 0.6806,
       "step": 980
     },
     {
-      "epoch": 1.0501600853788686,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0002854452817100096,
-      "loss": 0.6644,
+      "epoch": 1.007164790174002,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00028668039880565526,
+      "loss": 0.7057,
       "step": 984
     },
     {
-      "epoch": 1.0544290288153682,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0002853202817859522,
-      "loss": 0.6779,
+      "epoch": 1.0112589559877174,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0002865654627800212,
+      "loss": 0.6704,
       "step": 988
     },
     {
-      "epoch": 1.0586979722518677,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028519477499762213,
-      "loss": 0.6798,
+      "epoch": 1.015353121801433,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00028645005622890673,
+      "loss": 0.708,
       "step": 992
     },
     {
-      "epoch": 1.0629669156883672,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002850687618151256,
-      "loss": 0.6308,
+      "epoch": 1.0194472876151484,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002863341795499361,
+      "loss": 0.6166,
       "step": 996
     },
     {
-      "epoch": 1.0672358591248665,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028494224271046565,
-      "loss": 0.6534,
+      "epoch": 1.0235414534288638,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00028621783314235314,
+      "loss": 0.6911,
       "step": 1000
     },
     {
-      "epoch": 1.071504802561366,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002848152181575402,
-      "loss": 0.7001,
+      "epoch": 1.0276356192425793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002861010174070202,
+      "loss": 0.6379,
       "step": 1004
     },
     {
-      "epoch": 1.0757737459978656,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002846876886321406,
-      "loss": 0.6383,
+      "epoch": 1.0317297850562948,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002859837327464167,
+      "loss": 0.691,
       "step": 1008
     },
     {
-      "epoch": 1.0800426894343649,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002845596546119496,
-      "loss": 0.6803,
+      "epoch": 1.0358239508700102,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002858659795646375,
+      "loss": 0.6792,
       "step": 1012
     },
     {
-      "epoch": 1.0843116328708644,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00028443111657653947,
-      "loss": 0.7376,
+      "epoch": 1.0399181166837257,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000285747758267392,
+      "loss": 0.6846,
       "step": 1016
     },
     {
-      "epoch": 1.088580576307364,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00028430207500737054,
-      "loss": 0.6841,
+      "epoch": 1.0440122824974412,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00028562906926200225,
+      "loss": 0.7281,
       "step": 1020
     },
     {
-      "epoch": 1.0928495197438635,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0002841725303877892,
-      "loss": 0.6774,
+      "epoch": 1.0481064483111566,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002855099129574018,
+      "loss": 0.6851,
       "step": 1024
     },
     {
-      "epoch": 1.0971184631803628,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00028404248320302585,
-      "loss": 0.6982,
+      "epoch": 1.052200614124872,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028539028976413435,
+      "loss": 0.7062,
       "step": 1028
     },
     {
-      "epoch": 1.1013874066168623,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00028391193394019376,
-      "loss": 0.7055,
+      "epoch": 1.0562947799385876,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000285270200094352,
+      "loss": 0.6651,
       "step": 1032
     },
     {
-      "epoch": 1.1056563500533618,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002837808830882864,
-      "loss": 0.6633,
+      "epoch": 1.060388945752303,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002851496443618143,
+      "loss": 0.6657,
       "step": 1036
     },
     {
-      "epoch": 1.1099252934898614,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00028364933113817615,
-      "loss": 0.6493,
+      "epoch": 1.0644831115660185,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00028502862298188634,
+      "loss": 0.6647,
       "step": 1040
     },
     {
-      "epoch": 1.1141942369263607,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002835172785826125,
-      "loss": 0.661,
+      "epoch": 1.068577277379734,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00028490713637153786,
+      "loss": 0.7025,
       "step": 1044
     },
     {
-      "epoch": 1.1184631803628602,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.0002833847259162199,
-      "loss": 0.6788,
+      "epoch": 1.0726714431934494,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00028478518494934123,
+      "loss": 0.6967,
       "step": 1048
     },
     {
-      "epoch": 1.1227321237993597,
+      "epoch": 1.076765609007165,
       "grad_norm": 0.251953125,
-      "learning_rate": 0.00028325167363549606,
-      "loss": 0.7103,
+      "learning_rate": 0.0002846627691354705,
+      "loss": 0.7205,
       "step": 1052
     },
     {
-      "epoch": 1.127001067235859,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00028311812223881,
-      "loss": 0.6641,
+      "epoch": 1.0808597748208801,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00028453988935169954,
+      "loss": 0.6407,
       "step": 1056
     },
     {
-      "epoch": 1.1312700106723586,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00028298407222640055,
-      "loss": 0.7192,
+      "epoch": 1.0849539406345956,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000284416546021401,
+      "loss": 0.6704,
       "step": 1060
     },
     {
-      "epoch": 1.135538954108858,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002828495241003738,
-      "loss": 0.7105,
+      "epoch": 1.089048106448311,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002842927395695446,
+      "loss": 0.692,
       "step": 1064
     },
     {
-      "epoch": 1.1398078975453576,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00028271447836470174,
-      "loss": 0.6379,
+      "epoch": 1.0931422722620265,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0002841684704226955,
+      "loss": 0.6846,
       "step": 1068
     },
     {
-      "epoch": 1.144076840981857,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00028257893552522046,
-      "loss": 0.6747,
+      "epoch": 1.097236438075742,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002840437390090133,
+      "loss": 0.651,
       "step": 1072
     },
     {
-      "epoch": 1.1483457844183564,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00028244289608962773,
-      "loss": 0.666,
+      "epoch": 1.1013306038894575,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002839185457582502,
+      "loss": 0.7293,
       "step": 1076
     },
     {
-      "epoch": 1.152614727854856,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002823063605674816,
-      "loss": 0.6981,
+      "epoch": 1.105424769703173,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002837928911017496,
+      "loss": 0.6636,
       "step": 1080
     },
     {
-      "epoch": 1.1568836712913555,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00028216932947019817,
-      "loss": 0.7096,
+      "epoch": 1.1095189355168884,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002836667754724447,
+      "loss": 0.6393,
       "step": 1084
     },
     {
-      "epoch": 1.1611526147278548,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002820318033110499,
-      "loss": 0.724,
+      "epoch": 1.1136131013306039,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0002835401993048568,
+      "loss": 0.7271,
       "step": 1088
     },
     {
-      "epoch": 1.1654215581643543,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0002818937826051636,
-      "loss": 0.7105,
+      "epoch": 1.1177072671443193,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002834131630350942,
+      "loss": 0.6788,
       "step": 1092
     },
     {
-      "epoch": 1.1696905016008539,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00028175526786951825,
-      "loss": 0.6659,
+      "epoch": 1.1218014329580348,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028328566710085024,
+      "loss": 0.6806,
       "step": 1096
     },
     {
-      "epoch": 1.1739594450373532,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002816162596229436,
-      "loss": 0.6361,
+      "epoch": 1.1258955987717503,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.000283157711941402,
+      "loss": 0.7403,
       "step": 1100
     },
     {
-      "epoch": 1.1782283884738527,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028147675838611775,
-      "loss": 0.6774,
+      "epoch": 1.1299897645854657,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000283029297997609,
+      "loss": 0.6585,
       "step": 1104
     },
     {
-      "epoch": 1.1824973319103522,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002813367646815655,
-      "loss": 0.681,
+      "epoch": 1.1340839303991812,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028290042571191114,
+      "loss": 0.6178,
       "step": 1108
     },
     {
-      "epoch": 1.1867662753468518,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002811962790336561,
-      "loss": 0.7276,
+      "epoch": 1.1381780962128967,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002827710955283277,
+      "loss": 0.688,
       "step": 1112
     },
     {
-      "epoch": 1.191035218783351,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00028105530196860154,
-      "loss": 0.7113,
+      "epoch": 1.1422722620266121,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00028264130789245565,
+      "loss": 0.6886,
       "step": 1116
     },
     {
-      "epoch": 1.1953041622198506,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00028091383401445454,
-      "loss": 0.6788,
+      "epoch": 1.1463664278403276,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00028251106325146797,
+      "loss": 0.6451,
       "step": 1120
     },
     {
-      "epoch": 1.1995731056563501,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00028077187570110654,
-      "loss": 0.6695,
+      "epoch": 1.150460593654043,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002823803620541122,
+      "loss": 0.7498,
       "step": 1124
     },
     {
-      "epoch": 1.2038420490928496,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028062942756028553,
-      "loss": 0.6908,
+      "epoch": 1.1545547594677585,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00028224920475070905,
+      "loss": 0.6605,
       "step": 1128
     },
     {
-      "epoch": 1.208110992529349,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002804864901255545,
-      "loss": 0.7365,
+      "epoch": 1.158648925281474,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00028211759179315053,
+      "loss": 0.6827,
       "step": 1132
     },
     {
-      "epoch": 1.2123799359658485,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002803430639323089,
-      "loss": 0.6305,
+      "epoch": 1.1627430910951895,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028198552363489874,
+      "loss": 0.6841,
       "step": 1136
     },
     {
-      "epoch": 1.216648879402348,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00028019914951777497,
-      "loss": 0.662,
+      "epoch": 1.1668372569089047,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000281853000730984,
+      "loss": 0.6681,
       "step": 1140
     },
     {
-      "epoch": 1.2209178228388473,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00028005474742100786,
-      "loss": 0.6763,
+      "epoch": 1.1709314227226202,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002817200235380035,
+      "loss": 0.6497,
       "step": 1144
     },
     {
-      "epoch": 1.2251867662753468,
+      "epoch": 1.1750255885363357,
       "grad_norm": 0.25,
-      "learning_rate": 0.0002799098581828892,
-      "loss": 0.6827,
+      "learning_rate": 0.00028158659251411954,
+      "loss": 0.6667,
       "step": 1148
     },
     {
-      "epoch": 1.2294557097118464,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00027976448234612534,
-      "loss": 0.6805,
+      "epoch": 1.1791197543500511,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002814527081190583,
+      "loss": 0.6781,
       "step": 1152
     },
     {
-      "epoch": 1.2337246531483457,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027961862045524527,
-      "loss": 0.664,
+      "epoch": 1.1832139201637666,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002813183708141077,
+      "loss": 0.6829,
       "step": 1156
     },
     {
-      "epoch": 1.2379935965848452,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00027947227305659866,
-      "loss": 0.7206,
+      "epoch": 1.187308085977482,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00028118358106211635,
+      "loss": 0.6888,
       "step": 1160
     },
     {
-      "epoch": 1.2422625400213447,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00027932544069835354,
-      "loss": 0.6668,
+      "epoch": 1.1914022517911975,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002810483393274916,
+      "loss": 0.7212,
       "step": 1164
     },
     {
-      "epoch": 1.2465314834578443,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002791781239304947,
-      "loss": 0.6946,
+      "epoch": 1.195496417604913,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00028091264607619826,
+      "loss": 0.7186,
       "step": 1168
     },
     {
-      "epoch": 1.2508004268943438,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00027903032330482106,
-      "loss": 0.6449,
+      "epoch": 1.1995905834186285,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002807765017757565,
+      "loss": 0.6889,
       "step": 1172
     },
     {
-      "epoch": 1.255069370330843,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002788820393749442,
-      "loss": 0.6893,
+      "epoch": 1.203684749232344,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00028063990689524093,
+      "loss": 0.7395,
       "step": 1176
     },
     {
-      "epoch": 1.2593383137673426,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002787332726962858,
-      "loss": 0.645,
+      "epoch": 1.2077789150460594,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00028050286190527823,
+      "loss": 0.6695,
       "step": 1180
     },
     {
-      "epoch": 1.2636072572038421,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00027858402382607573,
-      "loss": 0.689,
+      "epoch": 1.2118730808597749,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00028036536727804606,
+      "loss": 0.6742,
       "step": 1184
     },
     {
-      "epoch": 1.2678762006403415,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002784342933233502,
-      "loss": 0.6507,
+      "epoch": 1.2159672466734903,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002802274234872713,
+      "loss": 0.6568,
       "step": 1188
     },
     {
-      "epoch": 1.272145144076841,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00027828408174894925,
-      "loss": 0.6714,
+      "epoch": 1.2200614124872058,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00028008903100822834,
+      "loss": 0.7136,
       "step": 1192
     },
     {
-      "epoch": 1.2764140875133405,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.000278133389665515,
-      "loss": 0.695,
+      "epoch": 1.2241555783009213,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002799501903177375,
+      "loss": 0.631,
       "step": 1196
     },
     {
-      "epoch": 1.2806830309498398,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002779822176374892,
-      "loss": 0.7099,
+      "epoch": 1.2282497441146367,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00027981090189416343,
+      "loss": 0.7038,
       "step": 1200
     },
     {
-      "epoch": 1.2849519743863393,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00027783056623111155,
-      "loss": 0.6568,
+      "epoch": 1.2323439099283522,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00027967116621741326,
+      "loss": 0.6591,
       "step": 1204
     },
     {
-      "epoch": 1.2892209178228389,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0002776784360144172,
-      "loss": 0.7091,
+      "epoch": 1.2364380757420674,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002795309837689352,
+      "loss": 0.7088,
       "step": 1208
     },
     {
-      "epoch": 1.2934898612593382,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027752582755723474,
-      "loss": 0.7185,
+      "epoch": 1.240532241555783,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002793903550317169,
+      "loss": 0.6818,
       "step": 1212
     },
     {
-      "epoch": 1.2977588046958377,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002773727414311842,
-      "loss": 0.6609,
+      "epoch": 1.2446264073694984,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00027924928049028337,
+      "loss": 0.7158,
       "step": 1216
     },
     {
-      "epoch": 1.3020277481323372,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027721917820967465,
-      "loss": 0.6336,
+      "epoch": 1.2487205731832138,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00027910776063069586,
+      "loss": 0.6467,
       "step": 1220
     },
     {
-      "epoch": 1.3062966915688368,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00027706513846790235,
-      "loss": 0.6845,
+      "epoch": 1.2528147389969293,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002789657959405498,
+      "loss": 0.6674,
       "step": 1224
     },
     {
-      "epoch": 1.3105656350053363,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00027691062278284835,
-      "loss": 0.6605,
+      "epoch": 1.2569089048106448,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00027882338690897327,
+      "loss": 0.6747,
       "step": 1228
     },
     {
-      "epoch": 1.3148345784418356,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002767556317332764,
-      "loss": 0.6922,
+      "epoch": 1.2610030706243602,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00027868053402662534,
+      "loss": 0.6999,
       "step": 1232
     },
     {
-      "epoch": 1.3191035218783351,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00027660016589973097,
-      "loss": 0.6919,
+      "epoch": 1.2650972364380757,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00027853723778569427,
+      "loss": 0.7059,
       "step": 1236
     },
     {
-      "epoch": 1.3233724653148347,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002764442258645347,
-      "loss": 0.6739,
+      "epoch": 1.2691914022517912,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00027839349867989587,
+      "loss": 0.714,
       "step": 1240
     },
     {
-      "epoch": 1.327641408751334,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00027628781221178655,
-      "loss": 0.7121,
+      "epoch": 1.2732855680655066,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00027824931720447194,
+      "loss": 0.712,
       "step": 1244
     },
     {
-      "epoch": 1.3319103521878335,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002761309255273595,
-      "loss": 0.6966,
+      "epoch": 1.277379733879222,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002781046938561882,
+      "loss": 0.6872,
       "step": 1248
     },
     {
-      "epoch": 1.336179295624333,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00027597356639889826,
-      "loss": 0.6589,
+      "epoch": 1.2814738996929376,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00027795962913333304,
+      "loss": 0.7098,
       "step": 1252
     },
     {
-      "epoch": 1.3404482390608323,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002758157354158173,
-      "loss": 0.6605,
+      "epoch": 1.285568065506653,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027781412353571544,
+      "loss": 0.6901,
       "step": 1256
     },
     {
-      "epoch": 1.3447171824973319,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00027565743316929824,
-      "loss": 0.7141,
+      "epoch": 1.2896622313203685,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00027766817756466334,
+      "loss": 0.6867,
       "step": 1260
     },
     {
-      "epoch": 1.3489861259338314,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002754986602522882,
-      "loss": 0.682,
+      "epoch": 1.293756397134084,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00027752179172302213,
+      "loss": 0.7032,
       "step": 1264
     },
     {
-      "epoch": 1.353255069370331,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002753394172594972,
-      "loss": 0.7279,
+      "epoch": 1.2978505629477994,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002773749665151525,
+      "loss": 0.6738,
       "step": 1268
     },
     {
-      "epoch": 1.3575240128068304,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002751797047873957,
-      "loss": 0.6412,
+      "epoch": 1.301944728761515,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00027722770244692924,
+      "loss": 0.6769,
       "step": 1272
     },
     {
-      "epoch": 1.3617929562433297,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00027501952343421323,
-      "loss": 0.6337,
+      "epoch": 1.3060388945752304,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002770800000257388,
+      "loss": 0.6839,
       "step": 1276
     },
     {
-      "epoch": 1.3660618996798293,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.000274858873799935,
-      "loss": 0.6944,
+      "epoch": 1.3101330603889458,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002769318597604784,
+      "loss": 0.6871,
       "step": 1280
     },
     {
-      "epoch": 1.3703308431163288,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002746977564863007,
-      "loss": 0.6983,
+      "epoch": 1.3142272262026613,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002767832821615534,
+      "loss": 0.7099,
       "step": 1284
     },
     {
-      "epoch": 1.374599786552828,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.0002745361720968016,
-      "loss": 0.6566,
+      "epoch": 1.3183213920163768,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002766342677408763,
+      "loss": 0.702,
       "step": 1288
     },
     {
-      "epoch": 1.3788687299893276,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00027437412123667833,
-      "loss": 0.6634,
+      "epoch": 1.3224155578300922,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002764848170118644,
+      "loss": 0.6578,
       "step": 1292
     },
     {
-      "epoch": 1.3831376734258272,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00027421160451291906,
-      "loss": 0.7026,
+      "epoch": 1.3265097236438077,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00027633493048943825,
+      "loss": 0.6473,
       "step": 1296
     },
     {
-      "epoch": 1.3874066168623265,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00027404862253425677,
-      "loss": 0.6845,
+      "epoch": 1.330603889457523,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027618460869002016,
+      "loss": 0.6512,
       "step": 1300
     },
     {
-      "epoch": 1.391675560298826,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002738851759111671,
-      "loss": 0.652,
+      "epoch": 1.3346980552712384,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00027603385213153186,
+      "loss": 0.7167,
       "step": 1304
     },
     {
-      "epoch": 1.3959445037353255,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00027372126525586614,
-      "loss": 0.6708,
+      "epoch": 1.3387922210849539,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002758826613333932,
+      "loss": 0.7231,
       "step": 1308
     },
     {
-      "epoch": 1.400213447171825,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00027355689118230823,
-      "loss": 0.6644,
+      "epoch": 1.3428863868986693,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00027573103681652,
+      "loss": 0.7133,
       "step": 1312
     },
     {
-      "epoch": 1.4044823906083244,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002733920543061832,
-      "loss": 0.6562,
+      "epoch": 1.3469805527123848,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002755789791033227,
+      "loss": 0.6972,
       "step": 1316
     },
     {
-      "epoch": 1.4087513340448239,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002732267552449146,
-      "loss": 0.644,
+      "epoch": 1.3510747185261003,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00027542648871770384,
+      "loss": 0.7027,
       "step": 1320
     },
     {
-      "epoch": 1.4130202774813234,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00027306099461765716,
-      "loss": 0.6693,
+      "epoch": 1.3551688843398157,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00027527356618505715,
+      "loss": 0.6562,
       "step": 1324
     },
     {
-      "epoch": 1.417289220917823,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002728947730452945,
-      "loss": 0.7215,
+      "epoch": 1.3592630501535312,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027512021203226507,
+      "loss": 0.6721,
       "step": 1328
     },
     {
-      "epoch": 1.4215581643543223,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002727280911504367,
-      "loss": 0.641,
+      "epoch": 1.3633572159672467,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00027496642678769717,
+      "loss": 0.7029,
       "step": 1332
     },
     {
-      "epoch": 1.4258271077908218,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.000272560949557418,
-      "loss": 0.6825,
+      "epoch": 1.3674513817809621,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002748122109812083,
+      "loss": 0.7028,
       "step": 1336
     },
     {
-      "epoch": 1.4300960512273213,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00027239334889229467,
-      "loss": 0.6791,
+      "epoch": 1.3715455475946776,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00027465756514413677,
+      "loss": 0.6865,
       "step": 1340
     },
     {
-      "epoch": 1.4343649946638206,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00027222528978284254,
-      "loss": 0.6332,
+      "epoch": 1.375639713408393,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00027450248980930264,
+      "loss": 0.7197,
       "step": 1344
     },
     {
-      "epoch": 1.4386339381003201,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002720567728585544,
-      "loss": 0.6817,
+      "epoch": 1.3797338792221086,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00027434698551100567,
+      "loss": 0.6694,
       "step": 1348
     },
     {
-      "epoch": 1.4429028815368197,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.000271887798750638,
-      "loss": 0.7123,
+      "epoch": 1.383828045035824,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002741910527850235,
+      "loss": 0.7057,
       "step": 1352
     },
     {
-      "epoch": 1.447171824973319,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002717183680920135,
-      "loss": 0.6692,
+      "epoch": 1.3879222108495395,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002740346921686101,
+      "loss": 0.7029,
       "step": 1356
     },
     {
-      "epoch": 1.4514407684098185,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002715484815173113,
-      "loss": 0.7111,
+      "epoch": 1.3920163766632547,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00027387790420049357,
+      "loss": 0.6723,
       "step": 1360
     },
     {
-      "epoch": 1.455709711846318,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0002713781396628693,
-      "loss": 0.6938,
+      "epoch": 1.3961105424769702,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002737206894208744,
+      "loss": 0.6895,
       "step": 1364
     },
     {
-      "epoch": 1.4599786552828176,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00027120734316673074,
-      "loss": 0.6775,
+      "epoch": 1.4002047082906857,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002735630483714236,
+      "loss": 0.6655,
       "step": 1368
     },
     {
-      "epoch": 1.464247598719317,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.000271036092668642,
-      "loss": 0.6807,
+      "epoch": 1.4042988741044011,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00027340498159528106,
+      "loss": 0.6728,
       "step": 1372
     },
     {
-      "epoch": 1.4685165421558164,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00027086438881004975,
-      "loss": 0.6702,
+      "epoch": 1.4083930399181166,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00027324648963705317,
+      "loss": 0.7074,
       "step": 1376
     },
     {
-      "epoch": 1.472785485592316,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.000270692232234099,
-      "loss": 0.6432,
+      "epoch": 1.412487205731832,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00027308757304281154,
+      "loss": 0.6722,
       "step": 1380
     },
     {
-      "epoch": 1.4770544290288155,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002705196235856303,
-      "loss": 0.7068,
+      "epoch": 1.4165813715455475,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00027292823236009056,
+      "loss": 0.6882,
       "step": 1384
     },
     {
-      "epoch": 1.4813233724653148,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002703465635111777,
-      "loss": 0.6775,
+      "epoch": 1.420675537359263,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0002727684681378861,
+      "loss": 0.6971,
       "step": 1388
     },
     {
-      "epoch": 1.4855923159018143,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00027017305265896596,
-      "loss": 0.6731,
+      "epoch": 1.4247697031729785,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000272608280926653,
+      "loss": 0.6815,
       "step": 1392
     },
     {
-      "epoch": 1.4898612593383138,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002699990916789084,
-      "loss": 0.7215,
+      "epoch": 1.428863868986694,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00027244767127830366,
+      "loss": 0.6712,
       "step": 1396
     },
     {
-      "epoch": 1.4941302027748131,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00026982468122260436,
-      "loss": 0.6321,
+      "epoch": 1.4329580348004094,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00027228663974620583,
+      "loss": 0.6738,
       "step": 1400
     },
     {
-      "epoch": 1.4983991462113126,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00026964982194333686,
-      "loss": 0.6468,
+      "epoch": 1.4370522006141249,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000272125186885181,
+      "loss": 0.6815,
       "step": 1404
     },
     {
-      "epoch": 1.5026680896478122,
+      "epoch": 1.4411463664278403,
       "grad_norm": 0.28515625,
-      "learning_rate": 0.00026947451449606984,
-      "loss": 0.6713,
+      "learning_rate": 0.00027196331325150217,
+      "loss": 0.6807,
       "step": 1408
     },
     {
-      "epoch": 1.5069370330843115,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002692987595374461,
-      "loss": 0.6192,
+      "epoch": 1.4452405322415558,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00027180101940289206,
+      "loss": 0.6822,
       "step": 1412
     },
     {
-      "epoch": 1.5112059765208112,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00026912255772578456,
-      "loss": 0.6668,
+      "epoch": 1.4493346980552713,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002716383058985213,
+      "loss": 0.6641,
       "step": 1416
     },
     {
-      "epoch": 1.5154749199573105,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00026894590972107807,
-      "loss": 0.6811,
+      "epoch": 1.4534288638689867,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00027147517329900636,
+      "loss": 0.7067,
       "step": 1420
     },
     {
-      "epoch": 1.51974386339381,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002687688161849906,
-      "loss": 0.7069,
+      "epoch": 1.4575230296827022,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00027131162216640774,
+      "loss": 0.6732,
       "step": 1424
     },
     {
-      "epoch": 1.5240128068303096,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002685912777808549,
-      "loss": 0.69,
+      "epoch": 1.4616171954964177,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002711476530642279,
+      "loss": 0.6609,
       "step": 1428
     },
     {
-      "epoch": 1.528281750266809,
+      "epoch": 1.4657113613101331,
       "grad_norm": 0.267578125,
-      "learning_rate": 0.00026841329517367023,
-      "loss": 0.7054,
+      "learning_rate": 0.0002709832665574093,
+      "loss": 0.7021,
       "step": 1432
     },
     {
-      "epoch": 1.5325506937033084,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00026823486903009957,
-      "loss": 0.6814,
+      "epoch": 1.4698055271238486,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00027081846321233273,
+      "loss": 0.6793,
       "step": 1436
     },
     {
-      "epoch": 1.536819637139808,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002680560000184673,
-      "loss": 0.6867,
+      "epoch": 1.473899692937564,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000270653243596815,
+      "loss": 0.6538,
       "step": 1440
     },
     {
-      "epoch": 1.5410885805763073,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002678766888087565,
-      "loss": 0.6085,
+      "epoch": 1.4779938587512795,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00027048760828010725,
+      "loss": 0.7086,
       "step": 1444
     },
     {
-      "epoch": 1.5453575240128068,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00026769693607260674,
-      "loss": 0.6843,
+      "epoch": 1.482088024564995,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00027032155783289274,
+      "loss": 0.6975,
       "step": 1448
     },
     {
-      "epoch": 1.5496264674493063,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00026751674248331125,
-      "loss": 0.6953,
+      "epoch": 1.4861821903787105,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00027015509282728525,
+      "loss": 0.7226,
       "step": 1452
     },
     {
-      "epoch": 1.5538954108858056,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002673361087158147,
-      "loss": 0.6817,
+      "epoch": 1.4902763561924257,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00026998821383682664,
+      "loss": 0.6931,
       "step": 1456
     },
     {
-      "epoch": 1.5581643543223054,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002671550354467104,
-      "loss": 0.7003,
+      "epoch": 1.4943705220061412,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00026982092143648537,
+      "loss": 0.624,
       "step": 1460
     },
     {
-      "epoch": 1.5624332977588047,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.000266973523354238,
-      "loss": 0.6627,
+      "epoch": 1.4984646878198566,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00026965321620265405,
+      "loss": 0.6446,
       "step": 1464
     },
     {
-      "epoch": 1.5667022411953042,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002667915731182806,
-      "loss": 0.6727,
+      "epoch": 1.5025588536335721,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002694850987131478,
+      "loss": 0.7135,
       "step": 1468
     },
     {
-      "epoch": 1.5709711846318037,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002666091854203628,
-      "loss": 0.7468,
+      "epoch": 1.5066530194472876,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002693165695472022,
+      "loss": 0.6878,
       "step": 1472
     },
     {
-      "epoch": 1.575240128068303,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002664263609436474,
-      "loss": 0.6985,
+      "epoch": 1.510747185261003,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00026914762928547097,
+      "loss": 0.6612,
       "step": 1476
     },
     {
-      "epoch": 1.5795090715048026,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00026624310037293354,
-      "loss": 0.7159,
+      "epoch": 1.5148413510747185,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00026897827851002457,
+      "loss": 0.6975,
       "step": 1480
     },
     {
-      "epoch": 1.583778014941302,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002660594043946537,
-      "loss": 0.7057,
+      "epoch": 1.518935516888434,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002688085178043475,
+      "loss": 0.673,
       "step": 1484
     },
     {
-      "epoch": 1.5880469583778014,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00026587527369687115,
-      "loss": 0.6788,
+      "epoch": 1.5230296827021494,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000268638347753337,
+      "loss": 0.6826,
       "step": 1488
     },
     {
-      "epoch": 1.592315901814301,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00026569070896927767,
-      "loss": 0.6561,
+      "epoch": 1.527123848515865,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002684677689433004,
+      "loss": 0.6603,
       "step": 1492
     },
     {
-      "epoch": 1.5965848452508005,
+      "epoch": 1.5312180143295804,
       "grad_norm": 0.2734375,
-      "learning_rate": 0.0002655057109031906,
-      "loss": 0.6514,
+      "learning_rate": 0.0002682967819619535,
+      "loss": 0.698,
       "step": 1496
     },
     {
-      "epoch": 1.6008537886872998,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002653202801915505,
-      "loss": 0.6549,
+      "epoch": 1.5353121801432958,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026812538739841833,
+      "loss": 0.7188,
       "step": 1500
     },
     {
-      "epoch": 1.6051227321237995,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0002651344175289185,
-      "loss": 0.6261,
+      "epoch": 1.5394063459570113,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00026795358584322135,
+      "loss": 0.7101,
       "step": 1504
     },
     {
-      "epoch": 1.6093916755602988,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002649481236114735,
-      "loss": 0.6311,
+      "epoch": 1.5435005117707266,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002677813778882911,
+      "loss": 0.68,
       "step": 1508
     },
     {
-      "epoch": 1.6136606189967981,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002647613991370101,
-      "loss": 0.7317,
+      "epoch": 1.547594677584442,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002676087641269566,
+      "loss": 0.7013,
       "step": 1512
     },
     {
-      "epoch": 1.6179295624332979,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002645742448049352,
-      "loss": 0.6584,
+      "epoch": 1.5516888433981575,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002674357451539448,
+      "loss": 0.6477,
       "step": 1516
     },
     {
-      "epoch": 1.6221985058697972,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.0002643866613162661,
-      "loss": 0.7089,
+      "epoch": 1.555783009211873,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00026726232156537886,
+      "loss": 0.6855,
       "step": 1520
     },
     {
-      "epoch": 1.6264674493062967,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00026419864937362746,
-      "loss": 0.6484,
+      "epoch": 1.5598771750255884,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000267088493958776,
+      "loss": 0.6618,
       "step": 1524
     },
     {
-      "epoch": 1.6307363927427962,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002640102096812487,
-      "loss": 0.6893,
+      "epoch": 1.563971340839304,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002669142629330455,
+      "loss": 0.6477,
       "step": 1528
     },
     {
-      "epoch": 1.6350053361792956,
+      "epoch": 1.5680655066530194,
       "grad_norm": 0.255859375,
-      "learning_rate": 0.0002638213429449617,
-      "loss": 0.7095,
+      "learning_rate": 0.00026673962908848654,
+      "loss": 0.7286,
       "step": 1532
     },
     {
-      "epoch": 1.639274279615795,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00026363204987219765,
-      "loss": 0.6641,
+      "epoch": 1.5721596724667348,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002665645930267862,
+      "loss": 0.6947,
       "step": 1536
     },
     {
-      "epoch": 1.6435432230522946,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002634423311719847,
-      "loss": 0.7056,
+      "epoch": 1.5762538382804503,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002663891553510174,
+      "loss": 0.6563,
       "step": 1540
     },
     {
-      "epoch": 1.647812166488794,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00026325218755494543,
-      "loss": 0.6934,
+      "epoch": 1.5803480040941658,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026621331666563665,
+      "loss": 0.6722,
       "step": 1544
     },
     {
-      "epoch": 1.6520811099252934,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00026306161973329374,
-      "loss": 0.663,
+      "epoch": 1.5844421699078812,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002660370775764822,
+      "loss": 0.6563,
       "step": 1548
     },
     {
-      "epoch": 1.656350053361793,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002628706284208328,
-      "loss": 0.6673,
+      "epoch": 1.5885363357215967,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000265860438690772,
+      "loss": 0.678,
       "step": 1552
     },
     {
-      "epoch": 1.6606189967982923,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00026267921433295165,
-      "loss": 0.7057,
+      "epoch": 1.5926305015353122,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00026568340061710124,
+      "loss": 0.6558,
       "step": 1556
     },
     {
-      "epoch": 1.664887940234792,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002624873781866233,
-      "loss": 0.6915,
+      "epoch": 1.5967246673490276,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002655059639654406,
+      "loss": 0.6684,
       "step": 1560
     },
     {
-      "epoch": 1.6691568836712913,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00026229512070040123,
-      "loss": 0.6858,
+      "epoch": 1.600818833162743,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.000265328129347134,
+      "loss": 0.6614,
       "step": 1564
     },
     {
-      "epoch": 1.6734258271077909,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002621024425944175,
-      "loss": 0.6542,
+      "epoch": 1.6049129989764586,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00026514989737489646,
+      "loss": 0.6763,
       "step": 1568
     },
     {
-      "epoch": 1.6776947705442904,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00026190934459037945,
-      "loss": 0.7185,
+      "epoch": 1.609007164790174,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00026497126866281223,
+      "loss": 0.6667,
       "step": 1572
     },
     {
-      "epoch": 1.6819637139807897,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00026171582741156725,
-      "loss": 0.6859,
+      "epoch": 1.6131013306038895,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002647922438263323,
+      "loss": 0.6783,
       "step": 1576
     },
     {
-      "epoch": 1.6862326574172892,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002615218917828312,
-      "loss": 0.6788,
+      "epoch": 1.617195496417605,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00026461282348227267,
+      "loss": 0.6843,
       "step": 1580
     },
     {
-      "epoch": 1.6905016008537888,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.00026132753843058896,
-      "loss": 0.7032,
+      "epoch": 1.6212896622313204,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00026443300824881174,
+      "loss": 0.6728,
       "step": 1584
     },
     {
-      "epoch": 1.694770544290288,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00026113276808282284,
-      "loss": 0.6972,
+      "epoch": 1.625383828045036,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00026425279874548883,
+      "loss": 0.6666,
       "step": 1588
     },
     {
-      "epoch": 1.6990394877267876,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00026093758146907703,
-      "loss": 0.682,
+      "epoch": 1.6294779938587514,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002640721955932013,
+      "loss": 0.6979,
       "step": 1592
     },
     {
-      "epoch": 1.7033084311632871,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00026074197932045496,
-      "loss": 0.7364,
+      "epoch": 1.6335721596724668,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00026389119941420323,
+      "loss": 0.6888,
       "step": 1596
     },
     {
-      "epoch": 1.7075773745997864,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00026054596236961655,
-      "loss": 0.7075,
+      "epoch": 1.6376663254861823,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002637098108321024,
+      "loss": 0.6875,
       "step": 1600
     },
     {
-      "epoch": 1.7118463180362862,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00026034953135077533,
-      "loss": 0.6649,
+      "epoch": 1.6417604912998978,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000263528030471859,
+      "loss": 0.6285,
       "step": 1604
     },
     {
-      "epoch": 1.7161152614727855,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00026015268699969594,
-      "loss": 0.6433,
+      "epoch": 1.6458546571136132,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002633458589597827,
+      "loss": 0.6737,
       "step": 1608
     },
     {
-      "epoch": 1.7203842049092848,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.000259955430053691,
-      "loss": 0.6143,
+      "epoch": 1.6499488229273287,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002631632969235311,
+      "loss": 0.6747,
       "step": 1612
     },
     {
-      "epoch": 1.7246531483457845,
-      "grad_norm": 0.25,
-      "learning_rate": 0.0002597577612516187,
-      "loss": 0.7164,
+      "epoch": 1.6540429887410442,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00026298034499210715,
+      "loss": 0.6784,
       "step": 1616
     },
     {
-      "epoch": 1.7289220917822838,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002595596813338801,
-      "loss": 0.7572,
+      "epoch": 1.6581371545547596,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00026279700379585724,
+      "loss": 0.6657,
       "step": 1620
     },
     {
-      "epoch": 1.7331910352187834,
+      "epoch": 1.6622313203684749,
       "grad_norm": 0.275390625,
-      "learning_rate": 0.00025936119104241577,
-      "loss": 0.6611,
+      "learning_rate": 0.000262613273966469,
+      "loss": 0.705,
       "step": 1624
     },
     {
-      "epoch": 1.737459978655283,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002591622911207036,
-      "loss": 0.656,
+      "epoch": 1.6663254861821903,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00026242915613696897,
+      "loss": 0.7061,
       "step": 1628
     },
     {
-      "epoch": 1.7417289220917822,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0002589629823137559,
-      "loss": 0.7315,
+      "epoch": 1.6704196519959058,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0002622446509417206,
+      "loss": 0.6685,
       "step": 1632
     },
     {
-      "epoch": 1.7459978655282817,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00025876326536811644,
-      "loss": 0.7098,
+      "epoch": 1.6745138178096213,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00026205975901642174,
+      "loss": 0.7255,
       "step": 1636
     },
     {
-      "epoch": 1.7502668089647813,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002585631410318577,
-      "loss": 0.6563,
+      "epoch": 1.6786079836233367,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000261874480998103,
+      "loss": 0.5933,
       "step": 1640
     },
     {
-      "epoch": 1.7545357524012806,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002583626100545782,
-      "loss": 0.6822,
+      "epoch": 1.6827021494370522,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00026168881752512517,
+      "loss": 0.7004,
       "step": 1644
     },
     {
-      "epoch": 1.75880469583778,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00025816167318739946,
-      "loss": 0.711,
+      "epoch": 1.6867963152507677,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00026150276923717693,
+      "loss": 0.6795,
       "step": 1648
     },
     {
-      "epoch": 1.7630736392742796,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002579603311829635,
-      "loss": 0.6557,
+      "epoch": 1.6908904810644831,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002613163367752729,
+      "loss": 0.6927,
       "step": 1652
     },
     {
-      "epoch": 1.767342582710779,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002577585847954297,
-      "loss": 0.6914,
+      "epoch": 1.6949846468781986,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00026112952078175146,
+      "loss": 0.6589,
       "step": 1656
     },
     {
-      "epoch": 1.7716115261472787,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00025755643478047224,
-      "loss": 0.6466,
+      "epoch": 1.699078812691914,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002609423219002722,
+      "loss": 0.6693,
       "step": 1660
     },
     {
-      "epoch": 1.775880469583778,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002573538818952771,
-      "loss": 0.7048,
+      "epoch": 1.7031729785056293,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002607547407758141,
+      "loss": 0.7185,
       "step": 1664
     },
     {
-      "epoch": 1.7801494130202775,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00025715092689853925,
-      "loss": 0.6707,
+      "epoch": 1.7072671443193448,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00026056677805467304,
+      "loss": 0.6349,
       "step": 1668
     },
     {
-      "epoch": 1.784418356456777,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025694757055045984,
-      "loss": 0.6981,
+      "epoch": 1.7113613101330603,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002603784343844597,
+      "loss": 0.7128,
       "step": 1672
     },
     {
-      "epoch": 1.7886872998932764,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00025674381361274345,
-      "loss": 0.6725,
+      "epoch": 1.7154554759467757,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00026018971041409715,
+      "loss": 0.6974,
       "step": 1676
     },
     {
-      "epoch": 1.7929562433297759,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.000256539656848595,
-      "loss": 0.7183,
+      "epoch": 1.7195496417604912,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002600006067938191,
+      "loss": 0.6253,
       "step": 1680
     },
     {
-      "epoch": 1.7972251867662754,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 0.0002563351010227171,
-      "loss": 0.6967,
+      "epoch": 1.7236438075742067,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00025981112417516693,
+      "loss": 0.688,
       "step": 1684
     },
     {
-      "epoch": 1.8014941302027747,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.000256130146901307,
-      "loss": 0.6597,
+      "epoch": 1.7277379733879221,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000259621263210988,
+      "loss": 0.6281,
       "step": 1688
     },
     {
-      "epoch": 1.8057630736392742,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002559247952520541,
-      "loss": 0.6943,
+      "epoch": 1.7318321392016376,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002594310245554333,
+      "loss": 0.7162,
       "step": 1692
     },
     {
-      "epoch": 1.8100320170757738,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00025571904684413643,
-      "loss": 0.6596,
+      "epoch": 1.735926305015353,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002592404088639549,
+      "loss": 0.7011,
       "step": 1696
     },
     {
-      "epoch": 1.814300960512273,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00025551290244821853,
-      "loss": 0.6709,
+      "epoch": 1.7400204708290685,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002590494167933042,
+      "loss": 0.6928,
       "step": 1700
     },
     {
-      "epoch": 1.8185699039487728,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.00025530636283644786,
-      "loss": 0.6656,
+      "epoch": 1.744114636642784,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002588580490015292,
+      "loss": 0.6628,
       "step": 1704
     },
     {
-      "epoch": 1.8228388473852721,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002550994287824523,
-      "loss": 0.6754,
+      "epoch": 1.7482088024564995,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025866630614797243,
+      "loss": 0.6604,
       "step": 1708
     },
     {
-      "epoch": 1.8271077908217717,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002548921010613374,
-      "loss": 0.6981,
+      "epoch": 1.752302968270215,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025847418889326867,
+      "loss": 0.7133,
       "step": 1712
     },
     {
-      "epoch": 1.8313767342582712,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00025468438044968296,
-      "loss": 0.6771,
+      "epoch": 1.7563971340839304,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002582816978993428,
+      "loss": 0.6795,
       "step": 1716
     },
     {
-      "epoch": 1.8356456776947705,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025447626772554064,
-      "loss": 0.6968,
+      "epoch": 1.7604912998976459,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002580888338294072,
+      "loss": 0.6114,
       "step": 1720
     },
     {
-      "epoch": 1.83991462113127,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00025426776366843065,
-      "loss": 0.7235,
+      "epoch": 1.7645854657113613,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00025789559734795984,
+      "loss": 0.6771,
       "step": 1724
     },
     {
-      "epoch": 1.8441835645677696,
+      "epoch": 1.7686796315250768,
       "grad_norm": 0.255859375,
-      "learning_rate": 0.0002540588690593391,
-      "loss": 0.6633,
+      "learning_rate": 0.0002577019891207816,
+      "loss": 0.6623,
       "step": 1728
     },
     {
-      "epoch": 1.8484525080042689,
+      "epoch": 1.7727737973387923,
       "grad_norm": 0.2734375,
-      "learning_rate": 0.000253849584680715,
-      "loss": 0.6679,
+      "learning_rate": 0.00025750800981493434,
+      "loss": 0.6542,
       "step": 1732
     },
     {
-      "epoch": 1.8527214514407684,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002536399113164671,
-      "loss": 0.711,
+      "epoch": 1.7768679631525077,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002573136600987584,
+      "loss": 0.6879,
       "step": 1736
     },
     {
-      "epoch": 1.856990394877268,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002534298497519615,
-      "loss": 0.694,
+      "epoch": 1.7809621289662232,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002571189406418702,
+      "loss": 0.6624,
       "step": 1740
     },
     {
-      "epoch": 1.8612593383137672,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00025321940077401814,
-      "loss": 0.661,
+      "epoch": 1.7850562947799387,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002569238521151603,
+      "loss": 0.7035,
       "step": 1744
     },
     {
-      "epoch": 1.865528281750267,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00025300856517090805,
-      "loss": 0.7393,
+      "epoch": 1.7891504605936541,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002567283951907908,
+      "loss": 0.6695,
       "step": 1748
     },
     {
-      "epoch": 1.8697972251867663,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00025279734373235055,
-      "loss": 0.7019,
+      "epoch": 1.7932446264073696,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000256532570542193,
+      "loss": 0.6326,
       "step": 1752
     },
     {
-      "epoch": 1.8740661686232656,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002525857372495102,
-      "loss": 0.6417,
+      "epoch": 1.797338792221085,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002563363788440652,
+      "loss": 0.6895,
       "step": 1756
     },
     {
-      "epoch": 1.8783351120597653,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.0002523737465149936,
-      "loss": 0.6431,
+      "epoch": 1.8014329580348005,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00025613982077237043,
+      "loss": 0.6529,
       "step": 1760
     },
     {
-      "epoch": 1.8826040554962646,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002521613723228468,
-      "loss": 0.6928,
+      "epoch": 1.805527123848516,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002559428970043338,
+      "loss": 0.6558,
       "step": 1764
     },
     {
-      "epoch": 1.8868729989327642,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0002519486154685521,
-      "loss": 0.6203,
+      "epoch": 1.8096212896622315,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00025574560821844066,
+      "loss": 0.7197,
       "step": 1768
     },
     {
-      "epoch": 1.8911419423692637,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00025173547674902524,
-      "loss": 0.6691,
+      "epoch": 1.813715455475947,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002555479550944338,
+      "loss": 0.6678,
       "step": 1772
     },
     {
-      "epoch": 1.895410885805763,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.00025152195696261206,
-      "loss": 0.6797,
+      "epoch": 1.8178096212896624,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002553499383133115,
+      "loss": 0.6602,
       "step": 1776
     },
     {
-      "epoch": 1.8996798292422625,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002513080569090859,
-      "loss": 0.6709,
+      "epoch": 1.8219037871033776,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002551515585573248,
+      "loss": 0.6831,
       "step": 1780
     },
     {
-      "epoch": 1.903948772678762,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00025109377738964455,
-      "loss": 0.6488,
+      "epoch": 1.825997952917093,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002549528165099755,
+      "loss": 0.6853,
       "step": 1784
     },
     {
-      "epoch": 1.9082177161152614,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00025087911920690695,
-      "loss": 0.6707,
+      "epoch": 1.8300921187308086,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00025475371285601356,
+      "loss": 0.6836,
       "step": 1788
     },
     {
-      "epoch": 1.912486659551761,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0002506640831649104,
-      "loss": 0.7067,
+      "epoch": 1.834186284544524,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00025455424828143473,
+      "loss": 0.6681,
       "step": 1792
     },
     {
-      "epoch": 1.9167556029882604,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002504486700691077,
-      "loss": 0.6498,
+      "epoch": 1.8382804503582395,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002543544234734786,
+      "loss": 0.6962,
       "step": 1796
     },
     {
-      "epoch": 1.9210245464247597,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002502328807263638,
-      "loss": 0.6911,
+      "epoch": 1.842374616171955,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00025415423912062557,
+      "loss": 0.6646,
       "step": 1800
     },
     {
-      "epoch": 1.9252934898612595,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00025001671594495316,
-      "loss": 0.6766,
+      "epoch": 1.8464687819856704,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00025395369591259503,
+      "loss": 0.647,
       "step": 1804
     },
     {
-      "epoch": 1.9295624332977588,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.0002498001765345563,
-      "loss": 0.7028,
+      "epoch": 1.850562947799386,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00025375279454034264,
+      "loss": 0.6682,
       "step": 1808
     },
     {
-      "epoch": 1.9338313767342583,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00024958326330625695,
-      "loss": 0.677,
+      "epoch": 1.8546571136131014,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00025355153569605823,
+      "loss": 0.6989,
       "step": 1812
     },
     {
-      "epoch": 1.9381003201707578,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002493659770725392,
-      "loss": 0.6952,
+      "epoch": 1.8587512794268168,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00025334992007316315,
+      "loss": 0.7076,
       "step": 1816
     },
     {
-      "epoch": 1.9423692636072571,
+      "epoch": 1.862845445240532,
       "grad_norm": 0.26953125,
-      "learning_rate": 0.0002491483186472842,
-      "loss": 0.6376,
+      "learning_rate": 0.00025314794836630807,
+      "loss": 0.6588,
       "step": 1820
     },
     {
-      "epoch": 1.9466382070437567,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00024893028884576725,
-      "loss": 0.6701,
+      "epoch": 1.8669396110542475,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002529456212713705,
+      "loss": 0.6762,
       "step": 1824
     },
     {
-      "epoch": 1.9509071504802562,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002487118884846546,
-      "loss": 0.7071,
+      "epoch": 1.871033776867963,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0002527429394854524,
+      "loss": 0.7133,
       "step": 1828
     },
     {
-      "epoch": 1.9551760939167555,
+      "epoch": 1.8751279426816785,
       "grad_norm": 0.2578125,
-      "learning_rate": 0.0002484931183820006,
-      "loss": 0.6462,
+      "learning_rate": 0.0002525399037068778,
+      "loss": 0.7536,
       "step": 1832
     },
     {
-      "epoch": 1.959445037353255,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002482739793572445,
-      "loss": 0.6398,
+      "epoch": 1.879222108495394,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00025233651463519045,
+      "loss": 0.6799,
       "step": 1836
     },
     {
-      "epoch": 1.9637139807897546,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0002480544722312074,
-      "loss": 0.6437,
+      "epoch": 1.8833162743091094,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00025213277297115124,
+      "loss": 0.6846,
       "step": 1840
     },
     {
-      "epoch": 1.9679829242262539,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002478345978260892,
-      "loss": 0.6784,
+      "epoch": 1.8874104401228249,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002519286794167359,
+      "loss": 0.6479,
       "step": 1844
     },
     {
-      "epoch": 1.9722518676627536,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00024761435696546544,
-      "loss": 0.7093,
+      "epoch": 1.8915046059365404,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00025172423467513267,
+      "loss": 0.6588,
       "step": 1848
     },
     {
-      "epoch": 1.976520811099253,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002473937504742844,
-      "loss": 0.7183,
+      "epoch": 1.8955987717502558,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002515194394507396,
+      "loss": 0.6707,
       "step": 1852
     },
     {
-      "epoch": 1.9807897545357525,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002471727791788637,
-      "loss": 0.7027,
+      "epoch": 1.8996929375639713,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00025131429444916247,
+      "loss": 0.6688,
       "step": 1856
     },
     {
-      "epoch": 1.985058697972252,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0002469514439068876,
-      "loss": 0.7189,
+      "epoch": 1.9037871033776868,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00025110880037721215,
+      "loss": 0.6671,
       "step": 1860
     },
     {
-      "epoch": 1.9893276414087513,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002467297454874036,
-      "loss": 0.6315,
+      "epoch": 1.9078812691914022,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00025090295794290214,
+      "loss": 0.6216,
       "step": 1864
     },
     {
-      "epoch": 1.9935965848452508,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.0002465076847508194,
-      "loss": 0.7154,
+      "epoch": 1.9119754350051177,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00025069676785544623,
+      "loss": 0.6815,
       "step": 1868
     },
     {
-      "epoch": 1.9978655282817503,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00024628526252889983,
-      "loss": 0.6901,
+      "epoch": 1.9160696008188332,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00025049023082525607,
+      "loss": 0.7354,
       "step": 1872
     },
     {
-      "epoch": 2.0021344717182497,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0002460624796547638,
-      "loss": 0.6671,
+      "epoch": 1.9201637666325486,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002502833475639386,
+      "loss": 0.6874,
       "step": 1876
     },
     {
-      "epoch": 2.0064034151547494,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002458393369628809,
-      "loss": 0.5825,
+      "epoch": 1.924257932446264,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0002500761187842937,
+      "loss": 0.6667,
       "step": 1880
     },
     {
-      "epoch": 2.0106723585912487,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002456158352890689,
-      "loss": 0.5568,
+      "epoch": 1.9283520982599796,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002498685452003118,
+      "loss": 0.6612,
       "step": 1884
     },
     {
-      "epoch": 2.014941302027748,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00024539197547048967,
-      "loss": 0.5848,
+      "epoch": 1.932446264073695,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002496606275271711,
+      "loss": 0.6915,
       "step": 1888
     },
     {
-      "epoch": 2.0192102454642478,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002451677583456469,
-      "loss": 0.6178,
+      "epoch": 1.9365404298874105,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002494523664812355,
+      "loss": 0.6593,
       "step": 1892
     },
     {
-      "epoch": 2.023479188900747,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00024494318475438254,
-      "loss": 0.5733,
+      "epoch": 1.940634595701126,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00024924376278005197,
+      "loss": 0.6741,
       "step": 1896
     },
     {
-      "epoch": 2.0277481323372464,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00024471825553787374,
-      "loss": 0.6131,
-      "step": 1900
+      "epoch": 1.9447287615148414,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000249034817142348,
+      "loss": 0.6915,
+      "step": 1900
     },
     {
-      "epoch": 2.032017075773746,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0002444929715386297,
-      "loss": 0.616,
+      "epoch": 1.9488229273285569,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002488255302880293,
+      "loss": 0.6599,
       "step": 1904
     },
     {
-      "epoch": 2.0362860192102454,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00024426733360048854,
-      "loss": 0.5947,
+      "epoch": 1.9529170931422724,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000248615902938177,
+      "loss": 0.64,
       "step": 1908
     },
     {
-      "epoch": 2.0405549626467447,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002440413425686141,
-      "loss": 0.6151,
+      "epoch": 1.9570112589559878,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00024840593581504567,
+      "loss": 0.6631,
       "step": 1912
     },
     {
-      "epoch": 2.0448239060832445,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00024381499928949275,
-      "loss": 0.613,
+      "epoch": 1.9611054247697033,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002481956296420603,
+      "loss": 0.6774,
       "step": 1916
     },
     {
-      "epoch": 2.049092849519744,
-      "grad_norm": 0.255859375,
-      "learning_rate": 0.0002435883046109304,
-      "loss": 0.596,
+      "epoch": 1.9651995905834188,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002479849851438142,
+      "loss": 0.6896,
       "step": 1920
     },
     {
-      "epoch": 2.0533617929562435,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.00024336125938204906,
-      "loss": 0.5233,
+      "epoch": 1.9692937563971342,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002477740030460663,
+      "loss": 0.6584,
       "step": 1924
     },
     {
-      "epoch": 2.057630736392743,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00024313386445328375,
-      "loss": 0.6059,
+      "epoch": 1.9733879222108497,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002475626840757386,
+      "loss": 0.6339,
       "step": 1928
     },
     {
-      "epoch": 2.061899679829242,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.0002429061206763795,
-      "loss": 0.6198,
+      "epoch": 1.9774820880245652,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000247351028960914,
+      "loss": 0.6797,
       "step": 1932
     },
     {
-      "epoch": 2.066168623265742,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00024267802890438783,
-      "loss": 0.6005,
+      "epoch": 1.9815762538382804,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002471390384308334,
+      "loss": 0.6773,
       "step": 1936
     },
     {
-      "epoch": 2.070437566702241,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00024244958999166386,
-      "loss": 0.6141,
+      "epoch": 1.9856704196519959,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024692671321589326,
+      "loss": 0.6526,
       "step": 1940
     },
     {
-      "epoch": 2.0747065101387405,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00024222080479386298,
-      "loss": 0.5931,
+      "epoch": 1.9897645854657113,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024671405404764335,
+      "loss": 0.6519,
       "step": 1944
     },
     {
-      "epoch": 2.0789754535752403,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00024199167416793753,
-      "loss": 0.5823,
+      "epoch": 1.9938587512794268,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002465010616587841,
+      "loss": 0.6989,
       "step": 1948
     },
     {
-      "epoch": 2.0832443970117396,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.0002417621989721338,
-      "loss": 0.6171,
+      "epoch": 1.9979529170931423,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002462877367831637,
+      "loss": 0.7191,
       "step": 1952
     },
     {
-      "epoch": 2.087513340448239,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00024153238006598872,
-      "loss": 0.5621,
+      "epoch": 2.0020470829068575,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002460740801557763,
+      "loss": 0.6218,
       "step": 1956
     },
     {
-      "epoch": 2.0917822838847386,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00024130221831032667,
-      "loss": 0.5689,
+      "epoch": 2.006141248720573,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002458600925127587,
+      "loss": 0.6015,
       "step": 1960
     },
     {
-      "epoch": 2.096051227321238,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00024107171456725618,
-      "loss": 0.5939,
+      "epoch": 2.0102354145342884,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002456457745913885,
+      "loss": 0.5645,
       "step": 1964
     },
     {
-      "epoch": 2.1003201707577372,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00024084086970016676,
-      "loss": 0.6377,
+      "epoch": 2.014329580348004,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00024543112713008104,
+      "loss": 0.6114,
       "step": 1968
     },
     {
-      "epoch": 2.104589114194237,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00024060968457372561,
-      "loss": 0.6324,
+      "epoch": 2.0184237461617194,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00024521615086838726,
+      "loss": 0.5825,
       "step": 1972
     },
     {
-      "epoch": 2.1088580576307363,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00024037816005387455,
-      "loss": 0.5936,
+      "epoch": 2.022517911975435,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00024500084654699065,
+      "loss": 0.6006,
       "step": 1976
     },
     {
-      "epoch": 2.113127001067236,
+      "epoch": 2.0266120777891503,
       "grad_norm": 0.296875,
-      "learning_rate": 0.00024014629700782654,
-      "loss": 0.5778,
+      "learning_rate": 0.0002447852149077053,
+      "loss": 0.5932,
       "step": 1980
     },
     {
-      "epoch": 2.1173959445037354,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00023991409630406258,
-      "loss": 0.5956,
+      "epoch": 2.030706243602866,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00024456925669347294,
+      "loss": 0.5681,
       "step": 1984
     },
     {
-      "epoch": 2.1216648879402347,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00023968155881232844,
-      "loss": 0.5668,
+      "epoch": 2.0348004094165812,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00024435297264836043,
+      "loss": 0.5712,
       "step": 1988
     },
     {
-      "epoch": 2.1259338313767344,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00023944868540363135,
-      "loss": 0.587,
+      "epoch": 2.0388945752302967,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00024413636351755736,
+      "loss": 0.5174,
       "step": 1992
     },
     {
-      "epoch": 2.1302027748132337,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00023921547695023683,
-      "loss": 0.5851,
+      "epoch": 2.042988741044012,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00024391943004737333,
+      "loss": 0.6106,
       "step": 1996
     },
     {
-      "epoch": 2.134471718249733,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.0002389819343256653,
-      "loss": 0.5977,
+      "epoch": 2.0470829068577276,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00024370217298523534,
+      "loss": 0.5328,
       "step": 2000
     },
     {
-      "epoch": 2.138740661686233,
+      "epoch": 2.051177072671443,
       "grad_norm": 0.267578125,
-      "learning_rate": 0.00023874805840468888,
-      "loss": 0.5932,
+      "learning_rate": 0.00024348459307968546,
+      "loss": 0.6033,
       "step": 2004
     },
     {
-      "epoch": 2.143009605122732,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00023851385006332817,
-      "loss": 0.5572,
+      "epoch": 2.0552712384851586,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024326669108037802,
+      "loss": 0.5893,
       "step": 2008
     },
     {
-      "epoch": 2.1472785485592314,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00023827931017884885,
-      "loss": 0.5813,
+      "epoch": 2.059365404298874,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00024304846773807708,
+      "loss": 0.6343,
       "step": 2012
     },
     {
-      "epoch": 2.151547491995731,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023804443962975843,
-      "loss": 0.5636,
+      "epoch": 2.0634595701125895,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002428299238046538,
+      "loss": 0.5784,
       "step": 2016
     },
     {
-      "epoch": 2.1558164354322304,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00023780923929580308,
-      "loss": 0.5689,
+      "epoch": 2.067553735926305,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00024261106003308408,
+      "loss": 0.562,
       "step": 2020
     },
     {
-      "epoch": 2.1600853788687298,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002375737100579641,
-      "loss": 0.5949,
+      "epoch": 2.0716479017400204,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00024239187717744567,
+      "loss": 0.5716,
       "step": 2024
     },
     {
-      "epoch": 2.1643543223052295,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00023733785279845477,
-      "loss": 0.5579,
+      "epoch": 2.075742067553736,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002421723759929157,
+      "loss": 0.6365,
       "step": 2028
     },
     {
-      "epoch": 2.168623265741729,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00023710166840071724,
-      "loss": 0.6293,
+      "epoch": 2.0798362333674514,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0002419525572357682,
+      "loss": 0.6373,
       "step": 2032
     },
     {
-      "epoch": 2.1728922091782286,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00023686515774941867,
-      "loss": 0.556,
+      "epoch": 2.083930399181167,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00024173242166337114,
+      "loss": 0.628,
       "step": 2036
     },
     {
-      "epoch": 2.177161152614728,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002366283217304485,
-      "loss": 0.5949,
+      "epoch": 2.0880245649948823,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00024151197003418427,
+      "loss": 0.5754,
       "step": 2040
     },
     {
-      "epoch": 2.181430096051227,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023639116123091487,
-      "loss": 0.6119,
+      "epoch": 2.092118730808598,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002412912031077562,
+      "loss": 0.584,
       "step": 2044
     },
     {
-      "epoch": 2.185699039487727,
-      "grad_norm": 0.263671875,
-      "learning_rate": 0.00023615367713914119,
-      "loss": 0.6332,
+      "epoch": 2.0962128966223132,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002410701216447219,
+      "loss": 0.6191,
       "step": 2048
     },
     {
-      "epoch": 2.1899679829242262,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00023591587034466302,
-      "loss": 0.6104,
+      "epoch": 2.1003070624360287,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0002408487264068,
+      "loss": 0.657,
       "step": 2052
     },
     {
-      "epoch": 2.1942369263607255,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.0002356777417382247,
-      "loss": 0.5749,
+      "epoch": 2.104401228249744,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00024062701815679032,
+      "loss": 0.595,
       "step": 2056
     },
     {
-      "epoch": 2.1985058697972253,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0002354392922117758,
-      "loss": 0.5365,
+      "epoch": 2.1084953940634596,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00024040499765857093,
+      "loss": 0.6026,
       "step": 2060
     },
     {
-      "epoch": 2.2027748132337246,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023520052265846823,
-      "loss": 0.6188,
+      "epoch": 2.112589559877175,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000240182665677096,
+      "loss": 0.5219,
       "step": 2064
     },
     {
-      "epoch": 2.207043756670224,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.0002349614339726523,
-      "loss": 0.5737,
+      "epoch": 2.1166837256908906,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023996002297839268,
+      "loss": 0.6293,
       "step": 2068
     },
     {
-      "epoch": 2.2113127001067236,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00023472202704987393,
-      "loss": 0.6577,
+      "epoch": 2.120777891504606,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023973707032955879,
+      "loss": 0.6043,
       "step": 2072
     },
     {
-      "epoch": 2.215581643543223,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002344823027868709,
-      "loss": 0.5555,
+      "epoch": 2.1248720573183215,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00023951380849875995,
+      "loss": 0.6114,
       "step": 2076
     },
     {
-      "epoch": 2.2198505869797227,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00023424226208156967,
-      "loss": 0.5833,
+      "epoch": 2.128966223132037,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00023929023825522715,
+      "loss": 0.5865,
       "step": 2080
     },
     {
-      "epoch": 2.224119530416222,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00023400190583308206,
-      "loss": 0.5973,
+      "epoch": 2.1330603889457525,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023906636036925396,
+      "loss": 0.6042,
       "step": 2084
     },
     {
-      "epoch": 2.2283884738527213,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00023376123494170178,
-      "loss": 0.5994,
+      "epoch": 2.137154554759468,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00023884217561219386,
+      "loss": 0.6083,
       "step": 2088
     },
     {
-      "epoch": 2.232657417289221,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00023352025030890094,
-      "loss": 0.6407,
+      "epoch": 2.1412487205731834,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023861768475645772,
+      "loss": 0.6202,
       "step": 2092
     },
     {
-      "epoch": 2.2369263607257204,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00023327895283732706,
-      "loss": 0.6193,
+      "epoch": 2.145342886386899,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00023839288857551095,
+      "loss": 0.5945,
       "step": 2096
     },
     {
-      "epoch": 2.2411953041622197,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00023303734343079927,
-      "loss": 0.6367,
+      "epoch": 2.1494370522006143,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00023816778784387094,
+      "loss": 0.6022,
       "step": 2100
     },
     {
-      "epoch": 2.2454642475987194,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00023279542299430513,
-      "loss": 0.5854,
+      "epoch": 2.15353121801433,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00023794238333710454,
+      "loss": 0.5561,
       "step": 2104
     },
     {
-      "epoch": 2.2497331910352187,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.00023255319243399737,
-      "loss": 0.5991,
+      "epoch": 2.1576253838280453,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00023771667583182498,
+      "loss": 0.5638,
       "step": 2108
     },
     {
-      "epoch": 2.254002134471718,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0002323106526571901,
-      "loss": 0.5852,
+      "epoch": 2.1617195496417603,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023749066610568968,
+      "loss": 0.5325,
       "step": 2112
     },
     {
-      "epoch": 2.258271077908218,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002320678045723558,
-      "loss": 0.5732,
+      "epoch": 2.1658137154554757,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023726435493739726,
+      "loss": 0.5708,
       "step": 2116
     },
     {
-      "epoch": 2.262540021344717,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00023182464908912177,
-      "loss": 0.5196,
+      "epoch": 2.169907881269191,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00023703774310668483,
+      "loss": 0.6038,
       "step": 2120
     },
     {
-      "epoch": 2.266808964781217,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00023158118711826662,
-      "loss": 0.599,
+      "epoch": 2.1740020470829067,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00023681083139432549,
+      "loss": 0.5861,
       "step": 2124
     },
     {
-      "epoch": 2.271077908217716,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0002313374195717171,
-      "loss": 0.5866,
+      "epoch": 2.178096212896622,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00023658362058212568,
+      "loss": 0.6176,
       "step": 2128
     },
     {
-      "epoch": 2.2753468516542155,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0002310933473625444,
-      "loss": 0.5709,
+      "epoch": 2.1821903787103376,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00023635611145292213,
+      "loss": 0.5727,
       "step": 2132
     },
     {
-      "epoch": 2.279615795090715,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00023084897140496102,
-      "loss": 0.6166,
+      "epoch": 2.186284544524053,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00023612830479057957,
+      "loss": 0.5516,
       "step": 2136
     },
     {
-      "epoch": 2.2838847385272145,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00023060429261431703,
-      "loss": 0.5986,
+      "epoch": 2.1903787103377685,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00023590020137998787,
+      "loss": 0.6297,
       "step": 2140
     },
     {
-      "epoch": 2.288153681963714,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.000230359311907097,
-      "loss": 0.5848,
+      "epoch": 2.194472876151484,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0002356718020070591,
+      "loss": 0.5994,
       "step": 2144
     },
     {
-      "epoch": 2.2924226254002136,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00023011403020091622,
-      "loss": 0.582,
+      "epoch": 2.1985670419651995,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00023544310745872532,
+      "loss": 0.591,
       "step": 2148
     },
     {
-      "epoch": 2.296691568836713,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022986844841451747,
-      "loss": 0.5702,
+      "epoch": 2.202661207778915,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00023521411852293545,
+      "loss": 0.6033,
       "step": 2152
     },
     {
-      "epoch": 2.300960512273212,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00022962256746776762,
-      "loss": 0.587,
+      "epoch": 2.2067553735926304,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023498483598865272,
+      "loss": 0.5993,
       "step": 2156
     },
     {
-      "epoch": 2.305229455709712,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00022937638828165397,
-      "loss": 0.5707,
+      "epoch": 2.210849539406346,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00023475526064585192,
+      "loss": 0.6182,
       "step": 2160
     },
     {
-      "epoch": 2.3094983991462112,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00022912991177828095,
-      "loss": 0.6133,
+      "epoch": 2.2149437052200613,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00023452539328551664,
+      "loss": 0.6129,
       "step": 2164
     },
     {
-      "epoch": 2.313767342582711,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00022888313888086672,
-      "loss": 0.5773,
+      "epoch": 2.219037871033777,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00023429523469963682,
+      "loss": 0.6295,
       "step": 2168
     },
     {
-      "epoch": 2.3180362860192103,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.0002286360705137395,
-      "loss": 0.602,
+      "epoch": 2.2231320368474923,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002340647856812055,
+      "loss": 0.5664,
       "step": 2172
     },
     {
-      "epoch": 2.3223052294557096,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00022838870760233436,
-      "loss": 0.5929,
+      "epoch": 2.2272262026612077,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002338340470242165,
+      "loss": 0.6029,
       "step": 2176
     },
     {
-      "epoch": 2.3265741728922094,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00022814105107318952,
-      "loss": 0.6213,
+      "epoch": 2.231320368474923,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00023360301952366166,
+      "loss": 0.6056,
       "step": 2180
     },
     {
-      "epoch": 2.3308431163287087,
+      "epoch": 2.2354145342886387,
       "grad_norm": 0.29296875,
-      "learning_rate": 0.00022789310185394315,
-      "loss": 0.5939,
+      "learning_rate": 0.0002333717039755279,
+      "loss": 0.577,
       "step": 2184
     },
     {
-      "epoch": 2.335112059765208,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00022764486087332967,
-      "loss": 0.5494,
+      "epoch": 2.239508700102354,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002331401011767946,
+      "loss": 0.6348,
       "step": 2188
     },
     {
-      "epoch": 2.3393810032017077,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0002273963290611762,
-      "loss": 0.5768,
+      "epoch": 2.2436028659160696,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00023290821192543083,
+      "loss": 0.6237,
       "step": 2192
     },
     {
-      "epoch": 2.343649946638207,
+      "epoch": 2.247697031729785,
       "grad_norm": 0.28125,
-      "learning_rate": 0.0002271475073483994,
-      "loss": 0.6515,
+      "learning_rate": 0.00023267603702039263,
+      "loss": 0.5623,
       "step": 2196
     },
     {
-      "epoch": 2.3479188900747063,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002268983966670018,
-      "loss": 0.6567,
+      "epoch": 2.2517911975435005,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002324435772616203,
+      "loss": 0.5739,
       "step": 2200
     },
     {
-      "epoch": 2.352187833511206,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00022664899795006818,
-      "loss": 0.608,
+      "epoch": 2.255885363357216,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00023221083345003544,
+      "loss": 0.5992,
       "step": 2204
     },
     {
-      "epoch": 2.3564567769477054,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00022639931213176226,
-      "loss": 0.5777,
+      "epoch": 2.2599795291709315,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00023197780638753847,
+      "loss": 0.5831,
       "step": 2208
     },
     {
-      "epoch": 2.360725720384205,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00022614934014732316,
-      "loss": 0.5906,
+      "epoch": 2.264073694984647,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023174449687700564,
+      "loss": 0.6073,
       "step": 2212
     },
     {
-      "epoch": 2.3649946638207044,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00022589908293306187,
-      "loss": 0.5678,
+      "epoch": 2.2681678607983624,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00023151090572228635,
+      "loss": 0.563,
       "step": 2216
     },
     {
-      "epoch": 2.3692636072572038,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0002256485414263578,
-      "loss": 0.564,
+      "epoch": 2.272262026612078,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002312770337282005,
+      "loss": 0.5428,
       "step": 2220
     },
     {
-      "epoch": 2.3735325506937035,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00022539771656565518,
-      "loss": 0.5967,
+      "epoch": 2.2763561924257933,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00023104288170053543,
+      "loss": 0.5443,
       "step": 2224
     },
     {
-      "epoch": 2.377801494130203,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022514660929045963,
-      "loss": 0.5272,
+      "epoch": 2.280450358239509,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002308084504460435,
+      "loss": 0.5721,
       "step": 2228
     },
     {
-      "epoch": 2.382070437566702,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00022489522054133448,
-      "loss": 0.5527,
+      "epoch": 2.2845445240532243,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00023057374077243884,
+      "loss": 0.5796,
       "step": 2232
     },
     {
-      "epoch": 2.386339381003202,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.0002246435512598975,
-      "loss": 0.5927,
+      "epoch": 2.2886386898669397,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00023033875348839526,
+      "loss": 0.6163,
       "step": 2236
     },
     {
-      "epoch": 2.390608324439701,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002243916023888172,
-      "loss": 0.6094,
+      "epoch": 2.292732855680655,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00023010348940354262,
+      "loss": 0.5588,
       "step": 2240
     },
     {
-      "epoch": 2.3948772678762005,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00022413937487180928,
-      "loss": 0.5965,
+      "epoch": 2.2968270214943707,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002298679493284648,
+      "loss": 0.6182,
       "step": 2244
     },
     {
-      "epoch": 2.3991462113127002,
+      "epoch": 2.300921187308086,
       "grad_norm": 0.28515625,
-      "learning_rate": 0.00022388686965363328,
-      "loss": 0.5922,
+      "learning_rate": 0.00022963213407469643,
+      "loss": 0.6273,
       "step": 2248
     },
     {
-      "epoch": 2.4034151547491995,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00022363408768008878,
-      "loss": 0.6087,
+      "epoch": 2.3050153531218016,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00022939604445472027,
+      "loss": 0.5913,
       "step": 2252
     },
     {
-      "epoch": 2.4076840981856993,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.0002233810298980121,
-      "loss": 0.6428,
+      "epoch": 2.309109518935517,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022915968128196443,
+      "loss": 0.605,
       "step": 2256
     },
     {
-      "epoch": 2.4119530416221986,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00022312769725527256,
-      "loss": 0.6142,
+      "epoch": 2.313203684749232,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00022892304537079945,
+      "loss": 0.5575,
       "step": 2260
     },
     {
-      "epoch": 2.416221985058698,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00022287409070076905,
-      "loss": 0.6185,
+      "epoch": 2.317297850562948,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00022868613753653565,
+      "loss": 0.6122,
       "step": 2264
     },
     {
-      "epoch": 2.420490928495197,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00022262021118442645,
-      "loss": 0.58,
+      "epoch": 2.321392016376663,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00022844895859542016,
+      "loss": 0.644,
       "step": 2268
     },
     {
-      "epoch": 2.424759871931697,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002223660596571921,
-      "loss": 0.5625,
+      "epoch": 2.325486182190379,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00022821150936463427,
+      "loss": 0.6069,
       "step": 2272
     },
     {
-      "epoch": 2.4290288153681963,
-      "grad_norm": 0.26171875,
-      "learning_rate": 0.00022211163707103215,
-      "loss": 0.5471,
+      "epoch": 2.329580348004094,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002279737906622905,
+      "loss": 0.5912,
       "step": 2276
     },
     {
-      "epoch": 2.433297758804696,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.000221856944378928,
-      "loss": 0.5963,
+      "epoch": 2.3336745138178094,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00022773580330742973,
+      "loss": 0.5825,
       "step": 2280
     },
     {
-      "epoch": 2.4375667022411953,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00022160198253487286,
-      "loss": 0.5698,
+      "epoch": 2.337768679631525,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00022749754812001856,
+      "loss": 0.6167,
       "step": 2284
     },
     {
-      "epoch": 2.4418356456776946,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00022134675249386808,
-      "loss": 0.5675,
+      "epoch": 2.3418628454452404,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002272590259209464,
+      "loss": 0.5872,
       "step": 2288
     },
     {
-      "epoch": 2.4461045891141944,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00022109125521191955,
-      "loss": 0.5796,
+      "epoch": 2.345957011258956,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022702023753202257,
+      "loss": 0.5571,
       "step": 2292
     },
     {
-      "epoch": 2.4503735325506937,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00022083549164603417,
-      "loss": 0.608,
+      "epoch": 2.3500511770726713,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002267811837759735,
+      "loss": 0.5777,
       "step": 2296
     },
     {
-      "epoch": 2.454642475987193,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0002205794627542163,
-      "loss": 0.6231,
+      "epoch": 2.3541453428863868,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022654186547644003,
+      "loss": 0.5692,
       "step": 2300
     },
     {
-      "epoch": 2.4589114194236927,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00022032316949546405,
-      "loss": 0.6001,
+      "epoch": 2.3582395087001022,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022630228345797435,
+      "loss": 0.5789,
       "step": 2304
     },
     {
-      "epoch": 2.463180362860192,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00022006661282976576,
-      "loss": 0.5945,
+      "epoch": 2.3623336745138177,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00022606243854603729,
+      "loss": 0.6044,
       "step": 2308
     },
     {
-      "epoch": 2.4674493062966913,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0002198097937180965,
-      "loss": 0.5861,
+      "epoch": 2.366427840327533,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002258223315669956,
+      "loss": 0.6143,
       "step": 2312
     },
     {
-      "epoch": 2.471718249733191,
+      "epoch": 2.3705220061412486,
       "grad_norm": 0.28125,
-      "learning_rate": 0.0002195527131224142,
-      "loss": 0.592,
+      "learning_rate": 0.00022558196334811873,
+      "loss": 0.5742,
       "step": 2316
     },
     {
-      "epoch": 2.4759871931696904,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00021929537200565642,
-      "loss": 0.5731,
+      "epoch": 2.374616171954964,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0002253413347175764,
+      "loss": 0.6163,
       "step": 2320
     },
     {
-      "epoch": 2.48025613660619,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00021903777133173635,
-      "loss": 0.6447,
+      "epoch": 2.3787103377686796,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00022510044650443547,
+      "loss": 0.612,
       "step": 2324
     },
     {
-      "epoch": 2.4845250800426895,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00021877991206553963,
-      "loss": 0.5882,
+      "epoch": 2.382804503582395,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00022485929953865714,
+      "loss": 0.5886,
       "step": 2328
     },
     {
-      "epoch": 2.4887940234791888,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021852179517292027,
-      "loss": 0.5735,
+      "epoch": 2.3868986693961105,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00022461789465109426,
+      "loss": 0.5864,
       "step": 2332
     },
     {
-      "epoch": 2.4930629669156885,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00021826342162069737,
-      "loss": 0.5654,
+      "epoch": 2.390992835209826,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00022437623267348823,
+      "loss": 0.6113,
       "step": 2336
     },
     {
-      "epoch": 2.497331910352188,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00021800479237665135,
-      "loss": 0.6594,
+      "epoch": 2.3950870010235414,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00022413431443846617,
+      "loss": 0.6025,
       "step": 2340
     },
     {
-      "epoch": 2.5016008537886876,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00021774590840952044,
-      "loss": 0.5994,
+      "epoch": 2.399181166837257,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022389214077953823,
+      "loss": 0.6034,
       "step": 2344
     },
     {
-      "epoch": 2.505869797225187,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002174867706889969,
-      "loss": 0.611,
+      "epoch": 2.4032753326509724,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022364971253109462,
+      "loss": 0.666,
       "step": 2348
     },
     {
-      "epoch": 2.510138740661686,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00021722738018572352,
-      "loss": 0.6217,
+      "epoch": 2.407369498464688,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00022340703052840257,
+      "loss": 0.5732,
       "step": 2352
     },
     {
-      "epoch": 2.5144076840981855,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021696773787128988,
-      "loss": 0.6403,
+      "epoch": 2.4114636642784033,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0002231640956076037,
+      "loss": 0.6258,
       "step": 2356
     },
     {
-      "epoch": 2.5186766275346852,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021670784471822878,
-      "loss": 0.6053,
+      "epoch": 2.4155578300921188,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00022292090860571108,
+      "loss": 0.5836,
       "step": 2360
     },
     {
-      "epoch": 2.5229455709711845,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00021644770170001256,
-      "loss": 0.6154,
+      "epoch": 2.4196519959058342,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00022267747036060627,
+      "loss": 0.5722,
       "step": 2364
     },
     {
-      "epoch": 2.5272145144076843,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00021618730979104954,
-      "loss": 0.5762,
+      "epoch": 2.4237461617195497,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00022243378171103636,
+      "loss": 0.5837,
       "step": 2368
     },
     {
-      "epoch": 2.5314834578441836,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00021592666996668022,
-      "loss": 0.5574,
+      "epoch": 2.427840327533265,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00022218984349661134,
+      "loss": 0.6311,
       "step": 2372
     },
     {
-      "epoch": 2.535752401280683,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00021566578320317376,
-      "loss": 0.611,
+      "epoch": 2.4319344933469806,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00022194565655780102,
+      "loss": 0.6041,
       "step": 2376
     },
     {
-      "epoch": 2.5400213447171827,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00021540465047772422,
-      "loss": 0.6169,
+      "epoch": 2.436028659160696,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00022170122173593206,
+      "loss": 0.5853,
       "step": 2380
     },
     {
-      "epoch": 2.544290288153682,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.000215143272768447,
-      "loss": 0.522,
+      "epoch": 2.4401228249744116,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00022145653987318536,
+      "loss": 0.6081,
       "step": 2384
     },
     {
-      "epoch": 2.5485592315901813,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021488165105437514,
-      "loss": 0.5935,
+      "epoch": 2.444216990788127,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0002212116118125929,
+      "loss": 0.5846,
       "step": 2388
     },
     {
-      "epoch": 2.552828175026681,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.00021461978631545558,
-      "loss": 0.6303,
+      "epoch": 2.4483111566018425,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00022096643839803482,
+      "loss": 0.5953,
       "step": 2392
     },
     {
-      "epoch": 2.5570971184631803,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0002143576795325455,
-      "loss": 0.6188,
+      "epoch": 2.452405322415558,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00022072102047423673,
+      "loss": 0.554,
       "step": 2396
     },
     {
-      "epoch": 2.5613660618996796,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00021409533168740888,
-      "loss": 0.5681,
+      "epoch": 2.4564994882292734,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002204753588867667,
+      "loss": 0.6111,
       "step": 2400
     },
     {
-      "epoch": 2.5656350053361794,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0002138327437627125,
-      "loss": 0.5694,
+      "epoch": 2.460593654042989,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002202294544820321,
+      "loss": 0.6074,
       "step": 2404
     },
     {
-      "epoch": 2.5699039487726787,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00021356991674202245,
-      "loss": 0.5456,
+      "epoch": 2.4646878198567044,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00021998330810727725,
+      "loss": 0.6092,
       "step": 2408
     },
     {
-      "epoch": 2.5741728922091784,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00021330685160980033,
-      "loss": 0.6229,
+      "epoch": 2.46878198567042,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00021973692061057987,
+      "loss": 0.5686,
       "step": 2412
     },
     {
-      "epoch": 2.5784418356456777,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00021304354935139966,
-      "loss": 0.6189,
+      "epoch": 2.472876151484135,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0002194902928408486,
+      "loss": 0.6034,
       "step": 2416
     },
     {
-      "epoch": 2.582710779082177,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0002127800109530622,
-      "loss": 0.5721,
+      "epoch": 2.4769703172978508,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002192434256478199,
+      "loss": 0.5837,
       "step": 2420
     },
     {
-      "epoch": 2.5869797225186764,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00021251623740191412,
-      "loss": 0.609,
+      "epoch": 2.481064483111566,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00021899631988205506,
+      "loss": 0.5938,
       "step": 2424
     },
     {
-      "epoch": 2.591248665955176,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002122522296859625,
-      "loss": 0.6232,
+      "epoch": 2.4851586489252817,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00021874897639493745,
+      "loss": 0.5849,
       "step": 2428
     },
     {
-      "epoch": 2.5955176093916754,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021198798879409146,
-      "loss": 0.6303,
+      "epoch": 2.4892528147389967,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00021850139603866946,
+      "loss": 0.6057,
       "step": 2432
     },
     {
-      "epoch": 2.599786552828175,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00021172351571605851,
-      "loss": 0.6504,
+      "epoch": 2.493346980552712,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0002182535796662696,
+      "loss": 0.555,
       "step": 2436
     },
     {
-      "epoch": 2.6040554962646745,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0002114588114424909,
-      "loss": 0.5888,
+      "epoch": 2.4974411463664277,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021800552813156947,
+      "loss": 0.5598,
       "step": 2440
     },
     {
-      "epoch": 2.608324439701174,
-      "grad_norm": 0.265625,
-      "learning_rate": 0.00021119387696488174,
-      "loss": 0.5791,
+      "epoch": 2.501535312180143,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00021775724228921108,
+      "loss": 0.6042,
       "step": 2444
     },
     {
-      "epoch": 2.6125933831376735,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021092871327558658,
-      "loss": 0.608,
+      "epoch": 2.5056294779938586,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00021750872299464358,
+      "loss": 0.5955,
       "step": 2448
     },
     {
-      "epoch": 2.616862326574173,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00021066332136781947,
-      "loss": 0.636,
+      "epoch": 2.509723643807574,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00021725997110412043,
+      "loss": 0.5756,
       "step": 2452
     },
     {
-      "epoch": 2.6211312700106726,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00021039770223564913,
-      "loss": 0.5648,
+      "epoch": 2.5138178096212895,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002170109874746967,
+      "loss": 0.5865,
       "step": 2456
     },
     {
-      "epoch": 2.625400213447172,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00021013185687399565,
-      "loss": 0.5866,
+      "epoch": 2.517911975435005,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021676177296422566,
+      "loss": 0.6014,
       "step": 2460
     },
     {
-      "epoch": 2.629669156883671,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00020986578627862627,
-      "loss": 0.6756,
+      "epoch": 2.5220061412487205,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00021651232843135617,
+      "loss": 0.5804,
       "step": 2464
     },
     {
-      "epoch": 2.6339381003201705,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00020959949144615199,
-      "loss": 0.6176,
+      "epoch": 2.526100307062436,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00021626265473552965,
+      "loss": 0.6267,
       "step": 2468
     },
     {
-      "epoch": 2.6382070437566703,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0002093329733740237,
-      "loss": 0.5987,
+      "epoch": 2.5301944728761514,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00021601275273697696,
+      "loss": 0.5653,
       "step": 2472
     },
     {
-      "epoch": 2.6424759871931696,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00020906623306052852,
-      "loss": 0.5734,
+      "epoch": 2.534288638689867,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00021576262329671568,
+      "loss": 0.5678,
       "step": 2476
     },
     {
-      "epoch": 2.6467449306296693,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00020879927150478588,
-      "loss": 0.6388,
+      "epoch": 2.5383828045035823,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00021551226727654696,
+      "loss": 0.5694,
       "step": 2480
     },
     {
-      "epoch": 2.6510138740661686,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0002085320897067441,
-      "loss": 0.6202,
+      "epoch": 2.542476970317298,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021526168553905265,
+      "loss": 0.5915,
       "step": 2484
     },
     {
-      "epoch": 2.655282817502668,
-      "grad_norm": 0.279296875,
-      "learning_rate": 0.00020826468866717637,
-      "loss": 0.6261,
+      "epoch": 2.5465711361310133,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00021501087894759227,
+      "loss": 0.5917,
       "step": 2488
     },
     {
-      "epoch": 2.6595517609391677,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00020799706938767697,
-      "loss": 0.5854,
+      "epoch": 2.5506653019447287,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021475984836629998,
+      "loss": 0.583,
       "step": 2492
     },
     {
-      "epoch": 2.663820704375667,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00020772923287065776,
-      "loss": 0.6162,
+      "epoch": 2.554759467758444,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002145085946600819,
+      "loss": 0.6088,
       "step": 2496
     },
     {
-      "epoch": 2.6680896478121667,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00020746118011934428,
-      "loss": 0.6165,
+      "epoch": 2.5588536335721597,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00021425711869461266,
+      "loss": 0.5994,
       "step": 2500
     },
     {
-      "epoch": 2.672358591248666,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00020719291213777208,
-      "loss": 0.643,
+      "epoch": 2.562947799385875,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021400542133633276,
+      "loss": 0.5656,
       "step": 2504
     },
     {
-      "epoch": 2.6766275346851653,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00020692442993078264,
-      "loss": 0.6155,
+      "epoch": 2.5670419651995906,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021375350345244557,
+      "loss": 0.6544,
       "step": 2508
     },
     {
-      "epoch": 2.6808964781216647,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020665573450402017,
-      "loss": 0.6108,
+      "epoch": 2.571136131013306,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00021350136591091415,
+      "loss": 0.5995,
       "step": 2512
     },
     {
-      "epoch": 2.6851654215581644,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00020638682686392734,
-      "loss": 0.6177,
+      "epoch": 2.5752302968270215,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021324900958045843,
+      "loss": 0.6408,
       "step": 2516
     },
     {
-      "epoch": 2.6894343649946637,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00020611770801774168,
-      "loss": 0.5889,
+      "epoch": 2.579324462640737,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00021299643533055214,
+      "loss": 0.62,
       "step": 2520
     },
     {
-      "epoch": 2.6937033084311635,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00020584837897349203,
-      "loss": 0.5726,
+      "epoch": 2.5834186284544525,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0002127436440314199,
+      "loss": 0.5754,
       "step": 2524
     },
     {
-      "epoch": 2.6979722518676628,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00020557884073999432,
-      "loss": 0.5945,
+      "epoch": 2.587512794268168,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000212490636554034,
+      "loss": 0.5731,
       "step": 2528
     },
     {
-      "epoch": 2.702241195304162,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0002053090943268482,
-      "loss": 0.6184,
+      "epoch": 2.5916069600818834,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00021223741377011178,
+      "loss": 0.582,
       "step": 2532
     },
     {
-      "epoch": 2.706510138740662,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00020503914074443302,
-      "loss": 0.6313,
+      "epoch": 2.595701125895599,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00021198397655211216,
+      "loss": 0.5946,
       "step": 2536
     },
     {
-      "epoch": 2.710779082177161,
-      "grad_norm": 0.2734375,
-      "learning_rate": 0.0002047689810039041,
-      "loss": 0.5739,
+      "epoch": 2.5997952917093143,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00021173032577323302,
+      "loss": 0.6236,
       "step": 2540
     },
     {
-      "epoch": 2.715048025613661,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00020449861611718896,
-      "loss": 0.5894,
+      "epoch": 2.60388945752303,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021147646230740814,
+      "loss": 0.6059,
       "step": 2544
     },
     {
-      "epoch": 2.71931696905016,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00020422804709698358,
-      "loss": 0.6277,
+      "epoch": 2.6079836233367453,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00021122238702930377,
+      "loss": 0.5637,
       "step": 2548
     },
     {
-      "epoch": 2.7235859124866595,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00020395727495674856,
-      "loss": 0.646,
+      "epoch": 2.6120777891504607,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00021096810081431628,
+      "loss": 0.5897,
       "step": 2552
     },
     {
-      "epoch": 2.727854855923159,
+      "epoch": 2.616171954964176,
       "grad_norm": 0.3046875,
-      "learning_rate": 0.0002036863007107052,
-      "loss": 0.6199,
+      "learning_rate": 0.00021071360453856866,
+      "loss": 0.5732,
       "step": 2556
     },
     {
-      "epoch": 2.7321237993596585,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.000203415125373832,
-      "loss": 0.5702,
+      "epoch": 2.6202661207778917,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00021045889907890763,
+      "loss": 0.6022,
       "step": 2560
     },
     {
-      "epoch": 2.736392742796158,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0002031437499618606,
-      "loss": 0.5644,
+      "epoch": 2.6243602865916067,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00021020398531290067,
+      "loss": 0.6029,
       "step": 2564
     },
     {
-      "epoch": 2.7406616862326576,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00020287217549127195,
-      "loss": 0.55,
+      "epoch": 2.6284544524053226,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00020994886411883297,
+      "loss": 0.6171,
       "step": 2568
     },
     {
-      "epoch": 2.744930629669157,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00020260040297929272,
-      "loss": 0.6404,
+      "epoch": 2.6325486182190376,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00020969353637570443,
+      "loss": 0.5882,
       "step": 2572
     },
     {
-      "epoch": 2.749199573105656,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00020232843344389145,
-      "loss": 0.6179,
+      "epoch": 2.6366427840327535,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002094380029632265,
+      "loss": 0.6273,
       "step": 2576
     },
     {
-      "epoch": 2.753468516542156,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0002020562679037744,
-      "loss": 0.5967,
+      "epoch": 2.6407369498464686,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00020918226476181935,
+      "loss": 0.5672,
       "step": 2580
     },
     {
-      "epoch": 2.7577374599786553,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.0002017839073783823,
-      "loss": 0.596,
+      "epoch": 2.6448311156601845,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020892632265260866,
+      "loss": 0.6087,
       "step": 2584
     },
     {
-      "epoch": 2.762006403415155,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00020151135288788607,
-      "loss": 0.6351,
+      "epoch": 2.6489252814738995,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00020867017751742266,
+      "loss": 0.5993,
       "step": 2588
     },
     {
-      "epoch": 2.7662753468516543,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0002012386054531831,
-      "loss": 0.611,
+      "epoch": 2.6530194472876154,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00020841383023878916,
+      "loss": 0.5903,
       "step": 2592
     },
     {
-      "epoch": 2.7705442902881536,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00020096566609589364,
-      "loss": 0.5849,
+      "epoch": 2.6571136131013304,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00020815728169993233,
+      "loss": 0.5986,
       "step": 2596
     },
     {
-      "epoch": 2.774813233724653,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00020069253583835677,
-      "loss": 0.5559,
+      "epoch": 2.661207778915046,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002079005327847699,
+      "loss": 0.6083,
       "step": 2600
     },
     {
-      "epoch": 2.7790821771611527,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0002004192157036265,
-      "loss": 0.6028,
+      "epoch": 2.6653019447287614,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00020764358437790994,
+      "loss": 0.5921,
       "step": 2604
     },
     {
-      "epoch": 2.783351120597652,
-      "grad_norm": 0.28515625,
-      "learning_rate": 0.00020014570671546828,
-      "loss": 0.6117,
+      "epoch": 2.669396110542477,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00020738643736464772,
+      "loss": 0.5782,
       "step": 2608
     },
     {
-      "epoch": 2.7876200640341517,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00019987200989835468,
-      "loss": 0.6328,
+      "epoch": 2.6734902763561923,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00020712909263096297,
+      "loss": 0.6629,
       "step": 2612
     },
     {
-      "epoch": 2.791889007470651,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00019959812627746198,
-      "loss": 0.5806,
+      "epoch": 2.6775844421699078,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00020687155106351661,
+      "loss": 0.6093,
       "step": 2616
     },
     {
-      "epoch": 2.7961579509071504,
+      "epoch": 2.6816786079836232,
       "grad_norm": 0.30859375,
-      "learning_rate": 0.00019932405687866616,
-      "loss": 0.6413,
+      "learning_rate": 0.00020661381354964762,
+      "loss": 0.5907,
       "step": 2620
     },
     {
-      "epoch": 2.80042689434365,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00019904980272853902,
-      "loss": 0.5669,
+      "epoch": 2.6857727737973387,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020635588097737015,
+      "loss": 0.5855,
       "step": 2624
     },
     {
-      "epoch": 2.8046958377801494,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00019877536485434435,
-      "loss": 0.5743,
+      "epoch": 2.689866939611054,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00020609775423537053,
+      "loss": 0.6578,
       "step": 2628
     },
     {
-      "epoch": 2.8089647812166487,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00019850074428403417,
-      "loss": 0.5599,
+      "epoch": 2.6939611054247696,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00020583943421300405,
+      "loss": 0.5752,
       "step": 2632
     },
     {
-      "epoch": 2.8132337246531485,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00019822594204624478,
-      "loss": 0.5746,
+      "epoch": 2.698055271238485,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002055809218002917,
+      "loss": 0.6257,
       "step": 2636
     },
     {
-      "epoch": 2.8175026680896478,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.000197950959170293,
-      "loss": 0.5978,
+      "epoch": 2.7021494370522006,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00020532221788791767,
+      "loss": 0.6225,
       "step": 2640
     },
     {
-      "epoch": 2.821771611526147,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00019767579668617219,
-      "loss": 0.5769,
+      "epoch": 2.706243602865916,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00020506332336722572,
+      "loss": 0.5765,
       "step": 2644
     },
     {
-      "epoch": 2.826040554962647,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00019740045562454848,
-      "loss": 0.6237,
+      "epoch": 2.7103377686796315,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00020480423913021636,
+      "loss": 0.6118,
       "step": 2648
     },
     {
-      "epoch": 2.830309498399146,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001971249370167569,
-      "loss": 0.6006,
+      "epoch": 2.714431934493347,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002045449660695439,
+      "loss": 0.5911,
       "step": 2652
     },
     {
-      "epoch": 2.834578441835646,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001968492418947975,
-      "loss": 0.5673,
+      "epoch": 2.7185261003070624,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00020428550507851313,
+      "loss": 0.5962,
       "step": 2656
     },
     {
-      "epoch": 2.838847385272145,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00019657337129133155,
-      "loss": 0.5596,
+      "epoch": 2.722620266120778,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00020402585705107617,
+      "loss": 0.5941,
       "step": 2660
     },
     {
-      "epoch": 2.8431163287086445,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.00019629732623967753,
-      "loss": 0.5872,
+      "epoch": 2.7267144319344934,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00020376602288182992,
+      "loss": 0.6027,
       "step": 2664
     },
     {
-      "epoch": 2.847385272145144,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00019602110777380737,
-      "loss": 0.6229,
+      "epoch": 2.730808597748209,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002035060034660123,
+      "loss": 0.5518,
       "step": 2668
     },
     {
-      "epoch": 2.8516542155816436,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00019574471692834257,
-      "loss": 0.6037,
+      "epoch": 2.7349027635619243,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00020324579969949964,
+      "loss": 0.6127,
       "step": 2672
     },
     {
-      "epoch": 2.855923159018143,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001954681547385503,
-      "loss": 0.6017,
+      "epoch": 2.7389969293756398,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020298541247880343,
+      "loss": 0.6011,
       "step": 2676
     },
     {
-      "epoch": 2.8601921024546426,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00019519142224033956,
-      "loss": 0.6481,
+      "epoch": 2.7430910951893552,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00020272484270106712,
+      "loss": 0.5692,
       "step": 2680
     },
     {
-      "epoch": 2.864461045891142,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00019491452047025714,
-      "loss": 0.6017,
+      "epoch": 2.7471852610030707,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0002024640912640633,
+      "loss": 0.6303,
       "step": 2684
     },
     {
-      "epoch": 2.8687299893276412,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.0001946374504654841,
-      "loss": 0.5909,
+      "epoch": 2.751279426816786,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002022031590661904,
+      "loss": 0.6613,
       "step": 2688
     },
     {
-      "epoch": 2.872998932764141,
-      "grad_norm": 0.267578125,
-      "learning_rate": 0.00019436021326383137,
-      "loss": 0.6528,
+      "epoch": 2.7553735926305016,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00020194204700646958,
+      "loss": 0.6369,
       "step": 2692
     },
     {
-      "epoch": 2.8772678762006403,
+      "epoch": 2.759467758444217,
       "grad_norm": 0.30078125,
-      "learning_rate": 0.0001940828099037364,
-      "loss": 0.6313,
+      "learning_rate": 0.0002016807559845418,
+      "loss": 0.6217,
       "step": 2696
     },
     {
-      "epoch": 2.88153681963714,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0001938052414242588,
-      "loss": 0.5678,
+      "epoch": 2.7635619242579326,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00020141928690066446,
+      "loss": 0.651,
       "step": 2700
     },
     {
-      "epoch": 2.8858057630736393,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00019352750886507684,
-      "loss": 0.5905,
+      "epoch": 2.767656090071648,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002011576406557087,
+      "loss": 0.5942,
       "step": 2704
     },
     {
-      "epoch": 2.8900747065101386,
-      "grad_norm": 0.27734375,
-      "learning_rate": 0.00019324961326648322,
-      "loss": 0.6025,
+      "epoch": 2.7717502558853635,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0002008958181511559,
+      "loss": 0.6106,
       "step": 2708
     },
     {
-      "epoch": 2.894343649946638,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00019297155566938142,
-      "loss": 0.5867,
+      "epoch": 2.775844421699079,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00020063382028909468,
+      "loss": 0.5713,
       "step": 2712
     },
     {
-      "epoch": 2.8986125933831377,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00019269333711528178,
-      "loss": 0.6047,
+      "epoch": 2.7799385875127944,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00020037164797221798,
+      "loss": 0.6318,
       "step": 2716
     },
     {
-      "epoch": 2.902881536819637,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00019241495864629737,
-      "loss": 0.6194,
+      "epoch": 2.7840327533265095,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00020010930210381973,
+      "loss": 0.5837,
       "step": 2720
     },
     {
-      "epoch": 2.9071504802561368,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00019213642130514036,
-      "loss": 0.6012,
+      "epoch": 2.7881269191402254,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019984678358779182,
+      "loss": 0.5899,
       "step": 2724
     },
     {
-      "epoch": 2.911419423692636,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.000191857726135118,
-      "loss": 0.5665,
+      "epoch": 2.7922210849539404,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000199584093328621,
+      "loss": 0.5779,
       "step": 2728
     },
     {
-      "epoch": 2.9156883671291354,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00019157887418012857,
-      "loss": 0.6171,
+      "epoch": 2.7963152507676563,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019932123223138573,
+      "loss": 0.6003,
       "step": 2732
     },
     {
-      "epoch": 2.919957310565635,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00019129986648465784,
-      "loss": 0.5935,
+      "epoch": 2.8004094165813713,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001990582012017531,
+      "loss": 0.5879,
       "step": 2736
     },
     {
-      "epoch": 2.9242262540021344,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00019102070409377474,
-      "loss": 0.6052,
+      "epoch": 2.8045035823950872,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019879500114597569,
+      "loss": 0.5881,
       "step": 2740
     },
     {
-      "epoch": 2.928495197438634,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001907413880531278,
-      "loss": 0.5968,
+      "epoch": 2.8085977482088023,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019853163297088843,
+      "loss": 0.5774,
       "step": 2744
     },
     {
-      "epoch": 2.9327641408751335,
-      "grad_norm": 0.283203125,
-      "learning_rate": 0.0001904619194089409,
-      "loss": 0.6177,
+      "epoch": 2.812691914022518,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019826809758390548,
+      "loss": 0.6252,
       "step": 2748
     },
     {
-      "epoch": 2.937033084311633,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00019018229920800966,
-      "loss": 0.6111,
+      "epoch": 2.816786079836233,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019800439589301715,
+      "loss": 0.65,
       "step": 2752
     },
     {
-      "epoch": 2.941302027748132,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00018990252849769733,
-      "loss": 0.6115,
+      "epoch": 2.8208802456499487,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019774052880678676,
+      "loss": 0.6235,
       "step": 2756
     },
     {
-      "epoch": 2.945570971184632,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00018962260832593086,
-      "loss": 0.5695,
+      "epoch": 2.824974411463664,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019747649723434732,
+      "loss": 0.5734,
       "step": 2760
     },
     {
-      "epoch": 2.949839914621131,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018934253974119716,
-      "loss": 0.5689,
+      "epoch": 2.8290685772773796,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019721230208539882,
+      "loss": 0.6161,
       "step": 2764
     },
     {
-      "epoch": 2.954108858057631,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0001890623237925389,
-      "loss": 0.6078,
+      "epoch": 2.833162743091095,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019694794427020461,
+      "loss": 0.5959,
       "step": 2768
     },
     {
-      "epoch": 2.95837780149413,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00018878196152955087,
-      "loss": 0.64,
+      "epoch": 2.8372569089048105,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001966834246995887,
+      "loss": 0.5986,
       "step": 2772
     },
     {
-      "epoch": 2.9626467449306295,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018850145400237578,
-      "loss": 0.6209,
+      "epoch": 2.841351074718526,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019641874428493223,
+      "loss": 0.612,
       "step": 2776
     },
     {
-      "epoch": 2.9669156883671293,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.0001882208022617005,
-      "loss": 0.5918,
+      "epoch": 2.8454452405322415,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019615390393817067,
+      "loss": 0.5609,
       "step": 2780
     },
     {
-      "epoch": 2.9711846318036286,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.00018794000735875208,
-      "loss": 0.6031,
+      "epoch": 2.849539406345957,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019588890457179035,
+      "loss": 0.6219,
       "step": 2784
     },
     {
-      "epoch": 2.9754535752401283,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.0001876590703452939,
-      "loss": 0.6224,
+      "epoch": 2.8536335721596724,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019562374709882564,
+      "loss": 0.6171,
       "step": 2788
     },
     {
-      "epoch": 2.9797225186766276,
-      "grad_norm": 0.287109375,
-      "learning_rate": 0.0001873779922736214,
-      "loss": 0.5607,
+      "epoch": 2.857727737973388,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019535843243285566,
+      "loss": 0.6393,
       "step": 2792
     },
     {
-      "epoch": 2.983991462113127,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0001870967741965586,
-      "loss": 0.6302,
+      "epoch": 2.8618219037871033,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00019509296148800093,
+      "loss": 0.5619,
       "step": 2796
     },
     {
-      "epoch": 2.9882604055496262,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00018681541716745388,
-      "loss": 0.6344,
+      "epoch": 2.865916069600819,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001948273351789207,
+      "loss": 0.6155,
       "step": 2800
     },
     {
-      "epoch": 2.992529348986126,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.000186533922240176,
-      "loss": 0.5858,
+      "epoch": 2.8700102354145343,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019456155442080928,
+      "loss": 0.6029,
       "step": 2804
     },
     {
-      "epoch": 2.9967982924226253,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00018625229046911033,
-      "loss": 0.5743,
+      "epoch": 2.8741044012282497,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019429562012939316,
+      "loss": 0.593,
       "step": 2808
     },
     {
-      "epoch": 3.001067235859125,
-      "grad_norm": 0.275390625,
-      "learning_rate": 0.0001859705229091548,
-      "loss": 0.6399,
+      "epoch": 2.878198567041965,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019402953322092805,
+      "loss": 0.6088,
       "step": 2812
     },
     {
-      "epoch": 3.0053361792956244,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.0001856886206157159,
-      "loss": 0.4803,
+      "epoch": 2.8822927328556807,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019376329461219516,
+      "loss": 0.64,
       "step": 2816
     },
     {
-      "epoch": 3.0096051227321237,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018540658464470485,
-      "loss": 0.539,
+      "epoch": 2.886386898669396,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019349690522049853,
+      "loss": 0.6028,
       "step": 2820
     },
     {
-      "epoch": 3.0138740661686234,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00018512441605253357,
-      "loss": 0.502,
+      "epoch": 2.8904810644831116,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019323036596366174,
+      "loss": 0.6205,
       "step": 2824
     },
     {
-      "epoch": 3.0181430096051227,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001848421158961107,
-      "loss": 0.5283,
+      "epoch": 2.894575230296827,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019296367776002466,
+      "loss": 0.6108,
       "step": 2828
     },
     {
-      "epoch": 3.022411953041622,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00018455968523283775,
-      "loss": 0.4959,
+      "epoch": 2.8986693961105425,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019269684152844037,
+      "loss": 0.63,
       "step": 2832
     },
     {
-      "epoch": 3.0266808964781218,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018427712512060493,
-      "loss": 0.5063,
+      "epoch": 2.902763561924258,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019242985818827198,
+      "loss": 0.6686,
       "step": 2836
     },
     {
-      "epoch": 3.030949839914621,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00018399443661778748,
-      "loss": 0.5275,
+      "epoch": 2.9068577277379735,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001921627286593894,
+      "loss": 0.5533,
       "step": 2840
     },
     {
-      "epoch": 3.0352187833511204,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00018371162078324147,
-      "loss": 0.4883,
+      "epoch": 2.910951893551689,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019189545386216625,
+      "loss": 0.5544,
       "step": 2844
     },
     {
-      "epoch": 3.03948772678762,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00018342867867629986,
-      "loss": 0.4861,
+      "epoch": 2.9150460593654044,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019162803471747667,
+      "loss": 0.6254,
       "step": 2848
     },
     {
-      "epoch": 3.0437566702241194,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0001831456113567687,
-      "loss": 0.4827,
+      "epoch": 2.91914022517912,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001913604721466922,
+      "loss": 0.5852,
       "step": 2852
     },
     {
-      "epoch": 3.048025613660619,
-      "grad_norm": 0.28125,
-      "learning_rate": 0.00018286241988492292,
-      "loss": 0.4858,
+      "epoch": 2.9232343909928353,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019109276707167839,
+      "loss": 0.5763,
       "step": 2856
     },
     {
-      "epoch": 3.0522945570971185,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.0001825791053215026,
-      "loss": 0.5651,
+      "epoch": 2.927328556806551,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019082492041479188,
+      "loss": 0.624,
       "step": 2860
     },
     {
-      "epoch": 3.056563500533618,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018229566872770886,
-      "loss": 0.5377,
+      "epoch": 2.9314227226202663,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019055693309887712,
+      "loss": 0.6114,
       "step": 2864
     },
     {
-      "epoch": 3.0608324439701176,
+      "epoch": 2.9355168884339817,
       "grad_norm": 0.296875,
-      "learning_rate": 0.00018201211116519977,
-      "loss": 0.4624,
+      "learning_rate": 0.00019028880604726316,
+      "loss": 0.6243,
       "step": 2868
     },
     {
-      "epoch": 3.065101387406617,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.00018172843369608673,
-      "loss": 0.4872,
+      "epoch": 2.939611054247697,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019002054018376052,
+      "loss": 0.595,
       "step": 2872
     },
     {
-      "epoch": 3.069370330843116,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00018144463738293005,
-      "loss": 0.5293,
+      "epoch": 2.943705220061412,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018975213643265799,
+      "loss": 0.5596,
       "step": 2876
     },
     {
-      "epoch": 3.073639274279616,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00018116072328873536,
-      "loss": 0.5263,
+      "epoch": 2.947799385875128,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018948359571871936,
+      "loss": 0.5698,
       "step": 2880
     },
     {
-      "epoch": 3.0779082177161152,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001808766924769493,
-      "loss": 0.5073,
+      "epoch": 2.951893551688843,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018921491896718048,
+      "loss": 0.6207,
       "step": 2884
     },
     {
-      "epoch": 3.0821771611526145,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00018059254601145586,
-      "loss": 0.5269,
+      "epoch": 2.955987717502559,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018894610710374574,
+      "loss": 0.6277,
       "step": 2888
     },
     {
-      "epoch": 3.0864461045891143,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00018030828495657218,
-      "loss": 0.4948,
+      "epoch": 2.960081883316274,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018867716105458506,
+      "loss": 0.5984,
       "step": 2892
     },
     {
-      "epoch": 3.0907150480256136,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00018002391037704447,
-      "loss": 0.5062,
+      "epoch": 2.96417604912999,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018840808174633088,
+      "loss": 0.5292,
       "step": 2896
     },
     {
-      "epoch": 3.0949839914621133,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00017973942333804436,
-      "loss": 0.527,
+      "epoch": 2.968270214943705,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018813887010607456,
+      "loss": 0.5835,
       "step": 2900
     },
     {
-      "epoch": 3.0992529348986126,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00017945482490516464,
-      "loss": 0.503,
+      "epoch": 2.972364380757421,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018786952706136343,
+      "loss": 0.6388,
       "step": 2904
     },
     {
-      "epoch": 3.103521878335112,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001791701161444153,
-      "loss": 0.5259,
+      "epoch": 2.976458546571136,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018760005354019765,
+      "loss": 0.6434,
       "step": 2908
     },
     {
-      "epoch": 3.1077908217716117,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00017888529812221964,
-      "loss": 0.5332,
+      "epoch": 2.9805527123848514,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018733045047102695,
+      "loss": 0.5781,
       "step": 2912
     },
     {
-      "epoch": 3.112059765208111,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00017860037190541023,
-      "loss": 0.5093,
+      "epoch": 2.984646878198567,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018706071878274718,
+      "loss": 0.6452,
       "step": 2916
     },
     {
-      "epoch": 3.1163287086446103,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001783153385612248,
-      "loss": 0.5255,
+      "epoch": 2.9887410440122824,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018679085940469763,
+      "loss": 0.6056,
       "step": 2920
     },
     {
-      "epoch": 3.12059765208111,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00017803019915730252,
-      "loss": 0.5086,
+      "epoch": 2.992835209825998,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018652087326665741,
+      "loss": 0.5462,
       "step": 2924
     },
     {
-      "epoch": 3.1248665955176094,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001777449547616796,
-      "loss": 0.4953,
+      "epoch": 2.9969293756397133,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018625076129884233,
+      "loss": 0.6397,
       "step": 2928
     },
     {
-      "epoch": 3.1291355389541087,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00017745960644278567,
-      "loss": 0.4607,
+      "epoch": 3.0010235414534288,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001859805244319017,
+      "loss": 0.5911,
       "step": 2932
     },
     {
-      "epoch": 3.1334044823906084,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00017717415526943958,
-      "loss": 0.4885,
+      "epoch": 3.0051177072671442,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018571016359691532,
+      "loss": 0.5014,
       "step": 2936
     },
     {
-      "epoch": 3.1376734258271077,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001768886023108455,
-      "loss": 0.5169,
+      "epoch": 3.0092118730808597,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018543967972539004,
+      "loss": 0.5192,
       "step": 2940
     },
     {
-      "epoch": 3.1419423692636075,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001766029486365887,
-      "loss": 0.5089,
+      "epoch": 3.013306038894575,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018516907374925651,
+      "loss": 0.5236,
       "step": 2944
     },
     {
-      "epoch": 3.146211312700107,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00017631719531663187,
-      "loss": 0.4862,
+      "epoch": 3.0174002047082906,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018489834660086624,
+      "loss": 0.5035,
       "step": 2948
     },
     {
-      "epoch": 3.150480256136606,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00017603134342131078,
-      "loss": 0.5473,
+      "epoch": 3.021494370522006,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018462749921298817,
+      "loss": 0.5226,
       "step": 2952
     },
     {
-      "epoch": 3.154749199573106,
+      "epoch": 3.0255885363357216,
       "grad_norm": 0.322265625,
-      "learning_rate": 0.00017574539402133058,
-      "loss": 0.4956,
+      "learning_rate": 0.0001843565325188055,
+      "loss": 0.5353,
       "step": 2956
     },
     {
-      "epoch": 3.159018143009605,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00017545934818776152,
-      "loss": 0.5392,
+      "epoch": 3.029682702149437,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018408544745191247,
+      "loss": 0.4981,
       "step": 2960
     },
     {
-      "epoch": 3.1632870864461045,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017517320699203515,
-      "loss": 0.4986,
+      "epoch": 3.0337768679631525,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018381424494631128,
+      "loss": 0.4917,
       "step": 2964
     },
     {
-      "epoch": 3.167556029882604,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.0001748869715059401,
-      "loss": 0.4916,
+      "epoch": 3.037871033776868,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018354292593640857,
+      "loss": 0.4919,
       "step": 2968
     },
     {
-      "epoch": 3.1718249733191035,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00017460064280161832,
-      "loss": 0.5525,
+      "epoch": 3.0419651995905834,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001832714913570126,
+      "loss": 0.5445,
       "step": 2972
     },
     {
-      "epoch": 3.176093916755603,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017431422195156082,
-      "loss": 0.5064,
+      "epoch": 3.046059365404299,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018299994214332962,
+      "loss": 0.5512,
       "step": 2976
     },
     {
-      "epoch": 3.1803628601921026,
+      "epoch": 3.0501535312180144,
       "grad_norm": 0.314453125,
-      "learning_rate": 0.00017402771002860383,
-      "loss": 0.4823,
+      "learning_rate": 0.00018272827923096095,
+      "loss": 0.483,
       "step": 2980
     },
     {
-      "epoch": 3.184631803628602,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00017374110810592465,
-      "loss": 0.5476,
+      "epoch": 3.05424769703173,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018245650355589964,
+      "loss": 0.4659,
       "step": 2984
     },
     {
-      "epoch": 3.188900747065101,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017345441725703767,
-      "loss": 0.5754,
+      "epoch": 3.0583418628454453,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018218461605452722,
+      "loss": 0.5279,
       "step": 2988
     },
     {
-      "epoch": 3.193169690501601,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.0001731676385557905,
-      "loss": 0.4915,
+      "epoch": 3.0624360286591608,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018191261766361053,
+      "loss": 0.5276,
       "step": 2992
     },
     {
-      "epoch": 3.1974386339381002,
-      "grad_norm": 0.29296875,
-      "learning_rate": 0.0001728807730763596,
-      "loss": 0.489,
+      "epoch": 3.0665301944728762,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001816405093202985,
+      "loss": 0.5401,
       "step": 2996
     },
     {
-      "epoch": 3.2017075773746,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001725938218932467,
-      "loss": 0.4872,
+      "epoch": 3.0706243602865917,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001813682919621189,
+      "loss": 0.5575,
       "step": 3000
     },
     {
-      "epoch": 3.2059765208110993,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017230678608127437,
-      "loss": 0.5863,
+      "epoch": 3.074718526100307,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018109596652697495,
+      "loss": 0.5572,
       "step": 3004
     },
     {
-      "epoch": 3.2102454642475986,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00017201966671558227,
-      "loss": 0.568,
+      "epoch": 3.0788126919140226,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018082353395314243,
+      "loss": 0.4961,
       "step": 3008
     },
     {
-      "epoch": 3.2145144076840984,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.000171732464871623,
-      "loss": 0.5618,
+      "epoch": 3.082906857727738,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00018055099517926625,
+      "loss": 0.5249,
       "step": 3012
     },
     {
-      "epoch": 3.2187833511205977,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00017144518162515813,
-      "loss": 0.4897,
+      "epoch": 3.0870010235414536,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018027835114435702,
+      "loss": 0.541,
       "step": 3016
     },
     {
-      "epoch": 3.223052294557097,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00017115781805225404,
-      "loss": 0.5517,
+      "epoch": 3.091095189355169,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00018000560278778825,
+      "loss": 0.5231,
       "step": 3020
     },
     {
-      "epoch": 3.2273212379935967,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00017087037522927806,
-      "loss": 0.5146,
+      "epoch": 3.0951893551688845,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017973275104929276,
+      "loss": 0.5053,
       "step": 3024
     },
     {
-      "epoch": 3.231590181430096,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001705828542328944,
-      "loss": 0.4868,
+      "epoch": 3.0992835209826,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001794597968689596,
+      "loss": 0.4864,
       "step": 3028
     },
     {
-      "epoch": 3.2358591248665953,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00017029525614005995,
-      "loss": 0.5151,
+      "epoch": 3.1033776867963154,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001791867411872308,
+      "loss": 0.4932,
       "step": 3032
     },
     {
-      "epoch": 3.240128068303095,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00017000758202802047,
-      "loss": 0.53,
+      "epoch": 3.107471852610031,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017891358494489805,
+      "loss": 0.5039,
       "step": 3036
     },
     {
-      "epoch": 3.2443970117395944,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001697198329743065,
-      "loss": 0.525,
+      "epoch": 3.1115660184237464,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017864032908309946,
+      "loss": 0.5224,
       "step": 3040
     },
     {
-      "epoch": 3.2486659551760937,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016943201005672917,
-      "loss": 0.5,
+      "epoch": 3.115660184237462,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017836697454331658,
+      "loss": 0.5166,
       "step": 3044
     },
     {
-      "epoch": 3.2529348986125934,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016914411435337644,
-      "loss": 0.5364,
+      "epoch": 3.119754350051177,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017809352226737075,
+      "loss": 0.5291,
       "step": 3048
     },
     {
-      "epoch": 3.2572038420490927,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00016885614694260877,
-      "loss": 0.5358,
+      "epoch": 3.1238485158648923,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001778199731974201,
+      "loss": 0.5146,
       "step": 3052
     },
     {
-      "epoch": 3.2614727854855925,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0001685681089030552,
-      "loss": 0.5054,
+      "epoch": 3.127942681678608,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001775463282759563,
+      "loss": 0.5323,
       "step": 3056
     },
     {
-      "epoch": 3.265741728922092,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00016828000131360939,
-      "loss": 0.5387,
+      "epoch": 3.1320368474923233,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017727258844580125,
+      "loss": 0.5511,
       "step": 3060
     },
     {
-      "epoch": 3.270010672358591,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00016799182525342553,
-      "loss": 0.5186,
+      "epoch": 3.1361310133060387,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017699875465010382,
+      "loss": 0.5502,
       "step": 3064
     },
     {
-      "epoch": 3.274279615795091,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001677035818019142,
-      "loss": 0.4767,
+      "epoch": 3.140225179119754,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017672482783233668,
+      "loss": 0.5313,
       "step": 3068
     },
     {
-      "epoch": 3.27854855923159,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001674152720387385,
-      "loss": 0.5128,
+      "epoch": 3.1443193449334697,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017645080893629298,
+      "loss": 0.4961,
       "step": 3072
     },
     {
-      "epoch": 3.2828175026680895,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00016712689704380978,
-      "loss": 0.5026,
+      "epoch": 3.148413510747185,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017617669890608305,
+      "loss": 0.5336,
       "step": 3076
     },
     {
-      "epoch": 3.287086446104589,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00016683845789728383,
-      "loss": 0.5144,
+      "epoch": 3.1525076765609006,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017590249868613137,
+      "loss": 0.5259,
       "step": 3080
     },
     {
-      "epoch": 3.2913553895410885,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00016654995567955667,
-      "loss": 0.5295,
+      "epoch": 3.156601842374616,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000175628209221173,
+      "loss": 0.5445,
       "step": 3084
     },
     {
-      "epoch": 3.295624332977588,
+      "epoch": 3.1606960081883315,
       "grad_norm": 0.31640625,
-      "learning_rate": 0.00016626139147126067,
-      "loss": 0.5135,
+      "learning_rate": 0.00017535383145625056,
+      "loss": 0.5006,
       "step": 3088
     },
     {
-      "epoch": 3.2998932764140876,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001659727663532603,
-      "loss": 0.5232,
+      "epoch": 3.164790174002047,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017507936633671093,
+      "loss": 0.4755,
       "step": 3092
     },
     {
-      "epoch": 3.304162219850587,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00016568408140664817,
-      "loss": 0.5302,
+      "epoch": 3.1688843398157625,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017480481480820195,
+      "loss": 0.5028,
       "step": 3096
     },
     {
-      "epoch": 3.3084311632870866,
+      "epoch": 3.172978505629478,
       "grad_norm": 0.32421875,
-      "learning_rate": 0.00016539533771274103,
-      "loss": 0.541,
+      "learning_rate": 0.0001745301778166691,
+      "loss": 0.4948,
       "step": 3100
     },
     {
-      "epoch": 3.312700106723586,
+      "epoch": 3.1770726714431934,
       "grad_norm": 0.330078125,
-      "learning_rate": 0.00016510653635307566,
-      "loss": 0.4886,
+      "learning_rate": 0.00017425545630835238,
+      "loss": 0.489,
       "step": 3104
     },
     {
-      "epoch": 3.3169690501600853,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00016481767840940482,
-      "loss": 0.525,
+      "epoch": 3.181166837256909,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.000173980651229783,
+      "loss": 0.518,
       "step": 3108
     },
     {
-      "epoch": 3.321237993596585,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001645287649636933,
-      "loss": 0.5047,
+      "epoch": 3.1852610030706243,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017370576352778008,
+      "loss": 0.5336,
       "step": 3112
     },
     {
-      "epoch": 3.3255069370330843,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00016423979709811367,
-      "loss": 0.5079,
+      "epoch": 3.18935516888434,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001734307941494474,
+      "loss": 0.5271,
       "step": 3116
     },
     {
-      "epoch": 3.3297758804695836,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00016395077589504233,
-      "loss": 0.526,
+      "epoch": 3.1934493346980553,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017315574404217017,
+      "loss": 0.5106,
       "step": 3120
     },
     {
-      "epoch": 3.3340448239060834,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00016366170243705563,
-      "loss": 0.5898,
+      "epoch": 3.1975435005117707,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017288061415361174,
+      "loss": 0.4721,
       "step": 3124
     },
     {
-      "epoch": 3.3383137673425827,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00016337257780692552,
-      "loss": 0.5152,
+      "epoch": 3.201637666325486,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017260540543171036,
+      "loss": 0.5574,
       "step": 3128
     },
     {
-      "epoch": 3.342582710779082,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00016308340308761557,
-      "loss": 0.5282,
+      "epoch": 3.2057318321392017,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017233011882467582,
+      "loss": 0.5063,
       "step": 3132
     },
     {
-      "epoch": 3.3468516542155817,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00016279417936227713,
-      "loss": 0.5333,
+      "epoch": 3.209825997952917,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017205475528098637,
+      "loss": 0.5305,
       "step": 3136
     },
     {
-      "epoch": 3.351120597652081,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016250490771424497,
-      "loss": 0.5153,
+      "epoch": 3.2139201637666326,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001717793157493852,
+      "loss": 0.5379,
       "step": 3140
     },
     {
-      "epoch": 3.355389541088581,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00016221558922703343,
-      "loss": 0.5634,
+      "epoch": 3.218014329580348,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017150380117887751,
+      "loss": 0.5002,
       "step": 3144
     },
     {
-      "epoch": 3.35965848452508,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00016192622498433225,
-      "loss": 0.5322,
+      "epoch": 3.2221084953940635,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017122821251872684,
+      "loss": 0.4996,
       "step": 3148
     },
     {
-      "epoch": 3.3639274279615794,
+      "epoch": 3.226202661207779,
       "grad_norm": 0.3203125,
-      "learning_rate": 0.00016163681607000258,
-      "loss": 0.5333,
+      "learning_rate": 0.00017095255071845206,
+      "loss": 0.5536,
       "step": 3152
     },
     {
-      "epoch": 3.368196371398079,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00016134736356807292,
-      "loss": 0.5325,
+      "epoch": 3.2302968270214945,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017067681672782416,
+      "loss": 0.5189,
       "step": 3156
     },
     {
-      "epoch": 3.3724653148345785,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00016105786856273504,
-      "loss": 0.5468,
+      "epoch": 3.23439099283521,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017040101149686264,
+      "loss": 0.5405,
       "step": 3160
     },
     {
-      "epoch": 3.3767342582710778,
+      "epoch": 3.2384851586489254,
       "grad_norm": 0.3125,
-      "learning_rate": 0.00016076833213833982,
-      "loss": 0.4629,
+      "learning_rate": 0.0001701251359758326,
+      "loss": 0.5288,
       "step": 3164
     },
     {
-      "epoch": 3.3810032017075775,
+      "epoch": 3.242579324462641,
       "grad_norm": 0.31640625,
-      "learning_rate": 0.0001604787553793934,
-      "loss": 0.4933,
+      "learning_rate": 0.00016984919111524136,
+      "loss": 0.5539,
       "step": 3168
     },
     {
-      "epoch": 3.385272145144077,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.0001601891393705529,
-      "loss": 0.5063,
+      "epoch": 3.2466734902763563,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016957317786583497,
+      "loss": 0.5721,
       "step": 3172
     },
     {
-      "epoch": 3.389541088580576,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00015989948519662254,
-      "loss": 0.5133,
+      "epoch": 3.250767656090072,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016929709717859525,
+      "loss": 0.4987,
       "step": 3176
     },
     {
-      "epoch": 3.393810032017076,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00015960979394254944,
-      "loss": 0.5361,
+      "epoch": 3.2548618219037873,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016902095000473637,
+      "loss": 0.4804,
       "step": 3180
     },
     {
-      "epoch": 3.398078975453575,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00015932006669341963,
-      "loss": 0.5147,
+      "epoch": 3.2589559877175027,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016874473729570148,
+      "loss": 0.5401,
       "step": 3184
     },
     {
-      "epoch": 3.402347918890075,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00015903030453445392,
-      "loss": 0.5254,
+      "epoch": 3.263050153531218,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016846846000315957,
+      "loss": 0.4868,
       "step": 3188
     },
     {
-      "epoch": 3.4066168623265742,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015874050855100398,
-      "loss": 0.5348,
+      "epoch": 3.2671443193449337,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016819211907900225,
+      "loss": 0.5596,
       "step": 3192
     },
     {
-      "epoch": 3.4108858057630735,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00015845067982854809,
-      "loss": 0.5104,
+      "epoch": 3.2712384851586487,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001679157154753402,
+      "loss": 0.5044,
       "step": 3196
     },
     {
-      "epoch": 3.4151547491995733,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001581608194526872,
-      "loss": 0.5197,
+      "epoch": 3.2753326509723646,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016763925014450008,
+      "loss": 0.5377,
       "step": 3200
     },
     {
-      "epoch": 3.4194236926360726,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015787092850914078,
-      "loss": 0.5313,
+      "epoch": 3.2794268167860796,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001673627240390214,
+      "loss": 0.5826,
       "step": 3204
     },
     {
-      "epoch": 3.423692636072572,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00015758100808374286,
-      "loss": 0.498,
+      "epoch": 3.2835209825997955,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001670861381116529,
+      "loss": 0.5723,
       "step": 3208
     },
     {
-      "epoch": 3.4279615795090717,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001572910592624379,
-      "loss": 0.5305,
+      "epoch": 3.2876151484135105,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016680949331534948,
+      "loss": 0.4922,
       "step": 3212
     },
     {
-      "epoch": 3.432230522945571,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00015700108313127659,
-      "loss": 0.5478,
+      "epoch": 3.291709314227226,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016653279060326883,
+      "loss": 0.5247,
       "step": 3216
     },
     {
-      "epoch": 3.4364994663820703,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001567110807764121,
-      "loss": 0.5547,
+      "epoch": 3.2958034800409415,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016625603092876824,
+      "loss": 0.5303,
       "step": 3220
     },
     {
-      "epoch": 3.44076840981857,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00015642105328409565,
-      "loss": 0.5129,
+      "epoch": 3.299897645854657,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016597921524540125,
+      "loss": 0.5386,
       "step": 3224
     },
     {
-      "epoch": 3.4450373532550693,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00015613100174067276,
-      "loss": 0.5963,
+      "epoch": 3.3039918116683724,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016570234450691436,
+      "loss": 0.5131,
       "step": 3228
     },
     {
-      "epoch": 3.449306296691569,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015584092723257897,
-      "loss": 0.5061,
+      "epoch": 3.308085977482088,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016542541966724374,
+      "loss": 0.473,
       "step": 3232
     },
     {
-      "epoch": 3.4535752401280684,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00015555083084633584,
-      "loss": 0.563,
+      "epoch": 3.3121801432958033,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000165148441680512,
+      "loss": 0.5059,
       "step": 3236
     },
     {
-      "epoch": 3.4578441835645677,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015526071366854687,
-      "loss": 0.5036,
+      "epoch": 3.316274309109519,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001648714115010248,
+      "loss": 0.4812,
       "step": 3240
     },
     {
-      "epoch": 3.462113127001067,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015497057678589348,
-      "loss": 0.5358,
+      "epoch": 3.3203684749232343,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001645943300832678,
+      "loss": 0.5404,
       "step": 3244
     },
     {
-      "epoch": 3.4663820704375667,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00015468042128513086,
-      "loss": 0.5037,
+      "epoch": 3.3244626407369497,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016431719838190287,
+      "loss": 0.5238,
       "step": 3248
     },
     {
-      "epoch": 3.470651013874066,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00015439024825308396,
-      "loss": 0.5663,
+      "epoch": 3.328556806550665,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00016404001735176549,
+      "loss": 0.4977,
       "step": 3252
     },
     {
-      "epoch": 3.474919957310566,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00015410005877664336,
-      "loss": 0.5418,
+      "epoch": 3.3326509723643807,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016376278794786087,
+      "loss": 0.5405,
       "step": 3256
     },
     {
-      "epoch": 3.479188900747065,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00015380985394276126,
-      "loss": 0.5644,
+      "epoch": 3.336745138178096,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016348551112536095,
+      "loss": 0.5508,
       "step": 3260
     },
     {
-      "epoch": 3.4834578441835644,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.0001535196348384474,
-      "loss": 0.514,
+      "epoch": 3.3408393039918116,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016320818783960105,
+      "loss": 0.5301,
       "step": 3264
     },
     {
-      "epoch": 3.487726787620064,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00015322940255076497,
-      "loss": 0.5324,
+      "epoch": 3.344933469805527,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016293081904607663,
+      "loss": 0.4835,
       "step": 3268
     },
     {
-      "epoch": 3.4919957310565635,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015293915816682645,
-      "loss": 0.5336,
+      "epoch": 3.3490276356192425,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016265340570043978,
+      "loss": 0.5369,
       "step": 3272
     },
     {
-      "epoch": 3.4962646744930628,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001526489027737898,
-      "loss": 0.5402,
+      "epoch": 3.353121801432958,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00016237594875849628,
+      "loss": 0.5258,
       "step": 3276
     },
     {
-      "epoch": 3.5005336179295625,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00015235863745885407,
-      "loss": 0.5029,
+      "epoch": 3.3572159672466735,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016209844917620207,
+      "loss": 0.579,
       "step": 3280
     },
     {
-      "epoch": 3.504802561366062,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015206836330925556,
-      "loss": 0.506,
+      "epoch": 3.361310133060389,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016182090790965988,
+      "loss": 0.5422,
       "step": 3284
     },
     {
-      "epoch": 3.509071504802561,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00015177808141226358,
-      "loss": 0.5351,
+      "epoch": 3.3654042988741044,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016154332591511623,
+      "loss": 0.5292,
       "step": 3288
     },
     {
-      "epoch": 3.513340448239061,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00015148779285517654,
-      "loss": 0.5121,
+      "epoch": 3.36949846468782,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016126570414895785,
+      "loss": 0.5345,
       "step": 3292
     },
     {
-      "epoch": 3.51760939167556,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00015119749872531774,
-      "loss": 0.5102,
+      "epoch": 3.3735926305015353,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001609880435677085,
+      "loss": 0.5141,
       "step": 3296
     },
     {
-      "epoch": 3.52187833511206,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00015090720011003142,
-      "loss": 0.5728,
+      "epoch": 3.377686796315251,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00016071034512802577,
+      "loss": 0.4843,
       "step": 3300
     },
     {
-      "epoch": 3.5261472785485592,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00015061689809667852,
-      "loss": 0.5172,
+      "epoch": 3.3817809621289663,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016043260978669763,
+      "loss": 0.514,
       "step": 3304
     },
     {
-      "epoch": 3.5304162219850586,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015032659377263278,
-      "loss": 0.5217,
+      "epoch": 3.3858751279426818,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016015483850063912,
+      "loss": 0.5275,
       "step": 3308
     },
     {
-      "epoch": 3.5346851654215583,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.0001500362882252766,
-      "loss": 0.5536,
+      "epoch": 3.389969293756397,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00015987703222688926,
+      "loss": 0.5134,
       "step": 3312
     },
     {
-      "epoch": 3.5389541088580576,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001497459825419969,
-      "loss": 0.5247,
+      "epoch": 3.3940634595701127,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015959919192260757,
+      "loss": 0.5475,
       "step": 3316
     },
     {
-      "epoch": 3.5432230522945574,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014945567781018122,
-      "loss": 0.5181,
+      "epoch": 3.398157625383828,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00015932131854507072,
+      "loss": 0.5782,
       "step": 3320
     },
     {
-      "epoch": 3.5474919957310567,
+      "epoch": 3.4022517911975436,
       "grad_norm": 0.322265625,
-      "learning_rate": 0.00014916537511721343,
-      "loss": 0.5136,
+      "learning_rate": 0.0001590434130516695,
+      "loss": 0.5299,
       "step": 3324
     },
     {
-      "epoch": 3.551760939167556,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00014887507555046974,
-      "loss": 0.5113,
+      "epoch": 3.406345957011259,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015876547639990518,
+      "loss": 0.549,
       "step": 3328
     },
     {
-      "epoch": 3.5560298826040553,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014858478019731485,
-      "loss": 0.5573,
+      "epoch": 3.4104401228249746,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001584875095473865,
+      "loss": 0.5149,
       "step": 3332
     },
     {
-      "epoch": 3.560298826040555,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001482944901450974,
-      "loss": 0.5207,
+      "epoch": 3.41453428863869,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001582095134518263,
+      "loss": 0.5442,
       "step": 3336
     },
     {
-      "epoch": 3.5645677694770543,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00014800420648114642,
-      "loss": 0.5296,
+      "epoch": 3.4186284544524055,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015793148907103802,
+      "loss": 0.5089,
       "step": 3340
     },
     {
-      "epoch": 3.568836712913554,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00014771393029276678,
-      "loss": 0.6069,
+      "epoch": 3.422722620266121,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00015765343736293263,
+      "loss": 0.5063,
       "step": 3344
     },
     {
-      "epoch": 3.5731056563500534,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00014742366266723566,
-      "loss": 0.556,
+      "epoch": 3.4268167860798364,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015737535928551528,
+      "loss": 0.508,
       "step": 3348
     },
     {
-      "epoch": 3.5773745997865527,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00014713340469179776,
-      "loss": 0.5044,
+      "epoch": 3.4309109518935514,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015709725579688197,
+      "loss": 0.5439,
       "step": 3352
     },
     {
-      "epoch": 3.5816435432230525,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014684315745366205,
-      "loss": 0.5,
+      "epoch": 3.4350051177072674,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015681912785521618,
+      "loss": 0.5304,
       "step": 3356
     },
     {
-      "epoch": 3.5859124866595518,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00014655292203999691,
-      "loss": 0.5212,
+      "epoch": 3.4390992835209824,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001565409764187857,
+      "loss": 0.5379,
       "step": 3360
     },
     {
-      "epoch": 3.590181430096051,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014626269953792677,
-      "loss": 0.4651,
+      "epoch": 3.4431934493346983,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00015626280244593937,
+      "loss": 0.5265,
       "step": 3364
     },
     {
-      "epoch": 3.594450373532551,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001459724910345274,
-      "loss": 0.5483,
+      "epoch": 3.4472876151484133,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00015598460689510342,
+      "loss": 0.5461,
       "step": 3368
     },
     {
-      "epoch": 3.59871931696905,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001456822976168223,
-      "loss": 0.5076,
+      "epoch": 3.4513817809621288,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015570639072477865,
+      "loss": 0.5123,
       "step": 3372
     },
     {
-      "epoch": 3.6029882604055494,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00014539212037177854,
-      "loss": 0.5367,
+      "epoch": 3.4554759467758442,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015542815489353687,
+      "loss": 0.5182,
       "step": 3376
     },
     {
-      "epoch": 3.607257203842049,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00014510196038630232,
-      "loss": 0.4982,
+      "epoch": 3.4595701125895597,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001551499003600175,
+      "loss": 0.5118,
       "step": 3380
     },
     {
-      "epoch": 3.6115261472785485,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00014481181874723557,
-      "loss": 0.5131,
+      "epoch": 3.463664278403275,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015487162808292454,
+      "loss": 0.5201,
       "step": 3384
     },
     {
-      "epoch": 3.6157950907150482,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00014452169654135115,
-      "loss": 0.5494,
+      "epoch": 3.4677584442169906,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015459333902102302,
+      "loss": 0.5019,
       "step": 3388
     },
     {
-      "epoch": 3.6200640341515475,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001442315948553494,
-      "loss": 0.5233,
+      "epoch": 3.471852610030706,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015431503413313594,
+      "loss": 0.5194,
       "step": 3392
     },
     {
-      "epoch": 3.624332977588047,
+      "epoch": 3.4759467758444216,
       "grad_norm": 0.326171875,
-      "learning_rate": 0.0001439415147758536,
-      "loss": 0.5378,
+      "learning_rate": 0.00015403671437814063,
+      "loss": 0.5196,
       "step": 3396
     },
     {
-      "epoch": 3.628601921024546,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014365145738940623,
-      "loss": 0.4954,
+      "epoch": 3.480040941658137,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015375838071496583,
+      "loss": 0.5081,
       "step": 3400
     },
     {
-      "epoch": 3.632870864461046,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001433614237824647,
-      "loss": 0.5358,
+      "epoch": 3.4841351074718525,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015348003410258813,
+      "loss": 0.5039,
       "step": 3404
     },
     {
-      "epoch": 3.637139807897545,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00014307141504139737,
-      "loss": 0.5616,
+      "epoch": 3.488229273285568,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015320167550002863,
+      "loss": 0.536,
       "step": 3408
     },
     {
-      "epoch": 3.641408751334045,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00014278143225247955,
-      "loss": 0.5346,
+      "epoch": 3.4923234390992834,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00015292330586634997,
+      "loss": 0.5372,
       "step": 3412
     },
     {
-      "epoch": 3.6456776947705443,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00014249147650188912,
-      "loss": 0.5462,
+      "epoch": 3.496417604912999,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015264492616065255,
+      "loss": 0.4794,
       "step": 3416
     },
     {
-      "epoch": 3.6499466382070436,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00014220154887570295,
-      "loss": 0.5239,
+      "epoch": 3.5005117707267144,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015236653734207158,
+      "loss": 0.5065,
       "step": 3420
     },
     {
-      "epoch": 3.6542155816435433,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00014191165045989235,
-      "loss": 0.5092,
+      "epoch": 3.50460593654043,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001520881403697738,
+      "loss": 0.4901,
       "step": 3424
     },
     {
-      "epoch": 3.6584845250800426,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00014162178234031935,
-      "loss": 0.5623,
+      "epoch": 3.5087001023541453,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015180973620295383,
+      "loss": 0.5332,
       "step": 3428
     },
     {
-      "epoch": 3.6627534685165424,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001413319456027324,
-      "loss": 0.5408,
+      "epoch": 3.512794268167861,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015153132580083116,
+      "loss": 0.4868,
       "step": 3432
     },
     {
-      "epoch": 3.6670224119530417,
+      "epoch": 3.5168884339815762,
       "grad_norm": 0.322265625,
-      "learning_rate": 0.00014104214133276257,
-      "loss": 0.5285,
+      "learning_rate": 0.00015125291012264684,
+      "loss": 0.55,
       "step": 3436
     },
     {
-      "epoch": 3.671291355389541,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00014075237061591907,
-      "loss": 0.5002,
+      "epoch": 3.5209825997952917,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015097449012765993,
+      "loss": 0.4989,
       "step": 3440
     },
     {
-      "epoch": 3.6755602988260403,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00014046263453758559,
-      "loss": 0.5481,
+      "epoch": 3.525076765609007,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001506960667751445,
+      "loss": 0.4985,
       "step": 3444
     },
     {
-      "epoch": 3.67982924226254,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001401729341830162,
-      "loss": 0.5803,
+      "epoch": 3.5291709314227226,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015041764102438618,
+      "loss": 0.5499,
       "step": 3448
     },
     {
-      "epoch": 3.6840981856990394,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001398832706373308,
-      "loss": 0.5124,
+      "epoch": 3.533265097236438,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00015013921383467884,
+      "loss": 0.5324,
       "step": 3452
     },
     {
-      "epoch": 3.688367129135539,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00013959364498551183,
-      "loss": 0.5015,
+      "epoch": 3.5373592630501536,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001498607861653212,
+      "loss": 0.5165,
       "step": 3456
     },
     {
-      "epoch": 3.6926360725720384,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013930405831239943,
-      "loss": 0.5209,
+      "epoch": 3.541453428863869,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001495823589756138,
+      "loss": 0.5517,
       "step": 3460
     },
     {
-      "epoch": 3.6969050160085377,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00013901451170268803,
-      "loss": 0.5044,
+      "epoch": 3.5455475946775845,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014930393322485548,
+      "loss": 0.548,
       "step": 3464
     },
     {
-      "epoch": 3.7011739594450375,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00013872500624092173,
-      "loss": 0.5419,
+      "epoch": 3.5496417604913,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014902550987234004,
+      "loss": 0.5582,
       "step": 3468
     },
     {
-      "epoch": 3.7054429028815368,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001384355430114908,
-      "loss": 0.5198,
+      "epoch": 3.5537359263050154,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00014874708987735316,
+      "loss": 0.5943,
       "step": 3472
     },
     {
-      "epoch": 3.7097118463180365,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.000138146123098627,
-      "loss": 0.5329,
+      "epoch": 3.557830092118731,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001484686741991688,
+      "loss": 0.512,
       "step": 3476
     },
     {
-      "epoch": 3.713980789754536,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00013785674758640012,
-      "loss": 0.4795,
+      "epoch": 3.5619242579324464,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00014819026379704614,
+      "loss": 0.527,
       "step": 3480
     },
     {
-      "epoch": 3.718249733191035,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00013756741755871352,
-      "loss": 0.5345,
+      "epoch": 3.566018423746162,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001479118596302262,
+      "loss": 0.5487,
       "step": 3484
     },
     {
-      "epoch": 3.7225186766275344,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001372781340993001,
-      "loss": 0.4707,
+      "epoch": 3.5701125895598773,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001476334626579284,
+      "loss": 0.5161,
       "step": 3488
     },
     {
-      "epoch": 3.726787620064034,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001369888982917185,
-      "loss": 0.5731,
+      "epoch": 3.574206755373593,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014735507383934748,
+      "loss": 0.5076,
       "step": 3492
     },
     {
-      "epoch": 3.7310565635005335,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00013669971121934874,
-      "loss": 0.5232,
+      "epoch": 3.5783009211873082,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014707669413365008,
+      "loss": 0.5335,
       "step": 3496
     },
     {
-      "epoch": 3.7353255069370332,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00013641057396538846,
-      "loss": 0.5134,
+      "epoch": 3.5823950870010233,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001467983244999714,
+      "loss": 0.476,
       "step": 3500
     },
     {
-      "epoch": 3.7395944503735326,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00013612148761284846,
-      "loss": 0.5263,
+      "epoch": 3.586489252814739,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001465199658974119,
+      "loss": 0.5597,
       "step": 3504
     },
     {
-      "epoch": 3.743863393810032,
+      "epoch": 3.590583418628454,
       "grad_norm": 0.310546875,
-      "learning_rate": 0.0001358324532445491,
-      "loss": 0.5561,
+      "learning_rate": 0.00014624161928503417,
+      "loss": 0.5411,
       "step": 3508
     },
     {
-      "epoch": 3.7481323372465316,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00013554347194311587,
-      "loss": 0.5115,
+      "epoch": 3.59467758444217,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001459632856218594,
+      "loss": 0.5302,
       "step": 3512
     },
     {
-      "epoch": 3.752401280683031,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00013525454479097556,
-      "loss": 0.5685,
+      "epoch": 3.598771750255885,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001456849658668641,
+      "loss": 0.5309,
       "step": 3516
     },
     {
-      "epoch": 3.7566702241195307,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001349656728703522,
-      "loss": 0.5418,
+      "epoch": 3.602865916069601,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000145406660978977,
+      "loss": 0.5123,
       "step": 3520
     },
     {
-      "epoch": 3.76093916755603,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001346768572632628,
-      "loss": 0.517,
+      "epoch": 3.606960081883316,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014512837191707552,
+      "loss": 0.5087,
       "step": 3524
     },
     {
-      "epoch": 3.7652081109925293,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0001343880990515135,
-      "loss": 0.5114,
+      "epoch": 3.611054247697032,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001448500996399825,
+      "loss": 0.5063,
       "step": 3528
     },
     {
-      "epoch": 3.7694770544290286,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0001340993993166955,
-      "loss": 0.5065,
+      "epoch": 3.615148413510747,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001445718451064631,
+      "loss": 0.5133,
       "step": 3532
     },
     {
-      "epoch": 3.7737459978655283,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00013381075914018097,
-      "loss": 0.5653,
+      "epoch": 3.619242579324463,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001442936092752213,
+      "loss": 0.5234,
       "step": 3536
     },
     {
-      "epoch": 3.7780149413020276,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.0001335221796031188,
-      "loss": 0.5156,
+      "epoch": 3.623336745138178,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014401539310489656,
+      "loss": 0.534,
       "step": 3540
     },
     {
-      "epoch": 3.7822838847385274,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00013323366178643106,
-      "loss": 0.5307,
+      "epoch": 3.6274309109518934,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001437371975540606,
+      "loss": 0.4773,
       "step": 3544
     },
     {
-      "epoch": 3.7865528281750267,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00013294520677080836,
-      "loss": 0.5329,
+      "epoch": 3.631525076765609,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014345902358121423,
+      "loss": 0.5007,
       "step": 3548
     },
     {
-      "epoch": 3.790821771611526,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00013265681563670633,
-      "loss": 0.4968,
+      "epoch": 3.6356192425793243,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001431808721447838,
+      "loss": 0.5236,
       "step": 3552
     },
     {
-      "epoch": 3.7950907150480258,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00013236848946434097,
-      "loss": 0.5156,
+      "epoch": 3.63971340839304,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000142902744203118,
+      "loss": 0.5174,
       "step": 3556
     },
     {
-      "epoch": 3.799359658484525,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013208022933368537,
-      "loss": 0.51,
+      "epoch": 3.6438075742067553,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001426246407144847,
+      "loss": 0.5517,
       "step": 3560
     },
     {
-      "epoch": 3.803628601921025,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00013179203632446496,
-      "loss": 0.5558,
+      "epoch": 3.6479017400204707,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014234656263706737,
+      "loss": 0.5203,
       "step": 3564
     },
     {
-      "epoch": 3.807897545357524,
+      "epoch": 3.651995905834186,
       "grad_norm": 0.3359375,
-      "learning_rate": 0.00013150391151615386,
-      "loss": 0.4999,
+      "learning_rate": 0.00014206851092896195,
+      "loss": 0.515,
       "step": 3568
     },
     {
-      "epoch": 3.8121664887940234,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00013121585598797075,
-      "loss": 0.5827,
+      "epoch": 3.6560900716479017,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014179048654817367,
+      "loss": 0.5206,
       "step": 3572
     },
     {
-      "epoch": 3.8164354322305227,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.0001309278708188747,
-      "loss": 0.4998,
+      "epoch": 3.660184237461617,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00014151249045261347,
+      "loss": 0.5397,
       "step": 3576
     },
     {
-      "epoch": 3.8207043756670225,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00013063995708756138,
-      "loss": 0.5295,
+      "epoch": 3.6642784032753326,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001412345236000948,
+      "loss": 0.5156,
       "step": 3580
     },
     {
-      "epoch": 3.824973319103522,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013035211587245874,
-      "loss": 0.5403,
+      "epoch": 3.668372569089048,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001409565869483305,
+      "loss": 0.5261,
       "step": 3584
     },
     {
-      "epoch": 3.8292422625400215,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.0001300643482517233,
-      "loss": 0.5708,
+      "epoch": 3.6724667349027635,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014067868145492928,
+      "loss": 0.5615,
       "step": 3588
     },
     {
-      "epoch": 3.833511205976521,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00012977665530323568,
-      "loss": 0.4615,
+      "epoch": 3.676560900716479,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001404008080773924,
+      "loss": 0.4976,
       "step": 3592
     },
     {
-      "epoch": 3.83778014941302,
-      "grad_norm": 0.2890625,
-      "learning_rate": 0.00012948903810459694,
-      "loss": 0.4947,
+      "epoch": 3.6806550665301945,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001401229677731107,
+      "loss": 0.5772,
       "step": 3596
     },
     {
-      "epoch": 3.84204909284952,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00012920149773312453,
-      "loss": 0.5208,
+      "epoch": 3.68474923234391,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013984516149936088,
+      "loss": 0.5169,
       "step": 3600
     },
     {
-      "epoch": 3.846318036286019,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00012891403526584782,
-      "loss": 0.4672,
+      "epoch": 3.6888433981576254,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013956739021330234,
+      "loss": 0.5636,
       "step": 3604
     },
     {
-      "epoch": 3.8505869797225185,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012862665177950473,
-      "loss": 0.5478,
+      "epoch": 3.692937563971341,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001392896548719742,
+      "loss": 0.5431,
       "step": 3608
     },
     {
-      "epoch": 3.8548559231590183,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00012833934835053694,
-      "loss": 0.5584,
+      "epoch": 3.6970317297850563,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001390119564322915,
+      "loss": 0.5717,
       "step": 3612
     },
     {
-      "epoch": 3.8591248665955176,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001280521260550867,
-      "loss": 0.5175,
+      "epoch": 3.701125895598772,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013873429585104215,
+      "loss": 0.5165,
       "step": 3616
     },
     {
-      "epoch": 3.863393810032017,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00012776498596899196,
-      "loss": 0.5283,
+      "epoch": 3.7052200614124873,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013845667408488377,
+      "loss": 0.5255,
       "step": 3620
     },
     {
-      "epoch": 3.8676627534685166,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001274779291677831,
-      "loss": 0.5144,
+      "epoch": 3.7093142272262027,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001381790920903401,
+      "loss": 0.5074,
       "step": 3624
     },
     {
-      "epoch": 3.871931696905016,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00012719095672667822,
-      "loss": 0.5655,
+      "epoch": 3.713408393039918,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013790155082379793,
+      "loss": 0.5742,
       "step": 3628
     },
     {
-      "epoch": 3.8762006403415157,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001269040697205797,
-      "loss": 0.5192,
+      "epoch": 3.7175025588536337,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001376240512415037,
+      "loss": 0.5069,
       "step": 3632
     },
     {
-      "epoch": 3.880469583778015,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012661726922406978,
-      "loss": 0.5378,
+      "epoch": 3.721596724667349,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013734659429956024,
+      "loss": 0.5224,
       "step": 3636
     },
     {
-      "epoch": 3.8847385272145143,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012633055631140663,
-      "loss": 0.5993,
+      "epoch": 3.7256908904810646,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001370691809539234,
+      "loss": 0.6127,
       "step": 3640
     },
     {
-      "epoch": 3.8890074706510136,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012604393205652042,
-      "loss": 0.5803,
+      "epoch": 3.72978505629478,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013679181216039898,
+      "loss": 0.5235,
       "step": 3644
     },
     {
-      "epoch": 3.8932764140875133,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00012575739753300927,
-      "loss": 0.4871,
+      "epoch": 3.7338792221084955,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013651448887463905,
+      "loss": 0.4979,
       "step": 3648
     },
     {
-      "epoch": 3.8975453575240127,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012547095381413513,
-      "loss": 0.5436,
+      "epoch": 3.737973387922211,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013623721205213916,
+      "loss": 0.5628,
       "step": 3652
     },
     {
-      "epoch": 3.9018143009605124,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00012518460197281982,
-      "loss": 0.5496,
+      "epoch": 3.742067553735926,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013595998264823454,
+      "loss": 0.5107,
       "step": 3656
     },
     {
-      "epoch": 3.9060832443970117,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012489834308164114,
-      "loss": 0.5225,
+      "epoch": 3.746161719549642,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013568280161809713,
+      "loss": 0.543,
       "step": 3660
     },
     {
-      "epoch": 3.910352187833511,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001246121782128285,
-      "loss": 0.5285,
+      "epoch": 3.750255885363357,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013540566991673226,
+      "loss": 0.4928,
       "step": 3664
     },
     {
-      "epoch": 3.9146211312700108,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00012432610843825934,
-      "loss": 0.4903,
+      "epoch": 3.754350051177073,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001351285884989752,
+      "loss": 0.4854,
       "step": 3668
     },
     {
-      "epoch": 3.91889007470651,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00012404013482945496,
-      "loss": 0.5382,
+      "epoch": 3.758444216990788,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000134851558319488,
+      "loss": 0.5305,
       "step": 3672
     },
     {
-      "epoch": 3.92315901814301,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012375425845757612,
-      "loss": 0.5086,
+      "epoch": 3.762538382804504,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013457458033275623,
+      "loss": 0.5603,
       "step": 3676
     },
     {
-      "epoch": 3.927427961579509,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012346848039341976,
-      "loss": 0.4784,
+      "epoch": 3.766632548618219,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013429765549308558,
+      "loss": 0.5247,
       "step": 3680
     },
     {
-      "epoch": 3.9316969050160084,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001231828017074143,
-      "loss": 0.5401,
+      "epoch": 3.7707267144319347,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001340207847545987,
+      "loss": 0.5073,
       "step": 3684
     },
     {
-      "epoch": 3.9359658484525077,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012289722346961612,
-      "loss": 0.5282,
+      "epoch": 3.7748208802456498,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013374396907123173,
+      "loss": 0.5365,
       "step": 3688
     },
     {
-      "epoch": 3.9402347918890075,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012261174674970516,
-      "loss": 0.4625,
+      "epoch": 3.7789150460593657,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013346720939673112,
+      "loss": 0.5316,
       "step": 3692
     },
     {
-      "epoch": 3.944503735325507,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00012232637261698137,
-      "loss": 0.5061,
+      "epoch": 3.7830092118730807,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001331905066846505,
+      "loss": 0.5118,
       "step": 3696
     },
     {
-      "epoch": 3.9487726787620065,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00012204110214036011,
-      "loss": 0.5255,
+      "epoch": 3.787103377686796,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001329138618883471,
+      "loss": 0.5699,
       "step": 3700
     },
     {
-      "epoch": 3.953041622198506,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00012175593638836878,
-      "loss": 0.5236,
+      "epoch": 3.7911975435005116,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013263727596097855,
+      "loss": 0.5433,
       "step": 3704
     },
     {
-      "epoch": 3.957310565635005,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012147087642914245,
-      "loss": 0.4608,
+      "epoch": 3.795291709314227,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001323607498554999,
+      "loss": 0.4993,
       "step": 3708
     },
     {
-      "epoch": 3.961579509071505,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012118592333041974,
-      "loss": 0.51,
+      "epoch": 3.7993858751279426,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001320842845246598,
+      "loss": 0.4852,
       "step": 3712
     },
     {
-      "epoch": 3.965848452508004,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.0001209010781595392,
-      "loss": 0.508,
+      "epoch": 3.803480040941658,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013180788092099775,
+      "loss": 0.5193,
       "step": 3716
     },
     {
-      "epoch": 3.970117395944504,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00012061634198343506,
-      "loss": 0.5114,
+      "epoch": 3.8075742067553735,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001315315399968404,
+      "loss": 0.5161,
       "step": 3720
     },
     {
-      "epoch": 3.9743863393810033,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012033171586863328,
-      "loss": 0.5341,
+      "epoch": 3.811668372569089,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001312552627042985,
+      "loss": 0.4898,
       "step": 3724
     },
     {
-      "epoch": 3.9786552828175026,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012004720088124757,
-      "loss": 0.5253,
+      "epoch": 3.8157625383828044,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00013097904999526363,
+      "loss": 0.5347,
       "step": 3728
     },
     {
-      "epoch": 3.982924226254002,
+      "epoch": 3.81985670419652,
       "grad_norm": 0.33203125,
-      "learning_rate": 0.00011976279808697542,
-      "loss": 0.4828,
+      "learning_rate": 0.00013070290282140472,
+      "loss": 0.5471,
       "step": 3732
     },
     {
-      "epoch": 3.9871931696905016,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011947850855109406,
-      "loss": 0.5087,
+      "epoch": 3.8239508700102354,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013042682213416503,
+      "loss": 0.4795,
       "step": 3736
     },
     {
-      "epoch": 3.991462113127001,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00011919433333845649,
-      "loss": 0.5251,
+      "epoch": 3.828045035823951,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013015080888475865,
+      "loss": 0.5442,
       "step": 3740
     },
     {
-      "epoch": 3.9957310565635007,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001189102735134875,
-      "loss": 0.5309,
+      "epoch": 3.8321392016376663,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012987486402416738,
+      "loss": 0.4921,
       "step": 3744
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.8828125,
-      "learning_rate": 0.00011862633014017965,
-      "loss": 0.5049,
+      "epoch": 3.8362333674513818,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012959898850313736,
+      "loss": 0.5354,
       "step": 3748
     },
     {
-      "epoch": 4.004268943436499,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00011834250428208943,
-      "loss": 0.5173,
+      "epoch": 3.8403275332650972,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00012932318327217585,
+      "loss": 0.4821,
       "step": 3752
     },
     {
-      "epoch": 4.008537886872999,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011805879700233293,
-      "loss": 0.431,
+      "epoch": 3.8444216990788127,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012904744928154792,
+      "loss": 0.5164,
       "step": 3756
     },
     {
-      "epoch": 4.012806830309499,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00011777520936358237,
-      "loss": 0.441,
+      "epoch": 3.848515864892528,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012877178748127313,
+      "loss": 0.529,
       "step": 3760
     },
     {
-      "epoch": 4.017075773745998,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00011749174242806152,
-      "loss": 0.4575,
+      "epoch": 3.8526100307062436,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001284961988211225,
+      "loss": 0.5858,
       "step": 3764
     },
     {
-      "epoch": 4.021344717182497,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00011720839725754237,
-      "loss": 0.4469,
+      "epoch": 3.856704196519959,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012822068425061476,
+      "loss": 0.5239,
       "step": 3768
     },
     {
-      "epoch": 4.025613660618997,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001169251749133405,
-      "loss": 0.4558,
+      "epoch": 3.8607983623336746,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012794524471901363,
+      "loss": 0.5538,
       "step": 3772
     },
     {
-      "epoch": 4.029882604055496,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.0001166420764563117,
-      "loss": 0.4459,
+      "epoch": 3.86489252814739,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012766988117532418,
+      "loss": 0.5425,
       "step": 3776
     },
     {
-      "epoch": 4.034151547491995,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00011635910294684747,
-      "loss": 0.4952,
+      "epoch": 3.8689866939611055,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012739459456828967,
+      "loss": 0.5078,
       "step": 3780
     },
     {
-      "epoch": 4.0384204909284955,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001160762554448715,
-      "loss": 0.4439,
+      "epoch": 3.873080859774821,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012711938584638823,
+      "loss": 0.4744,
       "step": 3784
     },
     {
-      "epoch": 4.042689434364995,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001157935350098355,
-      "loss": 0.4712,
+      "epoch": 3.8771750255885364,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00012684425595782984,
+      "loss": 0.5257,
       "step": 3788
     },
     {
-      "epoch": 4.046958377801494,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00011551094270071499,
-      "loss": 0.449,
+      "epoch": 3.881269191402252,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012656920585055263,
+      "loss": 0.4963,
       "step": 3792
     },
     {
-      "epoch": 4.0512273212379935,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011522847957600585,
-      "loss": 0.4341,
+      "epoch": 3.8853633572159674,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012629423647221992,
+      "loss": 0.5248,
       "step": 3796
     },
     {
-      "epoch": 4.055496264674493,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011494614669371991,
-      "loss": 0.4613,
+      "epoch": 3.889457523029683,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000126019348770217,
+      "loss": 0.5464,
       "step": 3800
     },
     {
-      "epoch": 4.059765208110993,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00011466394511138123,
-      "loss": 0.4788,
+      "epoch": 3.8935516888433983,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00012574454369164762,
+      "loss": 0.5515,
       "step": 3804
     },
     {
-      "epoch": 4.064034151547492,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011438187588602205,
-      "loss": 0.4853,
+      "epoch": 3.8976458546571138,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001254698221833309,
+      "loss": 0.5136,
       "step": 3808
     },
     {
-      "epoch": 4.068303094983992,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00011409994007417881,
-      "loss": 0.4509,
+      "epoch": 3.901740020470829,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012519518519179805,
+      "loss": 0.5189,
       "step": 3812
     },
     {
-      "epoch": 4.072572038420491,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00011381813873188827,
-      "loss": 0.4562,
+      "epoch": 3.9058341862845447,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012492063366328899,
+      "loss": 0.561,
       "step": 3816
     },
     {
-      "epoch": 4.07684098185699,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00011353647291468349,
-      "loss": 0.4604,
+      "epoch": 3.9099283520982597,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001246461685437494,
+      "loss": 0.4895,
       "step": 3820
     },
     {
-      "epoch": 4.0811099252934895,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00011325494367759,
-      "loss": 0.4389,
+      "epoch": 3.9140225179119756,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012437179077882693,
+      "loss": 0.5158,
       "step": 3824
     },
     {
-      "epoch": 4.08537886872999,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00011297355207512149,
-      "loss": 0.4375,
+      "epoch": 3.9181166837256907,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012409750131386858,
+      "loss": 0.5381,
       "step": 3828
     },
     {
-      "epoch": 4.089647812166489,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00011269229916127645,
-      "loss": 0.4594,
+      "epoch": 3.9222108495394066,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001238233010939169,
+      "loss": 0.5346,
       "step": 3832
     },
     {
-      "epoch": 4.093916755602988,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00011241118598953358,
-      "loss": 0.4653,
+      "epoch": 3.9263050153531216,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000123549191063707,
+      "loss": 0.5082,
       "step": 3836
     },
     {
-      "epoch": 4.098185699039488,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00011213021361284841,
-      "loss": 0.4702,
+      "epoch": 3.9303991811668375,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012327517216766327,
+      "loss": 0.4927,
       "step": 3840
     },
     {
-      "epoch": 4.102454642475987,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00011184938308364886,
-      "loss": 0.4141,
+      "epoch": 3.9344933469805525,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012300124534989616,
+      "loss": 0.5242,
       "step": 3844
     },
     {
-      "epoch": 4.106723585912487,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.0001115686954538318,
-      "loss": 0.5026,
+      "epoch": 3.9385875127942684,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001227274115541987,
+      "loss": 0.5478,
       "step": 3848
     },
     {
-      "epoch": 4.110992529348986,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00011128815177475857,
-      "loss": 0.4381,
+      "epoch": 3.9426816786079835,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012245367172404367,
+      "loss": 0.5329,
       "step": 3852
     },
     {
-      "epoch": 4.115261472785486,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00011100775309725151,
-      "loss": 0.4856,
+      "epoch": 3.946775844421699,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001221800268025799,
+      "loss": 0.5041,
       "step": 3856
     },
     {
-      "epoch": 4.119530416221985,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00011072750047158987,
-      "loss": 0.446,
+      "epoch": 3.9508700102354144,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012190647773262923,
+      "loss": 0.5311,
       "step": 3860
     },
     {
-      "epoch": 4.123799359658484,
+      "epoch": 3.95496417604913,
       "grad_norm": 0.328125,
-      "learning_rate": 0.0001104473949475056,
-      "loss": 0.4616,
+      "learning_rate": 0.0001216330254566834,
+      "loss": 0.5264,
       "step": 3864
     },
     {
-      "epoch": 4.128068303094984,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00011016743757417995,
-      "loss": 0.4653,
+      "epoch": 3.9590583418628453,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012135967091690053,
+      "loss": 0.4574,
       "step": 3868
     },
     {
-      "epoch": 4.132337246531484,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00010988762940023896,
-      "loss": 0.4561,
+      "epoch": 3.963152507676561,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012108641505510195,
+      "loss": 0.5259,
       "step": 3872
     },
     {
-      "epoch": 4.136606189967983,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00010960797147375007,
-      "loss": 0.5195,
+      "epoch": 3.9672466734902763,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012081325881276917,
+      "loss": 0.5375,
       "step": 3876
     },
     {
-      "epoch": 4.140875133404482,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010932846484221774,
-      "loss": 0.4554,
+      "epoch": 3.9713408393039917,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012054020313104037,
+      "loss": 0.5233,
       "step": 3880
     },
     {
-      "epoch": 4.145144076840982,
+      "epoch": 3.975435005117707,
       "grad_norm": 0.3203125,
-      "learning_rate": 0.00010904911055257994,
-      "loss": 0.4183,
+      "learning_rate": 0.0001202672489507072,
+      "loss": 0.5807,
       "step": 3884
     },
     {
-      "epoch": 4.149413020277481,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.0001087699096512038,
-      "loss": 0.4471,
+      "epoch": 3.9795291709314227,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011999439721221173,
+      "loss": 0.493,
       "step": 3888
     },
     {
-      "epoch": 4.15368196371398,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00010849086318388205,
-      "loss": 0.4709,
+      "epoch": 3.983623336745138,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011972164885564298,
+      "loss": 0.5157,
       "step": 3892
     },
     {
-      "epoch": 4.1579509071504805,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00010821197219582894,
-      "loss": 0.446,
+      "epoch": 3.9877175025588536,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011944900482073375,
+      "loss": 0.4709,
       "step": 3896
     },
     {
-      "epoch": 4.16221985058698,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010793323773167628,
-      "loss": 0.4398,
+      "epoch": 3.991811668372569,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011917646604685753,
+      "loss": 0.5762,
       "step": 3900
     },
     {
-      "epoch": 4.166488794023479,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00010765466083546973,
-      "loss": 0.4566,
+      "epoch": 3.9959058341862845,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011890403347302505,
+      "loss": 0.4861,
       "step": 3904
     },
     {
-      "epoch": 4.1707577374599785,
-      "grad_norm": 0.369140625,
-      "learning_rate": 0.00010737624255066458,
-      "loss": 0.4657,
+      "epoch": 4.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011863170803788112,
+      "loss": 0.495,
       "step": 3908
     },
     {
-      "epoch": 4.175026680896478,
+      "epoch": 4.004094165813715,
       "grad_norm": 0.306640625,
-      "learning_rate": 0.0001070979839201222,
-      "loss": 0.4301,
+      "learning_rate": 0.00011835949067970147,
+      "loss": 0.4461,
       "step": 3912
     },
     {
-      "epoch": 4.179295624332978,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00010681988598610572,
-      "loss": 0.4653,
+      "epoch": 4.008188331627431,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011808738233638947,
+      "loss": 0.4565,
       "step": 3916
     },
     {
-      "epoch": 4.183564567769477,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00010654194979027664,
-      "loss": 0.4484,
+      "epoch": 4.012282497441146,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011781538394547278,
+      "loss": 0.4521,
       "step": 3920
     },
     {
-      "epoch": 4.187833511205977,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010626417637369036,
-      "loss": 0.4923,
+      "epoch": 4.016376663254862,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00011754349644410038,
+      "loss": 0.4518,
       "step": 3924
     },
     {
-      "epoch": 4.192102454642476,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00010598656677679281,
-      "loss": 0.4887,
+      "epoch": 4.020470829068577,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011727172076903907,
+      "loss": 0.5021,
       "step": 3928
     },
     {
-      "epoch": 4.196371398078975,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00010570912203941605,
-      "loss": 0.4493,
+      "epoch": 4.024564994882293,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011700005785667038,
+      "loss": 0.4439,
       "step": 3932
     },
     {
-      "epoch": 4.2006403415154745,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00010543184320077488,
-      "loss": 0.4353,
+      "epoch": 4.028659160696008,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001167285086429874,
+      "loss": 0.4796,
       "step": 3936
     },
     {
-      "epoch": 4.204909284951975,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001051547312994626,
-      "loss": 0.4318,
+      "epoch": 4.032753326509724,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011645707406359143,
+      "loss": 0.4299,
       "step": 3940
     },
     {
-      "epoch": 4.209178228388474,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00010487778737344711,
-      "loss": 0.4578,
+      "epoch": 4.036847492323439,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011618575505368872,
+      "loss": 0.4309,
       "step": 3944
     },
     {
-      "epoch": 4.213447171824973,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010460101246006738,
-      "loss": 0.4882,
+      "epoch": 4.040941658137155,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00011591455254808753,
+      "loss": 0.4322,
       "step": 3948
     },
     {
-      "epoch": 4.217716115261473,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00010432440759602902,
-      "loss": 0.4627,
+      "epoch": 4.04503582395087,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011564346748119453,
+      "loss": 0.4437,
       "step": 3952
     },
     {
-      "epoch": 4.221985058697972,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00010404797381740096,
-      "loss": 0.4424,
+      "epoch": 4.049129989764586,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00011537250078701184,
+      "loss": 0.4282,
       "step": 3956
     },
     {
-      "epoch": 4.226254002134472,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00010377171215961106,
-      "loss": 0.4816,
+      "epoch": 4.053224155578301,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001151016533991337,
+      "loss": 0.4951,
       "step": 3960
     },
     {
-      "epoch": 4.230522945570971,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.0001034956236574427,
-      "loss": 0.4705,
+      "epoch": 4.0573183213920165,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011483092625074347,
+      "loss": 0.4729,
       "step": 3964
     },
     {
-      "epoch": 4.234791889007471,
+      "epoch": 4.061412487205732,
       "grad_norm": 0.345703125,
-      "learning_rate": 0.00010321970934503051,
-      "loss": 0.4949,
+      "learning_rate": 0.00011456032027460996,
+      "loss": 0.4921,
       "step": 3968
     },
     {
-      "epoch": 4.23906083244397,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00010294397025585676,
-      "loss": 0.4712,
+      "epoch": 4.0655066530194475,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011428983640308463,
+      "loss": 0.4696,
       "step": 3972
     },
     {
-      "epoch": 4.243329775880469,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00010266840742274735,
-      "loss": 0.4776,
+      "epoch": 4.0696008188331625,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011401947556809827,
+      "loss": 0.4852,
       "step": 3976
     },
     {
-      "epoch": 4.247598719316969,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010239302187786795,
-      "loss": 0.4343,
+      "epoch": 4.073694984646878,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00011374923870115769,
+      "loss": 0.4644,
       "step": 3980
     },
     {
-      "epoch": 4.251867662753469,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00010211781465272028,
-      "loss": 0.4418,
+      "epoch": 4.077789150460593,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011347912673334255,
+      "loss": 0.46,
       "step": 3984
     },
     {
-      "epoch": 4.256136606189968,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00010184278677813798,
-      "loss": 0.4419,
+      "epoch": 4.081883316274309,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011320914059530232,
+      "loss": 0.48,
       "step": 3988
     },
     {
-      "epoch": 4.2604055496264674,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001015679392842831,
-      "loss": 0.4555,
+      "epoch": 4.085977482088024,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011293928121725278,
+      "loss": 0.4841,
       "step": 3992
     },
     {
-      "epoch": 4.264674493062967,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00010129327320064179,
-      "loss": 0.4332,
+      "epoch": 4.09007164790174,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00011266954952897305,
+      "loss": 0.4967,
       "step": 3996
     },
     {
-      "epoch": 4.268943436499466,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.000101018789556021,
-      "loss": 0.4498,
+      "epoch": 4.094165813715455,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011239994645980233,
+      "loss": 0.4563,
       "step": 4000
     },
     {
-      "epoch": 4.273212379935966,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.000100744489378544,
-      "loss": 0.4756,
+      "epoch": 4.098259979529171,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00011213047293863659,
+      "loss": 0.4613,
       "step": 4004
     },
     {
-      "epoch": 4.277481323372466,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001004703736956471,
-      "loss": 0.47,
+      "epoch": 4.102354145342886,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011186112989392545,
+      "loss": 0.454,
       "step": 4008
     },
     {
-      "epoch": 4.281750266808965,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010019644353407556,
-      "loss": 0.4328,
+      "epoch": 4.106448311156602,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011159191825366912,
+      "loss": 0.4905,
       "step": 4012
     },
     {
-      "epoch": 4.286019210245464,
-      "grad_norm": 0.353515625,
-      "learning_rate": 9.99226999198795e-05,
-      "loss": 0.4681,
+      "epoch": 4.110542476970317,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011132283894541492,
+      "loss": 0.4467,
       "step": 4016
     },
     {
-      "epoch": 4.2902881536819635,
-      "grad_norm": 0.326171875,
-      "learning_rate": 9.964914387841055e-05,
-      "loss": 0.4508,
+      "epoch": 4.114636642784033,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011105389289625427,
+      "loss": 0.4868,
       "step": 4020
     },
     {
-      "epoch": 4.294557097118463,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.93757764343176e-05,
-      "loss": 0.463,
+      "epoch": 4.118730808597748,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00011078508103281952,
+      "loss": 0.4499,
       "step": 4024
     },
     {
-      "epoch": 4.298826040554963,
-      "grad_norm": 0.33203125,
-      "learning_rate": 9.910259861154322e-05,
-      "loss": 0.4611,
+      "epoch": 4.122824974411464,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00011051640428128062,
+      "loss": 0.4711,
       "step": 4028
     },
     {
-      "epoch": 4.303094983991462,
-      "grad_norm": 0.328125,
-      "learning_rate": 9.88296114333196e-05,
-      "loss": 0.4357,
+      "epoch": 4.126919140225179,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00011024786356734199,
+      "loss": 0.458,
       "step": 4032
     },
     {
-      "epoch": 4.307363927427962,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.855681592216503e-05,
-      "loss": 0.4493,
+      "epoch": 4.131013306038895,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00010997945981623944,
+      "loss": 0.5013,
       "step": 4036
     },
     {
-      "epoch": 4.311632870864461,
-      "grad_norm": 0.353515625,
-      "learning_rate": 9.828421309987966e-05,
-      "loss": 0.4656,
+      "epoch": 4.13510747185261,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010971119395273683,
+      "loss": 0.449,
       "step": 4040
     },
     {
-      "epoch": 4.31590181430096,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.801180398754204e-05,
-      "loss": 0.4473,
+      "epoch": 4.139201637666326,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010944306690112285,
+      "loss": 0.4601,
       "step": 4044
     },
     {
-      "epoch": 4.3201707577374595,
-      "grad_norm": 0.318359375,
-      "learning_rate": 9.773958960550509e-05,
-      "loss": 0.3999,
+      "epoch": 4.143295803480041,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010917507958520812,
+      "loss": 0.4678,
       "step": 4048
     },
     {
-      "epoch": 4.32443970117396,
-      "grad_norm": 0.353515625,
-      "learning_rate": 9.746757097339238e-05,
-      "loss": 0.4582,
+      "epoch": 4.147389969293757,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010890723292832163,
+      "loss": 0.4787,
       "step": 4052
     },
     {
-      "epoch": 4.328708644610459,
-      "grad_norm": 0.359375,
-      "learning_rate": 9.719574911009422e-05,
-      "loss": 0.488,
+      "epoch": 4.151484135107472,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010863952785330779,
+      "loss": 0.4761,
       "step": 4056
     },
     {
-      "epoch": 4.332977588046958,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.692412503376388e-05,
-      "loss": 0.4812,
+      "epoch": 4.155578300921187,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001083719652825233,
+      "loss": 0.4994,
       "step": 4060
     },
     {
-      "epoch": 4.337246531483458,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.665269976181389e-05,
-      "loss": 0.4514,
+      "epoch": 4.159672466734903,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00010810454613783376,
+      "loss": 0.5027,
       "step": 4064
     },
     {
-      "epoch": 4.341515474919957,
-      "grad_norm": 0.34375,
-      "learning_rate": 9.638147431091193e-05,
-      "loss": 0.4569,
+      "epoch": 4.163766632548619,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001078372713406106,
+      "loss": 0.4376,
       "step": 4068
     },
     {
-      "epoch": 4.345784418356457,
-      "grad_norm": 0.330078125,
-      "learning_rate": 9.611044969697745e-05,
-      "loss": 0.4836,
+      "epoch": 4.167860798362334,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010757014181172803,
+      "loss": 0.4762,
       "step": 4072
     },
     {
-      "epoch": 4.350053361792956,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.583962693517744e-05,
-      "loss": 0.4565,
+      "epoch": 4.171954964176049,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010730315847155966,
+      "loss": 0.4696,
       "step": 4076
     },
     {
-      "epoch": 4.354322305229456,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.556900703992299e-05,
-      "loss": 0.4488,
+      "epoch": 4.176049129989765,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010703632223997534,
+      "loss": 0.4367,
       "step": 4080
     },
     {
-      "epoch": 4.358591248665955,
-      "grad_norm": 0.330078125,
-      "learning_rate": 9.529859102486508e-05,
-      "loss": 0.4646,
+      "epoch": 4.18014329580348,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010676963403633828,
+      "loss": 0.4768,
       "step": 4084
     },
     {
-      "epoch": 4.362860192102454,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.502837990289129e-05,
-      "loss": 0.4815,
+      "epoch": 4.184237461617196,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010650309477950149,
+      "loss": 0.4386,
       "step": 4088
     },
     {
-      "epoch": 4.3671291355389545,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.475837468612162e-05,
-      "loss": 0.4595,
+      "epoch": 4.188331627430911,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010623670538780487,
+      "loss": 0.4399,
       "step": 4092
     },
     {
-      "epoch": 4.371398078975454,
-      "grad_norm": 0.36328125,
-      "learning_rate": 9.448857638590476e-05,
-      "loss": 0.4787,
+      "epoch": 4.1924257932446265,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010597046677907198,
+      "loss": 0.4919,
       "step": 4096
     },
     {
-      "epoch": 4.375667022411953,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.421898601281451e-05,
-      "loss": 0.4634,
+      "epoch": 4.1965199590583415,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001057043798706068,
+      "loss": 0.4441,
       "step": 4100
     },
     {
-      "epoch": 4.3799359658484525,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.394960457664564e-05,
-      "loss": 0.4688,
+      "epoch": 4.200614124872057,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010543844557919073,
+      "loss": 0.4588,
       "step": 4104
     },
     {
-      "epoch": 4.384204909284952,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.368043308641054e-05,
-      "loss": 0.4529,
+      "epoch": 4.2047082906857725,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010517266482107927,
+      "loss": 0.4188,
       "step": 4108
     },
     {
-      "epoch": 4.388473852721451,
-      "grad_norm": 0.3515625,
-      "learning_rate": 9.341147255033498e-05,
-      "loss": 0.4625,
+      "epoch": 4.208802456499488,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010490703851199903,
+      "loss": 0.5049,
       "step": 4112
     },
     {
-      "epoch": 4.392742796157951,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.314272397585482e-05,
-      "loss": 0.442,
+      "epoch": 4.212896622313203,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010464156756714434,
+      "loss": 0.4469,
       "step": 4116
     },
     {
-      "epoch": 4.397011739594451,
-      "grad_norm": 0.37109375,
-      "learning_rate": 9.287418836961173e-05,
-      "loss": 0.4462,
+      "epoch": 4.216990788126919,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010437625290117429,
+      "loss": 0.5029,
       "step": 4120
     },
     {
-      "epoch": 4.40128068303095,
-      "grad_norm": 0.330078125,
-      "learning_rate": 9.260586673744986e-05,
-      "loss": 0.4353,
+      "epoch": 4.221084953940634,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010411109542820963,
+      "loss": 0.4443,
       "step": 4124
     },
     {
-      "epoch": 4.405549626467449,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.233776008441183e-05,
-      "loss": 0.4606,
+      "epoch": 4.22517911975435,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010384609606182933,
+      "loss": 0.4548,
       "step": 4128
     },
     {
-      "epoch": 4.4098185699039485,
-      "grad_norm": 0.333984375,
-      "learning_rate": 9.206986941473496e-05,
-      "loss": 0.4685,
+      "epoch": 4.229273285568065,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010358125571506772,
+      "loss": 0.427,
       "step": 4132
     },
     {
-      "epoch": 4.414087513340448,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.180219573184765e-05,
-      "loss": 0.4664,
+      "epoch": 4.233367451381781,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010331657530041128,
+      "loss": 0.4632,
       "step": 4136
     },
     {
-      "epoch": 4.418356456776948,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.153474003836552e-05,
-      "loss": 0.4663,
+      "epoch": 4.237461617195496,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010305205572979536,
+      "loss": 0.48,
       "step": 4140
     },
     {
-      "epoch": 4.422625400213447,
-      "grad_norm": 0.337890625,
-      "learning_rate": 9.126750333608762e-05,
-      "loss": 0.4855,
+      "epoch": 4.241555783009212,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010278769791460118,
+      "loss": 0.4348,
       "step": 4144
     },
     {
-      "epoch": 4.426894343649947,
-      "grad_norm": 0.33203125,
-      "learning_rate": 9.10004866259928e-05,
-      "loss": 0.4489,
+      "epoch": 4.245649948822927,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010252350276565269,
+      "loss": 0.4493,
       "step": 4148
     },
     {
-      "epoch": 4.431163287086446,
-      "grad_norm": 0.369140625,
-      "learning_rate": 9.07336909082359e-05,
-      "loss": 0.4741,
+      "epoch": 4.249744114636643,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010225947119321326,
+      "loss": 0.4539,
       "step": 4152
     },
     {
-      "epoch": 4.435432230522945,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.046711718214386e-05,
-      "loss": 0.4946,
+      "epoch": 4.253838280450358,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00010199560410698284,
+      "loss": 0.4563,
       "step": 4156
     },
     {
-      "epoch": 4.439701173959445,
-      "grad_norm": 0.37890625,
-      "learning_rate": 9.020076644621231e-05,
-      "loss": 0.4543,
+      "epoch": 4.257932446264074,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010173190241609452,
+      "loss": 0.4621,
       "step": 4160
     },
     {
-      "epoch": 4.443970117395945,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.993463969810156e-05,
-      "loss": 0.4739,
+      "epoch": 4.262026612077789,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010146836702911154,
+      "loss": 0.445,
       "step": 4164
     },
     {
-      "epoch": 4.448239060832444,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.966873793463286e-05,
-      "loss": 0.4391,
+      "epoch": 4.266120777891505,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010120499885402429,
+      "loss": 0.4708,
       "step": 4168
     },
     {
-      "epoch": 4.452508004268943,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.940306215178488e-05,
-      "loss": 0.485,
+      "epoch": 4.27021494370522,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00010094179879824689,
+      "loss": 0.4508,
       "step": 4172
     },
     {
-      "epoch": 4.456776947705443,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.913761334468965e-05,
-      "loss": 0.4527,
+      "epoch": 4.274309109518936,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010067876776861423,
+      "loss": 0.4957,
       "step": 4176
     },
     {
-      "epoch": 4.461045891141943,
-      "grad_norm": 0.357421875,
-      "learning_rate": 8.887239250762926e-05,
-      "loss": 0.477,
+      "epoch": 4.278403275332651,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010041590667137899,
+      "loss": 0.4882,
       "step": 4180
     },
     {
-      "epoch": 4.465314834578442,
-      "grad_norm": 0.357421875,
-      "learning_rate": 8.860740063403169e-05,
-      "loss": 0.4881,
+      "epoch": 4.282497441146367,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010015321641220816,
+      "loss": 0.4791,
       "step": 4184
     },
     {
-      "epoch": 4.4695837780149414,
-      "grad_norm": 0.333984375,
-      "learning_rate": 8.834263871646747e-05,
-      "loss": 0.3817,
+      "epoch": 4.286591606960082,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.989069789618023e-05,
+      "loss": 0.4626,
       "step": 4188
     },
     {
-      "epoch": 4.473852721451441,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.807810774664562e-05,
-      "loss": 0.4121,
+      "epoch": 4.290685772773798,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.9628352027782e-05,
+      "loss": 0.4212,
       "step": 4192
     },
     {
-      "epoch": 4.47812166488794,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.781380871541027e-05,
-      "loss": 0.416,
+      "epoch": 4.294779938587513,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.93661797109053e-05,
+      "loss": 0.4547,
       "step": 4196
     },
     {
-      "epoch": 4.482390608324439,
-      "grad_norm": 0.373046875,
-      "learning_rate": 8.754974261273675e-05,
-      "loss": 0.4839,
+      "epoch": 4.298874104401229,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.910418184884408e-05,
+      "loss": 0.457,
       "step": 4200
     },
     {
-      "epoch": 4.48665955176094,
-      "grad_norm": 0.345703125,
-      "learning_rate": 8.72859104277278e-05,
-      "loss": 0.4641,
+      "epoch": 4.302968270214944,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.884235934429126e-05,
+      "loss": 0.424,
       "step": 4204
     },
     {
-      "epoch": 4.490928495197439,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.702231314861016e-05,
-      "loss": 0.476,
+      "epoch": 4.30706243602866,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.858071309933554e-05,
+      "loss": 0.4666,
       "step": 4208
     },
     {
-      "epoch": 4.495197438633938,
-      "grad_norm": 0.345703125,
-      "learning_rate": 8.675895176273049e-05,
-      "loss": 0.4549,
+      "epoch": 4.311156601842375,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.831924401545822e-05,
+      "loss": 0.4658,
       "step": 4212
     },
     {
-      "epoch": 4.4994663820704375,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.64958272565521e-05,
-      "loss": 0.5211,
+      "epoch": 4.3152507676560905,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.805795299353042e-05,
+      "loss": 0.4881,
       "step": 4216
     },
     {
-      "epoch": 4.503735325506937,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.623294061565081e-05,
-      "loss": 0.4672,
+      "epoch": 4.3193449334698055,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.77968409338096e-05,
+      "loss": 0.4483,
       "step": 4220
     },
     {
-      "epoch": 4.508004268943436,
-      "grad_norm": 0.375,
-      "learning_rate": 8.597029282471167e-05,
-      "loss": 0.4948,
+      "epoch": 4.3234390992835205,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.753590873593667e-05,
+      "loss": 0.4651,
       "step": 4224
     },
     {
-      "epoch": 4.512273212379936,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.570788486752487e-05,
-      "loss": 0.4638,
+      "epoch": 4.3275332650972365,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.727515729893288e-05,
+      "loss": 0.4345,
       "step": 4228
     },
     {
-      "epoch": 4.516542155816436,
-      "grad_norm": 0.353515625,
-      "learning_rate": 8.544571772698247e-05,
-      "loss": 0.4763,
+      "epoch": 4.3316274309109515,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.701458752119661e-05,
+      "loss": 0.4293,
       "step": 4232
     },
     {
-      "epoch": 4.520811099252935,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.518379238507444e-05,
-      "loss": 0.4538,
+      "epoch": 4.335721596724667,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.675420030050035e-05,
+      "loss": 0.4462,
       "step": 4236
     },
     {
-      "epoch": 4.525080042689434,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.492210982288497e-05,
-      "loss": 0.4507,
+      "epoch": 4.339815762538382,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.649399653398771e-05,
+      "loss": 0.4796,
       "step": 4240
     },
     {
-      "epoch": 4.5293489861259335,
-      "grad_norm": 0.333984375,
-      "learning_rate": 8.466067102058902e-05,
-      "loss": 0.4623,
+      "epoch": 4.343909928352098,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.623397711817012e-05,
+      "loss": 0.4458,
       "step": 4244
     },
     {
-      "epoch": 4.533617929562434,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.439947695744837e-05,
-      "loss": 0.5011,
+      "epoch": 4.348004094165813,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.597414294892379e-05,
+      "loss": 0.4275,
       "step": 4248
     },
     {
-      "epoch": 4.537886872998933,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.413852861180813e-05,
-      "loss": 0.4607,
+      "epoch": 4.352098259979529,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.571449492148686e-05,
+      "loss": 0.5074,
       "step": 4252
     },
     {
-      "epoch": 4.542155816435432,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.387782696109313e-05,
-      "loss": 0.4781,
+      "epoch": 4.356192425793244,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.545503393045605e-05,
+      "loss": 0.4946,
       "step": 4256
     },
     {
-      "epoch": 4.546424759871932,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.3617372981804e-05,
-      "loss": 0.4963,
+      "epoch": 4.36028659160696,
+      "grad_norm": 0.375,
+      "learning_rate": 9.519576086978357e-05,
+      "loss": 0.5084,
       "step": 4260
     },
     {
-      "epoch": 4.550693703308431,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.335716764951387e-05,
-      "loss": 0.5218,
+      "epoch": 4.364380757420675,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.493667663277424e-05,
+      "loss": 0.4977,
       "step": 4264
     },
     {
-      "epoch": 4.554962646744931,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.30972119388643e-05,
-      "loss": 0.4628,
+      "epoch": 4.368474923234391,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.467778211208231e-05,
+      "loss": 0.4551,
       "step": 4268
     },
     {
-      "epoch": 4.55923159018143,
-      "grad_norm": 0.31640625,
-      "learning_rate": 8.283750682356213e-05,
-      "loss": 0.4196,
+      "epoch": 4.372569089048106,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.441907819970826e-05,
+      "loss": 0.4583,
       "step": 4272
     },
     {
-      "epoch": 4.56350053361793,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.257805327637521e-05,
-      "loss": 0.4553,
+      "epoch": 4.376663254861822,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.416056578699593e-05,
+      "loss": 0.4288,
       "step": 4276
     },
     {
-      "epoch": 4.567769477054429,
-      "grad_norm": 0.353515625,
-      "learning_rate": 8.231885226912942e-05,
-      "loss": 0.4456,
+      "epoch": 4.380757420675537,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.39022457646294e-05,
+      "loss": 0.4798,
       "step": 4280
     },
     {
-      "epoch": 4.572038420490928,
-      "grad_norm": 0.365234375,
-      "learning_rate": 8.205990477270449e-05,
-      "loss": 0.4715,
+      "epoch": 4.384851586489253,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.364411902262982e-05,
+      "loss": 0.4408,
       "step": 4284
     },
     {
-      "epoch": 4.576307363927428,
-      "grad_norm": 0.328125,
-      "learning_rate": 8.180121175703076e-05,
-      "loss": 0.4911,
+      "epoch": 4.388945752302968,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.33861864503524e-05,
+      "loss": 0.4388,
       "step": 4288
     },
     {
-      "epoch": 4.580576307363927,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.154277419108515e-05,
-      "loss": 0.4513,
+      "epoch": 4.393039918116684,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.31284489364834e-05,
+      "loss": 0.48,
       "step": 4292
     },
     {
-      "epoch": 4.584845250800427,
+      "epoch": 4.397134083930399,
       "grad_norm": 0.349609375,
-      "learning_rate": 8.128459304288802e-05,
-      "loss": 0.4427,
+      "learning_rate": 9.287090736903701e-05,
+      "loss": 0.4625,
       "step": 4296
     },
     {
-      "epoch": 4.5891141942369265,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.1026669279499e-05,
-      "loss": 0.4801,
+      "epoch": 4.401228249744115,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.261356263535225e-05,
+      "loss": 0.483,
       "step": 4300
     },
     {
-      "epoch": 4.593383137673426,
-      "grad_norm": 0.37890625,
-      "learning_rate": 8.076900386701393e-05,
-      "loss": 0.4279,
+      "epoch": 4.40532241555783,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.235641562209006e-05,
+      "loss": 0.4827,
       "step": 4304
     },
     {
-      "epoch": 4.597652081109925,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.051159777056063e-05,
-      "loss": 0.4912,
+      "epoch": 4.409416581371546,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.209946721523007e-05,
+      "loss": 0.4702,
       "step": 4308
     },
     {
-      "epoch": 4.601921024546424,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.025445195429587e-05,
-      "loss": 0.4679,
+      "epoch": 4.413510747185261,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.184271830006764e-05,
+      "loss": 0.479,
       "step": 4312
     },
     {
-      "epoch": 4.606189967982925,
-      "grad_norm": 0.3515625,
-      "learning_rate": 7.999756738140148e-05,
-      "loss": 0.4619,
+      "epoch": 4.417604912998977,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.158616976121085e-05,
+      "loss": 0.4666,
       "step": 4316
     },
     {
-      "epoch": 4.610458911419424,
+      "epoch": 4.421699078812692,
       "grad_norm": 0.353515625,
-      "learning_rate": 7.974094501408056e-05,
-      "loss": 0.4972,
+      "learning_rate": 9.132982248257736e-05,
+      "loss": 0.4342,
       "step": 4320
     },
     {
-      "epoch": 4.614727854855923,
-      "grad_norm": 0.359375,
-      "learning_rate": 7.948458581355431e-05,
-      "loss": 0.5111,
+      "epoch": 4.425793244626408,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.107367734739135e-05,
+      "loss": 0.4567,
       "step": 4324
     },
     {
-      "epoch": 4.6189967982924225,
-      "grad_norm": 0.369140625,
-      "learning_rate": 7.9228490740058e-05,
-      "loss": 0.456,
-      "step": 4328
+      "epoch": 4.429887410440123,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.081773523818064e-05,
+      "loss": 0.4731,
+      "step": 4328
     },
     {
-      "epoch": 4.623265741728922,
-      "grad_norm": 0.375,
-      "learning_rate": 7.897266075283774e-05,
-      "loss": 0.4702,
+      "epoch": 4.433981576253839,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.05619970367735e-05,
+      "loss": 0.4324,
       "step": 4332
     },
     {
-      "epoch": 4.627534685165422,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.871709681014658e-05,
-      "loss": 0.4795,
+      "epoch": 4.438075742067554,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.030646362429553e-05,
+      "loss": 0.4642,
       "step": 4336
     },
     {
-      "epoch": 4.631803628601921,
-      "grad_norm": 0.333984375,
-      "learning_rate": 7.846179986924117e-05,
-      "loss": 0.4402,
+      "epoch": 4.4421699078812695,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.005113588116699e-05,
+      "loss": 0.5286,
       "step": 4340
     },
     {
-      "epoch": 4.636072572038421,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.820677088637793e-05,
-      "loss": 0.4795,
+      "epoch": 4.4462640736949846,
+      "grad_norm": 0.375,
+      "learning_rate": 8.979601468709933e-05,
+      "loss": 0.4724,
       "step": 4344
     },
     {
-      "epoch": 4.64034151547492,
-      "grad_norm": 0.349609375,
-      "learning_rate": 7.795201081680972e-05,
-      "loss": 0.4306,
+      "epoch": 4.4503582395087005,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.954110092109238e-05,
+      "loss": 0.4447,
       "step": 4348
     },
     {
-      "epoch": 4.644610458911419,
-      "grad_norm": 0.349609375,
-      "learning_rate": 7.769752061478213e-05,
-      "loss": 0.4304,
+      "epoch": 4.4544524053224155,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.928639546143135e-05,
+      "loss": 0.4771,
       "step": 4352
     },
     {
-      "epoch": 4.6488794023479185,
-      "grad_norm": 0.359375,
-      "learning_rate": 7.744330123352978e-05,
-      "loss": 0.4798,
+      "epoch": 4.458546571136131,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.903189918568372e-05,
+      "loss": 0.4557,
       "step": 4356
     },
     {
-      "epoch": 4.653148345784419,
-      "grad_norm": 0.365234375,
-      "learning_rate": 7.71893536252731e-05,
-      "loss": 0.4586,
+      "epoch": 4.462640736949846,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.877761297069622e-05,
+      "loss": 0.4514,
       "step": 4360
     },
     {
-      "epoch": 4.657417289220918,
-      "grad_norm": 0.33203125,
-      "learning_rate": 7.693567874121436e-05,
-      "loss": 0.4106,
+      "epoch": 4.466734902763562,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.85235376925919e-05,
+      "loss": 0.455,
       "step": 4364
     },
     {
-      "epoch": 4.661686232657417,
-      "grad_norm": 0.3359375,
-      "learning_rate": 7.668227753153444e-05,
-      "loss": 0.5,
+      "epoch": 4.470829068577277,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.826967422676698e-05,
+      "loss": 0.4382,
       "step": 4368
     },
     {
-      "epoch": 4.665955176093917,
-      "grad_norm": 0.349609375,
-      "learning_rate": 7.642915094538898e-05,
-      "loss": 0.4805,
+      "epoch": 4.474923234390992,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.801602344788783e-05,
+      "loss": 0.4872,
       "step": 4372
     },
     {
-      "epoch": 4.670224119530416,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.617629993090516e-05,
-      "loss": 0.4428,
+      "epoch": 4.479017400204708,
+      "grad_norm": 0.365234375,
+      "learning_rate": 8.776258622988823e-05,
+      "loss": 0.4487,
       "step": 4376
     },
     {
-      "epoch": 4.674493062966915,
-      "grad_norm": 0.3203125,
-      "learning_rate": 7.592372543517772e-05,
-      "loss": 0.4533,
+      "epoch": 4.483111566018424,
+      "grad_norm": 0.375,
+      "learning_rate": 8.750936344596597e-05,
+      "loss": 0.4817,
       "step": 4380
     },
     {
-      "epoch": 4.678762006403415,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.56714284042659e-05,
-      "loss": 0.488,
+      "epoch": 4.487205731832139,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.725635596858014e-05,
+      "loss": 0.4282,
       "step": 4384
     },
     {
-      "epoch": 4.683030949839915,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.541940978318952e-05,
-      "loss": 0.4891,
+      "epoch": 4.491299897645854,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.700356466944786e-05,
+      "loss": 0.4675,
       "step": 4388
     },
     {
-      "epoch": 4.687299893276414,
-      "grad_norm": 0.349609375,
-      "learning_rate": 7.516767051592553e-05,
-      "loss": 0.4062,
+      "epoch": 4.49539406345957,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.675099041954158e-05,
+      "loss": 0.4866,
       "step": 4392
     },
     {
-      "epoch": 4.691568836712913,
-      "grad_norm": 0.32421875,
-      "learning_rate": 7.49162115454047e-05,
-      "loss": 0.4521,
+      "epoch": 4.499488229273285,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.649863408908586e-05,
+      "loss": 0.5158,
       "step": 4396
     },
     {
-      "epoch": 4.695837780149413,
-      "grad_norm": 0.32421875,
-      "learning_rate": 7.466503381350762e-05,
-      "loss": 0.4607,
+      "epoch": 4.503582395087001,
+      "grad_norm": 0.375,
+      "learning_rate": 8.62464965475544e-05,
+      "loss": 0.427,
       "step": 4400
     },
     {
-      "epoch": 4.700106723585913,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.441413826106176e-05,
-      "loss": 0.4585,
+      "epoch": 4.507676560900716,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.599457866366725e-05,
+      "loss": 0.4479,
       "step": 4404
     },
     {
-      "epoch": 4.704375667022412,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.416352582783738e-05,
-      "loss": 0.4606,
+      "epoch": 4.511770726714432,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.574288130538736e-05,
+      "loss": 0.4869,
       "step": 4408
     },
     {
-      "epoch": 4.7086446104589115,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.391319745254448e-05,
-      "loss": 0.4655,
+      "epoch": 4.515864892528147,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.549140533991807e-05,
+      "loss": 0.4635,
       "step": 4412
     },
     {
-      "epoch": 4.712913553895411,
-      "grad_norm": 0.384765625,
-      "learning_rate": 7.366315407282888e-05,
-      "loss": 0.4625,
+      "epoch": 4.519959058341863,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.524015163369993e-05,
+      "loss": 0.4592,
       "step": 4416
     },
     {
-      "epoch": 4.71718249733191,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.341339662526908e-05,
-      "loss": 0.4787,
+      "epoch": 4.524053224155578,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.498912105240773e-05,
+      "loss": 0.4623,
       "step": 4420
     },
     {
-      "epoch": 4.72145144076841,
+      "epoch": 4.528147389969294,
       "grad_norm": 0.359375,
-      "learning_rate": 7.316392604537243e-05,
-      "loss": 0.4552,
+      "learning_rate": 8.473831446094733e-05,
+      "loss": 0.4546,
       "step": 4424
     },
     {
-      "epoch": 4.72572038420491,
-      "grad_norm": 0.37109375,
-      "learning_rate": 7.291474326757192e-05,
-      "loss": 0.5086,
+      "epoch": 4.532241555783009,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.448773272345298e-05,
+      "loss": 0.5183,
       "step": 4428
     },
     {
-      "epoch": 4.729989327641409,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.26658492252224e-05,
-      "loss": 0.4603,
+      "epoch": 4.536335721596725,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.423737670328432e-05,
+      "loss": 0.4457,
       "step": 4432
     },
     {
-      "epoch": 4.734258271077908,
-      "grad_norm": 0.36328125,
-      "learning_rate": 7.241724485059719e-05,
-      "loss": 0.4938,
+      "epoch": 4.54042988741044,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.398724726302301e-05,
+      "loss": 0.4337,
       "step": 4436
     },
     {
-      "epoch": 4.7385272145144075,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.21689310748848e-05,
-      "loss": 0.4472,
+      "epoch": 4.544524053224156,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.373734526447032e-05,
+      "loss": 0.4435,
       "step": 4440
     },
     {
-      "epoch": 4.742796157950907,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.192090882818502e-05,
-      "loss": 0.4911,
+      "epoch": 4.548618219037871,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.348767156864382e-05,
+      "loss": 0.4753,
       "step": 4444
     },
     {
-      "epoch": 4.747065101387407,
-      "grad_norm": 0.3359375,
-      "learning_rate": 7.167317903950591e-05,
-      "loss": 0.4696,
+      "epoch": 4.552712384851587,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.323822703577431e-05,
+      "loss": 0.4746,
       "step": 4448
     },
     {
-      "epoch": 4.751334044823906,
-      "grad_norm": 0.361328125,
-      "learning_rate": 7.142574263675983e-05,
-      "loss": 0.4716,
+      "epoch": 4.556806550665302,
+      "grad_norm": 0.37109375,
+      "learning_rate": 8.298901252530326e-05,
+      "loss": 0.44,
       "step": 4452
     },
     {
-      "epoch": 4.755602988260406,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.117860054676048e-05,
-      "loss": 0.4309,
+      "epoch": 4.560900716479018,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.274002889587954e-05,
+      "loss": 0.4366,
       "step": 4456
     },
     {
-      "epoch": 4.759871931696905,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.093175369521887e-05,
-      "loss": 0.4704,
+      "epoch": 4.564994882292733,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.249127700535643e-05,
+      "loss": 0.4715,
       "step": 4460
     },
     {
-      "epoch": 4.764140875133404,
-      "grad_norm": 0.3359375,
-      "learning_rate": 7.068520300674038e-05,
-      "loss": 0.4772,
+      "epoch": 4.569089048106449,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.224275771078889e-05,
+      "loss": 0.437,
       "step": 4464
     },
     {
-      "epoch": 4.7684098185699035,
-      "grad_norm": 0.328125,
-      "learning_rate": 7.043894940482106e-05,
-      "loss": 0.3981,
+      "epoch": 4.573183213920164,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.199447186843051e-05,
+      "loss": 0.4556,
       "step": 4468
     },
     {
-      "epoch": 4.772678762006404,
-      "grad_norm": 0.33203125,
-      "learning_rate": 7.019299381184396e-05,
-      "loss": 0.4228,
+      "epoch": 4.5772773797338795,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.174642033373037e-05,
+      "loss": 0.438,
       "step": 4472
     },
     {
-      "epoch": 4.776947705442903,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.994733714907614e-05,
-      "loss": 0.4329,
+      "epoch": 4.5813715455475945,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.149860396133048e-05,
+      "loss": 0.4682,
       "step": 4476
     },
     {
-      "epoch": 4.781216648879402,
-      "grad_norm": 0.36328125,
-      "learning_rate": 6.970198033666475e-05,
-      "loss": 0.4629,
+      "epoch": 4.58546571136131,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.125102360506255e-05,
+      "loss": 0.4744,
       "step": 4480
     },
     {
-      "epoch": 4.785485592315902,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.945692429363401e-05,
-      "loss": 0.4789,
+      "epoch": 4.5895598771750254,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.100368011794491e-05,
+      "loss": 0.5337,
       "step": 4484
     },
     {
-      "epoch": 4.789754535752401,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.921216993788137e-05,
-      "loss": 0.4654,
+      "epoch": 4.593654042988741,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.075657435218008e-05,
+      "loss": 0.4905,
       "step": 4488
     },
     {
-      "epoch": 4.794023479188901,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.896771818617441e-05,
-      "loss": 0.4655,
+      "epoch": 4.597748208802456,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.050970715915138e-05,
+      "loss": 0.4834,
       "step": 4492
     },
     {
-      "epoch": 4.7982924226254005,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.87235699541471e-05,
-      "loss": 0.418,
+      "epoch": 4.601842374616172,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.02630793894201e-05,
+      "loss": 0.4572,
       "step": 4496
     },
     {
-      "epoch": 4.8025613660619,
-      "grad_norm": 0.369140625,
-      "learning_rate": 6.847972615629668e-05,
-      "loss": 0.4384,
+      "epoch": 4.605936540429887,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.001669189272272e-05,
+      "loss": 0.441,
       "step": 4500
     },
     {
-      "epoch": 4.806830309498399,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.823618770598009e-05,
-      "loss": 0.4601,
+      "epoch": 4.610030706243603,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.977054551796792e-05,
+      "loss": 0.429,
       "step": 4504
     },
     {
-      "epoch": 4.811099252934898,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.799295551541035e-05,
-      "loss": 0.4756,
+      "epoch": 4.614124872057318,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.952464111323335e-05,
+      "loss": 0.458,
       "step": 4508
     },
     {
-      "epoch": 4.815368196371399,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.775003049565355e-05,
-      "loss": 0.4785,
+      "epoch": 4.618219037871034,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.927897952576326e-05,
+      "loss": 0.4573,
       "step": 4512
     },
     {
-      "epoch": 4.819637139807898,
-      "grad_norm": 0.3359375,
-      "learning_rate": 6.750741355662502e-05,
-      "loss": 0.4597,
+      "epoch": 4.622313203684749,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.903356160196522e-05,
+      "loss": 0.4269,
       "step": 4516
     },
     {
-      "epoch": 4.823906083244397,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.726510560708639e-05,
-      "loss": 0.4657,
+      "epoch": 4.626407369498464,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.878838818740711e-05,
+      "loss": 0.4751,
       "step": 4520
     },
     {
-      "epoch": 4.8281750266808965,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.70231075546416e-05,
-      "loss": 0.4471,
+      "epoch": 4.63050153531218,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.85434601268146e-05,
+      "loss": 0.4122,
       "step": 4524
     },
     {
-      "epoch": 4.832443970117396,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.67814203057341e-05,
-      "loss": 0.4173,
+      "epoch": 4.634595701125896,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.829877826406794e-05,
+      "loss": 0.4739,
       "step": 4528
     },
     {
-      "epoch": 4.836712913553895,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.654004476564297e-05,
-      "loss": 0.4853,
+      "epoch": 4.638689866939611,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.805434344219902e-05,
+      "loss": 0.4686,
       "step": 4532
     },
     {
-      "epoch": 4.840981856990394,
-      "grad_norm": 0.3671875,
-      "learning_rate": 6.629898183847983e-05,
-      "loss": 0.4331,
+      "epoch": 4.642784032753326,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.781015650338865e-05,
+      "loss": 0.4247,
       "step": 4536
     },
     {
-      "epoch": 4.845250800426895,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.605823242718543e-05,
-      "loss": 0.447,
+      "epoch": 4.646878198567042,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.756621828896363e-05,
+      "loss": 0.4744,
       "step": 4540
     },
     {
-      "epoch": 4.849519743863394,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.5817797433526e-05,
-      "loss": 0.4857,
+      "epoch": 4.650972364380758,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.732252963939369e-05,
+      "loss": 0.4481,
       "step": 4544
     },
     {
-      "epoch": 4.853788687299893,
-      "grad_norm": 0.349609375,
-      "learning_rate": 6.557767775809026e-05,
-      "loss": 0.4598,
+      "epoch": 4.655066530194473,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.707909139428889e-05,
+      "loss": 0.4948,
       "step": 4548
     },
     {
-      "epoch": 4.8580576307363925,
-      "grad_norm": 0.357421875,
-      "learning_rate": 6.53378743002857e-05,
-      "loss": 0.4475,
+      "epoch": 4.659160696008188,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.683590439239626e-05,
+      "loss": 0.5113,
       "step": 4552
     },
     {
-      "epoch": 4.862326574172892,
+      "epoch": 4.663254861821904,
       "grad_norm": 0.3515625,
-      "learning_rate": 6.509838795833553e-05,
-      "loss": 0.4592,
+      "learning_rate": 7.65929694715974e-05,
+      "loss": 0.4743,
       "step": 4556
     },
     {
-      "epoch": 4.866595517609392,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.485921962927493e-05,
-      "loss": 0.4802,
+      "epoch": 4.667349027635619,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.63502874689054e-05,
+      "loss": 0.4563,
       "step": 4560
     },
     {
-      "epoch": 4.870864461045891,
-      "grad_norm": 0.36328125,
-      "learning_rate": 6.462037020894817e-05,
-      "loss": 0.4664,
+      "epoch": 4.671443193449335,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.610785922046173e-05,
+      "loss": 0.4818,
       "step": 4564
     },
     {
-      "epoch": 4.875133404482391,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.438184059200474e-05,
-      "loss": 0.4688,
+      "epoch": 4.67553735926305,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.586568556153378e-05,
+      "loss": 0.4595,
       "step": 4568
     },
     {
-      "epoch": 4.87940234791889,
-      "grad_norm": 0.36328125,
-      "learning_rate": 6.41436316718965e-05,
-      "loss": 0.5069,
+      "epoch": 4.679631525076766,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.562376732651177e-05,
+      "loss": 0.4212,
       "step": 4572
     },
     {
-      "epoch": 4.883671291355389,
-      "grad_norm": 0.361328125,
-      "learning_rate": 6.390574434087385e-05,
-      "loss": 0.433,
+      "epoch": 4.683725690890481,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.53821053489057e-05,
+      "loss": 0.4709,
       "step": 4576
     },
     {
-      "epoch": 4.887940234791889,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.366817948998284e-05,
-      "loss": 0.4681,
+      "epoch": 4.687819856704197,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.514070046134281e-05,
+      "loss": 0.5149,
       "step": 4580
     },
     {
-      "epoch": 4.892209178228389,
-      "grad_norm": 0.359375,
-      "learning_rate": 6.343093800906142e-05,
-      "loss": 0.4435,
+      "epoch": 4.691914022517912,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.489955349556457e-05,
+      "loss": 0.4564,
       "step": 4584
     },
     {
-      "epoch": 4.896478121664888,
-      "grad_norm": 0.376953125,
-      "learning_rate": 6.31940207867365e-05,
-      "loss": 0.5192,
+      "epoch": 4.696008188331628,
+      "grad_norm": 0.365234375,
+      "learning_rate": 7.465866528242361e-05,
+      "loss": 0.4646,
       "step": 4588
     },
     {
-      "epoch": 4.900747065101387,
-      "grad_norm": 0.357421875,
-      "learning_rate": 6.295742871042025e-05,
-      "loss": 0.4732,
+      "epoch": 4.700102354145343,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.441803665188124e-05,
+      "loss": 0.4564,
       "step": 4592
     },
     {
-      "epoch": 4.905016008537887,
-      "grad_norm": 0.375,
-      "learning_rate": 6.27211626663071e-05,
-      "loss": 0.4491,
+      "epoch": 4.7041965199590585,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.417766843300443e-05,
+      "loss": 0.4589,
       "step": 4596
     },
     {
-      "epoch": 4.909284951974386,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.24852235393701e-05,
-      "loss": 0.4309,
+      "epoch": 4.7082906857727735,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.393756145396267e-05,
+      "loss": 0.4152,
       "step": 4600
     },
     {
-      "epoch": 4.913553895410886,
-      "grad_norm": 0.365234375,
-      "learning_rate": 6.224961221335802e-05,
-      "loss": 0.5054,
+      "epoch": 4.7123848515864895,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.369771654202563e-05,
+      "loss": 0.4353,
       "step": 4604
     },
     {
-      "epoch": 4.9178228388473855,
-      "grad_norm": 0.345703125,
-      "learning_rate": 6.201432957079159e-05,
-      "loss": 0.4712,
+      "epoch": 4.7164790174002045,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.345813452355999e-05,
+      "loss": 0.4508,
       "step": 4608
     },
     {
-      "epoch": 4.922091782283885,
-      "grad_norm": 0.349609375,
-      "learning_rate": 6.177937649296042e-05,
-      "loss": 0.5025,
+      "epoch": 4.72057318321392,
+      "grad_norm": 0.37109375,
+      "learning_rate": 7.321881622402648e-05,
+      "loss": 0.4222,
       "step": 4612
     },
     {
-      "epoch": 4.926360725720384,
-      "grad_norm": 0.380859375,
-      "learning_rate": 6.154475385991978e-05,
-      "loss": 0.4562,
+      "epoch": 4.724667349027635,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.297976246797742e-05,
+      "loss": 0.4815,
       "step": 4616
     },
     {
-      "epoch": 4.930629669156883,
-      "grad_norm": 0.380859375,
-      "learning_rate": 6.131046255048722e-05,
-      "loss": 0.4456,
+      "epoch": 4.728761514841351,
+      "grad_norm": 0.373046875,
+      "learning_rate": 7.274097407905361e-05,
+      "loss": 0.5069,
       "step": 4620
     },
     {
-      "epoch": 4.934898612593383,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.107650344223908e-05,
-      "loss": 0.505,
+      "epoch": 4.732855680655066,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.250245187998141e-05,
+      "loss": 0.5212,
       "step": 4624
     },
     {
-      "epoch": 4.939167556029883,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.084287741150765e-05,
-      "loss": 0.4764,
+      "epoch": 4.736949846468782,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.226419669257027e-05,
+      "loss": 0.4838,
       "step": 4628
     },
     {
-      "epoch": 4.943436499466382,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.060958533337736e-05,
-      "loss": 0.4722,
+      "epoch": 4.741044012282497,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.202620933770954e-05,
+      "loss": 0.5086,
       "step": 4632
     },
     {
-      "epoch": 4.9477054429028815,
-      "grad_norm": 0.376953125,
-      "learning_rate": 6.037662808168198e-05,
-      "loss": 0.4642,
+      "epoch": 4.745138178096213,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.178849063536572e-05,
+      "loss": 0.4991,
       "step": 4636
     },
     {
-      "epoch": 4.951974386339381,
-      "grad_norm": 0.37109375,
-      "learning_rate": 6.014400652900093e-05,
-      "loss": 0.4851,
+      "epoch": 4.749232343909928,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.155104140457982e-05,
+      "loss": 0.4491,
       "step": 4640
     },
     {
-      "epoch": 4.95624332977588,
-      "grad_norm": 0.384765625,
-      "learning_rate": 5.991172154665645e-05,
-      "loss": 0.4993,
+      "epoch": 4.753326509723644,
+      "grad_norm": 0.375,
+      "learning_rate": 7.131386246346439e-05,
+      "loss": 0.4606,
       "step": 4644
     },
     {
-      "epoch": 4.96051227321238,
-      "grad_norm": 0.33984375,
-      "learning_rate": 5.967977400470988e-05,
-      "loss": 0.4346,
+      "epoch": 4.757420675537359,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.107695462920057e-05,
+      "loss": 0.4968,
       "step": 4648
     },
     {
-      "epoch": 4.96478121664888,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.944816477195874e-05,
-      "loss": 0.4559,
+      "epoch": 4.761514841351075,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.084031871803559e-05,
+      "loss": 0.4662,
       "step": 4652
     },
     {
-      "epoch": 4.969050160085379,
-      "grad_norm": 0.37109375,
-      "learning_rate": 5.92168947159334e-05,
-      "loss": 0.4851,
+      "epoch": 4.76560900716479,
+      "grad_norm": 0.36328125,
+      "learning_rate": 7.060395554527977e-05,
+      "loss": 0.505,
       "step": 4656
     },
     {
-      "epoch": 4.973319103521878,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.898596470289363e-05,
-      "loss": 0.5008,
+      "epoch": 4.769703172978506,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.03678659253036e-05,
+      "loss": 0.476,
       "step": 4660
     },
     {
-      "epoch": 4.9775880469583775,
-      "grad_norm": 0.361328125,
-      "learning_rate": 5.875537559782569e-05,
-      "loss": 0.4251,
+      "epoch": 4.773797338792221,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.013205067153522e-05,
+      "loss": 0.4489,
       "step": 4664
     },
     {
-      "epoch": 4.981856990394878,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.852512826443875e-05,
-      "loss": 0.423,
+      "epoch": 4.777891504605937,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.989651059645743e-05,
+      "loss": 0.4559,
       "step": 4668
     },
     {
-      "epoch": 4.986125933831377,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.829522356516197e-05,
-      "loss": 0.5199,
+      "epoch": 4.781985670419652,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.966124651160479e-05,
+      "loss": 0.4769,
       "step": 4672
     },
     {
-      "epoch": 4.990394877267876,
-      "grad_norm": 0.361328125,
-      "learning_rate": 5.8065662361140965e-05,
-      "loss": 0.4578,
+      "epoch": 4.786079836233368,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.942625922756114e-05,
+      "loss": 0.4543,
       "step": 4676
     },
     {
-      "epoch": 4.994663820704376,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.783644551223491e-05,
-      "loss": 0.4284,
+      "epoch": 4.790174002047083,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.91915495539565e-05,
+      "loss": 0.4677,
       "step": 4680
     },
     {
-      "epoch": 4.998932764140875,
-      "grad_norm": 0.333984375,
-      "learning_rate": 5.760757387701295e-05,
-      "loss": 0.4219,
+      "epoch": 4.794268167860798,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.89571182994645e-05,
+      "loss": 0.4448,
       "step": 4684
     },
     {
-      "epoch": 5.003201707577374,
-      "grad_norm": 0.32421875,
-      "learning_rate": 5.7379048312751295e-05,
-      "loss": 0.4038,
+      "epoch": 4.798362333674514,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.872296627179943e-05,
+      "loss": 0.4758,
       "step": 4688
     },
     {
-      "epoch": 5.0074706510138745,
+      "epoch": 4.80245649948823,
       "grad_norm": 0.3515625,
-      "learning_rate": 5.715086967542995e-05,
-      "loss": 0.4555,
+      "learning_rate": 6.848909427771361e-05,
+      "loss": 0.463,
       "step": 4692
     },
     {
-      "epoch": 5.011739594450374,
-      "grad_norm": 0.32421875,
-      "learning_rate": 5.692303881972924e-05,
-      "loss": 0.4328,
+      "epoch": 4.806550665301945,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.825550312299432e-05,
+      "loss": 0.4602,
       "step": 4696
     },
     {
-      "epoch": 5.016008537886873,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.6695556599027046e-05,
-      "loss": 0.4424,
+      "epoch": 4.81064483111566,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.802219361246149e-05,
+      "loss": 0.4589,
       "step": 4700
     },
     {
-      "epoch": 5.020277481323372,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.646842386539517e-05,
-      "loss": 0.4536,
+      "epoch": 4.814738996929376,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.778916654996455e-05,
+      "loss": 0.4715,
       "step": 4704
     },
     {
-      "epoch": 5.024546424759872,
-      "grad_norm": 0.3203125,
-      "learning_rate": 5.624164146959656e-05,
-      "loss": 0.4217,
+      "epoch": 4.818833162743092,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.755642273837969e-05,
+      "loss": 0.4401,
       "step": 4708
     },
     {
-      "epoch": 5.028815368196371,
-      "grad_norm": 0.326171875,
-      "learning_rate": 5.601521026108172e-05,
-      "loss": 0.4335,
+      "epoch": 4.822927328556807,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.732396297960732e-05,
+      "loss": 0.4417,
       "step": 4712
     },
     {
-      "epoch": 5.033084311632871,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.57891310879859e-05,
-      "loss": 0.4477,
+      "epoch": 4.827021494370522,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.70917880745692e-05,
+      "loss": 0.4187,
       "step": 4716
     },
     {
-      "epoch": 5.0373532550693705,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.556340479712558e-05,
-      "loss": 0.4567,
+      "epoch": 4.8311156601842375,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.68598988232054e-05,
+      "loss": 0.4683,
       "step": 4720
     },
     {
-      "epoch": 5.04162219850587,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.5338032233995574e-05,
-      "loss": 0.4348,
+      "epoch": 4.835209825997953,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.662829602447207e-05,
+      "loss": 0.4654,
       "step": 4724
     },
     {
-      "epoch": 5.045891141942369,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.511301424276577e-05,
-      "loss": 0.4333,
+      "epoch": 4.8393039918116685,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.639698047633834e-05,
+      "loss": 0.4565,
       "step": 4728
     },
     {
-      "epoch": 5.050160085378868,
-      "grad_norm": 0.34375,
-      "learning_rate": 5.488835166627783e-05,
-      "loss": 0.368,
+      "epoch": 4.8433981576253835,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.616595297578346e-05,
+      "loss": 0.4464,
       "step": 4732
     },
     {
-      "epoch": 5.054429028815369,
-      "grad_norm": 0.326171875,
-      "learning_rate": 5.466404534604229e-05,
-      "loss": 0.3895,
+      "epoch": 4.847492323439099,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.59352143187945e-05,
+      "loss": 0.4892,
       "step": 4736
     },
     {
-      "epoch": 5.058697972251868,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.44400961222351e-05,
-      "loss": 0.4395,
+      "epoch": 4.851586489252814,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.57047653003632e-05,
+      "loss": 0.4535,
       "step": 4740
     },
     {
-      "epoch": 5.062966915688367,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.421650483369486e-05,
-      "loss": 0.4307,
+      "epoch": 4.85568065506653,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.547460671448333e-05,
+      "loss": 0.4681,
       "step": 4744
     },
     {
-      "epoch": 5.0672358591248665,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.3993272317919296e-05,
-      "loss": 0.4373,
+      "epoch": 4.859774820880245,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.524473935414807e-05,
+      "loss": 0.4673,
       "step": 4748
     },
     {
-      "epoch": 5.071504802561366,
-      "grad_norm": 0.34375,
-      "learning_rate": 5.377039941106229e-05,
-      "loss": 0.4661,
+      "epoch": 4.863868986693961,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.50151640113473e-05,
+      "loss": 0.4545,
       "step": 4752
     },
     {
-      "epoch": 5.075773745997865,
-      "grad_norm": 0.345703125,
-      "learning_rate": 5.354788694793093e-05,
-      "loss": 0.4519,
+      "epoch": 4.867963152507676,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.478588147706454e-05,
+      "loss": 0.4575,
       "step": 4756
     },
     {
-      "epoch": 5.080042689434365,
-      "grad_norm": 0.365234375,
-      "learning_rate": 5.332573576198194e-05,
-      "loss": 0.4807,
+      "epoch": 4.872057318321392,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.455689254127464e-05,
+      "loss": 0.4646,
       "step": 4760
     },
     {
-      "epoch": 5.084311632870865,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.3103946685319115e-05,
-      "loss": 0.4407,
+      "epoch": 4.876151484135107,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.43281979929409e-05,
+      "loss": 0.4741,
       "step": 4764
     },
     {
-      "epoch": 5.088580576307364,
-      "grad_norm": 0.353515625,
-      "learning_rate": 5.2882520548689615e-05,
-      "loss": 0.4046,
+      "epoch": 4.880245649948823,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.409979862001215e-05,
+      "loss": 0.4658,
       "step": 4768
     },
     {
-      "epoch": 5.092849519743863,
-      "grad_norm": 0.33203125,
-      "learning_rate": 5.266145818148142e-05,
-      "loss": 0.3884,
+      "epoch": 4.884339815762538,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.387169520942037e-05,
+      "loss": 0.5014,
       "step": 4772
     },
     {
-      "epoch": 5.0971184631803625,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.244076041171971e-05,
-      "loss": 0.4297,
+      "epoch": 4.888433981576254,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.364388854707786e-05,
+      "loss": 0.4532,
       "step": 4776
     },
     {
-      "epoch": 5.101387406616863,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.22204280660642e-05,
-      "loss": 0.4182,
+      "epoch": 4.892528147389969,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.341637941787433e-05,
+      "loss": 0.4745,
       "step": 4780
     },
     {
-      "epoch": 5.105656350053362,
-      "grad_norm": 0.369140625,
-      "learning_rate": 5.200046196980571e-05,
-      "loss": 0.4596,
+      "epoch": 4.896622313203685,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.318916860567447e-05,
+      "loss": 0.4742,
       "step": 4784
     },
     {
-      "epoch": 5.109925293489861,
-      "grad_norm": 0.33203125,
-      "learning_rate": 5.178086294686332e-05,
-      "loss": 0.4252,
+      "epoch": 4.9007164790174,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.296225689331523e-05,
+      "loss": 0.4955,
       "step": 4788
     },
     {
-      "epoch": 5.114194236926361,
+      "epoch": 4.904810644831116,
       "grad_norm": 0.345703125,
-      "learning_rate": 5.156163181978108e-05,
-      "loss": 0.415,
+      "learning_rate": 6.273564506260277e-05,
+      "loss": 0.4486,
       "step": 4792
     },
     {
-      "epoch": 5.11846318036286,
-      "grad_norm": 0.36328125,
-      "learning_rate": 5.134276940972513e-05,
-      "loss": 0.4309,
+      "epoch": 4.908904810644831,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.250933389431029e-05,
+      "loss": 0.4609,
       "step": 4796
     },
     {
-      "epoch": 5.122732123799359,
+      "epoch": 4.912998976458547,
       "grad_norm": 0.361328125,
-      "learning_rate": 5.112427653648039e-05,
-      "loss": 0.4162,
+      "learning_rate": 6.228332416817504e-05,
+      "loss": 0.4645,
       "step": 4800
     },
     {
-      "epoch": 5.1270010672358595,
-      "grad_norm": 0.359375,
-      "learning_rate": 5.090615401844774e-05,
-      "loss": 0.443,
+      "epoch": 4.917093142272262,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.205761666289548e-05,
+      "loss": 0.4939,
       "step": 4804
     },
     {
-      "epoch": 5.131270010672359,
-      "grad_norm": 0.369140625,
-      "learning_rate": 5.068840267264081e-05,
-      "loss": 0.4277,
+      "epoch": 4.921187308085978,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.183221215612904e-05,
+      "loss": 0.4646,
       "step": 4808
     },
     {
-      "epoch": 5.135538954108858,
-      "grad_norm": 0.35546875,
-      "learning_rate": 5.047102331468286e-05,
-      "loss": 0.4332,
+      "epoch": 4.925281473899693,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.16071114244891e-05,
+      "loss": 0.4811,
       "step": 4812
     },
     {
-      "epoch": 5.139807897545357,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.025401675880393e-05,
-      "loss": 0.4439,
+      "epoch": 4.929375639713409,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.138231524354229e-05,
+      "loss": 0.4903,
       "step": 4816
     },
     {
-      "epoch": 5.144076840981857,
-      "grad_norm": 0.359375,
-      "learning_rate": 5.003738381783754e-05,
-      "loss": 0.4421,
+      "epoch": 4.933469805527124,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.115782438780612e-05,
+      "loss": 0.4538,
       "step": 4820
     },
     {
-      "epoch": 5.148345784418357,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.982112530321791e-05,
-      "loss": 0.4589,
+      "epoch": 4.93756397134084,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.093363963074602e-05,
+      "loss": 0.4825,
       "step": 4824
     },
     {
-      "epoch": 5.152614727854856,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.9605242024976656e-05,
-      "loss": 0.4492,
+      "epoch": 4.941658137154555,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.070976174477281e-05,
+      "loss": 0.4081,
       "step": 4828
     },
     {
-      "epoch": 5.1568836712913555,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.938973479174002e-05,
-      "loss": 0.4433,
+      "epoch": 4.94575230296827,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.048619150124005e-05,
+      "loss": 0.4719,
       "step": 4832
     },
     {
-      "epoch": 5.161152614727855,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.917460441072552e-05,
-      "loss": 0.403,
+      "epoch": 4.949846468781986,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.026292967044121e-05,
+      "loss": 0.4626,
       "step": 4836
     },
     {
-      "epoch": 5.165421558164354,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.895985168773931e-05,
-      "loss": 0.4159,
+      "epoch": 4.9539406345957016,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.003997702160727e-05,
+      "loss": 0.4362,
       "step": 4840
     },
     {
-      "epoch": 5.169690501600853,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.8745477427172884e-05,
-      "loss": 0.4468,
+      "epoch": 4.958034800409417,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.981733432290399e-05,
+      "loss": 0.4964,
       "step": 4844
     },
     {
-      "epoch": 5.173959445037354,
-      "grad_norm": 0.388671875,
-      "learning_rate": 4.853148243200007e-05,
-      "loss": 0.4467,
+      "epoch": 4.962128966223132,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.959500234142904e-05,
+      "loss": 0.464,
       "step": 4848
     },
     {
-      "epoch": 5.178228388473853,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.831786750377425e-05,
-      "loss": 0.4175,
+      "epoch": 4.9662231320368475,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.937298184320967e-05,
+      "loss": 0.4479,
       "step": 4852
     },
     {
-      "epoch": 5.182497331910352,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.8104633442625026e-05,
-      "loss": 0.4225,
+      "epoch": 4.970317297850563,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.9151273593199924e-05,
+      "loss": 0.468,
       "step": 4856
     },
     {
-      "epoch": 5.1867662753468515,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.7891781047255564e-05,
-      "loss": 0.372,
+      "epoch": 4.974411463664278,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.892987835527809e-05,
+      "loss": 0.4376,
       "step": 4860
     },
     {
-      "epoch": 5.191035218783351,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.767931111493928e-05,
-      "loss": 0.4298,
+      "epoch": 4.9785056294779935,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.870879689224377e-05,
+      "loss": 0.441,
       "step": 4864
     },
     {
-      "epoch": 5.19530416221985,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.746722444151718e-05,
-      "loss": 0.4175,
+      "epoch": 4.982599795291709,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.84880299658157e-05,
+      "loss": 0.4317,
       "step": 4868
     },
     {
-      "epoch": 5.19957310565635,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.725552182139455e-05,
-      "loss": 0.3751,
+      "epoch": 4.986693961105424,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.8267578336628875e-05,
+      "loss": 0.4447,
       "step": 4872
     },
     {
-      "epoch": 5.20384204909285,
+      "epoch": 4.99078812691914,
       "grad_norm": 0.357421875,
-      "learning_rate": 4.704420404753825e-05,
-      "loss": 0.4366,
+      "learning_rate": 5.804744276423181e-05,
+      "loss": 0.4438,
       "step": 4876
     },
     {
-      "epoch": 5.208110992529349,
-      "grad_norm": 0.390625,
-      "learning_rate": 4.6833271911473674e-05,
-      "loss": 0.3922,
+      "epoch": 4.994882292732855,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.782762400708424e-05,
+      "loss": 0.4339,
       "step": 4880
     },
     {
-      "epoch": 5.212379935965848,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.6622726203281546e-05,
-      "loss": 0.4245,
+      "epoch": 4.998976458546571,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.760812282255433e-05,
+      "loss": 0.4601,
       "step": 4884
     },
     {
-      "epoch": 5.216648879402348,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.641256771159541e-05,
-      "loss": 0.4335,
+      "epoch": 5.003070624360286,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.7388939966915894e-05,
+      "loss": 0.4073,
       "step": 4888
     },
     {
-      "epoch": 5.220917822838848,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.62027972235982e-05,
-      "loss": 0.3568,
+      "epoch": 5.007164790174002,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.7170076195346144e-05,
+      "loss": 0.4749,
       "step": 4892
     },
     {
-      "epoch": 5.225186766275347,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.599341552501974e-05,
-      "loss": 0.4441,
+      "epoch": 5.011258955987717,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.695153226192293e-05,
+      "loss": 0.4516,
       "step": 4896
     },
     {
-      "epoch": 5.229455709711846,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.578442340013335e-05,
-      "loss": 0.4205,
+      "epoch": 5.015353121801433,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.6733308919621946e-05,
+      "loss": 0.4908,
       "step": 4900
     },
     {
-      "epoch": 5.233724653148346,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.557582163175336e-05,
-      "loss": 0.455,
+      "epoch": 5.019447287615148,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.651540692031448e-05,
+      "loss": 0.4131,
       "step": 4904
     },
     {
-      "epoch": 5.237993596584845,
-      "grad_norm": 0.357421875,
-      "learning_rate": 4.5367611001231755e-05,
-      "loss": 0.4524,
+      "epoch": 5.023541453428864,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.629782701476464e-05,
+      "loss": 0.4205,
       "step": 4908
     },
     {
-      "epoch": 5.242262540021345,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.515979228845562e-05,
-      "loss": 0.4531,
+      "epoch": 5.027635619242579,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.608056995262668e-05,
+      "loss": 0.4144,
       "step": 4912
     },
     {
-      "epoch": 5.2465314834578445,
-      "grad_norm": 0.34765625,
-      "learning_rate": 4.495236627184391e-05,
-      "loss": 0.4045,
+      "epoch": 5.031729785056295,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.586363648244261e-05,
+      "loss": 0.4081,
       "step": 4916
     },
     {
-      "epoch": 5.250800426894344,
-      "grad_norm": 0.380859375,
-      "learning_rate": 4.4745333728344804e-05,
-      "loss": 0.4356,
+      "epoch": 5.03582395087001,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.564702735163956e-05,
+      "loss": 0.4058,
       "step": 4920
     },
     {
-      "epoch": 5.255069370330843,
+      "epoch": 5.039918116683726,
       "grad_norm": 0.35546875,
-      "learning_rate": 4.453869543343257e-05,
-      "loss": 0.4126,
+      "learning_rate": 5.543074330652706e-05,
+      "loss": 0.4239,
       "step": 4924
     },
     {
-      "epoch": 5.259338313767342,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.433245216110468e-05,
-      "loss": 0.4596,
+      "epoch": 5.044012282497441,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.521478509229468e-05,
+      "loss": 0.4397,
       "step": 4928
     },
     {
-      "epoch": 5.263607257203842,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.4126604683879226e-05,
-      "loss": 0.4338,
+      "epoch": 5.048106448311157,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.499915345300936e-05,
+      "loss": 0.4284,
       "step": 4932
     },
     {
-      "epoch": 5.267876200640342,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.3921153772791526e-05,
-      "loss": 0.389,
+      "epoch": 5.052200614124872,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.478384913161277e-05,
+      "loss": 0.4047,
       "step": 4936
     },
     {
-      "epoch": 5.272145144076841,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.3716100197391704e-05,
-      "loss": 0.423,
+      "epoch": 5.056294779938588,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.456887286991891e-05,
+      "loss": 0.4198,
       "step": 4940
     },
     {
-      "epoch": 5.2764140875133405,
-      "grad_norm": 0.37109375,
-      "learning_rate": 4.3511444725741406e-05,
-      "loss": 0.4227,
+      "epoch": 5.060388945752303,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.435422540861151e-05,
+      "loss": 0.3665,
       "step": 4944
     },
     {
-      "epoch": 5.28068303094984,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.330718812441125e-05,
-      "loss": 0.4305,
+      "epoch": 5.064483111566019,
+      "grad_norm": 0.375,
+      "learning_rate": 5.413990748724129e-05,
+      "loss": 0.4719,
       "step": 4948
     },
     {
-      "epoch": 5.284951974386339,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.3103331158477855e-05,
-      "loss": 0.4355,
+      "epoch": 5.068577277379734,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.392591984422371e-05,
+      "loss": 0.4073,
       "step": 4952
     },
     {
-      "epoch": 5.289220917822838,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.289987459152073e-05,
-      "loss": 0.4461,
+      "epoch": 5.07267144319345,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.37122632168363e-05,
+      "loss": 0.3911,
       "step": 4956
     },
     {
-      "epoch": 5.293489861259339,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.2696819185619904e-05,
-      "loss": 0.4352,
+      "epoch": 5.076765609007165,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.349893834121593e-05,
+      "loss": 0.4298,
       "step": 4960
     },
     {
-      "epoch": 5.297758804695838,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.249416570135251e-05,
-      "loss": 0.4253,
+      "epoch": 5.080859774820881,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.32859459523566e-05,
+      "loss": 0.4137,
       "step": 4964
     },
     {
-      "epoch": 5.302027748132337,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.229191489779047e-05,
-      "loss": 0.4096,
+      "epoch": 5.084953940634596,
+      "grad_norm": 0.375,
+      "learning_rate": 5.3073286784106714e-05,
+      "loss": 0.439,
       "step": 4968
     },
     {
-      "epoch": 5.3062966915688365,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.20900675324972e-05,
-      "loss": 0.4336,
+      "epoch": 5.0890481064483115,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.2860961569166595e-05,
+      "loss": 0.3966,
       "step": 4972
     },
     {
-      "epoch": 5.310565635005336,
+      "epoch": 5.0931422722620265,
       "grad_norm": 0.33984375,
-      "learning_rate": 4.188862436152513e-05,
-      "loss": 0.4163,
+      "learning_rate": 5.264897103908599e-05,
+      "loss": 0.4607,
       "step": 4976
     },
     {
-      "epoch": 5.314834578441836,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.168758613941257e-05,
-      "loss": 0.4522,
+      "epoch": 5.0972364380757424,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.243731592426135e-05,
+      "loss": 0.4216,
       "step": 4980
     },
     {
-      "epoch": 5.319103521878335,
-      "grad_norm": 0.375,
-      "learning_rate": 4.148695361918117e-05,
-      "loss": 0.4418,
+      "epoch": 5.1013306038894575,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.222599695393368e-05,
+      "loss": 0.4576,
       "step": 4984
     },
     {
-      "epoch": 5.323372465314835,
-      "grad_norm": 0.3671875,
-      "learning_rate": 4.1286727552332796e-05,
-      "loss": 0.4095,
+      "epoch": 5.105424769703173,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.2015014856185796e-05,
+      "loss": 0.4706,
       "step": 4988
     },
     {
-      "epoch": 5.327641408751334,
-      "grad_norm": 0.369140625,
-      "learning_rate": 4.108690868884704e-05,
-      "loss": 0.4613,
+      "epoch": 5.109518935516888,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.1804370357939663e-05,
+      "loss": 0.4304,
       "step": 4992
     },
     {
-      "epoch": 5.331910352187833,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.088749777717818e-05,
-      "loss": 0.45,
+      "epoch": 5.113613101330604,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.15940641849543e-05,
+      "loss": 0.4476,
       "step": 4996
     },
     {
-      "epoch": 5.336179295624333,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.068849556425238e-05,
-      "loss": 0.4189,
+      "epoch": 5.117707267144319,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.138409706182299e-05,
+      "loss": 0.4314,
       "step": 5000
     },
     {
-      "epoch": 5.340448239060833,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.048990279546508e-05,
-      "loss": 0.4027,
+      "epoch": 5.121801432958035,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.1174469711970716e-05,
+      "loss": 0.4327,
       "step": 5004
     },
     {
-      "epoch": 5.344717182497332,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.029172021467794e-05,
-      "loss": 0.4234,
+      "epoch": 5.12589559877175,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.0965182857651964e-05,
+      "loss": 0.4191,
       "step": 5008
     },
     {
-      "epoch": 5.348986125933831,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.009394856421634e-05,
-      "loss": 0.438,
+      "epoch": 5.129989764585465,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.075623721994806e-05,
+      "loss": 0.4486,
       "step": 5012
     },
     {
-      "epoch": 5.353255069370331,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.989658858486628e-05,
-      "loss": 0.417,
+      "epoch": 5.134083930399181,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.05476335187645e-05,
+      "loss": 0.4605,
       "step": 5016
     },
     {
-      "epoch": 5.35752401280683,
-      "grad_norm": 0.369140625,
-      "learning_rate": 3.9699641015872016e-05,
-      "loss": 0.4407,
+      "epoch": 5.138178096212896,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.033937247282891e-05,
+      "loss": 0.4839,
       "step": 5020
     },
     {
-      "epoch": 5.36179295624333,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.9503106594932775e-05,
-      "loss": 0.4158,
+      "epoch": 5.142272262026612,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.013145479968824e-05,
+      "loss": 0.4389,
       "step": 5024
     },
     {
-      "epoch": 5.3660618996798295,
-      "grad_norm": 0.38671875,
-      "learning_rate": 3.93069860582005e-05,
-      "loss": 0.4035,
+      "epoch": 5.146366427840327,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.992388121570625e-05,
+      "loss": 0.4528,
       "step": 5028
     },
     {
-      "epoch": 5.370330843116329,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.9111280140276765e-05,
-      "loss": 0.4175,
+      "epoch": 5.150460593654043,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.9716652436061364e-05,
+      "loss": 0.4562,
       "step": 5032
     },
     {
-      "epoch": 5.374599786552828,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.89159895742101e-05,
-      "loss": 0.466,
+      "epoch": 5.154554759467758,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.950976917474393e-05,
+      "loss": 0.4521,
       "step": 5036
     },
     {
-      "epoch": 5.378868729989327,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.872111509149334e-05,
-      "loss": 0.422,
+      "epoch": 5.158648925281474,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.930323214455374e-05,
+      "loss": 0.3968,
       "step": 5040
     },
     {
-      "epoch": 5.383137673425827,
-      "grad_norm": 0.380859375,
-      "learning_rate": 3.8526657422060696e-05,
-      "loss": 0.4236,
+      "epoch": 5.162743091095189,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.909704205709785e-05,
+      "loss": 0.4224,
       "step": 5044
     },
     {
-      "epoch": 5.387406616862327,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.833261729428531e-05,
-      "loss": 0.4562,
+      "epoch": 5.166837256908905,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.889119962278786e-05,
+      "loss": 0.4055,
       "step": 5048
     },
     {
-      "epoch": 5.391675560298826,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.8138995434976147e-05,
-      "loss": 0.4252,
+      "epoch": 5.17093142272262,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.868570555083752e-05,
+      "loss": 0.4874,
       "step": 5052
     },
     {
-      "epoch": 5.3959445037353255,
-      "grad_norm": 0.330078125,
-      "learning_rate": 3.794579256937568e-05,
-      "loss": 0.421,
+      "epoch": 5.175025588536336,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.8480560549260394e-05,
+      "loss": 0.4371,
       "step": 5056
     },
     {
-      "epoch": 5.400213447171825,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.7753009421156776e-05,
-      "loss": 0.4225,
+      "epoch": 5.179119754350051,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.827576532486737e-05,
+      "loss": 0.4315,
       "step": 5060
     },
     {
-      "epoch": 5.404482390608324,
-      "grad_norm": 0.3671875,
-      "learning_rate": 3.756064671242039e-05,
-      "loss": 0.4786,
+      "epoch": 5.183213920163767,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.807132058326409e-05,
+      "loss": 0.4226,
       "step": 5064
     },
     {
-      "epoch": 5.408751334044824,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.736870516369245e-05,
-      "loss": 0.4224,
+      "epoch": 5.187308085977482,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.786722702884874e-05,
+      "loss": 0.4251,
       "step": 5068
     },
     {
-      "epoch": 5.413020277481324,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.7177185493921553e-05,
-      "loss": 0.4515,
+      "epoch": 5.191402251791198,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.766348536480954e-05,
+      "loss": 0.4411,
       "step": 5072
     },
     {
-      "epoch": 5.417289220917823,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.6986088420475885e-05,
-      "loss": 0.405,
+      "epoch": 5.195496417604913,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.7460096293122174e-05,
+      "loss": 0.3946,
       "step": 5076
     },
     {
-      "epoch": 5.421558164354322,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.6795414659140895e-05,
-      "loss": 0.4452,
+      "epoch": 5.199590583418629,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.725706051454759e-05,
+      "loss": 0.4129,
       "step": 5080
     },
     {
-      "epoch": 5.425827107790822,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.660516492411631e-05,
-      "loss": 0.4512,
+      "epoch": 5.203684749232344,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.705437872862955e-05,
+      "loss": 0.4273,
       "step": 5084
     },
     {
-      "epoch": 5.430096051227321,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.641533992801375e-05,
-      "loss": 0.4299,
+      "epoch": 5.20777891504606,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.685205163369197e-05,
+      "loss": 0.4656,
       "step": 5088
     },
     {
-      "epoch": 5.434364994663821,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.6225940381853726e-05,
-      "loss": 0.4892,
+      "epoch": 5.211873080859775,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.665007992683687e-05,
+      "loss": 0.4251,
       "step": 5092
     },
     {
-      "epoch": 5.43863393810032,
-      "grad_norm": 0.36328125,
-      "learning_rate": 3.6036966995063354e-05,
-      "loss": 0.4279,
+      "epoch": 5.2159672466734905,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.6448464303941824e-05,
+      "loss": 0.4032,
       "step": 5096
     },
     {
-      "epoch": 5.44290288153682,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.584842047547327e-05,
-      "loss": 0.4391,
+      "epoch": 5.220061412487206,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.6247205459657364e-05,
+      "loss": 0.4569,
       "step": 5100
     },
     {
-      "epoch": 5.447171824973319,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.566030152931547e-05,
-      "loss": 0.3981,
+      "epoch": 5.2241555783009215,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.604630408740498e-05,
+      "loss": 0.4433,
       "step": 5104
     },
     {
-      "epoch": 5.451440768409818,
-      "grad_norm": 0.37890625,
-      "learning_rate": 3.547261086122016e-05,
-      "loss": 0.4155,
+      "epoch": 5.2282497441146365,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.584576087937445e-05,
+      "loss": 0.4305,
       "step": 5108
     },
     {
-      "epoch": 5.455709711846318,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.528534917421355e-05,
-      "loss": 0.406,
+      "epoch": 5.232343909928352,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.5645576526521355e-05,
+      "loss": 0.3824,
       "step": 5112
     },
     {
-      "epoch": 5.459978655282818,
+      "epoch": 5.236438075742067,
       "grad_norm": 0.341796875,
-      "learning_rate": 3.509851716971487e-05,
-      "loss": 0.3866,
+      "learning_rate": 4.5445751718565165e-05,
+      "loss": 0.4235,
       "step": 5116
     },
     {
-      "epoch": 5.464247598719317,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.491211554753407e-05,
-      "loss": 0.4313,
+      "epoch": 5.240532241555783,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.524628714398645e-05,
+      "loss": 0.4215,
       "step": 5120
     },
     {
-      "epoch": 5.468516542155816,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.472614500586887e-05,
-      "loss": 0.4399,
+      "epoch": 5.244626407369498,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.504718349002447e-05,
+      "loss": 0.4246,
       "step": 5124
     },
     {
-      "epoch": 5.472785485592316,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.4540606241302464e-05,
-      "loss": 0.4731,
+      "epoch": 5.248720573183214,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.4848441442675154e-05,
+      "loss": 0.4044,
       "step": 5128
     },
     {
-      "epoch": 5.477054429028815,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.43554999488006e-05,
-      "loss": 0.4415,
+      "epoch": 5.252814738996929,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.4650061686688514e-05,
+      "loss": 0.4406,
       "step": 5132
     },
     {
-      "epoch": 5.481323372465315,
-      "grad_norm": 0.3828125,
-      "learning_rate": 3.4170826821709264e-05,
-      "loss": 0.4452,
+      "epoch": 5.256908904810645,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.445204490556618e-05,
+      "loss": 0.4196,
       "step": 5136
     },
     {
-      "epoch": 5.4855923159018145,
-      "grad_norm": 0.3671875,
-      "learning_rate": 3.398658755175183e-05,
-      "loss": 0.3971,
+      "epoch": 5.26100307062436,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.4254391781559336e-05,
+      "loss": 0.4163,
       "step": 5140
     },
     {
-      "epoch": 5.489861259338314,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.3802782829026685e-05,
-      "loss": 0.4456,
+      "epoch": 5.265097236438076,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.405710299566622e-05,
+      "loss": 0.454,
       "step": 5144
     },
     {
-      "epoch": 5.494130202774813,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.3619413342004556e-05,
-      "loss": 0.4326,
+      "epoch": 5.269191402251791,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.386017922762958e-05,
+      "loss": 0.441,
       "step": 5148
     },
     {
-      "epoch": 5.498399146211312,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.3436479777525796e-05,
-      "loss": 0.4265,
+      "epoch": 5.273285568065507,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.3663621155934724e-05,
+      "loss": 0.4363,
       "step": 5152
     },
     {
-      "epoch": 5.502668089647813,
-      "grad_norm": 0.3515625,
-      "learning_rate": 3.3253982820798115e-05,
-      "loss": 0.4393,
+      "epoch": 5.277379733879222,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.3467429457806965e-05,
+      "loss": 0.4327,
       "step": 5156
     },
     {
-      "epoch": 5.506937033084312,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.3071923155393656e-05,
-      "loss": 0.4317,
+      "epoch": 5.281473899692937,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.327160480920915e-05,
+      "loss": 0.4212,
       "step": 5160
     },
     {
-      "epoch": 5.511205976520811,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.289030146324681e-05,
-      "loss": 0.3967,
+      "epoch": 5.285568065506653,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.307614788483963e-05,
+      "loss": 0.4076,
       "step": 5164
     },
     {
-      "epoch": 5.5154749199573105,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.270911842465124e-05,
-      "loss": 0.4288,
+      "epoch": 5.289662231320369,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.2881059358129806e-05,
+      "loss": 0.4527,
       "step": 5168
     },
     {
-      "epoch": 5.51974386339381,
-      "grad_norm": 0.35546875,
-      "learning_rate": 3.252837471825779e-05,
-      "loss": 0.4006,
+      "epoch": 5.293756397134084,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.268633990124163e-05,
+      "loss": 0.4017,
       "step": 5172
     },
     {
-      "epoch": 5.524012806830309,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.234807102107153e-05,
-      "loss": 0.4299,
+      "epoch": 5.297850562947799,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.2491990185065625e-05,
+      "loss": 0.392,
       "step": 5176
     },
     {
-      "epoch": 5.528281750266809,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.216820800844952e-05,
-      "loss": 0.454,
+      "epoch": 5.301944728761515,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.22980108792184e-05,
+      "loss": 0.4646,
       "step": 5180
     },
     {
-      "epoch": 5.532550693703309,
-      "grad_norm": 0.353515625,
-      "learning_rate": 3.198878635409814e-05,
-      "loss": 0.4166,
+      "epoch": 5.30603889457523,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.2104402652040144e-05,
+      "loss": 0.4128,
       "step": 5184
     },
     {
-      "epoch": 5.536819637139808,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.180980673007055e-05,
-      "loss": 0.3965,
+      "epoch": 5.310133060388946,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.191116617059272e-05,
+      "loss": 0.4126,
       "step": 5188
     },
     {
-      "epoch": 5.541088580576307,
-      "grad_norm": 0.328125,
-      "learning_rate": 3.163126980676425e-05,
-      "loss": 0.4384,
+      "epoch": 5.314227226202661,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.1718302100657176e-05,
+      "loss": 0.4295,
       "step": 5192
     },
     {
-      "epoch": 5.545357524012807,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.1453176252918495e-05,
-      "loss": 0.4153,
+      "epoch": 5.318321392016377,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.15258111067313e-05,
+      "loss": 0.4273,
       "step": 5196
     },
     {
-      "epoch": 5.549626467449306,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.1275526735611896e-05,
-      "loss": 0.4423,
+      "epoch": 5.322415557830092,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.133369385202756e-05,
+      "loss": 0.42,
       "step": 5200
     },
     {
-      "epoch": 5.553895410885806,
-      "grad_norm": 0.37109375,
-      "learning_rate": 3.1098321920259736e-05,
-      "loss": 0.4317,
+      "epoch": 5.326509723643808,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.114195099847083e-05,
+      "loss": 0.4329,
       "step": 5204
     },
     {
-      "epoch": 5.558164354322305,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.0921562470611765e-05,
-      "loss": 0.4189,
+      "epoch": 5.330603889457523,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.0950583206695786e-05,
+      "loss": 0.4637,
       "step": 5208
     },
     {
-      "epoch": 5.562433297758805,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.074524904874935e-05,
-      "loss": 0.4512,
+      "epoch": 5.334698055271239,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.075959113604506e-05,
+      "loss": 0.4401,
       "step": 5212
     },
     {
-      "epoch": 5.566702241195304,
-      "grad_norm": 0.369140625,
-      "learning_rate": 3.056938231508333e-05,
-      "loss": 0.433,
+      "epoch": 5.338792221084954,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.056897544456673e-05,
+      "loss": 0.4225,
       "step": 5216
     },
     {
-      "epoch": 5.570971184631803,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.0393962928351373e-05,
-      "loss": 0.3729,
+      "epoch": 5.34288638689867,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.037873678901198e-05,
+      "loss": 0.4322,
       "step": 5220
     },
     {
-      "epoch": 5.5752401280683035,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.021899154561545e-05,
-      "loss": 0.4289,
+      "epoch": 5.346980552712385,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.018887582483306e-05,
+      "loss": 0.4224,
       "step": 5224
     },
     {
-      "epoch": 5.579509071504803,
-      "grad_norm": 0.361328125,
-      "learning_rate": 3.0044468822259594e-05,
-      "loss": 0.4024,
+      "epoch": 5.3510747185261005,
+      "grad_norm": 0.375,
+      "learning_rate": 3.9999393206180914e-05,
+      "loss": 0.4694,
       "step": 5228
     },
     {
-      "epoch": 5.583778014941302,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.987039541198719e-05,
-      "loss": 0.4107,
+      "epoch": 5.3551688843398155,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.98102895859028e-05,
+      "loss": 0.4165,
       "step": 5232
     },
     {
-      "epoch": 5.588046958377801,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.9696771966818735e-05,
-      "loss": 0.3989,
+      "epoch": 5.359263050153531,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.9621565615540325e-05,
+      "loss": 0.4271,
       "step": 5236
     },
     {
-      "epoch": 5.592315901814301,
-      "grad_norm": 0.35546875,
-      "learning_rate": 2.952359913708927e-05,
-      "loss": 0.3772,
+      "epoch": 5.3633572159672465,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.9433221945326985e-05,
+      "loss": 0.4569,
       "step": 5240
     },
     {
-      "epoch": 5.596584845250801,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.9350877571445924e-05,
-      "loss": 0.4425,
+      "epoch": 5.367451381780962,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.924525922418591e-05,
+      "loss": 0.4756,
       "step": 5244
     },
     {
-      "epoch": 5.6008537886873,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.91786079168457e-05,
-      "loss": 0.4352,
+      "epoch": 5.371545547594677,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.905767809972779e-05,
+      "loss": 0.4734,
       "step": 5248
     },
     {
-      "epoch": 5.6051227321237995,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.900679081855268e-05,
-      "loss": 0.4541,
+      "epoch": 5.375639713408393,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.887047921824858e-05,
+      "loss": 0.3928,
       "step": 5252
     },
     {
-      "epoch": 5.609391675560299,
-      "grad_norm": 0.36328125,
-      "learning_rate": 2.883542692013606e-05,
-      "loss": 0.4505,
+      "epoch": 5.379733879222108,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.868366322472704e-05,
+      "loss": 0.4826,
       "step": 5256
     },
     {
-      "epoch": 5.613660618996798,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.866451686346729e-05,
-      "loss": 0.4272,
+      "epoch": 5.383828045035824,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.849723076282308e-05,
+      "loss": 0.4614,
       "step": 5260
     },
     {
-      "epoch": 5.617929562433297,
-      "grad_norm": 0.388671875,
-      "learning_rate": 2.8494061288718033e-05,
-      "loss": 0.4172,
+      "epoch": 5.387922210849539,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.831118247487481e-05,
+      "loss": 0.4696,
       "step": 5264
     },
     {
-      "epoch": 5.622198505869797,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.8324060834357505e-05,
-      "loss": 0.4151,
+      "epoch": 5.392016376663255,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.812551900189694e-05,
+      "loss": 0.4323,
       "step": 5268
     },
     {
-      "epoch": 5.626467449306297,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.815451613715029e-05,
-      "loss": 0.3926,
+      "epoch": 5.39611054247697,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.794024098357826e-05,
+      "loss": 0.3972,
       "step": 5272
     },
     {
-      "epoch": 5.630736392742796,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.798542783215375e-05,
-      "loss": 0.463,
+      "epoch": 5.400204708290686,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.775534905827943e-05,
+      "loss": 0.4397,
       "step": 5276
     },
     {
-      "epoch": 5.6350053361792956,
-      "grad_norm": 0.3671875,
-      "learning_rate": 2.7816796552715893e-05,
-      "loss": 0.444,
+      "epoch": 5.404298874104401,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.7570843863030995e-05,
+      "loss": 0.426,
       "step": 5280
     },
     {
-      "epoch": 5.639274279615795,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.7648622930472725e-05,
-      "loss": 0.3938,
+      "epoch": 5.408393039918117,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.7386726033530995e-05,
+      "loss": 0.4399,
       "step": 5284
     },
     {
-      "epoch": 5.643543223052294,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.7480907595346098e-05,
-      "loss": 0.4214,
+      "epoch": 5.412487205731832,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.720299620414274e-05,
+      "loss": 0.4422,
       "step": 5288
     },
     {
-      "epoch": 5.647812166488794,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.7313651175541312e-05,
-      "loss": 0.4117,
+      "epoch": 5.416581371545548,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.701965500789287e-05,
+      "loss": 0.4196,
       "step": 5292
     },
     {
-      "epoch": 5.652081109925294,
-      "grad_norm": 0.35546875,
-      "learning_rate": 2.7146854297544602e-05,
-      "loss": 0.429,
+      "epoch": 5.420675537359263,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.68367030764689e-05,
+      "loss": 0.432,
       "step": 5296
     },
     {
-      "epoch": 5.656350053361793,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.698051758612107e-05,
-      "loss": 0.4168,
+      "epoch": 5.424769703172979,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.665414104021729e-05,
+      "loss": 0.448,
       "step": 5300
     },
     {
-      "epoch": 5.660618996798292,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.6814641664312004e-05,
-      "loss": 0.4305,
+      "epoch": 5.428863868986694,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.647196952814099e-05,
+      "loss": 0.435,
       "step": 5304
     },
     {
-      "epoch": 5.664887940234792,
-      "grad_norm": 0.341796875,
-      "learning_rate": 2.6649227153432916e-05,
-      "loss": 0.4034,
+      "epoch": 5.432958034800409,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.6290189167897526e-05,
+      "loss": 0.4299,
       "step": 5308
     },
     {
-      "epoch": 5.669156883671292,
-      "grad_norm": 0.376953125,
-      "learning_rate": 2.648427467307086e-05,
-      "loss": 0.4394,
+      "epoch": 5.437052200614125,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.6108800585796774e-05,
+      "loss": 0.4619,
       "step": 5312
     },
     {
-      "epoch": 5.673425827107791,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.6319784841082464e-05,
-      "loss": 0.4339,
+      "epoch": 5.441146366427841,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.5927804406798655e-05,
+      "loss": 0.4072,
       "step": 5316
     },
     {
-      "epoch": 5.67769477054429,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.6155758273591215e-05,
-      "loss": 0.4682,
+      "epoch": 5.445240532241556,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.574720125451118e-05,
+      "loss": 0.425,
       "step": 5320
     },
     {
-      "epoch": 5.68196371398079,
+      "epoch": 5.449334698055271,
       "grad_norm": 0.3515625,
-      "learning_rate": 2.599219558498557e-05,
-      "loss": 0.4467,
+      "learning_rate": 3.556699175118824e-05,
+      "loss": 0.3989,
       "step": 5324
     },
     {
-      "epoch": 5.686232657417289,
-      "grad_norm": 0.375,
-      "learning_rate": 2.5829097387916294e-05,
-      "loss": 0.4677,
+      "epoch": 5.453428863868987,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.538717651772733e-05,
+      "loss": 0.4356,
       "step": 5328
     },
     {
-      "epoch": 5.690501600853788,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.5666464293294418e-05,
-      "loss": 0.4077,
+      "epoch": 5.457523029682702,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.520775617366763e-05,
+      "loss": 0.4129,
       "step": 5332
     },
     {
-      "epoch": 5.6947705442902885,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.5504296910288868e-05,
-      "loss": 0.4685,
+      "epoch": 5.461617195496418,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.502873133718775e-05,
+      "loss": 0.4549,
       "step": 5336
     },
     {
-      "epoch": 5.699039487726788,
-      "grad_norm": 0.380859375,
-      "learning_rate": 2.534259584632405e-05,
-      "loss": 0.4629,
+      "epoch": 5.465711361310133,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.4850102625103504e-05,
+      "loss": 0.436,
       "step": 5340
     },
     {
-      "epoch": 5.703308431163287,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.5181361707077852e-05,
-      "loss": 0.4192,
+      "epoch": 5.469805527123849,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.4671870652866e-05,
+      "loss": 0.4197,
       "step": 5344
     },
     {
-      "epoch": 5.707577374599786,
-      "grad_norm": 0.39453125,
-      "learning_rate": 2.5020595096479084e-05,
-      "loss": 0.4105,
+      "epoch": 5.473899692937564,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.449403603455941e-05,
+      "loss": 0.4587,
       "step": 5348
     },
     {
-      "epoch": 5.711846318036286,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.4860296616705457e-05,
-      "loss": 0.4562,
+      "epoch": 5.4779938587512795,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.431659938289875e-05,
+      "loss": 0.409,
       "step": 5352
     },
     {
-      "epoch": 5.716115261472785,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.4700466868181106e-05,
-      "loss": 0.4119,
+      "epoch": 5.4820880245649946,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.413956130922797e-05,
+      "loss": 0.3965,
       "step": 5356
     },
     {
-      "epoch": 5.720384204909285,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.454110644957462e-05,
-      "loss": 0.4017,
+      "epoch": 5.4861821903787105,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.396292242351779e-05,
+      "loss": 0.4204,
       "step": 5360
     },
     {
-      "epoch": 5.7246531483457845,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.4382215957796454e-05,
-      "loss": 0.4441,
+      "epoch": 5.4902763561924255,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.3786683334363376e-05,
+      "loss": 0.4489,
       "step": 5364
     },
     {
-      "epoch": 5.728922091782284,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.4223795987996997e-05,
-      "loss": 0.4319,
+      "epoch": 5.494370522006141,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.36108446489826e-05,
+      "loss": 0.4325,
       "step": 5368
     },
     {
-      "epoch": 5.733191035218783,
+      "epoch": 5.498464687819856,
       "grad_norm": 0.33984375,
-      "learning_rate": 2.406584713356421e-05,
-      "loss": 0.4403,
+      "learning_rate": 3.3435406973213777e-05,
+      "loss": 0.4531,
       "step": 5372
     },
     {
-      "epoch": 5.7374599786552825,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.390836998612134e-05,
-      "loss": 0.4331,
+      "epoch": 5.502558853633572,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.326037091151342e-05,
+      "loss": 0.4556,
       "step": 5376
     },
     {
-      "epoch": 5.741728922091783,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.3751365135524887e-05,
-      "loss": 0.4297,
+      "epoch": 5.506653019447287,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.308573706695445e-05,
+      "loss": 0.4505,
       "step": 5380
     },
     {
-      "epoch": 5.745997865528282,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.359483316986216e-05,
-      "loss": 0.4406,
+      "epoch": 5.510747185261003,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.2911506041223984e-05,
+      "loss": 0.4272,
       "step": 5384
     },
     {
-      "epoch": 5.750266808964781,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.343877467544934e-05,
-      "loss": 0.4002,
+      "epoch": 5.514841351074718,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.273767843462113e-05,
+      "loss": 0.4392,
       "step": 5388
     },
     {
-      "epoch": 5.754535752401281,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.3283190236828993e-05,
-      "loss": 0.4169,
+      "epoch": 5.518935516888434,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.2564254846055177e-05,
+      "loss": 0.4276,
       "step": 5392
     },
     {
-      "epoch": 5.75880469583778,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.312808043676818e-05,
-      "loss": 0.4487,
+      "epoch": 5.523029682702149,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.2391235873043396e-05,
+      "loss": 0.4145,
       "step": 5396
     },
     {
-      "epoch": 5.76307363927428,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.2973445856255983e-05,
-      "loss": 0.4472,
+      "epoch": 5.527123848515865,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.221862211170883e-05,
+      "loss": 0.3965,
       "step": 5400
     },
     {
-      "epoch": 5.767342582710779,
-      "grad_norm": 0.341796875,
-      "learning_rate": 2.2819287074501614e-05,
-      "loss": 0.4155,
+      "epoch": 5.53121801432958,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.2046414156778674e-05,
+      "loss": 0.4354,
       "step": 5404
     },
     {
-      "epoch": 5.771611526147279,
-      "grad_norm": 0.37109375,
-      "learning_rate": 2.2665604668931974e-05,
-      "loss": 0.3958,
+      "epoch": 5.535312180143296,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.187461260158166e-05,
+      "loss": 0.4275,
       "step": 5408
     },
     {
-      "epoch": 5.775880469583778,
-      "grad_norm": 0.36328125,
-      "learning_rate": 2.2512399215189735e-05,
-      "loss": 0.4213,
+      "epoch": 5.539406345957011,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.170321803804649e-05,
+      "loss": 0.4575,
       "step": 5412
     },
     {
-      "epoch": 5.780149413020277,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.2359671287130975e-05,
-      "loss": 0.4596,
+      "epoch": 5.543500511770727,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.153223105669957e-05,
+      "loss": 0.3868,
       "step": 5416
     },
     {
-      "epoch": 5.784418356456777,
-      "grad_norm": 0.68359375,
-      "learning_rate": 2.220742145682312e-05,
-      "loss": 0.4457,
+      "epoch": 5.547594677584442,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.1361652246662944e-05,
+      "loss": 0.4461,
       "step": 5420
     },
     {
-      "epoch": 5.788687299893277,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.205565029454293e-05,
-      "loss": 0.4362,
+      "epoch": 5.551688843398158,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.1191482195652405e-05,
+      "loss": 0.4191,
       "step": 5424
     },
     {
-      "epoch": 5.792956243329776,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.1904358368774034e-05,
-      "loss": 0.4062,
+      "epoch": 5.555783009211873,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.102172148997543e-05,
+      "loss": 0.4842,
       "step": 5428
     },
     {
-      "epoch": 5.797225186766275,
-      "grad_norm": 0.359375,
-      "learning_rate": 2.1753546246205224e-05,
-      "loss": 0.4612,
+      "epoch": 5.559877175025589,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.085237071452898e-05,
+      "loss": 0.439,
       "step": 5432
     },
     {
-      "epoch": 5.801494130202775,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.16032144917279e-05,
-      "loss": 0.422,
+      "epoch": 5.563971340839304,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.068343045279779e-05,
+      "loss": 0.4481,
       "step": 5436
     },
     {
-      "epoch": 5.805763073639274,
-      "grad_norm": 0.330078125,
-      "learning_rate": 2.1453363668434304e-05,
-      "loss": 0.4251,
+      "epoch": 5.56806550665302,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.0514901286852177e-05,
+      "loss": 0.3919,
       "step": 5440
     },
     {
-      "epoch": 5.810032017075773,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.1303994337615283e-05,
-      "loss": 0.4051,
+      "epoch": 5.572159672466735,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.0346783797345936e-05,
+      "loss": 0.4635,
       "step": 5444
     },
     {
-      "epoch": 5.8143009605122735,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.1155107058758024e-05,
-      "loss": 0.4795,
+      "epoch": 5.576253838280451,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.0179078563514625e-05,
+      "loss": 0.3848,
       "step": 5448
     },
     {
-      "epoch": 5.818569903948773,
+      "epoch": 5.580348004094166,
       "grad_norm": 0.353515625,
-      "learning_rate": 2.1006702389544315e-05,
-      "loss": 0.4346,
+      "learning_rate": 3.0011786163173357e-05,
+      "loss": 0.4465,
       "step": 5452
     },
     {
-      "epoch": 5.822838847385272,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.0858780885848035e-05,
-      "loss": 0.4167,
+      "epoch": 5.584442169907881,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9844907172714767e-05,
+      "loss": 0.4399,
       "step": 5456
     },
     {
-      "epoch": 5.827107790821771,
-      "grad_norm": 0.365234375,
-      "learning_rate": 2.0711343101733496e-05,
-      "loss": 0.4471,
+      "epoch": 5.588536335721597,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.9678442167107242e-05,
+      "loss": 0.475,
       "step": 5460
     },
     {
-      "epoch": 5.831376734258271,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.056438958945299e-05,
-      "loss": 0.4174,
+      "epoch": 5.592630501535313,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.951239171989278e-05,
+      "loss": 0.4141,
       "step": 5464
     },
     {
-      "epoch": 5.835645677694771,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.0417920899445013e-05,
-      "loss": 0.4145,
+      "epoch": 5.596724667349028,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.9346756403184974e-05,
+      "loss": 0.4536,
       "step": 5468
     },
     {
-      "epoch": 5.83991462113127,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.027193758033196e-05,
-      "loss": 0.4621,
+      "epoch": 5.600818833162743,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9181536787667237e-05,
+      "loss": 0.4386,
       "step": 5472
     },
     {
-      "epoch": 5.8441835645677696,
-      "grad_norm": 0.373046875,
-      "learning_rate": 2.0126440178918285e-05,
-      "loss": 0.4449,
+      "epoch": 5.604912998976459,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.9016733442590683e-05,
+      "loss": 0.4425,
       "step": 5476
     },
     {
-      "epoch": 5.848452508004269,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.998142924018838e-05,
-      "loss": 0.4257,
+      "epoch": 5.6090071647901745,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.8852346935772107e-05,
+      "loss": 0.4275,
       "step": 5480
     },
     {
-      "epoch": 5.852721451440768,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.983690530730439e-05,
-      "loss": 0.425,
+      "epoch": 5.6131013306038895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.868837783359222e-05,
+      "loss": 0.4468,
       "step": 5484
     },
     {
-      "epoch": 5.856990394877268,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.9692868921604425e-05,
-      "loss": 0.3803,
+      "epoch": 5.6171954964176045,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8524826700993625e-05,
+      "loss": 0.4208,
       "step": 5488
     },
     {
-      "epoch": 5.861259338313768,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.954932062260029e-05,
-      "loss": 0.4228,
+      "epoch": 5.62128966223132,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8361694101478704e-05,
+      "loss": 0.4079,
       "step": 5492
     },
     {
-      "epoch": 5.865528281750267,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.940626094797571e-05,
-      "loss": 0.3995,
+      "epoch": 5.6253838280450355,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.8198980597107956e-05,
+      "loss": 0.4446,
       "step": 5496
     },
     {
-      "epoch": 5.869797225186766,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.9263690433584066e-05,
-      "loss": 0.4417,
+      "epoch": 5.629477993858751,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.8036686748497883e-05,
+      "loss": 0.4611,
       "step": 5500
     },
     {
-      "epoch": 5.874066168623266,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.9121609613446605e-05,
-      "loss": 0.411,
+      "epoch": 5.633572159672466,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.7874813114818994e-05,
+      "loss": 0.4355,
       "step": 5504
     },
     {
-      "epoch": 5.878335112059765,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.8980019019750264e-05,
-      "loss": 0.4317,
+      "epoch": 5.637666325486182,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.7713360253794143e-05,
+      "loss": 0.4381,
       "step": 5508
     },
     {
-      "epoch": 5.882604055496264,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.8838919182845847e-05,
-      "loss": 0.462,
+      "epoch": 5.641760491299897,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.7552328721696388e-05,
+      "loss": 0.4526,
       "step": 5512
     },
     {
-      "epoch": 5.886872998932764,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.8698310631245816e-05,
-      "loss": 0.4091,
+      "epoch": 5.645854657113613,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.7391719073347018e-05,
+      "loss": 0.4705,
       "step": 5516
     },
     {
-      "epoch": 5.891141942369264,
-      "grad_norm": 0.375,
-      "learning_rate": 1.8558193891622597e-05,
-      "loss": 0.408,
+      "epoch": 5.649948822927328,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.723153186211392e-05,
+      "loss": 0.4437,
       "step": 5520
     },
     {
-      "epoch": 5.895410885805763,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.8418569488806385e-05,
-      "loss": 0.4194,
+      "epoch": 5.654042988741044,
+      "grad_norm": 0.375,
+      "learning_rate": 2.707176763990944e-05,
+      "loss": 0.438,
       "step": 5524
     },
     {
-      "epoch": 5.899679829242262,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.8279437945783166e-05,
-      "loss": 0.4776,
+      "epoch": 5.658137154554759,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.69124269571885e-05,
+      "loss": 0.4156,
       "step": 5528
     },
     {
-      "epoch": 5.903948772678762,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.814079978369302e-05,
-      "loss": 0.4469,
+      "epoch": 5.662231320368475,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.6753510362946823e-05,
+      "loss": 0.4496,
       "step": 5532
     },
     {
-      "epoch": 5.908217716115262,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.800265552182778e-05,
-      "loss": 0.5029,
+      "epoch": 5.66632548618219,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.659501840471898e-05,
+      "loss": 0.4761,
       "step": 5536
     },
     {
-      "epoch": 5.912486659551761,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.7865005677629495e-05,
-      "loss": 0.4646,
+      "epoch": 5.670419651995906,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.643695162857638e-05,
+      "loss": 0.4506,
       "step": 5540
     },
     {
-      "epoch": 5.91675560298826,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.7727850766688108e-05,
-      "loss": 0.4514,
+      "epoch": 5.674513817809621,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.6279310579125562e-05,
+      "loss": 0.4432,
       "step": 5544
     },
     {
-      "epoch": 5.92102454642476,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.7591191302739866e-05,
-      "loss": 0.4539,
+      "epoch": 5.678607983623337,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.6122095799506394e-05,
+      "loss": 0.4281,
       "step": 5548
     },
     {
-      "epoch": 5.925293489861259,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.745502779766511e-05,
-      "loss": 0.4462,
+      "epoch": 5.682702149437052,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.596530783138983e-05,
+      "loss": 0.4595,
       "step": 5552
     },
     {
-      "epoch": 5.929562433297759,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.7319360761486566e-05,
-      "loss": 0.4291,
+      "epoch": 5.686796315250768,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.5808947214976428e-05,
+      "loss": 0.4131,
       "step": 5556
     },
     {
-      "epoch": 5.9338313767342585,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.7184190702367377e-05,
-      "loss": 0.4539,
+      "epoch": 5.690890481064483,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.5653014488994328e-05,
+      "loss": 0.4694,
       "step": 5560
     },
     {
-      "epoch": 5.938100320170758,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.704951812660908e-05,
-      "loss": 0.433,
+      "epoch": 5.694984646878199,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.5497510190697323e-05,
+      "loss": 0.4834,
       "step": 5564
     },
     {
-      "epoch": 5.942369263607257,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.6915343538649845e-05,
-      "loss": 0.5016,
+      "epoch": 5.699078812691914,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.5342434855863187e-05,
+      "loss": 0.4144,
       "step": 5568
     },
     {
-      "epoch": 5.9466382070437565,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.6781667441062607e-05,
-      "loss": 0.3977,
+      "epoch": 5.70317297850563,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.5187789018791703e-05,
+      "loss": 0.429,
       "step": 5572
     },
     {
-      "epoch": 5.950907150480256,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.6648490334553045e-05,
-      "loss": 0.4445,
+      "epoch": 5.707267144319345,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.5033573212302814e-05,
+      "loss": 0.4901,
       "step": 5576
     },
     {
-      "epoch": 5.955176093916756,
-      "grad_norm": 0.373046875,
-      "learning_rate": 1.6515812717957843e-05,
-      "loss": 0.4533,
+      "epoch": 5.711361310133061,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4879787967734878e-05,
+      "loss": 0.39,
       "step": 5580
     },
     {
-      "epoch": 5.959445037353255,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.638363508824271e-05,
-      "loss": 0.4409,
+      "epoch": 5.715455475946776,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.4726433814942814e-05,
+      "loss": 0.3819,
       "step": 5584
     },
     {
-      "epoch": 5.963713980789755,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.6251957940500636e-05,
-      "loss": 0.4325,
+      "epoch": 5.719549641760492,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.4573511282296126e-05,
+      "loss": 0.4443,
       "step": 5588
     },
     {
-      "epoch": 5.967982924226254,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.6120781767949897e-05,
-      "loss": 0.4231,
+      "epoch": 5.723643807574207,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4421020896677318e-05,
+      "loss": 0.4127,
       "step": 5592
     },
     {
-      "epoch": 5.972251867662753,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.5990107061932373e-05,
-      "loss": 0.396,
+      "epoch": 5.727737973387923,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4268963183479967e-05,
+      "loss": 0.488,
       "step": 5596
     },
     {
-      "epoch": 5.9765208110992525,
+      "epoch": 5.731832139201638,
       "grad_norm": 0.36328125,
-      "learning_rate": 1.5859934311911514e-05,
-      "loss": 0.4448,
+      "learning_rate": 2.4117338666606796e-05,
+      "loss": 0.4206,
       "step": 5600
     },
     {
-      "epoch": 5.980789754535753,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.5730264005470734e-05,
-      "loss": 0.4321,
+      "epoch": 5.7359263050153535,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.39661478684681e-05,
+      "loss": 0.3882,
       "step": 5604
     },
     {
-      "epoch": 5.985058697972252,
-      "grad_norm": 0.375,
-      "learning_rate": 1.560109662831134e-05,
-      "loss": 0.4532,
+      "epoch": 5.7400204708290685,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.3815391309979843e-05,
+      "loss": 0.3997,
       "step": 5608
     },
     {
-      "epoch": 5.989327641408751,
-      "grad_norm": 0.373046875,
-      "learning_rate": 1.547243266425095e-05,
-      "loss": 0.4304,
+      "epoch": 5.744114636642784,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.366506951056173e-05,
+      "loss": 0.4571,
       "step": 5612
     },
     {
-      "epoch": 5.993596584845251,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.5344272595221448e-05,
-      "loss": 0.415,
+      "epoch": 5.7482088024564995,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.3515182988135618e-05,
+      "loss": 0.4476,
       "step": 5616
     },
     {
-      "epoch": 5.99786552828175,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.5216616901267398e-05,
-      "loss": 0.4507,
+      "epoch": 5.7523029682702145,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.336573225912371e-05,
+      "loss": 0.4467,
       "step": 5620
     },
     {
-      "epoch": 6.00213447171825,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.5089466060544086e-05,
-      "loss": 0.4352,
+      "epoch": 5.75639713408393,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3216717838446565e-05,
+      "loss": 0.3617,
       "step": 5624
     },
     {
-      "epoch": 6.006403415154749,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.4962820549315808e-05,
-      "loss": 0.4194,
+      "epoch": 5.760491299897646,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3068140239521588e-05,
+      "loss": 0.4339,
       "step": 5628
     },
     {
-      "epoch": 6.010672358591249,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.4836680841954085e-05,
-      "loss": 0.4156,
+      "epoch": 5.764585465711361,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.2919999974261177e-05,
+      "loss": 0.4,
       "step": 5632
     },
     {
-      "epoch": 6.014941302027748,
-      "grad_norm": 0.349609375,
-      "learning_rate": 1.471104741093581e-05,
-      "loss": 0.4571,
+      "epoch": 5.768679631525076,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.2772297553070784e-05,
+      "loss": 0.4217,
       "step": 5636
     },
     {
-      "epoch": 6.019210245464247,
-      "grad_norm": 0.349609375,
-      "learning_rate": 1.4585920726841627e-05,
-      "loss": 0.4524,
+      "epoch": 5.772773797338792,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.262503348484745e-05,
+      "loss": 0.4739,
       "step": 5640
     },
     {
-      "epoch": 6.023479188900747,
-      "grad_norm": 0.326171875,
-      "learning_rate": 1.446130125835394e-05,
-      "loss": 0.4418,
+      "epoch": 5.776867963152508,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.247820827697789e-05,
+      "loss": 0.4524,
       "step": 5644
     },
     {
-      "epoch": 6.027748132337247,
-      "grad_norm": 0.373046875,
-      "learning_rate": 1.4337189472255434e-05,
-      "loss": 0.4257,
+      "epoch": 5.780962128966223,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.2331822435336644e-05,
+      "loss": 0.418,
       "step": 5648
     },
     {
-      "epoch": 6.032017075773746,
-      "grad_norm": 0.330078125,
-      "learning_rate": 1.4213585833427089e-05,
-      "loss": 0.3953,
+      "epoch": 5.785056294779938,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.2185876464284554e-05,
+      "loss": 0.4336,
       "step": 5652
     },
     {
-      "epoch": 6.036286019210245,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.409049080484661e-05,
-      "loss": 0.497,
+      "epoch": 5.789150460593654,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.2040370866666945e-05,
+      "loss": 0.4555,
       "step": 5656
     },
     {
-      "epoch": 6.040554962646745,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.396790484758653e-05,
-      "loss": 0.4486,
+      "epoch": 5.793244626407369,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.1895306143811768e-05,
+      "loss": 0.4541,
       "step": 5660
     },
     {
-      "epoch": 6.044823906083244,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.3845828420812672e-05,
-      "loss": 0.4221,
+      "epoch": 5.797338792221085,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.175068279552805e-05,
+      "loss": 0.4081,
       "step": 5664
     },
     {
-      "epoch": 6.049092849519744,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.372426198178228e-05,
-      "loss": 0.3952,
+      "epoch": 5.8014329580348,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.16065013201041e-05,
+      "loss": 0.4147,
       "step": 5668
     },
     {
-      "epoch": 6.0533617929562435,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.3603205985842319e-05,
-      "loss": 0.4418,
+      "epoch": 5.805527123848516,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.146276221430572e-05,
+      "loss": 0.423,
       "step": 5672
     },
     {
-      "epoch": 6.057630736392743,
-      "grad_norm": 0.380859375,
-      "learning_rate": 1.3482660886427893e-05,
-      "loss": 0.3987,
+      "epoch": 5.809621289662231,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.131946597337463e-05,
+      "loss": 0.4226,
       "step": 5676
     },
     {
-      "epoch": 6.061899679829242,
-      "grad_norm": 0.6328125,
-      "learning_rate": 1.3362627135060356e-05,
-      "loss": 0.4132,
+      "epoch": 5.813715455475947,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1176613091026716e-05,
+      "loss": 0.4253,
       "step": 5680
     },
     {
-      "epoch": 6.0661686232657415,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.3243105181345826e-05,
-      "loss": 0.4552,
+      "epoch": 5.817809621289662,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.1034204059450193e-05,
+      "loss": 0.4547,
       "step": 5684
     },
     {
-      "epoch": 6.070437566702241,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.3124095472973306e-05,
-      "loss": 0.4393,
+      "epoch": 5.821903787103378,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.0892239369304122e-05,
+      "loss": 0.4063,
       "step": 5688
     },
     {
-      "epoch": 6.074706510138741,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.3005598455713201e-05,
-      "loss": 0.4293,
+      "epoch": 5.825997952917093,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.0750719509716584e-05,
+      "loss": 0.4275,
       "step": 5692
     },
     {
-      "epoch": 6.07897545357524,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.2887614573415434e-05,
-      "loss": 0.4416,
+      "epoch": 5.830092118730809,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.0609644968283068e-05,
+      "loss": 0.4478,
       "step": 5696
     },
     {
-      "epoch": 6.08324439701174,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.2770144268007998e-05,
-      "loss": 0.4191,
+      "epoch": 5.834186284544524,
+      "grad_norm": 0.375,
+      "learning_rate": 2.046901623106476e-05,
+      "loss": 0.3858,
       "step": 5700
     },
     {
-      "epoch": 6.087513340448239,
+      "epoch": 5.83828045035824,
       "grad_norm": 0.345703125,
-      "learning_rate": 1.2653187979495111e-05,
-      "loss": 0.4396,
+      "learning_rate": 2.0328833782586735e-05,
+      "loss": 0.4287,
       "step": 5704
     },
     {
-      "epoch": 6.091782283884738,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.2536746145955744e-05,
-      "loss": 0.4466,
+      "epoch": 5.842374616171955,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.018909810583657e-05,
+      "loss": 0.4537,
       "step": 5708
     },
     {
-      "epoch": 6.096051227321238,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.2420819203541848e-05,
-      "loss": 0.4069,
+      "epoch": 5.846468781985671,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.0049809682262485e-05,
+      "loss": 0.4189,
       "step": 5712
     },
     {
-      "epoch": 6.100320170757738,
+      "epoch": 5.850562947799386,
       "grad_norm": 0.3515625,
-      "learning_rate": 1.2305407586476746e-05,
-      "loss": 0.3761,
+      "learning_rate": 1.9910968991771642e-05,
+      "loss": 0.4401,
       "step": 5716
     },
     {
-      "epoch": 6.104589114194237,
+      "epoch": 5.854657113613102,
       "grad_norm": 0.3515625,
-      "learning_rate": 1.2190511727053571e-05,
-      "loss": 0.3931,
+      "learning_rate": 1.977257651272869e-05,
+      "loss": 0.4498,
       "step": 5720
     },
     {
-      "epoch": 6.108858057630736,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.2076132055633536e-05,
-      "loss": 0.4263,
+      "epoch": 5.858751279426817,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.963463272195394e-05,
+      "loss": 0.4612,
       "step": 5724
     },
     {
-      "epoch": 6.113127001067236,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.1962269000644486e-05,
-      "loss": 0.3988,
+      "epoch": 5.8628454452405325,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.949713809472177e-05,
+      "loss": 0.4135,
       "step": 5728
     },
     {
-      "epoch": 6.117395944503735,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.184892298857908e-05,
-      "loss": 0.4063,
+      "epoch": 5.8669396110542475,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9360093104759035e-05,
+      "loss": 0.4503,
       "step": 5732
     },
     {
-      "epoch": 6.121664887940235,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.1736094443993422e-05,
-      "loss": 0.4328,
+      "epoch": 5.8710337768679635,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9223498224243395e-05,
+      "loss": 0.4107,
       "step": 5736
     },
     {
-      "epoch": 6.125933831376734,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.162378378950527e-05,
-      "loss": 0.4695,
+      "epoch": 5.8751279426816785,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9087353923801723e-05,
+      "loss": 0.4364,
       "step": 5740
     },
     {
-      "epoch": 6.130202774813234,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.151199144579255e-05,
-      "loss": 0.4899,
+      "epoch": 5.879222108495394,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8951660672508335e-05,
+      "loss": 0.4805,
       "step": 5744
     },
     {
-      "epoch": 6.134471718249733,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.1400717831591844e-05,
-      "loss": 0.4101,
+      "epoch": 5.883316274309109,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8816418937883615e-05,
+      "loss": 0.4737,
       "step": 5748
     },
     {
-      "epoch": 6.138740661686232,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.1289963363696631e-05,
-      "loss": 0.42,
+      "epoch": 5.887410440122825,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.86816291858923e-05,
+      "loss": 0.4202,
       "step": 5752
     },
     {
-      "epoch": 6.1430096051227325,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.117972845695595e-05,
-      "loss": 0.4035,
+      "epoch": 5.89150460593654,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8547291880941702e-05,
+      "loss": 0.4629,
       "step": 5756
     },
     {
-      "epoch": 6.147278548559232,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.1070013524272664e-05,
-      "loss": 0.4258,
+      "epoch": 5.895598771750256,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8413407485880394e-05,
+      "loss": 0.4427,
       "step": 5760
     },
     {
-      "epoch": 6.151547491995731,
-      "grad_norm": 0.3515625,
-      "learning_rate": 1.0960818976602059e-05,
-      "loss": 0.3922,
+      "epoch": 5.899692937563971,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8279976461996505e-05,
+      "loss": 0.4057,
       "step": 5764
     },
     {
-      "epoch": 6.1558164354322304,
-      "grad_norm": 0.375,
-      "learning_rate": 1.0852145222950148e-05,
-      "loss": 0.4279,
+      "epoch": 5.903787103377686,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.814699926901597e-05,
+      "loss": 0.4321,
       "step": 5768
     },
     {
-      "epoch": 6.16008537886873,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.0743992670372304e-05,
-      "loss": 0.434,
+      "epoch": 5.907881269191402,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8014476365101222e-05,
+      "loss": 0.4007,
       "step": 5772
     },
     {
-      "epoch": 6.164354322305229,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.0636361723971593e-05,
-      "loss": 0.4193,
+      "epoch": 5.911975435005118,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7882408206849446e-05,
+      "loss": 0.423,
       "step": 5776
     },
     {
-      "epoch": 6.168623265741729,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.0529252786897346e-05,
-      "loss": 0.4083,
+      "epoch": 5.916069600818833,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7750795249290944e-05,
+      "loss": 0.4179,
       "step": 5780
     },
     {
-      "epoch": 6.172892209178229,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.0422666260343675e-05,
-      "loss": 0.4413,
+      "epoch": 5.920163766632548,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7619637945887765e-05,
+      "loss": 0.4201,
       "step": 5784
     },
     {
-      "epoch": 6.177161152614728,
-      "grad_norm": 0.36328125,
-      "learning_rate": 1.0316602543547808e-05,
-      "loss": 0.4353,
+      "epoch": 5.924257932446264,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.748893674853205e-05,
+      "loss": 0.4133,
       "step": 5788
     },
     {
-      "epoch": 6.181430096051227,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.0211062033788819e-05,
-      "loss": 0.4319,
+      "epoch": 5.92835209825998,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7358692107544363e-05,
+      "loss": 0.4251,
       "step": 5792
     },
     {
-      "epoch": 6.1856990394877265,
-      "grad_norm": 0.33203125,
-      "learning_rate": 1.0106045126385904e-05,
-      "loss": 0.4079,
+      "epoch": 5.932446264073695,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7228904471672294e-05,
+      "loss": 0.4342,
       "step": 5796
     },
     {
-      "epoch": 6.189967982924227,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.000155221469714e-05,
-      "loss": 0.4086,
+      "epoch": 5.93654042988741,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.7099574288088906e-05,
+      "loss": 0.4283,
       "step": 5800
     },
     {
-      "epoch": 6.194236926360726,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.89758369011781e-06,
-      "loss": 0.4405,
+      "epoch": 5.940634595701126,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.697070200239103e-05,
+      "loss": 0.4336,
       "step": 5804
     },
     {
-      "epoch": 6.198505869797225,
-      "grad_norm": 0.36328125,
-      "learning_rate": 9.794139942079055e-06,
-      "loss": 0.4013,
+      "epoch": 5.944728761514841,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6842288058597946e-05,
+      "loss": 0.3919,
       "step": 5808
     },
     {
-      "epoch": 6.202774813233725,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.691221358046363e-06,
-      "loss": 0.4015,
+      "epoch": 5.948822927328557,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6714332899149764e-05,
+      "loss": 0.4539,
       "step": 5812
     },
     {
-      "epoch": 6.207043756670224,
-      "grad_norm": 0.345703125,
-      "learning_rate": 9.58882832351814e-06,
-      "loss": 0.4249,
+      "epoch": 5.952917093142272,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6586836964905775e-05,
+      "loss": 0.447,
       "step": 5816
     },
     {
-      "epoch": 6.211312700106723,
-      "grad_norm": 0.37890625,
-      "learning_rate": 9.486961222024287e-06,
-      "loss": 0.4196,
+      "epoch": 5.957011258955988,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6459800695143166e-05,
+      "loss": 0.4306,
       "step": 5820
     },
     {
-      "epoch": 6.215581643543223,
-      "grad_norm": 0.333984375,
-      "learning_rate": 9.385620435124691e-06,
-      "loss": 0.4236,
+      "epoch": 5.961105424769703,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6333224527555332e-05,
+      "loss": 0.3874,
       "step": 5824
     },
     {
-      "epoch": 6.219850586979723,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.28480634240788e-06,
-      "loss": 0.4004,
+      "epoch": 5.965199590583419,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.620710889825039e-05,
+      "loss": 0.3592,
       "step": 5828
     },
     {
-      "epoch": 6.224119530416222,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.184519321489509e-06,
-      "loss": 0.4866,
+      "epoch": 5.969293756397134,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6081454241749782e-05,
+      "loss": 0.4402,
       "step": 5832
     },
     {
-      "epoch": 6.228388473852721,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.084759748011095e-06,
-      "loss": 0.4178,
+      "epoch": 5.97338792221085,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.595626099098667e-05,
+      "loss": 0.4694,
       "step": 5836
     },
     {
-      "epoch": 6.232657417289221,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.985527995638386e-06,
-      "loss": 0.418,
+      "epoch": 5.977482088024565,
+      "grad_norm": 0.375,
+      "learning_rate": 1.583152957730447e-05,
+      "loss": 0.4011,
       "step": 5840
     },
     {
-      "epoch": 6.23692636072572,
-      "grad_norm": 0.361328125,
-      "learning_rate": 8.886824436060213e-06,
-      "loss": 0.4134,
+      "epoch": 5.981576253838281,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5707260430455413e-05,
+      "loss": 0.4348,
       "step": 5844
     },
     {
-      "epoch": 6.24119530416222,
-      "grad_norm": 0.357421875,
-      "learning_rate": 8.788649438986888e-06,
-      "loss": 0.4561,
+      "epoch": 5.985670419651996,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.558345397859893e-05,
+      "loss": 0.3799,
       "step": 5848
     },
     {
-      "epoch": 6.245464247598719,
-      "grad_norm": 0.33984375,
-      "learning_rate": 8.691003372148959e-06,
-      "loss": 0.3838,
+      "epoch": 5.9897645854657116,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.54601106483004e-05,
+      "loss": 0.4204,
       "step": 5852
     },
     {
-      "epoch": 6.249733191035219,
-      "grad_norm": 0.361328125,
-      "learning_rate": 8.593886601295736e-06,
-      "loss": 0.4071,
+      "epoch": 5.993858751279427,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.53372308645295e-05,
+      "loss": 0.4476,
       "step": 5856
     },
     {
-      "epoch": 6.254002134471718,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.497299490194004e-06,
-      "loss": 0.424,
+      "epoch": 5.9979529170931425,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.521481505065873e-05,
+      "loss": 0.4125,
       "step": 5860
     },
     {
-      "epoch": 6.258271077908217,
-      "grad_norm": 0.36328125,
-      "learning_rate": 8.401242400626601e-06,
-      "loss": 0.4589,
+      "epoch": 6.0020470829068575,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5092863628462093e-05,
+      "loss": 0.4496,
       "step": 5864
     },
     {
-      "epoch": 6.2625400213447175,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.305715692391069e-06,
-      "loss": 0.409,
+      "epoch": 6.006141248720573,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4971377018113617e-05,
+      "loss": 0.4461,
       "step": 5868
     },
     {
-      "epoch": 6.266808964781217,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.21071972329837e-06,
-      "loss": 0.3921,
+      "epoch": 6.0102354145342884,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4850355638185713e-05,
+      "loss": 0.3945,
       "step": 5872
     },
     {
-      "epoch": 6.271077908217716,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.116254849171427e-06,
-      "loss": 0.3989,
+      "epoch": 6.014329580348004,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.472979990564797e-05,
+      "loss": 0.4214,
       "step": 5876
     },
     {
-      "epoch": 6.2753468516542155,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.022321423843903e-06,
-      "loss": 0.3277,
+      "epoch": 6.018423746161719,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.460971023586565e-05,
+      "loss": 0.4268,
       "step": 5880
     },
     {
-      "epoch": 6.279615795090715,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.928919799158818e-06,
-      "loss": 0.4409,
+      "epoch": 6.022517911975435,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4490087042598147e-05,
+      "loss": 0.4359,
       "step": 5884
     },
     {
-      "epoch": 6.283884738527215,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.83605032496727e-06,
-      "loss": 0.4155,
+      "epoch": 6.02661207778915,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4370930737997722e-05,
+      "loss": 0.4153,
       "step": 5888
     },
     {
-      "epoch": 6.288153681963714,
-      "grad_norm": 0.37109375,
-      "learning_rate": 7.74371334912705e-06,
-      "loss": 0.4119,
+      "epoch": 6.030706243602866,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4252241732608004e-05,
+      "loss": 0.4195,
       "step": 5892
     },
     {
-      "epoch": 6.292422625400214,
-      "grad_norm": 0.3515625,
-      "learning_rate": 7.651909217501467e-06,
-      "loss": 0.421,
+      "epoch": 6.034800409416581,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4134020435362487e-05,
+      "loss": 0.3509,
       "step": 5896
     },
     {
-      "epoch": 6.296691568836713,
-      "grad_norm": 0.361328125,
-      "learning_rate": 7.56063827395787e-06,
-      "loss": 0.4199,
+      "epoch": 6.038894575230297,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4016267253583324e-05,
+      "loss": 0.4191,
       "step": 5900
     },
     {
-      "epoch": 6.300960512273212,
-      "grad_norm": 0.33203125,
-      "learning_rate": 7.469900860366573e-06,
-      "loss": 0.4212,
+      "epoch": 6.042988741044012,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3898982592979802e-05,
+      "loss": 0.4477,
       "step": 5904
     },
     {
-      "epoch": 6.3052294557097115,
-      "grad_norm": 0.365234375,
-      "learning_rate": 7.379697316599376e-06,
-      "loss": 0.4452,
+      "epoch": 6.047082906857728,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.378216685764686e-05,
+      "loss": 0.4107,
       "step": 5908
     },
     {
-      "epoch": 6.309498399146212,
-      "grad_norm": 0.326171875,
-      "learning_rate": 7.290027980528473e-06,
-      "loss": 0.3906,
+      "epoch": 6.051177072671443,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3665820450063898e-05,
+      "loss": 0.3961,
       "step": 5912
     },
     {
-      "epoch": 6.313767342582711,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.200893188024998e-06,
-      "loss": 0.419,
+      "epoch": 6.055271238485159,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3549943771093258e-05,
+      "loss": 0.4052,
       "step": 5916
     },
     {
-      "epoch": 6.31803628601921,
-      "grad_norm": 0.3515625,
-      "learning_rate": 7.1122932729579145e-06,
-      "loss": 0.446,
+      "epoch": 6.059365404298874,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3434537219978813e-05,
+      "loss": 0.389,
       "step": 5920
     },
     {
-      "epoch": 6.32230522945571,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.024228567192741e-06,
-      "loss": 0.4435,
+      "epoch": 6.06345957011259,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3319601194344698e-05,
+      "loss": 0.4063,
       "step": 5924
     },
     {
-      "epoch": 6.326574172892209,
-      "grad_norm": 0.337890625,
-      "learning_rate": 6.936699400590201e-06,
-      "loss": 0.4534,
+      "epoch": 6.067553735926305,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3205136090193923e-05,
+      "loss": 0.436,
       "step": 5928
     },
     {
-      "epoch": 6.330843116328708,
-      "grad_norm": 0.365234375,
-      "learning_rate": 6.84970610100517e-06,
-      "loss": 0.4267,
+      "epoch": 6.071647901740021,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3091142301906887e-05,
+      "loss": 0.4485,
       "step": 5932
     },
     {
-      "epoch": 6.335112059765208,
+      "epoch": 6.075742067553736,
       "grad_norm": 0.353515625,
-      "learning_rate": 6.763248994285297e-06,
-      "loss": 0.4479,
+      "learning_rate": 1.2977620222240165e-05,
+      "loss": 0.4521,
       "step": 5936
     },
     {
-      "epoch": 6.339381003201708,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.677328404269788e-06,
-      "loss": 0.4158,
+      "epoch": 6.079836233367452,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2864570242325133e-05,
+      "loss": 0.4712,
       "step": 5940
     },
     {
-      "epoch": 6.343649946638207,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.591944652788323e-06,
-      "loss": 0.454,
+      "epoch": 6.083930399181167,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2751992751666457e-05,
+      "loss": 0.4428,
       "step": 5944
     },
     {
-      "epoch": 6.347918890074706,
-      "grad_norm": 0.369140625,
-      "learning_rate": 6.507098059659705e-06,
-      "loss": 0.4538,
+      "epoch": 6.088024564994882,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2639888138141014e-05,
+      "loss": 0.4535,
       "step": 5948
     },
     {
-      "epoch": 6.352187833511206,
-      "grad_norm": 0.36328125,
-      "learning_rate": 6.422788942690748e-06,
-      "loss": 0.436,
+      "epoch": 6.092118730808598,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2528256787996372e-05,
+      "loss": 0.4473,
       "step": 5952
     },
     {
-      "epoch": 6.356456776947706,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.339017617675046e-06,
-      "loss": 0.4181,
+      "epoch": 6.096212896622313,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2417099085849468e-05,
+      "loss": 0.3956,
       "step": 5956
     },
     {
-      "epoch": 6.360725720384205,
-      "grad_norm": 0.3671875,
-      "learning_rate": 6.255784398391833e-06,
-      "loss": 0.4432,
+      "epoch": 6.100307062436029,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2306415414685366e-05,
+      "loss": 0.4278,
       "step": 5960
     },
     {
-      "epoch": 6.3649946638207044,
-      "grad_norm": 0.35546875,
-      "learning_rate": 6.173089596604724e-06,
-      "loss": 0.4172,
+      "epoch": 6.104401228249744,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.219620615585593e-05,
+      "loss": 0.445,
       "step": 5964
     },
     {
-      "epoch": 6.369263607257204,
-      "grad_norm": 0.37109375,
-      "learning_rate": 6.090933522060665e-06,
-      "loss": 0.4493,
+      "epoch": 6.10849539406346,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2086471689078353e-05,
+      "loss": 0.4198,
       "step": 5968
     },
     {
-      "epoch": 6.373532550693703,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.009316482488696e-06,
-      "loss": 0.4276,
+      "epoch": 6.112589559877175,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.1977212392434082e-05,
+      "loss": 0.4577,
       "step": 5972
     },
     {
-      "epoch": 6.377801494130202,
-      "grad_norm": 0.322265625,
-      "learning_rate": 5.92823878359876e-06,
-      "loss": 0.4382,
+      "epoch": 6.116683725690891,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1868428642367378e-05,
+      "loss": 0.4793,
       "step": 5976
     },
     {
-      "epoch": 6.382070437566703,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.8477007290807285e-06,
-      "loss": 0.413,
+      "epoch": 6.120777891504606,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1760120813684009e-05,
+      "loss": 0.4424,
       "step": 5980
     },
     {
-      "epoch": 6.386339381003202,
+      "epoch": 6.1248720573183215,
       "grad_norm": 0.349609375,
-      "learning_rate": 5.767702620602976e-06,
-      "loss": 0.4251,
+      "learning_rate": 1.1652289279550026e-05,
+      "loss": 0.4027,
       "step": 5984
     },
     {
-      "epoch": 6.390608324439701,
-      "grad_norm": 0.37109375,
-      "learning_rate": 5.6882447578116105e-06,
-      "loss": 0.4585,
+      "epoch": 6.1289662231320365,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1544934411490469e-05,
+      "loss": 0.3947,
       "step": 5988
     },
     {
-      "epoch": 6.3948772678762005,
-      "grad_norm": 0.328125,
-      "learning_rate": 5.609327438328992e-06,
-      "loss": 0.4004,
+      "epoch": 6.1330603889457525,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1438056579387966e-05,
+      "loss": 0.4369,
       "step": 5992
     },
     {
-      "epoch": 6.3991462113127,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.5309509577529165e-06,
-      "loss": 0.4432,
+      "epoch": 6.1371545547594675,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1331656151481654e-05,
+      "loss": 0.4243,
       "step": 5996
     },
     {
-      "epoch": 6.4034151547492,
+      "epoch": 6.141248720573183,
       "grad_norm": 0.34765625,
-      "learning_rate": 5.453115609655284e-06,
-      "loss": 0.4075,
+      "learning_rate": 1.1225733494365791e-05,
+      "loss": 0.4472,
       "step": 6000
     },
     {
-      "epoch": 6.407684098185699,
+      "epoch": 6.145342886386898,
       "grad_norm": 0.3515625,
-      "learning_rate": 5.37582168558115e-06,
-      "loss": 0.4164,
+      "learning_rate": 1.1120288972988445e-05,
+      "loss": 0.3859,
       "step": 6004
     },
     {
-      "epoch": 6.411953041622199,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.299069475047591e-06,
-      "loss": 0.4106,
+      "epoch": 6.149437052200614,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1015322950650408e-05,
+      "loss": 0.4321,
       "step": 6008
     },
     {
-      "epoch": 6.416221985058698,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.222859265542539e-06,
-      "loss": 0.4265,
+      "epoch": 6.153531218014329,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0910835789003785e-05,
+      "loss": 0.4198,
       "step": 6012
     },
     {
-      "epoch": 6.420490928495197,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.147191342523838e-06,
-      "loss": 0.3976,
+      "epoch": 6.157625383828045,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.0806827848050791e-05,
+      "loss": 0.4539,
       "step": 6016
     },
     {
-      "epoch": 6.4247598719316965,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.072065989418067e-06,
-      "loss": 0.3704,
+      "epoch": 6.16171954964176,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0703299486142541e-05,
+      "loss": 0.4371,
       "step": 6020
     },
     {
-      "epoch": 6.429028815368197,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.997483487619552e-06,
-      "loss": 0.4541,
+      "epoch": 6.165813715455476,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0600251059977854e-05,
+      "loss": 0.3967,
       "step": 6024
     },
     {
-      "epoch": 6.433297758804696,
+      "epoch": 6.169907881269191,
       "grad_norm": 0.341796875,
-      "learning_rate": 4.923444116489245e-06,
-      "loss": 0.4111,
+      "learning_rate": 1.0497682924601841e-05,
+      "loss": 0.4088,
       "step": 6028
     },
     {
-      "epoch": 6.437566702241195,
-      "grad_norm": 0.349609375,
-      "learning_rate": 4.849948153353755e-06,
-      "loss": 0.3754,
+      "epoch": 6.174002047082907,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0395595433404935e-05,
+      "loss": 0.4137,
       "step": 6032
     },
     {
-      "epoch": 6.441835645677695,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.776995873504241e-06,
-      "loss": 0.4393,
+      "epoch": 6.178096212896622,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.029398893812151e-05,
+      "loss": 0.4145,
       "step": 6036
     },
     {
-      "epoch": 6.446104589114194,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.704587550195404e-06,
-      "loss": 0.4263,
+      "epoch": 6.182190378710338,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.0192863788828654e-05,
+      "loss": 0.4265,
       "step": 6040
     },
     {
-      "epoch": 6.450373532550694,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.632723454644477e-06,
-      "loss": 0.3849,
+      "epoch": 6.186284544524053,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.0092220333945073e-05,
+      "loss": 0.4127,
       "step": 6044
     },
     {
-      "epoch": 6.454642475987193,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.561403856030188e-06,
-      "loss": 0.4042,
+      "epoch": 6.190378710337769,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.992058920229823e-06,
+      "loss": 0.4384,
       "step": 6048
     },
     {
-      "epoch": 6.458911419423693,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.490629021491815e-06,
-      "loss": 0.4332,
+      "epoch": 6.194472876151484,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.892379892781088e-06,
+      "loss": 0.4406,
       "step": 6052
     },
     {
-      "epoch": 6.463180362860192,
-      "grad_norm": 0.361328125,
-      "learning_rate": 4.420399216128051e-06,
-      "loss": 0.4284,
+      "epoch": 6.1985670419652,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.793183595035082e-06,
+      "loss": 0.4378,
       "step": 6056
     },
     {
-      "epoch": 6.467449306296691,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.350714702996122e-06,
-      "loss": 0.4086,
+      "epoch": 6.202661207778915,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.694470368764812e-06,
+      "loss": 0.3789,
       "step": 6060
     },
     {
-      "epoch": 6.471718249733191,
-      "grad_norm": 0.359375,
-      "learning_rate": 4.281575743110804e-06,
-      "loss": 0.4516,
+      "epoch": 6.206755373592631,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.596240554078838e-06,
+      "loss": 0.3834,
       "step": 6064
     },
     {
-      "epoch": 6.475987193169691,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.212982595443376e-06,
-      "loss": 0.4195,
+      "epoch": 6.210849539406346,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.4984944894202e-06,
+      "loss": 0.4451,
       "step": 6068
     },
     {
-      "epoch": 6.48025613660619,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.144935516920733e-06,
-      "loss": 0.4371,
+      "epoch": 6.214943705220062,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.40123251156527e-06,
+      "loss": 0.4477,
       "step": 6072
     },
     {
-      "epoch": 6.4845250800426895,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.077434762424308e-06,
-      "loss": 0.4639,
+      "epoch": 6.219037871033777,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.304454955622425e-06,
+      "loss": 0.401,
       "step": 6076
     },
     {
-      "epoch": 6.488794023479189,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.010480584789289e-06,
-      "loss": 0.3879,
+      "epoch": 6.223132036847493,
+      "grad_norm": 0.375,
+      "learning_rate": 9.208162155031074e-06,
+      "loss": 0.4438,
       "step": 6080
     },
     {
-      "epoch": 6.493062966915688,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.9440732348034296e-06,
-      "loss": 0.463,
+      "epoch": 6.227226202661208,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.112354441560476e-06,
+      "loss": 0.4212,
       "step": 6084
     },
     {
-      "epoch": 6.497331910352187,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.878212961206412e-06,
-      "loss": 0.3724,
+      "epoch": 6.231320368474924,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.017032145308483e-06,
+      "loss": 0.4082,
       "step": 6088
     },
     {
-      "epoch": 6.501600853788688,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.812900010688619e-06,
-      "loss": 0.3974,
+      "epoch": 6.235414534288639,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.9221955947005e-06,
+      "loss": 0.4046,
       "step": 6092
     },
     {
-      "epoch": 6.505869797225187,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.748134627890442e-06,
-      "loss": 0.4343,
+      "epoch": 6.239508700102354,
+      "grad_norm": 0.37890625,
+      "learning_rate": 8.82784511648838e-06,
+      "loss": 0.418,
       "step": 6096
     },
     {
-      "epoch": 6.510138740661686,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.6839170554011977e-06,
-      "loss": 0.4224,
+      "epoch": 6.24360286591607,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.733981035749193e-06,
+      "loss": 0.4293,
       "step": 6100
     },
     {
-      "epoch": 6.5144076840981855,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.620247533758325e-06,
-      "loss": 0.413,
+      "epoch": 6.247697031729785,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.640603675884194e-06,
+      "loss": 0.4244,
       "step": 6104
     },
     {
-      "epoch": 6.518676627534685,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.557126301446439e-06,
-      "loss": 0.3969,
+      "epoch": 6.2517911975435005,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.547713358617714e-06,
+      "loss": 0.4109,
       "step": 6108
     },
     {
-      "epoch": 6.522945570971185,
+      "epoch": 6.255885363357216,
       "grad_norm": 0.357421875,
-      "learning_rate": 3.494553594896482e-06,
-      "loss": 0.4197,
+      "learning_rate": 8.455310403995924e-06,
+      "loss": 0.4285,
       "step": 6112
     },
     {
-      "epoch": 6.527214514407684,
-      "grad_norm": 0.37890625,
-      "learning_rate": 3.432529648484722e-06,
-      "loss": 0.4544,
+      "epoch": 6.2599795291709315,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.363395130385908e-06,
+      "loss": 0.4134,
       "step": 6116
     },
     {
-      "epoch": 6.531483457844184,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.3710546945320383e-06,
-      "loss": 0.3835,
+      "epoch": 6.2640736949846465,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.27196785447446e-06,
+      "loss": 0.438,
       "step": 6120
     },
     {
-      "epoch": 6.535752401280683,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.3101289633029384e-06,
-      "loss": 0.4282,
+      "epoch": 6.268167860798362,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.181028891267017e-06,
+      "loss": 0.4141,
       "step": 6124
     },
     {
-      "epoch": 6.540021344717182,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.2497526830047084e-06,
-      "loss": 0.3845,
+      "epoch": 6.272262026612077,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.09057855408658e-06,
+      "loss": 0.4089,
       "step": 6128
     },
     {
-      "epoch": 6.544290288153682,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.1899260797866145e-06,
-      "loss": 0.405,
+      "epoch": 6.276356192425793,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.000617154572597e-06,
+      "loss": 0.4153,
       "step": 6132
     },
     {
-      "epoch": 6.548559231590182,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.130649377738953e-06,
-      "loss": 0.4102,
+      "epoch": 6.280450358239508,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.91114500267993e-06,
+      "loss": 0.4158,
       "step": 6136
     },
     {
-      "epoch": 6.552828175026681,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.0719227988923844e-06,
-      "loss": 0.4326,
+      "epoch": 6.284544524053224,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.82216240667784e-06,
+      "loss": 0.4268,
       "step": 6140
     },
     {
-      "epoch": 6.55709711846318,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.013746563216851e-06,
-      "loss": 0.3952,
+      "epoch": 6.288638689866939,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.733669673148768e-06,
+      "loss": 0.4038,
       "step": 6144
     },
     {
-      "epoch": 6.56136606189968,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.9561208886210275e-06,
-      "loss": 0.4046,
+      "epoch": 6.292732855680655,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.645667106987407e-06,
+      "loss": 0.4577,
       "step": 6148
     },
     {
-      "epoch": 6.565635005336179,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.899045990951271e-06,
-      "loss": 0.4215,
+      "epoch": 6.29682702149437,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.558155011399669e-06,
+      "loss": 0.412,
       "step": 6152
     },
     {
-      "epoch": 6.569903948772679,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.8425220839909892e-06,
-      "loss": 0.4504,
+      "epoch": 6.300921187308086,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.471133687901498e-06,
+      "loss": 0.3985,
       "step": 6156
     },
     {
-      "epoch": 6.574172892209178,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.7865493794597072e-06,
-      "loss": 0.4461,
+      "epoch": 6.305015353121801,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.384603436317993e-06,
+      "loss": 0.4508,
       "step": 6160
     },
     {
-      "epoch": 6.578441835645678,
-      "grad_norm": 0.34765625,
-      "learning_rate": 2.7311280870123687e-06,
-      "loss": 0.4312,
+      "epoch": 6.309109518935517,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.298564554782288e-06,
+      "loss": 0.4141,
       "step": 6164
     },
     {
-      "epoch": 6.582710779082177,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.6762584142385023e-06,
-      "loss": 0.4268,
+      "epoch": 6.313203684749232,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.213017339734506e-06,
+      "loss": 0.4444,
       "step": 6168
     },
     {
-      "epoch": 6.586979722518676,
-      "grad_norm": 0.341796875,
-      "learning_rate": 2.6219405666614402e-06,
-      "loss": 0.4287,
+      "epoch": 6.317297850562948,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.127962085920808e-06,
+      "loss": 0.4156,
       "step": 6172
     },
     {
-      "epoch": 6.591248665955176,
-      "grad_norm": 0.3515625,
-      "learning_rate": 2.5681747477375834e-06,
-      "loss": 0.4265,
+      "epoch": 6.321392016376663,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.043399086392343e-06,
+      "loss": 0.4466,
       "step": 6176
     },
     {
-      "epoch": 6.595517609391676,
-      "grad_norm": 0.369140625,
-      "learning_rate": 2.5149611588556053e-06,
-      "loss": 0.4699,
+      "epoch": 6.325486182190379,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.9593286325042185e-06,
+      "loss": 0.4325,
       "step": 6180
     },
     {
-      "epoch": 6.599786552828175,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.462299999335715e-06,
-      "loss": 0.4402,
+      "epoch": 6.329580348004094,
+      "grad_norm": 0.375,
+      "learning_rate": 6.875751013914516e-06,
+      "loss": 0.431,
       "step": 6184
     },
     {
-      "epoch": 6.6040554962646745,
-      "grad_norm": 0.380859375,
-      "learning_rate": 2.4101914664289125e-06,
-      "loss": 0.4118,
+      "epoch": 6.33367451381781,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.79266651858329e-06,
+      "loss": 0.4705,
       "step": 6188
     },
     {
-      "epoch": 6.608324439701174,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.358635755316218e-06,
-      "loss": 0.4176,
+      "epoch": 6.337768679631525,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.710075432771606e-06,
+      "loss": 0.4316,
       "step": 6192
     },
     {
-      "epoch": 6.612593383137673,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.307633059108044e-06,
-      "loss": 0.396,
+      "epoch": 6.341862845445241,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.627978041040488e-06,
+      "loss": 0.4479,
       "step": 6196
     },
     {
-      "epoch": 6.616862326574173,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.2571835688432737e-06,
-      "loss": 0.4517,
+      "epoch": 6.345957011258956,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.54637462624999e-06,
+      "loss": 0.3946,
       "step": 6200
     },
     {
-      "epoch": 6.621131270010673,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.2072874734887835e-06,
-      "loss": 0.4267,
+      "epoch": 6.350051177072672,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.465265469558256e-06,
+      "loss": 0.4489,
       "step": 6204
     },
     {
-      "epoch": 6.625400213447172,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.1579449599385403e-06,
-      "loss": 0.3821,
+      "epoch": 6.354145342886387,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.384650850420397e-06,
+      "loss": 0.4443,
       "step": 6208
     },
     {
-      "epoch": 6.629669156883671,
-      "grad_norm": 0.375,
-      "learning_rate": 2.1091562130130358e-06,
-      "loss": 0.444,
+      "epoch": 6.358239508700103,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.304531046587719e-06,
+      "loss": 0.4218,
       "step": 6212
     },
     {
-      "epoch": 6.6339381003201705,
-      "grad_norm": 0.345703125,
-      "learning_rate": 2.0609214154584717e-06,
-      "loss": 0.4546,
+      "epoch": 6.362333674513818,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.224906334106689e-06,
+      "loss": 0.4614,
       "step": 6216
     },
     {
-      "epoch": 6.638207043756671,
-      "grad_norm": 0.3359375,
-      "learning_rate": 2.0132407479462097e-06,
-      "loss": 0.4492,
+      "epoch": 6.366427840327534,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.145776987317891e-06,
+      "loss": 0.4615,
       "step": 6220
     },
     {
-      "epoch": 6.64247598719317,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.9661143890719877e-06,
-      "loss": 0.4657,
+      "epoch": 6.370522006141249,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.067143278855241e-06,
+      "loss": 0.4737,
       "step": 6224
     },
     {
-      "epoch": 6.646744930629669,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.9195425153553047e-06,
-      "loss": 0.4184,
+      "epoch": 6.3746161719549645,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.9890054796449875e-06,
+      "loss": 0.4364,
       "step": 6228
     },
     {
-      "epoch": 6.651013874066169,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.8735253012387207e-06,
-      "loss": 0.3755,
+      "epoch": 6.37871033776868,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.911363858904661e-06,
+      "loss": 0.4312,
       "step": 6232
     },
     {
-      "epoch": 6.655282817502668,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.8280629190872586e-06,
-      "loss": 0.3719,
+      "epoch": 6.3828045035823955,
+      "grad_norm": 0.375,
+      "learning_rate": 5.834218684142344e-06,
+      "loss": 0.4062,
       "step": 6236
     },
     {
-      "epoch": 6.659551760939167,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.783155539187736e-06,
-      "loss": 0.4165,
+      "epoch": 6.3868986693961105,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.757570221155638e-06,
+      "loss": 0.4327,
       "step": 6240
     },
     {
-      "epoch": 6.6638207043756665,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.7388033297480509e-06,
-      "loss": 0.4176,
+      "epoch": 6.3909928352098255,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.6814187340307125e-06,
+      "loss": 0.4012,
       "step": 6244
     },
     {
-      "epoch": 6.668089647812167,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.6950064568967137e-06,
-      "loss": 0.4079,
+      "epoch": 6.395087001023541,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.605764485141507e-06,
+      "loss": 0.3789,
       "step": 6248
     },
     {
-      "epoch": 6.672358591248666,
-      "grad_norm": 0.361328125,
-      "learning_rate": 1.6517650846820495e-06,
-      "loss": 0.3964,
+      "epoch": 6.399181166837257,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.530607735148762e-06,
+      "loss": 0.4342,
       "step": 6252
     },
     {
-      "epoch": 6.676627534685165,
-      "grad_norm": 0.345703125,
-      "learning_rate": 1.6090793750717145e-06,
-      "loss": 0.4027,
+      "epoch": 6.403275332650972,
+      "grad_norm": 0.37890625,
+      "learning_rate": 5.4559487429990615e-06,
+      "loss": 0.4502,
       "step": 6256
     },
     {
-      "epoch": 6.680896478121665,
-      "grad_norm": 0.369140625,
-      "learning_rate": 1.5669494879520128e-06,
-      "loss": 0.454,
+      "epoch": 6.407369498464687,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.381787765924056e-06,
+      "loss": 0.4161,
       "step": 6260
     },
     {
-      "epoch": 6.685165421558164,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.5253755811273638e-06,
-      "loss": 0.4305,
+      "epoch": 6.411463664278403,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.308125059439522e-06,
+      "loss": 0.4442,
       "step": 6264
     },
     {
-      "epoch": 6.689434364994664,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.4843578103195864e-06,
-      "loss": 0.4352,
+      "epoch": 6.415557830092118,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.234960877344491e-06,
+      "loss": 0.4584,
       "step": 6268
     },
     {
-      "epoch": 6.6937033084311635,
-      "grad_norm": 0.359375,
-      "learning_rate": 1.4438963291674988e-06,
-      "loss": 0.4269,
+      "epoch": 6.419651995905834,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.1622954717203514e-06,
+      "loss": 0.4413,
       "step": 6272
     },
     {
-      "epoch": 6.697972251867663,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.4039912892262195e-06,
-      "loss": 0.4617,
+      "epoch": 6.423746161719549,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.090129092929984e-06,
+      "loss": 0.396,
       "step": 6276
     },
     {
-      "epoch": 6.702241195304162,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.3646428399665676e-06,
-      "loss": 0.397,
+      "epoch": 6.427840327533265,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.018461989616928e-06,
+      "loss": 0.4253,
       "step": 6280
     },
     {
-      "epoch": 6.706510138740661,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.3258511287746299e-06,
-      "loss": 0.4148,
+      "epoch": 6.43193449334698,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.947294408704533e-06,
+      "loss": 0.4244,
       "step": 6284
     },
     {
-      "epoch": 6.710779082177162,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.2876163009510943e-06,
-      "loss": 0.4369,
+      "epoch": 6.436028659160696,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.876626595395039e-06,
+      "loss": 0.434,
       "step": 6288
     },
     {
-      "epoch": 6.715048025613661,
-      "grad_norm": 0.34375,
-      "learning_rate": 1.2499384997107842e-06,
-      "loss": 0.3841,
+      "epoch": 6.440122824974411,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.806458793168799e-06,
+      "loss": 0.4057,
       "step": 6292
     },
     {
-      "epoch": 6.71931696905016,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.2128178661820586e-06,
-      "loss": 0.4223,
+      "epoch": 6.444216990788127,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.736791243783427e-06,
+      "loss": 0.4292,
       "step": 6296
     },
     {
-      "epoch": 6.7235859124866595,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.176254539406346e-06,
-      "loss": 0.426,
+      "epoch": 6.448311156601842,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.667624187272917e-06,
+      "loss": 0.4297,
       "step": 6300
     },
     {
-      "epoch": 6.727854855923159,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.1402486563375613e-06,
-      "loss": 0.4266,
+      "epoch": 6.452405322415558,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.598957861946906e-06,
+      "loss": 0.4292,
       "step": 6304
     },
     {
-      "epoch": 6.732123799359658,
-      "grad_norm": 0.35546875,
-      "learning_rate": 1.1048003518416394e-06,
-      "loss": 0.4314,
+      "epoch": 6.456499488229273,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.53079250438978e-06,
+      "loss": 0.4307,
       "step": 6308
     },
     {
-      "epoch": 6.736392742796158,
-      "grad_norm": 0.353515625,
-      "learning_rate": 1.0699097586960192e-06,
-      "loss": 0.4188,
+      "epoch": 6.460593654042989,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.463128349459855e-06,
+      "loss": 0.4114,
       "step": 6312
     },
     {
-      "epoch": 6.740661686232658,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.0355770075891112e-06,
-      "loss": 0.3995,
+      "epoch": 6.464687819856704,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.395965630288628e-06,
+      "loss": 0.4695,
       "step": 6316
     },
     {
-      "epoch": 6.744930629669157,
-      "grad_norm": 0.37890625,
-      "learning_rate": 1.0018022271198634e-06,
-      "loss": 0.4489,
+      "epoch": 6.46878198567042,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.32930457827993e-06,
+      "loss": 0.4417,
       "step": 6320
     },
     {
-      "epoch": 6.749199573105656,
-      "grad_norm": 0.3359375,
-      "learning_rate": 9.685855437972457e-07,
-      "loss": 0.4017,
+      "epoch": 6.472876151484135,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.263145423109121e-06,
+      "loss": 0.4577,
       "step": 6324
     },
     {
-      "epoch": 6.7534685165421555,
-      "grad_norm": 0.36328125,
-      "learning_rate": 9.359270820398002e-07,
-      "loss": 0.4647,
+      "epoch": 6.476970317297851,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.197488392722348e-06,
+      "loss": 0.3964,
       "step": 6328
     },
     {
-      "epoch": 6.757737459978655,
-      "grad_norm": 0.380859375,
-      "learning_rate": 9.03826964175125e-07,
-      "loss": 0.4367,
+      "epoch": 6.481064483111566,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.132333713335689e-06,
+      "loss": 0.4551,
       "step": 6332
     },
     {
-      "epoch": 6.762006403415155,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.722853104394578e-07,
-      "loss": 0.4079,
+      "epoch": 6.485158648925282,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.067681609434426e-06,
+      "loss": 0.432,
       "step": 6336
     },
     {
-      "epoch": 6.766275346851654,
-      "grad_norm": 0.333984375,
-      "learning_rate": 8.413022389772595e-07,
-      "loss": 0.3718,
+      "epoch": 6.489252814738997,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.003532303772256e-06,
+      "loss": 0.4323,
       "step": 6340
     },
     {
-      "epoch": 6.770544290288154,
-      "grad_norm": 0.349609375,
-      "learning_rate": 8.108778658406645e-07,
-      "loss": 0.4257,
+      "epoch": 6.493346980552713,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.939886017370564e-06,
+      "loss": 0.3916,
       "step": 6344
     },
     {
-      "epoch": 6.774813233724653,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.810123049891648e-07,
-      "loss": 0.407,
+      "epoch": 6.497441146366428,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.876742969517538e-06,
+      "loss": 0.4087,
       "step": 6348
     },
     {
-      "epoch": 6.779082177161152,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.517056682891099e-07,
-      "loss": 0.4074,
+      "epoch": 6.501535312180144,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.8141033777675854e-06,
+      "loss": 0.4148,
       "step": 6352
     },
     {
-      "epoch": 6.783351120597652,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.229580655132905e-07,
-      "loss": 0.4029,
+      "epoch": 6.505629477993859,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.751967457940436e-06,
+      "loss": 0.4203,
       "step": 6356
     },
     {
-      "epoch": 6.787620064034152,
+      "epoch": 6.5097236438075745,
       "grad_norm": 0.357421875,
-      "learning_rate": 6.947696043405726e-07,
-      "loss": 0.418,
+      "learning_rate": 3.6903354241204886e-06,
+      "loss": 0.4405,
       "step": 6360
     },
     {
-      "epoch": 6.791889007470651,
+      "epoch": 6.5138178096212895,
       "grad_norm": 0.353515625,
-      "learning_rate": 6.671403903554473e-07,
-      "loss": 0.4573,
+      "learning_rate": 3.6292074886559995e-06,
+      "loss": 0.4195,
       "step": 6364
     },
     {
-      "epoch": 6.79615795090715,
-      "grad_norm": 0.33203125,
-      "learning_rate": 6.400705270476481e-07,
-      "loss": 0.3894,
+      "epoch": 6.5179119754350054,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.5685838621584804e-06,
+      "loss": 0.4599,
       "step": 6368
     },
     {
-      "epoch": 6.80042689434365,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.135601158118175e-07,
-      "loss": 0.446,
+      "epoch": 6.5220061412487205,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.5084647535017996e-06,
+      "loss": 0.4015,
       "step": 6372
     },
     {
-      "epoch": 6.80469583778015,
-      "grad_norm": 0.3828125,
-      "learning_rate": 5.87609255947008e-07,
-      "loss": 0.4481,
+      "epoch": 6.526100307062436,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.448850369821565e-06,
+      "loss": 0.3875,
       "step": 6376
     },
     {
-      "epoch": 6.808964781216649,
+      "epoch": 6.530194472876151,
       "grad_norm": 0.36328125,
-      "learning_rate": 5.622180446564151e-07,
-      "loss": 0.4419,
+      "learning_rate": 3.389740916514461e-06,
+      "loss": 0.4491,
       "step": 6380
     },
     {
-      "epoch": 6.8132337246531485,
-      "grad_norm": 0.36328125,
-      "learning_rate": 5.373865770469943e-07,
-      "loss": 0.4108,
+      "epoch": 6.534288638689867,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.331136597237377e-06,
+      "loss": 0.4337,
       "step": 6384
     },
     {
-      "epoch": 6.817502668089648,
-      "grad_norm": 0.341796875,
-      "learning_rate": 5.131149461290784e-07,
-      "loss": 0.4523,
+      "epoch": 6.538382804503582,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.2730376139068816e-06,
+      "loss": 0.4579,
       "step": 6388
     },
     {
-      "epoch": 6.821771611526147,
-      "grad_norm": 0.365234375,
-      "learning_rate": 4.89403242816011e-07,
-      "loss": 0.4355,
+      "epoch": 6.542476970317297,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.2154441666984164e-06,
+      "loss": 0.3983,
       "step": 6392
     },
     {
-      "epoch": 6.826040554962646,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.662515559238633e-07,
-      "loss": 0.4443,
+      "epoch": 6.546571136131013,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.158356454045602e-06,
+      "loss": 0.4155,
       "step": 6396
     },
     {
-      "epoch": 6.830309498399147,
-      "grad_norm": 0.36328125,
-      "learning_rate": 4.436599721710843e-07,
-      "loss": 0.4419,
+      "epoch": 6.550665301944729,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.101774672639684e-06,
+      "loss": 0.4145,
       "step": 6400
     },
     {
-      "epoch": 6.834578441835646,
-      "grad_norm": 0.35546875,
-      "learning_rate": 4.2162857617815127e-07,
-      "loss": 0.429,
+      "epoch": 6.554759467758444,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.045699017428671e-06,
+      "loss": 0.4317,
       "step": 6404
     },
     {
-      "epoch": 6.838847385272145,
-      "grad_norm": 0.376953125,
-      "learning_rate": 4.0015745046725336e-07,
-      "loss": 0.4691,
+      "epoch": 6.558853633572159,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.990129681616782e-06,
+      "loss": 0.4389,
       "step": 6408
     },
     {
-      "epoch": 6.8431163287086445,
-      "grad_norm": 0.34765625,
-      "learning_rate": 3.7924667546200826e-07,
-      "loss": 0.4274,
+      "epoch": 6.562947799385875,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.9350668566637982e-06,
+      "loss": 0.4404,
       "step": 6412
     },
     {
-      "epoch": 6.847385272145144,
-      "grad_norm": 0.359375,
-      "learning_rate": 3.5889632948716273e-07,
-      "loss": 0.4345,
+      "epoch": 6.567041965199591,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.880510732284297e-06,
+      "loss": 0.4248,
       "step": 6416
     },
     {
-      "epoch": 6.851654215581643,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.391064887682426e-07,
-      "loss": 0.4309,
+      "epoch": 6.571136131013306,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.8264614964470856e-06,
+      "loss": 0.4753,
       "step": 6420
     },
     {
-      "epoch": 6.855923159018143,
-      "grad_norm": 0.333984375,
-      "learning_rate": 3.1987722743135325e-07,
-      "loss": 0.4297,
+      "epoch": 6.575230296827021,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.7729193353745017e-06,
+      "loss": 0.4326,
       "step": 6424
     },
     {
-      "epoch": 6.860192102454643,
-      "grad_norm": 0.365234375,
-      "learning_rate": 3.012086175028461e-07,
-      "loss": 0.4073,
+      "epoch": 6.579324462640737,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.7198844335418637e-06,
+      "loss": 0.4478,
       "step": 6428
     },
     {
-      "epoch": 6.864461045891142,
-      "grad_norm": 0.326171875,
-      "learning_rate": 2.831007289090526e-07,
-      "loss": 0.4363,
+      "epoch": 6.583418628454452,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6673569736766876e-06,
+      "loss": 0.4379,
       "step": 6432
     },
     {
-      "epoch": 6.868729989327641,
-      "grad_norm": 0.38671875,
-      "learning_rate": 2.6555362947601746e-07,
-      "loss": 0.4251,
+      "epoch": 6.587512794268168,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.615337136758172e-06,
+      "loss": 0.4266,
       "step": 6436
     },
     {
-      "epoch": 6.8729989327641405,
-      "grad_norm": 0.333984375,
-      "learning_rate": 2.4856738492931573e-07,
-      "loss": 0.4395,
+      "epoch": 6.591606960081883,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.5638251020165812e-06,
+      "loss": 0.3951,
       "step": 6440
     },
     {
-      "epoch": 6.877267876200641,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.3214205889366954e-07,
-      "loss": 0.3945,
+      "epoch": 6.595701125895599,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.512821046932495e-06,
+      "loss": 0.4134,
       "step": 6444
     },
     {
-      "epoch": 6.88153681963714,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.162777128928317e-07,
-      "loss": 0.3737,
+      "epoch": 6.599795291709314,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4623251472363937e-06,
+      "loss": 0.4396,
       "step": 6448
     },
     {
-      "epoch": 6.885805763073639,
-      "grad_norm": 0.361328125,
-      "learning_rate": 2.0097440634931906e-07,
-      "loss": 0.3549,
+      "epoch": 6.60388945752303,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.412337576907858e-06,
+      "loss": 0.4173,
       "step": 6452
     },
     {
-      "epoch": 6.890074706510139,
-      "grad_norm": 0.357421875,
-      "learning_rate": 1.8623219658417953e-07,
-      "loss": 0.4497,
+      "epoch": 6.607983623336745,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.362858508175136e-06,
+      "loss": 0.4338,
       "step": 6456
     },
     {
-      "epoch": 6.894343649946638,
+      "epoch": 6.612077789150461,
       "grad_norm": 0.361328125,
-      "learning_rate": 1.7205113881674226e-07,
-      "loss": 0.4447,
+      "learning_rate": 2.313888111514395e-06,
+      "loss": 0.4044,
       "step": 6460
     },
     {
-      "epoch": 6.898612593383138,
-      "grad_norm": 0.375,
-      "learning_rate": 1.5843128616451762e-07,
-      "loss": 0.4265,
+      "epoch": 6.616171954964176,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2654265556493022e-06,
+      "loss": 0.4584,
       "step": 6464
     },
     {
-      "epoch": 6.9028815368196375,
-      "grad_norm": 0.37109375,
-      "learning_rate": 1.4537268964288085e-07,
-      "loss": 0.4346,
+      "epoch": 6.620266120777892,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2174740075502794e-06,
+      "loss": 0.4268,
       "step": 6468
     },
     {
-      "epoch": 6.907150480256137,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.328753981649222e-07,
-      "loss": 0.4533,
+      "epoch": 6.624360286591607,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.1700306324340334e-06,
+      "loss": 0.4085,
       "step": 6472
     },
     {
-      "epoch": 6.911419423692636,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.2093945854133035e-07,
-      "loss": 0.404,
+      "epoch": 6.628454452405323,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.123096593762974e-06,
+      "loss": 0.4266,
       "step": 6476
     },
     {
-      "epoch": 6.915688367129135,
-      "grad_norm": 0.3359375,
-      "learning_rate": 1.0956491548009261e-07,
-      "loss": 0.4195,
+      "epoch": 6.632548618219038,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.076672053244599e-06,
+      "loss": 0.4212,
       "step": 6480
     },
     {
-      "epoch": 6.919957310565635,
-      "grad_norm": 0.34765625,
-      "learning_rate": 9.87518115864283e-08,
-      "loss": 0.3697,
+      "epoch": 6.6366427840327535,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.03075717083096e-06,
+      "loss": 0.4263,
       "step": 6484
     },
     {
-      "epoch": 6.924226254002134,
-      "grad_norm": 0.3671875,
-      "learning_rate": 8.850018736260567e-08,
-      "loss": 0.4148,
+      "epoch": 6.640736949846469,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9853521047181963e-06,
+      "loss": 0.4296,
       "step": 6488
     },
     {
-      "epoch": 6.928495197438634,
-      "grad_norm": 0.3671875,
-      "learning_rate": 7.88100812077419e-08,
-      "loss": 0.3985,
+      "epoch": 6.6448311156601845,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9404570113458197e-06,
+      "loss": 0.3664,
       "step": 6492
     },
     {
-      "epoch": 6.9327641408751335,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.968152941771999e-08,
-      "loss": 0.4048,
+      "epoch": 6.6489252814738995,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8960720453963295e-06,
+      "loss": 0.4485,
       "step": 6496
     },
     {
-      "epoch": 6.937033084311633,
-      "grad_norm": 0.34765625,
-      "learning_rate": 6.111456618503874e-08,
-      "loss": 0.4433,
+      "epoch": 6.653019447287615,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8521973597946326e-06,
+      "loss": 0.4253,
       "step": 6500
     }
   ],
   "logging_steps": 4,
-  "max_steps": 6559,
+  "max_steps": 6839,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 7,
   "save_steps": 500,
@@ -11401,7 +11401,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.2506587099166147e+18,
+  "total_flos": 4.253726566678266e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null